035: Adoption of Patient Reported Outcomes in Early Clinical Trials for Oncology Drugs: Challenges & Opportunities

Statement of Purpose: Given the differences between oncology targeted therapies vs traditional chemotherapies, it has become critical to understand adverse event impact on patient function and quality of life (QoL). Consequently, there is currently a shift in the regulatory expectations for inclusion of Patient Reported Outcomes (PROs) in early-stage oncology trials and in dose optimization decision-making, consistent with the FDA Project Optimus. The need for PROs inclusion at early-stage was reiterated in the recent FDA guidance ‘Optimizing the Dosage of Human Prescription Drugs and Biological Products for the Treatment of Oncologic Diseases’. However, the adoption of PROs in oncology early-phase trials remains low. The objective of this work was to (1) review the current landscape for PROs inclusion in oncology early-phase clinical trials (2) review statistical and pharmacometrics analysis methods for PROs in early clinical trials and (3) provide recommendations for incorporation and interpretation of PROs in dose selection decision-making.

Description of Methods & Materials: A search was done to evaluate the adoption of PROs in recent oncology Phase 1 and 2 trials that were entered into clinicaltrials.gov between Jan 2021 and Jan 2023. A review of literature was also conducted to summarize statistical and pharmacometric analysis approaches for PRO data collected from early-stage clinical trials.

Data & Results: This work was ongoing at the time of abstract submission, but our findings should remain consistent with prior reviews which reported that only 3.4% of the trials (2015-2019) included a PRO endpoint (Coleman 2021) and that only 5.3% of the early phase, dose-finding oncology trials (2007-2020) had PROs or QoL outcomes (Lai-Kwon 2021). Limited reports of formal statistical and pharmacometrics model-based data analysis for PROs are available in early clinical development. In particular, standardization is lacking and numerous limitations are still present when it comes to evaluating PROs in Phase 1 oncology studies. Statistical analyses of PROs as primary and secondary endpoints included linear mixed models, linear regression, and analysis of covariance for multivariable analysis while binary and ordinal outcomes utilized logistic regression, and lastly some trials used repeated measures analysis (Qian 2021). Pharmacometric analyses were based on traditional exposure-response analyses (for instance based on scores of specific PRO items (Xia 2017)). More sophisticated approaches based on item response theory were also reported (Schindler 2018).

Interpretation, Conclusion or Significance: There are challenges associated with the inclusion of PROs in oncology early-phase trials such as burdening patients, difficulty in interpreting PRO data in single-arm trials but also skepticism of investigators towards the use of PROs endpoints at early-stage. But there are also opportunities to standardize the instruments and methods to analyze these endpoints in early-phase trials. The use of PROs in exposure/response analyses could also be a way to incorporate patient’s voice in dose selection in the context of Project Optimus. With a shift in the FDA’s regulatory expectations, there is a need to raise awareness and to recommend sponsors to incorporate PROs in oncology early-phase trials, and consider the approaches outlined in this poster to maximize PRO data utility in such trials.

Disclosures: