Treatment of bipolar depression with minocycline and/or aspirin: an adaptive, 2×2 double-blind, randomized, placebo-controlled, phase IIA clinical trial
Jonathan B Savitz,T Kent Teague,Masaya Misaki,Matt Macaluso,Brent E Wurfel,Matt Meyer,Douglas Drevets,William Yates,Ondria Gleason,Wayne C Drevets,Sheldon H Preskorn
DOI: https://doi.org/10.1038/s41398-017-0073-7
2018-01-24
Abstract:Given evidence of chronic inflammation in bipolar disorder (BD), we tested the efficacy of aspirin and minocycline as augmentation therapy for bipolar depression. Ninety-nine depressed outpatients with BD were enrolled in a 6 week, double-blind, placebo-controlled trial, and randomized to one of four groups: active minocycline (100 mg b.i.d.) + active aspirin (81 mg b.i.d.) (M + A); active minocycline + placebo aspirin (M + P); placebo-minocycline + active aspirin (A + P); and placebo-minocycline + placebo aspirin (P + P). A blinded interim analysis mid-way through the study led to the dropping of the M + P and A + P arms from further enrollment giving numbers per group who were included in the final analysis of: 30 (M + A), 18 (M + P), 19 (A + P), and 28 (P + P). When the study started, there were three primary outcome measures. Based on the results of the interim analysis, the primary outcome variable, response to treatment as defined by >50% decrease in Montgomery-Äsberg Depression Rating Scale (MADRS) score was maintained. The other two (i.e., the change in mean MADRS score from baseline to end of study and the remission rate, with remission being defined as a score of <11 on the MADRS) were reduced to exploratory outcome measures because the interim analysis indicated that the study was adequately powered to test differences in response rate but not the mean change in MADRS scores or remission rates. CRP and IL-6 were assayed to measure inflammation. Urinary thromboxane B2 (11-D-TXB2) concentrations, which were significantly increased at baseline in the combined BD sample (n = 90) vs. a healthy control group (n = 27), served as an indirect marker of cyclooxygenase (COX) activity. In a two-group analysis, the M + A group showed a greater response rate than the P + P group (p(one-tailed) = 0.034, OR = 2.93, NNT = 4.7). When all four arms were included in the analysis, there was a main effect of aspirin on treatment response that was driven by both the M + A and the A + P groups (p(two-tailed) = 0.019, OR = 3.67, NNT = 4.0). Additionally, there was a significant 3-way interaction between aspirin, minocycline, and IL-6, indicating that response to minocycline was significantly greater in participants in the M + P group with higher IL-6 concentrations. Further, participants in the M + P group who responded to treatment had significantly greater decreases in IL-6 levels between baseline and visit 7 vs. non-responders. Regarding the exploratory outcomes, there was a main effect for aspirin on the remission rate (χ12 = 4.14, p(2t) = 0.04, OR = 2.52, NNT = 8.0). There was no significant main effect of aspirin or minocycline on the mean change in MADRS score across visits. Aspirin and minocycline may be efficacious adjunctive treatments for bipolar depression. Given their potential import, additional studies to confirm and extend these findings are warranted.