The choice of outcomes for clinical trials in preclinical AD is influenced by factors such as trial goals (e.g., proof-of-concept, target engagement, demonstrating efficacy), participant characteristics, type of intervention, available infrastructure, data available to assess potential performance, Health Authority input). API and its partners have confronted these issues in conceiving and implementing its two active trials. API first proposed preclinical trials in cognitively unimpaired PSEN1 E280Amutation carriers and APOE4 homozygotes using biomarker outcomes to make go-no-go decisions. Based on input from prospective industry partners and Health Authorities, the focus was broadened to include clinical outcomes. We used existing prospective datasets to empirically derive a composite cognitive test score as a primary endpoint to track change in the ADAD trial, serving as a single measure of multiple cognitive domains, reducing the risk of Type 1 error. The primary aim of the API APOE4 homozygote clinical trial is to assess time to MCI or dementia due to AD and/or change in a different composite cognitive test score defined specifically for this trial. External to this trial, efforts will be undertaken to validate the clinical meaningfulness of change in the composite test scores. In both trials, key secondary clinical outcomes were selected, and biomarker aims were modified to address possible prognostic, predictive and theragnostic utility. The sample size for the API Colombia trial, conducted with the Neurosciences Group of Antioquia and Genentech/Roche, is 200 carriers and, to maintain genotype blind, 100 noncarriers; that for the 4-arm APOE4 homozygote trial, conducted in partnership with Novartis/Amgen, is approximately 1340. Sample size estimates will be provided for biomarker outcomes. In order to receive Health Authority approval, experimental therapies in preclinical trials are required to show a clinical effect. There is no consensus regarding the optimal approaches to do so. API and its partners are testing assumptions and refining methods, and are addressing the future potential for biomarker outcomes to serve as “reasonably likely“ therapeutic surrogates for decision-making in drug development. We hope these findings, which will be shared, will help advance the field’s efforts to find ways to assess clinically meaningful change in preclinical trials.
Read full abstract