Combinatory analysis of immune cell subsets and tumor-specific genetic variants predict clinical response to PD-1 blockade in patients with non-small cell lung cancer

Abstract

Objectives: Immunotherapy by blocking programmed death protein-1 (PD-1) or programmed death protein-ligand1 (PD-L1) with antibodies (PD-1 blockade) has revolutionized treatment options for patients with non-small cell lung cancer (NSCLC). However, the benefit of immunotherapy is limited to a subset of patients. This study aimed to investigate the value of combining immune and genetic variables analyzed within 3-4 weeks after the start of PD-1 blockade therapy to predict long-term clinical response.

Materials and methodology: Blood collected from patients with NSCLC were analyzed for changes in the frequency and concentration of immune cells using a clinical flow cytometry assay. Next-generation sequencing (NGS) was performed on DNA extracted from archival tumor biopsies of the same patients. Patients were categorized as clinical responders or non-responders based on the 9 months' assessment after the start of therapy.

Results: We report a significant increase in the post-treatment frequency of activated effector memory CD4⁺ and CD8⁺ T-cells compared with pre-treatment levels in the blood. Baseline frequencies of B cells but not NK cells, T cells, or regulatory T cells were associated with the clinical response to PD-1 blockade. NGS of tumor tissues identified pathogenic or likely pathogenic mutations in tumor protein P53, Kirsten rat sarcoma virus, Kelch-like ECH-associated protein 1, neurogenic locus notch homolog protein 1, and serine/threonine kinase 11, primarily in the responder group. Finally, multivariate analysis of combined immune and genetic factors but neither alone, could discriminate between responders and non-responders.

Conclusion: Combined analyses of select immune cell subsets and genetic mutations could predict early clinical responses to immunotherapy in patients with NSCLC and after validation, can guide clinical precision medicine efforts.

Keywords: KRAS; PD-1; TP53; effector memory T cells; non-small cell lung cancer.

Figure 1

Baseline frequencies of circulating B cells, NK cells and T cells in association with the clinical response to PD-1 blockade. NSCLC patients received PD-1 blockade at 2-3 week cycles and blood was drawn at baseline for later flow cytometry analysis. (A) Kaplan-Meier curves of progression-free survival in responders and non-responders at the 9 months cutoff. (B) Frequencies of B cells, NK cells and CD3⁺ T cells (C) Frequencies of CD4⁺ and CD8⁺ T cells and regulatory T cells with memory phenotype (CD3⁺CD4⁺CD25^highCD127^lowCD45RO⁺CD194⁺) in responder and non-responder patients as defined in the materials and methods section. Patient treated with anti-PD-L1 marked with red symbols. Dotted lines represent median frequencies of respective immune cell subsets for n=3 healthy individuals analyzed on one occasion.

Figure 2

Changes in frequencies of effector and memory T cell populations in the blood of responders and non-responders after PD-1 blockade. NSCLC patients received PD-1 blockade at 2-3 week cycles and blood was drawn at pre-(A) compared post-1^st treatment cycle before for later flow cytometry analysis. (A) The frequencies of activated effector memory T cells (CD3⁺CD4⁺/CD8⁺CD45RA^-CCR7^-CD38⁺HLA-DR⁺) in circulation of responders and non-responders. (B) The frequencies of central memory T cells (CD3⁺CD4⁺CD8⁺CD45RA^-CCR7⁺) in circulation of responders and non-responders. (C) The frequencies of effector T cells (CD3⁺CD4⁺/CD8⁺CD45RA^intCCR7^-) in circulation of responders and non-responders. Patient treated with anti-PD-L1 marked with red symbols.

Figure 3

Analysis of immune and genetic variables measured and their relation to clinical outcome. (A) Kaplan-Meier curves estimates comparing overall survival of patients based on KRAS or KRAS and TP53 mutational status of the tumor (B) Principal component analysis of variables measured (baseline frequencies of B cells, NK cells, CD3⁺ T cells, CD4⁺ and CD8⁺ T cells and Tregs, pre- and post- 1st cycle of treatment, fold-change activated effector memory T cells, central memory T cells, effector T cells and Tregs, gene score, PD-L1 status and TMB score). Large symbols, green (responder) and blue (non-responders) indicate weighted means of the groups (C) Score scatter plot and (D) loading column plot from an orthogonal partial least squares‐discriminant analysis (OPLS‐DA) of the X variables in (B), using a VIP cutoff > 0.8 and VIPcvSE cutoff <1.5, and Y variables (responders/non-responders). R2Y defines the goodness of fit, and Q2 the goodness of prediction. Total number of patients for the analysis (n=14) with responders (n=10) and non-responders (n=4).