Construction and validation of a nomogram model for the differential diagnosis of primary testicular benign and malignant tumors

Abstract

Abstract Testicular tumors often pose a significant threat to the health of young adult males, and inappropriate radical orchiectomy imposes a heavy physical and psychological burden on many patients.In order to accurately distinguish testicular benign and malignant tumors prior to treatment, we retrospectively enrolled 298 patients with testicular tumors treated at the First Affiliated Hospital of Zhengzhou University between June 2016 and June 2024, all of whom met the inclusion and exclusion criteria. Using a random number method, the patients were divided into a training set (209 cases, 70%) and an internal validation set (89 cases, 30%). Additionally, 55 patients from two different medical centers, treated between January 2019 and January 2024 and meeting the same criter-ia, were included as an external validation set. Tumor size (OR = 2.70, 95% CI: 1.80–4.05, P  < 0.001), boundary (OR = 3.99, 95% CI: 1.34–11.86, P  = 0.013), AFP (OR = 7.49, 95% CI: 1.92–29.25, P  = 0.004), hCG (OR = 5.21, 95% CI: 1.30–20.95, P  = 0.020), and CDFI (Grade II: OR = 4.23, 95% CI: 1.35–13.26, P  = 0.014; Grade III: OR = 16.26, 95% CI: 3.68–71.83, P  < 0.001) were identified as five independent discriminative factors for distinguishing benign from malignant testicular tumors. A nomogram model incorporating these factors was constructed and demonstrated favorable performance. The model was deployed online as a dynamic nomogram ( https://nomogram98.shinyapps.io/dynnomapp/ ).