Recurrent Miscarriage

N Engl J Med 2010; 363:1740-1747, 28/10/2010

A 32-year-old woman presents for evaluation after three consecutive miscarriages. Before these miscarriages, she had one successful pregnancy. The patient reports no medical problems or previous surgeries. In her first miscarriage, uterine bleeding and cramping started in the 7th week of gestation, and an ultrasonographic evaluation showed an empty gestational sac of 6 weeks' size. In her next pregnancy, she presented with a miscarriage in the 8th week. The third miscarriage was identified by ultrasonographic testing as a dead embryo measuring 7.5 weeks' size. How should her case be evaluated and managed?

The Clinical Problem

Miscarriage is the spontaneous loss of the conceptus before 20 weeks' gestation. In 10% of pregnancies, miscarriage is a clinically recognized event, and in another 20% of pregnancies, it is manifested only as a transient elevation of the level of human chorionic gonadotropin before or near menses.1

A majority of miscarriages that occur before 10 weeks' gestation are due to chromosomal aneuploidies arising from noninherited, new nondisjunctional events; these events are more frequent in very early miscarriages.2 The rate of miscarriage increases with a maternal age of less than 18 years or an age of 35 years or more, an increasing number of previous miscarriages, and increasing parity.3 The sharp increase in the rate of miscarriage in women 35 years of age or older is due in part to increasing rates of aneuploidy in association with older oocytes.2,3

Recurrent miscarriage, defined as the loss of three or more consecutive pregnancies, occurs in approximately 1% of couples attempting to bear children.4 This rate is higher than would be expected if recurrences were due solely to chance, suggesting an underlying predisposition in some couples. Many experts consider two consecutive losses as sufficient for the diagnosis of recurrent miscarriage5,6 because the recurrence rate is similar to that after three losses. Two consecutive miscarriages occur in up to 5% of reproductively active couples and an even higher proportion of women 35 years of age or older. For many couples, evaluation does not reveal a cause, and in some cases, it leads to an unrelated finding or unnecessary treatment.

Most women with recurrent miscarriage have recurrent early miscarriage — either the failure of the development of an ultrasonographically recognizable embryo or embryonic death before 10 weeks' gestation; however, signs and symptoms of the miscarriage (uterine bleeding and cramping) may appear after 10 weeks. A substantial proportion of recurrent early miscarriages are due to conceptus aneuploidy, although it is less common than with sporadic miscarriage.7 A smaller proportion of women with recurrent miscarriage have late miscarriage (death of the conceptus at or after 10 weeks' gestation), usually before 15 to 16 weeks.8 Causes and recurrence rates for late miscarriage may differ from those for early miscarriage.

Recurrent miscarriage is a distressing problem. Frequently, the cause is elusive or multifactorial, and misinformation abounds, giving rise to frustration for the couple and the physician.

Strategies and Evidence

Evaluation

The physician should determine the

1. nature of previous pregnancy losses, especially with regard to the actual gestational age at the time of the demise of the conceptus (rather than the onset of symptoms).

2.The medical and obstetrical history should cover any features suggestive of the antiphospholipid syndrome (e.g., thrombosis or fetal death)

3.or possible uterine malformation (e.g., breech presentation).

4.Poorly controlled diabetes9

5.or thyroid disease10 in early pregnancy is associated with miscarriage, but well-controlled disease is not; testing for these conditions is not recommended for the sole indication of recurrent miscarriage.

6. Obesity,11 cigarette smoking,12 alcohol use,13 and moderate-to-heavy caffeine use14 may be associated with sporadic miscarriage, but relationships between these conditions and recurrent miscarriage have not been extensively studied and are uncertain. There is no good evidence that physical activity, including sexual activity and exercise, causes miscarriage.

Five to 15% of women with recurrent miscarriage have clinically significant antiphospholipid antibody titers, as compared with 2 to 5% of unselected obstetrical patients.15

****Because results may be transiently positive after infection, the antiphospholipid syndrome should be diagnosed only when two tests performed 12 or more weeks apart are positive.

7. It also has become common to screen patients with recurrent miscarriage for thrombophilias, although it is uncertain whether this screening is warranted. Case–control studies have shown a modest association (odds ratios of 2 to 3) between recurrent miscarriage and thrombophilias such as the factor V Leiden mutation and the prothrombin G20210A mutation.16 However, associations have tended to be stronger for fetal deaths, such as stillbirths after 20 weeks' gestation, than for recurrent early losses, and several large, prospective cohort studies have not shown significant associations between thrombophilias and sporadic pregnancy loss.17-19 Furthermore, many thrombophilias are common in the general population, and the vast majority of women with thrombophilias have normal pregnancies.

8. Uterine malformations, most commonly arcuate and septate uteruses are detected in 10 to25% of women with recurrent miscarriage but in only 5% of controls,20 and evaluation of the uterine cavity (primarily to look for a septum) is recommended by professional organizations in women with recurrent miscarriage (Table 1Table 1Evaluation of a Woman with Recurrent Miscarriage.). Vascular insufficiency is thought to underlie miscarriage in the case of septate uterus.

8.One partner (more often the woman) has a balanced chromosomal rearrangement in 3 to 6% of cases of recurrent miscarriage.2,21 The most common abnormality is a translocation, either reciprocal or robertsonian (involving two homologous or nonhomologous acrocentric chromosomes [chromosomes 13, 14, 15, 21, and 22]) Parental karyotyping is expensive. Because available treatment (in vitro fertilization [IVF] with preimplantation genetic diagnosis) has not been shown to improve the outcome, as compared with that of spontaneous conception (see below), some couples may choose to forgo parental karyotype testing.

Though controversial,

9.some experts recommend karytoype analysis of the conceptus in couples with recurrent miscarriage to avoid unnecessary evaluation and treatment and because an aneuploid conceptus indicates a somewhat greater likelihood of success with a subsequent pregnancy.22 If karyotype analysis is not possible because of cell-culture failure, array comparative genomic hybridization may be considered. This test provides most of the information provided by cytogenetic analysis, as well as data regarding smaller regions of chromosomes than those identified by traditional karyotype analysis. Array comparative genomic hybridization cannot identify balanced translocations.

Although any bacterial or viral infections that spread to the uterus may cause sporadic miscarriage, there is no proven infectious cause of recurrent miscarriage. Thus, screening tests for ureaplasma, chlamydia, and other infectious agents are not recommended in the evaluation of recurrent miscarriage.

Similarly, although a deficiency of progesterone during the luteal phase (resulting in inadequate endometrial preparation) has been proposed as a cause of miscarriage, screening for a deficiency would be problematic, because there is no clear threshold for normal progesterone values, which vary considerably from one person to another, and progesterone measurements are not closely correlated with other markers of endometrial function.23,24 Thus, diagnosis of this putative disorder by means of timed endometrial histologic analysis or measurement of progesterone levels is not recommended.

Immunologic testing such as HLA typing tests for immune reactivity in the mother against the father or determination of leukocyte populations has been recommended by some clinicians on the basis of the hypothesis that recurrent miscarriage may be caused by maternal immunologic rejection of the conceptus as a semi-allograft.

10.However, most experts do not support such testing, since unequivocal evidence of an alloimmune cause of recurrent miscarriage in humans is lacking, and there is no good evidence that screening for alloimmunization, with treatment in identified cases, improves clinical outcomes.5,6,25

Management

The prognosis for women with a history of recurrent miscarriage is often favorable, even without intervention. In one of the largest treatment trials involving women with idiopathic recurrent miscarriage,26 which included women with 4.2 consecutive miscarriages and an average age of 32.7 years, the placebo group had a live-birth rate of 65%. Other studies suggest even better outcomes with frequent prenatal visits and visualization of the pregnancy on ultrasonographic examination, beginning early in pregnancy and extending throughout the first trimester,25,27 perhaps owing to maternal reassurance.

Anticoagulant Therapy

1. Repeatedly positive antiphospolipid antibody tests need,

prophylactic doses of unfractionated heparin (e.g., 5000 U subcutaneously twice daily) and low-dose aspirin, as compared with aspirin alone.28,29 , two clinical trials showed improved live-birth rates with the use of

This strategy has become standard treatment for recurrent miscarriage due to the antiphospholipid syndrome, yet two more recent randomized trials involving women with this syndrome showed no significant improvement in the live-birth rate with the use of prophylactic doses of low-molecular-weight heparin and low-dose aspirin, as compared with aspirin alone.30,31 Unlike the earlier positive studies, the negative studies included women with low levels of antibodies, used low-molecular-weight heparin (rather than unfractionated heparin), and showed high rates of live births among women who received only low-dose aspirin. Improved live-birth rates were also reported in some32 but not other33 studies of heparin thromboprophylaxis (with or without aspirin) in women with a heritable thrombophilia and recurrent miscarriage or fetal death. Thus, the role of this treatment specifically for the prevention of recurrent miscarriage remains controversial. Women with a history of thrombosis in whom the antiphospholipid syndrome or a heritable thrombophilia is diagnosed should receive an appropriate dose of unfractionated heparin or low-molecular-weight heparin.34

For women with unexplained recurrent miscarriage, it has become common practice to provide empirical treatment with low-dose aspirin, prophylactic doses of low-molecular-weight heparin, or both. However, two recent controlled, randomized trials involving women with unexplained recurrent miscarriage showed no improvement in live-birth rates with the use of low-dose aspirin alone35 or low-molecular-weight heparin and low-dose aspirin.35,36, Thus, this approach is not recommended outside of clinical trials.

Treatment of Uterine Anomalies

Most experts recommend hysteroscopic resection of a uterine septum in women with recurrent miscarriage; this recommendation is based on retrospective and uncontrolled observational studies and case series.37,38 However, data from well-designed trials supporting this practice are lacking, and septa are also detected among women with normal pregnancies.39 The high likelihood of successful pregnancy outcomes in nulliparous women with septa, as well as good outcomes without surgery, weigh against surgical correction of a uterine septum in nulligravid women.40 Since most women with bicornuate or unicornuate uteruses have good obstetrical outcomes, and surgical repair of these abnormalities is more invasive than septum resection and associated with a higher risk of complications, repair of these abnormalities is not generally advised.

Management of Genetic Abnormalities

Although the prognosis varies depending on the abnormality, pregnancy outcomes among couples with a balanced translocation are better than one might anticipate. Overall, there is up to a 70% chance of a live birth with no interventions.41 The risk of a live-born infant with a translocation trisomy is low, generally less than 1%. IVF with preimplantation genetic diagnosis has been used in an attempt to avert such outcomes. However, the high likelihood of a karyotypically normal or balanced offspring in the absence of this intervention41,42 makes its usefulness questionable.

To increase the likelihood of successful pregnancy, preimplantation genetic screening of biopsied blastomeres during IVF has been recommended in order to prevent the transfer of karyotypically abnormal embryos in couples with normal karyotypes. However, this approach is not supported by appropriately conducted randomized trials. Also, in a randomized trial involving women undergoing IVF who were at increased risk for miscarriage because of their age, the use of IVF with preimplantation genetic screening did not significantly increase the likelihood of successful live birth.43 This finding may be explained in part by chromosomal mosaicism in early embryos, with discordance between the chromosomes in a biopsied blastomere and those in the resulting fetus. Given these data, and in the absence of convincing evidence of a benefit, preimplantation genetic screening in couples with idiopathic recurrent miscarriage is not recommended.44 The use of array comparative genomic hybridization, which has finer resolution, with biopsied polar bodies and assessment of the karyotype contribution from the oocyte may potentially improve the success of preimplantation genetic screening of blastomeres, but trials are needed to evaluate its use in couples with recurrent miscarriage.

Immunologic Intervention

Although alloimmunity has been hypothesized to be a possible cause of recurrent miscarriage, a randomized trial of paternal leukocyte immunization showed no improvement in the rate of live births,26 and the Food and Drug Administration now requires investigational-drug approval for the use of this biologic product. In addition, a Cochrane review of seven trials showed that intravenous immune globulin had no significant benefit with regard to the live-birth rate among couples with recurrent miscarriage.45

Areas of Uncertainty

Some experts have questioned the association of recurrent miscarriage with antiphospholipid antibodies, and a recent international task force concluded that further studies using standardized antiphospholipid-antibody testing with confirmation in a central laboratory would be informative.46 Anticoagulant therapy is commonly prescribed for women with a history of recurrent miscarriage and antiphospholipid antibodies, and it is recommended in published guidelines, although discrepant results of recent trials have raised controversy in the field.

The treatment of women with recurrent miscarriage who also carry a heritable thrombophilia has not been adequately studied, and thus no recommendations regarding treatment efficacy can be made; results of a completed treatment trial (Low Molecular Weight Heparin and/or Aspirin in Prevention of Habitual Abortion; ClinicalTrials.gov number, NCT009596211) that included women with recurrent miscarriage and heritable thrombophilias are awaited. Recent studies did not show that treatment with low-molecular-weight heparin and low-dose aspirin in women with recurrent miscarriage was beneficial.35,36 Additional trials of low-molecular-weight heparin that include women with unexplained recurrent miscarriage are currently under way (Prevention of Unexplained Recurrent Abortion by Enoxaparine [NCT00740545] and Effectiveness of Dalteparin Therapy as Intervention in Recurrent Pregnancy Loss [NCT00400387]).

It is likely that genetic abnormalities that are not detectable by karyotype testing play a role in some cases of sporadic or recurrent miscarriage. Several single-gene disorders are associated with both sporadic and recurrent miscarriage. Genetic abnormalities that are reported to be associated with recurrent miscarriage include skewed X-chromosome inactivation, confined placental mosaicism, abnormalities in sperm DNA, and small deletions or additions detected by means of array comparative genomic hybridization (see Figure 1 in the Supplementary Appendix, available with the full text of this article at NEJM.org).47,48

Multiple intramural and submucosal fibroids are associated with an increased risk of pregnancies ending in miscarriage.49,50 The benefits of myomectomy are uncertain and supported only by reports of better outcomes after surgery than before surgery. More extensive surgery to remove uterine myomas involving laparoscopy or laparotomy poses hazards for subsequent pregnancies (e.g., a risk of uterine rupture), and its effect on the likelihood of subsequent miscarriage is uncertain.49,50

The risk of recurrent miscarriage may be increased among women with the polycystic ovary syndrome.51 However, the rate of subsequent miscarriage among women with recurrent miscarriage and polycystic ovaries detected on ultrasonographic examination is not higher than that among women with recurrent miscarriage who do not have polycystic ovaries,52 and rates of miscarriage after ovulation induction are not consistently higher among women with the polycystic ovary syndrome than among other women.53 Suppression of luteinizing hormone with long-acting agonists of gonadotropin-releasing hormone in women with the polycystic ovary syndrome, recurrent miscarriage, and luteinizing hormone hypersecretion did not improve the rate of successful pregnancy in a randomized, controlled trial.51 Some studies have shown greater insulin resistance in women with recurrent miscarriage than in women with a similar body-mass index and successful pregnancies, even in the absence of the polycystic ovary syndrome.54 Although data are lacking regarding the use of metformin in women with recurrent miscarriage, its use has not been shown to reduce the risk of miscarriage among women with the polycystic ovary syndrome.55

Guidelines

The American College of Obstetricians and Gynecologists,5 the Royal College of Obstetricians and Gynaecologists,25 and the American Society for Reproductive Medicine6 have published guidelines for the evaluation and management of recurrent miscarriage. The recommendations provided here are generally consistent with these guidelines (Table 1).

Conclusions and Recommendations

For a patient with recurrent miscarriage such as the woman described in the vignette, we recommend a medical history taking (with attention to a personal or family history of thrombosis, autoimmune disease, or clinically overt diabetes or thyroid disease) and testing for antiphospholipid antibodies, uterine abnormalities, and possibly parental karyotype abnormalities. We recommend the administration of heparin at a prophylactic dose and low-dose aspirin during the next pregnancy in women with the antiphospholipid syndrome, though we acknowledge that recent trials have raised questions regarding the need for heparin in women who do not have a history of thrombosis. Given recent trials showing no benefit of anticoagulant therapy in women with idiopathic recurrent miscarriage, we do not currently recommend anticoagulant therapy in other women with recurrent miscarriage. Most experts recommend that a uterine septum or intrauterine abnormalities be treated by means of hysteroscopy before a next pregnancy is attempted, but data from randomized trials are lacking to show that this strategy improves outcomes. Some experts recommend IVF with preimplantation genetic diagnosis for couples in whom one or both persons is a carrier of a balanced translocation, but this procedure is expensive and may not improve the live-birth rate, as compared with spontaneous conception. Obtaining karyotypes of the conceptus from subsequent miscarriages may have an emotional benefit and modest prognostic usefulness, but the effect of this information on clinical decision making is uncertain. Couples should be informed that good evidence is lacking to support several commonly used interventions for recurrent miscarriage and that observation is a reasonable strategy, since without intervention, a subsequent pregnancy will result in a live birth for about two thirds of couples.

An audio version of this article is available at NEJM.org.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.

No potential conflict of interest relevant to this article was reported.

Source Information

From the University of Utah Health Sciences Center (D.W.B., M.G., R.M.S.) and Intermountain Healthcare (D.W.B.) — both in Salt Lake City.

Address reprint requests to Dr. Branch at the University of Utah Health Sciences Center, 30 N. 1900 East, 2B200 SOM, Salt Lake City, UT 84132, or at ware.branch@hsc.utah.edu.