¹Pfizer, Inc, Ann Arbor, MI

Abstract

Background: Based on the results of the Phase 2 trials in insomnia patients, one out of 2 compounds was to be selected for further development. A CUI was developed based on data from Compound 1. The CUI is an integrated measure of clinical benefit/risk over the dose-range and was proposed as a tool to compare compounds.

Methods: The attributes of an ideal agent were identified and included 5 measures of efficacy (e.g. sleep onset, maintenance, quality, sleep stages) and 2 measures of carryover (morning) sedation. These attributes were ranked as a percentage for their relative importance based on market research. Dose-response analyses were conducted on each endpoint and each scale was normalized based on clinically meaningful differences. CUI was calculated over the dose-range. Sensitivity analyses were conducted to test for clinically meaningful changes in CUI. Bootstrap parameters of the dose-response curves were used to build confidence intervals.

Results: The shape of the CUI-dose relationship was different for Compound 1 and 2. Peak CUI (80% CI) was 0.345 (0.25-0.43) and 0.436 (0.35-0.52), respectively. Although CUI was greater for Compound 2, peak CUI was observed at doses that were not considered viable.

Conclusions: A CUI was used as a novel approach for decision making. CUI is helpful to characterize optimal dose range and is particularly useful when decisions are based on trade-offs amongst multiple clinical endpoints. Development of Compound 2 was discontinued based on these results.