Comparison of Amitriptyline, Cyclobenzaprine, and Placebo in the Treatment of Fibromyalgia

Abstract Objective . To compare the relative efficacy and tolerability of amitriptyline, cyclobenzaprine, and placebo in the treatment of fibromyalgia, and to identify predictors of response to amitriptyline and cyclobenzaprine. Methods , Two hundred eight patients who fulfilled the American College of Rheumatology criteria for the classification of fibromyalgia were entered into a 6‐month prospective, double‐blind, multicenter trial and were randomized to 1 of 3 treatment groups: amitriptyline, cyclobenzaprine, or placebo. Results . After 1 month, 21%, 12%, and 0% of the amitriptyline, cyclobenzaprine, and placebo patients, respectively, had significant clinical improvement (amitriptyline versus placebo P = 0.002, cyclobenzaprine versus placebo P = 0.02, amitriptyline versus cyclobenzaprine P not significant). These percentages increased to 36%, 33%, and 19%, respectively, at the 6‐month assessment ( P not significant). The nature and frequency of side effects reported by patients treated with amitriptyline and those reported by patients treated with cyclobenzaprine were similar. A normal Minnesota Multiphasic Personality Inventory (MMPI) profile at baseline was predictive of clinical improvement at the 1‐month evaluation (odds ratio 3.3, 95% confidence interval 1.2—9.0). However, neither the MMPI profile nor any of the demographic, clinical, or functional parameters evaluated at baseline predicted long‐term response. Conclusion . Our data confirm the short‐term efficacy of amitriptyline and cyclobenzaprine in a small percentage of patients with fibromyalgia. Long‐term efficacy could not be demonstrated because of a higher‐than‐expected placebo response. Predictors of response to these drugs could not be determined.