Longitudinal peripheral blood TCR tracking predicts response to immune checkpoint inhibitors

Background: While immune checkpoint inhibitor (ICI) therapy has revolutionized oncology, reliable biomarkers to predict clinical response are critically needed. Static, single-time-point analysis of the T-cell receptor (TCR) repertoire has proven inadequate. We hypothesized that the longitudinal tracking of TCR clonal dynamics would provide a more powerful and predictive biomarker of clinical outcome. Methods: To test this hypothesis, we employed a two-stage strategy. First, we performed an exploratory analysis on the public longitudinal TCR-seq dataset GSE212217 to investigate whether dynamic changes in TCR clones could distinguish responders from non-responders. Based on the observed temporal expansion patterns, we developed a novel classification model that categorizes TCR clones algorithmically by quantifying the direction and magnitude of change in their relative abundance over time. Clones are assigned to distinct behavioral types, Response, Super-response, Transient, and Quiescent, based on specific, quantitative criteria applied to their longitudinal trajectories. Second, we validated this model in an independent, prospective pilot cohort of advanced melanoma patients using TCR sequencing from serial blood samples. Results: Analysis of the public dataset confirmed that baseline TCR diversity metrics were unable to differentiate between clinical responders and non-responders. In contrast, our longitudinal model identified distinct clonal expansion trajectories. This novel response score was significantly elevated in responders compared to non-responders in the discovery cohort (Wilcoxon test, P = 0.011). Critically, patients with a high score exhibited a marked survival advantage (Log-rank P = 0.040), confirming the clinical prognostic value of our dynamic metric. Validation in our independent melanoma cohort confirmed the clinical utility of our model. Strikingly, the patient with disease progression exhibited a low proportion of active, expanding clones (5.5%), whereas the patient achieving a deep clinical response demonstrated a markedly higher proportion of active, expanding clones (10.5%). Conclusion: We have developed and validated a novel biomarker based on the dynamic expansion of T-cell clones that effectively predicts ICI response and survival. This work definitively moves beyond the limitations of static TCR metrics, establishing longitudinal clonal tracking as a crucial strategy for liquid biopsy. This approach provides a critical tool for real-time response monitoring and precision stratification of cancer immunotherapy.