Glioblastoma is the most common and aggressive primary brain tumor in adults that fails to meet survival endpoint criteria in current immunotherapy Phase III clinical trials. As a small subset of immunotherapy-responsive patients has been identified, biomarkers that can nominate patients for immunotherapy and inform treatment combinations have been increasingly urgent. Mechanistically-driven single biomarkers (e.g. PD-1/PD-L1 enrichment, tumor mutational burden, and MGMT methylation) are early opportunities for immunotherapy response prediction but have had inconsistent results across studies. In contrast, new context-dependent biomarker signals such as interferon signaling, immune cell population studies, and radiographic biomarkers show promise in vaccine-based therapies and in immune checkpoint blockade. Ongoing clinical trials are increasingly implementing early exploratory endpoints for biomarker discovery from several body compartments (e.g. CSF, blood, tissue). Present advancements in machine learning and liquid biopsy have provided nuanced biomarker signatures that can be composed of several biomarker categories. Progress in standardizing biomarkers in glioblastoma immunotherapy relies on the early implementation and uniformity of biomarker-driven endpoints in clinical trials.
Kim et al. (Tue,) studied this question.