Predictive Models of Hepatocellular Carcinoma Recurrence After Liver Transplantation.

Background: Liver transplantation (LT) is a curative treatment for patients with hepatocellular carcinoma who are not candidates for resection. Despite the generalized use of the Milan criteria and up-to-seven criteria, new markers have been proposed to predict recurrence after LT. Biomarkers such as neutrophil-lymphocyte ratio (NLR) and platelet-lymphocyte ratio (PLR), and scores such as the Model of Recurrence After Liver transplantation (MORAL) are used as predictors of post-LT recurrence.

Objective: We aim to compare NLR, PLR, and MORAL score with Milan criteria and up-to-seven criteria.

Methods: A descriptive study of 99 patients who underwent LT for hepatocellular carcinoma in our hospital between April 2010 and April 2016. The 5 prognostic models were applied to the patients to stratify them into risk groups. We used a Kaplan-Meier survival plot to measure recurrence-free survival in each model. Receiver operative curves were used to compare the models.

Results: Three-year recurrence-free survival in MORAL was 91.1% for the low-risk group, 89.8% for the moderate-risk group, 60% for the high-risk group, and 75% for the very high-risk group (P = .003). The combined MORAL score was superior in predicting 1- and 3-year recurrence with the area under the curve 0.684 (95% confidence interval [CI]: 0.52-0.85) compared with Milan (0.536 [95% CI: 0.37-0.70]), up-to-seven (0.601 [95% CI: 0.43-0.77]), PLR (0.452 [95% CI: 0.30-0.61]), and NLR (0.542 [95% CI: 0.37-0.71]).

Conclusions: A model based only on pre-LT radiological signs leads to underdiagnosis of tumor load; therefore, the risk of recurrence must be recalculated after LT. The combined MORAL score was the best prognostic model of 1- and 3-year recurrence after LT in our study.