The application of peripheral blood immune profiling in personalized treatment of locally advanced and advanced lung cancer: a nomogram approach

Purpose: Immunotherapy has revolutionized the treatment of lung cancer, yet many patients experience limited or transient benefits. Identifying those most likely to benefit remains a critical challenge. This study aims to establish a predictive model based on peripheral blood lymphocyte subsets to evaluate treatment responses in locally advanced and advanced lung cancer patients receiving chemotherapy with or without immunotherapy. Methods: We prospectively enrolled 171 patients, peripheral blood lymphocyte subsets were analyzed pre-treatment, post-treatment, and at disease progression using flow cytometry, focusing on CD3^–CD16⁺CD56⁺ cells, CD3^–CD19⁺ cells, CD3⁺CD4⁺ T cells, CD4⁺/CD8⁺ T-cell ratio, and CD3⁺CD8⁺ T cells. We assessed correlations between these subsets and treatment efficacy and constructed a nomogram to predict outcomes. Results: Baseline lymphocyte profiles were closely associated with treatment responses. Elevated CD3^–CD16⁺CD56⁺ cells, increased CD4⁺/CD8⁺ T cell ratio, and higher CD3^–CD19⁺ cells correlated with favorable treatment outcomes, particularly in patients receiving combined therapy. Conversely, higher CD3⁺ and CD3⁺CD8⁺ T cell counts were linked to poorer short-term efficacy. A nomogram integrating five immune parameters achieved an area under the receiver operating characteristic curve (AUC) of 0.778, outperforming individual marker. In the combination therapy subgroup, a four-parameter model achieved an AUC of 0.725. Furthermore, baseline and progression-stage lymphocyte profiles in responder and non-responder cohorts, exhibit no significant differences, indicating stable immune parameters over the disease course. Conclusion: Peripheral blood lymphocyte subsets are promising non-invasive biomarkers for predicting treatment responses in locally advanced and advanced lung cancer patients, particularly with immunotherapy. The developed nomogram models enhance predictive accuracy, supporting personalized treatment decisions.