A cluster of acute hepatitis C virus infection among men who have sex with men – results from contact tracing and public health implications

Abstract

Objective: 

An acute hepatitis C virus (HCV) infection in an HIV-positive man who had sex with men (MSM) was notified. In the period of his seroconversion he was also diagnosed with a rectal lymphogranuloma venereum (LGV) infection, and was part of a cluster of 15 LGV cases in 2003. Our aim was to investigate HCV transmission and to search for potential spread among sexual contacts and known LGV patients.

Methods: 

Our case series included the index, two recent contacts, and 14 LGV cases. They were interviewed about parenteral exposure for HCV, history of sexually transmitted diseases(STDs), sexual behaviour and drug use. Laboratory investigations included anti-HCV antibodies, HCV-polymerase chain reaction, and HCV genotyping.

Results: 

Seven out of 17 MSM recently seroconverted for HCV (41%). Three genotypes were found. Parenteral risk factors were excluded. Six out of seven had LGV proctitis coinciding with HCV seroconversion, six (86%) were HIV infected. Unprotected anal contact was practised by both HCV uninfected and infected cases. Unprotected active and passive fisting was reported by all seven HCV infected men, compared with two of nine uninfected men (P = 0.003). Non-intravenous drug use during sexual activities was common among all MSM. Numerous, often anonymous, sexual contacts in various European countries were reported.

Conclusions: 

A cluster of acute HCV infection is reported among mostly HIV-positive MSM, with multiple partners throughout Europe. Sexual techniques potentially leading to mucosal damage (fisting), concomitant STDs such as LGV and drug use seem facilitating factors for spread. Extensive case finding and partner tracing is advocated as well as targeted prevention messages.

Introduction

Transmission of hepatitis C (HCV) is mainly parenteral. Sexual transmission in heterosexual couples is proven ^[1], but rare ^[2,3]. Hepatitis C infection leads in about 70% of cases to chronic hepatitis with serious sequelae such as liver cirrhosis and hepatocellular carcinoma ^[4]. HIV–HCV co-infected patients exhibit a more rapid progression to liver cirrhosis and liver failure than patients with a HCV mono-infection ^[5].

In the Netherlands recent HCV infection is notifiable to enable source finding and prevent further spread.

Early in December 2003 a HCV infection was reported in a 29-year-old HIV-positive man. After the start of antiretroviral therapy in August 2003, elevated transaminases and subsequently HCV infection were detected. Evaluation of possible ways of transmission at the Municipal Health Service (MHS) revealed no parenteral risk factors. In March 2003 he had suffered from proctitis which, in May, was diagnosed and treated as lymphogranuloma venereum (LGV). Retrospectively, his HCV-seroconversion had been between January and May 2003. He was part of a cluster of LGV in mostly HIV-positive men having sex with men (MSM), which was reported to the MHS in December 2003 by the Sexually Transmitted Diseases Clinic of the Erasmus University Medical Centre ^[6–8].

We hypothesized that LGV may facilitate the spread of HCV. To investigate possible HCV transmission by sexual behaviour and to search for potential spread, intensive sexual contact tracing according to the Dutch Infectious Disease Act was initiated. Furthermore, we assessed HCV infection among the LGV cases diagnosed earlier in 2003.

Methods

Sexual contacts of index patient (A) in the 6 months preceding HCV seroconversion and thereafter were assessed. We obtained consent from 14 known LGV cases for HCV-testing in serum taken at moment of LGV diagnosis. In cases of HCV positivity, the time of seroconversion was established with stored blood samples. HCV-negative participants were retested and HCV-positive patients were referred for treatment.

A questionnaire was developed containing data on socio-demographic characteristics, known risk factors for HCV (medical invasive procedures; sharing toothbrushes, razors, needles and piercing/tattooing) as well as sexually transmitted infection history, number of partners in the previous 6 months, sexual techniques and drug use.

Informed consent was obtained for HIV testing and HCV sequencing.

LGV was diagnosed by C. trachomatis (CT) polymerase chain reaction (PCR) in a rectum and/or urine specimen, and confirmed by detection of serovar L₂ through genotyping. CT-positive cases without genotyping with high levels of CT IgG and IgA serum antibodies were defined as probable LGV and LGV-contacts with high antibodies as possible LGV ^[7].

The presence of antibodies against HCV was determined by using Axsym HCV version 3.0 (Abbott Diagnostic Division, Wiesbaden, Germany). Reactive samples were confirmed by use of the INNO-LIA HCV Ab III (Innogenetics, Ghent, Belgium).

In subsequent anti-HCV positive sera, an increasing number of immunoblot reactivities against HCV antigens were categorized as early infection. Genotyping of HCV-positive samples was performed using an Inno-LIPA HCV genotyping reverse hybridization assay (InnoGenetics). The PCR amplicons were generated using a COBAS AMPLICOR HCV 2.0 assay (Roche Diagnostics, Almese, the Netherlands).

The sequence analysis was carried out on a PCR segment of 222 base pairs in the NS5 region (nucleotides 7975 to 8196) as described before ^[9,10]. Sequences were aligned using BioEdit software (Tom Hall, Ibis, Therapeutics, Carlsbad, California, USA) and phylogenetic analysis was performed using MegAlign software (DNAStar, Madison, Wisconsin, USA). Reference strains publicly available were chosen for comparison.

Descriptive analysis was done. The chi-squared test was used to compare proportions.

Results

The index case A had about 30 partners in Europe in the 6 months previous to his HCV seroconversion – nine of them were traced in the LGV cluster – and four recent contacts.

Early in September 2003, case A participated in a sex party (I) with three MSM (B, C, Y), who were notified by A. Y refused investigation. A, B, C and a non-traceable person had a second encounter at the beginning of October. There had been multiple mutual unprotected anal intercourse as well as unprotected fisting. A, B and Y assumed each other to be HIV positive.

Case B, a HIV-positive MSM, had 50, mostly protected, anal receptive and fisting contacts. In mid-December 2003 reactivity against HCV antigen was found, confirmed by a positive HCV-PCR. It is possible that case B acquired his HCV infection at one of the parties.

C reported protected receptive anal and fisting contacts with less than 10 MSM. With the exception of a recent single unprotected receptive intercourse without ejaculation with case B, he reported unprotected fisting once in spring. In late September 2003 he developed fatigue and fever, and shortly after anal itching and discharge. He showed a three-fold increase of transaminases between 7 and 31 October, when he was found to be anti-HIV positive and anti-HCV positive. Stored serum from October 2002 was anti-HCV negative (insufficient for HIV testing). In December 2003, case C was treated for rectal chlamydial infection. Retrospective investigation revealed LGV serovar L2. It was concluded that C had acquired HCV, LGV and possibly HIV infection and that this most likely occurred in early September 2003 (Fig. 1).

Fig. 1: Cluster of seven recent hepatitis C virus (HCV) cases: time schedule including information on HCV, HIV and lymphogranuloma venereum (LGV). Light grey columns indicate the period of HCV seroconversion. The beginning of the period was HCV negative if not stated otherwise. Dark grey indicates the period of proctitis symptoms; at end of the period confirmed LGV diagnosis if not stated otherwise. HAART, highly active antiretroviral therapy.

Among the 14 LGV-infected cases, four other cases of HCV infection were found. In three of them (K, L, O), HCV seroconversion co-incided with proctitis/LGV diagnosis. Recent seroconversion could be proved in all but one case (K) (Fig. 1).

None of the seven HCV-infected cases in this group of 17 MSM had a history of intravenous drug use or other risks for parenteral transmission in the year before HCV seroconversion.

HCV genotype 4 and two subgroups of genotype 1 were detected (Fig. 2). All HCV cases but O were sexual contacts of A, yet O and A were closely homologous. The two cases with closely clustered HCV 4 virus (L, G) did not know each other. The investigated participants of party I had two different strains of HCV; B and C were infected with an identical strain of HCV virus. Viral load varied from 2.36 × 10³ to 2.12 × 10⁷ mEq/ml.

Fig. 2: Phylogenetic analysis of hepatitis C virus (HCV) cases. Cases described in our paper are numbered in bold with the case ID in brackets.

All MSM except one HCV-negative case were HIV infected; their mean age was 39 years. Concomitant sexually transmitted diseases (STDs) apart from LGV were frequent (two of seven HCV infected; six of 10 HCV uninfected; P = 0.2).

There was no difference in unprotected insertive anal intercourse between HCV-infected and uninfected MSM [86% (six of seven) versus 89% (eight of nine); P = 0.8), nor in unprotected receptive anal intercourse (100% (seven of seven) versus 89% (eight of nine); P = 0.4).

Unprotected receptive fisting in groups was performed by all seven HCV cases compared witho two out of nine men in the uninfected group (P = 0.003). Among HCV-negative men, two reported protected fisting, four never engaged in fisting, two unknown. Two of the HCV-infected men mentioned spanking practices with bleeding, one of them needle sharing during MSM practices.

Seventy-one percent of HCV-infected individuals reported more than10 partners compared with 40% of uninfected men (P = 0.2). Encounters took place at gay venues in the Netherlands, as well as in other European countries and New York (USA).

Non-intravenous drug use during sexual activities was reported frequently by all MSM: poppers, ecstasy, gammahydroxybutyrate (GHB), ketamine, cocaine, marijuana, amphetamine and Viagra. Intranasal drug equipment was not shared.

Discussion

Within this group of 17 MSM identified in a cluster of mainly HIV-positive LGV-cases, seven recent HCV infections were diagnosed, sexual behaviour techniques being the unique risk for transmission. Our findings indicate emerging transmission of hepatitis C among MSM with potential international spread and are in line with the recently reported increase of acute HCV among MSM in the UK and France ^[11–13]. Although we cannot assess independent HCV-determinants from this selected group, our data allow valuable indications for specific risk for transmission.

Although anal insertive intercourse with ejaculation has been reported as a risk factor for HCV ^[14], we found no difference in reported unprotected anal intercourse between HCV-infected and uninfected cases, and transmission by sperm is unlikely. This finding is supported by studies isolating HCV RNA, in which a minority of sperm samples were positive for HCV RNA and positive samples were of low titre ^[15].

However, unprotected fisting (active and passive) during group-sex was more common in HCV-infected than in uninfected MSM. None of the HCV-infected men used gloves. One can imagine that fisting and the use of rectal enema may cause mucosal damage, thereby increasing the risk of transmission.

We originally assumed that A infected B and C at party I, genotyping of the HCV cases clarified that A did not transmit his virus to B or C. A did not have passive fisting contacts on that occasion. B and C were both receptive in fisting and the fact that they are infected by an identical strain suggests that transmission occurred during fisting. Y reportedly engaged in active and receptive intercourse and fisting; however, as he refused investigation this potential source could not be proven.

Blood-borne infections can be transmitted from one to the next receptive partner without the active person being at risk. This spread may also occur in receptive anal intercourse without ejaculation.

There are controversial data about the association of number of sex partners and HCV-infection ^[14,16]. Our HCV cases reported relatively more sexual partners than uninfected men. Case C, however, who most likely was infected at party I, illustrates that one single risky encounter can be sufficient to contract HCV infection.

It has previously been reported that STDs may facilitate the spread of blood-borne disease ^[14,17–19]. Since seroconversion for HCV was concurrent with proctitis symptoms in most cases, LGV infection has probably facilitated transmission of HCV. However, a STD is not obligatory for contracting HCV (case B).

In our selected group HIV cannot be explored as an independent HCV-determinant, but immuno-suppression is not conditional for HCV infection. Earlier reports are controversial about the role of HIV, co-infection with HCV and HCV viral load on HCV susceptibility and transmission ^[20–23].

Our data show an ongoing spread of HCV among MSM within specific international sexual networks, in which LGV also emerged recently and possibly contributed to the transmission of HCV. The finding of different HCV genotypes in our group indicates several entries for HCV. Within this greatly anonymous network, still only the tip of the iceberg is traced.

To assess the spread of HCV among MSM, active case finding, reporting and contact tracing is necessary. HCV testing of HIV-positive MSM at initial presentation and subsequently according to risk and in unexplained liver disease may reveal more cases ^[24]. Screening MSM for HCV at STD clinics needs to be considered. Case–control studies may identify the specific factors that promote transmission during sexual encounters.

The majority of these MSM were searching contacts within HIV-positive groups (serosorting), suggesting HIV infection as an indicator of high-risk sexual behaviour ^[25,26] HIV-serosorting may be a strategy among HIV-positive MSM, assuming them to be only susceptible for treatable STDs.

MSM have to be aware of ongoing HCV spread, and that protection for HIV (condom use during anal sex) does not prevent transmission of blood-borne diseases in riskful practices such as fisting. Educational efforts should stress that barrier methods (condoms, gloves) and water-based lubricants are essential in the prevention of blood-borne diseases among MSM. Serosorting parties provide ideal conditions for the spread of highly virulent STDs and blood-borne diseases and form hazards for individuals and are reservoirs for infections in the community.

Contributions of authors

H.M.G. investigated the cluster, in co-operation with O.dZ., J.G.dH. and H.B.T. H.G. designed the questionnaire, collected and analysed data, wrote the first draft and finalized the report. Laboratory investigations were performed by G.vD. and H.G.M.N. All authors have contributed to interpretation of data, critically reviewed the draft and were involved in the final report.

Acknowledgements

We acknowledge the willingness of the patients to share aspects of their personal life for this investigation.