ECT VERSUS SHAM FOR CLOZAPINE RESISTANT SCHIZOPHRENIA: A SECONDARY ANALYSIS OF A PILOT STUDY BASED ON PANSS-30 FACTOR ANALYTIC DIMENSIONS AND THE PANSS-6

Abstract Background About 40% of patients with Treatment-Resistant Schizophrenia (TRS) do not respond to an adequate trial with clozapine and are termed Clozapine- Resistant Schizophrenia (CRS). A metanalysis showed that clozapine augmentation with ECT is efficacious for CRS, but no studies with Sham-ECT were reported (1). We published a pilot single-blinded study trial, which, to our knowledge, is the first to evaluate the efficacy of electroconvulsive therapy (ECT) in comparison with Sham–ECT in CRS (2) (NCT02049021) (PS). In this PS, patients with CRS received ECT (n=12) or SHAM-ECT (n=7) and after 12 sessions, but no differences were found between both groups as measured by the total PANSS and subscales at the endpoint. Factors, obtained by factor analytic studies (3), as well as individual items of shorter versions of the PANSS (4), are considered more valid predictors of TRS than subscales. The aim of the present study is a reanalysis of our PS based either PANSS factors obtained through factors analysis as well as on the individual items of the PANSS-30 and the PANSS-6. Methods We used individual participant data (IPD) at the symptom level, based on a statistical approach of a study already published (5). We used PANSS-30 items, traditional subscales, as well as factor scores (Positive, Negative and Cognitive) derived from our factor analysis of patients with TRS (4). Additionally, we analyzed also all the PANSS-6 items and the sum of scores (hallucinatory behavior –P3 and lack of spontaneity and flow of conversation- N6) which identify severe cases. STATA, Comprehensive Metanalysis, and R were used for calculations of effect sizes (fixed effects) and power Results We obtained significant differences between groups with high effect sizes (>0.8) for three symptoms at endpoint: P3 (-0.993, 95% CI= -1.977 to -0.009, p= 0.048), N6 (-1.036, 95% CI= -2.025 to -0.047 p= 0.040), and Disorientation (G10) (-1.445, 95%CI=-2.48—0.40, p=0.006). The between-groups effect-sizes of other symptoms or factors showed no statistically significant differences Discussion The present reanalysis showed that patients with CRS do benefit from ECT treatment, with an improvement of their Hallucinatory Behavior or Lack of Spontaneity and Flow of Conversation, but suffer from some cognitive impairments (Disorientation). These findings are limited due to the small sample. We are conducting a trial with a much larger sample size to test the efficacy and tolerability of ECT vs Sham in CRS.