Related Experiment Video
Updated: Jun 2, 2026

Selecting Multiple Biomarker Subsets with Similarly Effective Binary Classification Performances
Published on: October 11, 2018
Comparing biomarkers as principal surrogate endpoints.
1Fred Hutchinson Cancer Research Center, Vaccine & Infectious Disease Division, Seattle, Washington 98109, USA. yhuang@fhcrc.org
This study introduces a new way to evaluate principal surrogates, which are biomarkers predicting clinical outcomes. The methods allow comparing multiple biomarkers and risk models, advancing surrogate endpoint research.
Area of Science:
- Biostatistics
- Clinical Trials Methodology
- Epidemiology
Background:
- A new definition of principal surrogate endpoints, based on causal inference, has been proposed.
- Current methods for evaluating principal surrogates are limited, focusing on single biomarkers and specific risk models.
- Comparing the principal surrogate value of multiple biomarkers or complex risk models remains an open research question.
Purpose of the Study:
- To propose a novel framework for characterizing and comparing the principal surrogate value of biomarkers and risk models, especially those involving multiple biomarkers.
- To introduce a new summary measure, the standardized total gain, for comparing markers and assessing incremental value.
- To develop a flexible statistical method for estimating the joint surrogate value of multiple biomarkers.
Main Methods:
- Characterizing principal surrogate value using the distribution of risk differences between interventions.
- Developing a semiparametric estimated-likelihood method for estimating the joint surrogate value of multiple biomarkers.
- Accommodating two-phase biomarker sampling and incorporating continuous baseline covariates for broader applicability and robustness.
Main Results:
- The proposed methods provide a way to quantify and compare the principal surrogate value of individual markers and multi-marker risk models.
- The standardized total gain offers a metric for assessing the comparative and incremental value of biomarkers.
- The developed semiparametric method is shown to be applicable in settings with two-phase sampling and continuous covariates.
Conclusions:
- The proposed methodology extends the evaluation of principal surrogates to complex scenarios involving multiple biomarkers and risk models.
- The novel summary measure facilitates informed decisions about biomarker selection and development.
- The methods are validated using both simulated and real-world data, demonstrating their utility in contexts like HIV vaccine trials.
Related Concept Videos
Types 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, controlled...
Comparing the Survival Analysis of Two or More Groups
Bioequivalence of Drugs: Drugs with Multiple Indications
Blood Studies for Cardiovascular System II: CRP, Hcy, and Cardiac Natriuretic Peptide Markers
These markers indicate stress or strain on the heart muscle:
Natriuretic Peptides (BNP)
Cardiac myocytes produce these hormones in response to ventricular stretching...
Drug Products: Biologics, Biosimilars and Interchangeables
Measurement of Bioavailability: Pharmacodynamic Methods
