Duloxetine: painful diabetic neuropathy and fibromyalgia
Clinical bottom line
There is good evidence that duloxetine at doses of 60 or 120 mg daily helps some people with painful diabetic neuropathy or fibromyalgia. The NNT for one person to have at least 50% pain relief is about 6. Nausea, somnolence constipation and reduced appetite are common adverse events.
Sultan et al. Duloxetine for painful diabetic neuropathy and fibromyalgia pain: systematic review of randomised trials. BMC Neurology 2008 8:29. (http://www.biomedcentral.com/1471-2377/8/29) Lunn et al. Duloxetine for treating painful neuropathy or chronic pain. Cochrane Database Syst Rev. 2009 Oct 7;(4):CD007115.
Both reviews examined the same trial data
- Date review completed: 2008
- Number of trials included: 6
- Number of patients: 2216
- Control group: placebo
- Main outcomes: At least 50% pain relief, average pain scores, adverse events.
The trials were all properly done, with adequate quality and validity scores, in patients with at least moderate pain at baseline. The duration 12-13 weeks. Three trials in painful diabetic neuropathy involved about 30% of patients with a diagnosis of major depressive disorder. Duloxetine doses were 60 mg or 120 mg daily, with no important difference between 60 and 120 mg in any of the efficacy measures.
Painful diabetic neuropathy
Duloxetine dose made no difference to efficacy results, which were the same for both doses. Overall for duloxetine 60 or 120 mg daily the NNT was 5.1 (3.9 to 7.3) for at least 50% pain relief at 12 weeks. The Cochrane review also reported on the NNT for at least 30% pain relief at 12 weeks or less was 4.3 (3.2 to 6.6).
Duloxetine dose made no difference to efficacy results, which were the same for both doses. The NNT for at least 50% pain relief in trials of 12 weeks was 6.4 (4.7 to 9.9) for duloxetine 60 mg and 120 mg doses combined. The Cochrane review also reported on the NNT for at least 30% pain relief at 12 weeks or less was 5.8 (4.1 to 9.8).
Adverse event withdrawals were higher with duloxetine (15%) than placebo (8%). Particular adverse events higher with duloxetine than placebo were nausea (29% vs 10%), somnolence (14% vs 4%), constipation (13% vs 3%), and decreased appetite (7% vs 1%).
This is a set of good trials, of sufficient quality to avoid most known sources of bias, of sufficient size to avoid the effects of random chance, and of sufficient validity to be credible. The trials were long, at 12 weeks, and they show a reasonable level of efficacy for duloxetine in painful diabetic neuropathy and fibromyalgia. The single equivocation is that responders may well have been defined using last observation carried forward, and we do not yet know how using this imputation method may have affected results.