|
High incidence of HCV in HIV-negative men who
have sex with men using pre-exposure prophylaxis
|
|
|
Download the PDF here  
PrEP prevents HIV, but not HCV infection. Currently, sexually active MSM intending to use PrEP are recommended to be screened for HCV infection at PrEP-initiation, while the advised testing frequency during PrEP use varies with 3-month intervals, regularly, annually or even 'consider annual testing'.  
Based on the observed incidence rate, we suggest that all MSM be tested for HCV antibodies at PrEP-initiation, followed by a minimum of 6-month testing intervals during PrEP. More frequent testing (e.g. at 3-month intervals) for HCV-RNA should be considered for those with previous HCV infection and those reporting HCV risk factors.  
In conclusion, we demonstrate ongoing sexual transmission of HCV and high IR of primary and re-infection in a cohort of HIV-negative MSM using PrEP. Both rates and risk factors are comparable to those among HIV-positive MSM. Phylogenetic analysis suggests shared sexual networks of MSM with and without HIV. Our findings highlight the importance of an integrated sexual health approach, which includes professional education concerning HCV, frequent provider- or client-initiated testing and treatment of those with detectable HCV-RNA. Additional preventive interventions are certainly needed, such as campaigns to inform MSM about the risks of HCV and effective behavioural interventions, particularly for those at risk of (re-)infection.  
Among HIV-negative MSM enrolled in the Amsterdam PrEP project, we found a high incidence of HCV primary and re-infection, with a much lower IR for primary than re-infection. Factors associated with HCV infection were receptive CAS with casual partners, being diagnosed with an anal STI, IDU and sharing straws when snorting drugs. Phylogenetic analyses revealed that HCV sequences of HIV-negative MSM starting or using PrEP were highly interspersed with HCV sequences from HIV-positive MSM.  
The IRs of overall HCV infection and re-infection observed in HIV-negative MSM using PrEP are comparable to those observed in HIV-positive MSM.  
In contrast, HIV-negative MSM not on PrEP have a 19-fold lower HCV incidence compared to HIV-positive MSM according to a recent meta-analysis.  
The prevalence of HCV antibody-positive serology among HIV-negative MSM in Amsterdam participating in cross-sectional surveys at our STI clinic varied between 0.0% and 1.7% and did not change from 2007-2017.  
These findings imply that there is no overflow of HCV infections observed in HIV-positive MSM and HIV-negative MSM starting or using PrEP to the larger HIV-negative MSM population, at least for the time being. Taken together, this evidence adds to the need for routine HCV-screening among MSM if initiating or using PrEP, as well as continued surveillance of HCV infections within the broader HIV-negative MSM population.  
The reasons explaining why HIV-negative PrEP-using MSM acquire HCV may be twofold: their sexual networks are more likely to overlap with those of HIV-positive MSM, in whom HCV is more prevalent, and after initiating PrEP, the protection against HIV provided by PrEP makes the use of other prevention strategies, such as serosorting and condom use, less important. Several studies have indeed shown decreases in condom use during the course of PrEP use, which likely results in increased exposure to HCV and susceptibility to HCV-transmission when STI are present.  
We demonstrated that factors associated with incident HCV infection were related to both sex (receptive CAS with casual partners and a recent anal STI) and drug use (IDU and sharing straws when snorting drugs). These factors are largely comparable to those found in HIV-positive MSM.5  
HCV incidence was higher among those on daily PrEP than among those on event-driven PrEP, most likely resulting from higher levels of sexual risk behaviour in this group. Although several HCV infections were among participants reporting IDU, the majority (9/15) of infections was not related to IDU, indicating that sexual behaviour is, next to IDU, an important target for intervention.  
Similar to our study, 2 studies from France have reported that HCV isolates from HIV-negative MSM, of whom the majority was using PrEP, were infected with HCV strains also found in HIV-positive MSM.  
Our phylogenetic analysis, using longer HCV NS5B fragments and larger subsets of HCV isolates obtained from various risk groups of both MSM and non-MSM, confirms the existence of robust MSM-specific HCV clusters. This emphasizes the overlap between sexual networks of MSM with and without HIV.  
-----------------------  
High incidence of HCV in HIV-negative men who have sex with men using pre-exposure prophylaxis  
Jnl of Hepatology May 2020 - Elske Hoornenborg1,*, Liza Coyer1, Anders Boyd1,5, Roel Christiaan Alfons Achterbergh1,
Maarten Franciscus Schim van der Loeff1,2, Sylvia Bruisten1, Henry John Christiaan de Vries1,3,
Jelle Koopsen4, Thijs J.W. van de Laar6,7, Maria Prins1,2, on behalf of the Amsterdam PrEP
Project team in the HIV Transmission Elimination AMsterdam (H-TEAM) Initiative‡
1Public Health Service of Amsterdam, Department of Infectious Diseases, Amsterdam, the Netherlands; 2Amsterdam University medical
Centers, (UMC), Academic Medical Center, University of Amsterdam, Department of Infectious Diseases, Amsterdam Infection & Immunity
Institute (AI&II), Amsterdam, the Netherlands; 3Amsterdam UMC, University of Amsterdam, Academic Medical Center, Department of
Dermatology, Amsterdam Institute for Infection and Immunity (AI&II), Amsterdam, Netherlands; 4Amsterdam UMC, University of Amsterdam,
Academic Medical Center, Clinical Virology Laboratory, Amsterdam, the Netherlands; 5Stichting HIV Monitoring, Amsterdam, the Netherlands;
6Department of Donor Medicine Research, Laboratory of Blood-borne Infections, Sanquin Research, Amsterdam, the Netherlands; 7Laboratory
of Medical Microbiology, Onze Lieve Vrouwe Gasthuis, Amsterdam, the Netherlands  
Highlights
• HIV-negative men who have sex with men while on pre-exposure prophylaxis are at risk of incident HCV infection.
• High incidence rates of both HCV primary and re-infection were observed.
• Identified HCV risk-factors were similar to those in HIV-positive men who have sex with men.
• Specific clusters of HCV strains were identified in men who have sex with men, with and without HIV.
• HCV-testing for HIV-negative men who have sex with men while on pre-exposure prophylaxis is recommended.  
Background & Aims  
HCV has emerged as a sexually transmitted infection (STI) among HIV-positive men who have sex with men (MSM). We evaluated HCV incidence and its risk factors among HIV-negative MSM using HIV pre-exposure prophylaxis (PrEP).  
Methods  
Participants of the Amsterdam PrEP project were tested for HCV antibodies or HCV-RNA every 6 months. Participants used daily or event-driven PrEP and could switch regimens during follow-up. We calculated incidence rates (IRs) for overall HCV infection and separately for primary and re-infection. A univariable Bayesian exponential survival model was used to identify risk factors associated with incident HCV infection. The HCV NS5B gene fragment (709 bp) was sequenced and compared to HCV isolates from HIV-positive MSM and other risk groups (n = 419) using phylogenetic analysis.  
Results  
Among 350 participants contributing 653.6 person-years (PYs), we detected 15 HCV infections in 14 participants (IR = 2.30/100PY). There were 8 primary infections (IR = 1.27/100PY) and 7 re-infections (IR = 27.8/100PY). IR was 2.71/100PY in daily and 1.15/100PY in event-driven PrEP users.  
Factors associated with incident HCV infection were higher number of receptive condomless anal sex acts with casual partners (posterior hazard ratio [HR] 1.57 per ln increase; 95% credibility interval [CrI] 1.09–2.20), anal STI (posterior HR 2.93; 95% CrI 1.24–7.13), injecting drug use (posterior HR 4.69; 95% CrI 1.61–12.09) and sharing straws when snorting drugs (posterior HR 2.62; 95% CrI 1.09–6.02). We identified robust MSM-specific HCV clusters of subtypes 1a, 4d, 2b and 3a, which included MSM with and without HIV.  
Conclusions  
HIV-negative MSM using PrEP are at risk of incident HCV infection, while identified risk factors are similar to those in HIV-positive MSM. Regular HCV testing is needed, especially for those with a previous HCV infection and those reporting risk factors.  
Lay summary  
We report that hepatitis C virus infections are frequently acquired among HIV-negative men who have sex with men (MSM) using pre-exposure prophylaxis to prevent HIV infection. New infections occurred more frequently in those reporting receptive anal sex without using condoms, having an anal sexually transmitted infection, injecting drugs, and sharing straws when snorting drugs. The viruses found in HIV-negative men using pre-exposure prophylaxis are genetically similar to those in HIV-positive men, but not in other hepatitis C risk groups, suggesting that (sexual) transmission is occurring between HIV-positive MSM and HIV-negative MSM using pre-exposure prophylaxis.
![graphic](../images/051320/051320-2/graphic.gif)
Results  
Study population  
Of the 376 enrolled participants, 26 did not have at least 2 HCV-test results and were excluded. Of the 350 analyzed participants, 348 were MSM and 2 transgender women. At enrolment, median age was 40 years (IQR 33-48), 298 (85.1%) self-identified as white, 261 (74.6%) lived in Amsterdam and 10 (2.9%) reported IDU within 3 months before enrolment (Table 1). Initially, 257 (73.4%) participants chose daily and 93 (26.6%) event-driven PrEP. There were 17/350 (4.9%) participants with HCV antibodies at baseline, 15 of whom were HCV-RNA positive.  
HCV incidence  
Median cumulative follow-up was 2.0 years (IQR 1.9-2.0), totaling 653.6 PY at risk, during which time we identified 12 incident HCV infections and 3 possible incident HCV infections among 14 MSM (Table 2) (overall IR 2.30/100PY, Table S3). Six of 15 infections were diagnosed among 35 participants who reported IDU at least once during follow-up (IR 10.49/100PY). Nine infections were diagnosed among 314 participants who never reported IDU during follow-up (IR 1.51/100PY, Table S3). HCV was not diagnosed among transgender women. Primary infection occurred in 8 of 333 at-risk participants over 628.4 PY of follow-up (IR 1.27/100PY) and re-infection occurred in 7 of 21 at-risk participants over 25.2 PY of follow-up (IR 27.8/100PY). When excluding the 3 possible re-infections, the overall and re-infection IRs were 1.84/100PY and 16.6/100PY, respectively. No participant with incident HCV infection acquired HIV.  
The overall IRs for daily and event-driven PrEP regimens were 2.71/100PY and 1.15/100PY, respectively (Table S3). IRs of primary infection were 1.51/100PY in daily PrEP and 0.61/100PY in event-driven PrEP, and IRs of re-infection 41.4/100PY and 9.3/100 PY, respectively. One person with incident HCV infection (Table 2, participant 11) was likely infected during PrEP discontinuation, while PrEP was re-continued at HCV diagnosis.
Of the 14 MSM with incident HCV infection(s), 13 had chosen daily PrEP and one event-driven PrEP at baseline. At the estimated date of first incident HCV infection, 12 were using daily and 2 event-driven PrEP. For 7/15 incident HCV infections, HCV-RNA was detectable in stored samples 62 to 189 days prior to HCV diagnosis.  
Of the 4 proven re-infections, all had been treated with direct-acting antivirals for their primary HCV infection and had achieved SVR12. The 3 possible re-infections occurred in 2 persons. In these cases, we could not exclude relapse of a previous infection because the 2 viremic episodes were either separated by only one HCV-RNA negative visit (Table 2, participant 13, first infection) or 2 subsequent HCV-RNA negative visits were <24 weeks apart (participant 13 [second infection] and participant 14; Table S4) without a documented switch in genotype.  
Risk factors for incident HCV infection  
We found that HCV incidence was associated with increased number of receptive CAS acts with casual partners in the past 3 months (posterior HR 1.57 per ln increase, 95% CrI 1.09-2.20) and recent anal STI diagnosis (posterior HR 2.93, 95% CrI 1.24-7.13). Higher IRs were also found among those who shared straws when snorting drugs (posterior HR 2.62, 95% CrI 1.09-6.02) and those who injected drugs in the past 12 months (posterior HR 4.69, 95% CrI 1.61-12.09) (Table 3 and Fig. 1). Multivariable analysis was precluded by the small numbers of events.  
Using non-informative priors resulted in CrI with increased uncertainty (Table S5), yet all identified factors for incident HCV infection were retained with the exception of sharing straws when snorting drugs. A stronger prior for daily PrEP regimen and for reporting sex with at least one HIV-positive partner (prior HR 2 for both) did not substantially alter posterior HRs (1.78, 95% CrI 0.95-3.32 and 2.07, 95%CrI 0.92-4.83, respectively). HRs estimated from the non-penalized regression approach resulted in much higher HRs with larger 95% confidence intervals as expected (Table S6). We did not identify risk factors when we restricted the analysis to participants who did not report IDU during follow-up, although all factors had similar directions of effect compared to the main analysis (Table S7). The proportion engaging in HCV risk factors did not increase over time since initiation of PrEP (Table S8). Number of receptive condomless anal sex acts with casual partners, proportion having had anal sex with an HIV-positive partner, having a recent anal STI and sharing sex toys were all lower in the event-driven vs. daily PrEP group (Table S9).  
Hepatitis C virus genotyping and phylogenetic analysis  
Of the 8 primary infections, 6 were subtype 1a, one 2b and one 3a. All 4 proven re-infections were subtype 1a. Of the 3 possible re-infections, one was subtype 4d, one 1a and one could not be genotyped.  
Fig. 2 shows the maximum-likelihood phylogenies of HCV-1a, HCV-2b, HCV-3a and HCV-4d, including all incident and prevalent HCV infections among AMPrEP participants together with HCV strains from HIV-positive MSM, PWID and blood donors in the Netherlands. Phylogenetic trees revealed 10 robust MSM-specific clusters of HCV-1a (n = 7), HCV-2b (n = 1), HCV-3a (n = 1) and HCV-4d (n = 1), which ranged in size from 3 to 53 sequences. HCV sequences from HIV-negative MSM were highly interspersed with those of HIV-positive MSM: 27/29 (93.1%) strains of HIV-negative MSM were part of 10 MSM-specific clusters also containing HIV-positive MSM, one formed a homologous pair with a strain from an HIV-positive MSM, and one strain from an HIV-negative MSM had a singleton sequence.  
Six of 17 participants with prevalent or resolved HCV infection were diagnosed with a re-infection during follow-up. Of those, 2 had resolved their initial infection prior to enrolment, leaving 4 participants for whom we could compare the primary and re-infecting HCV strain. Phylogenetic analysis confirmed that 3 participants indeed had 2 different HCV strains over time: participant 10: HCV 1a-1 and HCV 1a-2, participant 12: HCV 1a-1 and HCV 1a-8, and participant 9: HCV 2b-1 and HCV 1a-1 (Fig. 2). Participant 14 was classified as possibly re-infected and had 2 viraemic episodes with 2 highly similar HCV 4d-1 strains separated by 2 HCV-RNA negative visits more than 2 months apart.
| |
|
|
|
|
|