Development and Validation of A Diagnostic Nomogram Based on PET/CT and Clinical Characteristics: Differentiating mediastinal lymphomas from aggressive thymic epithelial tumors

Purpose This study aims to establish and validate diagnostic nomogram integrating the PET/CT metabolic parameter with clinical features, to differentiate between aggressive thymic epithelial tumors(aTETs) and primary mediastinal lymphomas(PMLs). Methods A retrospective evaluation was conducted of 376 patients with anterior mediastinal masses treated at Peking University Cancer Hospital from January 2018 to December 2022. A total of 114 patients diagnosed pathologically with aTETs or PMLs were included. Of these, 79 patients were allocated to the training set and 35 to the validation set, in a ratio of approximately 7:3. Clinical and PET/CT data were collected. Univariate analysis and multivariate logistic regression analysis were conducted to identify clinical predictive factors and create diagnostic nomogram. Receiver Operating Characteristic (ROC)curves, calibration plots, and Decision Curve Analysis (DCA) were generated. The predictive accuracy and clinical value of the model were comprehensively evaluated using the validation set. Results Univariate analysis and logistic regression identified three key parameters associated with differentiating aTETs from PMLs, age(OR 0.75, 95%CI 0.65~0.87,P＜0.001),location(OR 35.66, 95%CI 2.44~521.98, P=0.009), and maximum standard uptake value(SUV max )(OR 1.30, 95%CI 1.10~1.54, P=0.002). The ROC of the nomogram for the training and validation sets were 0.981 and 0.977, respectively, demonstrating excellent predictive capability and appropriate performance. Based on DCA, the clinical prediction model shows strong clinical utility in distinguishing between aTETs and PMLs. Conclusion Diagnostic nomogram based on SUV max , age and location has been established and validated, which can be used to differentiate between patients with aTETs and PMLs.