Antiepileptic Drugs, 5th Edition



Clinical Efficacy and Use in Other Neurological Disorders

Bernd Schmidt MD, PhD

Head, Department of Neurology and Psychiatric Clinic, Wittnau, Germany

Gabapentin (GBP) is the first of the newer antiepileptic drugs (AEDs) that expanded its use, shortly after it was labeled and marketed for partial seizures with or without secondary generalization, into a broad range of neurologic and psychiatric indications. Now, most prescriptions for GBP worldwide are for conditions other than epilepsy. Following the lead of GBP, for most of the newer AEDs, marketing approval for additional indications, particularly for pain syndromes, bipolar disorder, and migraine, is sought early after the drug is first launched as an AED, or development programs for those other indications are pursued at the same time. In the future, it may be that the development of a drug as an AED may become only a sideline of other neuropsychiatric indications targeted first, because of the much larger market share and attractiveness of other central nervous system indications, compared with epilepsy, provided the mechanism of action of novel molecules suggests true broad-spectrum central nervous system activity.

On October 23, 2000, gabapentin (Neurontin) was approved by the German regulatory authorities for “neuropathic pain in adults,” particularly diabetic polyneuropathy and postherpetic neuralgia, following the lead of Austria, Switzerland, the United Kingdom, Ireland, Italy, Brazil, Chile, Columbia, Ecuador, Peru, Venezuela, Mexico, the Philippines, and Singapore. The drug's package insert recommends the following GBP dosage regimen: initial dose of three times 300 mg from day 1, or a titration to 900 mg/day within 3 days, followed by an increase to 1,800 mg/day within a week, not too exceed a maximum daily dose of 3,600 mg.


For neurologic disorders other than epilepsy, GBP's efficacy has been documented best by controlled randomized clinical trials of symptomatic treatment for neuropathic pain syndromes associated with diabetes and for postherpetic neuralgia. Based on these data, marketing approval for the indication of neuropathic pain was granted in some countries. In a double-blind trial, 165 patients with chronic painful diabetic neuropathies (diabetes type 1 and 2) were randomly assigned to receive ≤3,600 mg/day GBP or placebo for 8 weeks. At the 8-week study end point, for the intent-to-treat population, there was a significantly different (p = .001) mean daily pain score, used as a primary outcome variable, in favor of GBP over placebo. Consistent with the primary measure were significant improvements on total pain, mean sleep interference, and visual analogue scale (VAS) and present pain intensity (PPI) scores of the Short-Form McGill Pain Questionnaire. All other comedications that could have influenced pain or pain perception were excluded; antidiabetic medication had to be kept constant, and during the trial, hemoglobin A1C values remained unchanged in both groups. An interesting feature of the trial results is that the mean curves for the GBP and placebo group were already significantly separate in the first week for the parameter sleep interference and in the second treatment week for all pain rating scales, a finding suggesting that the doses used during the initial 4-week titration period may be effective (900 to 1,800 mg/day).

Overall, 67% of the patients taking GBP made it through the forced titration ≤3,600 mg/day, and 6% withdrew because of adverse events. Dizziness (23%) and somnolence (22%) were the two adverse events occurring significantly more frequently than with placebo (1).

In a study with a very similar design, GBP was investigated in postherpetic neuralgia in a much older and comorbid patient population than the patients in the previously mentioned study, who had diabetic neuropathy. In the postherpetic neuralgia trial, 229 patients, whose mean age was 73 years and who had a median history of postherpetic pain for 28 years, were randomized to receive either ≤3,600 mg/day GBP or placebo for 8 weeks. Two-thirds


of these patients were not taking any concomitant pain medication, and one-third received tricyclic antidepressants, opioids, or both, in stable doses throughout the study. After the first 4 weeks of forced GBP titration, 83% of the patients reached doses of ≥2,400 mg/day, and 65% received stable doses of 3,600 mg/day. The average daily pain score, the primary efficacy parameter, for the GBP-treated group was significantly reduced (p = .001) compared with the placebo-treated group from week 2 until the end of study week 8, without signs of tolerance. Significant differences in favor of the GBP-treated group were also achieved for secondary parameters such as daily sleep rating scores, total pain, or the Short-Form McGill Pain Questionnaire. Dizziness led to withdrawal in 5% and somnolence in 4% of the patients. In addition, somnolence was the most common adverse event (27%), followed by dizziness (23%). Serious adverse events were not reported in this older group of patients with herpetic neuralgia (2). Neither placebo-controlled study has any available data on long-term results .

A third randomized controlled trial compared two active regimens for diabetic peripheral neuropathy—GBP, 900 to 1,800 mg/day, versus amitriptyline hydrochloride (AMI), 25 to 75 mg/day—in a double-blind 6-week crossover design, with an intermediate 1-week washout. Only 25 patients were enrolled, 19 of whom completed both arms of the crossover trial and were evaluated. Both drugs (mean achieved doses: GBP, 1,565 mg/day; AMI, 59 mg/day) provided significant pain relief from 2 weeks onward, compared with baseline pain diary scores. However, at the end of treatment, there was no statistical difference in pain intensity scores between GBP and AMI. The authors agree that, to avoid a type II error, approximately 260 patients per paired crossover period would have been needed to draw valid conclusions on comparative efficacy. In the GBP-treated group, sedation and dizziness were the most frequent adverse events, and, as expected, in the AMI group, dry mouth and weight gain were the most common (3).

In an attempt to compare efficacy between studies and therapeutic regimens for treatment of diabetic neuropathic pain and postherpetic pain, numbers needed to treat (NNTs) were calculated for patients achieving >50% pain relief. In painful diabetic neuropathy, the NNT for use of all types of tricyclic antidepressants was 3.0 (confidence interval or CI, 2.4 to 4.0) , for tramadol it was 3.4 (CI, 2.3 to 6.4), for phenytoin it was 2.1 (CI, 1.5 to 3.6), for carbamazepine it was 3.3 (CI, 2.0 to 9.2), and for GBP it was 3.7 (CI, 2.4 to 8.3). For postherpetic neuralgia, a comparison is available only between tricyclic antidepressants achieving an NNT of 2.5 (CI, 1.7 to 3.3) and GBP 3.2 (CI, 2.4 to 5.0). Those comparisons do not take tolerability or dropout rates from adverse events into consideration (4).

Further to these randomized controlled trials is a large body of literature about GBP use in a variety of other neuropathic pain syndromes caused by peripheral nerve trauma or compression, infections, metabolic disturbances, or neurodegeneration. Most of these reports are based on open case series with heterogeneous patient populations.

Trigeminal neuralgia was treated in 13 patients with increasing GBP doses of between 300 and 2,000 mg/day; half of these patients had been unresponsive to previous carbamazepine therapy. In the de novo treatment group, GBP was found to be effective in 83% of the patients; 57% in the carbamazepine-refractory group responded (5). Seven patients with multiple sclerosis and chronic, severe , refractory trigeminal neuralgia received GBP doses of between 900 and 2,400 mg/day. Three patients with multiple sclerosis were receiving concomitant baclofen. Six of the seven patients experienced complete pain relief, and no patient discontinued GBP because of adverse events. The onset of the drug's effect occurred after 3 to 4 days, with a maximum effect reached from week 2 onward that was maintained for at least 1 year (6). Neuropathic pain related to acquired immunodeficiency syndrome was treated with GBP in addition to antiretroviral medication, and GBP demonstrated partial efficacy in this condition as well (7).

Neuropathic cancer pain was investigated openly in 22 patients who receiving GBP, 800 to 1,200 mg/day, as an adjunct to stable opioid doses. Assessment of a treatment effect was made 7 to 14 days after stable doses of GBP had been achieved, and the result was a significant reduction of global pain, burning pain, and shooting pain scores. No control group was included in the study design. No new adverse events, aside from mild sedation attributable to GBP, was reported. The authors of this study discussed the option of reducing concomitant opioids, to ease the drug burden of patients with cancer and to improve the overall side effects (8).

A more in-depth evaluation of an effect of GBP on the different components of neuropathic pain was undertaken in 18 patients with postherpetic neuralgia, phantom limb pain, postthoracotomy pain, idiopathic peripheral neuropathy, poststroke central pain, and pain from spinal injury. In this study, GBP doses were titrated up to a maximum of 2,400 mg/day and were maintained for 6 weeks. GBP induced a statistically significant pre-post effect on spontaneous pain, but particularly on paroxysmal pain and evoked pain (brush-induced allodynia, cold-induced allodynia, and hyperalgesia) (9).


In 1984, even before the use of GBP was explored in epilepsy, the manufacturer initiated 4-week pilot studies of the drug in spasticity of spinal and supraspinal origin,


because GBP was able to relieve muscular rigidity in mice. In those early studies, only individual patients had a moderate benefit, probably because of the low doses (900 mg/day maximum) used at that time. Based on the results, the development program for the use of this drug in spasticity was not continued.

However, a 6-day, placebo-controlled, blinded crossover study using 2,700 mg/day GBP in 21 evaluable patients with multiple sclerosis demonstrated significant beneficial effects in self-rating and observer rating scales. Values on the Kurtzke Expanded Disability Status Scale were not different between active treatment and placebo, a finding that could be attributed to the short-term nature of this trial. Tolerability was not found to be a problem, not even during the rapid forced titration (10).

A longer, but open, study looked at the effects of 600 to 1,200 mg/day GBP given for a follow-up period ≥3 and ≤9 months in 18 evaluable patients with multiple sclerosis who had painful tonic spasms, trigeminal neuralgia, dysesthetic or paresthetic symptoms, and ocular ataxia resistant to previous treatment. Fourteen of the 18 patients experienced complete and sustained recovery within 1 month of the start of treatment, as evaluated by means of a three-point score compared with the baseline value. Mild somnolence occurred in two patients (11). Larger and longer blinded randomized controlled trials with electrophysiologic and functional outcomes would be required to establish definitively GBP's place in disorders associated with spasticity.


Two open case series investigated the use of GBP in the restless leg syndrome and periodic limb movement disorder of sleep. Three of eight patients receiving a median dose of GBP of 1,163 mg/day for a minimum of 1 week reported ≥75% improvement, compared with pretreatment. Adverse events included dizziness, nausea, and drowsiness, causing two patients to discontinue the trial (12).

Of the initial 71 patients with periodic limb movement disorder of sleep, 48% were rated subjectively improved in terms of sleep depth and continuity, when the GBP dose was titrated ≤1,800 mg/day, with the larger of the divided doses given in the late afternoon. Otherwise rare adverse events, abnormal sensations, and movements are reported in a review of restless leg syndrome and periodic limb movement disorder of sleep (13).


In an acute challenge experiment, 19 patients with Parkinson's disease (Hoehn and Yahr stages 2.5 to 5.0) received six consecutive doses of 400 mg GBP, in addition to dopaminergics, followed by six doses of placebo, or vice versa, in a double-blind, crossover trial. A significant (p = .0028) improvement of total score on the Unified Parkinson's Disease Rating Scale was noted in favor of GBP, but not in individual subscales. The most frequent adverse events in patients between 41 and 78 years of age were unsteady gait, dizziness, and sleepiness. The authors of this study discussed the weaknesses of this brief study, such as a potentially unsuitable dose regimen, short exposure, and failure to discriminate among various forms of Parkinsonism. These investigators also stated that other patients who were not in the study did benefit from GBP over several years, whereas others observed a worsening of their disease (14).

In 152 patients with amyotrophic lateral sclerosis who underwent 6 months of treatment with 2,400 mg/day GBP, a nonsignificant trend toward a slower decline in arm strength megascores versus placebo was described, with a median difference of 37% and no effect on vital capacity. The study was initially powered to detect a 50% reduction in arm megascore decline. Compared with placebo, there were significantly more adverse effects, such as light-headedness, drowsiness, and limb swelling, in the GBP-treated group (15).

Later, an open, randomized, natural history-controlled study looked into the effects of GBP, 500 versus 1,000 mg/day, given for 6 to 12 months to a total of 110 patients with definite amyotrophic lateral sclerosis. An additional 121 patients who were receiving other symptomatic treatments were considered the control group. The authors of this study found dose- and treatment duration-dependent effects on average slopes for less decline of muscular limb strength, forced vital capacity, a disease severity score (Norris), and an activities of daily living scale (Functional Rating Scale). Statistical differences are significant when the patients who were treated for longer periods and with higher doses are compared with the natural history control group. No blinding or placebo was involved in this study (16).


Four patients with Huntington's disease were reported in an abstract to have received a low dose of GBP, 900 mg/day, for a mean duration of 3 years. Improvement in terms of reduction of choreic movements was measured, as well as severity of Huntington's disease, by the Abnormal Involuntary Movement Scale. In all patients from this open case series, a moderate to good effect on the Abnormal Involuntary Movement Scale without adverse events was noted (17).

Another study treated 12 outpatients with schizoaffective and bipolar disorders and tardive dyskinesia for 1 to 6 years and who had otherwise pharmacoresistant antipsychotic-induced movement disorders; GBP, 600 to 1,200 mg/day, was added to their baseline medication. Ten of 12


patients in this openly treated cohort experienced a clinically relevant benefit after 3 weeks. In additional single patients, a continued effect, particularly on blepharospasm and facial dyskinesias, outlasted the 24-month observation time point, without relevant adverse events, aside from some sedation (18).

A single case report deals with good results of GBP titrated ≤75 mg/kg/day after 6 months in a 13-year-old pediatric patient, presenting with hemichorea and hemiballismus after an embolic stroke. The patient did not complain of dizziness, nausea, or sedation at the time of peak GBP levels (19).

In contrast to the benefits of GBP in a variety of movement disorders, worsening or the new development of abnormal movement patterns when GBP is initiated has been noted and reviewed; all these effects are reversible after dose reduction or withdrawal of GBP. In the case of a 68-year-old patient, the development of paroxysmal dystonic movements in both hands was reported during combined therapy with propranolol and GBP 900 mg/day; the problem resolved immediately after reduction of the propranolol dose to 40 mg/day, a finding suggesting a pharmacodynamic interaction between the two drugs (20).


Besides pain syndromes, GBP is widely used at present in various types of tremors, mainly essential tremor. Eighteen patients with essential tremor completed a placebo-controlled crossover design with 2-week treatment periods of GBP, 1,800 mg/day, added to baseline antitremor medication. Tremor was assessed by examiners who rated tremors in different body parts, using Fahn-Tolosa-Marin Tremor Rating Scales, as well as a global disability rating. Statistical analysis for nonparametric data did not show a difference in any outcome parameters among the baseline, GBP, and placebo values. Two patients from the initial group of 20 randomized dropped out because of adverse events that occurred while they were taking GBP, and seven patients reported fatigue, nausea, or dizziness (21).

In contrast to the previously described add-on therapy, GBP was given as the only antitremor drug to 16 patients with essential tremor in a double-blind, three-way placebo-and propranolol-controlled crossover study (GBP, 1,200 mg/day; propranolol, 120 mg/day). Significant and similar improvements in all outcome measures (Tremor Clinical Rating Scale, self-reported disability scale) were found for both treatment strategies, GBP and propranolol, versus baseline. Accelometric assessments proved not to work methodologically within this trial. The authors of this study concluded that a larger and longer trial would be needed (22).

The call for blinded, well-controlled randomized controlled trials is reiterated in two further articles dealing with the use of GBP in orthostatic tremor in patients who unsuccessfully responded previously to clonazepam. Four, respectively seven, patients were treated with GBP individually at doses between 300 and 2,400 mg/day; immediate effects were mostly investigated, but one patient was observed for ≤22 months. In both open studies, orthostatic tremor was found to be reduced in most patients; the adverse effects were only mild to moderate and were much less severe than these patients had experienced with many other antitremor medications (23,24).


Results of studies of patients with migraine are available mainly through abstracts, which do not always describe the study in detail. The largest number of patients with migraine, with or without aura, is reported from a 12-week prophylaxis trial with two imbalanced groups of 99 patients taking between 1,800 and 2,400 mg GBP versus placebo. In a modified intent-to-treat population, the mean baseline migraine headache rate during the last month of treatment compared with baseline values, was 2.9 for the GBP-treated group and 3.4 for the placebo-treated group (p = .031). The proportion of patients who had a reduction of migraines of ≥50% was 36% with GBP and 14% with placebo; 16% of the patients taking GBP and 9% taking placebo discontinued the trial because of adverse events. Dizziness (25%), somnolence (24%), and asthenia (22%) were the leading adverse events within the GBP-treated group (25).

In one study, 145 Italian patients with migraine were openly randomized to GBP, either 1,200 or 2,000 mg/day, and were evaluated after weeks 4, 10, and 16. Only 34 of the patients completed the trial, but in the abstract, no reason for the dropout is provided. In both dosage groups, reduction of migraine crisis was the same after weeks 4 and 16 (40% and 65%, respectively) versus baseline. Adverse events included somnolence (26%), asthenia (10%), and dizziness (7%). Rightly, the authors concluded that other studies are necessary (26).

Another cohort of 20 Italian patients with migraine, with or without aura, openly received 600, 900, or 1,200 mg/day GBP for 6 months. A relevant improvement in number of attacks per month, pain intensity, and disability was observed for the 900- and 1,200-mg doses, but not for the 600-mg dose. Tolerability was not a problem (27).


In addition to the previously mentioned categories, case series or anecdotal reports note GBP use in rarer neurologic conditions such as reflex sympathetic dystrophy, central pain, myokymia-cramp syndrome, muscle cramps associated with various diseases, and hemifacial spasms and as adjunctive therapy for idiopathic chronic hiccup. The


reported patients usually improved within a dose range of GBP of 1,200 to 2,700 mg/day, in a thrice-daily dose regimen (28).


  1. Backonja M, Beydoun A, Edwards KR, et al. Gabapentin for the symptomatic treatment of painful neuropathy in patients with diabetes mellitus: a randomized controlled trial.JAMA1998;280:1831-1836.
  2. Rowbotham M, Harden N, Stacey B, et al. Gabapentin for the treatment of postherpetic neuralgia: a randomized controlled trial. JAMA1998;280:1837-1842.
  3. Morello CM, Leckband SG, Stoner CP, et al. Randomized double-blind study comparing the efficacy of gabapentin with amitriptyline on diabetic peripheral neuropathy pain.Arch Intern Med1999;159:1931-1937.
  4. Sindrup SH, Jensen TS. Efficacy of pharmacological treatments of neuropathic pain: an update on effect related to mechanism of drug action. Pain1999;83:389-400.
  5. Magnus L. Nonepileptic uses of gabapentin. Epilepsia1999;40[Suppl 6]:S66-S74.
  6. Khan OA. Gabapentin relieves trigeminal neuralgia in multiple sclerosis patients. Neurology1998;51:611-614.
  7. Neville MW. Gabapentin in the management of neuropathic pain. Am J Pain Manage2000;10:6-12.
  8. Caraceni A, Zecca E, Martini C, et al. Gabapentin as an adjuvant to opioid analgesia for neuropathic cancer pain. J Pain Symptom Manage1999;17:441-445.
  9. Attal N, Brasseur L, Parker F, et al. Effects of gabapentin on the different components of peripheral and central neuropathic pain syndromes: a pilot study. Eur Neurol1998;40:191-200.
  10. Cutter NC, Scott DD, Johnson JC, et al. Gabapentin effect on spasticity in multiple sclerosis: a placebo-controlled, randomized trial. Arch Phys Med Rehabil2000;81:164-169.
  11. Solaro C, Lunardi GL, Capello E, et al. An open-label trial of gabapentin treatment of paroxysmal symptoms in multiple sclerosis patients. Neurology1998;51:609-611.
  12. Alder CH. Treatment of restless legs syndrome with gabapentin. Clin Neuropharmacol1997;20:148-151.
  13. Williams DC. Periodic limb movements of sleep and restless legs syndrome. Va Med Q1996;123:260-265.
  14. Olson W, et al. Gabapentin for Parkinsonism: a double-blind, placebo-controlled, crossover trial. Am J Med1997;102:60-66.
  15. Miller RG, Moore D, Young LA, et al. Placebo-controlled trial of gabapentin in patients with amyotrophic lateral sclerosis. Neurology1996;47:1383-1388.
  16. Mazzini L, Mora G, Balzarini C, et al. The natural history and the effects of gabapentin in amyotrophic lateral sclerosis J Neurol Sci1998;160[Suppl 1]:S57-63.
  17. Cosentino C, Torres L, Cuba JM. Gabapentin for Huntington's disease. J Neurol1996;243[Suppl 2]:S75-S76.
  18. Hardoy MC, Hardoy MJ, Carta MG, et al. Gabapentin as a promising treatment for antipsychotic-induced movement disorders in schizoaffective and bipolar patients. J Affect Disord1999;54:315-317.
  19. Kothare SV, Pollack P, Kulberg AG, et al. Gabapentin treatment in a child with delayed-onset hemichorea/hemiballismus. Pediatr Neurol2000;22:68-71.
  20. Palomeras E, Sanz, P, Cano, A, et al. Dystonia in a patient treated with propranolol and gabapentin. Arch Neurol2000;57:570-571.
  21. Pahwa R, Lyons K, Hubble JP, et al. Double-blind controlled trial of gabapentin in essential tremor. Mov Disord1998;13:465-467.
  22. Gironell A, Kulisevsky J, Barbanoj M, et al. A randomized placebo-controlled comparative trial of gabapentin and propranolol in essential tremor. Arch Neurol1999;56:475-580.
  23. Onofrj M, Thomas A, Paci C, et al. Gabapentin in orthostatic tremor: results of a double-blind crossover with placebo in four patients. Neurology1998;51:880-882.
  24. Evidente VG, Adler CH, Caviness JN, et al. Effective treatment of orthostatic tremor with gabapentin. Mov Disord1998;13:829-831.
  25. Mathew NT, Magnus-Miller L, Saper J, et al. Efficacy and safety of gabapentin in migraine prophylaxis. Cephalgia1999;19:380.
  26. Jimenez MD, Friera G, Manjon MT. Efficacy and safety of gabapentin in prophylaxis of migraine headache. Cephalgia1999;19:376.
  27. Farinelli M, Betti E, Benedetti O, et al. Gabapentin in drug resistant migraine. Cephalgia1999;19:378.
  28. Merren MD. Gabapentin for treatment of pain and tremor:a large case series. South Med J1998;91:739-744.