Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Quantification of HIV in semen: correlation with antiviral treatment and immune status Pietro L. Vernazza*§, Bruce L. Gilliam†, John Dyer†, Susan A. Fiscus‡, Joseph J. Eron‡, Andreas C. Frank§ and Myron S. Cohen† Objective: This study examined the concentration of HIV in semen and the effects of biological factors on HIV excretion. Methods: Semen samples from 101 men at different stages of the disease were evaluated by quantitative HIV culture and HIV RNA detection. Blood plasma samples were available from 56 patients. The effects of CD4 and CD8 count, blood plasma RNA levels, treatment status and clinical staging on the shedding of HIV were evaluated. Results: HIV RNA levels in semen correlated with quantitative HIV culture of seminal cells and a strong association of positive seminal cell culture with high RNA levels was observed. CD4 count and antiviral treatment were inversely correlated with the concentration of HIV in semen. Blood plasma HIV RNA values were correlated with HIV RNA levels in semen, although some patients had highly discrepant results. Conclusions: The strong correlation between seminal cell culture and concentration of HIV RNA in seminal plasma suggested that HIV detected in seminal plasma was released by productively infected cells in the male genital tract. The study showed that the concentration of HIV in semen, which was likely to be correlated with HIV infectivity, was a function of the immune status of the HIV-infected individual. The results suggested that antiviral therapy may have reduced the concentration of HIV in semen. AIDS 1997, 11:987–993 Keywords: Semen, viral load, HIV RNA, infectivity, culture Introduction HIV-1 is primarily a sexually transmitted disease (STD) that affects more than 21 million people worldwide [1]. Our current knowledge regarding the efficiency of HIV-1 sexual transmission is limited. The spread of HIV in a population is described by the basic reproductive rate Ro (i.e., the average number of secondary cases infected by one index case) [2]. Ro is a function of the efficiency (β) and duration (D) of the transmissibility and the number of sexual partners (c), expressed by the formula Ro = β • c • D. Partner and observational studies have indicated large variability in the sexual transmissibility of HIV [3–7]. Transmissibility of HIV depends on the type of sexual exposure, the presence of other cofactors in the recipient (e.g., vaginal drying agents, mucosal disruption, resistance to infection), and the infectivity of the donor [8]. Several biological From the *Institute for Clinical Microbiology and Immunology, Kantonsspital, St Gallen, Switzerland, the Departments of † Medicine, and ‡Microbiology and Immunology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA and the §Department of Medicine, Kantonsspital, St Gallen, Switzerland. Sponsorship: This study was supported by grants from the Swiss National Science Foundation and the Swiss Federal Office of Public Health (32-38818.93 and 316.93.7159), The Australian Society for Infectious Diseases, the Cooperative Agreement SU01AI25868 from NIAID, General Clinical Research Centers program NIH RR00046 and grant numbers UOAI31496 and NIDDK RO149381. Requests for reprints to: Pietro L. Vernazza, Kantonsspital, 9007 St Gallen, Switzerland. Date of receipt: 2 August 1996; accepted: 16 March 1997. © Rapid Science Publishers ISSN 0269-9370 987 988 AIDS 1997, Vol 11 No 8 factors have the potential to modulate HIV infectivity. The status of the donor immune system, local mucosal factors, and the phenotypic characteristics of the viral quasispecies are all likely to influence the concentration of virus in genital secretions. The minimum quantity of HIV required for sexual transmission of the virus to a susceptible partner is not known. A better understanding of the biological factors influencing the shedding of HIV is critical for the design of novel pharmacological and immunotherapeutic approaches that would help to reduce the infectivity of individuals and limit the further spread of the epidemic. A critical inoculum is possibly required for transmission of HIV by any route of exposure. For example, higher concentration of HIV in blood is associated with increased risk of HIV transmission by transfusion, vertical transmission and after needle-stick exposure [9,10]. In order to understand how biological factors influence transmission, accurate quantification of HIV in semen may be important. HIV has been recovered ex vivo by culture of the cellular component of semen in 8–54% of infected individuals [11–14]. However, culture techniques used in these studies were not quantitative and HIV culture may not have been sensitive enough to detect the viral inoculum required to support efficient transmission. More recently, molecular amplification techniques have been used to detect HIV DNA and RNA in semen, but only in small studies [15–20]. Quantitative amplification techniques should allow better definition of the range of concentration of HIV in semen and how biological factors influence the concentration. In the present study, quantitative HIV-1 culture and quantitative HIV RNA amplification were used to examine HIV-1 shedding in semen of 101 HIVseropositive men and to investigate the effect of stage of disease, blood viral burden, CD4+ cell depletion, and antiretroviral therapy on the concentration of virus excreted. HIV culture of seminal cells and RNA amplification in cell-free seminal fluid were also compared to determine the correlation between the two compartments. (Switzerland). The study was approved by the local ethical committees and all patients gave informed consent for participation in the study. Clinical staging was performed according to the revised Centers for Disease Control and Prevention classification system [21]. Participants had to be free of signs or symptoms of STD. In addition, all Swiss patients were screened and found to be negative for asymptomatic chlamydia infection by polymerase chain reaction. Patients were classified according to their antiretroviral treatment status: treatment-naive, previous antiretroviral treatment and active treatment. However, treatment status had to be stable for at least 12 weeks prior to sampling. Patients in the active treatment group received either one or two nucleoside analogues but no protease inhibitors. CD4 cell counts of blood lymphocytes were determined by flow cytometry using standardized methods on samples obtained within 3 months of the date of semen collection. Results were expressed as absolute values and as the percentage of total lymphocytes. Blood plasma samples were taken by venipuncture in a cell-preparation Vacutainer tube (CPT; Becton Dickinson, Franklin Lakes, New Jersey, USA) using citrate as an anticoagulant. Semen samples were obtained by masturbation and collected in a sterile container and a virus transport medium (RPMI plus 1000 IU/ml penicillin plus 1 mg/ml streptomycin) was added to the sample. Semen samples were processed in the laboratory within 6 h of ejaculation. Processing of semen samples After liquefaction of the ejaculates, seminal cells were separated from the seminal fluid by centrifugation in a 15 ml conical tube for 10 min at 800–1000 g. The seminal plasma was removed, aliquotted and stored frozen at −75°C. One aliquot of each Swiss sample was shipped frozen in dry ice to the US site for RNA detection. One-half of the seminal cells was used for quantitative HIV culture. Methods Semen microculture assay HIV culture of seminal cells was performed immediately after cell separation following a quantitative adaptation of a method previously described in 24-well plates [14]. Quantification was performed by six duplicate three- or fivefold dilutions of seminal cells. The equivalent of one-sixth of the cells from one ejaculate was used for the first dilution. Quantitative results were obtained by reference to the number and pattern of serially diluted wells that were positive for p24 antigen, using a published algorithm [22] and results were expressed as infectious units per ejaculate (IUPE). Study population, clinical staging, semen and blood samples The study population consisted of HIV-1-positive men attending one of two collaborating HIV clinics at the University of North Carolina at Chapel Hill (North Carolina, USA) and at the Kantonsspital St Gallen Quantitative detection of HIV RNA in seminal plasma and blood plasma A commercially available technique for the quantification of HIV RNA in cell-free seminal plasma was used [nucleic acid sequence-based amplification (NASBA)] following previously described methods [18,23]. A total HIV in semen Vernazza et al. of 100 µl of the sample were used per assay. The lower limit of detection of the assay with this input volume is 1000 RNA copies/ml. The dilution factor of the semen samples with the known amount of transport medium was taken into account for the final calculation of the HIV RNA concentration in seminal plasma. Statistical analysis The χ2 test was used for comparisons of distribution ratios, and Mann–Whitney U-test for comparisons between groups. Correlation of non-parametric variables was examined using Spearman’s rank correlation tests. Multiple logistic regression was performed to define the relative contribution of different factors on the detection of HIV in seminal cells (by culture) and in seminal plasma (HIV RNA). The factors included in the multiple regression model were blood plasma RNA concentration, CD4 and CD8 counts, treatment status (currently treated: ‘yes/no’), and clinical stage (symptomatic versus asymptomatic). In addition, multiple linear regression was performed to define the correlation of these parameters with the quantitative HIV RNA results in seminal plasma. Results Demographics Semen samples from 101 consecutive patients were collected at the two study sites. Paired semen and blood samples were available from 56 patients. Clinical staging, CD4 count, status of antiviral treatment, and demographic information are summarized in Table 1. The median age of the study population was 34.0 years. Fifty-eight per cent of the men were asymptomatic and 44% had a CD4 count below 200 × 106/l (or 14% of lymphocytes). Fifty-three patients (52%) had never been treated with an antiretroviral agent at the time of semen donation. Twenty-nine patients were currently taking antiretroviral agents and 19 had stopped treatment more than 3 weeks prior to semen donation. The study populations from the two study sites were comparable in terms of disease stage and CD4 count. The Swiss population had a significantly higher proportion of men never treated with an antiretroviral drug (χ2 test, P = 0.0005) and a higher proportion of men infected by needle sharing. Quantitative aspects of HIV in semen: culture and RNA detection Of the 101 men studied, semiquantitative seminal cell culture was successfully performed in samples taken from 98 (three cultures had fungal contamination). Infectious HIV was recovered from 29 samples (30%). The infectious titre in positive semen cultures ranged from 2 to 3800 IUPE. Cell-free HIV RNA was detectable in 63 (62%) out of 101 seminal plasma samples. One patient Table 1. Patient characteristics at the two study sites. Study centre US Risk groups Homosexual 51 Heterosexual 5 Injecting drug use 1 All patients 57 Median age (years) 33.5 Absolute CD4 count (×106/l) >500 9 200–500 23 <200 25 Median (×106/l) 270 Relative CD4 count (% lymphocytes) >27 7 15–27 25 0–14 25 Mean (%) 16 Peripheral blood RNA level (copies/ml) (n = 56) <5000 9 5000–30000 3 30000–100000 4 >100000 9 Clinical staging (CDC) Stage A 27 Stage B 18 Stage C 12 Previous antiretroviral treatment at study entry* Never 21 (11) Current 17 (9) Discontinued 19 (5) Mean processing time of samples (h) 2.5 Swiss Total 18 14 12 44 35.0 69 19 13 101 34.0 6 18 20 300 15 41 45 280 4 21 19 17 11 46 44 17 6 10 7 8 15 13 11 16 32 4 8 59 22 20 32 (25) 12 (6) 0 (0) 53 (36) 29 (15) 19 (5) 3.6 3.0 *Number of patients for whom a blood sample was available is shown in parenthesis. Data are expressed as numbers of patients, unless otherwise indicated. with a past history of candida oesophagitis and a CD4 count of 10 × 106/l had the highest level of HIV RNA in semen [3.15 × 108 (or 8.5 log10) copies/ml] and also the highest titre of infectious virus by culture (3800 IUPE). No clinical findings suggestive of urethritis were present on examination. The RNA values of the remaining samples ranged from non-detectable to 9 × 106 (6.95 log10) copies/ml [median value of all samples, 5100 (3.71 log10) copies/ml]. HIV RNA levels above 10 000 copies/ml were detectable in 44% of the patients. Recovery rate of HIV RNA from the Swiss patients was lower than that from the US patients (70 versus 52%; P = 0.066). This lower recovery rate was restricted to samples from individuals with higher CD4 counts (above 14% of lymphocytes or 200 × 106/l). Spearman’s rank correlation testing between quantitative measurements of HIV in seminal cells and cell-free seminal plasma revealed a correlation coefficient of 0.60 (P < 0.001). In fact, in the 29 patients with a positive HIV culture from seminal cells, only one had no detectable HIV RNA in the seminal plasma and only three had an HIV RNA level in semen below 10 000 copies/ml (P < 0.000001, χ 2 test; Table 2). The 989 990 AIDS 1997, Vol 11 No 8 Table 2. Detection of HIV in semen by nucleic acid sequencebased amplification and cell culture. (a) HIV RNA detection in seminal plasma No. tested HIV culture in seminal cells Positive 29 Negative 69 No. (%) positive No. (%) >10000 copies/ml 28 (97)* 32 (46) 26 (90)* 18 (26) *P < 0.001. median HIV RNA level in seminal plasma for samples with positive seminal cell culture was 1.2 × 105 (5.1 log10) copies/ml as opposed to 2.5 × 103 (3.4 log10) for samples with negative culture (Fig. 1). (b) HIV in semen as a function of blood plasma HIV RNA level Seven out of 56 patients (13%) had no detectable HIV RNA in the blood, although one of these patients had detectable HIV RNA in semen [6500 (3.81 log 10 ) copies/ml]. HIV RNA was detectable in blood as well as semen from 29 patients (P = 0.03, χ2 test). HIV blood plasma level was highly predictive of detection of HIV RNA in semen by multiple logistic regression [odds ratio (OR), 3.07; 95% confidence interval (CI), 1.44–6.53; P = 0.003]. Spearman’s rank correlation of HIV RNA concentrations in semen and blood revealed a correlation coefficient of 0.56 (P < 0.001; Fig. 1). By multiple linear regression analysis, blood plasma RNA levels correlated well with the concentration of HIV RNA in semen (P = 0.0008). Despite the correlation between HIV RNA results in semen and blood, in some patients the difference in HIV RNA load far exceeded the variability of the assay. When the 28 patients with positive RNA values in both compartments were compared, only 12 patients had semen and blood RNA levels within 1 log 10. The HIV RNA level in semen was more than 10-fold (1.0 log10) lower than blood in nine subjects and 10-fold higher than blood in seven subjects. In addition, eight patients had no detectable HIV RNA in semen despite HIV RNA concentration in blood plasma above 30 000 copies/ml (4.5 log10). All patients with positive semen cultures had positive HIV RNA values in blood plasma. The median HIV RNA value in blood was 4.91 log10 copies/ml in the 12 patients with positive culture compared with 4.25 log10 copies/ml in the 42 patients with negative culture result. No correlation was observed between the level of blood plasma RNA and culture positivity (Spearman’s rank’s correlation and multiple logistic regression). HIV in semen as a function of blood CD4 and CD8 count Patients with low blood CD4 counts were more likely Fig. 1. Correlation of seminal plasma HIV RNA concentration with (a) quantitative seminal cell culture and (b) blood plasma HIV RNA concentration for patients with detectable HIV in semen by (a) culture or (b) RNA detection. The box plots indicate median and interquartile range of HIV RNA concentration in samples with negative or positive for (a) HIV culture or (b) HIV RNA in semen. (v), HIV RNA detectable in (a) semen or (b) blood. (O), HIV RNA not detectable. to have detectable HIV RNA in the seminal plasma (Table 3). On multiple logistic regression, CD4 count was independently and inversely predictive of RNA detection in semen (OR, 0.78; 95% CI, 0.63–0.93; P = 0.01). CD4 counts were significantly correlated with the amount of HIV RNA in semen (rs = 0.23, P = 0.019, Spearman’s rank correlation). The correlation remained significant in a multiple linear regression analysis that included CD4 and CD8 count, blood plasma RNA level, treatment status and clinical stage of disease (P = 0.04). An increased recovery rate of HIV by culture of seminal cells was observed at lower CD4 counts (Table 3). HIV was detected by culture in 33% of patients with CD4 counts below 200 × 106/l compared with 14% in patients with CD4 counts above 500 × 106/l. By multiple logistic regression, absolute CD4 count was inde- HIV in semen Vernazza et al. Table 3. Results of HIV seminal cell culture and HIV RNA detection in semen and clinical stage of disease. HIV RNA† Cell culture* CD4 count (×106/l) 0–199 200–500 >500 CD4 count (% of lymphocytes) 0–14 15–27 >27 CDC stage A B C Median HIV RNA† log10 copies/ml Total No. (%) positive P‡ Total No. (%) positive P‡ 43 40 14 14 (33) 14 (37) 1 (7) NS 45 41 15 32 (71) 25 (61) 6 (40) 0.05 3.71 4.03 3.24 41 46 11 16 (39) 13 (28) 0 (0) < 0.002 44 46 11 34 (77) 24 (52) 5 (45) < 0.01 4.15 3.47 3.47 59 21 18 16 (27) 7 (33) 6 (33) NS 59 22 20 33 (56) 15 (68) 15 (75) NS 3.61 3.62 3.90 *HIV culture of seminal cells. †HIV RNA detection in seminal plasma. ‡χ2 test for trends. NS, Not significant. Table 4. Results of HIV seminal cell culture and HIV RNA detection in semen and antiviral treatment. Treatment status Never treated Previously treated Currently treated Median HIV RNA (copies/ml) Seminal plasma RNA† Mean blood CD4 count (×106/l) Blood Semen Total No. (%) positive Total No. (%) positive 380 190 180 4.45 4.92 4.30 3.68 4.47 3.48 53 19 29 32 (60) 16 (84) 15 (51) 52 19 27 15 (29) 7 (37) 7 (26) Seminal cell culture* *HIV culture of seminal cells: differences not significant (χ2 test). †HIV RNA detection in seminal plasma: P = 0.02, previously treated versus currently treated; P = 0.03, treated versus not treated (χ2 test). pendently and inversely predictive of a positive HIV culture result (OR, 0.45; 95% CI, 0.23–0.87; P = 0.02). Similarly, the multiple logistic model identified CD8 counts as a predictor of positive HIV culture, although the association did not quite reach statistical significance (OR, 1.18; 95% CI, 0.99–1.39; P = 0.06). Spearman’s rank correlation revealed a significant inverse correlation between IUPE and CD4 percentage (r = 0.25, P = 0.015). HIV in semen as a function of clinical stage Asymptomatic patients were less likely to have detectable HIV RNA in semen than patients with advanced stages of disease, although the difference was not statistically significant (Table 3). Stage of disease was not an independent predictor of HIV RNA detection by multiple logistic regression. The median value of HIV RNA in semen was slightly but insignificantly higher in patients with an AIDS-defining diagnosis (3.9 versus 3.6; P = 0.31, Mann–Whitney U-test). Results of HIV culture of seminal cells were independent of the clinical stage of disease, although isolation of HIV from semen was somewhat less likely in asymptomatic patients. HIV in semen as a function of antiviral treatment Drug-naive patients had higher levels of CD4 counts and lower levels of HIV RNA in blood plasma than patients currently or previously treated (Table 4) reflecting the CD4 dependence of the indication for antiviral treatment. When patients with and without current antiviral treatment were compared, detection of HIV RNA in semen was significantly lower in the treated group of patients (Table 4; P = 0.03). Current antiviral treatment was an independent inverse predictor of HIV RNA detection by multiple logistic regression, but this association did not reach the 5% level of significance (OR, 0.38; 95% CI, 0.14–1.03; P = 0.054). Patients currently receiving antiviral treatment had lower HIV RNA levels in semen than patients without antiviral treatment, and antiviral treatment was inversely but not significantly correlated with the level of HIV RNA in semen by multiple linear regression (P = 0.07). Patients on current antiretroviral treatment had a trend towards lower detection rates of HIV by seminal cell culture, although the difference did not reach statistical significance. Similarly, multiple logistic regression analysis did not show a statistically significant association between treatment status and culture results (OR, 0.38; 95% CI, 0.12–1.22; P = 0.1). Discussion HIV is primarily an STD which is spread with greater efficiency by men than women [8]. Semen is a likely and important, if not critical, source of virus [24]. Transmission of infectious disease requires a critical inoculum, which is integral to the efficiency of trans- 991 992 AIDS 1997, Vol 11 No 8 mission. Therefore, the amount of HIV in semen will probably be one determinant of the infectivity of an individual. The current study was undertaken to determine the concentration of HIV in semen from a large group of men and to determine what factors influence this concentration. Earlier studies were limited by the small number of subjects studied or by assays available for detection of HIV [11–15]. Semen samples were collected from 101 men. The results demonstrated a strong correlation of seminal cell culture results with the level of HIV RNA in seminal plasma. These results expanded on our earlier studies [18] and supported the conclusion that HIV detected in seminal plasma is released by productively infected cells in the male genital tract. This conclusion was consistent with the observation that at steady-state, HIV RNA levels in blood plasma appeared to be directly proportional to the number of productively infected cells contributing to that pool [25]. Our observation that the chance of detecting a positive culture was 59% when copies of HIV RNA in seminal plasma exceeded 10 000 copies/ml compared with 6% when fewer copies were detected was also consistent with the hypothesis that HIV RNA in seminal plasma was representative of the number of infected cells in the genital tract releasing virus into semen. Ex vivo culture was likely to be positive only when there was a specific, relatively large number of these infected cells shed into the semen. A correlation was also observed between the concentration of HIV in blood and seminal plasma. Although this correlation may have been a result of direct communication between these compartments, differences in phenotype and genotypes that have been observed between HIV from blood and semen in individuals argued against this conclusion [14, 26]. An alternative hypothesis was that HIV-infected men that had poor control of systemic HIV-1 replication manifested by high HIV-1 blood plasma RNA also had poor control of genital tract HIV replication reflected by higher levels of seminal plasma HIV RNA. Exceptions to this correlation were likely to occur as manifest by our subject who had no detectable RNA in his blood plasma but detectable RNA in semen. Individuals such as this one, even if they were relatively uncommon, may have had an important role in the continuation of the HIV epidemic if they were destined for long disease-free survival as would be predicted by their low blood plasma HIV RNA [27]. The concentration of HIV in the semen was also related to the blood CD4 count and the individual’s stage of disease. These results must be considered in the light of a large number of epidemiological studies designed to better understand and predict HIV transmission. In particular, existing data have suggested that patients with late-stage disease or low CD4 count represented the greatest risk for transmission [6,28–30]. These results were consistent with the increased likelihood of finding a positive semen culture in patients with lower CD4 counts and the increased concentration of HIV RNA in seminal plasma noted in our study. Krieger and coworkers [12] failed to detect a correlation between blood CD4 count and HIV in semen but reported that positive culture was correlated with peripheral CD8 counts. The effect of CD8 count was also noted in our study but did not reach statistical significance. This difference might have been explained by different culture techniques and size of groups examined. In addition, the fact that CD4 and CD8 counts were related variables (reflecting a proportion of the total number of lymphocytes) reduced the power of the multiple logistic regression analysis to identify both factors as independent predictors. In addition, in a more recent study, Krieger and coworkers [31] had used amplification techniques and noted a correlation of HIV RNA levels in semen with the stage of disease, CD4 count, and positive HIV culture. Given our observation of a correlation between blood plasma RNA and seminal plasma RNA, future epidemiological studies of transmission that collect blood RNA data may find that this measure is a stronger predictor of transmission than disease stage or absolute CD4 count. Mathematical models of HIV transmission have predicted that patients with primary disease (who have very high blood RNA levels) may be most important for the continuation of the HIV epidemic [32]. Although consistent with our observations, our study did not include patients with primary HIV infection. However, specimens have been recently obtained from three patients with primary disease whose semen and blood viral burden were studied before therapy. None of these patients had concentrations of HIV in their semen outside the range of the entire group in this study despite having high levels of blood RNA [33]. Other factors, such as antiretroviral therapy, may also be important in determining the concentration of HIV in semen. Musicco et al. [34] reported a 50% reduction of the risk of transmission in men treated with zidovudine. Although our study was not designed to address this question, it was observed that patients who were currently on therapy had significantly lower HIV RNA in seminal plasma (approximately 10-fold lower) than patients who had discontinued therapy at least 3 weeks prior to the study. In addition, active antiretroviral treatment was an independent predictor of a negative culture. In an unpublished prospective trial of a nucleoside combined with a non-nucleoside agent or placebo, a median 1.4 log10 decline (25-fold) was observed in seminal plasma RNA after 8 weeks on therapy in 11 patients [35]. HIV in semen Vernazza et al. The effects of vaccines, antiretroviral therapy and immune modulating therapy on the concentration of HIV in semen are of considerable importance. Data collected in this study, including the demonstration that HIV-1 RNA could be quantified in seminal plasma in a majority of individuals, has set the stage for development of interventions specifically designed to reduce the concentration of HIV in semen below the level of efficient transmission. 14. 15. 16. 17. 18. Acknowledgement Special thanks to A. Tschopp for performing the statistical analysis and D. Feldman for editing the manuscript. 19. 20. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. World Health Organization: The HIV/AIDS Situation in Mid1996: Global and Regional Highlights. Fact Sheet 1. July 1996. Geneva: UNAIDS, WHO; 1996. Anderson RM: The transmission dynamics of sexually transmitted diseases: the behavioural component. In Infectious Diseases of Humans: Dynamics and Control. Edited by Anderson RM, May RM. Oxford: Oxford University Press; 1991. Clumeck N, Teilman H, Hermans P: A cluster of HIV Infection among heterosexual people without apparent risk factors. N Engl J Med 1989, 321:1460–1462. Wiley JA, Herschkorn SJ, Padian NS: Heterogeneity in the probability of HIV transmission per sexual contact: the case of maleto-female transmission in penile–vaginal intercourse. Stat Med 1989, 8:93–102. Kaplan E: Modeling HIV infectivity: must sex acts be counted? J Acquir Immune Defic Syndr 1990, 3:55–61. De Vincenzi I: Longitudinal study of human immunodeficiency virus transmission by heterosexual partners. N Engl J Med 1994, 331:341–346. Downs AM, De Vincenzi I, for the European Study Group in Heterosexual Transmission of HIV: Probability of heterosexual transmission of HIV: relationship to the number of unprotected sexual contacts. J Acquir Immune Defic Syndr Hum Retrovirol 1996, 11:388–395. Mayer KH, Anderson DJ: Heterosexual HIV transmission. Infect Agent Dis 1995, 4:273–284. Khouri YF, McIntosh K, Cavacini L, et al.: Vertical transmission of HIV-1: correlation with maternal viral load and plasma levels of CD4 binding site anti-gp120 antibodies. J Clin Invest 1995, 95:732–737. Centers for Disease Control and Prevention: Case–control study of HIV seroconversion in health-care workers after percutaneous exposure to HIV-infected blood: France, United Kingdom, and United States, January 1988–August 1994. MMWR 1995, 44:929–933. Anderson DJ, Obrien TR, Politch JA, 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. Krieger JN, Coombs RW, Collier AC, et al.: Recovery of human immunodeficiency virus type 1 from semen: minimal impact of stage of infection and current antiviral chemotherapy. J Infect Dis 1991, 163:386–388. Van Voorhis BJ, Martinez A, Mayer K, Anderson DJ: Detection of human immunodeficiency virus type 1 in semen from seropositive men using culture and polymerase chain reaction deoxyribonucleic acid amplification techniques. Fertil Steril 1991, 55:588–594. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 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. Hamed KA, Winters MA, Holodniy M, Katzenstein DA, Merigan TC: Detection of human immunodeficiency virus type 1 in semen: effects of disease stage and nucleoside therapy. J Infect Dis 1993, 167:798–802. Mermin JH, Holodniy M, Katzenstein DA, Merigan TC: Detection of human immunodeficiency virus DNA and RNA in semen by the polymerase chain reaction. J Infect Dis 1991, 164:769–772. Tindall B, Evans L, Cunningham P, et al.: Identification of HIV-1 in semen following primary HIV-1 infection. AIDS 1992, 6:949–952. Dyer JR, Gilliam BL, Eron JJ, Grosso L, Cohen MS, Fiscus SA: Quantitation of human immunodeficiency virus type 1 RNA in cell free seminal plasma: comparison of NASBA TM with AmplicorTM reverse transcription–PCR amplification and correlation with quantitative culture. J Virol Methods 1996, 60:161–170. Rasheed S, Li Z, Xu D: Human immunodeficiency virus load. Quantitative assessment in semen from seropositive individuals and in spiked seminal plasma. J Reprod Med 1995, 40:747–757. Jurriaans S, Vernazza PL, Goudsmit J, Boogaard J, Vangemen B: HIV-1 viral load in semen versus blood plasma. XI International Conference on AIDS. Vancouver, July 1996 [abstract 4036]. Centers for Disease Control: 1993 Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR 1992, 41 (RR17):1–19. Myers LE, McQuay LJ, Hollinger FB: Dilution assay statistics. J Clin Microbiol 1994, 32:732–739. Vangemen B, Vanbeuningen R, Nabbe A, et al.: A one-tube quantitative HIV-1 RNA NASBA nucleic acid amplification assay using electrochemiluminescent (ECL) labelled probes. J Virol Methods 1994, 49:157–167. Milman G, Sharma O: Mechanisms of HIV/SIV mucosal transmission. AIDS Res Hum Retroviruses 1994, 10:1305–1312. Perelson AS, Neumann AU, Markowitz M, Leonard JM, Ho DD: HIV-1 dynamics in vivo: virion clearance rate, infected cell lifespan, and viral generation time. Science 1996, 271:1582–1586. Zhu T, Wang N, Carr A, et al. : Genetic characterisation of human immunodeficiency virus type 1 in blood and genital secretions: evidence for viral compartmentalization and selection during sexual transmission. J Virol 1996, 70:3098–3107. Mellors JW, Rinaldo CR, Gupta P, White RM, Todd JA, Kingsley LA: Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science 1996, 272:1167–1170. Goedert JJ, Eyster ME, Biggar RJ, Blattner WA: Heterosexual transmission of human immunodeficiency virus: association with severe depletion of T-helper lymphocytes in men with hemophilia. AIDS Res Hum Retroviruses 1987, 3:355–361. Seage GR, Horsburgh CR Jr, Hardy AM, et al.: Increased suppressor T cells in probable transmitters of human immunodeficiency virus infection. Am J Public Health 1989, 79:1638–1642. Seage GR, Mayer KH, Horsburgh CR Jr: Risk of human immunodeficiency virus infection from unprotected receptive anal intercourse increases with decline in immunologic status of infected partners. Am J Epidemiol 1993, 137:899–908. Speck C, Coombs R, Koutsky L, et al.: Rates and determinants of HIV-1 shedding in semen. XI International Conference on AIDS. Vancouver, July 1996 [abstract 334]. Jacquez JA, Koopman JS, Simon CP, Longini IM: Role of the primary infection in epidemics of HIV infection in gay cohorts. J Acquir Immune Defic Syndr 1994, 7:1169–1184. Dyer JR, Gilliam BL, Eron Jr JJ, Cohen MS, Fiscus SA, Vernazza PL: Shedding of HIV-1 in semen during primary infection [letter]. AIDS 1997, 11:543–545. Musicco M, Lazzarin A, Nicolosi A, et al.: Antiretroviral treatment of men infected with human immunodeficiency virus type 1 reduces the incidence of heterosexual transmission. Italian Study Group on HIV Heterosexual Transmission. Arch Intern Med 1994, 154:1971–1976. Gilliam BL, Dyer J, Fiscus SA, et al.: Effects of reverse transcriptase inhibitor therapy on HIV-1 viral burden in semen. J Acquir Immune Defic Syndr Hum Retrovirol 1997 (in press). 993