Fig. 3: Flow cytometry-based quantification of CD57⁺CD8⁺T cells predicts immunotherapeutic response in two prospective NSCLC cohorts.

a–c Flow cytometry-based quantification of T cells, CD8⁺ T cells, CD57⁺CD8⁺ T cells and CD57^-CD8⁺ T cells in the prospective NSCLC cohorts. d CD57⁺CD8⁺ T cells/T cells ratio (p < 0.001), e CD57⁺CD8⁺ T cells/ CD8⁺ T cells ratio (p = 0.001) and f total CD57⁺ T cells/T cells (p = 0.001) as determined by flow cytometry predicted clinical response to immunotherapy (n = 27). g Total CD8⁺ T cells/T cells ratio was not of predictive value (p = 0.059, n = 27). h ROC curve analyzed the ability of CD57⁺CD8⁺ T cells/T cells ratio to identify responders (AUC = 0.9375, n = 27). Sensitivity (87.5%) refers to the proportion of true positive subjects with the disease among all subjects with disease. Specificity (100.0%) refers to the proportion of true negative subjects without the disease among subjects without disease. PPV (84.6%) refers to the proportion of patients with positive results among subjects with positive results. NPV (100.0%) refers to the proportion of subjects without disease with a negative result among subjects with negative results. Accuracy (92.6%) refers to the proportion of subjects correctly classified among all subjects. i DCB proportion comparing the ability of CD57⁺CD8⁺ T cells/T cells ratio as determined by flow cytometry, total CD8⁺ T cells/T cells ratio by flow cytometry, and PD-L1 status by conventional IHC to predict treatment response. j ROC curves for predicting treatment response using the CD57⁺CD8⁺ T cells/T cells ratio (AUC = 0.733), CD8⁺ T cells/T cells ratio (AUC = 0.631), and PD-L1 positivity (AUC = 0.560) (n = 48). ***p ≤ 0.001. NSCLC non-small cell lung cancer, ROC receiver operating characteristic, AUC area under the curve, PPV positive predictive value, NPV negative predictive value, DCB durable clinical benefit, PD-L1 programmed death-ligand 1, IHC immunohistochemistry.