Recurrent miscarriage syndrome (RMS) is a common obstetric problem, affecting over 500,000 women in the United States per year.[1] Infertility, although less well defined epidemiologically, is also a common clinical problem. While the World Health Organization WHO and the Royal College of Obstetricians and Gynaecologists (RCOG) define recurrent miscarriage as 3 or more consecutive miscarriages before the 20th week of pregnancy, the American Society for Reproductive Medicine (ASRM) and European Society of Human Reproduction and Embryology (ESHRE) definition is the spontaneous loss of two or more pregnancies.[2, 3]
Proper evaluation can determine the etiology of RMS in almost all women. The most common hemostasis-related cause is a thrombotic disorder, of which the most common is antiphospholipid syndrome (APLS). Hemorrhagic defects are very rare hemostasis-related causes of RMS, but these conditions are also treatable in many instances and should be investigated in appropriate cases.
Treatment of the common procoagulant defects consists of preconception low-dose aspirin, followed by the addition of immediate postconception low-dose unfractionated or low-molecular-weight heparin.
For patient education information, see the Women's Health Center and Pregnancy Center, as well as Miscarriage, Threatened Miscarriage, and Infertility.
RMS due to blood protein or platelet defects may come about through either of two mechanisms: (1) disorders associated with a hemorrhagic tendency or (2) defects associated with a thrombotic tendency. Hemorrhagic (bleeding) defects associated with RMS are rare, whereas thrombotic or hypercoagulable/thrombophilic defects are extremely common.[4, 5] The hemorrhagic defects associated with fetal wastage syndrome presumably lead to inadequate fibrin formation, thus precluding adequate implantation of the fertilized ovum into the uterus.
Fetal wastage in women with thrombotic defects results from thrombosis of early placental vessels. Fetal wastage peaks in the first trimester, but may also occur in the second and third trimesters.[1, 6] The earlier the pregnancy, the smaller the placental and uterine vessels and, therefore, the greater the propensity to undergo partial or total occlusion by thrombus formation. Thrombotic occlusion of placental vessels, both venous and arterial, preclude adequate nutrition and, thus, viability of the fetus.[1, 4]
The thrombotic hemostasis defects associated with recurrent miscarriage syndrome include the following:
As new procoagulant factor mutations associated with hypercoagulability and thrombosis are discovered, they, too, are anticipated to be found associated with placental thrombosis and RMS in many cases.
Antiphospholipid syndromes
Many clinicians consider APLS to be the most common prothrombotic disorder among both hereditary and acquired defects, as well as the most common thrombotic disorder causing recurrent miscarriage.[1, 4, 7, 8, 22, 23, 24, 25, 26, 27] It has also long been recognized that treatment for this condition is often successful.[28]
Various treatment programs have been advocated.[29] One difficulty in evaluating these programs has been that some of them have primarily addressed patients with secondary APLS and fetal wastage, in particular those with underlying systemic lupus erythematosus or other autoimmune disorders. Only a few investigators have addressed populations with primary APLS who have no known underlying disease.
When assessing causes of infertility alone, APLS is thought to account for about 30% of infertility cases; however, in one series, abnormal CD56+/CD16 cell ratios were the single most common defect found (40%) in infertility patients.[30] In another series, only 21% of patients with RMS had APLS; however, when assessing women with APLS historically, 80% had suffered at least one miscarriage.[31, 29]
In a series reported by Granger and Farquharson, 387 unselected patients were assessed for APLS. Of these, 16% harbored APLS, and of those with APLS, 56% of patients treated with low-dose aspirin (ASA) had a term delivery.[32] Borelli et al found that 60% of their studied patients with "habitual" unexplained miscarriage had APLS.[33] Although the great majority of cases of APLS are clearly acquired,[7, 25] familial APLS that is associated with RMS has been reported.[34] Clearly, however, screening for APLS is indicated in patients with RMS.[1, 4, 7, 8, 22, 23, 25, 35]
In addition, because APLS is very common and because many of the hereditary thrombophilias, such as factor V Leiden, are very prevalent in North America, it is not unexpected that some women with RMS have APLS in combination with other procoagulant defects.[36, 37] Aznar et al reported a case of RMS that was complicated by deep venous thrombosis (DVT) and thrombotic stroke in a patient with APLS, factor V Leiden, and congenital protein S deficiency.[38]
Many mechanisms have been proposed whereby APLS interferes with the hemostasis system and predisposes to thrombosis.[1, 4, 5, 7, 8, 22, 23, 24, 25, 35] However, some investigators have proposed mechanisms that are specific for RMS. These proposed mechanisms have included the hypothesis that APLSs induce acquired activated protein C resistance (APC-R),[38] as well as interfere with prothrombin (factor II), protein C and protein S, tissue factor, factor XI,[39] and the tissue factor/tissue factor pathway inhibitor (TF/TFPI) system.[40] Another study also found that patients with APLS harbored antibodies to prothrombin, protein C, and protein S.[41] Other investigators have proposed these patients may also develop antibodies to "thromboplastin" and thrombin.[42]
Another proposed mechanism is that APLSs interfere with annexin-V (also referred to as placental anticoagulant protein, serine only).[43] Three studies demonstrated immunoglobulin (Ig) fractions of antiphospholipid antibody (APLA) or beta2-glycoprotein-1 (B2GP1) decrease trophoblastic annexin-V[43, 44, 45] , but several studies have shown this anti–annexin-V activity to be limited to the antiphosphatidylserine subgroup antibody idiotype.[46, 47]
On rare occasions, APLS may be inherited (this author has seen three such families) and others have been reported.[34] For that reason, a positive maternal history may warrant evaluation at first pregnancy, as should a history of familial thrombosis.
Other thrombophilic states
Patients with other congenital or acquired thrombophilic states are also at high risk for placental thrombosis and RMS. In a study that assessed a variety of these defects in 46 selected women with RMS (anatomic and hormone defects were ruled out before the hemostasis assessment), the following was found: 76% had anticardiolipin antibodies (void of lupus anticoagulants), 3% had a lupus anticoagulant (void of anticardiolipin antibodies), 11% had congenital protein S deficiency (3 quantitative; 1 dysfunctional), 6.5% had sticky platelet syndrome (2 with type I; 1 with type II), 3% had dysfibrinogenemia, and 3% had congenital TPA deficiency.[6]
In a study that assessed the prevalence of hereditary and acquired defects in patients with RMS, 9.4% had isolated factor XII deficiency and 7.4% had APLS; fibrinolytic system defects, leading to hypofibrinolysis and hypercoagulability, were found in 42.6% of patients.[48] This study concluded that von Willebrand disease, fibrinogen deficiency, antithrombin deficiency, protein C and protein S deficiency, TPA deficiency, and PAI-1 defects played no role in RMS.[48]
However, in a similar study assessing hereditary hemostasis defects in 125 patients with RMS, quite different results were noted, and factor V Leiden mutation was found in 14%.[49] However, in another study of 50 patients with RMS, it was concluded that factor V Leiden, prothrombin G20210A mutation, and 5,10-MTHFR mutations were not causes of RMS.[50] Bokarewa et al also noted in their study that although factor V Leiden was responsible for a greater than 3-fold risk of DVT, there was no association with miscarriage.[51]
Yet Brenner et al revealed that factor V Leiden was responsible for 48% of recurrent miscarriages,[52] and Rai et al reported that factor V Leiden was associated with a high incidence of second-trimester miscarriages in their series.[53] An additional two studies clearly showed an association between factor V Leiden mutation and recurrent miscarriages.[54, 55] Thus, the preponderance of evidence certainly strongly suggests that heterozygous factor V Leiden mutation is a significant risk factor for recurrent miscarriage and increases the risk for miscarriage by at least 3.3-fold.[52]
Another common thrombophilic disorder, prothrombin G20210A gene mutation, was described by Poort and associates in 1996.[56] Although Kutteh et al found no association between this mutation and RMS,[50] a study by Brenner et al found a 2.2-fold increased risk of recurrent miscarriage in women with this genetic procoagulant defect.[57]
Another hereditary defect that leads to hypercoagulability and thrombosis is 5,10-MTHFR C677T mutation. Although Kutteh et al found no association between this defect and recurrent miscarriage,[50] Brenner et al showed a clear association between heterozygosity for this mutation and recurrent miscarriage, with those who harbor the mutation having a 2-fold enhanced risk of miscarriage.[57]
Finally, although hypofibrinolysis in general has been shown to be associated with recurrent miscarriage,[18] only recently has the role of PAI-1 elevation and PAI-1 polymorphism or polymorphisms been shown as a cause of RMS.[19] Potential and proposed mechanisms of antiphospholipid antibody-induced thrombosis (APL-T) include the following:
Hemorrhagic or bleeding defects are rare causes of recurrent miscarriage relative to thrombotic or thrombophilic disorders.[1, 4, 58] The hemorrhagic defects associated with fetal wastage syndrome include the following:
Management of these patients is generally plasma substitution therapy or, in appropriate disorders, DDAVP (vasopressin) therapy.[1, 4]
RMS affects more than 500,000 women in the United States per year.[1] Females of any race can be affected.
The incidence of recurrent miscarriage varies depending the definition used but RMS is generally thought to occur in 1-3% of women of reproductive age.[3]
The prognosis for a successful live birth in women with RMS is generally good. Poor prognostic factors include advanced maternal age and the number of previous losses.[74]
Patients have a history of two or more miscarriages. They may or may not have a history of thrombosis or bleeding.
The physical examination is usually normal, unless maternal thrombosis is present.
A coagulation defect is the single, most common cause of recurrent miscarriage (two or more miscarriages), after hormonal defects and anatomic defects have been eliminated as the cause. Recurrent miscarriage, based upon the available literature and our experience, is generally due to well-defined defects, as follows[1] :
The approximate prevalence of causes of recurrent miscarriage syndrome and infertility is summarized in the image below.[1] These are in contrast to first-time miscarriage, which in most cases is due to a chromosomal defect and may affect up to 25% of first pregnancies.[1]
If the patient has no history of bleeding, an antiphospholipid panel should be ordered. If the patient does have a history of bleeding, the following tests should be ordered:
For information on the workup of antiphospholipid syndrome in these patients, see Antiphospholipid Syndrome and Pregnancy.
Because fetal loss associated with clotting disorders is thought to occur due to interference with adequate fibrin formation for implantation of the fertilized ovum into the uterine lining, the authors choose not to use vigorous preconception antithrombotic therapy in those patients with recurrent miscarriage syndrome (RMS) due to thrombophilia; rather, we use low-dose aspirin at 81 mg/d. This issue may be of theoretical concern only, in view of the report by Sher et al, who used preconception low-dose heparin with a high success rate for in vitro fertilization techniques.[75] However, the authors remain concerned and continue to advocate low-dose aspirin as the preconception antithrombotic therapy in most instances.[1, 59, 6]
The regimen of a postconception addition of fixed, low-dose heparin at 5000 units every 12 hours is empirical, but higher doses seem to be associated with bleeding and a lower success rate.[76] It may be that even lower doses of heparin might suffice.
The authors do not advocate using corticosteroid therapy in this patient population, based upon the negative experience of others in fetal wastage syndrome and the authors' own experience of using steroids in conjunction with antithrombotics in patients with antiphospholipid syndrome (APLS) and other types of thrombosis, wherein the corticosteroid use could be shown to lower antiphospholipid antibody titers but failed to abort thrombotic events.[1, 6, 7, 8, 23, 24, 25] In addition, steroid use in patients with APLS is considered possibly detrimental.[28]
Various treatment programs have been used for women with APLS (anticardiolipin antibodies or lupus anticoagulants) and fetal wastage syndrome; however, many of these studies have examined only very small populations or failed to distinguish between primary or secondary APLS in the information provided. Brown reported a 90% failure rate (miscarriage) among untreated women,[77] Perino et al reported a 93% failure rate in untreated women,[78] and Many et al also reported a similar failure rate in untreated patients.[79]
Lubbe and Liggins noted an 80% successful term pregnancy rate in a small group of women with use of prednisone and aspirin[80] ; a similar success rate with this regimen was noted by Lin.[81] Cowchuck et al noted a 75% success rate with prednisone alone or with aspirin alone, but the investigators also noted more undesirable effects in the prednisone-treated population.[82] Landy et al, reported a 90% success rate in a small population with either aspirin alone or with prednisone alone.[83] However, Many et al only noted a 43% successful term pregnancy rate with aspirin and prednisone,[79] and Semprini et al noted only a 14% success rate with prednisone alone.[84]
Several studies have assessed the role of postconception addition of heparin; however, most have used higher doses than used in the authors' population. Rosove et al reported a 93% success rate with dose-adjusted subcutaneous (SC) heparin[76] ; the mean heparin doses were about 25,000 U/d. Kutteh noted a success rate of 76% in a population of 25 patients treated with aspirin plus dose-adjusted SC heparin[85] ; the mean heparin dose was 26,000 U/d. Many et al reported that patients treated with prednisone plus aspirin and heparin at 5000 U twice a day had a better outcome (69%) than did those who were treated with aspirin plus prednisone (43%) or with prednisone alone (7%).[79]
The authors' results suggest that fixed low-dose heparin is more effective than high-dose, dose-adjusted regimens[1, 6] ; more than 98% of the authors' RMS population with APLS or other prothrombotic propensity had a normal term delivery. Higher doses of heparin may somehow contribute to adverse outcomes, such as small periplacental hemorrhages.
In a meta-analysis of 19 trials involving 2391 patients with a history of recurrent miscarriage with or without thrombophilia and 543 patients with APS for patients with or without thrombophilia, low molecular weight heparin therapy had the greatest probability of live births; for patients with APS, the combination of unfractionated heparin and aspirin was the superior treatment.[86]
Parke reported on the combination of low-dose heparin used in conjunction with intravenous immunoglobulin (IVIG).[87] Her success rate, however, was only 27%, suggesting that IVIG has little role in antiphospholipid fetal wastage syndrome.
Treatment with hydroxychloroquine proved beneficial in a preliminary study of 170 pregnancies in 96 women with persistent antiphospholipid antibodies. Compared with a control group of 65 women who received conventional treatment, the 31 women who received hydroxychloroquine for at least 6 months before pregnancy and throughout gestation had a higher rate of live births (67% vs 57%; P = 0.05) and lower rates of antiphospholipid antibody–related pregnancy morbidity (47% vs 63%; P = 0.004), miscarriage at > 10 weeks of gestation (2% vs 11%; P = 0.05), and placenta-mediated complications (2% vs 11%; P = 0.05). Hydroxychloroquine-treated women also had longer pregnancy duration and a higher rate of spontaneous vaginal labor.[88]
Over the past 5 years, the authors have carefully assessed 351 women referred for thrombosis and hemostasis evaluation after recurrent miscarriages. In the Dallas/Fort Worth Metroplex (DFW Metroplex), composed of a population of about 6 million, a flow protocol is followed to maximize success and to keep the costs of evaluation for the etiology of RMS and infertility at a minimum while providing the best chances for defining an etiology and, thus, providing ideal therapy for a successful term-pregnancy outcome.[1, 5, 6] This protocol is presented in the image below.
In all instances, women with RMS and infertility are first seen by an obstetrician or reproductive specialist. Anatomic defects and hormonal defects are assessed and, if found, the workup stops at this point and treatment is initiated (about 25% of all women). If no anatomic or hormonal defect is found, the patient is then seen by referral for hemostasis evaluation; the positive yield among this selected population is about 92%. If these evaluation findings are negative (about 8%), then, if the patient desires, chromosomal evaluation is initiated (about a 7% yield).
Most of the obstetricians and reproductive specialists in the DFW Metroplex refer patients after two or more miscarriages; however, some specialists refer after one miscarriage in the face of a positive patient family history for miscarriage; occasionally, patients request a workup after only one miscarriage. The authors' practice has been to accommodate the desires of the patient after discussing the costs and other implications of evaluation.
At the time of this writing, all 322 patients with a defect have been monitored for at least 15 months; their results have been analyzed in detail, with the summary presented below.
The mean age of the patients referred for a hemostasis evaluation is 33.3 years, the mean number of miscarriages before referral is 2.9 (range = 2-9), and the percentage found to have a hemostasis defect is 92% (322 of 351). See Table 1, below.
Table 1. Characteristics of the First 351 Women Referred for Hemostasis Evaluation (Open Table in a new window)
Patient Characteristics (All 351 Patients) |
Mean |
Standard Deviation |
Maximum |
Minimum |
Age, y |
33.3 |
5.63 |
49 |
18 |
Number of Miscarriages |
2.9 |
2.39 |
9 |
2 |
All patients underwent a thorough evaluation for thrombophilia and, when indicated, a hemorrhagic disorder. Of the 351 patients, 29 (8%) had no defect. Of the remaining 322 patients, 10 (3%) had a bleeding disorder: three (1%) with platelet dysfunction, one (0.3%) with factor XIII deficiency, three (1%) with von Willebrand disease, and three (1%) with Osler-Weber-Rendu syndrome.
The remainder of the patients had a thrombophilia, as follows:
A total of 364 defects were found in the 312 patients with thrombophilia; thus, several had two, and a few had three, separate defects.
As has been found by most other investigators, the most common defect found in RMS has been APLS; however, unlike some groups, the authors assess for all phospholipid antibody subgroups, including the following:
Of note, by including all antiphospholipid subgroups, 29% of patients are found to have a subgroup antiphospholipid antibody but no anticardiolipin antibody or lupus anticoagulant; thus, 29% of patients would remain undiagnosed if an assessment of these subgroups were not performed. Interestingly, this finding is about the same as that noted in young-age patients (< 51 y) with thrombotic stroke.[89]
The particulars of the patients with APLS in the authors' population, with demonstration of the idiotypes found, are summarized in Table 2, below.
Table 2. Clotting Disorders Found in the Authors' Population (Open Table in a new window)
Antiphospholipid Found |
Patients With APLS, % |
ACLA-IgG only |
32.6 |
ACLA-IgM only |
23.4 |
ACLA-IgA only |
7 |
ACLA-IgG + IgM |
3 |
ACLA-IgG + IgA |
1 |
ACLA IgA + IgM |
0 |
Lupus anticoagulant only |
2 |
ACLA + lupus anticoagulant |
2 |
Subgroup Only (No ACLA or lupus anticoagulant present) |
|
Antiphosphatidylserine |
4 |
Antiphosphatidylinositol |
2 |
Antiphosphatidylethanolamine |
5 |
Antiphosphatidic acid |
5 |
Antiphosphatidylcholine |
7 |
Antiphosphatidylglycerol |
1 |
Anti-annexin-V |
5 |
B2GP1 |
0 |
Hexagonal phospholipid |
0 |
Total |
|
(9 Patients had ACLA + a subgroup antibody) |
|
Total with only a subgroup antibody |
|
APLS patients with only a subgroup antibody, % |
29 |
All patients with a thrombophilic defect were treated with preconception aspirin at 81 mg/d, and at documentation of conception, the women were treated with the addition of subcutaneous (SC) unfractionated heparin at 5000 U q12 hours by self-injection (first 120 patients) or SC low–molecular-weight (LMW) heparin (dalteparin [Fragmin], 5000 U q24 h by self-injection; subsequent 192 patients). Both drugs (aspirin and heparin or LMW heparin) are used to term.
All patients are instructed in the administration of heparin injections; they are also informed of all important side effects of heparin therapy and are extensively informed of the benefits and risks of heparin/LMW heparin therapy, including the fact that side effects, although rare, include the following:
Patients are also informed that about 5-10% of patients develop a transient transaminasemia during heparin/LMW heparin therapy, but this is without any known adverse clinical consequences.
Patients receive the following instructions about self-administration of SC medication:
All patients are instructed to return immediately if they note dark or black areas of the injection site, which are potentially indicative of skin necrosis. The methods of follow-up are summarized in the list below.
The DFW Metroplex Cooperative RMS Group follow-up protocol for fetal wastage syndrome that is associated with hypercoagulable blood protein/platelet defects is as follows:
Laboratory assessment
Clinicians considering the use of LMW heparin in pregnancy should be made aware of the US Food and Drug Administration (FDA) safety alert warning regarding the use of enoxaparin (Lovenox) in pregnancy and women of childbearing age.
All of the authors' 315 patients with a thrombophilic defect were treated with the aforementioned regimen of preconception low-dose aspirin plus postconception thromboprophylactic (low-dose) SC heparin or dalteparin. Patients with MTHFR mutations were also treated with folic acid at 5 mg/d plus pyridoxine at 50 mg/d.
Four pregnancy losses (2.6%) occurred in patients receiving antithrombophilic prophylaxis. One loss was during the second trimester and accompanied a cholecystectomy, and one loss was during the first trimester in a patient with APLS and a fetal chromosomal defect; neither of these were considered treatment failures. However, two patients suffered first-trimester loss, and placental thrombi and infarcts were present. Thus, those two losses clearly represented treatment failure.
The overall success in treating patients with RMS with procoagulant/platelet defects in the authors' program is, therefore, 99% (313/315) with respect to normal term delivery. All patients were monitored for a minimum of 3 months after delivery. No patient sustained a thrombotic episode during the pregnancy, delivery, or postpartum period except the two patients who experienced treatment failures, both of whom had placental vascular thrombi. In addition, no patient developed HIT/thrombocytopenia, and none had a clinically significant hemorrhage.
Almost all patients developed small ecchymoses at the injection sites, but these findings were considered insignificant by both the patient and physician. Ten percent of patients developed eosinophilia, which had abated by 3 months postpartum, and 7% developed mild to moderate elevations of hepatic transaminases; these laboratory findings also returned to normal by 3 months postpartum. Per the obstetricians, reproductive medicine specialists, and involved pediatricians, no neonatal or pediatric problems were associated with the administered therapy. No patient sustained a fracture during or after treatment.
Patients with bleeding disorders were not treated. No patient had a significant hemorrhage during pregnancy or delivery. None required any blood product therapy.
Management of patients with recurrent miscarriage due to hemorrhagic disorders is generally with plasma substitution therapy or, in appropriate disorders, DDAVP (vasopressin) therapy.[1, 4] Treatment of patients with thrombotic disorders is with aspirin, heparin, or low-molecular weight heparin (LMWH).
Antiplatelet effect indicated to decrease risk of thrombosis and pregnancy loss in pregnant women with antiphospholipid antibody (APS) syndrome. Although not proven effective when used alone, most clinicians use aspirin with subcutaneous heparin in pregnant patients with APS. Begin aspirin as soon as conception is attempted.
Inhibits platelet aggregation by inhibiting platelet cyclooxygenase. This, in turn, inhibits conversion of arachidonic acid to PGI2 (potent vasodilator and inhibitor of platelet activation) and thromboxane A2 (potent vasoconstrictor and platelet aggregate).
Indicated to decrease risk of thrombosis and pregnancy loss in pregnant women with APS.
Augments activity of antithrombin III and prevents conversion of fibrinogen to fibrin. Does not actively lyse but is able to inhibit further thrombogenesis. Prevents re-accumulation of clot after spontaneous fibrinolysis.
Produced by partial chemical or enzymatic depolymerization of unfractionated heparin (UFH). Binds to antithrombin III, enhancing its therapeutic effect. The heparin-antithrombin III complex binds to and inactivates activated factor X (Xa) and factor II (thrombin).
Does not actively lyse but is able to inhibit further thrombogenesis. Prevents reaccumulation of clot after spontaneous fibrinolysis.
Advantages include intermittent dosing and decreased requirement for monitoring. Heparin anti–factor Xa levels may be obtained if needed to establish adequate dosing.
LMWH differs from UFH by having a higher ratio of antifactor Xa to antifactor IIa compared to UFH.
Prevents DVT, which may lead to pulmonary embolism in patients undergoing surgery who are at risk for thromboembolic complications. Used for prevention in hip replacement surgery (during and following hospitalization), knee replacement surgery, or abdominal surgery in those at risk of thromboembolic complications, or in nonsurgical patients at risk of thromboembolic complications secondary to severely restricted mobility during acute illness.
Used to treat DVT or PE in conjunction with warfarin for inpatient treatment of acute DVT with or without PE or for outpatient treatment of acute DVT without PE.
No utility in checking aPTT (drug has wide therapeutic window and aPTT does not correlate with anticoagulant effect).
Average duration of treatment is 7-14 d.
Indicated to decrease the risk of thrombosis and pregnancy loss in pregnant women with APS.
Enhances inhibition of Factor Xa and thrombin by increasing antithrombin III activity. In addition, preferentially increases inhibition of Factor Xa.
Except in overdoses, no utility exists in checking PT or aPTT because aPTT does not correlate with anticoagulant effect of fractionated LMWH.
Average duration of treatment is 7-14 d.