To investigate the risk, risk factors, and dynamic changes in lymphocyte subsets and peripheral inflammatory cytokines of Pneumocystis jirovecii(PJ) infection in patients with advanced non-small cell lung cancer (NSCLC) receiving immune checkpoint inhibitors (ICIs) therapy.

This retrospective study included 34 patients with advanced NSCLC diagnosed with pulmonary PJ infection and treated with ICIs at our hospital between January 2019 and December 2024. Patients without pulmonary PJ infection during the same period served as the control group 47 patients. Peripheral blood mononuclear cells (PBMCs) were analyzed by cytometry by time-of-flight(CyTOF), including CD4⁺ and CD8⁺ T cell subsets and the expression of exhaustion markers T cell immunoglobulin and mucin domain-3 (TIM-3⁺ ), lymphocyte activation gene 3 (LAG-3⁺ ), and programmed death receptor 1 (PD-1⁺ ). Flow cytometry was used to detect peripheral blood T lymphocyte subsets at baseline (T0) and after one cycle of treatment (T1) in all subjects, and a series of serum cytokines were detected by enzyme-linked immunosorbent assay (ELISA).

Compared with the control group, the observation group exhibited significantly higher white blood cell counts [(6.50±2.10) × 109/L vs. (5.21±1.14) ×10⁹/L; t=3.249, P=0.002] and neutrophil-to-lymphocyte ratios (NLR) [(5.17±2.58) vs. (3.53±2.07); t=3.210, P=0.002]. Conversely, the observation group showed significantly lower lymphocyte counts [(1.28±0.56) ×10⁹/L vs. (1.55±0.67)×10⁹/L; t=-1.915, P=0.049], CD4⁺ T-cell counts [(420.54±102.65)×10⁶/L vs. (552.76±120.23)×10⁶/L; t=-5.170, P<0.001], and CD4⁺ /CD8⁺ T-cell ratios [(1.10±0.18) vs. (1.38±0.25); t=-5.910, P<0.001]. Multivariate analysis demonstrated that immunotherapy combined with chemotherapy (OR=1.910, P=0.011), negative PD-L1 expression (OR=2.333, P=0.033), NLR(OR=2.010, P=0.020), and a CD4⁺ /CD8⁺ T-cell ratio ≤1.0 (OR=2.578, P=0.036) were independent factors associated with an increased risk of pulmonary PJ infection. CyTOF analysis revealed that the relative abundances of exhausted CD8⁺ T cells and total CD8⁺ T cells in the control group were significantly higher than those in the observation group (Z-score difference of approximately 1.5~2.0), accompanied by the upregulated expression of CD366 (TIM-3), CD223 (LAG-3), and CD279 (PD-1) (Log2FC>1.0, P<0.1). Flow cytometry analysis demonstrated that PD-1⁺ CD8⁺ T-cell counts in the observation group significantly increased at T1 compared with T0 [(298.50±48.25)/μl vs. (418.29±67.43)/μl; t=-2.621, P=0.013]. Conversely, in the control group, significant decreases from T0 to T1 were observed in both LAG-3⁺ CD4⁺ T-cell counts [(685.50±104.83)/μl vs. (229.35±34.26)/μl; t=2.825, P=0.007] and PD-1⁺ CD8⁺ T-cell counts [(383.21±41.65)/μl vs. (108.21±37.43)/μl; t=2.083, P=0.043]. Analysis of cytokine levels revealed significant upward trends at T1 compared with T0 for serum interleukin-2 (IL-2) [(45.23±15.01) pg/ml vs. (135.12±20.21)pg/ml; t=-2.698, P=0.045], tumor necrosis factor-α(TNF-α) [(60.56±9.18) pg/ml vs. (186.25±53.47) pg/ml; t=-3.240, P=0.021], and interferon-γ(IFN-γ) [(50.56±10.34) pg/ml vs. (201.23±20.56) pg/ml; t=-3.274, P=0.019]. The fold changes (T1/T0) of these cytokines in the control group were significantly higher than those in the observation group. In contrast, interleukin-6 (IL-6) levels exhibited a downward trend, with a significantly smaller fold change observed in the control group compared with the observation group. With a median follow-up of 7.79 months, the median survival time of the observation group was significantly shorter than that of the control group (7.23 months vs. 14.12 months; P=0.047).

Reducing the number of exhausted T cells and increasing TNF-α may be potential mechanisms by which ICIs decreases susceptibility to pulmonary PJ infection. Exhausted T cells have potential predictive value for the risk of pulmonary PJ infection in patients with advanced NSCLC.