SU16. Predictors of Treatment Response of Adjunct Minocycline to Clozapine in Refractory Schizophrenia

Abstract

Background: Accumulating evidence suggests minocycline (MINO) is effective as an adjunct treatment in schizophrenia. We previously reported significant improvement in avolition, anxiety/depression and working memory and a trend for positive symptoms and total Brief Psychiatric Rating Scale (BPRS) scores with MINO in a double-blind, placebo-controlled trial. We also found a >30% reduction in total BPRS scores in 25% of the MINO group compared to 4% in placebo (P = .04). Here we examine variables associated with symptom improvements. Methods: Participants with persistent positive symptoms of schizophrenia and schizoaffective disorder, despite adequate clozapine dosing, were randomized to receive adjunct MINO (100-mg BID) (N = 29) or placebo (N = 23). Both the BPRS and the Schedule for the Assessment of Negative Symptoms (SANS) were assessed biweekly in this 10-week trial. We examined demographic variables, duration of illness, baseline symptoms, serum cytokine levels, and CLZ blood levels to BPRS and SANS changes using Spearman’s correlation coefficients. Results: Symptom improvements were not related to demographic variables except that a shorter duration of illness related to SANS improvement in the MINO group (P = .04). Baseline total BPRS was associated with greater BPRS reduction (r = −0.51, P < .001), stronger in the MINO group than in placebo (difference in line slopes (F(1,45) = 5.13, P = .03). A greater increase in IFN-gamma with MINO was associated with a greater decrease in BPRS positive symptoms (r = −0.45, P = .02). SANS total score improvement was associated with an increase in TNF-alpha (r = −0.4, P = .045) and a trend for increased IFN-gamma (r = −0.38, P = .053). Lastly, CLZ levels were not related to BPRS; however, a greater decrease in CLZ levels related to a improvement in the SANS total score in the MINO group (r = 0.48, P = .01). In the placebo group, higher total CLZ was associated with greater SANS (difference in line slopes F(1, 42) = 4.43, P = .04). Conclusion: Variables associated with improvement on BPRS and positive symptoms include a higher baseline BPRS score and an increase in IFN-gamma levels. Variables associated with negative symptom improvement include shorter duration of illness, a decrease in CLZ levels, an increase in TNF-alpha, and trend for increase in IFN-gamma levels. These data suggest that despite MINO increasing CLZ levels (data not shown here) this does not contribute to MINO’s treatment effect. More research is needed to help define predictors of treatment response to MINO in larger samples.