Preoperative MRI features predicting very early recurrence of intrahepatic mass-forming cholangiocarcinoma after R0 resection: a comparison with the AJCC 8th edition staging system

PurposeThis study aimed to establish a nomogram based on preoperative magnetic resonance imaging (MRI) features to predict the very early recurrence (VER, less than 6 months) of intrahepatic mass-forming cholangiocarcinoma (IMCC) after R0 resection.MethodsThis study enrolled a group of 193 IMCC patients from our institution between March 2010 and January 2022. Patients were allocated into the development cohort (n = 137) and the validation cohort (n = 56), randomly, and the preoperative clinical and MRI features were collected. Univariate and multivariate stepwise logistic regression assessments were adopted to assess predictors of VER. Nomogram was constructed and certificated in the validation cohort. The performance of the prediction nomogram was evaluated by its discrimination, calibration, and clinical utility. The performance of the nomogram was compared with the T stage of the American Joint Committee on Cancer (AJCC) 8th edition staging system.ResultsFifty-three patients (27.5%) experienced VER of the tumor and 140 patients (72.5%) with non-VER, during the follow-up period. After multivariate stepwise logistic regression, number of lesions, diffuse hypoenhancement on arterial phase, necorsis and suspicious lymph nodes were independently associated with VER. The nomogram demonstrated significantly higher area under the curve (AUC) of 0.813 than T stage (AUC = 0.666, P = 0.006) in the development cohort, whereas in the validation cohort, the nomogram showed better discrimination performance, with an AUC of 0.808 than T stage (0.705) with no significantly difference (P = 0.230). Decision curve analysis reflected the clinical net benefit of the nomogram.ConclusionThe nomogram based on preoperative MRI features is a reliable tool to predict VER for patients with IMCC after R0 resection. This nomogram will be helpful to improve survival prediction and individualized treatment.