Accurate prognostic assessment of primary pulmonary lymphoepithelioma-like carcinoma (PPLELC) remains challenging. We analyzed clinicopathological data from PPLELC patients across three centers to identify factors associated with overall survival, which were incorporated into a prognostic nomogram. The model was developed in a training cohort (70%) and internally validated using a split-sample validation cohort (30%) and bootstrap resampling. Model performance was evaluated using the concordance (C-index), time-dependent area under the receiver-operating characteristic curve (AUC), and calibration curves. Comparative performance with American Joint Committee on Cancer (AJCC) staging system was assessed using net reclassification improvement (NRI), integrated discrimination improvement (IDI), and decision curve analysis. Risk stratification was conducted using X-tile. A total of 526 patients were included and randomly assigned to the training (n = 368) and validation (n = 158) cohorts. Six prognostic factors were incorporated into the nomogram. The internally validated model demonstrated good discriminative ability, with C-index values of 0.827 in the training cohort and 0.803 in the validation cohort. Time-dependent AUCs for 1-, 3-, and 5-year overall survival were 0.944, 0.852, and 0.868 in the training cohort, and 0.750, 0.884, and 0.868 in the validation cohort, respectively. When compared with the AJCC staging system, the nomogram showed improved risk discrimination as reflected by NRI, IDI, and changes in the C-index. Risk stratification based on the nomogram effectively separated patients into distinct prognostic groups. This study presents an internally validated prognostic nomogram for estimating overall survival in patients with PPLELC. The model may serve as a complementary tool to conventional staging systems to support individualized prognostic assessment and clinical communication.
Wang et al. (Sun,) studied this question.