Abstract
BACKGROUND: Pulmonary invasive mucinous adenocarcinoma (IMA), a rare subtype of lung adenocarcinoma, exhibits distinct clinicopathological features. However, its prognostic determinants remain poorly understood due to its low incidence and the limited availability of longitudinal survival data. Diagnostic challenges complicate the management of patients with IMA, as it often mimics inflammatory or infectious pulmonary lesions. At the same time, the options for therapy are limited, as conventional systemic therapies frequently yield suboptimal outcomes. These issues highlight the urgent need for systematic investigations into survival-associated factors and the development of precision-driven prognostic tools. To address these deficiencies, we conducted a comprehensive analysis of clinical characteristics and prognostic factors that influence overall survival (OS) in patients with IMA. Additionally, we developed and validated a multivariable-based nomogram to facilitate individualized survival prediction, aiming to improve risk stratification and guide personalized clinical decision-making for this understudied malignancy. METHODS: A cohort of 310 patients with IMA from the First Affiliated Hospital of Zhengzhou University served as the training group, while data from 378 patients in the Surveillance, Epidemiology, and End Results (SEER) cohort were used for external validation. Survival analysis was performed to determine prognostic factors using univariate and multivariate Cox regression. A nomogram was constructed to estimate the 1-, 3-, and 5-year survival rates. The model's performance was evaluated using the concordance index (C-index), receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). RESULTS: Significant prognostic factors identified in the training cohort included age [70-79 years: hazard ratio (HR) =3.849; P=0.002], non-smoking (HR =0.334; P=0.02); advanced T stage (T4: HR =4.998; P=0.003), metastatic disease (M1: HR =2.073; P=0.02), and absence of surgery (HR =2.731; P=0.005). The observed 1-, 3-, and 5-year survival rates were 84.1%, 67.0%, and 51.5%, respectively. The nomogram exhibited high predictive accuracy, with a C-index of 0.879 and area under the curve (AUC) values of 0.897, 0.924, and 0.856 for predicting 1-, 3-, and 5-year survival rates, respectively. Calibration plots showed excellent concordance between predicted and actual outcomes, and DCA confirmed the model's clinical utility. The results from the validation cohort corroborated the model's robustness. CONCLUSIONS: This study identified several key prognostic factors associated with OS in patients with IMA and developed a robust nomogram for personalized survival prediction. Future multicenter studies should aim to incorporate molecular biomarkers and advanced imaging to further enhance the model's clinical utility.