Combination of two distinct subsets of peripheral blood CD8+T cells from patients with NSCLC predicts response outcome to immune checkpoint inhibitor therapy

BackgroundImmune checkpoint inhibitor (ICI) has achieved a great success as a promising regime for the treatment of patients with many types of solid malignancies, associated with predictive biomarker of PD-L1 expression. However, generally a low rate of ICI therapy response remains a critical hurdle to overcome for expanding its versatile therapeutic efficacy and necessitates importance of developing a biomarker better predicting response outcome after ICI.MethodsPeripheral blood CD8+ T cell compartment from patients with stage IV of non-small cell lung cancer (n=121) before ICI treatment targeting programmed cell death 1 (PD-1) or its ligand 1 (PD-L1) was analyzed and correlated with patients’ ICI treatment outcome.ResultsStrong correlation between the patients' response outcome after ICI and the proportion of two distinct subsets of blood CD8+ T cells, namely CD27+ CD28+ CD45RA- CCR7- and CD27+ CD28+ CD45RA+ CCR7- cells was observed. Using these two cellular parameters combined with machine learning based probability graph, we found that both initial discovery (n=72) and later validation cohorts of patients (n=49) showed a power of predicting approximately 55.6% of ICI responders (95% CI, 23.1-88%; PR based on the RECISTv1.1 criteria) compared to that of ∼15.3% (95% CI, 7-23.6%; no biomarker included) and of ∼22.2% (95% CI, 0-49.3%; biomarker based on tumor PD-L1 expression). Mechanistically, the observed strong correlation was due to the basal functional fitness and reactive capacity of responding CD8+ T cells, which was characterized by lower initial levels of perforin, granzyme B and interferon-γ expression in these cells before ICI treatment. As such, the patients with enhanced proportion of this subset in their bloods showed greater capacity to enhance ICI-driven upregulation of cytotoxic molecules and accordingly better ICI response outcomes.ConclusionsThese observations are in line with current notion that a relatively less differentiated subset of CD8+ T cells would be a major target for ICI and provide a potential of developing this non-invasive blood-based approach as an ICI response predictor for patients with cancer.Legal entity responsible for the studyChonnam National University Medical School and Chonnam National University Hwasun Hospital.FundingHas not received any funding.DisclosureAll authors have declared no conflicts of interest. BackgroundImmune checkpoint inhibitor (ICI) has achieved a great success as a promising regime for the treatment of patients with many types of solid malignancies, associated with predictive biomarker of PD-L1 expression. However, generally a low rate of ICI therapy response remains a critical hurdle to overcome for expanding its versatile therapeutic efficacy and necessitates importance of developing a biomarker better predicting response outcome after ICI. Immune checkpoint inhibitor (ICI) has achieved a great success as a promising regime for the treatment of patients with many types of solid malignancies, associated with predictive biomarker of PD-L1 expression. However, generally a low rate of ICI therapy response remains a critical hurdle to overcome for expanding its versatile therapeutic efficacy and necessitates importance of developing a biomarker better predicting response outcome after ICI. MethodsPeripheral blood CD8+ T cell compartment from patients with stage IV of non-small cell lung cancer (n=121) before ICI treatment targeting programmed cell death 1 (PD-1) or its ligand 1 (PD-L1) was analyzed and correlated with patients’ ICI treatment outcome. Peripheral blood CD8+ T cell compartment from patients with stage IV of non-small cell lung cancer (n=121) before ICI treatment targeting programmed cell death 1 (PD-1) or its ligand 1 (PD-L1) was analyzed and correlated with patients’ ICI treatment outcome. ResultsStrong correlation between the patients' response outcome after ICI and the proportion of two distinct subsets of blood CD8+ T cells, namely CD27+ CD28+ CD45RA- CCR7- and CD27+ CD28+ CD45RA+ CCR7- cells was observed. Using these two cellular parameters combined with machine learning based probability graph, we found that both initial discovery (n=72) and later validation cohorts of patients (n=49) showed a power of predicting approximately 55.6% of ICI responders (95% CI, 23.1-88%; PR based on the RECISTv1.1 criteria) compared to that of ∼15.3% (95% CI, 7-23.6%; no biomarker included) and of ∼22.2% (95% CI, 0-49.3%; biomarker based on tumor PD-L1 expression). Mechanistically, the observed strong correlation was due to the basal functional fitness and reactive capacity of responding CD8+ T cells, which was characterized by lower initial levels of perforin, granzyme B and interferon-γ expression in these cells before ICI treatment. As such, the patients with enhanced proportion of this subset in their bloods showed greater capacity to enhance ICI-driven upregulation of cytotoxic molecules and accordingly better ICI response outcomes. Strong correlation between the patients' response outcome after ICI and the proportion of two distinct subsets of blood CD8+ T cells, namely CD27+ CD28+ CD45RA- CCR7- and CD27+ CD28+ CD45RA+ CCR7- cells was observed. Using these two cellular parameters combined with machine learning based probability graph, we found that both initial discovery (n=72) and later validation cohorts of patients (n=49) showed a power of predicting approximately 55.6% of ICI responders (95% CI, 23.1-88%; PR based on the RECISTv1.1 criteria) compared to that of ∼15.3% (95% CI, 7-23.6%; no biomarker included) and of ∼22.2% (95% CI, 0-49.3%; biomarker based on tumor PD-L1 expression). Mechanistically, the observed strong correlation was due to the basal functional fitness and reactive capacity of responding CD8+ T cells, which was characterized by lower initial levels of perforin, granzyme B and interferon-γ expression in these cells before ICI treatment. As such, the patients with enhanced proportion of this subset in their bloods showed greater capacity to enhance ICI-driven upregulation of cytotoxic molecules and accordingly better ICI response outcomes. ConclusionsThese observations are in line with current notion that a relatively less differentiated subset of CD8+ T cells would be a major target for ICI and provide a potential of developing this non-invasive blood-based approach as an ICI response predictor for patients with cancer. These observations are in line with current notion that a relatively less differentiated subset of CD8+ T cells would be a major target for ICI and provide a potential of developing this non-invasive blood-based approach as an ICI response predictor for patients with cancer.