Early response or nonresponse at week 2 and week 3 predict ultimate response or nonresponse in adolescents with schizophrenia treated with olanzapine: results from a 6-week randomized, placebo-controlled trial

In adults with schizophrenia, early response/non-response (ER/ENR) to antipsychotics at 2 weeks robustly predicts ultimate response/non-response (UR/UNR). However, less data about the predictive value of ER/ENR exist in adolescents with schizophrenia. Post hoc analysis of a 6-week trial in adolescents aged 13–17 with schizophrenia were randomized 2:1 to olanzapine or placebo. ER was defined as ≥20 % reduction in Brief Psychiatric Rating Scale-children (BPRS-C) total score at week 2 (ER2) or 3 (ER3); UR was defined with increasing stringency as total BPRS-C score reduction ≥20, ≥30, ≥40 or ≥50 %; remission was defined cross-sectionally using Andreasen et al. (2005) criteria. By week 2 (n = 69) and 3 (n = 66), olanzapine-treated youth achieved 73.3 and 85.5 % of their overall BPRS-C score reduction at 6 weeks last observation carried forward. ER and ENR patients did not differ significantly regarding baseline demographic, illness and treatment variables. ER 2 (frequency = 68.1 %) and ER 3 (frequency = 65.2 %) significantly predicted UR and remission (p = 0.0044–p < 0.0001), with ER3 having more predictive power. A ≥ 20 % BPRS-C reduction threshold for ER had best predictive validity (area under the curve = 0.88–0.92). At 6 weeks, patients with ER had significantly greater improvements in BPRS-C, Clinical Global Impressions Improvement and Severity scores, greater cross-sectional remission and less all-cause discontinuation (p = 0.047–p < 0.0001). Adverse event profiles were similar in the ER and ENR groups. Adolescents with schizophrenia experienced the majority of symptomatic improvement early during olanzapine treatment. ER predicted UR and remission, with ER3 having best predictive power. A ≥ 20 % improvement threshold for defining ER was confirmed as a robust outcome indicator.