O10.4. LATENT INHIBITION AS A STRATIFICATION TOOL FOR SCHIZOPHRENIA DRUG DEVELOPMENT

Abstract

Background: Stratified medicine approaches have potential to dramatically improve the efficacy of drug development for schizophrenia and other psychiatric conditions, as they have for cancer. Such methods require accurate biomarkers that define populations of patients with a common neurobiological underpinning, for whom a particular treatment is more likely to be effective. Latent inhibition, a learning phenomenon known to be disrupted in schizophrenia, is a good candidate biomarker. Previous studies have shown that LI is typically impaired in positive symptomatic patients and remediated with dopamine blocking therapies, but accentuated in negative and cognitively symptomatic patients that anti‐psychotics are largely ineffective for. Computerised LI assessment may therefore be a simple noninvasive means to differentiate patients with differing neurochemical states and/or etiologies. Methods: We modified Granger et al's (2016) LI task to allow web‐based delivery on the CANTAB Connect software platform. We tested the validity and repeatability of the task in inducing LI behaviour in a total of 200 individuals recruited and assessed online via Prolific. To assess sensitivity of the task to a pro‐cognitive pharmacological manipulation, we compared LI in 20 healthy non‐smoking volunteers who received a dose of 2mg nicotine or placebo via mouth spray in a double‐blind crossover design with two‐day washout. To assess task sensitivity to a clinical model, we compared LI in 30 healthy volunteers administered either 7.5% CO2 or medical air in an anxiety induced dopamine release protocol, again in a double‐blind crossover design, with 30‐minute washout between gas inhalations. Results: LI was reliably demonstrated in a 6‐minute task administered via the web to remote participants. Administration of nicotine, but not placebo, enhanced latent inhibition (p