Methods to address functional unblinding of raters in CNS trials.

Abstract

Treatment-emergent adverse events (TEAEs) associated with the unique properties of a pharmaceutical product may functionally unblind clinician ratings, obscure true medication effects, and affect confidence about clinical trial results. Central nervous system studies are particularly susceptible to functional unblinding because they rely on relatively subjective symptom assessments. Two different methods were used to examine possible functional unblinding in pooled data from three recent five-week, double-blind, placebo-controlled trials of xanomeline and trospium chloride (formerly known as KarXT) in participants with schizophrenia experiencing acute psychosis. Xanomeline/trospium is an M₁/M₄ muscarinic receptor agonist that may produce cholinergic side effects. First, we compared the scores of remote (site-independent) raters, blinded to TEAEs, who listened to audio recorded, site-based Positive and Negative Syndrome Scale (PANSS) interviews. Second, we conducted a post hoc analysis of participant subgroups with or without reported cholinergic-related TEAEs to ascertain whether cholinergic TEAEs influenced trial outcome. Remote ratings closely replicated 575 available "paired" site-based PANSS total scores at baseline and endpoint (intraclass correlation coefficient = 0.88 and 0.93, respectively). Both site-based and remote PANSS scores yielded significant improvement favouring xanomeline/trospium over placebo (both p < 0.0001) and yielded significantly greater treatment response (≥30% improvement from baseline) than placebo (both p < 0.0001). The significant improvement of PANSS scores favouring xanomeline/trospium over placebo was comparable in magnitude for all subgroups regardless of whether participants reported cholinergic-related TEAEs, or any TEAEs at all (all p < 0.001). In sum, the two different methods used to assess functional unblinding in these studies found no impact of cholinergic TEAEs, or any TEAEs, on the trial results. These methods may have utility across all clinical trials.