Highly active antiretroviral therapy does not completely suppress HIV in semen of sexually active HIV-infected men who have sex with men

Introduction

Approximately 33.3 million people worldwide are living with HIV/AIDS, and 1.8 million deaths and 2.6 million new infections occur annually [1]. Unprotected intercourse is the most common route through which HIV-1 is transmitted, and semen of HIV-infected men is an important source of infectious HIV [2]. Whereas the HIV/AIDS epidemic in Sub-Saharan Africa is generalized with approximately equal percentages of infections occurring in men and women, the epidemic in the US and many other developed countries is concentrated in men who have sex with men (MSM) [3] Recent reports suggest that MSM populations in lower and middle income countries are also disproportionately infected with HIV in several locations [4].

MSM are highly susceptible to HIV transmission. In the pre-antiretroviral therapy (ART) era, the rate of HIV transmission among MSM was estimated to be 1 in 100 acts of unprotected anal intercourse [5]), whereas the overall rate of HIV sexual transmission from men to women was much lower (<1 in 1,000 acts of unprotected vaginal intercourse [6]). High peripheral blood and genital HIV viral loads are associated with an increased rate of HIV transmission [7, 8], and ART, which reduces blood and genital viral load, has been shown to decrease HIV transmission [7, 9-11]. Yet, evidence suggests that the MSM epidemic has had a resurgence in the era of potent antiretroviral therapy [12, 13].

Genital infections with common STI pathogens including HSV-2, Neisseria gonorrhoeae and Chlamydia trachomatis have been associated with increased HIV viral load in semen and enhanced HIV transmission in ART-naïve men [14]. Several recent studies have documented the persistence of HIV virions and infected cells in semen from men on stable ART regimens despite undetectable viral loads in peripheral blood {reviewed in [15]}, and HIV transmission events have also been reported [16, 17]. Viral persistence in semen may be attributable to HIV compartmentalization in the male genital tract [18], isolated viral replication due to incomplete penetration of antiretroviral drugs [15], and/or activation of resident or infiltrating HIV-infected cells by inflammatory mediators produced locally in the genital tract in response to either symptomatic or asymptomatic infections [19]. MSM on HAART may continue to shed HIV in semen because of continued high risk sex behavior and a high prevalence of STIs and genital inflammation [20-22].

To determine the prevalence of seminal HIV shedding in sexually active HIV-infected MSM on HAART, and risk factors associated with detection of HIV in semen, we recruited participants from a Lesbian/Gay/Bisexual/Transgender (LGTB) clinic in Boston, MA, that provides ART to a substantial population of HIV-infected MSM at high risk for STIs [22]. We quantified levels of HIV in paired blood and semen samples, and examined the relationship between HIV in semen and an array of clinical, behavioral and biological variables.

Methods

Results

Discussion

Because HIV transmission by semen is a major factor driving the AIDS epidemic, it is important to define variables that determine shedding of infectious HIV into semen. Concomitant STIs, genital inflammation and high peripheral blood viral loads have been associated with seminal HIV shedding in the past, largely in pre-HAART populations [14, 28]. As more men initiate ART and HAART, it is crucial to understand the interplay of therapy, which can suppress blood and genital HIV viral loads, with risk factors that can enhance HIV replication in the genital tract, and the potential for sexual transmission of drug-resistant HIV. Recent evidence indicates that HAART suppresses HIV transmission [7, 10]. The HPTN 052 clinical trial [9] demonstrated that early initiation of ART was associated with a 96% reduction in HIV transmission in HIV discordant heterosexual couples with a low prevalence of STIs. The only study on the effect of HAART on HIV transmission in an MSM population, an observational study conducted in San Francisco over a decade ago, concluded that the transmission rate was decreased by approximately 60% [29].

In our cross-sectional study of sexually active, ART-experienced MSM on a current HAART regimen for >3 mos, 30% of the men had detectable HIV RNA and/or DNA in semen. Eighteen percent of the men had detectable HIV in blood plasma despite being on a HAART regimen, consistent with earlier studies that have shown that substantial numbers (up to 25%) of men will experience virologic failure while on a HAART regimen [30]. These men, as expected, had a significantly higher prevalence as well as increased copy numbers of HIV in semen than did the BP-HIV⁻ men in our study, indicating, as has been shown before [2], that HIV levels in peripheral blood are an important predictor of seminal HIV.

Among the 83 men with undetectable HIV in blood plasma, 25% had HIV in semen. This is a higher prevalence of seminal HIV shedding than has been reported from other large recent studies of BP-HIV⁻ men on HAART, which have reported 2-3% shedding rates [31, 32]. This is likely due to the high prevalence of STIs and genital inflammation in our sexually active MSM cohort; 9.6% of the men were diagnosed with STI or urethritis, and 24% had genital inflammation (leukocytospermia). Our study provides evidence that genital infections and inflammation are common in HIV-infected MSM that engage in unprotected intercourse, and that these factors can promote compartmentalized shedding of HIV in the genital tract of men on suppressive HAART therapy. We also used a highly sensitive detection assay and measured both cell-free and cell-associated HIV. Cell-associated HIV was included because early reports suggested that HIV-infected cells persisted in semen longer than cell-free HIV after initiation of HAART [24, 33], and also because of a recent resurgence in interest in the sexual transmission of cell-associated HIV [34]. CF-RNA and HIV-infected cells were each detected in 13% of BP-HIV⁻ cases, but only 1/21 subjects with seminal HIV was positive for both forms of virus. This provides further evidence that cell-free and cell-associated HIV in semen may arise from different sources, as has been suggested in previous reports [35].

In our study, seminal HIV viral copy numbers ranged from 80-2,560 (median 200 copies) in men with undetectable HIV in blood. These seminal HIV levels are considerably lower that those commonly detected in ART-naïve men [24, 36, 37], but could represent an infectious innoculum in MSM since rectal intercourse is an especially effective route of HIV transmission due to the thin rectal epithelium [5, 6]. A series of recent studies substantiate this possibility. A study in macaques comparing the relative transmissibility of SHIV across different mucosal surfaces demonstrated that rectal transmission could be achieved with 5-fold less SHIV than needed for vaginal transmission [38]. A theoretical paper [39] and results from a clinical study [8] both conclude that <1,000 copies of HIV RNA in semen pose a low but real risk of male-to-female HIV transmission; a five-fold reduction in this copy number for rectal transmission (<200 copies) is within the range of values detected in our study. Furthermore, Butler et al. [40] reported a median seminal plasma viral load of 4,300 HIV RNA copies/ml for men transmitting HIV to their MSM partners with a minimum and maximum of 110 and 69,000 HIV RNA copies/ml, respectively. Therefore, the data from our study in combination with these recent reports suggest that sexually active HIV–infected MSM on ART that continue to shed low levels of HIV in semen may be infectious to their sexual partners.

One concern about seminal HIV shedding in men on HAART is that the virus can develop antiretroviral resistance mutations [24, 41, 42], which may contribute to the high prevalence of antiretroviral drug-resistant HIV in ART-naïve HIV-infected individuals in the United States (reported to range from 8 to 24% [43-45]). MSM are at higher risk of acquiring drug-resistant HIV than heterosexual men or women [43, 45]. Our data suggest that HIV replication commonly occurs in the genital tract of sexually active MSM under evolutionary pressure from HAART; this could increase the prevalence of drug-resistant HIV in semen and its transmission to seronegative partners. Although early attempts to culture and sequence HIV in semen from men in this study failed due to low viral copy numbers, further studies are warranted to determine whether HIV in semen from MSM on HAART is infectious in vitro and contains drug resistance mutations..

The multivariate analysis also revealed an independent association between seminal HIV and UIAS-HIV. This raises the possibility that seminal HIV in subjects engaging in UIAS with HIV-infected partners may be attributable to urethral superinfection or contamination with HIV from rectal secretions of sex partners. Reports of high numbers of HIV-target cells in the urethra [46], and HIV-infected cells in urethral secretions from HIV-seropositive men with and without urethritis [47] indicate that the urethra may be a primary HIV infection site. Furthermore, higher concentrations of HIV RNA have been reported in rectal mucosa secretions than in blood and seminal plasma among MSM, and may be independent of ART [48]. Given that many HIV⁺ MSM engage in unprotected anal intercourse with other HIV-infected partners as a way to experience pleasure without having to worry about transmitting HIV to others [20], the potential of acquiring HIV from an HIV⁺ partner during unprotected insertive sex, and subsequently transmitting it to an HIV⁻ partner must be considered. This finding needs to be confirmed, since at present “serosorting” (HIV-infected individuals only having unprotected sex with other infected persons), is considered by some to be a reasonable public health strategy to prevent transmission to uninfected partners [49].

In conclusion, we detected a high prevalence of seminal HIV shedding in a cohort of sexually active HIV-infected MSM on HAART. HIV in semen was associated with detection of HIV in blood (due to poor drug adherence and/or virologic HAART failure), and with STIs and genital inflammation in men who were fully suppressed (undetectable HIV in blood). In light of recent evidence that even low amounts of HIV in semen could pose a transmission risk in MSM, who are more vulnerable to HIV infection than heterosexual men, this information has potential clinical significance for the HIV epidemic in MSM. HIV-infected men who engage in unprotected intercourse may use HAART and viral load status in their sexual decision-making, and being on HAART or having an undetectable blood viral load may relax concerns about transmitting HIV [50]. Therefore, MSM at risk for transmitting HIV may believe that they have a low risk based on incorrect assumptions that HAART eliminates HIV from semen. Until more information on transmission risk in MSM is available, it would be prudent to advise sexually active HIV-infected MSM to use condoms and other risk-reduction strategies throughout all stages of HIV disease regardless of HIV treatment status, and to promote the aggressive diagnosis and treatment of STIs.