Drug Treatments for Diabetic Neuropathy: A Meta-Analysis

This is for our chronic pain providers.

Published in Ann Intern Med 2014 Nov 4; 161:639

Authors: Griebeler ML et al.

Randomized trials are plentiful, but the evidence as a whole has substantial limitations.

A new systematic review and network meta-analysis of drug therapies for painful diabetic neuropathy has been published. The reviewers discovered 65 relevant randomized trials involving 27 medications, with nearly 13,000 participants. Key findings are as follows:
• Serotonin-norepinephrine reuptake inhibitors (SNRIs), topical capsaicin, tricyclic antidepressants (TCAs), and anticonvulsants all reduced pain compared with placebo.
• As a group, SNRIs and TCAs reduced pain more than did anticonvulsants and capsaicin.
• Very few studies extended beyond 3 months’ duration.
• For individual drugs, carbamazepine, venlafaxine, duloxetine, amitriptyline, and pregabalin1 were statistically superior to placebo.
• Few head-to-head comparisons of individual drugs exist, and most showed insignificant differences between drugs; however, pregabalin was inferior to venlafaxine and duloxetine.
• Side-effect profiles differ among these drug classes; adverse effects were problematic with virtually all of them.

Comment
Although clinicians have many choices of drug therapies for diabetic neuropathy, this analysis reveals weaknesses in the evidence. Studies generally were short-term, and most of these drugs have substantial central nervous system side effects. Moreover, comparative results are somewhat skewed by the size of randomized trials: For example, although the mean reduction in pain was slightly more with gabapentin than with pregabalin, only pregabalin’s results were statistically significant, because its industry-sponsored trials were much larger than those of gabapentin. Currently, only duloxetine and pregabalin are FDA-approved for treating patients with diabetic neuropathy.