Development and Validation of a Nomogram for Preoperative Microvascular Invasion Prediction in Patients with Single Hepatocellular Carcinoma

Abstract

Objective: The presence of microvascular invasion (MVI) is associated with poor survival in hepatocellular carcinoma (HCC) patients. However, pathological diagnosis of MVI has little value in making tailored preoperative decisions. In this study, we aim at developing a preoperative prediction model for MVI based on noninvasive variables. Methods: Preoperative and pathological variables from 357 patients who underwent surgical resection between January 2015 and December 2017 were analyzed. Patients were divided into training cohort (n=257) and validation cohort (n=100) according to time interval. Preoperative predictors for MVI were analyzed using univariate and multivariate analyses. A nomogram incorporating independent predictors were constructed and validated. Results: Pathological examination confirmed MVI in 140 (39.2%) patients. Preoperative imaging features including larger tumor diameter, the presence of intra-tumoral artery, and tumor type together with higher serum AFP, were independent predictors for MVI in HCC patients. The nomogram incorporating these variables showed good performance with a C-index of 0.80 (95% CI, 0.75-0.86) and 0.81 (95% CI, 0.72-0.91) in the training and validation cohorts, respectively. Calibration curves and Hosmer-Lemeshow test demonstrated good agreement between prediction and pathological confirmation in both cohorts. The nomogram achieved sensitivity of 83.0% (75.9%-90.2%) and 82.4% (69.5%-95.2%), specificity of 68.9% (61.5%-76.8%) and 59.1% (47.2%-71.0%) in two cohorts, respectively. Survival analysis revealed that patients with positive predicted MVI suffered a higher risk of early recurrence after surgical resection. Conclusions: The nomogram incorporating preoperative variables obtains desirable results in predicting MVI and is of great importance in tailoring preoperative management in patients with single HCC. Funding: This work was supported by PUMC Youth Fund/Fundamental Research Funds for the Central Universities (3332016031). Declaration of Interest: The authors report no conflicts of interest. Ethical Approval: The retrospective research was approved by the institutional review board, and the requirement for informed consent was exempted.