| |
Persistent HIV RNA shedding in semen despite effective antiretroviral therapy
|
| |
| |
AIDS:
24 September 2009
"Isolated semen HIV shedding was detected at at least one visit for 12 of 25 (48%) participants, and was high level in four of 25 (16%)...despite suppression in blood"...."All but one participant reported excellent medication compliance and had consistent bVL suppression"
"The small sample size makes it difficult to draw firm conclusions regarding the effect of drug class and semen drug concentration. However, isolated semen shedding despite effective ART was observed in almost half the participants, and so we are confident regarding the validity of our major observation. Indeed, we only assessed cell-free virus (HIV RNA) and, as cell-associated HIV may persist in semen despite therapy [11], our study may have underestimated the infectious potential of semen on ART....
....In conclusion, effective ART is likely to substantially reduce HIV transmission at a population level, but substantial interindividual heterogeneity in sVL despite an undetectable bVL suggests that some individuals may remain sexually infectious. Further studies are urgently needed to assess whether isolated semen HIV RNA shedding may result in sexual transmission."
Sheth, Prameet M; Kovacs, Colin; Kemal, Kimdar S; Jones, R Brad; Raboud, Janet M; Pilon, Richard; la Porte, Charles; Ostrowski, Mario; Loutfy, Mona; Burger, Harold; Weiser, Barbara; Kaul, Rupert; the Toronto Mucosal Immunology Group
aDepartment of Medicine, Canada
bDalla Lana School of Public Health, Canada
cDepartment of Immunology, University of Toronto, Canada
dMaple Leaf Medical Clinic, Toronto, Canada
eWadsworth Center, New York State Department of Health, Albany, New York, USA
fCenter for Communicable Diseases and Infection Control, Public Health Agency of Canada
gUniversity of Ottawa, Ottawa Hospital and Ottawa Health Research Institute, Ottawa
hCanadian Immunodeficiency Research Collaborative
iDepartment of Medicine, Women's College Hospital
jUniversity Health Network, Toronto, Canada.
Abstract
Effective antiretroviral therapy (ART) may reduce HIV sexual transmission by lowering genital HIV levels. A prospective study of men starting ART (n = 25) demonstrated rapid, substantial reductions in semen HIV RNA. However, despite an undetectable blood viral load, isolated semen HIV shedding was detected at more than one visit in 12 of 25 (48%) participants, with semen HIV RNA levels exceeding 5000 copies/ml in four of 25 (16%). Isolates were drug-sensitive, and this phenomenon was not associated with semen drug levels or regimen.
There were an estimated 2.5 million new HIV infections in 2007 [1], most acquired through sex [2]. The blood HIV RNA viral load (bVL) correlates with both the probability of sexual HIV transmission [3] and with genital HIV levels [4,5]. Effective highly active antiretroviral therapy (ART) results in an undetectable bVL and may reduce HIV transmission at a population level [4-6]. However, the impact of ART on genital HIV levels can be heterogeneous [7-11] and so it is controversial whether effective ART will consistently render an individual sexually non-infectious [12,13]. Statistical models suggest that, in the absence of condom use, HIV incidence could increase despite substantial reductions in the genital viral load on ART [14], although real world data to inform such models are lacking. We performed a prospective, longitudinal study to better define the impact of ART initiation on semen HIV RNA levels.
Participants were HIV-infected, therapy-naive men starting ART through an HIV primary care clinic, based upon current clinical guidelines [15]. Blood and semen were collected at weeks 0, 2, 4, 8, 12, 16, 20 and 24. Semen was collected by masturbation into 10 ml of transport medium, semen/blood plasma isolated and HIV RNA assayed by Versant HIV-1 RNA 3.0 assay (bDNA; Bayer Diagnostics, Puteaux Cedex, France) [16,17]. A semen viral load more than 5000 RNA copies/ml was defined as 'high level'. Screening for Neisseria gonorrhea, Chlamydia trachomatis, syphilis, herpes simplex type 2 (HSV-2) and cytomegalovirus (CMV) was performed at each visit [16,17]. HIV RNA was extracted from blood/semen plasma, cDNA synthesized and the protease and reverse transcriptase regions of HIV-1 pol amplified [18,19]. Phylogenetic trees were constructed by neighbor-joining, after gap-stripping of columns. Genotypic drug resistance was determined through the Stanford HIV Drug Resistance Database (http://hivdb.stanford.edu). Concentrations of efavirenz, lopinavir, atazanavir, ritonavir, abacavir, zidovudine and lamivudine (3TC) were determined using a validated high performance liquid chromatography method [20,21]. Semen drug and HIV RNA levels were multiplied by six to correct for dilution [16,17]; viral load sensitivity cut-offs were less than 50 HIV RNA copies/ml for bVL and less than 300 HIV RNA copies/ml for semen HIV RNA viral load (sVL). SPSS, version 12 for Windows (SPSS Inc., Chicago, Illinois, USA) was used for statistical analysis using nonparametric comparisons, with values reported as median (range). The study protocol was approved by the University of Toronto Research Ethics Board.
Twenty-five participants were enrolled (Table 1) and none had a clinical/laboratory sexually transmitted infection (STI) or urethritis at any visit. HSV-2 seroprevalence was 9/25 (36%) and cytomegalovirus (CMV) seroprevalence 25/25 (100%). All participants achieved an undetectable bVL by week 16 of ART, and all but two participants had an undetectable sVL (Fig. 1a): one of these never suppressed sVL despite effective ART (Fig. 1b), whereas the other admitted suboptimal compliance after bVL rebound, with subsequent complete bVL and sVL suppression. However, there was significant heterogeneity in the consistency of sVL suppression. Isolated semen HIV shedding was detected at at least one visit for 12 of 25 (48%) participants, and was high level in four of 25 (16%; Fig. 1b-c). Isolated semen shedding was present during 19 of 116 (16.4%) study visits with an undetectable bVL, and at a high level (6672-16026 RNA copies/ml) in five of these nineteen. Phylogenetic analysis showed that all sequences were unique and that blood and semen sequences were closely related but distinct (data not shown).
No association was apparent between isolated semen HIV shedding and specific antiretroviral agents or classes (Table 1). Seven (7/12) participants with isolated HIV shedding were taking a nonnucleoside reverse transcriptase inhibitor (NNRTI) vs. five of 13 without (P = 0.32); three of four with high-level isolated semen shedding were taking a boosted protease inhibitor (P = 0.31). Baseline clinical parameters, including HSV-2 serostatus, did not predict subsequent isolated semen shedding. In participants with isolated HIV semen shedding, the median pretherapy bVL was 4.65 log10 RNA copies/ml (vs. 4.50; P = 0.86); the baseline CD4+ T cell count was 195 (vs. 230/µl; P = 0.97). However, the baseline sVL was 10-fold higher in those with isolated semen shedding than in those without detectable isolated shedding (median 4.42 vs. 3.41 log10 RNA copies/ml; P = 0.03).
No resistance mutations to the three major drug classes were detected at baseline in blood or semen in any participant, or in semen during high-level isolated shedding. Assay limitations did not permit analysis of semen isolates when sVL was less than 5000 RNA copies/ml, or in one participant despite high-level isolated shedding. Drug levels in paired blood-semen plasma samples were assayed for 19 of 25 participants, at the visit corresponding with the peak semen HIV RNA level (if isolated semen shedding) or after 5 months (for nonshedders). As expected based on a previous study [22], concentrations of efavirenz were substantially lower in semen than blood (0.30 vs. 1.19 mg/l; P < 0.001) and semen levels of the protease inhibitors were very low or undetectable; 3TC concentrations were more than 100-fold higher in semen (4.96 vs. 0.01 mg/l; P = 0.01). However, there was no association between isolated semen shedding and the semen detection and/or level of any drug (data not shown).
Semen HIV sVL was then assayed at a single visit in 13 STI-free, long-term ART-treated men, with an undetectable bVL for median 82 months (range: 48-216 months). Isolated semen HIV RNA shedding was detected in four of 13 participants (31%; median 564 RNA copies/ml, range: 336-828 copies/ml), both those on NNRTI-based and boosted protease inhibitor-based regimens (n = 2 for each).
The present prospective study confirms that effective ART dramatically reduces HIV semen RNA levels [8,23] and should substantially reduce sexual transmission at a population level. However, isolated semen HIV RNA shedding was detected in many participants despite suppression in blood. Standard baseline clinical tests, drug regimen and semen drug levels were not useful predictors, whereas the pretherapy semen (not blood) viral load was predictive. All but one participant reported excellent medication compliance and had consistent bVL suppression. Isolated semen HIV shedding was not related to a lack of semen antiretroviral activity, as sVL had initially become undetectable in most (11/12) participants. In this short-term study, isolated semen shedding was not associated with compartmentalized drug resistance.
The small sample size makes it difficult to draw firm conclusions regarding the effect of drug class and semen drug concentration. However, isolated semen shedding despite effective ART was observed in almost half the participants, and so we are confident regarding the validity of our major observation. Indeed, we only assessed cell-free virus (HIV RNA) and, as cell-associated HIV may persist in semen despite therapy [11], our study may have underestimated the infectious potential of semen on ART.
In conclusion, effective ART is likely to substantially reduce HIV transmission at a population level, but substantial interindividual heterogeneity in sVL despite an undetectable bVL suggests that some individuals may remain sexually infectious. Further studies are urgently needed to assess whether isolated semen HIV RNA shedding may result in sexual transmission.
|
|
| |
| |
|
|
|
