Directions for the Future
Much of the workshop was devoted to the question of what is known about the potential risks of oocyte donation. Those discussions were generally carried out with an eye to two other questions: What is still not known about the potential risks of oocyte donation (and how can one learn what one needs to know)? And how can the potential risks of oocyte donation best be minimized?
This chapter describes the discussions that centered on the latter two questions.
THE NEED FOR MORE AND BETTER DATA
One of the most striking facts about in vitro fertilization (IVF), Dr. Giudice commented, is just how little is known for sure about the long-term health outcomes for the women—and men—who undergo the procedures. Although more than a million IVF cycles have been performed in the United States over the past 20 years, and although there are registries that keep track of the various reproductive outcomes, such as the number of eggs retrieved and the number of children born, there are no registries that track the health of the people who have taken part. Without such registries to draw from, most of the studies of the health outcomes of IVF have been anecdotal or have focused on relatively small groups of people. Furthermore, Dr. Giudice noted, the studies vary quite a lot in terms of study design, the number of subjects, and outcome, so it is impossible to draw a consistent picture from them.
The situation is complicated by the fact that the available studies are not directly applicable to the question at hand—the safety of oocyte do-
nation for research. For one thing, Dr. Giudice pointed out, the available data come primarily from IVF patients and not from healthy subjects, yet it is healthy women and not those coping with infertility who will be donating eggs for research. This raises the possibility, for example, that the existing data will overstate the potential risks for healthy donors, given that IVF patients may be more likely to have a variety of conditions, such as pelvic adhesions and polycystic ovary syndrome, that increase the odds of complications from the ovarian stimulation or the retrieval surgery.
In addition, the available data come primarily from Caucasian women in middle to upper socioeconomic groups, because they are the women most likely to be able to pay for IVF on their own. Since fertility treatment is generally not covered, or not covered fully, by medical insurance, women in lower economic brackets are less often able to afford such treatments and so make up a relatively small percentage of women in IVF programs. But the pool of research donors is likely to be significantly broader than just Caucasian women in middle to upper socioeconomic groups, and it is difficult to infer just what potential risks these research donors may face when the only available data are from a collection of women who differ from them in age, race, and socioeconomic status.
One other complicating factor is that the potential risks from hormone therapy, from surgery, and from anesthesia seem to have been changing over the past 20 years. “Many of these risks,” Dr. Giudice said, “seem to have been greater early in the process of in vitro fertilization than they are currently.” The reason would seem to lie in the increasing experience that reproductive specialists have been accumulating, she said. By doing procedures over and over again, doctors hone their skills and learn to avoid certain mistakes, leading to a decline in potential risk. This decline is good news, of course, but it adds to the uncertainty about exactly what potential risks egg donors face now.
The bottom line is that there is a great deal of uncertainty about the potential risks of oocyte donation for research. David Guzick, dean of the University of Rochester School of Medicine and Dentistry, made this point when discussing the future fertility of egg donors:
“What we know about future fertility in connection with oocyte donation is really only by inference,” he said. “What were presented [at the workshop] were data on general IVF patients and a much smaller amount of data on donors. We learned that the incidence of infection, the incidence of adhesions, and the incidence of general surgical problems is
low. We inferred from that, therefore, that the likelihood that there should be problems with fertility is low, but we don’t really know that. We don’t really have data to tell us, if these individuals who donated their eggs were followed, how their fertility would compare with a matched control group.
“We also know a lot about the biology of follicle selection,” he continued, “and we know about the physiology of administering exogenous gonadotropins and the fact that, in repeated stimulation cycles, there does not appear to be a reduction in the number of eggs that are produced. And we might infer from that, therefore, that fertility may not be compromised in the future, but we don’t really know that. We don’t have data on individuals—healthy research subjects—who have undergone repeated stimulation cycles, and we don’t have any data on what the future fertility of these individuals might be.”
Similarly, he said, based on what is known about the biology of follicles over time, we do not think that even repeated donations will cause a woman to have an earlier menopause, but again there are no data that tell us that for sure. We don’t really know that.
The only way to completely resolve these issues, Dr. Guzick said, is to follow a cohort of oocyte donors and observe what happens to them over time—to monitor their fertility over the years and compare it with a control population that did not donate their eggs. “And I think until we know that, we won’t truly be reassured about future fertility,” he said.
More generally, Dr. Giudice said, it is important to accumulate health data over the years for all women whose eggs are harvested for various purposes and to monitor them for long-term effects. “Almost every speaker addressed the issue of some type of database,” she observed.
With more data—and more complete data—it will be possible to quantify the various potential risks of oocyte donation much better than can be done today and therefore to put numbers to the various potential risks that a potential donor faces. Doctors and medical researchers should be able to offer concrete answers to some questions: Does ovarian stimulation increase a woman’s lifetime risk of uterine cancer? What effect does a history of pelvic inflammatory disease have on a woman’s risk factors for retrieval surgery? A more complete database will also allow researchers to tease out the answers to other questions: What effect does having had children have on the risks of oocyte donation? Is there any reason to prefer one age range over another among women who are donating oocytes for research purposes? These are the sorts of questions
that are impossible to answer well without a great deal of data accumulated in a very deliberate and consistent way.
MINIMIZING POTENTIAL RISKS
Nearly all of the speakers cautioned against relying on probabilities because the most important strategy in collecting oocytes for stem cell research is to be cautious in relying on probabilities, because the most important strategy to minimize the potential risks to oocyte donors is to make decisions based on common sense on a case-by-case basis. Of course, physicians try not to subject any of their patients to unnecessary risks, but because research donors represent a special situation—women who are undergoing a procedure not for their own benefit but for the benefit of others—the workshop participants said that even greater care should be taken to make sure that these donors do not pay for their altruism with their own health.
“We absolutely want to minimize risk for our reproductive donors,” said Dr. Marcelle Cedars. “For example,” she said, “there might be risks in terms of a difficult position of an ovary or getting every last follicle that I might take for a patient that I wouldn’t take for a reproductive donor. And perhaps we should go even one step beyond that in terms of our donors for research.”
There was some discussion as to just how far to go to minimize potential risks, particularly with regard to the issue of excluding particular donors with certain risk factors. “I notice that we’ve all agreed we should be conservative,” said Kurt Barnhart, director of the Center for Clinical Research on Women’s Health at the University of Pennsylvania, “but we’ve all danced around the issue by not offering specifics on what should be excluded and what shouldn’t.” The issue, he said, is finding a balance—to minimize potential risk but “not exclude everybody who might want to participate. We need to be cautious, of course, to minimize potential risk, but we don’t want to eliminate something without evidence for eliminating it.”
Zev Rosenwaks had a somewhat different take. “The big difference between research donors and IVF patients,” he said, “is that whenever possible, if you identify any irregularity, whether it be infectious, anatomic, or otherwise, a research donor should be excluded. We have to be careful to think about statistics, but at the end of the day common sense in terms of potential complications should rule.”
Wherever the bar is set, everyone agreed that a great deal of effort should be taken to minimize potential risk for women who are donating their eggs for research. Therefore, much discussion at the workshop was devoted to how such potential risk can best be minimized.
There are two basic ways to minimize potential risk to egg donors. The first is to identify which potential donors have particular risk factors and to exclude them from the donor pool, since they have a higher than normal risk for complications. This is exactly how Cornell handles its egg donors, Dr. Rosenwaks said. “We eliminate patients with endometriosis, a history of PID [pelvic inflammatory disease], previous pelvic surgery, irregular menstrual bleeding, PCOS [polycystic ovarian syndrome], uterine myomas, familial thrombophilia, ovarian tumors, and any other medical condition that we feel may be a problem in terms of the stimulation. We feel that with stem cell research, we should have exactly the same criteria. We’re not dealing with IVF patients. We’re dealing with patients that are donating either altruistically or maybe for minimum pay. But at the end of the day, it is our responsibility to make sure that safety is paramount.”
In determining which potential donors to accept and which to exclude, Dr. Giudice said, the importance of a thorough medical history cannot be overstated. “It’s not enough just to ask about menstrual cycles,” she said. “You really need to nail down how regular they are.” And even things that might seem unrelated to oocyte donation—such as a pituitary tumor—can end up playing an important role. There have been rare cases, Dr. Giudice said, in which a woman with a pituitary tumor took gonadotropin-releasing hormone (GnRH) agonists as part of the hormone therapy and, as a result, suffered pituitary apoplexy.
Besides getting a thorough medical history, Dr. Giudice said, doctors should also use their diagnostic tools to identify potential risk factors. Ultrasound is particularly useful, she said, “in terms of assessing the pelvis for uterine fibroids, for possible endometriomas, for possible ovarian tumors, and also for the occasional malplaced ovary that may be in a place that may put the patient at risk for some of the surgical risks that Dr. Murphy discussed.”
Another potential way to screen research donors, Dr. Giudice said, is by age. “What is the optimal age group? Is it the reproductive age span, 18 to 50? Is it the ovum donor population currently used for reproduction, 21 to 34? This is something that we just don’t have any information on, but I think we are obliged to define that at some point for the safety of our donors.”
Finally, she said, it is possible that the exploding growth of knowledge in genetics and genomics could eventually help doctors pinpoint which women will be at greatest or the least risk from oocyte donation. “My hope is that over the next ten years we’ll have some information that will give us a bit more wisdom in terms of choosing our donors, not only for egg donation, but for other clinical trials as well.”
The downside of being careful to exclude any potential donors that may be at higher risk from the procedure is that it greatly reduces the donor pool, Dr. Rosenwaks noted. At Cornell, for example, potential donors are screened in a wide variety of areas: “The donors see the psychologist, they see a genetic counselor, they see the physician, they go through a multiphasic personality test, and so on and so forth.” The result is that over an 8-year period, of the 1,600 potential donors at Cornell who returned their questionnaires—and not counting the women who had called but never returned the questionnaires—only about 200 patients actually came in to donate. Only one out of eight women who had been interested enough to contact the program, get a questionnaire, fill it out, and return it actually ended up donating eggs.
This exclusionary approach can be applied to lessen any of the major potential risks from oocyte donation. In the case of potential surgical risks, for example, there are several factors that put women at higher risk from retrieval surgery, Dr. Murphy said. “I would not use someone who’s at risk for complications, such as those with endometriosis and their increased risk of adhesions and endometriomas, nor would I probably use those that have had previous infectious disease.”
Anesthesia, Dr. Barnhart said, has “the most identifiable and quantifiable differences” in risk thanks to the ASA (American Society of Anesthesiologists) classification. Only donors in the lowest risk category for anesthetic risk should be allowed to go through the retrieval surgery.
It should also be possible to screen women on the basis of risks to future fertility, said Dr. Nicholas Cataldo. One approach would be to look to older women who are sure that they have completed their families. But, he noted, it would be important to examine the relationship between donor age and research outcomes. For example, does a 32-year-old egg work as well for somatic cell nuclear transfer as a 22-year-old egg? It would also make sense to screen women for factors, such as a history of pelvic inflammatory disease or endometriosis, that would put them at a higher risk for fertility problems exacerbated by the retrieval surgery.
The second major approach to minimizing the potential risk of oocyte donation focuses on the process itself and asks what modifications
can be made to that process to make it less risky for the women who take part. This approach is particularly useful when it is done on a patient-by-patient basis, modifying the different procedures to take into account the particular medical characteristics of a donor. In theory, this technique can be applied to any of the potential risks of oocyte donation, but the workshop participants focused on its application to one potential risk in particular: the development of ovarian hyperstimulation syndrome caused by hormones used to stimulate the ovaries to produce more eggs.
PREVENTING OVARIAN HYPERSTIMULATION SYNDROME
Of all the risks facing women undergoing in vitro fertilization, the most common and the most threatening is ovarian hyperstimulation syndrome (OHSS). As described in Chapter 2, studies have found that a large percentage of women undergoing ovarian stimulation experience symptoms of OHSS ranging from mild to severe, and thus women donating their eggs for research could be expected to face similar complications.
According to several of the speakers at the workshop, it should be possible to prevent many cases of OHSS, including all or almost all of the most severe cases. That prevention will require a combination of the two basic risk minimization strategies: identifying and excluding from treatment those women most at risk and, for those women who do undergo ovarian stimulation, modifying the treatment according to the characteristics of the individual patient.
For some women, Dr. Cedars said, the risk of OHSS is just too great to allow them to be research donors. “I would recommend exclusion of women with polycystic ovarian syndrome,” she said, “because I believe their response, even with careful monitoring, is quite difficult to predict and control.” She would also exclude women who don’t have full-blown PCOS but who have polycystic-like ovaries according to the ultrasound pictures and also women with irregular menstrual cycles. Some doctors, she noted, would even exclude women if they have elevated levels of androgens of luteinizing hormone, but she is comfortable leaving these in the pool if they are otherwise asymptomatic and have a normal-appearing ovary on ultrasound.
Dr. Rosenwaks said he follows a similar protocol in his center. As donors, he excludes not only women with the classic polycystic ovarian
syndrome but also patients who exhibit polycystic ovaries on ultrasound but no biochemical changes—normal follicle-stimulating hormone and luteinizing hormone and normal menstrual cycles.
With the most highly at-risk women excluded, Dr. Cedars said, the second step is to tailor the stimulation protocol to the individual donors with the goal of avoiding ovarian hyperstimulation in each. So it is important to understand just what it is that triggers OHSS.
As Dr. Rosenwaks explained, the development of ovarian hyperstimulation syndrome depends on a large number of follicles in the ovary being exposed to human chorionic gonadotropin (hCG), which is used as a surrogate for luteinizing hormone in order to induce the follicles to ovulate. So there are at least two approaches that can be taken to avoid OHSS: controlling the number of follicles that develop in the ovary and modifying their exposure to hCG.
“For a young fertile donor who might have 20-plus follicles,” Dr. Cedars said, “you really don’t want all 20 of those follicles. I think what you shoot for is maybe 10 to 15.” To do this, she explained, the doctor uses ultrasound to examine the ovary and count the number of antral follicles before the start of the hormone treatment. The doctor then uses this information along with the patient’s age and weight to determine a starting dose.
“The main factor that goes into determining this initial start dose,” she said, “is the antral follicle count, because, remember, with the most aggressive stimulation we’re going to get plus or minus two of that number. So if we have a patient with very high number of antral follicles, we don’t want all those follicles to develop. We’re going to decrease the dose.” Over the course of the treatment, the physician continues to monitor the patient’s progress and will further decrease the dose if too many follicles are developing or if the estradiol levels are too high.
And if for some reason, after a week and a half of hormone treatment, too many follicles have developed, OHSS can still be controlled by manipulating the dose of hCG used to induce ovulation, Dr. Cedars said. “If you don’t give hCG, you will not get hyperstimulation.”
At Cornell, Dr. Rosenwaks said, of 841 egg donor cycles started over the 14-year period from 1992 through 2005, 20 were canceled because of the risk of hyperstimulation at the stage at which hCG would normally have been applied. “We did not take them to retrieval.” Instead the eggs were consigned to atresia, the reabsorption back into the body. “You withhold hCG, you do not get hyperstimulation,” he said, echoing Dr. Cedars. And, indeed, of the 800-plus egg donor cycles at Cornell from
1992 through 2005, there was not a single case of severe OHSS, Dr. Rosenwaks said.
There are other ways to modify the hCG ovulation trigger without completely cutting it out, Dr. Cedars said. “One is to decrease the dose of hCG, because part of the reason that you get that hyperstimulation three to seven days after the injection of hCG is because of the long half-life of hCG. So if you decrease the initial dose, you decrease the time in which hCG levels are circulating and high enough to cause the continued stimulation to the ovary, and there’s some evidence to suggest you can decrease the occurrence of hyperstimulation.”
It is also possible, she noted, to use recombinant LH, which has the same effect as hCG of inducing ovulation, but it has a shorter half-life and thus does not stay in the body as long. There is some preliminary evidence suggesting that this technique decreases the occurrence and the duration of ovarian hyperstimulation.
As doctors and medical researchers learn more about the way that these hormones work in the body, more options will undoubtedly open up as well, but today we already have the capability of avoiding severe OHSS almost completely, Dr. Rosenwaks concluded. “With careful donor selection, individualization of stimulation protocols, careful monitoring, and utilizing appropriate preventive measures, severe OHSS can be virtually eliminated,” he said.
ALTERNATIVE SOURCES FOR OOCYTES
Given that there is always going to be some potential risk to egg donation, it makes sense to look for alternate sources of eggs—sources that do not rely on the traditional process of ovary stimulation and surgical retrieval that was developed for in vitro fertilization. The workshop participants discussed several of these alternatives.
One possibility, as Dr. Guzick pointed out, is to take advantage of an existing resource—that is, couples who have undergone IVF and who may have embryos they don’t wish to use themselves and which might be available for donation. At this point, no one really knows how many such embryos there are, but it should be worth looking into.
Catherine Racowsky, associate professor of obstetrics, gynecology, and reproductive biology at Harvard Medical School, brought up a second alternative. “In a typical IVF cycle,” she said, “only about 80 percent of the eggs are mature. Of the remaining 20 percent, some of them are
completely immature, and some of them are partially mature.” At this point, in a typical IVF labs, those immature and partially mature eggs are discarded, but researchers are working on ways to mature such eggs in vitro, so that they can be recovered and used. If these techniques can be perfected, this would be a way to increase the supply of eggs without the need for more donors.
On a related note, Dr. Cataldo pointed out that techniques are being developed to mature eggs in vitro after only a very brief exposure to hormones, primarily hCG. This could greatly increase the potential donor pool. For example, in women who have a number of small antral follicles, it might be possible to retrieve a significant number of oocytes without putting the woman through the usual ovarian stimulation. Clinical studies have already shown that these oocytes can be successfully fertilized after being matured in vitro, Dr. Cataldo said. So women who might otherwise be excluded from donating their eggs for research—such as women with polycystic-appearing ovaries—could in this way provide eggs without the potential risks that would accompany such a donation done via the usual path.
Finally, there was some discussion about the possibility of retrieving oocytes from cadavers in much the same way that organs are now retrieved from the bodies of people who have signed organ donation cards. As Dr. Giudice put it, in addition to donating your organs to science you might want to donate your gametes.
For that to become a reality, it would be necessary to be able to store oocytes from cadavers in such a way that they remain viable until they can be used. At this time, Dr. Racowsky said, medical researchers are working to perfect the technique of oocyte freezing. “Some programs are having really quite good success rates with egg freezing now,” she said, “but it’s not universally the case. With a little bit more experience and technological advances, hopefully in the near future we’ll have that also as a useful tool to be able to store very valuable material for this work.” If so, it should open up one more alternative source of oocytes.
For now and for at least the near future, however, the major source of oocytes for research is likely to remain eggs donated by women specifically for use in research.