Sei sulla pagina 1di 107

This Week in Medicine

Corbis

Chinas cities growing At the end


of 2011, Chinas urban population
overtook its rural population for the
rst time. The National Bureau of
Statistics reports that 513% of the
countrys 135 billion people now live
in cities. For an economy at Chinas
stage of development, this ratio is still
low, possibly due to past constraints
on free movement.
Organ donation in Chile Chiles
Senate has discussed legislation to
reregister the nations organ donors.
2 years after an opt-out system of
registration was instated, a shortage
exists, with more than 2 million
Chileans unwilling to donate. The new
law would assume every adult a donor
unless documentation is produced to
the contrary, even if he or she opted
out under the 2010 system.

RCP and Malaysia The UKs Royal


College of Physicians (RCP) and
Newcastle University have signed
an agreement with the Malaysian
Ministry of Health to help develop
world-class medical education in
Malaysia. The RCP will deliver a series
of programmes in the new medical
education campus in Educity, Malaysia,
eventually oering internationally
recognised qualications.

Safety rst Actors in Los Angeless


pornographic lm industry are to be
legally required to wear condoms after
the city council voted 9-1 in favour
of the measure. The vote headed o
a public ballot, due in June, brought
about by a petition from the AIDS
Healthcare Foundation. Porn lm
producers have threatened to move
state to avoid the mandate.

Taking on depression A European


Parliament interest group met
this week to discuss how to tackle
depression in the European Union.
The group highlighted the scarcity
of specialist depression care services
in individual member states, called
for patients to have more of a say in
their treatment, and suggested that
more support be provided in the
workplace.

New at WFP The World Food Programme (WFP), the UNs rst response
organisation for natural disaster relief
and the ght against world hunger,
has appointed Ertharin Cousin as
its new Executive Director. Cousin,
currently US representative to the UN
agencies for food and agriculture, will
be taking over from current Executive
Director Josette Sheeran in April.

Roger Collier, CMAJ

Bursaries for rural students A bursary scheme for aspiring medical


professionals is being launched in
rural South Africa. Matriculants from
poor families in Limpopo province
who choose to study medicine,
pharmacy, and other medical subjects
can apply for funding for fees, books,
accommodation, and meals, in
return for a period working for the
municipality after graduation.
CMAJ gets new Editor John Fletcher
has been appointed Editor-in-Chief
of the Canadian Medical Association
Journal (CMAJ), taking over from
interim Editor Rajendra Kale. Fletcher,
a Cambridge and Harvard scholar,
has previously worked as an editor at
the BMJ. He intends to focus on the
journals digital presence to promote
the coverage of current medical topics
and address nancial issues.
www.thelancet.com Vol 379 January 28, 2012

Sexual exploitation gures A report


by French charity Fondation Scelles
suggests that 4042 million people are
involved in prostitution worldwide,
70% of whom are aged 1325 years.
Many are migrants, and 12 million
are thought to be tracked. The report
claims that major sporting events
see substantial increases in organised
prostitution in local areas.

Smart pills Proteus Biomedical has


announced a partnership with Lloyds
Pharmacy in the UK to launch a new
digital pill to check whether patients
are taking their prescribed drugs. The
product, called Helius, is a sensorenabled tablet that is swallowed
along with a patients medication. A
chemical reaction within the stomach
creates a signal detected via a sensor
patch worn on the body.
Kings India Institute The India
Institute at Kings College, London,
UK was inaugurated on Jan 26India
Republic Day. The Institute focuses
on activities across a range of
disciplines to address dilemmas and
opportunities that face contemporary
India. Students will examine issues
around the future of democracy,
intellectual property, science, the
environment, growth, and inequality.

For the Centre for Research on


the Epidemiology of Disasters
report on disasters in 2011 see
http://www.unisdr.org/
archive/24692
For more on John Fletcher and
the CMAJ see http://www.cmaj.
ca/site/earlyreleases/18jan12_
utility_stability_and_a_digital_
presence.xhtml
For the RCP/Malaysian
Government collaboration see
http://www.rcplondon.ac.uk/
press-releases/rcp-signsmalaysian-agreement
For more on the new Executive
Director of the WFP see http://
www.wfp.org/news/newsrelease/wfp-welcomesappointment-new-executivedirector
For more on the drug
compliance monitor Helius see
http://www.proteusbiomed.
com/category/press-releases/
For the Kings India Institute
see http://www.kcl.ac.uk/
aboutkings/worldwide/global/
indiainstitute/index.aspx

Paul Grundy

Major disasters A report by the Centre


for Research on the Epidemiology
of Disasters shows that, in 2011,
302 disasters claimed almost 30 000
lives, aected 206 million people,
and came at an economic cost of
US$366 billion. Major cities in seismic
zones should be prepared for repeat
events, the report warns, since twothirds of the total deaths were caused
by earthquakes in Japan and Turkey.

Editorial

In this weeks Perspectives section, Hilary and Steven Rose


describe the moving lm Playing Against Time. This lm
describes the life and music of Barbara Thompson, a jazz
saxophonist, and her eorts to continue with her music
despite the diagnosis of Parkinsons disease in 1996,
including her struggle to obtain care and treatment. The
diagnosis of a neurological disorder, such as Parkinsons
disease, multiple sclerosis, or motor neuron disease, is a
devastating one. Not only is there no curetreatment will
at best delay progressionthere is also the predicament
of a continuing decline in health, functioning, and quality
of life, and a long-term reliance on professionals within
the health and social care system.
An estimated 2 million people in the UK live with a
neurological disorder, excluding migraine, but there is
no centrally collated disease register. It has long been
recognised that there is a lack of prompt access to a
diagnosis, to appropriate care, and to long-term personal
plans for people with neurological disorders in the UK.
In 2005, the National Service Framework for Long-Term
Conditions highlighted that people with neurological
disorders faced lengthy delays in diagnosis, scarce
information on their condition and available services, little
integration of health and social services, and poor hospital
care. It identied 11 areas of relevant quality requirements,
such as patient-centred care and supporting family and
carers, but without clear targets and without a clear plan of
data collection and monitoring. These requirements were
to be fullled by 2015. A midpoint review was abandoned
by the new government. It is therefore perhaps not
surprising, but still deeply disappointing, that a National
Audit Oce report published shortly before the end of
2011 showed that: newly diagnosed people often do not
receive information about their disease, for example 35%
of people newly diagnosed with multiple sclerosis were
not given information about the disorder; the number of
emergency admissions to hospital has increased rather
than decreased; and many people with neurological
disorders admitted to hospital as an emergency do not
see a neurologist. In a study done by Oxford University,
only 22% had a personal care plan. Spending has increased
by 38% since 2005 but it was unclear from the available
data where and how this money was used.
Following these ndings, the Neurological Alliance,
an umbrella organisation representing charities, warned
www.thelancet.com Vol 379 January 28, 2012

of a neurology time bomb with projected increases in


prevalence in an ageing population. Without a focused,
nationally led strategy and strong clinical leadership
within the Department of Health, the service provision
might further deteriorate rather than improve. Steve Ford,
chief executive of the charity Parkinsons UK and chair
of the Neurological Alliance, told a House of Commons
public accounts committee last week that without a
clear outcome strategy and a national clinical director,
neurological disorders are likely to remain low priority for
the new commissioning groups. Unlike cancer, stroke, and
diabetes, which all have strategies and clinical champions,
degenerative neurological disorders are heterogeneous
and complex. Success or progress is not easily measured
in terms of decrease in mortality or incidence. New
treatments are urgently needed but costly and dicult to
develop. In 2011, many new initially promising treatment
avenues, for example for dementia, have been abandoned
by pharmaceutical companies because of futility or harm.
The new US initiative of developing new treatments
through national networks, the national Network for
Excellence in Neuroscience Clinical Trials (NeuroNEXT),
is to be applauded. But it will take many years for these
initiatives to translate into direct benets for patients.
In the meantime, urgent attention should be paid to
an integrated patient-centred approach to diagnosis,
treatment, support, and care for patients with neurological
disorders and their families. There are good examples in
some communities. Steve Pollock, consultant neurologist
in Kent, explained to the public accounts committee
how specialist multiple sclerosis nurses had been used
as navigators through the complex landscape of health
and social care services and achieved a drop in hospital
admissions. With the new Health and Social Care Bill
focused on competition and cost-eciency, an already
fragmented, unequal, and poorly functioning system
might become even more so for people with neurological
disorders. Without a clear targeted focus on neurological
disorders, the governments new long-term conditions
strategy that is currently being developed will fail this
growing group of patients yet again. A clinical champion
for neurology who will present a national strategy with
clear targets, an inbuilt monitoring and data collection
system, and priorities for research is urgently needed and
long overdue. The Lancet

Science Photo Library

Neurological diseases remain neglected and ignored

See Perspectives page 303

For the National Audit Oce


report see http://www.nao.org.
uk/publications/1012/
neurological_conditions.aspx
For more on NeuroNEXT
see Editorial Lancet Neurology
2012; 11: 119

287

Editorial

FPG/Getty Images

Tobacco in the USA: smoke and mirrors

For more on State of Tobacco


Control 2012 see http://www.
stateoftobaccocontrol.org/

In an episode of the science ction series Star Trek: Deep


Space Nine, a trio of capitalist aliens nd themselves
thrown back in time to 1947. They are overjoyed to
discover human beingsincluding medical stasmoking
tobacco. If theyll buy poison, theyll buy anything!
cackles one of the protagonists, smelling a prot.
When future historians look back on the long and
unhappy story of people and tobacco, what will be said
about our current era? The American Lung Associations
report, State of Tobacco Control 2012, gives some clues.
In retrospect, it will seem incredible that 60 years after
Richard Doll conrmed the link between smoking and
lung cancer, and 40 years since a US President declared
war on cancer, 2011 was an abysmal year for tobacco
control at the state level. Comprehensive smoke-free
laws were not passed by any state. In oneNevada
they were weakened. Tobacco taxes were not raised,
and smoking prevention programmes were cut. The
litigiousness of the tobacco industry amazes now and will

continue to amaze for decades to come. The report details


the lawsuits that have been led against the US Food and
Drug Administration in an eort to hamper smoking
prevention eorts, such as restricting sales to youths and
adding graphic warning labels to cigarette packets. Where
does the Federal Government stand? Although commendable in some areasfor example, in its ongoing eorts to
implement the Family Smoking Prevention and Tobacco
Control Actit continues to dawdle over the submission
of the Framework Convention on Tobacco Control to the
Senate, despite having signed it back in 2004.
Renewed eort at State and Federal levels is needed
to put tobacco where it belongs: in the history books,
as a sad and strange episode in the story of human
health. The colourful magazine advertisements of a few
decades ago, featuring doctors endorsing cigarettes,
seem alien to us now. Hopefully the current arguments
over cigarette packaging and sales restriction will seem
equally bizarre to future generations. The Lancet

Corbis

The Research Works Act: a damaging threat to science

For more on the Research


Works Act see http://www.
govtrack.us/congress/billtext.
xpd?bill=h112-3699

288

Academic publishers have become the enemies of


science. So wrote Dr Mike Taylor, a scientist at the UKs
University of Bristol. He, and many scientists like him, are
angry that publishers are supporting the Research Works
Act (RWA), a controversial Bill before the US Congress.
The RWA is co-sponsored by Republican representative
Darrell Issa and the Democrat Carolyn Maloney. It was
introduced to Congress in December, 2011. The Bill
states that its purpose is to ensure the continued
publication and integrity of peer-reviewed research
works by the private sector. If enacted, it would
prevent any Federal agency from requiring government
researchers to enhance the dissemination of their
taxpayer-funded research by posting it on the internet
free of charge without the publishers prior permission.
This provision would overturn the US National Institutes
of Health (NIH) 4-year-old public open-access policy,
which requires NIH grantees to submit copies of their
peer-reviewed papers to PubMed Central for posting no
later than a year after journal publication. On PubMed
Central, the public can freely access papers reporting
publicly funded research.

This short and hastily put together legislation is not


in the interests of either science or the public. Putting
limitations on the dissemination of a scientists own
work is a startlingly ill-considered strategy. Science
is a public enterprise. A scientic publishers primary
responsibility is to serve the research community. Their
own interestsnancial and reputationaldepend
upon the trust the public has in science. Obstructing
the dissemination of publicly funded science will
damage, not enhance, that trust. The RWA brings
publishers and publishing into disrepute. Already,
several academic publishers have spoken out against
this Bill, including the American Association for the
Advancement of Science.
The Lancet also strongly opposes this Bill. Medical
and scientic publishing benets society by creating
and sustaining research collaborations, identifying
and giving space to emerging and neglected concepts
in medicine, positively discriminating on behalf of
unheard voices in health, and opposing the forces
that undermine the values of our profession. The RWA
does none of these. This bad Bill should be rejected.
www.thelancet.com Vol 379 January 28, 2012

Comment

Optimisation of mass chemotherapy to control


soil-transmitted helminth infection
Leaders from international agencies, including WHO and
the World Bank, charities, and pharmaceutical companies,
together with politicians from donor and recipient
countries, will meet in London on Jan 30, 2012,1 to pledge
increased support and collaboration for the control of
neglected tropical diseases (NTDs).2 A London Declaration
from this meeting will mark an expanded vision from
WHO for the elimination of some NTDs and improved
implementation of control eorts for other NTDs by 2020.
Advocacy has been important for raising awareness
of NTDs,3 highlighting the low costs of treatments and
attracting donors. Governments in the USA and UK
have given generously, as have charities such as the
Bill & Melinda Gates Foundation and the Carter Centre.
Drug donations from pharmaceutical companies
have had a catalytic role, especially for the treatment
and control of helminth infections. Mobilisation of
the education sector has also provided a clear policy
framework for deworming delivered through schools.4,5
NTDs are found predominately among poor rural
populations in Africa, Asia, and Latin America and more
than a third of the worlds population is infected with
worms (helminths). Rapid progress to control helminth
infections is possible, however, given the availability of
cheap and eective drugs that can be administered to
entire at-risk populations. For soil-transmitted helminth
infections, the use of albendazole or mebendazole, and to
a lesser extent levamisole or pyrantel pamoate, provides
eective treatment for roundworms (Ascaris lumbricoides),
whipworms (Trichuris trichuria), and hookworms
(Necator americanus or Ancylostoma duodenale), which
are the most widespread infections.6
The increased availability of funds to control soiltransmitted helminth infections raises the question of how
best to use these resources, since reinfection means that
treatment must be administered repeatedly to individuals
in endemic areas. A crucial step for programmes is to
ensure that treatment is targeted to areas of greatest need.
The development of online resources, such as the Global
Atlas of Helminth Infection, provides countries with vital
information for planning and implementing treatment.
WHO guidelines for the interval between rounds of
treatment for common soil-transmitted helminths, and
www.thelancet.com Vol 379 January 28, 2012

for who should be treated and for how long, are based
on the consensus opinion of experts in helminth control5
and do not take account of analytical methods based on
current understanding of the transmission dynamics of
helminths. The key questions for control by chemotherapy
alone focus on how to achieve the greatest impact with
the minimum use of anthelmintic drugs (panel). Disease
transmission during drug treatment is a dynamic process,
with loss and acquisition of worms determined by many
factors, including coverage and ecacy. Models of these
processes provide predictive information about the eect
of intervention at the community level, whether by mass
or targeted chemotherapy.7 This knowledge, however, is
rarely applied in practice.
Helminths have much more predictable dynamics than
most viruses, bacteria, and protozoa. After treatment,
worm populations return to precontrol levels in a
monotonic way, because of density-dependent processes
that inuence parasite reproduction, infection, and mortality and are partly related to the build-up of a small
degree of acquired immunity; the long life expectancy of
established worms in the human host (years rather than
days) is also a factor.5 This understanding should be used
more widely to rene community-based guidelines and
to inform policy makers of what to expect from a given
intervention programme.

For Global Atlas of Helminth


Infection see http://www.
thiswormyworld.org

Panel: Key questions for control of soil-transmitted infections by chemotherapy


1 For a given transmission level, how often should mass or targeted chemotherapy
be administered to sustain infection prevalence and intensity below dened levels?
2 In terms of cost-eectiveness, is it best to target school children, those predisposed
to heavy infection, or the entire community?
3 As the prevalence and intensity fall after repeated rounds of treatment, can the
interval between treatments increase, and by how much?
4 How is the interval between treatments aected by the species mix in the community?
5 How do the demography of the population and the starting geographical distribution
of infection aect the structure of optimum treatment programmes when resources
are nite?
6 What level of infection across a community should trigger mass chemotherapy to
minimise morbidity?
7 Is elimination in a dened area possible by chemotherapy alone?
8 How might repeated mass treatment aect the evolution of drug resistance and how
can this risk be minimised?
9 What should be the target of control programmes?
10 What are the best indicators for assessing the impact of control?

289

Comment

See Online for webappendix

290

A series of important principles emerge from analytical


studies. First, prediction is possible given the relatively
simple dynamics of these infections. Analytical methods
can answer specic questions, such as who should be
treated, how frequently, and whether the frequency
can be decreased as repeated treatment lowers average
worm burdens.7 Second, prevalence is an inadequate
way to monitor the impact of control measures (see
webappendix).8 Third, the optimum interval between
rounds of treatment to keep worm loads to very low
levels depends on the species mix of parasites and the
prevailing reproductive success in a dened location. For
example, if roundworm is the major infection in areas of
high-to-medium reproductive success, treatment needs
to be administered every year, whereas if hookworm is
the major problem, treatment rounds can shift to every
2 years. For schistosomes, with much longer estimated
life expectancy in the human host, the equivalent interval
is 35 years, depending on reproductive success. In areas
of low transmission, intervals can be much longer for all
soil-transmitted helminth infections. In mixed-species
infections, which are typical in many regions, the species
with the shortest life expectancy will determine the
frequency of treatment to sustain low burdens.
Eradication in a given community is possible if
treatment coverage is high and frequent, and the basic
reproductive number R0 not too high (see webappendix
for typical dynamics of hookworm).7,8 In some communities where hookworm is endemic, such as rural
communities in Kenya and India,9 children younger than
15 years often constitute almost 40% of the population
and host perhaps up to 75% of the total worm population,
which is why treatment targeted at school children
alone can result in eective control. Again, based on
demographic and epidemiological measures, predictions
of impact can be made to guide the design of who
should be treated and how often. Human movement
and migration patterns will lead to reintroduction from
surrounding areas if treatment is not uniform and intense
across a large area. Simulations show that the interval
between rounds of treatment can be lengthened after
a few rounds of frequent and intense application (see
webappendix for simulations about eect of increasing
frequency of treatment for school children).
The use of predictive tools in the design of control
programmes for soil-transmitted helminth infections
is not widespread, but they should be used if the goals

of the London Declaration are to be met and value for


money achieved from new drug donations and increased
donor support for implementation. Such predictive
models should become part of WHOs guidelines for
control of soil-transmitted helminths.
Developing analytical approaches for the design of
community-based chemotherapy will provide scientic
precision and predictability. But, as always in public health,
other factors have a confounding role. Integrating control
of soil-transmitted helminth infections with schoolfeeding programmes has been proven to be eective,4,9
as has integration with other disease control activities (ie,
horizontal as opposed to vertical approaches). Optimum
treatment rounds will be tempered by opportunity,
logistics, and delivery costs. Mass and widespread
chemotherapy alone will not eliminate soil-transmitted
helminth infections in the longer term. Concomitant
improvements in sanitation and hygiene are essential.2
Perhaps the biggest challenge to eective control will be
to sustain the interest of governments and international
donors once worm burdens are reduced to very low levels
and obvious morbidity is averted.
*Roy Anderson, T Deirdre Hollingsworth, James Truscott,
Simon Brooker
Department of Infectious Disease Epidemiology, Imperial College,
London W2 1PG, UK (RA, TDH, JT); London School of Hygiene and
Tropical Medicine, London, UK (SB); and Kenya Medical Research
Institute-Wellcome Trust Research Programme, Nairobi, Kenya (SB)
roy.anderson@imperial.ac.uk
RA is a non-executive director of and shareholder in GlaxoSmithKline. TDH, JT,
and SB declare that they have no conicts of interest.
1

2
3
4
5

6
7
8
9

Department for International Development. UK to protect 140 million


from neglected tropical diseases. Jan 21, 2012. http://www.dd.gov.uk/
News/Latest-news/2012/Britain-to-protect-more-than-140-million-inglobal-eort-to-rid-the-world-of-neglected-tropical-diseases (accessed
Jan 23, 2011).
WHO. Working to overcome the global impact of neglected tropical
diseases. Geneva: World Health Organization, 2010.
Hotez PJ, Fenwick A, Savioli L, Molyneux DH. Rescuing the bottom billion
through control of neglected tropical diseases. Lancet 2009; 373: 157075.
Miguel E, Kremer M. Worms: identifying impacts on education and health
in the presence of treatment externalities. Econometrica 2004; 72: 159217.
Montresor A, Gyorkos TW, Crompton DW, Bundy DA, Savioli L. Monitoring
helminth control programmes: guidelines for monitoring the impact of
control programmes aimed at reducing morbidity caused by
soil-transmitted helminths and schistosomes, with particular reference
to school-age children. Geneva: World Health Organization, 1999.
Bethony J, Brooker S, Albonico M, et al. Soil-transmitted helminth infections:
ascariasis, trichuriasis, and hookworm. Lancet 2006; 367: 152132.
Anderson RM, May RM. Infectious diseases of humans: dynamics and
control. Oxford: Oxford University Press, 1991.
Anderson RM, Medley GF. Community control of helminth infections of
man by mass and selective chemotherapy. Parasitology 1985; 90: 62960.
Pullan RL, Gething PW, Smith JL, et al. Spatial modelling of soil-transmitted
helminth infections in Kenya: a disease control planning tool.
PLoS Negl Trop Dis 2011; 5: e958.

www.thelancet.com Vol 379 January 28, 2012

Comment

Adjuvant therapy for gastric cancer after D2 gastrectomy

www.thelancet.com Vol 379 January 28, 2012

disease-free survival with a hazard ratio of 056


(95% CI 044072), whereas in terms of overall survival
the benet was less explicit (072, 052100) at this
timepoint. In the trial of adjuvant TS-1,2 where the
primary endpoint was overall survival, the hazard ratios
for overall survival and disease-free survival were very
similar, and subsequent 5-year follow-up data conrmed
the improvement in both endpoints.5 Improvement of
disease-free survival after longer-term follow-up does
not necessarily mean that there will also be an increase
in overall survival, suggesting that follow-up to at least
5 years is needed.
The other issue is adherence and safety. More
than half of the patients in CLASSIC treated with
capecitabine and oxaliplatin had grade 3 or 4 adverse
events, and nearly 10% withdrew from therapy because
of adverse events. A further 20% of patients refused
treatment, and 33% (174 of 520) of patients withdrew
from treatment with capecitabine and oxaliplatin
or observation. In clinical practice, non-adherence
compromises disease outcomes after adjuvant
treatment, and good adherence to treatment could
benet more patients at high risk of recurrence. In the

Published Online
January 7, 2012
DOI:10.1016/S01406736(11)61928-4
See Articles page 315

Science Photo Library

Although adjuvant therapy for gastric cancer is known


to improve prognosis,1 controversies remain about
the choice of chemotherapeutic regimens, the use of
radiation therapy, and the selection of patients who
are likely to benet from dierent therapies. Adjuvant
therapy is used to eradicate potential micrometastases
after curative surgery and, therefore, therapeutic
strategies vary dependent on anticipated locoregional
disease control after surgery, and by institution
and country. It seems reasonable to apply adjuvant
chemoradiotherapy, providing a combination of radiation therapy for local control and chemotherapy to
achieve systemic control and radiosensitisation, to
patients with less locoregional control; and systemic
adjuvant chemotherapy to those with greater local
control after surgery.
Although guidelines advocate D2 dissection in
centres with specialist expertise, D2 surgery is not
always undertaken as the standard of care outside
east Asia. In Asian countries where D2 dissection is
standard, systemic adjuvant chemotherapy has become
conventional since the ndings of the ACTS-GC trial,
with TS-1, were reported.2 In non-Asian countries,
where D1 is the most common surgical procedure for
gastric cancer, more than half of patients with curative
resection have locoregional recurrence, and adjuvant
chemoradiation therapy has been shown to improve
not only disease-free survival but also overall survival by
reducing locoregional and distant recurrence.3
In The Lancet, Yung-Jue Bang and co-workers now
report4 intriguing results from CLASSIC, a phase 3 trial of
adjuvant chemotherapy in Asian patients after surgery
for gastric cancer. 1035 patients with stage IIIIIb
gastric cancer were randomly assigned to adjuvant
chemotherapy with capecitabine and oxaliplatin for
6 months or surgery alone. The data show improved
disease-free survival and overall survival in patients with
stage II and III gastric cancer, with moderate tolerability.
Disease-free survival is evidently improved across disease
stages II, IIIa, and IIIb, for which TS-1 adjuvant therapy
has not always been proven to have activity.2
There are, however, some issues that need to be
addressed. The primary endpoint of Bang and colleagues
trial was disease-free survival at 3 yearsthere was
a signicant benet of adjuvant chemotherapy on

Diuse gastric adenocarcinoma

291

Comment

adjuvant setting, however, a substantial proportion


of patients are already cured by surgery alone and
additional therapy is not required, which might, in turn,
decrease adherence. Thus, decision making by patients
and physicians might be practically dependent on
individual perceptions and judgment about recurrent
risk, therapeutic burden including adverse events and
cost, and benet from the therapy.
Together, the ndings of CLASSIC4 and ACTSGC2 suggest that uoropyrimidine-based adjuvant
therapies improve disease outcomes of patients with
gastric cancer undergoing D2 resection and that, in
Asia, we now have two options available for adjuvant
treatment. Identication of higher-risk patients
and prediction of drug ecacy by biomarkers, and
introduction of targeted agents such as trastuzumab,
should be considered for adjuvant therapy of gastric
cancer in the future.

Toshirou Nishida
Department of Surgery, Osaka Police Hospital, Osaka 543-0035,
Japan, and Department of Surgery, Osaka University Graduate
School of Medicine, Osaka, Japan
toshin@mvb.biglobe.ne.jp
I declare that I have no conicts of interest.
1

GASTRIC (Global Advanced/Adjuvant Stomach Tumor Research


International Collaboration) Group, Paoletti X, Oba K, Burzykowski T, et al.
Benet of adjuvant chemotherapy for resectable gastric cancer:
a meta-analysis. JAMA 2010; 303: 172937.
Sakuramoto S, Sasako M, Yamaguchi T, et al. Adjuvant chemotherapy
for gastric cancer with S-1, an oral uoropyrimidine. N Engl J Med 2007;
357: 181020.
Macdonald JS, Smalley SR, Benedetti J, et al. Chemoradiotherapy after
surgery compared with surgery alone for adenocarcinoma of the stomach
or gastroesophageal junction. N Engl J Med 2001; 345: 72530.
Bang Y-J, Kim Y-W, Yang H-K, et al, for the CLASSIC trial investigators.
Adjuvant capecitabine and oxaliplatin for gastric cancer after D2
gastrectomy (CLASSIC): a phase 3 open-label, randomised controlled trial.
Lancet 2012; published online Jan 7. DOI:10.1016/S0140-6736(11)61873-4.
Sasako M, Sakuramoto S, Katai H, et al. Five-year outcomes of a
randomized phase III trial comparing adjuvant chemotherapy with
S-1 versus surgery alone in stage II or III gastric cancer. J Clin Oncol 2011;
29: 438793.

Vitamin K antagonists: self-determination by self-monitoring?


Published Online
December 1, 2011
DOI:10.1016/S01406736(11)61748-0
See Articles page 322

292

Vitamin K antagonists have been successfully used


for more than 50 years as oral anticoagulant drugs in
patients at risk of venous or arterial thromboembolism.
Because of a narrow therapeutic window and an
unpredictable dose response, regular laboratory
monitoring is mandatory. A strong association exists
between the intensity of anticoagulation and adverse
clinical events, with an increased thrombotic risk at
an international normalised ratio (INR) of less than
20 and an enhanced risk of bleeding when the INR
exceeds 40.1 Even in the stringent environment
of clinical trials, a large proportion of patients are
outside their therapeutic range and hence exposed
to a risk of bleeding or thrombosis.2,3 Thus, eorts to
increase the amount of time the patient spends in the
therapeutic range have been made: establishment of
specialised anticoagulation clinics, use of computerassisted dosing systems, or patient education. Portable
monitoring devices have become available, which oer
the opportunity for patients to undertake INR tests
themselves (self-testing) and to make their own dose
adjustments (self-management). Self-monitoring is
attractive not only because it oers exibility and selfdetermination but also because it can improve the
quality of anticoagulation.

In The Lancet, Carl Heneghan and colleagues4


summarise results from 11 studies that compared
self-monitoring of oral anticoagulation therapy with
conventional care. They merged data of individual
patients rather than aggregating already published
information. This method allowed for a time-toevent analysis and a exible analysis of subgroups and
outcomes, together with increasing statistical power.
Self-monitoring was superior to usual care with 49%
(hazard ratio 051, 95% CI 031085) risk reduction of
thromboembolic events. Bleeding rates were similar
in both groups and self-monitoring had no major
eect on mortality, a nding that is contrary to that
of a recently published meta-analysis of aggregate
data,5 which reported a signicant 26% (odds ratio
074, 95% CI 063087) reduction in death among
patients who self-monitored. The results also showed
that self-management (patients test and adjust their
own dose) is better than self-testing (patients test,
but dosage is done by physicians). However, selfmonitoring does not provide the same advantage to
all patients. The benet over conventional care was
mainly seen in patients younger than 55 years and in
those with mechanical heart valves. In patients with
atrial brillation, however, the thromboembolic risk
www.thelancet.com Vol 379 January 28, 2012

in self-monitoring patients was reduced by a third


(hazard ratio 067), but with an upper 95% CI of 157 this
dierence was not signicant. A possible explanation
could be that patients younger than 55 years, especially
those with mechanical heart valves, are highly aware
of thromboembolic risks and are therefore prepared to
manage their medical treatments, including therapy
with vitamin K antagonists.
Advocacy for wide use of self-monitoring of
vitamin K antagonists raises the question of
whether they will remain the drugs of choice for
long-term anticoagulation in the future. One of
their main indications is atrial brillation. With an
ageing population, the number of candidates for
chronic anticoagulation and the demand for oral
anticoagulants will grow. However, several new oral
anticoagulantswith the advantage of a predictable
anticoagulant eect and no need to be monitored
have been or are about to be marketed. Dabigatran, a
thrombin inhibitor, and rivaroxaban, directed against
factor Xa, are both at least as safe and eective
as vitamin K antagonists in patients with venous
thrombosis.6,7 Importantly, dabigatran at a dose of
150 mg twice daily and apixaban, another factor Xa
inhibitor, are superior to vitamin K antagonists in
patients with atrial brillation in terms of reduction
of stroke and bleeding, and apixaban is superior in
terms of reduction of mortality.2,8 To what extent
these new agents will replace vitamin K antagonists is,
however, unclear. These drugs are partly cleared by the
kidneys and their use is restricted in patients with renal
insuciency; they do not have a direct antidote and are
expensive. Moreover, few dierences exist in ecacy
and safety between the new drugs and vitamin K
antagonists in patients who spend most of their time
within the therapeutic INR range.9 Whether clinical
outcomes with these new drugs are as good in patients
with mechanical heart valves as in patients with atrial
brillation or venous thrombosis is also not known.
So we agree with Heneghan and colleagues that selfmanagement (rather than self-testing) of treatment
with vitamin K antagonists should be oered to
patients with mechanical heart valves, especially those
younger than 55 years. However, we do not see a place
for self-monitoring in other areas of this treatment
except for individual patients for whom access to
routine usual anticoagulation care is restricted.
www.thelancet.com Vol 379 January 28, 2012

Science Photo Library

Comment

Thrombus occluding artery

*Paul Alexander Kyrle, Sabine Eichinger


Department of Medicine I, Medical University of Vienna,
1090 Vienna, Austria
paul.kyrle@meduniwien.ac.at
PAK and SE are consultants for Bayer Healthcare (manufacturer of
rivaroxaban) and Boehringer Ingelheim (manufacturer of dabigatran) and
have received speakers fees from Bayer Healthcare, Boehringer Ingelheim,
and Sano-Aventis. PAK is a consultant for Pzer (manufacturer of apixaban)
and has received speakers fees from Daiichi Sankyo. SE has received speakers
fees from Pzer.
1

7
8

Singer DE, Chang Y, Fang MC, et al. Should patient characteristics inuence
target anticoagulation intensity for stroke prevention in nonvalvular atrial
brillation? Circ Cardiovasc Qual Outcomes 2009; 2: 297304.
Connolly SJ, Ezekowitz MD, Yusuf S, et al, for the RE-LY Steering Committee
and Investigators. Dabigatran versus warfarin in patients with atrial
brillation. N Engl J Med 2009; 361: 113951.
Patel MR, Mahaey KW, Garg J, et al, for the ROCKET AF Steering
Committee and Investigators. Rivaroxaban versus warfarin in nonvalvular
atrial brillation. N Engl J Med 2011; 365: 88391.
Heneghan C, Ward, A, Perera R, and The Self-Monitoring Trialist
Collaboration. Self-monitoring of oral anticoagulation: systematic review
and meta-analysis of individual patient data. Lancet 2011; published online
Dec 1. DOI:10.1016/S0140-6736(11)61294-4.
Bloomeld HE, Krause A, Greer N, et al. Meta-analysis: eect of patient
self-testing and self-management of long-term anticoagulation on major
clinical outcomes. Ann Intern Med 2011; 154: 47282.
Schulman S, Kearon C, Kakkar AK, et al, for the RE-COVER Study Group.
Dabigatran versus warfarin in the treatment of acute venous
thromboembolism. N Engl J Med 2009; 361: 234252.
The EINSTEIN Investigators. Oral rivaroxaban for symptomatic venous
thromboembolism. N Engl J Med 2010; 363: 2499510.
Granger CB, Alexander JH, McMurray JJ, et al, for the ARISTOTLE
Committees and Investigators. Apixaban versus warfarin in patients with
atrial brillation. N Engl J Med 2011; 365: 98192.
Wallentin L, Yusuf S, Ezekowitz MD, et al, for the RE-LY investigators.
Efficacy and safety of dabigatran compared with warfarin at different
levels of international normalised ratio control for stroke prevention in
atrial fibrillation: an analysis of the RE-LY trial. Lancet 2010;
376: 97583.

293

Comment

Neonatal screening for lysosomal storage disorders


Published Online
November 30, 2011
DOI:10.1016/S01406736(11)61744-3

Scott Applewhite/AP/Press Association Images

See Articles page 335

294

Lysosomal storage disorders are a diverse group of more


than 50 serious, progressive diseases. Until recently,
treatment was symptomatic with the expected outcomes of great disability and premature death.
Improvements in bone-marrow transplantation and
development of recombinant enzyme replacement
therapies for some of these disorders have raised the
expectation that neonatal screening could enable early
treatment before irreversible damage occurs.1 The
greatly improved early outcomes of infants treated
soon after birth following identication through the
Taiwanese Pompes disease screening programme2 add
support to this contention, but long-term data are not
yet available.
In The Lancet, Thomas Mechtler and colleagues3
report the outcome of anonymised neonatal screening
in Austria for a group of four treatable lysosomal
storage disorders. During a 7 month period, just under
35 000 blood spots from neonates were tested for
levels of the enzymes responsible for Pompes disease,
Fabrys disease, Gauchers disease, and Niemann-Pick
disease types A and B. Mechtler and co-workers report,
as have other pilot screening programmes for such
disorders,4,5 a higher than expected birth incidence
(1 per 2315), with a high frequency of mutations that
are associated with late onset of symptoms, often in
adulthood. They rightly speculate that consideration of
neonatal screening for this group of disorders will raise
many ethical issues.

Technological and therapeutic advances during the


past 10 years have made possible a great expansion
of neonatal screening, and newborn babies are an
attractive target for screening tests because of the
inherent eciencies of screening this population.
However, views about which disorders should be
included in dried blood-spot screening vary substantially,
ranging from the conservative ve disorders sanctioned
in the UK to more than 50 recommended in the USA.6
Frameworks and systems for rational consideration
of the merits of screening for additional disorders are
developed or under consideration in Australia and
elsewhere.7,8 Assessment of the benets of neonatal
screening is particularly dicult, mainly because the
rarity of the target disorders requires huge numbers
of trial participants, usually running into millions, to
get signicant results. Additionally, the rm belief,
often backed by only low-order evidence, that early
treatment is benecial makes a randomised controlled
trial unethical for many. Only for cystic brosis has such
a trial been attempted.9
Another problem is the increased rate of diagnosis
by screening compared with clinical diagnosisthis
increase might be substantial, and makes comparison
of screened and unscreened groups dicult, particularly
for disorders that might present their rst symptoms
in adulthood. Scarcity of formal evidence of benet,
although some exists,10 and fears of the harm, possibly
overestimated,11 emanating mainly from false-positive
screening results have probably been the key reasons for
the very conservative approach to neonatal screening
in the UK and a few other countries. But clearly harm,
including unnecessary deaths,10 might also come from
not screening.
Many uncertainties remain for clinicians and
families confronted by positive results from screening
for lysosomal storage disorders. Precise genotypephenotype correlation is not possible for many of these
disorders, leaving the managing clinician uncertain
when counselling a childs family about the implications
of the result. Whereas the best possible treatment for
mucopolysaccharidosis type I is generally accepted to be
bone-marrow transplantation before the age of 2 years,
the appropriate age (or the need) for commencement
of enzyme replacement therapy for mild forms of
www.thelancet.com Vol 379 January 28, 2012

Comment

this disorder, or for others such as Fabrys disease, is a


matter of debate, particularly with mutations predicted
to have a late onset phenotype. Even with evidence of
ecacy, the very high costs of treatments3,12 are likely
to be beyond the capacity for health-system funding
in many jurisdictions. Additionally, opinions vary
between patients. Somewhat surprisingly, studies done
with hypothetical scenarios put to Australian and US
mucopolysaccharidosis family-support groups have
found much variation in support for neonatal screening
for untreatable disorders.13 A solution might be separate,
informed consent for such testing, but the complexity
of obtaining such consent could put routine screening
programmes at risk. Although there are currently many
uncertainties about neonatal screening for lysosomal
storage disorders, as suggested by Mechtler and
colleagues3 discussion needs to happen, because this
is but one example of what is to come; therapeutic and
technological possibilities inexorably advance, and the
ability to make an early diagnosis by screening eciently
and cheaply will be relevant to many disorders in the
foreseeable future. The anonymised study design used by
Mechtler and colleagues showed feasibility of screening
and established frequency of the disorders, but precluded
clinical assessment of neonates identied as having a
disorder by the screening process. Future research studies
without long-term follow up, informed consent, and
clinical assessment could be considered unethical.

JF has received funding to attend lysosomal storage disease meetings


sponsored by Genzyme, Shire (TKT), and Actelion, and a gift from Actelion of
miglustat for research studies; she has participated in advisory boards for
Actelion and Protalix. BW declares that she has no conicts of interest.
1
2

6
7

8
9

10

11

12

13

Nakamura K, Hattori K, Endo F. Newborn screening for lysosomal storage


disorders. Am J Med Genet C Semin Med Genet 2011; 157: 6371.
Chien YH, Lee NC, Thurberg BL, et al. Pompe disease in infants: improving
the prognosis by newborn screening and early treatment. Pediatrics
2009; 124: e111625.
Mechtler TP, Stary S, Metz TF, et al. Neonatal screening for lysosomal
storage disorders: feasibility and incidence from a nationwide study in
Austria. Lancet 2011; published online Nov 30. DOI:10.1016/S01406736(11)61266-X.
Spada M, Pagliardini S, Yasuda M, et al. High incidence of later-onset
Fabry disease revealed by newborn screening. Am J Hum Genet 2006;
79: 3140.
Hwu WL, Chien YH, Lee NC, et al. Newborn screening for Fabry disease
in Taiwan reveals a high incidence of the later-onset GLA mutation
c.936+919G>A (IVS4+919G>A). Hum Mutat 2009; 30: 1397405.
Pollitt RJ. Introducing new screens: why are we all doing dierent things?
J Inherit Metab Dis 2007; 30: 42329.
Calonge N, Green NS, Rinaldo P, et al. Committee report: method for
evaluating conditions nominated for population-based screening of
newborns and children. Advisory Committee on Heritable Disorders in
Newborns and Children. Genet Med 2010; 12: 15359.
UK National Screening Committee. Programme appraisal criteria.
http://www.screening.nhs.uk/criteria (accessed Nov 5, 2011).
Farrell PM, Lai HJ, Li Z, et al. Evidence on improved outcomes with early
diagnosis of cystic brosis through neonatal screening: enough is
enough! J Pediatr 2005; 147 (suppl 3): S3036.
Wilcken B, Haas M, Joy P, et al. Expanded newborn screening: outcome
in screened and unscreened patients at age 6 years. Pediatrics 2009;
124: e24148.
Prosser LA, Ladapo JA, Rusinak D, Waisbren SE. Parental tolerance of
false-positive newborn screening results. Arch Pediatr Adolesc Med
2008; 162: 87076.
Connock M, Burls A, Frew E, et al. The clinical eectiveness and
cost-eectiveness of enzyme replacement therapy for Gauchers disease:
a systematic review. Health Technol Assess 2006; 10: 1136.
Hayes IM, Collins V, Sahhar M, Wraith JE, Delatycki MB. Newborn
screening for mucopolysaccharidoses: opinions of patients and their
families. Clin Genet 2007; 71: 44650.

*Janice Fletcher, Bridget Wilcken


Genetics and Molecular Pathology, SA Pathology, Womens and
Childrens Hospital, North Adelaide, SA 5006 Australia (JF);
University of Adelaide, Adelaide, SA, Australia (JF); and Sydney
Childrens Hospitals Network and University of Sydney, Sydney,
NSW, Australia (BW)
janice.etcher@adelaide.edu.au

Oral azithromycin for treatment of yaws


Yawsan infectious disease caused by Treponema
pallidum subsp pertenueaects children and adults in
poor rural communities in tropical countries, causing
disguring lesions of the skin and bones. The yaws
elimination programme coordinated by WHO in the
1950s and 1960s screened more than 160 million
people, treated more than 50 million people with
intramuscular injections of benzathine benzylpenicillin,
www.thelancet.com Vol 379 January 28, 2012

and reduced the prevalence of yaws by more than 95%;


however, yaws was not eliminated.1 This disease is
now re-emerging (largely unnoticed) in parts of Africa,
southeast Asia, and the Pacic islands. Oriol Mitj and
colleagues study2 in The Lancet shows that a single oral
dose of azithromycin is as eective as intramuscular
benzathine benzylpenicillin for the treatment of yaws
in Papua New Guinea. 250 children aged 6 months

Published Online
January 11, 2012
DOI:10.1016/S01406736(11)61842-4
See Articles page 342

295

Comment

Science Photo Library

to 15 years with yaws were randomly allocated, in an


open-label trial, to receive either drug. At 6-month
follow-up 96% of patients in the azithromycin
group were cured, as were 93% in the benzathine
benzylpenicillin group. These data correspond to a
treatment dierence of 34% (95% CI 93 to 24),
meeting the prespecied criteria for non-inferiority.
This publication is perhaps the most important on
yaws in the past 50 years, and could facilitate the
elimination of this ancient scourge.
Single-dose oral treatment for yaws would have several
advantages. Injectable penicillin causes anaphylactic
shock in a small proportion of patients. Although the
risk is small (<1 patient in 50 000), the consequences can
be severe, and control programmes using penicillin need
to train sta and provide them with the means to treat
anaphylaxis. Although WHOs elimination programme
has not been implicated, attempts to control yaws by
mass treatment with injectable drugs have previously
led to the transmission of blood-borne viruses, such as
hepatitis C.3 By contrast, oral azithromycin is safe and
easy to administer. Community-based mass treatment
programmes using this antibiotic for the control of
trachoma have been well accepted by rural communities
in many parts of Africa, are highly eective in reducing
the prevalence of trachoma,4 and could have collateral
benets. In one study in Ethiopia, mass treatment every
year with azithromycin reduced all-cause mortality by
50% in children aged 15 years.5
Should oral azithromycin now replace injectable
penicillin in community-based treatment programmes
for the control and elimination of yaws? Penicillin has
stood the test of time: T pallidum remains fully sensitive
to the drug after 60 years; however, the same is not true
for macrolide antibiotics, such as azithromycin. An oral
dose of azithromycin was equivalent to intramuscular
penicillin for the treatment of syphilis in Africa,6,7
and was used to treat early and incubating syphilis in
North America in the early 1990s. However, treatment
failures were noted in men who have sex with men in
California shortly after introduction of the drug.8 A
mutation conferring resistance to both erythromycin
and azithromycin was present in a high proportion of
strains of T pallidum found in North America and Europe
in men who have sex with men.9 Macrolide resistance
has been documented in T pallidum in patients with
syphilis in China.10
296

The rapid spread of macrolide-resistant T pallidum


strains in sexual networks in developed countries
does not necessarily imply that resistance will become
prevalent in rural areas in the developing countries
where yaws remains endemic. Notably, no macrolide
resistant strains were found in patients with syphillis
in Madagascar.11 Perhaps such strains are less likely to
persist in resource-poor settings where these relatively
expensive antibiotics are rarely used. Communitybased mass treatment with azithromycin has been
widely used to control trachomalike yaws a disease
of poor rural communities in developing countries
for more than 10 years. Macrolide-resistant strains
of Streptococcus pneumoniae have been isolated from
nasopharyngeal swabs taken soon after azithromycin
mass treatment, but did not generally persist, except
in Australian aboriginal populations who had access
to macrolide antibiotics between rounds of mass
treatment.12,13 Macrolide-resistant T pallidum in
American patients with syphilis is associated with use
of macrolides for other disorders,14 which suggests that
azithromycin could remain eective for the treatment
of yaws in communities where macrolides are not
widely available. More studies, such as that done by
Mitj and colleagues, are needed in other regions, and
careful follow up will be needed to ensure that clinically
signicant resistance to azithromycin does not develop
in T pallidum subsp pertenue.
David Mabey
Clinical Research Department, London School of Hygiene and
Tropical Medicine, London WC1E 7HT, UK
david.mabey@lshtm.ac.uk
I declare that I have no conicts of interest.
1
2

3
4
5

Antal GM, Lukehart SA, Meheus AZ. The endemic treponematoses.


Microbes Infect 2002; 4: 8394.
Mitj O, Hays R, Ipai A, et al. Single-dose azithromycin versus benzathine
benzylpenicillin for treatment of yaws in children in Papua New Guinea: an
open-label, non-inferiority, randomised trial. Lancet 2012; published online
Jan 11. DOI:10.1016/S0140-6736(11)61624-3.
Nerrienet E, Pouillot R, Lachenal G, et al. Hepatitis C virus infection in
Cameroon: a cohort-eect. J Med Virol 2005; 76: 20814.
Solomon AW, Holland MJ, Alexander NDE, et al. Mass treatment with
single-dose azithromycin for trachoma. N Engl J Med 2004; 351: 196271.
Porco TC, Gebre T, Ayele B, et al. Eect of mass distribution of azithromycin
for trachoma control on overall mortality in Ethiopian children:
a randomized trial. JAMA 2009; 302: 96268.
Riedner G, Rusizoka M, Todd J, et al. Single dose azithromycin versus
penicillin G benzathine for the treatment of early syphilis. N Engl J Med
2005; 353: 123644.
Hook EW, Behets F, Van Damme K, et al. A phase III equivalence trial of
azithromycin versus benzathine penicillin for treatment of early syphilis.
J Infect Dis 2010; 201: 172935.

www.thelancet.com Vol 379 January 28, 2012

Comment

10
11

Centers for Disease Control and Prevention. Azithromycin treatment


failures in syphilis infectionsSan Francisco, California, 20022003.
Morb Mortal Wkly Rep 2004; 53: 19798.
Lukehart SA, Godornes C, Molini BJ, et al. Macrolide resistance in
Treponema pallidum in the United States and Ireland. N Engl J Med 2004;
351: 15458.
Zhou P, Li K, Lu H, et al. Azithromycin treatment failure among primary and
secondary syphilis patients in Shanghai. Sex Transm Dis 2010; 37: 72629.
Van Damme K, Behets F, Ravelomanana N, et al. Evaluation of
azithromycin resistance in Treponema pallidum specimens from
Madagascar. Sex Transm Dis 2009; 36: 77576.

12

13

14

Batt SL, Charalombous BM, Solomon AW, et al. Impact of azithromycin


administration for trachoma control on the carriage of antibiotic resistant
Streptococcus pneumoniae. Antimicrob Agents Chemother 2003;
47: 276569.
Leach AJ, Shelby-James TM, Mayo M, et al. A prospective study of the impact
of community-based azithromycin treatment of trachoma on carriage and
resistance of Streptococcus pneumoniae. Clin Infect Dis 1997; 24: 35662.
Marra CM, Colina AP, Godornes C, et al. Antibiotic selection may contribute
to increases in macrolide-resistant Treponema pallidum. J Infect Dis 2006;
194: 177173.

We invite you to submit your best research in respiratory


medicine for our annual respiratory theme issue, which
will be published to coincide with the 2012 European
Respiratory Society (ERS) conference in Vienna, Austria
on Sept 15.
High-quality randomised controlled trials that
advance knowledge about treatment and prevention
of respiratory disorders are particularly welcome, but
we will also consider other original research papers
that are clinically important or add substantially to
understanding of underlying disease processes or
susceptibility to disease.
If your paper is to be presented at the ERS conference
and you need to adhere to an embargo policy, please
let us know so that we can plan online publication
accordingly. Articles should be submitted via our online
submission system by the deadline of April 27, 2012. In
exceptional circumstances we might consider papers
after this deadline for our fast track process but please
let us know in advance about your submission in such
cases. When submitting, please state clearly in the
cover letter that your paper is in response to the call
for papers in respiratory medicine. Do not hesitate to

www.thelancet.com Vol 379 January 28, 2012

Siegfried Layda/Getty Images

Respiratory medicine: a call for research papers

contact us should you have any questions about this


call for papers.

To submit a paper go to
http://ees.elsevier.com/thelancet

Sabine Kleinert
The Lancet, London NW1 7BY, UK
Sabine.kleinert@lancet.com

297

Comment

Oine: A shot in the dark

Science Photo Library

Every 5 years in Britain, a gargantuan exercise in judging


the performance of our research-based universities takes
place. The Higher Education Funding Council of England
(HEFCE), for example, will spend 16 billion in 201112
on research in English universities. (A friend of mine in
Italy almost choked on his cappuccino when I told him this
gure.) But how should HEFCE, and similar bodies in Wales,
Scotland, and Northern Ireland, judge who gets what?
In the past, institutions have competed in a Research
Assessment Exercise. But the system of giving block
grants to universities is undergoing a minor revolution.
The new means of sorting intellectual wheat from cha
is called the Research Excellence Framework (REF), due to
report its ndings in December, 2014. University leaders
are anxious. The ranking their institution receives is akin
to the position of a football club in a league table (there
are only two rankings that matter: 3*, internationally
excellent, and 4*, world-leading). What happens to
football managersadmired hero one day, unemployed
failure the nextcan just as easily happen to university
Vice-Chancellors. REF2014 is their judgment day.
*

RIBA Library Photographs Collection

Katherine Rolfe

The revolutionary aspect of the REF is this: in addition


to pure research excellence, universities will have
20% of their score based on impact (65% of the REF
will depend on the originality and rigour of research
outputs, while 15% will be attributed to the vitality and
sustainability of the institutions research environment).
But what is impact? Thankfully, the REF will have
nothing to do with journal impact factors. Being
published in The Lancet wont help you one little bit.
Instead, impact will be measured by the eects of
research beyond the academic communitythe reach
and signicance of work on the economy, society,
culture, public policy or services, health, environment,
and quality of life. Although the REF is measuring the
period 200813 (the closing date for submissions is
Nov 29, 2013, and authors are already, mistakenly,
negotiating publication of papers so they can meet
this deadline), universities can submit work going back
to 1993. It takes time to generate true impact. How
an institution interprets impact will demand a lively
imagination. Judging cause and eect will be an act
of faith (one scientist involved in the REF has called
298

impact assessment a massive experiment; another


said that it might sound like a shot in the dark). But
then the entire REF is an act of faith, a shot in the dark.
*
The REF is about quality. The principal means of judging
quality will be peer review by panels of academics
(citation data will be used, but only at the margins).
Institutions can submit up to four research outputs per
scientist. But at a meeting last week of The 1942 Club (a
group of professors of medicine established to discuss the
eects of health-service reforms on academic medicine),
many questions were asked of those leading the REF.
How will the REF take account of elds that are relatively
poorly funded (eg, palliative medicine)? How will the
REF stop game playing? Peer review will mean highly
subjective assessments of value. In advance of The 1942
Clubs meeting, I asked (via Twitter) what quality means.
Answers included simplicity, elegance, the preparedness
for discovery (good research is about more than
results), as well as sound methods and statistics. One
medical scientist wrote that there is no Platonic ideal
of quality. So what criteria will the REF panels use? How
will they judge (consistently) importance, usefulness,
relevance, methods, ethics, completeness, and accuracy?
How will they decide whether work adds to the sum of
our knowledge or whether it is wasteful? Any assessment
of quality depends on our own personal judgment about
the purpose of research. Dierent funding bodies, for
example, emphasise dierent purposes. The MRC focuses
on human health alone, the Wellcome Trust includes
animal health, and the National Institute for Health
Research emphasises, in addition to health, wealth
creation. The director of one of these funding bodies has
said that it is not a concern for the patient that motivates
his interest in research, but the wonder of science. Will
REF panels be judging research submissions based on
their importance for patients and the public? Or will their
measure be the wonder that work inspires? Who knows.
The REF may be the least worst system we have to allocate
billions of pounds of public money to our universities. But
it is a awed and unreliable system all the same.
Richard Horton
richard.horton@lancet.com

www.thelancet.com Vol 379 January 28, 2012

World Report

CDC planning trial for mysterious nodding syndrome


While the cause of nodding syndrome in Africa continues to perplex scientists, they are hoping to
soon unravel one mystery: which treatments the disease might respond to. John Donnelly reports.

www.thelancet.com Vol 379 January 28, 2012

associated with malnourished children


or young adults who have a vitamin B6
deciency.
Some epidemiologists have suggested that climate change could have
played a part, especially if the disease
is denitively linked to the parasite
causing river blindness and the range
of the parasite is expanding. But Dowell
said many questions remain. One
question is if that parasite is connected
to the epidemic, why has it in the past
attacked peoples corneas, while now it
is attacking the brain.

All the children look like they


have cerebral palsy. So many of
the kids are malnourished...
Children who come down with
nodding syndrome often lose their
ability to interact normally, lose
their attention span, and cannot
feed themselves. Some children die,
but many of the cases are related to
seizures, including drownings and
getting burned by res. The children
have a real cognitive decline, Dowell
said. These were healthy, growing
kids, as young as 5 or 6 years old who
once they get this, they waste away,
they are malnourished, they drop out
of school, and they become completely
dependent on their caregivers.
Last year, when The Lancet visited
Witto Payam, a community in South
Sudan, nodding syndrome had
created a village of vacant children.
Some 70 children with the disease
gathered in a small clearing in the centre
of the community. They sat next to their
parents, or each other. Each seemed
distracted. Some used sticks to draw
lines in the ground. Some leaned on an
adult. One girl was asleep in the dirt.
My daughter was a normal girl,
and then at age 6 [years], she started

nodding whenever she saw food, said


Rasul Kegi, a mother. 3 years later, she
started convulsing. Her mental ability
is so diminished. Even when I call her
sometimes, she doesnt respond.
Fred Hartman, Management Sciences
for Healths Country Lead for South
Sudan and supervisor for the USAIDfunded Sudan Health Transformation
Project II, has worked for 30 years in
Africa, but when he rst saw children
with nodding syndrome in 2010 he was
shocked. Ive been all over the world,
and Ive never seen anything like it,
he said. All the children look like they
have cerebral palsy. So many of the kids
are malnourished. You bring them food
and they cant eat it. And by the time
you make the diagnosis, its too late for
the children. Its really sad.
Dowell called the CDCs upcoming
investigation important. We are in the
business of detecting outbreaks. For
the most part, when we investigate an
outbreak, the cause becomes clear, but
periodically we get these ones where
the cause isnt clear, he said. Asked
where nodding syndrome stood on the
list of mystery outbreaks, Dowell said:
Its right there at the top of the list.

Management Sciences for


Health provided funding for JDs
reporting trip to South Sudan
in 2011

John Donnelly

Dominic Chavez

A mysterious illness aecting mostly


children called nodding syndrome
is spreading across wide swaths of
northern Uganda and South Sudan,
stumping investigators who have failed
so far to determine what is causing
the disease that gives young people
epileptic-like seizures and impairs their
cognition.
A team of scientists from the US
Centers for Disease Control and
Prevention will be travelling to the
region in mid-February for their fourth
such exploration; this time, they hope
to lay the groundwork for a clinical trial
to test treatments for the disease.
It is a devastating disease from the
patients point of view and the villagers
point of view, said Scott Dowell,
director of CDCs Division of Global
Disease Detection and Emergency
Response, and the lead investigator
on nodding syndrome. Its been so
dicult to gure out what is causing it
and what should be done about it.
In northern Uganda, the current cases
are in the districts of Kitgum, Pader, and
Gulu. Ugandan ocials say that in Pader
District alone nodding syndrome is the
cause of death for 66 children. They
have diagnosed roughly 1000 cases
between August and December, 2011.
There is no known cure for the disease,
and investigators do not know if it is
communicable.
Researchers have been examining
several possible causes for nodding
syndrome, a name given because those
infected often nod their heads when
food is put in front of them. They have
found an unexplained association
with Ochocerca volvulus, a parasite that
can cause river blindness; tests have
shown that high numbers of those with
nodding syndrome also have antibodies
in their system for Ochocerca. Nearly
all the children live near rivers or fastmoving streams. The disease also is

Many of the children in Witto Payam village, South Sudan, have nodding syndrome

299

World Report

BRCA patent dispute may head to US Supreme Court

Science Photo Library

The long-running dispute over patents for the BRCA1 and BRCA2 genes granted to Myriad
Genetics may nally be laid to rest by the US Supreme Court. Sharmila Devi reports.

Chromosome 17 and the


BRCA1 gene (red)

300

US physicians, scientists, and patients


challenging the validity of commercial
patents on human genes are moving
closer to a possible hearing in the
US Supreme Court, which they hope
will uphold a general prohibition on
patenting laws of nature.
On Jan 13, Utah-based Myriad
Genetics, a company that makes tests
for breast and ovarian cancer, led a
response with the Supreme Court to
a petition made by the American Civil
Liberties Union (ACLU) and the Public
Patent Foundation (PUBPAT).
Since 2009, the two legal groups
have represented a host of plaintis
including womens groups and genetic
counsellors. They argue that the patents
granted to Myriad for the BRCA1 and
BRCA2 gene sequences eectively grant
the company a monopoly on medical
and research testing for familial breast
and ovarian cancer.
In March, 2010, a district court
ruled that human genes and Myriads
sequence comparison or method
claims were not patent-eligible. In
July, 2011, a divided Court of Appeals
for the Federal Circuit armed the
lower courts nding that the method
claims were invalid because they were
an unpatentable mental process. But
the appeals court did reverse the lower
court decision in holding that the gene
sequences at issue were patent-eligible
as isolated human DNA.
Daniel Ravicher, executive director
of PUBPAT, told The Lancet he had
read Myriads response and he was
surprised because the company
seemed to have given up on the
method claims because it had not
sought Supreme Court review of the
appeal courts invalidation of them. As
many in the biotechnology community
will tell you, those method claims are
likely much more signicant than the
isolated gene claims because to do

sequencing today, one need not isolate


specic genes, said Ravicher. As for
the isolated gene claims, they attempt
to dene them as having extremely
narrow scope, which is contrary to both
their assertions of those patents and
the patent language itself.

Many scientists believe this


case to be about critical
questions of patient care and
social policy...
A spokeswoman for Myriad, speaking
before the company had led its
Supreme Court response, told The Lancet
that it had never denied, opposed, or
impeded any research studies on the
gene sequences and that it actively
collaborates with researchers on studies
relating to them.
Myriad provided research testing
services at a fraction of the commercial
testing price to researchers conducting
research funded by NCI [National
Cancer Institute] or another institute
under the National Institute of Health,
said Rebecca Chambers, Myriads
director of investor relations and
corporate communications. During
this programme, 178 scientists received
the discounted research testing services
and 5932 individuals were tested for
BRCA mutations. She also pointed
out that around 20% of the human
genome had already been patented
by drug and biotech companies and
academic institutions. The US Patent
and Trademark Oce (USPTO) has
been awarding these types of patents
since 1980.
Ravicher said if the Supreme Court
decided in the next few months to
take the case, it would probably hold
hearings at its next sitting starting in
October. He said the court only ruled
on about 60 cases a year and took up

just 1% of all petitions but he believed


his plaintis had a good chance. Many
scientists believe this case to be about
critical questions of patient care and
social policy. The decisions made in the
lower court were highly fractious so this
is precisely the kind of case the Supreme
Court should hear, he said. Myriad has
been trying to get patents all over the
world but theyve been challenged in
Australia while other jurisdictions are
also more sceptical. So this is not just a
problem localised in the US.
The ACLU and PUBPAT originally led
the lawsuit on behalf of a coalition of
professional organisationsled by the
Association for Molecular Pathology
(AMP) and representing more than
150 000 physicians, scientists as well as
patientsagainst Myriad, some licenceholders, the University of Utah Research
Foundation, and USPTO.
Among their arguments, they
challenged the patents on constitutional grounds, saying they
encumbered scientic inquiry and
medical care and violated sections of
the rst amendment, which outlines
basic US freedoms.
Myriads patents gave it the right
to exclude others from sequencing
the BRCA1 or BRCA2 gene sequences
or doing other diagnostic tests. That
pathologists can be excluded from
looking at or reading a patients DNA
sequence to characterise or assess the
risk for disease is akin to prohibiting a
physician from taking a patients pulse
to see if his or her heart is beating,
said Mary Steele Williams, executive
director of AMP. I think that the fact
that patients can be prevented from
accessing the information contained in
their DNA would oend most peoples
conceptions of individual rights and
personal liberty.

Sharmila Devi
www.thelancet.com Vol 379 January 28, 2012

World Report

Anticoagulant loses its lustre

The US Food and Drug Administrations (FDA) approval of Boehringer


Ingelheims dabigatran for stroke
prevention in patients with atrial
brillation in October, 2010, was a
watershed event for cardiologists. It
marked an end to a 50 year hunt for
a replacement for warfarina widely
used but dicult to dose drugand
promised a new era in anticoagulant
therapy. Uptake was swift, and the
drug rapidly headed for blockbuster
status. But a string of safety ags,
including most recently data
suggesting the drug increases the
risk of heart attacks, have tempered
enthusiasm for the direct thrombin
inhibitor.
Given the widespread use of the
drug, these signals have to be taken
seriously, says Steven Nissen, a
cardiologist at the Cleveland Clinic,
OH, USA, and a member of the FDA
advisory panel that unanimously
recommended approval for dabigatran. But for now, he adds, further
analysis of the drugs risks and
benets is needed rather than a shoot
from the hip decision. A watch, wait,
study, and review approach is the
right approach.
The latest blow comes from a metaanalysis, published in the Archives of
Internal Medicine, that suggests the
drug is associated with an increase in
acute coronary events. Cardiologists
Ken Uchino and Adrian Hernandez,
both from the Cleveland Clinic, pooled
the results of seven randomised
controlled trials of the drug versus
various comparators and found that
patients treated with dabigatran
had a 2733% higher relative risk of
myocardial infarction.
This reminds us that we need to
keep the risk of myocardial infarction
in mind when using the drug, says
Uchino. A rst report of the phase 3
www.thelancet.com Vol 379 January 28, 2012

data from the RE-LY trial, published


in 2009 in the New England Journal of
Medicine, which paved the way for the
drugs approval, similarly suggested
that dabigatran was associated
with an increased risk of myocardial
infarction, although this nding
lost statistical signicance once
additional data were included in the
analysis. Despite the new reminder
of possible cardiotoxicity, Uchino
adds that the dierence in absolute
risk (014017%) of myocardial
infacrtion remains small and the
drugs overall benet to risk prole
seems to remain benecial.

Ive been in this business for


long enough to know that
when a new drug comes out
you need to take it on slowly...
You may nd out new things
about it that no one knew
when it was approved.
Last year, the drug also faced
headwinds as a trickle of case studies
highlighted the need for caution
when using it in elderly patients, who
are amongst the most likely to suer
from increased bleeding associated
with anticoagulants. A larger setback
also came as regulators in Australia,
Japan, and Europe warned doctors
about serious bleeding-related sideeects. The FDA followed suit in
December, 2011, noting that it is
investigating the incidence of serious
and fatal bleeding.
Although safety ags merit further
study, says cardiologist Mori Krantz,
of the Denver Health Medical Center,
CO, USA, they will not as yet aect his
prescribing habits. The jewel has lost
its lustre, but its still a jewel, he says.
Krantz also sits on the FDAs advisory
panel for heart drugs.

Boehringer Ingelheim, of course,


stands behind its drug. A reanalysis
of the RE-LY data focusing on
myocardial infarction, published in
Circulation only days before Uchinos
meta-analysis, did not nd an
increased incidence of the ischaemic
events. The company also says it has
completed a more comprehensive
meta-analysis that shows the drug is
safe, which it plans to submit for peer
review later this year. With regard
to the regulatory reminders about
serious bleeding, the incidence is still
lower than would have been expected
from the results seen in the pivotal
RE-LY trial, says Reinhard Malin, a
spokesman for Boehringer Ingelheim.
While
Boehringer
Ingelheim
and regulators dig deeper into the
data, the questions raised by the
safety ags are a boost to rival drug
developers with new anticoagulants
of their own. Late last year, both the
FDA and the European Medicines
Agency approved Bayers factor Xa
inhibitor rivaroxaban to prevent
stroke in patients with atrial
brillation. Bristol-Myers Squibb and
Pzers factor Xa inhibitor apixaban
also yielded promising results in
clinical trials, and has been led for
regulatory review in the same setting
in both the USA and Europe.
But given this dynamic landscape,
and the time it can take for the
hype surrounding some new drugs
to settle, Nissen sounds a general
note of caution over the novel
anticoagulants. Ive been in this
business for long enough to know
that when a new drug comes out
you need to take it on slowly, says
Nissen. You may nd out new things
about it that no one knew when it
was approved.

Science Photo Library

Safety ags for Boehringer Ingelheims antiblood clotting agent dabigatran serve as a reminder
about the risks of rapidly adopting newly approved drugs. Asher Mullard reports.

For Uchinos meta-analysis see


www.ncbi.nlm.nih.gov/
pubmed/22215856
archintermed.2011.1666
For more on the RE-LY trial see
N Engl J Med 2009;
361: 1139-1151
For the Circulation paper see
http://www.ncbi.nlm.nih.gov/
pubmed/22215856

Asher Mullard
301

World Report

Greeces nancial crisis dries up drug supply


The Greek Governments austerity drive has inadvertently triggered problems with the countrys
drug supply, causing shortages of hundreds of medicines. Eva Karamanoli reports from Athens.

AP

Greeces health sector seems to be


sinking day by day, without any sign of
recovery. The beginning of 2012 found
patients queuing outside pharmacies
to get their prescriptions lled. Instead
of holding receipts of their social
service funds in their hands, they were
holding money. We need to pay in
cash and then, after months, get the
money from social security, says
Aggeliki Stavrou. Still, its not certain
that they will get the drugs needed.
In the past few months, according
to the Panhellenic Association of
Pharmacists, 500 commonly used
drugs are in short supply. Among
them, there are drugs for hypertension, gastroenterological disorders, cancer, kidney diseases, and
painkillers. Pharmacists often turn
to each other to exchange medicines
to cover their clients needs. Every
day we come across red codes in our
computers in a variety of medicines
most of which arent in stock, or their
availability is very low. We try to serve
our clients by exchanging medicines
among colleagues. Last week, I gave
to a pharmacy in Perama Madopar
(used in the treatment of Parkinsons
disease) and they gave me Keppra
(used to treat epilepsy), says Georgia

Greek patients queue outside a pharmacy during the pharmacists strike in January

302

Fanourakis, a pharmacist in Pireaus,


implying that there are times when
even aspirin is in short supply.
A few months ago, the only reason
for a patient being unable to get the
medicine prescribed would be that the
pharmacists were on strike. Now the
drugstores are open but they are out
of stock. Greeces drug supply is drying
up because of several reasons, all
linked to the countrys nancial crisis.

...international drug companies


prefer to sell their products in
countries other than Greece,
due to the nations nancial
situation.
Last year, the government mandated
lower drug prices to cut down its
medical expenses. Medical expenses
were 24 billion in 2004, 52 billion
in 2009, and dropped to 165 billion
in 2011, after the measure was taken.
However, the governments decision
has fed a secondary market since
wholesalers prefer to sell their products
in other countries where the prot
is higher. The situation gets more
dicult because of how the Greek
health system operates. Social service
funds, according to the latest update,
owe pharmacists 400 million. As a
consequence pharmacists are unable
to pay their suppliers and thus begins
an endless cycle. That is the reason
that pharmacists announced in midJanuary that patients will be provided
with medicine only if they pay in cash.
But that is also dicult because there
are people that cannot aord to do so.
In addition to the secondary market,
and the social security blockage,
international drug companies prefer
to sell their products in countries
other than Greece, due to the nations
nancial situation. Also the companies

that import medicine demand to be


paid cash which is not possible nowadays. Greece is nancially unsafe
for the international pharmaceutical
companies,
says
Constantinos
Lourantos, president of the Attica
Pharmacists Association, implying that
it is impossible for the pharmacists
to pay their suppliers in cash. Due to
medicine shortage, the populations
health is in great danger, adds
Kostas Kouvaris, president of Pireaus
Pharmacists Association. He blames
the wholesalers and the drug makers,
who demand to be paid in cash in order
to supply the pharmacists. Among
12 000 pharmacies that operate,
800 are about to close and 200 have
severe operational diculties. The
measures takenserving prescriptions
paid in cashare absolutely necessary,
says Theodore Abatzoglou president
of the Panhellenic Association of
Pharmacists. He points out that the
recent government law for the prices
of medicine is inadmissible. It is the
same law that also caused hospitals
to run out of drug stocks. Because of
the extremely low prices set by the
government, international companies
are not interested in supplying Greek
hospitals with medicine. Health
institutes are now mainly being
supplied by companies that sell
generic drugs at lower prices.
In mid-January, the Ministry of
Health, due to patients being unable to
get the drug prescribed, re-introduced
a discount that pharmacists get from
the wholesalers and promised that a
percentage of the debt to pharmacists
will be paid. These changes managed
to end the pharmacists mobilisation
and, for now, prescriptions are served
as per usual, that is if the drugs are
available.

Eva Karamanoli
www.thelancet.com Vol 379 January 28, 2012

Perspectives

Film
Beats and the brain

www.thelancet.com Vol 379 January 28, 2012

a lm about Parkinsons
disease seen through the prism
of music
Ray Chaudhuri at Londons Kings
College Hospital and Oxford neurosurgeon Tipu Aziz. Chaudhuri proposes
an apomorphine pump, Aziz discusses
deep brain stimulation and the future
potential of stem cells. Barbara opts for
the pump, but it takes a year of ghting
through the bureaucracy and hassles
over who will pay before she can return
to Kings to have the pump tted in
2010. As the apomorphine diuses
through her body, her body relaxes and,
for a time, she is able to go on tour with
Jon and his band, Colosseum.
But this isnt a lm about the clinical
course of Parkinsons disease, or the

potential drugs and surgery. Rather,


it is about a remarkable woman and
her husband working to come to
terms with the disease and continue
with their lifes workmaking music.
We watch the nurse specialist at
Kings College Hospital patiently
training the two of them to set up the
apomorphine perfusion. The couple
share the arduous daily routine of xing
the pump and syringending an
injection site in Barbaras increasingly
recalcitrant bodyin anonymous hotel
bedrooms as they tour. We also see, in
stark contrast, the musician playing on
stage in front of a huge and adoring
audience in Vienna.
Above all, this is a lm about a team
Barbara and Jon, doctor and nurse, and
the lm makerviewed through the
prism of Dibbs own ways of seeing.
There is one moment that encapsulates
these relationships, when, after Barbara
has been answering questions from the
nurse about whether she hallucinates,
Jon jokes: Yes she does; she probably
imagines theres a lm crew in the room
with us right now. And with this the
viewer, too, is invited into the room.

Jon Hiseman Temple Music

functions perfectly, and a choir for


patients and their carers, Sing for Joy.
Her own experience as a patient is
portrayed without sentimentality: we
see her exhausted after performances,
scarcely able to stand, working against
the adverse eects and diminishing
benets of the drugs, and, with Jon,
visiting doctors and nurses in search
of treatment. She consults neurologist

Playing Against Time: A Film


about Parkinsons Disease
and Music
Directed by Mike Dibb, with
Barbara Thompson and
Jon Hiseman. Dibb Directions
Production for The Wellcome
Trust. To be broadcast on
Feb 19, 2012, on BBC4
See Editorial page 287

Hilary Rose, Steven Rose


S.P.R.Rose@open.ac.uk

Jon Hiseman Temple Music

The Wellcome Trust has a long


track record of fostering science-art
collaborations, but Playing Against Time
must be one of the best. It focuses on
the life and music of Barbara Thompson,
one of Europes nest jazz saxophonists,
who in 1996 was diagnosed with
Parkinsons disease. Since then
she has continued composing and
performing with the constant support
of her partner, jazz/rock drummer
Jon Hiseman. Film maker Mike Dibb
describes Playing Against Time as a lm
about Parkinsons disease seen through
the prism of music, and central to his
lm is Barbara and Jons collaboration
as wife and husband, patient and carer,
saxophonist and drummer.
To portray Parkinsons disease in
the way Dibb describes requires a lm
maker his subjects know they can trust.
Dibbs earlier credits include the BBC
series Ways of Seeing, with the artist and
critic John Berger, and a prole of the
trumpeter and composer Miles Davis.
In 1979 he made Jazz, Rock and Marriage
with Barbara and Jon, featuring Barbara
playing with her band Paraphernalia.
Playing Against Time picks up the story
in 2005, and follows the musicians on
tour over the succeeding 5 years, as
Barbaras Parkinsonian tremors worsen.
Parkinsons disease is cruel, but
through Barbaras determination
it has become almost miraculously
incorporated into her playing. The
saxophonist sways to the rhythm of
her music, and it is as if she is riding
the disease rather than it is riding her.
Flashbacks to the earlier lm and the
constant use of split screen, however,
make the changes over the years
painfully visible.
Barbara engages with other people
with Parkinsons disease and patients
groupsduetting with a chartered
accountant turned campaigner who
sings about the disease, a jazz pianist
whose left hand can no longer press
the keys but whose right hand still

303

Perspectives

Book
Joseph Rotblats conscientious life in science and politics

Keeper of the Nuclear


Conscience: The Life and Work
of Joseph Rotblat
Andrew Brown.
Oxford University Press, 2012.
Pp 327. 1899
ISBN 9780199586585

304

To the world at large, the late


Joseph Rotblat is known as the Polishborn, British-based, scientist who
worked on the Manhattan Project to
build the atomic bomb, and for the
rest of his life campaigned for the
abolition of nuclear weapons. He, and
the unocial federation of European,
American, Soviet, and other scientists
that he founded, known as Pugwash,
were jointly awarded the Nobel
Peace Prize in 1995. To physicians,
Rotblat is also known as a physicist
who specialised in understanding the
eects on living tissue of radiation and
radioactive fallout and in developing
nuclear medicine, through his longheld professorship of physics at
St Bartholomews Hospital in London.
Keeper of the Nuclear Conscience,
the rst major biography of Rotblat,
naturally devotes more space to the
rst, political, role than to the second,
medical, one. But since the books
author, Andrew Brown, is himself
a physician who practised for two
decades as a radiation oncologist,
Rotblats lifelong interest in medicine
runs as a thread through the narrative.
Indeed, Brown is ideally matched with
his subject, having previously written
biographies of two scientists heavily
involved in the cold war: the nuclear
physicist James Chadwick (a key
inuence on Rotblat), and the biologist
and Communist J D Bernal. The result is
a thorough study of a modest, humane,
and public-spirited individualbased
on Rotblats personal papers and the
Pugwash archives, as well as personal
interviews with Rotblats circle.
War formed Rotblat. As a boy Jo
experienced the fear, poverty, and
disruption of war, and he never
forgot them, writes Brown. Rotblat
himself, interviewed at the age of 91
for the British Library Sound Archive,
shied away from any discussion
of his childhood in Warsaw during
World War 1. By 1918, he said, his

once-prosperous Jewish family was


reduced to a state of complete and
abject penury, real hunger, disease
and squalor. For the rest of his life,
he rejected potatoes because they
reminded him of the bitter taste of
frost-damaged tubers in his wartime
diet. There was no money to complete
his schooling and he was forced to
work as an electrician, while attending
Warsaws Free University at night. He
left Poland for Britain in 1939, just in
time to avoid the Nazi invasion.

Rotblats lifelong interest in


medicine runs as a thread
through the narrative
During World War 2, Rotblat worked
on a meagre salary as a nuclear
physicist in heavily bombed Liverpool,
while his desperate family in Poland
struggled to survive Hitlers Holocaust;
his wife was an early victim, although
he did not learn her probable fate until
the autumn of 1945. Recruited by
Chadwick to Los Alamos in early 1944,
within months he learned that the
Nazi project to build an atomic bomb
had failed. He therefore resigned from
the Manhattan Projectthe only
physicist to do so on ethical grounds
and returned to Liverpool. That
decision brought him to the attention
of the Federal Bureau of Investigation
as a potential spya trunk containing
his personal and scientic papers
mysteriously vanished from the train
carrying Rotblat from Washington,
DC, to New York.
The dropping of the atomic
bombs in August, 1945, and grief
at the death of his wife, brought
about a personal crisis. Rotblat
decided not to return to Poland; to
move into medical physics, so as to
investigate techniques that might
be of immediate benet to human
beings; and to devote his energy

to ensuring that the atomic bomb


would never again be used in a war.
He even considered qualifying as
a physician, but was advised by a
medical collaborator that it would
take too many years to attain a senior
position. From 1950, when he moved
to St Bartholomews Hospital, Rotblat
began to establish a talented team
of medical physicists to undertake
research in medical electronics and
the biophysics of living cells, and
to invent new ways to monitor
patients undergoing surgery. But
there were continual administrative
problems: the atmosphere surrounding [Rotblats] department at
Barts was one of constant tension
spiked with pettiness, notes Brown,
without fully explaining the source
of the disagreement. No doubt part
was the friction inevitable between
established disciplines compelled
to work closely together. However,
Rotblats driving personality may also
have been a factor. A British Pugwash
admirer told Brown that Rotblat
resembled a great Russian general,
with the qualities of considerable
charm combined with ruthlessness in
a good cause, amazing stamina, and a
tendency to megalomania.
Rotblats
stamina
emanates
from each page of this engrossing
biography. When he was 95, he
wrote a vigorous piece for a book of
mine on Albert Einstein, concerning
Einsteins search for world peacean
inspiration for Pugwash in 1955.
Well into his eighties, Rotblat would
run down escalators in the London
Underground. Once, a group of idling
teenagers was in the way and told
him: Whats the hurry? Youre an old
man. An impatient Rotblat replied:
Dont you see? Its because Im an old
man that Im in a hurry.

Andrew Robinson
andrew.robinson33@virgin.net

www.thelancet.com Vol 379 January 28, 2012

Perspectives

Prole
Ndola Prata: ghting for womens reproductive health
Ever since I was a little girl, I always wanted to x things,
says Ndola Prata, Scientic Director of the Bixby Center for
Population, Health and Sustainability at the University of
California, Berkeley. Prata battles daily for more pragmatic
approaches to reproductive and sexual health care. I nd
myself questioning why we dont focus on interventions
that can be scaled up, thus reaching most of the women
in need, she says. Her conviction that its worth ghting
for what you believe in comes from her parents, she says.
Growing up in conict-ridden Angola in the 1970s wasnt
easy. Once Portugal acceded independence in 1975, some
Angolan citizens left but Pratas parents resolutely stayed
put, believing that Angola would become a great country.
Prata was already well versed in the challenges of
reproductive health in her country by the time she went
to medical school. In my extended family you couldnt go
5 minutes without someone talking about a problem related
to reproductive health, she says. Although she initially
wanted to be a neurosurgeon, her trademark practicality
kicked in, and she realised that Angola would be better
served if she tried to address its high fertility rates and poor
reproductive care. By the late 1990s, Prata had practised
medicine in Angola for a decade and had also completed
an MSc in medical demography from the London School of
Hygiene and Tropical Medicine in the UK. In 1998, she moved
to the School of Public Health at the University of California,
Berkeley, where she has stayed ever since, combining her
work for the Bixby Center with her role as Associate Professor
of Maternal and Child Health. Prata was fairly certain her
move to the USA would be long termshe fell in love with
an American scientist, who she would later marry.
Much of Pratas time, she says, is spent ghting for
realistically aordable methods of improving reproductive
health to be used more widely. Her battle lines are often
etched on that tricky demarcation between care that is ideal
and that which is good enough. In this, she says, I often
clash with proponents of gold standards of care. Some
interventions have such high standards, that they are just not
achievable on a large scale in resource-poor settings within
an acceptable timeframe. This practical streak is partly a
legacy, Prata says, of coming from a country like Angola
you do what you can with what you have. Community
health workers, such as traditional birth attendants, for
instance, can be useful in improving access to maternal care,
she says, but their use has stirred controversy. They play
an important role in family planning and maternal health
because they are there with the women during pregnancy
and birth. Like it or not, scientists need to engage with
these workers to ensure their advice is medically sound. Not
acknowledging their role is more damaging, she argues.
www.thelancet.com Vol 379 January 28, 2012

Pratas insistence on a real-world approach to research in


low-income countries can lead to some dicult questions.
In some situations it is possible, and even necessary, to
devise treatment policies without extensive randomised
trials, she argues. Prata and her colleagues showed that the
drug misoprostol could control post-partum haemorrhage
in low-resource settings, but it took two applications and
4 years before WHO included misoprostol in its Essential
Medicine List for post-partum haemorrhage. Data showed
that oxytocin is more eective than misoprostol; yet
oxytocin is not available for most poor women in developing
countries. Since misoprostol is also eective, it didnt seem
to make sense to wait so long, Prata told The Lancet.
Prata is by no means dismissive of the importance of good
evidence, however. Between 2001 and 2006, and again
in 2010, she worked concurrently at Berkeley and for the
US Centers for Disease Control and Prevention (CDC). At
the CDCs Division of Reproductive Health she did national
surveys in developing countries to gather evidence of the
impact of reproductive health programmes. But she also
points out, Yet ne-tuning the knowledge base through
extensive trials can have serious consequences for people
desperate for treatment. It can get to a point where it is no
longer ethical. Sometimes you have to use the information
that exists. Doing nothing might be more detrimental.
At the Bixby Center, she works on family planning, which
she thinks is one of the greatest unmet needs in developing
countries, in places such as Ethiopia and Mozambique.
Better access to contraception is required, but also an
understanding of factors that can improve maternal health,
such as spacing births, she says. Prata recounts how one
woman told her during a recent eld trip to Africa that It was
a really good year, because this year, I didnt get pregnant.
Pratas current research focus is on task shifting in maternal
health, nancing mechanisms to scale up community-based
contraceptives, and community-based systems to track
maternal mortality. Her insistence on pushing against
conventional thinking has made for some dicult times, and
she relies on the support of mentors like Professor Malcolm
Potts, former Medical Director of the International Planned
Parenthood Federation, who set up the Bixby Center. Potts
says that One of the greatest joys in my professional life in
the past 10 years has been to see Ndolas explosive impact
on family planning and reproductive health in low-resources
settings. Prata cant ever imagine settling for a quiet life. I
realised early on that I had a slightly dierent way of thinking
from my peers. I always feel a great responsibility to be vocal
and try to nd solutions, rather than just complain.

Priya Shetty
305

Comment

O that moral science were in as fair a way of improvement,


that men would cease to be wolves to one another, and
that human beings would at length learn what they now
improperly call humanity!
Benjamin Franklin, 17801

Are human volunteers in research sponsored by the


US Government treated safely and ethically? President
Barack Obama posed this fundamental question to the
Presidential Commission for the Study of Bioethical
Issues (Bioethics Commission) in response to chilling
accounts of unethical research undertaken by the US
Public Health Service in Guatemala during the 1940s,
in which vulnerable populations were deliberately
exposed to sexually transmitted diseases without their
knowledge or consent.24
In a report released on Dec 15, 2011, the Bioethics
Commission concluded that the current US system
provides substantial protections for the health, rights, and
welfare of research subjects and protects them from harm
or unethical treatment.5 Although condent that the
current system aords substantial protections to research
volunteers, the Commission identied 14 specic changes
to current policies that could further reduce the likelihood
of harm or unethical treatment. These recommendations
include development of a compensation system for
research participants who are injured during research
studies; respecting equivalent ethical standards in foreign
countries where US-funded research is conducted; and
more explicit justications for site selection to prevent
potential exploitation of foreign research volunteers.
As noted by the International Research Panel
that advised the Bioethics Commission, the USAunlike
many countrieslacks a comprehensive system to address
potential harms connected to research participation.6
Thus some research volunteers are left without any
assurance of protection from personal nancial and
physical risk when they take part in research that benets
society as a whole. Although current US regulations
require investigators to inform research participants
about any medical care or compensation to be provided
in the event of research-related injuries,7 no rule requires
free medical care or compensation to be provided. In
this regard, the US regulations dier from the research
www.thelancet.com Vol 379 January 28, 2012

injury compensation systems that exist in most other


industrialised nations.8 The Commission recommends
that the US Government adopt a carefully designated set
of standards for compensation.5
Research supported by the US Government is subject
to the same regulatory requirements domestically and
internationally. These requirements permit government
agencies to recognise foreign procedures if they provide
protections that are at least equivalent to federal
rules.9 For example, procedures for informed consent
in the UK might be deemed equivalent protections
to those required by the US Government. However,
US federal agencies have almost never exercised this
authority and often insist that all US rules be met even
in foreign countries where protections are equal to, or
more stringent, than those in the USA.10 The Commission
recommends that the federal government develop a
process for evaluating requests from foreign governments
and other non-US institutions to determine if local laws
and procedures can be recognised as providing equivalent
protections to research participants.
The increasingly globalised nature of medical research
presents ethical challenges when study sites do not oer
robust protections for volunteers or if the research
subjects [are]being systematically selected because of
their easy availability, their compromised position, or
their manipulability.11 The Commission recognised the
potential for exploitation in low-income communities
and countries and recommends that researchers and the
proposed research sites demonstrate the capacityor
the ability to achieve the capacity when the research is
to be conductedto protect all human participants.5
In addition, the Commission recommends that the US
Government develop criteria for ethical selection of sites,
taking into consideration the extent to which sites can
benet the broader community. The Commission found
that many of the risks of exploitation can be mitigated,
for example, when a study is designed specically to
respond to the health needs of the local community.5
Good science requires good ethics, and vice versa. To
do good science, individual investigators, along with
the institutions that support and sponsor them, must
assume responsibility to protect participants from

Presidential Commission for the Study of Bioethical Issues

Moral science and the Presidential Commission for the Study


of Bioethical Issues

Published Online
December 16, 2011
DOI:10.1016/S01406736(11)61884-9

e20

Comment

avoidable harm and unethical treatment and to respect


their inherent dignity. Benjamin Franklin1 reminds us
of the long-standing ethical principles at the heart of
modern laws, rules, and regulations that govern research
involving human beings. The challenge of pursuing
science in a morally justied way is one that every
generation must take up. Without sucient attention to
ethics, we risk losing sight of what is owed to our fellow
human beings who deserve special protection by virtue
of their willingness to participate in experiments so as
to benet others. The US Government should respond
swiftly to the recommendations of the Bioethics
Commissions report, or oer compelling reasons for
maintaining the status quo.

4
5

*Amy Gutmann, James W Wagner


Presidential Commission for the Study of Bioethical Issues,
Washington, DC 20005, USA
president@upenn.edu

10

AG is Chair and JWW Vice Chair of the Presidential Commission for the Study of
Bioethical Issues.

11

1
2

Franklin B, Sparks J. The works of Benjamin Franklin, vol VIII. Boston:


Hilliard, Gray, and Company, 1839.
Presidential Commission for the Study of Bioethical Issues. Ethically
impossible STD research in Guatemala from 1946 to 1948. Sept, 2011.
Washington, DC: Presidential Commission for the Study of Bioethical
Issues, 2011.

US Department of Health and Human Services. Information on the


19461948 United States Public Health Service STD Inoculation Study.
http://www.hhs.gov/1946inoculationstudy (accessed Dec 12, 2011).
Reverby S. Normal exposure and inoculation syphilis: a PHS Tuskegee
doctor in Guatemala, 19461948. J Policy History 2011; 23: 627.
Presidential Commission for the Study of Bioethical Issues. Moral science:
protecting participants in human subjects research. Dec 15, 2011.
Washington, DC: Presidential Commission for the Study of Bioethical
Issues, 2011.
Presidential Commission for the Study of Bioethical Issues. Research across
borders: proceedings of the International Research Panel of the Presidential
Commission for the Study of Bioethical Issues. Sept, 2011. Washington, DC:
Presidential Commission for the Study of Bioethical Issues, 2011: 11.
US Department of Health and Human Services. Code of Federal
Regulations. Basic HHS Policy for Protection of Human Research Subjects.
Revised Jan 15, 2009. Eective July 14, 2009. 45 CFR 46.116.
http://www.hhs.gov/ohrp/humansubjects/guidance/45cfr46.html
(accessed Dec 12, 2011).
Institute of Medicine. Responsible research: a systems approach to
protecting research participants. Washington, DC: National Academies
Press, 2003.
US Department of Health and Human Services. Code of Federal
Regulations. Basic HHS Policy for Protection of Human Research Subjects.
Revised Jan 15, 2009. 45 CFR 46.101(h). http://www.hhs.gov/ohrp/
humansubjects/guidance/45cfr46.html#46.101 (accessed Dec 12, 2011).
Equivalent Protections Working Group. Report of the Equivalent
Protections Working Group. July 17, 2003. Washington, DC: US Department
of Health and Human Services, 2003. http://www.hhs.gov/ohrp/
international/epwgreport2003.pdf (accessed Dec 12, 2011).
The National Commission for the Protection of Human Subjects of
Biomedical and Behavioral Research. The Belmont report: ethical principles
and guidelines for the protection of human subjects of research.
Department of Health, Education, and Welfare Publication OS 78-0012.
Washington, DC: Department of Health, Education, and Welfare, 1978: 10.

Tackling the spread of drug-resistant tuberculosis in Europe


Published Online
September 15, 2011
DOI:10.1016/S01406736(11)61428-1

e21

Multidrug-resistant (MDR) tuberculosis and extensively


drug-resistant (XDR) tuberculosis have become an
important health problem in many countries of the WHO
European region and currently threaten global eorts
to control tuberculosis.14 About 81 000 (184%) of the
440 000 patients worldwide with MDR tuberculosis live
in this region.2 The highest rates occur predominantly
in eastern Europe; however, population movement
means that drug-resistant tuberculosis is a priority
public health issue for all European countries.35 The
spread of MDR and XDR tuberculosis is driven by a
complex interplay of factors, including irrational
drug policies and social and clinical determinants,
such as HIV infection, imprisonment, migration, and
socioeconomic factors. In some parts of Europe, there
is evidence of poor tuberculosis programmes with low
case detection, notication, and treatment completion
rates compounded by suboptimum management.3,6,7

The greatest fear is that drug-resistant strains of


Mycobacterium tuberculosis could become the dominant
causes of tuberculosis in Europe.8
The consolidated action plan to prevent and combat
multidrug- and extensively drug-resistant tuberculosis in the
WHO European Region 20112015 is a roadmap designed
to guide the national and international response to
this ominous situation.9 The WHO action plan focuses
on six strategic directions: tackling determinants of
emergence and spread; addressing issues that relate
to the health system; working through partnerships;
managing and increasing resources; collaboration to
develop new diagnostics, medicines, and vaccines; and,
nally, improved surveillance and monitoring. The plan
also outlines seven intervention groups, each supported
by best practice examples, to be promoted by national
governments; which, if adopted, would lead to universal
access to diagnosis, treatment, and prevention of drugwww.thelancet.com Vol 379 January 28, 2012

resistant tuberculosis. The ultimate goal of eective


implementation of the plans recommendations is the
prevention of 250 000 new cases of MDR tuberculosis
and 13 000 new cases of XDR tuberculosis, as well as the
diagnosis of 225 000 patients with MDR tuberculosis to
ensure that 127 000 of these patients are successfully
treated, thereby preventing transmission.
In 2007, a European Ministerial declaration in Berlin
called for action against tuberculosis, but also raised
the issue of advocacy without provision of adequate
nancial resources and technical expertise to support
implementation of specic recommendations. In 2009
in Beijing, health ministers from 27 countries with a high
burden of MDR tuberculosis called for renewed action. This
was followed by World Health Assembly Resolution 6215
that called for universal access to diagnosis and treatment
for drug-resistant tuberculosis.10 Ambitious action plans
may not always be backed up with adequate resources. The
successful reversal of the epidemic of MDR tuberculosis in
New York during the early 1990s was achieved through
planning, adequate funding, political will, and admirable
leadership.11 There are examples of successes within Europe,
such as the centralised national tuberculosis programmes
in Latvia and Estonia. The challenge remains as to whether
these successes can be replicated in larger countries
through tackling factors such as those that facilitate
transmission in vulnerable groups, inadequate housing,
and provision of appropriate tuberculosis health services.
In England, a tuberculosis action plan was launched in
2004,12 yet although levels of drug-resistant disease have
remained stable, incidence rates of tuberculosis have
continued to rise.
WHOs action plan targets national tuberculosis
programmes, but will only succeed if it also tackles the
problems of inadequate drug susceptibility testing facilities,
poor availability and misuse of second-line tuberculosis
drugs, and the need for improved management and
follow-up services. With the lowest cure and completion
rate in the world,3 Europe should urgently address the
unacceptably low treatment success rates in the region.
However, there are reasons for being cautiously
optimistic about the prospect for control of MDR
and XDR tuberculosis in the region. First, before the
development of the WHO plan, WHO-Europe had assisted
countries with a high burden of MDR tuberculosis
within the region to develop national response plans.
Second, the plan includes a monitoring framework with
www.thelancet.com Vol 379 January 28, 2012

Bloomberg via Getty Images

Comment

11 core indicators that are designed to measure progress


towards decreasing the proportion of MDR tuberculosis
in previously treated patients by 20%, increasing MDR
case detection to 85%, and successfully treating at
least 75% of patients with MDR tuberculosis by 2015.
Third, intermediate milestones are outlined that should
allow national and international stakeholders to hold
governments and WHO to account. All the targets in
the plan relate to controlling acquired resistancean
approach resulting from pragmatism about what can
be achieved within the timeframe. Although this is an
appropriate short-term measure and will indirectly avert
spread, the eect of the plan on transmission should also
be assessed with medium-term and long-term objectives
targeted towards achieving global control of tuberculosis.
The current epidemiological picture of drug-resistant
tuberculosis in Europe is daunting, and WHOs ambitious
plan must be implemented if we are to stem this tide.
*Ibrahim Abubakar, Masoud Dara, Davide Manissero,
Alimuddin Zumla
Tuberculosis Section, Health Protection Agency, London NW9 5EQ,
UK, and Norwich Medical School, University of East Anglia, Norwich,
UK (IA); WHO Regional Oce for Europe, Copenhagen, Denmark
(MD); European Centre for Disease Control and Prevention,
Stockholm, Sweden (DM); and Division of Infection and Immunity,
University College London Medical School, London, UK (AZ)
ibrahim.abubakar@hpa.org.uk
IA chaired the WHO International Task Force that developed the monitoring
framework for the WHO plan and MD wrote the rst draft of the WHO plan.9
DM and AZ declare that they have no conicts of interest.

e22

Comment

2
3

e23

Gandhi NR, Nunn P, Dheda K, et al. Multidrug-resistant and extensively


drug-resistant tuberculosis: a threat to global control of tuberculosis.
Lancet 2010; 375: 183043.
WHO. Global tuberculosis control 2010. http://www.who.int/tb/
publications/global_report/2010/en/index.html (accessed Sept 5, 2011).
European Centre for Disease Prevention and Control/WHO. Tuberculosis
surveillance in Europe 2009. http://www.ecdc.europa.eu/en/publications/
Publications/Forms/ECDC_DispForm.aspx?ID=660 (accessed Sept 5, 2011).
Fears R, Kaufmann S, Ter Meulen V, Zumla A. Drug-resistant tuberculosis in
the European Union: opportunities and challenges for control.
Tuberculosis (Edinb) 2010; 90: 18287.
Kruijshaar ME, Watson JM, Drobniewski F, et al. Increasing antituberculosis
drug resistance in the United Kingdom: analysis of national surveillance
data. BMJ 2008; 336: 123134.
Wright A, Zignol M, Van Deun A, et al. Epidemiology of antituberculosis
drug resistance 200207: an updated analysis of the Global Project on
Anti-Tuberculosis Drug Resistance Surveillance. Lancet 2009; 373: 186173.
Veen J, de Colombani P. WHO regional oce for Europe. Review of the
national tuberculosis programme in Ukraine. Oct 1022, 2010. http://
www.euro.who.int/__data/assets/pdf_le/0007/142369/e95006.pdf
(accessed Sept 5, 2011).

8
9

10

11
12

Cohen T, Murray M. Modeling epidemics of multidrug-resistant


M tuberculosis of heterogeneous tness. Nat Med 2004; 10: 111721.
WHO European Region. The consolidated action plan to prevent and
combat multidrug- and extensively drug-resistant tuberculosis in the WHO
European Region 20112015. Sept 1215, 2011. EUR/RC61/Inf.Doc/3.
Copenhagen: WHO, 2011.
62nd World Health Assembly. Prevention and control of
multidrug-resistant tuberculosis and extensively drug-resistant
tuberculosis. WHA62.15. 8th plenary meeting, May 22, 2009. A62/VR/8.
http://www.who.int/tb/features_archive/wha62_15_tb_resolution/en/
index.html (accessed Sept 5, 2011).
Frieden TR, Fujiwara PI, Washko RM, Hamburg MA. Tuberculosis in
New York Cityturning the tide. N Engl J Med 1995; 333: 22933.
Department of Health. Stopping tuberculosis in England: Chief Medical
Ocers action plan. London: COI Communications for the Department
of Health, 2004.

www.thelancet.com Vol 379 January 28, 2012

Perspectives

The art of medicine


Shakespeare under water
Eorts to nd the real Shakespeare are fraught with
pitfalls. Although ample documentation exists about
his business dealings, evidence of the mans inner
psychological processesunltered through plays or
poetryis non-existent. Attempts to address the private
life of Shakespeare often result in the critic projecting his
or her own preoccupations onto the Bard and his works.
A case in point is Sigmund Freuds insistence that Hamlet
represented Shakespeares reaction to the death of his
father. So strong was Freuds belief, indeed, that he was
willing to subscribe to the idea that the play had not been
written at the time generally agreed, and the man called
William Shakespeare was not its author.
We do not know if the modern notion of creative writing
as a process of conscious or unconscious personal revelation
even applied in Elizabethan or Jacobean England. Still,
tantalising pieces of circumstantial evidencebiographical
details that may have found their way into the playsemerge
from time to time. The Oxford Dictionary of National Biography
states that a woman by the name of Katharine Hamlet
drowned at Tiddington, near Shakespeares home town
of Stratford-upon-Avon, in 1579 when the playwright
was 15 years old. The parallel with the fate of Ophelia,
Hamlets unfortunate sweetheart who goes mad and dies
by drowning, seems clear. The work of Dr Steven Gunn of
Merton College, Oxford University, UK, sheds more light on
this issue. His analysis of Tudor coroners records showed
that another drowning occurred near Stratford-uponAvon a decade before the death of Katharine Hamlet. The
deceased was a young girl who had fallen into a millpond
whilst picking owers. Her name was Jane Shaxspere:

Was Jane Shaxspere a relation of the then 5-year-old


William? And did the incident pass into Shakespeare family
legend to emerge in Hamlet? These are unanswerable
questions. But the fact that the case was recorded by the
coroners court is certainly of interest. The signicance of the
coroners verdict in cases of unexpected death is reected
in the gravedigger scene in Hamlet. The conversation
between the men who bury Ophelia includes a mangled
description of the chain of logic a coroner would follow: it is
underscored by the fear of a verdict of suicide.
The cause of death mattered to many dierent parties: to
the justice system, to the church, to the crown that may have
been entitled to the assets of the deceased, and, of course, to
the bereaved family. Much was at stake from the coroners
inquest. Tudor coroners records form a valuable record of
causes of accidental death in a predominantly rural society at
a pivotal moment of its social and economic development.
Perhaps the general patterns that emerge from such records
will allow scholars to build up a richer understanding of the
cultural milieu that Shakespeare inhabited.
Gunn tells me that with the funding of the Economic
and Social Research Council, he has begun to examine the
reports on around 9000 accidental deaths, whose details
were preserved for posterity as they were led with other
records sent from the provinces to the Court of the Kings
Bench. There are many colourful causes of death that
are both comic and tragic: cow attacks, exploding stage
props, and maypole accidents. Yet on a larger scale, the
reports provide a way into, as Gunn puts it, what people
were doing all day. Gunn describes, for example, how the
promotion of archery practice as a necessary part of the
loyal Englishmans weekly routine is reected in a large
number of arrow-related deaths.

The National Archives , ref. KB9/625 (241)

By reason of collecting and holding out certain owers called


yellow boddles growing on the bank of a certain small
channel at Upton aforesaid called Upton millpondthe same
Jane Shaxspere the said sixteenth day of June about the
eighth hour after noon of the same day suddenly and by
misfortune fell into the same small channel and was drowned

in the aforesaid small channel; and then and there she


instantly died. And thus the aforesaid owers were the cause
of the death of the aforesaid Jane.

Coroners inquest report on Jane Shaxspere, June 29, 1569

306

www.thelancet.com Vol 379 January 28, 2012

Another big killeraccounting for half of all mortalitywas


drowning. Gunn explains that men and women drowned for
dierent reasons: women would tend to meet with a mishap
whilst fetching water (and the presence of a bucket could
provide the vital piece of evidence to distinguish accidental
death from suicide). Men, on the other hand, often drowned
when bathing after physical work. Travelling across water to
get to and from work or the market could also be perilous.
So even if the cases of Katharine Hamlet and Jane Shaxspere
did not register with the young William Shakespeare, perhaps
drowning in general was a common concern for the average
Tudor man and woman, occupying the same mental space
that death on the roads does in modern Britain. Ben Crystal,
author of Shakespeare on Toast, agrees: I like the analogy of
a car crash. Other than being stabbed or falling illor some
other happenstanceI suppose death by drowning would
have been a similarly major worry. He sounds a note of
caution Its terribly hard to read Shakespeares own opinion
into anything in the canon. He keeps himself quite well
hidden, but adds that Sebastian and Viola in Twelfth Night
consider each other drowned, and Antonios prospects of
business success [in The Merchant of Venice] rest on the sea.
Water may also represent a medium of change in
Shakespeares playsa gateway to another type of existence.
In The Tempest, Prosperos spirit, Ariel, describes the transformation of a drowned corpse into precious treasures:
Full fathom ve thy father lies;
Of his bones are coral made;
Those are pearls that were his eyes;
Nothing of him that doth fade,
But doth suer a sea-change
Into something rich and strange.

A similar device is used in Richard III as Clarence, who will


meet his end submerged in a butt of wine, describes a
dream of falling overboard a ship and yet surviving to
experience a strange underwater world.
Methought I saw a thousand fearful wrecks;
Ten thousand men that shes gnawd upon;
Wedges of gold, great anchors, heaps of pearl,
Inestimable stones, unvalued jewels,
All scatterd in the bottom of the sea...

So drowning is an important part of the Shakespearean


world: both as a cause of death and separation, and as
a more magical, transformative process. But what was
Shakespeares attitude towards accidental death in general?
Gunn tells me that in the 16th century deaths that might
now be considered accidental could be ascribed to divine
intervention, for example, deaths related to maypole
dancing, or the collapse of houses in which heretical
meetings were held. Dr Martin Wiggins, Senior Lecturer
and Fellow at the Shakespeare Institute of the University of
Birmingham, Stratford-upon-Avon, UK, describes how the
gravediggers conversation in Hamlet reects the confusion
facing those trying to make sense of senseless events: The
www.thelancet.com Vol 379 January 28, 2012

Private Collection/Photo Peter Nahum at The Leicester Galleries, London/The Bridgeman Art Library

Perspectives

Ophelia (185152) by Sir John Everett Millais

interesting thing about Ophelias case is that it gives rise


to exactly the sort of rationalisation that doesnt work in
cases of genuine accident. People respond to the death by
wanting it not to have been merely accidental, so that the
gravediggers paint themselves into a corner whereby they
have to choose between suicide and malignant water which
seeks you out and kills you.
There is one more example of accidental death in
Shakespeare that Wiggins brings upthis one not involving
water, but rather a fall from a height. It is not as well known
as Ophelias death, but it is important nevertheless. In King
John, the titular monarchs nephew, Arthur, is saved from
a murder plot when both his uncle and a hired murderer
have second thoughts. Nevertheless, he dies through an
accidental fall from a castle wall. The whole sequence,
Wiggins explains, shows human actions failing to lead to
their intended results. The way the action is structured resists
any attempt to make human sense of it: its completely lefteld. The play was written very close in time to the death
of Shakespeares son, so its also got a lot to do with facing
bereavement and the senselessness of premature death. As
Wiggins says, Shakespeareat a key emotional moment
in his lifedidnt go in for the easy malign universe
rationalisation of accident. Accidental death, as studied
by Gunn, gives an idea of the social and cultural forces that
shaped Shakespeares work; and King John, produced in the
aftermath of a sudden and unexpected bereavement, is a
reminder of his complex and uncompromising nature as an
artist. But the nature of Shakespeare the man remains, as it
always will, submerged far below the surface.

Niall Boyce
The Lancet, London NW1 7BY, UK

For more on Steven Gunn see


http://www.merton.ox.ac.uk/
fellows_and_research/gunn.shtml
For more on Ben Crystal see
http://www.bencrystal.com
For the Shakespeare Institute see
http://www.birmingham.ac.uk/
schools/edacs/departments/
shakespeare/index.aspx

307

University of Rochester Medical Center

Obituary

Robert Ader
Psychologist and co-founder of the eld of
psychoneuroimmunology. He was born on
Feb 20, 1932, in the Bronx, NY, USA, and
died on Dec 20, 2011, in Pittsford, NY, USA,
aged 79 years.
The change that Robert Ader helped initiate in medical science
began with a serendipitous discovery. It happened in the early
1970s, when he and Nicholas Cohen were studying taste
aversion. The researchers had been giving rats a saccharin
solution accompanied by an injection of cyclophosphamide,
an immunosuppressive drug that induces gastrointestinal
upset. When the injections stopped, they found as expected
that the rats had become conditioned to avoid consuming
the sweet solution. To complete the experimental protocol,
they forced the rats to take the saccharin solution using eye
droppers. This is where the surprise arrived.
Ader and Cohen found that some of the animals they had
force-fed with the saccharin later died. The magnitude of the
avoidance response and the mortality rate of the rats was
directly related to the volume of solution consumed. What
could be going on? A hypothesis that seemed reasonable
to me was that, in addition to conditioning the avoidance
response, we were conditioning the immunosuppressive
eects (of cyclophosphamide), Ader said in a 2010
interview. In other words, the taste of saccharin alone was
enough to stimulate neural signals that suppressed the
rats immune systems, just as if they had been overdosed
with the immunosuppressant. Ader and Cohen went on
to conrm this hypothesis in a controlled experiment,
308

showing that the behavioural conditioning process could


suppress immune responses as measured by antibody
concentrationsthereby revealing connections between
the brain and the immune system.
Although such mindbody connections had been written
and talked about for centuries, most scientists at the time
considered the idea that emotions had an eect on health
to be nothing more than folklore. Had their experiments
been less rigorously conducted, Ader and Cohen could also
have been easily dismissed, notes Bruce Rabin, Professor
of Pathology and Psychiatry at the University of Pittsburgh
School of Medicine. The critical thing he did was very, very
good science. The study was so well controlled. One mistake
would have undermined the eld for years, but his work was
just as beautiful as can be. Cohen, now Professor Emeritus
of Microbiology and Immunology at the University of
Rochester Medical Center, agrees saying that Ader had a total
aversion to taking shortcuts in developing an experimental
protocol. The papers we wrote were airtight.
After graduating from Tulane University, New Orleans,
Ader received his PhD in psychology at Cornell University,
New York, in 1957. Soon after, he was appointed an
assistant professor in the Department of Psychology at the
University of Rochester, Rochester, NY, where he remained
for ve decades, until his retirement in July, 2011. He held
many titles during his tenure, including the George Engel
Professor of Psychosocial Medicine. After the seminal study
published in 1975, Ader continued to work in the eld he
dubbed psychoneuroimmunology, defending the idea
against doubters, and continuing to build evidence. But
his inuence extended beyond his academic contributions,
Rabin said: A lot of it had to do with his personality, his
wonderful communication skills and skills as a teacher. He
was an extremely kind, caring individual who wanted to
teach people, wanted to excite students about the area he
had developed. His personality attracted many students to
study with him.
Ader was the founder and past president of the
Psychoneuroimmunology Research Society, and also past
president of the Academy of Behavioral Medicine Research
and the American Psychosomatic Society. The eld he
helped inaugurate has found application in many medical
specialties. The ideas that you championed in your early
conditioning work were remarkably radical and heretical,
wrote Janice Kiecolt-Glaser, Distinguished University
Professor at the Ohio State University Medical Center, in a
collection of letters published in honour of Aders retirement.
Now I tell my students about how your courage and vision
were such an important force in helping shape the eld so
they can pursue their ideas today. Ader is survived by his
wife of 54 years, Gayle, four daughters, and a grandson.

Stephen Pincock
stephen.pincock@journalist.co.uk

www.thelancet.com Vol 379 January 28, 2012

Correspondence

While WHO undergoes a wide-ranging


reform sparked by a US$300 million
budget shortfall, the agency is facing
an exodus of qualied sta that is
aecting its ability to work.1 The
Executive Board is due to meet on
Jan 16 to agree long-term principles
and priorities for the organisation; it
must ensure, in particular, that core
functions are accorded the priority
they merit. Oxfam is especially
concerned that inadequate funding
will severely diminish the WHO
Essential Medicines Department,
which for more than three decades has
had an indispensable role in enabling
developing countries to access aordable medicines.
A key cause of WHOs nancial
predicament is its declining budget,
exacerbated by the adverse exchange
rate of the Swiss franc against the
US dollar.2 Money to pay for salaries
and management, as well as for its
vitally important core functions, has
diminished while voluntary funding
for projects has increased. Thus the
basic costs are squeezed while specic,
often vertical, activities that might be
limited to a number of countries only
are supported.
The planned reforms will take
time2015 is the estimated endpoint
for achieving nancial stability. In the
interim, continued sta losses will
put at risk essential WHO functions
that support public health, such as
the global norms, guidelines, and
standards produced through the
expert committees and similar bodies.
The Medicines Department, although
not the only unit under threat, is
a sobering illustration of the consequences of failing to protect core
functions of the organisation.
Between 2000 and 2010, income
for core medicines functions dropped
sharply. Experienced and specialised
www.thelancet.com Vol 379 January 28, 2012

sta in key posts lost through


mandatory redundancies and voluntary
losses cannot be replaced in the
current uncertain climate. Crucial work
is running late, or is made possible
only through cross-subsidisation taken
from project work.
Today only 10% of the Expert
Committee on Specications for
Pharmaceutical Preparations, which
underpins all quality assurance guidance for the development, production,
quality control, regulation, inspection,
and distribution of medicines, is paid
from the regular budget.3
Updating the Essential Medicines
List, which is the key to rational
management of pharmaceutical
systems, and the new Essential
Medicines List for Children, are paid
for by the Bill and Melinda Gates
Foundation (personal communication). Pain guidelines have been
put on hold. Work to monitor drug
safety (pharmacovigilance) and use,
antibiotic resistance, guidelines for
medicines pricing and availability,
and WHO support to the Viennabased International Narcotics Control
Board is faltering or in danger
of being discontinued (personal
communication).
WHO has been the global leader on
medicines since 1978. Since then it has
provided credible and evidence-based
guidelines for member states, nongovernmental organisations, industry,
and agencies such as the Global Fund,
UNITAID, and publicprivate partnerships. Yet this core work, which informs
all of public health involving medicines,
is now challenged by funding shortfalls
and survives through support from
voluntary contributions, with the
risk of donors dictating the agenda.
Allowing continuing erosion and
outsourcing of its component parts
would risk irreparable loss.
To restore them to optimal eciency
and reverse the brain drain, WHOs
core functions must be adequately
and securely funded once more from
the regular budgeta move that the
Executive Board should advance this

month. A rescue package will not wait


until 2015.
I declare that I have no conicts of interest.

Mohga M Kamal-Yanni
mkamalyanni@oxfam.org.uk

Published Online
January 13, 2012
DOI:10.1016/S01406736(12)60040-3

Oxfam, Oxford OX4 2JY, UK


1

Sridhar D, Gostin LO. Reforming the World


Health Organization. JAMA 2011; published
online March 29. DOI:10.1001/jama.2011.418.
WHO Executive Board. Financing of the World
Health Organization. http://apps.who.int/gb/
ebwha/pdf_les/EBSS/EBSS2_ID2-en.pdf
(accessed Jan 10, 2012).
WHO. WHO Expert Committee on
Specications for Pharmaceutical
Preparations: outcome of 44 meetings.
http://www.who.int/medicines/services/
expertcommittees/pharmprep/
ep_44meetingsreport/en/index.html
(accessed Jan 10, 2012).

Screening for congenital


heart disease with
newborn pulse oximetry
Andrew Ewer and colleagues
(Aug 27, p 785)1 claim that their testing
protocol has superior sensitivity to
that advocated by de-Wahl Granelli
and colleagues.2 This claim is incorrect
for several reasons.
First, Ewer and colleagues
sensitivity gure includes patients
known to have critical congenital
heart disease by fetal ultrasound
scanning. For infants screened blind,
as ours were, the sensitivity of Ewer
and colleagues protocol was 58%;
our protocol had a sensitivity of
62%.2 Second, Ewer and colleagues
claim that our protocol would have
missed one of their critical cases
(prenatally diagnosed hypoplastic left
heart syndrome). Our protocol would
class the rst reading of 92%/97% as
abnormal, but the second reading of
100%/97% is impossible in a patient
with aortic atresia who is breathing
air; presumably he or she was by then
receiving prostaglandin infusion and
probably oxygen.
Ewer and colleagues protocol
results in a signicantly lower positive
predictive value (92%) than does
ours (207%, 95% CI 128307), and

Science Photo Library

Action to preserve
WHOs core functions
cannot wait for
organisational reform

Submissions should be
made via our electronic
submission system at
http://ees.elsevier.com/
thelancet/

309

Correspondence

a ve-times higher false-positive rate


(085%), with substantial implications
for cost and workload; our falsepositive rate of 017% is low enough
to make screening cost-neutral.24
Ewer and colleagues do make some
important contributions. Their study
conrms earlier ndings that incorporation of both preductal and
postductal readings, and inclusion
of the dierence in the screening
cutos, improves detection rate
compared with postductal readings
only.2,5 Second, Ewer and colleagues
show that routine pulse-oximetry
screening before discharge signicantly improves the detection rate
of critical congenital heart disease
even in health districts that use
systematic antenatal cardiac anomaly
scanninga new ndingachieving
a total predischarge detection rate of
92%.1,2 Lastly, the results conrm that
screening should ideally be done after
the rst 6 h of life.1,5
We declare that we have no conicts of interest.

*Ingegerd stman-Smith,
Anne de-Wahl Granelli
ingegerd.ostman-smith@pediat.gu.se
Division of Paediatrics, Department of Clinical
Sciences, The Sahlgren Academy, Gothenburg
University, Section of Paediatric Cardiology, Queen
Silvia Childrens Hospital, Smrslottsgatan, stra
Sjukhuset, 416 85 Gothenburg, Sweden
1

310

Ewer AK, Middleton LJ, Furmston AT, et al.


Pulse oximetry screening for congenital heart
defects in newborn infants (PulseOx): a test
accuracy study. Lancet 2011; 378: 78594.
de-Wahl Granelli A, Wennergren M,
Sandberg K, et al. Impact of pulse oximetry
screening on the detection of duct-dependent
congenital heart disease: a Swedish
prospective screening study in 39 821
newborns. BMJ 2009; 338: a3037.
Griebsch I, Knowles RL, Brown J, Bull C, Wren C,
Dezateux CA. Comparing the clinical and
economic eects of clinical examination, pulse
oximetry and echocardiography in newborn
screening for congenital heart defects: a
probabilistic cost-eectiveness model and
value of information analysis.
Int J Technol Assess Health Care 2007;
23: 192204.
Mahle W, Koppel R. Screening with pulse
oximetry for congenital heart disease. Lancet
2011; 378: 74950.
de-Wahl Granelli A, Melander M,
Sunnegardh J, Sandberg K, stman-Smith I.
Screening for duct-dependent congenital
heart disease: a critical evaluation of strategies
to maximise sensitivity. Acta Paediatr 2005;
94: 159096.

The PulseOx study1 reports on


oximetry as a screening test for critical
and severe congenital heart disease. It
concludes that oximetry adds value to
existing screening. These results will
inform decision makers about national
screening recommendations.2
Oximetry should be judged as step
three of a stepwise screening programme for congenital heart disease
(where step one is midtrimester fetal
anomaly ultrasound scanning, and
step two newborn physical examination). Improved detection at step
one will reduce the additional yield
from later screening steps.
The PulseOx investigators do not
describe the detail of their ultrasound
scanning cardiac imaging protocol. In
2011, the UKs Fetal Anomaly Screening Programme added visualisation
of the ventricular outow tracts to
the standard four chamber view as
service standard to improve detection
rates.3 This standard might not have
been in place in all centres during the
PulseOx study.
Table 2 in the publication displays
results for all three screening steps
and hence allows analysis of the
incremental yield with each step. Of
24 critical cases, 18 are shown as true
positive, but close inspection shows
that only one case was not already
suspected at step one or two. Similarly,
in severe congenital heart disease,
of eight true positives, seven were
already identied before oximetry.
Hence, of 20 055 babies screened,
only two extra cases (critical
number 13, severe number 29) were
detected by addition of oximetry to
the existing two-step screening. This
proportion is likely to be even smaller
after the introduction of the national
scanning protocol that includes the
ventricular outow tracts.
The accompanying Comment4 also
notes the unsatisfactorily high falsepositive rate.
We declare that we have no conicts of interest.

*Peter Macfarlane, Rupa Talekar


peter.macfarlane@rothgen.nhs.uk

Department of Child Health, Rotherham NHS


Foundation Trust, Rotherham S60 2UD, UK
1

Ewer AK, Middleton LJ, Furmston AT, et al.


Pulse oximetry screening for congenital heart
defects in newborn infants (PulseOx): a test
accuracy study. Lancet 2011; 378: 78594.
UK National Screening Committee. The UK NSC
policy on congenital heart disease screening in
newborns. http://www.screening.nhs.uk/
congenitalheartdisease (accessed Aug 27, 2011).
NHS Fetal Anomaly Screening Programme.
18+0 to 20+6 weeks fetal anomaly scan
national standards and guidance for England.
http://fetalanomaly.screening.nhs.uk/
standardsandpolicies (accessed Aug 27, 2011).
Mahle W, Koppel R. Screening with pulse
oximetry for congenital heart disease. Lancet
2011; 378: 74950.

Although I agree that pulse oximetry


is useful as an important vital sign
monitor soon after birth, I have doubt
in its value as a screening instrument
for identifying congenital heart disease.
Table 2 of Ewer and colleagues paper1
shows that, of 53 cases of congenital
heart disease, 18 were identied
by antenatal ultrasound scanning,
26 by pulse oximetry, and 17 by both
methods. This essentially means an
additional eight diagnoses for every
20 000 babies screened.
Of the nine babies not identied
by antenatal screening, three had
oxygen saturations under 80% in
the rst 24 h (babies numbered 2,
6, and 13). Irrespective of the pulse
oximetry reading, such low oxygen
saturations would have been noticed by
the mother herself or by the nurses or
doctors, resulting in examination and
opportunistic pulse oximetry reading.
What happened to the remaining
six babies? Did they all survive after
intervention? Isnt it likely that some of
these would have also been identied
early enough opportunistically?
Therefore this study looks only at
the additional benet over antenatal
scanning, and ignores the fact that
some of these babies would be
identied opportunistically, including
by the use of pulse oximetry. It is also
important to note that antenatal
ultrasonography is becoming more
sophisticated and is likely to become
better at identifying congenital heart
disease. Without consideration of
www.thelancet.com Vol 379 January 28, 2012

Correspondence

a ve-times higher false-positive rate


(085%), with substantial implications
for cost and workload; our falsepositive rate of 017% is low enough
to make screening cost-neutral.24
Ewer and colleagues do make some
important contributions. Their study
conrms earlier ndings that incorporation of both preductal and
postductal readings, and inclusion
of the dierence in the screening
cutos, improves detection rate
compared with postductal readings
only.2,5 Second, Ewer and colleagues
show that routine pulse-oximetry
screening before discharge signicantly improves the detection rate
of critical congenital heart disease
even in health districts that use
systematic antenatal cardiac anomaly
scanninga new ndingachieving
a total predischarge detection rate of
92%.1,2 Lastly, the results conrm that
screening should ideally be done after
the rst 6 h of life.1,5
We declare that we have no conicts of interest.

*Ingegerd stman-Smith,
Anne de-Wahl Granelli
ingegerd.ostman-smith@pediat.gu.se
Division of Paediatrics, Department of Clinical
Sciences, The Sahlgren Academy, Gothenburg
University, Section of Paediatric Cardiology, Queen
Silvia Childrens Hospital, Smrslottsgatan, stra
Sjukhuset, 416 85 Gothenburg, Sweden
1

310

Ewer AK, Middleton LJ, Furmston AT, et al.


Pulse oximetry screening for congenital heart
defects in newborn infants (PulseOx): a test
accuracy study. Lancet 2011; 378: 78594.
de-Wahl Granelli A, Wennergren M,
Sandberg K, et al. Impact of pulse oximetry
screening on the detection of duct-dependent
congenital heart disease: a Swedish
prospective screening study in 39 821
newborns. BMJ 2009; 338: a3037.
Griebsch I, Knowles RL, Brown J, Bull C, Wren C,
Dezateux CA. Comparing the clinical and
economic eects of clinical examination, pulse
oximetry and echocardiography in newborn
screening for congenital heart defects: a
probabilistic cost-eectiveness model and
value of information analysis.
Int J Technol Assess Health Care 2007;
23: 192204.
Mahle W, Koppel R. Screening with pulse
oximetry for congenital heart disease. Lancet
2011; 378: 74950.
de-Wahl Granelli A, Melander M,
Sunnegardh J, Sandberg K, stman-Smith I.
Screening for duct-dependent congenital
heart disease: a critical evaluation of strategies
to maximise sensitivity. Acta Paediatr 2005;
94: 159096.

The PulseOx study1 reports on


oximetry as a screening test for critical
and severe congenital heart disease. It
concludes that oximetry adds value to
existing screening. These results will
inform decision makers about national
screening recommendations.2
Oximetry should be judged as step
three of a stepwise screening programme for congenital heart disease
(where step one is midtrimester fetal
anomaly ultrasound scanning, and
step two newborn physical examination). Improved detection at step
one will reduce the additional yield
from later screening steps.
The PulseOx investigators do not
describe the detail of their ultrasound
scanning cardiac imaging protocol. In
2011, the UKs Fetal Anomaly Screening Programme added visualisation
of the ventricular outow tracts to
the standard four chamber view as
service standard to improve detection
rates.3 This standard might not have
been in place in all centres during the
PulseOx study.
Table 2 in the publication displays
results for all three screening steps
and hence allows analysis of the
incremental yield with each step. Of
24 critical cases, 18 are shown as true
positive, but close inspection shows
that only one case was not already
suspected at step one or two. Similarly,
in severe congenital heart disease,
of eight true positives, seven were
already identied before oximetry.
Hence, of 20 055 babies screened,
only two extra cases (critical
number 13, severe number 29) were
detected by addition of oximetry to
the existing two-step screening. This
proportion is likely to be even smaller
after the introduction of the national
scanning protocol that includes the
ventricular outow tracts.
The accompanying Comment4 also
notes the unsatisfactorily high falsepositive rate.
We declare that we have no conicts of interest.

*Peter Macfarlane, Rupa Talekar


peter.macfarlane@rothgen.nhs.uk

Department of Child Health, Rotherham NHS


Foundation Trust, Rotherham S60 2UD, UK
1

Ewer AK, Middleton LJ, Furmston AT, et al.


Pulse oximetry screening for congenital heart
defects in newborn infants (PulseOx): a test
accuracy study. Lancet 2011; 378: 78594.
UK National Screening Committee. The UK NSC
policy on congenital heart disease screening in
newborns. http://www.screening.nhs.uk/
congenitalheartdisease (accessed Aug 27, 2011).
NHS Fetal Anomaly Screening Programme.
18+0 to 20+6 weeks fetal anomaly scan
national standards and guidance for England.
http://fetalanomaly.screening.nhs.uk/
standardsandpolicies (accessed Aug 27, 2011).
Mahle W, Koppel R. Screening with pulse
oximetry for congenital heart disease. Lancet
2011; 378: 74950.

Although I agree that pulse oximetry


is useful as an important vital sign
monitor soon after birth, I have doubt
in its value as a screening instrument
for identifying congenital heart disease.
Table 2 of Ewer and colleagues paper1
shows that, of 53 cases of congenital
heart disease, 18 were identied
by antenatal ultrasound scanning,
26 by pulse oximetry, and 17 by both
methods. This essentially means an
additional eight diagnoses for every
20 000 babies screened.
Of the nine babies not identied
by antenatal screening, three had
oxygen saturations under 80% in
the rst 24 h (babies numbered 2,
6, and 13). Irrespective of the pulse
oximetry reading, such low oxygen
saturations would have been noticed by
the mother herself or by the nurses or
doctors, resulting in examination and
opportunistic pulse oximetry reading.
What happened to the remaining
six babies? Did they all survive after
intervention? Isnt it likely that some of
these would have also been identied
early enough opportunistically?
Therefore this study looks only at
the additional benet over antenatal
scanning, and ignores the fact that
some of these babies would be
identied opportunistically, including
by the use of pulse oximetry. It is also
important to note that antenatal
ultrasonography is becoming more
sophisticated and is likely to become
better at identifying congenital heart
disease. Without consideration of
www.thelancet.com Vol 379 January 28, 2012

Correspondence

I declare that I have no conicts of interest.

Krishnarajah Nirantharakumar
k.nirantharan@bham.ac.uk
University of Birmingham, Birmingham B15 2TT, UK
1

Ewer AK, Middleton LJ, Furmston AT, et al.


Pulse oximetry screening for congenital heart
defects in newborn infants (PulseOx): a test
accuracy study. Lancet 2011; 378: 78594.

Authors reply
Dierences in results between
de-Wahl Granelli and colleagues
study1 and ours are likely to have arisen
because of antenatal screening and
timing of pulse oximetry. Detection
of critical congenital heart disease
by antenatal ultrasound is very
variable: we detected 50% of cases
of critical congenital heart disease
antenatally compared with only 3%
in de-Wahl Granelli and colleagues
study.1 Detection rates of between
15% and 50% have been reported
among UK health regions.2 We noted
a lower incremental value of pulse
oximetry than did de-Wahl Granelli and
colleagues, almost certainly because
many more cases were identied
by antenatal screening in our study
hospitals. Our detection rate for critical
congenital heart disease of 75% with
pulse oximetry for the full cohort
is higher than theirs. We agree that
routine visualisation of the outow
tracts should further improve antenatal detection in isolated cardiac
lesions; however, consistency of
detection has yet to be shown.
de-Wahl Granelli and colleagues
screened at a median of 38 h after
birth, compared with 12 h in our
study, which might explain our higher
false-positive rate. Discharge from
UK hospitals is frequently within
24 h, so delayed pulse oximetry is not
feasible and also increases risks of late
diagnosis. Importantly, half (28/57) of
eligible babies with critical congenital
heart disease in de-Wahl Granelli and
colleagues study required admission
www.thelancet.com Vol 379 January 28, 2012

to intensive care before screening took


place compared with none in our study.
We conrm that all babies in our study
were asymptomatic at testing, and the
individual case highlighted by stmanSmith and de-Wahl Granelli was not
receiving prostaglandin or oxygen.
We disagree that it is impossible to
have saturations of 100% and 97%
with hypoplastic left heart syndrome.
Notably, since the initial examination
for this baby was normal, the second
test result would have passed
according to the protocol described by
de-Wahl Granelli and colleagues.
We cannot deduce the value of
screening where physical examination
precedes pulse oximetry from our
study, since the results of physical
examination are likely to have been
aected by the prior knowledge of
the saturation result. We disagree
that all cyanosed babies would be
detected clinically, since the ability
of clinicians to detect cyanosis is
notably poor3in the non-screening
regions in de-Wahl Granelli and
colleagues study, 44% of babies with
transposition of great arteries were
discharged without diagnosis.
Our study, combined with the
evidence of others,1 supports the
introduction of routine pulse-oximetry
screening for neonates. Introduction
of pulse- oximetry screening will
particularly improve detection rates
in regions where antenatal diagnosis
rates are lower. The optimum timing
of screening requires careful consideration; delayed screening is likely to
minimise false-positive rates, whereas
earlier screening should improve
timely diagnosis.
We declare that we have no conicts of interest.

*A K Ewer, L J Middleton,
S Thangaratinam, K S Khan, J J Deeks
a.k.ewer@bham.ac.uk
School of Clinical and Experimental Medicine,
University of Birmingham, Birmingham, UK (AKE);
*Birmingham Womens Healthcare NHS Foundation
Trust, Birmingham B152TG, UK (AKE); Birmingham
Clinical Trials Unit, University of Birmingham,
Birmingham, UK (LJM); Queen Mary University of
London, Barts and the London School of Medicine,
London, UK (ST, KSK); and Public Health,

Epidemiology and Biostatistics, University of


Birmingham, Birmingham, UK (JJD)
1

de-Wahl Granelli A, Wennergren M,


Sandberg K, et al. Impact of pulse oximetry
screening on detection of duct dependent
congenital heart disease: a Swedish
prospective screening study in
39 821 newborns. BMJ 2009; 338: A3037.
National Institute for Cardiovascular
Outcomes Research. Percentage of infants
who required treatment who were
antenatally diagnosed, analysed by year and
by SHA/country. http://www.ccad.org.
uk/002/congenital.nsf/vwContent/
Antenatal%20Diagnosis?Opendocument
(accessed Jan 3, 2012).
ODonnell CPF, Kamlin COF, Davis PG, Carlin JB,
Morley CJ. Clinical assessment of infant colour
at delivery. Arch Dis Child 2007; 92: F46567.

Coronary artery calcium


for guiding statin
treatment
Although of interest and hypothesisgenerating, the MESA report by
Michael Blaha and colleagues (Aug 20,
p 684)1 is not a randomised trial and
thus it is impossible to conclude that
coronary artery calcication can be
used to target patients who derive
the most and the least absolute
benet from statin therapy.
Four majortrialsAURORA,CORONA,
4-D, and GISSI-HFdid not show a
benet for statin therapy compared
with placebo. What these four trials
share is the enrolment of very high-risk
individuals who are likely to have high
coronary artery calcication scores.
Moreover, heavily calcied plaques
are not the most likely to rupture and
cause infarction, nor do statins reduce
coronary artery calcication. Therefore,
until we have clear randomised
evidence, the aggressive coronary
artery calcication algorithm suggested
by Axel Schmermund and Thomas
Voigtlnder in their accompanying
Comment2 is premature and would
deny treatment to individuals in whom
completed trials have shown benet,
while providing it to groups never
tested.

Science Photo Library

these issues and with the lower limit


of the CI being 146% for sensitivity,
I wonder how one can justify it as a
screening instrument?

I am the principal investigator of the JUPITER trial,


which was funded by AstraZeneca, and am listed as a
co-inventor on patents held by the Brigham and

311

Correspondence

Womens Hospital in Boston that relate to the use of


inammatory biomarkers in cardiovascular disease
that have been licensed to Siemens and AstraZeneca.

Paul M Ridker
pridker@partners.org
Center for Cardiovascular Disease Prevention, Brigham
and Womens Hospital, Boston, MA 02215, USA
1

Blaha MJ, Budo MJ, Delippis AP, et al.


Associations between C-reactive protein,
coronary artery calcium, and cardiovascular
events: implications for the JUPITER
population from MESA, a population-based
cohort study. Lancet 2011; 378: 68492.
Schmermund A, Voigtlnder T. Predictive
ability of coronary artery calcium and CRP.
Lancet 2011; 378: 64143.

risk factor in itself. Besides, the cost and


complexity of methods of measuring
coronary artery calcium are severe
limitations on its recommendation
for public health purposes, particularly
from a global perspective. Lowhanging fruits should be harvested
before people are further medicalised
for the purpose of risk stratication.4
We declare that we have no conicts of interest.

*Andre Pascal Kengne,


Justin Basile Echouo-Tcheugui,
Eugene Sobngwi
andre.kengne@mrc.ac.za

The paper by Michael Blaha and


colleagues1 adds to the debate on
cardiovascular risk assessment by
suggesting that coronary artery calcium
could (1) further risk-stratify individuals
eligible for the JUPITER trial, (2) be
useful to target subgroups of patients
for prevention, and (3) be better than
high-sensitivity C-reactive protein
(hsCRP) alone. Such claims, however,
can hardly be solely based on eect sizes
for the association of coronary artery
calcium with cardiovascular outcomes.
The American Heart Association2
states that cardiovascular markers
should not be assessed in isolation for
their predictive abilities but rather on
their added predictive contribution
beyond established predictors. A
mere strong association between a
risk marker and an outcome is not a
guarantee of its high predictive ability.
Ideally, Blaha and colleagues should
have presented the readers with global
discrimination and reclassication
indices3 for models, including classical
cardiovascular risk factors (Framingham
model variables for instance) plus hsCRP
only, then coronary artery calcium only,
and nally coronary artery calcium
plus hsCRP. Furthermore, 950 people
(and few outcomes) are too small a
population on which to base claims of
risk stratication or reclassication.
Finally, it is not particularly appropriate to rely on coronary artery
calcium for risk prediction since it is a
measure of disease burden (showing
the eects of several risk factors), not a
312

South African Medical Research Council and University


of Cape Town, PO Box 19070, Tygerberg, 7505,
Cape Town, South Africa (APK); George Institute for
International Health, Sydney, NSW, Australia (APK);
Hubert Department of Global Health, Rollins School of
Public Health, Emory University, Atlanta, GA, USA
(JBE-T); and Institute of Health and Society, Medical
School, University of Newcastle-Upon-Tyne,
Newcastle-Upon-Tyne, UK (ES)
1

Blaha MJ, Budo MJ, Delippis AP, et al.


Associations between C-reactive protein,
coronary artery calcium, and cardiovascular
events: implications for the JUPITER
population from MESA, a population-based
cohort study. Lancet 2011; 378: 68492.
Hlatky MA, Greenland P, Arnett DK, et al.
Criteria for evaluation of novel markers of
cardiovascular risk: a scientic statement from
the American Heart Association. Circulation
2009; 119: 240816.
Pencina MJ, DAgostino RB Sr, DAgostino RB Jr,
Vasan RS. Evaluating the added predictive
ability of a new marker: from area under the
ROC curve to reclassication and beyond.
Stat Med 2008; 27: 15772.
Kengne AP, Patel A, Marre M, et al.
Contemporary model for cardiovascular risk
prediction in people with type 2 diabetes.
Eur J Cardiovasc Prev Rehabil 2011; 18: 39398.

Authors reply
Paul Ridker questions the ability of
coronary artery calcium to identify
appropriate patients for statin
therapy. However, his implication that
increased coronary artery calcium is a
marker for poor response to statins is
not supported by the available data.
Statins clearly reduce recurrent
coronary events in secondary prevention
trials in which patients have established
coronary disease and presumably
elevated coronary artery calcium. The
St Francis Heart Study1 tested this
hypothesis more directly, randomising
1005 patients with elevated coronary

artery calcium scores to atorvastatin


20 mg (and vitamins C and E) versus
placebo. Although atorvastatin produced a non-signicant reduction in all
atherosclerotic cardiovascular events
across the entire population (event rate
69% vs 99%, 5-year number needed
to treat [NNT] 39), there was a greater
eect in the subgroup with coronary
artery calcium scores of greater than
400 (event rate 87% vs 150%, 5-year
NNT 185). Coronary artery calcium was
not tested in the AURORA, CORONA,
4-D, and GISSI-HF trials. Instead, these
trials all enrolled highly specialised
populations with either heart failure or
advanced kidney disease.
There are no tests yet available that
predict a preferential benet with
statins (higher or lower relative risk
reduction based on the test result).2
Until such data from appropriately
designed biomarker trials are available,
serum biomarkers and imaging tests
are best used for identication of
increased absolute risk, enabling
targeted treatment decisions for
patients in whom treatment would be
expected to produce larger or smaller
absolute benet.
One must acknowledge that it is
dicult to prevent events if they rarely
happen. We showed that nearly 50% of
patients tting JUPITER criteria have no
coronary artery calcium, and have an
extremely low event rate of about 1 per
1000 patient-years (about 1% 10-year
event rate). Even if statins could
prevent every event in this population,
the 5-year NNT would still be more than
200! On the basis of the mountainous
data to suggest an excellent prognosis
for those with coronary artery calcium
scores of zero,35 would one argue that
most of the event reduction in trials
such as JUPITER occurred in patients
with coronary artery calcium scores of
zero, or in the higher-risk patients with
measurable coronary artery calcium?
Andre Kengne and colleagues
point out that, by integrating the
exposure to risk factors over a
lifetime, coronary artery calcium is
best regarded as a measure of disease
www.thelancet.com Vol 379 January 28, 2012

Correspondence

We declare that we have no conicts of interest.

*Michael J Blaha, Matthew J Budo,


Roger S Blumenthal, Khurram Nasir
mblaha1@jhmi.edu
Johns Hopkins Hospital, Baltimore, MD 21287, USA
(MJB, RSB); Division of Cardiology, Los Angeles
Biomedical Research Institute at Harbor-UCLA,
Torrance, CA, USA (MJB); and Section of
Cardiovascular Medicine, Yale University School of
Medicine, New Haven, CT, USA (KN)
1

Arad Y, Spadaro LA, Roth M, Newstein D,


Guerci AD. Treatment of asymptomatic adults
with elevated coronary calcium scores with
atorvastatin, vitamin C, and vitamin E: the
St Francis Heart Study randomized clinical trial.
J Am Coll Cardiol 2005; 46: 16672.
Heart Protection Study Collaborative Group.
C-reactive protein concentration and the
vascular benets of statin therapy: an analysis
of 20 536 patients in the Heart Protection
Study. Lancet 2011; 377: 46976.
Blaha M, Budo MJ, Shaw LJ, et al. Absence of
coronary artery calcication and all-cause
mortality. JACC Cardiovasc Imaging 2009;
2: 692700.
Sarwar A, Shaw LJ, Shapiro MD, et al.
Diagnostic and prognostic value of absence
of coronary artery calcication.
JACC Cardiovasc Imaging 2009; 2: 67588.
Budo MJ, McClelland RL, Nasir K, et al.
Cardiovascular events with absent or minimal
coronary calcication: the Multi-Ethnic Study
of Atherosclerosis (MESA). Am Heart J 2009;
158: 55461.

ADDITION-Europe and
the case for diabetes
screening

quite small. Were the benet signicant, it would have required screening
of 1829 people,2 and treatment of
78 screen-positive ones, for 53 years
to prevent one cardiovascular event.
Without comparison with an unscreened group, the study could
not explore the specic benets of
diabetic screening. Glycated haemoglobin (HbA1c) in screen-positive
people decreased by 0304%
(1920 mmol/mol) over 53 years,
whereas it would be expected to
increase. Meta-analyses have shown
that intensive glucose lowering,
although reducing rates of non-fatal
myocardial infarction and surrogate
measures of microangiopathy, does
not aect cardiovascular mortality,
stroke, or patient-important microvascular disease.3
But the downsides of screening are
evident. The diagnosis of diabetes is a
surrogate for challenges with employment and insurability; a need for clinic
visits, tests, self-monitoring, and
hypoglycaemic drug use (some 60%
in ADDITION-Europe); and inconvenience and cost. Moreover, although
ADDITION-Europe showed no psychological eect of screening itself, those
who screened positive had signicantly
worse general health, greater anxiety,
and were more depressed.4
Current evidence does not support
population screening for diabetes. This
chorus, coming from public health
advocates, professional societies, and
industry, is not without consequence:
Heath5 has eloquently described
how screening policies for noncommunicable diseases could result in
catastrophic spending by individuals
and institutions in low-income and
middle-income countries.
We declare that we have no conicts of interest.

Intensive risk-factor control did not


signicantly reduce cardiovascular
risk in people screened positive for
diabetes in the ADDITION-Europe trial
(July 9, p 156).1 Although the study
might have been too small or too brief
to show such an eect, the corollary
is that absolute benet is likely to be
www.thelancet.com Vol 379 January 28, 2012

*John S Yudkin, Victor M Montori,


Kasia J Lipska, Edwin A M Gale
j.yudkin@ucl.ac.uk
University College London, London N7 0AG, UK (JSY);
Knowledge and Evaluation Research Unit, Division of
Endocrinology and Diabetes, Departments of
Medicine and Health Sciences Research, Mayo Clinic,
Rochester, MN, USA (VMM); Department of

Medicine, Yale University School of Medicine,


New Haven, CT, USA (KJL); and University of Bristol,
Bristol, UK (EAMG)
1

Grin SJ, Borch-Johnsen K, Davies MJ, et al.


Eect of early intensive multifactorial therapy
on 5-year cardiovascular outcomes in
individuals with type 2 diabetes detected by
screening (ADDITION-Europe): a clusterrandomised trial. Lancet 2011; 378: 15667.
Sandbaek A, Grin SJ, Rutten G, et al. Stepwise
screening for diabetes identies people with high
but modiable coronary heart disease risk: the
ADDITION study. Diabetologia 2008; 51: 112734.
Hemmingsen B, Lund SS, Gluud C, et al.
Targeting intensive glycaemic control versus
targeting conventional glycaemic control
for type 2 diabetes mellitus.
Cochrane Database Syst Rev 2011; 6: CD008143.
Eborall H, Davies R, Kinmonth AL, Grin S,
Lawton J. Patients experiences of screening for
type 2 diabetes: prospective qualitative study
embedded in the ADDITION (Cambridge)
randomised controlled trial. BMJ 2007; 335: 490.
Heath I. Seeming virtuous on chronic diseases.
BMJ 2011: 343: d4239.

Authors reply
ADDITION-Europe is not a trial of
screening for diabetes. We reviewed
the evidence for screening and identied the key uncertainties: the eect
of early treatment in the lead time
after detection by screening and the
harms associated with screening.1
ADDITION-Europe addressed the
eect of early treatment. Inclusion
of a no-screening control group in
one centre enabled us to add to the
growing evidence of minimal harms
associated with screening.2 A diagnosis
of diabetes certainly has adverse consequences, which are brought forward
by screening. However, our study
suggests that stepwise screening
might attenuate these eects by
facilitating psychological adjustment.3
We share John Yudkin and colleagues
healthy scepticism about the eects
of rapid, intensive glucose lowering
in patients with longstanding poor
glycaemic control, but we question
its relevance in this context. Patients
at high cardiovascular risk, such as
those with screen-detected diabetes,
should not be denied access to eective
treatments such as antihypertensives,
statins, and metformin, which were
prescribed to 45%, 16%, and 04% of
ADDITION-Europe participants, respectively, before diagnosis and to 80%,

Science Photo Library

burden rather than a risk factor. We


agree, and believe that this explains
why coronary artery calcium is such
a powerful predictor of risk and such
a valuable negative predictor when
the coronary artery calcium score is
zero. Kengne and colleagues, however,
wish to consider this a weakness of
coronary artery calcium testing. We
nd direct measurement of the disease
for integrated, cumulative risk factor
exposure very appealing.

313

Correspondence

76%, and 52% 5 years later. ADDITIONEurope might have been too small
and too brief, but the most likely
explanation for the non-signicant
17% reduction in risk of cardiovascular
events was the quality of routine care
and hence the limited dierences in
treatment between study groups. This
in turn explains the number needed
to treat (NNT) of 78 screen-detected
patients treated for 53 years to prevent
one cardiovascular event.
However, it would be wrong to infer
from this nding that screening is not
worthwhile. In assessing the potential
benets of screening, the NNT should
reect changes in treatment before
and after diagnosis (as shown above),
the duration of the lead time between
detection by screening and clinical
diagnosis, the likelihood of treatment
irrespective of a diagnosis of diabetes,
and the eectiveness of treatment.
Our estimates suggest that around
ten patients with screen-detected
diabetes would need to be treated for
10 years to prevent one event.4
We have reduced the uncertainty,
but whether screening represents an
ecient use of scarce resources remains
unclear. Tackling the growing burden of
diabetes requires a balanced investment
in primary prevention, better care for
those with the disorder, and perhaps
also in some settings earlier detection
via screening in high-risk groups.5
Corbis

SJG has attended an advisory board for Colgate


Palmolive, has received honoraria for speaking from
Unilever, Eli Lilly, GlaxoSmithKline, MSD, and Novo
Nordisk, and has received travel expenses from Eli
Lilly; MJD has served on advisory boards for Novo
Nordisk, Eli Lilly, MSD, Bristol-Myers Squibb, and
Roche, and has received honoraria for speaking from
Novo Nordisk, Eli Lilly, Sano-Aventis, Novartis, and
MSD; GEHMR has served as a consultant and
participated in advisory boards for Novo Nordisk and
MSD, and has received honoraria for speaking from
Novo Nordisk; TL has received research funding from
Novo Nordisk, AstraZeneca, Pzer, GlaxoSmithKline,
Servier, and HemoCue, has received honoraria for
speaking from various pharmaceutical companies,
and holds stock in Novo Nordisk.

*Simon J Grin, Melanie J Davies,


Guy E H M Rutten, Torsten Lauritzen,
on behalf of the ADDITION-Europe
investigators
simon.grin@mrc-epid.cam.ac.uk

314

MRC Epidemiology Unit, Institute of Metabolic


Science, Cambridge CB2 0QQ, UK (SJG); Department
of Cardiovascular Sciences, University of Leicester,
Leicester, UK (MJD); Julius Centre for Health Sciences
and Primary Care, University Medical Centre Utrecht,
Utrecht, Netherlands (GEHMR); and School of Public
Health, Department of General Practice, University
of rhus, rhus, Denmark (TL)
1

Wareham NJ, Grin SJ. Should we screen for


type 2 diabetes? Evaluation against National
Screening Committee criteria. BMJ 2001;
322: 98688.
Eborall HC, Grin SJ, Prevost AT, Kinmonth AL,
French DP, Sutton S. Psychological impact of
screening for type 2 diabetes: controlled trial
and comparative study embedded in the
ADDITION (Cambridge) randomised controlled
trial. BMJ 2007; 335: 48689.
Eborall H, Davies R, Kinmonth AL, Grin S,
Lawton J. Patients experiences of screening
for type 2 diabetes: prospective qualitative
study embedded in the ADDITION
(Cambridge) randomised controlled trial. BMJ
2007; 335: 49093.
Echouo-Tcheugui JB, Sargeant LA, Prevost AT,
et al. How much might cardiovascular disease
risk be reduced by intensive therapy in people
with screen-detected diabetes? Diabet Med
2008; 25: 143339.
van den Donk M, Sandbaek A, Borch-Johnsen K,
et al. Screening for type 2 diabetes: lessons
from the ADDITION-Europe study. Diabet Med
2011; 28: 141624.

Cited or read?
The need to establish quantitative
and qualitative indices of scientic
production by a researcher or
institution is hotly debated. Worldwide, the impact factor and H-index
are regarded as the best available.1,2
The advantages and disadvantages
have been widely discussed.3 However,
the question might arise: is an article
written to be cited or to be read?
In 2009, I published an article in the
American Journal of Medical Genetics
on non-immune hydrops fetalis.4
The journal has an impact factor
of 2505 and, as of June, 2011, the
article had been cited seven times
(two self-citations), according to ISI
Web of Science. However, between
publication and June, 2011, the article
was downloaded 3167 times from
the journals website (data from
the publisher)the highest number
of any article during that period. I
assume that many colleagues dealing
with hydrops fetalis read the article

in the hope of nding help for their


patients.
No existing index establishes
whether an article is more useful
to physicians involved in everyday
practice or to those who need to be
cited to obtain funding or promotion.
Thus, an article could be broadly
read, but rarely cited, thereby falling
through the cracks of the validation
mechanism. The total number of
articles, total number of citations,
citations per article, and number of
highly visible items are currently the
most useful parameters in calculating
the quality of scientic production.
There is no way for authors to obtain
information about the downloading of
their articles, unlike what is possible for
citations. Most publishers consider this
proprietary information. I suggest that
publishers should provide a real-time
counter of article downloads and that
these data be taken into consideration
in the overall assessment of scientic
production. This further element could
be useful for building a fair, honest, upto-date, and eective way of working
to improve citation metrics.
I declare that I have no conicts of interest.

Carlo Bellini
carlobellini@ospedale-gaslini.ge.it
Neonatal Intensive Care Unit, Gaslini Childrens
Hospital, University of Genoa, Department of
Pediatrics, 16147 Genova, Italy
1

Hirsch JE. An index to quantify an individuals


scientic research output. Proc Natl Acad Sci USA
2005; 102: 1656972.
Poynard T, Thabut D, Munteanu M, et al.
Hirsch index and truth survival in clinical
research. PLoS One 2010; 5: e12044.
Saleem T. The Hirsch indexa play on numbers
or a true appraisal of academic output?
Int Arch Med 2011; 7: 25.
Bellini C, Hennekam RC, Fulcheri E, et al. Etiology
of nonimmune hydrops fetalis: a systematic
review. Am J Med Genet 2009; 149A: 84451.

Department of Error
Williams HC, Dellavalle RP, Garner S. Acne vulgaris.
Lancet 2012; 379: 36172In this Seminar
(Jan 28), an aliation was missing for Sarah
Garner. The aliation should be The
Commonwealth Fund, New York, NY, USA. This
correction has been made to the online version
as of Jan 27, 2012, and to the printed Seminar.

www.thelancet.com Vol 379 January 28, 2012

Correspondence

New antithrombotic
drugs for atrial
brillation: caution is
needed

We believe that dabigatran is not


safe in many elderly patients with
impaired renal function and the
European drug regulation authorities
are correct in proceeding carefully with
the licensing of these drugs.

The Correspondence letter by


Eloi Marijon and others (Aug 20,
p 662)1 chastises the drug regulation
authorities in Europe for the slowness
of their assessment processes in the
approval of new antithrombotic
drugs for use in atrial brillation. They
claim that the ROCKET-AF2 and RE-LY3
studies have shown these new drugs to
be superior to vitamin K antagonists.
Marijon and colleagues then go on to
extrapolate the claimed advantages to
the European population with atrial
brillation and calculate the number
of strokes that might have been saved
by earlier introduction of these drugs.
Such extrapolation is a common
and serious error. The studies recruited
carefully selected patients who are not
typical of the population with atrial
brillation. The RE-LY study recruited
very few patients older than 80 years
and excluded those with a creatinine
clearance of less than 30 mL/min.
The studies show ecacy but not
eectiveness and they certainly do not
show safety.
In New Zealand, dabigatran was
funded from July 1, 2011. At Auckland
City Hospital, in July and August, 2011,
we admitted 14 patients on dabigatran
for atrial brillation with bleeding.
Four of these patients died during
their admission. In three, surgery (for
fractured neck of femur in two and an
ischaemic limb in one) was delayed.
11 patients were older than 80 years
and three had creatinine clearance of
less than 30 mL/min. We have no data
on the number of prescriptions for this
drug in the community.
10 years ago The Lancet published
an account of the rapid introduction
of troglitazone to the market.4 The
early licensing of that drug in the USA
should stand as warning to all those
who wish to rush new drugs to the
market.

We declare that we have no conicts of interest.

www.thelancet.com Vol 379 January 28, 2012

*Sarah Bell, Jessica Nand,


David Spriggs
sarahbell@xtra.co.nz
Auckland District Health Board, Grafton, Auckland
1050, New Zealand (SB); and Auckland City
Hospital, Auckland, New Zealand (JN, DS)
1

Marijon E, Fauchier L, Le Heuzey J.


New antithrombotic drugs and European
approval process. Lancet 2011; 378: 66263.
The Executive Steering Committee, on behalf
of the ROCKET-AF Study Investigators.
Rivaroxabanonce daily, oral, direct factor Xa
inhibition compared with vitamin K
antagonism for prevention of stroke and
embolism trial in atrial brillation: rationale
and design of the ROCKET-AF study. Am Heart J
2010; 159: 34047.
Connolly SJ, Ezekowitz MD, Yusuf S, et al.
Dabigatran versus warfarin in patients with atrial
brillation. N Engl J Med 2009; 361: 113951.
Gale EAM. Lessons from the glitazones: a story
of drug development. Lancet 2001;
357: 187075.

Authors reply
We thank Sarah Bell and colleagues
for reinforcing the point that approval
processes should be optimised, in
terms of delay, without decreasing the
rigour of these processes.
We are aware of several recently
published cases emphasising that
prescription of dabigatran could be
deleterious, particularly in elderly
patients with poor renal function and
low bodyweight.1
Guideline developers should consider exclusion criteria as well as the
characteristics of patients enrolled in
randomised controlled trials. Bell and
colleagues legitimately underline that
patients with a calculated creatinine
clearance lower than 30 mL/min
were excluded from the RE-LY trial,
and that the study recruited very
few patients older than 80 years.2
We would add that this was the case
for all trials of new antithrombotic
drugs in atrial brillation,24 and that
additionally, in RE-LY, only 002% had
a bodyweight less than 50 kg,2 whereas

in other trials, a reduced dose of


antithrombotic drugs was considered
when the creatinine clearance was
between 30 and 49 mL/min,4 as well
as for patients who met two of the
following criteria: age 80 years or
older, bodyweight 60 kg or less, or
a serum creatinine concentration of
133 mol/L or higher.3
Taken together, there are also
substantial data to support the need
for special caution or even contraindication for their use in elderly
patients, those with a low bodyweight,
and, more importantly, in those with
renal failure, even it is mild. The eect
of each of these factors has been
shown.5
We totally agree that observational
studies, based on the real life use of
dabigatran, can add substantial complementary information about its
adverse eects, and should therefore
be started early. However, such
registries would only be possible if the
prescription of dabigatran is possible.
Finally, we have to admit that any
antithrombotic drugs have been and
will be responsible for haemorrhagic
complications, irrespective of adherence to the prescription guidelines.
To the best of our knowledge, the
delay since the publication of the rst
results of the RE-LY trial was not related
to the need for a further assessment of
the drug, particularly safety and benet/
risk concerns. The cause of the delay
was rather negotiation of the price of
the drug, and it seems likely that this
delay resulted in a loss of chance for a
substantial number of patients.
EM has no conicts of interest. LF has served as a
consultant for Bayer, Medtronic, and SanoAventis, and has received funding for conference
travel and educational symposia from Boehringer
Ingelheim, Bayer, Medtronic, and Sano-Aventis.
J-YLH has served as a consultant for Boehringer
Ingelheim, Bayer, BMS/Pzer, and Daiichi-Sankyo.

*Eloi Marijon, Laurent Fauchier,


Jean-Yves Le Heuzey
eloi_marijon@yahoo.fr
Paris Cardiovascular Research Center, Inserm U970,
European Georges Pompidou Hospital, 75737 Paris
Cedex 15, France (EM); Cardiology Department,
Trousseau University Hospital and Franois Rabelais
University, Tours, France (LF); and Cardiology

Submissions should be
made via our electronic
submission system at
http://ees.elsevier.com/
thelancet/

e24

Correspondence

Department, European Georges Pompidou Hospital


and Paris Descartes University, Paris, France (J-YLH)
1

Legrand M, Mateo J, Aribaud A, et al. The use


of dabigatran in elderly patients.
Arch Intern Med 2011; 171: 128586.
Connolly SJ, Ezekowitz MD, Yusuf S, et al.
Dabigatran versus warfarin in patients with atrial
brillation. N Engl J Med 2009; 361: 113951.
Connolly SJ, Eikelboom J, Joyner C, et al.
Apixaban in patients with atrial brillation.
N Engl J Med 2011; 364: 80617.
Patel MR, Mahaey KW, Garg J, et al. Rivaroxaban
versus warfarin in nonvalvular atrial brillation.
N Engl J Med 2011; 365: 88391.
Eikelboom JW, Wallentin L, Connolly SJ, et al.
Risk of bleeding with 2 doses of dabigatran
compared with warfarin in older and younger
patients with atrial brillation: an analysis of
the randomized evaluation of long-term
anticoagulant therapy (RE-LY) trial. Circulation
2011; 123: 236372.

IAMP tackles a void in


medical education:
leadership

Published Online
November 18, 2011
DOI:10.1016/S01406736(11)61756-X

e25

David Holmess Prole (Oct 22,


p 1455)1 quoted Steve Wesselingh, the
dean of Monash University, Melbourne,
VIC, Australia, as stating: academic
institutions have an enormous
amount of expertise and knowledge,
but rarely are they engaged in the
process of health and social policy.
The inaugural Inter Academy Medical
Panel (IAMP) Young Physician Leaders
(YPL) programme held during the
third World Health Summit in Berlin,
Germany, attempted to address this
decit.2 IAMP sent out a global call
for nominations for physicians aged
40 years or younger with demonstrated leadership skills in medicine or
public health, and 22 participants were
chosen, representing 18 countries
low-income, middle-income, and
high-incomeand diverse physician
specialties.
The inaugural YPL group discussed
personal and systemic leadership
challenges as well as the necessary
substrate for leadership development.3
The programme consisted of an interactive brainstorming session aimed
at helping participants to develop a
strong leadership style. Mentoring
and peer-learning relationships were

developed through peer and senior


faculty interactions.
Participants
from
developing
countries acknowledged challenges of
limited resources and infrastructure.
For example, one participant had been
asked to start a centre with US$30 of
funding. Common challenges in the
developing and developed world were
the diculty of breaking down silos,
and that politics, as well as preference
for seniority over talent, can often
get in the way of success. Participants
from all countries shared diculties in
gaining credibility as young leaders and
breaking into established hierarchies.
Most important from this experience
was the sense that the world is small;
that the goals, aspirations, and
challenges of our junior faculty remain
similar despite vastly dierent cultures
and access to resources.
Having pioneered this YPL programme for the young physician
leaders, the next issue to address is
the programmes future.4 The IAMP
executive board members will continue
to work as personal mentors for each
inaugural YPL member, providing
their insight and guidance on future
career decisions and linking them to
other professional leaders in their
elds. The inaugural class of IAMPs
YPL leadership programme felt that
the workshop was a success and the
IAMP asks for its member academies
to continue supporting these young
leaders as they return to their countries
and support an annual programme and
development of a growing network
of young physician leaders. These
eorts will help address the dearth of
leadership training programmes for
young academicians and nurture them
as they learn to shape global health
policy for millions in need.
All authors were participants in the YPL programme.
The other participants were: Laila Asmal,
Davaalkham Dambadarjaa, Shen Yang Lim,
Ushanka Arjuna Bandara Medagama,
Rose Muhindo, Mayowa Ojo Owolabi,
Anthony Adebukola Oyekunle, Gustavo Patio,
Alexandre da Costa Pereira, Anis Safura Ramli,
Katharine Victoria Sedano Rojas,
Seif Abdallah Shekalaghe, Antonio Lucio Teixeira Jr,
and Derya Tilki.

*Rebekah E Gee, Tero Jarvinen,


Tanvira A Sultana, Raul Destura,
Biljana Gjoneska
rgee@lsuhsc.edu
Louisiana State University Schools of Public Health
and Medicine, New Orleans, LA 70130, USA (REG);
University of Tampere, Tampere, Finland (TJ);
Bangladesh Institute for Research and
Rehabilitation in Diabetes, Endocrine and Metabolic
Disorders, Shahbag, Dhaka, Bangladesh (TAS);
University of the Philippines, Manila, Philippines
(RD); and Macedonian Academy of Sciences and
Arts, Skopje, Macedonia (BG)
1
2

3
4

Holmes, D. Steve Wesselingh: the wizard of Oz.


Lancet 2011; 378: 1455.
Adli M, Kleinert S, Bonk M, Wesselingh S,
Ganten D. Science to policy: M8 Alliance
invites policy makers to step in. Lancet 2011;
378: 144749.
Goleman D. What makes a leader?
Harvard Bus Rev 1998; 76: 93102.
Taylor C, Farver C, Stoller JK. Perspective: can
emotional intelligence training serve as an
alternative approach to teaching
professionalism to residents? Acad Med 2011;
86: 155154.

One more reason to


fund the Global Fund
In your Editorial (Oct 1, p 1198),1 you
rightly stress ve reasons to ensure
continued nancing for the Global
Fund as it undergoes necessary
reforms. However, there is another
compelling, often unremarked, reason
to further support the Global Fund:
the organisation accelerates the
availability of better health products at
better prices to developing countries,
generating a global public good
that maximises the value for money
achieved by all global health donors.
Since its inception, the Global Fund
has mobilised massive purchasing
power that has created the conditions
necessary to substantially improve the
supply, design, and price of products
from antiretroviral drugs to malaria
treatments.2,3 These better products and
prices are accessible to most developing
countries, irrespective of the funding
source, and the Global Funds increasingly
sophisticated price tracking systems
ensure that its recipients pay those prices
where possible.4
To date, the Global Fund has
achieved this eect through a laissezwww.thelancet.com Vol 379 January 28, 2012

Correspondence

165

Price
Pre-intervention
trend

Volume
Pre-intervention
trend

155

140
120

Average price (US$)

145
100

135

80

Global Fund and


WHO actively
facilitate switch
to ACTs

125
115
105

60
40

095

Volume (millions of doses)

faire approach. The one exception, in


which the Global Fund substantially
accelerated the shift to superior
malaria treatment (artemisinin-based
combination therapies) in 2004, shows
the power of a more active approach
(gure). In May, 2011, the Global
Fund Board accordingly adopted an
ambitious new market shaping
strategy that will greatly increase the
funds impact through more active
engagement in product markets that
present particular opportunities or
risks.5 Application of this strategy to
antiretroviral drugs can save the Global
Fund and other donors hundreds of
millions of dollars, with a greater eect
as the strategy is extended to other
crucial health products.5
In this time of scal constraint and
unprecedented biomedical innovation, the world needs institutions like
the Global Fund that can successfully
use market forces to achieve greater
return on itsand othersinvestments and accelerate the poors access
to the best new products.

20

085
075

0
2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

Year

Figure: Price and volume of artemether-lumefantrine in the public sector before and after the Global Fund and WHO actively facilitate
the switch to ACTs
ACTs=artemisinin-based combination therapies. In 2004, the Global Fund, working with WHO, provided many of its recipient countries with
the impetus and nancing necessary to rapidly change from suboptimal therapies (eg, chloroquine) to ACTs.
5

Global Fund to Fight AIDS, Tuberculosis and


Malaria. 23rd board meeting: report of the
Market Dynamics and Commodities Ad Hoc
Committee. http://www.theglobalfund.org/
en/board/meetings/twentythird/documents/
(accessed Nov 15, 2011).

OS and SH are the Chair and Vice Chair, respectively,


of the Market Dynamics Committee of the Global
Fund Board, which is responsible for developing and
overseeing the organisations strategies for using its
resources to improve market outcomes for health
products. OS and SH both work for organisations
that have received funding from the Global Fund.
We are grateful to Aaron Woolsey, Arvind Nagarajan,
Neel Lakhani, and Rose Martin-Weiss for the their
support to this analysis.

*Oliver Sabot, Shanelle Hall


osabot@clintonhealthaccess.org
Clinton Health Access Initiative, New York,
NY 10027, USA
1
2

The Lancet. Five reasons to fund the Global


Fund. Lancet 2011; 378: 1198.
Waning B, Kyle M, Diedrichsen E, et al.
Intervening in global markets to improve
access to HIV/AIDS treatment: an analysis of
international policies and the dynamics of
global antiretroviral medicines markets.
Globalization Health 2010; 6: 9.
Vasan A, Hoos D, Mukherjee J, Farmer P,
Roseneld AG, Perriens JH. The pricing and
procurement of antiretroviral drugs: an
observational study of data from the Global
Fund. Bull World Health Organ 2006;
84: 39398.
Global Fund to Fight AIDS, Tuberculosis and
Malaria. 22nd board meeting: report of the
Market Dynamics and Commodities Ad Hoc
Committee. http://www.theglobalfund.org/
en/board/meetings/twentysecond/
documents/ (accessed Oct 11, 2011).

www.thelancet.com Vol 379 January 28, 2012

e26

Articles

Adjuvant capecitabine and oxaliplatin for gastric cancer


after D2 gastrectomy (CLASSIC): a phase 3 open-label,
randomised controlled trial
Yung-Jue Bang*, Young-Woo Kim, Han-Kwang Yang, Hyun Cheol Chung, Young-Kyu Park, Kyung Hee Lee, Keun-Wook Lee, Yong Ho Kim, Sang-Ik Noh,
Jae Yong Cho, Young Jae Mok, Yeul Hong Kim, Jiafu Ji, Ta-Sen Yeh, Peter Button, Florin Sirzn, Sung Hoon Noh*, for the CLASSIC trial investigators

Summary
Background D2 gastrectomy is recommended in US and European guidelines, and is preferred in east Asia, for
patients with resectable gastric cancer. Adjuvant chemotherapy improves patient outcomes after surgery, but the
benets after a D2 resection have not been extensively investigated in large-scale trials. We investigated the eect on
disease-free survival of adjuvant chemotherapy with capecitabine plus oxaliplatin after D2 gastrectomy compared
with D2 gastrectomy only in patients with stage IIIIIB gastric cancer.

Lancet 2012; 379: 31521

Methods The capecitabine and oxaliplatin adjuvant study in stomach cancer (CLASSIC) study was an open-label,
parallel-group, phase 3, randomised controlled trial undertaken in 37 centres in South Korea, China, and Taiwan.
Patients with stage IIIIIB gastric cancer who had had curative D2 gastrectomy were randomly assigned to receive
adjuvant chemotherapy of eight 3-week cycles of oral capecitabine (1000 mg/m twice daily on days 1 to 14 of each
cycle) plus intravenous oxaliplatin (130 mg/m on day 1 of each cycle) for 6 months or surgery only. Block
randomisation was done by a central interactive computerised system, stratied by country and disease stage.
Patients, and investigators giving interventions, assessing outcomes, and analysing data were not masked. The
primary endpoint was 3 year disease-free survival, analysed by intention to treat. This study reports a prespecied
interim ecacy analysis, after which the trial was stopped after a recommendation by the data monitoring committee.
The trial is registered at ClinicalTrials.gov (NCT00411229).

*Both authors contributed


equally

Findings 1035 patients were randomised (520 to receive chemotherapy and surgery, 515 surgery only). Median follow-up
was 342 months (254417) in the chemotherapy and surgery group and 343 months (256419) in the surgery only
group. 3 year disease-free survival was 74% (95% CI 6979) in the chemotherapy and surgery group and 59% (5364) in
the surgery only group (hazard ratio 056, 95% CI 044072; p<00001). Grade 3 or 4 adverse events were reported in
279 of 496 patients (56%) in the chemotherapy and surgery group and in 30 of 478 patients (6%) in the surgery only
group. The most common adverse events in the intervention group were nausea (n=326), neutropenia (n=300), and
decreased appetite (n=294).
Interpretation Adjuvant capecitabine plus oxaliplatin treatment after curative D2 gastrectomy should be considered as a
treatment option for patients with operable gastric cancer.
Funding F Homann-La Roche and Sano-Aventis.

Introduction
Gastric cancer is the second most common cause of
cancer-related mortality worldwide, with 988 000 new
cases and 736 000 deaths per year.1 Surgery is the main
treatment for operable gastric cancer; however, recurrence
rates are high (about 4080% in advanced cases).2,3 In east
Asia, particularly in Japan and South Korea, D2 gastrectomy
is the standard surgical treatment for localised gastric
cancer.4,5 In Europe, two randomised controlled studies
done in the UK6 and the Netherlands7 showed little initial
dierence between D1 and D2 surgery. However, longterm follow-up of the Dutch trial showed a reduction in
gastric cancer-specic deaths with extended surgery,8 a
nding that has been supported by ndings from a smaller
Taiwanese study.9 As a result of these ndings, D2
gastrectomy is now recommended in European10 and US
treatment guidelines11 for resectable disease.
www.thelancet.com Vol 379 January 28, 2012

Adjuvant chemotherapy is a standard component of


resectable gastric cancer therapy and improves patient
outcomes,12,13 although the preferred treatment diers
by geographical region. The recommended adjuvant
treatment is chemoradiotherapy in the USA,11 perioperative chemotherapy in the UK and parts of Europe,10
and adjuvant chemotherapy in Japan.4 These recommendations are based on the US Intergroup-011614 and
UK Medical Research Council Adjuvant Gastric Infusional
Chemotherapy (MAGIC)15 trials, which showed survival
benets with postoperative chemoradiotherapy and
perioperative chemotherapy, respectively, compared with
surgery alone. However, the benets were evident only
after limited dissection of regional lymph nodes in both
studies, raising questions about the need for postoperative
radiotherapy or perioperative chemotherapy after D2
gastrectomy.16,17 The Japanese recommendation for

Published Online
January 7, 2012
DOI:10.1016/S01406736(11)61873-4
See Comment page 291

CLASSIC trial investigators


listed at the end of Article
Department of Internal
Medicine (Prof Y-J Bang MD),
and Department of Surgery
(Prof H-K Yang MD), Seoul
National University College of
Medicine, Seoul, South Korea;
Gastric Cancer Branch, Research
Institute and Hospital,
National Cancer Center,
Goyang-si, Gyeonggi-do,
South Korea (Y-W Kim MD);
Department of Surgery
(Prof S H Noh MD), and
Department of Medical
Oncology, Yonsei Cancer
Center, Cancer Metastasis
Research Centre
(Prof H C Chung MD,
Prof J Y Cho MD), Yonsei
University College of Medicine,
Seoul, South Korea;
Department of Surgery,
Chonnam National University
Hwasun Hospital, Jeonnam,
South Korea (Prof Y-K Park MD);
Yeungnam University College
of Medicine, Daegu, South
Korea (Prof K H Lee MD);
Department of Internal
Medicine, Seoul National
University Bundang Hospital,
Seongnam, Gyeonggi-do,
South Korea (Prof K-W Lee MD);
Department of Surgery, Kyung
Hee University School of
Medicine, Seoul, South Korea
(Prof Yo H Kim MD);
Department of Surgery, Seoul
Veterans Hospital, Seoul,
South Korea (S-I Noh MD);
Department of Surgery
(Prof Y J Mok MD) and,
Department of Internal
Medicine (Prof Ye H Kim MD),
Korea University College of
Medicine, Seoul, South Korea;
Beijing Cancer Hospital,

315

Articles

Beijing, China (Prof J Ji MD);


Department of Surgery, Chang
Gung Memorial Hospital at
Linkou, Taipei, Taiwan
(Prof T-S Yeh MD); Infopeople,
Sydney, NSW, Australia
(P Button MSc); and
F Homann-La Roche, Basel,
Switzerland (F Sirzn MD)
Correspondence to:
Prof Yung-Jue Bang, Department
of Internal Medicine, Seoul
National University College of
Medicine, Jongno-gu,
Seoul 110 744, South Korea
bangyj@snu.ac.kr

adjuvant therapy is based on the Adjuvant Chemotherapy


Trial of TS-1 for Gastric Cancer (ACTS-GC) study,13,18
which showed a survival benet with adjuvant
chemotherapy after D2 gastrectomy compared with
surgery alone. A subgroup analysis of the data showed a
survival benet for stages II and IIIA disease.
Increased acceptance of D2 gastrectomy raises new
questions about the optimum adjuvant therapy for
patients with operable gastric cancer. The Capecitabine
and Oxaliplatin Adjuvant Study in Stomach Cancer
(CLASSIC) study was designed to compare the eect
of adjuvant capecitabine plus oxaliplatin after D2
gastrectomy with surgery alone on disease-free survival
in patients with stage II or III gastric cancer. Although
several phase 2 trials have shown the activity of
capecitabine and oxaliplatin in advanced gastric
cancer,1922 phase 3 data for this regimen in gastric
cancer are absent. We summarise the results of the
prespecied interim analysis from the CLASSIC study,
which was done after a recommendation by the
independent data monitoring committee to fully assess
and report the study.

Methods
Study design and patients

See Online for webappendix

CLASSIC was a randomised, open-label, multicentre,


parallel-group, phase 3 study. The trial was done in
37 centres in South Korea, China, and Taiwan.
Eligible patients were ambulatory; aged 18 years or
older; had histologically conrmed, American Joint
Committee on Cancer/Union Internationale Contre le
Cancer23 stage II (T2N1, T1N2, T3N0), IIIA (T3N1, T2N2,
T4N0), or IIIB (T3N2) gastric adenocarcinoma with no
evidence of metastatic disease; and had had D2 surgery

1035 randomised

515 received surgery only

429 completed
observation phase

22 lost to
follow-up
84 died

515 analysed

86 did not complete


observation phase

520 received capecitabine


and oxaliplatin

346 completed
treatment phase

174 did not complete


treatment phase

18 lost to
follow-up
61 died

520 analysed

Figure 1: Trial prole


Numbers of patients eligible not available. Reasons for withdrawal during the study treatment or observation
phase are given in the webappendix.

316

and achieved R0 resection. Patients were included only if


they had a Karnofsky performance status of 70% or more.
Patients who had had chemotherapy, immunotherapy, or
radiotherapy for gastric cancer were excluded. Patients
had to have adequate renal function (creatinine clearance
>50 mL/min or serum creatinine 15 times the upper
limit of normal [ULN]), hepatic function (total bilirubin
15 times the ULN, aspartate or alanine aminotransferase
25 times the ULN, alkaline phosphatase 25 times the
ULN), and haematological function (absolute neutrophil
count 1510/L or platelet count 10010/L).
The protocol was approved by the institutional review
board at each participating institution, and the study was
done in accordance with the Declaration of Helsinki and
the Good Clinical Practice Guidelines dened by the
International Conference on Harmonization. All patients
provided written, informed consent.

Randomisation and masking


Patients were randomly assigned to receive capecitabine
and oxaliplatin or surgery alone in a 1:1 ratio. Randomisation was done after surgery with a centralised interactive
computerised system and stratied by country and disease
stage (II, IIIA, and IIIB). A random permuted block design
(with a block size of four) was used in each combination
of country and stage of disease stratum. Because study
centre was not a stratication factor, a particular study
centre could not predict the next allocation, or even know
how many numbers in the block had already being
allocated if they had correctly guessed the block size.
Patients, and investigators giving interventions, assessing
outcomes, and analysing data were not masked.

Procedures
All patients had curative D2 gastrectomy within 6 weeks
before randomisation. At least 15 lymph nodes were
examined to ensure adequate disease classication. All
surgeons had experience doing this type of surgery
(>50 procedures per year). To further ensure the quality of
surgery, standard operating procedures were predened
and given to all surgeons before the start of the study, and
surgery was photographed.
Patients assigned to the chemotherapy group received
eight 3-week cycles of oral capecitabine (1000 mg/m
twice daily on days 114 of each cycle) plus intravenous
oxaliplatin (130 mg/m on day 1 of each cycle). Dose
reductions or interruptions were allowed to manage
potentially serious or life-threatening adverse events.
Subsequent dose escalations for capecitabine were
allowed after two more cycles in the absence of
grade 24 toxic eects, if oxaliplatin was permanently
discontinued. For toxic eects judged to be due to only
one drug, the dose of the other drug was not modied. In
cases of oxaliplatin-related neurological adverse events,
capecitabine could be continued as monotherapy.
Oxaliplatin monotherapy was not allowed if capecitabine
was discontinued.
www.thelancet.com Vol 379 January 28, 2012

Articles

Prespecied tumour assessments to assess whether


patients were disease free were done by abdominal CT or
MRI every 6 months during the rst 3 years and yearly
thereafter, and by chest radiograph every 3 months for
the rst 2 years, every 6 months for the subsequent year,
and yearly thereafter. If signs or symptoms indicated a
possible recurrence or development of a new gastric
cancer, investigations were then done to verify whether
the patient was disease free. The same assessment was
used for each patient.
Adverse events were graded according to the National
Cancer Institutes Common Terminology Criteria for
Adverse Events (version 3.0). Adverse events were
documented during chemotherapy and for 28 days after
the last dose of study medication. In the surgery only
group, adverse events were recorded for up to 190 days
after randomisation. Relative dose intensity was dened
as the dose received divided by the planned dose for the
eight treatment cycles.
The primary endpoint was 3 year disease-free survival,
dened as the time from randomisation to the time of
recurrence of the original gastric cancer, development
of a new gastric cancer, or death from any cause.
Secondary endpoints were overall survival (dened as
the time from the date of randomisation to date of death
from any cause) and safety (any adverse event).

Statistical analysis
3 year disease-free survival in the oxaliplatin and
capecitabine and surgery only groups was predicted to be
650% and 562%, respectively (hazard ratio [HR] 075,
with the assumption of a 3 year dropout rate of 10% with
recruitment for 12 months). The null hypothesis (no
dierence in 3 year disease-free survival between study
groups) was tested with Cox proportional hazards
regression stratied by country and disease stage, with
covariates of age, sex, and nodal metastatic status. A
sample size of 512 patients in each group was planned to
record 385 disease-free survival events with 80% power
at a 005 signicance level by a two-sided log-rank test.
An interim ecacy analysis was scheduled for
after 257 events (668% of the number of events at
nal analysis), with stopping boundaries decided by
application of the Lan-DeMets method24 and OBrienFleming spending function25 using the actual number of
events. Interim ecacy analyses for disease-free survival
and overall survival were done with Cox proportional
hazards regression by intention to treat. All ecacy
analyses were repeated per protocol to test the sensitivity of
the results (data not shown). Safety was assessed per
protocol.
Time to event endpoints were analysed with KaplanMeier survival methods. Estimates of treatment eect
were calculated as HRs with 95% CIs. Study treatment
groups were compared with a two-sided log-rank test.
We also did a prespecied analysis of disease-free
survival in subgroups.
www.thelancet.com Vol 379 January 28, 2012

This Article reports a planned interim analysis of data,


which was triggered because the necessary number of
events had occurred. The trial was then stopped after a
recommendation by the independent data monitoring
committee, who reviewed the data in March, 2011.
The trial was registered at ClinicalTrials.gov
(NCT00411229).

Role of the funding source


The study sponsor helped to design the study, to interpret
data, and helped with the decision to submit the report
for publication in conjunction with the authors.
Employees of the sponsor collected, managed, and
analysed data. The sponsors funded writing assistance.
The two principal investigators (Y-JB and SHN) had full
access to all study data and had nal responsibility for
the decision to submit for publication.
Surgery only (n=515)
Age (years)
Men

558 (116)
358 (70%)

Karnofsky performance status (%)

Capecitabine and oxaliplatin


(n=520)
561 (111)
373 (72%)

90% (90100)

90% (90100)

Body surface area (m)

162 (015)

162 (015)

Time since surgery (months)

114 (017)

114 (017)

AJCC/UICC23 stage
IB

0 (0%)

II

261 (51%)

253 (49%)

IIIA

184 (36%)

193 (37%)

IIIB

69 (13%)

73 (14%)

IV

1 (<1%)

1 (<1%)

0 (0%)

Tumour stage
T1

3 (1%)

8 (2%)

T2

282 (55%)

282 (54%)

T3

229 (44%)

227 (44%)

T4

1 (<1%)

3 (1%)

Tumour location*
Antrum

234 (45%)

237 (46%)

Body

172 (33%)

166 (32%)

Body and antrum

29 (6%)

31 (6%)

Fundus

40 (8%)

46 (9%)

Fundus and body

13 (3%)

10 (2%)

Gastro-oesophageal junction

9 (2%)

15 (3%)

Whole gastric

6 (1%)

6 (1%)

Other

12 (2%)

Lymph nodes examined

436 (167)

9 (2%)
450 (174)

Nodal status
N0

56 (11%)

47 (9%)

N1

308 (60%)

313 (60%)

N2

151 (29%)

160 (31%)

Data are mean (SD), n (%), or median (IQR). Intention-to-treat population; all patients were Asian.
AJCC/UICC=American Joint Cancer Committee/Union Internationale Contre le Cancer. *Antrum is the lower third,
body the middle third, and fundus the upper third. Includes multiple localisations.

Table 1: Baseline patient characteristics

317

Articles

A
Surgery only
Capecitabine and oxaliplatin

10
09

Chance of being disease free

08
07
06
05
04
03
02
01
0
36

39

42

45

48

51

Number at risk
Surgery only 515 442 415 388 353 332 289 255 211 188 148 120 58
Capecitabine 520 462 442 425 409 379 332 295 246 218 166 147 73
and oxaliplatin

25
31

22
30

20
25

6
10

0
0

39

42

45

48

51

Number at risk
Surgery only 515 473 460 453 443 424 380 338 288 238 204 156 112 43
Capecitabine 520 481 468 461 451 425 396 352 306 255 217 169 120 39
and oxaliplatin

34
36

31
33

12
16

0
0

12

15

18

21

24

27

30

33

B
10
09
08

Chance of survival

07
06
05
04
03
02
01
0
0

12

15

18

21 24 27 30
Time (months)

33

36

Figure 2: 3 year disease-free survival (A) and preliminary overall survival (B) in the intention-to-treat population

Results
Figure 1 shows the trial prole. Between June, 2006, and
June, 2009, 1035 patients were randomly assigned to
receive either oxaliplatin and capecitabine (n=520) or
surgery only (n=515; intention-to-treat population). The
trial was stopped because accrual was complete and the
necessary 257 disease-free survival events had occurred.
The median duration of follow-up was 342 months
(254417) in the chemotherapy group and 343 months
(256419) in the surgery only group; 468 patients were
followed up in the chemotherapy group and 463 in the
surgery only group. Patient demographic and baseline
disease characteristics were similar in each group
(table 1).
At the cuto for the interim analysis, 106 patients (20%)
in the chemotherapy group had relapsed, developed a new
318

gastric cancer, or died, compared with 163 patients (32%)


in the surgery only group. 3 year disease-free survival was
higher in the chemotherapy group than in the surgery only
group (HR 056, 95% CI 044072; p<00001). 3 year
disease-free survival was 74% (95% CI 6979) in the
chemotherapy group and 59% (5364) in the surgery only
group. Kaplan-Meier curves for disease-free survival show
early separation between the two study groups (gure 2A).
65 patients (13%) died in the chemotherapy group
compared with 85 (17%) in the surgery only group. 3 year
overall survival was 83% (95% CI 7987) in the
chemotherapy group and 78% (7483) in the surgery only
group (HR 072, 95% CI 052100; p=00493; gure 2B),
although a robust estimate of median overall survival is
not yet available.
In the chemotherapy group, 96 patients (18%) developed
recurrences or new occurrences of gastric cancer
compared with 155 patients (30%) in the surgery only
group. The sites of gastric cancer recurrence or new
occurrence were the peritoneum (47 in the chemotherapy
group vs 56 in the surgery only group), locoregional sites
(21 vs 44), and distant sites (49 vs 78).
Subgroup analysis of 3-year disease-free survival showed
consistent benet for oxaliplatin and capecitabine
compared with surgery only for several factors (gure 3).
Survival was signicantly higher in the chemotherapy
group than in the surgery only group for all disease stages
(gure 3). 3 year disease-free survival for stage II disease
was 85% (95% CI 7990) for chemotherapy and 71%
(6478) for surgery alone, for stage IIIa disease 66% (5775)
for chemotherapy and 51% (4260) for surgery alone, and
for stage IIIb disease 61% (4873) for chemotherapy and
33% (95% CI 15%51%) for surgery alone.
For nodal status 1 and 2 subgroups 3 year disease-free
survival was signicantly improved with oxaliplatin and
capecitabine compared with surgery only, but not for
nodal status 0 (gure 3).
Of the 1035 randomised patients, 61 were excluded
from the safety population (25 in the chemotherapy
group vs 36 in the surgery only group). The main reasons
for exclusion were absence of follow-up information
(18 vs 36) and not receiving study treatment after
randomisation (22 in the chemotherapy group). Additionally, one patient in the surgery only group received
chemotherapy in error, and was included in the
chemotherapy group for the safety analysis. The safety
population included 496 patients in the chemotherapy
group and 478 patients in the surgery only group.
346 patients (67%) assigned to the chemotherapy group
received eight cycles as planned. 167 patients had
capecitabine dose reductions, 147 had cycle interruptions,
and 369 had cycle delays, and 163 patients needed
oxaliplatin dose reductions. The median relative dose
intensity was 85% for capecitabine and 98% for
oxaliplatin.
Table 2 shows adverse events reported by 10% or
more of patients. Almost nine times as many grade 3 or
www.thelancet.com Vol 379 January 28, 2012

Articles

4 adverse events were reported in the chemotherapy


group than in the surgery group. The most commonly
reported adverse events at any grade in the chemotherapy
group were nausea, neutropenia, decreased appetite,
peripheral neuropathy, diarrhoea, and vomiting (table 2).
The most common grade 3 or 4 adverse events in
the chemotherapy group were neutropenia, thrombocytopenia, nausea, and vomiting (table 2). Grade 3 or
4 peripheral neuropathy, a cumulative toxic eect
associated with oxaliplatin, occurred in 12 (2%) patients.
43 patients (9%) had serious adverse events related to
chemotherapy. Two patients died within 28 days of the
last dose of treatment in the chemotherapy group: one
because of sepsis (possibly related to treatment), and one
because of metastatic gastric cancer (judged unrelated to
the study treatment). One patient in the surgery only
group died because of cardiac failure during the
observation phase.
Adverse events led to chemotherapy dose modications
in 446 (90%) patients; neutropenia, nausea, vomiting,
thrombocytopenia, and decreased appetite were the
most common reasons (webappendix). Discontinuations
because of adverse events in the chemotherapy group
occurred in 50 (10%) patients, mainly as a result of
neutropenia (n=17; 3%), thrombocytopenia (n=5; 1%),
and vomiting (n=5; 1%).

China and Taiwan

125

066 (033132)

South Korea

910

056 (043073)

II

515

055 (036084)

IIIA

377

057 (039082)

IIIB

143

057 (035095)

<65

766

062 (047083)

65

269

048 (030078)

Women

304

083 (054127)

Men

731

049 (036066)

N0

103

090 (041197)

N1

621

062 (044089)

N2

311

045 (031066)

<57

508

063 (046088)

57

527

052 (036075)

1035

058 (045074)

Stage of disease

Age (years)

Sex

Nodal status

Weight (kg)

Overall

Discussion
This study shows that a 6 month course of chemotherapy
after D2 gastrectomy improves 3 year disease-free survival
compared with surgery only. Chemotherapy reduced the
relative risk of disease recurrence, new disease
occurrence, or death compared with surgery alone.
Moreover, a subgroup analysis suggested that adjuvant
capecitabine and oxaliplatin was benecial for all disease
stages (II, IIIA or IIIB). The overall survival data from
our study are not yet mature; however, the data suggest
an improvement in overall survival with capecitabine and
oxaliplatin compared with surgery only. An analysis after
a median follow-up of 5 years is planned to conclusively
establish the overall survival benet of capecitabine and
oxaliplatin in this setting.
We used 3 year disease-free survival as the primary
endpoint because most relapses occur within 3 years of
surgery.26 Although 3 year disease-free survival has not yet
been formally validated as a surrogate measure,
preliminary data from the Global Advanced/Adjuvant
Stomach Tumor Research International Collaboration
(GASTRIC) group indicate that 3 year disease-free survival
is strongly correlated with 5 year overall survival, the
benchmark for judging eectiveness of adjuvant therapy
in gastric cancer.27 The clinical relevance of disease-free
survival in gastric cancer is supported by the GASTRIC
group meta-analysis of 17 trials, which showed an 18%
relative risk reduction for both disease-free survival and
overall survival with adjuvant chemotherapy compared
with surgery only in patients with resectable disease.12
www.thelancet.com Vol 379 January 28, 2012

HR (95% CI)

Country

04
06
1
2
Favours capecitabine and oxaliplatin Favours surgery only

Figure 3: 3 year disease-free survival by stratication and prognostic factors in the intention-to-treat population

Surgery alone (n=478)

Capecitabine and oxaliplatin


(n=496)

All grades

Grade 3 or 4

All grades

Grade 3 or 4

253 (53%)

30 (6%)

490 (99%)

279 (56%)

20 (4%)

0 (0%)

326 (66%)

39 (8%)

1 (<1%)

300 (60%)

107 (22%)

18 (4%)

1 (<1%)

294 (59%)

23 (5%)

0 (0%)

0 (0%)

277 (56%)

12 (2%)

Diarrhoea

52 (11%)

1 (<1%)

236 (48%)

9 (2%)

Vomiting

16 (3%)

0 (0%)

191 (39%)

37 (7%)

Fatigue

Patients with 1 adverse event


Nausea
Neutropenia
Decreased appetite
Peripheral neuropathy

4 (<1%)

11 (2%)

0 (0%)

156 (31%)

23 (5%)

Thrombocytopenia

0 (0%)

0 (0%)

130 (26%)

40 (8%)

Handfoot syndrome

0 (0%)

0 (0%)

93 (19%)

5 (1%)

Asthenia

5 (1%)

0 (0%)

87 (18%)

10 (2%)
8 (2%)

Abdominal pain

42 (9%)

2 (<1%)

85 (17%)

Constipation

17 (4%)

0 (0%)

63 (13%)

1 (<1%)

Dizziness

25 (5%)

0 (0%)

64 (13%)

3 (<1%)
3 (<1%)

Stomatitis, all
Weight decreased
Peripheral sensory neuropathy

0 (0%)

0 (0%)

59 (12%)

13 (3%)

2 (<1%)

59 (12%)

1 (<1%)

0 (0%)

0 (0%)

50 (10%)

3 (<1%)

Data are n (%). *Patients who received at least one dose of capecitabine or oxaliplatin.

Table 2: Adverse events reported by 10% of patients (safety population*)

319

Articles

Before CLASSIC, the Japanese ACTS-GC trial was the


only large randomised trial of adjuvant chemotherapy
(1 year of the oral uoropyrimidine S-1) after curative
resection with D2 gastrectomy.13 Our ndings accord
with the ACTS-GC trial, and together these studies show
that adjuvant chemotherapy after D2 gastrectomy
improves outcomes in patients with resectable gastric
cancer (panel).
We used surgery only as the reference treatment for
our trial because it was the appropriate choice when the
CLASSIC trial was designed. At that time, no large-scale
randomised trials supporting the use of adjuvant therapy
after D2 surgery had been done. The ndings of the
ACTS-GC trial,13 which established uoropyrimidine
monotherapy as an eective adjuvant regimen after D2
surgery, were reported in 2007, after recruitment for the
CLASSIC trial had begun. Both postoperative chemoradiotherapy and perioperative chemotherapy have been
adopted as standard treatments in the USA and Europe,
but their relevance after D2 surgery is unknown.
The frequency, severity, and type of adverse events
documented with capecitabine and oxaliplatin in the
CLASSIC study were consistent with the safety prole
reported with adjuvant capecitabine and oxaliplatin
in colon cancer.28,29 In our study more than half of patients
who received chemotherapy had peripheral neuropathya
cumulative, dose-related toxic eect associated with
oxaliplatinalthough grade 3 or 4 events were infrequent.
Most patients needed dose modications because of
adverse events, the most common being neutropenia,
nausea, and vomiting. Capecitabine, should be started at
the recommended dose with prompt dose modications
as needed for emergent adverse events.
A key question about our trial, as with any trial done in
one geographical region, is how generalisable the ndings
Panel: Research in context
Systematic review
A meta-analysis from the GASTRIC group based on individual patient data from
17 randomised clinical trials12 showed that adjuvant chemotherapy with a uorouracil-based
regimen improves overall survival and disease-free survival compared with surgery only.
However, the type of gastrectomy used in the studies was not standardised. D2 gastrectomy
is now recommended in US11 and European10 treatment guidelines, and is widely used in east
Asia as the preferred type of surgery for patients with operable gastric cancer. However, the
benets of adjuvant chemotherapy after this more extensive type of resection were
uncertain. To address this question, two phase 3 trials were initiated (ACTS-GC and CLASSIC).
The ACTS-GC trial showed a survival benet with adjuvant uoropyrimidine monotherapy
after D2 gastrectomy compared with surgery only.13
Interpretation
The CLASSIC trial shows that adjuvant therapy with capecitabine and oxaliplatin improves
3 year disease-free survival after D2 gastrectomy compared with D2 gastrectomy only.
Improved overall survival was also evident after only 3 years, although the data are
immature and patient follow-up is ongoing. Together, the ACTS-GC and CLASSIC trials
show that uoropyrimidine-based adjuvant therapy improves outcomes after D2
gastrectomy in patients with operable gastric cancer.

320

are to other regions where disease management practices


might dier. The CLASSIC trial had good patient
outcomes; the 3 year overall survival rate in the surgery
only group was substantially higher than that in the US
Intergroup-0116 and UK MAGIC populations (78% in
CLASSIC vs 3040% in Intergroup-0116 and MAGIC).14,15
Although our patient population had fewer T3 and
T4 lesions than did US and European gastric cancer
populations (44% in CLASSIC vs 68% in Intergroup-0116
vs 64% in MAGIC),14,15 node-positive disease was more
frequent (90% vs 85% vs 72%).14,15 The dierence in
outcomes seems unlikely to be due to prognostic
disparities, although the possibility of intrinsic biological
dierences in gastric cancer by region has been
suggested.30 Instead, we suggest that the favourable
outcomes in our study are a result of the consistent use of
D2 surgery and the high quality of that surgery (ensured
by prospectively dened standard operating procedures
and sampling of at least 15 lymph nodes). Now that D2
gastrectomy is standard of care in both Europe10 and the
USA,11 our study ndings could be highly relevant for, and
might be generalisable to, other regions when D2 surgery
is done by experienced surgeons.
On the basis of our ndings, we suggest that a trial of
capecitabine and oxaliplatin plus trastuzumab after D2
gastrectomy in patients with human epidermal growth
factor receptor (HER)-2-positive operable disease is
warranted. Such a study would build on the ToGA trial,31
which showed improved overall survival with trastuzumab
plus chemotherapy in patients with HER-2-positive
advanced disease.
In conclusion, ndings from the CLASSIC trial support
the use of adjuvant capecitabine and oxaliplatin as a new
treatment option for patients with resectable disease.
Contributors
All authors had access to all the study data, analysed data, interpreted
data, and reviewed, edited and approved the report for publication. Y-JB
and SHN designed the study. Y-JB, Y-WK, H-KY, HCC, Y-KP, KHL,
K-WL, YHK, S-IN, JYC, YJM, YHK, JFJ, T-SY, and SHN recruited
patients and collected data. PB undertook the statistical analysis.
CLASSIC trial investigators
Y I Park (National Cancer Center, Goyang, South Korea); Y-J Bang
(Seoul National University College of Medicine, Seoul, South Korea);
H C Chung, J Y Cho (Yonsei Cancer Center, Younsei University College
of Medicine, Seoul, South Korea); Y-K Park (Chonnam National
University Hwasun Hospital, Jeonnam, South Korea); K H Lee
(Yeungnam University College of Medicine, Daegu, South Korea);
K W Lee (Seoul National University Bundang Hospital, Seongnam,
South Korea); S Y Kim (Kyung Hee University Hospital, Seoul, South
Korea); S I Noh (Veterans Health Service Medical Center, Seoul, South
Korea); Y J Mok, Y H Kim (Korea University College of Medicine, Seoul,
South Korea); J-H Choi (Ajou University Hospital, Suwon, South
Korea); H-M Chin (St Vincents Hospital, Suwon, South Korea); W Yu
(Kyungpook National University Hospital, Daegu, South Korea); J-I Lee
(Korea Cancer Center Hospital, Seoul, South Korea); D-B Shin (Gachon
University Gil Hospital, Incheon, South Korea); H S Park
(Soonchunhyang University Hospital, Seoul, South Korea); H S Song
(Keimyung University Hospital Dongsan Medical Center, Daegu, South
Korea); D H Yang (Chonbuk National University Hospital, Jeonju, South
Korea); S H Lee (Kosin University Gospel Hospital, Busan, South
Korea); S N Lee (Ewha Womans University School of Medicine, Seoul,
South Korea); JF Ji (Beijing Cancer Hospital, Beijing, China); L Han

www.thelancet.com Vol 379 January 28, 2012

Articles

(Tianjin Medical University Cancer Institute and Hospital, Tianjin,


China); Y Zhan (Sun Yat-Sen University Cancer Center, Guangzhou,
Guangdong, China); Y L He (The First Aliated Hospital, Sun Yat-Sen
University, Guangzhou, Guangdong, China); Y P Liu (First Clinical
College of China Medical University, Shenyang, Liaoning, China);
S Wang (Peking University Peoples Hospital, Beijing, China);
C M Huang (Fujian Medical University Union Hospital, Fujian, China);
Y H Sun (Zhongshan Hospital, Fudan University, Shanghai, China);
M Yan (Rudin Hospital, Shanghai Jiao Tong University School of
Medicine, Shanghai, China); Y Q Shi (Fudan University Shanghai
Cancer Center, Shanghai, China); T-S Yeh (Chang Gung Memorial
Hospital, Taoyuan, Taiwan); P-H Lee (National Taiwan University
Hospital, Taipei, Taiwan); C-S Chang (Changhua Christian Hospital,
Changhua, Taiwan); P-W Lin (National Cheng Kung University
Hospital, Tainan, Taiwan); R-K Hsieh (Mackay Memorial Hospital,
Taipei, Taiwan); and S-Y Chen (Veterans General Hospital, Cancer
Center, Taipei, Taiwan).
Conicts of interest
PB was employed by Roche Products Australia at the time of the study
and is currently employed by Infopeople. FS is employed by
F Homann-La Roche and owns stock in the company. Y-JB has acted as
a consultant for F Homann-La Roche and Sano-Aventis, and has
received honoraria from F Homan-La Roche. YHK has received
honoraria and acted as a consultant for F Homan-La Roche and
Sano-Aventis. The other investigators declare that they have no
conicts of interest.
Acknowledgments
We thank the patients, the patients families, and the investigators,
co-investigators, nurses, and study teams at the 37 centres in
South Korea, China, and Taiwan. CLASSIC was sponsored by
F Homann-La Roche and Sano-Aventis. Writing assistance was
provided for this report by Miller Medical Communications, funded by
F Homann-La Roche.
References
1
Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM.
GLOBOCAN 2008, Cancer Incidence and Mortality Worldwide:
IARC CancerBase No. 10. Lyon, France: International Agency for
Research on Cancer; 2010. Available from: http://globocan.iarc.fr/
factsheets/cancers/stomach.asp (accessed Oct 12, 2011).
2
Gallo A, Cha C. Updates on esophageal and gastric cancers.
World J Gastroenterol 2006; 12: 323742.
3
Gunderson LL. Gastric cancerpatterns of relapse after surgical
resection. Semin Radiat Oncol 2002; 12: 15061.
4
Japanese Gastric Cancer Society. Guidelines for Diagnosis and
Treatment of Carcinoma of the Stomach. http://www.jgca.jp/
PDFles/Guidelines2004_eng.pdf (accessed Oct 12, 2011).
5
Park JM, Kim YH. Current approaches to gastric cancer in Korea.
Gastrointest Cancer Res 2008; 2: 13744.
6
Cuschieri A, Weeden S, Fielding J, et al, and the Surgical Co-operative
Group. Patient survival after D1 and D2 resections for gastric cancer:
long-term results of the MRC randomized surgical trial. Br J Cancer
1999; 79: 152230.
7
Bonenkamp JJ, Hermans J, Sasako M, et al, and the Dutch Gastric
Cancer Group. Extended lymph-node dissection for gastric cancer.
N Engl J Med 1999; 340: 90814.
8
Songun I, Putter H, Kranenbarg EM, Sasako M, van de Velde CJ.
Surgical treatment of gastric cancer: 15-year follow-up results of the
randomised nationwide Dutch D1D2 trial. Lancet Oncol 2010;
11: 43949.
9
Wu CW, Hsiung CA, Lo SS, et al. Nodal dissection for patients with
gastric cancer: a randomised controlled trial. Lancet Oncol 2006;
7: 30915.
10 Okines A, Verheij M, Allum W, Cunningham D, Cervantes A, and
the ESMO Guidelines Working Group. Gastric cancer: ESMO
Clinical Practice Guidelines for diagnosis, treatment and follow-up.
Ann Oncol 2010; 21 (suppl 5): v5054.
11 NCCN. NCCN Clinical Practice Guidelines in Oncology. Gastric
Cancer. Version 2.2011. 2011; http://www.nccn.org (accessed Dec 5,
2011).

www.thelancet.com Vol 379 January 28, 2012

12

13

14

15

16

17

18

19

20

21

22

23
24
25
26

27

28

29

30

31

Paoletti X, Oba K, Burzykowski T, et al, and the GASTRIC (Global


Advanced/Adjuvant Stomach Tumor Research International
Collaboration) Group. Benet of adjuvant chemotherapy for
resectable gastric cancer: a meta-analysis. JAMA 2010; 303: 172937.
Sakuramoto S, Sasako M, Yamaguchi T, et al, and the ACTS-GC
Group. Adjuvant chemotherapy for gastric cancer with S-1, an oral
uoropyrimidine. N Engl J Med 2007; 357: 181020.
Macdonald JS, Smalley SR, Benedetti J, et al. Chemoradiotherapy
after surgery compared with surgery alone for adenocarcinoma of
the stomach or gastroesophageal junction. N Engl J Med 2001;
345: 72530.
Cunningham D, Allum WH, Stenning SP, et al, and the MAGIC Trial
Participants. Perioperative chemotherapy versus surgery alone for
resectable gastroesophageal cancer. N Engl J Med 2006; 355: 1120.
Ohtsu A, Sasako M. Overview of adjuvant therapy for resected
gastric cancer: dierences in Japan and the United States.
Semin Oncol 2005; 32 (suppl 9): S10104.
Ohtsu A, Yoshida S, Saijo N. Disparities in gastric cancer
chemotherapy between the East and West. J Clin Oncol 2006;
24: 218896.
Sasako M, Sakuramoto S, Katai H, et al. Five-year outcomes of a
randomized phase III trial comparing adjuvant chemotherapy with
S-1 versus surgery alone in stage II or III gastric cancer. J Clin Oncol
2011; 29: 438793.
Jatoi A, Murphy BR, Foster NR, et al, and the North Central Cancer
Treatment Group. Oxaliplatin and capecitabine in patients with
metastatic adenocarcinoma of the esophagus, gastroesophageal
junction and gastric cardia: a phase II study from the North Central
Cancer Treatment Group. Ann Oncol 2006; 17: 2934.
Liu C, Sun Q, Hang X, Zhong B, Wang D. Multicenter phase II
study of capecitabine plus oxaliplatin as a rst-line therapy in
Chinese patients with advanced gastric cancer. Anticancer Drugs
2008; 19: 82531.
Luo HY, Xu RH, Wang F, et al. Phase II trial of XELOX as rst-line
treatment for patients with advanced gastric cancer. Chemotherapy
2010; 56: 94100.
Park YH, Lee JL, Ryoo BY, et al. Capecitabine in combination with
oxaliplatin (XELOX) as a rst-line therapy for advanced gastric
cancer. Cancer Chemother Pharmacol 2008; 61: 62329.
American Joint Committee on Cancer. Cancer Staging Manual.
6th ed. Chicago: Springer; 2003.
Lan KKG, DeMets DL. Discrete sequential boundaries for clinical
trials. Biometrika 1983; 70: 65963.
Jenison C, Turnbull B. Group sequential methods with applications
to clinical trials. Boca Raton: Chapman and Hall/CRC, 2000.
Deng J, Liang H, Wang D, Sun D, Pan Y, Liu Y. Investigation of the
recurrence patterns of gastric cancer following a curative resection.
Surg Today 2011; 41: 21015.
Rougier P, Sakamoto J. Surrogate endpoints for overall survival in
resectable gastric cancer and in advanced gastric carcinoma:
analysis of individual data from the gastric collaboration. Ann Oncol
2011; 22 (suppl 5): v10v18 (abstr O-0013).
Haller DG, Tabernero J, Maroun J, et al. Capecitabine plus
oxaliplatin compared with uorouracil and folinic acid as adjuvant
therapy for stage III colon cancer. J Clin Oncol 2011; 29: 146571.
Schmoll HJ, Cartwright T, Tabernero J, et al. Phase III trial of
capecitabine plus oxaliplatin as adjuvant therapy for stage III colon
cancer: a planned safety analysis in 1,864 patients. J Clin Oncol 2007;
25: 10209.
Strong VE, Song KY, Park CH, et al. Comparison of gastric cancer
survival following R0 resection in the United States and Korea
using an internationally validated nomogram. Ann Surg 2010;
251: 64046.
Bang YJ, Van Cutsem E, Feyereislova A, et al, and the ToGA Trial
Investigators. Trastuzumab in combination with chemotherapy
versus chemotherapy alone for treatment of HER2-positive
advanced gastric or gastro-oesophageal junction cancer (ToGA):
a phase 3, open-label, randomised controlled trial. Lancet 2010;
376: 68797.

321

Articles

Self-monitoring of oral anticoagulation: systematic review


and meta-analysis of individual patient data
Carl Heneghan, Alison Ward, Rafael Perera, and The Self-Monitoring Trialist Collaboration

Summary
Lancet 2012; 379: 32234
Published Online
December 1, 2011
DOI:10.1016/S01406736(11)61294-4
See Comment page 292
Oxford University, Department
of Primary Care Health
Sciences, Oxford, UK
(C Heneghan DPhil, A Ward PhD,
R Perera DPhil)
Correspondence to:
Dr Carl Heneghan, Oxford
University, Department of
Primary Care Health Sciences,
2338 Hythe Bridge St, Oxford,
OX1 2ET, UK
carl.heneghan@phc.ox.ac.uk

Background Uptake of self-testing and self-management of oral coagulation has remained inconsistent, despite good
evidence of their eectiveness. To clarify the value of self-monitoring of oral anticoagulation, we did a meta-analysis
of individual patient data addressing several important gaps in the evidence, including an estimate of the eect on
time to death, rst major haemorrhage, and thromboembolism.
Methods We searched Ovid versions of Embase (19802009) and Medline (19662009), limiting searches to randomised
trials with a maximally sensitive strategy. We approached all authors of included trials and requested individual
patient data: primary outcomes were time to death, rst major haemorrhage, and rst thromboembolic event. We did
prespecied subgroup analyses according to age, type of control-group care (anticoagulation-clinic care vs primary
care), self-testing alone versus self-management, and sex. We analysed patients with mechanical heart valves or atrial
brillation separately. We used a random-eect model method to calculate pooled hazard ratios and did tests for
interaction and heterogeneity, and calculated a time-specic number needed to treat.
Findings Of 1357 abstracts, we included 11 trials with data for 6417 participants and 12 800 person-years of follow-up.
We reported a signicant reduction in thromboembolic events in the self-monitoring group (hazard ratio 051; 95% CI
031085) but not for major haemorrhagic events (088, 074106) or death (082, 062109). Participants younger
than 55 years showed a striking reduction in thrombotic events (hazard ratio 033, 95% CI 017066), as did
participants with mechanical heart valve (052, 035077). Analysis of major outcomes in the very elderly
(age 85 years, n=99) showed no signicant adverse eects of the intervention for all outcomes.
Interpretation Our analysis showed that self-monitoring and self-management of oral coagulation is a safe option for
suitable patients of all ages. Patients should also be oered the option to self-manage their disease with suitable
health-care support as back-up.
Funding UK National Institute for Health Research (NIHR) Technology Assessment Programme, UK NIHR National
School for Primary Care Research.

Introduction
Oral anticoagulation with vitamin K antagonists substantially reduces the incidence of thromboembolic
events.14 Although the number of patients receiving
oral anticoagulants has consistently increased, uptake
is limited by requirements to maintain the international
normalised ratio (INR) within a narrow target range,
which includes frequent testing and appropriate dose
adjustments. Benets shown in clinical trials might not
translate into routine practice: namely the risk of major
bleeding could be high in specic populations of
patients, especially in the elderly.2
Introduction of reliable and analytically accurate
point-of-care devices allows self-testing by the patient in
the home setting.1,2 Patients can have their test result
managed by their health-care provider (self-testing) or
they can interpret their INR result, and adjust their own
dose of anticoagulant accordingly (self-management).
Previous systematic reviews46 showed that selfmonitoring is a safe intervention, which gives rise to
signicant reduction in thromboembolic events, while
reducing the risk of death. Additionally, patients spend
more time in the therapeutic range of INR than they
322

would without self-monitoring. However, previous


conclusions were limited by methodological problems
and inadequate reporting of important outcome data
over time.46 Also important subgroup analyses,
stratied by age and indication for anticoagulation
therapy, have not been possible.
Uptake of self-testing and self-management has
remained inconsistent in and between countries,
despite good evidence of their eectiveness and
guidelines encouraging patients to discuss this option
with clinical sta.1,7,8 To clarify further the value of selfmonitoring of oral anticoagulation we did a metaanalysis of individual patient data, which updated our
previous systematic reviews and enabled more detailed
analysis than previously. Specically, we aimed to
address several important gaps in the evidence,
including obtaining an estimate of the eect of selfmonitoring on time to death, rst major haemorrhage,
and rst thromboembolic event. We also aimed to
investigate eects in important subgroups such as the
elderly and those with specic disease indications for
anticoagulation such as atrial brillation or mechanical
heart valve.
www.thelancet.com Vol 379 January 28, 2012

Articles

Methods

Statistical analysis

Search strategy and study selection

Primary outcomes were time to death, rst major


haemorrhage, and rst thromboembolic events. Major
haemorrhages included: 1) bleeding that was fatal, 2)
symptomatic bleeding in a critical area or organ such as
intracranial, intraspinal, intraocular, retroperitoneal,
intra-articular or pericardial, or intramuscular with
compartment syndrome; and 3) bleeding causing a fall
in haemoglobin concentrations of 20 g/L (124 mmol/L)
or more, or leading to transfusion of two units of packed
red-blood cells. Thromboembolic events were stroke,
arterial embolism, symptomatic deep-vein thrombosis,
or pulmonary embolism.11
One secondary outcome was time in therapeutic
range. For individual data, time-to-event outcomes were
summarised as log (hazard ratio), and time in range as
mean (SD). We used the survival-curve and hazard-ratio
programme SCHARP (version 4) for meta-analysis of
individual patient data. SCHARP is an SAS application
for meta-analysis of individual patient data with a pointand-click interface that produces publication-quality
graphs and appropriate summary statistics for time-toevent data.12
We used SPSS (version 17) for analysis of baseline
characteristics. We did prespecied subgroup analyses for
the primary outcomes with the following age bands (<55,
5564, 6574, and 75 years), the type of control-group
care, type of self-monitoring, and sex. We also analysed
patients who had mechanical heart valves, atrial brillation,
and other indications separately, stratifying them by age
(<65 years and 65 years).
Because the components of the interventions dier
somewhat (eg, in terms of their training and education)
and in assuming a dierent underlying eect for each
trial intervention, a random eects model was used to
calculate pooled hazard ratios. Random eects generally
lead to wider CIs than the xed eects; however, when
no heterogeneity is present the results of the xed and
random eects are equivalent. Time-to-event outcomes
were analysed with hazard ratios, which take into
account the number and timing of events, and the time
until last follow-up for each patient not experiencing an
event. We used a two-step process for meta-analysis: a
hazard ratio was estimated for each trial and then hazard
ratios were pooled in a meta-analysis. The log-rank
observed-minus-expected statistic and its variance were
calculated for each trial.13 We examined heterogeneity
with the I statistics.14,15 We also did tests for interaction
between subgroups of patients, partitioning the total
heterogeneity across all trials into within-group and
between-group heterogeneity (the test for interaction).
We calculated a time-specic number needed to treat at
various timepoints with the method outlined by Altman
and Andersen,16

The protocol methods have been previously published.9 We used the same search strategy as for
previous reviews.46 We searched Ovid versions of Embase
(19802009) and Medline (19662009), limiting searches
to randomised trials with a maximally sensitive strategy.10
A list of search terms is shown in webappendix pp 13.
We modied these searches for the Cochrane Central
Register of Controlled Trials, the Cochrane Library,
issue 2, 2009, and Cinahl (19822009). We also searched
for trials that are still underway or unpublished (eg, UK
National Research Register and Trials Central), and handsearched reference lists of retrieved papers.

Trial eligibility and quality assessment


We included randomised trials that compared the
eects of self-monitoring (self-testing) or self-management (self-testing and self-dosage) of anticoagulation
with control and dosage by personal physician,
anticoagulation management clinics, or managed
services, or reported the clinical outcomes of thromboembolic events and major bleeding episodes. We
included studies of adults on anticoagulant therapy
irrespective of the indication for treatment, with no
language restrictions.
As in our previous systematic review,5 we assessed the
quality of studies by the presence of randomisation,
allocation concealment, masked outcome assessments,
intention-to-treat analysis, and attrition rates. Two
reviewers (CH and AW) independently assessed the
articles for inclusion, and disagreements were resolved
by discussion.

Data extraction
We approached all authors whose trials met the inclusion
criteria and requested the following data for individual
patients: date of randomisation, age, indication for
treatment, type of care, demographic and psychosocial
characteristics at randomisation including quality-of-life
measures, treatment allocation, time to death, time to
rst major haemorrhage, time to rst thromboembolic
event, and INR measurements.

Data validation
We kept original data on a secure server with a backup copy according to a prespecied data-security-agreement policy. Two researchers (CB and AF) cross-checked
trial details, summary measures, and major outcomes
were cross-checked with prespecied outcome denitions
against published articles. Any inconsistencies were
discussed with the original trialist and corrections were
made when appropriate. Requirements for authorship
were those of the International Committee of Medical
Journal Editors and a representative of each trial was
invited to an investigators meeting before publication to
discuss analysis and results.
www.thelancet.com Vol 379 January 28, 2012

Number needed to treat =

See Online for webappendix

1
(Sc[t]hSc[t])
323

Articles

where at a specied timepoint (t), the survival probability


in the control group is Sc(t), then the survival probability
in the active group is Sc(t)h, where h is the hazard ratio
comparing the treatment groups. The number needed

1357 potential records identified

1313 excluded

45 full text articles reviewed


5 systematic reviews

23 excluded
8 duplicates
15 not randomised trials

to treat represents the number of patients treated in the


intervention group for one less primary outcome event
over the time stipulated. For subgroups, we calculated
the average eect over 5 years based on the control
event rate. All analyses were on an intention to treat
basis. We deviated from our original protocol in that we
did not present data on psychological factors, which we
hope to report elsewhere.
Sucient data from eight trials1724 were available for us
to calculate the mean time in therapeutic range at
timepoints of 7 days, 30 days, 90 days, 6 months, and
1 year by the method of linear interpolation set out by
Rosendaal and colleagues.25 We assessed publication bias
by constructing a funnel plot of precision (SE of the log
hazard ratio) against hazard ratio for the endpoints of
major haemorrhage and thromboembolic events,6 and
did Beggs and Eggers tests.15

Role of the funding source


21 randomised trials identified
(7598 participants)

10 excluded
10 did not have data available
(1181 participants)

Results

11 trials included
(6417 participants)

Of 1357 abstracts, we identied 21 trials (20 published,


one unpublished) that met the eligibility criteria
(gure 1). We were unable to obtain adequate data from

Figure 1: Flow chart of studies

Country

Dates of
Year of
recruitment publication
of primary
results

The sponsor of the study had no role in study design,


data collection, data analysis, data interpretation, or
writing of the report. The corresponding author had
full access to all the data in the study and had nal
responsibility for the decision to submit for
publication.

Age range,
Study
duration years (mean)
(months)

Total
number of
patients

Female

Atrial
brillation

Mechanical
valve

Other

Type of control
Self
manage- group care
ment

Beyth et al36*

USA

199295

2000

6594 (747)

325 (5%)

184 (57%)

54 (17%)

36 (11%)

235 (72%)

No

Primary Care

Cromheecke
et al17

Holland

1998

2000

2271 (423)

49 (1%)

21 (43%)

11 (22%)

23 (47%)

15 (31%)

Yes

Anticoagulation
clinic

Koertke et al18

Germany 199497

2001

24

1777 (597)

930 (14%)

293 (32%)

Sunderji et al37 Canada

19982002

2004

20

2085 (600)

139 (2%)

41 (29%)

47 (34%)

82 (59%)

MenndezJndula et al19

Spain

200102

2005

12

1990 (635)

737 (11%)

347 (47%)

296 (40%)

285 (39%)

Vller et al38

Germany 19992001

2005

19

3685 (644)

202 (3%)

53 (26%)

202 (100%)

Yes

Primary care

Fitzmaurice
et al20

UK

200102

2005

12

1887 (651)

617 (10%)

217 (35%)

343 (56%)

97 (16%)

177 (29%)

Yes

Both

Christensen
et al21

Denmark 200203

2006

2178 (507)

100 (2%)

33 (33%)

24 (24%)

35 (35%)

41 (41%)

Yes

Both

Siebenhofer
et al23

Austria

200205

2007

36

6089 (688)

195 (3%)

81 (42%)

89 (46%)

32 (16%)

74 (38%)

Yes

Both

Matchar et al22 USA

200306

2010

36

2390 (670)

2922 (46%)

51 (2%)

2236 (77%)||

684 (23%)

2 (<1%)

No

Anticoagulation
clinic

Kaatz et al24

USA

199899

2001

12

3087 (641)

201 (3%)

84 (42%)

86 (43%)

39 (19%)

76 (38%)

No

Anticoagulation
clinic

Totals

19922006

200010

1794 (650)

1405 (22%)

3388 (53%)

6417

930 (100%)

10 (7%)

Yes

Primary care

Yes

Primary care

154 (21%) Yes

2243 (350%) 784 (12%)

Anticoagulation
clinic

Data are range (mean) or number (%). *Coumatrack monitor.Coaguchek system. Pro time microcoagulation system.Two patients were unclassied for indication. Study stopped early. ||2236 with no
mechanical heart valve, 2422 had atrial brillation, mechanical heart valve, or both.

Table 1: Characteristics of studies

324

www.thelancet.com Vol 379 January 28, 2012

Articles

www.thelancet.com Vol 379 January 28, 2012

Probability of an event (%)

A
50

Control
Self-monitoring

40

Hazard ratio 051 (95% CI 031085; p=001)

30
20
10
0
0

Number at risk
Control 3151 (84)
Self-monitoring 3266 (43)

Time (years)
2137 (39)
2232 (38)

1214 (17)
1299 (19)

771 (7)
846 (12)

275 (4)
317 (2)

0
0

B
Probability of an event (%)

50
40

Hazard ratio 088 (95% CI 074106; p=018)

30
20
10
0
0

Number at risk
Control 3151 (131)
Self-monitoring 3266 (106)

Time (years)
2109 (56)
2189 (64)

1189 (30)
1273 (31)

742 (24)
820 (21)

251 (2)
296 (1)

0
0

C
50
Probability of an event (%)

ten trials.2635 These trials were small, ranging from


50320 participants (total 1181 participants). Of
21 original trials, including 7598 participants, we present
results for 6417 (84%) participants.
Table 1 shows 11 included trials: three in the USA,22,24,36
two in Germany,18,38 and one each from Austria and
Germany,23 Canada,37 Denmark,21 Netherlands,17 Spain19
and the UK.20 Participant recruitment into the trials
occurred from 1992 to 2006, and trials were published
between 2000 and 2010. The Coaguchek (Roche
Diagnostics, Basel, Switzerland), Pro time microcoagulation (ITC Nexus Dx, Edison, NJ USA), and the
Coumatrak monitor (Du Pont Pharmaceuticals,
Wilmington, DE, USA) systems were used (table 1). For
all trials, we veried clear methods for randomisation,
allocation concealment, and intention-to-treat analyses.
For publication bias, we saw no funnel-plot asymmetry
and no bias with Beggs test: p=035 for thromboembolic
events and p=100 for major haemorrhage;
corresponding results of Eggers test were p=005 for
thromboembolic events and p=092 for major
haemorrhage. 3266 (51%) participants were randomly
allocated to self-monitoring and 3151 to conventional
care. Participants in the intervention group were on
average 17 years (642 [SD 117] years vs 659 [SD 105]
years; p<00001) younger than those in the control
groups. A wide range of ages was included: from 17 to
94 years of age, with 99 participants aged 85 years or
older. 12 800 person years of follow-up were obtained
(mean 199 years [SD 122]), with a maximum follow-up
of 1888 days (517 years).
Over a third of participants had a mechanical heart
valve insertion; one trial18 included only participants
with this indication. Over half of participants had atrial
brillation; one trial38 included only participants with
this indication. For other disorders, over 10% of
participants from nine trials17,1924,36,37 were included. In
eight trials,1721,23,37,38 just under half (46%) of those in the
intervention group used self-management and in three
trials,22,24,36 just over half (54%) used self-testing only with
dose adjustments undertaken by their regular clinician
(table 1).
In four trials18,3638 including a quarter of participants,
primary care was used as the control and in another four
trials17,19,22,24 including more than half (61%) of participants,
specialist anticoagulation clinics were used (table 1).
In the three remaining trials,20,21,23 the control included
either primary care or specialist clinics (table 1).
A signicant reduction in thromboembolic events was
seen in the self-monitoring group (hazard ratio 051,
95% CI 031085; p=0010; I=526%; gure 2).
At 1 year, the number needed to treat to prevent one
thromboembolic event was 78 (95% CI 55253), and by
5 years it was 27 (1987; table 2). Individual study hazard
ratios are shown in webappendix p 4. No signicant
reduction in major haemorrhagic events (hazard ratio
088, 95% CI, 074106; p=018, I=0) or in deaths

40

Hazard ratio 082 (95% CI 062109; p=018)

30
20
10
0
0

Number at risk
Control 3151 (130)
Self-monitoring 3266 (96)

Time (years)
2197 (67)
2271 (76)

1264 (48)
1346 (46)

816 (19)
884 (24)

296 (5)
334 (4)

0
0

Figure 2: Hazard ratios for major outcomes


Hazard ratios for thrombotic events (152 events in the control group, 114 in the self-monitoring group; A),
haemorrhagic events (244 in the control group, 230 in the self-monitoring group; B), and death (274 in the
control group, 247 in the self-monitoring group; C).

(082, 062109; p=018; I= 370) were apparent with


self-monitoring (gure 2).
In prespecied subgroups the rate of thromboembolic
events in men was signicantly reduced in the selfmonitoring group (gure 3; p=0010; I=613) whereas
in women it was not (gure 3; p=046; I=266;).
However, the ratio of male to female participants was
four to one (5012 men vs 1405 women) and the interaction
test showed that these two subgroups did not dier
signicantly ( 001; p=094). Participants younger than
55 years of age who self-monitored had a striking
reduction in thromboembolic events (gure 3; p=0002;
I=0), whereas in other age groups non-signicant eects
were shown. In participants younger than 55 years, this
325

Articles

Number needed
to treat (95% CI)

Number of patients
still at risk

All participants (N=6417)

Number of patients
still at risk

(Continued from previous column)

Thrombosis

Death

Year 1

78 (55 to 253)

4369

Year 1

101 (50 to )

Year 2

47 (33 to 154)

2513

Year 2

56 (28 to )

1921

Year 3

36 (26 to 119)

1617

Year 3

37 (18 to )

1216

Year 4

32 (23 to 104)

592

Year 4

30 (15 to )

413

Year 5

27 (19 to 87)

Year 5

26 (13 to )

Year 1

205 (94 to )

4598

Year 2

123 (59 to )

2462

Year 3

96 (44 to )

1564

Year 4

74 (34 to )

547

Year 5

70 (32 to )

Major Haemorrhage

Death
Year 1

137 (65 to )

4468

Year 2

82 (39 to )

2610

Year 3

55 (26 to )

1700

Year 4

47 (22 to )

630

Year 5

42 (20 to )

Mechanical valve only (n=2243)


Thrombosis
Year 1

55 (41 to 116)

1721

Year 2

37 ( 28 to 78)

589

Year 3

33 ( 25 to 70 )

419

Year 4

31 ( 23 to 65)

186

Year 5

24 (18 to 50)

Year 1

127 (66 to )

1699

Year 2

65 (34 to )

565

Year 3

43 (22 to )

386

Year 4

31 (16 to )

160

Year 5

29 (15 to )

Year 1

156 (67 to )

1763

Year 2

65 (34 to )

618

Year 3

43 (23 to )

442

Year 4

31 (16 to )

202

Year 5

29 (15 to )

Major haemorrhage

Death

Atrial brillation only (n=3388)


Thrombosis
Year 1

185 (85 to )

2386

Year 2

91 (42 to )

1860

Year 3

65 (30 to )

1163

Year 4

54 ( 25 to )

393

Year 5

49 (23 to )

Major haemorrhage
Year 1

675 (79 to )*

2344

Year 2

453 (53 to )*

1824

Year 3

342 (40 to )*

1134

Year 4

268 (32 to )*

373

Year 5

255 (30 to )*

(Continues in next column)

326

Number needed to
treat (95% CI)

2428

The number needed to treat is estimated as:


Number needed to treat =

1
(Sc[t]hSc[t])

*Number needed to harm.

Table 2: Number needed to treat at various timepoints for all


self-monitoring participants with a mechanical valve and atrial
brillation compared with standard care

result corresponded to a number needed to treat


of 21 (95% CI 1742) to prevent one thromboembolic
event at 1 year. Non-signicant improvement in major
outcomes was seen in the self-monitoring group with
younger age ( 775; p=0052).
In terms of indication, participants with a mechanical
heart valve who self-monitored had signicant
reductions in thromboembolic events (gure 3;
p=0001; I=0). At 1 year the number needed to treat to
prevent one event was 55 (95% CI 41116) and by
5 years it was 24 (1850). Eects for both atrial
brillation (gure 3; p=035; I=409) and other
indications (gure 3; p=012; I=0) were not signicant.
An interaction test ( 688, p=0032) between
indications was signicant. Participants who selfmanaged oral anticoagulation also had signicantly
fewer thromboembolic events (gure 4; p<0001; I=0),
whereas participants self-testing alone did not (gure 4;
p=051; I=503). The interaction test between selftesting and self-management for this dierence was
signicant ( 981, p=0002). For participants selfmanaging, the number needed to treat to prevent
one thromboembolic event was 39 (95% CI 3165). For
major haemorrhage and death, we detected no
signicant eects or interactions by age, sex, indication,
or type of monitoring (gure 3).
Analysis of major outcomes in the very elderly
(85 years, n=99) showed no signicant adverse eects
of self-monitoring for all outcomes, and a reduction in
mortality was seen (hazard ratio 044, 95% CI 020098;
p=0044; I=0); however, the number of participants in
this analysis was small (n=75).
We postulated that type of control care in our prespecied subgroups might aect the overall eectiveness
of self-monitoring. Yet, little dierence was seen in terms
of anticoagulation clinic care versus primary care for
thromboembolic events (gure 3; 218, p=034); major
www.thelancet.com Vol 379 January 28, 2012

Articles

A
Sex
Male
Female
Age (years)
<55
5564
6574
75
Indication
Mechanical valve
Atrial brillation
Other
Self-monitoring
Self-testing only
Self-management
Type of control group care
Anticoagulation clinic
Primary care
Total

Self-monitoring

Control

OE

Variance

Hazard ratio (random)

HR (95% CI)

Interaction test

96/2453
18/539

124/2323
28/555

1757
328

5473
1113

040 (020080)
076 (036158)

p=094

11/496
33/895
44/1038
26/488

25/338
34/798
66/1117
27/538

917
217
988
102

837
1659
2706
1303

033 (017066)
080 (045142)
069 (048101)
096 (041226)

p=0052

36/1132
70/1629
8/261

63/1040
69/1623
20/237

1622
006
387

2458
3458
630

052 (035077)
067 (028157)
054 (025118)

p=0032

89/1566
25/1487

89/1557
63/1382

283
1852

4444
2157

074 (030182)
042 (028065)

p=0002

92/1959
16/995

112/1950
27/893

1146
634

5083
1044

044 (017114)
055 (028108)

p=034

114/3053

152/2939

201/2542
29/618

201/2403
43/639

986
515

9983
1756

091 (074110)
075 (047119)

p=045

25/500
66/812
78/1110
61/583

24/340
44/699
101/1185
75/640

271
667
1166
448

1179
2716
4386
3349

079 (045141)
128 (088186)
077 (057103)
087 (062123)

p=020

78/1111
126/1676
26/260

91/1015
119/1677
34/281

1011
235
690

4166
6060
1387

078 (058106)
104 (081134)
061 (036103)

p=013

190/1729
40/1487

200/1719
44/1382

1141
327

9681
2089

084 (064112)
086 (056131)

p=087

182/1959
42/1158

187/1950
46/1055

827
402

9158
2189

091 (074112)
083 (050136)

p=060

230/3216

244/3101

204/2440
43/618

219/2297
55/639

1706
281

10504
2411

085 (070103)
089 (044181)

p=084

14/495
46/874
98/1034
89/583

15/353
46/781
129/1112
84/640

356
284
1544
474

661
2282
5565
4291

058 (022157)
072 (036141)
076 (058099)
112 (083151)

p=019

64/1101
138/1555
45/334

75/1002
156/1549
43/349

966
1432
371

3454
7267
2153

080 (042152)
072 (043120)
119 (078181)

p=023

195/1729
52/1342

201/1719
73/1237

894
1204

9873
3105

091 (075111)
075 (042133)

p=015

159/1934
72/988

173/1926
90/884

1062
1302

8283
4020

082 (052128)
072 (044118)

p=020

247/3071

274/2956

051 (031085)

B
Sex
Male
Female
Age (years)
<55
5564
6574
75
Indication
Mechanical valve
Atrial brillation
Other
Self-monitoring
Self-testing only
Self-management
Type of control group care
Anticoagulation clinic
Primary care
Total

088 (074106)

C
Sex
Male
Female
Age (years)
<55
5564
6574
75
Indication
Mechanical valve
Atrial brillation
Other
Self-monitoring
Self-testing only
Self-management
Type of control group care
Anticoagulation clinic
Primary care
Total

082 (062109)
01

02

05

Favours self-monitoring

10

Favours control

Figure 3: Major outcomes by sex, age, indication, type of monitoring and control group care
Thrombosis (A), major haemorrhage (B), and death (C). OE=observed minus expected.

www.thelancet.com Vol 379 January 28, 2012

327

Articles

A
Self-monitoring
Self-management
Christensen21
Cromheecke17
Fitzmaurice20
Koertke18
Menndez-Jndula19
Siebenhofer23
Sunderji37
Vller38
Subtotal
Self-testing only
Kaatz24
Matchar22
Beyth36
Subtotal
Total

Control

0/50
0/25
4/337
12/488
3/368
6/99
0/69
0/101
25/1487

0/50
2/24
3/280
21/442
21/369
13/96
2/70
1/101
63/1382

2/101
87/1465
0/163
89/1566
114/3053

6/100
83/1457
0/162
89/1557
152/2939

Hazard ratio (random)

Hazard ratio (95% CI)

013 (001203)
163 (033812)
050 (025100)
027 (012062)
047 (019116)
014 (001218)
017 (000875)
042 (028065)

035 (009141)
098 (073133)
074 (030182)
051 (031085)
01

02

05

10

B
Self-management
Christensen21
Cromheecke17
Fitzmaurice20
Koertke18
Menndez-Jndula19
Siebenhofer23
Sunderji37
Vller38
Subtotal
Self-testing only
Kaatz24
Matchar22
Beyth36
Subtotal
Total

0/50
0/25
5/337
25/488
3/368
6/99
0/69
1/101
40/1487

0/50
1/24
4/280
20/442
7/269
11/96
1/70
0/101
44/1382

013 (000655)
103 (028384)
110 (061198)
054 (016189)
057 (022147)
014 (000692)
701 (01435401)
086(056131)

32/101
147/1465
11/163
190/1729
230/3216

36/100
143/1457
21/162
200/1719
244/3101

081 (050132)
096 (076121)
052(026105)
084 (064112)
088 (074106)
01

02

05

10

C
Self-management
Christensen21
Cromheecke17
Fitzmaurice20
Koertke18
Menndez-Jndula19
Siebenhofer23
Sunderji37
Vller38
Subtotal
Self-testing only
Kaatz24
Matchar22
Beyth36
Subtotal
Total

1/50
0/25
13/337
18/488
5/368
15/99
0/69
0/101
52/1342

0/50
0/24
11/280
36/442
15/369
11/96
0/70
0/101
73/1237

707 (01435658)

2/101
152/1465
41/163
195/1729
247/3071

1/100
157/1457
43/162
201/1719
274/2956

193 (0201860)
091 (073114)
090 (059139)
091 (075111)
082 (062109)

098 (044219)
045 (026077)
046 (019112)
140 (065302)

075 (042133)

01

02

05

Favours self-monitoring

10

Favours control

Figure 4: Comparison by type of monitoring (self-monitoring compared with self-testing only)


Thrombosis (A), major haemorrhage (B), and death (C).

328

www.thelancet.com Vol 379 January 28, 2012

Articles

A
Thrombosis
Sex
Male
Female
Age (years)
<65
65
Total
Major haemorrhage
Male
Female
Age (years)
<65
65
Total
Death
Male
Female
Age
<65
65
Total

Self-monitoring

Variance

Hazard ratio (95% CI) Interaction test

Control

OE

28/853
8/211

53/748
10/236

1507
002

1993
440

042 (024072)
100 (039253)

p=015

22/745
14/376
36/1132

32/552
31/473
63/1040

855
694

1300
1084

052 (030089)
053 (029096)
052 (035077)

p=097

65/860
13/226

77/743
14/249

1162
101

3486
642

072 (051100)
117 (054254)

p=025

44/665
34/400
78/1111

36/499
54/484
91/1015

036
878

1925
2131

098 (063153)
066 (043101)
078 (058106)

p=021

56/852
8/232

54/735
21/252

353
538

2711
717

088 (060128)
088 (019400)

p=014

24/701
40/395
64/1101

28/518
47/476
75/1002

657
054

1237
2102

062 (027143)
097 (064149)
080 (042152)

p=016

058 (014237)
071 (025204)

p=049

091 (048173)
089 (044178)
067 (028157)

p=071

105 (081136)
089 (033243)

p=076

135 (082223)
095 (071127)
104 (081134)

p=023

083 (065105)
094 (030297)

p=091

059 (013265)
076 (049119)
072 (043120)

p=053

01

Hazard ratio (random)

02

05

10

B
Thrombosis
Sex
Male
Female
Age (years)
<65
65
Total
Major haemorrhage
Male
Female
Age (years)
<65
65
Total
Death
Male
Female
Age (years)
<65
65
Total

65/1405
5/122

59/1382
10/129

152
118

3095
345

19/462
51/1060
70/1629

19/425
50/1093
69/1623

085
134

946
2490

119/1462
7/186

107/1438
12/200

257
043

5630
385

37/477
89/1087
126/1676

25/456
94/1120
119/1677

466
224

1537
4484

129/1341
9/186

143/1322
13/200

1263
128

6723
528

32/512
106/1030
138/1555

31/470
125/1067
156/1549

081
1304

1571
5662

01

02

05

Favours
self-monitoring

10

Favours
control

Figure 5: Major outcomes in mechanical valve and atrial brillation by age and sex
Patients with mechanical valve by sex and age (<65 years and 65 years; A) and patients with atrial brillation by sex and age (<65 years and 65 years; B).
OE=observed minus expected.

haemorrhage ( 101, p=060), and death outcomes


( 325, p=020).
A signicant reduction in thromboembolic events was
seen in men with a mechanical heart valve who were
self-monitoring (gure 5; p=0002, I=138), which was
not signicant in women (gure 5; p=099; I=58).
However, the number of women was small (n=447 for
thrombosis) and this interaction was not signicant
( 204, p=015). Men with a mechanical valve who were
www.thelancet.com Vol 379 January 28, 2012

self-monitoring also had a signicant reduction in major


haemorrhagic events (gure 5; p=0049; I=0), whereas
women did not (gure 5; p=069; I=0). However, the
interaction test was not signicant ( 131, p=025).
Participants younger than 65 years and those 65 years
or older with a mechanical heart valve who were
self-monitoring oral anticoagulation showed similar
signicant reductions with roughly a halving of
thrombotic events (gure 5). We saw no signicant
329

Articles

Country

7 days
Intervention

30 days
Control

Intervention

90 days

6 months

Control

Intervention

Control

578 (145)

478 (186)

1 year

Intervention

Control

Intervention

Control

Cromheecke
et al17

Holland

Krtke et al18

Germany

492% (429)

569% (413)

712% (287)

557% (355)

782% (220)

596% (325)

809% (200) 606% (312)

830% (186)

617 (308 )

MenndezJndula et al

Spain

667% (374)

625% (447)

662% (267)

682% (362)

662% (186)

685% (272)

680% (160) 690% (223)

679% (140)

688 (200)

Fitzmaurice
et al20

UK

666% (404)

532% (453)

697% (322)

639% (368)

707% (251)

636% (317)

710% (230)

634% (288)

717% (220)

638 (284)

Christensen21
et al

Denmark

802% (243)

617% (441)

783% (246)

712% (337)

775% (206)

677% (309)

755% (189)

673% (245)

755% (189)

674 (242)

Siebenhofer23
et al

Austria

436% (435 )

529% (440)

510% (366)

615% (358)

536% (262)

633% (297)

574% (245) 650% (265)

613% (199)

645 (211)

Matchar et al22

USA

633% (217)

523% (251)

635% (160)

530% (203)

641% (146)

551% (199)

651% (141)

577% (199)

672% (140)

610 (201)

Kaatz et al24

USA

597% (400)

611% (416)

566% (287)

669% (321)

629% (216)

700% (253)

647% (196)

716% (199)

659% (174)

708 (176)

Data are mean (SD).

Table 3: Percentage mean time in therapeutic range at 7 days to 1 year

Time in therapeutic range

7 days

Number of tests

Mean dierence between self-monitoring Heterogeneity


and control group (95% CI)

p value

1225% (899 to 1551)

Mean dierence between self-monitoring Heterogeneity


and control group (95% CI)

p value

<0001

025% (010 to 039)

77%

0001

72%

005

228% (159 to 297)

94%

<0001

513% (113 to 1140)

79%

011

1271% (933 to 1610)

96%

<0001

271% (610 to 1151)

94%

055

2422% (1840 to 3004)

93%

<0001

30 days

613% (009 to 1235)

6 months
1 year

Data % or % (95% CI).

Table 4: Mean dierence between self-monitoring and control group in time in therapeutic range and number of tests for participants with a
mechanical valve

Time in therapeutic range

7 days

Number of tests

Mean dierence between self-monitoring Heterogeneity


and control group (95% CI)

p value

1038% (856 to 1220)

Mean dierence between self-monitoring Heterogeneity


and control group (95% CI)

p value

0%

<0001

001% (025 to 028)

92%

091

30 days

316% (407 to 1039)

77%

039

178% (097 to 260)

97%

<0001

6 months

440% (086 to 967)

79%

010

1203% (746 to 1660)

99%

<0001

1 year

513% (097 to 928)

57%

002

2174% (1311 to 3037)

98%

<0001

Data % or % (95% CI).

Table 5: Mean dierence between self-monitoring and control group in time in therapeutic range and number of tests for participants with atrial brillation

eects or interaction in terms of major haemorrhage and


death for other subgroups of participants with a
mechanical valve. In participants with atrial brillation,
we saw no signicant eects across subgroups by sex or
age, and no signicant interactions (gure 5).
One study17 provided data only at 90 days for the mean
time in therapeutic range (table 3). The time in therapeutic
range improved and SDs decreased over time. By 1 year,
four trials18,2022 showed improvements in the intervention
group, whereas the three trials,19,23,24 which did not show
improvement, all had smaller SDs in the intervention
group. In the rst 7 days participants with atrial brillation
and a mechanical heart valve who self-monitored oral
330

coagulation spent signicantly more time in therapeutic


range than did those who did not self-monitor (table 4,
table 5), but over time the dierences between groups
reduced. Self-monitoring also led to an increase in the
number of tests undertaken. At 1 year, participants with a
mechanical valve or atrial brillation undertook more
tests per year than did those receiving usual care (table 4,
table 5). The substantial variation between studies was
illustrated by the high heterogeneity.

Discussion
Our study used individual patient data for assessment of
self-monitoring for oral anticoagulation. Overall, we
www.thelancet.com Vol 379 January 28, 2012

Articles

observed a signicant reduction in thromboembolic


events in the self-monitoring group. However, we did
not nd any signicant eects for major haemorrhage
or mortality.
Our ndings accord with those of previous systematic
reviews, in which patients who self-monitor or selfmanage could improve the quality of their oralanticoagulation therapy. However, despite the decrease
in the number of thromboembolic events without
concomitant increases in harms, we did not see the
reduction in mortality shown in previous systematic
reviews.46 The odds ratio in a meta-analysis by Bloomeld
and colleagues39 was similar to our result for reduction in
thromboembolic events (odds ratio 058, 95% CI
045075; p<0001). However, the result for death was
similar in eect size, but the observed result diered
signicantly (odds ratio 074, 95% CI 063087;
p<0001). This eect was highly heterogeneous (I=51%),
which was attributed to the largest study to date.22 Reasons
given for this high heterogeneity were that this large
study had substantially longer follow-up and higher
quality of control in the usual care group than did other
similar studies. The trend for reduction in mortality
favoured self-monitoring, yet our previous estimate for a
reliable and conclusive treatment eect would require
5150 participants in each study group.6 Potentially,
unavailable data from the ten studies that we were unable
to access, were sucient to remove the signicance of
this result.
Additionally, our previous estimate that self-monitoring
was feasible for only half of patients requiring
anticoagulant therapy might underestimate the true
numbers. In the largest trial,22 about 80% (2922 of 3643)
of trained patients were competent in the use of selfmonitoring equipment. Yet, even this estimate is
confounded by eligibility criteria: in several trials20,32 fewer
than 50% of the potentially eligible patients were
randomly assigned. Self-monitoring patients deemed not
competent had higher numbers of practice attempts and
higher cuvette wastage, and were less able to eciently
do a ngerstick procedure.40 Factors associated with
unsuccessful self-monitoring include refusal by patients,
exclusion by their family practitioner, failure to pass
training, old age, poor cognition, and poor manual
dexterity.6,20,40 One trial excluded people unable to attend
training,19 and in another trial20 of an unselected
population, young patients were more likely to
successfully self-monitor oral anticoagulatoin.
In Germany 20% (160 000) of patients on anticoagulation
undertook self-management, compared with only 1% of
those in USA who did self-testing at home. Reasons for
this dierence include reimbursement, motivation by
the patient, and willingness of the physician to support
self-monitoring.20 Limitations include the reluctance of
individuals to participate, but also the direct costs to
patients and the training required for eective
monitoring.
www.thelancet.com Vol 379 January 28, 2012

In patients younger than 55 years of age, two-thirds


reduction in thromboembolic events translated into
21 participants self-monitoring for 1 year to prevent one
thromboembolic event. For patients with a mechanical
heart valve, a 50% decrease in thromboembolic events
meant that the number needed to undertake selfmonitoring to prevent one event was 55 after 1 year and
24 over 5 years. By comparison, 63 patients are needed to
prevent one heart attack with daily statin therapy over
5 years.41,42
Patients who self-tested and adjusted their doses had
signicantly lower rates of thromboembolic events, which
suggests that patients should be given the opportunity,
and provided with training, to undertake self-management.
However, self-management does not mean that patients
are left to fend for themselves: for instance, in one trial
participants had 24 h back-up available,37 and good quality
control measures are needed. The type of control care did
not aect the overall eectiveness of self-monitoring. This
nding is often contradictory to the evidence, which shows
that patients from community practices have signicantly
worse anticoagulation control than do those from anticoagulation clinics. However, the same systematic review
highlighted that patients recruited to clinical trials tended
to spend more time in the therapeutic range than did
those in the community.43
For participants with atrial brillation we reported no
signicant eects across subgroups by sex or age, and no
signicant interactions. Participants with atrial brillation
were older than those with a mechanical heart valve, and
in this age group, rates of events tended to be low. In a
previous trial of 973 elderly patients in the community on
anticoagulation, thromboembolic events were 14%
a year.44
Mean time in therapeutic range tended to be better in
the self-monitoring groups. Importantly, even when the
time in therapeutic range showed worse control, the SDs
were less, which suggests lower variation and therefore
more stable control of oral anticoagulation than in the
control care group.45 Full analysis of this issue is beyond
the scope of this report, but is an important issue in
establishing optimum anticoagulation control.
We also reported a reduction in mortality in very elderly
patients who self-monitored oral anticoagulation. This
result, although potentially misleading owing to the
small numbers and number of analyses, warrants further
investigation. The evidence already supports the use of
anticoagulation for elderly patients unless contraindications apply or patients decide the benets are not
worth the inconvenience of such treatment.44 Our review
was restricted to adults, although increasing numbers of
children receive warfarin. But self-monitoring could be a
safe and eective management strategy for children and
clinical studies are recommended.46
Some limitations are worth noting. First, we could have
missed a study, especially because of non-publication.
The results dier for publication bias because of variation
331

Articles

in the methods for calculating Beggs and Eggers tests.


Yet, the results of both suggest a weak eect of publication
bias due to eects of small studies. Second, we were
unable to obtain data from ten studies, although this was
a small proportion of the overall dataset, which reduced
the overall sample size. However, we were able to receive
data from the largest trial to date, which was recently
published.22 Third, some heterogeneity in outcomes was
obsereved. Dierences occurred in the populations (ie,
the monitor and the intervention populations), which all
add to the inherent variability. Fourth, only a small
number of participants aged over 85 years were included,
and further research in this age group is needed. Finally,
we do not know why fewer women than men were
included and whether this is because women are reluctant
to participate in self-monitoring or the overall recruitment
strategies target men. One reason could be that in the
largest study,22 which comprised nearly half of the data,
only 17% of the included participants were women.
Furthermore, the study was done in a Veterans Aairs
population, which mainly includes men.
Adoption of self-monitoring will depend on ndings
from economic analyses, which in the past have produced
conicting results. In the UK, a review concluded in
general, patient self-management is unlikely to be more
cost-eective than the current specialised anticoagulation
clinics,47 whereas a Canadian study suggested: selfmanagement is a cost-eective strategy for patients
receiving long-term oral anticoagulation therapy for atrial
brillation or for a mechanical heart valve.48
We believe the results of our review will lead to a
systematic change in practice, in terms of the signicant
reduction in thromboembolic events in patients with a
mechanical heart valve requiring long-term anticoagulation. Such patients should be oered the option to
self-manage their disease with suitable health-care
support as back-up. Additionally, several reviews and our
study show that self-monitoring and self-management is
a safe option for suitable patients.5,6,49
Contributors
Members named in the writing committee contributed to the data
collection or the systematic review and data analysis, and to the
preparation of the published Article.
Collaborators
Writing committee Carl Heneghan, Alison Ward, Rafael Perera,
Clare Bankhead, Alice Fuller, Richard Stevens, Kairen Bradford,
Sally Tyndel (Oxford University, Oxford, UK), Pablo Alonso-Coello
(Institute of Biomedical Research [IIB Sant Pau], Barcelona, Spain),
Jack Ansell (Lenox Hill Hospital, New York, USA), Rebecca Beyth
(University of Florida, FL, USA), Artur Bernardo (Klinik Gais AG, Gais,
Switzerland), Thomas Decker Christensen (Aarhus University Hospital,
Skejby, Aarhus, Denmark), M E Cromheecke (Academic Medical Centre,
University of Amsterdam, Netherlands), Robert G Edson (VA Cooperative
Studies Program Coordinating Center, Palo Alto, CA, USA),
David Fitzmaurice (University of Birmingham, Birmingham, UK),
Alain P A Gadisseur (Antwerp University Hospital, Antwerp, Belgium),
Josep M Garcia-Alamino (Oxford University, Oxford, UK), Chris Gardiner
(University of Oxford, Oxford, UK), J Michael Hasenkam (Aarhus
University Hospital, Skejby, Aarhus, Denmark), Alan Jacobson (Loma
Linda University, CA, USA), Scott Kaatz (Henry Ford Hospital, Detroit,

332

USA), Farhad Kamali (Newcastle University, Newcastle upon Tyne, UK),


Tayyaba Irfan Khan (Newcastle University, Newcastle upon Tyne, UK),
Eve Knight (Anticoagulation Europe, Kent, UK), Heinrich Krtke
(Institute of Applied Telemedicine, Bad Oeynhausen, Germany),
Marcel Levi (Academic Medical Centre, University of Amsterdam,
The Netherlands), David Matchar (Duke-National University of Singapore
Graduate Medical School, Singapore), Brbara Menndez-Jndula
(Hospital de la Santa Creu i Sant Pau, Barcelona, Spain), Ivo Rakovac
(Institute for Biomedicine and Health Sciences, Graz, Austria),
Christian Schaefer (International Self-Monitoring Association of Oral
Anticoagulation Patients [ISMAAP], Geneva, Switzerland),
Andrea Siebenhofer (Goethe University, Frankfurt, Germany),
Juan Carlos Souto (Hospital de la Santa Creu i Sant Pau, Barcelona,
Spain), Rubina Sunderji (University of British Columbia and Vancouver
General Hospital, Vancouver, Canada), Kenneth Gin (University of
British Columbia and Vancouver General Hospital, Vancouver, Canada),
Karen Shalansky (University of British Columbia and Vancouver General
Hospital, Vancouver, Canada), Heinz Vller (Klinik am See, Rdersdorf/
Berlin, Germany), Otto Wagner (Ruhr Universitt Bochum, Bad
Oeynhausen, Germany), Armin Zittermann (Ruhr Universitt Bochum,
Bad Oeynhausen, Germany). Analysis committee Rafael Perera,
Clare Bankhead, Richard Stevens, Carl Heneghan, Alison Ward,
Alice Fuller. Data management committee Alison Ward, Alice Fuller,
Rafael Perera, Clare Bankhead, Richard Stevens. Original trialists
Rebecca Beyth, Thomas Decker Christensen, M E Cromheecke,
Robert G Edson, David Fitzmaurice, Alain P A Gadisseur, Chris Gardiner,
J Michael Hasenkam, Alan Jacobson, Scott Kaatz, Farhad Kamali,
Tayyaba Irfan Khan, Heinrich Krtke, Marcel Levi, David Matchar,
Brbara Menndez-Jndula, Ivo Rakovac, Andrea Siebenhofer,
Juan Carlos Souto, Rubina Sunderji, Kenneth Gin, Karen Shalansky,
Heinz Vller, Otto Wagner, Armin Zittermann.
Conicts of interest
Department of Primary Health Care, Oxford University (UK) received
funding from the National Institute for Health Research Technology
Assessment Programme (NIHR HTA; UK) and NIHR National School
for Primary Care Research (UK); Iberoamerican Cochrane Centre
received funding from Instituto de Salud Carlos III; Jack Ansell received
funding from Alere ITC; Rebecca Beyth received funding from the
5th International Patient/Physician Conference on Oral Anticoagulant
Therapy and the NIHR, University of Florida, Department of Veterans
Aairs, and US National Institutes of Health; Robert Edson received
funding from VA Cooperative Studies Program; Alain P A Gadisseur
received funding from Roche Diagnostics, Bayer Healthcare and
Boehringer Ingelheim; Department of Haemotology, Antwerp University
Hospital (Germany) received funding from Bayer Healthcare; University
College London Hospital (UK) received funding from Roche Diagnostics;
J Michael Hasenkam received funding from Roche Diagnostics, Johnson
& Johnson, Edwards Lifesciences, and Nycomed; Department of
Cardiothoracic Surgery, Aarhus University Hospital (Denmark) received
funding from National Board of Health; Alan Jacobson received funding
from Anticoagulation Forum, Loma Linda Veterans Association for
Research, Biosite Corporation, Boehringer Ingelheim, Daiichi Sankyo,
Farallon Medical, Hemosene, Inverness Medical, Pzer Medical, Quality
Assured Services, Roche Diagnostics, Sano-Aventis, Tapestry Medical,
GlaxoSmithKline, and VA Cooperative Studies Program and was the
co-chair of the THINRS study; Department of Internal Medicine, Loma
Linda University (USA) received funding from Loma Linda Veterans
Association for Research, Biosite Corporation, Boehringer Ingelheim,
Farallon Medical, Hemosene, Inverness Medical, and Sano-Aventis;
Scott Kaatz received funding from Biosite and Vox media; Department of
Internal Medicine, Henry Ford Hospital (USA) received funding from
Roche Diagnostics; Institute of Cellular Medicine, University of
Newcastle (UK) received funding from Heart Research UK; Wolfson Unit
of Clinical Pharmacology, School of Clinical and Laboratory Sciences,
University of Newcastle (UK) received funding from the BUPA
Foundation and Coaguchek; Anticoagulation Europe received funding
from Roche Diagnostics; David Matchar received funding from
Boerhinger Ingelheim, Roche, and VA Cooperative Studies Program and
was the Co-Chair of the THINRS study; Juan Carlos Souto is the
Scientic Director of Monitor Medical.

www.thelancet.com Vol 379 January 28, 2012

Articles

References
1
Fitzmaurice DA, Gardiner C, Kitchen S, Mackie I, Murray ET,
Machin SJ. An evidence-based review and guidelines for patient
self-testing and management of oral anticoagulation. Br J Haematol
2005; 131: 156165.
2
Szucs TD, Bramkamp M. Pharmacoeconomics of anticoagulation
therapy for stroke prevention in atrial brillation: a review.
J Thromb Haemost 2006; 4: 118085.
3
Poller L, Keown M, Chauhan N, et al. European concerted action on
anticoagulation. Quality assessment of the CoaguChek Mini and
TAS PT-NC point-of-care whole-blood prothrombin time monitors.
Clin Chem 2004; 50: 53744.
4
Christensen TD, Johnsen SP, Hjortdal VE, Hasenkam JM.
Self-management of oral anticoagulant therapy: a systematic
review and meta-analysis. Int J Cardiol 2007; 118: 5461.
5
Heneghan C, Alonso-Coello P, Garcia-Alamino JM, Perera R,
Meats E, Glasziou P. Self-monitoring of oral anticoagulation:
a systematic review and meta-analysis. Lancet 2006; 367: 40411.
6
Garcia-Alamino JM, Ward AM, Alonso-Coello P, et al.
Self-monitoring and self-management of oral anticoagulation.
Cochrane Database Syst Rev 2010; 4: CD003839.
7
Ansell J, Jacobson A, Levy J, Voller H, Hasenkam JM.
Guidelines for implementation of patient self-testing and patient
self-management of oral anticoagulation. International consensus
guidelines prepared by International Self-Monitoring Association
for Oral Anticoagulation. Int J Cardiol 2005; 99: 3745.
8
Lip GY, Rudolf M, Kakar P. Management of atrial brillation:
the NICE guidelines. Int J Clin Pract 2007; 61: 911.
9
Perera R, Heneghan C, Fitzmaurice D. Individual patient
meta-analysis of self-monitoring of an oral anticoagulation protocol.
J Heart Valve Dis 2008; 17: 23338.
10 Dickersin K, Manheimer E, Wieland S, Robinson KA, Lefebvre C,
McDonald S. Development of the Cochrane Collaborations
CENTRAL Register of controlled clinical trials. Eval Health Prof
2002; 25: 3864.
11 Schulman S, Kearon C. Denition of major bleeding in clinical
investigations of antihemostatic medicinal products in non-surgical
patients. J Thromb Haemost 2005; 3: 69294.
12 Tinazzi A, Tierney J, Stewart L, Parmar M, Torri V. Survival
curve and hazard ratio program (SCHARP) for meta-analysis
of individual patient data.
Syst Rev Evid Action Int Cochrane Colloq 6th 1998 Baltim Md 1998;
6: 65.
13 Tierney JF, Stewart LA, Ghersi D, Burdett S, Sydes MR. Practical
methods for incorporating summary time-to-event data into
meta-analysis. Trials 2007; 8: 16.
14 DerSimonian R, Laird N. Meta-analysis in clinical trials.
Control Clin Trials 1986; 7: 17788.
15 Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring
inconsistency in meta-analyses. BMJ 2003; 327: 55760.
16 Altman DG, Andersen PK. Calculating the number needed to
treat for trials where the outcome is time to an event. BMJ 1999;
319: 149295.
17 Cromheecke ME, Levi M, Colly LP, et al. Oral anticoagulation
self-management and management by a specialist anticoagulation
clinic: a randomised cross-over comparison. Lancet 2000;
356: 97102.
18 Krtke H, Minami K, Breymann T, et al. INR self-management
after mechanical heart valve replacement: ESCAT (Early
Self-Controlled Anticoagulation Trial). Z Kardiol 2001;
90 (suppl 6): 11824.
19 Menendez-Jandula B, Souto JC, Oliver A, et al. Comparing
self-management of oral anticoagulant therapy with clinic
management: a randomized trial. Ann Intern Med 2005; 142: 110.
20 Fitzmaurice DA, Murray ET, McCahon D, et al. Self management
of oral anticoagulation: randomised trial. BMJ 2005; 331: 1057.
21 Christensen TD, Maegaard M, Sorensen HT, Hjortdal VE,
Hasenkam JM. Self-management versus conventional management
of oral anticoagulant therapy: a randomized, controlled trial.
Eur J Intern Med 2006; 17: 26066.
22 Matchar DB, Jacobson A, Dolor R, et al. Eect of home testing
of international normalized ratio on clinical events. N Engl J Med
2010; 363: 160820.

www.thelancet.com Vol 379 January 28, 2012

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

Siebenhofer A, Rakovac I, Kleespies C, Piso B, Didjurgeit U.


Self-management of oral anticoagulation reduces major outcomes
in the elderly. A randomized controlled trial. Thromb Haemost 2008;
100: 108998.
Kaatz S, Elston-Lafata J, Gooldy S. Anticoagulation therapy home
and oce monitoring evaluation study. J Thromb Thrombolysis 2001;
12: 111.
Rosendaal FR, Cannegieter SC, van der Meer FJ, Briet E. A method
to determine the optimal intensity of oral anticoagulant therapy.
Thromb Haemost 1993; 69: 23639.
White RH, McCurdy SA, von Marensdor H, Woodru DE Jr,
Leftgo L. Home prothrombin time monitoring after the
initiation of warfarin therapy. A randomized, prospective
study. Ann Intern Med 1989; 111: 73037.
Horstkotte D, Piper C, Wiemer M. Optimal frequency of patient
monitoring and intensity of oral anticoagulation therapy in valvular
heart disease. J Thromb Thrombolysis 1998; 5 (suppl 1): 1924.
Sawicki PT. A structured teaching and self-management program
for patients receiving oral anticoagulation: a randomized controlled
trial. Working group for the study of patient self-management of
oral anticoagulation. JAMA 1999; 281: 14550.
Khan TI, Kamali F, Kesteven P, Avery P, Wynne H. The value
of education and self-monitoring in the management of warfarin
therapy in older patients with unstable control of anticoagulation.
Br J Haematol 2004; 126: 55764.
Fitzmaurice DA, Murray ET, Gee KM, Allan TF, Hobbs FD.
A randomised controlled trial of patient self management
of oral anticoagulation treatment compared with primary
care management. J Clin Pathol 2002; 55: 84549.
Sidhu P, OKane HO. Self-managed anticoagulation: results from
a two-year prospective randomized trial with heart valve patients.
Ann Thorac Surg 2001; 72: 152327.
Gadisseur AP, Breukink-Engbers WG, van der Meer FJ,
van den Besselaar AM, Sturk A, Rosendaal FR. Comparison
of the quality of oral anticoagulant therapy through patient
self-management and management by specialized anticoagulation
clinics in the Netherlands: a randomized clinical trial.
Arch Intern Med 2003; 163: 263946.
Gardiner C, Williams K, Longair I, Mackie IJ, Machin SJ, Cohen H.
A randomised control trial of patient self-management of oral
anticoagulation compared with patient self-testing. Br J Haematol
2006; 132: 598603.
Dauphin C, Legault B, Jaeux P, et al. Comparison of INR
stability between self-monitoring and standard laboratory method:
preliminary results of a prospective study in 67 mechanical heart
valve patients. Arch Cardiovasc Dis 2008; 101: 75361.
Soliman Hamad MA, van Eekelen E, van Agt T, van Straten AH.
Self-management program improves anticoagulation
control and quality of life: a prospective randomized study.
Eur J Cardiothorac Surg 2009; 35: 26569.
Beyth RJ, Quinn L, Landefeld CS. A multicomponent intervention
to prevent major bleeding complications in older patients receiving
warfarin. A randomized, controlled trial. Ann Intern Med 2000;
133: 68795.
Sunderji R, Gin K, Shalansky K, et al. A randomized trial of patient
self-managed versus physician-managed oral anticoagulation.
Can J Cardiol 2004; 20: 111723.
Vller H, Glatz J, Taborski U, Bernardo A, Dovifat C, Heidinger K.
Self-management of oral anticoagulation in nonvalvular atrial
brillation (SMAAF study). Z Kardiol 2005; 94: 18286.
Bloomeld HE, Krause A, Greer N, et al. Meta-analysis: eect of
patient self-testing and self-management of long-term
anticoagulation on major clinical outcomes. Ann Intern Med 2011;
154: 47282.
Dolor RJ, Ruybalid RL, Uyeda L, et al. An evaluation of patient
self-testing competency of prothrombin time for managing
anticoagulation: pre-randomization results of VA Cooperative Study
#481The Home INR Study (THINRS). J Thromb Thrombolysis 2010;
30: 26375.
Baigent C, Keech A, Kearney PM, et al. Ecacy and safety of
cholesterol-lowering treatment: prospective meta-analysis of data
from 90,056 participants in 14 randomised trials of statins. Lancet
2005; 366: 126778.

333

Articles

42

43

44

45

334

Thavendiranathan P, Bagai A, Brookhart MA, Choudhry NK.


Primary prevention of cardiovascular diseases with statin therapy:
a meta-analysis of randomized controlled trials. Arch Intern Med
2006; 166: 230713.
van Walraven C, Jennings A, Oake N, Fergusson D, Forster AJ.
Eect of study setting on anticoagulation control: a systematic
review and metaregression. Chest 2006; 129: 115566.
Mant J, Hobbs FD, Fletcher K, et al. Warfarin versus aspirin for
stroke prevention in an elderly community population with atrial
brillation (the Birmingham Atrial Fibrillation Treatment of the
Aged Study, BAFTA): a randomised controlled trial. Lancet 2007;
370: 493503.
Glasziou PP, Irwig L, Heritier S, Simes RJ. Monitoring cholesterol
levels: measurement error or true change? Ann Intern Med 2008;
148: 65661.

46

47

48

49

Bauman ME, Black K, Bauman ML, et al. EMPoWarMENT:


Edmonton pediatric warfarin self-management pilot study
in children with primarily cardiac disease. Thromb Res 2010;
126: e11015.
Connock M, Stevens C, Fry-Smith A, et al. Clinical eectiveness
and cost-eectiveness of dierent models of managing long-term
oral anticoagulation therapy: a systematic review and economic
modelling. Health Technol Assess 2007; 11: iii66.
Regier DA, Sunderji R, Lynd LD, Gin K, Marra CA.
Cost-eectiveness of self-managed versus physician-managed
oral anticoagulation therapy. CMAJ 2006; 174: 184752.
Siebenhofer A, Berghold A, Sawicki PT. Systematic review
of studies of self-management of oral anticoagulation.
Thromb Haemost 2004; 91: 22532.

www.thelancet.com Vol 379 January 28, 2012

Articles

Neonatal screening for lysosomal storage disorders:


feasibility and incidence from a nationwide study in Austria
Thomas P Mechtler*, Susanne Stary*, Thomas F Metz, Vctor R De Jess, Susanne Greber-Platzer, Arnold Pollak, Kurt R Herkner, Berthold Streubel,
David C Kasper

Summary
Background The interest in neonatal screening for lysosomal storage disorders has increased substantially because of
newly developed enzyme replacement therapies, the need for early diagnosis, and technical advances. We tested for
Gauchers disease, Pompes disease, Fabrys disease, and Niemann-Pick disease types A and B in an anonymous
prospective nationwide screening study that included genetic mutation analysis to assess the practicality and
appropriateness of including these disorders in neonatal screening panels.

Lancet 2012; 379: 33541

Methods Specimens from dried blood spots of 34 736 newborn babies were collected consecutively from January, 2010
to July, 2010, as part of the national routine Austrian newborn screening programme. Anonymised samples were
analysed for enzyme activities of acid -glucocerebrosidase, -galactosidase, -glucosidase, and acid sphingomyelinase
by electrospray ionisation tandem mass spectrometry. Genetic mutation analyses were done in samples with suspected
enzyme deciency.

*TPM and SS contributed equally

Findings All 34 736 samples were analysed successfully by the multiplex screening assay. Low enzyme activities were
detected in 38 babies. Mutation analysis conrmed lysosomal storage disorders in 15 of them. The most frequent
mutations were found for Fabrys disease (1 per 3859 births), followed by Pompes disease (1 per 8684), and Gauchers
disease (1 per 17 368). The positive predictive values were 32% (95% CI 1652), 80% (2899), and 50% (793),
respectively. Mutational analysis detected predominantly missense mutations associated with a late-onset phenotype.
Interpretation The combined overall proportion of infants carrying a mutation for lysosomal storage disorders was
higher than expected. Neonatal screening for lysosomal storage disorders is likely to raise challenges for primary
health-care providers. Furthermore, the high frequency of late-onset mutations makes lysosomal storage disorders a
broad health problem beyond childhood.
Funding Austrian Ministry of Health, Family, and Women.

Introduction
Starting in the early 1960s, neonatal screening was the
rst organised nationwide eort to identify inborn
errors of metabolism and endocrine defects at a
presymptomatic stage. It was revolutionised by the
introduction of electrospray ionisation tandem mass
spectrometry (ESI-MS) in 1993,1 which enabled for the
rst time the identication of several disorders by
simultaneous measurement of aminoacids and
acylcarnitines.2 The increased technological capacity
means that expanded neonatal screening programmes
can now identify a broad range of disorders in which
early detection and presymptomatic treatment result in
clinical benet. A further and more controversial benet
of such expanded programmes is the opportunity to
inform parents of future reproductive risks.3
Lysosomal storage disorders are an attractive candidate for an expanded neonatal screening programme.
These disorders result in the accumulation of macromolecular substrates that would normally be degraded
by enzymes involved in lysosomal metabolism.4 Although
individual lysosomal storage disorders are believed to be
rare, their combined incidence has been estimated at
1 per 7700 livebirths for white people,5 and a third of
www.thelancet.com Vol 379 January 28, 2012

inborn errors of metabolism in the Arabic population


are attributed to these disorders.6 Progressive lysosomal
substrate deposition can occur in cells throughout the
body and results in gradual deterioration of renal
function to end-stage renal disease, cerebrovascular,
cardiovascular, or neurological disease, muscle weakness,
and cardiomyopathy among others.79 Management by a
multidisciplinary team is necessary. Lysosomal storage
disorders have a progressive course, and can present
at any age and aect any number of tissues and
organ systems.10 In most cases, treatment is directed
toward symptomatic care of secondary complications.
Haemopoietic stem-cell transplantation and enzymereplacement therapy are eective in some patients,
although these therapies still have limitations.11 Nonetheless, early diagnosis and treatment are essential for
optimum outcomes, which encourages the incorporation
of lysosomal storage disorders into the neonatal
screening panel.12
The technology for simultaneous screening of several
enzyme activities related to lysosomal storage disorders
from one blood punch was initially complicated, timeconsuming, and laborious but new protocols and
technologies are now available that allow simplied

Published Online
November 30, 2011
DOI:10.1016/S01406736(11)61266-X
See Comment page 294

Department of Pediatrics and


Adolescent Medicine
(T P Mechtler MD, T F Metz MSc,
S Greber-Platzer MD,
Prof A Pollak MD,
K R Herkner PhD,
D C Kasper PhD), Department
of Pathology (S Stary PhD,
Prof B Streubel MD), Medical
University of Vienna, Vienna,
Austria; and Newborn
Screening and Molecular
Biology Branch, Centers
for Disease Control and
Prevention, Atlanta, USA
(V R De Jess PhD)
Correspondence to:
Prof Berthold Streubel,
Department of Obstetrics and
Gynecology, Medical University
of Vienna, Vienna A 1090,
Austria
berthold.streubel@
meduniwien.ac.at

335

Articles

Methods
Study design and population
Blood spot samples from 34 736 newborn babies were
collected consecutively from January, 2010, to July, 2010,
during the national Austrian newborn screening
programme. In Austria, neonatal screening is routine
and centralised with more than 99% coverage
(83 million inhabitants, mean 77 496 births per year,
SD 1183 in the past 10 years). We analysed all samples
that were screened successfully with the regular
screening panel for endocrine and metabolic disorders,
including cystic brosis, additionally for four dierent
lysosomal enzyme activities.19 These were acid
-glucocerebrosidase (GBA, decient in Gauchers
disease), -galactosidase (GLA, decient in Fabrys
disease), -glucosidase (GAA, decient in Pompes
disease), and acid sphingomyelinase (ASM, decient in
Niemann-Pick disease types A and B). The samples were
not corrected for white blood cell count because, in
Austria, a second blood spot sample is taken from
preterm babies, those of very low birthweight, and those
who are sick.20 Dried blood spots from potentially
enzyme-decient infants were retested in duplicates,
and positive results were diagnostically conrmed by
subsequent mutation analyses. The ethics committee of
the Medical University of Vienna and Vienna General
Hospital approved the study (EK 478/2009).

Sample preparation and ESI-MS

See Online for webappendix

Screening for lysosomal storage disorders with ESI-MS


technology is technically feasible.13,17,2125 In these previous
studies, all 212 patients tested (76 Fabrys disease,
71 Pompes disease, 48 Gauchers disease, and
17 Niemann-Pick disease types A and B) were identied
with 100% sensitivity and high specicity. We adapted
and modied the direct multiplex assay to screen for the
lysosomal storage disorders on dried blood spots by use
of automatised sample preparation and pipetting steps
with liquid handling stations to enable high-throughput
screening (webappendix pp 1331). The total analysis
time for 100 samples was about 3 h.

Quality control and data processing


We used low (enzyme activity below cuto) and medium
(enzyme activity above cuto) control materials provided
by the US Centers for Disease Control and Prevention
for each 96-well plate.26 Plates were judged valid when
336

the single enzyme activity of quality controls was in the


range recommended by the manufacturer. Additionally,
32 blood samples from aected individuals served as
controls (eight Pompes disease, ten Gauchers disease,
13 Fabrys disease, and one Niemann-Pick disease types A
and B). If quality control did not match, the complete
analysis for that batch was repeated. The retests were
done at least in duplicate from two separate blood spots
to validate the results and minimise the number of false
positive results with low enzyme activities. This practice
is in line with our neonatal screening procedures for
other metabolic disorders in which second-tier tests are
not available.

Sequence analysis
We isolated genomic DNA directly from the dried blood
spot card using QIAamp DNA micro Kit (Qiagen,
Carlsbad, CA, USA). We amplied exons and intron to
exon boundaries of the genes GBA, GAA, ASM, and
GLA by PCR using JumpStart RED Taq Ready Mix
(Sigma-Aldrich, St Louis, MO, USA) according to the
suppliers instructions. We sequenced PCR using a
3130 Genetic Analyzer (Applied Biosystems, Foster City,
CA, USA) following standard procedures. Primers used
for PCR amplication and sequencing are listed in the
32
30

n=34 732

28 n=34 733
26
24
Enzyme activity in mol/L per h

screening.1315 Experience of nationwide screening for


these disorders is scarce.1618 We implemented a multiplex
high-throughput screening assay for Gauchers disease,
Pompes disease, Fabrys disease, and Niemann-Pick
disease types A and B in an anonymous prospective
nationwide screening study that included genetic
mutation analysis to assess the practicality and
appropriateness of including these disorders in neonatal
screening panels.

22
20
18

16

n=34 727

14

12

n=34 735

10
8
6
4
2

n=3*
n=10

n=9 n=13

n=4 n=8

n=1

0
N S P
N P
Gauchers Niemann-Pick A/B
(GBA)
(ASM)

S P
Pompes
(GAA)

S P
Fabrys
(GLA)

Figure 1: Overview of lysosomal enzyme activity in the healthy Austrian


population, aected patients, and neonates as detected in the pilot
screening (n=34 736)
Error bars are 10th and 90th percentiles. Dotted lines are minimum values for
normal activity (01 percentile). N=neonates with normal lysosomal enzyme
activity. S=neonates with low lysosomal activity conrmed by mutation
analysis. P=patients with lysosomal storage disorders (controls).
GBA=-glucocerebrosidase. GAA=-glucosidase. GLA=-galactosidase.
ASM=acidsphingomyelinase. *Including one sample with a missense mutation
(modier) c.1093GA (p.Glu365Lys). Numbers indicate samples with normal
biochemical test for lysosomal enzyme activities. Detailed statistical information
is provided in the webappendix p 1.

www.thelancet.com Vol 379 January 28, 2012

Articles

webappendix pp 56. Unless otherwise noted, primers


used for PCR and sequencing reactions were identical.
Sequence analysis was pursued until two potentially
pathogenetic mutations were identied in the relevant
gene (or one mutation in the case of GLA analysis in
male patients). GenBank entries NM_000157.3 (GBA),
NM_000152.3 (GAA), and NM_000169.2 (GLA) were
used as references for coding regions, whereby
nucleotide A of the translation initiation codon ATG
constituted numbering +1 of the cDNA sequences. We
followed the standard naming conventions of the Human
Genome Variation Society, such that methionine encoded
by the translation initiation codon was designated as
position 1 in aminoacid numbering, by contrast with

Positive by rst-line
biochemical screening

common GBA variant naming, which omits aminoacid


residues of the GBA signal peptide.

Statistical analysis
We analysed all mass spectrometry data with Analyst 15
(AB Sciex, Foster City, CA, USA). We used SPSS version
160 for data analysis and we calculated all CIs according
to the Clopper-Pearson condence method.27

Role of the funding source


The sponsor of the study had no role in study design,
data collection, or analysis including interpretation. All
authors had full access to all the data used in this study,
take responsibility for the integrity and the accuracy of

Conrmed by
Positive predictive
mutation analysis value % (95% CI)

False-positive

False-positive rate
per million (95% CI)

Gauchers (GBA)

50 (793)

Pompes (GAA)

80 (2899)

30 (1160)

28

32 (1652)

19

550 (330850)

Fabrys (GLA)
Niemann-Pick types A/B (ASM)
Total

60 (10210)

0 (095)

30 (1160)

38

15

40 (2457)

23

660 (420990)

GBA=-glucocerebrosidase. GAA=-glucosidase. GLA= -galactosidase. ASM=acidsphingomyelinase.

Table 1: Detection of enzyme deciencies and corresponding lysosomal storage disorders (n=34 736)

Sex

Enzyme activity*

Gene

Female

10

GBA

Male

18

Male

cDNA

Protein

Mutation type

Phenotype

604CT
1226AG

Arg202X
Asn409Ser

Nonsense
Missense

Severe
Mild

GBA

680AT
680AT

Asn227IIe
Asn227IIe

Missense
Missense

Mild
Mild

04

GAA

896TC
896TC

Leu299Pro
Leu299Pro

Missense
Missense

Potentially less severe


Potentially less severe

Male

07

GAA

3213TG
1551+1GA

V480_I517del

Splicing
Splicing

Potentially mild
Very severe

Male

13

GAA

896TC
896TC

Leu299Pro
Leu299Pro

Missense
Missense

Potentially less severe


Potentially less severe

Female

13

GAA

3213TG
3213TG

Splicing
Splicing

Potentially mild
Potentially mild

Gauchers disease

Pompes disease

Fabrys disease
7

Male

05

GLA

335GA

Arg112His

Missense

Mild

Male

07

GLA

1076TC

Ile359Thr

Missense

Not known

Male

11

GLA

427GA

Ala143Thr

Missense

Cardiovascular or renal

10

Male

12

GLA

335GA

Arg112His

Missense

Mild

11

Male

13

GLA

427GA

Ala143Thr

Missense

Cardiovascular or renal

12

Female

14

GLA

427GA

Ala143Thr

Missense

Cardiovascular or renal

13

Female

17

GLA

427GA

Ala143Thr

Missense

Cardiovascular or renal

14

Male

22

GLA

427GA

Ala143Thr

Missense

Cardiovascular or renal

15

Female

25

GLA

427GA

Ala143Thr

Missense

Cardiovascular or renal

GBA=-glucocerebrosidase. GAA=-glucosidase. GLA=-galactosidase. *In mol/L per h. References are provided in webappendix pp 24.

Table 2: Mutations and enzyme activities of neonates

www.thelancet.com Vol 379 January 28, 2012

337

Articles

the data analysis, and had responsibility for the decision


to submit for publication.

Results
Minimum values for normal activity are above 40 mol/L
per h for Gauchers disease, 20 mol/L per h for Pompes
disease, 28 mol/L per h for Fabrys disease, and
13 mol/L per h for Niemann-Pick disease types A and B,
according to the 01 percentile of 5000 samples of each
enzymes activity. The coecient of variation was less
than 12% for GAA, GBA, and ASM, and less than 15% for
GLA with low and medium quality controls.
The adapted biochemical multiplex screening assay was
successful for all 34 736 samples. The 32 samples from
known aected patients had low enzyme activities that
could be clearly distinguished from the activities in
samples from healthy controls (gure 1). A detailed
overview of single enzyme activities of the controls is
given in the webappendix p 7.
This rst-line screening identied 124 samples with low
enzyme activity: 29 samples for GBA, 25 for GAA, 42 for

Samples with normal activity


Samples with positive biochemical result, not conrmed by
mutation analysis
Samples from babies with positive biochemical result conrmed by
mutation analysis
Samples from babies with positive biochemical result conrmed by
mutation analysis
50
n=34 727

100 percentile (46 mol/L per h)

45

40

Enzyme activity in mol/L per h

35
10 percentile (31 mol/L per h)
30

01 percentile (28 mol/L per h)

25

20

15
n=3
10
n=19
05
n=6

n=6
n=7

0
Normal

Male

Female

Male

Female

Neonatal screening
Aected patients
Fabrys disease

Figure 2: -galactosidase activity for Fabrys disease detected by neonatal


screening and in aected patients

338

GLA, and 28 for ASM. These samples were retested in


duplicates from the same dried blood spot card; 38 babies
were found to have low enzyme activity by ESI-MS (table 1).
Mutation analysis was done in these 38 cases and 15 babies
(1 per 2315 births; 95% CI 1 per 1403 to 1 per 4136) who
were positive by genetic testing were classed as conrmed
cases (gure 1). The positive predictive value for the
lysosomal storage disorders tested in this experimental
screening setting were greatest for GAA, followed by
GBA, and then GLA (table 1). The one sample that had a
positive ESI-MS screening result for ASM was negative
by genetic analysis.
Sequence analyses of the GBA, GLA, GAA, and ASM
genes were successful in all 38 neonates with low enzyme
activity detected by ESI-MS. 22 cases showed variants of
unproven clinical signicance and were reported as
negative (webappendix p 8). In one baby with suspected
Gauchers disease a DNA missense mutation 1093GA
(Glu365Lys) was detected. This case was regarded as
negative for Gauchers disease because this mutation is
believed to act as a modier28 and no additional mutation
was detected. 15 cases showed mutations that were judged
to be pathogenetic. Structural eects of these mutations
and previous reports on the pathogenetic potential are
cited in the webappendix pp 24. For two DNA mutations,
GAA 1551+1GA (case 4) and GBA 680AT (case 2),
other single-base substitutions aecting either the same
genomic position or the same aminoacid residue are
known to be pathogenetic. Comparing biochemical with
genetic data for the 15 conrmed cases, we did not nd
any correlation between mutation and low enzyme activity
(as measured by ESI-MS), nor could we estimate type of
mutation by the degree to which enzyme activity was
decreased. Most mutations (75%) were associated with a
mild phenotype related to later onset and slow disease
progression (table 2).
Fabrys disease is an X-linked recessive disorder. In our
study, the enzyme activities for GLA were in the range of
0308 mol/L per h for aected male controls and
0831 mol/L per h for aected female controls (webappendix p 7 and p 9). The median enzyme activity in
neonates identied as having low activity by ESI-MS
screening was lower in boys (median 12 mol/L
per h, IQR 0813) than in girls (median 17 mol/L
per h, IQR 1621). Figure 2 shows an overview of all
GLA enzyme activities. We concluded that these three
girls with heterozygote GLA mutation (cases 12, 13, 15)
and diminished GLA enzyme activity were likely to have
clinical manifestations. Use of a cuto value of 31 mol/L
per h (10 percentile) rather than 28 mol/L per h
(01 percentile) would result in an increase from 28 to
294 biochemically positive cases, and a decrease in
specicity (webappendix p 10).

Discussion
Lysosomal storage disorders are only one of a new
category of disorders that will confront clinicians with
www.thelancet.com Vol 379 January 28, 2012

Articles

dicult decisions for nationwide neonatal screening


programmes. Our results showed an unexpectedly
high number of babies with enzyme deciencies in a
predominantly white population in a central European
country (panel). We conrmed 15 cases with pathogenetic
mutations in addition to low lysosomal enzyme activities,
showing a high overall incidence of 1 per 2315 births
among the Austrian population.
Neonatal screening programmes are generally intended
to dierentiate asymptomatic babies who might have
a disease from those who do not, for disorders in
which early detection and urgent presymptomatic
treatment are necessary to avert serious clinical harm.29
Furthermore, a disorder must have a high prevalence to
justify inclusion in such programmes. Before 2002,
neonatal screening in Austria was provided for
six disordersphenylketonuria (1 per 12 000 births),
galactosaemia (1 per 80 000), congenital hypothyroidism
(1 per 3200), biotinidase deciency (1 per 80 000), cystic
brosis (1 per 3400), and congenital adrenal hyperplasia
(1 per 10 000).19 The introduction of ESI-MS increased the
capacity to test infants for more than 30 metabolic
disorders within the programme. The overall incidence
of disorders found by the Austrian neonatal screening
is about 1 per 800 births, and the incidences of
these disorders are similar to other countries with
predominantly white populations.19,30
For the detection of several lysosomal storage disorders
by routine neonatal screening, an additional tandem
mass spectrometry system and at least one extra
laboratory worker for sample preparation were needed.
The running costs were about 1 per sample for all four
multiplexed lysosomal storage disorders (webappendix
pp 1112), which is much the same as for other screening
assays such as 17-hydroxyprogesterone or isotopes for
aminoacid and acylcarnitine analysis. The positive
predictive values in our study were in the range of
3280%. These are similar to the positive predictive
values of other metabolites analysed in the neonatal
screening panel for metabolic and endocrine disorders.30,31
Thus, frequency, positive predictive value, and technical
practicability make nationwide neonatal screening for
lysosomal storage disorders technically feasible.
The high frequency of lysosomal storage disorders with
an overall incidence of 1 per 2315 births raises the issue of
whether and why the incidence has been underestimated.
Table 3 shows an overview of previous ndings. In our
study, positive cases were contributed predominantly by
Fabrys disease with an incidence of late-onset disease of
1 per 4100 in the Austrian population. Fabrys disease
occurs in all ethnic, racial, and demographic groups.9 Our
results accord with those of Spada and colleagues34 who
reported an incidence of 1 per 3100 for late-onset disease
and 1 per 37 000 for the classic phenotype. Furthermore,
results of several studies have shown that patients with
renal insuciency, cerebral infarctions, or left ventricular
hypertrophy of unknown cause might have Fabrys
www.thelancet.com Vol 379 January 28, 2012

Panel: Research in context


Systemic review
We searched the American College of Medical Genetics,
Orphanet, and PubMed, from January, 2000, to July, 2011,
with the term lysosomal storage disorders in combination
with nationwide OR high-risk population screening. We
identied several small studies, the results of which suggest
putative high frequencies among risk groups. However,
prospective nationwide studies for lysosomal storage
disorders are missing and consequently information on
feasibility and incidences are unknown. To identify
frequencies and the relevance of lysosomal storage disorders
for an entire population, we did nationwide neonatal
screening in Austria. We adapted established assays using
mass spectrometry for high-throughput screening.
Interpretation
Nationwide neonatal screening for lysosomal storage diseases
using electrospray ionisation tandem mass spectrometry is
technically feasible. Incidences were provided by the
combination of repeated biochemical screening and genetic
testing. The combined incidence was higher than expected
(1 per 2315 births).The mutation spectrum suggests a high
number of late-onset diseases. The ndings of our study add
relevant information for neonatal screening programmes and
draw attention to potential high-risk groups beyond
childhood including those with end-stage renal disease,
cardiovascular disorders, and cardiomyopathy.

Nationwide neonatal screening


study in Austria (n=34 736)

Previous selective studies

Gauchers disease

1 per 17 368 births

Ashkenazi Jews 1 per 800 births32

Pompes disease

1 per 8684

Infantile Pompes disease: Taiwan, 1 per


33 000 based on neonatal screening;16
Netherlands, 1 per 138 00033
Overall prevalence: Netherlands and New York
City, 1 per 40 000 births33

Fabrys disease

1 per 3859

General population: northern Italy, 1 per


3100 births for late-onset disease and 1 per
37 000 births for the classic phenotype;34
Taiwan, 1 per 2400 livebirths and 1 per
1600 in live male births35
High-risk population:33 male dialysis patients,
1 per 100 to 1 per 1000; idiopathic
hypertrophic cardiomyopathy, 1 per 20 to
1 per 30; cryptogenic strokes, 1 per 20 male
(1 per 40 female) patients

Niemann-Pick disease Not found


types A and B

Estimated 1 per 250 000 births36

Table 3: Reported incidences of lysosomal storage disorders

disease.37 Thus, neonatal screening could be benecial to


identify severe clinical cases, but has the drawbacks of
detecting mild forms, late-onset cases, and asymptomatic
cases. Fabrys disease is an X-linked recessive disorder.
The clinical manifestations in female heterozygotes range
from no signs or symptoms throughout a normal lifespan
339

Articles

to those as severe as in aected male patients.38 There is


no consensus as to whether random X-inactivation results
in the highly variable activity of GLA in female cases,
which causes diculties in screening studies.39 In our
study the enzyme activities for GLA were lower for aected
male controls than aected female controls, and in the
neonatal screening group, we found a small dierence
between boys and girls in GLA activity (webappendix p 7),
which might result in false-negative results21 even though
three of nine cases recorded were girls.
The high incidence of the late-onset phenotypes in
these diseases raises the question of when genetic
screening for these disorders should be undertakenin
the neonatal period or at early maturity? Clearly,
early detection, genetic counselling, and therapeutic intervention are benecial for the classic phenotype but the
time of screening for the late-onset variants of Fabrys
disease and other treatable diseases might raise concerns.40
The availability of enzyme replacement therapy for Fabrys
disease is an additional incentive for the introduction of
neonatal screening for this disorder.41 In one study, longterm treatment led to substantial and sustained clinical
benets; however, advanced cardiac and renal disease
cannot be reversed later, so early diagnosis is crucial.42
Neonatal screening and genetic analysis are less
controversial for infantile Pompes disease. In Taiwan,
screening for this disease and initiation of treatment
before onset of symptoms and irreversible muscle damage
showed a clear benet for infants.16 However, debates
about long-term clinical status in treated patients with
infantile-onset Pompes disease remain because little
information is available on the duration of improvement
in muscle and cardiac function. Furthermore, CNS
damage cannot be treated with enzyme replacement
therapies for neuronopathic lysosomal storage disorders
such as Gauchers II and Niemann-Pick A diseases.43 The
importance of consented genotyping and phenotype
prediction after biochemical rst-line screening is clear.
Apart from the potential clinical benet for patients,
neonatal screening for lysosomal storage disorders can
provide information on reproductive risk for parents and
future adults. This situation is common for screening of
metabolic disorders because most are inherited in a
recessive manner.
Currently, three US states have have passed legislation
to include several lysosomal storage disorders in their
neonatal screening programmes in the near future, and
Washington State is doing a pilot study to detect Pompes
disease, Fabrys disease, and mucopolysaccharidosis I
disease. In Europe, the regional neonatal screening
programme in Florence, Italy, will start in 2011.14,44
Population-based screening assays are already available
for lysosomal storage disorders, but new high-throughput
screening assays and novel treatment strategies are on the
horizon for many others. However, in the future we
will need: large population studies including people of
dierent ethnic backgrounds to obtain accurate
340

information about incidences; open studies to characterise


mild and insignicant clinical subtypes; and studies
including groups at high risk of developing the disorders
in adulthood (eg, patients with end-stage renal disease,
cardiomyopathy, cryptogenic stroke). This information
will be crucial to address important ethical issues including
the detection of mild or potentially insignicant clinical
phenotypes and counselling about reproductive risk.
Furthermore, better clinical characterisation should help
to improve therapy schemes and to justify potentially costintensive enzyme replacement therapies.
Contributors
TPM, TFM, SGP, AP, KRH, and DCK were responsible for the
biochemical screening testing. SS and BS were responsible for the
genetic testing. VRDJ was responsible for quality control materials. BS,
TPM, and DCK wrote the report. All authors approved the nal version.
Conicts of interest
We declare that we have no conicts of interest.
Acknowledgments
This work was supported by grants from the Austrian Ministry of
Health, Family, and Women (GZ 20.501/40-3 IV/A/2007). The ndings
and conclusions in this report are those of the authors and do not
necessarily represent the views of the Centers for Disease Control
and Prevention.
References
1
Chace DH, Millington DS, Terada N, Kahler SG, Roe CR,
Hofman LF. Rapid diagnosis of phenylketonuria by quantitative
analysis for phenylalanine and tyrosine in neonatal blood spots
by tandem mass spectrometry. Clin Chem 1993; 39: 6671.
2
Wilcken B, Wiley V, Hammond J, Carpenter K. Screening newborns
for inborn errors of metabolism by tandem mass spectrometry.
N Engl J Med 2003; 348: 230412.
3
Bombard Y, Miller FA, Hayeems RZ, Avard D, Knoppers BM.
Reconsidering reproductive benet through newborn screening:
a systematic review of guidelines on preconception, prenatal and
newborn screening. Eur J Hum Genet 2010; 18: 75160.
4
Wenger DA, Coppola S, Liu SL. Insights into the diagnosis
and treatment of lysosomal storage diseases. Arch Neurol 2003;
60: 32228.
5
Meikle PJ, Hopwood JJ, Clague AE, Carey WF. Prevalence
of lysosomal storage disorders. JAMA 1999; 281: 24954.
6
Moammar H, Cheriyan G, Mathew R, Al-Sannaa N. Incidence and
patterns of inborn errors of metabolism in the Eastern Province of
Saudi Arabia, 19832008. Ann Saudi Med 2010; 30: 27177.
7
Grabowski GA. Phenotype, diagnosis, and treatment of Gauchers
disease. Lancet 2008; 372: 126371.
8
van der Ploeg AT, Reuser AJJ. Pompes disease. Lancet 2008;
372: 134253.
9
Zarate YA, Hopkin RJ. Fabrys disease. Lancet 2008; 372: 142735.
10 Beck M. Variable clinical presentation in lysosomal storage
disorders. J Inherit Metab Dis 2001; 24 (suppl 2): 4751.
11 Beck M. Therapy for lysosomal storage disorders. IUBMB Life 2010;
62: 3340.
12 Beck M. New therapeutic options for lysosomal storage disorders:
enzyme replacement, small molecules and gene therapy.
Hum Genet 2007; 121: 122.
13 Metz TF, Mechtler TP, Orsini J, et al. Simplied newborn screening
protocol for lysosomal storage disorders. Clin Chem 2011;
57: 128694.
14 Duey TA, Bellamy G, Elliott S, et al. A tandem mass
spectrometry triplex assay for the detection of Fabry, Pompe, and
mucopolysaccharidosis-I (Hurler). Clin Chem 2010; 56: 185461.
15 Millington DS, Sista R, Eckhardt A, et al. Digital microuidics:
a future technology in the newborn screening laboratory?
Semin Perinatol 2010; 34: 16369.
16 Chien YH, Lee NC, Thurberg BL, et al. Pompe disease in infants:
improving the prognosis by newborn screening and early
treatment. Pediatrics 2009; 124: e111625.

www.thelancet.com Vol 379 January 28, 2012

Articles

17

18

19

20

21

22

23

24

25

26

27
28

29

30

Dajnoki A, Mhl A, Fekete G, et al. Newborn screening for Pompe


disease by measuring acid alpha-glucosidase activity using tandem
mass spectrometry. Clin Chem 2008; 54: 162429.
Duner PK, Caggana M, Orsini JJ, et al. Newborn screening for
Krabbe disease: the New York State model. Pediatr Neurol 2009;
40: 24552.
Kasper DC, Ratschmann R, Metz TF, et al. The national Austrian
newborn screening programeight years experience with mass
spectrometry. Past, present, and future goals. Wien Klin Wochenschr
2010; 122: 60713.
CLSI. Newborn screening for preterm, low birth weight, and sick
newborns; approved guideline. Wayne, PA: Clinical and Laboratory
Standards Institute 2009.
Dajnoki A, Fekete G, Keutzer J, et al. Newborn screening for Fabry
disease by measuring GLA activity using tandem mass
spectrometry. Clin Chim Acta 2010; 411: 142831.
Kasper DC, Herman J, De Jesus VR, Mechtler TP, Metz TF,
Shushan B. The application of multiplexed, multi-dimensional
ultra-high-performance liquid chromatography/tandem mass
spectrometry to the high-throughput screening of lysosomal
storage disorders in newborn dried bloodspots.
Rapid Commun Mass Spectrom 2010; 24: 98694.
Legini E, Orsini JJ, Hung C, et al. Analysis of glucocerebrosidase
activity in dry blood spots using tandem mass spectrometry.
Clin Chim Acta 2011; 412: 34346.
Li Y, Brockmann K, Turecek F, Scott CR, Gelb MH. Tandem mass
spectrometry for the direct assay of enzymes in dried blood spots:
application to newborn screening for Krabbe disease. Clin Chem
2004; 50: 63840.
Zhang XK, Elbin CS, Chuang WL, et al. Multiplex enzyme assay
screening of dried blood spots for lysosomal storage disorders by
using tandem mass spectrometry. Clin Chem 2008; 54: 172528.
De Jesus VR, Zhang XK, Keutzer J, et al. Development and
evaluation of quality control dried blood spot materials in
newborn screening for lysosomal storage disorders. Clin Chem
2009; 55: 15864.
Clopper CJ, Pearson ES. The use of condence or ducial limits
illustrated in the case of the binomial. Biometrika 1934; 26: 40413.
Park JK, Tayebi N, Stubbleeld BK, et al. The E326K mutation
and Gaucher disease: mutation or polymorphism? Clin Genet
2002; 61: 3234.
De Jess VR, Mei JV, Bell CJ, Hannon WH. Improving and
assuring newborn screening laboratory quality worldwide: 30-year
experience at the Centers for Disease Control and Prevention.
Semin Perinatol 2010; 34: 12533.
Schulze A, Lindner M, Kohlmller D, Olgemller K, Mayatepek E,
Homann GF. Expanded newborn screening for inborn errors of
metabolism by electrospray ionization-tandem mass spectrometry:
results, outcome, and implications. Pediatrics 2003; 111: 1399406.

www.thelancet.com Vol 379 January 28, 2012

31

32

33

34

35

36

37

38
39

40
41

42

43
44

Matern D, Tortorelli S, Oglesbee D, Gavrilov D, Rinaldo P.


Reduction of the false-positive rate in newborn screening by
implementation of MS/MS-based second-tier tests: the Mayo Clinic
experience (20042007). J Inherit Metab Dis 2007; 30: 58592.
Beutler E, Grabowski GA. Gaucher disease. In: Scriver CR,
Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular
bases of inherited diseases, Vol. 2, 8th edn. New York: McGraw-Hill,
2001: 363568.
Wang RY, Bodamer OA, Watson MS, Wilcox WR, for the ACMG
Work Group on Diagnostic Conrmation of Lysosomal Storage
Diseases. Lysosomal storage diseases: diagnostic conrmation
and management of presymptomatic individuals. Genet Med 2011;
13: 45784.
Spada M, Pagliardini S, Yasuda M, et al. High incidence
of later-onset Fabry disease revealed by newborn screening.
Am J Hum Genet 2006; 79: 3140.
Lin HY, Chong KW, Hsu JH, et al. High incidence of the cardiac
variant of Fabry disease revealed by newborn screening in the
Taiwan Chinese population. Circ Cardiovasc Genet 2009; 2: 45056.
Simonaro CM, Desnick RJ, McGovern MM, Wasserstein MP,
Schuchman EH. The demographics and distribution of type B
Niemann-Pick disease: novel mutations lead to new genotype/
phenotype correlations. Am J Hum Genet 2002; 71: 141319.
Monserrat L, Gimeno-Blanes JR, Marn F, et al. Prevalence of Fabry
disease in a cohort of 508 unrelated patients with hypertrophic
cardiomyopathy. J Am Coll Cardiol 2007; 50: 2399403.
Mehta A, Beck M, Eyskens F, et al. Fabry disease: a review
of current management strategies. Q JM 2010; 103: 64159.
Maier EM, Osterrieder S, Whybra C, et al. Disease manifestations
and X inactivation in heterozygous females with Fabry disease.
Acta Paediatr Suppl 2006; 95: 3038.
Fletcher JM. Screening for lysosomal storage disorders
a clinical perspective. J Inherit Metab Dis 2006; 29: 40508.
Eng CM, Guon N, Wilcox WR, et al. Safety and ecacy of
recombinant human alpha-galactosidase Areplacement therapy
in Fabrys disease. N Engl J Med 2001; 345: 916.
Mehta A, Beck M, Elliott P, et al, for the Fabry Outcome Survey
investigators. Enzyme replacement therapy with agalsidase alfa in
patients with Fabrys disease: an analysis of registry data. Lancet
2009; 374: 198696.
Hwu WL, Chien YH, Lee NC. Newborn screening for neuropathic
lysosomal storage disorders. J Inherit Metab Dis 2010; 33: 38186.
Zhou H, Fernho P, Vogt RF. Newborn bloodspot screening
for lysosomal storage disorders. J Pediatr 2011; 159: 713.e1.

341

Articles

Single-dose azithromycin versus benzathine benzylpenicillin


for treatment of yaws in children in Papua New Guinea:
an open-label, non-inferiority, randomised trial
Oriol Mitj, Russell Hays, Anthony Ipai, Moses Penias, Raymond Paru, David Fagaho, Elisa de Lazzari, Quique Bassat

Summary
Lancet 2012; 379: 34247
Published Online
January 11, 2012
DOI:10.1016/S01406736(11)61624-3
See Comment page 295
Lihir Medical Centre
International SOS, Newcrest
Mining, Lihir Island,
Papua New Guinea (O Mitj MD,
R Hays MD, A Ipai HEO,
M Penias HEO, R Paru BSc,
D Fagaho BSc); and Barcelona
Centre for International Health
Research, Hospital Clinic,
University of Barcelona,
Barcelona, Spain (O Mitj MD,
E de Lazzari MSc, Q Bassat PhD)
Correspondence to:
Dr Oriol Mitj, Department of
Medicine, Lihir Medical Center,
PO Box 34, Lihir Island, NIP,
Papua New Guinea
oriolmitja@hotmail.com

Background Yawsan endemic treponematosis and, as such, a neglected tropical diseaseis re-emerging in children
in rural, tropical areas. Oral azithromycin is eective for syphilis. We assessed the ecacy of azithromycin compared
with intramuscular long-acting penicillin to treat patients with yaws.
Methods We did an open-label, non-inferiority, randomised trial at Lihir Medical Centre, Papua New Guinea, between
Sept 1, 2010, and Feb 1, 2011. Children aged 6 months to 15 years with a serologically conrmed diagnosis of yaws
were randomly allocated, by a computer-generated randomisation sequence, to receive either one 30 mg/kg oral dose
of azithromycin or an intramuscular injection of 50 000 units per kg benzathine benzylpenicillin. Investigators were
masked to group assignment. The primary endpoint was treatment ecacy, with cure rate dened serologically as a
decrease in rapid plasma reagin titre of at least two dilutions by 6 months after treatment, and, in participants with
primary ulcers, also by epithelialisation of lesions within 2 weeks. Non-inferiority was shown if the upper limit of the
two-sided 95% CI for the dierence in rates was lower than 10%. The primary analysis was per protocol. This trial is
registered with ClinicalTrials.gov, number NCT01382004.
Findings We allocated 124 patients to the azithromycin group and 126 to the benzathine benzylpenicillin group. In the
per-protocol analysis, after 6 months of follow-up, 106 (96%) of 110 patients in the azithromycin group were cured,
compared with 105 (93%) of 113 in the benzathine benzylpenicillin group (treatment dierence 34%; 95% CI
93 to 24), thus meeting prespecied criteria for non-inferiority. The number of drug-related adverse events (all
mild or moderate) was similar in both treatment groups (ten [8%] in the azithromycin group vs eight [7%] in the
benzathine benzylpenicillin group).
Interpretation A single oral dose of azithromycin is non-inferior to benzathine benzylpenicillin and avoids the need
for injection equipment and medically trained personnel. A change to the simpler azithromycin treatment regimen
could enable yaws elimination through mass drug administration programmes.
Funding International SOS and Newcrest Mining.

Introduction
Yawsan endemic treponematosis and, as such, a
neglected tropical diseaseis re-emerging. 40 years after
a worldwide control programme almost eradicated the
disease, it has re-emerged in children in poor, rural, and
marginalised populations in parts of Africa, Asia, and
South America. Yaws is caused by Treponema pallidum
subsp pertenue, and aects mainly skin, bones, and
cartilage. The disease has a natural history in primary,
secondary, and tertiary stages. Unless diagnosed and
treated at an early stage, yaws can become a chronic,
relapsing disease, and can lead to severe deforming bone
lesions in the long term.1
Between 1952, and 1964, WHO and UNICEF led a
worldwide campaign to control and eventually eradicate
yaws and other endemic treponematoses.2 Yaws became
the second disease targeted for eradication, after
smallpox. Control programmes were established in
46 countries and, by the end of 1964, the number of cases
had reduced by 95%, from 50 million to 25 million.
However, control eorts were gradually abandoned in
342

most countries3 and the disease re-emerged in the late


1970s, prompting the adoption of WHOs assembly
resolution 3858.4 According to the last estimate by WHO
in 1995, more than 500 000 children were still aected in
Africa, Asia, and South America.5
Penicillin remains the drug of choice to treat endemic
treponematoses.6,7 WHO guidelines recommend one
intramuscular injection of long-acting benzathine
benzylpenicillin at a dose of 12 MU for adults and
06 MU for children;8 however, other guidelines recommend higher doses.9 This treatment is eective and has
several advantages, as described for venereal syphilis.10
Although this treatment is cheap and well tolerated, it
has drawbacks, including the operational and logistical
diculties related to treatment with drug injection, the
potential risk of transmission of blood-borne pathogens
with unsafe injection practices, the pain related to deep
intramuscular injection of a large volume (4 mL), and a
high rate of self-reported allergy to penicillin.
Oral phenoxymethylpenicillin for 710 days (50 mg/kg
daily in four doses; maximum dose 300 mg four times a
www.thelancet.com Vol 379 January 28, 2012

Articles

day) was eective in a yaws control programme.11 Such a


regimen overcomes the disadvantages of intramuscular
drug administration, but poor adherence to a multiday
treatment regimen is a risk. In pilot studies of the potential
of oral, single-dose treatment against several infectious
disorders, azithromycina macrolide antibiotic with a
long half-life in tissueseemed to be a valuable drug
against Chlamydia trachomatis,12 Neisseria gonorrhoeae,13 and
Haemophilus ducreyi14 infections. Promising results were
also reported from a large-scale study10 done in Tanzania,
with two regimens to treat early syphilis: one oral dose
(2 g) of azithromycin and one intramuscular dose of
benzathine benzylpenicillin 24 MU. A multicentre trial15
in North America and Madagascar had similar ndings.
The immediate-release formulation of azithromycin
given in one oral dose of 30 mg per kg of bodyweight has
been approved and widely used to treat acute otitis media
in children since 2001.16,17 The product is available as an
oral tablet or as syrup, which is easier to administer to
very young children.
We assessed the ecacy of a single oral dose of
azithromycin compared with the standard single
intramuscular dose of benzathine benzylpenicillin to
treat yaws.

Methods
Study setting and patients
We undertook a prospective, open-label, non-inferiority,
randomised controlled trial at Lihir Medical Centre in
Papua New Guinea between Sept 1, 2010, and
Feb 24, 2011. The Lihir islands are geographically
remote, and despite being host to a major gold-mining
operation since 1995, the living conditions and sanitation
remain basic in most areas. Yaws is still a substantial
cause of morbidity in Papua New Guinea.18,19 Monthly
reports for monitoring several indicators of infectious
diseases and maternal and child health are being
collected via forms from hospitals, health centres, and
aid posts throughout the country. The national health
department estimated the number of yaws cases to be
17 000 nationwide in 2003, and 23 000 in 2008, of which
5000 were in New Ireland province, where Lihir Island
is located, and another 5000 in the neighbouring
province of West New Britain (unpublished).
All patients examined in the outpatient medical
department and suspected to have primary-stage or
secondary-stage yaws were assessed for possible inclusion
in the study. Eligible patients were children aged 6 months
to 15 years with a rapid plasma reagin titre of at least 1 in
16 and a reactive T pallidum haemagglutination test.
Exclusion criteria were known allergy to penicillin or
macrolide antibiotics, use of antibiotics active against
T pallidum during the preceding month, and known or
suspected coexisting diseases for which additional
antibiotic treatment with drugs eective against T pallidum
would be needed (use of quinolones, sulphonamides,
trimethoprim, and metronidazole was allowed).
www.thelancet.com Vol 379 January 28, 2012

A diagnosis of yaws chancre (primary stage) was


established by dermatological examination on the
basis of chronic (symptomatic for >2 weeks), painless,
atraumatic ulcers with raised margins. Criteria for the
diagnosis of secondary yaws included the presence
of one or more of: multiple hyperkeratotic papules;
polyarthralgia; or bone pain and swelling aecting the
ngers or toes, forearm, tibia, or bula. When an overlap
between the stages occurred (ie, primary lesion
persisting after the appearance of secondary yaws
symptoms) we classied the case as secondary stage.
The study was approved by the National Medical
Research Advisory Committee of the Papua New Guinea
Ministry of Health. All patients, or their parents,
provided signed informed consent.

Randomisation and masking


Eligible participants were randomly assigned, by use of a
computer-generated random-numbers list, to receive
either 30 mg/kg (maximum 2 g) azithromycin orally or
50 000 units per kg (maximum 24 MU) benzathine
benzylpenicillin by intramuscular injection. Randomisation was done in permuted blocks of four and in a
1:1 ratio. The allocation was concealed from investigators
by use of opaque, sealed, and sequentially numbered
envelopes that were opened after the study team had
decided to enrol a patient. Laboratory technicians were
unaware of participants treatment allocation, treatment
response, and previous rapid plasma reagin results at all
times. All participants received directly observed treatment, but masking of patients was not possible for
logistical reasons.

Procedures
The primary endpoint was serological cure, dened as
a decrease in the rapid plasma reagin titre by at least
two dilutions at the 6-month follow-up examination,
compared with the titre at time of treatment. For ulcers,
improvement of lesions in 2 weeks after treatment was
also needed. Secondary endpoints were the individual
components of the primary endpoint, cure rate 3 months
after treatment, and cure rates according to stage of yaws,
rapid plasma reagin titre at treatment, and history of
household exposure.
To guarantee timely follow-up of participants, we
implemented a community-based follow-up strategy.
A eld team, consisting of a physician, a laboratory
technician, and a local health worker, located patients
twice a week (for follow-up visit) at their residence by
tracking detailed locator information. All participants
were re-examined 2 weeks after treatment to assess
clinical resolution. Photographic documentation of skin
lesions was obtained at diagnosis and at the 14-day followup visit for comparison over time. Patients with
worsening ulcers were retreated with benzathine
benzylpenicillin (at the same dose). We assessed all
participants at 3 months and 6 months after treatment. A
343

Articles

Statistical analysis
255 patients enrolled
5 ineligible
3 declined to participate
2 allergic to benzathine
benzylpenicillin
250 patients underwent
randomisation

124 assigned azithromycin

126 assigned benzathine


benzylpenicillin

12 lost to follow-up
2 adverse events

13 lost to
follow-up

110 completed

113 completed

Figure 1: Trial prole

Azithromycin Benzathine benzylpenicillin


(n=124)
(n=126)
Age (years)

92 (37)

84 (33)

Male sex

54 (44%)

59 (47%)

Exposure*

26 (21%)

20 (16%)

Primary stage

50 (40%)

56 (44%)

Secondary stage

74 (60%)

70 (56%)

Persisting ulcer

18 (15%)

15 (12%)

Secondary skin lesions

16 (13%)

11 (9%)

Arthralgias

68 (55%)

64 (51%)

Bone swelling or pain

12 (10%)

10 (8%)

Clinical presentation

RPR titre at treatment


1 in 32

47 (38%)

60 (48%)

1 in 64

77 (62%)

66 (52%)

Data are mean (SD) or number (%). RPR=rapid plasma regain. *Household
exposure to other children with open skin ulcers during the previous 3 months.

Table 1: Baseline characteristics of the intention-to-treat population

5 mL blood sample was obtained at each follow-up visit


for serological analysis for T pallidum. All rapid plasma
reagin tests were done in duplicate by two independent
trained technicians at the Lihir Medical Centre
microbiology department, and tests were done a third
time in cases of discrepant results.
Safety assessments included documentation of immediate adverse events and patient-reported adverse events.
So that immediate reactions could be recorded and
treated, patients stayed at the health centre for 30 min
after treatment. Patient-reported adverse events were
assessed at the 2 week examination. Patients (or their
parents or guardians) were explicitly asked about pain at
site of injection, rash, fever, vomiting, diarrhoea, and
stomach pain.
344

This study was based on the notion that azithromycin


would be non-inferior to benzathine benzylpenicillin
for the primary ecacy outcome, with use of a prespecied non-inferiority margin; the upper limit of the
95% CI for the dierence in cure rates between groups
would not exceed 10%. We calculated that a sample size
of 242 patients (121 per group) would give a power of
80% to test the hypothesis of non-inferiority. This
sample size accounted for an expected ecacy of
benzathine benzylpenicillin of 95%,11,20 a non-inferiority
margin of 10%, and a one-sided type 1 error rate of 005,
with the assumption that 10% of participants would be
lost to follow-up.
We selected the per-protocol population for the primary
analysis. This population included all patients who
underwent randomisation and who completed the study
procedures to month 6. We also did a supporting analysis
with the intention-to-treat (missing equals failure)
population, which included all eligible patients, and in
which patients with missing data were regarded as having
treatment failure.
For analysis of the primary endpoint (cure rate at
6 months), we estimated two-sided 95% CIs for the
dierence in cure proportions between the benzathine
benzylpenicillin group and the azithromycin group
according to Altman and colleagues method.21 We used
the same method to analyse secondary binary endpoints.
We did additional post-hoc analyses to assess the
consistency of treatment eects in subgroups dened
according to disease stage, rapid plasma reagin titre, and
household exposure, with Fishers exact test. To compare
baseline characteristics and adverse events between the
treatment groups, we used two-sided t and Fishers exact
tests with a signicance level of 005. We did all statistical
analyses with Stata (version 11.1).
This study is registered with ClinicalTrials.gov, number
NCT01382004.

Role of the funding source


The sponsor of the study had no role in study design,
data collection, data analysis, data interpretation, or
writing of the report. All authors had full access to all the
data in the study and had nal responsibility for the
decision to submit for publication.

Results
Figure 1 shows the trial prole. 250 patients with
serologically conrmed yaws were randomly assigned
to receive either azithromycin or benzathine benzylpenicillin. Baseline clinical and serological characteristics
of the two treatment groups were similar (table 1). Mean
age of the participants was 88 years (SD 36; range
8 months to 15 years). 42% of patients had primary yaws
(table 1). The rapid plasma reagin titre was less than 1 in
32 in 107 (43%) participants and 1 in 64 or more in
143 (57%; table 1).
www.thelancet.com Vol 379 January 28, 2012

Articles

25 (10%) of the 250 participants could not be traced: ten


(4%) of these children could not be located for any
follow-up visit and 15 (6%) were lost after the rst followup visit. We could not locate seven (3%) participants
because they provided an invalid address, and 18 (7%)
were originally from the study area but moved elsewhere
during follow-up. Two patients did not complete follow
up because of adverse events related to drug administration. The remaining 223 patients constituted the perprotocol population.
Adverse events in the rst 14 days of treatment
were reported by ten (8%) of 119 patients interviewed in
the azithromycin group, and by eight (7%) of 121 in
the benzathine benzylpenicillin group. Of participants
given azithromycin, six (5%) reported nausea, two (2%)
stomach pain, and two (2%) vomiting within 30 min of
taking the drug. We classed the two patients who vomited
as having had treatment failure in the intention-totreat analyses, and they were retreated with benzathine
benzylpenicillin. In patients given benzathine benzylpenicillin administration-related adverse eects were the
most common. Six (5%) patients in the benzathine
benzylpenicillin group reported persistent injection-site
pain, despite use of lidocaine 1% as a diluent, and two
(2%) had an injection-related abscess. No serious adverse
events were reported during treatment or for the entire
follow-up period.
129 participants with ulcers (in primary or secondary
stage) were re-examined 2 weeks after treatment; the
ulcers had resolved in 51 (40%) and were healing in
70 (54%; gure 2). The rates of healing did not dier
signicantly between the two treatment groups (data not
shown). We classed the remaining eight (6%) patients
(four in the azithromycin group and four in the
benzathine benzylpenicillin group) as having clinical
treatment failure.
In both the per-protocol and intention-to-treat
analyses the criteria for non-inferiority were met for the
composite primary endpoint of serological cure at
6 months and clinical healing of ulcers. In the perprotocol analysis, 106 of 110 patients assigned
azithromycin were cured at 6 months compared with
105 of 113 patients in the benzathine benzylpenicillin
group (risk dierence 34%, 95% CI 93 to 24;
table 2). Incidence of the individual components of the
primary endpoint and intermediate cure rates did not
dier signicantly between groups (table 2). In the
intention-to-treat population, 106 of 124 patients
assigned azithromycin and 105 of 126 patients assigned
benzathine benzylpenicillin met the criteria for the
primary endpoint (22%, 111 to 68; table 2).
In subgroup analyses, the cure rates at 6 months
according to yaws stage, rapid plasma reagin titre at
treatment, and household exposure did not dier
signicantly between treatments in the per-protocol
population (table 3). No participant in either treatment
group had recurrent clinical signs of yaws or serological
www.thelancet.com Vol 379 January 28, 2012

5 cm

5 cm

Figure 2: Ulcers in patients with primary-stage or secondary-stage yaws who were re-examined 2 weeks
after treatment
(A, B) Red, moist, bedded, 5 cm ulcer on the left leg of a 9-year-old patient with primary yaws. (C, D) Partially
epithelialised tumour 2 weeks after treatment with azithromycin.
Azithromycin %
(95% CI)

Risk dierence %
Benzathine
benzylpenicillin % (95% CI)
(95% CI)

Primary population (PP) analysis (n=223)*


Primary endpoint: cure at 6 months
Cure at 3 months

964% (910986) 929% (867964) 34% (93 to 24)


800% (716864) 805% (723868)

Serologically dened cure at 6 months


Clinical cure of ulcers 14 days after treatment

100% (965100)

05% (99 to 109)

963% (903988) 37% (72 to 01)

964% (913986) 965% (910986)

01% (48 to 50)

Secondary population (ITT) analysis (n=250)


Primary endpoint: cure at 6 months

855% (782906) 833% (759888)

Cure at 3 months

710% (625782) 722% (634797)

22% (111 to 68)


13% (99 to 124)

Serologically dened cure at 6 months

883% (814929) 861% (788911)

23% (107 to 61)

Clinical cure of ulcers 14 days after treatment

854% (782906) 865% (795914)

10% (76 to 96)

PP=per protocol. ITT=intention to treat. *Including only patients with complete follow-up and study endpoint.
Including all randomised patients; we regarded patients with missing data as having treatment failure.

Table 2: Incidence of clinical endpoints

Azithromycin %
(95% CI)

Benzathine
Risk dierence %
benzylpenicillin % (95% CI) (95% CI)

Cure at 6 months by yaws stage


Primary
Secondary

909% (788964)
100% (945100)

891% (782949)

18% (137 to 100)

966% (883991)

34% (81 to 12)

929% (830972)

71% (139 to -04)

930% (833972)

14% (99 to 70)

Cure at 6 months by RPR titre at treatment


1 in 32
1 in 64

100% (908100)
944% (866-978)

Cure at 6 months by household exposure


Positive
Negative

100% (851100)
955% (889985)

100% (816100)
920% (850959)

35% (103 to 34)

Data are for the per-protocol analysis. RPR=rapid plasma reagin.

Table 3: Subgroup analysis of the primary endpoint at 6 months

345

Articles

evidence of recurrence during the 6-month followup period.

Discussion
Our ndings show that azithromycin was non-inferior to
benzathine benzylpenicillin for the primary composite
endpoint of serological cure at 6 months and healing of
ulcers. Furthermore, the two treatment groups had
similar rates of cure at 3 month follow-up and in
subgroups dened according to demographic and
biological characteristics. These results add to previous
evidence of the suitability of use of a single dose of a drug
such as azithromycin to treat various infectious diseases
(panel). Rates of serologically dened cure at 6 months
were substantially higher than expected for both
treatments, which validates our non-inferiority hypothesis
for the estimated penicillin cure rate. All participants had
a lower rapid plasmin reagin titre at 6 months than at
3 months, including 150 (67%) patients who
seroconverted. Additionally, we did not identify any
clinical or serological relapse after cure at 6 month followup. The azithromycin regimen did not resolve active
primary lesions in four patients. However, in the three
cases of failure that could be investigated, an
immunoperoxidase stain from a skin biopsy specimen
was negative for spirochaetes; therefore, we could not
conrm the biological treatment failure.
Azithromycin was well tolerated and no major adverse
eects occurred. Of participants who were treated with
azithromycin and interviewed, 8% reported mild to
moderate side-eects that were mainly gastrointestinal.
Only two children vomited within 30 min of oral
azithromycin administration, thus negligible drug
absorption would have occurred. These children were
then re-treated with benzathine benzylpenicillin. The
small number of participants vomiting after administration emphasises the suitability of azithromycin for
mass treatment programmes.
Our study had several limitations. First, diagnostic
criteria for inclusion in the study of primary lesions
did not include a microbiological test (eg, darkeld
microscopy); therefore, the non-healing ulcers could have
had post-treatment infection by other pathogens.
However, darkeld microscopy is rarely used to diagnose
treponemal infections because rapid serological tests are
available. Second, the imprecise denition of serological
cure, which could lead to overestimations in true rates of
cure, is a major issue aecting all research on the
treatment of treponematoses. Because laboratory technicians in this study were unaware of participants
treatment assignments, this drawback should not have
biased the comparison of cure rates between the groups.
Third, 6 months of follow-up might not be sucient to
assess the results after antibiotic treatment for yaws. Four
participants in the benzathine benzylpenicillin group did
not achieve serological cure. This nding could represent
a slower than usual decline in non-treponemal test titres
346

Panel: Research in context


Systematic review
We searched PubMed from Jan 1, 1952, to Aug 1, 2010, with
the terms yaws, Treponema pallidum, penicillin, and
azithromycin. We searched for trials that assessed the
ecacy and safety of single-dose oral azithromycin to treat
infectious diseases in adults and children. We identied two
randomised controlled trials,10,12 which showed the ecacy of
oral azithromycin for the treatment of treponemal disease
(syphilis) in adults. However, we did not identify any study
that explored yaws treatment with azithromycin. We
searched only publications written in English.
Interpretation
Our results provide substantial evidence of the non-inferiority
of a single oral-dose of azithromycin compared with the
standard recommended therapybenzathine
benzylpenicillinfor treatment of yaws. This nding
represents a potentially useful advance in yaws control.

after treatment, rather than a true penicillin-resistant


infection. Finally, because our trial design required
patients to meet certain prespecied criteria, and because
the study was done in one centre, our ndings might not
be generalisable to all children with yaws.
Two important reasons for caution with use of
azithromycin are the sustained success of benzathine
benzylpenicillin treatment for yaws, and the emergence
of azithromycin-resistant T pallidum. Clinical treatment
failure with penicillin has been reported for yaws,20
although, because in-vitro culture for T pallidum has not
been achieved,22 penicillin resistance has not been proven
by microbiological methods. Moreover, in countries such
as Papua New Guinea, cases of reinfection are occurring,
which suggests increased tolerance of some T pallidum
subsp pertenue strains to penicillin treatment.23 Azithromycin resistance in the non-venereal treponemes
has not been investigated, but resistance in the syphilis
treponeme is geographically clusteredeg, more than
95% resistance in Shanghai versus 0% in Madagascar.24,25
In areas where mutations have been found (eg, Seattle,
USA) the frequency of resistance has increased substantially in the past 10 years.26 Use of azithromycin in
the Lihir Island community to treat other infections has
been scarce, which might explain why we did not
encounter a substantial problem with resistance in our
study since there had been very little selective pressure.
Nonetheless, T pallidum subsp pertenue, seems to have
two of the genes encoding 23S ribosomal RNA where the
mutation that confers high-level resistance to macrolides
is located, as does T pallidum subsp pallidum.27 Thus,
close monitoring for potential treatment failure should
be considered in future studies of azithromycin.
With yaws re-emerging, treatment with an eective drug
that can be easily administered on a large scale is the
www.thelancet.com Vol 379 January 28, 2012

Articles

preferred method for treatment, prevention, and,


eventually, elimination worldwide. Elimination programmes need to take account of all epidemiological,
biological, and pharmacological factors, and the practical
considerations of a mass campaign to deliver and
administer drugs in isolated and under-resourced
communities. The potential for treatment of yaws with an
oral, single-dose drug has been explored in this context.
Azithromycin overcomes the major logistical and medical
disadvantages of the present regimen: it avoids the need
for injection equipment and medically trained personnel,
which can be scarce in countries with few health resources;
it prevents all the injection-related risks and side-eects;
and it can be safely administered to individuals with
penicillin allergy (1% in our trial population). Although we
did not formally assess the relative costs related to drug
acquisition and administration, low-cost generic preparations of azithromycin are widely available and the
treatment could therefore be highly cost eective.
Our ndings provide clear evidence that one high dose
of azithromycin is non-inferior to benzathine benzylpenicillin for treatment of yaws. If further studies
conrm our ndings (a similar trial is in progress in
Ghana, West Africa (Kwakye-Maclean C, Ga West
Municipal Heath Directorate, personal communication),
the next step is to attempt elimination and possibly
eradication of the disease in the remaining endemic
countries with mass drug administration programmes
under WHOs leadership.
Contributors
OM conceived and designed the study. OM, RH, AI, MP, and RP
contributed to the recruitment, clinical care, and follow-up of patients.
OM, EdL, and QB analysed and managed the data. DF did all laboratory
tests. OM, RH, and QB wrote the article.
Conicts of interest
We declare that we have no conicts of interest.
Acknowledgments
The study was funded by International SOS (Australasia) Pty and
Newcrest Mining. We thank the eld teams for clinical follow-up and
specimen collection, laboratory sta of Lihir Medical Centre for serology
work, Kingsley Asiedu and Joan-Ramon Laporte for reviewing the
manuscript and providing valuable comments, and Maria de Ros Villar
for image editing.
References
1
Meheus A. Non-venereal treponematoses. Medicine 2005; 33: 8284.
2
Guthe T, Willcox RR. Treponematoses: a world problem.
WHO Chron 1954; 8: 37113.
3
Antal GM, Causse G. The control of endemic treponematoses.
Rev Infect Dis 1985; 7 (suppl 2): S22026.
4
WHO. Resolutions and decisions: WHA31.58 control of endemic
treponematoses. May 24, 1978. www.who.int/neglected_diseases/
mediacentre/WHA_31.58_Eng.pdf (accessed Nov 21, 2011).
5
WHO. Recent news from WHO. Bull World Health Organ 2007;
85: 74.
6
WHO Scientic Group. Treponemal infections. Technical Report
Series No 674. Geneva: World Health Organization, 1982.

www.thelancet.com Vol 379 January 28, 2012

7
8
9

10

11

12

13

14

15

16

17

18
19
20

21

22
23
24

25

26

27

Antal GM, Lukehart SA, Meheus AZ. The endemic treponematoses.


Microbes Infect 2002; 4: 8394.
WHO. Yaws: a forgotten disease. January, 2007. http://www.who.int/
mediacentre/factsheets/fs316/en/ (accessed Nov 21, 2011).
Meheus AZ, Narain JP, Asiedu KB. Endemic treponematoses.
In: Cohen J, Powderly SM, Opal WG, eds. Infectious diseases,
3rd edn. London: Elsevier (in press).
Riedner G, Rusizoka M, Todd J, et al. Single-dose azithromycin
versus penicillin G benzathine for the treatment of early syphilis.
N Engl J Med 2005; 353: 123644.
Scolnik D, Aronson L, Lovinsky R, et al. Ecacy of a targeted, oral
penicillin-based yaws control program among children living in
rural South America. Clin Infect Dis 2003; 36: 123238.
Martin DH, Mroczkowski TF, Dalu ZA, et al. A controlled trial of
a single dose of azithromycin for the treatment of chlamydial
urethritis and cervicitis. The Azithromycin for Chlamydial
Infections Study Group. N Engl J Med 1992; 327: 92125.
Handseld HH, Dalu ZA, Martin DH, Douglas JM Jr, McCarty JM,
Schlossberg D. Multicenter trial of single-dose azithromycin vs
ceftriaxone in the treatment of uncomplicated gonorrhea.
Sex Transm Dis 1994; 21: 10711.
Martin DH, Sargent SJ, Wendel GD Jr, McCormack WM, Spier NA,
Johnson RB. Comparison of azithromycin and ceftriaxone for the
treatment of chancroid. Clin Infect Dis 1995; 21: 40914.
Hook EW 3rd, Behets F, Van Damme K, et al. A phase III
equivalence trial of azithromycin versus benzathine penicillin
for treatment of early syphilis. J Infect Dis 2010; 201: 172935.
US Food and Drug Administration. Brieng document for
zithromax accelerated dosing; treatment of acute otitis media.
Nov 7, 2001. www.fda.gov/downloads/advisorycommittees/
committeesmeetingmaterials/drugs/anti-infectivedrugsadvisory
committee/ucm209921.pdf (accessed Nov 21, 2011).
Dunne MW, Khurana C, Mohs AA, et al Ecacy of single-dose
azithromycin in treatment of acute otitis media in children after
a baseline tympanocentesis. Antimicrob Agents Chemother 2003;
47: 266365.
Talwat E. Papua New Guinea yaws problems assessed.
Southeast Asian J Trop Med Public Health 1986; 17 (suppl 4): 5965.
Manning LA, Ogle GD. Yaws in the periurban settlements of
Port Moresby, Papua New Guinea. P N G Med J 2002; 45: 20612.
Backhouse JL, Hudson BJ, Hamilton PA, Nestero SI. Failure of
penicillin treatment of yaws on Karkar Island, Papua New Guinea.
Am J Trop Med Hyg 1998; 59: 38892.
Altman DG, Gore SM, Gardner MJ, Pocock SJ. Statistical guidelines
for contributors to medical journals. In: Altman DG, Machin D,
Bryant TN, Gardner MJ, eds. Statistics with condence: condence
intervals and statistical guidelines, 2nd edn. London: BMJ Books,
2000: 17190.
Stamm LV. Global challenge of antibiotic-resistant Treponema
pallidum. Antimicrob Agents Chemother 2010; 54: 58389.
Mitj O, Hays R, Ipai A, et al. Outcome predictors in
treatment of yaws. Emerg Infect Dis 2011; 17: 108385.
Martin IE, Gu W, Yang Y, Tsang RS. Macrolide resistance and
molecular types of Treponema pallidum causing primary syphilis
in Shanghai, China. Clin Infect Dis 2009; 49: 51521.
Van Damme K, Behets F, Ravelomanana N, et al. Evaluation of
azithromycin resistance in Treponema pallidum specimens from
Madagascar. Sex Transm Dis 2009; 36: 77576.
Lukehart SA, Godornes C, Molini BJ, et al. Macrolide resistance in
Treponema pallidum in the United States and Ireland. N Engl J Med
2004; 351: 15458.
Stamm LV, Bergen HL. A point mutation associated with bacterial
macrolide resistance is present in both 23s rRNA genes of an
erythromycin-resistant Treponema pallidum clinical isolate.
Antimicrob Agents Chemother 2000; 44: 80607.

347

Seminar

The cryoglobulinaemias
Manuel Ramos-Casals, John H Stone, Maria C Cid, Xavier Bosch
Lancet 2012; 379: 34860
Published Online
August 24, 2011
DOI:10.1016/S01406736(11)60242-0
Josep Font Laboratory of
Autoimmune Diseases
(M Ramos-Casals MD) and
Vasculitis Research Unit,
Department of Autoimmune
Diseases (M C Cid MD), and
Department of Internal
Medicine (X Bosch MD),
Institut Clnic de Medicina I
Dermatologia, Hospital Clnic,
Institut dInvestigacions
Biomdiques August Pi i
Sunyer, University of
Barcelona, Barcelona, Spain;
and Rheumatology Unit,
Massachusetts General
Hospital, Boston, MA, USA
(J H Stone MD)
Correspondence to:
Dr Xavier Bosch, Department
of Internal Medicine,
Hospital Clinic, Villarroel 170,
Barcelona 08036, Spain
xavbosch@clinic.ub.es

348

Cryoglobulins are immunoglobulins that precipitate in vitro at temperatures less than 37C and produce organ
damage through two main pathways: vascular sludging (hyperviscosity syndrome, mainly in type I
cryoglobulinaemia) and immune-mediated mechanisms (principally vasculitis, in mixed cryoglobulinaemia).
Cryoglobulinaemia is associated with many illnesses, which can be broadly grouped into infections, autoimmune
disorders, and malignancies; the most common cause is infection with hepatitis C virus. Mixed cryoglobulinaemic
syndrome is diagnosed when a patient has typical organ involvement (mainly skin, kidney, or peripheral nerve)
and circulating cryoglobulins. Cutaneous purpura is the most common manifestation of cryoglobulinaemic
vasculitis. The most frequently aected internal organs are the peripheral nerves, kidneys, and joints. The course
varies widely and prognosis is inuenced by both cryoglobulinaemic damage to vital organs and by comorbidities
associated with underlying diseases. More than 90% of cases of cryoglobulinaemia have a known underlying
cause; therefore treatment is focused on the cause of the disorder rather than merely symptomatic relief. Studies
suggest that both combined or sequential antiviral therapies and targeted biological treatments might be more
eective than monotherapy.

Denition and classication

Causes

Cryoglobulins are immunoglobulins that precipitate in


vitro at temperatures less than 37C and redissolve after
rewarming. Cryoglobulinaemia refers to the presence of
cryoglobulins in serum. However, the terms cryoglobulinaemic disease or cryoglobulinaemic vasculitis
are used to describe patients with symptoms related to
the presence of cryoglobulins. Many patients with cryoglobulinaemia remain asymptomatic.1
The pathological nature of cryoglobulins was
postulated in 1933, in a patient with multiple myeloma.2
The term cryoglobulin was introduced in 1947.3 Cryoglobulinaemic disease was described in 1966 by Meltzer
and colleagues,4 who reported 29 patients with cryoglobulins and a common clinical presentation (purpura,
arthralgia, and weakness), accompanied by organ
dysfunction and raised serum concentrations of
rheumatoid factor.
The composition of cryoglobulins is heterogeneous.
The most commonly used classication dates from
19745 (gure 1) but remains useful because of its
consistency for the clinical manifestations of the
three cryoglobulin subsets. Three basic types are
recognised according to the clonality and type of
immunoglobulins. Type I consists of monoclonal
immunoglobulin, generally either IgM or IgG. Type II
cryoglobulins are a mixture of monoclonal IgM and
polyclonal IgG. The IgM component of type II
cryoglobulins has rheumatoid factor activity (ie, these
immunoglobulins bind to the Fc portion of IgG). Type
III cryoglobulins are a mixture of polyclonal IgM and
IgG. Types II and III are referred to as mixed cryoglobulinaemias because they consist of both IgG and
IgM components. Some investigators have suggested
that type III cryoglobulinaemia is a transitional state
that evolves to a type II cryoglobulin prole (ie,
from a polyclonal to an oligomonoclonal B-cell population).6 Other forms of cryoglobulinaemia have been
reported.7

Infections
The discovery of the hepatitis C virus (HCV) in 19898
radically changed the focus of research from essential to
HCV-related cryoglobulinaemia.9,10 Ferri and colleagues11
conrmed the detection of circulating HCV-RNA in
nearly 90% of Italian patients with mixed
cryoglobulinaemia, although later studies found wide
geographical variations.1215 HCV is predominantly
associated with type II cryoglobulinaemia. The hepatitis
B virus is reported to be associated with mixed
cryoglobulinaemia.7,16 In patients infected with HIV, the
percentage with cryoglobulinaemia ranges from 7% to
17% but rises to between 35% and 64% in those coinfected with HCV.1721 Highly-active antiretroviral therapy
lowers the frequency of cryoglobulinaemia in HIV.22 Case
reports have associated cryoglobulinaemic syndrome
with a wide range of other infectious agents (table).

Autoimmune diseases
Patients with systemic autoimmune diseases can present with complications of mixed cryoglobulinaemia.

Search strategy and selection criteria


We searched Medline and Embase (from January, 1990, to
January, 2011). We used the search term cryoglobulinemia
in combination with epidemiology, diagnosis, virus,
cancer, autoimmune, pathogenesis, hyperviscosity,
vasculitis, prognosis, and therapy. We focused on
publications from the past 10 years but also included
commonly referenced and highly regarded older
publications. We searched the reference lists of articles
identied by this search strategy and selected those
publications we judged relevant. Review articles and book
chapters are cited to provide readers with more detail than
can be included in our Seminar. The date of the last search
was Jan 25, 2011.

www.thelancet.com Vol 379 January 28, 2012

Seminar

Type I

Type II

Type III

Monoclonal lg

Monoclonal lgM+polyclonal lgG

Polyclonal lgM+polyclonal lgG

Figure 1: Classication of cryoglobulinaemia5

Most frequent
causes

Less frequent
causes

Infrequent
causes

Infections

Hepatitis C
virus

HIV; Hepatitis B virus

Streptococcus spp; Brucella spp; Coxiella spp; Klebsiella spp; Leishmania spp; Chlamydia spp;
Mycobacterium tuberculosis; leprosy; hepatitis A virus; cytomegalovirus; parvovirus B-19;
chikungunya virus; Epstein-Barr virus; hantavirus; plasmodium; amoebaiasis; toxoplasmosis

Autoimmune
diseases

Sjgrens
syndrome

Systemic lupus
erythematosus;
Rheumatoid arthritis

Systemic sclerosis; antiphospholipid syndrome; inammatory myopathies; adult-onset Stills


disease; polyarteritis nodosa; giant-cell arteritis; Takayasus arteritis; ANCA-associated vasculitis;
autoimmune hepatitis

Cancer

B-cell
lymphoma

Multiple myeloma

Hodgkins lymphoma; chronic lymphocytic leukaemia; chronic myeloid leukaemia; myelodysplasia;


hepatocellular carcinoma; papillary thyroid cancer; lung adenocarcinoma; renal cell carcinoma;
nasopharyngeal carcinoma

Other causes

Alcoholic cirrhosis

Co-trimoxazole;* interferon alfa;* cocaine;* intravenous radiographic contrast;* influenza


vaccination; hepatitis B vaccination; intravesical BCG; moyamoya disease; endocarditis;
chilblains

ANCA=antineutrophil cytoplasmic antibodies. *Associated with cryoglobulinaemic exacerbation.

Table: Main causes associated with cryoglobulinaemia since 199023

In primary Sjgrens syndrome, cryoglobulinaemia is


associated with extraglandular involvement, an
enhanced risk of B-cell lymphoma, and poor survival.2426
The prevalence of cryoglobulinaemia is ve times higher
in patients with both Sjgrens syndrome and HCV
infection compared with those not infected with HCV.27
Cryoglobulins are detected in nearly 10% of patients
with systemic lupus erythematosus and rheumatoid
arthritis, but cryocrit values are generally lower
compared with those in patients with Sjgrens
syndrome, and the clinical manifestations of cryoglobulinaemic vasculitis are much less common.23,28,29
Cryoglobulins can be detected in a wide range of other
autoimmune diseases (table).
www.thelancet.com Vol 379 January 28, 2012

Malignancy
B-cell lymphoproliferative diseases are the major cause
of cryoglobulinaemia associated with malignancy.
Type I cryoglobulinaemia is reported predominantly in
patients with Waldenstrms macroglobulinaemia,
multiple myeloma, or chronic lymphocytic leukaemia.30
Mixed cryoglobulinaemias occur mainly in B-cell
lymphomas.31 Cryoglobulins can be detected in patients
with solid cancers.32

Essential cryoglobulinaemia
Nearly 10% of cases of mixed cryoglobulinaemia are
regarded as idiopathic or essential,7 a percentage that
rises to 25% in HCV-negative patients.32 The possibility
349

Seminar

19%
USA

0%

32%

44%

37%

China

Japan

Sweden
India

patients.12 Conversely, between 12% and 56% of HCVinfected patients have circulating cryoglobulins, again
with the highest frequency in Mediterranean patients
(gure 2). The reasons for this geographic variation are
not clear. Genetic studies have identied few consistent
risk factors. The risk of mixed cryoglobulinaemia has a
well-described relation with the duration of HCV
infection, with an annual incidence of cryoglobulinaemia
of 3%.15 Patients infected with HCV who have cryoglobulinaemia have a mean disease duration roughly
twice that of patients without cryoglobulinaemia,35 with
an odds ratio for cirrhosis of 487.36

51%

Pathophysiology
Lithuania

15%
19%

Ireland

Generation of cryoglobulins

55%

UK

28%
Poland
Germany
56%
France
37%

43%

55%

Spain

Italy

Bulgaria
17%

46%

Turkey
Greece
12%
29%

Israel

Egypt

16%
Iran
30%
Kuwait

Figure 2: Prevalence of cryoglobulinaemia in patients with chronic HCV infection


>40% in red, 2040% in green, <20% in blue.

of occult HCV infection should be investigated in


patients with cryoglobulinaemia who present with
persistently abnormal liver function tests of unknown
cause.33 Monoclonal gammopathies of unknown signicance can have cryoglobulin activity and might
account for an undetermined percentage of essential
cryoglobulinaemias.

Epidemiology
The prevalence of cryoglobulinaemia remains unknown.34
Reasons include the careful laboratory technique needed
to isolate and identify cryoglobulins and the absence of a
standard clinical assessment of patients with possible
cryoglobulinaemia. Even so, because HCV infects more
than 170 million individuals worldwide, the number of
patients at risk for the complications of mixed
cryoglobulinaemia is substantial.
The prevalence of HCV infection in patients with
mixed cryoglobulinaemia ranges from 30% to nearly
100%, with the highest prevalence in Mediterranean
350

Cryoglobulins are generated by the clonal expansion of


B cells, in the context of either lymphoproliferative
disorders or persistent immune stimulation triggered by
chronic infections or autoimmune diseases.5,7,37 Types I
and II cryoglobulinaemias result from the monoclonal
expansion of a clone that can be overtly malignant
(multiple myeloma), smouldering (Waldenstrms
macroglobulinaemia, plasmacytoid lymphoma), or
indolent (as in monoclonal gammopathy of unknown
signicance). By contrast, B-cell expansion is polyclonal
in type III cryoglobulinaemia.
HCV infection is the key model for studying
aetiopathogenesis.10,11 In this context, HCV lymphotropism
represents the rst step of B-cell clonal expansion,
irrespective of the presence or absence of cryoglobulinaemic disease.38 An HCV envelope protein, E2,
interacts with the major extracellular loop of
tetraspanin CD81, a signalling molecule expressed by
both hepatocytes and by B and T lymphocytes.39 This
interaction is believed to trigger chronic B-cell
stimulation.40 B-cell clones can be found within the
peripheral blood, bone marrow, and liver of patients with
HCV infection, particularly those with type II cryoglobulinaemia.4143 Bcell clones produce monoclonal
IgM that has a cross-reacting idiotype called WA, which
binds immunoglobulins directed to anti-HCV core
protein.44 Nearly 10% of patients who are asymptomatic
and HCV positive have circulating B cells that are positive
for WA.45 In-vivo and in-vitro cryoprecipitates from
patients with HCV-related cryoglobulinaemia contain
viral core proteins and RNA,4648 suggesting that cryoglobulin formation results from the host immune
response against chronic HCV infection.

Precipitation of cryoglobulins
Reversible precipitation on exposure to low temperatures
permits laboratory detection of cryoglobulins, but the
biochemical mechanisms of this process are not fully
understood.49 However, temperature is probably not the
only factor aecting the solubility of cryoglobulins. Some
cryoglobulins precipitate in vivo at temperatures well
above the 4C at which they are precipitated in the
www.thelancet.com Vol 379 January 28, 2012

Seminar

laboratory. Thus cold exposure could be a contributing


factor to clinical manifestations of cryoglobulinaemia in
the distal extremities. Internal organ involvement is more
dicult to explain on the basis of temperature changes
alone because of the tight regulation of core body
temperature. Protein solubility can depend on a range of
factors, including primary structure and steric conformation which, in turn, depend on temperature, pH,
and ionic strength.49 Scarcity of tyrosine residues, relative
abundance of hydrophobic aminoacids, and reduced
concentration of galactose and sialic acid in the glycosylated
portion of the molecule can increase precipitation.49,50 In
type II mixed cryoglobulinaemia, the formation of large,
complement bound, IgMIgG complexes is a major factor
inuencing cryoprecipitation.51

Pathogenesis of tissue injury


Two major mechanisms are at play to varying degrees
across the dierent types of cryoglobulinaemia:
cryoglobulin precipitation in the microcirculation, and
immune-complex-mediated inammation of blood
vessels. Vascular occlusion is more frequent in type I
cryoglobulinaemia, which is usually accompanied by
high cryoglobulin concentrations, and can be associated
with hyperviscosity syndrome and cold-induced acral
necrosis. Immune-complexmediated vasculitis is
more frequent in mixed cryoglobulinaemias,
particularly type II, in which the monoclonal IgM
component generates large immune complexes with
IgG and complement fractions, particularly C1q.
C1q can bind to receptors on endothelial cells,
facilitating immune complex deposition and subsequent vascular inammation.52

Clinical manifestations
The percentage of patients with circulating cryoglobulins
who develop symptoms varies from 2% to 50%.7 The
most common presentation, the triad of purpura,
arthralgia, and weakness, is reported in 80% of patients
at disease onset.7,16 The development of cryoglobulinaemic
symptoms is aected by age, underlying illness (such as
HCV infection) and the characteristics of the cryoglobulins
(type II subclass, high serum concentrations).7,53

Hyperviscosity syndrome
Hyperviscosity syndrome develops mainly in patients
with type I cryoglobulinaemia associated with haematological neoplasia,54 and is very uncommon in patients
with mixed cryoglobulinaemia (<3%).16,55 The key
symptoms are neurological (headache, confusion), ocular
(blurry vision, visual loss), and rhino-otological (epistaxis,
hearing loss). Massive intratubular cryoprecipitation
leading to rapidly progressive renal failure has been
reported. The physical examination should include
funduscopy to exclude hyperviscosity-related retinal
changes, especially haemorrhages. In patients in whom
hyperviscosity syndrome is suspected, it is helpful to
www.thelancet.com Vol 379 January 28, 2012

Figure 3: Cutaneous involvement in cryoglobulinaemia


(A) Purpura in legs, (B) atypical purpura, (C) cutaneous ulcers, (D) digital necrosis.

measure serum viscosity. Patients usually become


symptomatic at viscosity measurements that exceed
40 centipoise,56 but some patients are symptomatic with
lower viscosities.57 Symptomatic hyperviscosity requires
urgent treatment (eg, plasma exchange).

Cryoglobulinaemic vasculitis
Flares of cryoglobulinaemic vasculitis are often, but not
always, accompanied by general symptoms such as fever,
weakness, myalgia, and arthralgia. These symptoms,
particularly when combined with purpura, strongly
suggest vasculitis. Fevers of unknown origin can occur.
Articular involvement consists mainly of joint pain in the
hands, knees, and wrists, without clinical signs of
inammation (4471%).7,16,55,58 Arthritis is reported in
fewer than 10% of patients.59 Radiographs show no
evidence of bone erosions and anticitrullinated antibodies
are negative, by contrast with what is seen in rheumatoid
arthritis.59 Weakness and fatigue are reported by more
than 50% of patients. Fibromyalgia and endocrine
processes, such as hypothyroidism and diabetes, should
be investigated because of their higher frequency in
cryoglobulinaemic patients.6062
Cutaneous purpura is probably the most characteristic manifestation of cryoglobulinaemic vasculitis
(5482%).7,16,55,58,63 The typical presentation consists of
many small petechial lesions in the legs (gure 3A);
bullous or vesicular lesions are uncommon. Purpura
can extend to the abdominal region and, less frequently,
to the upper limbs and thorax (gure 3B). Involvement
at other sites is rare. Isolated purpura has a good
prognosis, and patients usually recover spontaneously,
often in less than 1 week, leaving discoloured skin
351

Seminar

concentration (>15 mg/dL) at diagnosis.7,68 In a series


of 105 patients with cryoglobulinaemic glomerulonephritis,69 15 (14%) evolved to chronic renal failure
after a mean follow-up of 6 years. Recurrence of
cryoglobulinaemia in the renal allograft has been
reported in patients receiving a kidney transplant.70,71
Between 17% and 60% of cryoglobulinaemic patients
present with peripheral neuropathy,7,16,53,55 which can be
the rst sign of cryoglobulinaemia.72 The main symptoms
are paresthesiae, with painful or burning sensations, or
both, in the lower limbs, which are often worse at night.
These generally precede motor involvement (gure 4B).
Electrodiagnostic studies, which show that polyneuropathy is more common than mononeuritis multiplex,7,58,73
often disclose subclinical peripheral nerve dysfunction.74
In a few patients, neuropathy can present as rapidlyprogressive sensory-motor involvement with severe
functional impairment.75,76

Less-common disease manifestations

Figure 4: Systemic cryoglobulinaemic vasculitis


(A) Membranoproliferative glomerulonephritis associated with HCV-related type II cryoglobulinaemia.
A glomerulus shows proliferative changes and subendothelial deposits within capillary loops (pink). These lesions
resemble pseudothrombi. (B) Multineuritis (radial and cubital paralysis), (C) intestinal ischaemia (diuse oedema
of intestinal wall), (D) pulmonary haemorrhage.

patches formed of haemosiderin deposits, although


recurrent episodes are frequent. Patients should be
questioned about exercise and depilation, which often
trigger episodes of purpura. Other factors, including
chronic venous insuciency, protracted standing, and
muggy weather should be considered.
The prognosis can be worsened by cutaneous ulcers
caused by coalescence of vasculitic lesions (usually around
the malleoli)7,16,55,63 or ischaemia in the distal regions
(hands, feet, lips, ears, and nose) (gure 3C and 3D).
Livedo reticularis and cutaneous ulcers usually indicate
involvement of medium-sized vessels (ie, small arteries or
arterioles). Cutaneous ulcers and digital necrosis signify a
high risk of infection, sepsis, and death.
About 20% of patients with cryoglobulinaemia present
with nephropathy at diagnosis and 30% have renal
complications during the disease course.1,16,51,55,58,6368
Renal features are indolent in nearly half the patients,
with proteinuria, microscopic haematuria, red-bloodcell casts, and varying degrees of renal failure. Nephrotic
(21%) or nephritic (14%) syndromes are less frequent.67
More than 70% of patients present with hypertension
and 4060% have a raised serum creatinine
352

26% of patients have gastrointestinal involvement.7,55,77


Intestinal ischaemia should be suspected in patients who
present with acute abdominal pain and general malaise.
Fever and bloody stools are reported by a third of patients
(gure 4C). Some patients present with intestinal
perforation and shock. Cryoglobulinaemic vasculitis
involving the gastrointestinal tract can mimic cholecystitis
pancreatitis.77 Cryoglobulinaemic vasculitis has been
identied within the gallbladder at cholecystectomy.
Pulmonary involvement occurs in less than 5% of
patients.7,16,78 In patients presenting with mild-to-moderate
eort dyspnoea and dry cough, interstitial lung brosis
should be suspected; bronchoalveolar lavage shows a
predominance of macrophages, with slightly increased
neutrophils or lymphocytes.79 A few patients manifest
with acute alveolar haemorrhage (haemoptysis, respiratory
failure, and diuse pulmonary inltrates)80 (gure 4D).
Pleural eusions are rare.58 CNS involvement, reported in
up to 6% of patients, is often dicult to conrm.7,58 The
most common clinical presentation is stroke (motor or
sensory decits, aphasia, or dysarthria), although diuse
cerebral involvement manifesting as encephalopathy has
also been described.80 MRI shows cerebral ischaemia in
half of these patients and, less often, haemorrhagic
lesions. A third of patients have diuse small lesions that
raise the spectre of cerebral vasculitis.80 Isolated cases of
transverse myelitis (often recurrent) and ischaemic
spinal-cord involvement have been reported.81
Fewer than ten cases of myocardial vasculitis have been
reported in cryoglobulinaemic patients. These patients can
have myocardial infarctions in the absence of cardiovascular
risk factors.80 There are reports of pericarditis or congestive
heart failure complicating cryoglobulinaemia.80,82
A small proportion of cryoglobulinaemic patients
present with multiorgan disease (skin, kidneys, lungs,
CNS, and gastrointestinal tract) involving the small
medium-sized arteries, capillaries, and venules.1
www.thelancet.com Vol 379 January 28, 2012

Seminar

Diagnosis
Cryoglobulin testing
There are no standardised or validated diagnostic or
classication criteria for cryoglobulinaemic vasculitis.83
Diagnosis is based on clinical, laboratory, and histopathological data. For most patients, cryoglobulinaemic
disease is diagnosed by the presence of typical organ
involvement (mainly skin, kidney or peripheral nerve)
and circulating cryoglobulins.
The diagnosis of cryoglobulinaemia requires demonstration of the presence of cryoglobulins in serum
(panel 1). Appropriate sample collection and handling is
crucial.84 Blood should be collected in prewarmed syringes
and tubes, transported, clotted, and centrifuged at
3740C, ensuring that the temperature never falls
below 37C. The serum should then be stored at 4C for
up to 7 days. Precipitation of type I cryoglobulins usually
occurs within hours. By contrast, mixed cryoglobulins,
particularly type III, can need days to precipitate.85
Expert laboratory interpretation that considers the
patient in the appropriate clinical context is essential.
Some healthy individuals have low concentrations of
cryoglobulins (<006 g/L),86,87 and mixed polyclonal
cryoglobulins often occur transiently during infection.49
On the other hand, a negative test for cryoglobulins does
not exclude cryoglobulinaemia because of the possibility
of false-negative results caused by improper sample
collection or inconsistent laboratory techniques.49
Moreover, cryoglobulin concentrations can uctuate,
depending on their in-vivo precipitation in target vessels.
Cryoglobulin should be assayed serially when there is a
high degree of suspicion of disease.49,85
Cryoglobulin concentration can be assayed indirectly
by measurement of the total protein concentration within
the cryoprecipitate and by estimation of the cryocrit.49,75
Cryoglobulin concentration is usually more than 5 g/L in
type I cryoglobulinaemia, but generally lower in types II
and III.49 Quantication of the cryocrit is important
because the amount of serum cryoprotein can correlate
with the severity of symptoms and is useful in monitoring
the response to treatment, particularly in patients with
hyperviscosity syndromes. Immunoxation of the
redissolved cryoprecipitate, if feasible, allows identication of the type of cryoglobulin. If the cryoprecipitate
is too small, the type of cryoglobulin can be estimated
indirectly by serum immunoxation. The monoclonal
component detected usually corresponds to the existing
monoclonal cryoglobulin.

Laboratory tests
Laboratory tests are a key instrument for evaluating
functional results of visceral involvement and causal
factors. Biochemical tests for renal and liver involvement
are mandatory. Low complement levels (particularly C4)
and raised titres of serum rheumatoid factor are usually
observed in mixed cryoglobulinaemias49,51,85 and can
correlate with some clinical symptoms.88
www.thelancet.com Vol 379 January 28, 2012

Panel 1: Clinical, laboratory, and histopathological red


ags advising cryoglobulin testing
Regard cryoglobulinaemia as highly probable when at least
two features of dierent subsets are present
Clinical ndings
Skin purpura in adults
Cutaneous necrotic ulcers
Glomerulonephritis
Peripheral neuropathy
Non-erosive arthritis
Acral ischaemia
Cold-induced acrocyanosis
Raynauds phenomenon
Laboratory abnormalities
Monoclonal gammopathy, particularly of IgM isotype or
with hyperviscosity
Unexplained low concentrations of complement
(especially C4)
Unexplained high titres of rheumatoid factor
Pseudothrombocytosis
Formation of erythrocyte rouleaux
Histopathological ndings
Leukocytoclastic vasculitis in adults
Membranoproliferative glomerulonephritis
Hyaline thrombi in capillaries in context of
glomerulonephritis or small-vessel vasculitis
Endoneural vasculitis
Unclassied systemic necrotising vasculitis involving
smallmedium-sized vessels

Cryoglobulin detection demands thorough evaluation


to identify potential underlying diseases. HCV testing is
mandatory (antibodies and serum HCV-RNA detection)
in patients with mixed cryoglobulinaemia. Testing for
other viruses (hepatitis B virus, HIV) and autoimmune
diseases (antinuclear, anti-DNA, anti-Ro/La, anticitrullinated antibodies) is recommended, even in
patients known to have HCV.49

Histopathological ndings
Histopathological examination is usually appropriate to
conrm the diagnosis in an aected organ. Precipitated
cryoglobulins appear in vivo as hyaline thrombi that
occlude small blood vessels, including the glomerular
tuft and endoneural microvessels. Hyaline thrombi are
especially likely when the monoclonal component in
type I and II cryoglobulinaemias is abundant.67,89
Vasculitis, the typical pathological lesion of mixed
cryoglobulinaemia, consists of mixed inammatory
inltrates involving small and, less often, medium sized
vessels. Fibrinoid necrosis can occur. The most commonly
aected organs are the skin, kidneys, and peripheral
nerves.51 Skin biopsies of purpuric lesions reveal
353

Seminar

Stratified treatment of HCV-related cryoglobulinaemic syndrome according to disease severity

Mild/moderate disease

Severe disease

Induction phase

Antiviral therapy
+/
Corticosteroids

Maintenance
phase

Purpura, articular,
general features
Mild neuropathy
GN without
renal failure

Antiviral therapy

Cutaneous ulcers,
ischaemia
Severe neuropathy
GN with renal failure/
nephrotic syndrome
GI involvement

Refractory

Rituximab
+
Corticosteroids

Antiviral therapy

Life-threatening

Rapidly progressive GN
CNS involvement
Intestinal ischaemia
Alveolar haemorrhage

Plasma exchanges
+
Corticosteroids pulses
Refractory
+
Cyclophosphamide
or
Rituximab

Antiviral therapy

Figure 5: Proposed therapeutic algorithm for patients with HCV-related cryoglobulinaemic syndrome
according to disease severity
GN=glomerulonephritis, GI=gastrointestinal. Limitations: there is little evidence for treatment of
cryoglobulinaemic features (available therapeutic data relies mainly on uncontrolled studies and case series). There
is no standardised classication of disease severity. Recommended strategies are based on personal experience and
expert opinions.1,94,98,105,106 Corticosteroids should be used at minimum dose and period necessary; short courses and
low doses (<30 mg/day) can be used. There are no available data on immunosuppressive maintenance therapy;
similar regimens and doses to those used in other autoimmune diseases (lupus, vasculitis) are suggested. Best dose
and therapeutic regimen of rituximab is unknown. Timing between rituximab and antiviral therapy administration
(simultaneous, or 14 weeks after starting rituximab, or after 46 months) is not clear.

leukocytoclastic vasculitis of the capillaries and


postcapillary venules. In cryoglobulinaemic glomerulonephritis, proliferative (highly cellular) glomerular
inltrates are recorded. Eosinophilic and periodic acidSchi test-positive subendothelial cryoprecipitate
material might be detected.
Renal biopsy detects type I membranoproliferative
glomerulonephritis in more than 70% of patients16,58,67,68
(gure 4A). More specic histopathological ndings are
hyaline intraluminal thrombi that contain IgM, IgG,
and C3 as endomembranous deposits and glomerular
inltration by monocytes. Glomerular crescents (1020%),
renal necrotising vasculitis (530%), and interstitial
inammation arise less often.67,68 Renal biopsy discloses
other histopathological patterns (focal, mesangioproliferative, or membranous glomerulonephritis) in
smaller percentages of patients.7,67 In these patients, the
possible role of cryoglobulinaemia-associated processes
(systemic lupus erythematosus, lymphoma, chronic viral
diseases),67,89,90 renal involvement associated with
cardiovascular risk factors (hypertension, diabetes), and
IgA nephropathy90,91 should be investigated. Renal biopsy
is recommended to conrm the diagnosis and assess the
extent of involvement.
354

Cryoglobulinaemic neuropathy is characterised by


vasculitis of the perineural and endoneural vessels, with
various degrees of axonal degeneration and demyelination. Endoneural microvessel involvement is more
common than in other systemic vasculitides. Endoneural
capillaries are thickened and inamed; extravasation of
erythrocytes and macrophages (endoneural purpura)
might be observed. Indirect immunouorescence
identies deposits of immunoglobulins (of the same
type as the serum cryoglobulins) and complement in
skin, kidney, and nerve biopsies.51,64,67,89,92,93

Outcome
The evolution of cryoglobulinaemic disease varies
widely. Roughly half of patients have chronic disease
with no involvement of vital organs. A third of patients
have moderate-to-severe disease, with chronic renal
failure or cirrhosis, and nearly 15% present with sudden
life-threatening disease.16 Patients with cryoglobulinaemic disease have a worse 10-year survival rate
compared with the rate in the general population.16,55
Risk factors for poor outcomes include male sex, age
more than 60 years, glomerulonephritis, gastrointestinal
or pulmonary involvement, chronic HCV infection and
type II cryoglobulinaemia.16,55,63,94,95
Prognosis is inuenced heavily by both cryoglobulinaemic damage to vital organs and by underlying diseases
and comorbidities. Chronic HCV infection resulting in
liver cirrhosis and cancer and haematological neoplasia are
associated with a poor prognosis. Therapeutic interventions
(glucocorticoids, immunosuppressive agents, plasma
exchange) raise the risk of infection, especially in patients
with chronic renal or liver failure. Patients with a
combination of severe processes (chronic organ failure,
infection, and vasculitis or neoplasia) are often seen and
need an integrated, multidisciplinary approach.96,97

Life-threatening cryoglobulinaemia
Glomerulonephritis results in acute renal failure in 10%
of patients67 or can evolve progressively to chronic renal
failure.16,67,94 10-year survival rates of glomerulonephritis
range between 33% and 49%,16,94 although a study reported
in 2007 recorded nearly 80% survival after 10 years67
resulting from improved therapeutic management. The
outlook is poor in men and patients with HCV infection,
high cryocrit values, low C3 values, and raised creatinine
at diagnosis.6769,94 Mortality rates associated with intestinal
ischaemia and alveolar haemorrhage are very high
(>80%),94 although a 2010 study77 suggested improved
prognosis for gastrointestinal involvement.

Risk of neoplasia
Mixed cryoglobulinaemic syndrome is traditionally
regarded as a crossroads between autoimmune disease
and cancer.98,99 Expansion of the peripheral B-lymphocyte
pool and lymphoid inltrates within the liver and bone
marrow, characteristic of overt B-cell lymphoproliferative
www.thelancet.com Vol 379 January 28, 2012

Seminar

processes, are often seen in patients with mixed


cryoglobulinaemia.93 B-cell lymphoma, the main
neoplastic complication in patients with mixed
cryoglobulinaemia,16,100 has been reported in 522% of
patients with mixed cryoglobulinaemia.7,16,32,101,102 Hypogammaglobulinaemia can be a marker for impending
lymphomagenesis.102 Associated HCV infection confers a
2030% increased risk of B-cell lymphoma.103 B-cell
lymphomas usually occur within 10 years of diagnosis of
cryoglobulinaemia. The most common types are diuse,
large, nodal marginal zone, and lymphoplasmacytoid
B-cell lymphomas.31 Treatment usually includes standard
chemotherapy regimens and rituximab.32,101
Liver cancer, the second most frequently diagnosed
neoplasia in cryoglobulinaemic patients, is attributable
to the close association with HCV.7,16,104 Hepatocellular
carcinoma occurs about a third as often as lymphoma in
patients infected with HCV.

Treatment

Panel 2: Key therapeutic points in patients with


cryoglobulinaemia
Treat underlying cause of cryoglobulinaemia when possible
Careful individual workup to assess how many organs are
aected and severity of involvement is essential when
planning therapeutic approaches
Carefully assess intensity and duration of conventional
immunosuppressive therapies, especially in patients
infected with HCV
Regard combination of pegylated interferon alfa and
ribavirin as therapeutic cornerstone of HCV-related
cryoglobulinaemic vasculitis
Plasma exchange is valuable therapeutic option in
life-threatening involvements, including
hyperviscosity syndrome
O-label use of B-cell-depleting therapies should be
reserved for severe refractory situations, and can
improve therapeutic responses when combined with
antiviral therapies

Conventional immunosuppression
Treatment should be modulated according to the
underlying aetiopathogenesis (hyperviscosity vs vasculitis)
and the severity of clinical presentation (gure 5).1,93,98,105,106
There are three broad strategies in the treatment of
cryoglobulinaemia: conventional immunosuppression,
antiviral treatments, and biological therapies (panel 2).
The immunosuppressive approaches used in cryoglobulinaemic vasculitis, based on high-dose glucocorticoids and cyclophosphamide, were derived mainly
from strategies used in other systemic vasculitides before
it was understood that most cases result from HCV
infection. No clinical trial has studied these agents in
cryoglobulinaemic vasculitis. However, they remain
essential to control severe disease quickly and can help to
alter the course of disease if used shrewdly as a bridge to
antiviral or biological agents. In patients with severe
cryoglobulinaemic vasculitis, one goal of therapy should
be to discontinue conventional immunosuppressive
agents within 23 months, once the major end-organ
eects have been controlled.
Patients with moderate to severe manifestations of
cryoglobulinaemic vasculitis can be treated with
glucocorticoids to control inammation rapidly.107
Regimens of methylprednisolone (0510 g/day) for
3 days followed by prednisone (10 mg/kg daily, not to
exceed 80 mg/day) are appropriate in the setting of skin
ulceration, sensorimotor neuropathy, glomerulonephritis, and other severe vasculitic manifestations.
When disease control is achieved, glucocorticoids should
be tapered to the minimum dose necessary to suppress
activity and discontinued altogether if possible.
Cyclophosphamide is used in the most severe cases of
cryoglobulinaemic vasculitis, either daily (oral, 2 mg/kg
daily) or intermittent (intravenous, 750 mg/m monthly)
regimens. Azathioprine (2 mg/kg daily) and mycophenolate mofetil (1 g twice daily) are often used instead
www.thelancet.com Vol 379 January 28, 2012

of cyclophosphamide or after cyclophosphamide for


remission maintenance, although neither has been
assessed in clinical trials for this aim.13,98,108
Immunosuppression requires close monitoring of
patients blood counts and other variables. Patients given
glucocorticoids and cyclophosphamide should receive
prophylaxis for Pneumocystis pneumonia and surveillance
for the development of other opportunistic infections.

Apheresis
Both plasma exchange and plasmapheresis remove
cryoglobulins from the circulation, thereby interrupting
the immune-complex-mediated pathogenesis of cryoglobulinaemic vasculitis. These interventions are useful in
patients with immediately life-threatening disease and for
those with hyperviscosity syndrome. However, apheresis
does not treat the underlying disease and can lead to
rebound in which cryoglobulin production increases after
the cessation of apheresis. Cyclophosphamide for up to
6 weeks might be needed to prevent this rebound.

Antiviral treatments
The most robust antiviral approach to the treatment of
HCV-related cryoglobulinaemic vasculitis involves the
use of both pegylated interferon alfa and ribavirin.109111
Patients with HCV genotypes 2 and 3 respond most
favourably to this regimen, with between 75% and 90%
achieving sustained virological responses at 24 weeks,
whereas those with genotypes 1 and 4 have a lower
likelihood of achieving sustained virological responses
(4552%).112114 The recommended regimen is shown in
panel 3.114 The dose of interferon should be carefully
adjusted in cirrhotic patients, whereas the dose of
ribavirin should be adjusted in patients with renal failure
(glomerular ltration rate <60 mL/min per m).
355

Seminar

Panel 3: Recommended treatment regimen for


combination interferon* and ribavirin therapy114
Therapeutic regimens
Pegylated interferon alfa-2a 180 g subcutaneously per week
+
Ribavirin
1000 mg orally per day (bodyweight <75 kg)
1200 mg orally per day (bodyweight >75 kg)
or
Pegylated interferon alfa-2b 15 g/kg subcutaneously per
week
+
Ribavirin
800 mg orally per day (bodyweight 65 kg)
1000 mg orally per day (bodyweight 6585 kg)
1200 mg orally per day (bodyweight 85105 kg)
1400 mg orally per day (bodyweight >105 kg)
Recommended duration of antiviral courses
Genotypes 1 and 4: 48 weeks
Genotypes 2 and 3: 24 weeks
*Use with extreme caution in cirrhotic patients. Dose adjustment in patients with
glomerular ltration rate <60 mL/min. Duration also determined by evaluation of
early virological response and clinical and immunological responses.

Interferon is currently contraindicated in recipients of


kidney, heart, and lung allografts.114
Failure to achieve a virological response at weeks 12
and 24 is associated with a low probability of achieving a
sustained virological response,111 and the guidelines of
the American Association for the Study of Liver Diseases
recommend discontinuation of antiviral therapy in this
setting. However, the eectiveness of antiviral therapy in
patients with cryoglobulinaemic vasculitis should be
assessed not only according to the virological response,
but also by the full eect of other clinical and immunological responses achieved.115 Maintenance of antiviral
therapies can be appropriate in patients with cryoglobulinaemia who seem to benet.
Drug intolerance and dose reductions are common with
both interferon alfa and ribavirin.116 Inuenza-like symptoms, cytopenia, depression, and autoimmune thyroiditis
are common with interferon alfa. Exacerbation of vasculitic
manifestations and delayed healing of necrotic ulcers have
also been reported with interferon alfa.117,118 Haemolytic
anaemia is the major adverse eect associated with
ribavirin. New generations of agents based on interferon
alfa and ribavirin are likely to improve dosing schedules
and lower the risk of adverse events.119,120

Biological therapies
The most-promising biological approach to cryoglobulinaemia used thus far is B-cell depletion with
rituximab. Peripheral B-lymphocyte depletion should
lead to a reduction in the B-cell clones that produce
356

cryoglobulins. This idea needs further testing in


mechanistic studies linked to clinical trials, but the
results of small series are promising.121126 These studies
suggest that disease relapses are associated with the
absence of virological control and the recovery of
peripheral B cells. In patients receiving conventional
therapy, a third had a relapse of vasculitis despite
sustained virological response.127 This nding suggests
that B-cell proliferation can become independent of HCV
over time and that a targeted B-cell approach combined
with pegylated interferon alfa and ribavirin might be
successful in deleting both virus-dependent and virusindependent clones.
Substantial clinical experience in the o-label use of
rituximab in patients with HCV-related cryoglobulinaemia
has been accumulated.122124 Three studies reported in
2010 deserve careful scrutiny.128130 A prospective study128
compared the outcomes of 55 patients given antiviral
therapy (pegylated interferon alfa and ribavirin for
48 weeks) to those of 38 patients given a sequential
regimen of rituximab (intravenous administration of
four infusions at 375 mg/m per week or two fortnightly
infusions of 1000 mg), followed 1 month later by antiviral
therapy. The rituximab-treated patients had a shorter
mean time to clinical remission (54 months vs
84 months, respectively), better renal response rates
(81% vs 40%), and higher rates of cryoglobulin clearance
(68% vs 44%) compared with those assigned to antiviral
therapy. Treatment discontinuations were similar in the
two groups (13% vs 9%).
The second study129 was a prospective trial of 41 patients
who were randomised to receive antiviral therapy alone or
combined with rituximab (375 mg/m once a week for
1 month, plus two 5-monthly infusions during the 48-week
course of antiviral treatment). A higher rate of complete
response was reported in patients given the combined
regimen (54% vs 33%). A third study, limited by small
sample size, reported excellent tolerance of rituximab in
cryoglobulinaemic patients with advanced HCV-related
liver disease, including improvement in cirrhosis.130
While awaiting the results of randomised trials, some
notes of caution about the o-label use of rituximab
should be highlighted. A reasonable assessment of the
benets and risks (especially severe infections)131,132 should
be made on an individual basis. Additionally, a substantial
risk of complex formation between rituximab and IgM-
with rheumatoid-factor activity has been reported,133
especially in patients with high baseline concentrations
of cryoglobulin.

Future perspectives
A new era in our understanding of and therapeutic
approach to cryoglobulinaemia began 20 years ago with
the discovery of the HCV. The progressive introduction
of antiviral therapies, with eradication of HCV currently
regarded as the therapeutic gold standard, has improved
survival rates.77,134 However, many aspects of the disease
www.thelancet.com Vol 379 January 28, 2012

Seminar

remain unresolved.1 The causes of essential cryoglobulinaemia are not known. Disease management
remains dicult in many patients, because many do not
respond or are intolerant to antiviral therapies. Few data
are available on the treatment of patients with essential131
or type I135,136 cryoglobulinaemia. The role of new antiviral
agents in HCV-related cryoglobulinaemia, such as
telaprevir,137 remains to be dened. Similarly, international eorts are needed to develop a set of criteria for
a homogeneous classication of patients in future
epidemiological and therapeutic studies.138
The new century has seen the emergence of B-celltargeted therapies, whose use is currently limited by the
absence of specic licensing. Two studies reported
in 2010128,129 suggest greater benet when rituximab is
added to antiviral therapy, although some researchers
support rituximab monotherapy.139 Dierences in study
designs and therapeutic schedules do not allow denitive
recommendations on the relative timing of antiviral
and B-cell-depletion strategies. B-cell-activating factor of
the family of tumour-necrosis-factor blocking agents140 and
Toll-like receptor agonists141 might be promising therapies.
Cryoglobulinaemia is characterised by a wide range of
causes, symptoms, and outcomes, with various aetiopathogenic pathways involved in organ damage. This
complex scenario results in equally complex management
considerations. The present emphasis on approaching
the dierent causes in a simultaneous or closely
sequenced manner could produce improved therapeutic
results for patients with this challenging disorder.
Contributors
All authors contributed to the literature search and writing of the
Seminar. MR-C, JHS, MCC contributed to the gures.
Conicts of interest
MCC has received consultancy fees, research grants and speakers fees
from Centocor; and speakers fees from Bristol-Myers Squibb and
Roche. JS has received consultancy fees from Genentech. XB and MR-C
declare that they have no conicts of interest.

10

11

12

13
14

15

16

17

18

19

20

21

22

Acknowledgments
We thank David Buss for his technical assistance.
References
1
Ferri C. Mixed cryoglobulinaemia. Orphanet J Rare Dis 2008; 3: 25.
2
Wintrobe MM, Buell MV. Hyperproteinemia associated with
multiple myeloma. Bull Johns Hopkins Hosp 1933; 52: 15665.
3
Lerner AB, Watson CJ. Studies of cryoglobulins I: unusual
purpura associated with the presence of a high concentration
of cryoglobulin (cold precipitable serum globulin). Am J Med Sci
1947; 214: 41015.
4
Meltzer M, Franklin EC. Cryoglobulinaemia: a study of 29 patients.
I: IgG and IgM cryoglobulins and factors eecting
cryoprecipitability. Am J Med 1966; 40: 82836.
5
Brouet JC, Clauvel JP, Danon F, Klein M, Seligmann M. Biologic
and clinical signicance of cryoglobulins: a report of 86 cases.
Am J Med 1974; 57: 77588.
6
Tissot JD, Schierli JA, Hochstrasser DF, et al. Two-dimensional
polyacrylamide gel electrophoresis analysis of cryoglobulins and
identication of an IgM-associated peptide. J Immunol Meth 1994;
173: 6375.
7
Trejo O, Ramos-Casals M, Garca-Carrasco M, et al.
Cryoglobulinaemia: study of etiologic factors and clinical and
immunologic features in 443 patients from a single center.
Medicine (Baltimore) 2001; 80: 25262.

www.thelancet.com Vol 379 January 28, 2012

23
24

25

26

27

28

Choo QL, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton M.
Isolation of a cDNA clone derived from a blood borne non-A, non-B
viral hepatitis genome. Science 1989; 244: 35962.
Bambara LM, Biasi D, Caramaschi P, Carletto A, Pacor ML.
Cryoglobulinaemia and hepatitis C virus (HCV) infection.
Clin Exp Rheumatol 1991; 9: 9697.
Ferri C, Greco F, Longombardo G, et al. Antibodies to hepatitis C
virus in patients with mixed cryoglobulinaemia. Arthritis Rheum
1991; 34: 160610.
Ferri C, Greco F, Longombardo G, et al. Association between
hepatitis C virus and mixed cryoglobulinemia. Clin Exp Rheumatol
1991; 9: 62124.
Sansonno D, Carbone A, De Re V, Dammacco F. Hepatitis C virus
infection, cryoglobulinaemia, and beyond. Rheumatology (Oxford)
2007; 46: 57278.
Tedeschi A, Barat C, Minola E, Morra E. Cryoglobulinaemia.
Blood Rev 2007; 21: 183200.
Persico M, De Marino FA, Di Giacomo Russo G, et al. Prevalence
and incidence of cryoglobulins in hepatitis C virus-related chronic
hepatitis patients: a prospective study. Am J Gastroenterol 2003;
98: 8848.
Adinol LE, Utili R, Attanasio V, et al. Epidemiology, clinical
spectrum and prognostic value of mixed cryoglobulinaemia in
hepatitis C virus patients: a prospective study. Ital J Gastroenterol
1996; 28: 19.
Ferri C, Sebastiani M, Giuggioli D, et al. Mixed cryoglobulinaemia:
demographic, clinical, and serologic features and survival in
231 patients. Semin Arthritis Rheum 2004; 33: 35574.
Cohen P, Roulot D, Ferrire F, et al. Prevalence of cryoglobulins
and hepatitis C virus infection in HIV-infected patients.
Clin Exp Rheumatol 1997; 15: 52327.
Dimitrakopoulos AN, Kordossis T, Hatzakis A, Moutsopoulos HM.
Mixed cryoglobulinaemia in HIV-1 infection: the role of HIV-1.
Ann Intern Med 1999; 130: 22630.
Fabris P, Tositti G, Giordani MT, et al. Prevalence and clinical
signicance of circulating cryoglobulins in HIV-positive patients
with and without co-infection with hepatitis C virus. J Med Virol
2003; 69: 33943.
Ramos-Casals M, Forns X, Brito-Zern P, et al. Cryoglobulinaemia
associated with hepatitis C virus: inuence of HCV genotypes,
HCV-RNA viraemia and HIV coinfection. J Viral Hepat 2007;
14: 73642.
Bonnet F, Pineau JJ, Taupin JL, et al. Prevalence of
cryoglobulinaemia and serological markers of autoimmunity
in human immunodeciency virus infected individuals:
a cross-sectional study of 97 patients. J Rheumatol 2003;
30: 200510.
Kosmas N, Kontos A, Panayiotakopoulos G, Dimitrakopoulos A,
Kordossis T. Decreased prevalence of mixed cryoglobulinaemia in
the HAART era among HIV-positive, HCV-negative patients.
J Med Virol 2006; 78: 125761.
Ramos-Casals M, Trejo O, Garca-Carrasco M, Cervera R, Font J.
Mixed cryoglobulinaemia: new concepts. Lupus 2000; 9: 8391.
Tzioufas AG, Boumba DS, Skopouli FN, Moutsopoulos HM. Mixed
monoclonal cryoglobulinaemia and monoclonal rheumatoid factor
cross-reactive idiotypes as predictive factors for the development of
lymphoma in primary Sjogrens syndrome. Arthritis Rheum 1996;
39: 76772.
Brito-Zern P, Ramos-Casals M, Bove A, Sentis J, Font J. Predicting
adverse outcomes in primary Sjgrens syndrome: identication of
prognostic factors. Rheumatology (Oxford) 2007; 46: 135962.
Baimpa E, Dahabreh IJ, Voulgarelis M, Moutsopoulos HM.
Hematologic manifestations and predictors of lymphoma
development in primary Sjgren syndrome: clinical and
pathophysiologic aspects. Medicine (Baltimore) 2009; 88: 28493.
Ramos-Casals M, Loustaud-Ratti V, De Vita S, et al, for the SS-HCV
Study Group. Sjgren syndrome associated with hepatitis C virus:
a multicenter analysis of 137 cases. Medicine (Baltimore) 2005;
84: 8189.
Garca-Carrasco M, Ramos-Casals M, Cervera R, et al.
Cryoglobulinaemia in systemic lupus erythematosus: prevalence
and clinical characteristics in a series of 122 patients.
Semin Arthritis Rheum 2001; 30: 36673.

357

Seminar

29

30

31

32

33
34
35

36

37

38

39
40

41

42

43

44

45

46

47

48

49
50

51

358

Perlemuter G, Cacoub P, Sba A, et al. Hepatitis C virus infection


in systemic lupus erythematosus: a case-control study. J Rheumatol
2003; 30: 147378.
Dammacco F, Lauletta G, Montrone M, Sansonno D. Mixed
cryoglobulinaemia: a model of virus-related disease in internal
medicine. Dig Liver Dis 2007; 39 (suppl 1): S812.
Trejo O, Ramos-Casals M, Lpez-Guillermo A, et al. Hematologic
malignancies in patients with cryoglobulinaemia: association with
autoimmune and chronic viral diseases. Semin Arthritis Rheum
2003; 33: 1928.
Saadoun D, Sellam J, Ghillani-Dalbin P, Crecel R, Piette JC,
Cacoub P. Increased risks of lymphoma and death among patients
with non-hepatitis C virus-related mixed cryoglobulinaemia.
Arch Intern Med 2006; 166: 210108.
Casato M, Lilli D, Donato G, et al. Occult hepatitis C virus infection
in type II mixed cryoglobulinaemia. J Viral Hepat 2003; 10: 45559.
Trendelenburg M, Schierli JA. Cryoglobulins are not essential.
Ann Rheum Dis 1998; 57: 35.
Lunel F, Musset L, Cacoub P et al. Cryoglobulinemia in chronic
liver diseases: role of hepatitis C virus and liver damage.
Gastroenterology 1994; 106: 1291300.
Kayali Z, Bochwold VE, Zimmermann B, Schmidt WN. Hepatitis C,
cryoglobulinemia, and cirrhosis: A meta-analysis. Hepatology 2002;
36: 97885.
Ramos-Casals M, De Vita S, Tzioufas AG. Hepatitis C virus,
Sjgrens syndrome and B-cell lymphoma: linking infection,
autoimmunity and cancer. Autoimmun Rev 2005; 4: 815.
Ferri C, Antonelli A, Mascia MT, et al. HCV-related autoimmune
and neoplastic disorders: the HCV syndrome. Dig Liver Dis 2007;
39 (suppl 1): S1321.
Pileri P, Uematsu Y, Campagnoli S, et al. Binding of hepatitis C
virus to CD81. Science 1998; 282: 93841.
Rosa D, Saletti G, De Gregorio E, et al. Activation of nave B
lymphocytes via CD81, a pathogenetic mechanism for hepatitis C
virus-associated B lymphocyte disorders. Proc Natl Acad Sci USA
2005; 102: 1854449.
Sansonno D, Lauletta G, De Re V et al. Intrahepatic B cell clonal
expansions and extrahepatic manifestations of chronic HCV
infection. Eur J Immunol 2004; 34: 12636.
Quartuccio L, Fabris M, Salvin S, et al. Bone marrow B-cell clonal
expansion in type II mixed cryoglobulinaemia: association with
nephritis. Rheumatology (Oxford) 2007; 46: 165761.
Charles ED, Green RM, Marukian S, et al. Clonal expansion
of immunoglobulin M+CD27+ B cells in HCV-associated mixed
cryoglobulinaemia. Blood 2008; 111: 134456.
Chen PP, Fong S, Goni F, et al. Cross-reacting idiotypes on
cryoprecipitating rheumatoid factor. Semin Immunopathol 1988;
10: 3555.
Knight GB, Gao L, Gragnani L, et al. Detection of WA B cells
in hepatitis C virus infection: a potential prognostic marker for
cryoglobulinemic vasculitis and B cell malignancies.
Arthritis Rheum 2010; 62: 215259.
Sansonno D, Cornacchiulo V, Iacobelli AR, Di Stefano R,
Lospalluti M, Dammacco F. Localization of hepatitis C virus antigens
in liver and skin tissues of chronic hepatitis C virus-infected patients
with mixed cryoglobulinaemia. Hepatology 1995; 21: 30512.
Sansonno D, Gesualdo L, Manno C, Schena FP, Dammacco F.
Hepatitis C virus-related proteins in kidney tissue from hepatitis C
virus-infected patients with cryoglobulinaemic
membranoproliferative glomerulonephritis. Hepatology 1997;
25: 123744.
Sansonno D, Lauletta G, Nisi L, et al. Non-enveloped HCV core
protein as constitutive antigen of cold-precipitable immune
complexes in type II mixed cryoglobulinaemia. Clin Exp Immunol
2003; 133: 27582.
Sargur R, White P, Egner W. Cryoglobulin evaluation: best practice?
Ann Clin Biochem 2010; 47: 816.
Mizuochi T, Pastore Y, Shikata K, et al. Role of galactosylation in
the renal pathogenicity of murine immunoglobulin G3 monoclonal
cryoglobulins. Blood 2001; 97: 353743.
Sansonno D, Dammacco F. Hepatitis C virus, cryoglobulinaemia,
and vasculitis: immune complex relations. Lancet Infect Dis 2005;
5: 22736.

52

53

54

55

56
57

58

59

60
61

62

63

64

65
66

67

68

69

70

71

72

73

74

Sansonno D, Tucci FA, Ghebrehiwet B, et al. Role of the receptor


for the globular domain of C1q protein in the pathogenesis of
hepatitis C virus-related cryoglobulin vascular damage. J Immunol
2009; 183: 601320.
Sene D, Ghillani-Dalbin P, Thibault V, et al. Longterm course
of mixed cryoglobulinaemia in patients infected with hepatitis C
virus. J Rheumatol 2004; 31: 2199206.
Della Rossa A, Tavoni A, Bombardieri S. Hyperviscosity syndrome
in cryoglobulinaemia: clinical aspects and therapeutic
considerations. Semin Thromb Hemost 2003; 29: 47377.
Della Rossa A, Tavoni A, DAscanio A, et al. Mortality rate and
outcome factors in mixed cryoglobulinaemia: the impact
of hepatitis C virus. Scand J Rheumatol 2010; 39: 16770.
Treon SP. How I treat Waldenstrm macroglobulinemia.
Blood 2009; 114: 237585.
Menke MN, Treon SP. Hyperviscosity syndrome. In: Sekeres MA,
Kalaycio ME, Bolwell BJ, eds. Clinical malignant hematology.
New York: McGraw-Hill, 2007: 93741.
Rieu V, Cohen P, Andr MH, et al. Characteristics and outcome
of 49 patients with symptomatic cryoglobulinaemia.
Rheumatology (Oxford) 2002; 41: 290300.
Wener MH, Hutchinson K, Morishima C, Gretch DR. Absence
of antibodies to cyclic citrullinated peptide in sera of patients with
hepatitis C virus infection and cryoglobulinaemia. Arthritis Rheum
2004; 50: 230508.
Poynard T, Cacoub P, Ratziu V, et al, for the Multivirc group. Fatigue
in patients with chronic hepatitis C. J Viral Hepat 2002; 9: 295303.
Antonelli A, Ferri C, Fallahi P, et al. Thyroid involvement in
patients with overt HCV-related mixed cryoglobulinaemia. Q JM
2004; 97: 499506.
Antonelli A, Ferri C, Fallahi P, et al. Type 2 diabetes in hepatitis
C-related mixed cryoglobulinaemia patients. Rheumatology (Oxford)
2004; 43: 23840.
Bryce AH, Kyle RA, Dispenzieri A, Gertz MA. Natural history and
therapy of 66 patients with mixed cryoglobulinaemia. Am J Hematol
2006; 81: 51118.
Matignon M, Cacoub P, Colombat M, et al. Clinical and
morphologic spectrum of renal involvement in patients with mixed
cryoglobulinaemia without evidence of hepatitis C virus infection.
Medicine (Baltimore) 2009; 88: 34148.
Alpers CE, Smith KD. Cryoglobulinaemia and renal disease.
Curr Opin Nephrol Hypertens 2008; 17: 24349.
Pioltelli P, Maldifassi P, Vacca A, et al. GISC protocol experience
in the treatment of essential mixed cryoglobulinaemia.
Clin Exp Rheumatol 1995; 13 (suppl): S18790.
Roccatello D, Fornasieri A, Giachino O, et al. Multicenter study
on hepatitis C virus related cryoglobulinaemic glomerulonephritis.
Am J Kidney Dis 2007; 49: 6982.
Beddhu S, Bastacky S, Johnson JP. The clinical and morphologic
spectrum of renal cryoglobulinaemia. Medicine (Baltimore) 2002;
81: 398409.
Tarantino A, Campise M, Ban G, et al. Long-term predictors
of survival in essential mixed cryoglobulinaemic
glomerulonephritis. Kidney Int 1995; 47: 61823.
Basse G, Ribes D, Kamar N, Esposito L, Rostaing L.
Life-threatening infections following rituximab therapy in renal
transplant patients with mixed cryoglobulinemia. Clin Nephrol
2006; 66: 39596.
Hiesse C, Bastuji-Garin S, Santelli G, et al. Recurrent essential
mixed cryoglobulinemia in renal allografts. Report of two cases
and review of the literature. Am J Nephrol 1989; 9: 15054.
Gemignani F, Melli G, Inglese C, Marbini A. Cryoglobulinaemia
is a frequent cause of peripheral neuropathy in undiagnosed
referral patients. J Peripher Nerv Syst 2002; 7: 5964.
Nemni R, Sanvito L, Quattrini A, Santuccio G, Camerlingo M,
Canal N. Peripheral neuropathy in hepatitis C virus infection with
and without cryoglobulinaemia. J Neurol Neurosurg Psychiatry 2003;
74: 126771.
Santoro L, Manganelli F, Briani C, et al; HCV Peripheral Nerve
Study Group. Prevalence and characteristics of peripheral
neuropathy in hepatitis C virus population.
J Neurol Neurosurg Psychiatry 2006; 77: 62629.

www.thelancet.com Vol 379 January 28, 2012

Seminar

75

76

77

78

79

80

81
82

83

84

85

86

87

88

89

90

91

92

93

94

95

96
97

Taieb G, Maisonobe T, Musset L, Cacoub P, Lger JM, Bouche P.


Cryoglobulinaemic peripheral neuropathy in hepatitis C virus
infection: Clinical and anatomical correlations of 22 cases.
Rev Neurol (Paris) 2010; 166: 50914.
Vidaller A, Cid MC. Tetraparesis and paresthesias in the lower
extremities in a 58-year-old man. Med Clin (Barc) 1999;
112: 47076 (in Spanish).
Terrier B, Saadoun D, Sne D, Scerra S, Musset L, Cacoub P.
Presentation and outcome of gastrointestinal involvement in
hepatitis C virus-related systemic vasculitis: a case-control study
from a single-centre cohort of 163 patients. Gut 2010; 59: 170915.
Amital H, Rubinow A, Naparstek Y. Alveolar hemorrhage
in cryoglobulinaemiaan indicator of poor prognosis.
Clin Exp Rheumatol 2005; 23: 61620.
Ferri C, La Civita L, Fazzi P, et al. Interstitial lung brosis and
rheumatic disorders in patients with hepatitis C virus infection.
Br J Rheumatol 1997; 36: 36065.
Retamozo S, Diaz-Lagares C, Bosch X, De Vita S, Ramos-Casals M.
Life-threatening cryoglobulinemia. In: Khamashta MA,
Ramos-Casals M, eds. Autoimmune diseases: acute and complex
situations London: Springer-Verlag, 2011: 13362.
Morgello S. The nervous system and hepatitis C virus.
Semin Liver Dis 2005; 25: 11821.
Safadi R, Ilan Y, Ashur Y, Shouval D. Hepatitis C-associated
cryoglobulinaemia presenting with pericardial eusion.
Am J Gastroenterol 1997; 92: 71012.
Basu N, Watts R, Bajema I, et al. EULAR points to consider in the
development of classication and diagnostic criteria in systemic
vasculitis. Ann Rheum Dis 2010; 69: 174450.
Bakker AJ, Slomp J, de Vries T, et al. Adequate sampling in
cryoglobulinaemia: recommended warmly. Clin Chem Lab Med
2003; 41: 8589.
Vermeersch P, Gijbels K, Marin G, et al. A critical appraisal
of current practice in the detection, analysis, and reporting
of cryoglobulins. Clin Chem 2008; 54: 3943.
Maire MA, Mittey M, Lambert PH. The presence of cryoprecipitable
immunoglobulins in normal human sera may reect specic
molecular interactions. Autoimmunity 1989; 2: 15564.
Shihabi ZK. Analysis and general classication of serum
cryoglobulins by capillary zone electrophoresis. Electrophoresis 1996;
17: 160712.
Vitali C, Ferri C, Nasti P, et al. Hypercomplementaemia as a marker
of the evolution from benign to malignant B cell proliferation in
patients with type II mixed cryoglobulinaemia. Br J Rheumatol 1994;
33: 79192.
DAmico G. Renal involvement in hepatitis C infection:
cryoglobulinaemic glomerulonephritis. Kidney Int 1998;
54: 65071.
Gopalani A, Ahuja TS. Prevalence of glomerulopathies in autopsies
of patients infected with the hepatitis C virus. Am J Med Sci 2001;
322: 5760.
Sumida K, Ubara Y, Hoshino J, et al. Hepatitis C virus-related
kidney disease: various histological patterns. Clin Nephrol
2010; 74: 44656.
Vallat JM, Leboutet MJ, Loubet A. Cryoglobulinaemic neuropathy:
ultrastructural study of the nerve capillaries. Acta Neuropathol 1981;
7: 25254.
Vallat JM, Desproges-Gotteron R, Leboutet MJ, Loubet A, Gualde N,
Treves R. Cryoglobulinaemic neuropathy: a pathological study.
Ann Neurol 1980; 8: 17985.
Ramos-Casals M, Robles A, Brito-Zeron P, et al. Life-threatening
cryoglobulinaemia: clinical and immunological characterizsation
of 29 cases. Semin Arthritis Rheum 2006; 36: 18996.
Monti G, Saccardo F, Pioltelli P, Rinaldi G. The natural history
of cryoglobulinaemia: symptoms at onset and during follow-up:
a report by the Italian Group for the Study of Cryoglobulinaemias
(GISC). Clin Exp Rheumatol 1995; 13 (suppl): S12933.
Monti G, Saccardo F. Emergency in cryoglobulinaemic syndrome:
what to do? Dig Liver Dis. 2007; 39 (suppl 1): S11215.
Ramos-Casals M, Trejo O, Garca-Carrasco M, et al. Triple
association between hepatitis C virus infection, systemic
autoimmune diseases, and B cell lymphoma. J Rheumatol 2004;
31: 49599.

www.thelancet.com Vol 379 January 28, 2012

98

99

100

101

102

103

104

105
106

107

108
109
110

111

112

113

114

115

116
117

118

119

Saadoun D, Landau DA, Calabrese LH, Cacoub PP.


Hepatitis C-associated mixed cryoglobulinaemia: a crossroad
between autoimmunity and lymphoproliferation.
Rheumatology (Oxford) 2007; 46: 123442.
La Civita L, Zignego AL, Monti M, Longombardo G, Pasero G,
Ferri C. Mixed cryoglobulinaemia as a possible preneoplastic
disorder. Arthritis Rheum 1995; 38: 185960.
McColl MD, Singer IO, Tait RC, McNeil IR, Cumming RL,
Hogg RB. The role of hepatitis C virus in the aetiology of
non-Hodgkins lymphomaa regional association? Leuk Lymphoma
1997; 26: 12730.
Monti G, Pioltelli P, Saccardo F, et al. Incidence and characteristics
of non-Hodgkin lymphomas in a multicenter case le of patients
with hepatitis C virus-related symptomatic mixed
cryoglobulinemias. Arch Intern Med 2005; 165: 10105.
Geri G, Terrier B, Semoun O, et al. Surrogate markers of B cell
non-Hodgkins lymphoma in patients with hepatitis C virus-related
cryoglobulinaemia vasculitis. Ann Rheum Dis 2010; 69: 217780.
Giordano TP, Henderson L, Landgren O, et al. Risk of non-Hodgkin
lymphoma and lymphoproliferative precursor diseases in US
veterans with hepatitis C virus. JAMA 2007; 297: 201017.
Saccardo E, Novati P, Sironi D, Castelnovo L, Rinaldi G, Monti G.
Causes of death in symptomatic cryoglobulinaemia: 30 years
of observation in a Department of Internal Medicine. Dig Liver Dis
2007; 39 (suppl 1): S5254.
Ferri C, Mascia MT. Cryoglobulinemic vasculitis.
Curr Opin Rheumatol 2006; 18: 5463.
Saadoun D, Delluc A, Piette JC, Cacoub P. Treatment
of hepatitis C-associated mixed cryoglobulinemia vasculitis.
Curr Opin Rheumatol 2008; 20: 2328.
Vassilopoulos D, Calabrese LH. Hepatitis C virus infection and
vasculitis. Implications of antiviral and immunosuppressive
therapies. Arthritis Rheum 2002; 46: 58597.
Ramos-Casals M, Font J. Mycophenolate mofetil in patients with
hepatitis C virus infection. Lupus 2005; 14 (suppl): S6472.
Feld JJ, Hoofnagle JH. Mechanism of action of interferon and
ribavirin in treatment of hepatitis C. Nature 2005; 436: 96772.
Cacoub P, Saadoun D, Limal N, Sene D, Lidove O, Piette JC.
PEGylated interferon alfa-2b and ribavirin treatment in patients
with hepatitis C virus-related systemic vasculitis. Arthritis Rheum
2005; 52: 91115.
Saadoun D, Resche-Rigon M, Thibault V, Piette JC, Cacoub P.
Antiviral therapy for hepatitis C virus associated mixed
cryoglobulinemia vasculitis: a long-term follow-up study.
Arthritis Rheum 2006; 54: 3696706.
Fried MW, Shiman ML, Reddy KR et al. Peginterferon alfa-2a plus
ribavirin for chronic hepatitis C virus infection. N Engl J Med 2002;
347: 97582.
Hadziyannis SJ, Sette H Jr., Morgan TR, et al, for the PEGASYS
International Study Group. Peginterferon-alpha 2a and ribavirin
combination therapy in chronic hepatitis C: a randomized study
of treatment duration and ribavirin dose. Ann Intern Med 2004;
140: 34655.
Ghany MG, Strader DB, Thomas DL, See LB, for the American
Association for the Study of Liver Diseases. Diagnosis,
management, and treatment of hepatitis C: an update.
Hepatology 2009; 49: 133574.
Joshi S, Kuczynski M, Heathcote EJ. Symptomatic and virological
response to antiviral therapy in hepatitis C associated with
extrahepatic complications of cryoglobulimia. Dig Dis Sci 2007;
52: 241017.
Russo MW, Fried MW. Side eects of therapy for chronic
hepatitis C. Gastroenterology 2003; 124: 171119.
Gordon AC, Edgar JD, Finch RG. Acute exacerbation of vasculitis
during interferon-alpha therapy for hepatitis C-associated
cryoglobulinaemia. J Infect 1998; 36: 22930.
Cid MC, Hernndez-Rodrguez J, Robert J, et al. Interferon-alpha
may exacerbate cryoblobulinemia-related ischemic manifestations:
an adverse eect potentially related to its anti-angiogenic activity.
Arthritis Rheum 1999; 42: 105155.
Nelson DR, Rustgi V, Balan V, et al. Safety and antiviral activity
of albinterferon alfa-2b in prior interferon non-responders with
chronic hepatitis C. Clin Gastroenterol Hepatol 2009; 7: 21218.

359

Seminar

120 Gish RG, Arora S, Rajender Reddy K, et al. Virological response


and safety outcomes in therapy-nave patients treated for chronic
hepatitis C with taribavirin or ribavirin in combination with
pegylated interferon alfa-2a: a randomizsed, phase 2 study.
J Hepatol 2007; 47: 5159.
121 Quartuccio L, Soardo G, Romano G, et al. Rituximab treatment for
glomerulonephritis in HCV-associated mixed cryoglobulinaemia:
ecacy and safety in the absence of steroids. Rheumatology (Oxford)
2006; 45: 84246.
122 Lamprecht P, Lerin-Lozano C, Merz H, et al. Rituximab induces
remission in refractory HCV-associated mixed cryoglobulinemic
vasculitis. Ann Rheum Dis 2003; 62: 123033.
123 Sansonno D, De Re V, Lauletta G, Tucci FA, Boiocchi M,
Dammacco F. Monoclonal antibody treatment of mixed
cryoglobulinemia resistant to interferon alpha with an anti-CD20.
Blood 2003; 101: 381826.
124 Zaja F, De Vita S, Mazzaro C, et al. Ecacy and safety of rituximab
in type II mixed cryoglobulinemia. Blood 2003; 101: 382734.
125 Saadoun D, Rosenzwaig M, Landau D, Piette JC, Klatzmann D,
Cacoub P. Restoration of peripheral immune homeostasis after
rituximab in mixed cryoglobulinemia vasculitis. Blood 2008;
111: 533441.
126 Roccatello D, Baldovino S, Rossi D, et al. Long-term eects of
anti-CD20 monoclonal antibody treatment of cryoglobulinaemic
glomerulonephritis. Nephrol Dial Transplant 2004; 19: 305461.
127 Landau DA, Saadoun D, Halfon P, et al. Relapse of hepatitis C
virus-associated mixed cryoglobulinemia vasculitis in patients with
sustained viral response. Arthritis Rheum 2008; 58: 60411.
128 Saadoun D, Resche-Rigon M, Sene D, et al. Rituximab plus
PEG-interferon alpha/ribavirin compared to PEG-interferon
alpha/ribavirin in hepatitis C-related mixed cryoglobulinemia.
Blood 2010; 116: 32634.
129 Dammacco F, Tucci FA, Lauletta G, et al. Pegylated interferon-alfa,
ribavirin, and rituximab combined therapy of hepatitis C
virus-related mixed cryoglobulinemia: a long-term study. Blood
2010; 116: 34353.
130 Petrarca A, Rigacci L, Caini P, et al. Safety and ecacy of rituximab
in patients with hepatitis C virus-related mixed cryoglobulinemia
and severe liver disease. Blood 2010; 116: 33542.

360

131 Terrier B, Launay D, Kaplanski G, et al. Safety and ecacy


of rituximab in nonviral cryoglobulinemia vasculitis: data from
the French Autoimmunity and Rituximab registry.
Arthritis Care Res (Hoboken) 2010; 62: 178795.
132 Gea-Banacloche JC. Rituximab-associated infections. Semin Hematol
2010; 47: 18798.
133 Sene D, Ghillani-Dalbin P, Amoura Z, Musset L, Cacoub P.
Rituximab may form a complex with IgM-kappa mixed cryoglobulin
and induce severe systemic reactions in patients with hepatitis C
virus-induced vasculitis. Arthritis Rheum 2009; 60: 384855.
134 Landau DA, Scerra S, Sene D, Resche-Rigon M, Saadoun D,
Cacoub P. Causes and predictive factors of mortality in a cohort
of patients with hepatitis C virus-related cryoglobulinemic vasculitis
treated with antiviral therapy. J Rheumatol 2010; 37: 61521.
135 Spizzo G, Mitterer M, Gunsilius E. Bortezomib for the treatment
of refractory type-1 cryoglobulinaemia. Br J Haematol 2010;
150: 23537.
136 Talamo G, Claxton D, Tricot G, Fink L, Zangari M. Response
to bortezomib in refractory type I cryoglobulinemia. Am J Hematol
2008; 83: 88384.
137 Hezode C, Forestier N, Dusheiko G, et al, for the PROVE2 Study
Team. Telaprevir and peginterferon with or without ribavirin for
chronic HCV infection. N Engl J Med 2009; 360: 183950.
138 De Vita S, Soldano F, Isola M, et al. Preliminary classication
criteria for the cryoglobulinemic syndrome. Ann Rheum Dis 2011;
70: 118390.
139 De Vita S, Quartuccio L. Rituximab monotherapy, rather than
rituximab plus antiviral drugs, for initial treatment of severe
hepatitis C virus-associated mixed cryoglobulinemia syndrome:
comment on the article by Terrier et al. Arthritis Rheum 2010;
62: 90809.
140 Sne D, Limal N, Ghillani-Dalbin P, Saadoun D, Piette JC,
Cacoub P. Hepatitis C virus-associated B-cell proliferationthe role
of serum B lymphocyte stimulator (BLyS/BAFF).
Rheumatology (Oxford) 2007; 46: 6569.
141 McHutchinson J, Bacon BR, Gordon SC, et al. Phase 1B,
randomized, double-blind, dose-escalation trial of CPG 10101 in
patients with chronic hepatitis C virus. Hepatology 2007; 46: 131718.

www.thelancet.com Vol 379 January 28, 2012

Seminar

Acne vulgaris
Hywel C Williams, Robert P Dellavalle, Sarah Garner

Acne is a chronic inammatory disease of the pilosebaceous unit resulting from androgen-induced increased sebum
production, altered keratinisation, inammation, and bacterial colonisation of hair follicles on the face, neck, chest,
and back by Propionibacterium acnes. Although early colonisation with P acnes and family history might have important
roles in the disease, exactly what triggers acne and how treatment aects the course of the disease remain unclear.
Other factors such as diet have been implicated, but not proven. Facial scarring due to acne aects up to 20% of
teenagers. Acne can persist into adulthood, with detrimental eects on self-esteem. There is no ideal treatment for
acne, although a suitable regimen for reducing lesions can be found for most patients. Good quality evidence on
comparative eectiveness of common topical and systemic acne therapies is scarce. Topical therapies including
benzoyl peroxide, retinoids, and antibiotics when used in combination usually improve control of mild to moderate
acne. Treatment with combined oral contraceptives can help women with acne. Patients with more severe inammatory
acne usually need oral antibiotics combined with topical benzoyl peroxide to decrease antibiotic-resistant organisms.
Oral isotretinoin is the most eective therapy and is used early in severe disease, although its use is limited by
teratogenicity and other side-eects. Availability, adverse eects, and cost, limit the use of photodynamic therapy. New
research is needed into the therapeutic comparative eectiveness and safety of the many products available, and to
better understand the natural history, subtypes, and triggers of acne.

Introduction
Acne is a disease of the pilosebaceous unithair follicles
in the skin that are associated with an oil gland (gure 1).2
The clinical features of acne include seborrhoea (excess
grease), non-inammatory lesions (open and closed
comedones), inammatory lesions (papules and pustules),
and various degrees of scarring. The distribution of acne
corresponds to the highest density of pilosebaceous units
(face, neck, upper chest, shoulders, and back). Nodules
and cysts comprise severe nodulocystic acne. This Seminar
summarises information relating to the clinical aspects of
common acne (acne vulgaris). Acne classication,
scarring, acne rosacea, chloracne, acne associated with
polycystic ovary syndrome, infantile acne, acne inversa,
and drug-induced acne have been reviewed elsewhere.310

Prevalence and natural history


Some degree of acne aects almost all people aged 15 to
17 years,1113 and is moderate to severe in about 1520%.8,12,14
Prevalence estimates are dicult to compare because
denitions of acne and acne severity have diered so
much between studies, and because estimates are
confounded by the availability and use of acne
treatments.15 Surveys of self-reported acne have proven
unreliable.16 Although perceived as a teenage disease,
acne often persists into adulthood.17,18 One population
study in Germany found that 64% of those aged 20 to
29 years and 43% of those aged 30 to 39 years had visible
acne.19 Another study of more than 2000 adults showed
that 3% of men and 5% of women still had denite mild
acne at the age of 40 to 49 years.20
Acne typically starts in early puberty with increased
facial grease production, and mid-facial comedones8
followed by inammatory lesions. Early-onset acne
(before the age of 12 years) is usually more comedonal
than inammatory, possibly because such individuals
have not yet begun to produce enough sebum to support
www.thelancet.com Vol 379 January 28, 2012

large numbers of Proprionibacterium acnes.21 One


prospective study of 133 children aged 55 to 12 years,
followed up for an average of 25 years, found
asynchronous facial sebum production initially, with
increasing numbers of glands switching on sebum
production over time.22 Subsequent expansion of the
propionibacterial skin ora (in the nares and then facial
skin) occurred earlier in children who developed acne
than in children of the same age and pubertal status who
did not, suggesting that postponement of sebum
production or expansion of propionibacterial skin ora
until after puberty could prevent acne or minimise
disease severity. Predictors of acne severity include early
onset of comedonal acne,8 and increasing number of
family members with acne history.14 Factors that can
cause acne to are include the menstrual cycle, picking,
and emotional stress.23,24 Beliefs about external factors
aecting acne vary according to ethnic group.25 Acne
vulgaris is a chronic disease that often persists for many
years.26 There is little research about what factors might
predict whether acne will last into adulthood.27 We could
not nd any good quality cohort studies summarising
the natural history of acne. Sequential prevalence surveys
of dierent populations showing a gradual decrease in

Lancet 2012; 379: 36172


Published Online
August 30, 2011
DOI:10.1016/S01406736(11)60321-8
This publication has
been corrected.
The corrected version rst
appeared at thelancet.com
on January 27, 2011
Centre of Evidence-Based
Dermatology, Nottingham
University Hospitals NHS
Trust, Nottingham, UK
(Prof H C Williams PhD);
Department of Dermatology,
University of Colorado Denver,
School of Medicine, Aurora and
VA Eastern Colorado Health
Care System, Denver, CO, USA
(R P Dellavalle MD); Center for
the Evaluation of Value and
Risk in Health, Tufts Medical
Centre, Boston, MA, USA
(S Garner PhD), and The
Commonwealth Fund,
New York, NY, USA (S Garner)
Correspondence to:
Prof Hywel C Williams, Centre of
Evidence-Based Dermatology,
Nottingham University
Hospitals NHS Trust,
Nottingham NG7 2UH, UK
hywel.williams@nottingham.
ac.uk

Search strategy and selection criteria


Our main sources of evidence included all systematic reviews
on acne published since 1999 which have been mapped by
NHS Evidenceskin disorders annual evidence updates,1
supplemented by specic searches on Medline for articles
published between January, 2003, and Jan 16, 2011, using the
search terms acne, comedones, vulgaris, and
aetiology, causes, natural history, pathophysiology,
treatment, management, and guidelines. We also
scrutinised citation lists from retrieved articles.

361

Seminar

Figure 1: Normal sebaceous follicle (A) and comedo (B), and inammatory acne lesion with rupture of follicular wall and secondary inammation (C)
Reproduced, with permission, from reference 2.

acne prevalence after the age of 20 years weakly underpin


our current understanding of the natural history of acne.
Mild inammatory acne declines or disappears in a large
proportion of those with acne in their teens. Cytokines
that induce comedogenic changes at the follicular
infundibulum might also inhibit lipid secretion from the
sebaceous gland, resulting in remission of individual
lesions.28 However, seborrhoea persists throughout adult
life, long after inammatory lesions have resolved.29
Adult acne related to circulating androgens goes by
several names, including post-adolescent or late-onset
acne, and occurs most commonly in women beyond the
age of 25 years.30

Cause
Risk factors and genes associated with acne prognosis
and treatment are unclear.31,32 Twin studies have pointed
to the importance of genetic factors for more severe
scarring acne.33 A positive family history of acne doubled
the risk of signicant acne in a study of 1002 Iranian
16-year-olds,14 and the heritability of acne was 78% in rstdegree relatives of those with acne in a large study of
Chinese undergraduates.34 Acne appears earlier in girls,
but more boys are aected during the mid-teenage years.35
Acne can occur at a younger age and be more comedonal
in black children than in white children, probably from
earlier onset of puberty.36 A study of 1394 Ghanaian
schoolchildren found that acne was less common in rural
locations, but the reasons for this are unclear.37
Although earlier observational studies suggested an
inverse association between smoking and acne,38 subsequent studies have shown that severe acne increases
with smoking.19,39 Increased insulin resistance and high
serum dehydroepiandrosterone might explain the
presence of acne in polycystic ovary syndrome.40,41
Occlusion of the skin surface with greasy products
(pomade acne),42 clothing, and sweating can worsen acne.
Drugs such as anti-epileptics typically produce a
362

monomorphic acne, and acneiform eruptions have been


associated with anti-cancer drugs such as getinib.10 The
use of anabolic steroids for increasing muscle bulk might
be underestimated, and can give rise to severe forms of
acne.43 Tropical acne can occur in military personnel
assigned to hot, humid conditions.44 Dioxin exposure can
result in severe comedonal acne (chloracne), but it is not
associated with common acne.
Diet, sunlight, and skin hygiene have all been
implicated in acne,45 but little evidence supports or
refutes such beliefs.46 One systematic review suggested
that dairy products (especially milk) increase acne risk,
but all the included observational studies had signicant shortcomings.47
Previous studies of giving young people large quantities
of chocolate to try and provoke acne were too small and
too short to claim no eect.48 The apparent absence of
acne in native non-Westernised people in Papua New
Guinea and Paraguay49 has led to the proposal that high
glycaemic loads in Western diet could have a role in acne,
perhaps through hyperinsulinaemia leading to increased
androgens, increased insulin-like growth-factor 1, and
altered retinoid signalling.50,51 A randomised controlled
trial showing that a low glycaemic load diet might
improve acne provides preliminary support for this
theory.52 Although acne has been associated with
increasing body mass,53 no evidence suggests that putting
people on restrictive diets reduces acne.

Disease mechanisms
Four processes have a pivotal role in the formation of
acne lesions: inammatory mediators released into the
skin; alteration of the keratinisation process leading to
comedones; increased and altered sebum production
under androgen control (or increased androgen receptor
sensitivity); and follicular colonisation by P acnes.27 The
exact sequence of events and how they and other factors
interact remains unclear.
www.thelancet.com Vol 379 January 28, 2012

Seminar

Immune-mediated inammatory processes might


involve CD4+ lymphocytes and macrophages that
stimulate the pilosebaceous vasculature precede follicular
hyperkeratinisation.54 Defective terminal keratinocyte
dierentiation leads to comedo formation under the
inuence of androgens and qualitative changes in the
sebum lipids that induce interleukin 1 (IL1) secretion.55
Sebaceous glands are an important part of the innate
immune system, producing a variety of antimicrobial
peptides, neuropeptides, and antibacterial lipids such as
sapienic acid. Each sebaceous gland functions like an
independent endocrine organ inuenced by corticotropinreleasing hormone, which might mediate the link
between stress and acne exacerbations.56 Vitamin D also
regulates sebum production, and insulin-like growthfactor 1 might increase sebum through sterol-responseelement-binding proteins.57 Oxidised lipids such as
squalene can stimulate keratinocyte proliferation and
other inammatory responses mediated by the
proinammatory leukotriene B4.58 Matrix metalloproteinases in sebum have an important role in
inammation, cell proliferation, degradation of the
dermal matrix, and treatment responsiveness.59
Sebaceous follicles containing a microcomedone
provide an anaerobic and lipid-rich environment in
which P acnes ourishes.60 Lipogenesis is directly
augmented by P acnes.61 Colonisation of facial follicles
with P acnes follows the asynchronous initiation of
sebum production,22 which might explain why treatment
with isotretinoin treatment too early can need to be
followed up with subsequent courses, as new previously
P acnes-naive follicles become colonised and inamed.
Unique P acnes strains with dierent bacterial resistance
proles colonise dierent pilosebaceous units and
induce inammation by the activation of toll-like
receptors in keratinocytes and macrophages.62 In-vitro
work suggests that P acnes could behave like a biolm
within follicles, leading to decreased response to
antimicrobial agents.63 P acnes resistance to commonly
used oral antibiotics for acne aects treatment response,
suggesting that direct antimicrobial eects might be
important in addition to the anti-inammatory actions
of antibiotics.64

How does acne aect people?


Acne results in physical symptoms such as soreness,
itching, and pain, but its main eects are on quality of life.
Psychological morbidity is not a trivial problem,65 and it is
compounded by multiple factors: acne aects highly
visible skina vital organ of social display; popular culture
and societal pressures dictate blemishless skin; acne can
be dismissed by health-care professionals as a trivial selflimiting condition; and acne peaks in teenage years, a
time crucial for building condence and self-esteem.
Case-control and cross-sectional studies assessing the
eect of acne on psychological health found a range of
abnormalities including depression, suicidal ideation,
www.thelancet.com Vol 379 January 28, 2012

anxiety, psychosomatic symptoms, shame, embarrassment, and social inhibition,66 which improve with eective
treatment.67 Anger inversely correlates with quality of life
in acne and satisfaction with acne treatment.68 Patients
might not volunteer depressive symptoms and need
prompting during consultation. UK teenagers with acne
twice as often scored in the borderline or abnormal range
on an age-appropriate validated questionnaire of emotional
wellbeing than did those who did not have acne, and had
higher levels of behavioural diculties.69 The presence of
acne was associated with unemployment in a case-control
study of young men and women.70 One community study
of 1417-year-old Australian students reported no
association between acne and subsequent psychological
or psychiatric morbidity, a surprising nding perhaps
explained by eective treatments or personality traits.71
Acne severity and degree of psychological impairment
do not necessarily correspondmild disease in one
person can cause high degrees of psychological disability,
whereas another with more severe disease can seem less
bothered by their acne.12 Most studies assessing
psychological morbidity in acne have been cross-sectional,
and therefore unable to establish causal direction. Few
studies report the direct and indirect costs of acne.72,73

How can acne be managed?


Skin hygiene
There is no good evidence that acne is caused or cured by
washing.46 Antibacterial skin cleansers might benet
mild acne, and acidic cleansing bars are probably better
than standard alkaline soaps. However, excessive washing
and scrubbing removes oil from the skin surface, drying
it and stimulating more oil production. Antibacterial skin
cleansers provide no additional benet to patients already
using other, potentially irritating topical treatments.46

Counselling and support


Spending time dispelling myths and explaining that
most treatments will not cure is worthwhile and might
improve adherence.74 Because acne treatments work by
preventing new lesions rather than treating existing ones,
an initial response might not appear for some weeks.
Most eective treatments can require months to work.75
Health-care providers should assess loss of self-esteem,
lack of condence, and symptoms of depression including
suicidal thoughts. Acnes emotional eect might not be
immediately evident or volunteered, but even mild acne
can cause signicant distress. Patients should also be
told that online acne information, including from some
support groups, varies in quality and can reect sponsor
bias, and clinicians have a role in guiding them to
trustworthy resources.

Treatment guidelines
The many over-the-counter and prescription treatments
for acne allow for a large number of potential combination treatments. A comprehensive systematic review
363

Seminar

Sebum excretion

Keratinisation

Inammation
Follicular
Proprionibacterium
acnes

Benzoyl peroxide

(+)

+++

(+)

Retinoids

++

(+)

Clindamycin

(+)

++

Antiandrogens

++

Azelaic acid

++

++

Tetracyclines

++

Erythromycin

++

Isotretinoin

+++

++

(++)

++

all are more eective than placebo, establishing the most


appropriate strategy for initial and maintenance treatment
requires further research.77,80 Topical treatments only work
where applied. Because topical therapies reduce new lesion
development they require application to the whole aected
areas, rather than individual spots. Most cause initial skin
irritation, and some people stop using them because of
this. The irritation can be minimised by starting with lower
strength preparations and gradually increasing frequency
or dose. Where irritation persists, a change in formulation
from alcoholic solutions to washes or gels to more
moisturising creams or lotions might help.

+++=very strong eect. ++=strong eect. +=moderate eect. (+)=indirect/weak eect. =no eect.

Table: Targets of acne treatments

in 1999 identied 274 trials of 140 treatments in


250 combinations.76 Most were placebo-controlled studies
of me-too products, and the authors found no basis from
controlled trials to judge the ecacy of any treatment in
relation to others, nor in the sequence of therapy. The
table shows how dierent treatment medications target
dierent aspects of acne pathology. The large number of
products and product combinations, and the scarcity of
comparative studies, has led to disparate guidelines with
few recommendations being evidence-based. Recent
acne guidelines include those from the Global Alliance
to Improve Outcomes in Acne,77 the American Academy
of Dermatology/American Academy of Dermatology
Association,78 and the European expert group on oral
antibiotics in acne.79 Because of the paucity of evidence,
these guidelines rely on the opinions of experts, many of
whom declare signicant potential conicts of interest.
Practical advice on how to manage acne based on a
systematic search of evidence by an independent team is
available in an online UK Clinical Knowledge Summary.75
All of these guidelines illustrate similar approaches on
which initial therapies should be basedie, acne severity
and whether the acne is predominantly non-inammatory
or inammatory. We propose an algorithm for treating
acne in gure 2 on the basis of our interpretation of the
clinical evidence. This interpretation diers slightly
from the Global Alliance recommendations by
suggesting slightly more initial use of topical benzoyl
peroxide than topical retinoids on the grounds of cost
and on a longer track record of ecacy and safety.
Assessment of treatment response in such a polymorphic
condition can be dicult and should include an
assessment of reduction of inammatory and noninammatory lesions in relation to baseline photographs,
plus an assessment of psychological wellbeing.

Benzoyl peroxide
Benzoyl peroxide is a safe and eective81 over-the-counter
preparation that has several mechanisms of action, and
should be applied to all the aected area.82 Single-agent
benzoyl peroxide works as well as oral antibiotics or a
topical antibiotic combination that included benzoyl
peroxide for people with mild-to-moderate facial acne.64 It
has greater activity than topical (iso)tretinoin against
inammatory lesions;83,84 the results of two further
underpowered trials were equivocal.85,86 Further studies
are needed, especially as combination therapy might be
better.86 Benzoyl peroxide causes initial local irritation.
Patients need to be counselled to expect irritation but
discontinue treatment if it becomes severe. Irritation will
decrease in most cases, especially if patients start applying
it every other day and then increase the frequency. Low
strength (25% or 5%) benzoyl peroxide is recommended,
since it is less irritating and there is no clear evidence that
stronger preparations are more eective.87

Topical retinoids
Treatment with tretinoin, adapalene, and isotretinoin
require medical prescriptions. Tazarotene is not licensed
in the UK for acne. All retinoids are contraindicated in
pregnancy, and women of childbearing age must use
eective contraception. Topical retinoids act on abnormal
keratinisation and are also anti-inammatory, so they
work for both comedonal and inammatory acne. Many
placebo-controlled or non-inferiority studies citing better
tolerability exist, but few trials guide practice. More trials
comparing retinoids against each other and against other
therapies are needed. Randomised controlled trials
(RCTs) have shown that higher-strength preparations
might have greater activity than lower-strength ones, but
at the expense of more irritation. All topical retinoids
induce local reactions, and should be discontinued if
severe. They do not seem to cause temporary worsening
of acne lesions,88 but can increase the sensitivity of skin
to ultraviolet light.

Topical treatments
Topical agents when used alone or in combination
eectively treat mild acne consisting of open and closed
comedones with a few inammatory lesions.77 The many
treatment options oer dierent modes of action. Although
364

Topical antibiotics
How topical antibiotics improve acne has not been
clearly dened, but they seem to act directly on P acnes
and reduce inammation. Topical antibiotics have less
www.thelancet.com Vol 379 January 28, 2012

Seminar

Mild acne

Mainly red spots

Mainly blackheads

Mixture of lesions

Take baseline photographs if possible to help assess subsequent treatment

Start 25% BP

Start topical retinoids

Start topical combination*

Build up over weeks until tolerance to irritation develops. If not enough benet when assessed at 68 weeks:

Add in topical retinoid


or topical antibiotic
or azelaic acid

Add in topical antibiotic


or topical BP if tolerable
or azelaic acid

Try dierent combination


product
or azelaic acid

Continue with topical therapy as long as benet continues. If not, progress to oral treatments as for moderate acne
Moderate acne
(or back acne or mild acne that fails to respond to topical therapy)

Women or older teenagers for whom


contraception needed or is acceptable

Women or older teenagers who do


not need or want contraception,
and men

Start combined oral contraceptive

Try topical combination product

Assess at 6 weeks. If no improvement then proceed to oral antibiotics plus


topical BP or retinoid (but not topical antibiotic)

Severe acne
(or moderate acne not responding to oral therapy)

If results are not good or are not sustained with the above treatments
eg, two 8-week courses of dierent oral antibiotics without signicant benet:

Proceed to oral isotretinoin early before scarring occurs (avoid wasting time
with several prolonged courses of oral antibiotics if ineective)

Counsel for adverse eects and ensure adequate contraception for women of
child-bearing potential

Assess at 6 weeks. If no improvement, try a dierent oral antibiotic plus


a topical

If there is a benecial response, carry


on for 46 months. Then stop and
use a 25% BP cream washout for
2 weeks to eradicate resistant
Propionbacterium acnes. Then try
further topicals as for mild acne as a
maintenance treatment, or if acne
relapses return to oral antibiotics.

If initial benet is lost within the


26-month period, stop oral
antibiotics and try another oral
antibiotic after a BP washout.

Figure 2: Suggested algorithm for treatment of mild, moderate, and severe acne based on our appraisal of current clinical evidence and uncertainties
Figures reproduced with permission from DermNet NZ. BP=benzoyl peroxide. *Topical combination could be benzoyl peroxide plus topical antibiotic, or topical
benzoyl peroxide plus topical retinoid.

activity than other agents against non-inamed lesions.


For more severe acne, topical antibiotics are usually
combined with other products such as topical retinoids
www.thelancet.com Vol 379 January 28, 2012

or benzoyl peroxide. Patients with back acne might


respond better to oral antibiotic therapy because of the
diculties of applying treatments to large areas that are
365

Seminar

dicult to reach. Topical antibiotics include clindamycin,


erythromycin, and tetracycline. Topical antibiotics are
also available in combination with benzoyl peroxide and
zinc acetate. Alcohol-based preparations are more
drying, and therefore more suitable for oilier skins. The
ecacy of erythromycin might be declining because of
bacterial resistance.89

Other topical therapies


Salicylic acid is an exfoliant and is a component of many
over-the-counter preparations. No studies support routine
use of salicylic acid in preference to other topical therapies.
The American Guidelines state that data from peerreviewed literature regarding the ecacy of sulphur,
resorcinol, sodium sulfacetamide, aluminium chloride,
and zinc are limited.78 Similarly there is no reliable
evidence to support the use of nicotinamide or combination
triethyl citrate and ethyl linoleate.90 Despite recent interest
in topical dapsone91 and taurine bromamine,92 neither is
licensed in the UK, and current comparative evidence
does not support a change in practice. A new vehicle,
emollient foam, containing sodium sulfacetaminde 10%
and sulphur 5% is now available for acne treatment in the
USA.93 Azelaic acid has both antimicrobial and anticomedonal properties but can cause hypopigmentation,
and darker-skinned patients should therefore be monitored
for signs. Anecdotal reports have suggested that azelaic
acid might reduce post-inammatory hyperpigmentation,
which is possibly attributable to its activity on abnormal
melanocytes. The American Guidelines note that its
clinical use, compared to other agents, has limited ecacy
according to experts.78

Combination topicals
There is accumulating information that combinations of
topical treatments with dierent mechanisms of action
work better than single agents.94 Few combinations have
been tested properly against the relevant monotherapy.
The trials tend to be methodologically awed by factors
such as suboptimal dose or frequency of monotherapy.82
Compliance can be increased with once-daily combination
products because of their convenience and faster speed
of onset,95,96 although individual generic preparations
used concomitantly might be more cost-eective.64
Benzoyl peroxide inactivates tretinoin, and the two agents
should not therefore be applied simultaneously; if used
in combination one should be applied in the morning
and one at night.

clearance rather than reduction in lesion counts. There is


no conclusive evidence that one antibiotic is more
eective than another (including rst and second
generation tetracyclines) or that oral antibiotics are more
eective than topical preparations for mild-to-moderate
facial acne.64 There is no evidence that higher doses are
more eective than lower doses or that controlled-release
preparations are necessary.64,76,82,97
The choice of antibiotic should therefore be based on
the patients preference, the side-eect prole, and cost.
The tetracyclines (tetracycline, oxytetracycline, doxycycline, or lymecycline) are the preferred options;
minocycline has signicant adverse eects.98 Cotrimoxazole should be avoided because the sulfamethoxazole component has signicant side-eects.
Quinolones are not recommended in adolescents due to
arthropathy risks and because oral ciprooxacin shows
rapid selectivity that promotes resistance.99 Aminoglycosides and chloramphenicol have very limited
eects79 and oral clindamycin, although eective, has
the potential for signicant adverse eects such as
pseudomembranous colitis. There is increasing
resistance to the macrolides (erythromycin and
azithromycin) and trimethoprim that is causing
worldwide concern.
The use of antibiotics for acne has been questioned
owing to resistance concerns, especially since they are
used for long periods at low doses.100 Concomitant
benzoyl peroxide can reduce problems with bacterial
resistance,101 whereas concomitant treatment with
dierent oral and topical antibiotics should be avoided.
Data from a large well-reported RCT indicated that
68 weeks is an appropriate time to assess response.64 If
an individual does not respond to antibiotics or stops
responding, there is no evidence that increasing the
frequency or dose is helpful. Such strategies increase
selective pressure without increasing ecacy.82
Antibiotics should be stopped if no further improvement
is evident. Antibiotics should not be routinely used for
maintenance because alternatives exist with similar
ecacy and preventative action.79,82 Benzoyl peroxide
protects against resistance by eliminating resistant
bacteria: the Global Alliance to Improve Outcomes in
acne (2003) recommends that if antibiotics must be
used for longer than 2 months, benzoyl peroxide should
be used for a minimum of 57 days between antibiotic
courses to reduce resistant organisms from the skin.77

Oral contraceptives

Oral treatments
Oral antibiotics
Oral antibiotics are usually reserved for more severe
acne, acne predominantly on the trunk, acne unresponsive
to topical therapy, and in patients at greater risk of
scarring. Although antibiotics have shown eectiveness
in terms of reducing the number of inammatory lesions,
none clear acne completely. Most patients seek acne
366

Combined oral contraceptives (COCs) contain an


oestrogen (ethinylestradiol) and a progestogen. COCs are
frequently prescribed for women with acne because
oestrogen suppresses sebaceous gland activity and
decreases the formation of ovarian and adrenal
androgens. Progestogen-only contraceptives often
worsen acne and should be avoided in women who
have no contraindications to oestrogen-containing
www.thelancet.com Vol 379 January 28, 2012

Seminar

preparations.102 Progestogens bind to both progesterone


and androgen receptors and their androgenic eects are
dependent on the type and dose of progestogen.
Third-generation progestogens such as desogestrel,
norgestimate, and gestodene bind more selectively to the
progesterone receptor than do the second-generation
progestogens (eg, levonorgestrel and norethisterone), but
at the cost of an increased risk of thromboembolism.
The Global Alliance Guidelines state that hormonal
therapy is an excellent choice for women who need oral
contraception and that it should be used as a component
for combination therapy in women with or without
endocrine abnormalities.77 Hormonal therapy should be
used early in women with moderate-to-severe acne, or in
those with seborrhoea, acne, hirsutism, and alopecia
symptoms. A Cochrane review found few important
dierences between dierent combined oral contraceptive
types in their eectiveness for treating acne, and how
they compare with alternative acne treatments is not
clear.102 Although preparations containing cyproterone
acetate have been traditionally used for acne treatment,
there is little evidence to show its superiority over other
progestins. The same applies to the antiandrogen actions
of spironolactone.103

Oral isotretinoin
When given for around 20 weeks, oral isotretinoin is the
most eective medication resulting in clinical cure in
around 85% of cases.76,94,104 Relapse rates are around 21%
and are dose-dependent, the best responses being seen
with daily doses of 1 mg/kg per day or a total of 150 mg/kg
over the treatment duration. Isotretinoin is usually
reserved for severe nodulocystic scarring acne or acne
resistant to other therapies (gure 3). Research is needed
to investigate whether isotretinoin could be benecial if
used sooner for moderate cases. Isotretinoin causes
cheilitis, dry skin, nose bleeds, secondary infection,
temporary worsening of lesions, photosensitivity, and
increased serum lipids, but these are rarely severe
enough to cause treatment withdrawal.105 Other less
common side-eects might include an increased risk of
ulcerative colitis.106 Due to teratogenicity, isotretinoin
should be given with adequate contraception for women
of childbearing age. Strict regulation in the USA has
decreased legal isotretinoin prescriptions and increased
illegal buying over the internet.107 A possible link between
isotretinoin and depression is discussed later.

Complementary and alternative medications


(CAMs)
The use of CAMs for acne is widespread. A systematic
review of CAM treatments for acne in 2006 identied
15 RCTs covering diverse approaches such as Aloe vera,
pyridoxine, fruit-derived acids, kampo (Japanese herbal
medicine), and ayurvedic herbal treatments.108 Although
mechanisms of potential benet for some of the CAM
therapies were biologically plausible, the included
www.thelancet.com Vol 379 January 28, 2012

Figure 3: Before (A) and after (B) view of a woman with severe acne treated with a course of isotretinoin
Reproduced with permission from Amy Derick; full patient consent was received.

studies were generally of poor quality and inconclusive.


Another systematic review found some benet for
acupuncture with moxibustion, but the quality of
included studies was limited.109 A systematic review of
four RCTs of tea-tree oil in 2000 did not nd conclusive
evidence of benet,110 although a recent well-reported
study of 60 people in Iran with mild-to-moderate acne
showed a modest reduction in lesion count and few
local adverse eects when compared with placebo,
suggesting that larger trials might be worthwhile.111
CAM cannot be recommended for acne treatment
because it is not supported by good evidenceCAMs
might work, but the key studies have not been done, or
when done, they have been inconclusive or reported
poorly. CAM therapy for acne is a research gap that
needs to be addressed given the high degree public
interest and spending on CAM approaches.

Special clinical problems


The depth and extent of acne scarring varies and can be
improved by multiple procedures including subcision,
punch excision, laser resurfacing, dermabrasion, and
chemical peels.27,112 Increasingly acne scarring is being
treated with fractionated laser treatmentsa technique
that produces thousands of microthermal areas of
dermal ablation separated by areas of untreated skin,
with fewer side-eects and a quicker healing period than
ablative lasers.113
Whereas open comedones can often be extracted with
minimal skin trauma, cysts and closed comedones
provide more challenging targets for acne surgery. Closed
comedones can be nicked with a bevelled needle before
expression with a comedone extractor, and large closed
comedones can be treated with electrocautery or laser.114,115
Injection of intralesional steroid (01 mL of 5 mg/mL
367

Seminar

triamcinolone acetonide) into cysts often rapidly improves


their appearance without extrusion.116
Body dysmorphic disorder (dysmorphophobia) is
dened as signicant psychosomatic distress caused by
imagined or minor defects in ones appearance. Roughly
14% of patients with acne have sucient distress related
to their facial appearance to be diagnosed with dysmorphophobia.117 It is notoriously dicult to treat, but
aggressive treatment of residual acne and cognitive
behavioural therapy can help.118,119
Severe acne with fever remains an important entity to
recognise and treat early to prevent extreme scarring and
patient suering.120,121
With the exception of avoidance of tetracyclines,
isotretinoin, and hormonal treatments, management of
prepubertal acne is similar to adult acne management
and is reviewed elsewhere.8,122
Trying to persuade teenagers to persist for many
months or years with potentially irritating topical
treatments that only prevent a proportion of new lesions
occurring is a challenge. Many teenagers are more
preoccupied by dealing with large spots as they appear,
prompting a range of home remedies, such as toothpaste,
publicised on social internet sites such as YouTube. The
use of social internet sites as a means of targeting
treatments for acne is potentially interesting.123 It is
possible that digital photography will be increasingly
used by physicians to manage concordance of acne
patients from home.124

that a rare idiosyncratic psychological reaction to


isotretinoin does occur,130 especially since there are
plausible biological mechanisms by which retinoids
might induce psychopathology.128 The picture is a complex
one as depression and suicidal ideation occur with severe
acne in the absence of isotretinoin treatment.131
A retrospective cohort study in Sweden found that
attempted suicide was increased in those taking
isotretinoin, although an increased risk was also present
before treatment. An increased risk of attempted suicide
was present 6 months after isotretinoin, which suggests
that patients should be monitored for suicidal behaviour
after treatment has ended.132 The generic form of
isotretinoin continues to be available.

Lasers, light sources, and photodynamic therapy


Two systematic reviews of 16 and 25 trials, respectively,133,134
assessed various forms of light sources including
photodynamic therapy, infrared lasers, broad-spectrum
light sources, pulsed dye lasers, intense pulsed light, and
potassium titanyl phosphate laser. Both reviews concluded
that optical treatments can improve inammatory acne in
the short-term, with the most consistent outcomes for
photodynamic therapy. Pain, redness, swelling, and
increased pigmentation were common adverse eects.
Although several forms of light therapies can improve
acne initially,135 longer-term outcomes and comparisons
with conventional acne therapies are needed.

Antibiotic resistance and other experimental therapies

Areas of controversy and uncertainty


Retinoid safety
Topical retinoids, a rst-line acne therapy in the USA,
have been associated with increased deaths in older
male veteran patients in a randomised controlled trial
of actinic keratosis.125 Although this nding has been
ascribed to chance, informing all topical retinoid users
of these results might be warranted until further data
are obtained.126 A branded form of oral isotretinoin
(Accutane) was introduced in the USA in 1982 and has
been used by more than 13 million patients, but has
now lost more than 95% of the market share and is
being discontinued by its manufacturer Roche
Pharmaceuticals.127 The business decision is based on
the high cost of defending the drug maker from
personal injury lawsuitsinitially suits alleged that
Accutane was associated with depression and suicide.
More recently at least 500 suits have been led alleging
that Accutane causes inammatory bowel disease.127
These disputed associations remain ripe areas for
future research.128
A systematic review of isotretinoin use and depression
and suicidal behaviour published in 2005 did not nd
any evidence to support the notion that depression
worsened after treatment, and some studies showed that
depression scores improved on treatment, although all
nine included studies had limitations.129 It is still possible
368

Concerns regarding the rise of antibiotic resistance have


increased the urgency for developing eective nonantibiotic therapies. Although two trials of subantimicrobial dosing (ie, the prescription of low doses that
are anti-inammatory but not antimicrobial) have shown
ecacy for lower doses,136,137 the studies are too small to
make reliable estimates of bacterial resistance that could
be promoted by the lower doses used. Relatively
inexpensive hand-held home lasers and heating devices
have been developed. Randomised controlled equivalency
trials of these new devices using patient-centred metrics
are urgently needed. In the more distant future,
vaccination with killed P acnes and sialidase-based
vaccines holds some promise.138

Natural history
Longitudinal studies that document the natural history
of acne are needed, especially with a view to identifying
risk factors for persistent disease. It is unknown whether
early treatment (eg, at prepubertal years) can alter the
natural history of P acnes colonisation and subsequent
inammatory acne.

So many treatments of unknown comparative ecacy


Treatment decisions for patients with acne and doctors
are compounded by the profusion of available treatments, most of which have been introduced through
www.thelancet.com Vol 379 January 28, 2012

Seminar

Panel: Important developments in understanding acne


and its treatment
Acne is a chronic disease that can persist into adulthood
Acne causes signicant psychological morbidity
Immune-mediated inammatory changes precede
follicular hyperkeratinisation and Propionibacterium acnes
colonisation
The possible association between acne and diet remains
uncertain
Comparative eectiveness research could help reduce the
plethora of current theraputic options for initiation and
maintenance treatment
Prolonged use of oral antibiotics might contribute to
bacterial resistance in the community
Oral isotretinoin results in signicant clearing of acne, but
it is limited by teratogenicity and other side-eects

10
11

12

13
14

15

16
17
18

19

placebo-controlled trials. Despite occasional exceptions,64


the absence of trials with active comparators is a
signicant handicap to shared clinical decision making.
Clinical trials of the cost-eectiveness of dierent
strategies for initial treatment and maintenance therapy
of acne are needed. Almost half of recently published
acne trials contain serious aws that could be overcome
by better reporting.139 The lack of agreement on suitable
outcome measures also hampers secondary research.140
Treatment uncertainties for acne are summarised in the
UK Database of Uncertainties about the Eects of
Treatments.141 Comparative-eectiveness research on
acne therapy has been targeted as a top 100 priority in the
USA by the Institute of Medicine. The panel summarises
key developments in understanding acne.

20
21

22

23

24

25

Contributors
All authors contributed to searching published works and took part in
writing the rst and subsequent drafts.

26

Conicts of interest
We declare that we have no conicts of interest.

27

References
1
Grindlay D, Williams HC. Systematic reviews on acne vulgaris
since 1999mapping by topic. http://www.library.nhs.uk/skin/
ViewResource.aspx?resID=280039&tabID=289&catID=8275
(accessed Sept 13, 2010).
2
Degitz K, Placzek M, Borelli C, Plewig G. Pathophysiology of acne.
J Dtsch Dermatol Ges 2007; 5: 31623.
3
Shalita AR. Acne: clinical presentations. Clin Dermatol 2004;
22: 38586.
4
Jacob CI, Dover JS, Kaminer MS. Acne scarring: a classication
system and review of treatment options. J Am Acad Dermatol 2001;
45: 10917.
5
van Zuuren EJ, Gupta AK, Gover MD, Graber M, Hollis S. Systematic
review of rosacea treatments. J Am Acad Dermatol 2007; 56: 10715.
6
Rosas Vazquez E, Campos Macias P, Ochoa Tirado JG,
Garcia Solana C, Casanova A, Palomino Moncada JF. Chloracne
in the 1990s. Int J Dermatol 1996; 35: 64345.
7
Roseneld RL. What every physician should know about polycystic
ovary syndrome. Dermatol Ther 2008; 21: 35461.
8
Lucky AW. A review of infantile and pediatric acne. Dermatology
1998; 196: 9597.
9
Meixner D, Schneider S, Krause M, Sterry W. Acne inversa.
J Dtsch Dermatol Ges 2008; 6: 18996.

www.thelancet.com Vol 379 January 28, 2012

28
29
30

31
32

33

34
35

Valeyrie-Allanore L, Sassolas B, Roujeau JC. Drug-induced skin,


nail and hair disorders. Drug Saf 2007; 30: 101130.
Rademaker M, Garioch JJ, Simpson NB. Acne in schoolchildren:
no longer a concern for dermatologists. Br Med J 1989;
298: 121719.
Law MP, Chuh AA, Lee A, Molinari N. Acne prevalence and
beyond: acne disability and its predictive factors among Chinese
late adolescents in Hong Kong. Clin Exp Dermatol 2010; 35: 1621.
Yahya H. Acne vulgaris in Nigerian adolescentsprevalence, severity,
beliefs, perceptions, and practices. Int J Dermatol 2009; 48: 498505.
Ghodsi SZ, Orawa H, Zouboulis CC. Prevalence, severity,
and severity risk factors of acne in high school pupils:
a community-based study. J Invest Dermatol 2009; 129: 213641.
Stathakis V, Kilkenny M, Marks R. Descriptive epidemiology
of acne vulgaris in the community. Australas J Dermatol 1997;
38: 11523.
Menon C, Gipson K, Bowe WP, Hostad OJ, Margolis DJ. Validity
of subject self-report for acne. Dermatology 2008; 217: 16468.
Collier CN, Harper JC, Cafardi JA, et al. The prevalence of acne
in adults 20 years and older. J Am Acad Dermatol 2008; 58: 5659.
Poli F, Dreno B, Verschoore M. An epidemiological study of acne
in female adults: results of a survey conducted in France.
J Eur Acad Dermatol Venereol 2001; 15: 54145.
Schafer T, Niehnaus A, Vieluf D, Berger J, Ring J. Epidemiology of
acne in the general population: the risk of smoking. Br J Dermatol
2001; 145: 100104.
Cunlie WJ, Gould DJ. Prevalence of facial acne vulgaris in late
adolescence and in adults. BMJ 1979; 1: 110910.
Friedlander SF, Eicheneld LF, Fowler JF Jr, Fried RG, Levy ML,
Webster GF. Acne epidemiology and pathophysiology.
Semin Cutan Med Surg 2010; 29: 24.
Mourelatos K, Eady EA, Cunlie WJ, Clark SM, Cove JH. Temporal
changes in sebum excretion and propionibacterial colonization in
preadolescent children with and without acne. Br J Dermatol 2007;
156: 2231.
Stoll S, Shalita AR, Webster GF, Kaplan R, Danesh S, Penstein A.
The eect of the menstrual cycle on acne. J Am Acad Dermatol 2001;
45: 95760.
Yosipovitch G, Tang M, Dawn AG, et al. Study of psychological
stress, sebum production and acne vulgaris in adolescents.
Acta Derm Venereol 2007; 87: 13539.
Cheng CE, Irwin B, Mauriello D, Liang L, Pappert A, Kimball AB.
Self-reported acne severity, treatment, and belief patterns across
multiple racial and ethnic groups in adolescent students.
Pediatr Dermatol 2010; 27: 44652.
Gollnick HP, Finlay AY, Shear N. Global alliance to improve
outcomes in acne. Can we dene acne as a chronic disease? If so,
how and when? Am J Clin Dermatol 2008; 9: 27984.
Thiboutot D, Gollnick H, Bettoli V, et al. Global Alliance to Improve
Outcomes in Acne. New insights into the management of acne:
an update from the Global Alliance to Improve Outcomes in Acne
group. J Am Acad Dermatol 2009; 60: S150.
Downie MM, Sanders DA, Kealey T. Modelling the remission of
individual acne lesions in vitro. Br J Dermatol 2002; 147: 86978.
Cunlie WJ, Shuster S. Pathogenesis of acne. Lancet 1969; ii: 68587.
Seira H, Farnaghi F, Vasheghani-Farahani A, et al. Assessment
of androgens in women with adult-onset acne. Int J Dermatol 2007;
46: 118891.
Ballanger F, Baudry P, NGuyen JM, Khammari A, Drno B.
Heredity: a prognostic factor for acne. Dermatology 2006; 212: 14549.
Wang F, Kwak HS, Elbuluk N, et al. Retinoic acid 4-hydroxylase
inducibility and clinical response to isotretinoin in patients with
acne. J Am Acad Dermatol 2009; 61: 25258.
Bataille V, Snieder H, MacGregor AJ, Sasieni P, Spector TD. The
inuence of genetics and environmental factors in the pathogenesis
of acne: a twin study of acne in women. J Invest Dermatol 2002;
119: 31722.
Wei B, Pang Y, Zhu H, et al. The epidemiology of adolescent acne
in North East China. J Eur Acad Dermatol Venereol 2010; 24: 95357.
Lucky AW, Biro FM, Huster GA, Leach AD, Morrison JA,
Ratterman J. Acne vulgaris in premenarchal girls. An early sign
of puberty associated with rising levels of dehydroepiandrosterone.
Arch Dermatol 1994; 130: 30814.

369

Seminar

36

37

38
39

40
41

42
43

44
45

46

47
48
49

50
51

52

53

54

55

56

57
58

59

60
61

370

Do JE, Cho SM, In SI, Lim KY, Lee S, Lee ES. Psychosocial aspects
of acne vulgaris: a community-based study with Korean adolescents.
Ann Dermatol 2009; 21: 12529.
Hogewoning AA, Koelemij I, Amoah AS, et al. Prevalence and risk
factors of inammatory acne vulgaris in rural and urban Ghanaian
schoolchildren. Br J Dermatol 2009; 161: 47577.
Mills CM, Peters TJ, Finlay AY. Does smoking inuence acne?
Clin Exp Dermatol 1993; 18: 10001.
Klaz I, Kochba I, Shohat T, Zarka S, Brenner S. Severe acne vulgaris
and tobacco smoking in young men. J Invest Dermatol 2006;
126: 174952.
Ehrmann DA. Polycystic ovary syndrome. N Engl J Med 2005;
352: 122336.
Chen MJ, Chen CD, Yang JH, et al. High serum
dehydroepiandrosterone sulfate is associated with phenotypic acne
and a reduced risk of abdominal obesity in women with polycystic
ovary syndrome. Hum Reprod 2011; 26: 22734.
Plewig G, Fulton JE, Kligman AM. Pomade acne. Arch Dermatol
1970; 101: 58084.
Melnik B, Jansen T, Grabbe S. Abuse of anabolic-androgenic
steroids and bodybuilding acne: an underestimated health problem.
J Dtsch Dermatol Ges 2007; 5: 11017.
Tucker SB. Occupational tropical acne. Cutis 1983; 31: 7981.
Green J, Sinclair RD. Perceptions of acne vulgaris in nal year
medical student written examination answers. Aust J Dermatol 2001;
42: 98101.
Magin P, Pond D, Smith W, Watson A. A systematic review
of the evidence for myths and misconceptions in acne
management: diet, face-washing and sunlight. Fam Pract 2004;
22: 6270.
Spencer EH, Ferdowsian HR, Barnard ND. Diet and acne: a review
of the evidence. Int J Dermatol 2009; 48: 33947.
Grant JD, Anderson PC. Chocolate and acne: a dissenting view.
Missouri Med 1965; 62: 45960.
Cordain L, Lindeberg S, Hurtado M, Hill K, Eaton SB,
BrandMiller J. Acne vulgaris: a disease of Western civilization
[comment]. Arch Dermatol 2002; 138: 158490.
Thiboutot DM, Strauss JS. Diet and acne revisited. Arch Dermatol
2002; 138: 159192.
Abulnaja KO. Changes in the hormone and lipid prole of obese
adolescent Saudi females with acne vulgaris. Braz J Med Biol Res
2009; 42: 50105.
Smith RN, Mann NJ, Braue A, Makelainen H, Varigos GA.
A low-glycemic-load diet improves symptoms in acne vulgaris
patients: a randomized controlled trial. Am J Clin Nutr 2007;
86: 10715.
Tsai MC, Chen W, Cheng YW, Wang CY, Chen GY, Hsu TJ. Higher
body mass index is a signicant risk factor for acne formation in
schoolchildren. Eur J Dermatol 2006; 16: 25153.
Jeremy AH, Holland DB, Roberts SG, Thomson KF, Cunlie WJ.
Inammatory events are involved in acne lesion initiation,
J Invest Dermatol 2003; 121: 2027.
Kurokawa I, Danby FW, Ju Q, et al. New developments in our
understanding of acne pathogenesis and treatment. Exp Dermatol
2009; 18: 82132.
Zouboulis CC, Bohm M. Neuroendocrine regulation of sebocytes
a pathogenetic link between stress and acne. Exp Dermatol 2004;
13: 3135.
Zouboulis CC, Baron JM, Bohm M, et al. Frontiers in sebaceous
gland biology and pathology. Exp Dermatol 2008; 17: 54251.
Alestas T, Ganceviciene R, Fimmel S, Muller-Decker K,
Zouboulis CC. Enzymes involved in the biosynthesis of
leukotriene B4 and prostaglandin E2 are active in sebaceous glands.
J Mol Med 2006; 84: 7587.
Papakonstantinou E, Aletras AJ, Glass E, et al. Matrix
metalloproteinases of epithelial origin in facial sebum of patients
with acne and their regulation by isotretinoin. J Invest Dermatol
2005; 125: 67384.
Brown SK, Shalita AR. Acne vulgaris. Lancet 1998; 351: 187176.
Iinuma K, Sato T, Akimoto N, et al. Involvement of
propionibacterium acnes in the augmentation of lipogenesis in
hamster sebaceous glands in vivo and in vitro. J Invest Dermatol
2009; 129: 211319.

62

63

64

65
66

67

68

69

70
71

72

73
74
75

76

77

78
79

80

81

82
83
84

85

86

Wang KC, Zane LT. Recent advances in acne vulgaris research:


insights and clinical implications. Adv Dermatol 2008;
24: 197209.
Coenye T, Honraet K, Rossel B, Nelis HJ. Biolms in skin
infections: Propionibacterium acnes and acne vulgaris.
Infect Disord Drug Targets 2008; 8: 15659.
Ozolins M, Eady EA, Avery AJ, et al. Comparison of ve
antimicrobial regimens for treatment of mild to moderate
inammatory facial acne vulgaris in the community: randomised
controlled trial. Lancet 2004; 364: 218895.
Ayer J, Burrows N. Acne: more than skin deep. Postgrad Med J 2006
Aug; 82: 50006.
Kubota Y, Shirahige Y, Nakai K, Katsuura J, Moriue T, Yoneda K.
Community-based epidemiological study of psychosocial eects
of acne in Japanese adolescents. J Dermatol 2010; 37: 61722.
Hahm BJ, Min SU, Yoon MY, et al. Changes of psychiatric
parameters and their relationships by oral isotretinoin in acne
patients. J Dermatol 2009; 36: 25561.
Rapp DA, Brenes GA, Feldman SR, et al. Anger and acne:
implications for quality of life, patient satisfaction and clinical care.
Br J Dermatol 2004; 151: 18389.
Smithard A, Glazebrook C, Williams HC. Acne prevalence,
knowledge about acne and psychological morbidity in
mid-adolescence: a community-based study. Br J Dermatol 2001;
145: 27479.
Cunlie WJ. Acne and unemployment. Br J Dermatol 1986;
115: 38687.
Magin P, Pond C, Smith W, Goode S. Acnes relationship with
psychiatric and psychological morbidity: results of a school-based
cohort study of adolescents. J Eur Acad Dermatol Venereol 2009;
24: 5862.
Inglese MJ, Fleischer AB Jr, Feldman SR, Balkrishnan R. The
pharmacoeconomics of acne treatment: where are we heading?
J Dermatolog Treat 2008; 19: 2737.
Lehmann HP, Andrews JS, Robinson KA, Holloway VL, Goodman SN.
Management of acne. Evid Rep Technol Assess 2001; 17: 13.
Thiboutot D, Drno B, Layton A. Acne counseling to improve
adherence. Cutis 2008; 81: 8186.
National Institute of health and Clinical Excellence. Clinical
knowledge summary: acne vulgaris. http://www.cks.nhs.uk/acne_
vulgaris (accessed Sept 15, 2010).
Lehmann HP, Andrews JS, Robinson KA, Holloway VL,
Goodman SN. Management of acne. US Agency for Healthcare
Research and Quality (AHRQ) Evidence Report/Technology
Assessment Number 17; Sept 2001.
Gollnick H, Cunlie W, Berson D, et al. Management of acne:
a report from a Global Alliance to Improve Outcomes in Acne.
J Am Acad Dermatol 2003; 49: S137.
Strauss JS, Krowchuck DP, Leyden JJ, et al. Guidelines of care for
acne vulgaris management. J Am Acad Dermatol 2007; 56: 65163.
Drno B, Bettoli V, Ochsendorf F, et al. European
recommendations on the use of oral antibiotics for acne.
Eur J Dermatol 2004; 14: 39199.
Thielitz A, Sidou F, Gollnick H. Control of microcomedone
formation throughout a maintenance treatment with adapalene gel,
0.1%. J Eur Acad Dermatol Venereol. 2007; 21: 74753.
Food and Drug Administration, HHS. Classication of benzoyl
peroxide as safe and eective and revision of labelling to drug facts
format; topical acne drug products for over-the-counter human use;
nal rule. Fed Regist 2010; 75: 976777.
Garner S. Acne vulgaris. In: Williams H, ed. Evidence-based
dermatology. London: BMJ Books, 2003, 87114.
Lyons RE. Comparative eectiveness of benzoyl peroxide
and tretinoin in acne vulgaris. Int J Dermatol 1978; 17: 24651.
Hughes BR, Norris JF, Cunlie WJ. A double-blind evaluation of
topical isotretinoin 005%, benzoyl peroxide gel 5% and placebo in
patients with acne. Clin Exp Dermatol 1992; 17: 16568.
Bucknall JH, Murdoch PN. Comparison of tretinoin solution
and benzoyl peroxide lotion in the treatment of acne vulgaris.
Curr Med Res Opin 1977; 5: 26668.
Handojo I. Retinoic acid cream (Airol cream) and benzoyl-peroxide
in the treatment of acne vulgaris.
Southeast Asian J Trop Med Public Health 1979; 10: 54851.

www.thelancet.com Vol 379 January 28, 2012

Seminar

87

88
89
90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107
108

109

Fakhouri T, Yentzer BA, Feldman SR. Advancement in benzoyl


peroxide-based acne treatment: methods to increase both ecacy
and tolerability. J Drugs Dermatol 2009; 8: 65761.
Yentzer BA, McClain RW, Feldman SR. Do topical retinoids cause
acne to are? J Drugs Dermatol 2009; 8: 799801.
Simonart T, Dramaix M. Treatment of acne with topical antibiotics:
lessons from clinical studies. Br J Dermatol 2005; 153: 395403.
Charakida A, Charakida M, Chu AC. Double-blind, randomized,
placebo-controlled study of a lotion containing triethyl citrate and
ethyl linoleate in the treatment of acne vulgaris. Br J Dermatol 2007;
157: 56974.
Stotland M, Shalita AR, Kissling RF. Dapsone 5% gel: a review of its
ecacy and safety in the treatment of acne vulgaris.
Am J Clin Dermatol 2009; 10: 22127.
Marcinkiewicz J, Wojas-Pelc A, Walczewska M, et al. Topical taurine
bromamine, a new candidate in the treatment of moderate
inammatory acne vulgaris: a pilot study. Eur J Dermatol 2008;
18: 43339.
Del Rosso JQ. The use of sodium sulfacetaminde 10%-sulfur 5%
emollient foam in the treatment of acne vulgaris.
J Clin Aesthetic Dermatol 2009; 2: 269.
Thiboutot D, Zaenglein A, Weiss J, Webster G, Calvarese B, Chen D.
An aqueous gel xed combination of clindamycin phosphate 12%
and benzoyl peroxide 25% for the once-daily treatment of moderate
to severe acne vulgaris: assessment of ecacy and safety in
2813 patients. J Am Acad Dermatol 2008; 59: 792800.
Yentzer BA, Ade RA, Fountain JM, et al. Simplifying regimens
promotes greater adherence and outcomes with topical acne
medications: a randomized controlled trial. Cutis 2010; 86: 10308.
Gollnick HP, Draelos Z, Glenn MJ, et al. Adapalene-benzoyl
peroxide, a unique xed-dose combination topical gel for the
treatment of acne vulgaris: a transatlantic, randomized,
double-blind, controlled study in 1670 patients. Br J Dermatol 2009;
161: 118089.
Simonart T, Dramaix M, De Maertelaer V. Ecacy of tetracyclines
in the treatment of acne vulgaris: a review. Br J Dermatol 2008;
158: 20816.
Garner SE, Eady EA, Popescu C, Newton J, Li WA. Minocycline for
acne vulgaris: ecacy and safety. Cochrane Database Syst Rev 2003;
1: CD002086.
Hoiby N, Jarlov JO, Kemp M, et al. Excretion of ciprooxacin in
sweat and multiresistant Staphylococcus epidermidis. Lancet 1997;
349: 16769.
Eady AE, Cove JH, Layton AM. Is antibiotic resistance in cutaneous
propionibacteria clinically relevant?: Implications of resistance for
acne patients and prescribers. Am J Clin Dermatol 2003; 4: 81331.
Patel M, Bowe WP, Heughebaert C, Shalita AR. The development
of antimicrobial resistance due to the antibiotic treatment of acne
vulgaris: a review. J Drugs Dermatol 2010; 9: 65564.
Arowojolu AO, Gallo MF, Lopez LM, Grimes DA, Garner SE.
Combined oral contraceptive pills for treatment of acne.
Cochrane Database Syst Rev 2009; 3: CD004425.
Brown J, Farquhar C, Lee O, Toomath R, Jepson RG. Spironolactone
versus placebo or in combination with steroids for hirsutism and/or
acne. Cochrane Database Syst Rev 2009; 15: CD000194.
Wessels F, Anderson AN, Kropman K. The cost-eectiveness of
isotretinoin in the treatment of acne. Part 1. A meta-analysis of
eectiveness literature. S Afr Med J 1999; 89: 78084.
Rademaker M. Adverse eects of isotretinoin: a retrospective review
of 1743 patients started on isotretinoin. Australas J Dermatol 2010;
51: 24853.
Crockett SD, Porter CQ, Martin CF, Sandler RS, Kappelman MD.
Isotretinoin use and the risk of inammatory bowel disease:
a case-control study. Am J Gastroenterol 2010; 105: 198693.
US Food and Drug Administration. Protecting yourself. http://
www.fda.gov/buyonline/accutane (accessed Feb 12, 2010).
Magin PJ, Adams J, Pond CD, Smith W. Topical and oral CAM in
acne: a review of the empirical evidence and a consideration of its
context. Complement Ther Med 2006; 14: 6276.
Li B, Chai H, Du YH, Xiao L, Xiong J. Evaluation of therapeutic
eect and safety for clinical randomized and controlled trials of
treatment of acne with acupuncture and moxibustion.
Zhongguo Zhen Jiu 2009; 29: 24751 (in Chinese).

www.thelancet.com Vol 379 January 28, 2012

110 Ernst E, Huntley A. Tea tree oil: a systematic review of randomized


clinical trials. Forsch Komplementarmed Klass Naturheilkd 2000;
7: 1720.
111 Enshaieh S, Jooya A, Siadat AH, Iraji F. The ecacy of 5% topical
tea tree oil gel in mild to moderate acne vulgaris: a randomized,
double-blind placebo-controlled study.
Indian J Dermatol Venereol Leprol 2007; 73: 2225.
112 Rivera AE. Acne scarring: a review and current treatment
modalities. J Am Acad Dermatol 2008; 59: 65976.
113 Tierney EP, Kouba DJ, Hanke CW. Review of fractional
photothermolysis: treatment indications and ecacy. Dermatol Surg
2009; 35: 114561.
114 Khunger N, IADVL Task Force. Standard guidelines of care for acne
surgery. Indian J Dermatol Venereol Leprol 2008; 74: S2836.
115 Bottomley WW, Yip J, Knaggs H, Cunlie WJ. Treatment of
closed comedonescomparisons of fulguration with topical
tretinoin and electrocautery with fulguration. Dermatol 1993;
186: 25357.
116 Taub AF. Procedural treatments for acne vulgaris. Dermatol Surg
2007; 33: 100526.
117 Bowe WP, Leyden JJ, Crerand CE, Sarwer DB, Margolis DJ. Body
dysmorphic disorder symptoms among patients with acne vulgaris.
J Am Acad Dermatol 2007; 57: 22230.
118 Hull SM, Cunlie WJ, Hughes BR. Treatment of the depressed
and dysmorphophobic acne patient. Clin Exp Dermatol 1991;
16: 21011.
119 National Institute of Health and Clinical Excellence.
Obsessive-compulsive disorder: core interventions in the treatment
of obsessive-compulsive disorder and body dysmorphic disorder.
NICE Clinical Guideline 31, November 2005.
120 Chua SL, Angus JE, Ravenscroft J, Perkins W. Synovitis, acne,
pustulosis, hyperostosis, osteitis (SAPHO) syndrome and acne
fulminans: are they part of the same disease spectrum?
Clin Exp Dermatol 2009; 34: e24143.
121 Galadari H, Bishop AG, Venna SS, Sultan E, Do D, Zeltser RJ.
Synovitis, acne, pustulosis, hyperostosis, and osteitis syndrome
treated with a combination of isotretinoin and pamidronate.
Am Acad Dermatol 2009; 61: 12325.
122 Antoniou C, Dessinioti C, Stratigos AJ, Katsambas AD. Clinical and
therapeutic approach to childhood acne: an update. Pediatr Dermatol
2009; 26: 37380.
123 Vance K, Howe W, Dellavalle RP. Social internet sites as a source
of public health information. Dermatol Clin 2009; 27: 13336.
124 Bergman H, Tsai KY, Seo SJ, Kvedar JC, Watson AJ. Remote
assessment of acne: the use of acne grading tools to evaluate digital
skin images. Telemed J E Health 2009; 15: 42630.
125 Weinstock MA, Bingham SF, Lew RA, et al. Topical tretinoin
therapy and all-cause mortality. Arch Dermatol 2009; 145: 1824.
126 Schilling LM, Dellavalle RP. Dealing with unanticipated mortality
in a large randomized clinical trial of topical tretinoin.
Arch Dermatol 2009; 145: 76.
127 Winnington P. Lessons from the untimely demise of a superstar.
Pract Dermatol 2009; 6: 5.
128 Kontaxakis VP, Skourides D, Ferentinos P, Havaki-Kontaxaki BJ,
Papadimitriou GN. Isotretinoin and psychopathology: a review.
Ann Gen Psychiatry 2009; 8: 2.
129 Marqueling AL, Zane LT. Depression and suicidal behavior in acne
patients treated with isotretinoin: a systematic review.
Semin Cutan Med Surg 2005; 24: 92102.
130 Magin P, Pond D, Smith W. Isotretinoin, depression and suicide:
a review of the evidence. Br J Gen Pract 2005; 55: 13438.
131 Halvorsen J, Stern R, Dalgard F, Thoresen M, Bjertness E, Lien L.
Suicidal ideation, mental health problems, and social impairment
are increased in adolescents with acne: a population-based study.
J Invest Dermatol 2010; 131: 36370.
132 Sundstrm A, Alfredsson L, Sjlin-Forsberg G, Gerdn B,
Bergman U, Jokinen J. Association of suicide attempts with acne
and treatment with isotretinoin: retrospective Swedish cohort study.
BMJ 2010; 341: c5812.
133 Haedersdal M, Togsverd-Bo K, Wulf HC. Evidence-based review of
lasers, light sources and photodynamic therapy in the treatment of
acne vulgaris. J Eur Acad Dermatol Venereol 2008; 22: 26778.

371

Seminar

134 Hamilton FL, Car J, Lyons C, Car M, Layton A, Majeed A. Laser and
other light therapies for the treatment of acne vulgaris: systematic
review. Br J Dermatol 2009; 160: 127385.
135 Sakamoto FH, Lopes JD, Anderson RR. Photodynamic therapy for
acne vulgaris: a critical review from basics to clinical practice: part I.
Acne vulgaris: when and why consider photodynamic therapy?
J Am Acad Dermatol 2010; 63: 18393.
136 Skidmore R, Kovach R, Walker C, et al. Eects of subantimicrobialdose doxycycline in the treatment of moderate acne. Arch Dermatol
2003; 139: 45964.
137 Toossi P, Farshchian M, Malekzad F, Mohtasham N,
Kimyai-Asadi A. Subantimicrobial-dose doxycycline in the
treatment of moderate facial acne. J Drugs Dermatol 2008;
7: 114952.

372

138 Kim J. Acne vaccines: therapeutic option for the treatment of acne
vulgaris? J Invest Dermatol 2008; 128: 235354.
139 Ingram JR, Grindlay JC, Williams HC. Problems in the reporting
of acne clinical trials: a spot check from the 2009 annual evidence
update on acne vulgaris. Trials 2010; 11: 77.
140 Barratt H, Hamilton F, Car J, Lyons C, Layton A, Majeed A.
Outcome measures in acne vulgaris: systematic review.
Br J Dermatol 2009; 160: 13236.
141 UK National Health Service. UK Database of Uncertainties about
the Eects of Treatments. http://www.library.nhs.uk/DUETs/
SearchResults.aspx?searchText=acne (accessed Sept 13, 2010).

www.thelancet.com Vol 379 January 28, 2012

Seminar

Thalassaemia
Douglas R Higgs, James Douglas Engel, George Stamatoyannopoulos

Thalassaemia is one of the most common genetic diseases worldwide, with at least 60 000 severely aected individuals
born every year. Individuals originating from tropical and subtropical regions are most at risk. Disorders of
haemoglobin synthesis (thalassaemia) and structure (eg, sickle-cell disease) were among the rst molecular diseases
to be identied, and have been investigated and characterised in detail over the past 40 years. Nevertheless, treatment
of thalassaemia is still largely dependent on supportive care with blood transfusion and iron chelation. Since 1978,
scientists and clinicians in this specialty have met regularly in an international eort to improve the management of
thalassaemia, with the aim of increasing the expression of unaected fetal genes to improve the deciency in adult
-globin synthesis. In this Seminar we discuss important advances in the understanding of the molecular and cellular
basis of normal and abnormal expression of globin genes. We will summarise new approaches to the development of
tailored pharmacological agents to alter regulation of globin genes, the rst trial of gene therapy for thalassaemia, and
future prospects of cell therapy.

Introduction
Abnormalities in the structure and synthesis of the -like
and -like globin chains that form tetramers of
haemoglobin (22) lead to the most common forms of
inherited anaemias.1 In thalassaemia, there are defects in
the production of either the -like (-thalassaemia2) or
the -like (-thalassaemia3) globin chains. From the 1970s,
these diseases, which specically aect red blood cells,
were among the rst to be analysed with the use of
molecular biology. Their detailed characterisation has
established many of the general principles supporting
our understanding of human molecular genetics.4
Furthermore, research into globin genes has greatly
contributed to the understanding of how human gene
expression is activated and silenced during dierentiation
and development.58 Despite these advances, manipulation
of globin gene expression to ameliorate or potentially
cure the common disorders of these genes is not yet
possible. Every 2 years since 1978, leading research
groups have met at the Hemaglobin Switching
Conference to report and discuss progress. In this
Seminar we will summarise the understanding of the
molecular and cellular pathophysiology, epidemiology,
and management of -thalassaemia, which is the main
clinical problem in this specialty.1 We also review the
developments reported at the 17th Hemaglobin Switching
Conference in Oxford, UK, which oer renewed hope for
novel approaches to treat these disorders.

the circulation to replace embryonic red blood cells.


During fetal development, HSCs migrate to the bone
marrow, which is the site of erythropoiesis for the rest of
normal adult life. In early postnatal life, adult red blood
cells from the marrow replace the fetal cells.9,10 At all
stages of development, senescent red blood cells are
continually replaced with new blood cells. These new
cells are derived from HSCs, which dierentiate into
mature red blood cells via erythroid progenitors and
precursors (erythroblasts). For an adult to maintain a
normal red blood cell count, about 2 million to 3 million
new cells must be produced every second. For severe
forms of thalassaemia, in which many erythroblasts and
mature red blood cells are damaged, erythropoiesis can
be increased by 2030 times.11,12 However, erythropoiesis
is ineective in severe cases of thalassaemia because the
increased numbers of erythroid precursors fail to develop
into mature red blood cells.

Published Online
September 12, 2011
DOI:10.1016/S01406736(11)60283-3
Medical Research Council
Molecular Haematology Unit,
Weatherall Institute of
Molecular Medicine, and
National Institute for Health
Research Biomedical Research
Centre, John Radclie Hospital,
Oxford, UK (Prof D R Higgs MD);
Cell and Developmental Biology,
University of Michigan Medical
School, Ann Arbor, MI, USA
(Prof J D Engel MD); and
Department of Medicine and
Department of Genome
Sciences, University
of Washington Medical School,
Seattle, WA, USA
(Prof G Stamatoyannopoulos MD)
Correspondence to:
Prof Douglas R Higgs, Medical
Research Council Molecular
Haematology Unit, Weatherall
Institute of Molecular Medicine,
John Radclie Hospital,
Headington, Oxford OX3 9DS, UK
doug.higgs@imm.ox.ac.uk

Normal expression of globin genes


Changes in the sites of erythropoiesis are associated with
changes in the types of haemoglobin produced. At the
molecular level, haemoglobin synthesis is controlled by
two multigene clusters on chromosome 16 (encoding the
-like globins) and on chromosome 11 (encoding the
-like globins). In the human clusters, the genes are
arranged along the chromosome in the order by which

Production of red blood cells

Search strategy and selection criteria

To fully understand the pathophysiology and management


of thalassaemia, how red blood cells are normally
produced (erythropoiesis), and how the globin genes are
normally expressed at each stage of development should
be considered. First, a transient cohort of embryonic red
blood cells originate in the blood islands of the yolk sac.
Denitive haemapoietic stem cells (HSCs), which persist
throughout fetal and adult life, then emerge from the
ventral wall of the dorsal aorta. These cells migrate from
the ventral wall to the fetal liver and, by about 60 days of
gestation, the rst fetal red blood cells are released into

We searched PubMed using the terms thalassaemia in


combination with molecular basis or treatment or
pathophysiology. We mostly selected publications from
from June 2006, to June 2011, but did not exclude frequently
referenced and highly regarded older publications. We also
searched the reference lists of articles identied by this
search strategy and selected the most relevant ones. Review
articles and book chapters are cited to provide readers with
more details and more references than can be addressed in
this Seminar.

www.thelancet.com Vol 379 January 28, 2012

Lancet 2012; 379: 37383

373

Seminar

Globin synthesis (%)

Site of
erythropoiesis

A
Yolk sac

Liver

Spleen

Bone marrow

100

60

20

0
0

3
6
Post-conceptual age (months)

Birth

12
Postnatal age (months)

Adult

B
Chromosome 11
HS

-LCR
3 2

22

Fetus
22

Adult
22

Embryo
22 22

Chromosome 16
HS-40

Telomere
Promoter
Regulatory element

Figure 1: Normal developmental switches in globin expression


(A) The sites of haemopoiesis at dierent stages of development and the levels of expression of the embryonic,
fetal, and adult globin chains at various gestational ages are shown. (B) The structure of the -like and -like
globin gene clusters are shown together with the types of haemaglobin produced at each developmental stage.
The promoters are regions that bind transcription factors, including polymerase II, which transcribes the gene into
RNA. Enhancers and related regulatory elements also bind transcription factors and interact with promoters to
increase the levels of RNA transcribed. HS=major upstream regulatory element. -LCR=-locus control region.

For the database of human


haemoglobin variants and
thalassaemias see http://globin.
cse.psu.edu/hbvar/menu.html

374

they are expressed during development to produce


dierent
haemaglobin
tetramers:
embryonic
(Hb Gower-I [22], Hb Gower-II [22], and
Hb Portland [22]), fetal (22), and adult (22 and 2 2).
In the late 1980s, expression of the globin genes was
found to be dependent on remote regulatory elements
upstream of the clusters. In the -globin cluster there are
ve such elements that are collectively referred to as the
-globin locus control region (-LCR).5,6 There are four
elements in the cluster, of which one (HS-40) seems to
have the dominant role (gure 1).8
Research has established how the globin genes are
activated in erythropoiesis. The key chromosomal
elements involved in this process are the upstream
regulatory elements and the promoters of the globin
genes themselves (gure 1). Each element is bound by
either repressing or activating proteins (transcription
factors). As HSCs become committed to the erythroid
lineage, the upstream elements are bound by activating
transcription factors, some of which are expressed
specically in erythroid cells (eg, GATA1, KLF1). The

associated chromatin then becomes activated. These


changes might be associated with alterations in DNA
methylation, with repressed genes being methylated and
active genes being unmethylated.1315 Associated changes
also occur in post-translational modications of the
histones, which can indicate activation (eg, acetylation
and methylation of specic lysine residues) or repression
(eg, methylation of other lysines). Individual promoters
and their associated chromatin might also become
activated or repressed by binding dierent transcriptionfactor complexes.1517 In the late stages of erythropoiesis,
the upstream elements and activated promoters seem to
physically interact via a looping mechanism, which is
mediated by the proteins bound to these elements
(gure 2). The interaction between upstream elements
and specic promoters of the globin genes recruit and
activate RNA polymerase II, which is needed to transcribe
the gene into RNA.18 This nuclear RNA is processed into
messenger RNA, transported into the cytoplasm of
erythroblasts, and translated into the globin chains.
Interpretations have suggested that the molecular
switches between embryonic, fetal, and adult globins rely
on competition between the globin promoters for access
to their activating upstream regulatory elements. The
ability of the promoters to compete might in turn rely on
changes in the activating or repressing transcription
factors that they bind (gure 2). Protein complexes that
are assembled at these elements might change the
anity of interactions between the upstream regions and
the various globin gene promoters during development
eg, in fetal life, the -globin promoter binds an activating
complex that preferentially interacts with the upstream
elements. By contrast, in adult life, the genes are bound
by repressive factors, whereas the genes are now bound
by activating factors (gure 2). This simple description
belies the complexity of this process, which involves
hundreds of proteins for whose involvement there is
good genetic and experimental evidence.1921 The activity
of the enhancer and promoter interactions should be
contained within the globin cluster and not propagated
into anking genes. The activity of such loci is delimited
by chromosomal boundary elements anking the globin
clusters.22 Although the precise role of such elements is
still under investigation, they have proven to be helpful
in strategies to develop gene therapy.

Variants that alter expression of globin genes


More than 200 -thalassaemia alleles have been described
in the database of human haemaglobin variants and
thalassaemias, which involve mutations in any of the
stages from transcription to RNA processing and
translation of -globin mRNA (gure 3).23 These
mutations are detectable by DNA analysis and provide
the basis for genetic counselling.24 Although most
-thalassaemias are caused by point mutations in the
gene or its immediate anking region, small deletions
removing the gene can also occur. When expression of
www.thelancet.com Vol 379 January 28, 2012

Seminar

globin is abolished by the mutation it is referred to as


-thalassaemia, whereas reduced output of normal
chains produces +-thalassaemia. Some structural
variants (particularly the E mutation) might also lead to
a thalassaemic eect because they are produced at
reduced levels, and their interactions with and +
thalassaemia (eg, E/T) lead to many forms of clinically
severe thalassaemia.25
Some rare forms of thalassaemia result from deletions
removing the upstream regulatory elements, but leaving
all of the globin genes intact.26,27 These deletions result in
a substantial reduction in expression of all of the linked
globin genes, and rst indicated the importance of longrange regulatory elements in controlling expression of
the -like globin genes. Other rare cis-acting mutations
(ie, mutations linked on the same chromosome) have
been important in the development of our understanding
of how the switch from -globin to -globin gene
expression is regulated. All these mutations result in
variable levels of increased -globin expression and
increased levels of fetal haemaglobinso called hereditary
persistence of fetal haemoglobin (HPFH). HPFH can
arise from mutations in the -globin promoters aecting
the binding of activating or repressive complexes
(gure 3).7,28,29 Other forms of HPFH result from deletions
removing the adult -globin and -globin genes, but
leaving at least one -globin gene intact (gure 3). These
deletions can lead to a moderate (-thalassaemia) or
considerable (HPFH) increase in -globin expression.7,28,29
Several models have been proposed to explain these
eects (gure 3). First, if the and genes normally
compete for the activity of the -LCR, removal of both
adult globin gene promoters ( and ) might abolish
competition with the fetal -globin genes. Second, these
deletions alter or remove binding sites for proteins that
repress -globin expression. Third, the large HPFH
deletions juxtapose new enhancer sequences (that
normally lie downstream of the cluster) next to the
-globin genes. Such enhancers can activate expression
from either side of a gene. Evidence suggests that all
three mechanisms might have a role in increased
-globin expression.28
In the past 10 years, several rare trans-acting mutations
(ie, mutations that are not physically linked to the locus)
have been identied that alter the pattern of globin gene
expression. These mutations include those in general
transcription factors and chromatin-associated factors,
and in transcription factors that specically aect the
erythroid lineage. Some mutations aect expression of
globin (eg, ATRX30), whereas others mainly aect
-globin or -globin expression (eg, GATA1 and KLF1).3133
Insights into the regulation of globin genes that are
obtained from these sporadic mutations have been
greatly enhanced by extensive observations from family
studies, twin studies, and genome-wide association
studies (GWAS). Findings from GWAS have identied
other transacting factors that can normally regulate the
www.thelancet.com Vol 379 January 28, 2012

-LCR

Stem cell

-LCR

BFUe

-LCR

RNA

-LCR

-RNA

Intermediate
erythroblast

Adult

Fetal
Promoter
Regulatory element

Histones acetylated
Histones unacetylated

DNA methylated
DNA unmethylated

Activating protein complexes


Repressive protein complexes

Figure 2: Expression of globin genes in fetal and adult life


(A) In multipotent stem cells, the -LCR and the genes are mostly unmodied and inert. (B) In early-committed
progenitor cells, the chromatin associated with the -LCR is acetylated and the underlying DNA becomes
unmethylated. (C) In fetal erythroblasts the genes are activated and associate with active protein complexes that
interact with the -LCR while the genes are silent. In adult erythroblasts the genes are now activated and the
-globin genes are silenced by repressive protein complexes. At each stage there might be competition between
the and genes for access to the -LCR. -LCR=-locus control region. BFUe=erythroid burst-forming unit.

patterns of globin gene expression.20,3437 Two new


regulatory pathways that lead to increased -globin
expression have been identied (BCL11A and
HBS1L-MYB) (gure 4), and both seem to act by directly
or indirectly aecting the production of repressor
proteins that specically target the -globin genes. These
ndings suggest that one or both of these pathways
might provide targets that are useful for therapeutic
intervention because patients with haemoglobinopathies
who co-inherit particular DNA sequence variants in these
pathways, or mutations in the regulatory protein-binding
sites, have high concentrations of haemaglobin F and
mild disease phenotypes.3840

The molecular and cellular pathology of


-thalassaemia
At each stage of development, the production of -like
and -like globins is balanced. -globin synthesis is
normally controlled by the two genes (one on each copy
of chromosome 11). A mutation aecting one gene
(/T or -thalassaemia trait) usually causes no clinically
signicant problem, whereas patients who inherit
deleterious mutations in both genes (T/T) frequently
have severe anaemia. The main pathophysiology in
-thalassaemia results from the synthesis of insucient
chains to partner the -globin chains to generate adult
375

Seminar

Point mutations
causing HPFH

HS

-LCR
3 2

About 200
point mutations

1
Small deletions
removing the gene
Large deletions causing
thalassaemia and HPFH

Upstream deletions

Relocation of enhancer

Trans-acting factors
identied in GWAS
and twin studies
BCL11A and HBS1L-MYB
Promoter

Regulatory element

Trans-acting mutations
TFIIH
GATA1
EKLF

Putative regulatory element

Figure 3: Mutations aecting the -globin locus


A summary of the mechanisms underlying -thalassaemia and HPFH. Dashed lines represent variation in the
amounts of anking DNA that are removed by 25 dierent deletions, which underly -thalassaemia and HPFH.
Enhancers downstream of the -globin cluster are relocated because of the associated deletions. HS=upstream
regulatory element. -LCR=-locus control region. HPFH=hereditary persistence of fetal haemoglobin.
GWAS=genome-wide association study.

BCL11A

HBS1L-MYB
DRED TRF2/TRF4

HS

-LCR
3 2

Promoter

SOX6

KLF1
GATA1
NFE2
BCL11A
NFE4

KLF1

KLF1
GATA1
NFE2

GATA1

MBD2

PRMT3

Regulatory element

Ikaros-PYR

BRG1

Putative regulatory element

Figure 4: New pathways regulating the switch from globin to globin


Many factors are involved in this process but we emphasise two identied
pathways. One pathway seems to result from variation in a DNA region
(HBS1L-cMYB on chromosome 6), which is needed for increased expression of
c-Myba proto-oncogene whose expression is crucial for erythropoiesis.
Reduction of c-Myb in the aected cells reduces the synthesis of two nuclear
receptorsTR2 and TR4that are thought to repress the globins by binding
to their promoters in the region where many HPFH point mutations occur.
Some of the variation in fetal haemaglobin associated with HBS1L-MYB might
be mediated via this pathway. A second pathway involves BCL11A (on
chromosome 2). BCL11A normally cooperates with other repressors (eg, Sox6)
to silence the -globin genes in erythroid cells. BCL11A binds to the -globin
locus within a large expanse of DNA containing putative regulatory regions
(dened by many HPFH deletions) lying between the fetal and adult genes.
These independant threads of information came together with the discovery
that the expression of BCLIIA is regulated by the transcription factor KLFI. In this
way another network of factors involved in regulating the -globin to -globin
switch was established. Predictions from this regulatory model have been
noted in patients with altered globin switching in vivo. bHS=upstream
regulatory element. -LCR=-locus control region. HPFH=hereditary
persistence of haemoglobin.

376

haemaglobin (22). Excess chains precipitate in


erythroid precursor cells causing dyserythropoiesis, and
in mature red blood cells causing membrane damage
and haemolysis (gure 5). These primary changes in
erythropoiesis lead to all the secondary changes (anaemia,
splenomegaly, marrow expansion and bone deformities,
hypermetabolic state, and iron accumulation) that lead to
long-term organ damage.12 Understanding how the
products of -chain degradation and iron result in
oxidative damage to the tissues,41 and how hypoxia and
bone-marrow dysfunction interact to result in
inappropriate levels of iron absorption is increasing.42
The severity of thalassaemia is determined partly by
the extent to which -globin synthesis is reduced by the
associated molecular defects (determined by the
combinations of + and thalassaemia). Nevertheless, for
more than 25 years researchers have questioned why
patients who are homozygous for identical molecular
defects in the -globin genes (T/T) can have such
remarkably dierent phenotypes. Some patients need
regular blood transfusion (-thalassaemia major), whereas
others are transfusion independent (-thalassaemia
intermedia). If we understood the genetic basis for this
natural variation, patients disease could be changed from
a severe to a mild phenotype. Two important modiers of
-thalassaemia have emerged from studies of clinical
genetics. First, the co-inheritance of -thalassaemia seems
to be associated with a mild phenotype43-thalassaemia
reduces the pool of free chains and thereby reduces
damage to the red blood cell and its precursors. Second,
the co-inheritance of any disorder associated with
increased synthesis of fetal -globin chains in adults (eg,
HPFH) is associated with a mild phenotype.12 In HPFH,
additional -globin chains partner excess chains to form
fetal haemoglobin (22), which increases the production
of fully functional haemaglobin and thereby reduces the
pool of free chains.

Epidemiology
WHO has estimated that about 15% of the worlds
population might be carriers of -thalassaemia (/T)
and that about 60 000 severely aected infants are born
every year.44 These individuals mostly originate from the
Mediterranean, Middle East, central Asia, India, and
southern China, which suggests that there could be a
selective advantage to carrying such a mutation in these
areas. Similar observations have been made for
thalassaemia,1 which is even more widely distributed
and more frequent than -thalassaemia. Findings from
extensive microepidemiological and case-controlled
studies45 strongly suggest that individuals with either
-thalassaemia or -thalassaemia trait are somewhat
protected in areas where Falciparum malaria is or has
been endemic, thus explaining the high carrier
frequency via natural selection. Frequent consanguinity
might also contribute to the prevalence of thalassaemia
in many of these areas.1,46
www.thelancet.com Vol 379 January 28, 2012

Seminar

Estimates of thalassaemia prevalence are based on


incomplete data and, at best, provide a minimum
estimate of the global problem. Most estimates of gene
frequency were made from a small number of centres
more than 20 years ago, when the geographical
distribution of these disorders was not fully known.
Furthermore, the extent to which the interaction between
thalassaemia and a structural variant E (which also
aects -globin synthesis) contributes to the range of
patients who have serious forms of thalassaemia was
unclear. In fact, 50% of seriously aected patients with
thalassaemia have the E/T genotype.1,47,48
Two additional points further emphasise the current
clinical importance of -thalassaemia. First, thalassaemia
originates mostly in low-income countries where infants
would not have survived to be diagnosed or treated;
however, many of these countries are rapidly undergoing
improvements in all aspects of public health. The burden
in treatment of such individuals with regular transfusion
could overwhelm the health budget of a developing
economy, as in Cyprus and Sardinia in the 1960s.1 Second,
with increasing migration and travel, a genetic disease
that was rare in northern Europe, Australia, and North
America is now becoming more common, and systems
for genetic counselling, prenatal diagnosis, and life-long
medical care are needed. According to the Thalassaemia
International Federation, at least 200 000 patients with
thalassaemia are registered as receiving regular treatment
throughout the world. Although the true burden of
thalassaemia is unknown, the actual number of patients
worldwide is probably underestimated and many do not
receive any treatment.

Clinical phenotypes and standard management


The three broad clinical phenotypes in patients with
thalassaemia are major, intermedia, and minor. These
phenotypes are associated with more than 200 dierent
mutations that either reduce (+-thalassaemia) or abolish
(-thalassaemia) expression of the aected -globin
genes. Thalassaemia major occurs in homozygotes (T/T)
or compound heterozygotes (eg, T/E) for such
mutations. Aected individuals usually present with
pallor, hepatosplenomegaly, and failure to thrive in the
rst year of life when the change from fetal to adult
haemoglobin is completed. The blood prole shows a
severe anaemia (<80 g/L) with hypochromic (mean
corpuscular haemaglobin [MCH] <20 pg) microcytic
(mean corpuscular volume [MCV] <70 fL) red blood cells.
Without blood transfusion, such children remain
anaemic and develop increasing hepatosplenomegaly.
They do not grow normally and show general features of
a hypermetabolic state. Children with untreated or
partially treated thalassaemia major die in the rst or
second decade of life.3,12,49,50 The most widely accepted
blood transfusion protocol for thalassaemia aims to
increase the concentration of haemaglobin to 130140 g/L
after transfusion, and to maintain the concentration at
www.thelancet.com Vol 379 January 28, 2012

A
genes

genes

mRNA

mRNA

globin

globin

Free -globin chains

HbA

Anaemia
Bone deformities

Bone-marrow expansion
Hypermetabolic state

Iron accumulation

Figure 5: The pathophysiology of -thalassaemia


(A) Outline of abnormal globin synthesis in thalassaemia; (B) electron micrographs showing precipitation of
-globin chains in an erythroblast (left), and a mature red cell (centre).The typical blood lm of a patient with
severe thalassaemia (right).

more than 90100 g/L at all times.3,12,47,49 By about 10 years


of age, children who are maintained on such a transfusion regimen develop the complications of iron
overload, including cardiomyopathy, liver brosis, and
endocrine dysfunction.
Until about 10 years ago, iron overload could be reduced
only by subcutaneous infusion of an iron chelator
(desferrioxamine) administered ve to seven nights per
week. Administration of this drug had demonstrable
benets for morbidity and mortality.5153 Such benets
were particularly so when good adherence to iron
chelation was achieved, with many patients with
thalassaemia surviving in good health to late adulthood.
More recently, the orally active iron chelators (deferiprone
and deferasirox) have been introduced. Deferiprone was
licensed in Europe in 1999 for adults with thalassaemia
major who could not be eectively treated with
desferrioxamine but is not yet licensed in North America.
Monotherapy with deferiprone controlled concentrations
of liver iron in only a few patients, but when desferrioxamine was also given two to four times a week,
more patients achieved iron balance.54
This regime also reduces myocardial iron more
eectively than desferrioxamine monotherapy given ve
nights a week at standard doses.50,51 The most serious
drawback of deferiprone is the occurrence of
agranulocytosis in about 1% of patients, necessitating
weekly blood counts. Defarasirox has been licensed in
North America and Europe since 2006 for the treatment of
several disorders with transfusional iron overload, has
been eective at improving liver iron, serum ferritin55 and

For more on the Thalassaemia


International Federation see
http://www.thalassemia.org.cy

377

Seminar

myocardial iron,56 and has an acceptable tolerability prole


in prospective studies involving more than 7000
patients.57
Although with good chelation, mortality from cardiac
disease is now decreasing and many patients with severe
thalassaemia now survive into late adulthood, for various
reasons, many others still suer the complications of
iron overload. The contribution of dierent chelation
modlities to improved outcomes has been debated and is
dicult to dene with retropsective analyses. Consequently, understanding of iron metabolism and
improvements in iron chelation in thalassaemia are still
important areas of future research.
Individuals with thalassaemia intermedia are also
homozygotes or compound heterozygotes for + and
thalassaemia. No strict criteria are available for
diagnosis of thalassaemia intermedia because this
diagnosis is dependent on how a patient develops without
transfusion. Patients with thalassaemia intermedia
generally present later in life (aged 26 years) than do
those with thalassaemia major. Some patients with
thalasaemia intermedia will not grow and develop
normally without regular blood transfusion and will
therefore be categorised as having thalassaemia major.
By contrast, other patients who present late might thrive
without regular blood transfusion (thalassaemia
intermedia). The blood prole shows a haemoglobin
concentration of 70100 g/L with an MCV of 5080 fL
and an MCH between 1624 pg.
The iron status in all patients with thalassaemia
intermedia should be carefully monitored (and chelation
considered) because excess gastrointestinal absorption
of iron almost always occurs as a direct result of
ineective erythropoiesis. Studies have also emphasised
the occurrence of other complications with increasing
age, including pulmonary hypertension, tumours
resulting from extramedullary erythropoiesis, silent
cerebral infarcts, thromboembolic disease, chronic leg
ulcers, and renovascular disease.58 Patients with either
thalassaemia major or thalassaemia intermedia might
benet from splenectomy. In patients with thalassaemia
major, splenectomy is indicated if the red blood cell
requirements to maintain haemaglobin at more than
100 g/L exceed about 200 mL/kg of packed red blood
cells per kg per year, with the assumption that other
reasons for increased consumption of red blood cells
(eg, immune haemolytic reactions) have been excluded.
Other indications for splenectomy in thalassaemia major
and intermedia include symptoms from splenic
enlargement, leucopenia or thrombocytopenia, and
increased iron overload despite iron chelation.49 Any
decision to undertake splenectomy in patients with
thalassaemia should be balanced against the substantial
risks associated with this treatment, particularly in
patients with thalassaemia intermedia. These risks
include sepsis, which might be reduced by presplenectomy immunisation against encapsulated
378

bacteria, and the use of prophylactic antibiotics.


Furthermore, there is a risk of thrombosis and possibly
an increased incidence of pulmonary hypertension.59,60
Individuals who inherit a single + or allele (/T)
generally have -thalassaemia minor. Carriers of
thalassaemia typically have only a mild hypochromic
microcytic anaemia with a raised concentration of the
minor HbA2. Such patients need no specic treatment,
but should avoid unnecessary iron supplements.

Genetic testing and prenatal diagnosis


Initial screening of populations and identication of
families at risk of producing infants who are aected by
thalassaemia has been achieved by examination of redblood-cell indices and analysis of haemaglobin. The
techniques to identify specic mutations underlying
thalassaemia in DNA from adults and fetuses are now
well established and extensively applied to genetic
counselling and prenatal diagnosis.24 New non-invasive
techniques to analyse fetal DNA in the maternal
circulation are being developed.61 Establishment of the
range of mutations in specic geographical regions has
enabled diagnostic centres to develop appropriate
strategies for eective counselling and diagnostic testing.
The application of such strategies in many regions of the
world has had a substantial eect on the birth rate of
infants with -thalassaemia major and intermedia.62
Genetic counselling can be complicated when the
clinical outcome cannot be accurately predicted from
the -globin genotypeeg, some patients with
E/ thalassaemia might have -thalassaemia major, and
others with exactly the same genotype might have a
very mild form of thalassaemia intermedia.39 The accuracy
of genetic counselling will rise as more genetic modiers
of thalassaemia are identied, their eects on phenotype
documented, and low-cost high-throughput DNA testing
becomes available. Therefore, in countries with no ethical
or religious objections, prenatal diagnostic programmes
should be established.

Pharmacological agents used to treat


thalassaemia
The aims of therapeutic interventions for thalassaemia
are to increase expression of globin or to decrease
expression of globin, thus restoring the balance between
-like and -like globin chains. Much evidence from
clinical genetic studies shows that either (or preferably
both) of these manipulations would have substantial
clinical belents in patients with thalassaemia.7,12,43,63
Pharmacological studies have all focused on increasing
expression of globin, but have been based on an
incomplete understanding of globin gene-regulation.6466
Various approaches have included use of cytostatic agents
(eg, hydroxycarbamide) to increase fetal haemaglobin by
altering the kinetics of erythropoiesis; DNA demethylating
agents (eg, azacitidine and decitabine) to hypomethylate
the -globin genes and increase their expression;
www.thelancet.com Vol 379 January 28, 2012

Seminar

deacetylase inhibitors (eg, sodium butyrate) to increase


-globin expression via histone acetylation; and shortchain fatty acids related to butyrate. Although some of
these experimental therapies have yielded notable results
in early-phase clinical trials (including some substantial
improvements) the results have not been suciently
encouraging to develop large-scale trials.6466 Furthermore,
despite extensive analysis, these agents have several
eects, and how any of these drugs work to increase
expression of fetal haemaglobin is not fully understood.
Improvement in the understanding of the transcriptional network causing the transregulation of the
change from -globin to -globin expression has provided
many new potential targets for manipulation of this
switch. As additional transcription factors, cofactors, and
chromatin-associated factors in this newly discovered
pathway are elucidated (gure 4), a detailed understanding
of this switch might be forthcoming and translated into a
mechanism-based approach to the reactivation of
production of haemaglobin-F in thalassaemia.

Stem-cell transplantation for thalassaemia


In the past 30 years, stem-cell transplantation has
substantially advanced treatment of thalassaemia major.67
Children who are identied before developing viral
hepatitis or severe iron overload and who receive HLAidentical related donor stem-cell transplants have a very
high likelihood of remission, with less than 10%
mortality and minimal morbidity, apart from impaired
fertility.67 Most groups report event-free survival of
8090% for thalassaemia.67,6972 By contrast, event-free
survival in adult patients with thalassaemia major who
receive bone-marrow transplantation is less than 70%.67,73
This nding is presumably because patients have already
developed the complications of thalassaemia and iron
A

Mobilisation

overload, thus compounding the clinical problems


associated with transplantation.
A major problem in development of a programme for
stem-cell transplantation in young patients with
thalassaemia is that HLA-matched stem cells are often
unavailable. Transplants from matched unrelated or
haploidentical donors are associated with a substantially
lower disease-free survival (2070%) and a higher
incidence of morbidity and mortality (2530%) than are
HLA-identical transplants from related donors.7478
Denitive HSCs can be obtained from cord blood and,
since the 1990s, several studies71 have assessed the use of
cord-blood transplantation for patients with thalassaemia.
The outcome of related cord-blood transplantation for
the treatment for thalassaemia is now approaching
that for conventional bone-marrow transplantation, with
rates of disease-free survival of about 90%. The main
benet from cord-blood transplantation will come from
the use of unrelated transplants that are derived from a
cord-blood bank; but, data are scarce.7981 Unpublished
results from the Eurocord cooperative group suggest
that transplantation of patients with cord-blood stem
cells from an unrelated donor is feasible; however, the
outcome is less succesful than that of patients given
HLA-compatible cord-blood transplants from a sibling
(Rocha V, personal communication).
The main complication of cord-blood transplantation
is graft rejection, which can be reduced by modication
of the protocols used for pretransplant immune
suppression.71,79 Stem-cell transplantation is an exciting
development for the treatment of otherwise t, young
patients with thalassaemia. Nevertheless, a ne clinical
judgment has to be made to balance a potential cure
that has a risk of mortality and morbidity (including
graft vs host disease), against a burdensome life-long

C
Viral vector
CD34+ cell
purification

PBMCs

Non-ablative myelosuppression

Quality control

Figure 6: An outline of the protocol used for gene therapy in patients with thalassaemia
(A) The patient is rst treated with G-CSF to mobilise HSCs into the peripheral blood. (B) Peripheral blood mononuclear cells are then collected and CD34+ cells
(enriched for HSCs) are harvested. (C) These cells are cultured with a viral vector that is designed to express high levels of normal human globin. (D) The patient
then undergoes non-ablative myelosuppression and is engrafted with the virally-modied HSCs (lled circles). PBMC=peripheral blood mononuclear cell.
G-CSF=granulocyte colony-stimulating factor. HSC=haemopoietic stem cell.

www.thelancet.com Vol 379 January 28, 2012

379

Seminar

Som
Somatic
matic cel
cell

Transplant
nspllant
gene-corrected
e-co
orreected
cellss

B
E

Reprogramme by viral
transduction with key
transcription factors

Dierentiate to HSC

Correct genetic defect

Patient-specic
iPS cell cones

Figure 7: An outline of the proposed use of iPS cells for therapy in patients with thalassaemia
Somatic cells from the patient (A) are reprogrammed by expression of key transcription factors that are associated
with pluripotency (B) to make patient-specic iPS cells (C). In thalassaemia, the defect in the -globin gene is then
corrected (D). These corrected iPS cells are then dierentiated into haemopoietic stem cells (E) and transplanted
(F) back into the patient after myelosuppression. iPS cell=induced pluripiotent stem cell.

treatment with blood transfusion and iron chelation


that has nevertheless converted -thalassaemia into a
chronic survivable disease with a life expectancy of
50 years or more. Stem-cell transplantation will probably
be available only to those with access to high-technology
medicine, and not generally available to most patients
with thalassaemia.

Gene therapy for thalassaemia


Thalassaemia was among the rst genetic diseases for
which gene therapy was proposed.82 This disorder is to
some extent a good target because defects in expression of
globin genes aect only the haemopoietic system and
specically aect erythropoiesis. Stem-cell transplantation
is well developed for the haemapoietic system; however,
unlike many genetic diseases, very high levels of tissuespecic gene expression are necessary to correct the globin
defect in thalassaemia. In principle, the procedure is
straightforward (gure 6). Haemopoietic stem cells from
the patient are harvested; modied viruses containing the
380

-globin gene linked to its upstream regulatory elements


are used to infect and insert the -gene construct into the
genome of HSCs. To reduce the disease burden and to
create space in the bone-marrow niche to accept the
modied HSCs, autologous stem cells are infused back
into the patient after myelosuppression. Such a protocol
overcomes the problems of donor availability and the
immunological complications associated with allogeneic
stem-cell transplantation.
Although seemingly simple, many challenges have
been encountered in the development of a suitable
protocol. Viruses can accommodate only small segments
of DNA and therefore, in view of the complexity of globin
gene expression, many years have been spent in the
development of viral constructs that contain all of the
sequences needed for adequate expression of the globin
genes. Even when such constructs have been obtained,
their expression might be positively or negatively aected
by their position of integration in the genome. To
overcome eects related to position, constructs now
frequently contain the chromosomal boundary elements
described above (gure 2). However, of great concern in
some patients treated with retroviral constructs (for
immune deciency disorders), insertions of viral
constructs next to crucial haemopoietic genes have led to
the development of leukaemia.8385
Despite these setbacks, considerable progress has been
made in the past 10 years. Rather than use of retroviruses,
safer lentiviral vectors have been developed, which have
expressed long-term, therapeutic levels of globin in
preclinical trials, and have corrected the anaemia in a
mouse model of thalassaemia.8690 Encouraged by these
preclinical studies,91 the rst clinical trial using a lentiviral
construct of globin to treat patients with thalassaemia
was started in 2007.92,93 Furthermore, at least two other
groups94,95 have announced plans to undertake similar
trials. In the 2007 trial, two patients received gene therapy.
A delayed haematological recovery in the rst patient
needed a rescue with back-up cells that had not been
manipulated. The second patient with E-thalassaemia
major (E/T) also had a delayed recovery, but eventually
established 10% engraftment with modied HSCs. At
present, this patient maintains a haemoglobin concentration of 90100 g/L, partly because of the anticipated
increase in -globin expression, but also because of an
unexplained increase in fetal haemaglobin. Consequently,
this patient has remained well and transfusion
independent for almost 2 years. However, a detailed
analysis has shown that in a partially dominant
haemopoietic clone, the viral vector has integrated into a
proto-oncogene, which could be a harbinger of a
leukaemic transformation as seen previously in genetherapy protocols with retroviruses.92

Beyond gene therapy


In 2007, a landmark report96 described how human
somatic cells (eg, skin broblasts) could be
www.thelancet.com Vol 379 January 28, 2012

Seminar

reprogrammed to form multipotent cells resembling


embryonic stem cells. These reprogrammed cells are
called induced pluripotent stem (iPS) cells.96,97 Generation
of iPS cells involves the introduction and expression of
four transcription factors (Oct4, Sox2, KLF4, and c-Myc)
in somatic cells. These transcription factors are normally
needed to establish and maintain pluripotency. iPS cells
have enabled new research possibilities to be explored,
establishing new models of human disease, providing
new methods for drug screening, and oering the
potential for new approaches to cell therapy.76 A proof of
principle for use of iPS cells was shown when these
cells were generated from the somatic cells of a mouse
model of sickle-cell disease.98,99 The S mutation in the
mouse iPS cells was corrected (with use of homologous
recombination), and the cells were then dierentiated
into multipotent HSCs before being transplanted back
into the irradiated mouse to regenerate a haemopoietic
system in which the -globin mutation was corrected.
In a move towards the development of a similar strategy
in human beings, skin broblasts from a patient with
homozygous thalassaemia were reprogrammed to
form iPS cells; these cells could then be successfully
dierentiated to form red blood cells.100
Although all the necessary steps for cell therapy could
be done with human cells (gure 7), some major
challenges still need to be overcome before this
approach can be considered for clinical use. The
reagents for generation of iPS cells will need to be
improved, as will the eciency with which somatic
cells are reprogrammed. Additionally, the generation of
genuine HSCs from iPS cells is not yet possible. Finally,
reprogrammed, genetically engineered iPS cells will
need to be shown to have no potential to undergo
subsequent malignant transformation.

Conclusions
Despite intensive clinical and scientic investigation of
thalassaemiaa molecular disease that is perhaps better
understood than any otherattempts to improve its
management and to develop targeted drug therapy have
not yielded a clear breakthrough. Stem-cell transplantation is an eective cure but still has a substantial
risk of mortality and morbidity. Supportive results
from cord-blood transplantation should encourage the
development of cord-blood banking to address this issue.
Gene therapy is now entering early-phase clinical trials,
and cell therapy with iPS cells is an exciting prospect
with many challenges to overcome. Even if these
approaches are successful, such therapy might not be
immediately applied to most patients in low-income
countries. Developments in the understanding of the
molecular circuitry involved in the transition from fetal
to adult globin-gene expression might facilitate the
development of new drugs to manipulate this switch.
Even so, the cost of what would presumably be lifelong
medication could rule out therapy for all but those who
www.thelancet.com Vol 379 January 28, 2012

are supported by wealthy economies. The most important


development for the next few years might be the use of
high-throughput DNA analysis to provide cheap accurate
genotyping and prenatal testing, which has been eective
when applied systematically.
Contributors
Each author contributed equally to reviewing the published works and
writing the Seminar. DRH had full access to all the data in the study
and had nal responsibility for the decision to submit for publication.
Conicts of interest
GS has received payment for board membership and holds stock for
HemaQuest, and has received royalties from University of Washington.
All other authors declare that they have no conicts of interest.
Acknowledgments
We thank Vijay Sankaran, Noemi Roy, Deborah Hay, Bill Wood,
Irene Roberts, and David Weatherall for reading and commenting on
the manuscript. DRH thanks the National Institute for Health
Research Biomedical Research Centre (Oxford) and Medical Research
Council for support.
References
1
Weatherall DJ. The inherited diseases of hemoglobin are an
emerging global health burden. Blood 2010; 115: 433136.
2
Harteveld CL, Higgs DR. -thalassaemia. Orphanet J Rare Dis 2010;
5: 13.
3
Galanello R, Origa R. Beta-thalassemia. Orphanet J Rare Dis 2010;
5: 11.
4
Weatherall DJ. Towards molecular medicine; reminiscences of the
haemoglobin eld, 19602000. Br J Haematol 2001; 115: 72938.
5
Li Q, Peterson KR, Fang X, Stamatoyannopoulos G. Locus control
regions. Blood 2002; 100: 307786.
6
Palstra RJ, de Laat W, Grosveld F. -globin regulation and
long-range interactions. Adv Genet 2008; 61: 10742.
7
Stamatoyannopoulos G, Grosveld F. Hemoglobin switching. In:
Stamatoyannopoulos G, Majerus PW, Perlmutter RM, Varmus H,
eds. The molecular basis of blood diseases, 3rd edn. Philadelphia:
WB Saunders Company, 2001: 13682.
8
Higgs DR, Wood WG. Long-range regulation of alpha globin gene
expression during erythropoiesis. Curr Opin Hematol 2008;
15: 17683.
9
Dzierzak E. A developmental approach to hematopoiesis. In:
Steinberg MH, Forget BG, Higgs DR, Weatherall DJ, eds. Disorders
of hemoglobin: genetics, pathophysiology and clinical management,
2nd edn. Cambridge: Cambridge University Press, 2009: 323.
10 Palis J. Ontogeny of erythropoiesis. Curr Opin Hematol 2008;
15: 15561.
11 Philipsen S, Wood WG. Erythropoiesis. In: Steinberg MH,
Forget BG, Higgs DR, Weatherall DJ, eds. Disorders of hemoglobin,
2nd edn. Cambridge: Cambridge University Press, 2009: 2445.
12 Weatherall DJ, Clegg JB. The thalassaemia syndromes, 4th edn.
Oxford: Blackwell Science, 2001.
13 Goren A, Simchen G, Fibach E, et al. Fine tuning of globin gene
expression by DNA methylation. PLoS One 2006; 1: e46.
14 Mabaera R, Richardson CA, Johnson K, Hsu M, Fiering S,
Lowrey CH. Developmental- and dierentiation-specic patterns
of human - and -globin promoter DNA methylation. Blood 2007;
110: 134352.
15 Kiefer CM, Hou C, Little JA, Dean A. Epigenetics of -globin gene
regulation. Mutat Res 2008; 647: 6876.
16 Miles J, Mitchell JA, Chakalova L, et al. Intergenic transcription,
cell-cycle and the developmentally regulated epigenetic prole
of the human beta-globin locus. PLoS One 2007; 2: e630.
17 Strahl BD, Allis CD. The language of covalent histone
modications. Nature 2000; 403: 4145.
18 de Laat W, Klous P, Kooren J, et al. Three-dimensional organization
of gene expression in erythroid cells. Curr Top Dev Biol 2008;
82: 11739.
19 Cantor AB, Orkin SH. Transcriptional regulation of
erythropoiesis: an aair involving multiple partners. Oncogene
2002; 21: 336876.

381

Seminar

20
21

22
23

24
25
26

27

28
29

30

31

32
33

34

35

36

37

38

39

40

41

42

43

382

Sankaran VG, Xu J, Orkin SH. Advances in the understanding


of haemoglobin switching. Br J Haematol 2010; 149: 18194.
Stamatoyannopoulos G. Control of globin gene expression during
development and erythroid dierentiation. Exp Hematol 2005;
33: 25971.
Gaszner M, Felsenfeld G. Insulators: exploiting transcriptional and
epigenetic mechanisms. Nat Rev Genet 2006; 7: 70313.
Giardine B, Borg J, Higgs DR, et al. Systematic documentation and
analysis of human genetic variation in hemaglobinopathies using
the microattribution approach. Nat Genet 2011; 43: 295301.
Old JM. Screening and genetic diagnosis of haemoglobin disorders.
Blood Rev 2003; 17: 4353.
Rees DC, Styles J, Vichinsky EP, Clegg JB, Weatherall DJ. The
hemoglobin E syndromes. Ann N Y Acad Sci 1998; 850: 33443.
Driscoll MC, Dobkin CS, Alter BP. -thalassemia due to a de novo
mutation deleting the 5' -globin gene activation-region
hypersensitive sites. Proc Natl Acad Sci USA 1989; 86: 747074.
Kioussis D, Vanin E, deLange T, Flavell RA, Grosveld FG. -globin
gene inactivation by DNA translocation in -thalassaemia. Nature
1983; 306: 66266.
Forget BG. Molecular basis of hereditary persistence of fetal
hemoglobin. Ann N Y Acad Sci 1998; 850: 3844.
Thein SL, Wood WG. The molecular basis of thalassemia,
thalassemia and hereditary persistence of fetal hemoglobin. In:
Steinberg MH, Forget BG, Higgs DR, Weatherall DJ, eds. Disorders
of hemoglobin, 2nd edn. Cambridge: Cambridge University Press,
2009: 32356.
Law MJ, Lower KM, Voon HP, et al. ATR-X syndrome protein
targets tandem repeats and inuences allele-specic expression
in a size-dependent manner. Cell 2010; 143: 36778.
Borg J, Papadopoulos P, Georgitsi M, et al. Haploinsuciency
for the erythroid transcription factor KLF1 causes hereditary
persistence of fetal hemoglobin. Nat Genet 2010; 42: 80105.
Cantor AB. GATA transcription factors in hematologic disease.
Int J Hematol 2005; 81: 37884.
Zhou D, Liu K, Sun CW, Pawlik KM, Townes TM. KLF1 regulates
BCL11A expression and - to -globin gene switching. Nat Genet
2010; 42: 74244.
Galarneau G, Palmer CD, Sankaran VG, Orkin SH, Hirschhorn JN,
Lettre G. Fine-mapping at three loci known to aect fetal
hemoglobin levels explains additional genetic variation. Nat Genet
2010; 42: 104951.
Menzel S, Garner C, Gut I, et al. A QTL inuencing F cell
production maps to a gene encoding a zinc-nger protein
on chromosome 2p15. Nat Genet 2007; 39: 119799.
Thein SL, Menzel S, Peng X, et al. Intergenic variants of
HBS1L-MYB are responsible for a major quantitative trait locus
on chromosome 6q23 inuencing fetal hemoglobin levels in adults.
Proc Natl Acad Sci USA 2007; 104: 1134651.
Thein SL, Menzel S, Lathrop M, Garner C. Control of fetal
hemoglobin: new insights emerging from genomics and clinical
implications. Hum Mol Genet 2009; 18: R21623.
Lettre G, Sankaran VG, Bezerra MAC, et al. DNA polymorphisms
at the BCL11A, HBSIL-MYB, and -globin loci associate with fetal
hemoglobin levels and pain crises in sickle cell disease.
Proc Natl Acad Sci USA 2008; 105: 1186974.
Nuinoon M, Makarasara W, Mushiroda T, et al. A genome-wide
association identied the common genetic variants inuence
disease severity in 0-thalassemia/hemoglobin E. Hum Genet 2010;
127: 30314.
Uda M, Galanello R, Sanna S, et al. Genome-wide association study
shows BCL11A associated with persistent fetal hemoglobin and
amelioration of the phenotype of -thalassemia.
Proc Natl Acad Sci USA 2008; 105: 162025.
Fibach E, Rachmilewitz EA. The role of antioxidants and iron
chelators in the treatment of oxidative stress in thalassemia.
Ann N Y Acad Sci 2010; 1202: 1016.
Gardenghi S, Ramos P, Follenzi A, et al. Hepcidin and Hfe in
iron overload in -thalassemia. Ann N Y Acad Sci 2010;
1202: 22125.
Voon HP, Vadolas J. Controlling -globin: a review of -globin
expression and its impact on -thalassemia. Haematologica 2008;
93: 186876.

44

45

46
47
48

49
50

51

52

53

54

55

56

57
58

59

60

61

62

63
64

65

66
67

68

Modell B, Darlison M. Global epidemiology of haemoglobin


disorders and derived service indicators. Bull World Health Organ
2008; 86: 48087.
Weatherall DJ, Williams TN, Allen SJ, ODonnell A. The population
genetics and dynamics of the thalassemias.
Hematol Oncol Clin North Am 2010; 24: 102131.
Bittles AH, Mason WM, Greene J, Rao NA. Reproductive behavior
and health in consanguineous marriages. Science 1991; 252: 78994.
Cohen AR, Galanello R, Pennell DJ, Cunningham MJ, Vichinsky E.
Thalassemia. Hematology Am Soc Hematol Educ Program 2004: 1434.
Olivieri NF, Muraca GM, ODonnell A, Premawardhena A,
Fisher C, Weatherall DJ. Studies in haemoglobin
E beta-thalassaemia. Br J Haematol 2008; 141: 38897.
Olivieri NF. Thalassaemia: clinical management.
Baillieres Clin Haematol 1998; 11: 14762.
Olivieri NF, Weatherall DJ. Clinical aspects of beta thalassemia
and related disorders. In: Steinberg MH, Forget BG, Higgs DR,
Weatherall DJ, eds. Disorders of hemoglobin, 2nd edn. Cambridge:
Cambridge University Press; 2009: 357416.
Borgna-Pignatti C, Rugolotto S, De Stefano P, et al. Survival
and complications in patients with thalassemia major treated with
transfusion and deferoxamine. Haematologica 2004; 89: 118793.
Brittenham GM, Grith PM, Nienhuis AW, et al. Ecacy
of deferoxamine in preventing complications of iron overload in
patients with thalassemia major. N Engl J Med 1994; 331: 56773.
Olivieri NF, Nathan DG, MacMillan JH, et al. Survival in medically
treated patients with homozygous -thalassemia. N Engl J Med 1994;
331: 57478.
Aydinok Y, Ulger Z, Nart D, et al. A randomized controlled 1-year
study of daily deferiprone plus twice weekly desferrioxamine
compared with daily deferiprone monotherapy in patients with
thalassemia major. Haematologica 2007; 92: 1599606.
Cappellini MD, Cohen A, Piga A, et al. A phase 3 study of deferasirox
(ICL670), a once-daily oral iron chelator, in patients with
beta-thalassemia. Blood 2006; 107: 345562.
Pennell DJ, Porter JB, Cappellini MD, et al. Ecacy of deferasirox in
reducing and preventing cardiac iron overload in beta-thalassemia.
Blood 2010; 115: 236471.
Porter JB. Deferasiroxcurrent knowledge and future challenges.
Ann NY Acad Sci 2010; 1202: 8793.
Taher AT, Musallam KM, Cappellini MD, Weatherall DJ. Optimal
management of thalassaemia intermedia. Br J Haematol 2011;
152: 51223.
Phrommintikul A, Sukonthasarn A, Kanjanavanit R,
Nawarawong W. Splenectomy: a strong risk factor for pulmonary
hypertension in patients with thalassaemia. Heart 2006;
92: 146772.
Singer ST, Kuypers FA, Styles L, Vichinsky EP, Foote D,
Rosenfeld H. Pulmonary hypertension in thalassemia: association
with platelet activation and hypercoagulable state. Am J Hematol
2006; 81: 67075.
Lo YM, Chiu RW. Noninvasive approaches to prenatal diagnosis
of hemoglobinopathies using fetal DNA in maternal plasma.
Hematol Oncol Clin North Am 2002; 24: 117986.
Cao A, Rosatelli MC, Monni G, Galanello R. Screening for
thalassemia: a model of success. Obstet Gynecol Clin North Am;
29: 30528.
Thein SL. Genetic modiers of the -haemoglobinopathies.
Br J Haematol 2008; 141: 35766.
Pace BS, Zein S. Understanding mechanisms of -globin gene
regulation to develop strategies for pharmacological fetal
hemoglobin induction. Dev Dyn 2006; 235: 172737.
Perrine SP. Fetal globin stimulant therapies in the
-hemoglobinopathies: principles and current potential.
Pediatr Ann 2008; 37: 33946.
Trompeter S, Roberts I. Haemoglobin F modulation in childhood
sickle cell disease. Br J Haematol 2009; 144: 30816.
Krishnamurti L, Bunn HF, Williams AM, Tolar J. Hematopoietic
cell transplantation for hemoglobinopathies.
Curr Probl Pediatr Adolesc Health Care 2008; 38: 618.
Gaziev J, Lucarelli G. Stem cell transplantation and gene therapy
for hemoglobinopathies. Curr Hematol Rep 2005; 4: 12631.

www.thelancet.com Vol 379 January 28, 2012

Seminar

69

70

71
72

73

74

75

76

77

78
79

80

81

82
83

84

85

Lawson SE, Roberts IA, Amrolia P, Dokal I, Szydlo R, Darbyshire PJ.


Bone marrow transplantation for -thalassaemia major: the UK
experience in two paediatric centres. Br J Haematol 2003; 120: 28995.
Lucarelli G, Galimberti M, Polchi P, et al. Bone marrow
transplantation in patients with thalassemia. N Engl J Med 1990;
322: 41721.
Pinto FO, Roberts I. Cord blood stem cell transplantation
for haemoglobinopathies. Br J Haematol 2008; 141: 30924.
Smiers FJ, Krishnamurti L, Lucarelli G. Hematopoietic stem cell
transplantation for hemoglobinopathies: current practice and
emerging trends. Pediatr Clin North Am 2010; 57: 181205.
Gaziev J, Sodani P, Polchi P, Andreani M, Lucarelli G. Bone
marrow transplantation in adults with thalassemia: treatment and
long-term follow-up. Ann N Y Acad Sci 2005; 1054: 196205.
La Nasa G, Giardini C, Argiolu F, et al. Unrelated donor bone
marrow transplantation for thalassemia: the eect of extended
haplotypes. Blood 2002; 99: 435056.
La Nasa G, Caocci G, Argiolu F, et al. Unrelated donor stem cell
transplantation in adult patients with thalassemia.
Bone Marrow Transplant 2005; 36: 97175.
Yannaki E, Emery DW, Stamatoyannopoulos G. Gene therapy for
-thalassaemia: the continuing challenge. Expert Rev Mol Med 2010;
12: e31.
Gaziev D, Galimberti M, Lucarelli G, et al. Bone marrow
transplantation from alternative donors for thalassemia:
HLA-phenotypically identical relative and HLA-nonidentical sibling
or parent transplants. Bone Marrow Transplant 2000; 25: 81521.
Lucarelli G, Gaziev J. Advances in the allogeneic transplantation
for thalassemia. Blood Rev 2008; 22: 5363.
Boncimino A, Bertaina A, Locatelli F. Cord blood transplantation
in patients with hemoglobinopathies. Transfus Apher Sci 2010;
42: 27781.
Hall JG, Martin PL, Wood S, Kurtzberg J. Unrelated umbilical cord
blood transplantation for an infant with -thalassemia major.
J Pediatr Hematol Oncol 2004; 26: 38285.
Jaing TH, Hung IJ, Yang CP, Chen SH, Sun CF, Chow R. Rapid
and complete donor chimerism after unrelated mismatched cord
blood transplantation in 5 children with -thalassemia major.
Biol Blood Marrow Transplant 2005; 11: 34953.
Friedmann T, Roblin R. Gene therapy for human genetic disease?
Science 1972; 175: 94955.
Hacein-Bey-Abina S, Von Kalle C, Schmidt M, et al. LMO2-associated
clonal T cell proliferation in two patients after gene therapy for
SCID-X1. Science 2003; 302: 41519.
Hacein-Bey-Abina S, Hauer J, Lim A, et al. Ecacy of gene therapy
for X-linked severe combined immunodeciency. N Engl J Med
2010; 363: 35564.
Kohn DB. Update on gene therapy for immunodeciencies.
Clin Immunol 2010; 135: 24754.

www.thelancet.com Vol 379 January 28, 2012

86

Imren S, Payen E, Westerman KA, et al. Permanent and


panerythroid correction of murine thalassemia by multiple
lentiviral integration in hematopoietic stem cells.
Proc Natl Acad Sci USA 2002; 99: 1438085.
87 May C, Rivella S, Callegari J, et al. Therapeutic haemoglobin
synthesis in -thalassaemic mice expressing lentivirus-encoded
human -globin. Nature 2000; 406: 8286.
88 Persons DA, Hargrove PW, Allay ER, Hanawa H, Nienhuis AW.
The degree of phenotypic correction of murine -thalassemia
intermedia following lentiviral-mediated transfer of a human
-globin gene is inuenced by chromosomal position eects and
vector copy number. Blood 2003; 101: 217583.
89 May C, Rivella S, Chadburn A, Sadelain M. Successful treatment
of murine -thalassemia intermedia by transfer of the human
-globin gene. Blood 2002; 99: 190208.
90 Rivella S, May C, Chadburn A, Rivire I, Sadelain M. A novel
murine model of Cooley anemia and its rescue by
lentiviral-mediated human -globin gene transfer. Blood 2003;
101: 293239.
91 Bank A, Dorazio R, Leboulch P. A phase I/II clinical trial
of -globin gene therapy for -thalassemia. Ann N Y Acad Sci 2005;
1054: 30816.
92 Cavazzana-Calvo M, Payen E, Negre O, et al. Transfusion
independence and HMGA2 activation after gene therapy of human
-thalassaemia. Nature 2010; 467: 31822.
93 Kaiser J. Gene therapy. -thalassemia treatment succeeds, with
a caveat. Science 2010; 326: 146869.
94 Persons DA. Hematopoietic stem cell gene transfer for the
treatment of hemoglobin disorders.
Hematology Am Soc Hematol Educ Program 2009; 1: 69097.
95 Sadelain M, Boulad F, Galanello R, et al. Therapeutic options
for patients with severe -thalassemia: the need for globin gene
therapy. Hum Gene Ther 2007; 18: 19.
96 Takahashi K, Yamanaka S. Induction of pluripotent stem cells from
mouse embryonic and adult broblast cultures by dened factors.
Cell 2006; 126: 66376.
97 Yu J, Vodyanik MA, Smuga-Otto K, et al. Induced pluripotent stem
cell lines derived from human somatic cells. Science 2007;
318: 191720.
98 Hanna J, Wernig M, Markoulaki S, et al. Treatment of sickle cell
anemia mouse model with iPS cells generated from autologous
skin. Science 2007; 318: 192023.
99 Higgs DR. A new dawn for stem-cell therapy. N Engl J Med 2008;
358: 96466.
100 Ye L, Chang JC, Lin C, Sun X, Yu J, Kan YW. Induced pluripotent
stem cells oer new approach to therapy in thalassemia and sickle
cell anemia and option in prenatal diagnosis in genetic diseases.
Proc Natl Acad Sci USA 2009; 106: 982630.

383

Case Report

Acute dyspnoeanot always above the diaphragm


Anil Chalisey, Shahzad Shah, Mahzuz Karim
Lancet 2012; 379: 384
Department of Renal Medicine,
Norfolk and Norwich
University Hospital, Norwich,
UK (A Chalisey MRCP,
S Shah MRCP, M Karim FRCP)
Correspondence to:
Dr Mahzuz Karim, Department of
Renal Medicine, Norfolk and
Norwich University Hospital,
Colney Lane,
Norwich NR4 7UY, UK
mkarim@doctors.org.uk

In October, 2010, a 55-year-old woman was admitted to


our department for investigation of recurrent ascites.
She had a medical history of adult polycystic kidney
disease with hepatic cysts. 9 months earlier, she had
started peritoneal dialysis, which was changed to
haemodialysis after 5 months because of poor catheter
ows and underdialysis. Subsequently, she developed
ascites which required frequent drainage. The uid was
borderline for transudate or exudate with no malignant
cells and was initially attributed to peritoneal irritation
from her dialysis catheter. Her ascites persisted despite
removal of the catheter and she was then admitted for
further investigation.
Blood tests showed hypoalbuminaemia but normal
liver function and coagulation. CT abdomen conrmed
polycystic kidneys and massive polycystic hepatomegaly
with large-volume ascites. Before further drainage could
be done, she developed sudden-onset dyspnoea,
tachypnoea, tachycardia, and hypoxia while straining at
stool. Bedside echocardiogram showed normal left and
right heart function with no right heart strain or
pericardial eusion. Chest radiograph (gure A) showed
a large right pleural eusion that had not been present
2 days earlier. An intercostal drain was inserted and 10 L
of uid were drained over 3 days. Repeat abdominal CT
showed no residual ascites. Review of the previous CT
showed focal herniation of the right hemidiaphragm into
the right pleural cavity (gure B) that was absent in the
later study. We believe that the diaphragmatic defect had
ruptured while our patient was straining, leading to
translocation of ascites into the pleural cavity. After
intercostal drain removal, she underwent regular
paracentesis to prevent another similar episode. Further
investigations conrmed portal hypertension; she had
a combined liver and kidney transplant in February, 2011.
At last follow-up in October, 2011, she was well with good
renal and hepatic function and no ascites.
B

Non-traumatic diaphragmatic rupture is uncommon


(28 published reports between 1956 and 2009) but has
been described after coughing, vomiting, exercise,
massage, and vaginal delivery.1 It can lead to translocation
of abdominal viscera into the thoracic cavity and most
commonly presents with dyspnoea, vomiting, and chest
and abdominal pain. Surgical repair is required in most
cases. Patients with ascites can also develop pleural
eusions as a result of diaphragmatic defects (often
small) in combination with the transdiaphragmatic
pressure gradient generated during respiration. These
eusions usually develop over weeks or months rather
than (as in our patient) acutely. In the context of portal
hypertension this disorder is referred to as hepatic
hydrothorax; it occurs in 610% of patients with
cirrhosis.2 Management is as for ascites but videoassisted thoracoscopic surgery and pleurodesis may also
be needed. Some patients have large pleural eusions
with little ascitesperhaps a result of translocation of
most of the uid caused by negative intrathoracic
pressure.3 Fluid leakage into the chest cavity can also
occur in patients on peritoneal dialysis;4 it is also
generally a chronic process but can occur acutely. Rightsided eusions are more common. The diagnosis can be
suggested by a high pleural-uid glucose concentration,
and radioisotope or contrast imaging might be needed
for conrmation. In some patients temporary cessation
of peritoneal dialysis is sucient to allow resolution and
subsequent resumption of peritoneal dialysis. However,
in others surgical intervention or a permanent switch to
haemodialysis is necessary. The dierential diagnosis of
sudden-onset dyspnoea is wide and includes conditions
such as pulmonary embolus, pneumothorax, and pulmonary oedema. However, in patients with ascites the
possibility of diaphragmatic rupture should also be
considered, particularly when there is a precipitant such
as a Valsalva manoeuvre.
Contributors
AC, SS, and MK looked after the patient and wrote the report. Written
consent to publish was obtained.
References
1
Losano JE, Edelman DA, Salwen WA, Basson MD. Spontaneous
rupture of the diaphragm: case report and comprehensive review
of the world literature. J Thorac Cardiovasc Surg 2010; 139: e12728.
2
Kiafar C, Gilani N. Hepatic hydrothorax: current concepts of
pathophysiology and treatment options. Ann Hepatol 2008; 7: 31320.
3. Roussous A, Phillipou N, Mantazris GJ, Gourgouliannis KI.
Hepatic hydrothorax: pathophysiology diagnosis and management.
J Gastroenterol Hepatol 2007; 22: 138893.
4. Leblanc M, Ouimet D, Pichette V. Dialysate leaks in peritoneal
dialysis. Semin Dial 2001; 14: 5054.

Figure: Right hemidiaphragm rupture and pleural eusion


(A) Chest radiograph showing large right pleural eusion; (B) CT showing focal herniation of the right
hemidiaphragm (arrow).

384

www.thelancet.com Vol 379 January 28, 2012