Division of Gastroenterology and Hepatology, University of Colorado Denver, Aurora, CO
Division of Gastroenterology and Hepatology, University of Colorado Denver, Anschutz Outpatient Pavilion, 7th Floor, 1635 Aurora Court, Mail Stop B-154, Aurora, CO 80045. Telephone: 720-848-2293; FAX: 303-724-1891;
This work was funded in part by grants from the National Center for Research Resources (KL2 RR024130), the National Institute of Diabetes and Digestive and Kidney Diseases (DK076565), and the Agency for Healthcare Research and Quality (DK076565).
Hepatitis C virus (HCV) is the most common blood-borne infection and is a leading cause of liver disease in the United States.1 An estimated 1.3% of the total US population is chronically infected with HCV.1 Among individuals who have been infected with chronic HCV for 20 to 30 years, 10% to 20% will develop cirrhosis, and 1% to 5% will develop hepatocellular cancer.2 This high burden of HCV disease in the United States has made HCV the leading indication for liver transplantation (LT) in the United States.3
Several studies have characterized the epidemiology and estimated the future burden of HCV disease in the United States. The leading modes of transmission are injection drug use and the transfusion of HCV-infected blood products. The peak incidence of new cases occurred during the calendar years 1970-1989, with an observed 85% drop in the incidence after 1989 that was attributable to the improved screening of blood products and interventions targeting the prevention of human immunodeficiency virus.1,4 The peak HCV antibody prevalence of 4% occurred in persons born in the calendar years 1940-1965.5 This birth cohort passed through its high-risk period (the ages of 20-35 years) during the period in which there was a high incidence of HCV infections (1970-1989).5 As the duration of HCV infections in this birth cohort with a high prevalence of HCV increases, the prevalence of complication-related liver disease is projected to increase dramatically during the next 10 to 20 years.2,5 On the basis of US population data, Davis et al.2 projected that from 2000 to 2030, the number of patients with HCV-related cirrhosis will climb nearly 2-fold from 472,000 to more than 879,000. Age-specific trends were not included in these projections of the burden of HCV disease in the United States. In a subsequent study, Wise et al.6 reported age-specific trends in HCV-associated mortality in the United States and showed increased rates in 55- to 64-year olds as of 2004. Importantly, HCV is the leading risk factor associated with the development of hepatocellular carcinoma (HCC) in the United States and has been implicated in up to 47% of cases.7 Model simulations of the complications of chronic HCV infections predict a peak incidence of cirrhosis in 2020 with a continued increase in the incidence of hepatic decompensation and HCC for the next 10 to 13 years.8
In this article, we examine observed and projected age-specific trends in the burden of HCV and HCV complicated by HCC on LT in the United States. Because of the dire projected trends for HCV-related complications in the general population, further understanding of age-specific trends in the demand for LT is needed.
Data on all adult patients (≥18 years) who were registered for primary LT on the waiting list in the United States from 1995 to 2010 were obtained from the Organ Procurement and Transplantation Network. These data were available from the Standard Transplant Analysis and Research File created on June 3, 2011. New adult registrants for the calendar years 1995-2010 were used to calculate observed annual trends, and the complete data were used to generate projections. Retransplants were excluded. Demographics (including the age at registration) and the liver disease diagnosis were evaluated. New registrants with a primary or secondary diagnosis of HCV or another text field entry for HCV infection at registration were analyzed as having HCV-related liver disease. Similarly, new registrants with a primary or secondary diagnosis of HCC or another text field entry for HCC at registration were analyzed as having HCC. Through this process, we identified primary LT registrants with HCV and HCV complicated by HCC. The year of birth was calculated on the basis of the age at registration and the calendar year of registration. To facilitate the comparison of data for new LT registrants to published US population data, we chose age categories based on those selected by Wise et al.6 The study was approved a priori by the institutional review boards.
Study populations were described with counts, proportions, and medians with ranges. Age-specific rates for new registrants by calendar year were calculated. A Poisson regression was used to examine the associations between LT registration and birth cohorts among patients with HCV-related liver disease with or without HCC. Projections of predicted counts of new registrants for calendar years outside the range of observed data were based on the extrapolation of natural cubic splines fitted to observed trends in Poisson regression models. All statistical analyses were performed with SAS version 9.2 (SAS Institute, Cary, NC) and R version 2.12 (R Foundation for Statistical Computing).
During the study period (1995-2010), there were 126,862 new registrants for LT, and 52,540 (41%) had HCV-associated liver disease. New registrants with HCV-associated liver disease had a median age of 52 years (range = 18-83 years), 71% were male, and 10,345 (20%) also had HCC.
HCV-Related Liver Disease: US Mortality Rates and Demand for LT
Previously, Wise et al.6 evaluated age-specific mortality rates for HCV-infected persons in the general US population between 1995 and 2004.6 Mortality rates were increasing for persons who were 55 to 64 years old in 2004 but were flat or decreasing in persons who were 35 to 44 years old, 45 to 54 years old, or 65 years old or older. We evaluated these same age groups for the frequency of new registrants for LT with HCV-related liver disease (Fig. 1A). In our analyses, we found age-specific trends in the demand for LT for HCV-related liver disease that were strikingly similar to those found by Wise et al. for HCV-associated mortality. In the calendar years 2004-2009, the number of new registrants with HCV increased among patients who were 55 to 64 years old, but the number remained flat or decreased among patients who were 35 to 44 years old, 45 to 54 years old, or 65 years old or older. Next, we examined the rates of new registrants with or without HCC. The age-specific trends for new registrants with HCV but without HCC were similar to those seen in the whole group (Fig. 1B). However, in patients whose HCV was complicated by HCC, this trend was more evident, with a dramatic rise in the rate of new registrants among patients between the ages of 55 to 64 years and a decline among patients between the ages of 45 and 54 years (Fig. 1C). This pattern of a decline in the 45- to 54-year-old age group and a rise in the 55- to 64-year-old age group is suggestive of a birth cohort effect by which patients transition between age categories across calendar years.
In Fig. 2, we use 5-year age cohorts to present the age-specific trends for new registrants for LT with HCV-related liver disease. Overall, the absolute number of new registrants with HCV climbed each year from 1995 to 1999, and a relatively stable rate was reached apart from a brief decline during 2002 and 2003. Similarly to the mortality data reported by Wise et al.,6 in recent calendar years, the proportion of new registrants with HCV who were 50 to 54 years old or 55 to 59 years old increased, and the proportion of new registrants with HCV who were 40 to 44 years old or 45 to 49 years old stabilized or decreased. The trends for age-specific rates of new registrants with HCV were significantly different in these age groups (P < 0.01), and this again was consistent with a birth cohort effect.
Peak New HCV LT Registrants: 1941-1960 Birth Cohort
We stratified new registrants for LT with HCV-related liver disease by the birth cohort and the age at registration (Fig. 3). More than 82% of the patients were between the ages of 40 and 59 years at the time of registration. The frequency of HCV-related liver disease among the birth cohorts was highest for patients born from 1951 to 1955, who were followed by patients born from 1956 to 1960, then by patients born from 1946 to 1950, and finally by patients born from 1941 to 1945. These 4 birth cohorts, spanning 1941 to 1960, accounted for 81% of the new registrants with HCV-related liver disease during the calendar years 1995-2010. Across the calendar years, the 1941-1960 birth cohort (representing individuals who were 50-69 years old in the year 2010) consistently dominated among the new registrants for LT with HCV. Interestingly, the 1961-1965 birth cohort showed a small yet increasing number of new registrants with HCV-related liver disease in more current calendar years.
HCC and Observed Age-Specific Trends
Among new registrants with HCV-related liver disease but without HCC, the 1941-1960 birth cohort was dominant, yet the absolute number for this birth cohort was relatively stable (Fig. 4A). When we examined the rates in patients with HCC, we found a dramatic increase in both the number and proportion of new registrants with HCV and HCC in the 1941-1960 birth cohort (Fig. 4B), with an approximately 4-fold increase between the calendar years 2000 and 2010.
Projected Age-Specific Trends in New Registrants for LT
Next, we plotted observed and projected age-specific trends by birth cohort and stratified patients by the absence or presence of a concurrent diagnosis of HCC (Fig. 5A, B). According to the rates observed up to 2010, the rates of new registrations for patients without HCC who were born from 1941 to 1955 are expected to decline, and the rates for those born from 1956 to 1960 are projected to be stable. However, for those with HCC, the rates of new registrations are expected to be steady in patients born from 1941 to 1950, and they are projected to increase in patients born from 1951 to 1960. By 2015, patients born in or before 1955 will be 60 years old or older. Figure 6 shows the projected rates of new HCV registrants with or without HCC born in or before 1955. Largely on the basis of a projected decline in new HCV registrants without HCC who are ≥60 years old in 2015, the relative proportion of new registrants with HCC is expected to increase. By 2015, up to 40% of all new registrants with HCV who are ≥60 years old are expected to have HCC.
Our analyses have identified the 1941-1960 US birth cohort with HCV-related disease as generating the greatest demand for LT. Additionally, within this birth cohort of individuals with HCV-related liver disease, we have found a dramatic increase in the rate of new registrants for LT due to HCC. Interestingly, our analyses of observed and projected rates suggest that older patients (≥60 years old) with HCC will increasingly contribute to the proportion of HCV-infected LT candidates unless current patterns of care change dramatically.
Prior studies of HCV epidemiology and HCV age-specific mortality have demonstrated that the peak prevalence of HCV and mortality associated with HCV occurs in the 1940-1965 US birth cohort.5,6 Our current study has identified a similar US birth cohort effect among LT registrants with HCV-related liver disease. Specifically, individuals born from 1941 to 1960 dominate the demand for LT in the United States. HCC is increasing in frequency in the United States, and HCV infection is the leading contributing risk factor (implicated in up to 47% of cases).7 The incidence rate of HCC in the United States more than doubled between the calendar years 1985 and 1998, with point estimates increasing from 1.3 to 3.0 per 100,000 persons and reaching 4.1 per 100,000 persons in 2000.9,10 Adjusting for age, Kim et al.11 found an increasing incidence of HCC in new registrants for LT with HCV. Using a birth cohort analysis, our study has confirmed this finding and has demonstrated that the increasing incidence of HCC is a significant contributor to the increase in the demand for LT in the 1941-1960 birth cohort of patients infected with HCV. Additionally, in a recent study, Asahina et al.12 reported that increasing age in a cohort of patients treated for HCV had a strong independent association with the incidence of HCC, particularly in those more than 65 years old. In this context, as the 1941-1960 birth cohort observed in our study ages, the demand for LT among patients with HCV and HCC is unlikely to decline until other age-specific comorbidities preclude transplantation.
When complications of end-stage liver disease occur in the setting of HCV, 1 of 2 results is likely: death or LT. Unlike Wise et al.,6 who used mortality with HCV as their measure of the HCV disease burden in the United States, we used listing for LT as our HCV disease burden measure. These 2 HCV disease burden measures showed similar age-specific trends, which we have demonstrated are likely birth cohort effects occurring in the United States and previously attributed to HCV transmission due to injection drug use and the unavailability of tests for adequately screening blood products for HCV during the years 1970-1990.5 In our study using listing for LT as the HCV disease burden measure, we observed different birth cohort patterns in the rates of new registration for LT among individuals with HCV-related disease, and these patterns depended on the HCC status. There are at least 2 potential epidemiological explanations for this: (1) the incidence of HCC is higher in older patients, and (2) older patients with non-HCC indications for LT (ie, ascites, hepatic encephalopathy, or portal hypertensive bleeding) are less likely to be referred or listed for LT. We have been unable to determine which explanation or to what degree these or other explanations explain our observations. Another potential influence on our findings is a possible ascertainment bias related to the improved documentation of HCC after 2002 with the implementation of Model for End-Stage Liver Disease–based allocation, which included additional priority for patients with HCC. Importantly, dependent on the knots in the natural cubic spline calculations, the projections of the observed data are largely based on data from the most recent 5- to 7-year period (ie, since 2003-2005).
The characterization of the observed trends and projected changes in the demographics of patients seeking LT in the United States may allow proactive planning by the US LT community to adapt current treatment approaches and policies to future needs. Prior epidemiological projections of HCV-related mortality and the need for LT predicted peak event rates in the calendar years 2014 and 2015, respectively.13 Using more contemporary data, our analyses have demonstrated a steady rise in the demand for LT in an increasingly older population with HCV infections that will be driven primarily by patients with HCC. Absent an abrupt reversal of our observed rate of new registrants through 2010, the peak demand stemming from the 1941-1960 birth cohort is likely to extend beyond 2015, but increasing age and other age-related comorbidities may have a significant influence on LT candidacy assessments in this birth cohort of patients, who will be 60 to 79 years old in 2020.14 The Centers for Disease Control and Prevention15-17 and others18,19 are currently considering the expansion of HCV screening to all persons in the 1945-1965 birth cohort in the United States, and this approach may increase the HCV testing rate.19,20 In the near term, such a policy may increase HCV diagnosis rates and potentially HCC diagnosis rates, but subsequently, it would be expected to reduce the occurrence of HCV-related liver disease19 and associated complications such as HCC. Additionally, advances in the treatment of HCV or HCC that have the potential to alter the disease course could result in a lower observed HCV-related disease burden, particularly over longer time horizons and if the interventions have improved tolerability in elderly patients.
Although LT can be performed safely in highly select candidates who are more than 70 years old,21 alternative treatment strategies may be more appropriate, particularly for patients infected with HCV. The strong adverse effect of donor age on graft and patient survival after LT is amplified in HCV-infected recipients.22,23 However, the practice of preferentially using younger donors for HCV recipients23-26 may not be appropriate in the elderly. Among patients with HCV who develop HCC, locoregional therapies or surgical resection for patients with compensated cirrhosis may be a more efficacious use of resources than LT.27-33 During the coming decade, the projected increase in the demand for LT from an aging HCV-infected population will challenge the LT community to reconsider current treatment paradigms.