Safety of sperm washing and ART outcome in 741 HIV-1-serodiscordant couples

Abstract

BACKGROUND: To evaluate the safety of sperm washing and assisted reproduction technique (ART) outcome offered to serodiscordant couples with a human immunodeficiency virus-1 (HIV-1)-positive male. METHODS: Sperm washing was performed and checked by RT–PCR on each semen sample before its fresh usage. Intrauterine insemination (IUI) or IFV/ICSI was offered according to fertility profile of each couple. Non-infected women underwent HIV testing 2 weeks before each procedure and for up to 6 months after. RESULTS: Seven hundred and forty-one couples entered the study of a possible 2011 serodiscordant couples counselled over 4 years. Superovulation and IUI were performed in 581 couples, where the pregnancy rate per cycle and pregnancy rate per couple were 19 and 78%, respectively, with multiple pregnancy rate being 4%. One hundred and sixty couples were treated by IVF/ICSI, where pregnancy rate per cycle and per couple were 22 and 41%, respectively, with multiple pregnancy rate being 10%. All female partners were still HIV-1 negative at follow-up. CONCLUSION: Sperm washing within a programme of reproductive counselling was proved to be safe in this large series of serodiscordant couples. The overall pregnancy rate (70.3%), independent of the procedure used (IUI or IVF/ICSI), justifies the effort of the medical team in setting up and implementing dedicated centres and of the individual patient in seeking a safe pregnancy.

Introduction

Sperm-washing techniques have substantially changed the paradigm of fathering children in serodiscordant couples for male human immunodeficiency virus (HIV) infection. The clinical value of sperm washing and its negligible or null risk were first reported in 1992 (Semprini et al., 1992), and since then, it has been reinforced by different articles both on methodological (Marina et al., 1998; Anderson, 1999; Gilling-Smith, 2000) and on clinical issues (Sauer and Chang, 2002; Ohl et al., 2003; Pena et al., 2003; Bujan et al., 2004; Garrido et al., 2004; Nicopoullos et al., 2004; Mencaglia et al., 2005). Furthermore, an estimate of the use of sperm washing in Europe (Gilling-Smith, 2000; Savasi et al., 2006) provided additional evidence of the clinical value of this procedure. These results brought assisted reproduction into a new era from the report of the Centers for Disease Control (CDC, 1990). The seroconversion in a woman after an artificial intrauterine insemination (IUI) from her HIV-infected husband was reported, but indeed the semen was not processed to separate lymphocytes from spermatozoa. This unwanted outcome was later reported by Ann Duerr in her review to stress the importance of sperm washing (Duerr and Jamieson, 2003).

In 2002, the American College of Obstetricians and Gynecologists (ACOG Committee on Ethics, 2001) and the American Society for Reproductive Medicine (ASRM) (Ethics Committee of the ASRM, 2002) recommended sperm washing to be offered to HIV-serodiscordant couples as a standard of care. Sauer (2005) reviewed 3019 IUI cycles after sperm washing in 1111 serodiscordant couples with 361 babies born and 543 IVF/ICSI cycles in 352 serodiscordant couples with 131 babies born.

Dramatic changes in life expectancy and quality of life in HIV-infected patients are putting this procedure among the expected medical care for these patients in western countries, whereas biological, clinical and epidemiological issues are still under scrutiny. Although the popularity achieved by this methodology is wide, the large number of reported cycles is still the sum of either small series or larger ones on either IUI (Vernazza et al., 1997; Marina et al., 1998) or ICSI in different centres (Ohl et al., 2003; Pena et al., 2003). Although safety issues need large multicentre follow-up, biological and clinical problems could be addressed by large series. For instance, in some centres the post–sperm-washing control is not routinely used, whereas in others frozen sperm is used just to allow time for the PCR control. Alternatively, ICSI is often used to avoid sperm washing. Ovulation induction in couples undergoing IUI is not universally adopted—some centres recommend and perform ICSI in all male serodiscordant couples, including those with proven fertile partners (Garrido and Meseguer, 2006) and so forth. Obviously, the best answer for each issue is not represented by a single option. However, safety, efficacy and cost efficiency must be considered to rank each one of them.

We report in this article our experience on a series of 741 serodiscordant couples treated in the last 4 years in one University Centre that provided an assisted reproduction programme with a thorough diagnostic triage and a therapeutical offer from IUI to IVF to ICSI, according to the fertility profile of each serodiscordant couple. All treatments were performed with fresh spermatozoa and checked using real-time PCR, after sperm washing.

Materials and methods

Assisted reproduction technology (ART) programme was offered to serodiscordant couples, with HIV-infected male partner, seeking medical assistance. Inclusion criteria were set to protect the couple and eventually the child; partners were to engage only in protected sexual acts. HIV infection was monitored and treated, and long-term compliance was assessed by the infectious disease physician. Standard laboratory criteria were adopted: (i) CD4⁺ lymphocytes >200/mm³ at least twice in the 4 months before treatment; (ii) stable viral load, with no increase >0.5 log in two successive samples during the 4 months before treatment; (iii) infection by a quantifiable amplifiable strain of HIV-1. Each couple was interviewed by a psychologist at inclusion and thereafter whenever necessary.

Both members of the couple signed an informed consent for ARTs, as required by national regulations, including items specifically addressing the risk of viral transmission to the female partner and its possible consequences.

Female fertility was assessed by standard procedures. Each woman underwent a gynaecological examination, a smear test, a vaginal sample for bacteriological testing, a cervical swab for chlamydia and mycoplasma and a vaginal ultrasound examination of the uterus and ovaries combined with a sonohysterography. Assessment of tubal patency was performed either by hysterosalpingogram or sonosalpingography. FSH, LH, thyroid-stimulating hormone (TSH) and estradiol (E₂) levels were determined from blood samples on day 3 of the menstrual cycles, with prolactin (PRL) and progesterone on days 22 and 24 of the same cycles. Laparoscopy was performed in selected patients with endometriosis or unexplained pelvic pain. Semen analysis and culture were systematically performed, and treatment was given for any infection by common bacteria, mycoplasma or chlamydia. Male partners were considered fertile when the total number of motile spermatozoa after capacitation was >10⁶/ml.

The ART laboratory used for the procedure was considered a ‘viral risk’ area, separated from laboratory facilities used for couples negative for HIV, hepatitis B virus (HBV) and hepatitis C virus (HCV). The ART laboratory complied with standard recommended safety precautions (World Health Organization, 1999). Specific precautions were implemented against the risk of HIV, HCV and HBV contamination as recommended by the French decree of 10 May 2001 (Journal Officiel de la République Française, 15 May 2001), and the potentially infected gametes and embryos were handled separately. A special biosafety cabinet workstation was used for all tasks that involved handling sperm, oocytes and embryos.

Semen processing

Semen samples were obtained by masturbation after 4–7 days of sexual abstinence. After liquefaction at room temperature, semen parameters were assessed as outlined by the WHO criteria. Semen analyses were performed, and samples were processed using a 40–80% density gradient (Pureception kit, Sage) to separate motile spermatozoa from non-sperm cells, immotile spermatozoa and seminal plasma. The ejaculate was layered over the gradient and centrifuged at 400 g for 30 min. After centrifugation, the supernatant was removed, and the sperm pellet recovered and resuspended in 3 ml of fresh medium (Sperm-washing medium, Sage). A washing at 400 g for 10 min was performed, and the supernatant was discarded, 1 ml of medium was subsequently gently layered on the pellet, and the tube was incubated at 37°C for 1 h.

After swim-up, a supernatant volume of ∼500 µl was recovered, and an aliquot of this volume (100 µl) was tested for detectable HIV RNA. Nucleic acid extraction and HIV-1 RNA quantification were performed using a real-time PCR assay (NucliSens EasyQ HIV-1 v1.1, bioMérieux sa. Marcy l’Etoile, France). The assay detection limit was 50 RNA copies/ml.

The remaining washed sperm (400 µl) was stored at 4°C for ∼22 h and used for IUI and ICSI/IVF procedures, if the PCR test for HIV was negative.

IUI

Fertile couples were offered IUI with ovulation induction using gonadotrophins. A low dose (50–75 IU) (Gonal F®, Ares-Serono, UK or Puregon®, Organon, France) of recombinant FSH was given from day 3 to the day of ovulation induction (HCG 5000 IU) based on follicle ultrasound monitoring from day 8 to a dominant follicle of 18 mm mean diameter. Insemination was performed 36 h after HCG administration. For the IUI, we use only fresh sperm, with a PCR control of semen sample within 12 h.

IVF/ICSI

Ovarian stimulation protocols were chosen according to clinical data, patients’ age and hormonal profile and the result of any previous stimulation. The standard long protocol was adopted for most IVF cycles. The ‘short protocol’ or ‘flare-up protocol’ was used for older or poor-responder patients. A GnRH analogue (Decapeptyl®, Ipsen, France; or Suprefact®, Aventis Pharma, Germany) was given s.c. each day. Recombinant FSH (Gonal F® or Puregon®) was given together after desensitization. The GnRH analogue was continued up to the day that HCG 5000 IU (Gonadotrophine Chorionique Endo 5000®, Organon) was administered.

Cycles were monitored every other day by serial transvaginal sonography and E₂ assays. Serum LH was measured to improve ovulation induction timing. Oocytes were retrieved under ultrasound guidance by flushing ovarian follicles with ASP medium (Vitrolife), incubated in G-FERT medium (Vitrolife) and subsequently fertilized by conventional IVF or ICSI and cultured in G-1 medium (Vitrolife) until the day of transfer (day 2 or 3). Supernumerary embryos on day 3 were frozen and thawed according to the standard technique involving 1,2-propanediol and sucrose as cryoprotectants (Testart et al., 1986). We have cryopreserved 51 embryos before the Italian act on March 2004, which banned embryo cryopreservation. After March 2004, and in accordance with Italian law, we fertilized three oocytes per cycle only after careful oocyte selection, whereas no embryo selection was allowed.

A sample of PCR-negative frozen semen was obtained in all couples treated by IVF/ICSI. This was used only in cases where fresh semen collected at the time of oocyte retrieval tested positive for HIV after sperm washing.

The status of the female partner was confirmed by HIV antibody testing and viral load measurements in the 2 weeks before and 2–3 weeks after each ART attempt. These tests were repeated 3 and 6 months after treatment and again at delivery. The children born were tested once after birth for the presence of HIV-1 antibodies.

Statistical analysis

Descriptive, parametric and non-parametric statistics were used where appropriate. Skewness was checked before using parametric tests.

A multivariate analysis was performed to assess independent and significant variables associated to IUI outcome after categorical transformation. All statistical analysis was performed using STATA 9 software (Statacorp, TX, USA).

Results

From January 2002 to January 2006, 2011 couples serodiscordant for HIV-positive male partner were counselled for ART at our centre. Seven hundred and forty-one of these couples completed the immuno-virological and fertility triage, met the inclusion criteria and were treated according to our protocols. Three hundred and ninety-four additional couples completed the same workup and signed the informed consent to be allocated on a waiting list. One hundred and ninety-four couples were not eligible according to our criteria. Six hundred and eighty-two couples declared their wishes to undergo ART procedures but either have not yet completed the diagnostic triage or were lost to follow-up. Table I summarizes demographic data and the immuno-virological profile of the male patients of this series.

Efficacy of sperm-washing procedures

The total number of negative sperm-washing procedures was 2871. All procedures were performed on the day of semen collection (2400 procedures in patients undergoing IUI and 283 in patients undergoing IVF/ICSI) In addition, we performed 188 sperm-washing procedures for a prudent cryopreservation of semen samples in couples undergoing IVF/ICSI.

The positive rate after sperm washing (assessed using real-time PCR with a threshold at 500 copies/ml) was 4%. This high positive rate could be due to contamination during RT–PCR analysis. In additional 2% of cases, it was not possible to give a results because the test failed.

Couples treated by IUI

Five hundred and eighty-one (78%) couples underwent superovulation and IUI. The mean age (±SD) was 38 ± 4 for male and 33.9 ± 4.1 for female partners. Basal FSH in women was 6.9 ± 2.9 IU/l. Basal FSH in women who conceived (6.9 ± 1.9) and did not conceive (6.5 ± 2.5) after IUI was not significantly different (P = n.s.). The mean numbers of treatments per couple was 4.13. The results of our IUI programme are summarized in Table II.

Seventy per cent of pregnancies were conceived in the first three cycles. Eighteen (4%) pregnancies were of multiple orders: 14 twins, 2 triplets and 2 quadruplets.

Table III summarizes the clinical outcome of IUI stratified for maternal age, antiretroviral therapy and semen characteristics. A multivariate analysis proved that maternal age was the only significant and independent predictor of IUI success (P = 0.002; CI 0.54–0.87). Semen total motile cell count inseminated was of borderline significance (P = 0.07).

All HIV-1 blood tests performed by PCR assays from the inseminated women were negative 3 months after the last IUI. Women who delivered infants were tested again for HIV-1 after delivery. The results were negative in all cases and for all the children.

Couples treated by IVF/ICSI

One hundred and sixty couples were treated by second-level ART procedures (mean number of cycles per couple 1.8). The mean age (±SD) was 40 (±4) for male and 36 (±4) for female partners. Basal FSH in women who conceived was 7.2 ± 2.2 and in women who did not conceive was 7.4 ± 2.3 (P = n.s.). The results of our IVF/ICSI programme are summarized in Table IV.

Discussion

This article reports the largest single series examining ART outcome in serodiscordant couples with an HIV-1-positive male. According to their fertility profile and the criteria adopted by our centre, 581 couples underwent IUI achieving a pregnancy rate per couple of 78% (19% per IUI cycle), with multiple pregnancy rate being 4%, and 160 couples underwent IVF/ICSI achieving a pregnancy rate per couple of 41% (24% per embryo transfer) with multiple pregnancy rate being 10%. As expected, maternal age and FSH were lower in the IUI group. Male partners were ∼4 years older than female partners in both groups. The semen of all male HIV-infected partners was treated by the sperm-washing procedure (Persico et al., 2006). After a 6-month follow-up, no female partner proved positive for HIV1.

In this experience, pregnancy rate per couple by IUI was higher than the average 45% overall pregnancy rate reviewed on different series by Sauer (2005), and the pregnancy rate per IUI, as well, was higher than the average 14% briefly reported on the European experience by Gilling-Smith (2000). These results could be explained by the routine adoption of ovulation induction with low doses of recombinant FSH and timing of ovulation with recombinant LH according to Marina et al. (1998), the standard usage of fresh sperm after real-time PCR and possibly a good selection of IUI cases, with an average of four attempts per couple. The ovulation induction policy was not a standard of care in those reported series, and the number of attempts was not homogeneously reported by each author. In addition to this, the use of frozen semen (Marina et al., 1998; Ohl et al., 2003; Bujan et al., 2004; Manigart et al., 2006) might have impacted negatively on the number of available motile sperm after freezing, as already reported (Leruez-Ville et al., 2002; Marcus-Braun et al., 2004; Oneta et al., 2004; Desrosiers et al., 2006) and on the pregnancy rate per IUI. Multiple pregnancy rate was 4% with two triplets and two quadruplets. All these couples conceived at their second, third or fourth attempt. These couples were counselled on the use of superovulation and its risks. Yet, they decided to accept these risks and face the possible consequences including multiple pregnancies.

In socioeconomic areas where HIV infection has turned into a chronic illness for the vast majority of patients (Englert et al., 2001), the desire for fathering a child has become a legitimate ethical and medical issue (Klein et al., 2003; Sauer, 2005). All couples seeking medical counselling in our centre were engaged in long-term relationships, and 69% of these were married. Counselling unprotected intercourses on the day of ovulation to fertile couples reduces the risk of horizontal transmission of HIV, but this ‘reduction’ would condemn ∼5% of women to be infected by their partners (Mandelbrot et al., 1997). Highly active antiretroviral therapy could further reduce this risk but does not guarantee an undetectable virospermia. We still do not support the idea that these couples should be allowed to try to conceive naturally, just focusing on the best ovulation window (Barreiro et al., 2004). This is an unacceptable option when considered on an epidemiological scale and not on a single medical practice risk perception. Adoption and heterologous insemination could be an alternative (Bujan et al., 2002). However, in our counselling experience, these options may clash with personal or ethical values. In Italy adoption can be very difficult, and heterologous insemination is not legally acceptable since 2004.

Our findings on safe pregnancy with no sexual transmission of HIV-1 to the female partner after sperm washing add up to >3000 cases so far reported and to the consistent biological results published on the efficiency of sperm washing (Sauer, 2005). According to the experience of our laboratory, this definition should be applied only to procedures where the sperm washing is completed by the swim-up phase of sperm cell selection (Hanabusa et al., 2000; Persico et al., 2006).

The efficiency of IUI and its relatively low cost make this first-level procedure the technique of choice in serodiscordant couples with an HIV-positive male partner, except for women over 40 years of age and when no other infertility problems are involved including borderline number of total motile sperm cells. These findings and recommendations are well in agreement with data and opinions reported, among others, by Nicopoullos and Bujan (Nicopoullos et al., 2004; Bujan et al., 2006).

When the female partner was suffering from infertility factors or the male partner had <1 × 10⁶ total motile cells in the final fraction after sperm washing or both partners had a combination of sub-fertility conditions, we performed IVF/ICSI. The pregnancy rate per embryo transfer was in agreement with both similar smaller series (Ohl et al., 2003) and larger non-HIV series (ESHRE, 2006). Other markers of outcome were as good in these couples treated after sperm washing as in other infertility series of comparable age: fertilization rate was 65% by IVF and 88% by ICSI. IVF/ICSI results are lower in terms of pregnancy rate when compared with those series in which fertile serodiscordant couples underwent ICSI (Garrido et al., 2004). However, pregnancy rate achieved by IUI and IVF/ICS altogether in our series sums up to 70.3% which is higher than the results obtained by ICSI in fertile couples. The problem with ICSI as first-line therapy in all serodiscordant couples is also the high multiple pregnancy rate [14% (Garrido et al., 2004) and 57.1% (Pena et al., 2003)] and the possible obstetrical and neonatal complications associated with these pregnancies. Finally, we should also consider the cost issue determined by ICSI to all serodiscordant couples. In fact, in addition to procedure related costs, we should also consider the possible additional costs determined by prenatal and neonatal care in multiple pregnancies (Oliviennes, 2000; Nakhuda and Sauer, 2005). These costs do not even benefit from a higher pregnancy rate.

The answer to the question posed by Sauer (2005) as to which is the preferred technique, between IUI and IVF/ICSI, is just to be found in the fertility profile of the serodiscordant couple seeking medical advices. The safety issue definitely requires large multicentre observational trials and until then ‘participants need to understand that no procedure is risk free, as all carry a possibility for transmitting infection’.

The persistence of this remaining area of uncertainty on safety is probably the reason to extend prudent testing for HIV-1 by PCR after sperm washing not only to IUI and IVF but also to ICSI until final biological evidence is provided. In our experience on 48 samples (Persico et al., 2006), HIV-1 RNA tested positive in 13% of seminal plasma and 3% non-sperm cell, and HIV-1 DNA was tested positive in 15% of non-sperm cell. So far, we agree with Garrido (2006) and Gilling-Smith (2000) that to protect patient from technical errors during semen washing, viral detection sampling before ART is the method of choice. We could also argue that this procedure accounted for a negligible fraction of costs in ICSI, and this could temporarily settle the discussion on clinical protocols notwithstanding ongoing research on biological issues (Garrido and Meseguer, 2006; Piomboni et al., 2006).

Vernazza (Vernazza et al., 2006) added further advice on how to reduce the risk of transmission beyond this problem of testing semen after sperm washing for ART procedures. It is a minor limitation of his letter that the large series quoted from a private Milan centre has never been presented in detail. In his letter, Vernazza et al. (2006) addressed the issue every centre is facing nowadays: the long waiting lists that tempt some couples to try for a spontaneous conception by unprotected intercourses. In our centre, ∼400 couples are on the waiting list, with an average 6-month delay at the time of writing. In our opinion, although large comprehensive follow-up studies should be undertaken to assess the relative risks of the possible scenarios described by Vernazza, healthcare policymakers should be made aware of the changing paradigm of HIV-infected patients in western countries to allocate resources to preventing HIV transmission within serodiscordant couples which, like other more lucky ones, are just trying to live a possible significant part of human life: childbearing.

Acknowledgements

We acknowledge the role of Giorgio Pardi, Chairman at the DMCO San Paolo Institute where he first encouraged and funded the first researches on sperm washing started in Milan at San Paolo DMCO in the early 1990s, Dr Enrico A. Semprini and Dr Paolo Levi-Setti, for the pioneer-dedicated researches and clinical work throughout a decade at the same Institute.

References