Intratumoral heterogeneity of CT enhancement for component prediction and prognostic significance in combined hepatocellular carcinoma‑cholangiocarcinoma.

OBJECTIVES

To construct a combined nomogram using CT enhancement ratio-based habitat imaging and radiological features in predicting the main component of combined hepatocellular carcinoma‑cholangiocarcinoma (cHCC-CCA), and to assess its ability for stratifying the prognosis.

MATERIALS AND

METHODS

Patients with pathologically diagnosed cHCC-CCA who underwent contrast-enhanced CT examinations were retrospectively included and randomized into the training and validation cohorts. Tumors were grouped into high hepatocellular carcinoma (HCC) component (high-HCC%) and low-HCC component (low-HCC%) according to pathology. Voxels of tumor from early enhancement ratio and late enhancement ratio maps were clustered into different habitats through the k-means algorithm. The volume fractions of different habitats were quantified. Logistic regression analyses were utilized to identify independent predictors for high-HCC%, construct prediction models, and visualize them as a nomogram. The predictive performance was assessed by receiver operating characteristic analysis. Survival analysis was conducted using the Kaplan-Meier method.

RESULTS

165 patients were finally included, and 78 (47.27%) patients were grouped as high-HCC%. Four tumor habitats were determined. The fraction of habitat 1 (f1) was significantly higher, while the fraction of habitat 4 (f4) was significantly lower in the high-HCC% group than in the low-HCC% group. Tumor capsule, corona enhancement, delayed enhancement, f1, and f4 were used to construct the combined nomogram with AUCs of 0.927 and 0.923 in training and validation cohorts, respectively. The combined nomogram predicted-high-HCC% exhibited better prognoses than the predicted-low-HCC% groups in terms of recurrence-free survival and overall survival.

CONCLUSION

Enhancement-based CT habitat imaging exhibited potential for predicting the main component cHCC-CCA, and provided a tool for prognosis stratification.

KEY POINTS: Question The component of combined hepatocellular-cholangiocarcinoma (cHCC-CCA) significantly affected the prognosis, but there is no effective method for predicting the main component of cHCC-CCA. Findings The combined nomogram integrated habitat parameters and radiological features can predict the main component of cHCC-CCA and help stratify the prognosis after hepatectomy. Clinical relevance The habitat-based combined nomogram offers an effective tool for personalized and appropriate treatment in cHCC-CCA patients.