Prevalence of HIV infection among non-elderly individuals with hepatitis C in Japan: a population-based cohort study using a health insurance claim data

An estimated 1.0% of people in the world are infected with hepatitis C virus (HCV), leading to 475,000 deaths in 2015 [1]. Although HCV is now treatable with direct-acting antivirals (DAAs), it is still a significant public health concern because only around 20% of HCV patients were estimated to be diagnosed and the accessibility of DAAs is still low in low and middle income countries [1].

In patients with chronic HCV infection, concurrent HIV infection is known to be associated with more rapid progression of liver fibrosis, and higher mortality [2‐4]. The prevalence of HCV in people living with HIV (PLWH) has been evaluated in many countries, and the global prevalence was estimated to be 2.4% [5]. On the other hand, the prevalence of HIV among patients infected with HCV has been only analyzed in specific populations, such as psychiatric patients [6].

HCV and HIV share similar routes of transmission: injection drug use, contaminated blood products, and sexual contact [7]. In Japan, for example, it is believed that, after the World War II, HCV was spread by illicit methamphetamine due to the post-war confusion [8]. Many hemophilia patients got HIV, HCV, and hepatitis B virus (HBV) due to contaminated blood products, until the 1990s when the screening of these infections became a standard worldwide. Nowadays, injection drug use is regarded as the main mode of transmission of HCV partly because HCV had been considered rarely transmitted by sexual contact; the transmission rate of HCV between non-HIV heterosexual partners were estimated to be one infection per 190,000 sexual contacts [9]. However, in Japan, where there are fewer injection drug users (IDUs) than in the United States and Europe [10], the route of HCV infection in younger people was not well understood.

In the past two decades, however, the sexual transmission of HCV among men who have sex with men (MSM) has become recognized worldwide [11‐14].Traumatic sex practice, especially receptive anal intercourse without a condom was reported as a risk for HCV transmission [15]. A characteristic of MSM who are co-infected with HCV/HIV is that they tend to be infected with HIV prior to HCV, implying that HIV is more sexually transmittable and more prevalent than HCV as sexually transmitted infections (STIs) among MSM [7, 16]. Some hypothesized that HIV infection may increase susceptibility to HCV, but recent reports showed that incidence of HCV among non-HIV MSM was comparable to that of HIV-positive MSM [17, 18].

Other STIs are also helpful to assess the sexual activity of the group. For example, syphilis is one of the most common STIs among MSM [19]. Amebiasis and hepatitis A are transmitted by fecal–oral contact and are more common in MSM than in heterosexuals [20, 21]. HBV, chlamydia, and gonorrhea are also seen among heterosexuals, but more common among MSM [22, 23].

Here, we performed a health insurance claim-based cohort study including a general non-elderly adult population across the nation to evaluate the prevalence of HIV among HCV patients. We also analyzed the incidence of STIs among these people to assess their sexual activity.

Between April 2012 and August 2018, a total of 16,517 patients with HCV diagnosis were identified (Fig. 1). Among them, 7865 patients were performed at least one HCV-RNA test during the study period, and 1383 patients were excluded by the age criterion. Sixty hemophilia patients were also excluded.

A total of 6422 patients were analyzed. Median age was 51 years (interquartile range [IQR], 44–55 years), and 48.1% (3089/6422) were male. The age distribution of HCV patients was 6.8% (438/6422) in the 20–29, 11.0% (709/6422) in the 30–39, 27.0% (1737/6422) in the 40–49, and 55.1% (3538/6422) in the 50–59 age group, while the decadal age distribution of the whole population in the database was nearly equal across all the age groups. Male was 35.8% (157/438) in the 20–29 s, 39.8% (282/709) in the 30–39, 48.2% (837/1737) in the 40–49, and 51.2% (1813/3538) in the 50–59 age group. In Tokyo, male was 39.0% (23/59) in the 20–29, 44.2% (46/104) in the 30–39, 51.1% (117/229) in the 40–49, and 53.4% (198/371) in the 50–59 age group.

HIV prevalence was 0.48% (31/6422, 95% CI: 0.33–0.68%) among HCV patients. HIV was diagnosed after HCV in 3.2% (1/31), before HCV in 58.1% (18/31), and concurrently in 38.7% (12/31). Characteristics of PLWH and non-PLWH are shown in Table 1. Compared with non-PLWH, PLWH were younger (median age, 37 vs 51 years, p < 0.001), more likely to be male (30/31 [96.8%] vs 3,059/6391 [47.9%], p < 0.001), and the residents of Tokyo (67.7% [21/31] vs 11.6% [742/6391], p < 0.001). Although the number was small, PLWH were more likely to be drug addicts (3.2% [1/31] vs 0.05% [3/6391], p = 0.02). In Tokyo, the HIV prevalence among 20–30 s male HCV patients was 18.6% [13/70, 95% CI, 10.3–29.7%] (Fig. 2).

Syphilis was the most common STI, and 29.0% (9/31) of PLWH were infected with syphilis during the study period (non-PLWH, 0.05% [3/6391], p < 0.001), only one of which was infected with syphilis before HIV diagnosis, and the incidence of syphilis among PLWH was 85.9/1,000 person-years (95% CI, 37.1–169.2). PLWH were more likely to be infected with amebiasis (6.5% [2/31] vs 0% [0/6391], p < 0.001), HAV (3.2% [1/31] vs 0.03% [2/6391], p = 0.01). There was no statistically significant difference between the number of PLWH and non-PLWH who infected with chlamydia (3.2% [1/31] vs 0.27% [17/6391], p = 0.08), gonorrhea (0% [0/31] vs 0.06% [4/6391], p = 1.00), and HBV (3.2% [1/31], 0.49% [31/6391], p = 0.14). In total, 38.7% (12/31) of PLWH experienced at least one STI, while 0.9% (56/6391) of non-PLWH experienced STI (p < 0.001).

We showed that the prevalence of HIV was 0.48% among non-elderly HCV patients in Japan. HCV/HIV co-infected patients were more likely to be young, male, and residents of Tokyo, the capital and most populous prefecture. The prevalence of HIV among 20 s and 30 s HCV male patients in Tokyo were astonishingly high at 18.6%. HCV/HIV co-infected patients had high incidence of STIs and were more likely to acquire HIV before HCV, suggesting that HCV might be prevalent as an STI among MSM particularly in Tokyo.

In Japan, almost all legal residents are under the universal health coverage insurance scheme, provided either by the Employee’s Health Insurance System or the National Health Insurance System. In this study, we analyzed one of the Employee’s Health Insurance Systems, covering about seven-million workers and their dependents. Although the results of laboratory tests are unavailable, we can accurately estimate the number of diseases which have specific laboratory tests or treatment, such as HIV. Since HCV RNA testing is usually performed only on HCV antibody-positive individuals, the health insurance claim database is reliable source of information for estimating the prevalence of HIV among HCV patients.

This is the first report estimating the HIV prevalence among non-elderly HCV patients in Japan, and showed extremely high HIV prevalence in male living in Tokyo. The seroprevalence of HCV was reported as 2.1% in HIV-infected MSM living in Tokyo [25]. In the same report, 21 of 753 PLWH were newly diagnosed with HCV over follow-up period of median 984 (IQR, 539–1557) days, of which only four (19%) were infected by IDU. In addition, Tokyo Metropolitan Infectious Diseases Surveillance Center reported 10 cases of acute hepatitis C in 2018, in which all patients were male, seven were through sexual contact among MSM, one was through injection drug use, one was both, and one was unknown [26]. In this study, HCV/HIV co-infected patients were more likely diagnosed with HIV before HCV, which also supports the idea that HCV is prevalent as an STI among MSM in Tokyo.

In this study, many HCV/HIV co-infected patients were infected with other STIs. Syphilis was the most common STI in the present study, and the incidence rate was 88.3/1000 person-years, which was higher than a previous estimate (43.7/1000 person-years) for HIV-infected MSM living in Tokyo from 2008 to 2015 [19]. Given that one-third (10/31) of the PLWH in our cohort do not live in Tokyo, where the prevalence of syphilis is known to be higher than other areas [19], it is possible that HCV/HIV co-infected PLWH are more sexually active than non-HCV PLWH. In the present study, PLWH were more likely to be infected with amebiasis and HAV, which are usually infected by oral-fecal contact among MSM. In contrast, the number of HBV genotype testing (i.e., the supposedly newly diagnosed HBV infection) among PLWH was not higher than non-PLWH, probably because ART with anti-HBV activity, such as tenofovir or lamivudine, could prevent new HBV infections [27].

There were more female than male HCV patients in younger age even in Tokyo. This can be partly due to differential health-seeking behavior and screening opportunity by sex: males can be underdiagnosed. In general, females are more likely to visit hospital compared with males [28], and are more likely to be diagnosed with HCV earlier than men because of HCV screening in pregnancy. Nonetheless, the HCV transmission route among young female in Japan remains unclear. Among female patients, heterosexual contact is considered not to be the main mode of infection considering the low transmission rate of HCV through heterosexual sex [9]. The proportion of female methamphetamine users is also low in Japan compared with other countries [10]. Tattooing and piercing could be one of the transmission routes [29]. Since HCV antibody testing for pregnant women started in the 1990s, people in their 20 s and 30 s in the early 2010s might have been vertically infected. However, as mentioned above, most acute hepatitis C patients are male in Tokyo recently [26], so the ratio of male to female HCV patients may change in the next few decades.

There were several limitations in our study. Firstly, the diagnosis may not be accurate in some diseases, because the diagnoses are made for health care insurance claim data. For example, in case of window period, HCV-RNA tests could be performed for patients with hepatitis with unknown origin, but this is rare. Therefore, we believe that most of the patients who underwent HCV-RNA test were HCV antibody-positive (i.e., confirmed HCV infection). In the present study, approximately nine thousand patients with HCV diagnosis but without HCV RNA testing were excluded, however, even when they were included, age distribution (data not shown), HIV prevalence (0.41%, 68/16,378), and the order of HCV/HIV diagnoses (HIV was diagnosed after HCV in 1.5% [1/68), before HCV in 45.6% [31/68], and concurrently in 52.9% [36/68]) were almost same. Secondly, PLWH needs regularly visit to hospital for antiretroviral treatment and might have had more opportunities of receiving screening test of HCV and other STIs. Furthermore, it is possible that HIV-induced immunocompromised status made some STIs (e.g., amebiasis [20]) more severe and more detectable. However, even if the prevalence of HCV and incidence of STIs in non-PLWH was underestimated, the very high prevalence of HCV and incidence of STIs in PLWH could not have been overestimated. Thirdly, this database does not cover all the types of health insurance system, and this could cause a selection bias (known as “healthy worker effect”). However, our database mainly covered employed workers and their dependents, so this database is suitable for assessing the prevalence among non-elderly adults. Lastly, we could not follow the data of patients who had changed the types of health care insurance. Therefore, we excluded patients aged > 60 years who were likely to change their insurance after retirement. We assumed the population in the cohort remained stable during the study period.

In conclusion, the prevalence of HIV was 0.48% among patients with HCV in Japan, and 15–20% among non-elderly male in Tokyo. The incidence of STIs was very high among HCV/HIV co-infected patients and were more likely to be diagnosed with HIV prior to HCV. These results might suggest a possibility of the outbreak of HCV among MSM in Tokyo. When non-elderly man living in urban area is diagnosed with HCV, it is necessary to check whether he is already infected with HIV or other STIs. In addition, it may be beneficial to screen PLWH for HCV in areas where an outbreak of HCV infection among MSM is suspected. Healthcare insurance claim data might be useful to uncover the transmission route of HCV and patients at risk of infection.