The impact of lymphocyte count dynamics on the predictive value of tumor mutational burden (TMB) for immune checkpoint inhibitors (ICI) outcomes in patients (pts) with cancer.

2663 Background: TMB has become a reliable biomarker for ICI response in several cancer types, leading to the pan-cancer FDA approval of the PD-1 inhibitor pembrolizumab in tumors with a TMB ≥10 mut/Mb. Lymphocyte count dynamics (lymphocyte stability LS) have been reported to be associated with both increased immune-related adverse events and improved overall survival (OS) on ICI. We aimed to investigate the role of LS as a risk stratification tool, in combination with TMB, in patients with cancer treated with ICIs. Methods: We identified 1215 pts from the Dana-Farber Cancer Institute, who had received ICI between 2015 and 2024. The change in relative blood lymphocyte counts was calculated between pre- (up to 30 days) and post-treatment (between 21 and 49 days). Lymphocyte counts were considered stable (LS ≥80%) if the drop in lymphocyte count did not exceed 20% after ICI-exposure. TMB was assessed from targeted panel sequencing and categorized into high and low, based on a cutoff of 10 mut/Mb. Overall survival was assessed using a multivariable Cox regression model, adjusting for several baseline characteristics. Results: The most prevalent cancer types were non-small cell lung cancer (n = 190), breast carcinoma (n = 160), glioma (n = 104), and melanoma (n = 100). Median TMB was 6.8 (IQR: 0–265.4). In total, 846 (69.6%) patients had LS (drop < 20%), while 369 (30.4%) did not. Higher TMB (≥10) and LS were both independently associated with better survival on ICI after adjusting for age, sex, tumor purity, line of therapy, ICI type and cancer type (HR 0.58, CI: 0.50–0.67; p < 0.001) and 0.75 (CI: 0.63–0.90; p = 0.002) respectively. Overall, patients with LS and high TMB demonstrated the longest median OS while the combination of unstable lymphocyte counts and low TMB was associated with the worst survival (Table). Conclusions: LS provides additional value, irrespective of TMB, in predicting response to ICI, suggesting distinct underlying immunological pathways. These findings emphasize the need for further research to validate LS and explore its integration into clinical decision-making in ICI-treated pts. Results from multivariable Cox regression. Groups Median OS (mo) N HR 95% CI P-value Stable lymphocytes (LS ≥80%) & High TMB (Cutoff = 10) 34.27 245 Ref. Ref. Ref. Stable (LS ≥80%) & Low TMB 18.99 601 1.3 1.04 - 1.63 0.019 Unstable (LS <80%) & High TMB 17.05 96 1.66 1.22 - 2.26 0.001 Unstable (LS <80%) & Low TMB 9.59 273 2.28 1.79 - 2.90 <0.001 Model is adjusted for age, sexe, tumor purity, lines of treatment, PD1/PDL1 therapy, and CTLA-4 therapy. Abbreviations: TMB: Tumor mutational burden; LS: Lymphocyte Stability; OS: Overall Survival; mo: Months.