Validating posttransplant hepatocellular carcinoma recurrence data in the United Network for Organ Sharing database

The Organ Procurement and Transplantation Network (OPTN)/United Network for Organ Sharing (UNOS) database is the most comprehensive collection of liver transplantation data, but the quality of these data with respect to hepatocellular carcinoma (HCC) recurrence has not been well assessed. In this study, we compared observed HCC recurrence rates in the UNOS database to expected rates calculated with a hierarchical model for recurrence adjusted for recipient and tumor characteristics. We used the UNOS Standard Transplant Analysis and Research data set for adult transplant patients with an initial exception for an HCC diagnosis granted between January 1, 2006 and September 30, 2010 who underwent transplantation within the same time window. We developed a risk-adjusted Poisson model with patients as the unit of analysis, random effects for transplant centers, and years of follow-up as an offset to predict expected recurrences for each center. To further investigate the possibility of underreporting, we imputed expected recurrences for non-HCC deaths. In all, 5034 HCC liver transplant recipients were identified, and 6.8% experienced recurrence at a median of 1 year after transplantation. The covariate-adjusted shrinkage estimates of the observed/expected HCC recurrence ratios by transplant center ranged from 0.6 to 1.76 (median = 0.97). The 95% confidence intervals for the shrinkage ratios included unity for every center, and this indicated that none could be unambiguously identified as having lower or higher than expected HCC recurrence rates. Imputing outcomes for patients potentially experiencing unreported recurrence changed the center-specific shrinkage ratios to 0.72 to 1.39 (median = 0.98), with no centers having a shrinkage ratio significantly different from 1. The observed HCC recurrence rate was not significantly lower than the expected rate at any center, and this suggests that no systematic underreporting has occurred. This study validates the OPTN HCC recurrence data and supports their potential for further analysis.