Neuropathic Pain Pharmacotherapy

Elon Eisenberg

Simon Vulfsons

David M. Peterson

Neuropathic pain is pain caused by a lesion or disease affecting the somatosensory nervous system.¹^,² According to a recent systematic review of epidemiologic studies on neuropathic pain, the estimated prevalence of neuropathic pain is between 6.9% and 10% of the general population.³ Neuropathic pain may result from a large variety of insults, examples of which are listed in Table 81.1. Common examples of peripheral neuropathic pain include lumbar radiculopathy, painful diabetic neuropathy (PDN), postherpetic neuralgia (PHN), and posttraumatic/surgical neuropathy. Neuropathic pain of central origin includes central poststroke pain, pain due to multiple sclerosis, and post-spinal cord injury pain.

Neuropathic pain may be continuous or intermittent and is often described as burning or hot, electric shocks or shooting, pricking or pins, and needles. Pain evoked by nonpainful stimuli such as light touching or cold (allodynia) and accompanying nonpainful sensations such as numbness and tingling are all suggestive of neuropathic pain. The combination of several descriptors is a strong indicator of neuropathic pain. The neurologic examination is aimed primarily at detecting reproducible negative sensory signs, such as partial or complete loss to one or several sensory modalities (e.g., light touch, cold sensation) with distinct neuroanatomical borders. Positive sensory signs, especially when accompanying negative signs, are also supportive of neuropathic pain.¹

Neuropathic pain has negative effects on multiple domains of life resulting in poor quality of life, comparable to that experienced by patients suffering from cancer or chronic heart failure. This is explained, at least in part, by misdiagnosis, undertreatment, and lack of substantial analgesic efficacy of most existing treatments.⁴^,⁵

Neuropathic pain pharmacotherapy is a key therapeutic element, which involves different drug classes including, but not limited to, antidepressants, anticonvulsants, opioids, and topical agents. However, even with the newest of these drugs, effective pain relief occurs in less than half of patients with chronic neuropathic pain.¹ Meta-analyses of trials in neuropathic pain commonly report outcomes in terms of the number needed to treat (NNT) to provide 50% pain relief for one patient. Even highly recommended first-line agents result in NNTs ranging from 3.6 to 7.7, meaning that only 1 out 3.6 to 7.7 patients will experience a 50% reduction in pain.⁶ Most clinical trials related to neuropathic pain pharmacotherapy were conducted with a specific agent in patients who had a specific cause of the underlying disorder, with PDN or PHN being the most frequently tested disorder. At the same time, meta-analyses show that the efficacy of any given drug is generally not dependent on the cause of the underlying disorder.¹ In patients with refractory neuropathic pain, combination therapy with two or more agents possessing different mechanisms has been suggested.⁷

TABLE 81.1 Causes of Neuropathic Pain (Examples)

Trauma (surgery, frostbite, amputation)
Inflammation (Guillain-Barré syndrome)
Infection (AIDS neuropathy, acute zoster, postherpetic neuralgia)
Degenerative spine disease
Ischemic disorders
Metabolic disorders (diabetes mellitus)
Neoplastic disorders (tumor invasion, paraneoplastic)
Congenital disorders (Fabry’s disease)
Toxicity (chemotherapy)
Immunologic disorders

A different approach of selecting patients to clinical trials, not based on the specific causes of the underlying disorder, has recently been introduced. According to this approach, standardized quantitative sensory testing (QST) is used for subgrouping patients with peripheral neuropathic pain of different etiologies into different clusters. For example, patients with sensory loss, patients with thermal hyperalgesia and those with mechanical hyperalgesia, regardless of the underlying cause. It is proposed that these profiles may be related to different pathophysiologic mechanisms and may be useful in future clinical trial design.⁸

This chapter reviews pharmacotherapy for neuropathic pain and emphasizes the strengths and the limitations of different treatments. It also addresses unresolved issues related to pharmacologic treatment of neuropathic pain.

Antidepressants

Four classes of antidepressant medications have been studied in neuropathic pain treatment: tricyclic antidepressants (TCAs), selective serotonin and norepinephrine reuptake inhibitors (SNRIs), selective serotonin reuptake inhibitors (SSRIs), and monoamine oxidase inhibitors (MAOIs). Certain drugs in the first two classes are commonly considered first-line recommended treatments for neuropathic pain.⁶^,⁹^,¹⁰ Table 81.2 lists major properties of common antidepressants.

TRICYCLIC ANTIDEPRESSANTS

Table 81.3 summarizes the placebo-controlled trials of various drugs for neuropathic pain and shows efficacy for amitriptyline, imipramine, nortriptyline, desipramine, clomipramine, and maprotiline at daily doses ranging from 30 to 200 mg for PHN, PDN, postmastectomy pain, central poststroke pain, and mixed type neuropathies. These results are commonly extrapolated to all types of neuropathic pain, and clinical experience suggests some broad utility for the drugs. A recent systematic review and meta-analysis found TCAs the most efficacious antidepressants for neuropathic pain with an overall NNT of 3.6 (95% confidence interval [CI], 3.0 to 4.4) and the number needed to harm (NNH) was 13.4 (95% CI, 9.3 to 24.4). The level of evidence was considered moderate.⁶ A Cochrane collaboration review of the efficacy of amitriptyline for neuropathic pain found evidence for a statistically significant benefit with an NNT of 4.6 (95% CI, 1.8 to 3.1) and NNH of 4.1 (95% CI, 3.2 to 5.7),²⁹ and finally, another Cochrane collaboration review found the NNT of TCAs for neuropathic pain to be 3.6 (95% CI, 3.0 to 4.5).³⁰ However, a few trials failed to demonstrate TCA efficacy for spinal cord injury pain,¹³ cisplatin-induced neuropathy,³¹ HIV neuropathy,¹⁶^,¹⁷ phantom limb pain,¹⁵ lumbar radiculopathy,²⁰ and neuropathic cancer pain.³²
Inability to demonstrate effectiveness of the TCAs in these neuropathic pain types may represent a true lack of therapeutic benefit, or it may be due to study design weaknesses or other factors. More studies are needed to clarify the value of specific TCAs in different types of neuropathic pain.

TABLE 81.2 Comparison of Antidepressants in Neuropathic Pain¹¹^,¹²^,²³³^,²³⁴^,²³⁵^,²³⁶^,²³⁷^,²³⁸

Agent	Normal Adult Daily Dose in Psychiatric Disorders^a	Usual Number of Doses per Day	Elimination Half-life (t_1/2)	Dosing Adjustments	Serotonin: Norepinephrine Selectivity Ratio^b
*Selective Serotonin Receptor Inhibitors*
Citalopram	20-40 mg	1	35 h	20 mg/d in hepatic impairment or elderly	3,500-3,900
Escitalopram	10-20 mg	1	27-32 h	10 mg/d in hepatic impairment or elderly	7,100
Fluoxetine	20-80 mg	1	24-72 h (acute), 96-144 h (chronic), norfluoxetine 96-384 h	Decrease dose in hepatic impairment.	300-545
Fluvoxamine	100-300 mg	1-2	16 h	Decrease dose in hepatic impairment and elderly.	580-620
Paroxetine HCL	20-60 mg	IR: 1 XR: 1	21 h	Maximum dose = 40 mg/d in hepatic impairment or elderly	300-450
Paroxetine mesylate	20-60 mg	1	33 h	Maximum dose = 40 mg/d in renal or hepatic impairment or elderly	300-450
Sertraline	50-200 mg	1	62-104 h	Decrease dose in hepatic impairment and elderly.	1,400-2,750
*Serotonin and Norepinephrine Reuptake Inhibitors*
Desvenlafaxine	50-100 mg	1	11 h	Decrease dose in moderate or severe renal disease.	85
Duloxetine	40-120 mg	1-2	8-17 h	Do not give to patients with hepatic impairment. Decrease dose in severe renal disease.	9
Venlafaxine	75-375 mg	IR: 2-3 XR: 1	3-7 h, ODV 9-13 h	Decrease dose in hepatic or renal impairment.	115-120
*Tricyclic Antidepressants^c*
Amitriptyline	100-300 mg	1-4	9-27 h	Lower doses in elderly and hepatic impairment	8
Clomipramine	100-300 mg	1-3	15-60 h	Lower doses in elderly and hepatic impairment	130
Desipramine	75-300 mg	1-3	10-30 h	Lower doses in elderly and hepatic impairment	0.05
Imipramine	100-300 mg	1-4	5-30 h	Lower doses in elderly and hepatic impairment	27
Maprotiline	100-225 mg	1-3	25-50 h	Lower doses in elderly and hepatic impairment	0.002
Nortriptyline	50-150 mg	1-4	20-55 h	Lower doses in elderly and hepatic impairment	0.24
^aDosing is per product U.S. approved dosing; dosing in pain syndromes may vary from that listed. ^bThese are estimates of selectivity ratio; actual selectivity ratios are concentration-dependent; numbers <1 indicate greater affinity for norepinephrine than serotonin. ^cTricyclic antidepressants may be safely administered once daily; some practitioners prefer to dose them more frequently.
IR, immediate release; ODV, O-desmethylvenlafaxine; XR, extended release.

In many countries, TCAs may be overlooked as first-line neuropathic pain drugs because they are generic and are not actively marketed. Newer, more expensive drugs for neuropathic pain may be selected despite comparatively lower effectiveness (higher NNTs). The analgesic efficacy of the TCAs is likely to be independent of their antidepressant effect.¹⁴ In a large survey, over one-fifth (21%) of the patients with chronic pain have been diagnosed with depression associated with their pain.³³ Concomitant chronic pain and depression favor the use of an antidepressant (TCA or another) over other medication classes.

TCAs are associated with dose-dependent adverse events, the most common being sedation, constipation, dry mouth, urinary retention, and orthostatic hypotension. TCAs can be administered once daily, usually at bedtime, exploiting their sedating properties. The dose used for neuropathic pain, typically 75 mg a day or less, is below the antidepressant dose for most patients, lowering the risk of side effects. Secondary amine TCAs (e.g., desipramine, nortriptyline) are better tolerated than tertiary amine TCAs (e.g., amitriptyline, imipramine) but are equally effective.²⁴^,²⁸ Nonetheless, TCAs may not be tolerated by many patients. The anticholinergic effects of TCAs are a relative contraindication in patients with benign prostatic hyperplasia, cardiac conduction defects, and other morbidities sensitive to parasympatholytic action. Desipramine has only one-quarter of the anticholinergic and sedative activity of amitriptyline at comparable doses, making desipramine a TCA of choice.³⁴ Use of a low initial dose and slow dose titration of all TCAs is important to minimize premature discontinuation due to side effects. Dry mouth is the most common adverse event, and it is often best managed with sugarless candies or chewing gum, especially sugarless lemon
candy to stimulate serous saliva flow. Sips of water provide only transient relief. The most common starting dose for TCAs is 25 mg, but frail elderly and other highly sensitive patients may tolerate an initial dose of 5 to 10 mg better. Increase the dose by the same number of milligram as the starting dose every 3 to 5 days until some diminution of pain complaints occurs or the daily dose totals 100 mg. The maximal effect often occurs within 3 weeks at that dose, generally before antidepressant effects peak.

TABLE 81.3 Summary of Randomized Controlled Trials of Antidepressants in Treatment of Neuropathic Pain Published as Peer-Reviewed Articles

Study	Drug	Diagnosis	Design	Number of Patients Treated with Active Drug	Maximal Dosage (mg) per Day	Treatment Duration (wk)	Results
Leijon and Boivie²³⁹	Amitriptyline	CPSP	Crossover	15	75	4	A > P
Cardenas et al.¹³	Amitriptyline	SCI	Parallel	44	125	6	A = P
Max et al.¹⁴	Amitriptyline	PDN	Crossover	29	150	6	A > P
Vrethem et al.²⁴⁰	Amitriptyline	PDN and other polyneuropathies	Crossover	33	75	4	A > P
Watson et al.²⁴¹	Amitriptyline	PHN	Crossover	24	137.5	3	A > P
Max et al.²⁴²	Amitriptyline	PDN	Crossover	34	150	6	A > P
Robinson et al.¹⁵	Amitriptyline	Phantom limb	Parallel	39	125	6	A = P
Kalso et al.²⁴³	Amitriptyline	Postmastectomy	Crossover	15	100	4	A > P
Kieburtz et al.¹⁶	Amitriptyline	HIV neuropathy	Parallel	46	100	10	A = P
Shlay et al.¹⁷	Amitriptyline	HIV neuropathy	Parallel	58	75	14	A = P
Kvinesdal et al.²⁴⁴	Imipramine	PDN	Crossover	12	100	5	A > P
Sindrup et al.²⁴⁵	Imipramine	PDN	Crossover	18	150	2	A > P
Sindrup et al.¹⁸	Imipramine	PDN	Crossover	29	150	4	A > P
Gomez-Perez et al.¹⁹	Nortriptyline	PDN	Crossover	18	60	4	A > P
Raja et al.²⁴⁶	Nortriptyline	PHN	Crossover	46	140	8	A > P
Panerai et al.²⁴⁷	Nortriptyline	Mixed neuropathies	Crossover	24	100	3	A > P
Khoromi et al.²⁰	Nortriptyline	Radiculopathy	Crossover	34	100	9	A = P
Sindrup et al.²⁴⁸	Clomipramine	PDN	Crossover	19	75	2	A > P
Panerai et al.²⁴⁷	Clomipramine	Mixed neuropathies	Crossover	24	100	3	A > P
Sindrup et al.²⁴⁸	Desipramine	PDN	Crossover	19	200	2	A > P
Max et al.²⁴⁹	Desipramine	PDN	Crossover	20	250	6	A > P
Kishore-Kumar et al.²⁵⁰	Desipramine	PHN	Crossover	19	250	6	A > P
Raja et al.²⁴⁶	Desipramine	PHN	Crossover	13	160	8	A > P
Vrethem et al.²⁴⁰	Maprotiline	PDN and other polyneuropathies	Crossover	33	75	4	A > P
Goldstein et al.²¹	Duloxetine	PDN	Parallel	342	120	12	A > P
Raskin et al.²²	Duloxetine	PDN	Parallel	232	120	12	A > P
Wernicke et al.²³	Duloxetine	PDN	Parallel	226	120	12	A > P
Sindrup et al.¹⁸	Venlafaxine	Mixed neuropathies	Crossover	30	225	4	A > P
Rowbotham et al.²⁴	Venlafaxine	PDN	Crossover	163	225	6	A > P
Tasmuth et al.²⁵	Venlafaxine	Postmastectomy	Crossover	13	75	4	A = P
Yucel et al.²⁵¹	Venlafaxine	Mixed neuropathies	Parallel	8	150	8	A = P^a
Sindrup et al.²⁶	Paroxetine	PDN	Crossover	20	40	2	A = P
Sindrup et al.²⁷	Citalopram	PDN	Crossover	20	40	3	A > P
Max et al.²⁸	Fluoxetine	PDN	Crossover	46	40	6	A > P
^aThe study showed significant effect of venlafaxine in the manifestations of hyperalgesia and temporal summation but not on the ongoing pain intensity.
CPSP, central poststroke pain; PDN, painful diabetic neuropathy; PHN, postherpetic neuralgia; SCI, spinal cord injury; A, active drug; P, placebo; > indicates that active drug was superior to the comparator in terms of pain reduction; < indicates that active drug was not superior to the comparator in terms of pain reduction; = indicates that active drug was equal to the comparator in terms of pain reduction.

The anticholinergic effects of TCAs can cause cardiac toxicity including ventricular ectopic activity, prolonged QT interval, myocardial infarction, and sudden death.³⁵ A screening electrocardiogram (ECG) might be considered in patients over 40 years of age or who have other risk factors prior to initiating a TCA. Use TCAs with caution in patients with a history of ischemic heart disease or increased risk of sudden cardiac death; if TCA treatment is selected in these patients, consider limiting the maximum dose to 100 mg per day or less.⁶^,¹¹^,³⁶

SELECTIVE SEROTONIN AND NOREPINEPHRINE REUPTAKE INHIBITORS

The SNRIs duloxetine and venlafaxine are newer antidepressants with effectiveness for neuropathic pain. Results with venlafaxine for neuropathic pain have been inconsistent. Venlafaxine was effective for PDN and for various forms of polyneuropathy at daily doses of 150 to 225 mg.¹⁸^,²⁴ In one placebo-controlled trial, perioperative venlafaxine at a daily dose of 75 mg prevented the development of postmastectomy pain syndrome,³⁷ but in other trials on patients with postmastectomy pain²⁵ syndrome and mixed type neuropathies, efficacy at a dose range of 75 to 150 mg could not be demonstrated.

Three large randomized controlled trials²¹^,²²^,²³ and one open-label 52-week extension trial³⁸ showed that duloxetine was effective for PDN at daily doses of 60 to 120 mg. Duloxetine 60 mg per day has similar efficacy to duloxetine 120 mg per day, but the lower dose is far more tolerable. It can be administered once daily. At that dose, the drug significantly improved sleep and quality of life. The most common adverse effects reported
in these clinical trials were nausea, somnolence, dizziness, and constipation, all of which tended to decrease over time. Initiating treatment at a daily dose of 30 mg for 1 week followed by an increase to 60 mg per day during the second week is likely to improve tolerability.³⁹ Duloxetine has not been associated with cardiac toxicity. It should not be used concomitantly with MAOIs or in patients with markedly impaired liver function.

A meta-analysis of trials for the SNRIs (i.e., duloxetine, venlafaxine, desvenlafaxine) across neuropathic pain types resulted in an NNT of 6.4 (95% CI, 5.2 to 8.4), indicating that they are less effective than the far less expensive TCAs.⁶ In the systematic review and meta-analysis cited earlier, the final quality of evidence for SNRI efficacy in the treatment of neuropathic pain was deemed high. Combined NNT was 6.4 (95% CI, 5.2 to 8.4), and NNH was 11.8 (95% CI, 9.5 to 15.2).⁶

SELECTIVE SEROTONIN REUPTAKE INHIBITORS

Reports of the effectiveness of SSRIs in neuropathic pain management are generally not favorable. Some reports may not have separated analgesic effect from the effects of mood elevation on pain perception. One well-controlled trial documented that fluoxetine was no more effective than placebo, whereas both amitriptyline and desipramine were efficacious.²⁸ Two other trials support efficacy of citalopram²⁷ and paroxetine²⁶ (both at 40 mg per day) in PDN. Although SSRIs show a favorable safety profile compared to TCAs and are generally well tolerated, a meta-analysis of available trials in neuropathic pain did not detect significant overall pain reduction with the SSRIs, and therefore, they should not be regarded as first-line agents.⁶

Antiepileptics

Interest in the antiepileptics for pain management dates back to the 1940s.⁴⁰ Dozens of randomized controlled trials have attempted to describe the role of different antiepileptic drugs in neuropathic pain treatment. The majority of antiepileptic trials are in PHN or PDN, with less data available for other neuropathic pain types. Recent meta-analyses strongly support the use of the calcium channel α₂δ ligands pregabalin and gabapentin (including gabapentin ER and gabapentin enacarbil) as first-line therapy for PHN and PDN.⁶^,⁴¹ Data also support the use of pregabalin for treating central neuropathic pain.⁴¹ Carbamazepine is likely effective for trigeminal neuralgia and possibly for PDN and central poststroke pain, but studies of carbamazepine were generally of short duration (i.e., less than 4 weeks) and poor methodologic quality.⁴² Although other antiepileptics may be beneficial in individual patients or specific types of neuropathic pain, pooled data either suggested minimal effectiveness or were inconclusive.⁶^,⁴¹ Table 81.4 summarizes trials of the antiepileptics in neuropathic pain.

PREGABALIN

Pregabalin is the most extensively studied drug of any type for treating neuropathic pain.⁶ Pregabalin is believed to exert its analgesic effect by binding to the α₂δ subunit of voltage-gated calcium channels on primary afferent neurons, reducing the release of neurotransmitters from their central terminals.¹³² A recent meta-analysis of 25 placebo-controlled trials of pregabalin in various types of neuropathic pain reported an NNT of 7.7 (95% CI, 6.5 to 9.4).⁶ Pregabalin was effective for treating PHN and PDN in numerous multicenter, randomized, controlled trials.⁷⁴^,⁹⁷^,¹⁰¹^,¹⁰²^,¹⁰³^,¹⁰⁴^,¹⁰⁵^,¹⁰⁶^,¹⁰⁷^,¹⁰⁸^,¹⁰⁹^,¹¹⁰^,¹¹¹^,¹¹²^,¹¹⁶^,¹³³^,¹³⁴ Pregabalin was also effective for treating spinal cord injury pain,¹¹⁴^,¹¹⁵ posttraumatic peripheral neuropathy,¹¹³ neuropathic cancer pain,⁶¹^,¹⁰⁰ and mixed types of neuropathic pain.¹¹⁶^,¹¹⁷^,¹¹⁹^,¹²⁰^,¹²¹^,¹²² Pregabalin did not significantly improve acute herpetic neuralgia or prevent progression to PHN in one clinical trial,⁹⁶ but it did significantly reduce acute herpetic neuralgia and subacute herpetic neuralgia in another trial.¹¹⁸ Pregabalin did not significantly improve neuropathic pain associated with HIV⁹⁸^,⁹⁹ or central poststroke pain.⁹⁷ The effective daily dose of 300 to 600 mg reduces pain and improves sleep, functioning, and quality of life. Response rates are higher when a maintenance dose of 600 mg per day is used.⁶ Pregabalin has several advantages compared to other anticonvulsants: It is usually administered twice daily, can be rapidly titrated, has early onset of analgesic effect, and has linear pharmacokinetics. No common drug interactions occur with pregabalin. The most commonly reported adverse events were dose-dependent and included dizziness, somnolence, dry mouth, abnormal vision, confusion, weight gain, and edema.¹¹

GABAPENTIN

There is also strong evidence of gabapentin’s effectiveness for PHN and PDN at doses of 900 to 3600 mg per day.⁷^,⁵⁴^,⁵⁵^,⁵⁶^,⁶²^,⁶³^,⁶⁴^,⁷⁰^,⁷¹^,⁷²^,⁷³^,⁷⁴^,⁷⁵ Gabapentin is available as immediate-release (IR), extendedrelease (ER), and gabapentin enacarbil formulations, and available data suggest similar effectiveness for the different formulations.⁶ Gabapentin IR is normally dosed thrice daily and is commonly the least expensive dosage form. Gabapentin enacarbil is dosed twice daily, and gabapentin ER may be dosed once daily. Across 14 studies, for all gabapentin dosage forms, the overall NNT to reduce pain intensity by 50% was 7.2 (95% CI, 5.9 to 9.1). For gabapentin IR, the NNT was 6.3 (95% CI, 5.0 to 8.3). The NNT for gabapentin enacarbil and gabapentin ER was 8.3 (95% CI, 6.2 to 13.0).⁶ In addition to PHN and PDN, gabapentin has demonstrated efficacy in HIV-associated painful neuropathy, pain in Guillain-Barré syndrome, and phantom limb pain.⁵⁷^,⁵⁸^,⁵⁹^,⁶⁵ Gabapentin was not effective for peripheral nerve injury pain in one trial.⁶⁶ An 8-day trial of gabapentin was effective for treating neuropathic cancer pain in one trial,⁶⁰ but in a smaller 4-week trial, gabapentin was no more effective than placebo.⁶¹ Results for gabapentin in spinal cord injury were mixed in three small trials.⁶⁷^,⁶⁸^,⁶⁹ Gabapentin is believed to have similar mechanism of action to pregabalin. However, unlike pregabalin, it has nonlinear pharmacokinetics and may take days or weeks to reach an effective dose. Gabapentin is relatively safe, with few clinically relevant drug interactions. The main adverse effects are somnolence, dizziness, and peripheral edema.¹¹ Some clinicians report that some patients who fail to respond to gabapentin may respond to pregabalin and vice versa.

CARBAMAZEPINE

Carbamazepine is commonly used to treat trigeminal neuralgia. Most carbamazepine studies in trigeminal neuralgia were conducted in the 1960s and 1970s using small samples sizes. In these studies, pain relief was superior with carbamazepine compared to placebo.⁴⁵^,⁴⁶^,⁴⁷^,¹³⁵ In a more recent trial, carbamazepine was more effective than lamotrigine for treating pain associated with trigeminal neuralgia.⁴⁸ Carbamazepine was also one of the first anticonvulsants used in PDN and was superior to placebo in two small trials.⁴³^,⁴⁴ Pain reduction was similar between carbamazepine and the TCA nortriptyline in a more recent trial.¹⁹ Carbamazepine was more effective than placebo in patients with mixed types of neuropathic pain in one controlled trial.⁴⁹ Effective doses of carbamazepine ranged from 800 to 2,400 mg per day in trigeminal neuralgia, but lower doses of 200 to 600 mg per day were effective for treating other neuropathic pain types. NNT to achieve 50% pain relief with carbamazepine in a meta-analysis of trials in trigeminal neuralgia, PDN, and central poststroke pain was 1.9 (95% CI, 1.6 to 2.5).⁴² However, this figure is based primarily on old trials, most of which were conducted in small patient groups for relatively short treatment periods. All of the carbamazepine trials were classified as third tier trials, and heterogeneity associated with pooling the trials was moderately high (I² = 50%).⁴² The analgesic mechanism of carbamazepine is related to voltage-dependent sodium

channel blocking, which results in decreased ectopic nerve discharges and neural membrane stabilization.¹³⁶ Adverse events are common and include dizziness, nausea, drowsiness, blurred vision, and ataxia.⁴⁴ Carbamazepine can cause Stevens-Johnson syndrome (SJS), toxic epidermal necrolysis (TEN), and drug reaction with eosinophilia and systemic symptoms (DRESS). Rare, but serious, side effects include blood dyscrasias, impairment of liver function, and reduction of sodium plasma levels and require routine blood tests to monitor them. Serum carbamazepine concentration monitoring is recommended to maximize efficacy, monitor compliance, and reduce toxicity.¹¹

TABLE 81.4 Summary of the Randomized Controlled Trials of Antiepileptic Drugs in Treatment of Neuropathic Pain Published as Peer-Reviewed Articles

Study	Drug	Diagnosis	Design	Number of Patients Treated with Active Drug	Maximal Dosage (mg) per Day	Treatment Duration (wk)	Results
Rull et al.⁴³	Carbamazepine	PDN	Crossover	30	400	2	A > P
Wilton⁴⁴	Carbamazepine	PDN	Crossover	40	400	2	A > P
Gomez-Perez et al.¹⁹	Carbamazepine	PDN	Crossover	16	200	4	A = nortriptyline
Nicol⁴⁵	Carbamazepine	TN	Crossover	20	2,400	2	A > P
Campbell et al.⁴⁶	Carbamazepine	TN	Crossover	77	800	2	A > P
Killian and Fromm⁴⁷	Carbamazepine	TN	Crossover	27	1,000	5 d	A > P
Vilming²⁵²	Carbamazepine	TN	Parallel	6	900	3	A > tizanidine
Lechin et al.²⁵³	Carbamazepine	TN	Crossover	48	1,200	8	A < pimozide
Lindstrom²⁵⁴	Carbamazepine	TN	Crossover	12	Maximal tolerated	2	A = tocainide
Shaikh et al.⁴⁸	Carbamazepine	TN	Crossover	21	1,200	6	A > lamotrigine^a
Harke et al.⁴⁹	Carbamazepine	Mixed types	Parallel	43	600	8 d	A > P
Dogra et al.⁵⁰	Oxcarbazepine	PDN	Parallel	69	1,800	16	A > P
Grosskopf et al.⁵¹	Oxcarbazepine	PDN	Parallel	71	1,200	16	A = P
Beydoun et al.⁵²	Oxcarbazepine	PDN	Parallel	258	1,800	16	A = P
Demant et al.⁵³	Oxcarbazepine	Mixed types	Crossover	83	2,400	6	A > P
Backonja et al.⁵⁴	Gabapentin	PDN	Parallel	84	3,600	8	A > P
Morello et al.²⁵⁵	Gabapentin	PDN	Crossover	26	1,800	6	A = amitriptyline
Dallocchio et al.⁵⁵	Gabapentin	PDN	Parallel	13	2,400	12	A > amitriptyline
Simpson et al.⁵⁶	Gabapentin	PDN	Parallel	30	3,600	8	A > P
Pandey et al.⁵⁷	Gabapentin	GBS	Crossover	18	15 mg/kg	1	A > P
Pandey et al.⁵⁸	Gabapentin	GBS	Parallel	12	900	1	A > P
Hahn et al.⁵⁹	Gabapentin	HIV-N	Parallel	15	2,400	4	A > P
Caraceni et al.⁶⁰	Gabapentin	NCP	Parallel	79	1,800	8 d	A > P
Mishra et al.⁶¹	Gabapentin	NCP	Parallel	30	1,800	4	A = P A = amitriptyline A < pregabalin^a
Rice and Maton⁶²	Gabapentin	PHN	Parallel	223	2,400	7	A > P
Rowbotham et al.⁶³	Gabapentin	PHN	Parallel	113	3,600	8	A > P
Chandra et al.⁶⁴	Gabapentin	PHN	Parallel	34	2,700	8	A = nortriptyline
Bone et al.⁶⁵	Gabapentin	PLP	Crossover	19	2,400	6	A > P
Gordh et al.⁶⁶	Gabapentin	PNI	Crossover	98	2,400	7	A = P
Levendoglu et al.⁶⁷	Gabapentin	SCI	Crossover	20	3,600	8	A > P
Tai et al.⁶⁸	Gabapentin	SCI	Crossover	7	1,800	4	A > P^b
Rintala et al.⁶⁹	Gabapentin	SCI	Crossover	32	3,600	8	A < amitriptyline A = diphenhydramine (active control)
Gilron et al.⁷	Gabapentin	PDN + PHN	Crossover	57	3,200	5	A > P; A = morphine A < A + morphine
Gilron et al.⁷⁰	Gabapentin	PDN + PHN	Crossover	46	3,600	6	A < A + nortriptyline A = nortriptyline
Smith et al.²⁵⁶	Gabapentin	PLP + RLP	Crossover	24	3,600	6	A = P
Serpell²⁵⁷	Gabapentin	Mixed types	Parallel	153	2,400	8	A > P
Sandercock et al.²⁵⁸	Gabapentin ER	PDN	Parallel	96	3,000	4	A > P
Wallace et al.⁷¹	Gabapentin ER	PHN	Parallel	269	1,800	10	A = P
Sang et al.⁷²	Gabapentin ER	PHN	Parallel	221	1,800	10	A > P
Irving et al.⁷³	Gabapentin ER	PHN	Parallel	107	1,800	4	A > P
Jensen et al.²⁵⁹	Gabapentin ER	PHN	Parallel	102	1,800	4	A = P^c
Rauck et al.⁷⁴	Gabapentin Enacarbil	PDN	Parallel	235	3,600	13	A = P A = pregabalin^a
Backonja et al.²⁶⁰	Gabapentin Enacarbil	PHN	Parallel	47	1,200	2	A > P
Zhang et al.⁷⁵	Gabapentin Enacarbil	PHN	Parallel	263	3,600	14	A > P
Rauck et al.⁷⁶	Lacosamide	CPSP	Parallel	60	400	6	A > P
Shaibani et al.⁷⁷	Lacosamide	PDN	Parallel	403	600	18	A = P
Wymer et al.⁷⁸	Lacosamide	PDN	Parallel	277	600	18	A = P
Ziegler et al.⁷⁹	Lacosamide	PDN	Parallel	281	600	18	A = P
Breuer et al.⁸⁰	Lamotrigine	CPMS	Crossover	17	400	11	A = P
Vestergaard et al.⁸¹	Lamotrigine	CPSP	Crossover	30	200	8	A > P
Vinik et al.⁸²	Lamotrigine	PDN	Parallel	360	400	10	A = P
Vinik et al.⁸²	Lamotrigine	PDN	Parallel	360	400	10	A = P
Eisenberg et al.⁸³	Lamotrigine	PDN	Parallel	29	400	8	A > P
Simpson et al.⁸⁴	Lamotrigine	HIV-N	Parallel	20	300	14	A > P
Simpson et al.⁸⁵	Lamotrigine	HIV-N	Parallel	150	600	11	A = P; A > P^d
Finnerup et al.⁸⁶	Lamotrigine	SCI	Crossover	22	400	8	A = P; A > P^e
Zakrzewska et al.⁸⁷	Lamotrigine	TN	Crossover	14	400	2	A > P
Shaikh et al.⁴⁸	Lamotrigine	TN	Crossover	21	400	6	A < carbamazepine^a
McCleane⁸⁸	Lamotrigine	Mixed types	Parallel	50	200	8	A = P
McCleane⁸⁹	Lamotrigine	Mixed types	Parallel	36	200	8	A = P
Silver et al.⁹⁰	Lamotrigine	Mixed types	Parallel	111	400	14	A = P
Jungehulsing et al.²⁶¹	Levetiracetam	CPSP	Crossover	42	3,000	8	A = P
Rossi et al.⁹¹	Levetiracetam	CPMS	Parallel	12	3,000	12	A > P
Falah et al.⁹²	Levetiracetam	CPMS	Crossover	30	3,000	6	A = P
Vilholm et al.⁹³	Levetiracetam	Postmastectomy	Crossover	26	3,000	4	A = P
Finnerup et al.⁹⁴	Levetiracetam	SCI	Crossover	34	3,000	5	A = P
Holbech et al.⁹⁵	Levetiracetam	Mixed types	Crossover	35	3,000	6	A = P
Krcevski and Kamenik⁹⁶	Pregabalin	AHN	Parallel	14	300	3	A = P^f
Kim et al.⁹⁷	Pregabalin	CPSP	Parallel	110	600	13	A = P, A > P^g
Simpson et al.⁹⁸	Pregabalin	HIV-N	Parallel	151	600	14	A = P
Simpson et al.⁹⁹	Pregabalin	HIV-N	Parallel	183	600	17	A = P
Mishra et al.⁶¹	Pregabalin	NCP	Parallel	30	600	4	A > P A > gabapentin^a A > amitriptyline
Raptis et al.¹⁰⁰	Pregabalin	NCP	Parallel	60	600	4	A > fentanyl transdermal
Richter et al.¹⁰¹	Pregabalin	PDN	Parallel	161	600	6	A > P
Lesser et al.¹⁰²	Pregabalin	PDN	Parallel	240	600	5	A > P
Rosenstock et al.¹⁰³	Pregabalin	PDN	Parallel	76	300	8	A > P
Tolle et al.¹⁰⁴	Pregabalin	PDN	Parallel	299	600	12	A > P
Arezzo et al.¹⁰⁵	Pregabalin	PDN	Parallel	82	600	12	A > P
Satoh et al.¹⁰⁶	Pregabalin	PDN	Parallel	179	600	13	A > P
Rauck et al.⁷⁴	Pregabalin	PDN	Parallel	56	300	13	A = P A = gabapentin enacarbil^a
Irving et al.²⁶² and Tanenberg et al.²⁶³	Pregabalin	PDN	Parallel	134	300	12	A = duloxetine A = duloxetine + gabapentin
Smith et al.¹⁰⁷	Pregabalin	PDN	Parallel	99	300	15	A = P A = carisbamate
Raskin et al.¹⁰⁸	Pregabalin	PDN	Crossover	301	300	6	A = P
Sabatowski et al.¹⁰⁹	Pregabalin	PHN	Parallel	157	300	8	A > P
Dworkin et al.¹¹⁰	Pregabalin	PHN	Parallel	89	600	8	A > P
van Seventer et al.¹¹¹	Pregabalin	PHN	Parallel	273	600	13	A > P
Stacey et al.¹¹²	Pregabalin	PHN	Parallel	179	600	4	A > P
Achar et al.²⁶⁴	Pregabalin	PHN	Parallel	25	150	8	A > amitriptyline
van Seventer et al.¹¹³	Pregabalin	PNI	Parallel	127	600	8	A > P
Siddall et al.¹¹⁴	Pregabalin	SCI	Parallel	70	600	12	A > P
Cardenas et al.¹¹⁵	Pregabalin	SCI	Parallel	108	600	16	A > P
Freynhagen et al.¹¹⁶	Pregabalin	PDN + PHN	Parallel	273	600	12	A > P
Vranken et al.¹¹⁷	Pregabalin	CPSP + SCI	Parallel	20	600	4	A > P
Liang et al.¹¹⁸	Pregabalin	AHN + SHN	Parallel	150	600	4	A > P
Guan et al.¹¹⁹	Pregabalin	PDN + PHN	Parallel	206	600	8	A > P
Moon et al.¹²⁰	Pregabalin	Mixed types	Parallel	162	600	9	A > P
Holbech et al.²⁶⁵	Pregabalin	Mixed types	Crossover	61	300	5	A > P A < imipramine A < imipramine + pregabalin
Haanpaa et al.¹²¹	Pregabalin	Mixed types	Parallel	277	600	8	A = capsaicin 8% patch
Gatti et al.²⁶⁶	Pregabalin	Mixed types	Parallel	134	290^h	13	A < oxycodone CR A < A + oxycodone CR
Gilron et al.¹²²	Pregabalin	Mixed types	Parallel	80	600	4-9	A > P
Kochar et al.¹²³	Sodium valproate	PDN	Parallel	29	1,200	4	A > P
Kochar et al.¹²⁴	Sodium valproate	PDN	Parallel	21	500	12	A > P
Kochar et al.¹²⁵	Valproic acid + sodium valproate	PHN	Parallel	22	1,000	8	A > P
Drewes et al.¹²⁶	Sodium valproate	SCI	Parallel	20	2,400	3	A = P
Otto et al.¹²⁷	Valproic acid	Mixed types	Crossover	31	1,500	4	A = P
Thienel et al.¹²⁸	Topiramate	PDN	Parallel	878	400	22	A = Pⁱ
Raskin et al.¹²⁹	Topiramate	PDN	Parallel	214	400	12	A > P
Khoromi et al.¹³⁰	Topiramate	LR	Crossover	42	400	8	A = P; A > P^j
Atli and Dogra¹³¹	Zonisamide	PDN	Parallel	13	600	12	A = P
^aTrial compared two antiepileptics to each other, and the same trial is displayed twice in this table (once for each antiepileptic). ^bA significant decrease of “unpleasant feeling” and a trend toward a decrease in both the “pain intensity” and “burning sensation.” ^cSignificant difference in “itchy pain sensations” but no difference in global pain intensity or other measures of pain. ^dLamotrigine was superior to placebo in patients who received antiviral neurotoxic therapy but not in patients who did not receive this therapy. ^eLamotrigine was superior to placebo in patients with incomplete spinal cord injury and evoked pain but equal to placebo in patients with complete injury and without evoked pain. ^fPregabalin was not effective for preventing subacute herpetic neuralgia or postherpetic neuralgia. ^gNo significant difference between pregabalin and placebo in pain score (primary outcome measure) but significant difference in sleep, anxiety, and clinical global impression scores. ^hMean dose reported rather than maximum. ⁱFindings from three double-blind, placebo-controlled trials. ^jNo significant difference between topiramate and placebo in pain score (primary outcome measure) but significant difference in global pain relief score.
AHN, acute herpetic neuralgia; CPMS, central pain due to multiple sclerosis; CPSP, central poststroke pain; ER, extended release; GBS, Guillain-Barré syndrome; HIV-N, HIV neuropathy; LR, lumbar radiculopathy; NCP, neuropathic cancer pain; PDN, painful diabetic neuropathy; PHN, postherpetic neuralgia; PLP, phantom limb pain; PNI, peripheral nerve injury; RLP, residual limb pain; SCI, spinal cord injury; SHN, subacute herpetic neuralgia; TN, trigeminal neuralgia; A, active drug; P, placebo; > indicates that active drug was superior to the comparator in terms of pain reduction; < indicates that active drug was not superior to the comparator in terms of pain reduction; = indicates that active drug was equal to the comparator in terms of pain reduction.