BackgroundSchizophrenia is a highly heterogeneous disorder, and despite extensive research progress approximately 30% of patients with schizophrenia show poor response to first-line antipsychotics, denoted as treatment-resistant schizophrenia (TRS). Meta-analytic evidence showed that clozapine is the most effective antipsychotic for TRS, although 40% of TRS patients do not respond even to clozapine (ultra-treatment-resistant schizophrenia -UTR). Recent studies indicated that TRS is neurobiologically and categorically distinct from treatment-responsive schizophrenia, being associated with elevated glutamate levels in the anterior cingulate cortex and unaltered striatal dopamine synthesis. Moreover, the striking majority of TRS patients do not respond to first-line antipsychotic therapy since disease onset and present more severe cognitive deficits since first episode of psychosis, further suggesting the presence of a distinct and more disrupted neurobiological substrate. It is widely known that cognitive impairment is a core feature of schizophrenia and determines a significant detrimental impact on long-term functional outcome, which represents the ultimate treatment goal. However, despite the central role of cognition in schizophrenia, to date no study has investigated longitudinal cognitive outcome among TRS patients.Based on these evidences, the aim of this study is to evaluate longitudinal cognitive trajectories in a sample of clinically stabilized patients with schizophrenia, stratified according to antipsychotic response. We hypothesized that treatment-resistance is associated with a more severe long-term cognitive decline.MethodsWe enrolled 93 patients with schizophrenia (DSM-V), stratified as follows: 32 first-line responders (FLR), 42 TRS and 19 UTR. Cognition was longitudinally assessed at baseline and at least after 6 years of follow-up (mean: 9.3±2.8 years) using the Brief Assessment of Cognition in Schizophrenia (BACS). From BACS subscores we calculated for each patient a Cognitive Index, as a measure of overall cognitive functioning. In order to quantify global cognitive functioning changes during the course of illness, we estimated effect size score for Cognitive Index using Cohen’s d. Finally, General linear Models (GLM) were performed with overall cognitive index effect size as dependent variable, treatment (FLR/TRS/UTR) as categorical variable and age, duration of illness and education as covariates.ResultsThe first GLM (FLR vs TRS+UTR) showed a significant main effect of treatment (F=7.34, p=0.01), with worse cognitive outcome between resistant patients. Consistently, the second GLM (FLR/TRS/UTR) resulted significant as well (F=17.90, p<0.001), with UTR group showing worse cognitive trajectory (Fisher’s post-hoc: p<0.001, UTR Cognitive Index effect size = -0.7).DiscussionThis is the first study to longitudinally evaluate cognitive trajectories of patients with schizophrenia according to their antipsychotic response. We showed that treatment resistance is associated with a more severe cognitive decline, with worse outcome among UTR patients. These data suggest that greater severity of treatment resistance in schizophrenia is associated with greater cognitive impairment, possibly due to the presence of a distinct and more disrupted neurobiological substrate that affects both cognition and antipsychotic response. These findings further highlight the necessity of early individuation and tailored pharmacological treatment for TRS patients, in order to improve long-term clinical, cognitive and functional outcome.