Gabapentin in postamputation phantom limb pain: A randomized, double-blind, placebo-controlled, cross-over study

Abstract

BACKGROUND AND OBJECTIVES: Severe phantom limb pain after surgical amputation affects 50% to 67% of patients and is difficult to treat. Gabapentin is effective in several syndromes of neuropathic pain. Therefore, we evaluated its analgesic efficacy in phantom limb pain. METHODS: Patients attending a multidisciplinary pain clinic with phantom limb pain were enrolled into this randomized, double-blind, placebo-controlled, cross-over study. Other anticonvulsant therapy was discontinued. Each treatment was 6 weeks separated by a 1-week washout period. Codeine/paracetamol was allowed as rescue analgesia. The daily dose of gabapentin was titrated in increments of 300 mg to 2400 mg or the maximum tolerated dose. Patients were assessed at weekly intervals. The primary outcome measure was visual analog scale (VAS) pain intensity difference (PID) compared with baseline at the end of each treatment. Secondary measures were indices of sleep interference, depression (Hospital Anxiety and Depression [HAD] scale), and activities of daily living (Bartel Index). RESULTS: Nineteen eligible patients (mean age, 56 years; range, 24 to 68 years; 16 men) were randomized, of whom 14 completed both arms of the study. Both placebo and gabapentin treatments resulted in reduced VAS scores compared with baseline. PID was significantly greater than placebo for gabapentin therapy at the end of the treatment (3.2 +/- 2.1 v 1.6 +/- 0.7, P =.03). There were no significant differences between placebo and gabapentin therapy in terms of the number of tablets of rescue medication required, sleep interference, HAD scale, or Bartel Index. The medication was well tolerated with few reports of adverse effects. CONCLUSIONS: After 6 weeks, gabapentin monotherapy was better than placebo in relieving postamputation phantom limb pain. There were no significant differences in mood, sleep interference, or activities of daily living, but a type II error cannot be excluded for these variables.