A Population-Based Nomogram for Predicting Overall Survival in Pulmonary Sarcomatoid Carcinoma With Real-World...

Area of Science:

• Oncology
• Thoracic Surgery
• Cancer Research

Background:

• Pulmonary sarcomatoid carcinoma (PSC) is a rare, aggressive thoracic malignancy.
• Limited understanding of prognostic factors and survival prediction in PSC.

Purpose of the Study:

• Investigate prognostic factors for overall survival (OS) in PSC.
• Develop and validate a predictive risk classifier (nomogram) for PSC patient OS.

Main Methods:

• Retrospective analysis of 765 PSC patients (SEER database and single institution).
• Univariate/multivariate Cox regression and Lasso-Cox models for prognostic factor identification.
• Nomogram construction and validation using ROC curves, C-index, calibration plots, and DCA.

Main Results:

• Identified independent prognostic factors for PSC OS: sex, TNM stage, surgery, radiation, chemotherapy, liver metastasis.
• Developed nomogram demonstrated excellent predictive performance and clinical utility.
• Lasso-Cox revealed surgery and immunotherapy as protective; ECOG status, comorbidities, and distant metastasis as risk factors.

Conclusions:

• The developed nomogram reliably predicts individual OS probability for PSC patients.
• Immunotherapy is associated with significantly improved survival in real-world PSC cohorts.