Recurrent Pregnancy Loss

Approximately 15 to 20% of clinically recognized pregnancies result in a spontaneous abortion (SAB). If all pregnancies are taken into consideration, including those that are too early to be clinically recognized, it is estimated that 30% fail. Recurrent pregnancy loss (RPL) is defined as the occurrence of two consecutive spontaneous abortions or, more strictly, three or more consecutive losses. About 1 to 2% of couples have 3 or more consecutive losses and as many as 5% of couples have at least 2 losses. Primary RPL refers to couples that have never had a live birth, while secondary RPL refers to those who have had repetitive losses following a successful pregnancy.

Genetic Causes

At least 50%, and possibly as many as 77% of first trimester abortions are chromosomally abnormal. Autosomal trisomies account for most of those abnormalities, while monosomy X is the single most common chromosomal abnormality found in spontaneous abortions (20-25%). No apparent cause is found in most of the remainder. In couples with RPL the causes may be categorized as genetic, anatomic, endocrine, immune, age-related, environmental, and unexplained.

Although trisomies account for the majority of first trimester losses, the risk of recurrent aneuploidy in subsequent pregnancies is 1% or less. Therefore, recurrent losses due to genetic abnormalities are usually related to abnormalities of chromosome structure rather than abnormalities of chromosome number. The most common structural abnormality found in RPL is a translocation. Individuals with a balanced translocation, a condition in which material is exchanged between two or more chromosomes, may be phenotypically normal, but the fetus may have genetic deficiencies or duplications after meiotic segregation. This may result in a genetic complement incompatible with successful fetal development. Translocations may be robertsonianor reciprocal. In robertsonian translocations, acrocentric chromosomes (13, 14, 15, 21, 22, and 24) are fused at their centromeres. In reciprocal translocations, the centromeres are not involved. About 60% of translocations are reciprocal and 40% are robertsonian. Women are twice as likely as men to have a balanced translocation. Inversions are rearrangements within the chromosome in which a sequence of genes is reversed. Inversions are much less likely than translocations to be found in couples experiencing RPL.

Uterine Abnormalities

The incidence of Mullerian anomalies in the general population has been found to be about 2 to 3%, but 14% in women with RPL. The septate uterus is by far the most common. Abnormalities other than those induced by DES generally occur before the 12th week of fetal life, as a result of medial fusion defects. The standard classification system currently in use is that of the American Society of Reproductive Medicine:

Class I Hypoplasis/agenesis
Class II Unicornuate
Class III Didelphus
Class IV Bicornuate
Class V Septate
Class VI Arcuate
Class VII DES drug related

There is an increase in SAB in women with unicornuate and didelphic uteri. With unicornuate uteri, the risk of SAB is 44% and with didelphic uteri, the risk is 36%. Less than 1% of fertile women have bicornuate uteri, with the reported risk of SAB ranging from 24 to 84%. It is difficult to determine the risk with accuracy since most studies do not carefully distinguish between bicornuate and septate uteri. Although the two conditions may have identical appearances on hysterosalpingography (HSG), many studies base the diagnosis on HSG findings alone. The septate uterus is the most common congenital uterine anomaly and is reported to have the highest risk of SAB, over 60%. Mullerian anomalies are diagnosed and distinguished from one another by various modalities including HSG, ultrasonography, MRI, laparoscopy, and hysteroscopy. Transvaginal ultrasonography is capable of defining the abnormality, with more invasive means such as laparoscopy and hysteroscopy sometimes being required.

Most women with uterine fibroids have successful pregnancies, however, the effect of the fibroids on the course of pregnancy and the risk of loss depends on the size, number, and location of the fibroids. Submucosal and intracavitary fibroids are felt to induce the greatest risk of loss, possibly due to pressure effect on the endometrial cavity and alteration of blood flow.

Endocrine Causes

Luteal phase defect, or deficient progesterone production in the luteal phase, has long been suspected to be a cause of RPL. Recent data suggests, however, that the presence of a low luteal phase progesterone level in a woman with RPL does not predict a greater likelihood of future loss than women with a normal progesterone level. If a progesterone level is low during pregnancy, it is likely to be a result rather than the cause of a failing pregnancy. Because of normal cycle variation and the poor specificity of diagnostic tests such as serum progesterone levels and endometrial biopsies, the condition tends to be overdiagnosed and over treated.

Another endocrine abnormality suggested to be a cause of RPL is polycystic ovarian syndrome (PCOS), with an abortion rate of up to 50% having been reported in women with PCOS. Although women who hypersecrete LH or are hyperandrogenic have been reported to be at increased risk for SAB, it has been shown that suppression of LH levels prior to pregnancy in those women does not improve the live birth rate. A recent ultrasound study showed that the presence of PCO morphology does not increase a woman’s risk of SAB.


The presence of two antiphospholipid antibodies, anticardiolipin antibodies (ACA) or lupus anticoagulant (LAC), is well established to increase the risk of RPL. The role of autoantibodies to other phospholipid antigens is much less clear. ACA and LAC cause placental insufficiency due to placental thrombosis and decidual vasculopathy. About 17% of women with RPL have ACA as compared to 4% of controls. ACA’s are detected by an ELISA assay, while LAC is identified by aPTT, or a Russell viper venom time. No other autoantibodies have been consistently shown to be associated with RPL. Although the presence of thyroid autoantibodies was suggested to be more common in women with RPL, larger studies failed to confirm this finding.

Other Coagulation Abnormalities

A point mutation in the factor V gene (factor V Leiden) is associated with a significant increase in the risk of venous thromboembolism and has been shown to increase the risk of late spontaneous abortion and stillbirth. The role of factor V Leiden in first trimester loss is less clear, with recent studies showing no increased incidence of the mutation in women with early losses. The mutation is present in 5% of normal individuals.

Hyperhomocysteinemia and a factor II (prothrombin) mutation have also been shown to increase the risk of thromboembolism. Both of these have been shown to be found no more commonly in women with RPL than in controls.

Alloimmune Causes

Because the cause of the majority of recurrent pregnancy losses is unknown, it has been suggested that some may be due to alloimmune causes. It has been speculated that HLA compatibility of the couple, absence of maternal leukocytotoxic antibodies, or the absence of maternal blocking antibodies may be responsible. All of these have been investigated and treatment directed toward these abnormalities has been studied. Mowbray, et al performed a double-blind trial of treatment with husband’s purified lymphocytes in 1985, and the results suggested improvement in pregnancy outcome. A more recent multicenter study by Ober, et al showed, in a much larger group of patients, that paternal lymphocyte immunization resulted in no beneficial effect.

Others have shown that increased levels of natural killer (NK) cells are present in women with chromosomally normal miscarriages and that increased levels of NK cells are present in non-pregnant women with a higher chance of SAB in a subsequent pregnancy. In contrast, more recent studies have shown no relationship between the risk of pregnancy loss and the presence or absence of HLA compatibility, leukocytotoxic antibodies, blocking antibodies, or elevated NK cell activity. At this point, the role of alloimmunity in the mechanism of recurrent pregnancy loss is unclear, so the treatment of these problems is still speculative and investigational.

Oocyte Abnormalities

The risk of pregnancy loss has long been known to increase with advancing age. The data obtained from oocyte donation has shown that a recipient’s risk of SAB is directly related to the age of the oocyte donor rather than to the recipient’s age. This implies that the risk of SAB is a function of oocyte properties rather than uterine receptivity.

In a study of couples with RPL, a negative evaluation for RPL, and poor response to gonadotropins, oocyte donation resulted in a delivery rate of 85.7% and an SAB rate of 11.1%. Again, this indicates that oocyte quality is a determinant of pregnancy outcome and may play a causative role in RPL.

Evaluation of Recurrent Pregnancy Loss

The evaluation of couples with two or more SABs should be directed toward those abnormalities, which have been proven to be causative:

  • Peripheral blood karyotyping of both partners
Uterine abnormalities:
  • Hysterosalpingogram
  • Pelvic ultrasound
  • Luteal phase serum progesterone?
  • Endometrial biopsy?
  • Anticardiolipin antibody
  • APTT or RVVT
  • Factor V Leiden
  • Cyle day 2, 3, or 4 serum FSH and estradiol levels


The treatment for couples with a balanced translocation who want to use their own eggs and sperm would be IVF (in vitro fertilization) with genetic testing of the embryos (PGD) before they are placed in the uterus.  This allows the couple to choose to implant only the completely normal embryos or the ones with a balanced, rather than unbalanced, translocation.  Other options would include using donated eggs or sperm, depending on which partner has the translocation.  The couple may also choose to keeping trying on their own, knowing that they do have a small chance of having a normal child.

The current treatment of the septate uterus is hysteroscopic metroplasty. The division of the septum is equally successful with hysteroscopic scissors, electrocautery, or laser, depending upon the preference and experience of the surgeon. The need for postoperative adjunctive therapy with an intrauterine stent and estrogen therapy is unclear but generally felt to be unnecessary. The chance of successful pregnancy after hysterscopic metroplasty has varied widely in many studies, depending upon the nature of the group studied and the size of the septum. The majority of studies, however, report a 70 to 90% chance of live birth after surgery.

If a luteal phase defect is suspected, treatment is generally initiated with clomiphene ovulation induction (which increases luteal phase progesterone production), or addition of progesterone in the early luteal phase. Progesterone may be administered orally, vaginally, or intramuscularly. Many practitioners treat empirically with progesterone because of the difficulty of making a firm diagnosis, low treatment cost, and lack of risk.

A combination of low dose aspirin and heparin has been shown to enhance the chance of successful pregnancy in women with antiphospholipid antibodies. Treatment is usually begun when fetal cardiac activity is detected. The addition of intravenous immunoglobulin to this regimen was shown in a multicenter, randomized pilot study to have no beneficial effect.

In women of advanced reproductive age and in those with poor ooctye quality indicated by elevated early follicular phase FSH and estradiol levels, consideration should be given to oocyte donation.

It should be recognized that in most studies of couples with unexplained RPL, control groups have a 60 to 70% chance of successful live birth. With this in mind, the practitioner should resist the temptation to treat these couples with unproven, expensive, or potentially deleterious modalities. The lessons learned from the diethylstilbestrol experience should not be forgotten.

Physician Resources

Sorry, the comment form is closed at this time.