Predictive biomarkers for immunotherapy response in renal cell carcinoma: why have they failed?
TL;DR
Current limitations in the development of consistent predictive biomarkers are outlined and future progress will require biomarker-stratified prospective trial designs with prespecified endpoints and rigorous external validation to establish clinically meaningful biomarkers.
Abstract
Purpose
OF REVIEW The heterogeneity of response to immunotherapies in renal cell carcinoma has created a strong need for predictive biomarkers to guide treatment decisions. However, no robust predictive biomarker has been adopted in clinical practice. This review aims to outline current limitations in the development of consistent predictive biomarkers. RECENT
Findings
Current biomarkers, including programmed cell death 1 ligand 1 expression, tumor mutational burden, blood-based biomarkers, and histopathological and molecular features, are primarily prognostic rather than predictive. Spatial and temporal tumor heterogeneity, methodological limitations, and trial design constraints are increasingly recognized as key factors limiting progress in the field. In addition, combination therapies further complicate biomarker interpretation. To better reflect the complexity and dynamic nature of renal cell carcinoma, biomarker development is increasingly exploring multiomics approaches, longitudinal assessments, and artificial intelligence-driven composite models. SUMMARY Establishing predictive biomarkers in clinical practice has not yet succeeded due to biological complexity, methodological and trial design limitations, and lack of sufficient validation. Single-biomarker approaches are insufficient to capture spatial and temporal heterogeneity and the impact of combination therapies. Future progress will require biomarker-stratified prospective trial designs with prespecified endpoints and rigorous external validation to establish clinically meaningful biomarkers.