Title: Decline in Annual Mortality of Hepatitis C Virus–Related Hepatocellular Carcinoma in the United States, From 2009 to 2018
Abstract: Hepatocellular carcinoma (HCC) is the fourth leading cause of cancer-related deaths in the world, accounting for 810,000 fatalities in 2015.1Global Burden of Disease Cancer Collaboration et al.JAMA Oncol. 2017; 3: 524-548Crossref PubMed Scopus (2735) Google Scholar With the introduction of direct-acting antiviral (DAA) agents in late 2013, the overall sustained virologic response for HCV infection increased significantly, and it allowed treatment of patients over a wide spectrum of clinical disease severity, including those with end-stage liver disease who were in the past deemed poor candidates for antiviral therapy. Therefore, it is plausible to hypothesize that a highly potent DAA-based antiviral regimen may reduce the risk of HCC by eradicating the HCV infection followed by regression in hepatic fibrosis and an improvement in survival partly due to reduction in the incidence of HCC.2AASLD/IDSA HCV Guidance Panel.Hepatology. 2015; 62: 932-954Crossref PubMed Scopus (972) Google Scholar Furthermore, smaller studies have proposed an association between the use of DAA agents and HCC.3Reig M. et al.Semin Liver Dis. 2017; 37: 109-118Crossref PubMed Scopus (63) Google Scholar In patients with a history of successful response to HCC treatment in the setting of HCV-related cirrhosis, hepatic decompensation may have a higher contribution to mortality than HCC recurrence.4Cabibbo G. et al.J Hepatol. 2017; 67: 65-71Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar We studied the temporal trends in etiology-based HCC- and cirrhosis-related mortality in the United States from 2009 to 2018. We analyzed the trends in HCC- and cirrhosis-related mortality in the United States for adults aged ≥20 years using deidentified national Multiple Cause of Death files from the National Vital Statistics System. The detailed methodology is shown in the Supplementary Methods. We analyzed a total of 25,907,886 deaths among adults in the United States from 2009 to 2018. The study population consisted of 109,937 deaths from HCC and 579,680 deaths from cirrhosis. As indicated in Figure 1A and Supplementary Table 1, age-standardized HCC-related mortality increased from 4.14 per 100,000 persons in 2009 to 4.90 per 100,000 persons in 2018, with an average annual percent change (APC) increase of 1.9% (95% confidence interval [CI], 1.5–2.3). In terms of etiology-based HCC-related deaths (Figure 1B and Supplementary Table 1), age-standardized HCC-related mortality from HCV infection increased from 0.96 per 100,000 persons in 2009 to 1.21 per 100,000 persons in 2014. However, a sharp decrease was observed after the introduction of DAA agents in late 2013, from 1.21 per 100,000 persons in 2014 to 1.05 per 100,000 persons in 2018. Using joinpoint analysis, there was an initial increase at an annual rate of 5.4% (95% CI, 3.6–7.3) from 2009 to 2014, followed by a decrease from 2014 to 2018 at an annual rate of −3.5% (95% CI, −5.9 to –1.1). On the contrary, there was a linear increase in the age-standardized HCC-related mortality from alcoholic liver disease (ALD) (7.8%; 95% CI, 6.5–9.1) and nonalcoholic fatty liver disease (NAFLD) (21.1%; 95% CI, 16.9–25.4]). The HCC-related mortality from hepatitis B virus infection remained stable between 2009 and 2013 and then markedly declined from 2013 to 2018 at an annual rate of –4.9% (95% CI, –8.3 to –1.3). When we redefined HCC by using International Classification of Diseases, 10th revision codes C22.0 and C.22.9 for sensitivity analysis (Supplementary Table 1), the results remained mostly identical. Age-standardized HCC-related mortality from HCV reached a plateau in 2014 (APC, 6.1%; 95% CI, 4.3–7.9) and then declined rapidly from 2014 to 2018 (APC, –4.7%; 95% CI, –7.0 to –2.4). As shown in Figure 1C and Supplementary Table 2, the age-standardized cirrhosis-related mortality increased from 21.4 per 100,000 persons in 2009 to 27.1 per 100,000 persons in 2018 with a statistically significant average APC increase of 2.7% (95% CI, 2.3–3.0). For mortality stratified by etiology of liver disease (Figure 1D and Supplementary Table 2), the age-standardized cirrhosis-related mortality from HCV infection reached a plateau in 2014 (APC, 4.1%; 95% CI, 3.1–5.1) and decreased steeply between 2014 and 2018 (APC, –7.5%; 95% CI, –8.7 to –6.2). There was a linear increase in the age-standardized cirrhosis-related mortality in ALD and NAFLD subcohorts. In this nationally representative population-based study, we report a significant decrease in national HCC- and cirrhosis-related mortality due to HCV infection coinciding with the introduction of DAA agents in late 2013. Cirrhosis- and HCC-related mortality from hepatitis B virus infection declined steadily during the study. In contrast, HCC- and cirrhosis-related mortality due to ALD and NAFLD increased markedly during the 10-year study. The increase in HCC after sustained virologic response with DAA agents can be attributed to the selection process where patients with increased pretreatment risk of HCC due to a higher proportion of advanced disease were treated compared to the interferon era.5Cabibbo G. et al.Liver Int. 2018; 38: 2108-2116Crossref PubMed Scopus (18) Google Scholar,6Li D.K. et al.Hepatology. 2018; 67: 2244-2253Crossref PubMed Scopus (107) Google Scholar In addition, the US Preventive Services Task Force recommendations to screen birth cohorts for HCV infection in June 2013 may have resulted in improved identification of HCV-related liver disease in the US population, resulting in higher rates of diagnosis of cirrhosis, leading to improvement in HCC screening and, thereby, mitigation of complications of HCV infection. A recent study using the Surveillance, Epidemiology, and End Results database documented that HCC incidence rates plateaued in 2013 and started declining in 2016.7Shiels M.S. et al.Gastroenterology. 2020; 158: 1503-1505Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar The North American cohort study showed a survival benefit related to DAA therapy in patients with complete response to HCC treatment,8Singal A.G. et al.Gastroenterology. 2019; 157: 1253-1263Abstract Full Text Full Text PDF PubMed Scopus (83) Google Scholar suggesting hepatic decompensation, not HCC recurrence, as the primary driver of death in patients with HCV-related cirrhosis after successful treatment of HCC.4Cabibbo G. et al.J Hepatol. 2017; 67: 65-71Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar If the current trends in cirrhosis-related mortality from HCV infection continue, our data suggest that the decline in the mortality from HCV-related cirrhosis will be much more pronounced in the future. Our study has several limitations. First, the underlying or contributing cause of death based on the death certificate may have the potential for misclassification and underestimation. Although the International Classification of Diseases, 10th revision code underestimated the true prevalence of NAFLD, the high proportion of NAFLD in the general population and the reports of HCC developing in the absence of cirrhosis warrant dedicated prospective studies to evaluate this major risk factor for HCC. In addition, we were unable to assess all etiologies of liver disease among patients with HCC and cirrhosis because of the inherent limitation of this database. The decline in HCV-related HCC is a testament to our aggressive national policies, the stance taken by subspecialty societies, educated patient advocacy groups, and the pursuit to completely eradicate HCV infection in the United States. Additional contributors from the Hepatocellular Carcinoma Research Committee for Chronic Liver Disease Foundation: Sammy Saab (Division of Gastroenterology and Hepatology, David Geffen School of Medicine at the University of California–Los Angeles, Los Angeles, California), Renumathy Dhanasekaran (Division of Gastroenterology and Hepatology, Stanford University School of Medicine, Stanford, California), Neil Mehta (Division of Gastroenterology and Hepatology, School of Medicine at the University of California–San Francisco, San Francisco, California), Fasiha Kanwa (Division of Gastroenterology and Hepatology, Baylor College of Medicine, Houston, Texas), and Hashem El-Serag (Division of Gastroenterology and Hepatology, Baylor College of Medicine, Houston, Texas). Donghee Kim, MD, PhD, (Conceptualization: Lead; Formal analysis: Lead; Investigation: Lead; Methodology: Lead; Writing – original draft: Lead; Writing – review & editing: Equal). Peter Konyn, MD (Investigation: Equal; Methodology: Equal; Writing – review & editing: Equal). George Cholankeril, MD (Investigation: Equal; Methodology: Equal; Writing – review & editing: Equal). Robert J Wong, MD (Investigation: Equal; Methodology: Equal; Writing – review & editing: Equal). Zobair M Younossi, MD (Investigation: Equal; Methodology: Equal; Writing – review & editing: Equal). Aijaz Ahmed, MD (Conceptualization: Lead; Investigation: Lead; Methodology: Lead; Supervision: Lead; Writing – original draft: Equal; Writing – review & editing: Lead). We performed analyses of trends in HCC- and cirrhosis-related mortality in the United States for adults aged ≥20 years using deidentified national Multiple Cause of Death files from the National Vital Statistics System, which captured more than 99% of deaths in all 50 states and the District of Columbia.1Murphy S.L. et al.Natl Vital Stat Rep. 2017; 66: 1-75PubMed Google Scholar These national mortality records used the International Classification of Diseases, 10th revision (ICD-10) to define the cause of death from death certificates.1Murphy S.L. et al.Natl Vital Stat Rep. 2017; 66: 1-75PubMed Google Scholar The National Vital Statistics System produces an annual national mortality record data set on multiple-cause mortality, in which each observation is 1 death outlined with accompanying demographic characteristics.1Murphy S.L. et al.Natl Vital Stat Rep. 2017; 66: 1-75PubMed Google Scholar In this study, we analyzed the record axis for underlying or contributing causes of death.2Ly K.N. et al.Ann Intern Med. 2012; 156: 271-278Crossref PubMed Scopus (607) Google Scholar Because this national mortality database is deidentified and publicly available, an approval from institutional review board was waived for the study. The methods used for this cohort have been described in detail elsewhere.3Reig M. et al.Semin Liver Dis. 2017; 37: 109-118Crossref PubMed Scopus (63) Google Scholar,4Cabibbo G. et al.J Hepatol. 2017; 67: 65-71Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar In brief, HCC was defined based on ICD-10 code C22.0. For sensitivity analysis, we represented HCC as C22.0 and C22.9 (malignant neoplasm of liver, not specified as primary or secondary), which may include a substantial number of HCCs. Among patients with HCC, ICD-10 diagnostic codes were used to identify HCV infection (B17.1, B18.2, and B19.2). Chronic hepatitis B virus infection was defined using codes B16, B17.0, B18.0, B18.1, and B19.1. ICD-10 diagnostic codes were used to identify NAFLD (K76.0 and K75.81) and ALD (K70.0, K70.1, K70.2, K70.3, K70.4, and K70.9). For cirrhosis-related mortality, we classified patients with cirrhosis complicated by portal hypertension: either cirrhosis (K70.3, K74.0, K74.1, K74.2, K74.3, K74.4, and K74.6) and/or portal hypertension (K76.6) and/or one of its manifestations of cirrhosis-related hepatic decompensation, including hepatic encephalopathy (K72.11 and K72.91), spontaneous bacterial peritonitis (K65.2), hepatorenal syndrome (K76.7), or variceal bleeding (I85.0 and I85.1). To calculate age-specific mortality, the number of deaths from HCC or cirrhosis was divided by the total US census population for each year. We calculated age-specific mortality per 100,000 persons by age group (20–29, 30–39, 40–49, 50–59, 60–69, 70–79, and ≥80 years). We standardized calculated age-specific mortality according to the age distribution of the 2010 US standard population using the direct method. We presented the demographic and clinical characteristics of the study population as frequencies with percentages. We examined changes in temporal trends over time using the National Cancer Institute’s (Bethesda, MD) Joinpoint regression program, version 4.7.0.0. This joinpoint regression determines whether single or multiple trend segments explain age-standardized mortality over time by fitting joined straight lines to trend data.5Kim H.J. et al.Stat Med. 2000; 19: 335-351Crossref PubMed Scopus (3314) Google Scholar For each trend segment, we obtained the APC and the average APC, which is a summary measure of trend accounting for transitions within each trend segment.5Kim H.J. et al.Stat Med. 2000; 19: 335-351Crossref PubMed Scopus (3314) Google Scholar The Joinpoint regression program identifies a set of the timepoints at which the change in the trend of the mortality is statistically significant and calculates the year-to-year percent change in annual age-standardized mortality and the 95% CI over each trend segment.5Kim H.J. et al.Stat Med. 2000; 19: 335-351Crossref PubMed Scopus (3314) Google Scholar,5Kim H.J. et al.Stat Med. 2000; 19: 335-351Crossref PubMed Scopus (3314) Google ScholarSupplementary Table 1Age-Standardized HCC-Related Mortality and APC Among US Adults ≥20 Years, 2009–2018Age-standardized mortality rate (per 100,000 persons)Average APC (95% CI)Trend segment 1Trend segment 2200920182009–2018YearAPC (95% CI)YearAPC (95% CI)HCC (defined by C22.0)4.14 (8953)4.90 (13,342)1.9 (1.5–2.3)aP < .05.2009–20181.9 (1.5–2.3)aP < .05. EtiologyHepatitis C0.961.051.3 (0.2–2.5)aP < .05.2009–20145.4 (3.6–7.3)aP < .05.2014–2018–3.5 (–5.9 to –1.1)aP < .05.ALD0.220.447.8 (6.5–9.1)aP < .05.2009–20187.8 (6.5–9.1)aP < .05.Hepatitis B0.150.11–2.2 (–4.5 to 0.1)2009–20131.2 (–4.0 to 6.5)2013–2018–4.9 (–8.3 to –1.3)aP < .05.NAFLD0.010.0821.1 (16.9–25.4)aP < .05.2009–201821.1 (16.9–25.4)aP < .05.HCC (defined by C22.0 and C22.9)7.51 (16,256)8.30 (22,559)1.2 (0.8–1.7)aP < .05.2009–20142.9 (2.2–3.5)aP < .05.2014–2018–0.7 (–1.6 to 0.1) EtiologyHepatitis C1.211.321.1 (0.0–2.2)aP < .05.2009–20146.1 (4.3–7.9)aP < .05.2014–2018–4.7 (–7.0 to –2.4)aP < .05.ALD0.290.536.9 (5.8–8.1)aP < .05.2009–20186.9 (5.8–8.1)aP < .05.Hepatitis B0.190.14–2.5 (–4.1 to –0.9)aP < .05.2009–2018–2.5 (–4.1 to –0.9)aP < .05.NAFLD0.010.1022.7 (14.7–31.3)aP < .05.2009–201144.9 (1.1–107.7)aP < .05.2011–201817.0 (11.5–22.8)aP < .05.a P < .05. Open table in a new tab Supplementary Table 2Age-Standardized Cirrhosis-Related Mortality and APC among US Adults ≥20 Years, 2009–2018Age-standardized mortality rate (per 100,000 persons)Average APC (95% CI)Trend segment 1Trend segment 2200920182009-2018YearAPC (95% CI)YearAPC (95% CI)Cirrhosis21.4 (46,542)27.1 (68,401)2.7 (2.3–3.0)aP < .05. EtiologyHepatitis C3.533.11–1.2 (–1.8 to –0.6)aP < .05.2009-20144.1 (3.1–5.1)aP < .05.2014–2018–7.5 (–8.7 to –6.2)aP < .05.ALD6.319.625.0 (4.5–5.4)aP < .05.2009-20185.0 (4.5–5.4)aP < .05.Hepatitis B0.300.28–0.9 (–1.9 to 0.1)2009-2018–0.9 (–1.9 to 0.1)NAFLD0.301.1516.2 (15.4–17.0)aP < .05.2009-201816.2 (15.4–17.0)aP < .05.a P < .05. Open table in a new tab
Publication Year: 2020
Publication Date: 2020-05-08
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
Cited By Count: 21
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot