Dissertation Literature Mastery Seminar – Uchechukwu Anyaduba, PhD Student – “Surrogate Endpoints in Oncology: Using Real-World Patient Data to Test and Improve Validation Methods”

Please join us for PhD student, Uchechukwu Anyaduba’s presentation on “Surrogate Endpoints in Oncology: Using Real-World Patient Data to Test and Improve Validation Methods”

Overall-Survival (OS) remains the gold standard endpoint used in trials to establish the clinical benefit of cancer treatments. For many cancers, OS requires that trials have an extended follow-up period to obtain a statistically-meaningful number of death events. Surrogate endpoints of OS might mediate this limitation. A surrogate endpoint is a biomarker known or expected to predict a clinical endpoint, with a reduced follow-up period required for the same degree of statistical power. A potential accelerated trial using a surrogate endpoint for death might provide an earlier opportunity for market approval while awaiting confirmatory OS data. However, the predictive accuracy of the surrogate for death must be demonstrated in validation studies, which depends on the: epidemiological, pathophysiological, therapeutic, and pharmacologic evidence provided and the overall benefit-to-risk to patients with the disease of interest.

This research aims to evaluate the existing validation framework (Prentices’ criteria) of meta-analysis of randomized trials. Subsequent investigation will explore alternative frameworks, such as Freedman’s Proportion Explained, decision tree or, machine learning-based prediction models for the joint assessment of DFS and OS in certain cancers to determine the best performing model for validation of surrogate endpoints for these cancers.

Date: November 29, 2021

Time: 12:00 – 1:00 p.m.

Location: 121 Bldg. – Classroom 1111 and Via Zoom (Zoom link will be on Outlook Invite)