Share this article on:

Factors of intermittent HIV-1 excretion in semen and efficiency of sperm processing in obtaining spermatozoa without HIV-1 genomes

Bujan, Louisa,b; Daudin, Myriama,b; Matsuda, Tomohirob; Righi, Laurenced; Thauvin, Laurencea,b; Berges, Laetitiad; Izopet, Jacquesd; Berrebi, Alainc; Massip, Patricee; Pasquier, Christophed

Clinical Science

Objective: To study the risk factors for HIV-1 in semen according to the localization of HIV-1 in sperm cell fractions and to assess the efficiency of sperm processing in obtaining spermatozoa without HIV-1 genomes.

Methods: Ninety-four HIV-infected patients provided 281 paired blood and semen samples. Sperm cell separation was performed using two successive methods. HIV-1 RNA was quantified in blood and seminal plasma. HIV-1 RNA and DNA were detected in cell fractions.

Results: HIV-1 RNA was found in 14% of seminal plasma samples and up to 8.7% of native semen cells were positive for HIV-1 RNA and DNA. Ten seminal plasma samples had detectable RNA although blood viral load was undetectable. Antiretroviral treatment reduced the likelihood of RNA detection in seminal plasma. For semen with polynuclear cells and HIV-1 RNA in seminal plasma, the likelihood of detecting HIV-1 genomes in semen cells was increased fourfold and sixfold, respectively. In 25% of patients, HIV-1 excretion was intermittent. In the group of patients with systematic negative seminal plasma, HIV-1 genomes were detected in up to 10% of sperm cell samples. Our method of sperm processing always enabled us to obtain spermatozoa without detectable HIV-1 genomes.

Conclusions: Polynuclear cells in semen are a risk factor for seminal HIV-1 excretion. Blood viral load was the only predictive factor for the intermittence of HIV-1 excretion in semen over time. Sperm processing using two successive methods was effective in obtaining spermatozoa without detectable HIV-1 genomes regardless of the viral load level in native semen.

From the aCECOS Midi-Pyrénées, the bResearch Group on Human Fertility and the cFédération de Gynécologie-Obstétrique, University Hospital Paule de Viguier, the dVirology Sevice and the eService for Infectious and Tropical Diseases, University Hospital Purpan, Toulouse, France.

Correspondence to L. Bujan, CECOS Midi-Pyrénées and Research Group on Human Fertility, University Hospital Paule de Viguier, TSA 70034, 31059 Toulouse Cedex 9, France.

Received: 30 July 2003; revised: 13 October 2003; accepted: 4 November 2003.

Back to Top | Article Outline


HIV-1 infection is a sexually transmitted disease that affects at least 42 million individuals in the world [World Health Organization (WHO) figures for 2002]. HIV-1 may be present in the human male genital tract [1]. Transmissibility of HIV-1 depends on the type of sexual exposure, the level of infectivity of the infected partner and the receptivity of the non-infected partner (mucosal disruption, genital infection, etc.) [2,3]. Sexual transmission is also related to blood viral load [4,5]. Since the virus has been found in semen [6,7], several studies have reported the presence of HIV-1 genomes or infectious virus in seminal fluid or sperm cells from HIV-1-infected men [8–12]. The seminal viral load has been correlated with blood viral load in many studies [11–14] but not in all [12,15]. Evidence for compartmentalization of HIV-1 between semen and blood has been found [16–20], suggesting that HIV-1 production in semen is regulated in a different manner to that in blood plasma.

Antiretroviral therapy can reduce blood and semen viral load. Zidovudine decreased the frequency of HIV-1 detection in semen [8]. More recently, highly active antiretroviral treatment (HAART), which has changed the prognosis of HIV infection, has been shown to have an effect on the seminal tract, reducing the semen viral load [10,14,21–23]. However, it is noteworthy that replication-competent viruses can be recovered from seminal cells of HIV-1-infected men who are receiving HAART and who have undetectable levels of HIV-1 RNA in blood plasma [18,24].

Reproductive tract infections are associated with susceptibility to HIV-1 sexual transmission [3]. Detection of HIV-1 in semen has been associated with increased numbers of leukocytes in semen [8,25], urethritis and asymptomatic prostatitis [26,27]. Moreover, treatment of urethritis has been reported to decrease HIV-1 concentration in semen [28–30].

The question of intermittent or continuous shedding of HIV-1 in semen is important in terms of sexual transmission risk. Krieger et al. [31] reported intermittent shedding of HIV-1 in semen but used low sensitivity methods, which could explain the negative results. More recently, a 10-week longitudinal study of 18 untreated patients showed that HIV-1 RNA was detected intermittently in the semen of eight patients [32].

Finally, HIV-1 infection affects a young population of reproductive age. For serodiscordant couples who do not wish to infect the partner, use of condoms makes natural conception impossible (behaviour-induced sterility). Following improvement of HIV-1 infection prognosis owing to HAART, several couples have requested medical assistance to conceive with risk of HIV-1 transmission reduced to a minimum. In a study of a large nationally representative sample of HIV-1-infected adults in the United States, 28% of HIV-1-infected men stated that they wished to have children in the future [33]. The presence of HIV-1 in spermatozoa has been a matter of debate [34], with several reports arguing against this [35–37]. Since 1993, spermatozoa isolated from semen of HIV-1-positive men have been used for medically assisted procreation in serodiscordant couples [38]. The processing methods must make it possible to obtain spermatozoa without detectable HIV-1 genomes [12].

Studies of HIV-1 excretion in semen and efficiency of sperm processing have implications for health policy concerning avoidance or decrease of the HIV-1 sexual transmission risk and for techniques to enable serodiscordant couples, with a male infected partner, to have a child without infection of the female partner. The present study of 94 men, who provided 281 semen and blood samples, examined (i) the relationship between blood and semen viral load as well as between sperm parameters and the detection of HIV-1 RNA and DNA in the various fractions of semen; (ii) the prognostic factors for excretion of HIV-1 in semen and its pattern of intermittence; and (iii) the efficiency of sperm preparation methods in obtaining spermatozoa without detectable HIV-1 genomes.

Back to Top | Article Outline



Ninety-four HIV-1-infected men attending the CECOS Midi-Pyrénées and Male Sterility Center (Hospital La Grave, Toulouse) provided 281 paired blood and semen samples between April 1998 and January 2001. The number of samples provided by the patient was variable: one sample for 25 patients, two for 19, three for 14, four for 20, five for 10, six for 2, seven for 2, nine for 1 and 11 for 1 patient. At the first examination, median age was 37 years (range, 25–50) and the median duration of HIV-1 infection was 144.6 months (range, 10.3–238.7). They were all clinically asymptomatic. The mode of transmission was intravenous drug use in 36 (38.3%), blood transfusion in 10 (10.6%), homosexual or heterosexual intercourse in 36 (38.3%) and unknown for 12 (12.8%). Eighty-seven (92.5%) patients were taking antiretroviral therapy, 16 (17.0%) were receiving two nucleoside inhibitors and 71 (75.5%) were receiving three or more drugs.

This study was carried out in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board (Comité de Protection des Personnes dans la Recherche Biomédicale Toulouse II, CHU Toulouse, France). All patients gave their informed consent.

Back to Top | Article Outline

Clinical specimens

Semen samples (281) were obtained by masturbation, after a recommended 3-day abstinence period, and collected in sterile containers. The samples were processed in the laboratory within 2 h of ejaculation, according to a previously published method [12]. Seminal plasma and whole sperm cells were isolated from a sample by centrifugation at 11 000 × g. Spermatozoa were prepared from whole semen by differential density gradient centrifugation over 50, 70 and 90% PureSperm (JCD S.A., Lyon, France). The pellet (90% PureSperm fraction) was washed twice with BM1 medium (Elios Biomedia, Paris, France). The 50% fraction contained mostly low-motility, dead or abnormal spermatozoa and seminal non-spermatozoa cells. The motile fraction was further separated by the ‘swim-up’ method from the washed 90% fraction. Briefly, the spermatozoa pellet was overlaid with 1.1 ml medium and incubated at 37°C in 5% CO2 for 60 min in a tube at an angle of 45° to allow the motile spermatozoa to swim up. The swim-up fraction contained only viable motile spermatozoa. The whole semen cells, 50% fraction cells and the spermatozoa obtained after swim-up were collected and analysed for viral genome. All samples were stored as dry pellets at −80°C.

Back to Top | Article Outline

Sperm analysis

Sperm parameters were assessed according to the WHO laboratory recommendations [39]. Before May 2000, the semen polymorphonuclear granulocyte count was performed using the peroxidase staining method when semen round cell count was > 1 × 109 cells/l, as recommended by WHO. After May 2000 it was performed systematically. As the total cell count in the ejaculate is the best marker of cell production, it was calculated from the sperm parameters. Total sperm count was the product of the sperm count per unit volume multiplied by the ejaculate volume; total round cell count was a product of round cell count per unit volume multiplied by the ejaculate volume, and the total polynuclear count was the product of polynuclear cells count per unit volume multiplied by the ejaculate volume.

Back to Top | Article Outline

Assessment of HIV-1 in blood

Plasma HIV-1 RNA was quantified with the Amplicor HIV-1 Monitor v1.5 assay (Roche Diagnostic Systems, Meylan, France) using the ultrasensitive protocol (detection limit 20 copies/ml).

Back to Top | Article Outline

HIV-1 nucleic acid detection in seminal plasma

The Amplicor HIV-1 Monitor v1.5 assay (Roche Diagnostic Systems) was used to quantify HIV-1 RNA in seminal plasma as previously described [12]. The blood plasma Internal Quality Standard (IQS) was added to 100 μl seminal plasma before HIV-1 RNA extraction. RNA was then extracted from seminal plasma using Nuclisens NASBA Diagnostics extraction kit (BioMérieux SA, Marcy l'Etoile, France) as recommended by the manufacturer. All extracted HIV-1 RNA was processed using the Monitor ultrasensitive assay protocol. Controls were included in each run. The detection limit of the assay in seminal plasma was 100 copies/ml. Specificity of the assay was 100% on a HIV-1-negative seminal plasma panel of 20 samples.

Back to Top | Article Outline

HIV-1 nucleic acid detection in seminal cells

HIV-1 RNA in seminal cells was detected using a modified HIV-1 Monitor v1.5 assay (Roche Diagnostic Systems), as previously described [12]. Since no systematic DNAase treatment was performed on extracted nucleic acids, both HIV-1 RNA and DNA could be detected. Frozen pellets of 2 × 106 cells were suspended in 600 μl Monitor lysis buffer containing IQS. Extraction was performed using Nuclisens extraction kit as recommended by the manufacturer. RNA was processed by the reverse transcriptase polymerase chain reaction (RT-PCR) and products were detected using the Monitor kit. The assay detection limit for HIV-1 genome in seminal cells was 20 copies/106 cells. The results of this test are presented as ‘DNA + RNA’ detection in the tables.

The Amplicor assay (Roche Diagnostic Systems) was used for the specific detection of HIV-1 DNA. Frozen pellets of 2 × 106 cells were suspended in 600 μl Monitor Lysis Buffer, submitted to thermal shock (three rounds of 15 s in liquid nitrogen and 30 s at 60°C) and incubation at 60°C for 1 h before inactivation of proteinase K at 95°C for 30 min. A total of 10 μl lysis solution was used for PCR amplification and detection by hybridization on microplaque, as recommended by the manufacturer. Positive and negative controls were included in each reaction set. Presence of PCR inhibitors was systematically tested by performing a home-made β-globin PCR on each extracted DNA. The assay detection limit for HIV-1 DNA in seminal cells was 5 copies/106 cells. The results of this test are presented as ‘DNA detection’ in the tables.

HIV-1 genome detection was performed in whole sperm cells (native sperm cells) obtained during sperm processing in the 50% fraction of the density gradient centrifugation and in the final spermatozoa fraction obtained after the second method of sperm processing, the swim-up method. As a sample could be positive for all tests or for one test in one cell fraction, the results are also presented as ‘at least one detection’ in a sample whatever its localization in the semen cells or the test used.

Back to Top | Article Outline

Statistical analysis

The non-parametric Mann–Whitney test was used to compare all the quantitative semen quality data, such as RNA levels in semen and blood plasma, CD4 cell count or age. Fisher's exact test was used to compare qualitative data between subgroups. The odds ratio and 95% confidence interval were also calculated to observe changes in the likelihood of HIV-1 nucleic acid detection in proportion to the semen quality. The degree of correlation between quantitative variables was examined using Spearman's rank correlation coefficient. All statistical analysis used (1 − β) = 80% and an α level of 5% according to the STATA 6.0 software (Stata Corp., College Station, Texas, USA).

Back to Top | Article Outline


HIV-1 detection in blood samples

Sixty-eight (72.2%) of the 94 patients had detectable HIV-1 RNA levels in blood. HIV-1 RNA was detected in 151 of 281 blood samples (53.7%). The median HIV-1 RNA level in blood plasma was 123 copies/ml (range, 3–130 000). Among the 151 positive blood samples, the viral load was < 1000 copies/ml for 114 (75.5%) samples, 1000–10 000 copies/ml for 30 (19.9%) and ≥ 10 000 copies/ml for seven (4.6%) samples.

Back to Top | Article Outline

HIV-1 detection in seminal plasma

HIV-1 RNA was not detected in 233 seminal plasma samples. In 10 samples, HIV-1 RNA detection was impossible because of the presence of PCR inhibitors (n = 8) or insufficient volume of seminal plasma (n = 2). HIV-1 RNA was detected in 38 samples. The median level of HIV-1 RNA in seminal plasma was 201 copies/ml (range, 5–277 500 copies/ml). The seminal viral load was < 1000 copies/ml in 28 (73.7%), 1000–10 000 in seven (18.4%) and ≥ 10 000 copies/ml in three (7.9%) samples.

HIV-1 RNA concentrations in blood and seminal plasma were not correlated (r = −0.02; P > 0.05). When blood viral load was detectable, 19.4% of seminal plasma samples were positive for HIV-1 RNA, although when blood viral load was undetectable 7.9% of seminal plasma were positive for HIV-1 RNA (P < 0.01) (Fig. 1). The median seminal viral load was 272.5 copies/ml (range, 18–277 500) and 143.5 copies/ml (range, 5–5600) when the blood viral load was detectable or undetectable, respectively (P > 0.05).

Fig. 1.

Fig. 1.

Ten seminal samples had detectable viral load when the blood viral load was undetectable (Fig. 1). All the patients with detectable HIV-1 RNA in seminal plasma and undetectable blood levels were receiving HAART. Twenty-eight blood and seminal paired plasmas were RNA positive, and in 10 the seminal viral load was higher than the corresponding blood plasma viral load.

Back to Top | Article Outline

HIV-1 genomes in sperm cells

Different sperm fractions were studied (Table 1): whole semen cells, intermediate 50% fraction cells and final spermatozoa fraction obtained after use of the two preparation methods. In order to distinguish between DNA and RNA, a second PCR technique for specific detection of HIV-1 DNA was used.

Table 1

Table 1

In cell fractions (native sperm or the intermediate 50% fraction), HIV-1 genome (DNA + RNA or DNA) was detected in 15.3% of samples.

Using the first technique, HIV-1 DNA + RNA was detected in 24 (8.7%) of cell samples from native semen before sperm processing. Using the second technique, HIV-1 DNA was detected in 15 (5.5%) of these cell fractions (Table 1). In the 50% fraction, which is the intermediate fraction obtained during density gradient centrifugation, the first step of sperm processing, HIV-1 DNA + RNA and DNA were detected in 5.6 (n = 269) and 5.2% (n = 270), respectively, of the samples analysed (Table 1).

Back to Top | Article Outline

HIV-1 genomes in spermatozoa obtained after sperm processing

Whatever the HIV-1 genome detection method employed, neither HIV-1 RNA nor DNA were detected in spermatozoa from the final fraction obtained after swim-up (Table 1). Absence of detectable HIV-1 RNA and DNA was the rule in our prepared spermatozoa samples, regardless of the results of HIV-1 RNA detection in seminal plasma or blood plasma, or of HIV-1 RNA and DNA detection in native or 50% fraction sperm cells. It was noteworthy that the patient who had the highest seminal HIV-1 RNA level (277 500 copies/ml) had no detectable HIV-1 genomes in the final spermatozoa fraction.

Back to Top | Article Outline

Semen quality

In the 281 semen samples, median volume was 3.1 ml (range, 0.6–9.6), sperm count was 79 × 109 cells/l (range 2–550) and total sperm count was 243 × 109 cells/l (range, 4.54–3150). Nine samples from seven patients showed a sperm count < 10 × 109 cells/l. Median spermatozoa viability and motility were 74% (range, 12–96) and 40% (range, 5–81), respectively. The median percentage of round cells was 1% (range, 0.08–17.2) and the total round cells per ejaculate was 3.2 × 106 (range, 0.13–46.0). Polynuclear cells were observed in 81 (40.1%) of 202 samples. In these samples, the median total polynuclear cell count per ejaculate was 0.096 × 106 (range, 0.001–8.256).

Back to Top | Article Outline

Predictive factors of for HIV-1 nucleic acid in seminal plasma or sperm cells

Patient age, the interval between the estimated time of HIV-1 infection and the examination and the number of samples provided by the patient during the study did not influence the results of HIV-1 genome detection in sperm.

The median blood CD4 cell count tended to be lower when HIV-1 RNA was detectable in seminal plasma (470.5 × 106 cells/l; range, 170–1118) than when it was undetectable (567.5 × 106 cells/l; range, 158–1601) (P < 0.07; Table 2). Positive seminal plasma HIV-1 RNA detection was more frequent when blood viral load was > 1000 copies/ml (38.2%) than when it was < 1000 copies/ml (13.6%; P < 0.003).

Table 2

Table 2

Antiretroviral treatment significantly reduced HIV-1 RNA detection in blood plasma (P < 0.05) and in seminal plasma (P < 0.05). The rate of positive seminal plasma HIV-1 RNA decreased according to the treatment: 30% in the group without treatment, 28.6% in the group treated with two drugs and 10.2% in the group taking three or more drugs (P < 0.01).

Several semen parameters were linked to HIV-1 genome detection in semen. As shown in Table 2, sperm count and total sperm count in the ejaculate were globally higher in the ejaculate without detectable HIV-1 RNA and DNA than in the samples with HIV-1 genomes in the sperm cell fractions. Similarly, viability and motility of spermatozoa were better when the seminal viral load was negative than when it was positive (Table 2). A similar pattern existed in different cell fractions according to HIV-1 RNA or DNA detection, but the difference between the positive and the negative groups of RNA detection was significant only in the 50% fraction cells.

The percentage of polynuclear cells and the polynuclear cell count per unit volume or per ejaculate were higher in samples with than without detectable HIV-1 genomes (Table 2). The presence of polynuclear cells in semen led to a fourfold increase in the likelihood of at least one positive HIV-1 genome detection in sperm cells. The effect of the presence of polynuclear cells was evident even with < 0.5 × 106 cells per ejaculate, and greatly intensified, up to 10–15 times higher risk of RNA positive detection when the number was > 0.5 × 106 cells per ejaculate. By comparison, the likelihood of HIV-1 genome detection in sperm cells decreased by one third to one tenth as the total sperm count increased (Table 3).

Table 3

Table 3

When seminal plasma was positive the likelihood of HIV-1 genome (DNA and/or RNA) detection in sperm cells was increased sixfold (odds ratio, 6.45; 95% confidence interval, 3.05–13.66).

Back to Top | Article Outline

Intermittent detection

Of the 67 patients who provided more than one blood and semen sample, two patients consistently showed the same HIV-1 RNA positive detection in each seminal plasma sample (Table 4). Forty-eight patients consistently had HIV-1 RNA-negative seminal plasma samples over time (concordant group); whereas 17 patients had discordant results in their samples (discordant group). The median number of samples obtained was 4 (range, 2–9) in the concordant and 4 (range, 2–11) in the discordant group.

Table 4

Table 4

Although blood samples positive for HIV-1 RNA were 88.2% and 68.8% in discordant and concordant patients, respectively, this difference was not statistically significant. However, when the blood viral load was positive, the median blood viral load was higher in discordant than in concordant patients. In another respect, it is important to note that 6–10% of sperm cell samples were positive for HIV-1 genome in patients whose seminal plasma samples were always negative for HIV-1 RNA.

Back to Top | Article Outline


HAART has radically changed the prognosis of HIV infection. It is effective in the majority of patients and reduces the plasma viral load below the detection limit. While decreased seminal viral load has been reported with HAART, RNA and/or DNA has been detected in the sperm of patients receiving this treatment. In this context, the factors predicting the presence of viral genome in sperm need to be identified. Our study of a large number of samples, using two methods of viral genome detection in different sperm cell fractions, enabled such predictive factors to be sought. The number of serial samples and the long follow-up for a given patient showed that excretion of HIV-1 in sperm was intermittent in some patients. To the best of our knowledge, this is the only study in which viral genome detection was carried out in different sperm fractions (seminal plasma, whole sperm cells, 50% fraction cells obtained after density gradient separation and in the final spermatozoa fraction obtained after preparation) in a large number of samples.

The HIV-1 genome was detected in seminal plasma and/or in sperm cells. In seminal plasma, viral load showed a wide variation (5–277 500 copies/ml). Several authors found a correlation between blood and seminal levels of HIV-1 RNA [11,14,40]. In our study, no such correlation was found, probably because of the great variation in seminal RNA levels and the number of positive samples. However, when blood viral load was positive, the seminal viral load was more likely to be positive. Antiretroviral treatment, especially with three or more drugs, has a significant effect on RNA detection in seminal plasma according to other studies, which showed decreased HIV-1 in semen following antiretroviral treatment [8,10,14,21,22]. A correlation between the decrease in blood viral load and the decrease in seminal viral load has been reported [10]. However, it is noteworthy that HIV-1 RNA was detected in semen although blood levels were undetectable for 10 samples from our patients receiving HAART. When blood and semen viral loads were both positive, viral load was found to be higher in seminal plasma than in blood in 26% of paired samples. These findings can be linked to the hypothesis of compartmentalization of blood and sperm [16–18,20]. Because of variable drug penetration, HIV-1 in the genital tract may be underexposed to antiretroviral drugs [22], allowing HIV to replicate in this compartment and to be selected for resistance [41,42]. Therefore, in any individual, the seminal viral load cannot be deduced from the plasma viral load. This particularly important point should be borne in mind in campaigns for information and infection prevention in patients receiving HAART, who on seeing that their plasma viral load is undetectable might relax their vigilance in prevention of sexual transmission.

The viral genome can be found in sperm cells. In 15.8% of samples, HIV-1 DNA and/or RNA was found in native sperm cells or in the intermediate fraction during spermatozoa processing. Several factors may be related to the presence of HIV-1 genome in sperm cells. The likelihood of detecting the viral genome is multiplied by 6.5 if seminal plasma has a positive viral load. However, certain seminal plasma samples without detectable RNA may nonetheless contain positive cells.

The presence of polynuclear cells seems be the most important parameter linked with HIV-1 genome detection in semen cells in our asymptomatic patients. The fact that a polynuclear count equal to or above 0.5 × 106 per ejaculate increases more than tenfold the likelihood of detecting HIV-1 genomes in semen cells argues that inflammatory cells are a source of HIV in semen. This was particularly striking in our patient with the highest seminal viral load and a concurrent increase of polynuclear count [43]. These cells in semen could be related to inflammation/infection involving other cells such as macrophages or lymphocytes, which are known to carry HIV. It is noteworthy that seminal leukocytosis [8,25], asymptomatic prostatitis, urethritis or gonorrhoea [26,44] increase the level of HIV recovery from semen. Moreover, treatment of urethritis can decrease seminal HIV levels [28–30].

Our results and the published studies raise the question of whether it is necessary to give treatment to reduce the inflammatory cells in asymptomatic patients in order to lower the seminal viral load, in particular before sperm preparation for medically assisted procreation.

In our series, several patients supplied more than one paired semen and blood sample, allowing us to study the intermittence of HIV excretion in semen. One quarter of patients who provided serial samples showed intermittent RNA in seminal plasma. In a recent study [32] of untreated HIV-1 patients followed for 10 weeks, 44% had intermittent HIV-1 in semen with no change in blood viral load. During a 2-month follow-up of 84 HIV-1-infected men who provided three semen samples at 4-week intervals, Coombs et al. [16] found that 39.3% had intermittent HIV-1 excretion in semen.

Our study has the longest published follow-up of semen samples from HIV-1-infected patients (34 months). In the group of patients with intermittent HIV-1 excretion, blood viral load was greater than in the group with no intermittent excretion. Seminal HIV-1 RNA levels were seen to show greater variation over time than blood levels, as was particularly noteworthy in our previously published patient with the highest seminal plasma levels [43]. In his case, shedding of HIV-1 in semen was associated with markedly increased seminal polymorphonuclear granulocyte count.

Undetectable seminal plasma RNA levels do not mean absence of HIV-1 genomes in sperm cells; in our group of patients whose seminal plasma results were always negative, 6–10% of sperm cell samples contained HIV-1 genomes.

Desire for children in serodiscordant couples does not differ from that of couples free from HIV [33]. As HIV infection affects a large population of reproductive age, many serodiscordant couples in which the male partner is infected by the virus seek medical assistance in order to conceive with reduced risk of transmission to the partner. The first insemination programme for serodiscordant couples was reported in 1993 [38], an era when study of viral load in blood and sperm was not possible. Before starting an insemination programme in these couples, we wished to validate the method of spermatozoa preparation. In this study, we used two successive preparation techniques, density gradient migration followed by the swim-up method. Detection of HIV-1 genome (RNA or DNA) was always negative in the spermatozoa obtained after using the two preparation techniques, regardless of the seminal plasma viral load or the sperm cell result. While the possible presence of HIV-1 in spermatozoa has been discussed, several reports argue against these findings [36,45]. In one insemination programme [46], the fractions of motile spermatozoa prepared with two methods were positive in 6 of 107 cases, but the PCR technique was performed with primers allowing amplification of gag, which according to the authors could lead to false-positive results. In another study, one DNA-positive and six RNA-positive motile fractions were found, but here only one method of sperm preparation was used (density gradient method) allowing possible contamination of the spermatozoa fraction by non-spermatozoal cells [11]. In an experimental study, Hanabusa et al. [47] showed that the use of two methods was more efficient than a single method in preparing motile spermatozoa without detectable HIV genomes. The efficiency of our method is illustrated by the large number of samples we examined and by the case of the patient with the highest HIV seminal viral load found in our study, who always had negative genome detection in the motile spermatozoa fraction.

To-date, we have performed 213 intrauterine inseminations with prepared spermatozoa, obtaining 37 pregnancies resulting in 28 deliveries and seven miscarriages. Two pregnancies are on-going. The 28 deliveries resulted in 33 children. No women were infected (L. Bujan, C. Pasquier, E. Labeyrie, P. Lanusse, B. Mercadier, M. Morucci M, et al., unpublished data). Our results are in agreement with the three published studies using prepared and virologically tested spermatozoa in insemination programmes for the non-infected female partner in HIV-1 serodiscordant couples [46,48,49].

In conclusion, our study demonstrates in a large number of paired blood and semen samples the relationship between the two compartments and the predictive factors for HIV-1 genome detection in semen. It brings out several points that are relevant to health policies for the prevention of sexual transmission of HIV. Although HAART may reduce blood RNA to undetectable levels, this does not mean that there are no viral genomes in semen. Negative results for RNA in seminal plasma do not necessarily mean an absence of viral genomes in sperm cells; in addition negative results for DNA and RNA in semen one day do not allow prediction of the following day's results, as we demonstrated intermittent viral excretion. The presence of seminal polynuclear cells increases the risk of finding viral genomes in semen. The methods of sperm preparation used in this study were effective in obtaining spermatozoa without detectable HIV-1 genomes.

Back to Top | Article Outline


The authors thank Dr R. Mieusset and Dr P. Thonneau for advice and support; the team of the Centre de Stérilité Masculine and Centre d'Etudes et de Conservation des Oeufs et du Sperme Humain and particularly Mss Françoise Cendres, Corinne Fourtané, Andrée Lamartre and Claudine Sémézies for technical assistance; the team of the Service de Virologie for their help and cooperation and Nina Crowte for language editing.

Sponsorship: This study was supported by grant ANRS 096 from the Agence Nationale de Recherche sur le SIDA (ANRS), Paris, France.

Note: The preliminary results of this study, concerning the efficiency of sperm processing in obtaining spermatozoa fractions without detectable HIV-1 genomes, were the subject of an oral presentation at the 56 Annual Meeting of the American Society of Reproductive Medicine, San Diego, October 2000.

Back to Top | Article Outline


1. Dejucq N, Jegou B. Viruses in the mammalian male genital tract and their effects on the reproductive system. Microbiol Mol Biol Rev 2001, 65:208–231.
2. Mayer KH, Anderson DJ. Heterosexual HIV transmission. Infect Agents Dis 1995, 4:273–284.
3. Royce RA, Sena A, Cates W, Jr, Cohen MS. Sexual transmission of HIV. N Engl J Med 1997, 336:1072–1078.
4. Lee TH, Sakahara N, Fiebig E, Busch MP, O'Brien TR, Herman SA. Correlation of HIV-1 RNA levels in plasma and heterosexual transmission of HIV-1 from infected transfusion recipients. J Acquir Immune Defic Syndr Hum Retrovirol 1996, 12: 427–428.
5. Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li C, Wabwire-Mangen F, et al. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group. N Engl J Med 2000, 342:921–929.
6. Zagury D, Bernard J, Leibowitch J, Safai B, Groopman JE, Feldman M, et al. HTLV-III in cells cultured from semen of two patients with AIDS. Science 1984, 226:449–451.
7. Ho DD, Schooley RT, Rota TR, Kaplan JC, Flynn T, Salahuddin SZ, et al. HTLV-III in the semen and blood of a healthy homosexual man. Science 1984, 226:451–453.
8. Anderson DJ, O'Brien TR, Politch JA, Martinez A, Seage GR, III, Padian N, et al. Effects of disease stage and zidovudine therapy on the detection of human immunodeficiency virus type 1 in semen. JAMA 1992, 267:2769–2774.
9. Vernazza PL, Eron JJ, Cohen MS, van der Horst CM, Troiani L, Fiscus SA. Detection and biologic characterization of infectious HIV-1 in semen of seropositive men. AIDS 1994, 8:1325–1329.
10. Vernazza PL, Gilliam BL, Flepp M, Dyer JR, Frank AC, Fiscus SA, et al. Effect of antiviral treatment on the shedding of HIV-1 in semen. AIDS 1997, 11:1249–1254.
11. Tachet A, Dulioust E, Salmon D, de Almeida M, Rivalland S, Finkielsztejn L, et al. Detection and quantification of HIV-1 in semen: identification of a subpopulation of men at high potential risk of viral sexual transmission. AIDS 1999, 13:823–831.
12. Pasquier C, Daudin M, Righi L,, Berges L, Thauvin L, Berrebi A, et al. Sperm washing and virus nucleic acid detection to reduce HIV and hepatitis C virus transmission in serodiscordant couples wishing to have children. AIDS 2000, 14:2093–2099.
13. Ball JK, Curran R, Irving WL, Dearden AA. HIV-1 in semen: determination of proviral and viral titres compared to blood, and quantification of semen leukocyte populations. J Med Virol 1999, 59:356–363.
14. Gupta P, Mellors J, Kingsley L, Riddler S, Singh MK, Schreiber S, et al. High viral load in semen of human immunodeficiency virus type 1-infected men at all stages of disease and its reduction by therapy with protease and nonnucleoside reverse transcriptase inhibitors. J Virol 1997, 71:6271–6275.
15. Liuzzi G, Chirianni A, Clementi M, Bagnarelli P, Valenza A, Cataldo PT, et al. Analysis of HIV-1 load in blood, semen and saliva: evidence for different viral compartments in a cross-sectional and longitudinal study. AIDS 1996, 10:F51–F56.
16. Coombs RW, Speck CE, Hughes JP, Lee W, Sampoleo R, Ross SO, et al. Association between culturable human immunodeficiency virus type 1 (HIV-1) in semen and HIV-1 RNA levels in semen and blood: evidence for compartmentalization of HIV-1 between semen and blood. J Infect Dis 1998, 177:320–330.
17. Byrn RA, Kiessling AA. Analysis of human immunodeficiency virus in semen: indications of a genetically distinct virus reservoir. J Reprod Immunol 1998, 41:161–176.
18. Kiessling AA, Fitzgerald LM, Zhang D, Chhay H, Brettler D, Eyre RC, et al. Human immunodeficiency virus in semen arises from a genetically distinct virus reservoir. AIDS Res Hum Retroviruses 1998, 14 Suppl 1:S33–S41.
19. Ping LH, Cohen MS, Hoffman I, Vernazza P, Seillier-Moiseiwitsch F, Chakraborty H, et al. Effects of genital tract inflammation on human immunodeficiency virus type 1 V3 populations in blood and semen. J Virol 2000, 74:8946–8952.
20. Nunnari G, Otero M, Dornadula G, Vanella M, Zhang H, Frank I, et al. Residual HIV-1 disease in seminal cells of HIV-1-infected men on suppressive HAART: latency without on-going cellular infections. AIDS 2002, 16:39–45.
21. Gilliam BL, Dyer JR, Fiscus SA, Marcus C, Zhou S, Wathen L, et al. Effects of reverse transcriptase inhibitor therapy on the HIV-1 viral burden in semen. J Acquir Immune Defic Syndr Hum Retrovirol 1997, 15:54–60.
22. Taylor S, Ferguson NM, Cane PA, Anderson RM, Pillay D. Dynamics of seminal plasma HIV-1 decline after antiretroviral treatment. AIDS 2001, 15:424–426.
23. Leruez-Ville M, Dulioust E, Costabliola D, Salmon D, Tachet A, Finkielsztejn L, et al. Decrease in HIV-1 seminal shedding in men receiving highly active antiretroviral therapy: an 18 month longitudinal study (ANRS EP012). AIDS 2002, 16:486–488.
24. Zhang H, Dornadula G, Beumont M, Livornese L, Jr, van Uitert B, Henning K, et al. Human immunodeficiency virus type 1 in the semen of men receiving highly active antiretroviral therapy. N Engl J Med 1998, 339:1803–1809.
25. Xu C, Politch JA, Tucker L, Mayer KH, Seage GR, III, Anderson DJ. Factors associated with increased levels of human immunodeficiency virus type 1 DNA in semen. J Infect Dis 1997, 176:941–947.
26. Moss GB, Overbaugh J, Welch M,, Reilly M, Bwayo J, Plummer FA, et al. Human immunodeficiency virus DNA in urethral secretions in men: association with gonococcal urethritis and CD4 cell depletion. J Infect Dis 1995, 172:1469–1474.
27. Winter AJ, Taylor S, Workman J, White D, Ross JD, Swan AV, et al. Asymptomatic urethritis and detection of HIV-1 RNA in seminal plasma. Sex Transm Infect 1999, 75:261–263.
28. Cohen MS, Hoffman IF, Royce RA, Kazembe P, Dyer JR, Daly CC, et al. Reduction of concentration of HIV-1 in semen after treatment of urethritis: implications for prevention of sexual transmission of HIV-1. AIDSCAP Malawi Research Group. Lancet 1997, 349:1868–1873.
29. Fiscus SA, Vernazza PL, Gilliam B, Dyer J, Eron JJ, Cohen MS. Factors associated with changes in HIV shedding in semen. AIDS Res Hum Retroviruses 1998, 14(Suppl 1):S27–S31.
30. Sadiq ST, Taylor S, Kaye S, Bennett J, Johnstone R, Byrne P, et al. The effects of antiretroviral therapy on HIV-1 RNA loads in seminal plasma in HIV-positive patients with and without urethritis. AIDS 2002, 16:219–225.
31. Krieger JN, Coombs RW, Collier AC, Ho DD, Ross SO, Zeh JE, et al. Intermittent shedding of human immunodeficiency virus in semen: implications for sexual transmission. J Urol 1995, 154:1035–1040.
32. Gupta P, Leroux C, Patterson BK, Kingsley L, Rinaldo C, Ding M, et al. Human immunodeficiency virus type 1 shedding pattern in semen correlates with the compartmentalization of viral quasi species between blood and semen. J Infect Dis 2000, 182: 79–87.
33. Chen JL, Philips KA, Kanouse DE, Collins RL, Miu A. Fertility desires and intentions of HIV-positive men and women. Fam Plann Perspect 2001, 33:144–152, 165.
34. Baccetti B, Benedetto A, Collodel G, di Caro A, Garbuglia AR, Piomboni P. The debate on the presence of HIV-1 in human gametes. J Reprod Immunol 1998, 41:41–67.
35. Pudney J, Nguyen H, Xu C, Anderson DJ. Microscopic evidence against HIV-1 infection of germ cells or attachment to sperm. J Reprod Immunol 1998, 41:105–125.
36. Quayle AJ, Xu C, Tucker L, Anderson DJ. The case against an association between HIV-1 and sperm: molecular evidence. J Reprod Immunol 1998, 41:127–136.
37. Kim LU, Johnson MR, Barton S, Nelson MR, Sontag G, Smith JR, et al. Evaluation of sperm washing as a potential method of reducing HIV transmission in HIV-discordant couples wishing to have children. AIDS 1999, 13:645–651.
38. Semprini AE. Insemination of HIV-negative women with processed semen of HIV-positive partners. Lancet 1993, 341: 1343–1344.
39. WHO. WHO Laboratory Manual for the Examination of Human Semen and Sperm–Cervical Mucus Interaction, 4th edn. Cambridge, UK: Cambridge University Press; 1999.
40. Vernazza PL, Dyer JR, Fiscus SA, Eron JJ, Cohen MS. HIV-1 viral load in blood, semen and saliva. AIDS 1997, 11:1058–1059.
41. Eyre RC, Zheng G, Kiessling AA. Multiple drug resistance mutations in human immunodeficiency virus in semen but not blood of a man on antiretroviral therapy. Urology 2000, 55:591.
42. Lafeuillade A, Solas C, Chadapaud S, Hittinger G, Poggi C, Lacarelle B. HIV-1 RNA levels, resistance, and drug diffusion in semen versus blood in patients receiving a lopinavir-containing regimen. J Acquir Immune Defic Syndr 2003, 32:462–464.
43. Bujan L, Daudin M, Alvarez M, Massip P, Puel J, Pasquier C. Intermittent human immunodeficiency type 1 virus (HIV-1) shedding in semen and efficiency of sperm processing despite high seminal HIV-1 RNA levels. Fertil Steril 2002, 78: 1321–1323.
44. Atkins MC, Carlin EM, Emery VC, Griffiths PD, Boag F. Fluctuations of HIV load in semen of HIV positive patients with newly acquired sexually transmitted diseases. Br Med J 1996, 313:341–342.
45. Gil T, Castilla J, Hortas M, Molina J, Redondo M, Samaniego F, et al. CD4+ cells in human ejaculates. Hum Reprod 1995, 10:2923–2927.
46. Marina S, Marina F, Alcolea R, Exposito R, Huguet J, Nadal J, et al. Human immunodeficiency virus type 1–serodiscordant couples can bear healthy children after undergoing intrauterine insemination. Fertil Steril 1998, 70:35–39.
47. Hanabusa H, Kuji N, Kato S, Tagami H, Kaneko S, Tanaka H, et al. An evaluation of semen processing methods for eliminating HIV-1. AIDS 2000, 14:1611–1616.
48. Semprini AE, Levi-Setti P, Bozzo M, Ravizza M, Taglioretti A, Sulpizio P, et al. Insemination of HIV-negative women with processed semen of HIV-positive partners. Lancet 1992, 340:1317–1319.
49. Gilling-Smith C. HIV prevention. Assisted reproduction in HIV-discordant couples. AIDS Read 2000, 10:581–587.

HIV RNA; HIV DNA; spermatozoa; semen; sperm cells; HIV-1 intermittent excretion

© 2004 Lippincott Williams & Wilkins, Inc.