Abstract
OBJECTIVES: Primary tumor resection has been associated with potential enhanced survival in select patients with stage IV metastatic lung squamous cell carcinoma (mLUSCC). However, identifying suitable candidates for surgery remains challenging. This study aimed to develop and validate a prognostic model to select mLUSCC patients likely to benefit from primary tumor resection. METHODS: Data on stage IV mLUSCC patients were retrieved from the SEER database. Patients were divided into surgery and nonsurgery groups based on primary tumor resection. Propensity score matching (PSM) using covariates including age, T/N stage, grade, metastases, chemotherapy, and others mitigated selection bias. Survival was assessed using Kaplan-Meier curves and Cox proportional hazards models. The surgery group was split randomly (70:30) into training and validation cohorts for model development via multivariate Cox regression. RESULTS: Among 18,701 eligible patients, 3.57% (n = 668) underwent primary tumor resection. Post-PSM, median overall survival (mOS) was superior in the surgery group (11 [95% CI: 9–12] vs. 6 [5–6] months; P < 0.0001). The model demonstrated good discriminative ability, with AUCs for 6-, 9-, and 12-month survival of 0.799, 0.746, and 0.744 in the training cohort, and 0.708, 0.727, and 0.667 in the validation cohort. Calibration plots showed excellent agreement between predicted and observed probabilities. CONCLUSIONS: This internally validated prognostic nomogram aims to identify patients with stage IV metastatic lung squamous cell carcinoma who are most likely to benefit from primary tumor resection, thereby supporting individualized multidisciplinary decision-making. Further external and prospective validation is warranted.