The Impact of Liver Transplantation on Hepatocellular Carcinoma Mortality in the United States

Abstract

Background: Hepatocellular carcinoma (HCC) carries a poor prognosis. Liver transplantation (LT) is potentially curative for localized HCC. We evaluated the impact of LT on U.S. general population HCC-specific mortality rates.

Methods: The Transplant Cancer Match Study links the U.S. transplant registry with 17 cancer registries. We calculated age-standardized incidence (1987-2017) and incidence-based mortality (IBM) rates (1991-2017) for adult HCCs. We partitioned population-level IBM rates by cancer stage and calculated counterfactual IBM rates assuming transplanted cases had not received a transplant.

Results: Among 129,487 HCC cases, 45.9% had localized cancer. HCC incidence increased on average 4.0% annually [95% confidence interval (CI) = 3.6-4.5]. IBM also increased for HCC overall (2.9% annually; 95% CI = 1.7-4.2) and specifically for localized stage HCC (4.8% annually; 95% CI = 4.0-5.5). The proportion of HCC-related transplants jumped sharply from 6.7% (2001) to 18.0% (2002), and further increased to 40.0% (2017). HCC-specific mortality declined among both nontransplanted and transplanted cases over time. In the absence of transplants, IBM for localized HCC would have increased at 5.3% instead of 4.8% annually.

Conclusions: LT has provided survival benefit to patients with localized HCC. However, diagnosis of many cases at advanced stages, limited availability of donor livers, and improved mortality for patients without transplants have limited the impact of transplantation on general population HCC-specific mortality rates.

Impact: Although LT rates continue to rise, better screening and treatment modalities are needed to halt the rising HCC mortality rates in the United States.See related commentary by Zhang and Thrift, p. 435.

Figure 1:. Characteristics of the study population.

The flow chart describes the characteristics of the 129,487 HCC cases from 12 states participating in the Transplant Cancer Match Study (1987–2017): 108,042 (83.4%) cases that were captured by cancer registries but were not registered in the SRTR (i.e., not entered on the waitlist or transplanted), 17,206 (13.3%) cases that were present in both cancer registries and the SRTR (i.e., entered on the waitlist and/or transplanted), and 4,239 (3.3%) cases that were on the waitlist or received a transplant but were not captured by the cancer registries. These cases were then further classified by the cancer stage at diagnosis as localized (confined to the liver), regional (involvement of regional lymph nodes), distant (metastasized), or unstaged. We reclassified some transplant recipients who had regional (N=1,847), distant (N=267), or unstaged (N=522) HCCs, as recorded by cancer registries, to localized stage, as the advanced stages may have been miscoded due to progression between diagnosis and transplantation. Localized HCCs were classified according to their transplant status. The number of HCC-specific deaths (1991–2017) for each stage are also specified. Abbreviations: HCC, hepatocellular carcinoma; SRTR, Scientific Registry of Transplant Recipients

Figure 2:. Incidence and IBM rates of HCC, overall and partitioned by cancer stage at diagnosis.

The figures present the age-standardized rates (2000 US population) for HCC incidence (panel A) and IBM (panel B) according to calendar year of diagnosis or death in the 12 participating US states (1987–2017) The height of each bar represents the overall incidence or IBM rate for a specific calendar year. Each bar is partitioned to represent rates contributed by localized, regional, distant, or unstaged HCCs. Abbreviations: HCC, hepatocellular carcinoma; IBM, incidence-based mortality

Figure 3:. Liver transplants in the 12 US states participating in the study.

Panel A, shows the proportion of total adult liver transplants that were conducted for HCC in the 12 participating US states (left y-axis) and the rate of HCC-related transplants (number of transplants divided by the general population; right y-axis) by calendar year of transplant (x-axis). Panel B shows the proportion of total HCC cases that were local stage, and the proportion of total HCC cases who received a transplant within 5 years of diagnosis (y-axis) by calendar year of diagnosis (x-axis). Data are shown for cases diagnosed during 1987–2013 to include transplants through 2017. Abbreviations: HCC, hepatocellular carcinoma

Figure 4:. Impact of liver transplants on HCC-specific mortality in the 12 US states participating in the study.

Panel A shows the age-standardized mortality rate within 5 years of diagnosis among localized HCC cases (on y-axis) by calendar year of diagnosis (x-axis), according to their transplant status. Data has been shown for calendar years 1987–2013 to include deaths that have occurred till 2017. Panel B presents the observed (solid line) and counterfactual (dotted line) IBM rates for localized HCCs by calendar year of death. Abbreviations: HCC, hepatocellular carcinoma; IBM, incidence-based mortality