Fertility treatment can use semen from men with HIV
By Genevra Pittman|
NEW YORK | Thu Mar 24, 2011
NEW YORK (Reuters Health) - Fertility treatments can be done safely and effectively in couples where the man is infected with the AIDS virus and the women isn't, according to a new review of past studies.
Over the last 2 decades, researchers have improved methods of "washing" the semen of men infected with HIV, the virus that causes AIDS. Unwashed semen could pass HIV to the woman or their baby.
"I think the procedure is getting safer and safer," said Dr. Deborah Anderson, a scientist at the Boston University School of Medicine who studies HIV. She was not involved in the current research, but she told Reuters Health that washing the man's semen lowers the risk of transmission enough that "it's an acceptable ... procedure for couples that really want to have children."
In the new review, published in the journal Fertility and Sterility, researchers from the Evandro Chagas Clinical Research Institute in Rio de Janeiro, Brazil looked at 17 earlier studies involving a total of about 1,800 couples in which only the male partner had HIV.
In each of the studies, researchers performed one of two common types of fertility treatments after washing the semen. Then they recorded how often women became pregnant after the procedures. They also monitored the women and any babies they had as a result of the procedures, to see whether HIV had been passed on from the semen.
About a third of the women had a procedure in which a single sperm is injected into a single egg; then the fertilized egg is placed into the woman's womb. This kind of fertility treatment is assumed to be safer for couples in which the male partner has HIV because it is easier to ensure that the sperm being used does not have the HIV virus.
The rest of the women had sperm injected directly into the womb, when their eggs were most likely to be there.
Ultimately, roughly half the women became pregnant, and about 80 to 85 percent of the pregnancies resulted in the birth of a baby.
The success rates for pregnancy were comparable to what has been shown in other studies of fertility treatment in couples without HIV. If anything, couples in the current study may have been more likely to get pregnant using fertility treatments because many of them had no underlying fertility problems, the authors say.
None of the women in the study, or babies that were born after fertility treatments, tested positive for HIV. However, in a few of the studies in which researchers tested semen after it was washed, between two and eight of every 100 samples tested positive for HIV - indicating that it still may be possible, if unlikely, for the virus to be passed either to the woman or to the fetus.
However, the findings are "very reassuring," according to Dr. Elizabeth Ginsburg of the Brigham and Women's Hospital Center for Reproductive Medicine in Boston.
Ginsburg, who was not involved in the study, said that even if some of the samples did test positive for HIV, the amount of the virus was probably so small that it wasn't likely to be passed to the mother or baby. In addition, she said, HIV transmission requires some sort of trauma to the woman's body because the virus is passed from semen to blood, and although there's a chance of that in intercourse, it's not likely in fertility treatment.
Despite mounting evidence of its safety, fertility procedures are not very common in couples in which the male partner has HIV. In part that's because the procedures aren't often covered by insurance, Ginsburg said. Although some fertility procedures may be as inexpensive as $1,000, others run many times higher.
"One of the things that is a shame is that when couples can't afford fertility treatment, they're stuck with the other option, which is having timed intercourse, and that puts the woman at risk," Ginsburg said.
Anderson said a new option for these couples might become available in the future -- medications that the woman can take to avoid getting the virus from her partner who has HIV. And, "if the mom doesn't get it, the baby's not going to get it," she said. "I think that's going to be the future of this field."
So far, only a couple of early studies have been done on the drugs' effectiveness at preventing transmission of the virus, and for now, Anderson said, fertility treatment is the safest possible option for these couples.
Systematic review of the effectiveness and safety of assisted reproduction techniques in couples serodiscordant for human immunodeficiency virus where the man is positive
To evaluate the effectiveness and safety of assisted reproduction techniques (ART) in human immunodeficiency virus (HIV) serodiscordant couples.
Systematic review of five databases of noncomparative open intervention and observational studies of serodiscordant couples undergoing ART, with study selection and data extraction performed independently and in duplicate.
Tertiary fertility centers.
HIV serodiscordant couples where the man is HIV positive.
Intrauterine insemination (IUI), in vitro fertilization (IVF), or intracytoplasmic injection (ICSI) performed after washed semen viral testing.
Main Outcome Measure(s)
Pregnancy rates per cycle, cumulative pregnancy, abortion rate, and HIV seroconversion in newborns or women.
Of the 658 abstracts retrieved, 41 were selected for review, and 17 full articles were included (3,900 IUI cycles in 1,184 couples in 11 aggregated studies and 738 ICSI/IVF cycles in 579 couples across 10 studies). The IUI and ICSI results were, respectively: pregnancy rates per cycle, 18% and 38.1%; cumulative pregnancy, 50% and 52.9%; and abortion rate, 15.6% and 20.6%. No seroconversions in women or newborns were detectable at birth or after 3 to 6 months.
Cumulative evidence suggests that ART is safe and effective for avoiding horizontal and vertical transmission in HIV serodiscordant couples.
Infection with human immunodeficiency virus (HIV) affects approximately 33 million people worldwide (1), over 80% of whom are of reproductive age. Until the 1990s, HIV infection was an absolute contraindication to pregnancy, and couples where the man was HIV positive were not considered eligible for assisted reproduction technology (ART). The emergence of more effective antiretroviral treatment engendered important reductions in viral replication and significant improvement in the immune systems of patients. These advances in treatment resulted in extended life expectancy, improved quality of life, and decreased risk of sexual and vertical transmission (1% to 2%) 2, 3, 4. Advances in semen preparation (5) were followed by the use of intrauterine insemination (IUI) in HIV serodiscordant couples, which resulted in a low risk of transmission. Other studies reported the use of alternative techniques such as in vitro fertilization (IVF) and intracytoplasmic sperm injection (ICSI) to reduce the risk of HIV exposure to the woman and to the gamete 5, 6.
Because HIV is both freely present in semen and associated with the accompanying cells, sperm washing and testing, along with ART in which only one spermatozoid is used (ICSI), appear to significantly decrease the chances of transmission (7). As a result, women of childbearing age with associated infertility factors wishing to have children can undergo ART with low rates of vertical transmission, and serodiscordant couples can also use these techniques to reduce the risk of transmission to the seronegative partner. In an HIV serodiscordant couple where the man is HIV positive, the risk of the woman becoming infected through unprotected sex is 0.1% to 0.2% (8). The couple should be informed of this risk (8) and of the treatment options available to provide a reasonable chance of pregnancy with minimal risk of viral transmission.
Although some studies have suggested that ART can be used 5, 9, there is still controversy regarding the safety and effectiveness of these methods; to date, there has been no systematic review to support the use of ART in HIV serodiscordant couples (7). We evaluated the effectiveness and safety of ART for HIV serodiscordant couples where the index case is the man. We describe the frequency of semen preparation techniques, the rates of HIV positivity in washed semen, the frequency of HIV seroconversion, and the pregnancy and abortion rates in ART.
Materials and methods
A systematic review of literature was performed using studies on HIV serodiscordant couples undergoing ART. The review was conducted in accordance with the Meta-analysis of Observational Studies in Epidemiology(MOOSE) (10) and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (11). We selected studies with HIV serodiscordant couples of childbearing age in which the man was the index case and who were assisted at tertiary infertility clinics, regardless of infertility factors.
We included several types of studies: noncomparative open intervention, observational cohort, case controls, and case series studies of HIV serodiscordant couples undergoing IUI or ICSI-IVF after viral testing of the seminal wash. Comparative randomized intervention studies, such as clinical trials, were not included because these were not available, possibly due to the ethical implications of conducting such a trial (7). The exclusion criteria were studies in which the main objective was to test postwash semen for HIV positivity without evaluating seroconversion; studies that involved oocyte donation cycles; and studies with a small sample size (fewer than five couples).
A literature search was performed for articles published before December 2007 on remote databases such as PubMed, LILACS, SciELO, Scirus, the Cochrane Library, and Scopus, on university websites for master's dissertations, in doctoral theses, and in abstracts of congresses and symposia in the field of assisted reproduction and gynecology. The search strategy included the following key words and subject headings: artificial insemination (AI), intracytoplasmic sperm injection (ICSI), HIV infection (HIV), assisted reproductive techniques (reproduction techniques), and infertility (infertility therapy). No language restrictions were imposed. Where applicable, the investigators were contacted by e-mail to complement article data or to provide an English version of the manuscript.
Primary outcomes were ART effectiveness, as measured by the frequency of pregnancies per cycle (pregnancy after embryo visualization) and by cumulative pregnancy; and ART safety, as measured by HIV seroconversion after IUI or ICSI procedure, according to postwash seminal testing. The female partner's HIV seroconversion was evaluated by polymerase chain reaction (PCR) or serology during pregnancy, childbirth, after abortion, or after the ART procedure when pregnancy had not yet occurred.
Secondary outcomes included the abortion and fertilization rates. We also examined possible confounding variables, including  female infertility factors, such as hormone assays and ultrasonography for ovulatory function and hysterosalpingography to investigate tubal patency; and  sexually transmitted diseases, measured by cervical mucus culture, serological testing, CD4+ T-cell count, and HIV viral load in the male partner before ART and possible associations with HIV positivity in postwash semen.
Two authors (R.L.V. and S.R.L.P.), who were blinded to the investigator names and journal titles of the reviewed literature, independently reviewed the selected abstracts and articles. Hand searching was performed in the reference lists of the selected articles.
Our search retrieved 1,049 abstracts, of which 391 were duplicates and 619 did not meet the inclusion criteria. We selected 39 abstracts from articles published in Portuguese (1), Hebrew (1), Spanish (1), English (35), French (1), Dutch (1), Czech (1), and German (1). Two additional studies were selected from cross-references for a total of 41 full articles, of which 17 were eligible for the present study on HIV serodiscordant couples, including eight open trials, four cohorts, two case series, and three case-control studies. The majority of these studies were performed in Europe, particularly in Spain, or in the United States. None of the studies were conducted in South America, Africa, or Asia. Of the 24 excluded articles, 12 had restricted inclusion criteria, nine had small sample sizes (fewer than 5 couples), and three had insufficient information even after e-mail contact.
Eligibility criteria for male partners in most of the selected articles included stable viral load in the previous 4 to 6 months; CD4+ T-cell counts above 200 cells per mm3; stable clinical condition despite antiretroviral therapy; regular outpatient follow-up evaluations; semen culture; consistent use of condoms; and agreement with the clinical procedures. The median ages (interquartile range) for IUI and ICSI/IVF studies were 33.2 years (range: 30.0 to 35.5 years) and 34.6 years (range: 33.4 to 36.6 years), respectively, for women, and 33.3 years (range: 31.9 to 38.0 years) and 37.2 years (range: 33.7 to 39.0 years), respectively, for men.
The maximum viral load was 49.000, and the minimum CD4+ T-cell count was 400 6, 12, 13, 14, 15, 16, 17. Four studies reported the use of antiretroviral treatment with no mention of the type of the drug prescribed 5, 6, 13, 17. Those reporting the AIDS clinical classification limited the participants to the A1 to B2 categories 5, 9, 13, 14.
Table 1 summarizes the associated infertility factors 5, 9, 13, 14, 18, 19, 20, 21, the drugs used for ovarian stimulation or to prevent LH surge (ICSI-IVF), and the procedure used to confirm pregnancy. Ten studies did not report the associated infertility factors. The most frequent drug used for ovarian stimulation was recombinant follicle-stimulating hormone (FSH), but natural cycles also were used 5, 20. A gonadotropin-releasing hormone (GnRH) agonist was used to prevent luteinizing hormone (LH) surge 6, 14, 15, 17, 21, 22, but one study reported the use of both a GnRH agonist and an antagonist (13).
The most frequently employed technique in semen preparation combined density gradient and the swim-up technique. Only 4 of the 11 studies reported postwash semen viral testing, and the HIV positivity rates ranged from 2.5% to 7.7% 9, 13 (Table 2). The PCR techniques in the postwash semen testing varied, but there was no seroconversion despite the differences in the sensitivity of the methods (Table 3, Table 4). In the 3,900 IUI cycles and the 738 ICSI-IVF cycles of HIV serodiscordant couples included in this review, there were no cases of seroconversion in female partners or newborns, either at birth or at the 3- or 6-month follow-up evaluation after ART, regardless of the semen preparation or the technique used to the testviral load.
Eleven studies included cycles of IUI, 10 of which also presented the results of ICSI-IVF cycles.
A total of 3,900 cycles were evaluated in 1,184 HIV serodiscordant couples undergoing IUI (see Table 3). The median of the pregnancy rate per cycle was 18.0% (14.5% to 23.0%), while the median of the cumulative pregnancy rate was 50.0% (from 40.0% to 63.1%). The median abortion rate was 15.6% (9.5% to 24.7%).
Table 4 summarizes the results of the 10 retrieved articles evaluating the performance of ICSI-IVF in HIV serodiscordant couples. A total of 738 cycles in 579 serodiscordant couples were assessed. The median of the pregnancy rates per cycle, the cumulative pregnancy rates, and the abortion rates were 38.1% (24.8% to 46.2%), 52.9% (41.0% to 67.5%), and 20.6% (9.3% to 29.5%), respectively. The number of cycles was likely underestimated, as three studies 13, 16, 22 provided the number of cycles for all participants, including hepatitis C virus (HCV) and coinfected patients, rather than for the HIV serodiscordant couples alone. The median number of collected oocytes was 13 (range: 9.6 to 15.8), and the median number of transferred embryos was 2.9 (range: 2.5 to 3.5). The fertilization rate was 71.5% (range: 50.1% to 77.1%).
This is the first systematic review to evaluate the effectiveness and safety of ART in HIV serodiscordant couples. The absence of seroconversion suggests that the use of ART avoids HIV sexual and vertical transmission. The choice of these techniques depends solely on the infertility factors. The effectiveness rates for successful pregnancy were similar to those of the general population with equivalent infertility and age profiles.
The majority of the articles did not contain information concerning the HIV status in postwash semen, which would be relevant to ascertain the safety of ART. Although the clinical and immunologic profiles of the male partners were not clearly defined in many studies, in nine of the studies 6, 9, 12, 14, 15, 17, 18, 22, 23 ART was restricted to a group of patients with improved HIV clinical status, low HIV viral load, and high CD4+ T-cell counts. Thus, the reported absence of HIV seroconversion in female partners and newborns may only apply to HIV serodiscordant couples with similar index case profiles. It would be useful to establish which parameters should be considered determinants of the safety of ART to not limit the access to ART for couples who are living with HIV/AIDS and want to conceive using these techniques.
Despite similar frequencies of ICSI-IVF and IUI in the included studies, there have been persistent doubts in several countries regarding the use of these different techniques for HIV serodiscordant couples. Although ICSI seems intuitively safer because it uses one single spermatozoid from a previously prepared and tested sample of semen, neither mothers nor newborns underwent HIV seroconversion when any of the three techniques was used. Our review highlights the significance of these findings because the majority of the studies, totaling 1,184 couples and 3,900 cycles, employed IUI (compared with 539 couples and 741 cycles of ICSI-IVF), which theoretically offers a higher risk and is less effective.
Publication bias must be considered. We were unable to identify studies reporting HIV seroconversion despite our efforts to retrieve all ART published studies without using language restriction in the search strategy.
The effectiveness of IUI techniques was consistent with other reports for the general female population. The cumulative pregnancy rate (40% to 63.1%) was within the range described by Tay et al. (16% to 60%) (24), similar to females without tubal occlusion (42.5%) (25) and improved (24.2%) relative to a large cohort study in women younger than 35 years (26). The highest cumulative pregnancy rate (78.0%), reported by Savasi et al. (15), may be partially explained by the greater number of cycles per couple.
The median pregnancy rate per cycle of IUI was 18.0%, which is higher than the rate reported in the general population of the same age group (11.5%) (26) and may be due to the fact that there was no evidence of infertility in half of these patients. In one study of younger women (9), the pregnancy rate per cycle was elevated (28.8%). The rate of abortion varied from 8.6% to 42.8%, which was more frequent than the 7.8% rate reported by Zadehmodarres et al. (27).
The range (17.7% to 49.7%) of pregnancy rates per cycle encompassed the ranges described for IVF (26.9% to 30.3%) and ICSI (28.5% to 30.9%) by the European Society of Human Reproduction and Embryology (28). The interval of the cumulative pregnancy rate (36.4% to 67.5%) was higher (28.3% to 38%) than that indicated in a recent report (29). This higher rate may be due to the exclusion of canceled cycles (17). The median abortion rate was 20.6%, possibly due to the age of the women studied (33 to 38 years), which was slightly higher compared with other articles on ICSI-IVF techniques.
Several studies included the results of ICSI and IVF cycles without distinguishing between the two methods. The fertilization rate is higher in ICSI cases; this factor may influence the number of embryos collected and, therefore, the number of embryos transferred. In addition, the majority of the women included in these studies presented with no infertility factors. A further search of the literature retrieved two additional ICSI articles with similar pregnancy rates results compared with a matched seronegative control group 30, 31. However, Queiroz et al. (32) reported a lower pregnancy rate in HIV serodiscordant couples (12.5%) compared with seronegative couples (52.9%).
A meta-analysis was not performed because evidence from the observational studies was better described in terms of frequency and range than by a summary measure 33, 34. The studies were not homogeneous regarding induction protocols, the age of the women, the factors associated with infertility, or the exclusion of cycles with poor responses. All of the articles included in our review were observational, possibly because comparisons between ART and unprotected sex would not be ethical. Recent descriptive studies included HIV seronegative couples as a control group 30, 31, 32. Standardized outcomes and measures of association would enable a meta-analysis study.
It is possible that our findings are applicable to HIV serodiscordant couples from middle to low income countries with similar clinical profiles because there is no reason to suspect that these populations would greatly differ. The high cost of these techniques must be balanced against considerations for the safety in countries with higher rates of HIV infection.