Biomarker-Based N-of-1 Trials for Addressing Patient Heterogeneity in Personalized Medicine

• Activity Number: 658
• Type: Contributed
• Date/Time: Thursday, August 7, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Yining Du*+ and Jack Lee
• Companies: MD Anderson Cancer Center and MD Anderson Cancer Center
• Keywords: N-of-1 ; meta-analysis ; bayesian ; hierarchical ; personalized ; biomarker

Abstract

By sharing the statistical strength, the key idea of the hierarchical model is that the inference about one unobserved quantity could affect inference about another unobserved quantity. This work is contributed to the analysis of the personalized treatment effects through biomarkers. We developed a statistical model in the hierarchical Bayesian scheme to estimate the treatment and biomarker effects. The Bayesian framework applying the hierarchical structure with adding the prior assumptions on the model parameters could be used to provide the estimations of parameters with limited data or sparse data. Furthermore, depending on the hierarchical model, we can borrow information across different biomarker groups within a treatment. Moreover, we applied the meta-analysis to combine the single patient (N-of-1) trials to investigate the personalized treatment effects. By providing more rigorous assessment of treatments effectiveness for an individual, single patient (N-of-1) trials offer a structured approach to provide the patient-focused and evidence-based treatment designs. Generally, such single-patient trials are designed to evaluate individual patient responses to the treatment(s)