Diagnoses in Search of Treatments

New surgical procedures and stimulated IUI demonstrate how the existence of treatment procedures—and the search for indications to use them—can shape what doctors recommend to patients with a range of differing fertility problems. Another pattern prominent in fertility medicine can similarly result in patients receiving inappropriate and risky fertility interventions. In these cases, clinicians have a diagnosis but no corresponding therapy. A problem is named; then comes unscientific experimentation with treatments.

Luteal Phase Deficiency

I have had the feeling for some time that luteal insufficiency as a recurring phenomenon in a given woman is often overdiagnosed. However, the definitive study has still not been performed.
Jaffe 1993, 484

Many couples sitting in their doctor's waiting room have already been initiated into one fertility ritual. Every morning, as soon as the woman wakes up, she reaches to the nightstand and places a special thermometer under her tongue for five minutes. She records this early morning temperature with a dot for each day of her menstrual cycle on graphed paper known as a basal body temperature (BBT) chart. So many women have spent so many mornings taking their temperature before doing anything else that the activity has become a not very private joke among fertility patients. The woman watches for a mid-cycle temperature rise indicating ovulation has occurred. If the temperature drops back before fourteen or so days, marking the end of that monthly cycle (and usually the start of her period), this woman is likely to hear a diagnosis of "luteal phase deficiency" or "defect" (LPD). Her doctor may then prescribe treatment with hormonal medication to correct this condition.

The luteal phase is the time between ovulation and the start of a woman's menstrual period, often thought of as the last half of her menstrual cycle. Luteal phase deficiency is a menstrual cycle abnormality generally defined as inadequate production of the hormone progesterone by the corpus luteum (the ovarian follicle from which a mature egg has just been released) following ovulation. If there is not enough progesterone and/or if production of this hormone does

not continue over enough days, the uterine lining will not receive the hormonal stimulation needed to allow implantation of an embryo. Patients diagnosed with luteal phase deficiency need to understand two basic facts about this condition: first, the diagnosis itself is uncertain, a focus of controversy among fertility specialists; and second, this uncertain diagnosis leads to treatments that are similarly controversial.

To judge by the arguments and questions among physicians, LPD might best be considered an infertility diagnosis in name only. Some specialists dispute the existence of this fertility problem altogether, placing women so labeled within the category of unexplained infertility. Although critics might grant that a condition dubbed LPD does exist in some women, they question whether this condition lowers a couple's fertility. To be defined as a factor contributing to subfertility, the condition must be consistent and repetitive, with an impact on the woman's chance for pregnancy. Instead, LPD is sporadic throughout women's reproductive life, a condition found in normally fertile women at the same rate as in the infertile, a rate expected by chance alone. As one study indicates, LPD is often observed to coincide with common events in a woman's life—including the first years of menstruation, times of stress or of very strenuous exercise, the postpartum period, the last years of menstruation.^[36] Patients need to know that critics within the profession are asking if the diagnosis and treatment of luteal phase deficiency improves the likelihood of pregnancy. So far, the answer appears to be no.

As with the elephant described by the blind men, a major problem is the absence of ways to identify accurately this entity called luteal phase deficiency. All three diagnostic tools that physicians use have serious flaws. The first and easiest tool is the basal body temperature (BBT) chart indicating whether ovulation has occurred and, if so, the length of the luteal phase. However, this daily record of a woman's morning temperature, so well known among fertility patients (see Figure 2), does not correlate well with other diagnostic techniques and does not distinguish patients ultimately thought to have a luteal deficiency.

The second diagnostic tool is a blood test measuring the woman's progesterone level during the luteal phase of her menstrual cycle in an attempt to determine whether she is producing enough of this hormone to support an embryo. Physicians do not know, however, on which day, or days, to draw a blood sample in order to gain information most useful for treating that woman. Further complicating blood tests, no matter the day, are the "frequent and rather wide fluctuations in serum [progesterone] levels over 24 hours."^[37] Because this hormone is released into the bloodstream in an intermittent pulsatile pattern,

Figure 2
Basal body temperature chart (BBT). This sample chart shows a typical
"normal" pattern of body temperature taken before rising in the morning.
The mid-cycle rise and plateau (luteal phase) indicate that ovulation occurred.
The temperature then drops if no pregnancy occurs; the next cycle begins
with the first day of menstruation. If there is no sustained temperature rise,
ovulation probably has not occurred. Fewer than 11–12 days of elevated
temperature may be interpreted as a sign of luteal phase defect (LPD).
However, basal body temperature can also be affected
by such factors as insomnia or illness.

serum levels will naturally vary at different times of day and night. The level of progesterone circulating in a woman's bloodstream, moreover, may not even reflect information actually needed about hormonal stimulation of endometrial tissue. That stimulation does not depend only on amounts of hormone produced and circulating in blood throughout her body. Hormones must also interact adequately with hormone receptors—in this instance, located in the uterine lining—to achieve their needed effect.^[38]

The third diagnostic approach, endometrial biopsy, is intended to analyze uterine tissue, in order to assess more directly the endometrium's preparation for implantation of an embryo. Though frequently performed, recent studies question the usefulness of this more invasive, painful, and costly test for predicting pregnancy success. Again, there are unresolved questions about when, during the second half of a woman's menstrual cycle, this test should be performed and about the need for testing on more than one day. The procedure itself can cause an early menstrual period, artificially reinforcing a picture that leads the doctor to a diagnosis of luteal phase deficiency.

At best, medical discussions of methods for identifying LPD describe evidence that is "suggestive," but not "diagnostic," of "rough approximations" rather than accurate or valid data. The incidence of LPD, many specialists now argue, is

considerably lower than previously claimed on the basis of the above tests. Even more important, women identified through such tests do not appear to have lowered fertility. Simply put, one review of diagnostic methods concluded: "The diagnosis of LPD in the clinical setting remains problematic and controversial primarily because there is no practical diagnostic method that has been validated."^[39] And a study of endometrial biopsy in particular found "firstly that diagnosis of LPD in both infertile and fertile women represents only a chance event; secondly, histological [from biopsy tissue analysis] endometrial adequacy or inadequacy in the cycle of conception or in previous cycles is not related to the outcome of pregnancy in infertile patients. Finally, treatment of LPD does not improve pregnancy outcome in infertile women. Thus, luteal phase evaluation [through biopsy] of the endometrium is not worthwhile."^[40]

This final conclusion points to the crucial added problem attached to the diagnosis of LPD: diagnosis does lead to treatment . Not only do women undergo unpleasant, often painful, and costly tests that are not worthwhile, but these tests then become the basis for undergoing treatments never demonstrated to be worthwhile. Doctors commonly prescribe progesterone supplements following ovulation, yet there is no evidence this treatment improves pregnancy rates. Doctors try pre-ovulation clomiphene citrate in an effort to attain a "better" ovulation that will produce greater amounts of progesterone, though some specialists advise that clomiphene should not be given to women who ovulate.^[41] After clomiphene, the doctor may prescribe the more powerful, difficult, and risky hMG; however, this treatment can itself shorten the woman's luteal phase. Confusion has come full circle—a treatment is tried on women to correct a particular condition whose very existence is a bone of contention; the treatment may cause symptoms of the diagnosed condition and must certainly obscure what is cause, what is effect, why the woman is not pregnant.

All such interventions bring risk—in these cases, with no known benefits. And, there is the further risk of initiating a self-perpetuating cycle of worthless diagnoses and futile treatments, with women and their doctors caught by the irresistible momentum of doing something. One can well imagine the following scenario: a couple has now accumulated three months of basal body temperature charts. One month the elevated dots cover thirteen days, another month ten, then thirteen again. The doctor orders a blood test for progesterone, which measures in the low-normal range. The doctor prescribes progesterone. After several months, he adds Clomid. Two months later the woman comes to the office for an endometrial biopsy, a procedure far more painful than the nurse described when she advised the woman to arrive well-dosed with Motrin. A

shorter luteal phase follows. The doctor escalates to hMG injections; the woman must schedule her days around the injections and frequent ultrasound monitoring of her ovaries to check for ovarian hyperstimulation syndrome. Still not pregnant, the couple asks, "What now?"

There is no good way for the couple or the doctor to answer this question. From the very first step, the months of BBT charts, the missing element was knowledge about the biology of this woman's fertility problem, if indeed she has one. Without such knowledge, no specialist can determine whether any treatment will help the woman conceive. As matters now stand, patients and their doctors are dealing with what the American College of Obstetricians and Gynecologists officially describes as "a term applied to a poorly understood group of subtle hormonal alterations."^[42] The diagnosis encompasses a heterogeneous group. Some women have never conceived, others miscarry very early in pregnancy. Some women ovulate regularly, others do not. Hormonal evaluations show varying patterns of various hormones. For some subgroup of women, medical intervention may improve fertility, but there is not way of knowing for whom or with what intervention.

The entity called LPD illustrates the need for more concise and discriminating diagnostic methods before applying treatments to individual women, as well as for treatments that target more specifically an identified fertility problem. In the meantime, the large, heterogeneous group of women diagnosed with LPD undergoes the same therapies when, in fact, their underlying physiological conditions call for differing medical treatment—or perhaps none at all. Indeed, this particular diagnostic label may become a historical remnant in this era of cost concerns. In the words of one specialist, LPD will be a "so what?" diagnosis. Why perform tests at all—especially flawed, often costly tests—when women will be given the same broadscale treatments anyway? Rather than improve diagnostic abilities to better determine individual patient needs, the label will fade—but interventions will continue to engulf a number of vaguely defined fertility problems.

Although physician enthusiasm for the LPD diagnosis may wane, two areas of fertility medicine attracting active and increasing attention—male factor and immunologic infertility—demonstrate how the existence of a diagnosis fuels precipitous intervention in the absence of adequate biological knowledge. These diagnoses may eventually identify well-understood physiological conditions that impair fertility and respond to treatment in some couples, affecting more definable and limited populations of women and men than does the amorphous luteal deficiency; at present, however, male factor and immunologic diagnoses

display a sense of groping in the dark and the risks such groping can bring. Patients need to be all the more cautious about new and unproven procedures on the one hand and outdated and disproven procedures on the other.

Male Factor Infertility

We currently remain in a position where clinical practice has run ahead of scientific analysis, allowing treatment of previously untreatable semen samples by means of techniques we do not yet fully understand.
Tucker et al. 1993, 324

One clear sign of progress in fertility medicine during recent years is that the woman seeing the doctor is often not alone; the appointment is for a couple. The doctor will more likely order a basic test of the man's semen and sperm before recommending extensive diagnostic procedures or treatments for the woman. Although male infertility is not a new diagnosis, infertility has until recently been defined and treated largely as a woman's problem. Frequently physicians failed to check the fertility of a woman's partner even in cases of "unexplained infertility." She might undergo years of medications, pelvic surgeries, and inseminations, unaware that the explanation resided with him. Knowledge about male fertility remained meager, as did treatment options. Now, specialists generally describe fertility problems as divided fairly evenly between the woman and man (approximately 40 percent each), with both partners contributing to lowered fertility in many couples trying to conceive. Indeed, the most common cause of infertility involves some type of sperm abnormality, many of which may not be detectable through sperm counts.^[43] Physicians continue to characterize this area of fertility medicine as "frustrating," an "enigma" with treatment results that are "discouraging"; by the 1990s, however, male fertility problems could no longer be ignored. In fact, male factor infertility had engendered its own subspecialty among urologists: andrologists now join the ranks of fertility specialists, along with their gynecologist colleagues.

This greatly increased attention to male fertility is yet another spin-off of in vitro fertilization. With attempts to join egg and sperm in a laboratory dish, physicians observed before their very eyes numerous instances in which apparently normal sperm were unable to fertilize an egg. Conversely, successful fertilization came as a surprise in some cases where sperm or semen characteristics predicted failure according to standard diagnostic tests. Experience with assisted reproductive technologies enhanced appreciation of the complexity of sperm and of the sperm's interaction with an egg. New data forced doctors and scientists to rethink the fertilization process. At the same time, the growth of as-

sisted reproduction opened new possibilities for overcoming male infertility through the use of various in vitro techniques. "Assisted fertilization" became a substep of assisted reproduction and the focus for a cascade of new fertility interventions. News reports began announcing triumphs with new treatments for male infertility.

Certainly, newfound medical interest in male fertility holds potential for expanded biological understanding and effective treatments for some conditions. If the history of fertility medicine is any guide, however, patients will need to avoid an equal potential for widespread unscientific experimentation with and dissemination of medical procedures claimed to solve the male fertility problem. Particularly during these initial years, during the first flush of professional enthusiasm over changing concepts of male fertility, doctors and patients may be all too eager to attach some kind of treatment to the reformulated diagnosis.

Enthusiasm needs to be tempered by the recognition that the diagnosis itself brings the first frustration. Physicians have few reliable methods for answering the most basic question of whether a man is fertile. Even watching a man's sperm perform with his partner's egg in a laboratory dish does not always answer this question; a failed in vitro fertilization one month may be followed by conception—in vitro or in vivo—the next. At the same time couples could read front-page newspaper articles lauding rapid development of high-tech treatments, they were almost certainly not reading a 1994 review article in Lancet stating that "the truth is [that male] diagnoses in most cases cover up our ignorance."^[44]

Furthermore, proponents are pushing costly and invasive interventions despite major gaps in knowledge about the problem treated or the impact of these new treatments. The mainstay fertility test for a man uses a microscope to observe the number, shape, and movement of sperm in a semen sample. A low sperm count or immobile or "sluggish" sperm have portended bleak prospects for pregnancy, as did some other semen characteristics (e.g., presence of white blood cells). However, men with poor results on standard semen analysis can have normal fertility; men whose semen test appears normal can have sperm disorders that severely impair fertility. According to one overview of fertility treatments, this basic, ubiquitous diagnostic tool "is at best only a weak predictor of fertility."^[45]

During the 1980s, specialists tried to develop a test of sperms' ability to fertilize an egg. Though now widely used, recent studies suggest this popular fertility test (the sperm penetration assay, or "hamster egg test," measuring how

successfully a man's sperm penetrate this animal's egg) does not accurately discriminate fertile from infertile men.^[46] Naturally conceived pregnancies occur for men whose sperm failed dismally with the hamster's egg in this laboratory test.

Reliance on the common diagnostic tests means that couples spend time, money, and emotional resources gaining unreliable information about their chances for a pregnancy. Test results "inform" their decisions about whether to undergo fertility treatment and the type of intervention to consider. And because the tests do not accurately identify men who are or are not fertile, ineffective treatments may be given credit for subsequent pregnancies, reinforcing use of the treatments for more and more couples. In reality, the discouraging status of diagnosis for male infertility is matched by a discouraging treatment picture.^[47] No therapy for men—including surgery, hormonal medications such as clomiphene, or steroid medications—has effectively restored fertility when sperm disorders are accurately identified. Nor has intrauterine insemination or standard IVF overcome fertility problems in men.

No therapy can be adequately evaluated, because the definition and diagnosis of male infertility are so questionable in the first place. Conditions once thought to impair fertility probably do not (e.g., asymptomatic varicocele, a testicular vein varicosity), and some (certain types of white blood cells found in semen) may actually play a fertility-enhancing role.^[48] The in vitro spotlight on male contributions to infertility displayed vividly how mysterious sperm and the fertilization process remains. Male infertility became all the more enigmatic as the heterogeneous nature of sperm disorders became evident. Ironically, how little is known about male fertility is revealed most dramatically through experimental interventions that are particularly troubling because they are performed with such little biological knowledge. The approach presently attracting excitement among fertility specialists is micromanipulation .

Although the couple asking a doctor for help might never put it this way, they may be asking a brand-new question if the diagnosis is male infertility: Can you help our sperm get into our egg? In this post-IVF era, micromanipulations (procedures requiring use of a microscope) of egg and sperm are experimental treatments that attempt to do just that. Driven by the observation that some sperm—including some with no apparent defect—just cannot fertilize their partner's eggs, fertility specialists have been trying to assist this process more directly. One approach uses chemical and/or mechanical procedures to thin or cut through the zona pellucida surrounding the ovulated egg. This approximately 4-micromillimeter-thick corona-like barrier must be traversed by sperm before

entering the egg cell itself. The hope was that sperm that for some reason cannot otherwise navigate successfully could follow the laboratory-created path of least resistance and that one would penetrate the egg for fertilization. Greater assistance takes the form of injecting sperm under the zona pellucida (called subzonal or underzonal insemination). The microscopic manipulation that currently has the most potential for achieving fertilization, especially with severe male problems, injects a single sperm past all barriers into an egg (called intracytoplasmic sperm injection, or ICSI). Held at the tip of a needle, the sperm is "introduced as atraumatically as possible into the egg" with the hope it will remain lodged inside the egg after the needle is withdrawn.^[49] As this country's most experienced ICSI practitioner bragged, "The presence of just a few, weakly twitching spermatozoa in the semen is all that is required" of men with extremely low sperm counts.^[50] Indeed, by the mid-1990s, the injected sperm need not even come from an ejaculated semen sample or have completed its maturation process. For men with no sperm in the semen (azoospermia) due to an obstruction, the mere head of an immature sperm could be withdrawn from the epididymis or from testicular tissue. And specialists were beginning to "tackle the problem" of infertility caused by "maturational arrest"^[51] —a failure of sperm to develop fully, often traced to a defect on the Y chromosome. Specialists predict that someday assisted reproduction will require only the man's genetic material, which could be taken, for example, from blood cells.

Medical reports during these early years of micromanipulation leave no doubt that physicians and their patients are forging ahead into uncharted territory.^[52] The fertilization process remains a black box, medical intervention a stab into its darkness. No matter how far from understanding, however, treatment proceeds. According to a 1992 summary of assisted reproduction, the few existing comparisons between micromanipulation and standard in vitro procedures suggested that, for most people, there is "little or no benefit [from] extreme attempts to facilitate entry or inject spermatozoa through the zona pellucida."^[53] This summary concluded that such attempts should "be focused on only the most severe cases" and that varying micromanipulation techniques should perhaps be applied to differing sperm disorders. Unfortunately, this conclusion immediately raises two problems. First, it brings us right back to the unknowns in defining and diagnosing male infertility of varying types and severity. And second, even if these technologies do prove, through scientifically controlled trials, to benefit certain severe types of male infertility, prospects for limiting their use to these certain conditions would not seem good. Judging from past experience, the more likely outcome is blanket application to diverse

but poorly understood fertility problems, followed by unsubstantiated claims of success.

In 1994 the American Society for Reproductive Medicine issued revised ethical guidelines for use of ICSI and similar new technologies. The aims of the guidelines are "to prevent the exploitation of procedures" for personal gain or of procedures that pose significant risk to patients and to offset confusion and distortion produced by intense media coverage. The society's ethics committee stated these major reservations about micromanipulation: low probability of success; possible damage to gametes, which may result in abnormal offspring; and potential for exploitation and abuse because results are difficult to measure.^[54]

International alarms also sounded. In 1995 French medical ethicists investigating the impact of ICSI on assisted reproduction wrote to Lancet: "The potential risks related to the implementation of ICSI should not be neglected—the genetic risk, the risk linked to abnormalities associated with male infertility, the risk of penetration of the oocyte, and the risk of introducing foreign material. Many are concerned about the uncertainty surrounding the fate of children born by means of this technique. . . . Injection [of immature sperm] has made this uncertainty even greater. A review of published work shows that ICSI constitutes human experimentation not preceded by adequate work in animals . . . done without any ethics committee approval of research, as provided for by [French law] . . . and its implementation has not been accompanied by any national evaluation protocol to follow the biological risks and societal effects, nor by any epidemiological surveys."^[55] Similarly, Italian specialists warned that "application of [ICSI] is burgeoning and we think that this progression has been premature: there is an absence of experimental and clinical evidence for the safety of this technique." Of particular concern is "the risk of fertilisation by abnormal sperm."^[56] Swedish physicians, reporting on another use of ICSI (reinseminating an egg as a last-ditch effort after a standard in vitro fertilization failure), concluded that "considering both the extra work involved and the potential genetic risk, it is doubtful whether [this use of] ICSI should be recommended as a routine procedure."^[57] They also noted that attaining pregnancy during a subsequent IVF cycle did not always require ICSI.

If micromanipulation of egg and sperm has aroused some skepticism, this emerging technology—like previous innovations in fertility medicine—also has strong advocates. In the 1990s, a new vocabulary proliferates in reports on male infertility: zonal drilling, partial zonal dissection, subzonal insertion, microinsemination, microsurgical epididymal sperm aspiration (MESA), testicular sperm extraction (TESE), direct egg injection, re-insemination, assisted hatch-

ing. The reports are mixed—some claim improved fertilization, others do not. Most of the experimental treatments lack scientific controls. Once again, interventions are becoming "established" so rapidly and unsystematically that there will be no way to determine what benefits exist and for whom.

Early results after use of ICSI are leading advocates to claim success rates for infertile men that are comparable to standard IVF in couples with no male fertility problems; early reports claim no increase in birth defects or chromosome abnormalities as well.^[58] An American group reporting on ICSI pregnancies—after cautioning that this advanced procedure should be limited to cases of severe male infertility—also expresses "hope" that understanding the relevance of sperm disorders to fertility will follow the clinical success already being claimed for this reproductive intervention.^[59] And, in a generally enthusiastic overview of ICSI in 1996, specialists from one of this country's largest fertility clinics acknowledge that "further investigation must address several major concerns: (1) oocyte damage; (2) the arbitrary selection of gametes; (3) the genetic integrity of the embryo; and (4) ensuing healthy live birth." They conclude, however, that success with the "one egg–one sperm" approach to treating severe male infertility "suggests that in the future ICSI may enjoy wider application in all IVF related treatments."^[60]

Other specialists are less sanguine about the early results; regarding offspring, in particular, one group notes that "normal appearance does not preclude infants' carrying significant genetic defects that may not be clinically apparent until adulthood."^[61] Indiscriminate and widespread use of ICSI—especially without improved diagnostic abilities—is eliminating the opportunity for properly designed studies to determine which men need this extreme assistance to attain fertilization and to document the long-term outcomes for offspring. Also neglected will be research into causes of male infertility and development of treatments aimed more directly at restoring male fertility and of strategies to prevent at least some of these problems in the first place.^[62]

What is certain is that neither patients nor doctors can ignore this technology. They can, however, be wary. They can be wary first of testimonials, based on personal observations and vague descriptions, that substitute for rigorous evidence. For example, a 1993 report echoes refrains heard time and time again about fertility treatments, phrases that did not bode well for patients in the past: doctors' impressions and patients' hopes are disturbingly prominent, hard data and potential risks barely evident. The published article opens with mention of unknowns as basic as identifying "which spermatozoa are capable of fertilization on a consistent basis." Then—turning from these unknowns with a

quick "be that as it may"—the authors launch into a description of their clinic's experience treating seventy-three couples by subzonal insemination and/or direct injection of sperm into an egg. Acknowledging that "these approaches may seem unwarranted" and that "we can be accused of being overzealous in our reporting of these procedures," the authors assert that "for many couples these techniques are their only hope for successful fertilization." With hope raised, however, reliance on the practitioners' clinical observations and impressions, rather than on any scientific evaluation, becomes evident. "Because of the conflicts of clinical and scientific interests [i.e., deciding on treatment for individual patients rather than randomly assigning patients in clinical trials], well-controlled comparisons with conventional insemination techniques are not feasible." They do conclude, however, that micromanipulation has been "of great value to many of the couples reported here."^[63]

By the final paragraph, this nebulous claim of great value for many patients takes on additional troubling dimensions. Not only do the authors dismiss potential risks, but they encourage dissemination of these untested (and in their view untestable) technologies. Although not simple, the procedures "are well within the grasp" of most IVF laboratories—the necessary equipment is not prohibitively expensive or exotic; personnel can be easily trained. Although the "lax and lack" of knowing which sperm to select for injection might result in too many genetic disorders in the offspring, counseling and prenatal screening will be the answer for many couples who see these assisted fertilization techniques as "their only chance for pregnancy using their own gametes."^[64] These clinicians are in effect telling their colleagues, "We think these treatments work, at least for some, but they cannot be scientifically evaluated. Sure there are unknowns and there may be adverse outcomes . . . but you can do it too."

No doubt they will, and no doubt there will be some pregnancies that would not have otherwise occurred. Fertility specialists are learning as they are doing. Whether investigators report positive results with assisted fertilization or no improvement, they agree that results do not correlate with traditional semen analyses and sperm tests; therefore, doctors cannot identify which couples do require this extra, and extra costly, procedure in order to use the man's sperm for IVF. Long-held notions about sperm, the fertilization process, and the diagnosis of male infertility are crumbling. One specialist noted the "dogma" refuted by the initial Belgian study reporting high rates of fertilization and embryo implantation after directly injecting single sperm into single eggs. Existing wisdom might have predicted the technique could not work—that sperm must undergo various physiological changes in the woman's reproductive tract before

fertilization is possible, and that semen characteristics influence the fertilization process. Though impressed by these revelations, this commentator does convey some ambivalence, sardonically predicting another likely outcome of this study: "Members of assisted reproductive technology groups from around the world will be flocking to Brussels to learn this technique as if it were Lourdes."^[65]

In fact, a pilgrimage of doctors to Brussels—and of patients seeking treatment—was already under way. As the decade progressed, more and more fertility clinics added ICSI to their treatment repertoire. With preliminary studies showing greater fertilization success than did the earlier attempts to assist sperm entry into an egg, ICSI has become the only micromanipulation specialists recommend. Ironically, ICSI's great popularity may actually turn male infertility into another "so what?" diagnosis; development of less invasive therapies for male infertility may be thwarted. Fertility specialists do not know why fertilization fails either in vivo or in vitro. In some IVF cycles, a problem with the ovarian stimulation or other unknown transient factors may adversely affect the egg-sperm interaction. Yet, although a failed in vitro fertilization is sometimes followed by success during the next cycle, many clinics now proceed to ICSI after fertilization failure. And others are jumping to ICSI even sooner, directing all couples straight to IVF with ICSI.^[66]

Better understanding of sperm maturation and the interaction of egg and sperm could lead to new targeted treatments, such as hormonal or other types of medications for men that improve sperms' ability to fertilize an egg (though researchers seem to experiment more gingerly with hormones for men than for women). Men would avoid treatments that cannot improve their fertility—e.g., surgery to remove varicoceles—and women would avoid undergoing IVF with or without ICSI.^[67]

Considerable doubt must therefore surround the potential benefit of current directions for overcoming male infertility. The search for treatments to match this diagnosis illustrates particularly well—almost in caricature—four significant issues of fertility medicine more generally, issues demanding additional wariness.

First, treatments presently generating the greatest interest among fertility specialists are all linked with IVF and superovulation of a man's partner. Infertility may no longer be attributed primarily to women, but women bear the burden of treatment, even for male fertility problems. These treatments are presented to subfertile couples as the choices they have, and many women agree to try. Early efforts with ICSI already place responsibility for treatment success upon the woman: proponents define the major determinant of ICSI success in

terms of egg quantity and quality.^[68] That is, by injecting a sperm in an egg, previously abysmal success rates for treatment of severe male infertility can be brought up to rates for IVF with a fertile male—if a woman is willing to undergo superovulation and embryo transfer and if she does not have "defective eggs" (due primarily to age) or is willing to use donated eggs. No other proven option exists, except artificial insemination with donor sperm—in fact, a more successful procedure that is much simpler and safer for the woman.

Second, early medical reports place more emphasis on how micromanipulation is done—and the relative ease of learning the techniques—than on determining when such extreme interventions are necessary for conceiving a pregnancy. In the absence of knowing how to predict which couples need ICSI in order to achieve fertilization, women again end up bearing the burden of diagnosis. IVF has become the ultimate diagnostic determinant of male fertility—by observing in a laboratory dish the sperms' ability to fertilize an egg—a diagnostic method that already involves superovulation and egg retrieval. Although several studies suggest couples can conceive on a next try following an in vitro fertilization failure, that failure becomes the reason for escalating to ICSI.^[69]

Third, as with assisted reproduction for other conditions, treatment to bypass male infertility brings crucial questions about "success." Achieving "successful fertilization," or even embryo implantation, following gamete micromanipulation is not equivalent to achieving a successful pregnancy. Once again, a healthy "take home baby" who would not have otherwise been born is the outcome of ultimate concern. Valid claims of success require evidence that the intervention was responsible for the birth. Without better understanding of male fertility, and scrupulous attention to the type and severity of a couple's subfertility, patients offered sophisticated and expensive micromanipulation might do just as well with standard IVF; patients offered IVF might conceive on their own.^[70]

Finally, if of less immediate impact on today's patients, is a broader dimension characterizing male fertility treatments—a dimension that in the long run affects many more women and men than those presently seeking reproductive assistance. Medical attention, resources, and technology now focus on individual sperm. From a specialty that usually deals with sperm by the millions, a sophisticated therapeutic arsenal aims at guiding a single sperm into a single egg. This approach is consistent with the emphasis of American medicine. While scientists have learned that male fertility can be significantly lowered by such environmental factors as chemical exposures (occupational and environmental), cigarette smoking, heavy alcohol consumption, and recreational and therapeutic

drug use, resources allocated to preventing sperm disorders are pitifully meager. While researchers investigate a decline in sperm counts internationally, perhaps caused by man-made chemicals in the environment, the medical emphasis in this country remains a highly technological and invasive fix that may benefit only a few, after the widespread damage to men's fertility is done.

Immunologic Infertility

Early claims [about immunology and infertility] were magnified because of unreliable assays and naïveté concerning the complexities of the immune response. . . . The dearth of well-designed and controlled experimental studies and the lack of effective therapy resulted in confusion [over immune processes] in human infertility.
Naz and Menge 1994, 1001

By the mid-1990s, some fertility specialists could recognize and acknowledge earlier "naïveté" about ways the immune system might contribute to infertility. How any pregnancy manages to succeed has long been an intriguing scientific question. After all, a woman's immune system normally attacks "foreign" cells or tissue—whether a viral or bacterial invader, or a transplanted organ. That an embryo can implant and grow without provoking a similar rejection is something of a mystery. Equally mysterious are reasons the normal reproductive process sometimes falters—yet it certainly makes biological sense that an immune response may be implicated. For any woman, sperm or embryos are different from her own body tissue. What if a woman's immune system reacts against them? Some breakdown of a special immunologic relationship that normally prevents this reaction may be the reason for a portion of unexplained infertility and of the agonizing cases in which women, for unknown reasons, suffer repeated early miscarriages.

Today, as in past years, patients receive treatments, the dearth of scientifically controlled studies and lack of effective therapy notwithstanding. Many couples with unexplained fertility problems visit the doctor innumerable times, devouring months, then years, with all manner of diagnostic tests and treatment attempts. They may be more likely than other fertility patients to try new therapies—anything that might, for whatever reason, remedy their mysterious condition. But today's patients need to ask whether new diagnostic tests can accurately determine if an irregular immune response is lowering their fertility. Are therapies the doctor is now recommending likely to overcome this problem, increasing chances for having a healthy baby? Or, in years to come, will today's immunologic information look naive, the interventions prescribed unjustified?

To assess the current moment, it helps to understand how diverse medical developments reinvigorated both clinicians' excitement about the diagnosis of immunologic infertility and their experiments with treatments—for patients are susceptible to both enthusiasms. Although relatively few couples are directly affected and although the following discussion of this complex, rapidly changing field must be limited to basic concepts, this diagnosis provides an illustration of the general need for caution in order to avoid premature, potentially harmful reproductive interventions.

As with increased attention to male fertility problems, development of IVF also spurred interest in the broad phenomenon called immunologic infertility. These technologies divided the reproductive process into discrete, observable steps, with each step holding the potential, at least in theory, for an immunologic stumble that might thwart pregnancy. If male fertility was considered enigmatic, male and female entwined—a couple's immunologic fit in vitro or vivo—appeared all the more so.

A faulty immune response may impede the earliest steps in which sperm travel through a woman's vagina and uterus to reach and penetrate an egg. If an egg is fertilized, an embryo's initial growth may also require certain immunologic responses, although the physiology of such a process is not yet understood. The next steps seem to be of critical importance. The embryo must establish itself in the uterine lining—as a "transplant" of sorts—accepted and nourished there throughout the months of fetal development until birth. As IVF practitioners and their patients learned all too well, even following a "successful" in vitro fertilization and transfer of embryos, some women consistently miscarry. With known causes (e.g., chromosomal anomalies, uterine structure defects) ruled out, immunologic "rejection" of the fetus became a prime if inscrutable suspect.

Assisted reproduction was not the only medical development fueling interest in the immune system's impact on fertility. In the 1980s two complementary explosions in knowledge and biotechnology reverberated throughout the medical world. First was the AIDS epidemic, which focused scientific and medical resources on the immune system, trying to delineate how it works and how it breaks down. In the wake of AIDS came new research and diagnostic tools aimed at unraveling immunologic mysteries. At the same time, study of human genes—particularly as culprits held responsible for specific health problems—yielded additional research techniques. Soon these new techniques were filtering into diverse medical specialties, including fertility medicine, allowing specialists to observe and measure previously hidden facets of the immune system as it participates in conception and pregnancy.

Numerous laboratory tests now generate numerous measures of various immune system activities. Measurements of differing types of antibodies (protein molecules the body produces to fight organisms or tissue not recognized as one's own) circulating in a woman's bloodstream may reflect an immune response to sperm, embryo, or fetus. DNA analyses can identify genetic characteristics that might influence a couple's fertility. Translation into safe and effective reproductive treatments, however, is not a foregone conclusion. Certainly, any such progression will not be quick or easy. As medical reports accumulate, their more immediate results translate into problematic choices for patients whose doctor suggests a diagnosis of immunologic infertility.

A first problem is the variable quality of laboratories now jumping into the lucrative business of immunologic testing and the still early stage in the development of the test themselves. For the couple seeking a doctor's help, diagnosis relies on a series of assays run on samples of their blood. However, different laboratories come up with differing measurements on the same blood test; out of the many labs to which doctors send blood samples, few have a reliable track record for measuring the immune activities in question.^[71] Laboratories define results as normal or abnormal differently, according to their own arbitrary cutoff points. To appreciate how difficult the technical issues involved in newly evolving immunologic tests are, remember that quality control has been a serious and ongoing problem with the long-established, standard Pap test for cervical cancer. Yet laboratories promote their services to doctors—for example, through journal advertisements—despite significant technical problems, not to mention vast unknowns regarding what these tests might reflect about fertility. Some advertisements suggest conditions for which antibody testing has been "recommended" although these recommendations are a focus of medical dispute (see Figure 3).

This lack of knowledge about the immune system's role in fertility is an even more basic concern. Tests now exist, but tests—even accurate, reliable ones—must be interpreted. With tools already in hand and continuing to develop, specialists are now debating just what the tests are measuring, what the numbers mean, what relevance the measures have to a couple's likelihood of pregnancy. Confusion surrounds the origin and function of particular antibodies circulating in a woman's bloodstream and, most important, whether they affect her chances for pregnancy. Little is known about the normal distribution of these antibodies in women who do not have problems with conception or early pregnancy loss. Nor has research determined whether such antibodies are a cause of infertility or a result. Elevated levels could be physiological residue of

Figure 3
One laboratory's ad seen regularly in ob-gyn journals, seeking physician customers
for diagnostic tests that remain controversial. A list of conditions for which these
tests have been "recommended" appears in fine print at the bottom of the ad (not
shown here), including unexplained infertility, endometriosis, and IVF prescreening—all
hotly disputed recommendations.

the earliest stages of a previous pregnancy—including unrecognized conceptions—or of the miscarriage process itself.^[72]

The lag in knowledge leads easily to overdiagnosis and overtreatment of patients as the new techniques gain popularity among doctors. Since test results themselves may not be reliable, the diagnosis given to patients by their doctor may not be accurate. The sheer number of diagnostic tests now available raises the likelihood of some abnormal result. The more tests run, the more chance of an inaccurate positive finding (a "false positive"). Or a measurement will fall outside the range a particular laboratory defines as normal yet not, in fact, affect the couple's fertility. One outcome of diagnosing and treating an immunologic fertility problem without adequate knowledge is that the intervention will be

futile, a waste of time, energy, resources. Even if a couple does have some type of immunologic problem, treatment may be ineffective because it targets poorly understood immune processes.^[73]

Futility might seem a minor concern, however, compared to the very real potential for harm. Beyond the risk any medical procedure brings are additional risks from tinkering with a physiological system that is central to the overall health of a woman and her desired offspring. Troubling questions arise, first, about assisted reproductive technologies now frequently recommended for women diagnosed with either immunologic or unexplained infertility who do indeed have some mild, perhaps symptomless immune system irregularity. Procedures such as IVF or intrauterine insemination following superovulation expose women to unusual concentrations of sperm and/or early embryos; these exposures could aggravate conditions involving abnormal immune reactions. No one knows the immune response to repeated hormonal superovulation or punctures of the ovary during egg retrieval. In addition to jeopardizing a woman's health, such intervention may jeopardize her best chance for pregnancy—conceiving spontaneously, though it may take her a longer than average time. Each treatment attempt brings a high likelihood of miscarriage and may further exacerbate the problems she is experiencing.

More direct immunotherapies attempt to alter responses between maternal, paternal, and fetal components of an intricate, but poorly understood, immunogenetic relationship. One prominent treatment is based on the idea that the woman's immune system is not adequately mounting the response that normally protects an embryo from immunologic rejection. According to this theory, the problem may be that the woman and her partner are too similar in some genetically crucial way, a "compatibility" reflected in the combined genetics of an early embryo. To stimulate the needed protective response, the woman is injected with white blood cells (leukocytes or lymphocytes) drawn from her partner. If this "paternal lymphocyte immunization" fails to provoke an adequate response, white blood cells are added from a less genetically compatible donor. Immunization begins before attempts to conceive and—depending on periodic test results—may continue at intervals throughout the first two trimesters of pregnancy.

A different treatment gaining attention in recent years is based on different test results, and attempts to subdue a different immunologic process—an excessive , "autoimmune" attack in response to a developing embryo. The woman receives a cocktail of aspirin, other blood-thinning and, perhaps, steroid medications to dampen her immunologic response and ameliorate its damaging physiological consequences. Some specialists start this treatment before conception, while others begin treatment when pregnancy is confirmed. Another therapy of unproven efficacy and safety attempts to supply a woman directly with immune system proteins that might improve her response by infusing immunoglobulins at regular intervals during pregnancy (for example, every three weeks until the final month): each unit of blood products required for this process (intravenous gamma or immunoglobulin therapy, IVIG) is pooled from approximately eighty anonymous donors.

Although these specific interventions vary, each brings its own dangers. In many instances, not only the woman but, if she does conceive, a developing fetus will be exposed to treatment. Certain risks are well documented and can be life-threatening to the pregnant woman—for example, bleeding problems and bone demineralization from blood-thinning medications (e.g., heparin, aspirin); viral contamination from intravenous immunoglobulin; ulcers, diabetes, and bone demineralization from corticosteroids (as well as the possibility of fetal abnormalities). Questions also loom about any manipulation of an individual's immune response. No one knows whether immunization with a different person's lymphocytes, for example, may have long-term consequences for the woman or an exposed fetus. As with other fertility treatments, most patients considering immunotherapy are women whose present good health could be compromised. A subgroup of women may have a mild or latent irregularity that these procedures could aggravate—a condition that may or may not be lowering their fertility. Even if it is, preliminary data indicate that, at best, present immunotherapies might improve chances of successful pregnancy for a very small proportion of these women. With no reliable tests to identify which couples are likely to be helped, many more women than can actually benefit will be taking on the risks of treatment.^[74]

In addition, explanations of immunologic infertility are noticeably fluid. Factors once described as significant in lowering fertility—for example, antisperm antibodies or genetic "sharing"—do not appear to be as important when investigated more thoroughly.^[75] At this still early stage in understanding the diagnosis called immunologic infertility, specialists are saying, in very loose translation, that there does seem to be something going on here, at least for a small percentage of couples, but we can't yet grasp what that something is or identify

who those couples are. Yet the immunologic diagnosis continues to expand, now reaching back to explain the failure of fertilization or implantation, as well as repeated miscarriage, during the first or second trimester. In earlier years doctors typically recommended immunologic approaches only after a woman's third miscarriage; the wait has now been chiseled to two losses, or one, or none (sometimes called "subclinical" autoimmunity, a diagnosis based on test results alone). Obviously, no one wants the experience of miscarriage to be a diagnostic requirement before receiving treatment, if that treatment is beneficial. However, the immunologic tests, themselves controversial, now bear a greater diagnostic burden. Not only is the chance for misdiagnosis enlarged, but the relationship of abnormal test results to fertility remains uncertain. Increased reliance on the tests means that yet again more and more women end up receiving unnecessary, unproven, and risky treatments.

In years to come, research will undoubtedly contribute to more sophisticated and accurate explanations. The persistent need for caution about interventions that proceed while knowledge is lacking or in considerable flux is reflected all too well by women who have experienced repetitive early miscarriages (also called recurrent spontaneous abortions). During the 1990s, "IVF failures" joined the diagnostic pool of couples offered treatments for immunologic infertility. Assisted reproductive technologies helped to reveal early miscarriages that can occur before women know they have conceived. However, couples hearing an immunologic explanation for their difficult and perplexing experience of several miscarriages should know that this long-recognized condition has resisted an extensive list of therapeutic efforts. Treatment prospects may sound different to couples who know that longer story, particularly if they also hear today's critics echoing the same cautions expressed in the past. A 1984 letter to Lancet warned: "Thirty years ago chaos arose from unscientific use of hormones. The only clear results were the tragic effects of diethylstilboestrol and virilizing progestagens [synthetic forms of progesterone] on surviving offspring. Now with immunological approaches, a new generation may fall into the trap of unnecessary treatment; for, even without specific therapy, women who have had a sequence of unexplained abortions have a good prospect of carrying a child to [term]. Immunological measures may be justified in some cases, but evaluation must be conducted at the highest level of academic stringency."^[76] The main difference with today's immunotherapies is the passage of a dozen and more years.^[77]

Faced with this history and with the introduction of ever-changing and ever more biologically sophisticated tests and therapies, what perspective can patients maintain on the diagnosis and treatment of immunologic infertility? They

can remember, first, that more couples are now labeled with this diagnosis, on the basis of immunologic testing, often when no other cause can be identified. This whittling away of the "unexplained" category during the past ten to fifteen years, however, has not been matched by improved pregnancy rates.^[78] Giving infertility an explanation, however valid, does not mean an effective treatment exists. A major problem with claims of immunotherapy success is that even after three consecutive first-trimester miscarriages, a substantial proportion of couples, possibly as high as 70 percent, will go on to have a successful pregnancy without treatment. This comparison becomes all the more important as clinicians shortcut the wait, testing and intervening in couples who may have had at most one early miscarriage (a relatively frequent natural occurrence).

Another complication is the intriguing suggestion that immunotherapies may be particularly susceptible to placebo effects. That is, when the therapies do improve pregnancy outcome, it is some psychoimmunologic boost from the act of intervening itself, not its medical content, that accounts for success. Several studies report similar increases in pregnancy rates with psychotherapy, "tender loving care," use of a placebo (injections or pills that appear the same as the medical therapy but with no active medical substance), or immunotherapy. Just as health and illness, more generally, are affected by mind-body interactions that are not well understood, so too may immunologic aspects of fertility be influenced.^[79] Why psychological support might be an effective intervention remains unknown, as does the reason untreated couples follow three miscarriages with a successful pregnancy. Placebo effects are not unreal or "all in a person's mind." Measurable physiological changes do occur, but it is the support, attention, and/or individual's belief in the treatment that stimulate the physiological response. Given the risks attached to immunotherapies, it is imperative to sort out this process.

The various immunotherapies also require scientifically controlled investigation to determine whether any do improve pregnancy outcomes and, if so, for what conditions. Preliminary results from adequately designed studies, performed at a few research centers, are beginning to piece together evidence that may help define immune responses that significantly impair fertility, as well as develop treatments in which benefits outweigh risks. Couples diagnosed with immunologic infertility should know they join a highly heterogeneous group; although they are offered the same treatments, the risks and benefits of a particular intervention will differ, depending on the specifics of a couple's fertility problem. Presently, medical knowledge about such specifics remains inadequate. The diagnosis encompasses abnormalities of sperm moving through a woman's reproductive tract, failure of fertilization or implantation, miscarriages, intrauterine growth retardation, and stillbirths. Not only do these conditions differ from each other, but each can itself result from a range of immunologic, genetic, endocrine, structural, and/or infectious disorders. Immunological treatment that may be promising and biologically plausible for one group (e.g., women with second-trimester pregnancy loss) cannot necessarily be transposed to another (e.g., IVF failures).^[80]

Like other reproductive interventions, immunotherapies can gain a foothold and become widely adopted despite concerns and unknowns. Debate may flare within the profession over the science and its application while clinical practice proceeds on its own steam. Unfortunately, with the diagnosis and treatment of immunologic infertility, clinical excitement and advocacy, verging at times on proselytizing, outpaces demonstrable achievements. Discussion provoked by one published study offers a view of how professional debate over scientific merit coexists with advocacy of a treatment. The study compared two ways of using blood-thinning (heparin) and steroid medications aimed at suppressing a woman's immune response to a fetus: starting medications before the woman attempts to become pregnant or waiting until she has conceived.^[81]

The report's authors acknowledge that "what constitutes an autoimmune abnormality in women with recurrent spontaneous abortion is not yet defined or agreed on. . . . How often the abnormalities discussed in this study are seen in normally reproducing women before and during pregnancy is unknown." Some women in their study, who show abnormal test results and experience repeated miscarriages over more than twelve years, have not had detectable autoimmune symptoms or diseases. The authors also acknowledge that, although lymphocyte immunization accompanied the study's experimental treatment of the combined steroid and blood-thinning medications, they do not know, and

are only now studying, whether the immunization is helpful. Despite these limitations, they conclude that women who have experienced recurrent spontaneous abortion, with diagnostic tests showing autoimmune abnormalities, should be given both immunosuppressive and blood-thinning medications and that treatment should begin before attempting another pregnancy rather than after pregnancy is confirmed.

Comments by other physicians are included in the publication of this article. In this medical format, where etiquette tends toward complimentary and supportive discussion, the responding specialists are noticeably unenthusiastic. The first commentator notes that the "timely topic" of potential immunologic aspects of recurrent spontaneous abortion "remains a confusing and controversial area clearly in need of well-controlled studies." He clearly does not think the present study meets this need; rather, it "requires cautious interpretation" for several reasons. His criticisms could not be more basic: the diagnosis of autoimmune infertility is questionable, the study design is faulty, and the advocated treatment brings significant risks to the woman.

Regarding diagnosis, the large number of blood tests run on each patient increases the likelihood that at least one test result will be abnormal yet not reflect any condition that affects the individual's fertility. Furthermore, the study's arbitrary cut-off defining normal and abnormal levels of antibodies measured in the blood probably means that many women who could have successful pregnancies without treatment are diagnosed as having immunologic infertility.

A fundamental problem with the design of the study is that women were assigned to treatment subgroups based on consultations with physicians rather than randomly and blindly (i.e., without patients or doctors knowing who was receiving which treatment); these non-controlled conditions cast doubt on comparisons of results between the groups. The study also did not include a no-treatment group, given only placebo. In addition to these flaws, the commentator lists more than a dozen other criticisms of the way the study was conducted, the statistical analyses used, and the conclusions drawn. His own analysis, in fact, suggests the drug combination advocated may actually be less effective than a safer alternative started after conception occurs. Combining corticosteroids (which suppress immune functions) and heparin (a blood thinner) during pregnancy can cause serious complications for the pregnant woman, such as bleeding or a severe bone disorder leading to fractures. "We have recommended against this treatment in the absence of evidence that it is better than [steroids] or heparin alone. . . . The effectiveness and safety of this regimen must be clearly shown before it is widely adopted."

A second commentator echoes the concern with risks, emphasizing that longer treatment increases the rate of complications. Since several other studies show no benefit from starting treatment before conception, he concludes that "this strategy may unnecessarily expose patients to potentially dangerous therapy for longer than is needed." And there are other weaknesses. The specific group of women studied were so-called IVF failures who lost embryos repeatedly, very soon after they were transferred to the uterus (failures that can result from a range of technical and physiological causes);^[82] yet the authors' conclusions take a more general form, with no further mention of this limited application.

Limitations similarly go unmentioned when one of the authors of the study speaks, eighteen months after its publication, to an audience mainly of gynecologists as part of a lecture series at a prominent hospital. This clinician-researcher is spreading the word about new techniques for diagnosing and treating immunologic infertility, as well as about his (and his laboratory's) availability for patient referrals. He displays charts and graphs tallying his patients. This presentation has none of the striking sense of criticism so apparent in the journal publication. Definitive statements—for example, "autoimmune therapy started after the pregnancy is established is far less effective than started preconception"—sound impressive. He does not discuss the controversy and confusion characterizing this area of fertility medicine. He does not mention risks, beyond a brief response to one question from the audience about the transient side effects of the treatment. His only allusion to critics of his work is a quotation he reads in closing: "Innovators are rarely received with joy and established authorities launch into condemnation of newer truths, for at every crossroads to the future are a thousand self-appointed guardians of the past."