Issues in Design of Translational Research Trials
Dr. Susan Hilsenbeck gave an insightful presentation titled “Issues in Design of Translational Research Trials” where she thoughtfully outlined translational research goals and biomarker measures in the early clinical trial space. She posted a number of questions on how best to include predictive biomarkers, such as which outcomes should be measured, the study design, power and sample size, and whether the study should be able to adapt. She defined biomarkers in terms of characteristics that are objectively measured as an indicator of normal biological processes, pathogenic processes or response to a therapeutic intervention.
Clinical trials have different aims with biomarkers and can usually be generalized by the phase of the clinical trial. For example, phase I clinical trials with targeted therapies tend to explore biomarkers, determine safety, and provide trends in efficacy. Within biomarkers, there also were a number of categories including (1) diagnostic, (2) monitoring, (3) predictive, (4) prognostic, (5) pharmacodynamic response, (6) safety and (7) susceptibility or risk, that she described. The role of the biomarker in the study can be either ancillary/exploratory where trial data is used to develop assays or identify biomarkers or biomarkers can be used in treatment assignment or eligibility. With the latter, this then needs to be a validated and standardized assessment for the study to proceed.
Dr. Hilsenbeck also compared trial designs, including those with adaptive or multi-stage designsas well as biomarker-stratified design, enrichment design, and biomarker-strategy design. Enrichment designs or biomarker-stratified designs allow for the biomarker to guide patient assignment. However, a positive study cannot distinguish between the biomarker guiding asuccessful treatment selection versus that the treatment options for those on the experimental arm are better than control for all patients. When designing these early clinical trials, it is important to pre-specify the hypotheses to be tested so that an appropriate sample size is planned. Further “interim” looks can be complex especially with a need to estimate false discovery rates. With any study, post-hoc should be treated with caution. Challenges can include missing biomarker status. This presentation gave an insightful look into early phase clinical trial design and considerations of biomarker sampling and analyses.
Applications for this presentation are directly relevant in cancer research. Biomarkers offer an exciting application to help guide new and existing treatments and identify those patients that would most benefit. Prognostic biomarkers offer the ability to classify patients into subgroups with different expected clinical outcomes to standard of care whereas predictive biomarkers can help identify patients that are more likely to respond or be resistant to a specific treatment. By effectively designing trials that incorporate biomarkers, this gives the field the possibility to identify, validate and improve novel prognostic and predictive biomarkers. However, careful consideration into the trial design as well as measured outcomes is necessary for there to be meaningful results and conclusions. It can be expected in the years to come that more biomarkers will emerge and help guide our future oncologic treatments.