Decrease in HIV-1 seminal shedding in men receiving highly active antiretroviral therapy: an 18 month longitudinal study (ANRS EP012)

Abstract

Thirty-nine HIV-1-infected men were prospectively tested for HIV seminal shedding after the initiation of highly active antiretroviral therapy. HIV RNA drastically decreased in the semen of all men, but low levels remained detectable in three men at month 18. Proviral HIV DNA became undetectable in the seminal cells of all men after 18 months. HIV-infected cells in the male genital compartment may come from the intermittent passage of blood lymphocytes, rather than constituing a major local reservoir. HIV-1 is recovered as free particles (HIV RNA) in the semen plasma and as proviral DNA in infected non-spermatic cells of 90 and 50% of untreated infected men, respectively. After a 6 month regimen of highly active antiretroviral therapy (HAART), HIV-RNA and HIV-DNA detection rates in semen were reported to be less than 15% [1–3]. The persistence of HIV RNA or proviral DNA in the semen of some men after long-term HAART has been reported in cross-sectional studies [4,5]. However, the dynamics of HIV-1 suppression in the male genital tract in response to long-term HAART is poorly documented. This is probably caused by difficulties in obtaining sequential semen samples from volunteers over long periods of time. The objective of this study was to evaluate prospectively HIV-1 shedding in the male genital compartment in response to long-term HAART. This study (ANRS EP12) was conducted between October 1996 and February 2000 at Cochin and Necker Hospitals, Paris, France. The Cochin Hospital Ethics Committee approved the study. HIV-1-infected men in whom a triple therapy including a protease inhibitor was initiated were asked to participate. Thirty-nine patients were included after providing written informed consent. Their median age was 37.7 years (range 32.1–41.0) and their median CD4 cell count was 323 cells/mm3 (3–711). The Centers for Disease Control and Prevention disease status was A, B, and C for 21, 12 and six men, respectively. Twenty-six patients were antiretroviral naive, and 13 patients had received mono or dual nucleoside analogue therapy. Twenty-eight men were followed for at least 6 months and 18 completed the entire study. The men who interrupted their participation in the study elected to do so because they had to stop therapy (n = 4), did not respond to therapy (n = 2), suffered an adverse drug event (n = 1), or were unable to tolerate the protocol burden (n = 14). Paired blood and semen samples were obtained for each man before the initiation of HAART and after 1, 3, 6, 12 and 18 months of treatment. Seminal fluid, spermatozoa pellet and non-spermatic cell (NSC) fractions were obtained by semen centrifugation over density gradient [6]. Quantifications of HIV RNA in blood and semen plasma and of HIV DNA in peripheral blood mononuclear cells (PBMC) and in NSC were performed as described previously [6,7]. The HIV-RNA detection threshold was 200 copies/ml in blood plasma, and varied from 20 to 400 copies/ml in seminal plasma, as it depended on the polymerase chain reaction inhibitor effect of each sample. The HIV-DNA detection threshold in NSC varied from 5 to 50 copies/106 cells, as it depended on the number of cells available for testing. When analysing RNA detection in the semen fluid as a categorical variable, a threshold value of 400 copies/ml was used because it was the highest threshold value recovered among all seminal plasma tested. MacNemar and Wilcoxon paired tests were used with a significant level of 0.05. HAART drastically reduced the HIV-RNA concentration in blood and in male genital compartments. At inclusion, HIV RNA was detectable in the blood and seminal plasma in 100 and 92% of men, respectively (Table 1). HIV-RNA loads decreased significantly both in the blood and seminal plasma within one month, with a median HIV-RNA decrease from baseline to month 1 in blood and seminal plasma of −2.0 log10 copies/ml (P < 0.001) and −1.8 log10 copies/ml (P < 0.001), respectively. At month 6, only 11% of the patients had detectable HIV-RNA levels in the semen (Table 1). These results are consistent with those of earlier series [2,3,8]. Furthermore, in our study we demonstrated that after 18 months of HAART all patients had HIV-RNA levels below 400 copies/ml. However, low levels of HIV RNA (292, 100 and 10 copies/ml) were still detectable in the semen of three men at month 18. In a cross-sectional study, Dornadula et al. [5] also reported the persistence of low levels of HIV RNA in the seminal fluid of eight men (out of 22) successfully treated with long-term HAART.Table 1: Evolution of HIV seminal and blood shedding after introduction of highly active antiretroviral therapy. HIV-DNA loads in PBMC decreased progressively from baseline to 18 months, although as expected, HIV-DNA detection in PBMC remained positive for all patients throughout the study [9]. At inclusion, HIV DNA was detected in 25 of the 38 NSC fractions tested (66%), with a median viral load of 1.1 log10 copies/million (range < 1 log10–2.7 log10). Comparable positive detection rates of HIV-infected cells in the semen ranging from 40 to 60% with low HIV-DNA loads were reported in others series [10–13]. The rate of positive HIV-DNA detection in NSC decreased during the study; however, a significant decrease was only achieved at month 18 when HIV DNA was undetectable in all NSC fractions tested (Table 1). Vernazza et al. [1] also showed that patients with detectable HIV-DNA levels in semen had a shorter duration of therapy (median 5.3 months) than patients with negative HIV-DNA detection in the semen (median 9.2 months). Zhang et al. [4] reported the persistence of proviral DNA in the seminal cells of two patients treated with triple therapy for more than 2 years. However, we demonstrate here that in most men, HAART is efficient at suppressing HIV-infected cells to undetectable levels in the male genital compartment, although long-term therapy is necessary to achieve this result. These observations suggest that in most men the male genital tract may not be a major reservoir of HIV latently infected cells. HIV-infected cells in the semen may be mainly the result of the intermittent passage of lymphocytes from blood to the male genital tract. However, it cannot be excluded that latently infected lymphocytes are sequestered in the prostate or in the seminal vesicles, as it had been suggested [14]. In conclusion, even after successful long-term HAART, low residual levels of HIV shedding may be detected in some men. Therefore recommendations for consistent condom use cannot be changed. However, long-term HAART induces the suppression of HIV shedding in the semen of most men. This is likely to reduce drastically HIV sexual transmission and should thus have a significant public health impact. Marianne Leruez-Villea Emmanuel Dulioustb Dominique Costabliolac Dominique Salmond Anne Tacheta Laurent Finkielsztejnd Marta De Almeidab Benjamin Silbermannd Didier Sicardd Pierre Jouannetb Christine Rouziouxa