A computed tomography-based radiomic score to predict survival in patients with adrenocortical carcinoma

Abstract

Abstract Background: Adrenocortical carcinoma (ACC) is a rare condition with a poor and hardly predictable prognosis. This study aims to build and evaluate a preoperative computed tomography (CT)-based radiomic score (Radscore) using features previously reported as biomarkers in adrenocortical carcinoma (ACC) to predict overall survival (OS) in patients with ACC. Methods: In this retrospective study, a Radscore based on preoperative CT examinations combining shape elongation, tumor maximal diameter, and the European Network for the Study of Adrenal Tumors (ENSAT) stage and was built using a logistic regression model to predict OS duration in a development cohort. An optimal cut-off of the Radscore was defined and the Kaplan-Meier method was used to assess OS. The Radscore was then tested in an external validation cohort. The C-index of the Radscore for the prediction of OS was compared to that of ENSAT stage alone. Findings: The Radscore was able to discriminate between patients with poor prognosis and patients with good prognosis in both the the validation cohort (54 patients; mean OS, 69·4 months; 95% CI: 57·4–81·4 months vs. mean OS, 75·6 months; 95% CI: 62·9–88·4 months, respectively; P = 0·022). In the validation cohort the C-index of the Radscore was significantly better than that of the ENSAT stage alone (0.62 vs. 0.35; P = 0·002). Conclusion: A Radscore combining morphological criteria, radiomics, and ENSAT stage on preoperative CT examinations allow a stratification of prognosis in patients with ACC compared with ENSAT stage alone.