New scoring system for prediction of microvascular invasion in patients with hepatocellular carcinoma

Abstract

Background & aims: The microvascular invasion of cancer cells (mvi) is a good prognostic factor after hepatic resection (HR) and liver transplantation for hepatocellular carcinoma (HCC). This study aimed to predict mvi in patients with HCC.

Methods: We studied 63 hepatectomized patients with HCC who had HCC without any extrahepatic metastases and vascular invasion, which were detected during preoperative evaluation. The preoperative clinicopathological data of these patients were analysed to predict presence of mvi. A scoring system was designed using significant risk factors. This system was applied to another series of 34 patients with HCC who underwent HR, and was evaluated for validation.

Results: Tumour size, serum des-gamma-carboxy prothrombin (DCP) levels and the maximum standardized uptake value (SUVmax) on 2-[18F]-fluoro-2-deoxy-D-glucose positron emission tomography were independent clinical predictors for mvi after multivariate analyses. Tumour size, serum DCP levels, and values of SUVmax were used to plot a receiver operating characteristic curve for predicting mvi. Areas under the curve of tumour size, serum DCP levels and SUV max values, were 0.8652, 0.8027 and 0.7848 respectively. Maximal sensitivity and specificity were obtained when the tumour size was 3.6 cm, SUVmax was 4.2, and the serum DCP level was 101 mAU/ml. A scoring system was designed using these three variables. The sensitivity and specificity of our scoring system were 100% and 90.9%, respectively, in the validation test.

Conclusion: Our scoring system for mvi, consisting of tumour size, serum DCP levels, and SUV max, provides a precise prediction of mvi.

Keywords: hepatocellular carcinoma; microvascular invasion; prediction; scoring system.