Long-Term Treatment of Bipolar Disorder with Valproate: Updated Systematic Review and Meta-analyses

Abstract

Supplemental digital content is available in the text. Abstract Learning objective After participating in this activity, learners should be better able to: • Evaluate the evidence regarding the effectiveness of long-term treatment of bipolar disorder with valproate Background Prophylactic treatment is critical for bipolar disorder (BD) patients. Valproate is commonly used for this purpose but lacks regulatory approval and carries appreciable risks. Methods Systematic literature searching through June 2020 sought prospective trials lasting ≥12 months with adults diagnosed with BD to support comparisons of risk of new illness episodes with valproate versus placebo or other agents. Results Included were 13 reports involving 9240 subjects treated for an average of 29.1 months (range, 12–124) in 21 trials: 9 were blinded, randomized trials (RCTs) of valproate versus placebo (n = 3), lithium (5), or olanzapine (1); 2 were unblinded RCTs versus lithium (1) or quetiapine (1); and 10 were open-label trials versus lithium (5), quetiapine (2), carbamazepine (1), lamotrigine (1), or olanzapine (1). Random-effects meta-analysis found valproate superior to placebo in 3 trials (odds ratio [OR] = 0.42 [95% confidence level (CI), 0.30–0.60]; p < .0001). In 11 trials, protective effects with valproate and lithium were similar (OR = 1.20 [CI, 0.81–1.79]; p = .36), as well in 5 comparisons versus antipsychotics quetiapine and olanzapine (OR = 0.96 [CI, 0.66–1.40]; p = .84), and 2 versus other mood-stabilizing anticonvulsants (carbamazepine and lamotrigine) (OR = 1.30 [CI, 0.75–2.26]; p = .34). Valproate was nonsignificantly more effective versus new mania than depression (χ2 = 3.03; p = .08). Conclusions Valproate was more effective than placebo in preventing new BD episodes of mania or depression, and not significantly different from lithium, second-generation antipsychotics, or other anticonvulsants. Overall benefits were nonsignificantly greater versus mania than bipolar depression.