Gabapentin in orthostatic tremor: Results of a double-blind crossover with placebo in four patients

An unusual tremor of legs and trunk during standing that disappears on walking and sitting was described by Heilmann in 1984¹ as orthostatic tremor (OT). A recent review by Thompson² describes more than 50 cases of OT, often appearing in association with tremor during walking, with essential tremor (ET), or with parkinsonian tremor.

Thompson's review and a recent consensus conference on tremor classification³ point out that the defining characteristics of OT are its appearance during standing and the electromyography (EMG) frequency varying between 14 and 24 Hz, which is higher than frequencies of other tremors.

Although several treatment approaches to OT, like primidone, clonazepam, valproic acid, and phenobarbital,^2,3 were described in monotherapies or in combinations, the consensus conference agreed on the inefficacy of various therapies, confirming that only clonazepam is an effective drug. They also suggest that responsiveness to clonazepam might be considered, with tremor frequency, the criterion that enables the identification of OT.³ Here we suggest that gabapentin (Gbp), a new antiepileptic agent with a lipophilic acid structure(probably acting through GABA inhibition potentiation),⁴ which also was used to treat pendular nystagmus⁵ and restless legs syndrome,⁶ also reduces OT.

Case reports. Case 1. Patient is a 71-year-old man with insulin-dependent diabetes. He had been confined at home for the past 6 years because of severe OT with an EMG frequency of 16 Hz, with increasing amplitude in 1 to 2 minutes after standing. A typical slow-frequency (3.5 to 4 Hz), levodopa-responsive (300 mg/d) parkinsonian resting tremor of both hands had appeared 5 years after the first occurrence of OT. Clonazepam, 0.5 mg three times daily, had reduced OT for 2 weeks but was discontinued because of ataxia and somnolence. MRI and CT scan showed two symmetric lesions, which were hyperintense on T2-weighted images of the globus pallidus medialis.

Case 2. Patient 2 is a 73-year-old woman confined at home for the past 5 years by an incapacitating OT at 14 to 17 Hz, which disappeared only during sitting and lying down. This patient also was affected by alcohol-responsive 8-Hz ET of the head and hands. Clonazepam reduced both OT and ET but was not tolerated because of somnolence; primidone, 62.5 mg twice daily, also was not tolerated.

Case 3. Patient 3 is a 67-year-old woman with a 2-year history of incapacitating OT at 18 Hz. Unsteadiness and shaking of the legs appeared after a few seconds in the standing position and disappeared when leaning against a wall, sitting, or lying down. Clonazepam, 0.5 mg three times daily, effectively reduced OT but was not tolerated because of excessive somnolence. Primidone therapy also was discontinued after 3 weeks of treatment at a maximum dosage of 125 mg/d in two doses.

Case 4. Patient 4 is a 76-year-old man who was unable to walk unaided because of a 5-year history of OT, which appeared within 1 minute after standing, increased in amplitude the longer he stood, and was not relieved by walking. OT frequency is 14 Hz with a subharmonic at 24 Hz. Clonazepam initially had reduced OT with severe somnolence. For 1 year, the dose of clonazepam had been increased to 3 mg/d to subjectively obtain the same tremor control. Clonazepam eventually was discontinued because of somnolence and recurrence of tremor.

Methods. Before treatment, each patient gave informed consent to undergo three separate drug efficacy evaluations, one of which could include a placebo.

OT was evaluated according to the following rating suggested by Fahn et al.⁷: 0, none; 1, slightly perceivable or intermittent; 2, modest amplitude <2 cm; 3, marked, nonintermittent, amplitude 2 to 4 cm; and 4, severe, amplitude above 4 cm. Rating scales were obtained weekly for each patient when lying down, standing, and walking.

A patient self-rating scale also was used,⁸ which asked each patient whether the tremor was unchanged (0), better (slightly, 1; considerably, 2), or worse (slightly, +1; considerably, +2).

Tremor was rated by an examiner unaware of the patient's self-rating every week for 9 months. Questionnaires for unwanted or adverse effects were administered weekly. Gbp, in 300-mg capsules, was administered in doses increasing from 300 to 900 mg (in three administrations) in 3 days. After 4 weeks, the dose was increased by 300 mg every 2 days until a total dose of 600 mg given three times daily (1,800 mg) was reached in each patient. In Patient 2, the Gbp dose was further increased after 4 weeks to 2,400 mg/d(two 400-mg capsules three times daily).

After 8 weeks of treatment, Gbp was tapered over 1 week and then substituted with two capsules given three times daily containing 250 mg of L-carnitine. A second crossover was performed 9 weeks after placebo treatment and tapering; then each patient was returned to the Gbp daily dose(1,800 mg for Patients 1, 3, and 4; 2,400 mg for Patient 2) for 9 more weeks.

Statistical evaluations on tremor rating scales and self-rating scales included Student's t-test and paired Wilcoxon test cross-relations between scores obtained at weeks 1 and 3 (basal); weeks 4, 5, and 7(Gbp, 900 mg); weeks 9, 13, and 16 (Gbp, 1,800 to 2,400); week 18(Gbp tapering); weeks 14, 21, 25, and 27 (placebo and placebo tapering); and weeks 28, 31, and 35 (Gbp crossover from placebo).

The surface EMG was recorded in each patient simultaneously from both tibialis anterioris, gastrocnemii mediales, vasti mediales, bicipites femuris, and the extensor and flexor carpi ulnaris of the left arm by bipolar Ag/AgCl electrodes placed over the muscle belly. In Patient 2, the EMG also was recorded from cervical paraspinal muscles referenced to mastoids. Bandpass filters were set at 5 to 5,000 Hz, and gain was 20,000; acquired signals were processed with a Brain Surveyor Unit (EB Neuro, Florence, Italy) performing fast Fourier transform frequency analysis of 32 repeated 8-second epochs with a 256-Hz sampling rate. EMG amplitude changes, in square millivolts measured at the specific tremor frequencies, were recorded and statistically compared (Student's t-test, paired Wilcoxon test) on weeks 3, 7, 15, 24, and 35.

Results. The left half of the figure shows the EMG frequency analysis of the four patients, with OT frequencies of 16, 15 to 17, 18, 14 and 24 Hz appearing only while standing. Thefigure also shows a parkinsonian resting tremor recorded from the flexor carpi ulnaris in Patient 1, and ET recorded from paraspinal neck muscles and flexor carpi ulnaris in Patient 2. Statistically, the OT amplitude reduction-basal versus weeks 15 and 35-was significant at the 0.001 level and was not significant versus week 7 (900 mg/d) or week 24 (placebo). The figure shows OT reduction or disappearance during treatment.

• Download figure
• Open in new tab
• Download powerpoint

Figure. Electromyography frequency analysis of 32 epochs lasting 8 seconds each, recorded in the four patients while they were lying down, standing, and standing during Gbp treatment of 1,800 to 2,400 mg/d. Horizontal calibration is 32 Hz, and vertical calibration is 300 mV². In all patients, orthostatic tremor is recorded when standing and is reduced or disappears with treatment. In Patient 1, a slow arm tremor is recorded at rest, which was not modified by Gbp. In Patient 2, the head-and-arm 8-Hz tremor was not modified by Gbp. FCU = flexor carpi ulnaris; GM = gastrocnemius medialis; CP = cervical paraspinal.

Tremor scores were 3 ± 0.5 during basal evaluation and placebo treatment, 2 ± 0.6 during 900-mg Gbp treatment, and 1 ± 0.5 during 1,800- to 2,400-mg Gbp treatment. Patients' self-rating scales were congruent with tremor rating scores, and cross-correlation between the two rating systems was highly significant (p < 0.01). Student's t-test and paired Wilcoxon tests confirmed the significant improvement of tremor with p < 0.01 during the 1,800- to 2,400-mg/d treatment and p ≥ 0.05 during the 900-mg/d treatment. OT tremor scores decreased to 2.00 ± 0.5 during week 13, 1.75 ± 0.5 in week 16, and 1.5 ± 0.5 in week 19, suggesting a dose-dependent OT reduction.

In Patients 1, 3, and 4, the Wilcoxon test showed significant reduction of OT tremor and of self-assessment scores at weeks 13 through 18 and 30 through 35. In Patient 2, OT reductions (from 4 to 2) were observed only in weeks 16 and 17, and in weeks 31 through 35.

Complete disappearance of tremor was obtained only in Patients 3 and 4 at weeks 16 to 17 and weeks 31 through 35. The self-assessment scales showed that all patients were satisfied with the treatment, and at the end of the study, when patients were informed of the crossover scheme, each patient requested to continue and is still receiving treatment. The satisfaction of all patients was confirmed by the fact that all reported being able to resume independent daily life, with no need of help when walking.

Subjectively, no patient reported sudden improvement after taking a single dose of Gbp (600 or 800 mg); slight amelioration was reported after 2 or 3 days of treatment. During the entire study, only Patient 2 complained of vertigo as a side effect, but vertigo disappeared after 1 week of treatment and also was reported during the first week of placebo treatment.

Discussion. OT is a relatively rare type of tremor scarcely responding to treatment.^2,3 A recent PET study shows abnormal activation of the cerebellum, lentiform nuclei, and thalamus.⁹ OT therefore appears to be a possible target for a drug-potentiating GABAergic inhibition.

The current study showed that OT almost disappeared during Gbp treatment in three of the patients, whereas in one patient, only the 2,400-mg dose induced tremor reduction.

Gbp was well tolerated, and side effects were noted in only one patient and were indistinguishable from the effects of placebo. All patients continued treatment after our study was completed.

Here we report the results of a new treatment and underscore the fact that OT is an incapacitating tremor because it interferes with the patient's ability to stand and, often, as in other reported cases,² to walk.

The recently proposed tremor rating scale does little to account for the effect of OT on daily living abilities⁷ and does not suggest a way to rate its incapacitating effect.