Distinct prognostic value of circulating anti-telomerase CD4+ Th1 immunity and exhausted PD-1+/TIM-3+ T cells in lung cancer

Abstract

Background: Despite the critical roles of Th1-polarised CD4⁺ T cells in cancer immunosurveillance, the translation of their potential to clinical use remains challenging. Here, we investigate the clinical relevance of circulating antitumor Th1 immunity in non-small cell lung cancer (NSCLC).

Methods: The circulating antitumor Th1 response was assessed by the ELISpot assay in 170 NSCLC patients using a mixture of HLA class II-restricted peptides from telomerase (TERT). Phenotyping of blood immune cells was performed by flow cytometry.

Results: TERT-reactive CD4 T-cell response was detected in 35% of NSCLC patients before any treatment. Functional analysis showed that these cells were effector memory and Th1 polarised capable to produce effector cytokines, such as IFN-γ, TNF-α and IL-2. The presence of anti-TERT Th1 response was inversely correlated with the level of exhausted PD-1⁺/TIM-3⁺CD4 T cells. The level of these two immune parameters differentially affected the survival, so that increased level of anti-TERT Th1 response and low rate of exhausted PD-1⁺TIM-3⁺CD4⁺ T cells were associated with a better prognosis.

Conclusions: Systemic anti-TERT Th1 response plays a strong antitumor protective role in NSCLC. This study underlines the potential interest of monitoring circulating antitumor Th1 response for patients' stratification and therapy decision.

Fig. 1

Distribution and functional characterisation of TERT-specific CD4⁺ T cells in patients with NSCLC. a TERT-specific CD4⁺ T-cell responses were evaluated in 170 naïve-NSCLC patients by IFN-γ ELISpot assay performed after an in vitro stimulation of PBMC with HLA class II peptides derived from TERT. b, c Distribution of anti-TERT IFN-γ CD4⁺ T cells in NSCLC patients (n = 170), shown as the number of spots (b), and ratio of spots (c). Grey lines indicate the positivity thresholds, and blue lines indicate the median of spots calculated in responders patients. d Frequency of patients with negative (NEG) and positive (POS) anti-TERT Th1 responses. e–g Phenotypic and functional characterisation of anti-TERT CD4⁺ T cells detected by flow cytometry. e Dot plots of one representative patient show CCR7 and CD45RA and ICOS staining in IFN-γ⁻/IFN-γ⁺ CD4⁺ T cells. f Dot plots of one representative patient show CXCR3 and CCR6 staining in IFN-γ⁺ CD4⁺ T cells. g Dot plots of one representative patient show IFN-γ, TNF-α, IL-2, IL-4 and IL-17 cytokines production in response to TERT stimulation. The data are representative of three independent experiments

Fig. 2

Relationship between TERT-specific CD4⁺ Th1 response and blood immune factors in patients with NSCLC. a Heatmap illustrating hierarchical clustering (Euclidean distance) of 22 blood immune parameters (in rows) from NSCLC patients (n = 110) (in columns). b, c Unsupervised principal component analysis (PCA), including frequency (b) and magnitude (c) of anti-TERT Th1 response in relation to 21 blood immune parameters. d, e Representative dot plots (top row) show expressions of PD-1 and/or TIM-3 among CD4⁺ T cells (d) and CD8⁺ T cells (e). Histograms show peripheral T-cell expression levels of PD-1, TIM-3 and PD-1⁺ TIM-3⁺ among CD4⁺ T cells (d) and CD8⁺ T cells (e) from healthy donors (HD, n = 35) and NSCLC patients (n = 109). Median and interquartile range (IQR) are indicated (Mann–Whitney test). *P < 0.05; **P < 0.01. ns not significant

Fig. 3

Inverse correlation between the presence of anti-TERT CD4⁺ Th1 response and the level of exhausted PD-1⁺TIM-3⁺ T cells. a, b Levels of circulating PD-1⁺ and/or TIM-3⁺ CD4⁺ T cells (a) and CD8⁺ T cells (b) in patients with anti-TERT Th1 response (n = 49) and in non-responders (n = 96) (Mann–Whitney test). c, d Levels of circulating PD-1⁺ and/or TIM-3⁺ CD4⁺ T cells (c) and CD8⁺ T cells (d) in patients with antiviral Th1 response (n = 116) and in non-responders (n = 28) (Mann–Whitney test). Box spans indicate median and 25th–75th percentile, whiskers indicate the highest/lowest datapoints. e Dot plots show Ki-67 staining of unstimulated PBMC from one representative patient. f, g Blood lymphocytes from patients were stimulated with TERT-derived peptides with or without blocking mAb against PD-L1, PD-1 and/or TIM-3. TERT-specific T cells were measured by ICS or ELISpot. f Histograms show IFN-γ spot-forming cells from three representative patients. g In left, representative dot plot of TNF-α and IFN-γ-producing CD4⁺ T cells; In right, percentage of IFN-γ and TNF-α-secreting anti-TERT CD4 + T cells (n = 6). The data are representative of three independent experiments. *P < 0.05; **P < 0.01. ns not significant

Fig. 4

Distribution of circulating exhausted PD-1⁺/TIM-3⁺ T cells and anti-TERT Th1 response across NSCLC stages. a, b Levels of circulating PD-1⁺ and/or TIM-3⁺ CD4⁺ T cells (a) and CD8⁺ T cells (b) in localised NSCLC (stages I–III, n = 77) and metastatic NSCLC (stage IV, n = 68) (Mann–Whitney test). Box span indicates 25th–75th percentiles. Whiskers indicate the highest and lowest datapoints. c Frequency of circulating anti-TERT Th1 response in localised versus metastatic NSCLC (χ² test). d Ratio of anti-TERT IFN-γ spots to exhausted PD-1⁺/TIM-3⁺ CD4⁺ T cells in localised NSCLC (n = 77) and metastatic NSCLC (n = 68) (Mann–Whitney test). e Frequency of circulating IFN-γ antitumor Th1 response against WT-1, and NYESO-1, in localised versus metastatic NSCLC (χ² test). f Frequency of antiviral T-cell responses in localised NSCLC (n = 87) versus metastatic NSCLC (n = 83) (χ² test). Histograms indicate mean ± SD. *P < 0.05; **P < 0.01. ns not significant

Fig. 5

Prognostic value of systemic anti-TERT Th1 response and exhausted PD-1⁺/TIM-3⁺ CD4⁺ T cells in NSCLC. a–c Association between the level of circulating anti-TERT CD4 Th1 response and overall survival. A threshold (low < 3.7 < high) was defined based on the ratio of TERT-specific IFN-γ spots. Kaplan–Meier curves according to anti-TERT Th1 ratio: in all TERT responders (n = 59) (a), in localised stages (n = 39) (b) and in metastatic stages (n = 20) (c) (log-rank tests). d Association between the level of circulating PD-1⁺TIM-3⁺ CD4⁺ T-cell subsets and overall survival. Two groups were determined based on the median rate of exhausted PD-1⁺TIM-3⁺ CD4⁺ T-cell (0.9). Kaplan–Meier curves according to PD-1⁺TIM-3⁺ T cell (log-rank tests). e, f Patients were classified into distinct groups based on the anti-TERT CD4 Th1 ratio and the median level of PD-1⁺TIM-3⁺CD4⁺T cells. e, f Kaplan–Meier curves for the following groups: anti-TERT Th1^high/CD4⁺PD1⁺TIM3^low (green), anti-TERT Th1^high/CD4⁺PD1⁺TIM3^high (blue), anti-TERT Th1^low/CD4⁺PD1⁺TIM3^low (black), anti-TERT Th1^low/CD4⁺PD1⁺TIM3^high (red) (log-rank test). Patients in the “blue” and “dark” groups are pooled in (f). g Schema of the relationship between anti-TERT Th1 immunity- exhausted PD-1⁺TIM-3⁺ CD4⁺ T cells and NSCLC progression