Assisted reproductive technology in HIV serodiscordant couples☆

Background 

As recently as 1990, the Centers for Disease Control (CDC) published recommendations advising against reproductive assistance in HIV-1 serodiscordant couples, specifically when the man was HIV positive. These recommendations were based on noncurrent evidence and served to discourage many US practitioners. In fact, the CDC decision was based on a single case of transmission where a woman was inseminated with semen from her HIV-positive husband, that was simply washed in a buffered solution (4). Current sperm treatment techniques include a multistep process of gradient centrifugation, washing, and a swim-up procedure. In an Italian clinic using this technique, no reported cases of transmission occurred in more than 1,080 inseminations (5). In February 2002, after 12 years of debate, the American Society for Reproductive Medicine revised its guidelines to state that HIV-1 serodiscordant couples may seek reproductive assistance. During this time period, European physicians have performed more than 3,000 inseminations in HIV discordant couples without a single case of seroconversion (Table 1), whereas less than 7% of US centers even offer services to HIV seropositive individuals (6).

Table 1. Summary of World Data on ART of HIV-1 Serodiscordant Couples

	First Author	Cycles	No. Pts	Preg	Births	Ong	Infect
	IUI
	Semprini (5)	1,954	623	272	242	-	0
	Marina (49)	458	233	116	86	20	0
	Tur (50)	155	67	32	-	-	0
	Gilling-Smith (51)	66	27	12	3	3	0
	Vernazza (52)	46	16	5	3	1	0
	Weigel (53)	143	64	19	14	-	0
	Bujan (54)	62	28	14	2	11	0
	Daudin (55)	93	39	18	-	-	0
	Brechard (56)	11		5			0
	Total	2,988	1,097	493	350	35	0
	IVF
	Semprini (5)	70	62	20	-	-	0
	Gilling-Smith (51)	12	7	2	-	-	0
	Weigel (53)	11	6	6	4	-	0
	Total	93	75	28	4	-	0
	IVF-ICSI
	Marina (49)	58	40	27	11	11	0
	Weigel (53)	32	20	11	4	1	0
	Jounnet (57)	97	68	33	22/7	-	0
	Loutradis (58)	2	2	2	2	-	0
	Pena (59)	113	61	35	26	12	0
	Morshedi (60)	12	6	6	3	1	0
	Batzer	2	1	2	1	0	0
	Total	316	198	116	68	26	0
	Total for all methods	3,397	1,370	637	422+	61	0

No. Pts = number of patients; Preg = clinical pregnancy rate (including miscarriage, ectopic, and normal pregnancy); Ong = ongoing, undelivered pregnancies; Infect = infections with HIV-1; seven additional pregnancies occurred with the transfer of frozen embryos. For complete references please visit: www.srmjournal.org.

There are obvious arguments raised against reproductive assistance of HIV-infected individuals. These include concern about viral transmission to the uninfected partner, embryo, or fetus, the future well-being of a child infected from birth with this chronic disease, and the loss of an HIV-positive parent, leaving the child to be raised by others. In addition, fear exists regarding laboratory contamination and danger to medical personnel. Although these concerns obligate us to proceed with caution, they are certainly not absolute contraindications to reproduction. Considering policies in similar long-term debilitating diseases, and a consensus that supports reproductive assistance based on informed decision-making, it seems that the exclusion of the HIV-infected individual may have more to do with the social stigma of this disease than true medical or ethical concern.

Treatment options 

The reproductive strategies to prevent perinatal or partner transmission for HIV discordant couples varies. For HIV-positive women, preconceptual diagnosis and treatment following the current National Institutes of Health (NIH) guidelines to reduce viral loads to fewer than 1,000 or undetectable levels (7), have resulted in no demonstrated perinatal transmissions (3) when appropriate delivery procedures are followed. Timed insemination, rather than unprotected intercourse, prevents male transmission.

For HIV-positive men, the issue is to prevent transmission to the woman with resultant potential perinatal transmission. In spontaneous conception between a seropositive male and a seronegative female, the transmission rate has been suggested as being from 4% to 11% (8). This is in sharp contrast to reported rates of 0 to 0.4% using techniques of sperm washed IUI or IVF–intracytoplasmic sperm injection (ICSI) (8). It should be noted that we can never assure 0% transmission. But with these techniques, clinical experience has so far shown no cases of transmission. This is in marked contrast to the risk reported with spontaneous conception.

Sperm preparation: laboratory technique 

The technique originally proposed by Semprini (9) in Milan, involves removing HIV-infected leukocytes in the semen from motile sperm by gradient centrifugation in combination with “swim-up.” Semen is processed by centrifuging through a three-layer discontinuous Percoll density gradient. After centrifugation of the gradient tubes, the pellet is resuspended in human tubal fluid (HTF), washed and again suspended in HTF. The specimen is then subjected to a swim-up procedure to recover the most motile component for use in ICSI (9).

Semprini has shown that when high titers of HIV are added to HIV-negative semen, testing with a quantitative enzyme-linked immunosorbent assay, demonstrated more than a 99% reduction in cell-free and cell-associated HIV after Percoll gradient separation. More than 99.9% was removed by either Percoll gradient or swim up. By combining these two techniques, the HIV titer in the motile sperm fraction was found to be less than 0.1% by quantitative enzyme-linked immunosorbent assay (9). In addition, the sperm fraction was not demonstrated to be infectious to peripheral blood lymphocyte HIV targets in vitro (10).

With this technique, Semprini et al. (5) have successfully completed more than 2,000 IUIs involving more than 800 women, without a single seroconversion. Semprini's work with Percoll-separated sperm has been reproduced by many other European physicians to date. With IUI using a prepared specimen, the sensitivity of the laboratory assays for detection of HIV virus has been quoted at 200–800 copies/mL (10).

Recently, Politch et al. (11) have reported a new sperm processing device using a double tube system and density gradient. This process is advocated by the researchers, as it tested significantly better in removing HIV-1 from the motile sperm fraction, as detected by reverse transcriptase-polymerase chain reaction (RT-PCR) (P<.01). In addition, the double-tube gradient method produced a significantly higher sperm yield than the traditional method of sperm preparation, and involved less time and technical expertise than the gradient swim-up technique. Obviously, as a new technique, it should be re-evaluated with prospective studies to prove its efficacy before it can be widely accepted in clinical practice.

Although the virus has rarely been directly detected on spermatozoa, even when HIV-RNA RT-PCR is used to screen specimens before IUI, small numbers of virus particles may go undetected according to Sauer and Chang (12). Given this information, it is imperative that couples be informed that although the infection risk may be reduced with ART, it cannot be excluded altogether.

Treatment of male seropositive discordant couples 

It is unclear whether IUI with processed semen is any less safe than the more expensive IVF-ICSI for male seropositive discordant couples. Arguments have been raised in favor of each method. First, it would seem that IVF-ICSI would be the preferable method given that most research has confirmed the existence of HIV-1 in association with cell-free seminal plasma rather than germ cells or viable spermatoza 13, 14. The IVF-ICSI provides a way to deliberately isolate and use the population of cells least likely to carry the virus, involving exposure from a single sperm cell rather than millions of sperm, as in IUI (12). In addition, ICSI pregnancy rates per cycle in HIV-1 discordant couples are approximately twice that of IUI rates (39.5 vs. 18.2) (13). However, as Sauer and Chang have pointed out, there is no test that exists to assess the status of a single sperm selected for ICSI (12). Because of the problem with laboratory testing at this level, the conservative approach mandating antiretroviral application during the infertility treatment days for both partners is perhaps warranted even when the viral load is undetectable (15).

In the United States, there has been resistance to IUI for assisting the reproduction of HIV seropositive males. In fact, in several states it is considered a criminal act (16). Arguments waged against the use of IVF-ICSI point out that there have been rare reports of HIV-1 virons or DNA in association with spermatozoa (17). The obvious concern is that the spermatozoa would then infect the oocyte and subsequent embryo. Politch and Anderson (13) suggest that IUI may be preferable to IVF-ICSI given that it does not bypass two important built-in mechanisms that potentially protect against sperm-borne pathogens and DNA. The first is the acrosome reaction, which involves shedding of the sperm outer membrane and any attached material. The second is penetration through the zona pellucida (ZP), which theoretically traps infectious particles and exogenous DNA.

Concerns have also been raised over the increased risk of birth defects and developmental defects associated with IVF-ICSI pregnancies. Hanson et al (18) found that children conceived with ICSI have twice the risk of major birth defects, in addition to increased risk of Y-linked and de novo sex chromosome abnormalities. However, as ISCI is usually used for male-factor infertility, it should be noted that these statistics might not be relevant to HIV-positive men who do not have the same issues with sperm quality.

Another clear advantage of IUI over ICSI is the high cost of ICSI cycles, an important issue, given that HIV is more prevalent among disadvantaged populations.

Treatment of female seropositive discordant couples 

The argument for treatment of HIV-positive women is not as clearcut. First, these patients should be adequately counseled on all of their options including the risks and benefits of IVF- ICSI, IUI, donor sperm, adoption, and even the use of a willing surrogate carrier. When the female partner is HIV positive, the current risk of vertical transmission, quoted by the International Perinatal HIV Group, is approximately 2% (19). Although this risk is not negligible, it is still low and may not be grounds for exclusion from active intervention by ART practitioners (12). In fact, newer data have shown that the level of plasma HIV-1 RNA largely determines the risk of perinatal transmission. Women with undetectable viral loads, or those with less than 1,000 copies per milliliter have close to a 0% transmission risk 7, 20. With the exception of adoption, virtually all reproductive options for HIV-positive couples involve some risk of vertical transmission.

As previously noted, the risk for vertical transmission for these women is lower than the general risk of serious congenital abnormalities in any given pregnancy, a risk that is accepted by reproductive practitioners throughout the country. Considering that fertility services are commonly offered to patients with cancer, women past menopause, and even those with complex medical illnesses, such as cystic fibrosis, cardiac disease, many autoimmune disorders, or insulin-requiring diabetes, it seems discriminatory that HIV-positive women with such a low chance of transmission should be excluded from the safer option that these services provide.

The unfortunate stigma of HIV was effectively demonstrated by Rozenberg et al. (21) who conducted a survey to assess acceptance of IVF treatment in women with HIV infection. The survey was sent to 3,450 Belgian physicians with 1,175 responses. Each physician was sent one case study. The first involved a 30-year-old HIV-positive, medically stable woman with normal immunological status and an undetectable viral load, who was on antiviral therapy and had an HIV-negative partner. It was clearly stated that she had a good chance of normal life expectancy, with a vertical transmission rate of less than 2% if she remained on her antiviral therapy.

The second case involved a woman in an identical scenario except that she was infected with hepatitis C instead of HIV.

The third case involved a patient with chronic active hepatitis C. Her liver enzymes were elevated despite viral treatment, and it was indicated that her life expectancy might be reduced. The vertical transmission rate was estimated to be 7%.

The survey results showed that a greater percentage of physicians were in favor of assisting hepatitis C patients than HIV patients. Of the physicians 40.2%–59.8% were in favor of IVF for the HIV patient versus 73.3%–86% in the hepatitis C patients with the same clinical situation and a transmission rate more than three times that of the HIV patient. There is no medical reason to explain the discrepancy in response to these very different clinical scenarios. Despite improvements in life expectancy, treatment, and vertical transmission, the investigators concluded that “the word HIV has a much stronger (negative) impact than the word hepatitis C” (21).

Ethical issues 

The social stigma of HIV has long prevented it from being approached as “just another” infectious disease. Opponents have argued that by assisting HIV serodiscordant couples with reproduction, the physician assumes the risk of complicity in transmitting the virus to the uninfected partner and possibly even to the embryo or child. Although there is a small, undefined risk of viral transmission, we have demonstrated that reproductive assistance using insemination techniques including ART can be provided safely. When opponents to assistance to HIV-infected individuals claim physician complicity, they show a clear disregard for patient autonomy. Given appropriate counseling, if a patient and his or her partner make the informed decision to accept the small risk of possible infection, and provided they make provisions for the care of the child, then the physician should not necessarily be culpable. In fact, by excluding these patients from reproductive assistance, we force them to make the difficult decision of forgoing the protection of condoms and attempting procreation by intercourse, thereby assuming an unacceptably high risk of transmission or, alternatively, abstaining altogether and forgoing parenthood.

It should be noted that this argument applies specifically to those individuals who desire their own genetic offspring. Obviously donor insemination and adoption are viable options that must be thoroughly explored, especially in those couples who desire the safest methods of conception and are not comfortable even with the low, undefined risk of transmission. Regarding the motivations and desires of HIV-1 serodiscordant couples, Klein et al. (37) noted as many as 50% of 31 highly informed couples claimed that they would not consider artificial insemination with donor sperm as an alternative means of reproduction. In addition, approximately 9% of the individuals surveyed claimed that they would engage in unprotected intercourse to achieve pregnancy if no safer alternative was available. These issues of childbearing are paramount, considering that women between the ages of 13 and 24 years represent the majority of women newly diagnosed with AIDS (38).

Additional concerns have been raised regarding the concept of nonmalfeasance. Are physicians doing harm by assisting in the conception of a child who may contract HIV, a disease that is at best, chronic, or potentially debilitating and even fatal? What about the possibility of losing one or both parents prematurely to this disease? In both instances, the concern of nonmalfeasance is not particularly valid when informed counseling has taken place and the parents are willing to care for the child or make arrangements regardless of medical outcome. If the practitioner has been vigilant in taking steps to ensure the lowest possible risk of viral transmission at conception and at birth, this should not be considered malfeasance.

Analogies have recently been made between HIV and other chronic disease states in which reproductive assistance has not met with such resistance 1, 38. A recent publication by the Ethics Committee of the American Society for Reproductive Medicine eloquently challenges this bias by comparing the ethical issues surrounding HIV to those in couples who are known carriers of certain chronic or equally devastating autosomal recessive diseases (1). On the basis of their argument, HIV should be considered no different than diseases such as Tay-Sachs, sickle-cell anemia, or cystic fibrosis, which carry transmission rates of 25%, regardless of medical involvement (unless PGD is used). Shouldn't physicians who find it ethically justifiable to aid in the reproduction of these individuals also be willing to assist HIV-positive individuals who are seeking a responsible and safe means of reproduction with a much lower risk of transmission.

Other comparisons have been made with the reproductive assistance of women of advanced maternal age and their increased risk of irreversible genetic disease. As illustrated by Lyerly and Anderson (38), when a woman of advanced maternal age seeks infertility treatment and plans to decline prenatal testing because she is committed to caring for any child she conceives, regardless of genetic outcome, she is never excluded from care based on concern for a potentially disabled child. It is equally unjustifiable that practitioners consider a 2% or lower risk of vertical transmission of HIV cause to exclude these patients from treatment. Rojansky and Schenker (39) also discuss risk factors for transmission of disease

Concerns related to premature loss of a parent are relevant. New research has shown that deaths resulting from AIDS as the primary cause have been on a steady decline during recent years in the United States. Instead, these patients, like their HIV-negative counterparts, are dying from heart disease, liver problems, cancer, and other illnesses unrelated to their HIV infections (40). As previously mentioned, many reproductive clinics also offer treatment to women with cancer and other advanced medical diseases. A recent survey by Stern et al. (41) found that up to 79% of ART clinics offer treatment to single women. In either case, concerns for potential single parenting are not routinely raised.

It should be made clear, however, that although these concerns should not serve as barriers to treatment, they are important, and should be routinely addressed when counseling serodiscordant couples.

Other arguments against therapy have centered on occupational exposure and cross-contamination within the laboratory. Although these are valid concerns, they are argument for continued maintenance of strict universal precautions, and possibly the use of a separate laboratory area, but not sufficient reason to deny access to care. In 1998, the United States Supreme Court ruled that individuals with AIDS are to be protected under the Americans with Disabilities Act (ADA), entitling them to medical services without discrimination. Since the decision on Bragdon v Abbott in 1990, a case that involved denial of dental services to a woman with asymptomatic HIV infection, both HIV infection and infertility have been classified as disabilities (42). Hence, the only lawful manner in which a physician can withhold treatment from individuals with disabilities under the ADA is when the provision of services poses “a direct threat to the health and safety of others” (43). Although some argue that the provision of ART to HIV-infected individuals might qualify as a “direct threat,” there is no objective scientific evidence existing to prove this point. There is no data that viral particles are transmitted under a laboratory hood or between cultured tissues of embryos in separate containers. Similar concerns regarding hepatitis B and C and other viral organisms have reinforced implementation of universal precautions within laboratories (appendix).

The infectious risk for a laboratory worker is approximately 10 times that of the general public, and almost three times that of other hospital employees (44). For this reason, strict adherence to universal precautions is essential in the treatment of all body fluids including semen, sperm, follicular fluid, and oocytes.

Though rare cases of nonsocomial transmission of HIV have been reported (45) no reported cases of HIV transmission to partner or fetus exist with the current assisted reproductive techniques.

The occurrence of viral cross-contamination during liquid nitrogen storage of biological material has been described by Tedder et al. (46). The specific incident was related to leakage of a cryopreservation bag containing bone marrow from a patient acutely infected with hepatitis B virus. This led to contamination of the entire tank and transmission to six patients from subsequent transplantation. Although the specimens in this case were bone marrow and peripheral blood stem cells, and the Hemofreeze bags used for storage are no longer used in such cases (46), there is potential for the same problem to arise with preservation of semen, oocytes, and embryos. There have been no reported cases of cross-contamination occurring with HIV when appropriate precautions are taken. Nevertheless, it is advisable to use such simple procedures as using separate storage and laboratory space for the embryos of viral positive-infected patients.

Conclusions 

The most rapidly accelerating route of transmission of AIDS among women is heterosexual intercourse. Moreover, perinatal transmission is the primary source of HIV infection among infants in the United States (47). The use of insemination techniques and the application of ART with appropriate testing, drug therapy, and implementation of universal precautions, should be applied to prevent viral transmission of all kinds.

In addition to the limited availability of treatment, the high cost of reproductive services also serves as an impediment for HIV-1 serodiscordant couples in the United States (15). The lower cost of procedures like IUI with washed sperm compared to IVF-ICSI need to be considered. At this time there is no evidence as to which procedure is safer. According to Sauer and Chang (12), who published the largest single series of IVF-ICSI cases among HIV-1 serodiscordant couples, IVF-ICSI is theoretically the safer procedure. However, this trial involves only 113 IVF cycles. Although the cumulative data are much more impressive (see Table 1), more experience is needed to ensure the safety and cost-effectiveness of this method. Although the European IUI reports are promising, according to the CDC, follow-up of treated couples is incomplete and therefore we cannot exclude the possibility of HIV infection in women and children lost to follow-up (48).

The ultimate goal of this type of therapy is not only to implement procedures that are effective in the United States but on a larger scale to develop cost-effective protocols that can potentially be replicated in community health centers of developing countries, which house greater than 90% of the world's HIV-infected population.

Ethical concerns must be addressed, including counseling before treatment and informed consent. Issues of autonomy, beneficence, and nonmalfeasance must be evaluated individually. In addition, strict protocols should be developed to decrease the risk of transmission of other infections that immunosuppressed HIV-1 infected men may carry in their semen. These include cytomegalovirus, herpes simplex virus, and hepatitis C virus (15).

The option of adoption as an alternative should be strongly encouraged. However, there have been numerous articles advocating this idea and not one has commented on the sort of discrimination and hardship that HIV-positive individuals might face when attempting to adopt.

As laboratory contamination appears to be preventable, why not the risk of “social contamination?” Reproduction in HIV discordant couples deserves to be approached in a manner similar to treating couples with other chronic medical disease. The battle against the AIDS epidemic in the United States should be not only scientific, but social.