Preoperative prediction of microvascular invasion in hepatocellular carcinoma using quantitative image analysis

Abstract

Objective: Microvascular invasion (MVI) is a significant risk factor for early recurrence after resection of hepatocellular carcinoma (HCC). Knowledge of MVI status preoperatively would optimize patient selection for resection or transplant. This study proposes quantitative imaging predictors of MVI. Methods: 121 patients who underwent resection of HCC at 2 institutions from 2003 to 2015 were included in this retrospective study. Patients were included based on the availability of contrast-enhanced CT imaging within 3 months preoperatively. We employed quantitative imaging analysis, including local binary patterns, which utilized regional pixel variations in CT image of tumors to automatically categorize tumor morphology and detect MVI. The analysis was performed on the index tumor, which is the largest tumor, and the analysis was repeated with 10fold cross-validation. Results: Given the diverse morphology of HCC tumors, tumors were first categorized into uniform (n = 40) or heterogeneous (n = 81) groups by two blinded radiologists, and it was validated by quantitative image analysis with area under curve (AUC) 0.94 and accuracy 86%. Thirteen tumors (33%) in the uniform group had MVI whereas 41 (51%) in the heterogeneous group had MVI. Within each group, imaging features measured by local binary patterns were most predictive of MVI by univariate analysis. Local binary patterns predicted MVI with AUC 0.78 and accuracy 75% among uniform tumors and with AUC 0.76 and accuracy 74% among heterogeneous tumors (Figure). Conclusion: Quantitative image analysis is a promising preoperative predictor of MVI. Validation in an external dataset is needed to elucidate the utility of this novel imaging marker.

Preoperative Prediction of Microvascular Invasion in Hepatocellular Carcinoma via Multi-Parametric MRI Radiomics

“…Various researches based on preoperative imaging features of HCC were performed to predict the MVI, but showed no consensus ( 14 – 16 ). Moreover, the reproducibility and practicality were still controversial due to the overreliance on the subjective judgment of radiologists ( 17 ). All of these lack characteristic evaluation on tumor heterogeneity that reflects different biological behaviors of HCC.…”

Influence of CT acquisition and reconstruction parameters on radiomic feature reproducibility

“…With respect to the liver, for example, publications have emerged linking quantitative imaging features to clinicopathological and outcome variables in single institution retrospective series. [4][5][6][7][8][9][10][11][12] While clinically promising, the successful clinical implementation of radiomics as a trusted biomarker requires reproducibility experiments studying the effect of varying image acquisition and reconstruction parameters on imaging features.…”

Predictive factors for hepatocellular carcinoma recurrence after curative treatments

“…These authors [71] found that peri-tumoral radiomics was better in predicting HCC early recurrence than tumoral radiomics [73] . Other authors using radiomics on pre-operative CT-scans found a good correlation with MVI (AUC 0.80) [74] . Beyond texture, using 3D MRI was also possible to evaluate tissue stiffness; a multicenter study [75] recently found that HCCs with subsequent recurrence had higher tumor stiffness.…”