J. B. Liu1,2, T. B. Baker4, N. Suss3, M. S. Talamonti2,3, K. K. Roggin2, D. J. Winchester2,3, M. S. Baker2,3 1American College Of Surgeons,Chicago, IL, USA 2University Of Chicago,Chicago, IL, USA 3Northshore University Health System,Evanston, IL, USA 4Northwestern University,Chicago, IL, USA
Introduction:
Prior studies from large national datasets comparing transplantation and resection for hepatocellular cancer (HCC) have not appropriately controlled for liver function. Previous multi-institutional series comparing transplantation and resection have included small numbers of patients with preserved liver function while also including those with decompensated cirrhosis. The benefit of transplantation relative to resection in patients with preserved liver function and potentially resectable HCC continues to be subject of considerable debate.
Methods:
We evaluated patients from the National Cancer Data Base (NCDB) undergoing treatment for HCC between 2010 and 2013 with calculated MELD scores <11. Patients undergoing resection were 1:1 propensity-matched to patients undergoing liver transplantation based on age, gender, comorbidity burden, tumor size, tumor multiplicity, pathologic stage, margin status and MELD score. Logistic regression models with robust standard errors were constructed to examine 30- and 90-day mortality. Unadjusted and adjusted survival analyses were conducted using Kaplan-Meier and shared frailty models.
Results:
2,463 patients underwent operative management for HCC. Patients undergoing resection were more likely to have positive resection margins than those undergoing transplantation (7.0% vs. 0.3%, p <0.0001). After propensity matching, 854 patients were included in our study: 427 underwent resection and 427 underwent transplantation. Rates of 30- (1.9% vs 1.9%, p = 1.00) and 90-day mortality (3.3% vs 3.0%, p = 0.85) were identical between matched cohorts. Median follow-up was 551 days for those undergoing resection and 607 days for those undergoing transplantation. Patients undergoing resection demonstrated lower rates of overall survival relative to those undergoing transplantation in unadjusted analysis (median overall survival 39% vs not reached, p < 0.0001, log-rank test) and an increased risk of death in shared frailty models (hazard ratio 2.21 [95% confidence interval 1.54-3.17]).
Conclusion:
Individualized care models are the cornerstone of treatment pathways for patients with HCC. In the subset of those with preserved liver function, there is active controversy as to whether resection or transplant offer superior overall survival rates for these patients. This propensity matched analysis of a large national database demonstrates a clear survival advantage for transplantation. Further prospective randomized clinical trials are needed to validate these findings.
