Cell-based therapies such as CAR T cells, tumor-infiltrating lymphocytes, and mesenchymal stromal cells elicit highly variable immune responses. Predictive biomarkers fall into patient and tumor characteristics, features of the therapeutic product, and early immune-system signals that together indicate likely benefit or toxicity. Evidence from clinical groups has identified reproducible signals clinicians use to stratify risk and guide management.
Predictive patient and tumor biomarkers
Baseline tumor burden and circulating tumor-derived DNA are strong predictors of both efficacy and toxicity because higher antigen load drives greater cellular expansion and inflammatory responses. Sattva Neelapu MD Anderson Cancer Center has documented relationships between tumor burden and severity of cytokine-mediated toxicities in CAR T programs. Preexisting immune contexture including levels of regulatory T cells and myeloid-derived suppressor cells in the tumor microenvironment influences engraftment and function of infused cells. Genetic diversity in HLA and population differences in immune repertoires can modulate antigen presentation and response rates, which matters for equitable outcomes across territories.Product and immune biomarkers
Characteristics of the manufactured cell product predict durability. Carl H. June University of Pennsylvania reports that greater peak expansion and long-term persistence of CAR T cells correlate with durable remissions. Composition skewed toward memory-like T cell phenotypes rather than highly differentiated effector cells is associated with superior persistence and is emphasized in work by Michel Sadelain Memorial Sloan Kettering Cancer Center. Early post-infusion dynamics provide pragmatic biomarkers: rapid rises in interleukin-6, interleukin-15, and acute-phase reactants such as C-reactive protein and ferritin reliably herald cytokine release syndrome and inform interventions, as described by Neelapu MD Anderson Cancer Center. Conversely, upregulation of exhaustion markers including PD-1, TIM-3, and LAG-3 on infused cells predicts functional decline and failure to control disease.Understanding causes and consequences of these biomarkers is clinically actionable. Predictive signatures permit preemptive toxicity mitigation, patient selection for alternative strategies, and adaptive manufacturing to enrich favorable cell subsets. Resource-limited settings face additional challenges because biomarker-driven personalization depends on advanced laboratory capacity and consistent cold-chain logistics, potentially widening access gaps. As trials increasingly incorporate circulating tumor DNA, immune phenotyping, and product analytics, these biomarkers form the evidence base that regulators and clinicians use to optimize safety and long-term benefit.