Construction of a novel clinical nomogram to predict cancer-specific survival in patients with primary malignant adrenal tumors: a large population-based retrospective study

Background: Primary malignant adrenal tumors were rare and had a poor prognosis. This investigation aimed to create a useful clinical prediction nomogram to anticipate cancer-specific survival (CSS) of patients with a primary malignant adrenal tumor. Method: This study included 1748 patients with malignant adrenal tumor diagnoses subjects from 2000 to 2019. These subjects were allocated randomly into training (70%) and validation (30%) cohorts. Patients with adrenal tumors underwent univariate and multivariate Cox regression analyses to identify the CSS-independent predictive biomarkers. Therefore, a nomogram was created depending on those predictors, and calibration curves, receiver operating characteristic (ROC) curves, and decision curve analysis (DCA) were used to assess the calibration capacity of the nomogram, discriminative power, and clinical efficiency, respectively. Afterward, a risk system for categorizing patients with adrenal tumors was established. Result: The univariate and multivariate Cox analysis demonstrated the CSS-independent predictive factors, including age, tumor stage, size, histological type, and surgery. As a result, a nomogram was developed using these variables. For the 3-, 5-, and 10-year CSS of this nomogram, the values of the area under the curve (AUC) of the ROC curves were 0.829, 0.827, and 0.822, respectively. Furthermore, the AUC values of the nomogram were higher than those of the individual independent prognostic components of CSS, indicating that the nomogram had stronger prognostic prediction reliability. A novel risk stratification method was created to further improve patient stratification and give clinical professionals a better reference for clinical decision-making. Conclusion: Through the developed nomogram and risk stratification method, the CSS of patients with malignant adrenal tumors could be predicted more precisely, assisting physicians to differentiate patients better and creating personalized treatment strategies to optimize patient benefits.