Topical review

Translation of symptoms and signs into mechanisms in neuropathic pain

Troels S. Jensen
a,b,
*
, Ralf Baron
c
a
Department of Neurology, Aarhus University, DK 8000 Aarhus C, Denmark
b
Danish Pain Research Center, Aarhus University, DK 8000 Aarhus C, Denmark
c
Klinik fur Neurologie, Christian-Albrechts-Universitat, Kiel, Germany
Keywords: Neuropathic pain; Symptoms; Signs; Mechanisms; Assessment
1. Introduction
For centuries, clinicians have been taught to examine and
classify patients on the basis of topographical lesion and the
underlying pathology. In most clinical specialities, such an
approach has been a key element in understanding the
pathophysiology of diseases and has led to progress in
terms of nding disease modifying or even disease curing
therapies. Examples are multiple including bacterial menin-
gitis, painful neuroborrelosis, osteoarthrosis, cancer, rheu-
matoid arthritis, ischaemic heart disease etc. In most of
these disorders, pain can be a major complaint, which
then rapidly disappears after the relevant therapy has been
given.
But what happens when the symptom itself becomes a
disease? When pain persists and becomes a chronic problem
and when the underlying diseases such as diabetes, cancer,
vasculitis are known, or cannot be cured? Are we then
helped by the classical Sherlock Holmes approach, rst,
to look for the crime site (topography of lesion) and
second, for the criminal (the disease) that caused this
pain? The short answer is: no. Clinical experience and
decades of rather discouraging systematic research mainly
related to therapy in chronic pain have shown that a strategy
directed at examining, classifying and treating pain on basis
of anatomy or underlying disease is of limited help to these
patients and their pain. These observations have then raised
the question whether an entirely different strategy in which
pain is analysed on the basis of underlying mechanisms
could be an alternative approach to examine and classify
patients to obtain a better outcome. Our increasing under-
standing of mechanisms underlying chronic pain together
with the discovery of new molecular targets for modifying
pain has strengthened the demand for other ways to treat
pain. Woolf and other authors (Woolf et al., 1998; Woolf
and Decosterd, 1999; Sindrup and Jensen, 1999) have
emphasised the rational for a treatment approach directed
at mechanism(s) rather than at diseases because new treat-
ments are being developed on basis of the biological
mechanisms that underlie the pain. One area that needs
such a new approach is neuropathic pain.
2. Neuropathic pain classication problems
According to the current International Association for
the Study of Pain (IASP) denition of neuropathic pain,
these disorders are characterised by lesions or dysfunction
of the system(s) that under normal conditions transmit
noxious information to the central nervous system. Thus
in theory, neuropathic pain should be easy to distinguish
from other conditions, but in practise, they are both dif-
cult to identify and to treat and there are several reason
why this is the case:
There is rarely one diagnostic test that can conrm or
refute the hypothesis of nerve lesion/dysfunction.
The perception of neuropathic (and other types of) pain
is a pure subjective phenomenon, which despite use of
the most sophisticated equipment cant be measured;
only correlates to the perceived content can be obtained.
The borderland between denite, probable, possible and
unlikely diagnoses is not clear. Prevalent disorders such
as cancer, low back pain, traumatic injuries may contain
a considerable (although as yet undetermined) neuro-
pathic component.
In contrast to other sensory systems, the pain system is
not static, but changes in a dynamic and somewhat
unpredictable fashion whenever the system has been
activated.
Pain 102 (2003) 18
0304-3959/03/$20.00 q 2003 International Association for the Study of Pain. Published by Elsevier Science B.V. All rights reserved.
doi:10.1016/S0304-3959(03)00006-X
www.elsevier.com/locate/pain
* Corresponding author. Tel.: 145-8949-3283; fax: 145-8949-3300.
E-mail address: tsj@akhphd.au.dk (T.S. Jensen).
Signs and symptoms of neuropathic pain may change
during the course of the disease and if it becomes
chronic.
There is at present no agreement on whether a restrictive
or a broader denition of neuropathic should be used;
the latter including dysfunctional disorders.
Systematic reviews of neuropathic pain treatment
(excluding simple entrapment disorders such as carpal
tunnel syndrome, meralgia paraesthetica or radial nerve
compression on the dorsal forearm) show that a moder-
ate or better pain relief is found only in approximately
one-third of the patients regardless of underlying disease
or anatomy of such disease (McQuay and Moore, 1998;
Sindrup and Jensen, 1999).
These constraints and limitations in our current concept
of neuropathic pain calls for another approach: to group
suspected neuropathic pain patients. There are specic
requirements to such a grouping as emphasised previously
(Woolf et al., 1998): it should be valid (the grouping
correspond to a specic pathological mechanism) and reli-
able (correspondence between examiners and between
results from one time point to the next), and the classica-
tion should be universal (applicable to all conditions,
mild as well as severe) The latter point may be difcult
to ascertain because of the dynamic nature of the nocicep-
tive system particularly under abnormal conditions. It
should be stressed that at this point of time, no studies
have provided a classication of symptoms and signs and
a scoring system, where the requirements are accounted
for.
When diseases and disorders are dominated by symp-
toms, which are merely subjective and the associated clin-
ical signs are few or non-existing the requirement for
validity and reliability becomes even more demanding. If
the correspondence between symptoms and signs on one
side and mechanism(s) on the other is understood, the
possibility for targeting neuropathic pain in a rational
way should be better. We will, in the forthcoming para-
graphs, briey review the possibilities for such an approach
and consider the practical requirements on how to pursue
in this avenue.
3. Mechanisms of neuropathic pain
Pathophysiological mechanisms underlying neuropathic
pain have been reviewed extensively within recent years
(Baron, 2000; Besson, 1999; Woolf and Salter, 2000;
Jensen et al., 2001; Hansson et al., 2001; Koltzenburg
and Scadding, 2001). The peripheral mechanism occurring
after peripheral nerve damage has been characterised in
great (although still incomplete) detail. Injured peripheral
nerve bres give rise to an intense and prolonged input of
ectopic activity to the central nervous system and in some
cases also secondary changes of the excitability of dorsal
horn neurons. At the cellular level formation of new chan-
nels, upregulation of certain receptors and downregulation
of others, altered local or descending inhibition are some of
the biological features that can contribute to a hyperexcit-
ability, which is assumed to be a sine qua non for chronic
pain. The neuronal hyperexcitability has a wide spectrum
of manifestations including increase in cellular excitability,
expansion of neuronal receptive elds, change of modality
to which neurons respond, recruitment of silent neurons or
circuits and a neuronal reorganisation in the dorsal horn
and further upstream (for review, see Woolf and Salter,
2000). It is, therefore, not surprising that these cellular
alterations subsequently give rise to neuroplastic changes
in which the somatosensory information can be distorted in
several ways including reorganisation of structures partici-
pating in the processing of noxious information. At this
point of time, we do not know the exact sequence of
changes in mechanisms and how they may inuence each
other. After lesioning central pathways, neurons in the
spinal cord and the brain with lost normal input may also
change their response characteristics and exhibit signs of
hyperexcitability in a fashion, mimicking that seen after
peripheral nerve injury (Vierck et al., 2000).
4. Symptoms, signs and mechanisms in neuropathic pain
The core in neuropathic pain is a lesion of the afferent
transmission system resulting in partial or complete loss of
input to the nervous system and a corresponding sensory
loss with negative sensory symptoms. The reduction of
afferent input caused by the nerve lesion is at the same
time the starting point for regeneration and disinhibition
with secondary development of hypersensitivity resulting
in various positive symptoms. Negative and positive
phenomena can be demonstrated either at the bedside or
in the laboratory. Combined with a pharmacological modu-
lation aimed at specic sites or at specic molecular targets,
it will be possible to gain some insight into mechanisms
participating in neuropathic pain.
Most likely symptoms, signs and mechanisms are
related, but probably not directly. For example, a diabetic
patient may have: steady pain, touch evoked pain, parox-
ysms and non-painful paraesthesia. In these cases, several
mechanisms can be involved such as tissue injury due to
ischaemia, sensitisation of peripheral receptors, ectopic
activity in sprouting regenerating bres, phenotypic
changes in dorsal root ganglion (DRG) cells, spinal reor-
ganisation etc. In other pain states, one mechanism may
give rise to different symptoms and signs. For example, a
peripheral nerve entrapment may cause paroxysms due to
ectopic activity from the lesioned nerve and an associated
(extra)-territorial brush evoked pain due to C-bre evoked
central sensitisation. Thus symptoms and mechanisms
involved in a particular pain condition cannot always be
predicted. Is it then impossible to dissect pathophysiologi-
T.S. Jensen, R. Baron / Pain 102 (2003) 18 2
cal mechanisms? It is probably not possible. As patients
report symptoms and not mechanisms and assessors
disclose signs and not mechanisms, the starting point
should be the patient with a focus on how symptoms and
signs translate into mechanisms rather than the vice versa.
4.1. Symptoms
An important point concerns the possible classication
of pain just on the basis of symptoms. There are at present
no data documenting such a classication. It has been
claimed that certain symptoms such as burning, smarting,
shock-like pains are characteristic for neuropathic pain, but
studies are not unanimous on this issue. The symptoms and
signs resulting from one particular mechanism cannot
always (if at all) be predicted because the plasticity gener-
ated in the nervous system implies an unpredictable chain
of events.
More recently, there has been suggestions that symptoms
also may have some bearing value in clarifying pain
mechanisms. Paroxysms are generally considered to be of
peripheral origin due to spontaneous ring in peripheral
nociceptive afferents. In tic doulourex, it has been
suggested that compression of the trigeminal root leads
to a hyperexcitability in a group of trigeminal ganglion
cells which then sets off an ignition focus that spreads
to other parts of the ganglion (Devor et al., 2002). New
observations (Otto et al., 2003) suggest that paroxysms also
could represent more central disturbance. These authors
found in patients with painful neuropathy that paroxysms
were related to a reduced function of small bre activity.
Other studies have been less successful in demonstrating
relationship between symptoms and mechanisms. Two
introduced neuropathic pain scales: the neuropathic pain
scale (NPS) (Galer and Jensen, 1997) and the Leeds assess-
ment of neuropathic symptoms and signs (LANSS) scale
(Bennett, 2001) need further evaluation in neuropathic pain
patients.
4.2. Signs
4.2.1. Peripheral sensitisation
Classical studies in patients with complex regional pain
syndrome (CRPS) type II have shown that a local anaes-
thetic block of an injured stump can reduce severe allody-
nia and signs of autonomic dysfunction (Livingston, 1998).
A large body of evidence obtained within the last decades
show that sensitisation of peripheral nociceptors can be the
fuel for generating activity in central connecting noxious
responding systems (Fields et al., 1998; Koltzenburg and
Scadding, 2001). Nystrom and Hagbarth (1981) in micro-
electrode recordings from transected nerves in amputees
showed spontaneous activity from stumps and blockade
of such activity with lidocain at the stump also relieves
the pain. Gracely et al. (1992) demonstrated in patients
with peripheral nerve injury that local anaesthetic block
of presumed ongoing C-bre activity in the injured region
abolished touch-evoked A-b allodynia. The functional
remodelling of synaptic organisation produced by afferent
C-bre activity was considered reversed by the peripheral
C-bre block. Moreover, in CRPS I, physiological sympa-
thetic activity and norepinephrine release enhances pain
indicating a pathological adrenergic sensitivity of nocicep-
tive bres (Baron et al., 2002).
Fields et al. (1998), along a similar line of thinking in
patients with postherpetic neuralgic (PHN), proposed the
existence of sensitised C-nociceptors as being a major
contributing factor to brush-evoked allodynia seen in
these patients. They observed an inverse relationship
between heat pain decit and ongoing pain indicating
that C-nociceptors may be a contributing factor to ongoing
pain (Rowbotham and Fields, 1996). In other neuropathic
pain states such as traumatic nerve injury and postmastect-
omy syndrome, similar peripheral sensitisation can be
demonstrated (Gottrup et al., 2000) showing that sensitised
C-nociceptors may indeed be the culprits for generating
certain types of spontaneous and evoked pains. Chabal et
al. (1989a) showed that modulating peripheral output by
local anaesthetizing stump neuromas with lidocaine
produced a reduction of tap evoked stump pain. In contrast,
perineuromal injection of a potassium channel blocker,
gallamine produced clear exacerbation of pain (Chabal et
al., 1989b). Local anaesthetics applied to the painful skin
in PHN patients produced signicant pain relief, supporting
the notion of an abnormal input from peripheral nocicep-
tors as an important pain generator.
4.2.2. Central sensitisation
A series of animal studies have shown the importance of
central mechanisms in maintaining pain and in generating
pain by non-noxious input. Human studies conrm that
activity in large myelinated A-b bres can maintain neuro-
pathic pain. After nerve injury, innocuous tactile stimuli get
access to dorsal horn neurons via low-threshold mechanor-
eceptors from A-b bres. If this A-b bre input is blocked,
the allodynia disappears, but burning spontaneous pain
persists indicating that the latter probably is mediated by
C-nociceptors.
While there is evidence that increasing activity in C-
nociceptors in certain neuropathic pain conditions produces
an increase in central sensitisation, there are also indica-
tions that some patients can have severe loss of C-bre
functions despite the presence of extensive allodynia in
the area corresponding to C-bre loss and pain. Baron
and Saguer (1993) in a study of PHN patients using C-
bre mediated histamine axon reexes to determine C-
bre activity showed abolished responses in areas with
marked allodynia suggesting that in this case, the allodynia
is a pure central phenomenon. In central pain conditions,
observations are less easy to interpret because of the
increasing complexity in ascending transmitting somato-
sensory systems and the multitude of modulating circuits.
Findings in post-stroke pain show that pain occurs in body
T.S. Jensen, R. Baron / Pain 102 (2003) 18 3
areas that have lost their normal patterned input and that
areas of pain occupy only a fraction of the area with abnor-
mal sensory function. These ndings indicate similar to
peripheral neuropathic pain conditions that lost input
causes a secondary sensitisation in populations of central
neurons, which then are both responsible for the pain and
the combination of negative and positive sensory ndings
(Boivie et al., 1999; Finnerup et al., 2003). In agreement
with such central sensitisations either in the spinal cord or
more rostrally pharmacological agents with an action on
cellular hyperexcitability such as opioids, gabapentin, lido-
caine and lamotrigine (Attal et al., 2000, 2002; Vestergaard
et al., 2001; Finnerup et al., 2002) can reduce pain and
abnormal sensitivity.
4.3. Pharmacological tests
An additional approach for classifying pain involves the
use of specic pharmacological agents. Based on their mode
of action and their different molecular targets, pharmacolo-
gical drugs may be used to determine whether a particular
symptom can be modulated by a drug with a specic action
(Attal et al., 2000; Jensen et al., 2001). By using different
modes of administration of the same agent (topical,
systemic, epidural, intrathecal etc.), the site of action can
be determined and the pain-generating site identied.
Several agents have been used. In particular the use of
opioids, N-methyl d-aspartate (NMDA) antagonist and
anticonvulsants have provided insight into the pharmacol-
ogy of various phenomena encountered in neuropathic pain.
The NMDA antagonist ketamine in subanaesthetic
concentrations has by systemic administration been shown
to reduce ongoing pain, brush-, pinprick-evoked pain in
traumatic nerve injuries in amputees, posttraumatic nerve
injuries and PHN (Eide et al., 1994; Nikolajsen et al., 1996).
Interestingly, the NMDA antagonist dextromethorphane
was effective in diabetic painful neuropathy and not in
PHN indicating different underlying mechanisms (Sang et
al., 2002).
NMDA receptors have also been demonstrated in the skin
with a location on unmyelinated and myelinated axons
(Carlton et al., 1995). Topical administration has in some
studies been able to demonstrate a reduction in experimental
mechanical hyperalgesia, while others have failed to nd
such an effect. There are no indications at this point that
peripheral NMDA receptors are involved in clinical neuro-
pathic pain.
Local and intravenous administered fast acting opioids
such as fentanyl are likewise useful in clarifying the possi-
ble involvement of peripheral as opposed to systemic opioid
receptors in neuropathic pain (Eide et al., 1994; Attal et al.,
2002).
Local anaesthetics are widely used in treating neuropathic
pain. Local anaesthetics, in addition to their ability to block
nerve impulse trafc, have an effect on damaged nociceptive
neurons without affecting nerve conduction and the percep-
tual responses to noxious stimuli. Intravenously adminis-
tered lidocaine can reduce ongoing and different types of
evoked pain both in experimental conditions and in various
neuropathic pain conditions peripheral as well as central.
The use of topical administration of lidocaine has been
one of the tools used to classify mechanisms in PHN (Fields
et al., 1998).
5. Assessment of neuropathic pain
Assessment of neuropathic pain involves a series of
systematic steps. A key problem in the existing literature
is that the diagnostic work-up of patients apparently varies
from one laboratory to another. Trivial as it may be, simple
differences in questioning, differences in equipment used
and changes in sequence of testing are obvious sources for
variability among examiners in the same patient (Hansson
et al., 2001; Jensen et al., 2001). Other confounding factors
may contribute to an unclear picture of what is and what is
not neuropathic pain. The diagnostic work-up of patients
can be exceedingly complex including sophisticated quan-
titative sensory testing, neurophysiological studies,
imaging and pharmacological tests. While this may be of
interest in some cases, the key question is whether we can
obtain a meaningful classication of patients also in terms
of mechanisms in a large group of patients in a reasonable
time using simple bedside equipment.
We propose that this is possible. Below, we have listed a
framework on how this can be done in the clinic. The propo-
sal is for obvious reasons open to criticism. The important
point here is not whether this is complete or not, or whether
it is correct or not. The important thing is that basically,
scientists and clinicians need to join together and nd a
consensus about how these patients should be assessed
both for clinical and research purposes. By doing so, a clas-
sication can be established and proper epidemiological,
clinical, genetic, therapeutic and other trials can be
performed also on a large scale.
5.1. Medical history
The history should clarify pain location, distribution,
intensity, quality and time course as well as the underlying
disease and possibly document the nervous system lesion
responsible for pain. Separation into stimulus-independent
and stimulus-dependent pain is useful because it allows
separating ongoing activity from provoked activity. Careful
history will allow such separation. Patients may describe
their pains in a variety of ways: unpleasant, pricking, stick-
ing, burning, scalding, aching or deep sore pain. Therefore,
comparison across patients assessors and laboratories
requires similarities in applied denitions, questioning
and symptom presentation. A characteristic in many neuro-
pathic pain conditions is the presence of allodynia follow-
ing exposure to non-painful cold. In such cases, patients
may describe their pain in a variety of ways: cold, wet, ice-
T.S. Jensen, R. Baron / Pain 102 (2003) 18 4
like or even in a paradox manner like burning-hot or ice
burning (like holding a snow-ball in the hand). Some
patients with central pain complain of pain by movement
in which the movement itself elicit a tightening, squeezing
or burning sensation in the skin. At other times, the pain is
one of paroxysms with stabbing, shooting, lancinating
types of pain. Unless patients are questioned in the same
manner, comparison will not be possible. The use of
various scales such as the McGill pain questionnaire,
NPS and the LANSS scale are undoubtedly important
steps in the direction of nding the most suitable neuro-
pathic pain questionnaire. Table 1 presents a suggestion of
history parameters that need to be recorded. The underly-
ing disease is, of course, also important even in terms
mechanisms of pain. As recently pointed out in an animal
experimental neuropathic study, the pain behaviour asso-
ciated with a nerve crush is different form that seen after a
nerve ligation (Decosterd et al., 2002.)
5.2. Clinical examination
The sensory examination should assess negative sensory
symptoms and ndings as well as positive sensory symp-
toms and signs (Tables 2 and 3). By using careful sensory
testing, the characteristic sensory ndings corresponding to
such symptoms can be detected and quantied. The stimu-
lus-evoked pains are classied according to its dynamic or
static character. A sensory examination at bedside will often
include: pinprick, touch, pressure, cold, heat and vibration.
Pinprick sensation can be assessed by the response to
pinprick stimuli, touch by gently applying cotton wool to
the skin, deep pain by gentle pressure, cold and warm sensa-
tion by measuring the response to a specic cold or warm
thermal stimulus e.g. thermorollers kept at 20 and 458C,
respectively. Cold sensation can also be assessed by the
response to acetone or menthol. Vibration can be assessed
by a tuning fork placed at strategic points (malleol, inter-
phalangeal joints etc.). At present, there is no consensus
about what, where and how to measure and what to compare
with. It is generally agreed that assessment should be carried
out in the area with maximal pain using the contralateral
area as control. However, contralateral segmental changes
following a unilateral nerve or root lesion cannot be
excluded so an examination at mirror sites may not neces-
sarily represent a true control site.
At bedside, the response can be graded as: normal,
decreased or increased (Andersen et al., 1995). This is a
simple way to determine whether negative or positive
phenomena are involved.
If hyperaesthetic, the response is classied as dysaes-
thetic, hyperalgesic or allodynic. A correlation of sponta-
neous pain and sensory response in the painful area suggest
that the two phenomena are reections of the same phenom-
enon: a central sensitisation of dorsal horn neurons (Rowbo-
tham and Fields, 1996; Gottrup et al., 2003).
A more sophisticated sensory testing has been proposed
using neurophysiological and imaging techniques to assess
the various positive phenomena (Table 3). When present,
allodynia or hyperalgesia can be quantitated by measuring
intensity, threshold for elicitation, duration and area of allo-
dynia.
5.2.1. Windup-like pain and aftersensations
Windup-like pain or abnormal temporal summation is the
clinical equivalent to increasing neuronal activity following
repetitive C-bre stimulation .0.3 Hz (Mendell and Wall,
T.S. Jensen, R. Baron / Pain 102 (2003) 18 5
Table 2
Assessment of negative sensory symptoms or signs in neuropathic pain
Negative sensory symptoms/
signs
Bedside examination Laboratory examination Mechanism
Reduced touch Touch skin with cotton wool Graded von Frey hair A-b bres
Reduced pin prick Prick skin with a pin single stimuli von Frey hair specic (e.g. 100g) A d bres
Reduced cold/warm Thermal response to cold 20 and 45 Detection/pain threshold cold warm A d/C-bres
Reduced vibration Tuning fork on malleol Vibrameter A-b bres
Table 1
Medical history in neuropathic pain
Pain complaint
(most prevalent)
Duration Interference with
daily activity
(010)
Character
(burning, shock-like, pins
and needles, aching etc.)
Average intensity
last week/last day
(010)
Nerve territory
(based on anatomic
drawing)
Extraterritorial
spread (based on
anatomic drawing)
1
2
3
4
T
.
S
.
J
e
n
s
e
n
,
R
.
B
a
r
o
n
/
P
a
i
n
1
0
2
(
2
0
0
3
)
1

8
6
Table 3
Assessment of positive sensory symptoms or signs in neuropathic pain
Positive sensory symptoms/signs Bedside exam Laboratory test Specic test (topical, systemical) Mechanism(s)
Spontaneous
Paraesthesia Grade (010) Area in cm
2
grade (010) None Spontaneous activity in LT A-b afferents
Dysaesthesia Grade (010) Area in cm
2
grade (010) Pharmacological Spontaneous activity in C/A-d afferents
Paroxysms Number Grade (010) Threshold for evocation Pharmacological Spontaneous activity in C-nociceptors
Supercial burning pain Grade (010) Area in cm
2
grade (010) Capsaicin provocation pharmacological Spontaneous activity in C-nociceptors?
Deep pain Grade (010) Area in cm
2
grade (010) Pharmacological Spontaneous activity in joint/muscle nociceptors?
Evoked
Touch evoked hyperalgesia Stroking skin with painters brush Block (ischaemia, compression)
pharmacological
Central sensitisation:
1. C-bre input
2. Lost C-bre input
Static hyperalgesia Gentle mechanical pressure Evoked pain to pressure Peripheral sensitisation
Punctate hyperalgesia Pricking skin with pin von Frey hair Pharmacological Central sensitisation:
A-d bre input
Punctate repetitive hyperalgesia
(windup-like pain)
Pricking skin with pin 2/s for
30 s
von Frey hair Block (ischaemia, compression)
pharmacological
Central sensitisation: A-d bre input
Aftersensation Measure pain duration after
stimulation
Measure pain duration after
stimulation
Block (ischaemia, compression)
pharmacological
Central sensitisation
Cold hyperalgesia Stim skin with cool metal roller Evoked pain to cold stimuli Pharmacological 1. Central sensitisation
2. Central disinhibition
Heat hyperalgesia Stim skin with warm metal roller Evoked pain to heat stimuli Pharmacological Peripheral sensitisation
Chemical hyperalgesia Topical capsaicin Topical capsaicin Menthol/Capsaicin/histamine test Peripheral sensitisation
Sympathetic maintained pain none Sympathetic blockade,
Modulation of sympathetic
outow
Sympathetic-afferent coupling
1965). The windup-like pain can be produced by a variety of
stimuli including mechanical, thermal and electrical types.
Aftersensations the persistence of pain long after termina-
tion of a painful stimulus is another characteristic feature
of neuropathic pain, which is closely related to a coexistent
dynamic or static hyperalgesia (Gottrup et al., 2003).
Consistent with this notion, windup-like pain and aftersen-
sations may both reect neuronal discharges in wide
dynamic range neurons.
By combining symptoms and signs and disease cate-
gories, it can be tested whether certain symptom clusters
across disease categories or within the same disease cate-
gory are linked to specic mechanisms (see above).
5.3. Pharmacological testing
Acute pharmacological tests using a double-blind placebo
controlled technique with either different drugs or different
administration forms of the same substance permit an exam-
ination of the location of the pain generator and the mole-
cular mechanism involved in pain. If testing further is
separated into an examination of ongoing pain and various
evoked responses (see Table 3) insight can also be obtained
into the molecular mechanisms involved in a particular
phenomenon or mechanism.
6. Conclusion
A mechanism-based analysis and eventually classica-
tion of neuropathic pain is an attractive approach for several
reasons. However, at present it is unclear if this in the daily
clinic will result in a better treatment outcome of pain
patients. It will be of interest to determine the possible addi-
tional yield provided by a hierarchical structured system
that classies pain on the basis of: (1) symptoms, (2)
symptoms 1signs, (3) symptoms 1signs 1mechanisms
and (4) symptoms 1signs 1mechanisms 1
pharmacological analysis (Rasmussen et al. unpublished
observations). This can be done once there is consensus
about what the content should be in the examination pack-
age.
Acknowledgements
The studies on which the present paper is based have in
part been supported by grants from Danish Pain Research
Center, Danish Medical Research Council (No. 9502209),
the Cancer Society (No. 78004), Institute for Experimen-
tal. Clinical Research University of Aarhus, Karen Elise
Jensens Foundation, the German Research Foundation (Ba
1921), the German Research Network on Neuropathic Pain
(BMBF) and the GermanDanish Commission of Kiel
University.
References
Andersen G, Vestergaard K, Ingeman-Nielsen M, Jensen TS. The incidence
of central post-stroke pain. Pain 1995;61:187193.
Attal N, Gaude V, Brasseur L, Dupuy M, Guirimand V, Parker F, Bouhas-
sira D. Intravenous lidocaine in central pain: a double-blind, placebo-
controlled psychophysical study. Neurology 2000;54:564574.
Attal N, Guirimand F, Brasseur L, Gaude V, Chauvin M, Bouhassira D.
Effects of iv morphine in central pain: a randomised placebo-controlled
study. Neurology 2002;58:554563.
Baron R, Schattschneider J, Binder A, Siebrecht D, Wasner G. Relation
between sympathetic vasoconstrictor activity and pain and hyperalgesia
in complex regional pain syndrome: a casecontrol study. Lancet
2002;359:16551660.
Baron R, Saguer M. Postherpetic neuralgia. Are C-nociceptors involved in
signalling and maintenance of tactile allodynia. Brain 1993;116:1477
1496.
Baron R. Peripheral neuropathic pain: from mechanisms to symptoms. Clin
J Pain 2000;16:s12s20.
Bennett M. The LANSS pain scale: the Leeds assessment of neuropathic
symptoms and signs. Pain 2001;92:147157.
Besson JM. The neurobiology of pain. Lancet 1999;353:16101615.
Boivie J. Central pain. In: Wall PD, Melzack R, editors. Textbook of pain,
3rd ed.. Edinburgh: Churchill Livingstone, 1999. pp. 871902.
Chabal C, Russell LC, Burchiel KJ. The effect of intravenous lidocaine,
tocainide and mexiletine on spontaneously active bers originating in
rat sciatic neuromas. Pain 1989a;38:333338.
Chabal C, Jacobson L, Russell LC, Burchiel KJ. Pain responses to perineur-
omal injection of normal saline, galamine and lidocaine in humans. Pain
1989b;36:321325.
Carlton SM, Hargett GL, Coggeshall RE. Localization and activation of
glutamate receptors in unmyelinated axons of rat glabrous skin.
Neurosci Lett 1995;197:2528.
Decosterd I, Allchorne A, Woolf CJ. Progressive tactile hypersensitivity
after a peripheral nerve crush: non-noxious mechanical stimulus-
induced neuropathic pain. Pain 2002;100:155162.
Devor M, Amir R, Rappaport ZH. Pathophysiology of trigeminal neuralgia:
the ignition hypothesis. Clin J Pain 2002;18:413.
Eide PK, Jrum E, Stubhaug A. Relief of post-herpetic neuralgia with the
N-methyl-d-aspartic receptor antagonist ketamine: a double-blind,
cross-over comparison with morphine and placebo. Pain
1994;54:347354.
Fields HL, Rowbotham M, Baron R. Postherpetic neuralgia: irritable noci-
ceptors and deafferentation. Neurobiol Dis 1998;5:209227.
Finnerup BN, Sindrup SH, Bach FW, Johannesen IL, Jensen TS. Lamotri-
gine in spinal cord injury pain: a randomised controlled clinical trial.
Pain 2002;96:375383.
Finnerup NB, Johannesen IL, Fuglsang-Federiksen A, Bach FW, Jensen
TS. Sensory function in spinal cord injury patient with and without
central pain. Brain 2003;126:5770.
Galer BS, Jensen MP. Development and preliminary validation of a pain
measure specic to neuropathic pain. The neuropathic pain scale.
Neurology 1997;48:332338.
Gottrup H, Andersen J, Arendt-Nielsen L, Jensen TS. Psychophysical
examination in patients with postmastectomy pain. Pain
2000;87:275284.
Gottrup H, Kristensen AD, Bach FW, Jensen TS. Aftersensations in experi-
mental and clinical hyperalgesia. Pain 2003; (in press).
Gracely RH, Lynch SA, Bennett GJ. Painful neuropathy: altered central
processing, maintained dynamically by peripheral input. Pain
1992;51:175194.
Hansson P, Lacerenza M, Marchettini P. Aspects of clinical and experi-
mental neuropathic pain: the clinical perspective. In: Hansson PT,
Fields HL, Hill RG, Marchettini P, editors. Neuropathic pain: patho-
physiology and treatment, progress in pain research and management
vol. 21. Seattle, WA: IASP Press, 2001. p. 118.
T.S. Jensen, R. Baron / Pain 102 (2003) 18 7
Jensen TS, Gottrup H, Bach FW, Sindrup SH. The clinical picture of
neuropathic pain. Eur J Pharmacol 2001;429:111.
Koltzenburg M, Scadding J. Neuropathic pain. Curr Opin Neurol
2001;14:641647.
Livingston WK. In: Fields HL, editor. Pain and suffering. Seattle, WA:
IASP Press, 1998. p. 1250.
McQuay HJ, Moore RA. An evidence-based resource for pain relief.
Oxford: Oxford University Press, 1998.
Mendell LM, Wall PD. Responses of single dorsal cord cells to peripheral
cutaneous unmyelinated bers. Nature. 1965;206:9799.
Nikolajsen L, Hansen C., Nielsen J, Keller J, Arendt-Nielsen L, Jensen TS.
The effect of ketamine on phantom pain: a neuropathic disorder main-
tained by peripheral input. Pain 1996;67:6977.
NystromB, Hagbarth KE. Microelectrode recordings fromtransected nerves
in amputees with phantom limb pain. Neurosci Lett 1981;27:211216.
Otto et al., 2003Otto M, Bak S, Bach FW, Jensen TS, Sindrup SH. Pain
phenomena and possible mechanisms in patients with painful poly-
neuropathy. Pain 2003;101:187192.
Rowbotham, MC, Fields HL. The relationship of pain, allodynia and ther-
mal sensation in post-herpetic neuralgia. Brain 1996;119:347354.
Sang CN, Booher S, Gilron I, Parada S, Max MB. Dextromethorphan
and memenatine in painful diabetic neuropathy and postherpetic neur-
algia: efcacy and doseresponse trials. Anesthesiology 2002;96:
10531061.
Sindrup SH, Jensen TS. Efcacy of pharmacological treatments of neuro-
pathic pain: an update and effect related to mechanism of drug action.
Pain 1999;83:389400.
Vestergaard K, Andersen G, Gottrup H, Kristensen BT, Jensen, TS. Lamo-
trigine for central post-stroke pain: a randomised controlled trial.
Neurology 2001;56:184190.
Vierck CJ, Siddall P, YezierskiRP. Pain following spinal cord injury:
animal studies and mechanistic studies. Pain 2000;89:15.
Woolf CJ, Bennett GJ, Doherty M, Dubner, R, Kidd, B, Koltzenburg, M,
Lipton, R, Loeser JD, Payne R, Torebjork E. Towards a mechanism-
based classication of pain? Pain 1998;77:227229.
Woolf CJ, Decosterd I. Implications of recent advances in the understand-
ing of pain pathophysiology for the assessment of pain in patients. Pain
1999;(Suppl. 6):s141148.
Woolf CJ, Salter MW. Neuronal plasticity: increasing the gain in pain
Science 2000;288:17651768.
T.S. Jensen, R. Baron / Pain 102 (2003) 18 8