Predictive model based on multiple immunofluorescence quantitative analysis for pathological complete response to neoadjuvant immunochemotherapy in lung squamous cell carcinoma

Objective This study aims to establish a prediction model for neoadjuvant immunochemotherapy (NICT) in lung squamous cell carcinoma to guide clinical treatment.Methods This retrospective study included 50 patients diagnosed with lung squamous cell carcinoma who received NICT. The patients were divided into the pathological complete response (PCR) group and the non-PCR group. HE staining and multiple immunofluorescence (mIF) techniques were utilized to analyze the differences in the immune microenvironment between these groups. LASSO regression and optimal subset regression were employed to identify the most significant variables and construct a prediction model.Results The PCR group showed higher densities of lymphocyte nuclei and karyorrhexis based on HE staining. Furthermore, based on mIF analysis, the PCR group showed higher cell densities of CD8+, PD-L1+, and CD8+PD-L1+ in the tumor region, while showing lower cell densities of CD3+Foxp3+, Foxp3+, and CD163+. Logistic univariate analysis revealed CD8+PD-L1+, PD-L1+, CD8+, CD4+LAG-3+, lymphocyte nuclei, and karyorrhexis as significant factors influencing PCR. By using diverse screening methods, the three most relevant variables (CD8+, PD-L1+, and CD8+PD-L1+ in the tumor region) were selected to establish the prediction model. The model exhibited excellent performance in both the training set (AUC=0.965) and the validation set (AUC=0.786). In the validation set, In comparison to the conventional TPS scoring criteria, the model attained superior accuracy (0.85), specificity(0.67), and sensitivity (0.92).Conclusion NICT treatment might induce anti-tumor effects by enriching immune cells and reactivating exhausted T cells. CD8+, PD-L1+, and CD8+PD-L1+ cell abundances within the tumor region have been closely associated with therapeutic efficacy. Incorporating these three variables into a predictive model allows accurate forecasting of treatment outcomes and provides a reliable basis for selecting NICT treatment strategies.