Comparative Efficacy and Safety of Six Antidepressants and Anticonvulsants in Painful Diabetic Neuropathy: A Network Meta-analysis

Abstract

Background: Anticonvulsants and antidepressants are mostly used in management of painful diabetic neuropathy (PDN). However there are few direct comparisons between drugs of these classes, making evidence-based decision-making in the treatment of painful diabetic neuropathy difficult. Objectives: This study aimed to perform a network meta-analysis and benefit-risk analysis to evaluate the comparative efficacy and safety of these drugs in PDN treatment. Study Design: Comparative effectiveness study. Setting: Medical Education and Research facility in India. Methods: A comprehensive data search was done in PubMed, Cochrane, and Embase up to August 2012. We then systematically reviewed the studies which compared any of 6 drugs for the management of PDN: amitriptyline, duloxetine, gabapentin, pregabalin, valproate, and venlafaxine or any of their combinations. We performed a random-effects network meta-analysis to rank treatments in terms of efficacy and safety. We chose the number of patients experiencing ≥ 50% reduction in pain and number of patient withdrawals due to adverse events (AE) as primary outcomes for efficacy and safety, respectively. We also performed benefit-risk analysis, taking efficacy outcome as benefit and safety outcome as risk. Analysis was intention-to-treat. Results: We included 21 published trials in the analysis. Duloxetine, gabapentin, pregabalin, and venlafaxine were shown to be significantly efficacious compared to placebo with odds ratios (OR) of 2.12, 3.98, 2.78, and 4.43, respectively. Amitriptyline (OR: 7.03, 95% confidence interval [CI]: 1.87, 29.05) and duloxetine (OR: 3.26, 95% CI: 1.04, 9.97) caused more withdrawals than gabapentin. The ranking order of efficacy was gabapentin, venlafaxine, pregabalin, duloxetine/ gabapentin, duloxetine, amitriptyline, and placebo and the ranking order of safety was placebo, gabapentin, pregabalin, venlafaxine, duloxetine/gabapentin combination, duloxetine, and amitriptyline. Benefit-risk balance favored the order: gabapentin, venlafaxine, pregabalin, duloxetine/gabapentin combination, duloxetine, placebo, and amitriptyline. Limitations: We could not include valproate in our analysis owing to the lack of studies reporting the dichotomous efficacy and safety outcomes. Conclusion: Gabapentin was found to be most efficacious and amitriptyline to be least safe among the treatments included in the study. Gabapentin showed most favorable balance between efficacy and safety. Key words: Amitriptyline, diabetic neuropathy, duloxetine, gabapentin, network meta-analysis, pain, pregabalin, valproate, valproic acid, venlafaxine