Related Experiment Video
Updated: Mar 19, 2026
Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances
Published on: October 11, 2018
SIMBA-a Bayesian decision framework for the identification of optimal biomarker subgroups for cancer basket clinical
Shijie Yuan1, Jiaxin Liu2, Zhihua Gong3
1Department of Statistics and Data Sciences, The University of Texas at Austin, Austin, TX 78712, United States.
Abstract:
Motivated by a multi-indication basket trial aiming to assess the efficacy of a novel biomarker-targeted therapy in gastric or gastroesophageal junction (G/GEJ), pancreatic, and other related cancers, we consider a statistical design and decision-making framework for such trials. Typically, the investigational therapy in the trial targets a biomarker that is present in multiple cancer indications, and patients with higher biomarker expression tend to exhibit higher response rates, assuming the targeting biomarkers are over-expressed in tumor cells. To enable information sharing across indications, the proposed SIMBA method introduces a Bayesian hierarchical model that defines positive and negative biomarker subgroups and identifies optimal go/no-go decisions. The operating characteristics of SIMBA are assessed via simulations and compared against existing methods in the literature. Overall, SIMBA is constructed to improve the identification of patient sub-populations who may benefit from biomarker-targeted therapeutics.
Insights
This study introduces the SIMBA method for biomarker-targeted cancer therapy trials. SIMBA improves identification of patient subgroups likely to benefit from novel therapeutics across multiple cancer types.
Area of Science:
- Oncology
- Biostatistics
- Clinical Trial Design
Background:
- Biomarker-targeted therapies show promise across multiple cancer types, including gastric/gastroesophageal junction (G/GEJ), pancreatic, and others.
- Patient response rates often correlate with biomarker expression levels, necessitating precise patient stratification.
- Existing trial designs may not fully leverage cross-indication data for targeted therapy assessment.
Purpose of the Study:
- To develop a statistical framework for multi-indication basket trials evaluating biomarker-targeted therapies.
- To introduce the SIMBA (Statistical Inference for Biomarker-guided Adaptive trials) method for improved decision-making.
- To enhance the identification of patient sub-populations who benefit from targeted therapeutics.
Main Methods:
- Proposed the SIMBA method, a Bayesian hierarchical model designed for information sharing across indications.
- Defined positive and negative biomarker subgroups within the model.
- Developed an optimal go/no-go decision-making framework based on the Bayesian model.
- Assessed SIMBA's operating characteristics through extensive simulations.
Main Results:
- SIMBA enables effective information sharing across different cancer indications.
- The method facilitates the identification of distinct biomarker subgroups.
- Simulations demonstrated SIMBA's advantages compared to existing statistical approaches.
- SIMBA supports informed go/no-go decisions in clinical trial progression.
Conclusions:
- The SIMBA method offers a robust statistical design for multi-indication basket trials.
- It improves the precision in identifying patient subgroups likely to respond to biomarker-targeted therapies.
- SIMBA has the potential to optimize the development of novel targeted therapeutics in oncology.
Related Concept Videos
01:23Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast,...
01:21Cancer Survival Analysis
02:50Combination Therapies and Personalized Medicine
The combination of the drug acetazolamide and sulforaphane is a good example of combination therapy to treat cancer. The cells in the interior of a large tumor often die due to the hypoxic and...