Share this article on:

Sexual Transmission of HIV-1 Among Serodiscordant Couples in Porto Alegre, Southern Brazil

Melo, Marineide Gonçalves MD, MSc*; Santos, Breno Riegel MD*; De Cassia Lira, Rita*; Varella, Ivana S. PHD*; Turella, Maria Lourdes RN*; Rocha, Tauí Melo BS*; Nielsen-Saines, Karin MD, MPH

doi: 10.1097/OLQ.0b013e31817e2491

Background: A cohort of 93 heterosexual HIV serodiscordant couples with no prior antiretroviral use were identified in a large referral center from February 2000 to January 2006 in southern Brazil.

Methods: Review of clinic records retrospectively identified 56 cases of untreated index cases whereas 37 couples were identified prospectively. Demographics, medical, and laboratory data were obtained. During follow-up, 41/93 index cases (44%) initiated antiretrovirals (ARVs) and from 52 without ARV use, 4 were lost to follow-up. Median viral loads were used to compare transmitters versus nontransmitters (Mann-Whitney test).

Results: Sixty-seven (72%) index cases were female (49% identified during ante-natal care). Unprotected sexual intercourse as a risk factor for HIV-1 infection was significantly higher as compared to intravenous drug use (P < 0.0001) in female index partners but not in male index cases. Sexually transmitted diseases were identified in 22 cases (24%). Six HIV-1 seroconversions occurred (6.5%). In all cases index partners were not using ARVs at the time of seroconversion. Among 26 couples with a male index case, there were 4 seroconversions (15%) and among 67 female index cases there were 2 seroconversions (3%). All seroconversions occurred with virus loads >1000 copies/mL. Eight female index cases (22%) reported no condom use.

Conclusions: Heterosexual transmission occurred more frequently from HIV-infected males to females (rate ratio 3.5; CI, 95% 0.8–16.5 P = 0.259), although without statistical significance, probably because of the small sample. Transmitters showed significantly higher median viral loads (P = 0.042) suggesting that heterosexual transmission of HIV is more a function of viral load than gender of index case. ARV use may play a role in the prevention of HIV-1 heterosexual transmission. Other factors may be involved and should be further evaluated in larger cohorts.

Study of HIV-1 serodiscordant couples demonstrated transmission rates of 6.5%, 15% men-women and 3% women-men. All transmitters had virus load >1000 copies and 0/41 patients on antiretrovirals transmitted.

From the *Department of Infectious Diseases, Hospital Nossa Senhora da Conceição/GHC, Porto Alegre, Brazil; and †Department of Pediatrics, David Geffen UCLA School of Medicine, Los Angeles, California

The authors are very grateful to Drs. Eduard J. Beck (WHO and Department of Epidemiology and Biostatistics, McGill University) and Sundhiya Mandalia (Department of Medicine, Imperial College, Chelsea and Westminster Hospital, London) for statistical advice.

Correspondence: Karin Nielsen-Saines, MD, MPH, Department of Pediatrics, David Geffen UCLA School of Medicine, 10833 LeConte Avenue, Los Angeles, CA 90095. E-mail:

Received for publication August 21, 2007, and accepted April 29, 2008.

THE RISK OF HIV HETEROSEXUAL TRANSMISSION through unprotected intercourse depends on myriad aspects, including the frequency of intercourse, sexual practices (i.e., vaginal vs. anal intercourse), presence of concurrent STDs such as genital ulcerative disease, HIV-1 virus load in plasma, semen and vaginal secretions, disease stage and CD4 cell count, use of antiretroviral therapy, particularly highly active antiretroviral therapy (HAART), and other factors such as personal hygiene and male circumcision.1–7 In Latin America, the HIV epidemic is fueled by several factors, including early initiation of sexual activity, overall lack of adherence to condoms and the frequent presence of multiple sexual partners.8 HAART has been clearly demonstrated to be of profound benefit to patients with HIV by prolonging and improving the quality of life. Nevertheless, in some instances it has been associated with loosening of safe sexual practices, such as carelessness in the use of condoms and other high risk behaviors. These can often result in an increased incidence of sexually transmitted diseases (STDs) and transmission of HIV-1 multidrug resistant virus to the uninfected partner, particularly in casual encounters.3 On the other hand, the benefit of the use of antiretrovirals (ARVs) for prevention of HIV-1 acquisition has been clearly demonstrated through the successful prevention of mother-to-child HIV-1 transmission9–11 and also through anecdotal reports of reduction in heterosexual transmission of HIV-1 among serodiscordant couples where the infected partner was receiving antiretroviral therapy.4

A cohort of 93 heterosexual HIV-1 serodisordant couples at a large urban HIV/AIDS referral center at the Hospital Nossa Senhora da Conceição in the city of Porto Alegre, Southern Brazil was followed from February 2000 to January 2006. Clinical and laboratory data were analyzed and epidemiologic and demographic characteristics of the cohort were obtained. The presence of STDs in both partners, incidence of HIV heterosexual transmission according to index case gender and the difference between median viral loads of transmitters versus nontransmitters was evaluated.

Back to Top | Article Outline

Materials and Methods

Study Population and Laboratory Assays

In preparation for a large multicenter randomized clinical trial of HIV-1 serodiscordant couples, we researched the number of antiretroviral naïve HIV-1 infected individuals receiving medical care at our institution with a steady HIV-1 uninfected partner of the opposite sex. Among 4500 patients, 56 (1.2%) fulfilled this criteria and were retrospectively enrolled. In addition, 37 couples were enrolled prospectively in the first year of observation. Such additional cases were identified during regular clinic appointments and from a network of basic health units in the vicinity of our medical center supported by our institution. All new patients were interviewed and a questionnaire with sexual practices was presented. To the uninfected partner, pretest counseling was offered before a new anti-HIV ELISA test was performed. Anti-HIV testing was performed following local standards requiring two enzyme-based tests with confirmation through immunofluorescence and repetition of the enzyme-based tests in a second sample. For the index case a physical exam to exclude any AIDS defining diagnosis was performed in the initial assessment. Laboratory tests including CD4, viral load (bDNA according to local standards) were ordered as part of the routine health evaluation. Gynecological examination, also part of our routine assessment, was performed every 6 months. Retrospectively identified cases were defined as serodiscordant if at least one negative anti-HIV test of a regular partner was available at the last appointment. All other relevant information (viral load, ARV initiation because of disease progression) was extracted. All patients provided informed consent for study participation.

Back to Top | Article Outline

Statistical Analysis

Epidemiologic and demographic aspects of the 93 couples were described. For comparison of proportions the 2-sided χ2 test or Fisher exact test was used. Mann-Whitney test was used to compare viral load and time of observation medians between transmitters versus nontransmitters and between index case genders. Incidence density was used to evaluate transmission rates among serodiscordant couples. Analyses were performed using the Statistical Package for the Social Sciences (SPSS Version12.0) e PEPI version 4.0.

Back to Top | Article Outline


Ninety-three HIV-1 serodiscordant couples in whom the index case was not using ARV at the moment of case identification (February 1, 2000 to January 31, 2006) were included. From the 93 couples, 67 (72%) index cases were women (95% CI: 62.3%–80.4%) and 33 of these (49.2%) were identified during ante-natal care. Unprotected sexual intercourse as a risk factor for HIV-1 infection in female index cases was significantly higher as the major risk factor for HIV-1 acquisition as compared with intravenous drug use (IDU) (P < 0.0001). Among male index cases, there was no difference between rates of IDU and unprotected sexual intercourse in terms of risk factors for HIV-1 acquisition (P = 0.405) (Table 1). During follow-up, 41 of the 93 index cases (44.1%) initiated ARV. There was a statistically significant difference in ARV use in female index cases for whom the major indication was mother to child prevention of HIV transmission (MTCT) as opposed to other causes (P < 0.0001). STDs were identified in 22 cases (23.6%). No men in the cohort were circumcised.



A sexual assessment interview of 37 prospectively identified couples, comprised by 13 men (35.1%) and 24 women (64.9%) index cases, indicated that 8 female index cases (21.6%) reported not using protection at any time, whereas all male index cases reported regular condom use.

Treatment with HAART was initiated in 41 index cases. The main reason for initiating ARVs was pregnancy in 33 female patients, and CD4 count <350 in 8 cases (5 male index cases and 3 females). The most commonly used HAART regimen consisted of zidovudine, lamivudine, and nelfinavir in pregnant female patients and zidovudine, lamivudine, and efavirenz in the 5 male index cases who initiated therapy. All patients who initiated HAART had a complete virologic response and developed an undetectable virus load as measured by nucleic acid sequence based amplification (NASBA). No treatment switches were required in any patients because they were all treatment naive at baseline and had a complete virologic response. No significant toxicities were observed in the treated cohort. None of the patients who initiated antiretroviral treatment during the observation period transmitted infection to their partners. There were 6 seroconversions during the time of observation (Incidence: 6.45; 95%CI: 2.65–12.93). All 6 transmissions occurred in the 52 patients who were not receiving HAART and 0 occurred in the 41 patients on ARV treatment (Incidence: 11.5%; 95%CI: 4.81–22.45). The incidence of transmission by gender is shown in Table 2. A comparison between median follow-up time of observation between transmitting and nontransmitting couples did not show any statistical significance: transmitters: 25.5 months of observation; range, 3 to 71 months; nontransmitters: 22.34 months; range, 10 to 24 months (P = 0.367). The comparison between median follow-up time of observation by gender of the index case also did not show any statistically significant difference: men: 23.8 months; range, 5 to 71 months; women: 25 months; range, 3 to 56 months (P = 0.948). The median viral load of transmitters was greater than that of nontransmitters and the difference was statistically significant (P = 0.042). Conversely, when viral load between genders was compared, there was no statistically significant difference (P = 0.142) (Table 3).





Back to Top | Article Outline


HIV serodiscordant couples in stable relationships are not usually the target of public health interventions for reduction of HIV transmission. Nevertheless, these couples benefit from ongoing counseling and testing. Studies have demonstrated that systematic efforts to identify serodiscordant couples in stable relationships have an impact on HIV transmission and may increase antiretroviral adherence.12,13 In the present cohort, the majority of HIV-1 infected index cases in serodiscordant relationships were women who acquired infection through unprotected sexual activity and were identified during regular prenatal care, a common feature in many settings.14 In men, acquisition of HIV-1 infection through IDU occurred in similar proportions as acquisition of infection through sexual intercourse. ARV use was much more frequently initiated for MTCT purposes than for a low CD4 count or disease progression in our study. Although it was not possible to correlate an STD diagnosis with an HIV transmission event, a high incidence of STDs (23.6%) was observed in our patients.

Although we could not find a correlation between condom use and heterosexual HIV transmission, it is interesting to note that all couples reporting no condom use were serodiscordant couples with a female index case, whereas couples with male index cases reported consistent practice of safe sex. Despite these reports, 4 male index patients transmitted infection to their partners, and in 1 of these cases, pregnancy also occurred. Thus it is debatable whether condom use was truly consistent. The timing of seroconversion was associated with development of pregnancy in half of our seroconversion cases. This suggests that the desire to have children may influence risk taking behaviors in serodiscordant couples. Other studies of serodiscordant couples have reported a high rate of unsafe sexual practices among individuals in stable relationships.15 Among uninfected women in such relationships, seroconversion during pregnancy is a major risk factor for HIV-1 transmission to the infant.16 Nevertheless, the rate of transmission seems to vary by geographic area, with a very low HIV incidence observed among serodiscordant couples in Pune, India,17 as opposed to African cohorts from Uganda, Tanzania, and Zambia2,18–20. These findings may be due to differences in condom use across cohorts, and variations in median CD4 cell counts, the magnitude of virus load, duration of the observation period, and rate of STIs.

In our study, heterosexual transmission occurred more frequently from an HIV infected male to a female (15% vs. 3% transmission). This difference was not statistically significant likely because of our small sample size, a limitation of this study. Studies conducted in North America and Europe demonstrated higher transmission rates from men to women.21,22 However, other studies in large African cohorts have shown the opposite, with a higher efficiency of HIV sexual transmission from female to male partners.1,2 Others have not found a statistically significant difference in sexual HIV transmission risk by gender.23 It seems that the most important predictor of transmission is the magnitude of the virus load, a risk factor of greater significance than patient gender.23 In studies conducted in Uganda, virus load was significantly associated with HIV transmission and much more frequently elevated among individuals in seroconcordant couples than infected individuals in serodiscordant relationships.24 HIV transmission seems to occur as a function of virus load in serum, semen, and genital secretions.2,25 With the widespread use of ARVs, transmission events seem to have been drastically reduced in the HAART era.4,6 In our cohort, we did not observe any transmission events in patients receiving ARVs or with undetectable virus loads.

Our data suggests a possible protective role of ARV use in the prevention of HIV-1 heterosexual transmission between sexual partners who are HIV-1 serodiscordant. Large randomized clinical studies evaluating the role of ARV in prevention of sexual transmission of HIV-1 infection are ongoing and should further elucidate the role of virus load and multiple additional risk factors involved in the pathogenesis of HIV-1 transmission.

Back to Top | Article Outline


1. Gray RH, Wawer MJ, Brookmeyer R, et al. Probability of HIV-1 transmission per coital act in monogamous, heterosexual, HIV-1 serodiscordant couples in Rakai, Uganda. Lancet 2001; 357:1149–1153.
2. Quinn TC, Wawer MJ, Sewankambo N, et al. Viral load and heterosexual transmission of human Immunodeficiency virus type 1. N Engl J Med 2000; 342:921–929.
3. Straten A, Gomez CA, Saul J, et al. Sexual risk behaviors among heterosexual HIV serodiscordant couples in the era of post-exposure prevention and viral suppressive therapy. AIDS 2000; 14:47–54.
4. Castilla J, Romero J, Hernando V, et al. Effectiveness of highly active antiretroviral therapy in reducing heterosexual transmission of HIV. J Acquir Immun Defic Syndr 2005; 40:96–101.
5. Desquilbet L, Deveau C, Goujard C, et al. Increase in at-risk sexual behaviour among HIV-1-infected patients followed in the French PRIMO cohort. AIDS 2002; 16:2329–2333.
6. Hosseinipour M, Cohen MS, Vernazza PL, et al. Can antiretroviral therapy be used to prevent sexual transmission of human immunodeficiency virus type 1? Clin Infect Dis 2002; 34:1391–1395.
7. Gupta P, Mellors J, Kingsley L, 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.
8. Schwartländer B, Garnett G, Walker N, et al. AIDS in a new millennium. Science 2000; 289:64–66.
9. Connor EM, Sperling RS, Gelber R, et al. (PACTG 076). Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994; 331:1173–1180.
10. Wortley PM, Lindegren ML, Fleming PL. Successful implementation of perinatal HIV prevention guidelines. MMWR Recomm Rep 2001; 50:15–28.
11. Cooper ER, Charutat M, Mofenson L, et al. Combination antiretroviral therapy for the treatment of pregnant HIV-1-infected women and prevention of perinatal HIV-1 transmission. J Acquir Immune Defic Syndr 2002; 29:484–494.
12. Were WA, Mermin JH, Wamai N, et al. Undiagnosed HIV infection and couple HIV discordance among household members of HIV-infected people receiving antiretroviral therapy in Uganda. J Acquir Immune Defic Syndr 2006; 43:91–95.
13. Bunnell R, Opio A, Musinguzi J, et al. HIV transmission risk behavior among HIV-infected adults in Uganda: results of a nationally representative survey. AIDS 2008; 22:617–624.
14. Siriwasin W, Shaffer N, Roongpisuthipong A, et al. HIV prevalence, risk and partner serodiscordance among pregnant women in Bangkok. JAMA 1998; 280:49–54.
15. Bouhnik AD, Preau M, Lert F, et al. Unsafe sex in regular partnerships among heterosexual persons living with HIV: evidence from a large representative sample of individuals attending outpatients services in France (ANRS-EN12-VESPA Study). AIDS 2007; 21(suppl 1):S57–S62.
16. Nielsen-Saines K, Melo M, Varella I, et al. Primary HIV-1 infection during pregnancy: high rate of HIV-1 MTCT in a cohort of patients in southern Brazil. From the Fourth Dominique Dormont International Conference. Maternal chronic viral infections transmitted to infants: from mechanisms to prevention and care. Retrovirology 2008; 5(suppl 1): O1. Available at:
17. Mehendale SM, Ghate MV, Kumar BK, et al. Low HIV-1 incidence among married serodiscordant couples in Pune, India. J Acquir Immune Defic Syndr 2006; 41:371–373.
18. Carpenter LM, Kamali A, Ruberantwari A, et al. Rates of HIV-1 transmission within marriage in rural Uganda in relation to the HIV sero-status of the partners. AIDS 1999; 13:1083–1089.
19. Hugonnet S, Mosha F, Todd J, et al. Incidence of HIV infection in stable sexual partnerships: a retrospective cohort study of 1802 couples in Mwanza region, Tanzania. J Acquir Immune Defic Syndr 2002; 30:73–80.
20. Fideli US, Allen SA, Musonda R, et al. Virologic and immunologic determinants of heterosexual transmission of human immunodeficiency virus Type I in Africa. AIDS Res Human Retrovirus 2001; 17:901–910.
21. Royce RA, Sena A, Cates W, et al. Sexual transmission of HIV. N Engl J Med 1977; 336:1072–1078.
22. Nicolosi A, Leite MLC, Musicco M, et al. For the Italian Study Group on HIV Heterossexual Transmission. The efficiency of male-to-female and female-to-male sexual transmission of human immunodeficiency virus: a study of 750 stable couples. Epidemiology 1994; 5:570–575.
23. Wawer MJ, Gray RH, Sewarkambo NK, et al. Rates of HIV-1 transmission per coital act, by stage of HIV-1 infection, in Rakai, Uganda. J Infect Dis 2005; 191:1403–1409.
24. Malamba SS, Mermin JH, Bunnell R, et al. Couples at risk: HIV-1 concordance and discordance among sexual partners receiving voluntary counseling and testing in Uganda. J Acquir Immune Defic Syndr 2005; 39:576–580.
25. Nielsen K. HIV-1 transmission and prevention: safe sex, PREP and PEP. Inf Med 2005; 22:175–186.
© Copyright 2008 American Sexually Transmitted Diseases Association