Influencing Trial Design Through Use of the Alzheimer's Disease Clinical Trial Simulator

Abstract

Specific Aims: To demonstrate how the Alzheimer disease Clinical Trial Simulator can help influence clinical trial design by demonstrating if historical trials failed due to poor trial design or because the study drugs do not work via observed separation of treatment and placebo groups and changes in ADAS-Cog 11 scores from baseline to end. Methods: Clinical trials that evaluated a drug intended to treat mild to moderate Alzheimer disease were identified through a systematic literature search on ClinicalTrials.gov database. Studies included met the three inclusion criteria: any ADAS-Cog scale as an outcome, a disease modifying drug as an intervention, participants with mild to moderate Alzheimer’s disease. Sample size and trial duration were parameters extracted from each study and inputted into the trial simulator tool. The drug effect was then set at intervals (20%, 40%, 60%, 80%) until separation of the treatment group from placebo was noted at that drug effect. Results: Changes in ADAS-Cog 11 scores from the baseline to end demonstrates a lack of disease progression at each drug effect percentage for the treatment group (ADAS-Cog 11 Mean ± SD; 1 ± 2, 1 ± 1, 0 ± 1, 1 ± 2 for ≥20%, ≥40%, ≥60%, ≥80%, respectively. Changes in ADAS-Cog 11 scores from baseline to end (ADAS-Cog 11 Mean ± SD; 4 ± 4, 3 ± 3, 3 ± 1, 3 ± 4 for ≥20%, ≥40%, ≥60%, ≥80%, respectively) demonstrates progression of disease in the placebo group for each drug effect percentage. Conclusions: Most historical trials require drug effects larger than normally expected in clinical trials, indicating that the design of these trials could have been improved.