What is Killing People With Hepatitis C Virus Infection?

JasonGrebely, B.Sc., Ph.D.; Gregory J.Dore, M.B.B.S., Ph.D., F.R.A.C.P., M.P.H.


Semin Liver Dis. 2011;31(4):331-339. 

In This Article

Causes of Mortality Among People With HCV Infection

The distribution of causes of death within a population with a chronic disease will depend on several factors: (1) disease-specific natural history and mortality risk, (2) distribution of duration of chronic disease within the population, (3) access to effective therapeutic intervention that alters natural history, and (4) age distribution and competing mortality risk within the population.

The three major disease-specific groupings for mortality among people with HCV infection are drug-related, liver disease-related, and HIV-related. Drug-related mortality includes drug overdose and suicide. Liver disease-related mortality includes decompensated cirrhosis and HCC. The mortality distribution based on these major groupings in population-based HCV notification—death registry linkage studies in Australia (New South Wales),[8] Sweden,[6] Scotland[7] and Denmark[2] (Lars Omland, personal communication, August 11, 2011) is shown in Fig. 2. In these four countries, the proportion of liver disease-related deaths varied from 19 to 24%, and for drug-related from 18 to 27%. A high proportion of drug-related mortality is consistent with injection drug use (IDU) being the major mode of HCV acquisition in all four settings. The proportion of HIV-related deaths was highest in Scotland (7.9%)[7] where 4% of the HCV-notified population was HIV co-infected, and lowest in Australia (0.4%)[8] where only 0.5% was HIV co-infected. Settings in which the HIV co-infection rate is even higher than Scotland, such as in developed countries in North America and Europe, would be expected to have larger proportions of deaths related to HIV disease.

Figure 2.

Contribution of human immunodeficiency virus- (HIV-) related, liver-related, drug-related, and other cause-related mortality (percentage of total number of deaths) in large population-based studies of people diagnosed with HCV infection in Australia (New South Wales),8 Sweden,6 Scotland,7 and Denmark2 (Lars Omland, personal communication, August 11, 2011).

Temporal trends in mortality rates and distribution among people with HCV infection are clearly important to monitor. From 1992 to 2006 in New South Wales, Australia, there has been a steady increase in the number of people with HCV dying from liver-related causes (Scott Walter, personal communication, August 11, 2011) (Fig. 3). In contrast, the number of deaths from drug-related causes increased rapidly during the 1990s, but has declined since 1999 due to the well-documented heroin "drought" starting in late 1999 and its likely subsequent reductions in IDU. The number of liver disease-related deaths reflects the expanding pool of chronic HCV including larger numbers with prolonged duration of infection (aging cohort effect). From 1997 to 2006 the age-adjusted liver disease mortality rate was stable (around 15 deaths per 10,000 person years), indicating no impact of improved HCV therapy.[8] The lack of an effect of HCV treatment on individual risk of liver-disease mortality probably relates to the generally low-treatment uptake rate, suboptimal efficacy (particularly among those with advanced liver disease), and the relatively short period of follow-up since improvements.

Figure 3.

Total number of deaths due to liver-related, drug-related, and other causes in a large population-based study of people diagnosed with hepatitis C virus (HCV) infection in New South Wales, Australia, from 1992 to 2006 (Scott Walter, personal communication, August 11, 2011).

Consistent with increasing numbers of people with HCV dying from liver-related causes, the proportion of all liver disease deaths with underlying HCV is increasing in many settings, as demonstrated in a population-based study in Scotland.[4] The burden of HCV-related advanced liver disease is also seen in increasing numbers of HCV-related liver transplants in many countries.[16]

Mortality Among Injection Drug Users With HCV Infection

The prevalence of HCV among regular IDUs and people receiving opioid substitution therapy (OST) is 60 to 80%;[61] thus, mortality studies in these populations are likely to reflect mortality among IDUs with HCV infection. Overall, mortality rates in these two populations are 1 to 2 per 100 person years,[62,63] although there is evidence that OST reduces drug-related mortality.[64,65] A study among the OST population in Australia demonstrated a considerably higher drug-related mortality rate compared with liver-disease mortality, with the ratio varying from threefold when receiving OST to 18-fold when not receiving OST.[65] However, this study covered a period during the 1990s with very high rates of drug-related mortality. A more recent mortality linkage study in New South Wales that included people on OST in 1980 to 1984 has demonstrated increasing rates of liver disease mortality, which in recent years is the leading cause of death overtaking drug-related mortality.[66] This study is of great importance as it demonstrates the impact of liver disease on mortality within an aging cohort, particularly when rates of IDU decline.

In Canada, the Vancouver community-based CHASE cohort (81% and 42% have used illicit and injection drugs in the past 6 months; HCV prevalence is 64%) has also examined all-cause and liver-related mortality through data linkage to a death registry.[49] Between 2003 and 2007, the rate of mortality was 1.9 per 100 person-years, with causes of death being 7% liver-related, 20% drug-related, 21% HIV-related, and 52% other cause-related. All-cause mortality was associated with age >50 years and HIV infection. Further, those >50 years of age were at significant risk of liver-related mortality. Given that many communities of IDUs were infected with HCV in the 1970s and 1980s, there will inevitably be greater incidence of liver disease over the next decade.

The potential future burden of advanced liver disease within aging cohorts is also reflected in an autopsy study among individuals dying from opioid toxicity in New South Wales, Australia.[67] Among 841 deaths over a 5-year period (1998–2002), the HCV prevalence was 71% and cirrhosis was present in 7%.[67] However, in those aged >44 years at death (n = 75), cirrhosis prevalence was 25%.

Mortality in Other Populations With HCV Infection

Injection drug use has been the major mode of HCV acquisition in North America, Europe, and Australia. However, in other settings, HCV transmission has largely been through non-IDU modes and the contribution of drug-related mortality is therefore considerably reduced. Thus, the impact of HCV-related disease on mortality rates and distribution is more evident.

In one study from the United Kingdom, 924 individuals who had acquired HCV infection via blood transfusion were traced during a look-back program.[68] By the end of 2004, 28% had died (255 of 924), with 26% dying of liver-related causes. The risk of liver-related mortality in those with HCV infection was three times higher than the control group of anti-HCV negative transfusion recipients. In Taiwan, 23,785 persons (aged 30 to 65 years, HCV prevalence 4.5%) were recruited from seven townships between 1991 and 1992 and followed through 2004.[69,70] Among participants with HCV mono-infection (n = 1,040), 171 died by 2004; 28% of deaths were liver-related. After adjusting for gender, age, cigarette smoking, and alcohol consumption, those with HCV mono-infection were two times more likely to die of any cause and five times more likely to die of chronic liver disease and cirrhosis compared with those without HCV. These data suggest that HCV still leads to excess mortality when drug-related and HIV effects are removed.

Mortality in People With HIV/HCV Co-infection

As reviewed elsewhere,[71] co-infection with HIV decreases spontaneous clearance of HCV infection,[72] increases HCV RNA levels,[73] increases HCV-related liver disease progression,[46,74] and reduces response to IFN-based therapy.[75,76] Since the introduction of triple-combination antiretroviral therapy in the mid-1990s, overall mortality rates among HIV-infected populations have declined dramatically.[77] Further, the distribution of causes of death have altered considerably, with a declining proportion of acquired immunodeficiency syndrome- (AIDS-) related mortality and increasing proportions of cardiovascular and liver disease mortality.[78,79] The contribution of liver disease to mortality is particularly high in settings with a high HIV/HCV co-infection prevalence;[78] however, even in Australia where only 10 to 15% of people with HIV are HCV co-infected, liver disease contributes to 11% of deaths (Kathy Petoumenos, personal communication, August 12, 2011).

Within the HIV/HCV co-infected population, factors that influence rates and distribution of mortality are access to antiretroviral therapy, access to and effectiveness of HCV therapy, drug use, and age distribution.[71] In Australia, within the HIV/HCV co-infected population there is universal access to antiretroviral therapy and high levels of uptake, relatively limited regular ICU (the vast majority are men who have sex with men), and an aging population. Among HIV/HCV co-infected patients enrolled in the Australian HIV Observational Database (AHOD; n = 3,531), liver disease has been the underlying cause in 26% of deaths (14 of 55 deaths) as compared with only 9% of deaths (16 of 181) in those with HIV alone (Kathy Petoumenos, personal communication, August 12, 2011).

Further, although the lifespan of those with HIV infection has been improved through the availability of contemporary antiretroviral therapy, the lives of those with HCV/HIV co-infection remain much shorter.[71] In Denmark, one study compared the mortality rates of 3,990 HIV-infected persons and the general population.[80] The study demonstrated that although mortality has dropped significantly in HIV-infected persons (from a high of 124 per 1000 person years in the era preceding HIV antiviral therapy to 25 per 1000 person years in 2000–2005), the impact was less pronounced among those co-infected with HCV (57 per 1000 person-years in those with HCV/HIV vs 19 per 1000 person-years among those with HIV alone). In a large study of 23,441 HIV-infected persons (76,893 person-years of follow-up), the frequency of and risk factors associated with liver-related deaths were assessed (66% with HCV co-infection, 17% active HBV co-infection).[78] Among 1246 deaths (5.3%; 1.6 per 100 person-years), liver-related death was the most frequent cause of non-AIDS related death (14.5% were from liver-related causes). Predictors of liver-related deaths were latest CD4 cell count, older age, IDU, HCV infection, and active HBV infection. Given the underreporting of liver-related disease, the actual impact is probably even greater.[71]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: