The value of contrast-enhanced magnetic resonance imaging for diagnosis of extrahepatic cholangiocarcinoma

Rationale and objectives: To establish a diagnostic model based on contrast-enhanced magnetic resonance imaging (MRI) and clinical characteristics for diagnosing extrahepatic cholangiocarcinoma (eCCA).Materials and methods: From April 2014 to September 2021, consecutive patients with extrahepatic bile duct lesions who underwent contrast-enhanced MRI within 1 month before pathological examination were retrospectively enrolled. Two radiologists blinded to clinicopathological information independently evaluated MR images. Univariable and multivariable logistic regression analyses were performed to identify significant clinicoradiological features associated with eCCA, which were subsequently incorporated into a diagnostic model. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve.Results: A total of 182 patients (mean age, 60.8 +/- 10.0 years, 117 men) were included, 144 (79 %) of whom had pathologically confirmed eCCA. Diffusion restriction (odds ratio [OR], 8.32; 95 % confidence interval [CI]: 2.88, 25.82; P < 0.001), indistinct outer margin (OR, 4.01; 95 % CI: 1.40, 11.84; P = 0.010), cholelithiasis (OR, 0.34; 95 % CI: 0.12, 1.00; P = 0.049), serum ln (carbohydrate antigen 125) (OR, 4.95; 95 % CI: 1.61, 18.55; P = 0.010), and serum ln(direct bilirubin) (OR, 1.82; 95 % CI: 1.29, 2.63; P < 0.001) were independently associated with eCCA. Incorporating the above 5 variables, a diagnostic model achieved an AUC of 0.912 (95 % CI: 0.859, 0.965), with well-fitted calibration curve (P = 0.815) and good clinical utility. Additionally, the sensitivity, specificity and accuracy of the model were 83.33 %, 86.84 %, and 84.07 %, respectively.Conclusion: The proposed model integrating two MRI features (i.e., indistinct outer margin and diffusion restriction) and three clinical characteristics (i.e., cholelithiasis, lnCA125 and lnDBIL) enabled accurate diagnosis of eCCA. This tool holds the potential to facilitate an early diagnosis and thereby allow timely treatment interventions and improved clinical outcomes for patients with eCCA.