Investigating the endemic transmission of the hepatitis C virus

Abstract

The hepatitis C virus (HCV) infects at least 3% of people worldwide and is a leading global cause of liver disease. Although HCV spread epidemically during the 20th century, particularly by blood transfusion, it has clearly been present in human populations for several centuries. Here we attempt to redress the paucity of investigation into how long-term endemic transmission of HCV has been maintained. Such transmission not only represents the ‘natural’ route of infection but also contributes to new infections today. As a first step, we investigate the hypothesis that HCV can be mechanically transmitted by biting arthropods. Firstly, we use a combined bioinformatic and geographic approach to build a spatial database of endemic HCV infection and demonstrate that this can be used to geographically compare endemic HCV with the range distributions of potential vector species. Second, we use models from mathematical epidemiology to investigate if the parameters that describe the biting behaviour of vectors are consistent with a proposed basic reproduction number (R₀) for HCV, and with the sustained transmission of the virus by mechanical transmission. Our analyses indicate that the mechanical transmission of HCV is plausible and that much further research into endemic HCV is needed.

Introduction

The hepatitis C virus (HCV) is an important viral pathogen of humans, infecting an estimated 120–180 million people globally (CDC, 1998) and causing 3–4 million new infections each year. HCV causes substantial morbidity and mortality worldwide – chronic infection can lead to liver damage (cirrhosis) and hepatocellular carcinoma, resulting in 8000–10,000 deaths per year in the United States alone. HCV is a positive-sense single-stranded RNA virus, taxonomically classified as the sole member of the genus Hepacivirus in the family Flaviviridae. It therefore shares genomic similarities with the flaviviruses, a genus of about 70 species that contains several important human pathogens, including Dengue virus and West Nile virus (Table 1). The vast majority of flaviviruses are transmitted by arthropod vectors, either mosquitoes or ticks, within which they are able to replicate (Gaunt et al., 2001).

HCV is a genetically diverse virus that is classified by phylogenetic analysis into six major genotypes (denoted 1–6), each of which contains many different subtypes (denoted alphabetically, 1a, 2c, 3d, etc.; Simmonds et al., 2005). The majority of HCV infections worldwide are caused by a subset of subtypes, notably subtypes 1a, 1b, 2a, 2b, 2c and 3a (Simmonds, 2004). These strains are both highly prevalent and globally distributed and consequently have been termed ‘epidemic’ subtypes (Smith et al., 1997, Pybus et al., 2001). Their existence stems from the recent transmission history of HCV, a blood-borne infection that remained unidentified until 1989. During the 20th century, the ‘epidemic’ strains rode a rising tide of human behaviours that inadvertently promoted the rapid global transmission of HCV: blood transfusion, use of blood products, haemodialysis, non-sterile administration of medicines by injection and intravenous drug use. Unsurprisingly, most research interest has been directed towards the development of drug, vaccine or preventative strategies against these epidemic subtypes because they cause the bulk of HCV morbidity. Following the identification of the virus in 1989, transmission via blood transfusion and blood products all but ceased.

These factors may explain why few attempts have been made to understand the apparent long history of HCV transmission in some regions, particularly Africa and South-East Asia. Many HCV strains show an ‘endemic’ pattern of transmission, characterised by a relatively low prevalence and by a high virus genetic diversity in a geographically restricted area (e.g. Mellor et al., 1995, Jeannel et al., 1998, Candotti et al., 2003, Ndjomou et al., 2003). Phylogenetic analyses using molecular clock methods estimate that the genotypes of HCV are in the region of 500–2000 years old (Smith et al., 1997, Pybus et al., 2001), indicating that endemic strains have been present and circulating in human populations for centuries before the introduction of medical injections, surgery and transfusions. The ‘epidemic’ subtypes of HCV were therefore originally endemic strains that became associated, most likely by chance, with efficient transmission networks during the 20th century (see Pybus et al., 2005).

Previous explanations for the long-term “endemic” transmission of HCV have proposed a panoply of culturally or religiously conditioned routes of transmission, including circumcision, ritual scarification, female circumcision and genital mutilation, as well as acupuncture (e.g. Shepard et al., 2005). While all these possibly contribute to HCV transmission to some degree in their respective locations of practice, the maintenance of HCV endemicity over many centuries and across continents, cultures and religions warrants a more widespread and ubiquitous mechanism.

We suggest two potentially ubiquitous routes of endemic HCV transmission. The first comprises the combined effects of domestic, sexual, vertical and intra-familial transmission. Sexual and vertical transmission of HCV are well studied, but not currently thought to contribute greatly to transmission. About 5% of children of infected mothers are perinatally infected with HCV (Shepard et al., 2005) and the per-year risk of transmission among individuals in long-term monogamous partnerships is just 0–0.6% (Terrault, 2002). Very little is known about other, presumably varied, blood-to-blood contact events in the home that might lead to HCV transmission. The global genetic diversity of vertically transmitted viruses should geographically mirror that of the human populations within which they have evolved, as has been suggested for GB virus C. However, HCV does not show any such evidence (although little co-evolution would be expected if HCV has only recently entered human populations).

The second ubiquitous route of endemic HCV transmission is by vector. As we show later, endemic HCV appears to be concentrated in the tropics and sub-tropics, where human populations are subject to higher biting rates by a wide range of abundant arthropods. Furthermore, the rest of the human pathogenic flaviviruses are vector borne (Table 1). HCV has been isolated from bodies or heads of mosquitoes collected from the houses of HCV-infected individuals (Chang et al., 2001, Hassan et al., 2003) and from mosquitoes experimentally fed with infected blood (Silverman et al., 1996, Bellini et al., 1997, Chang et al., 2001, Hassan et al., 2003). These experimental studies have failed to demonstrate that HCV replicates in mosquitoes, although one in vitro investigation reports that HCV can bind to and replicate within the mosquito AP61 cell line (Germi et al., 2001). Evolutionary analysis also indicates that HCV is unable to replicate in arthropod vectors; the unconstrained molecular evolution of the HCV envelope gene is much more consistent with virus replication in just one host species (Woelk and Holmes, 2002). However, this does not preclude the mechanical transmission of HCV on the mouthparts of biting arthropods. Mechanical transmission is important in the epidemiology of several viruses (Carn, 1996) and is non-specific, with single or multiple vector taxa contributing to transmission. In addition to transmission via mouthparts, pathogens could remain intact and infectious in an insect’s foregut and be regurgitated into a new host at a later bloodmeal.

The possibility of mechanical transmission has also been considered for other blood-borne human viruses, particularly human immunodeficiency virus (HIV) and the hepatitis B virus (HBV). Jupp and Lyons (1987) report HIV survival for several hours in bedbugs (Cimex sp.), but did not observe onward virus transmission, and also found no survival of HIV virions in Aedes aegypti mosquitoes. In contrast, Eigen et al. (2002) reported that infectious HIV can be regurgitated by the stable fly Stomoxys calcitrans. For HBV, several studies indicate that the virus can survive for some time in bedbugs and mosquitoes (e.g. Fouche et al., 1990, Silverman et al., 2001). Irrespective of these experimental results, mechanical transmission of HIV and HBV is not documented and even if it does occur, does not contribute appreciably to infection. However, HBV and HIV are transmitted effectively through sex, whereas HCV has no known efficient transmission route prior to the 20th century.

Here we aim to redress the comparative lack of investigation into the endemic transmission of HCV. Firstly, we combine bioinformatic and geographic approaches to estimate, to our knowledge, the first high-resolution global maps of endemic HCV infection. Second, we use epidemiological models to test the plausibility of the hypothesis that HCV can be mechanically transmitted by arthropod vectors.