Is molecular profiling ready for use in clinical decision making?

Area of Science:

• Oncology
• Genomics
• Biomarkers

Background:

• Molecular profiling, analyzing gene expression for cancer diagnosis and prognosis, offers potential for personalized patient management.
• Current clinical application of molecular profiling faces significant hurdles in assay standardization, data analysis, and bias mitigation.

Purpose of the Study:

• To critically evaluate the current state and challenges of molecular profiling in cancer management.
• To highlight the need for robust validation and evidence before widespread clinical adoption.

Main Methods:

• Review of current practices in molecular profiling assay development and data analysis.
• Assessment of classifier performance reporting, validation methodologies, and evidence for incremental value over traditional risk factors.
• Examination of challenges in clinical integration, utility, and cost-effectiveness.

Main Results:

• Assay development and data analysis are largely exploratory, prone to bias, and lack standardization.
• Classifier performance is often overinterpreted, with modest absolute sensitivity and specificity, especially in large validation cohorts.
• Limited evidence exists for the incremental benefit of molecular profiles beyond classic risk factors, except in breast cancer.

Conclusions:

• Molecular profiling is a promising technology, but its clinical utility and net benefit remain largely unproven outside specific contexts like breast cancer.
• Careful planning, robust evidence from clinical trials, and addressing standardization and validation issues are crucial for optimal incorporation into clinical practice.
• Cost-effectiveness cannot be fully assessed until these fundamental challenges are resolved.