THERAPY IN PRACTICE Drugs Aging 2003; 20 (1): 23-57

1170-229X/03/0001-0023/$30.00/0

Adis International Limited. All rights reserved.

Demographics, Assessment and

Management of Pain in the Elderly
Mellar P. Davis and Manish Srivastava
Harry R. Horvitz Center for Palliative Medicine, Cleveland, Ohio, USA

Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1. Demographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.1 Cancer Pain in the Elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.2 Benign Pain in the Elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.3 Vitamin D Deficiency in the Elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2. Pain Management in the Elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1 Unique Barriers to Pain Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2 Consequences of Uncontrolled Pain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3 Physiological Changes with Aging that Influence Pain Management . . . . . . . . . . . . . 26
2.4 Principles of Pain Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.5 Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.6 Biophysical and Biopsychosocial Therapies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3. Nonpharmacological Management of Pain in the Elderly . . . . . . . . . . . . . . . . . . . . . . . 29
4. Approaches to Pharmacological Management of Pain in the Elderly . . . . . . . . . . . . . . . . 29
4.1 Whether To Separate Malignant from Nonmalignant Pain . . . . . . . . . . . . . . . . . . . . 29
4.2 Principles of Good Prescribing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5. Paracetamol (Acetaminophen) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
6. NSAIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
6.1 Gastrointestinal Toxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.2 Renal Toxicity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.3 Cyclo-Oxygenase-2 Inhibitors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
7. Corticosteroids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
7.1 Prednisone and Prednisolone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
7.2 Dexamethasone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
8. Agents Preventing Bone Resorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
8.1 Calcitonin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
8.2 Bisphosphonates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
9. Opioids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
9.1 Adverse Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
9.2 Other Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
9.3 Opioid Classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
9.4 Opioids in Hepatic and Renal Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
9.5 Drug Interactions with Opioids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
10.Pharmacological Treatment of Neuropathic Pain in the Elderly . . . . . . . . . . . . . . . . . . . . 46
10.1 Antiepileptic Drugs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
10.2 Tricyclic Antidepressants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
10.3 Local Anaesthetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
10.4 Other Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
11.Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
24 Davis & Srivastava

Abstract The prevalence of pain increases with each decade of life. Pain in the elderly
is distinctly different from pain experienced by younger individuals. Cancer is a
leading cause of pain; however, other conditions that cause pain such as facet
joint arthritis (causing low back pain), polymyalgia rheumatica, Pagets disease,
neuropathies, peripheral vascular disease and coronary disease most commonly
occur in patients over the age of 50 years. Poorly controlled pain in the elderly
leads to cognitive failure, depression and mood disturbance and reduces activities
of daily living. Barriers to pain management include a sense of fatalism, denial,
the desire to be the good patient, geographical barriers and financial limitations.
Aging causes physiological changes that alter the pharmacokinetics and phar-
macodynamics of analgesics, narrowing their therapeutic index and increasing
the risk of toxicity and drug-drug interactions. CNS changes lead to an increased
risk of delirium.
Assessment among the verbal but cognitively impaired elderly is satisfactorily
accomplished with the help of unidimensional and multidimensional pain scales.
A comprehensive physical examination and pain history is essential, as well as a
review of cognitive function and activities of daily living. The goal of pain man-
agement among the elderly is improvement in pain and optimisation of activities
of daily living, not complete eradication of pain nor the lowest possible drug
dosages. Most successful management strategies combine pharmacological and
nonpharmacological (home remedies, massage, topical agents, heat and cold
packs and informal cognitive strategies) therapies.
A basic principle of the pharmacological approach in the elderly is to start
analgesics at low dosages and titrate slowly. The WHOs three-step guideline to
pain management should guide prescribing. Opioid choices necessitate an under-
standing of pharmacology to ensure safe administration in end-organ failure and
avoidance of drug interactions. Adjuvant analgesics are used to reduce opioid
adverse effects or improve poorly controlled pain. Adjuvant analgesics (NSAIDs,
tricyclic antidepressants and antiepileptic drugs) are initiated prior to opioids for
nociceptive and neuropathic pain. Preferred adjuvants for nociceptive pain are
short-acting paracetamol (acetaminophen), NSAIDs, cyclo-oxygenase-2 inhibi-
tors and corticosteroids (short-term). Preferred drugs for neuropathic pain include
desipramine, nortriptyline, gabapentin and valproic acid. Drugs to avoid are pen-
tazocine, pethidine (meperidine), dextropropoxyphene and opioids that are both
an agonist and antagonist, ketorolac, indomethacin, piroxicam, mefenamic acid,
amitriptyline and doxepin. The type of pain, and renal and hepatic function, alter
the preferred adjuvant and opioid choices. Selection of the appropriate analgesics
is also influenced by versatility, polypharmacy, severity and type of pain, drug
availability, associated symptoms and cost.

1. Demographics functionally dependent or require assistance with

1.1 Cancer Pain in the Elderly
Advanced cancer is common in the geriatric
population. Approximately 80% of patients with
cancer experience pain,[1] and 55% of patients with
cancer who are aged between 70 and 91 years are
activities of daily living.[2] In addition, 23% of
these patients are significantly depressed and 30%
have fair to poor nutrition. Unlike younger pa-
tients, 15% have underlying dementia.[2] Most are
taking multiple medications: 57% are on three or
more medications, 35% on five or more and 7% on

Adis International Limited. All rights reserved. Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
ten or more.[2,3] The mean number of associated
comorbid conditions is 3.5; these include cardio-
vascular disease, congestive heart failure, chronic
obstructive lung disease, diabetes, arthritis, hyper-
tension and osteoporosis.[2] Many have chronic
nonmalignant pain predating the cancer pain. Most
patients with advanced cancer have multiple pains,
with 67% of pains related to the underlying malig-
nancy and 25% as a result of treatment. The most
common cancer-related pain arises from bone me-
tastases. Neuropathic pain is experienced in 25
30% of patients.
1.2 Benign Pain in the Elderly
Age-specific morbidity rates for pain increase
per decade of life.[4] The prevalence of pain com-
plaints among the elderly (aged >65 years) is as
high as 6780%.[1,4,5] Pain that is functionally
impairing affects 4580% of nursing home resi-
dents;[6] half of these have daily pain. Common
comorbid conditions causing pain include: low
back pain from facet joint arthritis; osteoarthritis
and osteoporosis; previous bone fractures; rheu-
matoid arthritis and polymyalgia rheumatica;
Pagets disease; peripheral neuropathies and
neuropathic pain associated with stroke, shin-
gles, diabetes, trigeminal neuralgia and nutri-
tional neuropathies; peripheral vascular disease;
and coronary artery disease.
1.3 Vitamin D Deficiency in the Elderly
Home-bound elderly, particularly those on anti-
epileptic drugs or who malabsorb fat, are at high
risk for vitamin D deficiency, which causes deep
musculoskeletal and/or superficial light touch
pain. One-third of nursing home patients confined
indoors for 6 months develop vitamin D defi-
ciency, and as many as one-half of community
dwellers have reduced vitamin D levels.[3] A single
dose of 100 000IU restores vitamin D levels and
reduces pain.[3]
Adis International Limited. All rights reserved.
25
2. Pain Management in the Elderly
2.1 Unique Barriers to Pain Management
Cognitive impairment is one of the major bar-
riers to assessment of pain in the elderly.[7,8] Most
elderly patients with cognitive impairment who are
verbal can be assessed by a unidimensional pain
scale.[9] Cultural barriers among geriatric patients
from minority groups impair the ability to elicit a
pain history and determine pain severity because
of language difficulties and culturally unique ex-
pressions of pain.[8]
The elderly have fewer resources to pay for an-
algesics, which results in failure to fill prescrip-
tions. They usually have fixed incomes, lack Medi-
care reimbursement for prescription medication in
the US, and experience formulary limitations by
managed-care insurance companies and delays
from national mail-order pharmaceutical compa-
nies. Many pharmacies have limited opioid avail-
ability or do not carry opioids.[3,10]
Comorbidities complicate the management of
pain, particularly with disease-related decrease in
end-organ function.[1,3] Additional barriers in-
clude fatalism, denial, the philosophy of the good
patient, geography (home-bound), fear of adverse
effects, fear of loss of independence, fear of ad-
diction and fear of tolerance.[5,10-12] Some elderly
patients believe that pain medications may need to
be saved until pain becomes intolerable or that
opioids are inappropriate for severe cancer pain
until death is imminent.[12] The fear of pain as an
indicator of either serious disease or progressive
cancer leads patients to under-report pain.[12]
Negative attitudes toward the elderly by young
professional caregivers may be sensed by older
patients, which leads to avoidance. Depression
often associated with pain in the elderly may lead
to social isolation and failure to seek help with
intractable pain.[10]
Physician-generated barriers include failure to
assess patients fully because of lack of knowledge
and time.[13] Physicians frequently lack knowledge
of nonpharmacological approaches to pain man-
agement and the pharmacology of analgesics.[14]
Drugs Aging 2003; 20 (1)
26
Physicians fear opioid toxicity and fail in the pro-
active prevention of opioid adverse effects, which
increase with age.[15] Fear of opioid dependence
and legal scrutiny are also significant barriers
among inexperienced prescribers.[16]
2.2 Consequences of Uncontrolled Pain
Unrelieved pain increases the risk of cognitive
failure, specifically memory and attention span.[17]
Uncontrolled pain leads to diminished daytime en-
ergies, loss of normal sleep architecture, less deep
slow-wave sleep and more brief arousals through
the night.[18] Sleep-deprived individuals have
lower pain thresholds and thereby experience
greater pain.[5,18] Uncontrolled pain is associated
with loss of physical function, increased depres-
sion and greater mood disturbances.[5,18] Memory
complaints are not only related to depression with
chronic pain, but also increase with the chronicity
of pain.[19] Attention and the ability to perform
complex tasks are diminished with severe chronic
pain.[20] Fear is frequently associated with pain,
and the consequences of such fear are prevention
of evaluation, intensification of pain, the promo-
tion of functional disability and depression.[5] Poor
pain control is an important determinant of quality
of life, especially among those in long-term care
facilities and near the end of life. The elderly can
also use pain for secondary gain.[3] Finally, there
is increased healthcare expenditure associated
with unrelieved pain.[21]
2.3 Physiological Changes with Aging that
Influence Pain Management
The elderly are more likely to have adverse ef-
fects from the pharmacological management of
pain, because the drugs used have a narrower ther-
apeutic index as a result of reduced or impaired
renal and hepatic function compared with that
in younger patients. As a result, the elderly are
more sensitive to the analgesic properties and ad-
verse effects of opioids because of age-associated
hepatic, renal and CNS pharmacodynamic alter-
ations.[10]
Adis International Limited. All rights reserved.
Davis & Srivastava
Physiological changes with age include a de-
creased cardiac index of 1% per year and reduced
renal function of 1% per year after the age of 50
years, decreased hepatic blood flow, decreased he-
patic mass, reduced hepatic mono-oxygenases and
cytochrome enzymes with relative preservation of
conjugases, and decreased plasma protein binding.
Hepatic metabolism is reduced by 3040%.[22]
Such alterations in liver function increase oral bio-
availability of certain opioids by reducing hepatic
extraction.[22] Aging is also associated with in-
creased body fat, which increases the volume of
distribution of lipophilic medications and delays
both elimination and the onset of drug action with-
out changing serum concentrations. Age-related
reduced volume of distribution increases plasma
concentrations of hydrophilic drugs, which sec-
ondarily increases drug diffusion to receptor sites.
Polypharmacy due to multiple comorbidities in-
creases the potential for pharmacokinetic and
pharmacodynamic drug-drug interactions.
The elderly experience enhanced response to
CNS-active medications.[22] Subcortical and cor-
tical atrophy decreases receptor sites, receptor
affinity and synthesis of the neurotransmitters
acetylcholine, dopamine, GABA and noradrena-
line (norepinephrine). The elderly have a 28%
reduction in cerebral blood flow. Neuronal loss
is greatest in the neocortex and hippocampus,
particularly in the region of the locus ceruleus
and substantia nigra, and may influence pain. G-
proteins, which are secondary mediators of the
action on many receptors, including opioid re-
ceptors, have decreased coupling. There is re-
duced catalytic activity of adenylate cyclase,
which is also a secondary mediator of pain.[22]
Many of these changes predispose the elderly to
drug-induced delirium and reduced drug toler-
ance.[23] Overt delirium of any cause contributes
to relative opioid resistance and uncontrolled
pain.[24]
The expression of pain is altered with advanced
underlying dementia, even though the sensory reg-
istration of pain is well preserved. Executive func-
tions that govern the expression of pain are signif-
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
icantly reduced by dementia and cognitive failure,
leading to atypical behaviours and reactions with
pain.[12]
There is decreased saliva formation in the el-
derly, which is further reduced by multiple drugs
that have anticholinergic properties; this can de-
lay absorption of sublingual tablets and matrix-
embedded medications.[22] Stomach acid produc-
tion is reduced, though this does not appear to
greatly influence drug absorption.[22] Peristalsis is
reduced, increasing the risk of constipation, nau-
sea and vomiting from opioids.[22] Fortunately,
small-bowel absorption is not influenced to a sig-
nificant extent by aging and is not rate limiting.[22]
Decrease in sympathetic innervation of jux-
taglomerular cells within the kidney in the elderly
reduces aldosterone and renin release, leading to in-
creased risk for hyperkalaemia, particularly with
NSAIDs, by further reducing renin levels.[22] The
elderly are at increased risk for orthostatic hypo-
tension with adjuvants such as tricyclic antidepres-
sants or phenothiazines because of age-related di-
minished baroreceptor response.
In summary, senescence causes pharmacoki-
netic alterations that lead to higher concentrations
of drugs at receptor sites and delayed elimination
of lipophilic medications, as a result of increased
body fat, and of hydrophilic medications, as a re-
sult of reduced renal and/or hepatic function. Re-
duced receptor sites are offset by pharmacokinetic
enhancement by both reduced volume of distribu-
tion and reduced renal and hepatic clearance, par-
ticularly with hydrophilic medications.[22] Com-
pared with younger patients, the elderly develop a
high frequency of unusual manifestations of ad-
verse drug reactions, such as cognitive failure, loss
of bladder or bowel control and anorexia.
2.4 Principles of Pain Management
The approach to chronic pain management is
two-fold: defining the underlying pathology of
pain and measuring its consequences, and treating
both.[5] There are three principal elements to pain
management:[25]
comprehensive assessment;
Adis International Limited. All rights reserved.
27
the use of combined nonpharmacological and
pharmacological therapy;
age- and organ function-adjusted pharmacol-
ogy.
The goal of pain management is pain reduction
associated with improved function, reflected in
activities of daily living, sleep and socialisation,
and not necessarily complete absence of pain or
the use of minimal analgesic medication.
2.5 Assessment
Pain assessment tools for younger patients
are also likely to work in the elderly, even in the
presence of mild to moderate cognitive impair-
ment.[7,26,27] The patient is the only reliable source
for pain assessment. He or she must describe the
quality, location and nature of the pain and factors
that relieve or worsen pain.[10,28] Unidimensional
scales, and sometimes multidimensional scales
(McGill Pain Questionnaire or Brief Pain Inven-
tory), are useful in patients who are verbal but
cognitively impaired.[28] The McGill Pain Ques-
tionnaire evaluates sensory, affective, evaluative
and miscellaneous pain experiences and provides
additional information to unidimensional scales.
Pain is sometimes best characterised by patient-
generated descriptors. A comprehensive pain his-
tory seeks clues to the nature and aetiology and the
type of pain (nociceptive, neuropathic).[28] This is
confirmed by a good physical examination and
comprehensive neurological examination.
Pain is multidimensional and modulated by
personal meaning and psychological state.[22,28]
Social context, belief and environment shape the
pain experience.[29] Pain severity is modulated
negatively by depression, anxiety, delirium and the
patients psychological make-up (helplessness,
anger, denial, fear, vulnerability and depend-
ence).[25] Intensity, quality, time course, effect on
functional status and personal meaning are unique
to the individual and are not appreciably measured
by standard pain assessment tools.[11] Therefore,
pain scales are limited in their ability to fully assess
pain and should not be relied upon without a com-
plete pain history or physical examination.[28]
Drugs Aging 2003; 20 (1)
28
Quantitative pain scales do provide a means of
sequentially assessing pain during pain manage-
ment. Pain intensity scales cannot be used inter-
changeably with pain relief scales, since there is a
poor relationship between response as measured
by changes in the two scales.[26] Therefore, physi-
cians should use only one unidimensional pain
scale in the frequent evaluation of pain response
and it should be consistently used throughout the
course of treatment.[30] Pain scales that use a Likert
scale may be easier for the elderly to comprehend
than numerical or visual analogue scales. Pain
management is most successful when the under-
lying cause for pain is known, which occasionally
requires radiographs and ancillary studies. Treat-
ment of pain should not wait until the results of the
tests.[10]
There are additional assessments unique to the
elderly: screening for depression, screening for cog-
nitive impairment, assessment of activities of daily
living, assessment of gait and balance, and assess-
ment for visual and auditory impairment. Severely
demented patients may develop a particular pattern
of behaviour (grimacing, rocking, withdrawing,
hyperkinesis, hypokinesis) as a unique pain signa-
ture or behaviour, which may be recognised by
close caregivers.[5] In contrast, physiological res-
ponses to pain, which include increased heart rate
and decreased variability in respiration, are poor
indicators of pain in the nonverbal patient.[29,31]
There are no reliable objective biological or phys-
iological markers for the presence of pain.[10]
Pain diaries, which include the date, severity,
time course during the day, and nonpharmacologi-
cal and pharmacological treatments, are helpful in
the ongoing management of pain and evaluation of
response, although this should not lead to a preoc-
cupation with pain.[3,10]
2.6 Biophysical and Biopsychosocial
Therapies
The experience of pain is multifactorial, involv-
ing not only the senses but also the cognition, af-
fect, motivation and behaviour characteristics of
individuals with chronic pain.[32] Recognition of
Adis International Limited. All rights reserved.
Davis & Srivastava
the cognitive, behavioural and sociocultural di-
mensions of pain means that nonpharmacological
measures that improve these components of pain
will benefit patients with pain.[33] It is very difficult
to separate the consequences of the pain experience
(i.e. depression) from the sensory experience; both
are intertwined in the experience of suffering and
both require treatment.[34]
Depression is more common in patients with
chronic pain than in healthy controls, either predat-
ing or as a consequence of pain. Pain may kindle
depression in predisposed individuals.[35] The
strongest influence on pain, depression and coping
is the meaning patients give to their pain.[32] Con-
trol of depression and anxiety greatly improves
pain, and identifying and developing appropriate
coping skills, adaptive skills and meaning may de-
crease the crippling nature of chronic pain.[36]
Therefore, to manage chronic pain only in the con-
text of a biophysical model rather than a bio-
psychosocial experience will fail to fully manage
cancer-related and nonmalignant pain.[37]
The biopsychosocial model of pain emphasises
the importance of nonpharmacological therapies
in combination with pharmacotherapy tailored to
the individual.[38] Such an approach requires an
interdisciplinary approach to nonmalignant and
cancer pain in the elderly.[38] Cognitive, psycho-
logical and pharmacotherapies are tailored both to
patients and to pain characteristics.[39] Four factors
have been shown to predict the variability of out-
come measures in chronic geriatric pain:[40]
changes in patient activity;
changes in pain and related behaviour;
changes in constant pain; and
changes in attitude to pain centre goals.
Both physical and occupational therapy im-
prove activities of daily living and adaptation to
limitations. Psychotherapy, broadly including re-
laxation techniques, hypnosis, cognitive behaviou-
ral techniques and biofeedback, improves skills for
coping with chronic pain.[41] Spiritual develop-
ment and cognitive behavioural therapy reduce
the suffering of the pain experience and improve
coping while reducing behavioural expressions of
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
pain.[42] Therefore, to successfully manage non-
malignant or cancer pain in the geriatric patient,
a multidisciplinary programme that enhances ac-
tivities of daily living and improves healthy adap-
tive and coping skills while using appropriate an-
algesics has the greatest chance of success.[43-46]
3. Nonpharmacological Management
of Pain in the Elderly
The strategies least preferred by the elderly to
manage pain are most commonly prescribed by
physicians: medications, physical therapy and ex-
ercise.[11] Preferred by the elderly are home reme-
dies, massage, topical agents, physical modalities
such as heat and cold, and informal cognitive strat-
egies (social gatherings, visiting neighbours, mu-
sic, prayer and humour).[3,11]
Physical therapy may provide only time-limited
benefits for chronic pain in some patients, and may
fail to provide additional adaptive coping skills if
not continued at home by the patient. Psychother-
apy is a helpful support in this regard. Geographi-
cal barriers to physical therapy centre on transpor-
tation to and from the medical facilities, with a
limited number of family members available to
provide transport.[11] Exercise is usually avoided
by the elderly, particularly the frail, because of fear
of falling, personal safety (sense of vulnerability
in an unfamiliar environment), and limiting com-
orbidities such as chronic obstructive lung disease
and congestive heart failure.[11] Unconventional
therapies are perceived to be low risk, whereas
conventional therapies including medications are
perceived as high risk, with associated adverse
effects that ultimately may cost patients their in-
dependence.[11]
Transcutaneous electric nerve stimulation
(TENS) is a low-risk procedure, as is ultrasound;
both produce analgesia, although temporarily.
TENS units have been used for spine-related pain,
diabetic neuropathy, phantom limb pain and post-
herpetic neuralgia, and occasionally for arthritis.[47]
Psychological methods include hypnosis, med-
itation, relaxation, guided imagery, biofeedback,
prayer and music therapy.[47] These need to be ap-
Adis International Limited. All rights reserved.
29
plied on the basis of interests, beliefs, skills and
cognition. These behavioural therapies aim to en-
hance healthy lifestyles and bolster coping be-
haviours. Cognitive therapy can be combined with
behavioural approaches in a complementary fash-
ion, facilitating pharmacotherapy.
4. Approaches to Pharmacological
Management of Pain in the Elderly
4.1 Whether To Separate Malignant
from Nonmalignant Pain
Malignant and benign pain are frequently con-
sidered separately. Cancer pain is highly associ-
ated with physical pathology, whereas nonmalig-
nant pain is less so.[48] Cancer pain is seen more as
biophysical pathology than as a biopsychosocial
experience. Patients with cancer experience pain
that overlaps in aetiology (nociceptive, neuro-
pathic) with that seen in patients without the dis-
ease. The anatomical, physiological and biochem-
ical substrates and mechanisms of nociception are
not different between the two groups.[48] Chronic
pain, whether nonmalignant or malignant, is mod-
ulated by depression, anxiety, cognition, coping
skills and meaning. Also, patients with cancer pain
frequently have nonmalignant pain unrelated to the
underlying cancer or as a result of treatment, and
the two cannot be easily separated. These patients
usually have additional symptom burdens and
struggle with the finality of their existence. The
elderly benefit from a biopsychosocial model of
care and multimodality pain management. As the
pharmacology of treatment of both nonmalignant
and cancer pain completely overlaps, there appears
to be no advantage in separating cancer and non-
cancer pain.[48]
4.2 Principles of Good Prescribing
Clinicians must balance the risk and benefits of
drug therapy and should follow the WHO three-
step guideline whether the pain is nonmalignant or
malignant in origin.[49] Two major questions need
to be addressed when channelling a patients
drug course: does the drug bring a certain problem
Drugs Aging 2003; 20 (1)
30
to the patient, and does the patient bring a certain
problem to the drug?[50] Incremental goals for ob-
taining pain relief are pain relief that allows for
restful sleep, pain relief at rest, and tolerable pain
with activity. Improved function may occur to a
greater extent than reduced pain intensity as mea-
sured by unidimensional pain scales.[5]
Pain responses are better assessed by physical
functioning, psychosocial functioning, and im-
proved mood and cognition.[5] A trial period is nec-
essary and continuity of care is a must for assessing
treatment trials. Ineffective drugs are tapered and
discontinued to minimise polypharmacy.
Pain relief with certain analgesics (particularly
morphine) occurs at lower doses with age, presum-
ably because of delayed clearance or elimination.
Final doses necessary for pain relief do not correl-
ate with bodyweight or initial severity of pain. In
the elderly, the duration of pain relief is longer than
predicted by peak analgesic effect because of de-
layed clearance and reduced volume of distribu-
tion.[51,52] Therefore it is better to start low and go
slow with most analgesics.
Chronological age is not the same as physiolog-
ical age, and this accounts in part for the wide inter-
individual variation and unpredictable nature of
adverse effects as a result of divergent individual
patient therapeutic indexes. Interindividual differ-
ences exceed ethnic differences. This is further
complicated by the psychosocial and spiritual con-
text of the experience of pain.[10] The proper use of
adjuvant analgesics reduces existing opioid ad-
verse effects, widens therapeutic indexes and im-
proves pain relief in patients with incomplete or
suboptimal pain relief from opioids.
The principles of good prescribing are based on
the following: known evidence-based analgesic
benefits; a wide therapeutic index; fewest potential
adverse drug effects; fewest potential drug inter-
actions; treatment of several symptoms with one
drug; least cost burden; compatibility with patient
comorbidity and end-organ function; versatility
(multiple routes of administration); and prescrib-
ing limited to one drug per drug class. Factors in-
fluencing adverse drug reactions are age, sex, race,
Adis International Limited. All rights reserved.
Davis & Srivastava
polypharmacy, dosage, duration of treatment,
compliance and underlying end-organ function.[22]
5. Paracetamol (Acetaminophen)
Paracetamol (acetaminophen) is preferred to
NSAIDs in the elderly because of its low gastroin-
testinal and renal toxicity.[3,5] Paracetamol is not a
peripheral anti-inflammatory agent but is centrally
acting, and in this sense is a weaker analgesic than
cyclo-oxygenase (COX) inhibitors. It is well ab-
sorbed and metabolised by glucuronidation and
therefore has few drug interactions.[5] There are no
age-related reductions in paracetamol clearance.
The elderly may prefer an elixir form. Dosages
above 2.6 g/day are unlikely to produce further
analgesic benefit. Hepatic toxicity increases with
dosages above 4 g/day, but can occur at lower
dosages with coexisting liver disease, fasting or
alcohol consumption on a regular basis.[3] Long-
term use of paracetamol can eventually lead to
renal toxicity, particularly if combined with
NSAIDs.[47]
6. NSAIDs
NSAIDs have a major role in the management
of acute and chronic pain syndromes associated
with cancer or nonmalignant disease, particularly
with somatic or visceral nociceptive pain. NSAIDs
reduce joint swelling, tenderness and stiffness and
improve physical function in inflammatory arthri-
tis to a greater extent than in mechanical arthritis
(osteoarthritis). The choice of NSAID depends on do-
sage schedule, response, cost and physician as well
as patient preference and experience. Extended-
release NSAIDs are preferred for chronic pain for
the sake of compliance, although this is offset by
increased renal dysfunction when compared with
short-acting NSAIDs. As-needed administration
is preferred, particularly for intermittent pain.
NSAIDs should be used one at a time and should
not be combined with corticosteroids or paraceta-
mol because of the significant risk for gastrointes-
tinal toxicity, therapeutic redundancy and, in the
case of paracetamol, renal toxicity. Analgesia with
NSAIDs occurs within 4 hours, much sooner than
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain 31

the anti-inflammatory effects, which are not appar- renal failure.[47] Fortunately, renal toxicity is not
ent for several days to weeks.[3] as common as gastrointestinal toxicity. Diuretics,
ACE inhibitors and advanced age are major risk
6.1 Gastrointestinal Toxicity factors for renal toxicity.[3] Renal dysfunction is
usually reversible if recognised early and NSAIDs
Despite their known benefit, NSAIDs are also
are withdrawn. Interstitial nephritis is rare but
associated with adverse effects, the prevalence of
leads to permanent renal failure. Clinical factors
which increases with age. NSAIDs in the elderly
associated with NSAID-induced renal toxicity are
produce a higher risk of peptic ulcer disease, un-
reduced effective circulating volume, diuretic use,
controlled hypertension, renal failure and worsen-
congestive heart failure, cirrhosis, diabetes, infec-
ing congestive heart failure than in younger pa-
tion and pre-existing renal disease.[3]
tients.[10] Patients over the age of 60 years have a
Adverse renal reactions to NSAIDs are related
34% risk of severe gastrointestinal bleeding, and
to age over 75 years, congestive heart failure, de-
if there is a history of gastrointestinal bleeding, the
hydration, history of peptic ulcer disease, history
risk is 9%.[10,27] Elderly women are at particular
of renal disease, hypertension, female sex and con-
risk because of a higher dose per bodyweight.[3]
comitant use of diuretics.[47] Concurrent diuretics
Also, prevalence is increased in those with Heli-
and NSAIDs lead to reduced renal blood flow, re-
cobacter pylori gastritis.[3] The toxicity risk with
duced glomerular filtration rate and increased so-
NSAIDs is dependent on both dose and duration.
dium retention.[3] NSAIDs reduce renal prosta-
An inverse relationship exists between duration of
glandin and renin secretion and may cause a
therapy and gastrointestinal toxicity, supposedly
sudden rise in serum potassium while patients are
because of mucosal adaptation.[3]
on ACE inhibitors or potassium-sparing diuret-
NSAIDs associated with less gastrointestinal
ics.[3] Elderly patients with preserved renal func-
toxicity are etodolac, ibuprofen, nabumetone and
tion usually tolerate both long- and short-acting
meloxicam, although confidence levels of safety
NSAIDs; however, long-acting NSAIDs produce
overlap with those of other NSAIDs.[3] Routine use
a small decrement in renal function if patients have
of indomethacin, ketorolac, piroxicam and mefe-
pre-existing mild renal insufficiency.
namic acid is to be avoided because of increased
risk of toxicity.[53] Approximately 40% of patients Recommendations when prescribing NSAIDs
discontinue NSAIDs because of gastrointestinal are as follows: pretreatment measurement of elec-
intolerance.[47] trolytes and creatinine; repeat electrolytes in 14
weeks; start with low dosages and go slowly with
The use of misoprostol, double-dosage hista-
titration; consider as-needed schedule if pain is
mine H2 antagonists and proton pump inhibitors
only mild or intermittent. Less frequent toxicities
only partially reduces the risk of gastrointestinal
include anxiety, dizziness, tinnitus, drowsiness,
toxicity, potentiates the risk for drug interactions
confusion, bronchospasm in aspirin-sensitive indi-
and adverse effects and does not significantly
viduals, urticaria, pruritus and photophobia.[47]
reduce renal toxicity.[10] Alternatives in high-risk
individuals include non-acetylated salicylates,
COX-2 inhibitors, low-dosage corticosteroids and 6.3 Cyclo-Oxygenase-2 Inhibitors
low-dosage opioids, depending on the clinical sit-
Rofecoxib and celecoxib are the highly selec-
uation.
tive COX-2 inhibitors available in the US (etori-
6.2 Renal Toxicity
coxib is available in Europe). Both rofecoxib and
celecoxib have an equal risk of adverse renal
The renal toxicities of NSAIDs include revers- events when compared with nonselective COX in-
ible renal insufficiency, sodium and water reten- hibitors.[3] Neither significantly alters the pharma-
tion, hyperkalaemia, interstitial nephritis and acute cokinetics of warfarin or adversely alters platelet

Adis International Limited. All rights reserved. Drugs Aging 2003; 20 (1)
32
function, and therefore there is a significantly
lower risk for bleeding complication than nonse-
lective agents.[3] Celecoxib is much more depend-
ent on cytochrome P450 (CYP) for its metabolism
than is rofecoxib.[3]
Rofecoxib may be prescribed for individuals
with aspirin-sensitive asthma and, unlike cele-
coxib, can be used in patients with sulfa drug al-
lergies and is less likely to lead to reactions in pa-
tients with NSAID allergies.[3] Rofecoxib has an
onset of action that is more rapid and more durable
than that of celecoxib.[3] Rofecoxib is administered
once daily, which improves compliance. Therapy
with either COX-2 inhibitor should start at the
lowest dosage and be titrated slowly. Both cancer
pain and nonmalignant pain respond to COX-2 in-
hibitors; however, their use in cancer pain is not
well studied.
7. Corticosteroids
Corticosteroids are useful alternatives to NSAIDs
in certain clinical situations. Corticosteroids are
disease-modifying and analgesic for patients with
polymyalgia rheumatica and reduce pain associ-
ated with rheumatoid arthritis, although they are
not disease-modifying in this condition. Cortico-
steroids reduce peritumoral swelling, oedema and
pain associated with compressive neuropathies and
cerebral metastases and are preferred in these situ-
ations. Relief of pain and nausea occur in malig-
nant bowel obstruction, pain from soft tissue tu-
mour infiltration and cancer-associated visceral
pain.[21]
Corticosteroids can replace NSAIDs for malig-
nant bone pain in individuals who are at high risk
for NSAID-induced gastrointestinal toxicity, who
have renal dysfunction or who have several addi-
tional symptoms that may respond to corticoste-
roids (anorexia, nausea). The nonspecific benefits
of corticosteroids are improved appetite and mood
and a feeling of well-being which, although it tends
to be of short duration (measured in weeks), may
be meaningful to patients.
Toxicity appearing shortly after starting corti-
costeroids includes dysphoria, glucose intolerance,
Adis International Limited. All rights reserved.
Davis & Srivastava
hypertension, infection and psychosis.[5,21] Long-
term adverse effects include accelerated cataracts,
osteoporosis, aseptic necrosis of joints, and thin-
ning of the skin with associated risk of skin tears
and bruising. Therefore, dosage should be reduced
to the minimal effective level after maximum ben-
efit has been achieved. Various administration
schemes can be used: dexamethasone 48mg two
to three times daily; methylprednisolone 1632mg
two to three times daily; or prednisone 2040mg
two to three times daily.[21] Epidural cord compres-
sion is treated with doses as high as 100mg of dex-
amethasone, resulting in analgesia within 24 hours.
Corticosteroids can be given orally, intravenously,
subcutaneously, rectally, intra-articularly, topi-
cally and spinally.
Relative contraindications to corticosteroids in-
clude HIV infection, recent intestinal anastomosis,
uncontrolled congestive heart failure, severe renal
impairment, recent varicella infection, active or
latent oesophageal, gastric or duodenal ulcer, dia-
betes, systemic fungal infections, ocular herpes
simplex, myasthenia gravis, active or latent tuber-
culosis, and concurrent use of NSAIDs.[21]
7.1 Prednisone and Prednisolone
Prednisone has nonlinear pharmacokinetics and
prednisolone has linear pharmacokinetics, but
these differences are not used as a guide to deter-
mining administration strategies.[54,55] Pharma-
cokinetics and pharmacodynamics differ for both
agents depending on race and sex, but again these
differences are not clinically relevant.[56] In adults,
regardless of age, response to prednisone and
methylprednisolone is not influenced by the kine-
tics.[57] There is hepatic interconversion between
prednisone and prednisolone. The oral bioavaila-
bility of prednisolone is 62%, which is greater than
that of prednisone.[58]
Combined use with proton pump inhibitors
does not influence prednisone or prednisolone
concentrations.[59] Itraconazole inhibits methyl-
prednisolone metabolism by blocking CYP3A4
metabolism, thus delaying elimination.[60] Methyl-
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
prednisolone is more expensive in the US, and as
a result prednisone is preferred.
Patients with hepatic or renal failure or renal
transplant, those older than 65 years, women tak-
ing estrogen-containing oral contraception, and
patients taking ketoconazole have increased un-
bound prednisone. Increased unbound concentra-
tions results in a greater volume of distribution and
delayed elimination. Patients with hyperthyroid-
ism or Crohns disease, individuals taking medica-
tions that induce the CYP system and patients tak-
ing enteric-coated prednisolone have decreased
plasma concentrations of prednisolone.[61]
7.2 Dexamethasone
The oral bioavailability of dexamethasone is
6165%,[62-64] with wide interindividual variation
in bioavailability unrelated to age, sex or body-
weight. Obesity increases lag time to absorption
and positively correlates with the area under the
concentration-time curve.[65] There is diurnal cy-
clic variation in oral absorption, which is not dra-
matic and not clinically relevant.[65] Dexametha-
sone is 75% bound to albumin. Protein binding is
reduced by 24% in uraemic patients, which in-
creases bioavailability. Renal failure increases
dexamethasone clearance in part through reduced
circulating unbound drug.[66,67] The terminal half-
life of dexamethasone is 4 hours.[68,69]
Dexamethasone is principally metabolised by
CYP3A4 to an inactive metabolite. Liver dysfunc-
tion reduces clearance by 53%.[70] Phenytoin, car-
bamazepine and phenobarbital (phenobarbitone)
induce CYP3A4 activity and increase dexametha-
sone clearance. Dexamethasone in turn induces
CYP3A4 activity and may reduce drug concentra-
tions of antiepileptic drugs. Alcohol may increase
or decrease dexamethasone clearance. For un-
known reasons, dexamethasone clearance is de-
layed by depression and improves with resolution
of depression.[66,71,72]
Adis International Limited. All rights reserved.
33
8. Agents Preventing Bone Resorption
8.1 Calcitonin
Calcitonin relieves pain due to excessive bone
resorption. Pain from bone metastases, noncancer-
associated osteoporosis and Pagets disease may
respond. Phantom limb pain is relieved by calcito-
nin, independently of the effect of the drug on bone
turnover.[21] Calcitonin is reported to ameliorate
migraine headaches.[21] Adverse effects include
local cutaneous reactions, gastrointestinal upset
and allergic reactions. Nausea occurs in 10% of
patients and cutaneous flushing, particularly of
hands and feet, in 2.5%. Nasal sprays produce
rhinitis in 12%, general achiness in 10% and epi-
staxis in 3.5%. Dosage is 8 IU/kg subcutaneously
or intravenously every 612 hours; nasal spray
dosage is 200 IU/day.[21]
8.2 Bisphosphonates
The bisphosphonates, of which three genera-
tions are available, are analogues of pyrophos-
phate. Structurally they have a P-C-P backbone,
with potency related to the side chain attached to
the carbon located between the two phosphates.
Bisphosphonates bind to hydroxyapatite crystals,
preventing resorption directly and inhibiting os-
teoclastic activity. Intermittent use prevents osteo-
porotic complications. Pamidronic acid and now
zolendronic acid are indicated for cancer-related
hypercalcaemia. Bisphosphonates have antitumour
activity in some cancers.[21] Pamidronic acid re-
lieves pain associated with bone metastases and
reduces opioid requirements by 3050% of base-
line. Pamidronic acid is approved for long-term
monthly administration in myeloma and breast
cancer for reducing or delaying orthopaedic com-
plications. Other solid tumours may benefit, al-
though data are meagre.
Adverse effects include post-treatment arth-
ralgias and fevers in a minority of patients; both
resolve after several days. Rare complications in-
clude hypocalcaemia, hypokalaemia, hypophos-
phataemia and hypomagnesaemia. The usual dos-
age of pamidronic acid is 90mg intravenously
Drugs Aging 2003; 20 (1)
34
every 34 weeks; the dosage of zolendronic acid is
4mg intravenously every 4 weeks.[21]
9. Opioids
Opioids are the drugs of choice for moderate to
severe pain unrelieved by NSAIDs (steps 2 and 3
according to WHO guidelines). These recommen-
dations are based on severity of pain, not type of
pain or whether the pain is related to cancer or is
nonmalignant in origin. The potency of the various
opioids parallels opioid receptor affinity and clin-
ical efficacy (the number of receptors needing to
be occupied before pain relief). For morphine, un-
like fentanyl, analgesia does not correlate with
plasma concentrations.[73] The weak opioids co-
deine and tramadol have ceiling effects, resulting
in limited analgesia and dose-related adverse ef-
fects despite dose escalation. Both tramadol and co-
deine are converted to -agonists by CYP2D6.[74]
Patients receiving medications that block CYP2D6
or who are CYP2D6 poor metabolisers fail to ex-
perience analgesia with either of these opioids.[75]
Step 2 analgesics also consist of combinations
of opioids plus NSAIDs in a single formulation,
which have a ceiling dose because of the NSAID.
These are usually a convenient step 2 medication
and improve compliance. Alternatives to the opioid-
NSAID combinations are separately administered
potent opioids (morphine or oxycodone) and
NSAIDs, tailoring the regimen to the patients in-
dividual needs. Pentazocine produces neuropsy-
chiatric toxicity and is to be avoided in the elderly,
as is pethidine (meperidine) because of the seizure
risk from accumulation of its metabolite norpeth-
idine. The mixed agonist-antagonists nalbuphine,
Table I. Opioid routes of administration
Oral Rectal Sublingual
Codeine Codeine Fentanyl
Hydrocodone Methadone Methadone
Hydromorphone Morphine Morphine
Methadone Oxycodone
Morphine Tramadol
Oxycodone
Tramadol
Adis International Limited. All rights reserved.
Davis & Srivastava
butorphanol and buprenorphine have a limited
role in the management of pain, since they cause
withdrawal reactions when given to patients tak-
ing pure opioid agonists and also have a ceiling
dose.[47] Propoxyphene should also be avoided in
the elderly because of the risk of neural and cardiac
toxicity.
Scheduled doses of opioids are tailored to the pain
pattern of the patient, and formulaic approaches are
to be avoided. There is no standard schedule that
fits all patients, particularly with incident pain.
However, there are general guidelines for admin-
istration.[75,76] Frail elderly patients need extended
administration intervals or dosage reductions be-
cause of delayed opioid (morphine) clearance.[3]
The use of extended-release opioids increases
compliance, and these drugs are used once stable
doses of normal-release opioids are established.
Some physicians start with low doses of extended-
release opioids and provide breakthrough normal-
release opioids for interim pain. End-of-dose fail-
ure is treated by increases in the regular dose with
maintained intervals, or by shortening the intervals
and maintaining the doses. Table I shows the routes
of administration for the opioids used in this pop-
ulation.
Patients require different opioid regimens for
continuous baseline pain and incident pain.[76] The
dosage needed for breakthrough incident pain
may be higher than predicted based on baseline
maintenance dosage. Alternatives for manage-
ment of breakthrough or incident pain are patient-
controlled analgesia, sublingual fentanyl or spi-
nal opioids with or without bupivacaine or
clonidine.[76] A standard initial breakthrough op-
Subcutaneous Spinal: intrathecal/epidural Intravenous
Fentanyl Fentanyl Fentanyl
Hydromorphone Hydromorphone Hydromorphone
Methadone Morphine Methadone
Morphine Morphine
Oxycodone (Europe)
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
ioid dose is usually the same as the 4-hourly dose
when patients are on a 4-hour normal-release op-
ioid schedule, and one-third of a 12-hourly dose
given as needed at any time during a 4-hour inter-
val when on a 12-hourly extended-release op-
ioid.[76,77] The baseline dose of opioid is not ad-
justed if baseline pain is under control; if incident
pain is still a problem, only the dose for incident
pain is increased. The incident-type breakthrough
pain is then treated independently if it is signifi-
cantly unrelieved with usual doses.
Normal baseline dosage titrations for chronic
pain are 50100% of the baseline dosage, depend-
ing on the severity of pain. Baseline dosage in-
creases of less than 25% are usually ineffective;
dosage increments should be based on baseline
dosage.[3] Some patients require asymmetrical
long-term opioid administration because of worse
pain at night. In some instances, opioid dosage re-
quirements are higher with initial therapy than
when maintaining pain relief.[3]
Morphine is the opioid of choice whether pain
is nonmalignant or cancer pain. The starting oral
dosage in opioid-naive frail elderly patients is
5mg every 46 hours, with the same dose every
2 hours as needed. In the healthy (opioid-naive)
elderly without comorbidities, 10mg every 4 hours
is a reasonable starting dosage.
9.1 Adverse Effects
Opioid adverse effects are independent of type
of pain (nociceptive or neuropathic) or whether the
opioid is used for nonmalignant or cancer pain.
9.1.1 Constipation
To prevent opioid constipation, prophylactic
laxatives and stool softeners are started at the ini-
tiation of opioid administration. Bulk laxatives are
ineffective and should be avoided. It is important
to take adequate fluids to allow laxatives and stool
softeners to work. Softeners such as docusate so-
dium 100mg two to three times daily are combined
with a stimulant laxative or osmotic laxative (bi-
sacodyl, senna or sorbitol).[10] Dosage and sched-
ule must be individualised and are not linearly
related to opioid titration. The elderly are particu-
Adis International Limited. All rights reserved.
35
larly at risk for constipation because of ancillary
medications (calcium channel antagonists, antichol-
inergics), a sedentary existence and age-related re-
duced bowel motility. Patients with ileostomies do
not require laxatives, but those with colostomies
should receive them.
9.1.2 Sedation
Mild sedation and transient cognitive impair-
ment are common with initiation of opioids in the
elderly; they usually do not require dosage reduc-
tion, but rather reassurance that they usually re-
solve after 34 days.[5] Poor pain control leads to
cognitive failure independent of opioid use in older
individuals, and impaired cognition is not a con-
traindication to starting opioids.[75,78] If confusion
develops with opioid treatment and is accompa-
nied by hallucinations or myoclonus, opioid reduc-
tion, opioid rotation, altered route of administra-
tion, or discontinuation of adjuvants such as
tricyclic antidepressants are necessary, depending
on the clinical situation and pain control. If death
is imminent, opioids are usually not switched; hal-
operidol is used for hallucinations and benzodia-
zepines for myoclonus.
Driving is permissible when patients are taking
long-term stable dosages of opioid and cognition
returns to baseline.[10] Both patient and family
members should feel comfortable about operating
a motor vehicle. A test drive on an isolated country
road with a relative may be prudent before resum-
ing driving.
9.1.3 Nausea
Nausea from opioids arises from three potential
mechanisms: stimulation of the chemoreceptor
trigger zone located in the postrema of the fourth
ventricle; vertigo from cochlear stimulation; and
gastroparesis.[3] Nausea is not usually treated pro-
phylactically, as it occurs in a minority of patients.
Haloperidol is provided to those who are nause-
ated, at a dosage of 0.51mg every 4 hours as
needed. If nausea is associated with vertigo, pro-
chlorperazine 10mg orally or 25mg rectally every
46 hours or transdermal scopolamine are reason-
able choices. If early satiety is the cause, metoclo-
pramide 10mg should be given every 4 hours. An-
Drugs Aging 2003; 20 (1)
36
tiemetics have a significant risk of inducing extra-
pyramidal reactions and anticholinergic adverse
effects. Nausea usually abates spontaneously with-
in several days. Persistent nausea requires dosage
reduction, an alternative route of administration or
opioid switch.[5,10]
9.1.4 Myoclonus and Delirium
Myoclonus is a common, usually mild, though
not well recognised adverse reaction to opioids. It
is frequently missed by physicians because of lack
of understanding or recognition or because of its
mild nature in some patients. Myoclonus is not a
harbinger of seizure; it originates from the spinal
cord. Elimination of an offending antipsychotic
agent (phenothiazines or tricyclic antidepressants)
or opioid dosage reduction if pain is under control
is a logical first step.[75] If this fails to resolve my-
oclonus and the patient remains in pain, an opioid
switch may be necessary. Clonazepam, valproic
acid or gabapentin are used to treat myoclonus
when patients cannot be switched or death is im-
minent.[10]
There are several ways of managing confusion
and delirium associated with opioids. Dosage can
be reduced if pain is under control. Benzodiaze-
pines, ciprofloxacin and NSAIDs should be discon-
tinued. An antipsychotic agent such as haloperidol
is used if the patient is agitated. Opioid-sparing
adjuvant analgesics allow for opioid reduction and
may resolve confusion or delirium. The choice of
adjuvant depends on the type of pain. Opioid
switch rather than opioid reduction is a reasonable
option if pain is not well controlled.[79] In our ex-
perience, an opioid switch for delirium or confu-
sion was necessary in 15% of patients, all of whom
benefited from the switch. Changing the route of
administration from oral to subcutaneous, intrave-
nous, epidural or intrathecal reduces opioid dosage
and allows for continued analgesia with resolution
of neuropsychiatric symptoms. Opioid antagonists
such as naloxone do not reverse confusion, delir-
ium or myoclonus and are not used. Delirium that
predates opioid therapy is a contributing factor to
crescendo pain in cancer patients and needs to be
Adis International Limited. All rights reserved.
Davis & Srivastava
treated in addition to pain in order to improve pain
control.[24]
9.1.5 Respiratory Depression
Respiratory depression is rare in opioid-naive
individuals if low starting dosages are used and
titrated properly. It occurs under various condi-
tions:[3]
rapid dosage escalation of opioid over a short
period;
in patients who are placed on adjuvant analge-
sics while opioids are titrated;
after a successful neurolytic block performed
without opioid dosage reduction;
with rapid relief of pain by a single fraction of
radiation or rapid cord compression while on a
stable opioid dosage;
with the onset of acute renal failure and opioid
accumulation.
Patients who are sedated from opioids have mio-
sis.[3] Low doses of naloxone are used for suspec-
ted opioid-induced respiratory depression. Because
of the short half-life (30 minutes) of naloxone, con-
tinuous infusion of naloxone is necessary if the
patient is taking sustained-release opioids or meth-
adone.[76]
9.2 Other Considerations
9.2.1 Opioid Overlap with Conversion to
Sustained-Release or Transdermal Fentanyl
Switching from normal-release to sustained-
release morphine or oxycodone or from parenteral
to sustained-release morphine requires a 2-hour
overlap to maintain analgesia. Switching from
normal-release opioids to a transdermal fentanyl
patch requires a 1218-hour overlap, and small
doses of the normal-release opioid may be neces-
sary to avoid withdrawal reactions as well as to
maintain analgesia.[3] When switching from con-
tinuous subcutaneous or intravenous fentanyl to a
transdermal patch of equivalent dose, a graded
dose reduction of the infusion over 1218 hours is
necessary.
9.2.2 Opioid Switch
The choice of a second opioid in an opioid
switch depends on the severity and type of pain,
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
the performance status of the patient, the nature of
associated symptoms, regional alternative opioid
availability, versatile routes of administration, and
end-organ (i.e. renal and hepatic) function. A 25
50% reduction in equianalgesic dose should be
used when switching from one opioid to another
except in the case of methadone, where 10% of the
equianalgesic dose is recommended because of its
long half-life and intrinsic efficacy.[80]
9.2.3 Ideal Opioid for Acute Pain Management
The characteristics and administration strate-
gies of an ideal opioid for use in acute pain man-
agement may differ from those for chronic pain
management, and are independent of age. Impor-
tant characteristics include:[73]
rapid onset of action;
no pain at the injection site;
short duration of action;
inactive metabolites;
minimum risk of respiratory suppression;
wide therapeutic index;
limited drug interactions;
availability of a specific antagonist.
Small doses frequently administered without a
baseline dose better reflect the pattern of acute
pain with expected pain resolution within several
days to a week.
9.2.4 Opioids and Tolerance
Specific tolerance to analgesia is rare and is
not age-related. Tolerance can be adaptive (condi-
tioned) or pharmacological.[18] If pain increases in
a patient with cancer who is receiving stable doses
of analgesics, the most likely cause is advancing
cancer. Proposed mechanisms of tolerance include
G-protein modulation, subcellular alterations in
kinases, sodium channel adaptation to opioid-
induced neuronal hyperpolarisation, NMDA re-
ceptor activation, and release of dynorphin.[76-79]
Chronic nonmalignant pain may be associated
with tolerance more frequently than cancer pain,
but this is not well studied.
Adis International Limited. All rights reserved.
37
9.3 Opioid Classes
9.3.1 Opioid Receptor Pharmacology
There are three classes of opioids: lipophilic
agents, the codeine family and morphine-related
agents. Lipophilic agents, metabolised predomi-
nantly by CYP3A4, include methadone and fen-
tanyl; methadone is also metabolised by CYP1A2
and CYP2D6. The codeine family includes hydro-
codone, oxycodone, codeine and tramadol. These
agents are hydrophilic and metabolised by CYP-
2D6. Both codeine and tramadol require conver-
sion of the parent drug to the active metabolite
(morphine and desmethyl-tramadol), whereas the
parent drugs oxycodone and hydrocodone are the
principal analgesics. The morphine-related family
includes morphine and hydromorphone, both of
which are glucuronidated and hydrophilic.
The main site of action of opioids is on
(MOR), (KOR) or (DOR) opioid receptors in
the CNS. Access to the CNS is limited by protein
binding, drug ionisation and hydrophilicity.[73]
Methadone is also analgesic through NMDA and
monoamine receptors.[80]
9.3.2 Tramadol
Tramadol is one of the weak opioids, acting
on MOR through the O-desmethyl-tramadol meta-
bolite and as a monoamine reuptake inhibitor
through the unmodified parent drug.[3] Potency
compared with morphine is 10 : 1.[81]
Equivalent doses of morphine have more ad-
verse effects than tramadol, particularly constipa-
tion, neuropsychiatric symptoms and pruritus.[81]
Laxatives, antiemetics and antipsychotics are re-
quired less frequently with tramadol than with
morphine, which is an important consideration in
the elderly. Dosages can be increased to a maxi-
mum of 400 mg/day. Dosage reduction is neces-
sary in renal or hepatic failure; reductions are rou-
tine for creatinine clearance less than 30 ml/min
and/or age greater than 75 years. [5]
Adverse effects include constipation in 35%,
dizziness in 5%, nausea in 5%, sedation in 3%,
vomiting in 1% and sweating and headache in a
few.[3] Analgesia is not experienced in CYP2D6
poor metabolisers or with medications that block
Drugs Aging 2003; 20 (1)
38
CYP2D6. Conversion from tramadol to equi-
analgesic doses of morphine in a poor metaboliser
can lead to respiratory depression. There is little
advantage of this agent over codeine except that it
may be less constipating. Tramadol is favoured in
countries where access to potent opioids is se-
verely limited.
9.3.3 Pethidine (Meperidine)
Pethidine should simply be avoided in elderly
patients, and perhaps in all patients. Norpethidine
accumulates both with long-term use and in those
with reduced renal function, and can lead to sei-
zures. Norpethidine concentrations are also in-
creased by classical antiepileptic drugs through
stimulation of CYP3A4.[3] Pethidine increases the
risk of psychotomimetic adverse effects, falls and
sedation in elderly patients when compared with
other opioids.[3] Combinations of hydroxyzine and
pethidine increase the risks of orthostatic symp-
toms and sedation compared with each drug alone,
and this further predisposes frail elderly patients
to falls and reduced cognitive function.
9.3.4 Morphine
Morphine is well absorbed by mouth but highly
extracted (70%) by the liver, and as a result over-
all oral bioavailability is 2030% but with wide
interindividual variation.[76] Hepatic extraction of
morphine is predominantly dependent on hepatic
blood flow rather than glucuronidation rate.[82]
Serum protein binding is 30%, so reduced protein
binding plays little part in drug kinetics. Morphine
is glucuronidated by uridinediphosphoglucur-
onosyltransferase (UGT) 2B7 to an inactive me-
tabolite, morphine-3-glucuronide, which may be
neurotoxic, and to an analgesic metabolite, mor-
phine-6-glucuronide, which is 1060 times more
potent than the parent drug but much more hy-
drophilic.[76]
Both metabolites undergo enterohepatic circu-
lation and are subsequently renally excreted. Some
30% of morphine is conjugated at extrahepatic
sites (brain, intestinal tract and kidney). The parent
drug and particularly morphine-6-glucuronide are
quite hydrophilic and as a result have delayed hys-
Adis International Limited. All rights reserved.
Davis & Srivastava
teresis (peak plasma concentration to analgesia)
compared with lipophilic opioids.
Morphine glucuronidation is not altered with
age or dramatically influenced by hepatic fail-
ure.[76] However, clinically, hepatic function and
age better predict adverse effects than renal func-
tion.[83] The age-related decrease in clearance of
morphine and especially morphine metabolites
means that lower doses of morphine are usually
adequate in the elderly compared with younger
patients. Morphine is one of the safest opioids to
use in hepatic failure, but its clearance (parent drug
and metabolites) is quite sensitive to renal func-
tion.
Morphine given rectally has a 1 : 1 equi-
analgesic ratio to oral morphine, even though its
pharmacokinetics are slightly different. Partial he-
patic bypass results from the inferior and middle
haemorrhoidal vein drainage into systematic (ca-
val) veins and bypassing the portal vein. As a re-
sult, morphine concentrations are increased but
morphine-6-glucuronide concentrations are re-
duced.
In our experience, the optimal ratio of intrave-
nous or subcutaneous dose to oral dose of mor-
phine is 1 : 3. We have found that subcutaneous-
to-oral dose ratios of 1 : 2 are inadequate in a
significant number of patients
Topical morphine mixed in Intrasite1 gel can
be applied to painful ulcerative malignant and
nonmalignant skin wounds to relieve pain without
significant systemic exposure and with a probable
reduction in adverse events.[84-86] This may be par-
ticularly appealing to the elderly, who fear mor-
phine adverse effects. The mechanism of analgesia
is through binding to peripheral opioid receptors
on sensory afferent neurons.
Convenience is an important consideration to
the elderly. Controlled-release morphine allows
for 8-, 12- or 24-hour administration intervals,
which significantly improves compliance com-
pared with 4-hourly administration. Improved out-
comes from increased compliance may offset the
1 Use of tradenames is for product identification only and
does not imply endorsement.
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
cost of sustained-release morphine. If a 4-hourly
dosage schedule is preferred, a double dose can be
safely given at bedtime, which allows the patient
to sleep through the night undisturbed, although
sustained opioid release at night would be pre-
ferred. Routes of administration are oral, rectal,
subcutaneous, intravenous, epidural and intrathe-
cal as well as topical.
9.3.5 Methadone
Commercially available methadone consists of
two isomers: the L-isomer, which is analgesic, and
the D-isomer, which is a monoamine reuptake in-
hibitor. Both isomers are NMDA receptor antago-
nists.[87,88] The L-isomer is responsible for most of
the adverse effects of methadone.
Oral bioavailability is 80%. Methadone is 90%
protein bound, mostly to 1-acid glycoprotein,
which is increased in inflammatory states, there-
by delaying clearance.[88] It is metabolised by
CYP3A4, CYP1A2 and CYP2D6. Methadone is
quite lipophilic and readily crosses the blood-brain
barrier and oral mucosa. Plasma clearance is de-
pendent on the patients phenotype for expression
of CYP isoenzymes and on concomitant drug ther-
apy. Drug interactions occur with greater fre-
quency with methadone than with morphine.
Wide interindividual variation in pharmacoki-
netics results in a half-life that can range from
8.558 hours.[88] Methadone induces its own me-
tabolism with long-term administration, which
shortens its half-life. Pharmacokinetics are not
significantly altered by renal failure, unlike those
of most other opioids.[88]
Nonopioid analgesic receptor activities of me-
thadone include NMDA receptor binding, mono-
amine reuptake inhibition and DOR binding,
which further complicate equianalgesic adminis-
tration. A linear equianalgesic dose ratio rather
than a single dose ratio with morphine complicates
opioid rotation. Equianalgesic conversions pro-
portionally increase the dose ratios of morphine to
methadone when switching from progressively
higher doses of morphine. There are several dosage
strategies for methadone, including: (i) a sliding-
scale equianalgesic dosage schedule given every 8
Adis International Limited. All rights reserved.
39
hours based on total daily oral morphine equi-
analgesic dose; or (ii) an every-3-hours as-needed
dose based on 10% of the daily morphine equiva-
lent up to 30mg per dose.[80,88]
Drug accumulation occurs after several days of
as-needed administration and dosage intervals tend
to lengthen at day 46 from the start. A twice- or
three-times-daily schedule may be possible after a
week. The total daily dosage is added on days 5
and 6 and divided by 4, and this amount is given
every 12 hours beginning on day 7 when using an
as-needed dosage schedule. Some patients may re-
quire more frequent dosing intervals, every 8 or 6
hours. Doses should be adjusted accordingly.
Dose ratios do not differ between neuropathic
and non-neuropathic pain.[89] All previous opioids
are usually stopped before starting methadone. The
oral-to-rectal bioequivalent dose ratio is 1 : 1
and the intravenous- or subcutaneous-to-oral dose
ratio is 1 : 2.
Methadone is associated with a reduced inci-
dence of visual hallucinations, myoclonus and con-
stipation compared with morphine. Caution is ne-
cessary when using methadone in elderly patients
because the unpredictably long half-life may be
even longer in the elderly. Methadone should be
used with caution in individuals with somatised
existential pain or who have a low tolerance to
other opioids. Contraindications to methadone in-
clude acute uncontrolled asthma and severe
chronic obstructive lung disease with carbon dio-
xide retention. Methadone can be used in the rare
patient taking monoamine oxidase inhibitors
(MAOIs).
Methadone is given orally, rectally, sublingu-
ally, subcutaneously and intravenously. The cost
of methadone is one-tenth that of other sustained-
release opioids and it is quite cost effective, partic-
ularly for elderly patients with limited income.[88]
Methadone is a good second-line analgesic but re-
quires some experience for comfortable prescrib-
ing. It is an ideal analgesic for individuals with a
history of substance abuse.
Drugs Aging 2003; 20 (1)
40
9.3.6 Fentanyl
Fentanyl has a rapid onset of action due to its
lipophilic properties, and a short duration of action
after initial administration as a result of redistri-
bution from spinal cord to peripheral fat, muscle,
lung and gastrointestinal tract, all of which act as
storage sites. With continuous long-term adminis-
tration, storage sites become saturated and the
drug accumulates.[82,89] Drug elimination is by
CYP3A4, resulting in norfentanyl, which to all in-
tents and purposes is a nonactive metabolite. Elim-
ination by CYP3A4 is rate limiting. Since 85% of
fentanyl is protein bound, pharmacokinetics are
altered in low protein states.[89]
Analgesia is influenced by the plasma concen-
trations of fentanyl, the type of pain, the age of the
patient, coadministered medications and interindi-
vidual pharmacokinetics.[89] Fentanyl has a short
hysteresis, with a mean onset of analgesia 6 min-
utes after peak plasma concentrations, in contrast
to 15 minutes for parenteral morphine. Unlike mor-
phine, fentanyl does not cause histamine release
and is less likely to produce pruritus.[73] Also un-
like morphine, plasma concentrations correlate
closely with analgesia.[73] Delayed ventilatory de-
pression associated with long-term fentanyl treat-
ment is due to reabsorption and redistribution of
sequestered drug from the gastrointestinal tract,
muscle, fat and lung.[73,89] Hypotension and respi-
ratory depression are potentiated when fentanyl is
combined with benzodiazepines.[73]
Fentanyl can be given sublingually, intrana-
sally, subcutaneously, intravenously, transderm-
ally, epidurally and intrathecally. Parenteral and
epidural fentanyl are popular in the management
of acute postoperative pain because of quick on-
set of action and short half-life. Indications for
transdermal fentanyl or sublingual fentanyl for
breakthrough are when the oral route is no longer
feasible (transdermal), compliance is a problem
(transdermal), there is a need for rapid relief of
breakthrough pain and incident pain (sublingual)
or prior to dressing change (sublingual).[90-92]
Adis International Limited. All rights reserved.
Davis & Srivastava
Transdermal Fentanyl
The rate of absorption of the transdermal fen-
tanyl matrix patch is influenced by body tempera-
ture, body fat and age. Absorption is delayed in the
elderly. Subcutaneous fat acts as a secondary res-
ervoir which dampens drug absorption and contri-
butes to ongoing fentanyl release after removal of
the patch. Drug can be released for as long as 24
hours after removal of the patch. The transdermal
reservoir is designed to deliver drug for 72 hours,
which may be extended in frail elderly patients.
Transdermal fentanyl should not be started in
opioid-naive patients at dosages greater than 25
g/h and is not indicated for acute pain manage-
ment.[3] Patches should not be placed over dam-
aged skin. High body temperatures increase ab-
sorption, which increases the risk of opioid
toxicity.
The fentanyl transdermal patch is an inflexible
system and is costly for the elderly. However,
transdermal fentanyl is less constipating than mor-
phine and produces less daytime drowsiness and
less disruption to daily life. Transdermal delivery
assures compliance.[90]
Converting from morphine to fentanyl is a two-
step process: step 1 calculates the dose in oral mor-
phine equivalents at 4-hour intervals, and step 2
selects the appropriate fentanyl transdermal patch.
If the oral morphine equivalent dose is 520mg
4-hourly, a 25 g/h patch is selected; if 2535mg
4-hourly, a 50 g/h patch; if 4050mg 4-hourly, a
75 g/h patch; and if 5565mg 4-hourly, a 100
g/h patch.[93] Intravenous morphine at approxi-
mately 4 mg/h is equivalent to a 100 g/h fentanyl
patch.[80]
Sublingual Fentanyl
Sublingual (transmucosal) fentanyl is available
in a lozenge form designed for breakthrough pain.
In a double-blind trial comparing transmucosal
fentanyl with immediate-release morphine for
cancer-associated breakthrough pain, fentanyl was
more effective.[91] Sublingual administration par-
tially bypasses hepatic and intestinal CYP3A4
metabolism, which results in a peak fentanyl con-
centration twice that of oral fentanyl.[69] Sublin-
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
gual bioavailability is 4652%. Much of the sub-
lingual drug is swallowed and metabolised by in-
testinal and hepatic CYP3A4, which accounts in
part for the variable degree of analgesia with the
same dose.[94] The analgesic response is also influ-
enced by saliva pH and individual level of hepatic
and intestinal CYP3A4 activity.[94] As a result,
there is no association between sublingual fentanyl
dose, morphine breakthrough dose or long-term
opioid requirements, and the sublingual fentanyl
dose should be titrated independently of the
around-the-clock opioid regimen.[95] The mean
time to onset of analgesia is 4 minutes, and the
duration of analgesia is dose dependent: 159 min-
utes for the 200g dose and 220 minutes for the
800g dose.[92]
However, sublingual fentanyl is expensive and
difficult for the elderly to afford. Also, the xerosto-
mia that accompanies aging makes it difficult for
some to dissolve the lozenge. Alternatively, the
parenteral solution of fentanyl may be given sub-
lingually for breakthrough pain or prior to painful
dressing changes. Doses of parenteral fentanyl
given sublingually range from 2.515g, and this
form is much less expensive than the commercial
lozenge (table II).[92] A 10g sublingual dose of
the parenteral form costs $US0.22, and relieves
breakthrough pain cost effectively.[92]
9.3.7 Oxycodone
Oxycodone is a MOR and KOR agonist with an
elimination half-life longer than that of morphine
(3.7 vs 1.9 hours), and as a result the duration of
analgesia tends to be longer than that of mor-
phine.[96] The oral bioavailability of oxycodone is
5060%, with less interindividual variation than
morphine.[97,98] Oxycodone is metabolised through
CYP2D6, but unlike codeine the parent drug is the
active analgesic; drugs that block CYP2D6 do not
influence analgesia but prolong its duration by
delaying oxycodone elimination.[74,99] Oxymor-
phone, a minor metabolite, is 14 times more potent
than oxycodone, but accounts for only 10% of oxy-
codone metabolites and contributes little to anal-
gesia. There is a claim of fewer adverse reactions
with sustained-release oxycodone compared with
Adis International Limited. All rights reserved.
41
Table II. Fentanyl cost analysis: cost per dose in US dollars
($US-2002)
Dose (g) Cost ($US)
Oral transmucosal lozenge
200 9.28
400 11.76
600 14.22
800 15.90
1200 17.89
1600 22.12
Fentanyl injection
100 from 2mg vial 2.16
equivalent doses of normal-release oxycodone, al-
though this is controversial.[100]
The analgesic potency of intrathecal oxycodone
is markedly less than that of intrathecal mor-
phine, which may be related to the distribution of
KORs.[97,101] Oxycodone is as poorly absorbed
sublingually as morphine, as its lipid solubility is
equivalent to that of morphine.[97,101] The oxyco-
done-to-tramadol ratio is 1 : 8 and the morphine-
to-oxycodone ratio ranges between 1 : 1 and
1.5 : 1 because of the high variability in absorp-
tion of morphine and unequal cross-tolerance be-
tween morphine and oxycodone.[101-104] Oxyco-
done produces greater analgesic cross-tolerance
when switching to morphine from oxycodone than
from morphine to oxycodone.[105]
Women eliminate oxycodone more slowly (by
25%) than men.[106] Drug elimination is affected
more by sex than by age, which makes oxycodone
an attractive opioid in the elderly.[106] The coef-
ficient of variation in oral bioavailability for
oxycodone as measured by the area under the
concentration-time curve is 33% less than that of
equivalent morphine doses.[107] Oxycodone has
the same protein binding characteristics and vol-
ume of distribution as morphine and, like mor-
phine, its kinetics are not significantly altered in
low-protein states. Unlike methadone, protein
binding of oxycodone is not influenced by acute-
phase reactants such as 1-acid glycoprotein.[97]
Switching from morphine to oxycodone can im-
prove morphine-associated neurotoxicity.[108]
Drugs Aging 2003; 20 (1)
42
Oxycodone is as effective as morphine in the
relief of cancer-associated pain and causes fewer
drug-related hallucinations than morphine.[79] It
has been used for chronic nonmalignant pain asso-
ciated with osteoarthritis. Adverse effects are sim-
ilar to those of morphine, including nausea, som-
nolence, dizziness, pruritus and headaches, and do
not occur with greater frequency in the elderly than
in younger patients.[106] In the US, the oral formu-
lation of oxycodone can be given rectally with
equivalent bioavailability.[109,110] The parenteral
solution may also be given intranasally, with a bio-
availability of approximately 46%.[109]
9.3.8 Hydromorphone
Hydromorphone, a hydrogenated ketone ana-
logue of morphine, is an MOR agonist and to a
lessor extent a DOR agonist.[111] There is biex-
ponential absorption kinetics for oral and rectal
hydromorphone, with an initial rapid phase of 2
hours and a slow sustained release thereafter.[111]
Hydromorphone has a high first-pass clearance
(62 33%) with wide interindividual variation,
similar to that of morphine. Oral bioavailability
ranges between 10 and 65%.[112-115] Rectal bio-
availability is 33% with a range of 1066%.
Plasma concentrations appear to correlate with
analgesia in experimental trials, although this
requires confirmation.[113] The oral-to-parenteral
bioequivalent ratio is 5 : 1 but ranges between
3 : 1 and 9 : 1.[111] Hydromorphone is slightly
more lipophilic than morphine.
Hydromorphone is glucuronidated to hydro-
morphone-3-glucuronide and hydromorphone-
6-glucuronide by conjugases and reduced to
hydromorphinone by ketone reductase.[111] The
metabolites hydromorphone-6-glucuronide and
hydromorphinone are analgesic but produced at
low concentrations, and do not contribute signifi-
cantly to analgesia. Hydromorphone-3-glucuron-
ide accumulates in renal failure with a ratio to the
parent compound of 100 : 1 (normal ratio 27 : 1)
and may induce neurotoxicity with renal fail-
ure.[111,116,117] Like that of morphine, the clearance
of hydromorphone is mainly dependent on hepatic
blood flow and, to a lesser extent, protein binding.
Adis International Limited. All rights reserved.
Davis & Srivastava
Clearance is reduced with reduced hepatic blood
flow.
The equianalgesic ratio between morphine and
hydromorphone is 5 : 1, with some variation.[118]
Sublingual absorption is poor because of drug
ionisation at the usual saliva pH of 6. Drug inter-
actions are similar to those with morphine, al-
though not formally studied and rarely reported.
The onset of action with oral administration is 38
minutes, similar to that of morphine.[113] The hys-
teresis of parenteral hydromorphone is 8 minutes,
which is between that of fentanyl and morphine,
reflecting relatively greater lipophilicity than
morphine. There is incomplete analgesic cross-
tolerance to morphine, and switching to hydromor-
phone from morphine because of morphine-related
adverse effects successfully controls pain and re-
solves toxicity in 7075% of cases.[111] Morphine-
associated pruritus resolves by switching to hy-
dromorphone.[119] Hydromorphone is more soluble
than morphine, so large amounts can be given sub-
cutaneously in small volumes, which is similar to
diamorphine.[111] The adverse effects of hydromor-
phone are comparable to those of morphine. Abuse
potential is not greater than that of morphine, nor
does addiction occur more frequently with rapid
administration.[111,120]
Hydromorphone is as effective as morphine in
the relief of cancer pain and chronic nonmalignant
pain.[121] Sustained-release hydromorphone is as
effective as normal-release morphine in control-
ling pain.[122] Hydromorphone is given orally, rec-
tally, subcutaneously, intravenously, epidurally
and intrathecally. Overall, hydromorphone has lit-
tle advantage over morphine as a first-choice po-
tent opioid.
9.4 Opioids in Hepatic and Renal Failure
Aging is associated with a gradual reduction in
hepatic and renal function. Comorbid conditions
accentuate and accelerate loss of renal and hepatic
function and further complicate pain management.
Analgesics must be carefully and cautiously pre-
scribed in the elderly with hepatic and renal failure.
The therapeutic index of most opioids is reduced
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
with end-organ failure. Opioid subclasses are dif-
ferently affected by hepatic and renal failure and
choices can be made on the basis of an under-
standing of altered pharmacokinetics of each op-
ioid in patients with hepatic and renal fail-
ure.[82,123]
The onset and time course of analgesia are re-
lated to the concentration of the drug and the du-
ration of drug concentration at receptor sites. The
total clearance of a drug is the sum of partial
clearances (hepatic and renal). The contribution of
end-organ failure to drug clearance is related to the
dominant site of metabolism or excretion and the
particular failing organ.[82] Reduced end-organ
function, which reduces drug clearance, in effect
increases the magnitude and duration of analgesia
such that administration intervals are extended and
doses are reduced.
9.4.1 Hepatic Failure
Hepatic opioid clearance is dependent on he-
patic blood flow, residual hepatic enzyme capac-
ity, degree of plasma protein drug binding, hepatic
extraction ratio and shunting (portal and in-
trahepatic).[82] With extensive hepatic extraction
(first-pass clearance), drug metabolism is predom-
inantly dependent on hepatic blood flow. Drug
clearance is reduced by both intra- and extrahep-
atic shunting and reduced hepatic blood flow,
which increases oral bioavailability.[82]
When hepatic extraction is low, metabolism de-
pends on both residual enzyme capacity (mono-
oxygenases or conjugases) and drug protein bind-
ing. Reduced drug binding increases the volume
of distribution and drug concentrations at recep-
tor sites and increases elimination without increas-
ing plasma concentrations; drug half-life is pro-
longed.[82]
Morphine
The half-life of morphine is increased from 24
hours in severe hepatic dysfunction. Oral bioavail-
ability increases as a result of shunting and reduced
hepatic blood flow. This is counterbalanced by an
increase in extrahepatic glucuronidation. Bilirubin
displaces morphine from binding sites but this is
not clinically significant because morphine has
Adis International Limited. All rights reserved.
43
low protein binding.[82] In severe hepatic failure,
drug half-life is doubled and oral bioavailability
increased. Morphine is one of the safest opioids to
use in cirrhosis and hepatic failure.
Methadone
The pharmacokinetics of methadone are rela-
tively unaltered in mild to moderate cirrhosis. In
severe hepatic disease, protein binding is reduced
and the volume of distribution increased, produc-
ing new stable plasma concentrations and in-
creased tissue concentrations.[82] Long-term al-
cohol consumption stimulates CYP3A4, which
increases methadone clearance and offsets im-
paired methadone clearance.[82] The kinetics of
methadone are relatively unchanged except in se-
vere disease.
Fentanyl
Fentanyl disposition is relatively unaffected by
mild to moderate hepatic disease. However, elim-
ination is prolonged from 3.58.7 hours when he-
patic blood flow is severely reduced, as demon-
strated by aortic cross-clamp pharmacokinetic
studies during aortic aneurysm repair. It may be
assumed that half-life will be prolonged in severe
hepatic disease because of reduced hepatic blood
flow and reduced CYP3A4 enzyme activity.[82]
Oxycodone
The elimination half-life of oxycodone is sig-
nificantly prolonged in hepatic failure because of
decreased CYP2D6 activity, which does not have
the hepatic reserve of CYP3A4, or glucuronida-
tion. CYP2D6 is extremely rate limiting and easily
saturated and is only a small fraction of the entire
hepatic cytochrome enzyme system.[74] The mean
half-life in end-stage liver disease (before liver
transplant) is 13.9 (range 4.624.4) hours; the half-
life of oxycodone is normally 3.4 (range 2.65.1)
hours. In severe liver disease, oxycodone accumu-
lates if a standard 4-hour dosage schedule is main-
tained.[124] Morphine, fentanyl and perhaps meth-
adone are preferred over oxycodone in severe liver
disease.
Drugs Aging 2003; 20 (1)
44
Opioids That Should Be Avoided in Hepatic Failure
The following opioids should not be used in
liver disease: codeine, tramadol, dextropropoxy-
phene, pethidine and pentazocine. Both codeine
and tramadol have reduced conversion to the anal-
gesic metabolite. Dextropropoxyphene has a sig-
nificant risk of causing hepatic toxicity. Pethidine
is converted to norpethidine, which accumulates
in liver failure. Pentazocine has an increased risk
of neurotoxicity and confusion in patients with
end-stage liver disease.[82]
9.4.2 Renal Failure
Renal failure can affect both the metabolism
and elimination of opioids or their metabolites. Re-
duced elimination and prolonged half-life of inac-
tive metabolites are clinically irrelevant, but pro-
longed elimination of active opioids or their active
metabolites leads to prolonged analgesia and in-
creased opioid adverse effects.
Codeine
Codeine is principally metabolised to codeine-
6-glucuronide. Demethylation of codeine produces
morphine, which accounts for 10% of codeine me-
tabolites. The half-life of codeine in patients with
renal failure is markedly prolonged to 18.7 9
hours compared with 4 0.6 hours in healthy indi-
viduals. Codeine produces sedation in renal failure
if doses are not reduced or administration intervals
prolonged.[125]
Morphine
Morphine clearance is slightly prolonged by
renal failure, presumably because of reduced
renal glucuronidation. However, elimination of
the potent MOR agonist metabolite morphine-6-
glucuronide is significantly prolonged. An in-
crease in plasma concentration of this metabolite
prolongs analgesia. A linear relationship exists
between creatinine clearance and renal clear-
ance of glucuronide metabolites.[125] CNS con-
centrations of morphine-6-glucuronide are repor-
ted to be 15 times greater in patients with renal
failure than in healthy individuals. [125] Perito-
neal dialysis does not improve morphine-6-glu-
curonide clearance. Haemofiltration and haemo-
Adis International Limited. All rights reserved.
Davis & Srivastava
dialysis reduce morphine plasma concentrations
by 75% and 48%, respectively. [125] Morphine
doses need to be significantly reduced in renal fail-
ure.
Fentanyl
It has been assumed that renal failure does not
influence fentanyl pharmacokinetics, since fen-
tanyl is metabolised to an inactive metabolite by
the liver and norfentanyl is not a significant anal-
gesic even though it accumulates with renal fail-
ure. However, renal CYP3A4 accounts for some of
the extrahepatic metabolism of fentanyl and may
contribute to elevated fentanyl concentrations with
renal failure.[126] In addition, uraemia reduces he-
patic metabolism of certain drugs with high hepatic
extraction ratios.[127] Binding of fentanyl to albu-
min and 1-acid glycoprotein is reduced by renal
failure, further affecting its volume of distribution
and elimination. Variation in clearance rates for
fentanyl increases from 4068% in individuals
with renal failure, resulting in less predictable dose
selection. Blood urea nitrogen levels above 60
mg/dl are associated with reduced fentanyl elimi-
nation.[127] Renal failure has been associated with
prolonged respiratory depression if fentanyl doses
are not reduced.
Methadone
Methadone does not have known neuroactive
metabolites. Renal excretion accounts for 20% of
the drug. However, pharmacokinetics are not sig-
nificantly altered in renal failure.[88] Some authors
do suggest a 50% dose reduction in severe renal
failure.[125]
Hydromorphone
Hydromorphone clearance is not affected by
modest renal dysfunction. Switching from mor-
phine to hydromorphone for morphine-induced ad-
verse effects as a result of renal impairment brings
improvement in 80% of patients.[128] On the other
hand, myoclonus, hallucinations, agitation and
confusion have been reported in patients with ad-
vanced renal failure while receiving hydromor-
phone, indicating that doses need to be reduced.[129]
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain 45

In summary, methadone and hydromorphone 9.5.3 Antiepileptic Drugs

are relatively safe in moderate renal failure. Phenobarbital and phenytoin reduce methadone
Methadone is relatively safe in severe renal failure. concentrations by 50% through induction of
Elimination of codeine, fentanyl, oxycodone, mor- CYP3A4, potentially causing opioid withdrawal
phine and morphine glucuronide is prolonged by despite stable methadone dosages.[88] Carbamaze-
renal failure and doses need to be reduced or ad- pine does the same by inducing CYP3A4 activity.
ministration intervals extended. Doses of all op- Thiopental (thiopentone) reduces the analgesia of
ioids need to be reduced in patients with severe fentanyl and alfentanil and increases sedation as-
renal failure. Dialysis corrects morphine clearance sociated with both.[130] Carbamazepine, phenytoin
to a certain extent, depending on the type of renal and phenobarbital increase the neurotoxicity of
replacement therapy. pethidine by accelerating the metabolism of the
parent drug to norpethidine.[130]
9.5 Drug Interactions with Opioids 9.5.4 Benzodiazepines
Methadone concentrations are increased by di-
Potential drug interactions (table III) are not a azepam through competitive inhibition of drug me-
contraindication to drug combinations but require
cautious prescribing. We frequently take advan- Table III. Drug interactions with opioids
tage of certain drug interactions to improve pain Decrease concentrations Increase concentrations
control (for example, morphine and adjuvant anal- Morphine
gesics). The best known drug interactions occur as Kaolin (oral) Phenothiazines
a result of induction or inhibition of CYP isoen- Phenytoin Tricyclic antidepressants
zymes.[130] Barbiturates Ranitidine (questionable clinical
Carbamazepine significance)
9.5.1 Antibacterials Oral contraceptives Cimetidine
Erythromycin inhibits CYP3A4, reducing clear- Metoclopramide
Diclofenac
ance of alfentanil. Information about the influence
(morphine-6-glucuronide but not
of erythromycin on the kinetics of other CYP3A4- morphine)
dependent opioids is relatively sparse. The anti- Methadone
fungal drug ketoconazole is a potent inhibitor of Nevirapine Fluvoxamine
CYP3A4 and reduces methadone clearance. Flu- Ritonavir Fluoxetine
conazole to a lesser extent delays methadone me- Phenobarbital Ketoconazole
(phenobarbitone)
tabolism.[88] Rifampicin (rifampin) induces CYP- Fluconazole
3A4 and increases methadone clearance. It also Carbamazepine Alcohol (acute)
Risperidone Diazepam
increases morphine glucuronidation and morphine
Alcohol (chronic)
clearance. Oxycodone clearance is also increased Cigarette smoking
by rifampicin. Rifampicin (rifampin)
Fusidic acid
9.5.2 Histamine H2 Antagonists
Oxycodone
Cimetidine impairs hepatic CYP isoenzymes Rifampicin Quinidine
and reduces hepatic blood flow, leading to delayed Fluoxetine
morphine clearance.[76] Cimetidine increases the Ketoconazole
terminal half-life of fentanyl. Ranitidine binds to Cimetidine
the CYP system with one-tenth the potency of ci- Ritonavir
metidine and has a reduced risk of impairing mor- Fentanyl
phine clearance. Famotidine has little drug inter- Thiopental (thiopentone) Cimetidine
action with the opioids. Midazolam

Adis International Limited. All rights reserved. Drugs Aging 2003; 20 (1)
46
tabolism. Dextropropoxyphene delays the clear-
ance of alprazolam. Lorazepam and morphine are
glucuronidated through the same isoenzyme
(UGT2B7) and may theoretically compete for con-
jugation.[76] Benzodiazepines added to opioids in-
crease both sedation and the risk of delirium, and
in general do not improve analgesia unless pain is
due to muscle spasm. Midazolam and diazepam
accentuate the respiratory depression caused by
fentanyl and methadone.[130]
9.5.5 Tricyclic Antidepressants
Methadone increases desipramine concentra-
tions through competition at mutually shared CYP
sites.[88] Amitriptyline and nortriptyline increase
morphine concentrations but do not influence oxy-
codone clearance or metabolism.[130]
9.5.6 Phenothiazines
Phenothiazines potentially worsen the hypoten-
sion, respiratory depression and lethargy of op-
ioids. The risk of hypotension is greater with the
combination of chlorpromazine and morphine than
with either drug alone.[130] Phenothiazines may in-
terfere with morphine metabolism.[76]
9.5.7 Metoclopramide
Metoclopramide accelerates the absorption of
sustained-release morphine, resulting in increased
sedation.[76,130]
9.5.8 Selective Serotonin Reuptake Inhibitors
Selective serotonin reuptake inhibitors are po-
tent inhibitors of several CYP isoenzymes, partic-
ularly CYP3A4 and CYP2D6. Methadone clear-
ance is delayed by fluvoxamine and oxycodone
metabolism is inhibited by fluoxetine.[88] The an-
algesia of codeine and tramadol is lost, since the
analgesic metabolites of both drugs are not pro-
duced.[74] The analgesia of oxycodone and hy-
drocodone is preserved, since the MOR agonist is
the parent drug.
9.5.9 Monoamine Oxidase Inhibitors
MAOIs used with either pethidine or dextropro-
poxyphene can precipitate serotonin syndrome.
Morphine, methadone and fentanyl can be safely
given with MAOIs.[130]
Adis International Limited. All rights reserved.
Davis & Srivastava
10. Pharmacological Treatment of
Neuropathic Pain in the Elderly
Individual variation in response to neuropathic
pain is the rule, and doses of analgesics and adju-
vant analgesics needed to achieve relief should be
individualised.[131] The goal is the least amount of
medication necessary for pain control, and should
be accomplished through titration of one drug at a
time.[131] Many patients require polypharmacy to
achieve adequate relief, and sequential dosage ad-
justments of individual medications are necessary
to avoid pharmacological confusion. A topical
agent with regional effects (e.g. topical lidocaine
[lignocaine]) is preferred in the treatment of pe-
ripheral neuropathies in the elderly because of
safety and tolerability.[131] Long-term safety is a
concern in chronic nonmalignant pain, but is a
lesser concern for patients with a short survival
expectancy. Adjuvant analgesics should be used
first, before opioids, particularly in nonmalignant
pain. Classification of drugs for neuropathic pain
is as follows: topical analgesics, adjuvant analge-
sics, opioids. Adjuvant analgesics are tricyclic an-
tidepressants, antiepileptic drugs, local anaesthe-
tics, 2-adrenergic agonists, corticosteroids and
capsaicin.
Common features of neuropathic pain are dam-
age to thermonociceptive neuronal pathways by
disease, hyperalgesia and allodynia in areas of sen-
sory deficit with subsequent development of sec-
ondary hyperalgesia in nonaffected contiguous
areas, and increased likelihood of a poor response
to opioids.[132] Nonmalignant neuropathic pain is a
common experience among the elderly, and is
found in one-third of patients with advanced can-
cer pain. The fundamental principles of successful
treatment are as follows:
reduction of peripheral sensory input through
sodium channel blockade and prevention of
spontaneous neuronal depolarisation;
pre- and postsynaptic blockade through NMDA
inhibitors, opioid agonists or calcium channel
antagonists;
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
facilitation of inhibitory monoamine pathways
that descend through the periaqueductal grey
(PAG);
inhibition of neuroactive amino acid release
(glutamate and aspartate);
potentiation of the neuronal inhibiting amino
acids (GABA and glycine).
Therefore, both antiepileptic drugs (sodium
channel blockers and GABA agonists) and tricy-
clic antidepressants (PAG monoamine reuptake
inhibitors) are analgesics for neuropathic pain and
are used before opioids, particularly if the pain is
nonmalignant.[132]
Low dosages are started in the elderly (gaba-
pentin 100300 mg/day, valproic acid 250mg at
night, nortriptyline 10mg at night, desipramine
10mg at night) and slowly titrated at 47 day in-
tervals. Response to pain frequently occurs at sub-
therapeutic dosages, for both anticonvulsant (val-
proate, gabapentin) and antidepressant activity
(nortriptyline, desipramine), and within days. Pa-
tients with neuropathic cancer pain may require
lower dosages than patients with nonmalignant
neuropathic pain (diabetes or post-herpetic neural-
gia), although this is not well documented. Anti-
epileptic drugs are selected for patients with car-
diac conduction defects or arrhythmia, patients on
anticholinergic medications, or those experiencing
dry mouth, urinary retention, constipation or other
anticholinergic adverse effects.[10]
The administration frequency influences com-
pliance with the drug regimen. Carbamazepine
requires multiple daily doses, as does gabapentin,
whereas valproic acid and secondary amine tricy-
clic antidepressants are given once nightly. Sev-
eral symptoms may be treated simultaneously with
either class of medication, and drug selection
should be based on comorbidities or associated
symptoms such as bipolar affective disorder, uni-
polar depression, colic, myoclonus, nocturnal uri-
nary incontinence, seizures and tenesmus.
Combining tricyclic antidepressants and anti-
epileptic drugs for refractory pain is often associ-
ated with drug interactions, except with gaba-
pentin. These pharmacokinetic drug interactions
Adis International Limited. All rights reserved.
47
occur through the CYP system. Opioids also re-
lieve neuropathic pain in a significant number of
patients and should be considered as a second-line
medication. Methadone, in particular, may be ef-
fective, although this has not been formally con-
firmed.[88] There is no analgesic cross-tolerance
between tricyclic antidepressants and antiepileptic
drugs, and sequential trials with each class may be
successful. Both are equally effective and reduce
pain in two-thirds of cases.[3]
10.1 Antiepileptic Drugs
10.1.1 Carbamazepine
Carbamazepine relieves trigeminal neuralgia
and is one of several reasonable choices for dia-
betic and cancer-related neuropathic pain. How-
ever, carbamazepine does not reduce central (tha-
lamic) mediated pain. Carbamazepine relieves
burning and lancinating pain and allodynia. It is
structurally related to imipramine and for this rea-
son should not be prescribed for patients with hy-
persensitivity to tricyclic antidepressants. Carba-
mazepine induces CYP3A4 and accelerates the
clearance of multiple drugs, including methadone
and dexamethasone.[5,88]
Carbamazepine is usually given orally, but is
also well absorbed rectally. The initial dosage is
100mg twice daily, with titration up to 1200
mg/day depending on response and tolerability.
Plasma drug concentrations do not correlate with
analgesia.
Carbamazepine should be avoided in patients
who are neutropenic. The elderly are particularly
susceptible to the adverse effects of carbamaze-
pine, which include nystagmus, dizziness, diplo-
pia, light-headedness, lethargy, cognitive changes
and the syndrome of inappropriate antidiuretic
hormone secretion.[21] Because of its adverse ef-
fect profile and drug interactions, carbamazepine
is a second-line adjuvant, whereas gabapentin is a
preferred first-line adjuvant.
10.1.2 Gabapentin
Gabapentin relieves cancer-associated neuro-
pathic pain, diabetic neuropathy, post-herpetic
neuralgia, reflex sympathetic dystrophy and tri-
Drugs Aging 2003; 20 (1)
48
geminal neuralgia.[21,133-135] Absorption from the
gastrointestinal tract is mediated by a carrier sys-
tem in the small bowel, and therefore the drug
should not be rectally administered. The half-life
can exceed 24 hours in some elderly patients and
steady-state concentrations may take as long as 4
7 days to achieve.[5] Gabapentin has fewer drug
interactions than carbamazepine and valproic acid.
It is superior to amitriptyline in the treatment of
diabetic neuropathy.[134]
The initial dosage of gabapentin is 100mg three
times daily if renal function is normal, gradually
titrated to 3600 mg/day or even higher depending
on response. Increasing the dosage does not neces-
sarily increase adverse effects. Bioavailability di-
minishes with increasing dosage because of satu-
ration of the gabapentin carrier. Administration of
morphine with gabapentin increases gabapentin
serum concentrations by reducing its renal clear-
ance.[76] Gabapentin pharmacodynamically en-
hances the analgesic effects of morphine.[134] Dos-
age reduction is necessary for patients with renal
dysfunction. Doses are as low as 100300mg after
dialysis in patients on intermittent renal dialysis,
but must be individually determined.
The adverse effects of gabapentin are somno-
lence, ataxia, fatigue, tremor, diplopia and nystag-
mus. Gabapentin is available only for oral admin-
istration. Cost is a major drawback; otherwise
gabapentin is ideal for neuropathic pain.
10.1.3 Valproic Acid
Valproic acid enhances GABA synthesis and
prevents GABA degradation.[136,137] Besides neu-
ropathic pain, valproic acid relieves migraine
headaches and cluster headaches.[137,138] It reduces
neuropathic pain associated with advanced cancer
in 55% of patients at dosages of 200600mg twice
daily.[139] Valproic acid is less expensive than
gabapentin, does not interact with opioids, partic-
ularly methadone, and differs from classical anti-
convulsant medications in that it does not induce
CYP3A4 activity.[88]
Valproic acid modestly inhibits mono-oxygen-
ases, UGTs and epoxide hydrolase, but has fewer
drug-drug interactions than classical antiepileptic
Adis International Limited. All rights reserved.
Davis & Srivastava
drugs.[140,141] Hepatic toxicity associated with
valproic acid is due to 4-ene-valproic acid forma-
tion by CYP2B1.[142,143] Valproic acid is largely
glucuronidated. Plasma concentrations of amitri-
ptyline and nortriptyline are increased 42% by
valproic acid; reduction of the dosage of these two
antidepressants may be necessary if they are com-
bined with valproic acid.
The initial dosage of valproic acid is 250mg at
night for individuals aged over 65 years and 500mg
at night for those aged under 65 years. Valproic acid
is taken orally, rectally and intravenously. In our
experience, oral valproic acid is cost effective with-
in hospice (capitated) reimbursement plans. Val-
proic acid needs to be administered less frequently
than gabapentin, and is more versatile in terms of
administration routes than gabapentin.
10.1.4 Lamotrigine
Lamotrigine blocks sodium channels and in-
hibits release of glutamate in the CNS. Very little
is known about its benefits in cancer pain, but
it reduces phantom limb pain, AIDS-associated
neuropathy, trigeminal neuropathy, poststroke
pain and painful diabetic neuropathy.[132] There
are large interindividual differences in the phar-
macokinetics of lamotrigine as well as the analge-
sia obtained. Lamotrigine is a weak inducer of
UGTs, but has fewer drug interactions than other
antiepileptic drugs except for gabapentin. Adverse
effects are ataxia, incoordination, blurred vision,
diplopia and, rarely, cutaneous reactions, particu-
larly when doses are rapidly titrated or the drug
is combined with valproic acid. [21]
The initial dosage of lamotrigine is 50 mg/day
for 14 days, then 50mg twice daily for 14 days, then
increased by 100mg per day each week to a maxi-
mum dosage of 600 mg/day. In general, dosages
range between 50 and 400 mg/day. Lamotrigine
can be administered orally or rectally.[21] Further
studies are necessary before it becomes a standard
drug for neuropathic pain.
10.2 Tricyclic Antidepressants
Tricyclic antidepressants facilitate noradrena-
line and serotonin neurotransmission in the PAG
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
by inhibiting monoamine reuptake. They are also
sodium channel blockers.[21] Neuropathic pain
associated with cancer and diabetes, and post-
herpetic neuralgia, are relieved by several tricyc-
lic antidepressants.[132] Pain reduction occurs in
5765% of these patients.[3] Responses are seen
with amitriptyline dosages as low as 25mg
daily.[144] Neuropathic pain usually responds to
one-third to one-half of the usual antidepressant
dosage and has an onset of action as early as 310
days, although longer duration of treatment and
higher dosages are necessary in some patients.[3]
A dose-response relationship is seen between 25
and 75 mg/day, but also an increased risk for ad-
verse effects.[144,145] Therapeutic antidepressant
blood concentrations are 50300 g/L, but there
is no clear therapeutic plasma concentration for
neuropathic pain.
Tricyclic antidepressants are lipophilic and
highly protein bound and distribute widely to
brain, liver, lung and heart.[146] Plasma concentra-
tions peak by 28 hours, and the drugs undergo
extensive first-pass hepatic clearance. Imipramine
is metabolised to desipramine and amitriptyline to
nortriptyline, and these metabolites are subse-
quently conjugated to glucuronic acid.[146] The sec-
ondary amines desipramine or nortriptyline con-
tribute to the analgesia of the parent drug and are
even more lipophilic.[146] The half-life of the sec-
ondary amines is twice that of the parent tertiary
amine, and more than a week may be required to
clear tricyclic antidepressants. These drugs are
metabolised more slowly by individuals over the
age of 60 years and in those receiving medica-
tions that competitively or noncompetitively block
CYP metabolism. Amitriptyline dosages correlate
with CYP2D6 phenotype.[74] Dosage adjustments
are needed according to age and polypharm-
acy.[146]
There is little analgesic cross-tolerance be-
tween the various tricyclic antidepressants. The
secondary amines are preferred because of fewer
anticholinergic adverse effects, fewer associated
falls in the elderly, less mental cloudiness and re-
duced risk for orthostatic hypotension. Amitripty-
Adis International Limited. All rights reserved.
49
line and its demethylated metabolite nortriptyline
increase plasma morphine concentrations and de-
lay morphine clearance.[76] The long half-life of all
four tricyclic antidepressants allows for once-daily
administration.
Tricyclic antidepressants should be used cau-
tiously in patients with closed angle glaucoma,
benign prostatic hypertrophy, urinary retention,
constipation, cardiovascular disease or second and
third degree heart block, conduction defects or
prolonged QT intervals, and severe liver disease.
Gabapentin and valproic acid are preferred to tri-
cyclic antidepressants in these patients because
of fewer drug interactions, lack of potential ar-
rhythmias, no constipation, dry mouth and other
anticholinergic adverse effects, and little risk for
urinary retention and orthostatic hypotension. Sud-
den withdrawal of tricyclic antidepressants may
produce abdominal pain, anorexia, apathy, bad
dreams, diarrhoea, drowsiness, headaches, insom-
nia, irritability, malaise, mania, movement disor-
ders and profuse sweating.[21]
A single bedtime dose produces fewer adverse
effects and takes advantage of the sedative benefits
of tricyclic antidepressants. Paradoxically, in rare
circumstances, tricyclic antidepressants cause in-
somnia, and if this occurs a single morning dose is
given. Other benefits of tricyclic antidepressants
include relief of tenesmus and nocturnal urinary
incontinence.
Standard dosages are 1025mg at night in the
elderly, with dose increments of 1025mg every
47 days until response. Once maximum pain re-
lief is obtained, the dosage is maintained for 1
month. A slow taper over 36 months is possible
thereafter without loss of benefit.[3] Most tricyclic
antidepressants are given orally; doxepin may be
given rectally and amitriptyline intramuscularly.
10.3 Local Anaesthetics
Local anaesthetics relieve pain, regardless of
their route of administration.[147] All are sodium
channel blockers. Transdermal lidocaine relieves
regional (dermatomal) pain only in the area in
which it is placed. Mexiletine is an oral congener
Drugs Aging 2003; 20 (1)
50
of lidocaine.[101] The third local anaesthetic fre-
quently used is bupivacaine, usually by neuroaxis
infusion in conjunction with an opioid.
Local anaesthetics are classified as amino esters
and amino amides. Both types are lipid-soluble
drugs with acid-base ionisation constant (pKa) val-
ues that determine potency.[147] Local anaesthetics
rapidly enter the CNS, with concentrations in the
cerebrospinal fluid usually 90% of serum concen-
trations.[147] Distribution of the drug consists of
two phases: a rapid uptake in highly vascularised
tissues and a slow redistribution, biotransforma-
tion and elimination phase.[147] Elimination is de-
termined by hepatic and cardiac function for most
local anaesthetics, and by renal function for lido-
caine.[147]
Local anaesthetics are useful in painful diabetic
neuropathy, post-stroke pain, cancer-associated
neuropathic pain, radiculopathies, arachnoiditis,
trigeminal neuralgia, phantom pain, thalamic pain
and lightning pains associated with tabes dor-
salis.[21,148] Doses of lidocaine are 15 mg/kg in-
fused over 535 minutes.[149] The initial dosage of
mexiletine is 150200mg once to twice daily ti-
trated to 1200mg. Usual dosages are 150300mg
every 8 hours.[21] Responses to parenteral lido-
caine have been used as a predictor of response to
other sodium channel blockers such as mexiletine
or carbamazepine.[114]
Contraindications to the use of local anaesthet-
ics include heart block. A cardiologist should see
the patient if there are any cardiac conduction ab-
normalities before placing patients on systemic li-
docaine or mexiletine. All local anaesthetics are
negative inotropic agents and must be used cau-
tiously in patients with decompensated heart fail-
ure. Adverse effects include dizziness, gastrointes-
tinal upset, headaches, irritability, nervousness and
tremors. Gastrointestinal upset is minimised by
taking mexiletine with food and by slow dose titra-
tion. Seizures have been reported with high doses
of mexiletine.[21]
Lidocaine is available parenterally or by trans-
dermal patch, bupivacaine by the parenteral, epi-
dural and intrathecal routes, and mexiletine orally.
Adis International Limited. All rights reserved.
Davis & Srivastava
10.3.1 Lidocaine (Lignocaine) Transdermal Patch
The lidocaine 5% patch releases drug that binds
to regional neuronal sodium channels in peripheral
damaged nerves, thereby reducing spontaneous
impulses. The patch itself, by protecting the skin,
may prevent allodynia; the patch does not produce
significant serum concentrations.[131] The lido-
caine patch does not work in centrally mediated
neuropathic pain syndromes or radicular pain from
spinal cord pathology. Adverse effects are mini-
mal, but local skin irritation can occur.
The elderly are ideal candidates for the patch
because of a low risk of adverse effects and drug
interactions and improved compliance. Individuals
need to be instructed to place the patch in the
proper location over the area of pain if it is to be
effective. Up to three patches are placed in the af-
fected area for up to 12 hours per day.[131]
10.4 Other Agents
10.4.1 Clonidine
Clonidine is an 2-adrenergic agonist that in-
hibits noradrenaline release through binding to ad-
renergic autoreceptors found on sympathetic neu-
rons and additionally by increases in GABAA
receptor activity.[132] Clonidine adds to morphine
analgesia when taken orally or by transdermal
patch and is supra-additive when infused with mor-
phine into the neuroaxis. Oral and transdermal
clonidine relieves diabetic neuropathy and post-
herpetic neuralgia. Neuroaxis clonidine with op-
ioids is an acceptable combination for refractory
cancer pain. Adverse effects include hypotension,
constipation, dizziness, dry mouth, rebound hyper-
tension, sedation and sexual dysfunction. Trans-
dermal preparations can produce cutaneous ery-
thema and pruritus.
The dosage of oral clonidine is 0.1mg twice
daily increased weekly by 0.3mg twice daily up to
a maximum dosage of 2.4 mg/day depending on
response and tolerability.[21] Transdermal clonid-
ine needs to be changed weekly. Clonidine is avail-
able orally, transdermally, epidurally and intra-
thecally.
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
10.4.2 Capsaicin
Capsaicin binds to vanilloid receptors, which in
time exhausts neuronal substance P, a major neu-
rotransmitter of pain for C fibres.[132] Starting
doses are 0.0250.075% topical cream applied
five times daily. Initial administration produces
burning, which is relieved when capsaicin is com-
bined with topical anaesthetics. Compliance is a
challenge because of multiple daily applications
and worsening burning pain with initial use. On
the other hand, there are few systemic adverse ef-
fects. Cost is an issue.
10.4.3 Baclofen
Baclofen is a GABAB analogue that inhibits
release of presynaptic excitatory amino acids, de-
creases presynaptic calcium conductance and in-
creases postsynaptic potassium conductance and
thereby maintains afferent neurons in a hyper-
polarised state. Baclofen effectively relieves tri-
geminal neuralgia, glossopharyngeal neuralgia
and ophthalmic-post-herpetic neuralgia with
fewer adverse effects than the alternative, carbam-
azepine.[21,148] Baclofen has been successfully
combined with carbamazepine in refractory pain,
but with increased risk for drowsiness, nausea and
vomiting.
Oral baclofen is rapidly absorbed and has a
specialised transport process necessary for intes-
tinal absorption.[148] Peak serum concentrations
occur within 2 hours, and plasma half-life varies
between 3 and 7 hours. Baclofen is 80% excreted
unchanged in the urine.[148]
Common adverse effects of baclofen are ataxia,
dizziness, drowsiness, mental confusion, nausea
and vomiting.[21] Abrupt discontinuation of bac-
lofen leads to withdrawal reactions (hallucina-
tions, seizures, tachycardia), so the drug should be
tapered by 510mg per day at weekly intervals.[21]
Baclofen cannot be tolerated by 10% of patients
because of gastrointestinal upset. Patients with re-
nal disease may develop drug toxicity at low dos-
ages because of impaired elimination.[148]
The dosage of baclofen is usually 510mg two
to three times daily, increased by 510mg daily
Adis International Limited. All rights reserved.
51
every 23 days as tolerated to a 5060mg maxi-
mum dose per day.
11. Conclusions
The principles of geriatric pain management
can be summarised by the following points.
1. Use self-assessment tools to measure pain se-
verity. Use the same tool consistently on repeat
assessment. Unidimensional tools are used for pain
response, multidimensional tools as part of pain
evaluation.
2. Be aware that pain scales do not avoid the
need for a good pain history and physical exami-
nation.
3. Make a reasonably accurate diagnosis to fa-
cilitate pain management.
4. Assess the effect of pain on activities of daily
living and use functional status and daily activities
as one of the measures of pain relief.
5. Screen for depression, delirium and demen-
tia.
6. Use diagnostic procedures for the purpose of
improving pain management.
7. Treat pain in order to facilitate diagnostic
procedures.
8. Combine pharmacological and non-
pharmacological therapies. Use nonpharmacolog-
ical therapies that are acceptable to the patient.
9. Mobilise patients physically and psycholog-
ically by use of orthopaedic aids, surgery, radiation
and other supportive services.
10. Goal orient patients to pain control with im-
proved function and use incremental goals such as
pain relief that allows for sleep, pain relief at rest,
pain control with activities.
11. Start analgesics at a low dose and titrate
slowly, particularly in the frail or those with renal
or hepatic dysfunction. Table IV provides initial
and maximum dosages for analgesics in the el-
derly.
12. For opioid administration strategies, take
into account the severity of the chronic baseline
pain and the severity of incident pain. Doses and
schedule may need to be independently deter-
mined for chronic underlying pain and incident
Drugs Aging 2003; 20 (1)
52 Davis & Srivastava

pain. Dosage needs to be tailored to asymmetrical
pain patterns.
13. Use adjuvant analgesics to reduce opioid ad-
verse effects and improve pain relief in patients
with poorly controlled pain. Use medications
whose metabolism is less influenced by age and
that have few adverse effects and drug interactions
(table III and table V).
14. The preferred nociceptive adjuvants in the
elderly are paracetamol, short-acting NSAIDs,
COX-2 inhibitors and sometimes corticosteroids.
15. Rofecoxib is the preferred COX-2 inhibitor.
16. Avoid ketorolac, indomethacin, piroxicam
and mefenamic acid.
17. Do not combine NSAIDs and corticoste-
roids.
18. Combinations of NSAIDs and opioids are
convenient and improve compliance, but are dose-
limited by the NSAID.
19. The preferred adjuvant drugs for neuro-
pathic pain in the elderly are gabapentin, valproic
acid, desipramine and nortriptyline.

Table IV. Equianalgesics in the elderly

Analgesic Initial dosage (oral) Maximum/therapeutic Comments
dosagea
Paracetamol 500mg q4h 4000mg Avoid combination with NSAIDs; lower dosage with
(acetaminophen) weight loss, liver disease and regular alcohol use
Naproxen 250mg q812h 500mg tid Avoid combining with corticosteroids; do not use in
aspirin-sensitive asthma; avoid in renal failure and
congestive heart failure; combine with double-dose
histamine H2 antagonists in high-risk patients; delirium
can occur when combined with ciprofloxacin
Ibuprofen 200400mg q4h 800mg tid As for naproxen
Rofecoxib 12.525mg od 50mg od Second line; expense may prohibit its use in geriatric
patients (US)
Morphine sulphate 5mg q4h No ceiling Start low, go slow in frail elderly; proactively start
laxatives; watch dosage in renal failure
Hydromorphone 12mg q4h No ceiling Metabolism similar to morphine; expense may prohibit use
Oxycodone 5mg q4h No ceiling Dosage is less age dependent but very dependent on
hepatic and renal function
Methadone 35mg q3h as required or No ceiling Requires experience; drug interactions are common
35mg q8h regularly (SSRIs, antiepileptic drugs, benzodiazepines); very
inexpensive
Fentanyl 25 g/h transdermal patch No ceiling Not versatile; expensive; requires other short-acting
q72h sublingual opioids or fentanyl for breakthrough; improved
compliance
Desipramine 1025mg qhs 100150mg Elderly require lower dosages; avoid with cardiac
conduction defects; withdrawal reactions if abruptly
stopped
Nortriptyline 1025mg qhs 100150mg As for desipramine
Methylphenidate 5mg every morning to 5mg 10mg bid (morning and Potentiates opioid analgesia; avoid in heart failure and
bid (morning and afternoon) delirium
afternoon)
Valproic acid 250mg qhs 10001500mg od or Fewer drug interactions than classic antiepileptic drugs;
1000mg bid less expensive than gabapentin; rare hepatic toxicity
Gabapentin 100mg q8h 24003600mg in divided Clearance is renal dependent; few drug interactions;
doses withdrawal reactions can occur if stopped abruptly
Dexamethasone 48mg bid 100mg Interactions with antiepileptic drugs; taper to lowest
effective dosage
Prednisone 1020mg bid to tid 60100mg Same as dexamethasone
a Daily doses unless otherwise indicated.
bid = twice daily; od = once daily; qhs = at night; qxh = every x hours; SSRI = selective serotonin reuptake inhibitor; tid = three times daily.

Adis International Limited. All rights reserved. Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain 53

Table V. Drugs where metabolism is or is not affected by age[22] Acknowledgements

Affected Not affected
Pethidine (meperidine) Paracetamol (acetaminophen) The authors deeply appreciate the expertise of Michele
Morphine Oxazepam Wells who prepared this manuscript. No sources of funding
Ibuprofen Lorazepam were used to assist in the preparation of this manuscript. The
Naproxen Temazepam authors have no conflicts of interest that are directly relevant
Alprazolam Phenytoin to the content of this manuscript.
Desipramine Valproic acid
Diazepam
References
Imipramine 1. Cleary JR, Carbone PP. Palliative medicine in the elderly. Can-
Nortriptyline cer 1997 Oct; 80 (7): 1335-47
Trazodone 2. Overcash J. Symptom management in the geriatric patient. Can-
Triazolam cer Control 1998 May; 5 (3): 46-7
3. Gloth FM. Pain management in older adults: prevention and
treatment. J Am Geriatr Soc 2001 Feb; 49 (2): 188-99
4. Medling PS. Is there such a thing as geriatric pain? Pain 1991
Aug; 46 (2): 119-21
20. Avoid dextropropoxyphene, pethidine, pen- 5. Weiner DK, Hanlon JT. Pain in nursing home residents: man-
tazocine and mixed agonist-antagonists and ami- agement strategies. Drugs Aging 2001; 18 (1): 13-29
6. Ferrell BA. Pain evaluation and management in the nursing
triptyline. Avoid combining classical antiepileptic home. Ann Intern Med 1995 Nov; 123 (9): 681-7
drugs with tricyclic antidepressants. 7. Ferrell BA, Ferrell BR, Rivera L. Pain in cognitively impaired
nursing home patients. J Pain Symptom Manage 1995 Nov;
21. Choose opioids based on versatility, poly- 10 (8): 591-8
pharmacy, severity and type of pain, availability, 8. Severn AM, Dodds C. Cognitive dysfunction may complicate
assessment of pain in elderly patients [letter]. BMJ 1997 Aug;
associated symptoms, and renal and hepatic func- 315 (7107): 551
tion. 9. Ferrell BA, Stein WM, Beck JC. The Geriatric Pain Measure:
22. Tramadol and codeine have ceiling dosages, validity, reliability and factor analysis. J Am Geriatr Soc 2000
Dec; 48 (12): 1669-73
are liable to drug interactions and are greatly influ- 10. American Geriatrics Society. The management of chronic pain
enced by hepatic and renal failure; they are not in older persons: AGS panel on chronic pain in older persons.
J Am Geriatr Soc 1998 May; 46 (5): 635-51
first-line analgesics in elderly patients. 11. Lansbury G. Chronic pain management: a qualitative study of
23. Morphine is least influenced by hepatic fail- elderly peoples preferred coping strategies and barriers to
management. Disabil Rehabil 2000 Jan; 22 (1-2): 2-14
ure but its glucuronide metabolites are greatly af- 12. Ward SE, Goldberg N, Miller-McCauley V, et al. Patient-
fected by renal failure. related barriers to management of cancer pain. Pain 1993
Mar; 52 (3): 319-24
24. Oxycodone is significantly affected by he- 13. Cleeland CS. Undertreatment of cancer pain in elderly patients.
patic and renal failure. JAMA 1998 Jun; 279 (23): 1914-5
14. Von Roenn JH. Are we the barrier? J Clin Oncol 2001 Dec; 19
25. Paradoxically, fentanyl accumulates in se- (23): 4273-4
vere renal failure but is relatively safe in hepatic 15. Bernabei R, Gambassi G, Lapane K, et al. Management of pain
in elderly patients with cancer. SAGE Study Group System-
failure. atic Assessment of Geriatric Drug Use via Epidemiology.
26. Methadone requires minor dosage reduction JAMA 1998 Jun; 279 (23): 1877-82
in renal and hepatic failure. 16. Nikolaus T. Assessment of chronic pain in elderly patients.
Ther Umsch 1997 Jun; 54 (6): 340-4
27. Change one drug at a time in patients taking 17. Kewman DG, Vaishampayan N, Zald D, et al. Cognitive im-
adjuvants and opioids. Simultaneous drug and pairment in musculoskeletal pain patients. Int J Psychiatry
Med 1991; 21 (3): 253-62
dose changes lead to confusion, particularly if ad- 18. Lamberg L. Chronic pain linked with poor sleep; exploration
verse effects occur. of causes of treatment. JAMA 1999 Feb; 281 (8): 691-2
19. Schnurr RF, MacDonald MR. Memory complaints of chronic
28. Always reassess, particularly if there is a pain. Clin J Pain 1995; 11 (2): 103-11
change in pain pattern and severity. Continuity and 20. Eccleston C. Chronic pain and attention: a cognitive approach.
Br J Clin Psychol 1994; 33 (4): 535-47
persistence are major factors in successful pain 21. Guay DR. Adjunctive agents in the management of chronic
management. pain. Pharmacotherapy 2001 Sep; 21 (9): 1070-81

Adis International Limited. All rights reserved. Drugs Aging 2003; 20 (1)
54
22. Tumer N, Scarpace PJ, Lowenthal DT. Geriatric pharmacology:
basic and clinical considerations. Annu Rev Pharmacol Tox-
icol 1992; 32: 271-302
23. Flacker JM, Lipsitz LA. Neural mechanisms of delirium: cur-
rent hypotheses and evolving concepts. J Gerontol A Biol Sci
Med Sci 1999 Jun; 54 (6): B239-46
24. Coyle N, Breitbart W, Weaver S, et al. Delirium as a contrib-
uting factor to crescendo pain: three case reports. J Pain
Symptom Manage 1994 Jan; 9 (1): 44-7
25. Gibson MC, Schroder C. The many faces of pain for older,
dying adults. Am J Hosp Palliat Care 2001 Jan-Feb; 18 (1):
19-25
26. Bergh I, Sjostrom B, Oden A, et al. An application of pain rating
scales in geriatric patients. Aging 2000 Oct; 12 (5): 380-7
27. Freedman G, Peruvemba R. Geriatric pain management. The
anesthesiologists perspective. Anesthesiol Clin North Am
2000; 18 (1): 123-41
28. Rosenberg SK, Boswell MV. Pain management in geriatric. In:
Weiner RS, editor. Pain management: a practical guide for
clinicians. Boca Raton (FL): St Lucie Press, 1998: 683-92
29. Farrell MJ, Katz B, Helme RD. The impact of dementia on the
pain experience. Pain 1996; 67: 7-15
30. Angst MS, Brose W, Dyck JB. The relationship between the
visual analog pain intensity and pain relief scale changes dur-
ing analgesic drug studies in chronic pain patients. Anesthe-
siology 1999; 91 (1): 34-41
31. Porter FL, Malhotra KM, Wolf CM, et al. Dementia and res-
ponse to pain in the elderly. Pain 1996; 68: 413-21
32. Barkwell DP. Ascribed meaning: a critical factor in coping and
pain attenuation in patients with cancer-related pain. J Palliat
Care 1991; 7 (3): 5-14
33. Kappauf HW. Oncologic pain therapy: are there alternatives
to pill and high-tech? MMW Fortschr Med 1999; 142 (42):
38-41
34. Pincus T, Williams A. Models of psychosomatic research. J
Psychosom Res 1999; 47 (3): 211-9
35. Fishbain DA, Cutler R, Rosomoff HL, et al. Chronic pain-
associated depression: antecedent or consequence of chronic
pain? A review. Clin J Pain 1997; 13 (2): 116-37
36. Proctor T, Gatchel RJ, Robinson RC. Psychosocial factors and
risk of pain and disability. Occup Med 2000; 15 (4): 803-12
37. Zaza C, Stolee P, Prkachin K. The application of goal attain-
ment scaling in chronic pain settings. J Pain Symptom Man-
age 1999; 17 (1): 55-64
38. Gatchel RJ. A biopsychosocial overview of pretreatment
screening of patients with pain. Clin J Pain 2001; 17 (3): 192-9
39. Eccelston C. Role of psychology in pain management. Br J
Anaesth 2001; 87 (1): 144-52
40. Cutler RB, Fishbain DA, Lu Y, et al. Prediction of pain center
treatment outcome for geriatric chronic pain patients. Clin J
Pain 1994; 10 (1): 1-2
41. Anonymous. Integration of behavioral and relaxation ap-
proaches into the treatment of chronic benign pain and insom-
nia: NIH technology assessment panel on integration and
behavioral and relaxation approaches into the treatment of
chronic pain and insomnia. JAMA 1996; 276 (4): 313-8
42. Morley S, Eccelston C, Williams A. Systematic review and
meta-analysis of randomized controlled trials of cognitive
behaviour therapy and behaviour therapy for chronic pain in
adults, excluding headache [online]: Database of Abstracts of
Reviews of Effectiveness 2002. Available from URL:
http://nhscrd.york.ac.uk/online/dare/990758.htm [Accessed
2002 Nov 11]
Adis International Limited. All rights reserved.
Davis & Srivastava
43. Flor H, Birbaumer M, Schugens MM, et al. Symptom-specific
psychophysiological responses in chronic pain patients. Psy-
chophysiology 1992; 29 (4): 452-60
44. Flor H, Fydrich T, Turk DC. Efficacy of multidisciplinary pain
treatment centers: a meta-analytic review. Pain 1992; 49 (2):
221-30
45. Morley S, Eccelston C, Williams A. Systematic review and
meta-analysis of randomized controlled trials of cognitive
behaviour therapy and behaviour therapy for chronic pain in
adults, excluding headache. Pain 1999; 80: 1-13
46. Ciccone DS, Grezesiak RC. Cognitive dimensions of chronic
pain. Soc Sci Med 1984; 19 (12): 1339-45
47. Roche RJ, Forman WB. Pain management for the geriatric pa-
tient. Clin Podiatr Med Surg 1994; 11 (1): 41-53
48. Turk DC. Remember the distinction between malignant and
benign pain?: well, forget it. Clin J Pain 2002; 18: 75-6
49. Sorge J. The lesson from cancer pain. Eur J Pain 2000; 4
Suppl. A: 3-7
50. Egberts ACG, Lenderink AW, de Koning FHP. Channeling of
three newly introduced antidepressants to patients not res-
ponding satisfactorily to previous treatment. J Clin Pharmacol
1997; 17 (3): 149-55
51. Bellville JW, Forrest WH, Miller E. Influence of age on pain
relief from analgesics: a study of postoperative patients.
JAMA 1971 Sep; 217 (13): 1835-41
52. Kaiko RF. Age and morphine analgesia in cancer patients with
postoperative pain. Clin Pharmacol Ther 1980; 28(6): 823-6
53. McLeod PJ, Huang AR, Tamblyn RM, et al. Defining inappro-
priate practices in prescribing for elderly people: a national
consensus panel. Can Med Assoc J 1997 Feb; 156 (3): 385-91
54. Ferry JJ, Wagner JG. The non-linear pharmacokinetics of pred-
nisone and prednisolone. Biopharm Drug Dispos 1986 Jan-
Feb; 7 (1): 91-101
55. Greenberger PA, Chow MJ, Atkinson AJ, et al. Comparison of
prednisolone kinetics in patients receiving daily or alternate-
day prednisone for asthma. Clin Pharmacol Ther 1986 Feb;
32 (9): 163-8
56. Magee MH, Blum RA, Lates CD, et al. Clinical pharmacoki-
netics and pharmacodynamics in relation to sex and race. J
Clin Pharmacol 2001 Nov; 41 (11): 1180-94
57. Faure C, Andre J, Pelatan C, et al. Pharmacokinetics of intrave-
nous methylprednisolone and oral prednisone in paediatric
patients with inflammatory bowel disease during the acute
phase and in remission. Eur J Clin Pharmacol 1998; 54 (7):
555-60
58. Garg V, Jusko WJ. Bioavailability and reversible metabolism
of prednisone and prednisolone in man. Biopharm Drug Dis-
pos 1994 Mar; 15 (2): 162-72
59. Cavanaugh JH, Karol MD. Lack of pharmacokinetic interaction
after administration of lansoprazole or omeprazole with pred-
nisone. J Clin Pharmacol 1996 Nov; 36 (11): 1064-71
60. Leburn-Vignes B, Archer VC, Diquet B, et al. Effect of
itraconazole on the pharmacokinetics of prednisolone and
methylprednisolone and cortisol secretion in healthy subjects.
Br J Clin Pharmacol 2001 May; 51 (5): 443-50
61. Frey BM, Frey FJ. Clinical pharmacokinetics of prednisone and
prednisolone. Clin Pharmacokinet 1990 Aug; 19 (2): 126-46
62. Toth GG, Kloosterman C, Uges DR, et al. Pharmacokinetics of
high-dose oral and intravenous dexamethasone. Ther Drug
Monit 1999 Oct; 21 (5): 532-5
63. OSullivan BT, Cutler DJ, Hunt GE, et al. Pharmacokinetics of
dexamethasone and its relationship to dexamethasone sup-
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
pression test outcome in depressed patients and healthy con-
trol subjects. Biol Psychiatry 1997 Mar; 41 (5): 574-84
64. Elliott CL, Read GF, Wallace EM, et al. The pharmacokinet-
ics of oral and intramuscular administration of dexametha-
sone in late pregnancy. Acta Obstet Gynecol Scand 1996
Mar; 75 (3): 213-6
65. Lamiable D, Vistelle R, Sulmont V, et al. Pharmacokinetics of
dexamethasone administered orally in obese patients. The-
rapie 1990 Jul-Aug; 45 (4): 311-4
66. Guthrie SK, Heidt M, Pande A, et al. A longitudinal evaluation
of dexamethasone pharmacokinetics in depressed patients
and normal controls. J Clin Psychopharmacol 1992 Jun; 12
(3): 191-6
67. Miller LG. Cigarettes and drug therapy: pharmacokinetic and
pharmacodynamic considerations. Clin Pharm 1990 Feb; 9
(2): 125-35
68. Cummings DM, Larijani GE, Conner DP, et al. Charac-
terization of dexamethasone binding in normal and uremic
human serum. DICP 1990 Mar; 24 (3): 29-31
69. Brady ME, Sartiano GP, Rosenblum SL, et al. The pharmaco-
kinetics of single high doses of dexamethasone in cancer pa-
tients. Eur J Clin Pharmacol 1987; 32 (6): 593-6
70. Kutemeyer S, Schurmeyer TH, von zur Muhlen A. Effect of
liver damage on the pharmacokinetics of dexamethasone. Eur
J Endocrinol 1994 Dec; 131 (6): 594-7
71. Gupta SK, Ritchie JC, Ellinwood EH, et al. Modeling the phar-
macokinetics and pharmacodynamics of dexamethasone in
depressed patients. Eur J Clin Pharmacol 1992; 43 (1): 51-5
72. Maguire KP, Tuckwell VM, Schweitzer I, et al. Dexamethasone
kinetics in depressed patients before and after clinical res-
ponse. Psychoneuroendocrinology 1990; 15 (2): 113-23
73. Lewis KP, Stanley GD. Pharmacology. In: Gilbertson LI, edi-
tor. Conscious sedation. Philadelphia (PA): Lippincott Wil-
liams & Wilkins, 1999: 73-86
74. Davis MP, Homsi J. The importance of cytochrome P450
monooxygenase CYP2D6 in palliative medicine. Support
Care Cancer 2001; 9: 442-51
75. American Society of Anesthesiologists Task Force on Pain
Management, Cancer Pain Section. Practice guidelines for
cancer pain management: a report. Anesthesiology 1996; 84
(5): 1243-57
76. Donnelly S, Davis MP, Walsh D, et al. Morphine in cancer pain
management: a practical guide. Support Care Cancer 2002;
10: 13-35
77. Hanks GW, de Conno F, Cherny N, et al. Morphine and alter-
native opioids in cancer pain: the EAPC recommendations.
Br J Cancer 2001; 84 (5): 587-93
78. Duggleby W, Lander J. Cognitive status and postoperative pain:
older adults. J Pain Symptom Manage 1994; 9 (1): 19-27
79. De Stoutz ND, Bruera E, Surarez-Alamzor M. Opioid rotation
for toxicity reduction in terminal cancer patients. J Pain
Symptom Manage 1995 Jul; 10 (5): 378-84
80. Indelicato RA, Portenoy RK. Opioid rotation in the manage-
ment of refractory cancer pain. J Clin Oncol 2002; 20 (1):
348-52
81. Tawfik MO, Elborolossy K, Nasr F. Tramadol hydrochloride
in relief of cancer pain: a double blind comparison against
sustained release morphine [abstrac]. Pain 1990; Suppl. 5: S377
82. Tegeder I, Lotsch J, Geisslinger G. Pharmacokinetics of opioid
in liver disease. Clin Pharmacokinet 1999 Jul; 37 (1): 17-40
83. Mercadante S, Portenoy RK. Opioid poorly-responsive cancer
pain. Part 1: clinical considerations. J Pain Symptom Manage
2001; 21 (2): 144-50
Adis International Limited. All rights reserved.
55
84. Dionne RA, Lepinski AM, Gordon SM, et al. Analgesic effects
of peripherally administered opioids in clinical models of
acute and chronic inflammation. Clin Pharmacol Ther 2001
Jul; 70 (1): 66-73
85. Twillman RK, Long TD, Cathers TA, et al. Treatment of painful
skin ulcers with topical opioids. J Pain Symptom Manage
1999 Apr; 17 (4): 288-92
86. Krajnik M, Zylicz Z, Finlay I. Potential uses of topical opioids
in palliative care: report of 6 cases. Pain 1999 Mar; 80 (1-2):
121-5
87. Boulton DW, Arnaud P, DeVane CL. Pharmacokinetics and
pharmacodynamics of methadone enantiomers after a single
oral dose of racemate. Clin Pharmacol Ther 2001 Jul; 70 (1):
48-57
88. Davis MP, Walsh D. Methadone for relief of cancer pain: a
review of pharmacokinetics, pharmacodynamics, drug inter-
actions and protocols of administration. Support Care Cancer
2001; 9: 73-83
89. Peng PW, Sandler AN. A review of the use of fentanyl analgesia
in the management of acute pain in adults. Anesthesiology
1999 Feb; 90 (2): 576-99
90. Breitbart W, Chandler S, Eagel B, et al. An alternative algo-
rithm for dosing transdermal fentanyl for cancer-related pain.
Oncology 2000 May; 14 (5): 695-705
91. Coluzzi PH, Schwartzberg L, Conroy JD, et al. Breakthrough
cancer pain: a randomized trial comparing oral transmucosal
fentanyl citrate (OTFC (r)) and morphine sulfate immediate
release (MSIR (r)). Pain 2001; 91: 123-30
92. Gardner-Nix J. Oral transmucosal fentanyl and sufentanil for
incident pain. J Pain Symptom Manage 2001 Aug; 22 (2):
627-30
93. Hanks GW, Fallon MT. Transdermal fentanyl in cancer pain:
conversion from oral morphine [letter]. J Pain Symptom
Manage 1995 Feb; 10 (2): 87
94. Labroo RB, Paine MF, Thummel KE, et al. Fentanyl metabo-
lism by human hepatic and intestinal cytochrome P450 3A4:
implications for interindividual variability in disposition,
efficacy, and drug interactions. Drug Metab Dispos 1997; 25
(9): 1072-80
95. Mercadante S. Recent progress in the pharmacotherapy of can-
cer pain. Expert Review Anticancer Ther 2001; 1 (3): 487-94
96. Poyhia R, Olkkola KT, Seppala T. The pharmacokinetics of
oxycodone after intravenous injection in adults. Br J Phar-
macol 1991; 32 (4): 516-8
97. Poyhia R, Seppala T. solubility and protein binding of oxy-
codone in vitro. Pharmacol Toxicol 1994; 74: 23-7
98. Curtis GB, Johnson GH, Clark P, et al. Relative potency of
controlled-release oxycodone and controlled-release mor-
phine in a postoperative pain model. Eur J Clin Pharmacol
1999; 55 (6): 425-9
99. Heiskanen T, Olkkola KT, Kalso E, et al. Effects of blocking
CYP2D6 on the pharmacokinetics and pharmacodynamics of
oxycodone. Clin Pharmacol Ther 1998 Dec; 64 (6): 603-11
100. Kaplan R, Parris WC, Citron ML, et al. Comparison of controlled-
release and immediate release oxycodone tablets in patients
with cancer pain. J Clin Oncol 1998 Oct; 16 (10): 3230-7
101. Poyhia R, Vainio A, Kalso E. A review of oxycodones clinical
pharmacokinetics and pharmacodynamics. J Pain Symptom
Manage 1993 Feb; 8 (2): 63-7
102. Bruera E, Belzile M, Pituskin E, et al. Randomized, double-
blind, cross-over trial comparing safety and efficacy of oral-
release oxycodone with controlled-release morphine in patients
with cancer pain. J Clin Oncol 1998 Oct; 16 (10): 3222-9
Drugs Aging 2003; 20 (1)
56
103. Gagnon B, Bielech M, Watanabe S, et al. The use of intermittent
subcutaneous injections of oxycodone for opioid rotation in
patients with cancer pain. Support Care Cancer 1999; 7 (4):
265-70
104. Zhukovsky DS, Walsh D, Doona M. The relative potency be-
tween high dose oral oxycodone and intravenous morphine:
a case illustration. J Pain Symptom Manage 1999 Jul; 18 (1):
53-5
105. Nielsen CK, Ross FB, Smith MT. Incomplete, asymmetric, and
route-dependent cross-tolerance between oxycodone and
morphine in the Dark Agouti rat. J Pharmacol Exp Ther 2000
Oct; 295 (1): 91-9
106. Kaiko RF, Benzinger DP, Fitzmartin RD, et al. Pharmacokinetic-
pharmacodynamic relationships of controlled-release oxyco-
done. Clin Pharmacol Ther 1996 Jan; 59 (1): 52-61
107. Colucci RD, Swanton RE, Thomas GB, et al. Relative variabil-
ity in bioavailability of oral controlled-release formulations
of oxycodone and morphine. Am J Ther 2001 Jul-Aug; 8
(4): 231-6
108. Maddocks I, Somogyi A, Abbott F, et al. Attenuation of
morphine-induced delirium in palliative care by substitution
with infusion of oxycodone. J Pain Symptom Manage 1996
Sep; 12 (3): 182-9
109. Takala A, Kaasalainen V, Seppala T, et al. Pharmacokinetic
comparison of intravenous and intranasal administration of
oxycodone. Acta Anaethesiol Scand 1997; 41: 309-12
110. Leow KP, Smith MT, Watt JA, et al. Comparative oxycodone
pharmacokinetics in humans after intravenous, oral, and rec-
tal administration. Ther Drug Monit 1992; 14: 479-84
111. Sarhill N, Walsh D, Nelson KA. Hydromorphone: pharmacol-
ogy and clinical applications in cancer patients. Support Care
Cancer 2001; 9: 84-96
112. Ritschel WA, Parab PV, Denson DD, et al. Absolute bioavail-
ability of hydromorphone after peroral and rectal administra-
tion in humans: saliva/plasma ratio and clinical effects. J Clin
Pharmacol 1987 Sep; 29 (9): 567-53
113. Agnst MS, Drover DR, Lotsch J, et al. Pharmacodynamics of
orally administered sustained-release hydromorphone in hu-
mans. Anesthesiology 2001 Jan; 94 (1): 63-73
114. Parab PV, Ritschel WA, Coyle DE, et al. Pharmacokinetics of
hydromorphone after intravenous, peroral and rectal admin-
istration in humans. Biopharm Drug Dispos 1988 Mar-Apr; 9
(2): 187-99
115. Coda B, Tanaka A, Jacobson RC, et al. Hydromorphone anal-
gesia after intravenous bolus administration. Pain 1997 May;
71 (1): 41-8
116. Wright AW, Mather LE, Smith MT. Hydromorphone-3-
glucuronide: a more potent neuro-excitant than its structural
analogue, morphine-3-glucuronide. Life Sci 2001 Jun; 69
(4): 409-20
117. Smith MT. Neuroexcitatory effects of morphine and hydromor-
phone: evidence implicating the 3-glucuronide metabolites.
Clin Exp Pharmacol Physiol 2000 Jul; 27 (7): 524-8
118. Lawlor P, Turner K, Hanson J, et al. Dose ratio between mor-
phine and hydromorphone in patients with cancer pain: a
retrospective study. Pain 1997 Aug; 72 (1-2): 79-85
119. Katcher J, Walsh D. Opioid-induced itching: morphine sulfate
and hydromorphone hydrochloride. J Pain Symptom Manage
1999 Jan; 17 (1): 70-2
120. Abreu ME, Bigelow GE, Fleisher L, et al. Effect of intravenous
injection speed on responses to cocaine and hydromorphone
in humans. Psychopharmacologia 2001 Feb; 154 (1): 76-84
Adis International Limited. All rights reserved.
Davis & Srivastava
121. Miller MG, McCarthy N, OBoyle CA. Continuous subcutane-
ous infusion of morphine vs hydromorphone: a controlled
trial. J Pain Symptom Manage 1999 Jul; 18 (1): 9-16
122. Bruera E, Sloan P, Mount B, et al. A randomized, double-blind,
double-dummy, crossover trial comparing the safety and effi-
cacy of oral sustained-release hydromorphone with immediate-
release hydromorphone in patients with cancer pain. Canadian
Palliative Care Clinical Trials Group. J Clin Oncol 1996 May;
14 (5): 1713-7
123. Davies G, Kingswood C, Street M. Pharmacokinetics of opioids
in renal dysfunction. Clin Pharmacokinet 1996 Dec; 31 (6):
410-22
124. Tallgren M, Olkkola KT, Seppala T, et al. Pharmacokinetics and
ventilatory effects of oxycodone before and after liver trans-
plantation. Clin Pharmacol Ther 1997 Jun; 61 (6): 655-61
125. Davies G, Kingswood C, Street M. Pharmacokinetics of opioids
in renal dysfunction. Clin Pharmacokinet 1996 Dec; 31 (6):
410-22
126. Lohr JW, Willsky GR, Acara MA. Renal drug metabolism.
Pharmacol Rev 1998; 50 (1): 107-41
127. Koehntop DE, Rodman JH. Fentanyl pharmacokinetics in pa-
tients undergoing renal transplantation. Pharmacotherapy
1997; 17 (4): 746-52
128. Lee MA, Leng ME, Tiernan EJ. Retrospective study of the use
of hydromorphone in palliative care patients with normal and
abnormal urea and creatinine. Palliat Med 2001 Jan; 15 (1):
26-34
129. Babul N, Darke AC, Hagen N, et al. Hydromorphone metabolite
accumulation in renal failure. J Pain Symptom Manage 1995
Apr; 10 (3): 184-6
130. Maurer PM, Bartkowski RR. Drug interactions of clinical sig-
nificance with opioid analgesics. Drug Saf 1993; 8 (1): 30-48
131. Galer BS, Dworkin RH. A clinical guide to neuropathic pain.
New York: McGraw Hill Companies, 2000
132. Cox JM, Pappagallo M. Contemporary and emergent pharma-
cological therapies for chronic pain: nonopioid analgesia.
Expert Rev Neurotherapeutics 2001; 1 (1): 1-21
133. McQuay H, Carroll D, Jadad AR, et al. Anticonvulsant drugs
for management of pain: a systematic review. BMJ 1995 Oct;
311 (7012): 1047-52
134. Eckhardt K, Ammon S, Hofmann U, et al. Gabapentin enhances
the analgesic effect of morphine in health volunteers. Anesth
Analg 2000 Jul; 91 (1): 185-91
135. Dallocchio C, Buffa C, Mazzarello P, et al. Gabapentin vs ami-
triptyline in painful diabetic neuropathy: an open-label pilot
study. J Pain Symptom Manage 2000 Oct; 20 (4): 280-5
136. Cutrer FM, Moskowitz MA. Wolff Award 1996. The actions of
valproate and neurosteroids in a model of trigeminal pain.
Headache 1996 Nov-Dec; 36 (10): 579-85
137. Loscher W. Valproate: a reappraisal of it pharmacodynamic
properties and mechanisms of action. Prog Neurobiol 1999
May; 58 (1): 31-59
138. Hering R, Kuritzky A. Sodium valproate in the treatment of
cluster headache: an open clinical trial. Cephalalgia 1989 Sep;
9 (3): 195-8
139. Hardy JR, Rees EJA, Gwilliam B, et al. A phase II study to
establish the efficacy and toxicity of sodium valproate in pa-
tients with cancer-related neuropathic pain. J Pain Symptom
Manage 2001 Mar; 21 (3): 204-9
140. Tanaka E. Clinically significant pharmacokinetic drug interac-
tions between antiepileptic drugs. J Clin Pharm Ther 1999
Apr; 24 (2): 87-92
Drugs Aging 2003; 20 (1)
Demographics, Assessment and Management of Pain
141. Anderson GD. A mechanistic approach to antiepileptic drug
interactions. Ann Pharmacother 1998 May; 32 (5): 554-63
142. Neumann MG, Shear MH, Jacobson-Brown PM, et al.
CYP2E1-mediated modulation of valproic acid-induced
hepatocytotoxicity. Clin Biochem 2001 May; 34 (3): 211-8
143. Rogiers V, Akrawi M, Vercruysse A, et al. Effects of the anti-
convulsant, valproate, on the expression of components of the
cytochrome-P-450-mediated monooxygenase system and glu-
tathione S-transferases. Eur J Biochem 1995; 231 (2): 337-43
144. McQuay HJ, Carroll D, Glynn CJ. Low dose amitriptyline in
the treatment of chronic pain. Anesthesia 1992 Aug; 47 (8):
646-52
145. McQuay HJ, Carroll D, Glynn CJ. Dose-response for analgesic
effect of amitriptyline in chronic pain. Anaesthesia 1993 Apr;
48 (4): 281-5
Adis International Limited. All rights reserved.
57
146. Watson CPN. Antidepressant drugs in adjuvant analgesics. J
Pain Symptom Manage 1994 Aug; 9 (6): 392-405
147. Backonja MM. Local anesthetics as adjuvant analgesics. J Pain
Symptom Manage 1994 Nov; 9 (8): 491-9
148. Fromm GH. Baclofen as an adjuvant analgesic. J Pain Symptom
Manage 1994 Nov; 9 (8): 500-9
149. Marchettini P, Lacerenza M, Marangoni C, et al. Lidocaine test
in neuralgia. Pain 1992; 48: 277-82
Correspondence and offprints: Dr Mellar P. Davis, Harry R.
Horvitz Center for Palliative Medicine, Cleveland Clinic
Foundation, 9500 Euclid Avenue, Cleveland, OH 44195,
USA.
E-mail: davism6@ccf.org
Drugs Aging 2003; 20 (1)