Biomarker studies: a call for a comprehensive biomarker study registry

Area of Science:

• Oncology
• Biostatistics
• Biomarker Research

Background:

• Tumor biomarker studies offer insights into cancer biology and clinical behavior.
• Publication bias and hidden multiple hypothesis testing distort biomarker value assessment.
• Positive findings are preferentially reported, while negative results remain hidden, inflating biomarker utility.

Purpose of the Study:

• To address biases in tumor biomarker research.
• To propose a solution for more accurate biomarker evaluation.
• To accelerate the identification of clinically relevant biomarkers.

Main Methods:

• Review of existing biases in biomarker research, including publication bias and hidden multiple hypothesis testing.
• Proposal for the creation of a biomarker study registry.
• Suggestion for initial focus on studies using specimens from randomized trials, with future expansion to nonrandomized studies and raw data deposition.

Main Results:

• Biases lead to an overestimation of biomarker clinical validity and utility.
• Negative results are often not published, creating a skewed perception of biomarker effectiveness.
• A registry would facilitate a more balanced evaluation of biomarkers.

Conclusions:

• A dedicated biomarker study registry is crucial for mitigating biases.
• Such a registry would improve the reliability of biomarker research.
• Implementation could lead to more efficient discovery of impactful cancer biomarkers.