Recurrent pregnancy loss.

Recurrent pregnancy loss is a frustrating clinical dilemma for both patients and physicians because, in most cases, causes are nebulous and few treatments with proven benefit can be offered. Involved, expensive tests have frequently been proposed and their use has often filtered into clinical practice before their utility has been firmly demonstrated. Proposed causes of recurrent pregnancy loss include genetic and environmental etiologies, infectious agents, maternal congenital and acquired anatomic abnormalities, and immunologic and endocrinologic dysfunction. Appropriate management relies upon a realistic understanding of the often substantial limitations of currently available therapies.

Recurrent pregnancy loss Harry H. Hatasaka* and Michael W. Varnert ♦Division of Reproductive Endocrinology and tDivision of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Utah School of Medicine, Salt Lake City, Utah, USA Recurrent pregnancy loss is a frustrating clinical dilemma for both patients and physicians because, in most cases, causes are nebulous and few treatments with proven benefit can be offered. Involved, expensive tests have frequently been proposed and their use has often filtered into clinical practice before their utility has been firmly demonstrated. Proposed causes of recurrent pregnancy loss include genetic and environmental etiologies, infectious agents, maternal congenital and acquired anatomic abnormalities, and immunologic and endocrinologic dysfunction. Appropriate management relies upon a realistic understanding of the often substantial limitations of currently available therapies. Current Opinion in Obstetrics and Gynecology 1994, 6:503-509 Introduction Human reproduction is an inefficient process. Only ap­proximately 20% of attempted conception cycles will result in clinical pregnancies. Furthermore, 50-75% of conceptions will result in miscarriage. Of women aspir­ing to conceive, 25% will suffer at least one clinically recognized early pregnancy loss during their reproduc­tive careers. Losses comprise 15-20% of all clinically ev­ident pregnancies, and 0.5-1% of women will meet the classical definition for recurrent pregnancy loss (RPL) by having three or more clinically recognized consec­utive pregnancy losses. These women represent a pro­portion who simply have sporadic bad luck three times in a row, as well as women who have underlying repro­ductive problems predisposing to losses. Proposed causes of recurrent pregnancy loss After having a first miscarriage, couples inevitably ask why this happened. The most frequent known cause of sporadic spontaneous abortion (SAB) is cytogenetic abnormalities of the conceptus. Half of preimplantation embryos examined and one-third of embryos that have implanted are abnormal morphologically, which is most often the result of an underlying genetic defect. Aside from genetic abnormalities, there are environ­mental causes of sporadic early pregnancy loss which could theoretically be associated with RPL if a preg­nant woman was persistently exposed. Recently, chronic exposure to video display terminals of more than 20 h each week has been proposed to be a cause of SAB. This claim has been refuted [1], Recent studies examining the association, if any, between caffeine intake and RPL have come to opposite conclusions [2,3], Definitive roles for cigarette smoke, alcohol, anesthetic agents, heavy metals, chloroquine, oral hypoglycemics and industrial organic solvents as contributors to preg­nancy loss remain conjectural. There is some epidemio­logic evidence that smoking more than 14 cigarettes/day increases the risk of aborting a euploid conceptus 1.7 times [4]. Likewise, moderate drinking during 2 days of the week may double the risk over nondrinkers of euploid abortion [5], Genetic Embryonic aneuploidy is the only proposed etiology of RPL for which there is general agreement. To have genetically induced RPL, couples must have abnormal genetic complements which are compatible with nor­mal parental phenotypes and the ability to conceive, yet which persistently result in genetically abnormal concep- tuses that are unable to endure. These conditions can be fulfilled when at least one member of a couple has chro­mosomal aberrations with either an abnormal number of chromosomes (numerical abnormalities) or rearrange­ments of the genetic material within the chromosomes (structural abnormalities). Moreover, to maintain nor­mal phenotypes, these individuals must maintain com­plete (balanced) complements of chromosomal material, Abbreviations ACL-anticardiolipin; APS-antiphospholipid syndrome; APTT-activated partial thromboplastin time; DES-diethylstiIbestrol; LAC lupus anticoagulant; LPI-luteal phase inadequacy; RPL-recurrent pregnancy loss; SAB-spontaneous abortion. © Current Science Ltd ISBN 1-85922-124-6 ISSN 1040-872X 503504 Office gynecology or have only minimally unbalanced parental genomes. The production of unbalanced gametes during meio- sis as a result of the parental structural chromosomal abnormalities yields either duplication or deficiency of chromosomal segments in the resulting conceptuses as the cause of repetitive abortions. In a compilation of 79 cytogenetic studies of couples with two or more SABs, karyotypes from a total of 8298 women and 7834 men were reviewed [6], At least one member of 2.9% of the couples had a detectable chromosomal abnormality, which is about five-six times the prevalence in the general population. The female partners were twice as likely to have a chromosomal ab­normality than their male counterparts. Among all de­tectable chromosome abnormalities, 50% had balanced reciprocal translocations, 25% had Robertsonian trans­locations and 12% had X-chromosome mosaicism; and inversions and sporadic abnormalities made up the re­mainder (Fig. 1). Normally, only one in 700 phenotypi­cally normal people will have a translocation. Data from an even larger database including 199 studies and over 29 000 persons having two or more miscarriages, found similar distributions [7], Thus, parental chromosomal problems can predispose to RPL in three major ways, translocations, inversions and mosaicism, with transloca­tions being the most likely. Fig. 1. Approximate proportions of karyotypic abnormalities found in individual parents when couples have two or more pregnancy losses [6]. A distinction should be made between the sporadic ge­netic causes of SAB and the causes of RPL attributed to parental chromosomal abnormalities. Genetic causes are easily the most common demonstrable source of sporadic abortions. About 50% of first trimester SABs have chromosomal abnormalities. The distribution of cytogenetic abnormalities found in abortus tissue reveals that most are abnormalities of chromosome numbers (Fig. 2). Abortus specimens from RPL are expected to have a smaller proportion of chromosomal abnormalities than sporadic SABs because RPL abortuses are a select population mostly resulting from predisposing etiologies other than sporadic genetic causes. Even when abortus material from couples with RPL has karyotypic abnor­malities, the abnormalities may be sporadic and do not necessarily imply an increased risk for future pregnan­cies. Therefore, routine cytogenetic analysis of abortus material is not currently indicated. A more fruitful ap­proach is to evaluate parental karyotypes in couples with RPL. Unfortunately, there are no conclusive historical elements to ascertain whether a parent has a translocation or not, including having had a previous phenotypically normal child. Only karyotyping can make the diagnosis. Fig. 2. Approximate distribution of cytogenetic abnormalities found in clinically recognized sporadic spontaneous abortions [8**]. * most common single chromosomal abnormality. To date, only cytogenetic abnormalities have been de­tected in the search for genetic causes of RPL. Highly sensitive molecular genetic techniques now available of­fer the hope of uncovering subchromosomal genetic mutations as sources of RPL, as has been done in an­imal models. Adding support for such single gene pos­sibilities, is a report of couples with RPL and normal karyotypes, who nonetheless have a risk higher than couples without RPL of having offspring with chromo­somal abnormalities [9]. A novel example of a possible single gene defect has come to light based on case re­ports. Evidence shows that faulty centromere function may prematurely allow centromere division. This could be associated with chromosome instability and cell divi­sion errors as a source of RPL [10*]. Further support for the possibility of chromosome instability is provided by studies that have found methotrexate and other clasto-Recurrent pregnancy loss Hatasaka and Varner 505 gens to have increased chromosomal breakage rates in lymphocytes from couples with RPL compared with control couples [11]. Currently there are no tests for sub- chromosomal genetic mechanisms that may cause RPL. Infectious For an infectious agent to cause repetitive pregnancy loss, it would have to be able to exist in a chronic state. Such infections are unusual in otherwise asymp­tomatic women. The T strain mycoplasmas (Ureaplasma urealyticum and Mycoplasma hominis) have been proposed to cause RPL. However, mycoplasmas are highly preva­lent in the cervices of normal women, and the results of antibiotic trials have been mixed. Also, there are still no studies with culture-proven mycoplasma in abortus material. A number of agents including viruses, bacteria, my­coplasmas, and parasites (such as toxoplasma and Chlamy­dia trachomatis) appear to be involved in sporadic losses and potentially severe pregnancy disorders, but their in­volvement in RPL remains unproven. Definitive roles for the utility of cervical and endometrial cultures, serologi­cal testing and alternatively, empiric antibiotic treatment for couples with RPL are lacking. These procedures should only be considered individually if there is clinical suspicion of an infectious agent precipitating recurrent losses. Maternal anatomic abnormalities Both congenital and acquired uterine abnormalities have been associated with RPL. Uterine anomalies have been reported to be present in up to 15% of couples with RPL. The scarcity of specific uterine abnormalities, the urgency to offer treatment when they are discov­ered, and the ethical problem of not offering a treatment that may have benefit, have all slowed the determination of causality of RPL. We are left with trials of therapy which are largely uncontrolled, and therefore we cannot completely assess their attributable benefit. Nevertheless, it is highly probable that incompetent cervix is a cause of RPL. By definition, an incompetent cervix does not allow a pregnancy to reach term, and this implies per­sistent risk with each pregnancy. Prevalence rates from <1% all the way up to 15% have been observed for couples with RPL [12]. Most studies of treatment for incompetent cervix have used women who have already experienced morbidity from an episode of cervical in­competence to serve as their own controls. This study design is vulnerable to selection bias. Nonetheless, a review of 61 such studies suggests that the chance of successful gestation in women with cervical incompe­tence is approximately one in four before having a cer­clage procedure, and about three in four after cerclage [12], Benefit from cerclage has not yet been confirmed in the few randomized, prospective studies conducted. Among the numerous congenital uterine anomalies, septa have commonly been regarded as presenting the highest risk for miscarriage. However, a series of 182 women with uterine anomalies indicates that the unicor- nuate uterus is nearly twice as likely to be associated with SAB (approximately 50%) than septate and bicornuate uteri [13]. However, because septate uteri appear to be four or five times more common than unicornuate uteri, the septate uterus is probably the most likely Mullerian abnormality to be associated with pregnancy loss. There are few data to suggest how many SABs, if any, should be tolerated before reducing a septum. Moreover, there is still no randomized study of expectant management compared with septum reduction. Intrauterine adhesions It seems plausible that intrauterine adhesions can cause RPL. The degree of intrauterine adhesions required to precipitate RPL is unknown, however, so making the decision of which women with RPL to treat is subjec­tive. No single best treatment protocol has surfaced, but March and Israel [14] observed an increased rate of preg­nancy success from a baseline of 16.7% to 87.2% after hysteroscopic lysis of adhesions using microscissors fol­lowed by high-dose estrogen treatment for 2 months. Leiomyomata uteri As with intrauterine adhesions, a cause and effect rela­tionship between leiomyomata and RPL is difficult to confirm. The impression that myomas may be involved in pregnancy loss again comes from nonrandomized studies which present SAB rates after myomectomy com­pared with the before therapy rates in select populations. Data assembled by Buttram and Reiter [15] from 1941 such women reported in seven studies show a reduction in the SAB rate from 41% to 19% following myomec­tomy. In practice, there are no reliable means of predict­ing which clinical presentations are likely to respond to myomectomy. At this time, because of serious pos­sible morbidity, myomectomy for the indication of RPL should only be considered when all other potential eti­ologies have been excluded and suspicion is high that intracavitary distortion from leiomyomata are causative. Diethylstilbestrol Up to 1.5 million women carrying female fetuses are estimated to have taken diethylstilbestrol (DES) to bat­tle against pregnancy loss until 1971 when the US Food and Drug Administration withdrew its approval. Inves­tigations into the effect of DES on reproductive perfor­mance indicate an increased likelihood (up to two times) of a DES-exposed daughter to miscarry [16]. However, data are too few to implicate DES as a certain cause for RPL. Moreover, the presence of cervical abnormalities and hysterosalpingographic abnormalities attributable to506 Office gynecology DES exposure do not predict SAB and cannot be used to dictate management [17]. Preconceptional cerclage to improve reproductive performance in DES-exposed women has not proven cost-effective. The number of DES-exposed women of reproductive age is declining and the determination of their optimal management concerning RPL will probably not be possible. The immune factor Because pregnancies are immunologically privileged, re­searchers have sought to confirm that a breach of this intricate and poorly understood phenomenon may lead to RPL. The bulk of recent research activity into the pathophysiology of RPL has focused on the immune factor. Two different categories of immune dysfunc­tion have been scrutinized. Alloimmune refers to im­munologic disparity between individuals of the same species, whereas autoimmune refers to immunologic ac­tivity against one's self. Alloimmune Although conceptually attractive, there are no definitive tests to prove that altered alloimmunity even exists as a cause of RPL. A popular hypothesis has been that when partners share too many human leukocyte anti­gens, which include the transplantation matching anti­gens, proper maternal recognition of their conceptus may be thwarted such that immunological protection of the fetus cannot be afforded. However, studies look­ing for human leukocyte antigen sharing among couples with RPL have found inconsistent associations. Fur­thermore, none of the alloimmune immunologic tests such as the mixed lymphocyte reaction and antileuko- cytotoxic antibody testing have demonstrated predictive value for pregnancy outcome in couples with RPL [18], The tests are expensive, and presently have no practical value in the management of RPL. Despite the lack of concrete methods to detect, predict or follow the presence of alloimmune RPL, empiric therapies have been attempted. The principal therapy for unexplained RPL, presumed to be of alloimmune origin, has been to immunize the women with partner or third-party leukocytes. The risk of transfusion reac­tions, alloimmunization, infectious disease transmission and untoward fetal effects, implore the use of caution with this treatment. Until recently, there has been no standardization of admission criteria and immunization protocols, so efficacy has been difficult to assess. Meta­analysis of four published, randomized, placebo-con­trolled trials of immunotherapy by Fraser and colleagues [19**] involving 117 treated women and 129 control women demonstrates that there is fair evidence against its use. Another randomized placebo-controlled trial of im­munotherapy for a selected group of women with RPL has since been published and found that therapy bor­dered on but did not achieve a statistical benefit [20*]. An extensive meta-analysis is now being conducted by the American Society of Reproductive Immunology in­cluding both published and unpublished data to further investigate the value of immunotherapy [21*]. Until fur­ther elucidation of the safety and efficacy of immuno­therapy, interested couples should be referred to research centers, and proper informed consent provided. Autoimmune Repetitive pregnancy loss is one presentation of the antiphospholipid syndrome (APS). Other clinical man­ifestations may include venous or arterial thrombo­sis as well as autoimmune thrombocytopenia. At least one elevation of an antiphospholipid antibody as well as a clinical manifestation confirms the diagnosis. The two pertinent antiphospholipid antibodies are lu­pus anticoagulant (LAC) and anticardiolipin (ACL). Whereas ACL can be measured accurately by enzyme- linked immunoabsorbent assays using standards available from the Antiphospholipid Standardization Laboratory (Louisville, Kentucky), LAC requires more fastidious as­says. The results of ACL testing are reported in the semi- quantitative terms ‘negative, low-positive, medium-posi­tive, and high-positive'. Only medium- or high-positive values are associated with APS. When an automated activated partial thromboplastin time (APTT) prolongation is used to detect the pres­ence of LAC, the assay must be depleted of proco­agulant phospholipids for precision. A sample with an APTT prolongation should be mixed in a 1 :1 ratio with normal serum and retested. If the problem is a clotting factor deficiency rather than the presence of LAC, then the APTT will correct. Finally, to confirm that LAC is the cause of a prolonged APTT, the addi­tion of excess phospholipid to the assay should correct the APTT. Antinuclear antibody titers and other autoan­tibody titers do not consistently predict miscarriage. At this time, among the autoantibodies, only LAC and ACL should be obtained during the evaluation of RPL. Autoimmunity may play a role in up to 5-10% of women with RPL. While the pathophysiology is un­clear, decidual vasculopathy and placental thrombosis are common observations in women suspected of au­toimmune RPL. An intriguing line of investigation has demonstrated (in a murine model of APS with RPL), that interleukin-3 serum concentrations are diminished. The addition of exogenous interleukin-3 to these mice appears to avert ischemic placental atrophy and almost completely prevents RPL [22*]. The selective use of cytokines may one day prove to be the most effective treatment for APS in humans. In the meantime, treatment of APS-mediated RPL has been attempted with various combinations of heparin, low-dose aspirin (81 mg/day), glucocorticoids, and in­travenous immunoglobulins. To date, the lack of appro­priate controls and the inconsistent inclusion criteria and treatment protocols has made a treatment consensus im­ Recurrent pregnancy loss Hatasaka and Varner 507 possible. In has been assumed that it is rare for a woman with APS and RPL to reproduce successfully without treatment. Various treatment protocols report successful pregnancies 55-85% of the time. Prednisone and low- dose aspirin, and heparin with or without aspirin have been popular regimens. A critical consideration is that women with APS and their fetuses remain at high risk for complications despite treatment. Branch et al. [23] found that one-half of treated women with APS devel­oped pre-eclampsia and one-half suffered fetal distress, while over one-third delivered prematurely and another one-third had small-for-gestational-age infants. Five per­cent of the cohort also developed thrombosis which in­cluded a stroke. One treatment combination that may be particularly hazardous is a combination of heparin and glucocorticoids due to the potential additive risk of osteoporosis [24*]. _ Endocrine factors Poorly controlled systemic endocrinopathies, such as diabetes and thyroid dysfunction, can probably cause pregnancy loss. However, because severe manifestations of these disorders are found infrequently in populations of women with RPL and are clinically obvious, their overall contribution to RPL is negligible. As there is no evidence that asymptomatic endocrine problems lead to RPL, testing serum thyroid, glucose and prolactin con­centrations is unproductive. Luteal phase inadequacy (LPI) refers to an endometrium that is physiologically incapable of allowing normal im­plantation or pregnancy maintenance, and has been re­ported to be present in as many as 40% of cases of RPL. Diagnosis is perplexing because neither serum proges­terone concentrations or late luteal phase endometrial biopsies are diagnostically sensitive or specific. Up to 3-4% of fertile controls will have false-positive diag­noses of LPI by the standard endometrial biopsy criteria of late luteal phase biopsies out of phase by at least 2 days in two separate cycles. Because LPI is sometimes diag­nosed by endometrial biopsy while concomitant serum progesterone concentrations are normal, and because a myriad of other factors beside progesterone deficiency may contribute to an inadequate endometrium, the en­dometrial biopsy remains the gold standard test [25]. At this time there are no randomized controlled trials which have indicated a benefit from progesterone treat­ment for LPI. Balasch et al. [26] in a small, non-ran- domized, non-placebo-controlled trial, found a statis­tically significant advantage of progesterone supposito­ries over no treatment for RPL with LPI. Progesterone is usually given as 25 mg suppositories twice daily be­ginning on the third day after ovulation. If progesterone supplementation is considered, it should only be used in women with RPL who have demonstrated two out-of­phase endometrial biopsies, with the understanding that it is still considered empiric therapy. Counseling Several generalizations are pertinent when counseling couples with RPL. First, the older the woman, the higher the risk of even chromosomally normal losses. Second, the later in gestation that previous losses occur, the greater the risk for future pregnancy loss. Third, the outcomes of previous pregnancies influence the risk of future miscarriage. A previous live birth confers a better prognosis than when all previous pregnancies have been losses. Fourth, early ultrasonographic confirmation of fe­tal heart activity is not as reliable in predicting successful outcomes for couples with RPL compared with cou­ples without a history of repetitive losses. Recent reports find that in women with RPL, early ultrasound detec­tion of fetal heart activity is associated with only about an 80%) chance of a live-born whereas ordinarily a live embryo noted by first trimester ultrasonography is asso­ciated with a >96% rate of a live birth [27,28*]. Fifth, a recent report finds that a history of oligomenorrhea may predict a higher risk for miscarriage [29*]. This adds credibility to the claims that elevated preconcep- tional follicular phase luteinizing hormone serum con­centrations may signal a higher risk of miscarriage, be­cause oligomenorrheic women commonly have higher circulating luteinizing hormone concentrations. How­ever, Quenby and Farquharson [29*] could not confirm that higher luteinizing hormone profiles correspond to RPL, and the point remains controversial. Sixth, it ap­pears that the more pregnancy losses a couple have, the higher their chance may be of birth defects in future live- born children [30*]. This adds to the dispiriting evidence that couples with RPL may also have increased risks for infertility, prematurity and small-for-gestational-age in­fants [31]. Even when the prognosis is not good, couples will find the information provided by their physicians helpful in guiding their future decision-making. Evaluation and management A thorough history should include a detailed obstetri­cal summary, noting the gestational ages of pregnancy losses and interspersed live births. Family genetic and miscarriage histories are important as well as a com­plete history of environmental exposures. The possibility of APS should also be explored. Depending on the indi­vidual clinical situation, it is not unreasonable to begin an evaluation following two consecutive SABs. A judicious evaluation is outlined in Table 1. Still, the cost of a ‘com­plete' workup may be over US$ 2000, and the chance of finding an abnormality is reported to be less than 50% in some series. When RPL is unexplained, frustration may tempt care­takers to offer some form of treatment. Unfortunately, treatments for most suspected etiologies of RPL have not been validated with scientific rigor, so optimal manage­ment includes resisting the urge to offer treatments that508 Office gynecology Table 1. A suggested evaluation for recurrent pregnancy loss. History 1) Obtain a precise obstetric history noting documentation of gestational ages and patterns of pregnancy losses. 2) Search for historical evidence of antiphospholipid syndrome such as arterial or venous thromboses and thrombocytopenia. 3) Scrutinize family histories for patterns of RPL, and investigate consanguinity among family members. 4) Ascertain potential adverse environmental exposures. 5) Inquire about possible gynecologic and obstetrical infections. Physical examination 1) Perform a general physical examination noting in particular signs of autoimmune dysfunction and advanced endocrinopathy. 2) During the pelvic examination, look for evidence of Mullerian abnormalities and diethylstilbestrol exposure. Testing 1) Hysterosalpingogram or hysteroscopy. 2) Late luteal phase endometrial biopsy (repeat in a subsequent cycle if out of phase). 3) Karyotyping for both partners. 4) Lupus anticoagulant and anticardiolipin antibodies. 5) Other tests if indicated by history and physical examination. may carry undue risks. It is wise to factor in the strength of association of a presumed cause of a couple's RPL before recommending an empiric treatment. Couples with RPL need compassionate counseling. The fact that couples with even three consecutive unexplained losses have a reasonable chance of having a live-born child on a subsequent attempt, 50-75%, despite the lack of specific treatment, can be quite comforting. Several uncontrolled studies indicate that emotional support and counseling alone may favor successful outcomes [29*]. Conclusion Systematic evaluation will identify a cause for RPL ap­proximately half the time. Many treatments in current use for specific causes of RPL have been incompletely validated, and therefore require judgement and cau­tion before implementation. When a cause for RPL cannot be determined, complex, expensive testing for possible alloimmune causes has not been useful. Like­wise, the value of empiric immunotherapy still awaits confirmation and should be relegated to research centers. Attention to previous menstrual and reproductive histo­ries adds invaluable information for counseling couples with RPL. A realistic appraisal of unexplained RPL cur­rently favors conservative management based on accurate counseling combined with the unlimited availability of emotional support. References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as: • of special interest •• of outstanding interest 1. Schnoor TM, Crajewski BA, Hornung RW, Thun MJ, Egeland CM, Murray WE, Conover DL, Halperin WE: Video Display Terminals and the Risk of Spontaneous Abortion. N Engl J Med 1991, 324:727-733. 2. Infante-Rivard C, Fernandez A, Gauthier R, David M, Etienne Rivard G: Fetal Loss Associated with Caffeine Intake Before and During Pregnancy. JAMA 1993, 270:2940-2943. 3. Mills JL, Holmes LB, Aarons JH, Simpson JL, Brown ZA, Jovanovic-Peterson LG, Conley MR, Graubard Bi, Knopp RH, Metzger BE: Moderate Caffeine Use and the Risk of Sponta­neous Abortion and Intrauterine Growth Retardation. JAMA 1993, 269:593-597. 4. Kline J, Stein ZA, Susser M, Warburton D: Smoking: A Risk Factor for Spontaneous Abortion. N Engl J Med 1977, 297:793-796. 5. Kline J, Shrout P, Stein ZA, Susser M, Warburton D: Drinking During Pregnancy and Spontaneous Abortion. Lancet 1980, ii:176-180. 6. Tharapel AT, Tharapel SA, Bannerman RM: Recurrent Preg­nancy Losses and Parental Chromosome Abnormalities: A Review. Br J Obstet Gynaecol 1985, 92:899-914. 7. De Braekeleer M, Dao TN: Cytogenetic Studies in Couples Experiencing Repeated Pregnancy Losses. Hum Reprod 1990, 5:519-528. 8. Mishell DR, Jr: Recurrent Abortion. J Reprod Med 1993, •• 38:250-259. An up to date review of the causes and treatment of RPL which empha­sizes practical clinical management. 9. Drugan A, Koppitch FC, Williams JC, Johnson MP, Moghissi KS, Evans Ml: Prenatal Genetic Diagnosis Following Recurrent Early Pregnancy Loss. Obstet Gynecol 1990, 75:381-384. 10. Bajnoczky K, Gardo S: 'Premature Anaphase' in a Couple with • Recurrent Miscarriages. Hum Genet 1993, 92:388-390. This study finds an increased frequency over controls of premature cen­tromere separation during mitosis in lymphocyte cultures from both partners of couples with RPL. This finding is a good example of a po­tential subchromosomal abnormality which could lead to RPL. Single gene mutations may contribute to the large proportion of unexplained RPL. Research into molecular abnormalities causing RPL promises to be the next advance in understanding RPL. 11. Velissariou V, Lyberatou E, Antonopoulou E, Polymilis C: Chro­mosome Breakage in Individuals with Single-Cell Structural Aberrations and Habitual Abortions. Gynecol Obstet Invest 1993, 36:71-74. 12. Shortle B, Jewelewicz R: Clinical Aspects of Cervical Incompe­tence. Chicago: Year Book Medical Publishers; 1989. 13. Heinonen PK, Saarikoski S, Pystynen P: Reproductive Perfor­mance of Women with Uterine Anomalies. Acta Obstet Gy­necol Scand 1982, 61:157-162. 14. March CM, Israel R: Gestational Outcome Following Hysterop- scopic Lysis of Adhesions. Fertil Steril 1981, 36:455-459. 15. Buttram VC, Reiter RC: Uterine Leiomyomata: Etiology, Symp­tomatology, and Management. Fertil Steril 1981, 36:433-445. 16. Herbst AL, Senekjian EK, Frey KW: Abortion and Pregnancy Loss Among Diethylstilbestrol-Exposed Women. Semin Reprod Endocrinol 1989, 7:124-239. 17. Kaufman RH, Noller K, Adam E, Irwin J, Gray M, Jefferies JA, Hilton J: Upper Genital Tract Abnormalities and PregnancyRecurrent pregnancy loss Hatasaka and Varner 509 Outcome In Diethylstibestrol-Exposed Progeny. Am J Obstet Gynecol 1984, 148:973-984. 18. Coulam CB: Immunologic Tests in the Evaluation of Reproduc­tive Disorders: A Critical Review. Am I Obstet Gynecol 1992, 167:1844-1851. 19. Fraser EJ, Grimes DA, Schulz KF: Immunization as Therapy for •• Recurrent Spontaneous Abortion: A Review and Meta-Analysis. Obstet Gynecol 1993, 82:854-859. This meta-analysis does not support the current use of immunotherapy for RPL. However, only four published randomized controlled studies were included, each of which used different protocols, thus uncertainty re­mains regarding the value of immunotherapy. 20. Gatenby PA, Cameron K, Simes RJ, Adelstein S, Bennett MJ, • Jansen RPS, Shearman RP, Stewart G), Whittle M, Doran TJ: Treatment of Recurrent Spontaneous Abortion by Immuniza­tion With Paternal Lymphocytes: Results of a Controlled Trial. Am J Reprod Immunol 1993, 29:88-94. Near statistical benefit was shown in this randomized controlled trial of paternal cell immunotherapy involving 19 sequentially paired couples. Their meta-analysis of all four published randomized controlled trials us­ing intravenous paternal cell immunization showed an overall beneficial effect with a livebirth odds ratio of 2.6 with a 95% confidence interval of 1.6 to 4.4. 21. Coulam CB: News and Views: Report From the Ethics Commit- • tee for Immunotherapy. Am J Reprod Immunol 1993 30:45-47. The American Society of Reproductive Immunology has initiated a meta­analysis of all available trials of immunotherapy for RPL. The inclusion of unpublished trials is hoped to decrease the possibility of publication bias, because trials showing benefit generally have a better chance of publication. Also, the increased size of the database is hoped to clarify the value of immunotherapy for RPL. 22. Fishman P, Falach-Vaknine E, Zigelman R, Bakimer R, Sredni • B, Djaldetti M, Shoenfeld V: Prevention of Fetal Loss in Experimental Antiphospholipid Syndrome by in Vivo Admin­istration of Recombinant lnterleukin-3. J Clin Invest 1993 91:1834-1837. Mice immunized with anticardiolipin antibodies acquire the classic man­ifestations of APS, including RPL. Exogenous interleukin-3 reverses the manifestations. The potential importance of cytokines or growth factors, or both, in the pathophysiology of APS is highlighted. The use of cy­tokine therapy in humans remains a tempting yet premature undertaking for APS-induced RPL. 23. Branch DW, Silver RM, Blackwell JL, Reading JC, Scott JR: Out­come of Treated Pregnancies in Women with Antiphospholipid Syndrome: An Update of the Utah Experience. Obstet Gynecol 1992, 80:614-620. 24. Menashe Y, Ben-Baruch G, Greenspoon JS, Carp HJA, Rosen • DJD, Mashiach S, Many A: Successful Pregnancy Outcome with Combination Therapy in Women with the Antiphospho­lipid Antibody Syndrome. J Reprod Med 1993, 38:625-629. This case series includes four patients with APS who had previously mis­carried despite low-dose aspirin and heparin. A more aggressive regimen of aspirin, dipyridamole, prednisone and heparin (or warfarin) resulted in all four women delivering live births, but one developed a vertebral com­pression fracture. 25. Wentz AC: Endocrine Aspects of Recurrent Early Fetal Wastage. The Role of Luteal Phase Inadequacy. Acta'Eur Fertll 1992, 23:263-269. 26. Balasch J, Creus M, Marquez M, Burzaco I, Vanrell JA: The Sig­nificance of Luteal Phase Deficiency on Fertility: A Diagnostic and Therapeutic Approach. Hum Reprod 1986, 1:145-147. 27. van Leeuwen I, Branch DW, Scott JR: First-Trimester Ultra­sonography Findings in Women with a History of Recurrent Pregnancy Loss. Am J Obstet Gynecol 1993, 168:111-114. 28. Opsahl MS, Pettit DC: First Trimester Sonographic charac- • teristics of Patients with Recurrent Spontaneous Abortion. J Ultrasound Med 1993, 12:507-510. This is one of a number of recent studies which make the important prac­tical point that first trimester confirmation of fetal heart activity in women who have RPL is not a reliable predictor of a successful pregnancy out­come. 29. Quenby SM, Farquharson RG: Predicting Recurring Miscar- • riage: What is Important? Obstet Gynecol 1993, 82:132-138. A total of 203 consecutive couples attending a Liverpool miscarriage clinic were studied to identify predictive factors for future miscarriages. The most significant predictor was a history of oligomenorrhea but more than three previous miscarriages, maternal age 30 years or more, no pre­vious live births, low luteal phase estradiol levels and the presence of antiphospholipid antibodies also helped predict future miscarriage. 30. Khoury MJ, Erickson JD: Recurrent Pregnancy Loss as an • Indicator for Increased Risk of Birth Defects: A Population- Based Case-Control Study. Paediatr Perinat Epidemiol 1993, 7:404-416. This case-control study finds a significantly higher risk of birth defects among offspring of couples who have had RPL over normal control cou­ples. However, no specific birth defects associated with RPL could be distinguished. Furthermore, both the cases and controls had birth defect rates within the usual range of 3-4% reported for the general population. Thus, the clinical significance of the finding is debatable. 31. Beard RW: Clinical Associations of Recurrent Miscarriage. In: Early Pregnancy Loss: Mechanisms and Treatment. Edited by Beard RW, Sharp F. London: Springer-Verlag; 1988:3-8. Harry H. Hatasaka, Division of Reproductive Endocrinology, and Michael W. Varner, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Utah School of Medicine, 50 N. Medical Drive, 2B200 Salt Lake City, UT 84132, USA.