ABSTRACT Objectives To develop efficacious assessment tools to individualize the evaluation of overall survival (OS) and cancer‐specific survival (CSS) in patients with testicular cancer. Methods A total of 30,689 patients diagnosed with testicular cancer between 2004 and 2021 were selected from the Surveillance, Epidemiology, and End Results database. The study population was randomly divided into a training cohort and a validation cohort. Univariate and multivariate Cox analyses were conducted to identify significant predictors, which were subsequently utilized to construct nomograms for predicting 1‐, 3‐, 5‐, and 10‐year OS and CSS. The predictive performance of the nomograms underwent internal and external testing with the application of the concordance index (C‐index), receiver operating characteristic curves, and calibration curves. We developed a prognostic scoring system based on the coefficients within the Cox models for each subgroup. Results The significant predictors included age, race, marital status, TNM stage, radiation, chemotherapy, surgery and pathology. Age emerged as the most potent factor associated with overall and cancer‐specific death (≥ 60 vs. < 30 years old: HR = 12.19 for OS and HR = 5.94 for CSS, p < 0.001). Among all pathological subtypes, choriocarcinoma exhibited the worst OS and CSS (reference seminoma: HR = 2.79 for OS and HR = 5.02 for CSS, p < 0.001). The favorable internal validation (C‐index: 0.799 for OS and 0.859 for CSS; area under the curve = 0.773–0.892), external validation (C‐index: 0.784 for OS and 0.867 for CSS) and calibration curves indicated the nomograms possessed good predictive ability. We developed a prognostic scoring system for the first time, which is more accurate than the traditional TNM system in evaluating patients' survival outcomes. Conclusion The prognostic nomograms and scoring systems are capable of effectively evaluating the 1‐, 3‐, 5‐, and 10‐year OS and CSS of testicular cancer patients and providing a reliable tool for optimizing clinical treatment decisions and follow‐up management.
Huang et al. (Fri,) studied this question.