Berek Novak's Gyn 2019. Chapter 33 Recurrent Pregnancy Loss

 Chapter 33 Recurrent Pregnancy Loss

KEY POINTS

1 Isolated spontaneous pregnancy loss is common and approximately 15% to 30% of recognized pregnancies end in miscarriage. Recurrent pregnancy loss defined by three or more miscarriages is rare, occurring in 1% of women.

2 Clinical evaluation of a couple with recurrent miscarriage is warranted after two miscarriages caused by similar prevalence of abnormal results for couples with two or three pregnancy losses.

3 The most common cause of first trimester miscarriage is fetal aneuploidy.

4 Evaluation of patients with recurrent miscarriage should include a detailed patient and family history and examination of anatomic, genetic, endocrine, and antiphospholipid syndrome abnormities.

5 Much of the traditionally accepted testing, such as thrombophilia screening and several immunologic screening tests, are not supported by current evidence.

6 At least 50% of couples will have no specific abnormal finding after an evaluation for recurrent miscarriage.

7 After several pregnancy losses, there remains a greater chance of having a viable birth than another miscarriage, even without treatment. Prognosis can improve with treatment of a known underlying etiology.

8 Controversial treatments for recurrent miscarriage, including in vitro fertilization with chromosomal screening of embryos, immunosuppression, and anticoagulation, remain under investigation.

9 Treatment for recurrent pregnancy loss should include close monitoring with β- human chorionic gonadotropin levels and ultrasound, psychological support, and karyotypic analysis of tissues from subsequent miscarriage.

Despite advances in research and technology, recurrent pregnancy loss remains a clinical and emotional challenge for patients and providers. Sporadic miscarriage is the most common complication of pregnancy affecting approximately 1 in 4 couples in their reproductive years. Approximately 70% of human conceptions fail to achieve viability, and an estimated 50% are lost before the first missed menstrual period (1). Studies using sensitive assays for human chorionic 1961gonadotropin (hCG) indicate that the actual rate of pregnancy loss after implantation is 31% (2). [1] Miscarriage occurs in 15% to 25% of pregnancies that are clinically recognized before 20 weeks of gestation from last menstrual period (3,4). Traditionally, recurrent miscarriage has been defined as the occurrence of three or more clinically recognized (ultrasound or histopathologic evidence) pregnancy losses before 20 weeks from the last menstrual period (5).

Using this definition, recurrent pregnancy loss occurs in approximately 1 in 300 pregnancies (2) and in less than 1% of women (337). [2] Clinical investigation of pregnancy loss, however, may be initiated after two consecutive spontaneous miscarriages (338). A study of more than 1,000 patients with recurrent pregnancy loss reported no difference in the prevalence of abnormal results for evidence-based and investigative diagnostic tests when the diagnostic workup was initiated after two versus three or more losses (6). If clinical intervention is undertaken in the form of investigation after two spontaneous miscarriages, approximately 1% of pregnant women will require evaluation (3). Even with a history of recurrent pregnancy loss, a patient is more likely to carry her next pregnancy successfully to term than to miscarry. For patients with a history of recurrent pregnancy loss, the risk of subsequent pregnancy loss is estimated to be 24% after two clinically recognized losses, 30% after three losses, and 40% to 50% after four losses (7). These data make clinical study of recurrent pregnancy loss and its treatment difficult because very large groups of patients must be studied to demonstrate the effects of any proposed therapeutic intervention.

ETIOLOGY

Recurrent pregnancy loss has been attributed to genetic, anatomic, immunologic, endocrine, thrombophilic, and infectious causes but etiology remains under investigation. [6] Although the exact proportion of patients diagnosed with a particular abnormality varies by study, parental genetic abnormalities (2% to 5%), anatomic issues (10% to 12%), antiphospholipid syndrome (APS) (15%), and other identifiable issues (10%) have been found in couples with recurrent miscarriage. This leaves approximately 50% of couples with a diagnosis of unexplained recurrent miscarriage (Table 33-1) (3,6,8,338). The lack of abnormal findings in the couple having recurrent pregnancy loss is not surprising given how common fetal aneuploidy is with conception.

[3] At least 60% of preclinical and early clinical pregnancy losses are the result of de novo fetal aneuploidy (9). Among women aged 35 or greater who experience recurrent pregnancy loss, spontaneous fetal chromosomal abnormalities are responsible for over 80% of losses (10). This is thought to be the cause of anembryonic pregnancy losses, whereas pregnancy losses occurring after 10 weeks of fetal development are much less likely to derive from fetal aneuploidy. Pregnancy losses resulting from de novo fetal aneuploidy, whether early and undocumented or documented through evaluation of chromosomal content in fetal tissues, cloud the results of many published studies. Their presence or absence must be documented in all investigations on recurrent pregnancy loss patients and their potential as a confounding factor discussed. The timing of fetal demise and tissue chromosomal analysis of any collected fetal tissues should be carefully weighed when diagnostic and therapeutic investigations into causes of recurrent pregnancy loss are being considered.

Genetic Factors

Approximately 3% to 5% of recurrent pregnancy loss can be associated with a parental balanced structural chromosome rearrangement and can be diagnosed with blood tests for karyotype for both partners (338). Chromosome rearrangements are most commonly balanced reciprocal but can include robertsonian translocations, chromosomal inversions, insertions, and mosaicism (11–14). A balanced reciprocal translocation is a rearrangement in chromosomes in which breaks occur in two different chromosomes and the pieces are exchanged so that no genetic material is lost or gained in the process. Balanced rearrangements can only be detected with standard G-banding karyotype. Highdensity microarray technology and next-generation sequencing can detect smaller deletions and duplications, but are not able to differentiate balanced rearrangements from normal because the total genetic content is the same. When screening parents for chromosome rearrangement, the best test is a metaphase karyotype with G-banding. Embryos from a parent with a balanced translocation with a normal phenotype have either two normal chromosomes or inherit the balanced translocation; however, some embryos can have abnormal chromosome content and lead to babies born with birth defects if they inherit an unbalanced translocation. A robertsonian translocation is a form of chromosomal rearrangement with one of the short-arm chromosomes (13–15) in which there is a loss of part or all of the short arms and fusion of the long arms of the chromosomes. A balanced robertsonian translocation causes no health issues but unbalanced forms can result in trisomies or monosomies that usually result in miscarriage. Some chromosome inversions are benign and do not impact prognosis for patients with recurrent pregnancy loss. However, others may have significant reproductive phenotypes. All patients with balanced chromosome rearrangements should undergo genetic counseling prior to attempting conception to better understand their individual risks. Neither family history alone nor a history of prior term births is sufficient to rule out a potential parental chromosomal abnormality. Whereas the frequency of detecting a parental chromosomal abnormality is inversely related to the number of previous spontaneous losses, the chance of detecting a parental chromosomal abnormality is increased among couples who have never experienced a live birth (13). Testing for parental karyotopes in couples with recurrent miscarriage is recommended but some argue it has a high cost for discovery of a rare cause for recurrent pregnancy loss with limited treatment options. Proponents of testing indicate it promotes benefits that include permitting patient autonomy in choosing options for treatment and identification of patients at risk for offspring with unbalanced translocations who may have birth defects.

A couple with a balanced translocation has a high chance of live birth attempting to conceive via intercourse (339); or they have the option of screening embryos for chromosome arrangements with in vitro fertilization (IVF) (338). Studies to date evaluating the use of IVF with chromosomal screening for couples with recurrent pregnancy loss and translocations do not show a significantly higher chance of success compared to natural conception (235,340). These studies are limited in patient number and use older technology and techniques like fluorescent in situ hybridization (FISH) and single cell day 3 embryo biopsy (235,340). With advancements such as blastocyst biopsy, alternate chromosome analysis techniques and others, future studies may show a benefit for use of IVF in this population. Based on the existing data, IVF with embryo screening is not strongly recommended for a couple with a balanced translocation and recurrent miscarriage.

Thrombophilias

After combining a known association between the presence of antiphospholipid antibodies, placental dysfunction, and miscarriage with evidence showing subsequent increase in live birth with anticoagulation medications in these patients with recurrent miscarriage (299), attention was turned to investigating inherited thrombophilias as a cause for recurrent pregnancy loss. The clotting cascade is a delicate balance of prothrombotic (see Fig. 33-1) and antithrombotic pathways (see Fig. 33-2) and dysregulation in either or both of these pathways can cause significant health consequences. Because alterations in coagulation occur naturally in pregnancy with relative maternal hypercoagulability secondary to increases in several procoagulant activities, decreases in select antithrombotic pathways and venous stasis, the delicate balance between pro- and antithrombotic would seem to be even more tenuous throughout gestation (see Fig. 33-3). [5]

Despite the biologic plausibility that patients with recurrent miscarriage would have a higher prevalence of inherited thrombophilia and that anticoagulation of patients with these diagnoses would decrease risk of subsequent miscarriage, the evidence does not support the theories (341).

Table 33-1 Proposed Etiologies and Testing for Recurrent Spontaneous Miscarriage

FIGURE 33-1 The clotting cascade. Physiologic clotting is initiated by endothelial cell damage or abnormal exposure of negatively charged phospholipids to serum and blood components. Procoagulant pathways in black are part of the clotting cascade and are prothrombotic. Both lead to activation of cascades of proteolytic enzymes and cleavage of clotting factors. The extrinsic and intrinsic pathways are initiated by distinct mechanisms. They merge at the activation of factor X to form the common pathway. For all coagulation factors, the subscript letter “a” denotes the activated form of the factor.

FIGURE 33-2 Physiologic mechanisms that counteract the clotting cascade. The procoagulant cascade is inhibited by several physiologic mechanisms. The balance between pro- and antithrombotic pathways determines the state of coagulability.

Antithrombotic mechanisms include the action of antithrombin (AT) and of proteins C and S. The sites of inhibition by these substances are depicted by an “X.” Plasminogen activator inhibitors-1 and -2 (PAI-1 and PAI-2) indirectly inactivate plasmin. Plasmin plays an important role in the dissolution of coagulated blood. For all coagulation factors, the subscript letter “a” denotes the activated form of the factor. FSP, fibrin split products; FDP, fibrin degradation products.

FIGURE 33-3 Alterations coagulation and fibrinolysis during normal pregnancy.

Pregnancy is a state of hypercoagulability. Levels of factors VII, VIII, X, and XII are elevated throughout pregnancy, levels of factors II, V, and XIII rise in the first trimester, then return to normal values. The antithrombotic activity mediated by protein S decreases in pregnancy. The placenta produces plasminogen activator inhibitor-2 (PAI-2). For all coagulation factors, the subscript letter “a” denotes the activated form of the factor. FSP, fibrin split products; FDP, fibrin degradation products.

Inherited thrombophilia is a genetic predisposition to venous thromboembolism and traditionally includes factor V Leiden (FVL) mutation, prothrombin gene mutation, protein C deficiency, protein S deficiency, and antithrombin III deficiency. It is estimated that up to 15% of the Caucasian population carries an inherited thrombophilic mutation (15). FVL is a clotting protein that works in conjunction with activated factor X to convert prothrombin to thrombin. One comprehensive review shows that carriers for FVL have an increased relative risk of recurrent pregnancy loss (342), however, the absolute risk of miscarriage in this population is low (343).

Prothrombin gene mutation results in a deficiency of thrombin and resulting increased concentration of prothrombin in the plasma. Some small studies have shown a potential association between a mutation in the prothrombin gene and recurrent pregnancy loss but overall recommendations do not support an association with this mutation and any pregnancy complication including miscarriage, intrauterine growth restriction, or preeclampsia (342,343). Protein S and Protein C in combination are required for factor V and VIII activation and thus deficiency in either can lead to a hypercoagulable state. Patients with protein S or protein C deficiencies may be at slightly higher risk of thrombosis in pregnancy but associations with miscarriage risk have not been definitively demonstrated (344). Antithrombin is a small protein that regulates clot formation by inactivating factor Xa and thrombin. Deficiency in this protein leads to high risk of thrombosis and increased risk of miscarriage (344). Because of its low prevalence (1/2,500), antithrombin deficiency is not recommended for routine screening for recurrent miscarriage in the absence of significant family history of thrombosis (345).

Although not usually considered an inherited thrombophilia, a methylenetetrahydrofolate reductase (MTHFR) mutation may be associated with increased thrombosis risk and some suggest an association with recurrent pregnancy loss. MTHFR is an essential enzyme for folic acid metabolism and is responsible for converting homocysteine to methionine. A mutation in the enzyme may cause increased serum levels of homocysteine that may lead to a hypercoagulable state (see Fig. 33-4). There are two predominant mutations (C677T and A1298C) and they are common in certain populations with up to 40% of Caucasian and Hispanic populations being heterozygote for the C677T mutation (346). Some suggest that a mutation leads to lower function of the MTHFR enzyme resulting in higher homocysteine levels and increased risk of thrombosis and miscarriage but the association is controversial (347).


FIGURE 33-4 Homocysteine metabolism. Dietary methionine is metabolized either to cysteine or back into methionine. Homocysteine, a prothrombotic metabolite, is an

intermediate in this process. Conversion of homocysteine to methionine requires transfer of

a methyl group from 5-methyltetrahydrofolate. The conversion of folate to 5-

methyltetrahydrofolate is a multistep process requiring vitamin B2 as a cofactor for the

enzyme, methylene tetrahydrofolate reductase (MTHFR). Vitamin B12 is a required

cofactor for the enzyme methionine synthase. Vitamin B6 is also required for metabolism

of sulfur-containing amino acids such as methionine.

The proposed mechanistic basis for the association between adverse fetal

outcomes and heritable thrombophilias has focused on impaired placental

development and function, secondary to venous and/or arterial thrombosis at

the maternal–fetal interface. These findings have been noted in the placentas of

women with adverse fetal outcomes and known inherited thrombophilias and

have been demonstrated in patients with similar outcomes, but lacking inherited

thrombophilic risk (18–22). Detractors of placental thrombosis as causal in early

pregnancy losses (<10 weeks’ gestation) cite an elegant series of experiments

demonstrating that maternal blood flow into the intervillous spaces of the human

placenta does not occur until approximately 10 weeks of gestation (23–26). Prior

to the establishment of intervillous circulation, nutrient transfer from maternal

1972blood to fetal tissues appears to be dependent on transudation that, in turn, relies

on flow through the uterine vasculature. This suggests that maternal or fetal

thrombotic episodes in the developing placental vasculature could be equally

devastating prior to or after the establishment of intervillous circulation near 10

weeks of gestation. Very early pregnancy losses (biochemical, anembryonic) and

known aneuploid fetal losses are unlikely to be altered by the presence of, or

treatment for, an underlying thrombophilic state.

[5] Neither the American College of Obstetrician Gynecologists (ACOG) nor the American Society of Reproductive Medicine (ASRM) recommend screening for inherited thrombophilias or MTHFR mutations for patients with recurrent pregnancy loss or any adverse pregnancy outcome (338,345) unless other risk factors are present, such as a history of thrombosis in the patient or a close family member. Despite the lack of evidence and guidelines against routine screening for thrombophilia in recurrent pregnancy loss, many clinicians routinely order these tests (347,348).

Anatomic Abnormalities

Anatomic abnormalities of both the uterine cervix and the uterine body have been associated with recurrent pregnancy loss (28,29). These anatomic causes may be either congenital or acquired. During development, the uterus forms via the

apposition of a portion of bilateral hollow tubes called the müllerian ducts. The

dissolution of the walls of these ducts along their site of apposition allows

formation of the intrauterine cavity, the intracervical canal, and the upper vagina.

Congenital uterine anomalies may therefore include incomplete müllerian duct

fusion, incomplete septum resorption, and uterine cervical anomalies. Although

the causes underlying many of the congenital anomalies of the female

reproductive tract are unclear, it has been well-documented that prenatal exposure

to maternally ingested diethylstilbestrol (DES) results in complex congenital

uterine, cervical, and vaginal changes.

Historically, all congenital reproductive tract abnormalities have been linked to

isolated spontaneous pregnancy loss and recurrent pregnancy loss, although the

presence of an intrauterine septum and prenatal exposure to DES demonstrate the

strongest associations (30–32). Women with an intrauterine septum may have as

high as a 60% risk for spontaneous miscarriage (33). Uterine septum–related

losses most frequently occur during the second trimester (34). If an embryo

implants into the poorly developed endometrium overlying the uterine septum,

abnormal placentation and resultant first trimester losses may occur (35). The

most common uterine congenital anomaly associated with in utero DES exposure

is hypoplasia, which may contribute to first- or second-trimester pregnancy

losses, incompetent cervix, and premature labor (36,37). Congenital anomalies of

1973the uterine arteries may contribute to pregnancy loss via adverse alterations in

blood flow to the implanted blastocyst and developing placenta (38).

Acquired anatomic anomalies have likewise been linked to isolated and

recurrent pregnancy losses. These abnormalities include such disparate conditions

as intrauterine adhesions, uterine fibroids, and endometrial polyps. Endometrium

that develops over an intrauterine synechiae or over a fibroid that impinges in the

intrauterine cavity (submucous) may be inadequately vascularized (39). This may

promote abnormal placentation for any embryo attempting to implant over such

lesions. Although data supporting these concepts are limited, this abnormal

placentation may lead to spontaneous pregnancy loss. Less clear is the association

between intramural fibroids and recurrent pregnancy loss, but it is suggested that

large (≥5 cm) intramural fibroids are associated with pregnancy loss and that

removal improves outcomes (see Chapter 11) (32,40).

Endocrine Abnormalities

The endocrinology of normal pregnancy is complex. Because spontaneous

pregnancy is critically dependent on appropriately timed endocrinologic

changes of the menstrual cycle, it is not surprising that those endocrine

abnormalities that alter pregnancy maintenance may have their effects

during the follicular phase of the cycle in which conception occurs, or even

earlier. Modifications in follicular development and ovulation, in turn, may be

reflected in abnormalities of blastocyst transport and development, alterations in

uterine receptivity to the implanting blastocyst, and improper functioning of the

corpus luteum. Endocrinologic factors suggested to be associated with recurrent

miscarriage include luteal phase insufficiency, hypersecretion of luteinizing

hormone (LH), thyroid disease, and, potentially, insulin resistance and polycystic

ovarian syndrome (PCOS), diabetes mellitus, hyperprolactinemia, and decreased

ovarian reserve.

Beginning with ovulation and lasting until approximately 7 to 9 weeks of

gestation, maintenance of early pregnancy depends on the production of

progesterone by the corpus luteum. Normal pregnancies are characterized by a

luteal–placental shift at about 7 to 9 weeks of gestation, during which the

developing placental trophoblast cells take over progesterone production and

pregnancy maintenance (41). Spontaneous pregnancy losses occurring before 10

weeks of gestation may result from a number of alterations in normal

progesterone production or utilization. These include failure of the corpus luteum

to produce sufficient quantities of progesterone, impaired delivery of

progesterone to the uterus, or inappropriate utilization of progesterone by the

uterine decidua. Pregnancy failures may occur near the time of the expected

luteal–placental shift if the trophoblast is unable to produce biologically active

1974progesterone following demise of the corpus luteum.

Luteal phase insufficiency or luteal phase defects are characterized by

inadequate luteal milestones and most likely relate to adverse pregnancy outcome

via inadequate or improperly timed endometrial development at potential

implantation sites. An elegant description of abnormalities at the site of

implantation, which may be responsible for some cases of recurrent pregnancy

loss, describes impaired decidualization of the endometrium as a mechanism for

natural selection of human embryos (42). Luteal phase defect has many causes,

some of which are associated with hypersecretion of LH. Although the

mechanism underlying the association of elevated LH levels with recurrent

pregnancy loss is incompletely understood, abnormal LH secretion may have

direct effects on the developing oocyte (premature aging), on the endometrium

(dyssynchronous maturation), or both. Many patients with elevated LH levels also

display physical, endocrinologic, and metabolic characteristics of PCOS.

Some studies report ovarian radiologic evidence of PCOS in as many as 40% to

80% of recurrent pregnancy loss patients (43,44). In addition to inappropriately

elevated LH levels, PCOS patients are frequently obese and often have elevated

circulating androgen levels. Although controversial, both changes have been

linked to recurrent pregnancy loss, and elevated androgen levels have been shown

to adversely affect markers of uterine receptivity in women with a history of

recurrent pregnancy loss (43–46). Many women with PCOS have metabolic

alterations in glycemic control characterized by insulin resistance. This too may

be directly or indirectly related to adverse pregnancy outcome, and it may explain

increases in the rate of spontaneous pregnancy loss among women with type 2

diabetes mellitus (47). Women with overt insulin-dependent diabetes mellitus

(IDDM) appear to exhibit a threshold of pregestational glycemic control above

which spontaneous pregnancy loss is increased (48,49). Hyperglycemia has been

directly linked to embryonic damage (50). In cases of advanced IDDM with

accompanying vascular complications, compromised blood flow to the uterus

may be mechanistically involved in subsequent pregnancy loss.

Patients with thyroid disease often have concomitant reproductive

abnormalities, including ovulatory dysfunction and luteal phase defects. The

metabolic demands of early pregnancy mandate an increased requirement for

thyroid hormones. It is therefore not surprising that hypothyroidism has been

associated with isolated spontaneous pregnancy loss and with recurrent pregnancy

loss (51). The definition of hypothyroidism is itself now under scrutiny with

many investigators suggesting that cutoff thyroid-stimulating hormone (TSH)

values during pregnancy should be less than 2.5 mIU/mL (52). Others have

suggested even lower TSH cutoff values (53). Although it has long been debated

whether clinically euthyroid patients with antithyroid antibodies have higher rates

1975of miscarriage and recurrent pregnancy loss, thyroid hormone supplementation

was shown to reduce miscarriage in infertility patients with positive antithyroid

antibodies and otherwise normal TSH levels undergoing IVF (54–60). The

mechanism for an association between antithyroid antibody positivity and

recurrent pregnancy loss remains unclear; however, these antibodies could be

markers of more generalized autoimmunity or may predict an impaired ability of

the thyroid gland to respond to the demands of pregnancy.

Elevated prolactin levels have been associated with miscarriage and recurrent

miscarriage. Hyperprolactinemia has been associated with ovulatory dysfunction

and may result in miscarriage via alterations in the hypothalamic–pituitary–

ovarian axis resulting in impaired folliculogenesis, oocyte maturation or luteal

phase dysfunction (61–63). One study in patients with hyperprolactinemia and

recurrent pregnancy loss showed a higher live birth rate in a subsequent

pregnancy after normalization of prolactin levels with a dopamine agonist (63).

Some studies have shown an association between diminished ovarian reserve

(DOR) and recurrent miscarriage (64,65). Shahine et al. suggests that this link

may result from an association between DOR and increased risk of aneuploid

embryos (349). Aneuploidy (chromosomal aberrations) is the most common cause

of first trimester miscarriage and women with DOR have a higher percentage of

aneuploid embryos compared to age-matched controls in both the recurrent

pregnancy loss (349) and the nonrecurrent pregnancy loss patient populations

(350). Ovarian reserve testing (early follicular phase follicle stimulating hormone

level, estradiol, and anti-müllerian hormone levels) is not currently recommended

by ASRM for the evaluation of recurrent miscarriage. DOR can explain higher

risk of miscarriage caused by higher risk of aneuploidy and help guide patients

regarding the next steps in care. Increasingly, IVF with chromosomal screening of

embryos is offered as a treatment option for women with recurrent pregnancy loss

(351) but success is minimal in the setting of DOR and may not be higher than

success without intervention (352).

Maternal Infection

The association of infection with recurrent miscarriage is among the most

controversial and poorly explored of the potential causes for pregnancy loss.

Reproductive tract infection with bacterial, viral, parasitic, zoonotic, and

fungal organisms have all been theoretically linked to pregnancy loss;

however, mycoplasma, ureaplasma, chlamydia, and a-streptococcus are the

most commonly studied pathogens (66,67). There is data that directly addressed

the roles of some of these proposed organisms in recurrent pregnancy loss. One

prospective comparison trial involving 70 recurrent pregnancy loss patients

reported no elevations in any markers for present or past infection with

1976Chlamydia trachomatis when compared with controls (68). In contrast, a very

large, prospective trial demonstrated a link between the detection of bacterial

vaginosis and a history of second trimester pregnancy loss among 500 recurrent

pregnancy loss patients (69). The risk of bacterial vaginosis detection was

positively correlated with cigarette smoking in this study.

The etiologic mechanism linking specific organisms to either isolated or

recurrent pregnancy loss remains unclear and must certainly differ among

infectious organisms. Certain viral organisms, such as herpes simplex virus

(HSV) and human cytomegalovirus (HCMV), can directly infect the placenta

and fetus (70,71). The resulting villitis and related tissue destruction may lead to

pregnancy disruption. Another theoretic possibility warranting study is that

infection-associated–early pregnancy loss may result from immunologic

activation in response to pathologic organisms. A large body of evidence supports

the role of this mechanism in adverse events later in gestation, such as intrauterine

growth restriction, premature rupture of membranes, and preterm birth (72,73).

Alternatively, mechanisms that protect the fetus from autoimmune rejection may

protect virally infected placental cells from recognition and clearance. This could

potentially promote periods of unfettered infectious growth for some of the

pathogenic organisms gaining entry to the reproductive tract (74).

Chronic endometritis, defined as the asymptomatic presence of histologically

documented plasma cells in an endometrial biopsy, may be associated with

recurrent miscarriage. Studies have shown a prevalence of 7% to 13% based on

morphologic assessment alone (353,354) and up to 56% in patients with recurrent

pregnancy loss (354). Chronic endometritis is associated with previous infections

and retained pregnancy tissue and, despite the difficulty of determining cause and

effect, treatment with antibiotics shows a higher live birth rate in subsequent

pregnancies in some studies (353,354). Endometrial biopsy for chronic

endometritis diagnosis and treatment are not recommended by expert groups and

larger studies are needed before implementing evaluation and treatment in routine

practice.

Immunologic Phenomena

[5] Embryo implantation and pregnancy maintenance is a delicate interaction and

balance between maternal and embryonic/fetal immune, hormonal, and

coagulation systems. Maternal rejection of paternal antigens and immune system

dysfunction as a cause of recurrent miscarriage has been explored extensively but

remains controversial (75). Laboratory testing for immune dysfunction can be

misleading and the treatments aimed at maternal immunosuppression can have

significant side effects. Expert groups have omitted immune testing and

recommended against extensive immunosuppression as a treatment for

1977unexplained recurrent miscarriage (338).

A brief review of some of the important concepts in basic immunology is

warranted to place testing and treatment discussions in context. Immune

responses classically are divided into innate and acquired responses. Innate

responses represent the body’s first line of defense against pathogenic

invasion. They are rapid and are not antigen specific. Cell types and mechanisms

typically considered vital to innate immunity include complement activation,

phagocytosis by macrophage, and lysis by natural killer (NK) and natural killer T

(NKT) cells and possibly by TCR γ δ+ T cells. Acquired immune responses, in

contrast, are antigen specific and are largely mediated by T cells and B cells.

Acquired responses can be further divided into primary (response associated

with initial antigen contact) and secondary (rapid and powerful amnestic

responses associated with subsequent contact to the same antigen).

Antigen specificity is regulated by two sets of genes in the major

histocompatibility complex (MHC), located on chromosome 6 in humans. MHC

class I molecules (human leukocyte antigen [HLA]-A, -B, and -C) are present

on the surface of nearly every cell in the human body and are important in

defense against intracellular pathogens, such as viral infection and oncogenic

transformation. MHC class I molecules act as important ligands for both the Tcell receptor on CD8+ cytotoxic/suppressor T cells and for a variety of receptors

on NK cells (76). MHC class II molecules (HLA-DR, HLA-DP, and HLADQ), in contrast, are present on the surface of a limited number of antigenpresenting cells, including dendritic cells, macrophage and monocytes, B

cells, and tissue-specific cells such as the Langerhans cells in the skin. These

molecules are important in defense against extracellular pathogens, such as

bacterial invaders. The major ligand for MHC class II is the T-cell receptor on

CD4+ T helper cells.

One very important concept in immunology that has particular

application to pregnancy is that of immune tolerance. The passage of bone

marrow–derived T cells through the fetal thymus during early development has

been well described. In this developmental interval, the T cells encounter a

process termed thymic education. During thymic education, T cells that express

either the CD4 or the CD8 coreceptor are chosen, and autoreactive cells are

effectively eliminated. This education promotes T-cell tolerance, allowing

selection and survival exclusively of those T cells that recognize nonself and will

not react against self (80,81).

Cellular Immune Mechanisms

Three main questions summarize much of the theoretical thinking surrounding

pregnancy maintenance and reproductive immunology:

19781. Which immune cells populate the reproductive tract, particularly at

implantation sites?

2. How do the characteristics of antigen presentation differ at the maternal–fetal

interface?

3. What regulatory mechanisms specifically affect reproductive tract immune

cells?

Resident Cells: Natural Killer Cells and T Regulatory Cells

Immune cells populating the reproductive tract exhibit many characteristics

that distinguish them from their peripheral counterparts. In particular, the

human endometrium is populated by T cells, macrophage, and NK-like cells,

but very few B cells are present. The relative proportions of these resident cells

vary with the menstrual cycle and change dramatically during early pregnancy.

Surrounding the time of implantation, one particular cell type comprises between

70% and 80% of the total endometrial lymphocyte populations (82,83). This cell

type is called a variety of names, including decidual granular lymphocytes

(DGLs), large granular lymphocytes (LGLs), and decidual NK cells. This

heterogeneity of names reflects the fact that this particular cell type differs from

similar cells isolated from the periphery, although most believe it to be an NK cell

variant. While most peripheral NK cells have low cell surface expression of CD56

(CD56dim) and express high levels of CD16, the immunoglobulin receptor

responsible for NK-mediated, antibody-dependent cellular cytotoxicity, those in

the uterine decidua and at the placental implantation site, are largely CD56bright

and CD16dim or CD16− (84). If these unusual endometrial cells are considered

NK cells, the implantation site represents the largest accumulation of NK cells in

any state of human health or disease. These decidual NK cells display fairly poor

cytotoxic function but are robust cytokine secretors (85,86). The balance of

activating and inhibitory receptors expressed on their cell surfaces determines

their ultimate killing versus secretion patterns (87,88).

The role of NK cells in reproduction and recurrent pregnancy loss is

controversial. Some studies show increased levels of CD56 NK cells (those with

killer-type activating receptors) in peripheral blood of women with recurrent

pregnancy loss compared to controls (89,90) and other studies do not show a

difference (355,356). The studies showing a difference may be inherently flawed.

Most of the recurrent miscarriage patients included in the study were nulliparous

and the controls were multiparous (357); a previous successful pregnancy can

induce permanent changes in NK subsets (358). Some studies have shown that a

higher level of NK cytotoxicity in the peripheral blood before pregnancy resulted

in a higher rate of pregnancy loss (359), while other studies showed similar NK

1979cytotoxicity in recurrent pregnancy loss patients with subsequent pregnancy loss

or live birth (360). Results from studies of NK cell levels in endometrial tissues in

recurrent miscarriage vary; some showing higher levels of CD56 NK cells in the

endometrial tissue from recurrent pregnancy loss patients (94) and others showing

lower levels (95). Interpretation of NK cell levels in endometrial tissue can be

limited by the inherent variability of levels throughout the different menstrual

cycle stages and the questionable testing accuracy using either immunochemistry

or flow cytometry (361). A strong argument against using any NK cell testing in

the evaluation of recurrent miscarriage patients comes from a large prospective

study showing that high NK toxicity before pregnancy did not impact subsequent

pregnancy loss rate (362).

A subpopulation of CD4+ T cells has been described that, like all activated

lymphocytes, strongly expresses CD25 on their cell surface (77). These CD4+

CD25+ cells are more specifically identified by the intracellular presence of the

forkhead box P3 (Fox P3) transcription factor and have been called regulatory T

lymphocytes (Treg cells). Treg cells, when activated by autoantigens, can

suppress activated inflammatory cells. They secrete regulatory cytokines,

including interleukin-10 (IL-10) and transforming growth factor-β (TGF-β) (77–

79). They may impact avoidance of tissue destruction associated with

inflammation and may play a central role in pregnancy maintenance. In pregnant

mice and women, these specialized CD4+ cells are systemically expanded in an

alloantigen independent fashion and can suppress adverse maternal responses to

the fetus (92) and to self (93).

The human decidua contains characteristic immune effector cells. These

immune cell populations are reported to be altered in recurrent pregnancy loss

patients but not in patients experiencing isolated spontaneous pregnancy losses

(91,94–96). Most investigations into whether alterations in these cells (including

T cells, decidual NK cells, and NKT cells) determine pregnancy outcome are

hampered by insufficient patient numbers to allow meaningful conclusions. [5]

Further research is needed to understand the role of NK cells and T regulatory

cells in reproduction and miscarriage. Expert groups do not recommend testing

recurrent pregnancy loss patients for NK cell cytotoxicity or Treg cells and any

treatment targeted toward suppression of these immune cells should be conducted

only in the setting of a clinical trial.

Antigen Presentation at the Maternal–Fetal Interface

Historically, it was proposed that the implanting trophoblastic allograft could

make itself antigenically invisible to avoid immune detection by the maternal

host. It could downregulate its expression of the MHC-encoded transplantation

antigens (some of which would be of paternal origin) and thereby avoid

1980recognition as nonself. Although immunology studies rendered this theory

obsolete, the implanting fetus does utilize this strategy to the extent (97) that

placental trophoblast cells do not express MHC class II molecules (98,99). Unlike

nearly every other cell in the human body, trophoblast cells do not express

the classical MHC class I transplantation antigens HLA-A and -B. Rather, a

subpopulation of placental cells, specifically the extravillous cytotrophoblast

cells, express the classical MHC class I HLA-C products, and the nonclassical

HLA-E, -F, and -G products (74,100–104). These extravillous cytotrophoblast

cells are of particular interest because they are characterized by remarkable

invasive potential (105,106). These cells move from the tips of the anchoring

villae of the human placenta, invading deeply into the maternal decidua, and some

replace cells within the walls of decidual arterial vessels (105–107). Although the

invasive characteristics of extravillous cytotrophoblast may reflect non-MHC–

related mechanisms, including well-described integrin switching, the intimate

contact of these fetal-derived cells with maternal immune effector cells exposes

the fetus to recognition as nonself (108) and it is now apparent that the pregnant

woman recognizes and responds to paternally-derived antigens. These responses

appear to be essential to pregnancy maintenance but must be tightly regulated.

It is not known why all placental cells downregulate expression of HLA-A and

-B, whereas invasive extravillous cytotrophoblast express HLA-C, -E, and -G.

This area of investigation is ripe with hypotheses and study. Because NK cells of

the innate immune system recognize and kill cells that express no MHC, the

complete downregulation of MHC should cause trophoblast cells to act as targets

for those NK cells that are pervasive at sites of implantation (88,97). In addition

to possible protection from direct NK cell–mediated killing, the expression of

HLA-C, -E, and -G by trophoblast cells may serve a variety of alternate purposes.

NK cell receptor–mediated interactions with extravillous cytotrophoblast MHC

products modulate cytokine expression profiles at the maternal–fetal interface

(85,86). MHC expression aids in decidual and vascular invasion by the

trophoblast, an activity essential for proper placental development (109).

Whereas definitive correlations between placental MHC class I expression

patterns and recurrent pregnancy loss remain contentious, trophoblast expression

of HLA-G was linked to other disorders of placental invasion, such as

preeclampsia (109,110). Genetic mutations at the HLA-G locus have been linked

to recurrent pregnancy loss in some, but not all studies (111–114). Soluble or

secreted trophoblast MHC products may aid in the development of maternal

immune tolerance toward the placenta (115). Soluble HLA-G has been shown to

suppress T lymphocyte and NK cell function and to induce the expansion of Treg

cells in humans (116).

Aberrant expression of class II MHC determinants, or enhanced expression of

1981MHC class I on syncytiotrophoblast in response to IFN-γ (117) could mediate

pregnancy loss by enhancing cytotoxic T cell attack (118). This theory appears

unlikely, because the expression of classical MHC antigens does not seem to be

induced on aborted tissues from women experiencing one or more pregnancy

losses (118). MHC class II genotypes appear to affect susceptibility to a variety of

diseases, including diabetes and other autoimmune diseases. A similar link

between MHC class II typing and adverse pregnancy outcome was reported

for recurrent pregnancy loss (119,120).

Regulation of Decidual Immune Cells

The characteristics of the interactions between decidual immune effector cells and

the implanting fetus may be determined by factors other than those already

mentioned. As might be predicted, these regulatory effects are often targets for

investigative efforts, because they may offer more direct insight into potential

therapies for immune-mediated disorders of pregnancy maintenance. Three such

regulatory mechanisms will be discussed here: (a) alterations in T helper cell

phenotypes; (b) reproductive hormones and immunosuppression; and (c)

tryptophan metabolism.

Antigen-stimulated immune responses involving CD4± T cells can be divided

into two major classes: T helper 1 (TH1) responses and T helper 2 (TH2)

responses. This subclassification is overly simplified, but it has been useful in

broadly defining types of immune responses based on the characteristics of the

CD4± cells present, and their associated cytokines. Cytokines are signaling

molecules secreted from immune cells that bind to receptors on other immune

cells to either stimulate or inhibit function. TH1 and TH2 immune cells express

different cytokines with different functions and studies suggest that successful

embryo implantation ultimately occurs in a TH2 dominant environment and that

poor pregnancy outcomes may result if the balance favors a TH1 dominant

environment (363). Studies have suggested that TH1-type cytokines can be

harmful to implanting developing embryo (121,122) and that some patients with

recurrent pregnancy loss exhibit a local dysregulation of their T helper cellular

immune response to antigens, with typical shifts toward TH1 inflammatory

responses at the maternal–fetal interface (123,124). This remains difficult to study

and prove given that the dominant immune cell population (TH1 or TH2) is

usually determined by peripheral blood levels of cytokines and cytokines display

their results at close range. Measurements of cytokines in peripheral blood,

endometrial biopsies, or flushing of decidual tissue are subject to technical

difficulties and conclusions about cytokines and their role in recurrent pregnancy

loss are yet to be determined. TNF-α is one cytokine that can usually be detected

easily in the blood with most assays and may be a good marker for inflammation.

1982Studies have shown higher levels of TNF-α in women with recurrent pregnancy

loss in early pregnancy (364).

Although many mechanisms are aimed at avoiding maternal immune

recognition of the implanting fetus, research in humans and animals

indicates that immune responses to fetal antigens are present (126–128).

Thus, the regulation of this response at the maternal–fetal interface may be

critical. Reproductive hormones have dramatic effects on peripheral cell–

mediated immunity, as demonstrated by well-documented and notable gender

differences in immune responsiveness (129). The levels of these potentially

immunosuppressive hormones are quite elevated in the circulation of pregnant

women. The fact that the levels of these hormones at the maternal–fetal interface

may be dramatically higher than those in the maternal circulation during

pregnancy may help to explain an apparent inconsistency: with some exceptions,

overall immune responsiveness during pregnancy appears to change a little, while

local modulation at the maternal–fetal interface may be vital (130).

It has been suggested that the immunosuppressive effects of progesterone

within the reproductive tract are at least partially responsible for the maintenance

of the semiallogeneic implanting fetus (131). In vitro studies have shown that

progesterone mediates its suppression of T cell effector function by altering

membrane-resident potassium channels and cell membrane depolarization. This

action, in turn, affects intracellular calcium signaling cascades and gene

expression and may be mediated by nonclassical steroid receptors or may not

involve a receptor at all (132–134). Progesterone-mediated changes in T cell gene

expression have been associated with the development of TH2-type T helper cell

responses and with increased leukemia inhibitory factor (LIF) expression (135).

Because a shift in the intrauterine immune environment from TH2 to TH1 has

been linked with early spontaneous pregnancy loss, the elevated intrauterine

concentrations of progesterone characteristic of early pregnancy may promote an

immune environment favoring pregnancy maintenance (123). To this point, in

vitro evidence indicates that progesterone can inhibit mitogen-induced

proliferation of and cytokine secretion by CD8+ T cells and can alter the

expression of a transcription factor that drives the development of TH1 cells

(136,137).

Levels of estrogen rise markedly during pregnancy, and attention has focused

on the role of estrogen in immune modulation. A group of animal studies showed

that estrogens improve immune responses in males after significant trauma and

hemorrhage, suppress cell-mediated immunity after thermal injury, and protect

against chronic renal allograft rejection (138–140). In vitro, estrogens appear to

downregulate delayed-type hypersensitivity (DTH) reactions and promote the

development of TH2-type immune responses, particularly at the elevated estrogen

1983concentrations typical of pregnancy (141,142). In mice, elevations in estrogen

have been shown to recruit Treg cells to the uterus (390).

One additional regulatory mechanism proposed for the induction of maternal

tolerance to the fetal allograft involves the amino acid tryptophan and its

catabolizing enzyme indoleamine 2,3-dioxygenase (IDO). The IDO hypothesis of

tolerance in pregnancy rests on data showing that T cells need tryptophan for

activation and proliferation (143) and local alterations in tryptophan metabolism

at the maternal–fetal interface could either activate or fail to suppress maternal–

antifetal immunoreactivity (144). Studies in mice have shown that the inhibition

of IDO leads to loss of allogeneic, but not syngeneic, fetuses, and that this effect

is mediated by lymphocytes (145). Further support lies in studies demonstrating

that hamsters fed diets high in tryptophan have increased rates of fetal wastage

(146). Extending this theory to humans requires further investigation. However,

the demonstration of IDO expression in human uterine decidua, and the

documentation of alterations in serum tryptophan levels with increasing

gestational age during human pregnancy support further interest in this potential

local immunoregulatory mechanism (147,148).

Humoral Immune Mechanisms and Antiphospholipid Syndrome

Humoral responses to pregnancy-specific antigens exist, and patients with

recurrent pregnancy loss can display altered humoral responses to

endometrial and trophoblast antigens (123,149). Nevertheless, most literature

surrounding humoral immune responses and recurrent pregnancy loss focus on

organ nonspecific autoantibodies associated with the antiphospholipid syndrome

(APS). Historically, these immunoglobulin-G (IgG) and IgM antibodies were

considered to be directed against negatively charged phospholipids. Those

phospholipids most often implicated in recurrent pregnancy loss are cardiolipin

and phosphatidylserine. However, antiphospholipid antibodies often are directed

against a protein cofactor, β2 glycoprotein 1, which assists antibody association

with phospholipid (150–154). Antiphospholipid antibodies were originally

characterized solely by prolonged phospholipid-dependent coagulation tests in

vitro (activated partial thromboplastin time [aPTT], Russell viper venom time)

and by thrombosis in vivo. The association of these antiphospholipid antibodies

with thrombotic complications has been termed the APS, and although many of

these complications are systemic, some are pregnancy specific—spontaneous

miscarriage, stillbirth, intrauterine growth retardation, and preeclampsia

(155,156). A reassessment of the criteria used to diagnose APS resulted in

additions to the prior Sapporo criteria for diagnosis of APS and continues to

include adverse pregnancy outcomes. These criteria, which have been validated

clinically, are as follows (156–158):

1984For a patient to be diagnosed with antiphospholipid syndrome, one or more

clinical and one or more laboratory criteria must be present:

Clinical

1. One or more confirmed episode of vascular thrombosis of any type:

Venous

Arterial

Small vessel

2. Pregnancy complications:

Three or more consecutive spontaneous pregnancy losses at less than

10 weeks of gestation with exclusion of maternal anatomic and

hormonal abnormalities and exclusion of paternal and maternal

chromosomal abnormalities

One or more unexplained deaths of a morphologically normal fetus at

or beyond 10 weeks of gestation (normal fetal morphology documented

by ultrasound or direct examination of the fetus)

One or more premature births of a morphologically normal neonate at

or before 34 weeks of gestation secondary to severe preeclampsia or

placental insufficiency

Laboratory

Testing must be positive on two or more occasions with evaluations 12 or more

weeks apart:

1. Positive plasma levels of anticardiolipin antibodies of the IgG or IgM isotype

at medium to high levels

2. Positive plasma levels of lupus anticoagulant

3. Anti-β2 glycoprotein-1 antibodies of the IgG or IgM isotype in titers greater

than the 99th percentile

The presence of antiphospholipid antibodies (anticardiolipin or lupus

anticoagulant) and anti-β2 glycoprotein-1 antibodies during pregnancy is a

major risk factor for an adverse pregnancy outcome (150,151,159). In a large

series of couples with recurrent pregnancy loss, the incidence of the APS was

between 3% and 5% (66). The presence of anticardiolipin antibodies among

patients with known systemic lupus erythematosus portends less favorable

pregnancy outcomes (160).

Several mechanisms have been proposed by which antiphospholipid

1985antibodies might mediate pregnancy loss (161). Antibodies against

phospholipids could increase thromboxane and decrease prostacyclin synthesis

within placental vessels. The resultant prothrombotic environment could promote

vascular constriction, platelet adhesion, and placental infarction (162–164).

Alternatively, in vitro evidence from trophoblast cell lines indicates that IgM

action against phosphatidylserine inhibits formation of syncytial trophoblast

(165). Syncytialization is required for proper placental function. One study

demonstrated that both extravillous cytotrophoblast and syncytiotrophoblast cells

synthesize β2 glycoprotein-1, the essential cofactor for antiphospholipid antibody

binding (166). Although it gives insight into pathophysiology, the prognostic

value of serum levels of specific antibodies against β2 glycoprotein-1 with respect

to pregnancy outcome among recurrent pregnancy loss patients is contentious and

may be poorer than that of standard anticardiolipin antibodies (167–169). Some

have proposed that sera from antibody positive recurrent pregnancy loss patients

are particularly adept at inhibiting trophoblast adhesion to endothelial cells in

vitro (170). Others noted rapid development of atherosclerosis in the decidual

spiral arteries of patients who test positive for antiphospholipid antibodies (171).

Still others have demonstrated that levels of the placental antithrombotic molecule

—annexin V—are reduced within the placental villa from those women with

recurrent pregnancy loss who are antiphospholipid antibody positive (172).

However, placental pathologic evidence that supports causal involvement of the

antiphospholipid syndrome in pregnancy loss is equivocal. The characteristic

lesions for this syndrome (placental infarction, abruption, and hemorrhage) are

typically missing in women with antiphospholipid antibodies, and these same

pathologic lesions can be found in placentae from women with recurrent

miscarriages who do not have biochemical evidence of antiphospholipid

antibodies (161,173–175).

One additional group of autoantibodies that have been linked to recurrent

pregnancy loss is the antithyroid antibodies. Although the data remain somewhat

controversial, several investigators demonstrated an increased prevalence of these

antibodies among women with a history of recurrent pregnancy loss, even in the

absence of thyroid endocrinologic abnormalities (52,56,57,176–178).

Other antibody-mediated mechanisms for recurrent pregnancy loss have

been proposed, including antisperm and antitrophoblast antibodies, and

blocking antibody deficiency but none have withstood the test of time.

It is important to suggest during this in-depth discussion of the immunemediated mechanisms of isolated and recurrent pregnancy loss that pregnancy

may not require an intact maternal immune system. Supporting this concept are

data showing that agammaglobulinemic animals and women can successfully

reproduce (180). Further, viable births occur among women with severe immune

1986deficiencies and in murine models that lack T and B cells (severe combined

immunodeficiency [SCID] mice) and those that display a congenital absence of

their thymus (nude mice).

[5] Despite the knowledge that most first trimester miscarriages are caused by

aneuploidy and that the only immune-related issue recommended for routine

evaluation and treatment for recurrent miscarriage by expert groups is APS,

providers around the world still test and treat for many different immune issues in

couples with recurrent pregnancy loss. Patients and providers feel pressured for

answers and interventions in the setting of unexplained recurrent pregnancy loss

and immune issues play into the woman’s feelings of guilt and worry that there is

something wrong with her that is causing her to reject her pregnancy. Immunerelated dysfunction in relationship to miscarriage and recurrent pregnancy loss

remains up for debate, additional research, and better understanding.

Immunosuppressive treatments can result in harm and until more is understood,

intervention should be limited to investigations in the setting of Institutional

Review Board (IRB)-approved clinical trials.

Male Factors

Most publications that review testing and treatment for recurrent pregnancy loss

couples, including this chapter, recommend only a single test for the male partner

in the couple—a peripheral blood karyotype. The role of the male partner in the

etiology of recurrent pregnancy loss is understudied, but there is a growing body

of literature suggesting that the development and validation of novel testing and

treatment regimens for the male may prove beneficial (181,182). Detailed

peripheral chromosomal testing of men whose partners experienced recurrent

pregnancy loss revealed an increased incidence of Y chromosome microdeletions

when compared to male partners in fertile and infertile couples (183). Small

studies demonstrated that male partners in couples experiencing recurrent

pregnancy loss have an increased incidence of sperm chromosomal aneuploidy,

particularly sex chromosome disomy, when compared to fertile men (184,365–

367). The addition of specialized testing to a standard semen analysis among men

in couples with recurrent loss revealed reductions in sperm functional testing

(hypo-osmotic swelling, acrosome status, nuclear chromatin decondensation) and

increased DNA fragmentation and lipid peroxidation when compared to fertile

men or historical controls (185–187). The latter results suggest that these men

may have abnormal levels of reactive oxygen species in their semen or that their

sperm cells are particularly sensitive to these compounds. Paternal carriage of the

MTHFR C677T mutation and hyperhomocysteinemia were associated with DNA

damage and recurrent pregnancy loss (188). A single, small, uncontrolled

treatment study using antioxidants among male partners in recurrent pregnancy

1987loss couples who had high levels of sperm DNA damage or semen lipid

peroxidation suggested favorable treatment outcomes (186). The only evidencebased recommended test for a male partner in a couple with recurrent pregnancy

loss is a peripheral blood karyotype and the other tests and interventions remain

under investigation. With more focus on male factor in all aspects of

reproduction, more information may surface about the male factor in miscarriage.

Other Factors

It is increasingly evident that the implantation of the blastocyst within the

uterine decidua involves an exquisitely scripted crosstalk between embryo

and mother. Alterations in this dialogue often result in improper implantation

and placental development. For instance, recurrent pregnancy loss has been linked

to a dysregulation in the expression patterns of vascular endothelial growth

factors (VEGFs) in the developing placenta and their requisite receptors within

the maternal decidua (189). Cellular and extracellular matrix adhesion properties

may be involved in this dialogue. The concept of uterine receptivity has been

emboldened by the description of endometrial integrins and the timing of integrin

switching during implantation (190). Others have reported decreased levels of

endometrial mucin secretion and reductions in the endometrial release of soluble

intercellular adhesion molecule I among women with histories of recurrent

pregnancy loss (191,192). Programmed cell death (apoptosis) may play an

essential role in normal placental development. Alterations in two important

apoptotic pathways—Fas-Fas ligand and bcl2—have been linked to recurrent

pregnancy loss and poor pregnancy outcome (75,193).

Environmental Factors

A variety of environmental factors have been linked to sporadic and recurrent

early spontaneous pregnancy loss. These are difficult studies to perform, because,

in humans, they all must be retrospective and all are confounded by alternative or

additional environmental exposures. Nevertheless, the following factors have

been linked to pregnancy loss: exposure to medications (e.g., antiprogestogens,

antineoplastic agents, anti-inflammatory agents, and inhalation anesthetics),

exposure to ionizing radiation, prolonged exposure to organic solvents, and

exposure to environmental toxins, especially phthalates, bisphenol-A, and heavy

metals (194–197). The latter two exposures have been demonstrated to have

endocrine and immune effects that could lead to poor placentation and subsequent

pregnancy loss (198,199). Lathi et al. showed that maternal BPA levels were

directly associated with increased risk of miscarriage (368). Associations between

spontaneous pregnancy loss and exposures to video display terminals, microwave

1988ovens, high-energy electric power lines, and high altitudes (e.g., flight attendants)

are not substantiated (200,201). There is no compelling evidence that moderate

exercise during pregnancy is associated with spontaneous miscarriage.

Association between stress and miscarriage risk is controversial (369). In the

absence of cervical anatomic abnormalities or incompetent cervix, coitus does not

appear to increase the risk for spontaneous pregnancy loss (202,203).

Exposure to four particular substances—alcohol, cigarettes, marijuana, and

caffeine—deserves specific attention; all have been linked with increased risk of

miscarriage through studies with widely varied findings. Although some

conflicting data exist, one very large epidemiologic study has shown that alcohol

consumption during the first trimester of pregnancy, at levels as low as three

drinks per week, is associated with an increased incidence of spontaneous

pregnancy loss (204–206). Cigarette smoking has been linked to early

spontaneous pregnancy loss; however, this is not without controversy (207–

209). Alcohol and tobacco intake in the male partner correlates with the

incidence of domestic violence, which in turn is associated with early

pregnancy loss (210). Marijuana has not been shown to necessarily increase

the risk of miscarriage in the one study investigating association found in the

literature to date (370), however studies have shown poor effects on female

(371) and male (372) fertility and patients should be counseled to refrain

until further studies can evaluate risk further. Evidence adds to a growing

body of literature that suggests consumption of coffee and other caffeinated

beverages during early pregnancy is linked to adverse pregnancy outcomes

(207,211). One report casts doubt on the definition of a lower limit for safe use of

caffeine in the first trimester of pregnancy (207).

Maintaining a healthy weight may decrease risk of miscarriage as being

underweight (373) and overweight (374) have been associated with increased risk

of miscarriage. Risk of sporadic and recurrent miscarriage are increased with

obesity and etiologic theories include impact on both gamete quality and uterine

receptivity in autologous conception (375) and donor egg pregnancies (212).

PRECONCEPTION EVALUATION

[4] Investigative measures that are potentially useful in the evaluation of recurrent

spontaneous miscarriage include obtaining a thorough history from both partners,

performing a physical assessment of the woman (with attention to the pelvic

examination), and a limited amount of laboratory testing (Table 33-2).

Table 33-2 Investigative Measures Useful in the Evaluation of Recurrent Early

Pregnancy Loss

1989History

1. Pattern, trimester, and characteristics of prior pregnancy losses

2. History of subfertility or infertility

3. Menstrual history

4. Prior or current gynecologic or obstetric infections

5. Signs or symptoms of thyroid, prolactin, glucose tolerance, and hyperandrogenic

disorders (including polycystic ovarian syndrome)

6. Personal or familial thrombotic history

7. Features associated with the antiphospholipid syndrome (thrombosis, false positive

test for syphilis)

8. Other autoimmune disorders

9. Medications

10. Environmental exposures, illicit and common drug use (particularly caffeine,

alcohol, cigarettes, marijuana, and in utero diethylstilbestrol exposure)

11. Genetic relationship between reproductive partners

12. Family history of recurrent spontaneous abortion, of obstetric complications, or of

any syndrome associated with embryonic or fetal losses

13. Previous diagnostic tests and treatments, including, if available, chromosome

testing on products of conception

Physical Examination

1. General physical examination with attention to:

a. Obesity

b. Hirsutism/acanthosis

c. Thyroid examination

d. Breast examination/galactorrhea

e. Pelvic examination

1. Anatomy

2. Infection

3. Trauma

4. Estrogenization

5. Masculinization

Laboratory

19901. Parental peripheral blood karyotype (both partners)

2. Chromosome testing on products of conception

3. Hysterosalpingography, three-dimensional transvaginal sonography,

sonohysterography, or office hysteroscopy, followed by hysteroscopy/laparoscopy, if

indicated

4. Serum thyroid stimulating hormone level

5. Serum prolactin level

6. Serum HbA1c

7. Antiphospholipid screening

a. Anticardiolipin antibody levels (IgG and IgM)

b. Lupus anticoagulant (activated partial thromboplastin time or Russell viper

venom)

c. Anti-β2-glycoprotein-1 antibodies (IgG and IgM)

History

A description of all prior pregnancies, their sequence and whether histologic

assessment and karyotype determinations were performed on previously aborted

tissues are important aspects of the history. Approximately 60% of pregnancies

lost before 8 weeks of gestation are reported to be chromosomally abnormal;

most of these pregnancies are affected by some type of trisomy, particularly

trisomy 16 (215,216). The most common single chromosomal abnormality is

monosomy X (45X), especially among anembryonic conceptuses (217).

Aneuploid losses are particularly prevalent among women with recurrent

pregnancy loss who are over the age of 35 (8). Research suggests a higher rate of

aneuploid miscarriage in recurrent pregnancy loss women with DOR at a younger

age (349). Although somewhat controversial, the detection of aneuploidy in

miscarriage specimens may be lower when the couple experiencing recurrent

miscarriages is euploidic. Alternatively, some investigators have suggested that,

because aneuploidy is common among miscarriage specimens from patients

experiencing isolated and recurrent spontaneous pregnancy losses, if aneuploidy

is documented in fetal tissues from a recurrent pregnancy loss patient, this loss

does not affect their prognosis for future pregnancy maintenance (13).

Most women with recurrent pregnancy loss tend to experience

spontaneous miscarriage at approximately the same gestational age in

sequential pregnancies (218). Unfortunately, the gestational age when

pregnancy loss occurs, as determined by last menstrual period, may not be

informative, because there is often a 2- to 3-week delay between fetal demise and

signs of pregnancy expulsion (219). If patients have had ultrasounds in the first

1991trimester, careful history and detailed record review can reveal a more accurate

time of fetal demise than gestational age by last menstrual period or patient’s

recall. The designation into either primary or secondary categories is not helpful

in the diagnosis or management of couples experiencing recurrent miscarriage.

Approximately 10% to 15% of couples cannot be classified into either the

primary or secondary category because, although their first pregnancy resulted in

a loss, it was followed by a term delivery prior to subsequent losses. Expert

groups exclude biochemical miscarriages, defined as positive pregnancy test but

miscarriage before ultrasound or histopathologic evidence can be obtained, from

guidelines. This recommendation is not without controversy. Biochemical

miscarriages can be distressing for patients and providers can feel obligated or

unsure about investigating further. Some evidence suggests women with recurrent

biochemical miscarriages or a mix of biochemical and clinical miscarriages have

similar poor prognosis for subsequent pregnancies (376,377). Investigation is

encouraged for recurrent clinical miscarriage (ultrasound or histopathologic

evidence) (338) but not biochemical miscarriages.

It is important to glean any history of subfertility or infertility among

couples with recurrent pregnancy loss. This is defined as the inability to

conceive after 12 months of unprotected intercourse. By definition, 15% of

all couples will meet these criteria; this number increases to 33% among

couples with recurrent pregnancy losses. Because many pregnancies are lost

before or near the time of missed menses, subfertility among recurrent pregnancy

loss patients may, in some cases, reflect recurrent preclinical losses. Menstrual

cycle history may provide information about the possibility of oligo-ovulation or

other relevant endocrine abnormalities in recurrent pregnancy loss patients. A

personal and family history of thrombotic events or renal abnormalities may

provide vital information. A family history of pregnancy losses and obstetric

complications should be addressed specifically. Detailed information about drug

and environmental exposure should be obtained.

Physical Examination

A general physical examination should be performed to detect signs of

metabolic illness, including PCOS, diabetes, hyperandrogenism, and thyroid

or prolactin disorders. During the pelvic examination, signs of infection, DES

exposure, and previous trauma should be ascertained. Estrogenization of mucosal

tissues, cervical and vaginal anatomy, and the size and shape of the uterus should

be determined.

[4] Recommended Testing

1992[5] Laboratory assessment of couples with recurrent pregnancy losses should

include the following:

1. Chromosome analysis of the products of conception

2. Parental peripheral blood karyotyping with banding techniques

3. Assessment of the intrauterine cavity with either office hysteroscopy,

sonohysterography, three-dimensional transvaginal sonography, or

hysterosalpingography, followed by operative hysteroscopy if a

potentially correctable anomaly is found (220)

4. Thyroid function testing, including serum TSH levels

5. Anticardiolipin antibodies, anti-β2 glycoprotein-1 antibodies, and lupus

anticoagulant testing (aPTT or Russell viper venom testing)

6. Prolactin

7. HbA1c

Tests with Unproven Utility

A number of laboratory assessment tools are under investigation for use in

patients with a history of recurrent pregnancy loss. Results are either too

preliminary to warrant unfettered recommendation or studies of their use have

been too contradictory to allow final determination of their value.

Tests under investigation with evidence in favor of utilizing:

1. Evaluation of ovarian reserve using day 3 serum follicle–stimulating

hormone, estradiol, and anti-müllerian hormone levels. It appears that

decreased ovarian reserve may predict higher risk of aneuploid embryos

and miscarriage in all patients, including those with recurrent pregnancy

loss (64,65,349).

2. Testing for antithyroid antibodies among women with recurrent

pregnancy loss remains controversial, but is rapidly gaining support (52–

54,57,176–178). Investigators have demonstrated an increased prevalence

of these antibodies among women with a history of recurrent pregnancy

loss, even in the absence of thyroid endocrinologic abnormalities

(53,54,56,177).

3. Endometrial biopsy for plasma cells/chronic endometritis

Tests under investigation and the majority of research shows No clinical

benefit

a. Thrombophilia testing (FVL, prothrombin gene mutation, protein C

deficiency, protein S deficiency, antithrombin III deficiency,

homocysteine levels, MTHFR mutation). Once popular for patients with

1993recurrent pregnancy loss, evidence shows limited value in testing and or

treatment of thrombophilia testing to prevent miscarriage or poor

pregnancy outcome unless patients have a personal or family history of

thrombosis

b. Testing for peripheral evidence of TH1/TH2 cytokine dysregulation.

c. NK cell testing in peripheral blood or endometrium

d. Cervical cultures for mycoplasma, ureaplasma, and chlamydia may be

considered.

e. Testing for serologic evidence of PCOS using LH or androgen values may

be useful (43–46).

Tests not recommended for evaluation in patients with recurrent

pregnancy loss:

1. Evaluations that involve extensive testing for serum or site-specific auto- or

alloantibodies (including antinuclear antibodies and antipaternal cytotoxic

antibodies) are expensive and unproven. Their use often verifies the statistical

tenet that if the number of tests performed reaches a critical limit, the results of

at least one will be positive in every patient.

2. Testing for parental HLA profiles is never indicated in outbred populations.

Findings that HLA sharing is associated with poor pregnancy outcomes are

strictly limited to those specific populations studied, which have very high and

sustained levels of marriage within a limited community (179).

3. Use of mixed lymphocyte cultures has not proved useful. Use of other

immunologic tests is unnecessary unless these studies are performed, with

informed consent, under a specific study protocol in which the costs of these

experimental tests are not borne by the couple or their third-party payers.

4. Further work is necessary before suppressor cell or factor determinations,

cytokine, oncogene, and growth factor measurements, or embryotoxic factor

assessment can be clinically justified.

5. Endometrial biopsy for luteal phase defect

POSTCONCEPTION EVALUATION

Following conception, close monitoring of patients with histories of recurrent

pregnancy loss is advised to provide emotional support and to confirm

intrauterine pregnancy and its viability. Although controversial, some

studies recommend close monitoring in the first trimester of women with a

history of recurrent pregnancy loss resulting from a higher incidence of

ectopic pregnancy and complete molar gestation in this population (56,222–

226). Determining serum levels of a-hCG may be helpful in monitoring early

1994pregnancy until an ultrasonographic examination can be performed;

however, not all failing pregnancies will be preceded by inadequate a-hCG

levels (227). If used, serum a-hCG levels should be serially monitored from

the time of a missed menstrual period until the level is approximately 1,200

to 1,500 mIU/mL, at which time an ultrasonographic scan is performed and

blood sampling is discontinued. Other hormonal determinations are rarely of

benefit because levels are often normal until fetal death or miscarriage occurs

(228).

If a pregnancy has been confirmed, but clinical follow-up has met newly

defined criteria for a failed pregnancy (378), intervention is recommended to

expedite pregnancy termination and to obtain tissue for karyotype analysis. [9]

The importance of obtaining karyotypic analysis from tissues obtained after

pregnancy demise in a woman experiencing recurrent losses cannot be

overemphasized. Results may suggest karyotypic anomalies in the parents

and patients report their appreciation of receiving the information (379). The

documentation of aneuploidy may have important prognostic implications

and may direct future interventions. Cost analysis has demonstrated that

karyotypic analysis is financially prudent among patients with histories of

recurrent pregnancy loss (229). Analysis by traditional karyotype fails 20% to

40% of the time because of difficulties in growing cells from products of

conception and interpretation of results can be limited as a result of maternal cell

contamination (380). Techniques for analysis like comparative genomic

hybridization provide results in more than 90% of cases and can differentiate

46XX results from maternal cell contamination (230,380). This technology can be

used without fresh samples needed for cell growth and can even be used in

specimens that have been in formalin or paraffin (231).

In continuing pregnancies, first trimester screening using cell-free fetal DNA,

maternal chemistries, and fetal nuchal lucency measurement or chorionic villus

sampling is recommended for obstetrical indications. Amniocentesis may be

recommended to assess the fetal karyotype based on prior screening results. In the

future, fetal karyotype assessment may be performed using DNA isolated from

nucleated fetal erythrocytes in maternal blood (232) or from sampling of the

lower genital tract (381).

THERAPY

Advances in the treatment of patients with recurrent pregnancy loss have

been slow. Despite a rapid expansion in understanding the molecular and

subcellular mechanisms involved in implantation and early pregnancy

maintenance, extension of these concepts to prevention of recurrent early

1995pregnancy loss has lagged. Progress toward treatment of most causes of recurrent

pregnancy loss has been hampered by a variety of factors. The condition itself has

been inconsistently defined. The results of clinical trials involving recurrent

pregnancy loss patients are nearly impossible to compare and evaluate. Trial

design is frequently substandard, with lack of rationale, lack of appropriate

control groups, and poor statistical analysis, limiting the ability to draw concrete

conclusions from reported results. Epidemiologic data indicate that most patients

with a history of recurrent pregnancy loss will, in fact, have a successful

pregnancy the next time they conceive (7). For these reasons, with few

exceptions, most therapies for recurrent pregnancy loss must be considered

experimental. Until further study is completed, treatment protocols involving

these therapies should be undertaken only with informed consent and in the

setting of a well-designed, double-blind, placebo-controlled clinical trial.

[9] Common but controversial therapeutic options for patients with

recurrent pregnancy loss include the use of IVF with preimplantation

chromosomal screening of embryos, use of donor oocytes or sperm,

antithrombotic interventions, the repair of anatomic anomalies, the

correction of any endocrine abnormalities, the treatment of infections, and a

variety of immunologic interventions and drug treatments. Treatment for

patients with recurrent miscarriage should include close monitoring of

pregnancy, psychological counseling and support, and karyotypic analysis of

tissues from subsequent miscarriage.

Genetic Abnormalities

Evidence suggests that, in women with a history of three or more spontaneous

pregnancy losses, a subsequent pregnancy loss has a 58% chance of chromosomal

abnormality (382). Among women with recurrent pregnancy loss who are age 35

or older, the aneuploidy rate is much higher at 80% (8). Most chromosomal

abnormalities identified in miscarriages are autosomal trisomies and considered to

result from maternal nondisjunction. Maternal age appears as a consistent and

important risk factor for trisomy in most studies. DOR can be associated with

higher aneuploidy rate in young women with recurrent pregnancy loss (349).

There are several options for patients who suffer from recurrent pregnancy loss

who have an identified miscarriage caused by trisomy. The first is to conceive

again without any specific change in medical management, as these abnormalities

are sporadic and unlikely to recur. Studies examining patients with recurrent

pregnancy loss show that women who miscarry chromosomally abnormal

conceptions are more likely to achieve a live birth with subsequent pregnancy

than those who miscarry chromosomally normal conceptions (13,233). A second

treatment option is chromosomal screening of embryos before implantation,

1996commonly referred to as preimplantation genetic screening (PGS), and a third

option involves the use of donor gametes.

Because chromosomal abnormalities are the most commonly identified cause

of miscarriage, some have argued that the use of chromosomal screening of

embryos is indicated for patients with recurrent pregnancy loss. This technique

involves biopsying a single cell from a cleavage stage embryo (day 3) or many

labs are biopsying several cells from blastocysts (day 5 or 6). Genetic testing can

be performed on these cells to determine abnormalities in chromosome number

and morphology. Embryos that are diagnosed with genetic abnormalities would

be discarded and only those embryos with euploid results considered appropriate

for transfer into the uterus. The biopsying of cells from embryos to screen for

specific disease causing mutations like cystic fibrosis or sickle cells disease,

commonly referred to as preimplantation genetic diagnosis (PGD), is used in

many internationally recognized assisted reproductive technology centers,

however the use of chromosomal screening via biopsying embryos remains

controversial, especially for recurrent pregnancy loss patients.

The use of chromosomal screening of embryos has the potential to reduce the

incidence of pregnancy loss arising from a chromosome issue within the embryo.

However, definitive studies in this population have not yet been done. The use of

this technology requires the patient to undergo an IVF cycle to obtain embryos for

biopsy at significant cost and medical intervention. Although there are several

retrospective studies showing reduced miscarriage rates with this technique,

several prospective trials using the outcome of successful pregnancy per started

cycle fail to show significant benefit (234–244). IVF is invasive and costly and

many patients with recurrent pregnancy loss conceive quickly without

intervention and have a high likelihood of live birth with their subsequent

pregnancy. Therefore, the optimal control group for a study of IVF with

chromosomal screening in the recurrent pregnancy loss population is debated.

Should it be natural conceptions or IVF without testing of embryos? Finally, the

prognosis for a patient with recurrent pregnancy loss does seem to be linked to the

chromosome analysis of prior miscarriages. Recurrent pregnancy loss patients

who miscarry chromosomally abnormal embryos, seem to have better prognosis

than those who miscarry chromosomally normal conceptions, again arguing for

expectant management for recurrent pregnancy loss patients with a history of

aneuploid loss. On the other hand, patients with poorer prognosis are those who

miscarry chromosomally normal embryos and therefore would not benefit from

IVF and chromosomal screening of embryos. One intent to treat analysis showed

similar pregnancy outcomes in recurrent pregnancy loss patients who chose

expectant management compared to those who chose IVF with chromosomal

screening (383). The efficacy of chromosomal screening of embryos in the

1997treatment of patients with recurrent pregnancy loss continues to be investigated,

and the methods of embryo biopsy and genetic testing continue to evolve

(242,243). As these techniques improve and the understanding of aneuploidy and

recurrence improves, there may be a subset of recurrent pregnancy loss patients,

such as carriers of parental translocations, who might benefit from this

intervention. IVF with chromosomal screening of embryos can be considered, but

not strongly recommended, for patients with recurrent pregnancy loss.

An additional option for patients with recurrent miscarriage is the use donor

eggs or sperm. This treatment is useful for recurrent pregnancy loss patients with

specific parental genetic factors and couples with DOR. Patients with

robertsonian translocations involving homologous chromosomes have a genetic

anomaly which always results in unbalanced gametes, and the use of donor

oocyte or donor sperm is recommended. Use of donor gametes among patients

with a history of recurrent pregnancy loss can be useful in other cases where

couples are at higher risk for unbalanced offspring because one or both

prospective parents carries other chromosomal rearrangements, such as reciprocal

translocations. In these cases, use of donor gametes was demonstrated to be as

effective as its use in matched patients without such a history (245). In all cases of

balanced translocations or embryonic aneuploidy, genetic counseling is

recommended. Couples with DOR and/or advanced reproductive age have a

higher risk of miscarriage, aneuploidy (349), and limited success with IVF with

chromosomal screening and therefore should be counseled about the option of

using donor eggs.

Anatomic Anomalies

Hysteroscopic resection represents state-of-the-art therapy for submucous

leiomyomas, intrauterine adhesions, and intrauterine septa. Although its efficacy

has been debated, this approach appears to limit postoperative sequelae while

maintaining improved reproductive outcomes (30,31,245–249). Use has been

safely extended to patients with DES exposure, hypoplastic uteri, and

complicated septal anomalies (30,31,250). Attempts to improve on standard

hysteroscopic metroplasty, which is typically performed in the operating room

using general anesthesia, often with laparoscopic guidance, are under

investigation. Ultrasonographically guided transcervical metroplasty is reported to

be safe and effective (246). Ambulatory, office-based procedures, including

septum resection under fluoroscopic or ultrasound guidance, are attractive options

(247).

For patients with a history of loss secondary to cervical incompetence,

placement of a cervical cerclage is indicated. This is usually performed early in

the second trimester after documentation of fetal viability. Cervical cerclage

1998should be considered as a primary intervention for women with DES-associated

uterine anomalies.

Endocrine Abnormalities

Some investigators have proposed the use of ovulation induction for the treatment

of recurrent pregnancy loss (248,249). The theory behind its use in these patients

rests on hypotheses that ovulation induction is associated with healthier oocytes.

Healthier oocytes, in turn, may decrease the incidence of luteal phase

insufficiency through a more receptive hormonal environment, which should

result in improved pregnancy maintenance. This approach grossly oversimplifies

the mechanisms involved in implantation and early pregnancy maintenance. Until

appropriately studied, use of empiric ovulation induction for treatment of

unexplained recurrent pregnancy loss should be viewed with caution. Evidence

from small studies indicates such use is not effective (248). Still, use of ovulation

induction in some subsets of patients with recurrent pregnancy loss could be of

benefit. For instance, stimulating folliculogenesis with ovulation induction or

luteal phase support with progesterone should be considered for women with

luteal phase insufficiency. The efficacy of these therapies, however, is not

substantiated (250). Ovulation induction might be beneficial for women with

androgen and LH hypersecretion disorders, especially following pituitary

desensitization with gonadotropin-releasing hormone agonist therapy (66). This

treatment remains controversial because the only large, prospective, randomized

controlled trial reports no therapeutic efficacy; none for prepregnancy pituitary

suppression nor for luteal phase progesterone supplementation (251).

Links between PCOS, hyperandrogenism, hyperinsulinemia, and recurrent

pregnancy loss make use of insulin-sensitizing agents in the treatment of recurrent

pregnancy loss associated with PCOS attractive (43–46). Although further study

is needed, there are an increasing number of reports that support its use for this

application (252,253). Prepregnancy glycemic control is particularly important for

women with overt diabetes mellitus (47,49). Thyroid hormone replacement may

be helpful in cases of hypothyroidism. There are data indicating that thyroid

hormone therapy may be of some benefit in euthyroid recurrent loss patients with

antithyroid antibodies and possibly even in all pregnant women with “euthyroid”

TSH levels between 2.5 and 5.0 mIU/L (52–54,177). Use of dopamine agonists to

normalize prolactin levels in patients with recurrent pregnancy loss has improved

subsequent pregnancy outcomes in limited studies (63), but remains controversial

(384). There does not appear to be a place in the medical management of

recurrent pregnancy loss for administering bromocriptine to women who do not

have a prolactin disorder.

1999Infections

Empiric antibiotic treatment has been used for couples with recurrent miscarriage

but its efficacy is unproven. Elaborate testing for infectious factors among

recurrent pregnancy loss patients and use of therapeutic interventions is not

justified unless a patient is immunocompromised or a specific infection has been

documented (67). For cases in which an infectious organism has been identified,

appropriate antibiotics should be administered to both partners, followed by

posttreatment culture to verify eradication of the infectious agent before

attempting conception.

Chronic endometritis, defined as the asymptomatic presence of histologically

documented plasma cells in an endometrial biopsy, may be associated with

recurrent miscarriage. In suspected cases of chronic endometritis, it is difficult to

determine cause and effect but treatment with antibiotics shows a higher live birth

rate in subsequent pregnancies in some studies (353,354). Treatment for chronic

endometritis is not recommended by expert groups and larger studies are needed

before implementing evaluation and treatment in routine practice.

Immunologic Factors

[5] Immune-mediated recurrent pregnancy loss has received more attention than

any other single etiologic classification of recurrent pregnancy loss. Nevertheless,

the diagnosis and subsequent treatment of the majority of cases remain unclear

(56,118,254–256). Most therapies for proposed immune-related recurrent

pregnancy loss can have harmful side effects and must be considered

experimental.

Immunostimulating Therapies: Leukocyte Immunization

Stimulation of the maternal immune system using alloantigens on either paternal

or pooled donor leukocytes has been promoted for patients with immunologic

recurrent pregnancy loss, but data do not support its use and side effects can be

dangerous. One of the largest trials evaluating the efficacy of leukocyte

immunization in patients with unexplained recurrent pregnancy loss is a part of

the Recurrent Miscarriage (REMIS) study (258). This investigation was large

(over 90 patients per treatment arm), prospective, placebo controlled, randomized,

and double blinded. It demonstrated no efficacy for paternal leukocyte

immunization in couples with unexplained recurrent pregnancy loss. The best of

the meta-analyses definitively rejects use of this therapy in patients with recurrent

loss (259). Leukocyte immunization poses a significant risk to the mother and her

fetus (231,232,260). Several cases of graft-versus-host disease, severe intrauterine

growth retardation, and autoimmune and isoimmune complications have been

2000reported (257,260–264). Alloimmunization to platelets contained in the paternal

leukocyte preparation is associated with cases of potentially fatal fetal

thrombocytopenia. The routine use of this therapy for recurrent miscarriage is not

clinically justified.

Other immunostimulating therapies have been proposed and abandoned.

Intravenous preparations consisting of syncytiotrophoblast microvillus plasma

membrane vesicles have been used to mimic the fetal cell contact with maternal

blood that normally occurs in pregnancy (265). The efficacy of this therapy has

not been established (259,265,266). The use of third-party seminal plasma

suppositories has been attempted, based on the misconception that TLX was part

of an idiotype–anti-idiotype control system (267,268). Third-party seminal

plasma suppositories for recurrent miscarriage have no scientifically credible

rationale and should not be used.

Immunosuppressive Therapies

Immunosuppressive and other immunoregulating therapies have been advocated

for cases in which miscarriage was believed to result from antiphospholipid

antibodies or inappropriate cellular immunity toward the implanting fetus. Study

design problems, including small numbers of recruited patients, lack of

prestratification by maternal age and number of prior losses before randomization,

and other methodologic and statistical inaccuracies preclude definitive statements

regarding therapeutic efficacy for most of the proposed immunosuppressive

approaches.

Intravenous Immunoglobulin

Intravenous immunoglobulins (IVIgs) are composed of pooled samples of

immunoglobulins harvested from a large number of blood donors. Studies on the

use of IVIg therapy in the treatment of recurrent pregnancy loss are based on the

theory that some recurrent pregnancy loss patients have an overzealous immune

reactivity to their implanting fetus. IVIgs do have immunosuppressive effects, but

the mechanisms underlying this immune modulation are only partially

understood. Based on a large number of relatively small studies using a variety of

treatment protocols, there remains no conclusive evidence to suggest that use of

IVIg in the treatment of patients with unexplained (and presumed immunologic)

recurrent pregnancy loss has any benefit (261,269–273). This includes a recent

trial using IVIg in women with secondary recurrent pregnancy loss that was

ended early because interim analyses revealed no effect (269). The Cochrane

review of immune therapy for recurrent pregnancy loss addressed IVIg therapy

and reported that its use did not alter pregnancy outcomes in patients with

otherwise unexplained recurrent pregnancy loss (259,266). Improved

2001posttreatment pregnancy rates may be seen, however, when IVIg is used in those

specific patients with autoimmune-mediated pregnancy loss associated with APS

(274,275). Therapy with IVIgs for recurrent pregnancy loss is expensive,

invasive, and time-consuming, requiring multiple intravenous infusions over the

course of pregnancy (276). Use of IVIgs for recurrent pregnancy loss is not

supported by current evidence and may result in significant side effects such as

nausea, headache, myalgias, and hypotension (262). More serious adverse effects

include anaphylaxis (particularly in patients with IgA deficiency) (277).

Progesterone

As mentioned earlier, progesterone has known immunosuppressive effects and

empiric progesterone supplements for women with a history of recurrent

pregnancy loss is a relatively common practice (129–132,136,137). A number of

studies using in vitro cellular systems relevant to the maternal–fetal interface have

demonstrated that progesterone either inhibits TH1 immunity or causes a shift

from TH1- to TH2-type responses that are more favorable for successful

pregnancies (136,137,278). Clinical trials evaluating progesterone use in recurrent

miscarriage patients show varying results. In 2013, a Cochrane review of multiple

studies showed a benefit of progesterone supplementation in women with

recurrent pregnancy loss; however, the studies differed in sample size, timing,

dose, and type of progesterone supplementation (385). In 2015, Coomarasamy et

al. performed a randomized controlled trial of vaginal progesterone in women

with three or more first trimester miscarriages. Women were randomized to 400

mg vaginal micronized progesterone twice daily or matched placebo starting with

positive pregnancy test but no later than 6 weeks of gestation (386). Similar live

birth rates were achieved in the two groups and arguments against routine

progesterone use in recurrent pregnancy loss patients could be made despite the

late start of progesterone (up to 6 weeks gestation) in some patients. Some

research suggests an earlier start to progesterone supplementation is beneficial for

outcomes despite the side effects of luteal phase progesterone, including delayed

onset of menses, which can be distressing for women trying to conceive (387).

Progesterone has been administered intramuscularly and intravaginally for the

treatment of recurrent pregnancy loss. It is thought that vaginal administration

may increase local, intrauterine concentrations of progesterone better than

systemic administration. Vaginal formulations may provide a better method of

attaining local immunosuppressive levels of progesterone while decreasing

adverse systemic side effects.

Intralipid Infusion

The relative paucity of inflammatory diseases among the Greenland Inuit

2002population, who consume a diet high in fish oils, led investigators to study the

immune modulatory effects of lipid emulsions in total parenteral nutrition

preparations for preoperative patients and for burn and trauma victims (279–282).

The wide range of demonstrated effects, including lipid preparations that reduced

NK cell activity, reduced monocyte proinflammatory cytokine production and

increased susceptibility to infection, led investigators to hypothesize, as early as

1994, that lipid infusions might promote an immune environment that would

favor pregnancy maintenance (283). Since that time, a small number of

publications have addressed the effects of lipid infusions (intralipids) in women

with a history of pregnancy loss (284,285). These investigations have

demonstrated a decrease in peripheral NK cell activity in women treated with one

to three infusions of intralipids. This effect lasted from 4 to 9 weeks after the last

infusion (284). The authors did not address NK cell cytokine secretion patterns

nor did they assay decidual NK cell function. Despite this paucity of data,

intralipid infusions are being administered to recurrent pregnancy loss patients

with increasing frequency. The existing data do not support this practice and side

effects can include flushing, dizziness, myalgias, nausea, anaphylactic reactions,

kidney and liver dysfunction, infection, and thrombosis. Intralipid infusions in

recurrent pregnancy loss patients is not recommended clinically and should be

administered only under an institutional review board–approved protocol and in a

study setting.

TNF-α Inhibition

Interest in the potent proinflammatory cytokine, TNF-α, as a mediator of

pregnancy loss came out of the description of the TH1 and TH2 paradigm (125).

There have been several publications that link maternal serum TNF-α levels and

activating TNF-α gene promoter polymorphisms to recurrent pregnancy loss

(286–288). The development of antagonists of TNF-α in the form of blocking

antibodies (adalimumab, infliximab) and inhibitory recombinant proteins

(etanercept) has permitted successful treatment of several autoimmune disorders,

including rheumatoid arthritis, psoriasis, and Crohn disease. Their use has not

been associated with universally positive outcomes and may worsen some

disorders, including multiple sclerosis (289). These products are associated with

rare but worrisome side effects, including liver failure, aplastic anemia, interstitial

lung disease, and anaphylaxis (290). Although there exists only a single, small,

retrospective, observational, nonrandomly assigned case series that involved

treatment of recurrent pregnancy loss patients with inhibitors of TNF-α, these

positive preliminary results have led a growing number of clinics to offer this

therapy to patients, often at a significant cost (291). The safety of these

compounds in pregnancy has not been appropriately studied and preliminary

2003reports associating exposure to TNF-α inhibitors during early pregnancy to fetal

VACTERL syndrome is concerning (292). As with intralipid therapy, use of

TNF-α inhibition for the treatment of recurrent pregnancy loss should be

implemented only under an institutional review board–approved protocol in a

study setting and should not generate clinical income.

Other immunoregulating therapies theoretically useful in treating recurrent

pregnancy loss include the use of cyclosporine, pentoxifylline, and nifedipine,

although maternal and fetal risks with these agents preclude their clinical use.

Plasmaphoresis has been used to treat women with recurrent miscarriage and

antiphospholipid antibodies (293). Generalized immunosuppression with

corticosteroids, such as prednisone, has been advocated during pregnancy for

women with recurrent losses and chronic intervillositis and those with recurrent

pregnancy loss and APS (294). Although corticosteroids have shown some

treatment promise in these patients, maternal and fetal side effects, such as

gestational diabetes and intrauterine growth restriction, and the availability of

alternative therapies have limited their use (294,295). The efficacy and side

effects of prednisone plus low-dose aspirin was examined in a large, randomized,

placebo-controlled trial treating patients with autoantibodies and recurrent

pregnancy losses. Pregnancy outcomes for treated and control patients were

similar; however, the incidence of maternal diabetes and hypertension and the risk

of premature delivery were all increased among those treated with prednisone and

aspirin (296).

Antithrombotic Therapy

Therapy for patients with recurrent pregnancy losses associated with APS has

shifted toward the use of antithrombotic medications. Unlike immunosuppressive

treatments, this approach appears to address the effect (hypercoagulability), but

not the underlying cause (e.g., genetic, APS) of recurrent pregnancy loss. There

are reports that heparin, one typical anticoagulant, may exert direct

immunomodulatory effects by binding to antiphospholipid antibodies and may

decrease movement of inflammatory cells to sites of alloantigen exposure

(297,298). The combined use of low-dose aspirin (75 to 81 mg per day) and

subcutaneous unfractionated heparin (5,000 to 10,000 units twice daily) during

pregnancy has been best studied among women with APS and appears to be

efficacious (299–303). A typical regimen for women with antiphospholipid

syndrome would include use of aspirin (81 mg every day) beginning with any

attempts to conceive. After pregnancy has been confirmed, 5,000 IU

unfractionated sodium heparin is administered subcutaneously twice daily,

throughout gestation. Patients using this therapy should be treated in conjunction

with a perinatologist because of their increased risks for preterm labor, premature

2004rupture of the membranes, intrauterine growth restriction, intrauterine fetal

demise, and preeclampsia. Other potential risks include gastric bleeding,

osteopenia, and abruptio placenta.

Attempts have been made to extend the finding that antithrombotic therapy is

efficacious when used to treat patients with APS and recurrent pregnancy loss.

These attempts include the use of low–molecular-weight heparins (LMWHs), the

use of antithrombotic therapy in non-APS patients with thrombophilia and

recurrent pregnancy loss, and even its use among recurrent pregnancy loss

patients without thrombophilia (unexplained recurrent losses).

New formulations of heparin, termed LMWHs, have been demonstrated to

be superior to unfractionated heparin in the treatment of many clotting

disorders (304–306). LMWH has the advantage of an increased antithrombotic

ratio when compared with unfractionated heparin. This results in improved

treatment of inappropriate clotting with fewer bleeding side effects and LMWH

has been associated with a decreased incidence of thrombocytopenia and

osteoporosis when compared with its unfractionated counterpart. LMWH has a

long half-life and requires less frequent dosing and monitoring, thereby

improving patient compliance. LMWH appears to be safe for use in

pregnancy, and has shown promise when combined with low-dose aspirin in

the treatment of recurrent pregnancy loss associated with APS (300,305,307).

A few studies have compared the use of unfractionated heparin and aspirin to

LMWH and aspirin in the treatment of women with APS and adverse pregnancy

outcomes (303,308). The therapies had similar effects in one study (308). A metaanalysis suggested that unfractionated heparin was superior to LMWH, however,

there was significant heterogeneity between studies (303).

Anticoagulation for recurrent pregnancy loss patients with thrombophilia but

no personal or family history of thrombosis is controversial. Early studies showed

efficacy for LMWH treatment for patients with recurrent pregnancy loss

associated with other thrombophilias, including activated protein C resistance

associated with FVL, mutations in the promoter region of the prothrombin gene,

and decreases in protein C and protein S activities (309–311,317). The use of

LMWH for this indication appears to have a reassuring safety profile for mother

and fetus (307,312,313). Studies and expert groups argue that, despite the

biologic plausibility that patients with recurrent pregnancy loss would have a

higher chance of inherited thrombophilia and that anticoagulation of patients with

these diagnoses would decrease risk of subsequent miscarriage, the evidence does

not support the theories (338,341).

The prophylactic use of daily low-dose aspirin has become common practice

within the lay public based on its perceived cardiovascular effects combined with

its low incidence of side effects. Its sole use in the treatment of recurrent

2005pregnancy loss has likewise gained momentum, and many patients with histories

of recurrent loss will either be self-prescribing this therapy or will inquire about

its usefulness. There are no good data supporting its use either in patients with

heritable thrombophilias or in the general recurrent pregnancy loss population.

Although studies are small, the use of low-dose aspirin alone has not been shown

to be effective in the treatment of recurrent pregnancy loss associated with APS

(301,314,315). When used in these patients, it should be in combination with

unfractionated or LMWH. Large randomized prospective trials examining the

empiric use of aspirin alone or in combination with prophylactic doses of heparin

have shown no benefit of these therapies in unexplained recurrent pregnancy loss

(315). The use of aspirin in early pregnancy has been called into question with

reports of an increased incidence of isolated spontaneous pregnancy loss among

women who used this medication (213,214). These reports are poorly designed

and do not adequately address the level of aspirin exposure 81 mg versus 325 mg.

Although reviews have touted the overall safety of aspirin in pregnancy, outside

of use in combination with heparin for patients with recurrent pregnancy loss and

APS, this medication should be used only with justification in the well-informed

patient during early pregnancy (316).

As mentioned previously, vitamins B6, B12, and folate are important in

homocysteine metabolism, and hyperhomocysteinemia is linked to recurrent

pregnancy loss in some studies (15,17,27,388). Women with recurrent pregnancy

loss and isolated fasting hyperhomocysteinemia should be offered supplemental

folic acid (0.4 to 1.0 mg per day), vitamin B6 (6 mg per day), and possibly

vitamin B12 (0.025 mg per day) (318–321). Fasting homocysteine levels should

be retested after treatment. If levels are normalized or remain only marginally

elevated, no further therapy is necessary. Homocysteine levels will predictably

decrease during pregnancy.

Treatment of women with recurrent pregnancy loss and an identified

inherited or acquired thrombophilia should be based on accompanying

history. There are no published large prospective controlled trials examining the

benefit of anticoagulation for the prevention of miscarriage in the absence of

APS, and, therefore, anticoagulation recommendations for patients with inherited

thrombophilias are based on individualized risk of venous thromboembolic events

in pregnancy (16).

If a venous thromboembolic event occurs during the index pregnancy,

posthospitalization management requires therapeutic anticoagulation.

UFH: 10,000 to 15,000 U subcutaneous every 8 to 12 hours (monitor to

keep aPTT 1.5 to 2.5 times normal) OR

2006LMWH: enoxaparin 40 to 80 mg subcutaneous twice a day or

dalteparin 5,000 to 10,000 U subcutaneous twice a day). Consider

monitoring trough factor Xa levels in the third trimester.

If there is a personal history of venous thromboembolic events

(particularly in a previous pregnancy or with hormonal contraceptive

use) or a strong thrombophilic family history, treat with therapeutic

anticoagulation. Thrombotic risk is greatest during the postpartum

period.

Anticoagulation should be reinitiated after delivery in doses reflecting

predelivery treatment regimens. Postpartum anticoagulation should be

continued for 6 to 12 weeks postpartum (305). Women may continue

injectable therapy or transition to oral anticoagulants (e.g., coumarin). Use

of heparin or of coumarin derivatives does not prohibit breastfeeding.

Psychological Support

There is no doubt that experiencing isolated and recurrent losses can be

emotionally devastating. The risk of major depression is increased greater than

twofold among women with spontaneous pregnancy loss; in most women it arises

in the first weeks following delivery (322). A caring and empathetic attitude is

prerequisite to all healing. Acknowledgment of the pain and suffering couples

have experienced because of recurrent miscarriage can be a cathartic catalyst

enabling them to incorporate their experience of loss into their lives rather than

focusing their lives on their experience of loss (66). Referrals to support groups

and counselors should be offered. Self-help measures, such as meditation, yoga,

exercise, and biofeedback may also be useful.

PROGNOSIS

[7] The prognosis for successful pregnancy depends on the potential underlying

cause of pregnancy loss and (epidemiologically) on the number of prior losses.

Epidemiologic surveys indicate that the chance of a viable birth even after four

prior losses may be as high as 60%. Depending on the study, the prognosis for

successful pregnancy in couples with a cytogenetic etiology for reproductive

loss varies from 20% to 80% (323–325). Women with corrected anatomical

anomalies may expect a successful pregnancy in 60% to 90% of cases

(28,323,326–329). A success rate higher than 90% has been reported for

women with corrected endocrine abnormalities (324). Between 70% and

90% of pregnancies reported among women receiving therapy for

antiphospholipid antibodies have been viable (330,331).

Many forms of pre- or postconceptional tests have been proposed to help

2007predict pregnancy outcome (221,332,333,389); none have been fully

substantiated in large, prospective trials. The documentation of fetal cardiac

activity on ultrasound may offer prognostic value; however, it appears that

its predictions may be greatly affected by any underlying diagnosis. In one

study, the live birth rate following documentation of fetal cardiac activity between

5 and 6 weeks from the last menstrual period was approximately 77% in women

with two or more unexplained spontaneous miscarriages (334). It may be

important to note that most of the patients in this study had evidence of

inappropriate antitrophoblast cellular immunity. Others have shown that 86% of

patients with antiphospholipid antibodies and recurrent pregnancy loss had fetal

cardiac activity detected prior to subsequent demise (335). A prospective,

longitudinal, observational study of 325 patients with unexplained recurrent

pregnancy losses demonstrated that only 3% of 55 miscarriages occurred

following the detection of fetal cardiac activity using transvaginal

ultrasonography (336). The majority of couples with recurrent pregnancy loss

will go on to have live birth and reminding patients of this throughout evaluation

and treatment can be an invaluable tool for psychological and emotional support

through the process


Nhận xét