A meta-analysis of pain response in the treatment of fibromyalgia
Neil S Roskell1, Stephen M Beard, Yang Zhao
1Statistics, RTI Health Solutions, Manchester, U.K. nroskell@rti.org
Objective:
This meta-analysis compared efficacy (pain response) of drugs that are licensed or commonly used in the treatment of fibromyalgia. A meta-analysis of safety measured via discontinuation because of adverse events was also performed.
Methods:
We conducted a meta-analysis of 21 clinical trials to estimate treatment differences vs. placebo, separately, for duloxetine, fluoxetine, gabapentin, milnacipran, pramipexole, pregabalin, either of two tricyclic antidepressants, and tramadol plus paracetamol. Indirect treatment comparisons using mixed treatment comparisons methodology were conducted for all pairwise comparisons. Pain response was analyzed as improvement of at least 30%, and separately of 50%, from baseline.
Results:
When compared with placebo, statistically significant pain responses (improvement of 30% and 50%) were observed for patients treated with duloxetine, milnacipran 200 mg/day, pregabalin 300 or 450 mg/day, and tramadol plus paracetamol. Treatment with fluoxetine, gabapentin, or milnacipran 100 mg/day resulted in significant findings for the 30% improvement in pain response. The meta-analysis showed a statistically increased risk of discontinuation because of adverse events for milnacipran 100 and 200 mg/day (both P < 0.001), and pregabalin 300 and 450 mg/day (P = 0.009 and P < 0.001, respectively). All other treatments, except fluoxetine, showed numerically increased risk over placebo for discontinuation because of adverse events. In the indirect comparisons, no pairwise comparison of active treatments reached statistical significance for either pain response end point.
Conclusion:
All eight active treatments displayed evidence suggesting improvement over placebo in the treatment of pain in patients suffering from fibromyalgia. Indirect comparison of active treatments found no strong differences.
