These disorders can be separated into those associated with and those without autoantibody formation. So-called rheumatoid factor (RF) is an autoantibody found in many autoimmune inflammatory conditions such as SLE, rheumatoid arthritis, systemic sclerosis (scleroderma), mixed connective-tissue disease, dermatomyositis, polymyositis, and various vasculitis syndromes. The RF-seronegative spondyloarthropathies are strongly associated with expression of the HLA-B27 antigen and include ankylosing spondylitis, psoriatic arthritis, Reiter disease, and other arthritis syndromes.
Pregnancy may mitigate activity in some of these syndromes as a result of the immunosuppression that also allows successful engraftment of fetal and placental tissues. These changes are discussed in detail in Chapters 4 and 5 (Immunological Functions). One example is pregnancy-induced predominance of T2 helper cells compared with cytokine-producing T1 helper cells (Keeling, 2009). Pregnancy hormones alter immune cells, for example, estrogens upregulate and androgens downregulate T-cell response, and progesterone is immunosuppressive (Cutolo, 2006; Häupl, 2008a; Robinson, 2012).
Some immune-mediated diseases may either be caused or activated as a result of previous pregnancies. To explain, fetal cells and free fetal DNA are detectable in maternal blood beginning early in pregnancy (Simpson, 2013; Sitar, 2005; Waldorf, 2008). Fetal cell microchimerism is the persistence of fetal cells in the maternal circulation and organs following pregnancy. These persistent fetal cells may stimulate autoantibodies, or they may become engrafted in maternal tissues. This raises the possibility that fetal cell microchimerism is related to the predilection of autoimmune disorders for women (Adams, 2004; Lissauer, 2009). Evidence for this includes findings of fetal stem cells engrafted in maternal tissues in women with autoimmune thyroiditis and systemic sclerosis (Jimenez, 2005; Srivatsa, 2001). Such microchimerism has also been described in women with SLE and those with rheumatoid arthritis-associated HLA alleles (Johnson, 2001; Lee, 2010; Rak, 2009a). Similarly, engrafted maternal cells may provoke autoimmune conditions in a woman’s offspring (Ye, 2012). Perhaps related, women with SLE have a 0.6 male:female offspring ratio suggesting excessive male fetal loss in these women (Aggarwal, 2013).
Systemic Lupus Erythematosus
Lupus is a heterogeneous autoimmune disease with a complex pathogenesis that results in interactions between susceptibility genes and environmental factors (Hahn, 2012; Tsokos, 2011). Immune system abnormalities include overactive B lymphocytes that are responsible for autoantibody production. These result in tissue and cellular damage when autoantibodies or immune complexes are directed at one or more cellular nuclear components (Tsokos, 2011). In addition, immunosuppression is impaired, including regulatory T-cell function (Tower, 2013). Some autoantibodies produced in patients with lupus are shown in Table 59-1.
TABLE 59-1Some Autoantibodies Produced in Patients with Systemic Lupus Erythematosus (SLE) ||Download (.pdf) TABLE 59-1 Some Autoantibodies Produced in Patients with Systemic Lupus Erythematosus (SLE)
|Antibody ||Prevalence (%) ||Clinical Associations |
|Antinuclear (ANA) ||84–98 ||Best screening test, multiple antibodies; a second negative test makes SLE unlikely |
|Anti-double-stranded (ds)-DNA ||62–70 ||High titers SLE-specific; may correlate with disease activity, nephritis, and vasculitis |
|Anti-Sm (Smith) ||25–38 ||Specific for SLE |
|Anti-RNP ||33–40 ||Not SLE-specific, high titers associated with rheumatic syndromes |
|Anti-Ro (SS-A) ||30–49 ||Not SLE-specific; associated with Sjögren syndrome, predisposes to cutaneous lupus, neonatal lupus with heart block, reduced risk of nephritis |
|Anti-La (SS-B) ||10–35 ||Associated with anti-Ro; possible decreased nephritis risk |
|Antihistone ||70 ||Common in drug-induced lupus (95%) |
|Antiphospholipid ||21–50 ||Lupus anticoagulant and anticardiolipin antibodies associated with thrombosis, fetal loss, thrombocytopenia, valvular heart disease; false-positive test for syphilis |
|Anti-erythrocyte ||60 ||Small number develop hemolysis |
|Antiplatelet ||30 ||Thrombocytopenia in 15%; poor clinical test |
Almost 90 percent of lupus cases are in women, and its prevalence in those of childbearing age is approximately 1 in 500 (Lockshin, 2000). Accordingly, the disease is encountered relatively frequently during pregnancy. The 10-year survival rate is 70 to 90 percent (Hahn, 2012; Tsokos, 2011). Infection, lupus flares, end-organ failure, hypertension, stroke, and cardiovascular disease account for most deaths.
Genetic influences are implicated by a higher concordance with monozygotic compared with dizygotic twins—25 versus 2 percent, respectively. Moreover, there is a 10-percent frequency in patients with one affected family member. The relative risk of disease is increased if there is inheritance of the “autoimmunity gene” on chromosome 16 that predisposes to SLE, rheumatoid arthritis, Crohn disease, and psoriasis (Hahn, 2012). Susceptibility genes such as HLA-A1, B8, DR3, DRB1, and TET3 explain only a portion of the genetic heritability (Hom, 2008; Tsokos, 2011; Yang, 2013).
Clinical Manifestations and Diagnosis
Lupus is notoriously variable in its presentation, course, and outcome (Table 59-2). Findings may be confined initially to one organ system, and others become involved later. Or, the disease may first manifest with multisystem involvement. Common findings are malaise, fever, arthritis, rash, pleuropericarditis, photosensitivity, anemia, and cognitive dysfunction. At least half of patients have renal involvement. There is evidence that lupus is associated with decline in attention, memory, and reasoning (Kozora, 2008). Although Libman-Sacks endocarditis was described with lupus, it is likely due to presence of subsequently discussed anticardiolipin antibodies (Hojnik, 1996).
TABLE 59-2Clinical Manifestations of Systemic Lupus Erythematosus ||Download (.pdf) TABLE 59-2 Clinical Manifestations of Systemic Lupus Erythematosus
|Organ System ||Clinical Manifestations ||Percent |
|Systemic ||Fatigue, malaise, fever, weight loss ||95 |
|Musculoskeletal ||Arthralgias, myalgias, polyarthritis, myopathy ||95 |
|Hematological ||Anemia, hemolysis, leukopenia, thrombocytopenia, lupus anticoagulant, splenomegaly ||85 |
|Cutaneous ||Malar (butterfly) rash, discoid rash, photosensitivity, oral ulcers, alopecia, skin rashes ||80 |
|Neurological ||Cognitive dysfunction, mood disorder, headache, seizures ||60 |
|Cardiopulmonary ||Pleuritis, pericarditis, myocarditis, endocarditis, pneumonitis, pulmonary hypertension ||60 |
|Renal ||Proteinuria, casts, nephrotic syndrome, renal failure ||30–50 |
|Gastrointestinal ||Nausea, pain, diarrhea, abnormal liver enzyme levels ||40 |
|Vascular ||Thrombosis: venous (10%), arterial (5%) ||15 |
|Ocular ||Conjunctivitis ||15 |
Identification of antinuclear antibodies (ANA) is the best screening test. However, a positive result is not specific for lupus. For example, low titers are found in normal individuals, other autoimmune diseases, acute viral infections, and chronic inflammatory processes. Several drugs can also cause a positive reaction. Antibodies to double-stranded DNA (dsDNA) and to Smith (Sm) antigens are relatively specific for lupus, whereas other antibodies are not (see Table 59-1). Although hundreds of autoantibodies have been described in SLE, only a few have been shown to participate in tissue injury (Sherer, 2004; Tsokos, 2011). Microarray profiles are being developed for customized and more accurate SLE diagnoses (Lin, 2013; Yeste, 2013).
Anemia develops frequently, and there may be leukopenia and thrombocytopenia. Proteinuria and casts are found in the half of patients with glomerular lesions. Lupus nephritis can also cause renal insufficiency, which is more common if there are antiphospholipid antibodies (Moroni, 2004). Other laboratory findings include false-positive syphilis serology, prolonged partial thromboplastin time, and higher rheumatoid factor levels. Elevated serum d-dimer levels often follow a flare or infection, but unexplained persistent elevations are associated with a high risk for thrombosis (Wu, 2008).
The diagnostic criteria for SLE are listed in Table 59-3. If any four or more of these 11 criteria are present, serially or simultaneously, the diagnosis of lupus is made.
TABLE 59-3Criteria of the American Rheumatism Association for Systemic Lupus Erythematosus (SLE) ||Download (.pdf) TABLE 59-3 Criteria of the American Rheumatism Association for Systemic Lupus Erythematosus (SLE)
|Criteriaa ||Comments |
|Malar rash ||Malar erythema |
|Discoid rash ||Erythematous patches, scaling, follicular plugging |
|Photosensitivity ||Exposure to UV light causes rash |
|Oral ulcers ||Usually painless oral and nasopharyngeal ulcers |
|Arthritis ||Nonerosive involving two or more peripheral joints with tenderness, swelling, or effusion |
|Serositis ||Pleuritis or pericarditis |
|Renal ||Proteinuria greater than 0.5 g/day or > 3+ dipstick, or cellular casts |
|Neurological ||Seizures or psychosis without other cause |
|Hematological ||Hemolytic anemia, leukopenia, lymphopenia, or thrombocytopenia |
|Autoantibodies ||Anti-dsDNA or anti-Sm antibodies, or false-positive VDRL, abnormal level of IgM or IgG anticardiolipin antibodies, or lupus anticoagulant |
|ANA ||Abnormally elevated ANA titers |
Importantly, numerous drugs can induce a lupus-like syndrome. These include procainamide, quinidine, hydralazine, α-methyldopa, phenytoin, and phenobarbital. Drug-induced lupus is rarely associated with glomerulonephritis and usually regresses when the medication is discontinued (Rubin, 1997).
Of nearly 16.7 million pregnancies in the United States from 2000 to 2003, 13,555 were complicated by lupus—an incidence of approximately 1 in 1250 pregnancies (Clowse, 2008). During the past several decades, pregnancy outcomes in women with SLE have improved remarkably. Important factors for pregnancy outcome include whether disease is active at the beginning of pregnancy, age and parity, coexistence of other medical or obstetrical disorders, and whether antiphospholipid antibodies are detected (Antiphospholipid Antibody Syndrome). There is evidence that newly diagnosed lupus during pregnancy tends to be severe (Zhao, 2013).
During pregnancy, lupus improves in a third of women, remains unchanged in a third, and worsens in the remaining third. Thus, in any given pregnancy, the clinical condition can worsen or flare without warning (Khamashta, 1997). Petri (1998) reported a 7-percent risk of major morbidity during pregnancy. Women who have confined cutaneous lupus do not usually have adverse outcomes (Hamed, 2013). Frequent complications in a cohort of 13,555 women with SLE during pregnancy are shown in Table 59-4. The maternal mortality and severe morbidity rate was 325 per 100,000. In a review of 13 studies with 17 maternal deaths attributable to SLE and lupus nephritis, all occurred in those with active disease (Ritchie, 2012).
TABLE 59-4Complications in 13,555 Pregnancies in Women with Systemic Lupus Erythematosus ||Download (.pdf) TABLE 59-4 Complications in 13,555 Pregnancies in Women with Systemic Lupus Erythematosus
|Complications ||Percent |
|Comorbid illness || |
| Pregestational diabetes ||5.6 |
| Thrombophilia ||4.0 |
| Hypertension ||3.9 |
| Renal failure ||0.2 |
| Pulmonary hypertension ||0.2 |
|Pregnancy complications || |
| Preeclampsia ||22.5 |
| Preterm labor ||20.8 |
| Fetal-growth restriction ||5.6 |
| Eclampsia ||0.5 |
|Medical complications || |
| Anemia ||12.6 |
| Thrombocytopenia ||4.3 |
| Thrombotic—stroke, pulmonary embolism, deep-vein thrombosis ||1.7 |
|Infections—pneumonia, sepsis syndrome ||2.2 |
|Maternal morbidity-mortality rate ||325/100,000 |
From the foregoing, it is certain that lupus can be life threatening for both mother and fetus. Generally speaking, pregnancy outcome is best in those women in whom: (1) lupus activity has been quiescent for at least 6 months before conception; (2) there is no lupus nephritis manifest by proteinuria or renal dysfunction; (3) there is no evidence of the antiphospholipid antibody syndrome or lupus anticoagulant; and (4) superimposed preeclampsia does not develop (Peart, 2014; Stojan, 2012).
Active nephritis has been associated with particularly bad pregnancy outcomes, although these have improved remarkably during the past 30 years (Moroni, 2005; Stojan, 2012). Women with renal disease have a high incidence of gestational hypertension and preeclampsia. However, if their disease remains in remission, they usually have good pregnancy outcomes (Huong, 2001; Moroni, 2002). Of the 125 pregnancies reported by Lockshin (1989), 63 percent of women with preexisting renal disease developed preeclampsia compared with only 14 percent of those without underlying renal disease. Moroni and Ponticelli (2005) reviewed results from a total of 309 pregnancies complicated by established lupus nephritis. Of these, 30 percent suffered a flare, and 40 percent of these had associated renal insufficiency. The maternal mortality rate was 1.3 percent.
In two studies describing pregnancy outcomes in women with lupus nephritis, Wagner and coworkers (2009) compared outcomes of 58 women cared for during 90 pregnancies at the Mayo Clinic. Active nephritis was associated with a significantly higher incidence of maternal complications compared with women without nephritis—57 versus 11 percent. Quiescent nephritis had a nonsignificant effect on preeclampsia rates compared with lupus patients without renal impairment. The fetal death rate with active maternal nephritis was 35 percent compared with 9 percent in those with quiescent nephritis. In another study, Imbasciati and associates (2009) described outcomes in 113 pregnancies in 81 women with known lupus nephritis. During a third of pregnancies, there was a renal flare. After excluding nine miscarriages, of the 104 remaining pregnancies, a third were delivered preterm, a third of infants weighed < 2500 g, and the perinatal mortality rate was 6 percent.
Most recommend continuation of immunosuppressive therapy for nephritis during pregnancy. It is not clear whether the dose should be increased peripartum. Although it is often stated that this is the time that activation or exacerbations are most likely to develop, the evidence is not conclusive.
Lupus versus Preeclampsia-Eclampsia
Chronic hypertension complicates up to 30 percent of pregnancies in women with SLE (Egerman, 2005). And as discussed, preeclampsia is common, and superimposed preeclampsia is encountered even more often in those with nephritis or antiphospholipid antibodies (Bertsias, 2008). It may be difficult, if not impossible to differentiate lupus nephropathy from severe preeclampsia if the kidney is the only involved organ (Petri, 2007). Central nervous system involvement with lupus may culminate in convulsions similar to those of eclampsia. Thrombocytopenia, with or without hemolysis, may further confuse the diagnosis because of its similarity to the hemolysis, elevated liver enzymes, low platelet count (HELLP) syndrome. Management is identical to that for preeclampsia-eclampsia, described in Chapter 40 (Management).
Management During Pregnancy
Lupus management consists primarily of monitoring maternal clinical and laboratory conditions as well as fetal well-being (Lateef, 2012). Pregnancy-induced thrombocytopenia and proteinuria resemble lupus disease activity, and the identification of a lupus flare is confounded by the increase in facial and palmar erythema of normal pregnancy (Lockshin, 2003). Some authorities have advocated a number of numerical scales to emphasize ongoing disease activity. Components are weighted for severity, both with the SLE-Pregnancy Disease Activity Index (SLEPDAI) and with the Lupus Activity Index (Buyon, 1999; Ruiz-Irastorza, 2004). We have not found these to be useful.
Lupus activity monitoring and identification of lupus flares by various laboratory techniques has been recommended. The sedimentation rate may be misleading because of pregnancy-induced hyperfibrinogenemia. Serum complement levels are also normally increased in pregnancy (Chap. 4, Immunological Functions and Appendix, Serum and Blood Constituents). And, although falling or low levels of complement components C3, C4, and CH50 are more likely to be associated with active disease, higher levels provide no assurance against disease activation. Our experiences, as well as those of Varner and colleagues (1983) and Lockshin and Druzin (1995), are that there is no correlation between clinical manifestations of disease and complement levels.
Serial hematological studies may detect changes in disease activity. Hemolysis is characterized by a positive Coombs test, anemia, reticulocytosis, and unconjugated hyperbilirubinemia. Thrombocytopenia, leukopenia, or both may develop. According to Lockshin and Druzin (1995), chronic thrombocytopenia in early pregnancy may be due to antiphospholipid antibodies. Later, thrombocytopenia may indicate preeclampsia.
Serum aminotransferase activity reflects hepatic involvement, as does a rise in serum bilirubin levels. Azathioprine therapy also may induce enzyme elevations. Urine is tested frequently to detect new-onset or worsening proteinuria. Overt proteinuria that persists is an ominous sign, even more so if accompanied by other evidence of the nephrotic syndrome or abnormal serum creatinine levels.
The fetus should be closely observed for adverse effects such as growth restriction and oligohydramnios. Many recommend screening for anti-SS-A (anti-Ro) and anti-SS-B (anti-La) antibodies, because of associated fetal complications described subsequently. As discussed in Chapter 17 (Fetal Movements), antepartum fetal surveillance is done as outlined by the American College of Obstetricians and Gynecologists (2012a). Unless hypertension develops or there is evidence of fetal compromise or growth restriction, pregnancy is allowed to progress to term. Peripartum corticosteroids in “stress doses” are given to women who are taking these drugs or who recently have done so.
There is no cure, and complete remissions are rare. Approximately a fourth of women have mild disease, which is not life threatening, but may be disabling because of pain and fatigue. Arthralgia and serositis can be managed by occasional doses of nonsteroidal antiinflammatory drugs (NSAIDs). However, chronic or large intermittent dosing is avoided due to pregnancy side effects with these drugs described in Chapter 12 (Antifungal Medications) (Briggs, 2011). Low-dose aspirin can be used throughout gestation. Severe disease is managed with corticosteroids such as prednisone, 1 to 2 mg/kg orally per day. After the disease is controlled, this dose is tapered to a daily morning dose of 10 to 15 mg. Corticosteroid therapy can result in the development of gestational diabetes.
Immunosuppressive agents such as azathioprine are beneficial in controlling active disease (Contreras, 2004; Hahn, 2012). In nonpregnant patients, these are usually reserved for lupus nephritis or disease that is corticosteroid resistant. Azathioprine has a good safety record during pregnancy (Fischer-Betz, 2013; Petri, 2007). Its recommended daily oral dose is 2 to 3 mg/kg. According to Buhimschi and Weiner (2009), cyclophosphamide is teratogenic, and although not usually recommended during pregnancy, severe disease may be treated after 12 weeks’ gestation. As discussed in Chapter 12 (Lead), medications to be avoided include mycophenolate mofetil and methotrexate (Anderka, 2009; Briggs, 2011; Food and Drug Administration, 2008). In some situations, mycophenolate is the only treatment that achieves disease stability. In these cases, counseling regarding fetal risks is essential (Bramham, 2012).
Antimalarials help control skin disease. Although these agents cross the placenta, hydroxychloroquine has not been associated with congenital malformations. Because of the long half-life of antimalarials and because discontinuing therapy can precipitate a lupus flare, most recommend their continuation during pregnancy (Borden, 2001; Harris, 2002). Levy and associates (2001) randomly assigned 20 pregnant women to receive hydroxychloroquine or placebo and reported improved SLEPDAI scores with hydroxychloroquine.
When severe disease supervenes—usually a lupus flare—high-dose glucocorticoid therapy is given. Petri (2007) recommends pulse therapy consisting of methylprednisolone, 1000 mg given intravenously over 90 minutes daily for 3 days, then a return to maintenance doses if possible.
Perinatal Mortality and Morbidity
Adverse perinatal outcomes are increased significantly in pregnancies complicated by lupus. These include preterm delivery, fetal-growth restriction, stillbirths, and neonatal lupus syndrome (Madazli, 2014). Outcomes are worse with a lupus flare, significant proteinuria, or renal impairment, and with chronic hypertension, preeclampsia, or both (Aggarwal, 1999; Bramham, 2012; Scott, 2002; Wagner, 2009). The observations of Lee and coworkers (2009) are worrisome. In a mouse SLE model, they showed that autoantibodies directed against the N-methyl-d-aspartate neuroreceptor caused fetal neurotoxicity. This suggests an underlying etiology for the CNS manifestations and learning disorders in children of affected mothers. Ongoing work on anti-dsDNA antibodies has identified peptides that can protect target organs from antibody-mediated damage (Diamond, 2011).
The reasons at least partially responsible for adverse fetal consequences include decidual vasculopathy with placental infarction and decreased perfusion (Hanly, 1988; Lubbe, 1984). Placental pathology is discussed in more detail in Chapter 6 (Maternal Blood Flow Disruption).
This unusual constellation is characterized by newborn skin lesions—lupus dermatitis; a variable number of hematological and systemic derangements; and occasionally congenital heart block (Boh, 2004; Lee, 2009). Although usually associated with anti-SS-A and -SS-B antibodies, McGeachy and Lam (2009) described an affected infant in whom only anti-ribonucleoprotein (RNP) antibodies were found. Thrombocytopenia and hepatic involvement are seen in 5 to 10 percent of affected infants. One report suggests that neonatal lupus may appear up to 4 weeks after birth (Stirnemann, 2002). Lockshin and colleagues (1988) prospectively followed 91 infants born to women with lupus. Eight were possibly affected—four had definite neonatal lupus and four had possible disease. Clinical manifestations, which include cutaneous lupus, thrombocytopenia, and autoimmune hemolysis, are transient and clear within a few months (Lee, 1984). This may not be so for congenital heart block. In subsequent offspring, the recurrence risk for neonatal lupus is up to 25 percent (Julkunen, 1993).
Fetal and neonatal heart block results from diffuse myocarditis and fibrosis in the region between the atrioventricular (AV) node and bundle of His. Buyon and associates (1993) reported that congenital heart block developed almost exclusively in fetuses of women with antibodies to the SS-A or SS-B antigens. These antibodies may also cause otherwise unexplained stillbirths (Ottaviani, 2004). Even in the presence of such antibodies, however, the incidence of myocarditis is only 2 to 3 percent but increases to 20 percent with a prior affected child (Bramham, 2012; Lockshin, 1988). Fetal cardiac monitoring should be performed between 18 and 26 weeks’ gestation in pregnancies with either of these antibodies. The cardiac lesion is permanent, and a pacemaker is generally necessary. Long-term prognosis is poor. Of 325 infants with cardiac neonatal lupus, nearly 20 percent died, and of these, a third were stillborn (Izmirly, 2011).
Maternal administration of corticosteroids, plasma exchange, or intravenous immunoglobulin have not been found to reduce the risk of congenital heart block. Corticosteroid therapy to treat fetal heart block has not been subjected to randomized trials. There is some evidence that early treatment may mitigate fetal myocarditis. Shinohara and coworkers (1999) reported no heart block in 26 neonates whose mothers received corticosteroids before 16 weeks’ gestation as part of SLE maintenance therapy. By contrast, 15 of 61 neonates with heart block were born to women in whom corticosteroid therapy was begun after 16 weeks for SLE exacerbation. Rein and colleagues (2009) followed 70 fetuses prospectively of mothers positive for anti-SS-A or -SS-B antibodies with serial fetal kinetocardiography (FKCG) to measure AV conduction time. In six fetuses, first-degree heart block developed at 21 to 34 weeks, and maternal dexamethasone treatment was associated with normalization of AV conduction in all within 3 to 14 days. There were no recurrences, and the infants all were well at a median follow-up of 4 years. However, as noted in Chapter 16 (Bradyarrhythmia), corticosteroids may be less effective for higher degrees of heart block. Moreover, potential benefit should be weighed against the risks of chronic corticosteroid treatment, including impaired fetal growth (Friedman, 2009).
Long-Term Prognosis and Contraception
In general, women with lupus and chronic vascular or renal disease may limit family size because of morbidity associated with the disease as well as increased adverse perinatal outcomes. Two large multicenter clinical trials have shown that combination oral contraceptives (COCs) did not increase the incidence of lupus flares (Petri, 2005; Sánchez-Guerrero, 2005). Still, the American College of Obstetricians and Gynecologists (2013) recommends that COC use be avoided in women who have nephritis, antiphospholipid antibodies, or vascular disease. Progestin-only implants and injections provide effective contraception with no known effects on lupus flares (Chabbert-Buffet, 2010). Concerns that intrauterine device (IUD) use and immunosuppressive therapy lead to increased infection rates in these patients are not evidenced-based. Tubal sterilization may be advantageous and is performed with greatest safety postpartum or whenever the disease is quiescent. These options are discussed in further detail in Chapters 38 and 39.
Antiphospholipid Antibody Syndrome
Phospholipids are the main lipid constituents of cell and organelle membranes. There are proteins in plasma that associate noncovalently with these phospholipids. Antiphospholipid antibodies are directed against these phospholipids or phospholipid-binding proteins (Erkan, 2011; Giannakopoulos, 2013; Tsokos, 2011). This antibody group may be of IgG, IgM, and IgA classes, alone or in combination. Antiphospholipid antibodies are most common with lupus and other connective-tissue disorders, although a small proportion of otherwise normal women and men are found to have these antibodies in various forms.
The stimulus for autoantibody production is unclear, but it possibly is due to a preceding infection. The pathophysiology encountered in the antiphospholipid antibody syndrome—variably referred to as APAS or APS—is mediated by one or more of the following: (1) activation of various procoagulants, (2) inactivation of natural anticoagulants, (3) complement activation, and (4) inhibition of syncytiotrophoblast differentiation (Moutsopoulos, 2012; Tsokos, 2011). Clinically, these result in arterial or venous thromboses or pregnancy morbidity, and virtually every organ system may be involved as shown in Table 59-5.
TABLE 59-5Some Clinical Features of Antiphospholipid Antibody Syndrome ||Download (.pdf) TABLE 59-5 Some Clinical Features of Antiphospholipid Antibody Syndrome
|Venous thrombosis—thromboembolism, thrombophlebitis, livedo reticularis |
|Arterial thrombosis—stroke, transient ischemic attack, Libman-Sacks cardiac vegetations, myocardial ischemia, distal extremity and visceral thrombosis and gangrene |
|Hematological—thrombocytopenia, autoimmune hemolytic anemia |
|Other—neurological manifestations, migraine headaches, epilepsy; renal artery, vein, or glomerular thrombosis; arthritis and arthralgia |
|Pregnancy—preeclampsia syndrome, recurrent miscarriage, fetal death |
Central nervous system involvement is one of the most prominent clinical manifestations. In addition to cerebrovascular arterial and venous thrombotic events, there may be psychiatric features and even multiple sclerosis (Binder, 2010; Sanna, 2003). Renovascular involvement may lead to renal failure that may be difficult to differentiate from lupus nephritis (D’Cruz, 2009). Peripheral and visceral thromboses are also a feature. For example, Ahmed and associates (2009) reported a postpartum woman who developed spontaneous cecal perforation associated with a mesenteric vessel infarction. As discussed in Chapter 18 (Immunological Factors), antiphospholipid antibodies have been associated with excessive pregnancy loss (Branch, 2010). These and other detrimental effects on pregnancy outcomes are discussed subsequently.
A small proportion of these patients develop the catastrophic antiphospholipid antibody syndrome—CAPS. This is defined as a rapidly progressive thromboembolic disorder simultaneously involving three or more organ systems or tissues (Moutsopoulos, 2012).
Specific Antiphospholipid Antibodies
Several autoantibodies have been described that are directed against a specific phospholipid or phospholipid-binding protein. First, β2-glycoprotein I—also known as apolipoprotein H—is a phospholipid-binding plasma protein that inhibits prothrombinase activity within platelets and inhibits ADP-induced platelet aggregation (Giannakopoulos, 2013; Shi, 1993). Thus, its normal action is to inhibit procoagulant binding and thereby prevent coagulation cascade activation. Logically, antibodies directed against β2-glycoprotein I would inhibit its anticoagulant activity and promote thrombosis. This is important from an obstetrical viewpoint because β2-glycoprotein I is expressed in high concentrations on the syncytiotrophoblastic surface. Complement activation may contribute to it pathogenesis (Avalos, 2009; Tsokos, 2011). Teleologically, this seems appropriate because the decidua intuitively should be a critical area to prevent coagulation that might result in intervillous space thrombosis. Another possibility is that β2-glycoprotein I may be involved in implantation, and it may result in pregnancy loss via an inflammatory mechanism (Iwasawa, 2012; Meroni, 2011).
Second, lupus anticoagulant (LAC) is a heterogeneous group of antibodies directed also against phospholipid-binding proteins. LAC induces prolongation in vitro of the prothrombin, partial thromboplastin, and Russell viper venom times (Feinstein, 1972). Thus, paradoxically, this so-called anticoagulant is actually powerfully thrombotic in vivo.
Last, there are anticardiolipin antibodies (ACAs). These are directed against one of the many phospholipid cardiolipins found in mitochondrial membranes and platelets.
Antibodies against Natural Anticoagulants
Antiphospholipid antibodies have also been described that are directed against the natural anticoagulant proteins C and S (Robertson, 2006). Another is directed against the anticoagulant protein annexin V, which is expressed in high concentrations by the syncytiotrophoblast (Chamley, 1999; Giannakopoulos, 2013). That said, criteria to diagnosis antiphospholipid-antibody syndrome require testing for elevated levels of the following antiphospholipid antibodies: LAC, ACA, and anti-β2 glycoprotein I. Testing for other antibodies is not recommended (American College of Obstetricians and Gynecologists, 2012b).
Antiphospholipid Antibody Syndrome Diagnosis
Clinical classification criteria shown in Table 59-5 provide indications for testing. By international consensus, the syndrome is diagnosed based on laboratory and clinical criteria (Miyakis, 2006). First, one of two clinical criteria—which are vascular thrombosis or certain pregnancy morbidity—must be present. In addition, at least one laboratory criterion that includes LAC activity or medium- to high-positive levels of specific IgG- or IgM-ACAs must be confirmed on two occasions 12 weeks apart.
Tests for LAC are nonspecific coagulation tests. The partial thromboplastin time is generally prolonged because the anticoagulant interferes with conversion of prothrombin to thrombin in vitro. Tests considered more specific are the dilute Russell viper venom test and the platelet neutralization procedure. There is currently disagreement as to which of these two is best for screening. If either is positive, then identification of LAC is confirmed.
Branch and Khamashta (2003) recommend conservative interpretation of results based on repeated tests from a reliable laboratory that are consistent with each clinical case. Only approximately 20 percent of patients with APS have a positive LAC reaction alone. Thus, anticardiolipin enzyme-linked immunosorbent assay (ELISA) testing should also be performed. Many laboratories have a premade panel to use for APS testing.
Efforts have been made to standardize ACA assays by using ELISA. Values are reported in units and expressed as negative, low-positive, medium-positive, or high-positive (Harris, 1987a,b). Despite this, these assays remain without international standardization (Adams, 2013; Branch, 2003; Capuano, 2007). Interlaboratory variation can be large, and agreement between commercial kits is poor.
Pregnancy and Antiphospholipid Antibodies
Nonspecific low levels of antiphospholipid antibodies are identified in approximately 5 percent of normal adults (Branch, 2010). When Lockwood and coworkers (1989) first studied 737 normal pregnant women, they reported that 0.3 percent had lupus anticoagulant and 2.2 percent had elevated concentrations of either IgM or IgG anticardiolipin antibodies. Later studies confirmed this, and taken together, they totaled almost 4000 normal pregnancies with an average prevalence for APAs of 4.7 percent. This is the same as for normal nonpregnant individuals (Harris, 1991; Pattison, 1993; Yasuda, 1995).
In women with high levels of ACAs, and especially when lupus anticoagulant is identified, there are increased risks for decidual vasculopathy, placental infarction, fetal-growth restriction, early-onset preeclampsia, and recurrent fetal death. Some of these women, like those with lupus, also have a high incidence of venous and arterial thromboses, cerebral thrombosis, hemolytic anemia, thrombocytopenia, and pulmonary hypertension (American College of Obstetricians and Gynecologists, 2012b; Clowse, 2008). In 191 LAC-negative women with antiphospholipid-antibody syndrome, women with antibodies to cardiolipin and β2-glycoprotein I had significantly higher miscarriage rates than if either one alone was positive (Liu, 2013).
It is not precisely known how these antibodies cause damage, but it is likely that their actions are multifactorial (Tsokos, 2011). Platelets may be damaged directly by antiphospholipid antibody or indirectly by binding β2-glycoprotein I, which causes platelets to be susceptible to aggregation (Chamley, 1999; Giannakopoulos, 2013). Rand and colleagues (1997a,b, 1998) propose that phospholipid-containing endothelial cell or syncytiotrophoblast membranes may be damaged directly by the antiphospholipid antibody or indirectly by antibody binding to either β2-glycoprotein I or annexin V. This prevents the cell membranes from protecting the syncytiotrophoblast and endothelium. It results in exposure of basement membrane to which damaged platelets can adhere and form a thrombus (Lubbe, 1984).
Pierro and associates (1999) reported that antiphospholipid antibodies decreased decidual production of the vasodilating prostaglandin E2. Decreased protein C or S activity and increased prothrombin activation may also be contributory (Ogunyemi, 2002; Zangari, 1997). Amengual and coworkers (2003) presented evidence that thrombosis with APS is due to activation of the tissue factor pathway. Finally, uncontrolled placental complement activation by antiphospholipid antibodies may play a role in fetal loss and growth restriction (Holers, 2002).
Complications from APS cannot be completely explained by thrombosis alone. Animal models suggest that these are due to inflammation rather than thrombosis (Cohen, 2011). Some investigators theorize that APS-associated clotting is triggered by a “second hit” from innate inflammatory immune responses and recommend antiinflammatory agents (Meroni, 2011).
Adverse Pregnancy Outcomes
Antiphospholipid antibodies are associated with increased rates of fetal wastage (Chap. 18, Immunological Factors). In most early reports, however, women were included because they had had repeated adverse outcomes. Both are common—recall that the incidence of antiphospholipid antibodies in the general obstetrical population is about 5 percent and early pregnancy loss approximates 20 percent. Accordingly, data currently are too limited to allow precise conclusions to be drawn concerning the impact of these antibodies on adverse pregnancy outcomes. Fetal deaths, however, are more characteristic with APS than are first-trimester miscarriages (Oshiro, 1996; Roque, 2001). It has also been shown that women with higher antibody titers have worse outcomes compared with those with low titers (Nodler, 2009; Simchen, 2009).
Looking at the issue another way, the frequency of antiphospholipid antibodies may be increased with associated adverse obstetrical outcomes. Polzin and colleagues (1991) identified antiphospholipid antibodies in a fourth of 37 women with growth-restricted fetuses, however, none had evidence for lupus anticoagulant. Approximately a third of women with APS will develop preeclampsia during pregnancy (Clark, 2007b). Moodley and associates (1995) found antiphospholipid antibodies in 11 percent of 34 women with severe preeclampsia before 30 weeks.
When otherwise unexplained fetal deaths are examined, the data are mixed. Haddow and coworkers (1991) measured ACAs in 309 pregnancies with fetal death and found no differences compared with those in 618 normal pregnancies. In women with a history of recurrent pregnancy loss, those with antiphospholipid antibodies had a higher rate of preterm delivery (Clark, 2007a). In a case-control study of 582 stillbirths and 1547 live births, Silver and colleagues (2013) found a three- to fivefold increased risk for stillbirth in women with elevated anticardiolipin and anti-β2-glycoprotein I levels. Due to the risk of fetal-growth abnormalities and stillbirth, serial sonographic assessment of fetal growth and antepartum testing in the third trimester are recommended by the American College of Obstetricians and Gynecologists (2012b).
Because of the heterogeneity of studies, current treatment recommendations for women with antiphospholipid antibodies can be confusing (Branch, 2003; Robertson, 2006). As discussed, antiphospholipid antibodies that bind to immunoglobulins G, M, and A are semiquantified, and GPL, MPL, and APL binding units are expressed as negative, low-positive, medium-positive, or high-positive (American College of Obstetricians and Gynecologists, 2012b). Of the three, higher titers for GPL and MPL anticardiolipin antibodies are clinically important, whereas low-positive titers are of questionable clinical significance. And any titer of APL antibodies has no known relevance at this time.
As discussed in Chapter 52 (Acquired Thrombophilias), women with prior thromboembolic events who have antiphospholipid antibodies are at risk for recurrence in subsequent pregnancies. For these women, prophylactic anticoagulation with heparin throughout pregnancy and for 6 weeks postpartum with either heparin or warfarin is recommended (American College of Obstetricians and Gynecologists, 2012b). For those without history of thromboembolic events, recommendations for management from the American College of Obstetricians and Gynecologists (2012b) and the American College of Chest Physicians (Bates, 2012) are varied and listed in Table 52-8 (Some Recommendations for Thromboprophylaxis during Pregnancy). Some acceptable schemes include close antepartum observation with or without prophylactic or intermediate-dose heparin, and some form of postpartum anticoagulation for 4 to 6 weeks.
Recent trials have caused the need for heparin to be questioned in women with antibodies but no history of thrombosis (Branch, 2010). Although this is less clear, some recommend that women be treated if they have medium- or high-positive ACA titers or LAC activity and a previous second- or third-trimester fetal death not attributable to other causes (Dizon-Townson, 1998; Lockshin, 1995). Some report that women with recurrent early pregnancy loss and medium- or high-positive titers of antibodies may benefit from therapy (Robertson, 2006). Some individual agents are now considered for pregnant women with antiphospholipid antibody syndrome who have no prior thromboembolic event. These agents are described in the following paragraphs.
Aspirin, given in doses of 60 to 80 mg orally daily, blocks the conversion of arachidonic acid to thromboxane A2 while sparing prostacyclin production. This reduces synthesis of thromboxane A2, which usually causes platelet aggregation and vasoconstriction, and simultaneously spares prostacyclin, which normally has the opposite effect. There appear to be no major side effects from low-dose aspirin other than a slight risk of small-vessel bleeding during surgical procedures. Low-dose aspirin does not appear to reduce adverse pregnancy outcomes in antiphospholipid antibody-positive women without the complete syndrome (Del Ross, 2013).
Unfractionated heparin is given subcutaneously in dosages of 5000 to 10,000 units every 12 hours. Some prefer low-molecular-weight heparin, such as 40 mg enoxaparin once daily, because of its ease of administration and smaller risk of osteoporosis and heparin-induced thrombocytopenia (Kwak-Kim, 2013). With therapeutic dosing, some recommend measurement of heparin levels because clotting tests may be altered by lupus anticoagulant (Cowchock, 1998). The rationale for heparin therapy is to prevent venous and arterial thrombotic episodes. Heparin therapy also prevents thrombosis in the microcirculation, including the decidual-trophoblastic interface (Toglia, 1996). As discussed, heparin binds to β2-glycoprotein I, which coats the syncytiotrophoblast. This prevents binding of anticardiolipin and anti-β2-glycoprotein I antibodies to their surfaces, which likely prevents cellular damage (Chamley, 1999; Tsokos, 2011). Heparin also binds to antiphospholipid antibodies in vitro and likely in vivo. It is problematic that heparin therapy is associated with a number of complications that include bleeding, thrombocytopenia, osteopenia, and osteoporosis. A detailed description of the use of various heparins and their doses and adverse effects is found in Chapter 52 (Unfractionated Heparin).
Glucocorticosteroids generally should not be used with primary APS—that is, without an associated connective-tissue disorder. For women with lupus or those being treated for APS who develop lupus, corticosteroid therapy is indicated (Carbone, 1999). In such cases of secondary APS seen with lupus, the dose of prednisone should be maintained at the lowest effective level to prevent flares.
Immunoglobulin therapy is controversial and has usually been reserved for women with either overt disease or heparin-induced thrombocytopenia, or both (Alijotas-Reig, 2013). It is used when other first-line therapies have failed, especially in the setting of preeclampsia and fetal-growth restriction (Cowchock, 1996, 1998; Heilmann, 2003; Silver, 1997). Intravenous immunoglobulin (IVIG) is administered in doses of 0.4 g/kg daily for 5 days—total dose of 2 g/kg. This is repeated monthly, or it is given as a single dose of 1 g/kg each month. Treatment is expensive, and at $75 per gram, one course for a 70-kg woman costs more than $10,000. A Cochrane review of seven trials found no improvement in the livebirth rate for immunotherapy given to women with recurrent pregnancy loss (Porter, 2006). Xiao and associates (2013) studied prednisone and aspirin in 87 patients compared with prednisone, aspirin, low-molecular-weight heparin, and IVIG in 42 women with antiphospholipid antibody syndrome. They reported livebirth rates of 84 and 98 percent and obstetrical morbidity rates of 23 and 7 percent, respectively. We are of the view that randomized trials are needed before there is widespread application of this expensive and cumbersome therapy.
Immunosuppression has not been well evaluated, but azathioprine and cyclosporine do not appear to improve standard therapies (Silver, 1997). As mentioned with lupus on Pharmacological Treatment, methotrexate, cyclophosphamide, and mycophenolate mofetil are contraindicated because of teratogenic potential (Briggs, 2011; Buhimschi, 2009). Hydroxychloroquine therapy may be beneficial with APS (Albert, 2014).
Aspirin plus heparin is the most efficacious regimen (de Jesus, 2014). Low-dose unfractionated heparin—7500 to 10,000 units is used subcutaneously twice daily. Concurrently, low-dose aspirin—60 to 80 mg orally once daily—is given. Although typically given in addition to heparin in women with a prior thromboembolic event, the benefit of aspirin is unclear (American College of Obstetricians and Gynecologists, 2012b; Bouvier, 2014).
Fetal loss is common in women with antiphospholipid antibodies if untreated (Branch, 1992; Rai, 1995). Even with treatment, recurrent fetal loss rates remain at 20 to 30 percent (Branch, 2003; Empson, 2005; Ernest, 2011). Importantly, some women with lupus and antiphospholipid antibodies have normal pregnancy outcomes without treatment. Again, it is emphasized that women with lupus anticoagulant and prior bad pregnancy outcomes have had liveborns without treatment (Trudinger, 1988).
In a manner similar to neonatal lupus syndrome (Perinatal Mortality and Morbidity), up to 30 percent of neonates demonstrate passively acquired antiphospholipid antibodies, and thus there is concern for their adverse effects. For example, Tincani and coworkers (2009) found increased learning disabilities in these children. Simchen and colleagues (2009) reported a fourfold increased risk for perinatal strokes. Of 141 infants followed in a European registry, the rate of preterm birth was 16 percent; low birthweight, 17 percent; and behavioral abnormalities in 4 percent of the children. There were no cases of neonatal thrombosis (Motta, 2010).
This is a chronic multisystem disease of unknown cause with an immunologically mediated pathogenesis. Infiltrating T cells secrete cytokines that mediate inflammation and systemic symptoms. Its worldwide prevalence is 0.5 to 1 percent, women are affected three times more often than men, and peak onset is from 25 to 55 years (Shah, 2012). The cardinal feature is inflammatory synovitis that usually involves the peripheral joints. The disease has a propensity for cartilage destruction, bony erosions, and joint deformities. Criteria for classification are shown in Table 59-6, and a score of 6 or greater fulfills the requirements for definitive diagnosis.
TABLE 59-6Criteria for Classification of Rheumatoid Arthritis ||Download (.pdf) TABLE 59-6 Criteria for Classification of Rheumatoid Arthritis
|Factor ||Criteria ||Score |
|Joint involvement ||1 large—shoulder, elbow, hip, knee, ankle ||0 |
| ||2–10 large ||1 |
| ||1–3 small—MCP, PIP, thumb IP, MTP, wrists ||2 |
| ||4–10 small ||3 |
| ||> 10—at least 1 small ||5 |
|Serological testing ||Negative RF and negative ACPA ||0 |
| ||Low-positive RF or anti-CCP ||2 |
| ||High-positive RF or anti-CCP ||3 |
|Acute-phase reactants ||Normal CRP and ESR ||0 |
| ||Abnormal CRP or ESR ||1 |
|Duration of symptoms ||Less than 6 weeks ||0 |
| ||6 weeks or longer ||1 |
There is a genetic predisposition, and heritability is estimated at 15 to 30 percent (McInnes, 2011). Genome-wide associated studies have identified more than 30 loci involved in rheumatoid arthritis pathogenesis (Kurkó, 2013). There is an association with the class II major histocompatibility complex molecule HLA-DR4 and HLA-DRB1 alleles (McInnes, 2011; Shah, 2012). In addition, there is a genetic component to joint destruction in rheumatoid arthritis, as alterations in the IL2RA gene are associated with less destructive disease (Knevel, 2013). Cigarette smoking appears to increase the risk of rheumatoid arthritis (Papadopoulos, 2005). As discussed subsequently, a protective effect of pregnancy has been reported for development of rheumatoid arthritis, and this may be related to HLA-disparate fetal microchimerism (Guthrie, 2010; Hazes, 1990).
Rheumatoid arthritis is a chronic polyarthritis with symptoms of synovitis, fatigue, anorexia, weakness, weight loss, depression, and vague musculoskeletal symptoms. The hands, wrists, knees, and feet are commonly involved. Pain, aggravated by movement, is accompanied by swelling and tenderness. Extraarticular manifestations include rheumatoid nodules, vasculitis, and pleuropulmonary symptoms. The American College of Rheumatology criteria for rheumatoid arthritis diagnosis have been revised and are shown in Table 59-6 (Aletaha, 2010). The revision focuses on features at earlier stages to permit earlier therapy initiation with a goal of preventing or decreasing long-term sequelae.
Treatment is directed at pain relief, inflammation reduction, protection of articular structures, and preservation of function. Physical and occupational therapy and self-management instructions are essential. Although aspirin and other NSAIDs are the cornerstone of symptomatic therapy, they do not retard disease progression. Conventional NSAIDs nonspecifically inhibit both cyclooxygenase-1 (COX-1)—an enzyme critical to normal platelet function, and COX-2—an enzyme that mediates inflammatory response mechanisms. Gastritis with acute bleeding is an unwanted side effect common to conventional NSAIDs. Specific COX-2 inhibitors have been used to avert this complication. However, there are now concerns that their long-term use is associated with increased risk for myocardial infarction, stroke, and heart failure (Solomon, 2005). A systematic review by Adams and colleagues (2012) identified two studies of women treated for inflammatory arthritis. They reported a higher rate of cardiac malformations in infants exposed to NSAIDs in the first trimester. In addition, NSAIDs are associated with early spontaneous abortions, ductus constriction, and neonatal pulmonary hypertension (Briggs, 2011). Thus, risks versus benefits of these medications must be considered.
Glucocorticoid therapy may be added to NSAIDs. Of these, prednisone, 7.5 mg orally daily for the first 2 years of active disease, substantively reduces progressive joint erosions (Kirwan, 1995). Although corticosteroids are usually avoided if possible, low-dose therapy is used by some along with salicylates.
Because NSAIDs and glucocorticoids are primarily for symptomatic relief, the American College of Rheumatology (2012) recommends disease-modifying antirheumatic drugs (DMARDs). These have the potential to reduce or prevent joint damage and are given within 3 months of diagnosis. There are several DMARDs, and many have considerable toxicity. For example, methotrexate and leflunomide are teratogenic (Briggs, 2011). Sulfasalazine and hydroxychloroquine are safe for use in pregnancy (Partlett, 2011; Raza, 2010). These, combined with COX-2 inhibitors and with relatively low-dose prednisone—7.5 to 20 mg daily—for flares are often all the pharmacotherapy needed. Azathioprine treatment may be used if immunosuppressive drugs are warranted. From their review, Kuriya and associates (2011) found that a fourth of women with rheumatoid arthritis took a DMARD within 6 months of conception. During pregnancy, 4 percent of 393 pregnant women were given a category D or X medication. Methotrexate was the most common—2.9 percent. These authors advocate continued efforts to counsel women and their physicians regarding potential risks of many rheumatoid arthritis therapies during pregnancy.
Cytokine inhibitors, which include etanercept (Enbrel), adalimumab (Humira), and infliximab (Remicade), are also being used (McInnes, 2011). The newer immune-targeted drugs for rheumatoid arthritis have shown great promise. But, their use in pregnancy is limited, and fetal safety issues are a concern (Makol, 2011; Ojeda-Uribe, 2013). In the review by Kuriya and coworkers (2011) discussed earlier, 13 percent of 393 women were given a biological cytokine-inhibiting DMARD—primarily etanercept. Experience with these agents in pregnancy has been accruing (Verstappen, 2011). Berthelot and colleagues (2009) found more than 300 reported cases in their review and noted no fetal effects from exposure. Patients considered for etanercept treatment can be entered into a registry at 877-311-8972. There is also little known regarding pregnancy effects of the interleukin-1 receptor antagonist, anakinra (Kineret), and the antagonist to the B-cell CD20 antigen, rituximab (Rituxan).
Other drugs used less frequently include penicillamine, gold salts, minocycline, and cyclosporine. Orthopedic surgery for joint deformities, including replacement, is frequently performed outside of pregnancy.
Pregnancy and Rheumatoid Arthritis
Up to 90 percent of women with rheumatoid arthritis will experience improvement during pregnancy (de Man, 2008). Animal studies suggest this may be due to regulatory T-cell alterations. Transfer of regulatory T cells from pregnant mice conferred protection to those not pregnant (Munoz-Suano, 2012). That said, some women develop disease during pregnancy, and others become worse (Nelson, 1997).
A downside to this respite during pregnancy is that postpartum exacerbation is common (Østensen, 2007). This may be due to alterations in innate immunity postpartum (Häupl, 2008b). Barrett and coworkers (2000a) reported that a postpartum flare was more common if women were breast feeding. These same investigators performed a prospective United Kingdom study involving 140 women with rheumatoid arthritis (Barrett, 2000b). During 1 to 6 months postpartum, there was only a modest fall in objective disease activity, and only 16 percent had complete remission. At least 25 percent had substantive levels of disability. They observed that although overall disease actually did not exacerbate postpartum, the mean number of inflamed joints increased significantly.
There are some studies that report a protective effect of pregnancy for women developing subsequent new-onset rheumatoid arthritis. Silman and associates (1992) performed a case-control study of 88 affected women. They found that although there was a protective effect of pregnancy in the long term, the likelihood of new-onset rheumatoid arthritis was increased sixfold during the first 3 postpartum months. Pikwer and colleagues (2009) reported a significant reduction in the risk of subsequent arthritis in women who breast fed longer than 12 months.
As an explanation for these findings, and as discussed previously, sex hormones supposedly interfere with a number of putative processes involved in arthritis pathogenesis, including immunoregulation and interactions with the cytokine system (Häupl, 2008a,b). First, Unger and associates (1983) reported that amelioration of rheumatoid arthritis correlated with serum levels of pregnancy-associated α2-glycoprotein. This compound has immunosuppressive properties. Second, Nelson and coworkers (1993) reported that amelioration of disease was associated with a disparity in HLA class II antigens between mother and fetus. They suggested that the maternal immune response to paternal HLA antigens may play a role in pregnancy-induced remission of arthritis. In addition to monocyte activations, there also may be T-lymphocyte activation (Förger, 2008).
Juvenile Rheumatoid Arthritis
This is a group of diseases that are the most frequent cause of chronic arthritis in children. They persist into adulthood. Østensen (1991) reviewed outcomes of 76 pregnancies in 51 affected Norwegian women. Pregnancy had no effects on presentation of disease, but disease activity usually became quiescent or remained so during pregnancy. Postpartum flares were common as was discussed for rheumatoid arthritis. Joint deformities often developed in these women, and 15 of 20 cesarean deliveries were done for contracted pelves or joint prostheses. These observations are supported by the summary of similar results in 39 Polish women with juvenile rheumatoid arthritis (Musiej-Nowakowska, 1999).
There are no obvious adverse effects of rheumatoid arthritis on pregnancy outcome, but an increased risk for preterm birth is debated (Klipple, 1989; Langen, 2014; Wallenius, 2011). In a population-based cohort study, Skomsvoll and associates (2002) reported that women with rheumatic disease with two previous poor pregnancy outcomes were at high risk for recurrent adverse outcomes. Kaplan (1986) reported that women who later develop the disease had a prior higher-than-expected incidence of spontaneous abortion. However, Nelson and colleagues (1992) did not corroborate this.
Management During Pregnancy
Treatment of symptomatic women during pregnancy is with aspirin and NSAIDs. These are used with appropriate concerns for first-trimester effects, impaired hemostasis, prolonged gestation, premature ductus arteriosus closure, and persistent pulmonary circulation (Chap. 12, Antifungal Medications) (Briggs, 2011). Low-dose corticosteroids are also prescribed as indicated. Gold compounds have been administered in pregnancy. Of 14 women experiencing 20 pregnancies on gold therapy, 75 percent delivered healthy liveborn neonates (Almarzouqi, 2007).
Immunosuppressive therapy with azathioprine, cyclophosphamide, or methotrexate is not routinely used during pregnancy. Of these, only azathioprine should be considered during early pregnancy because the other agents are teratogenic (Briggs, 2011; Buhimschi, 2009). As noted earlier, DMARDs including sulfasalazine and hydroxychloroquine are acceptable for use in pregnancy.
If the cervical spine is involved, particular attention is warranted during pregnancy. Subluxation is common, and pregnancy, at least theoretically, predisposes to this because of joint laxity (Chap. 4, Musculoskeletal System). Importantly, there are anesthesia concerns regarding endotracheal intubation.
Following pregnancy, contraceptive counseling may include combination oral contraceptives. These are a logical choice because of their effectiveness and their potential to improve disease (Bijlsma, 1992; Farr, 2010). That said, all methods of contraception are appropriate.
This is a chronic multisystem disorder of unknown etiology. It is characterized by microvascular damage, immune system activation leading to inflammation, and excessive deposition of collagen in the skin and often in the lungs, heart, gastrointestinal tract, and kidneys. It is uncommon, displays a 3-to-1 female dominance, and typically affects those aged 30 to 50 years.
This strong prevalence of scleroderma in women and its increased incidence in the years following childbirth contribute to the hypothesis that microchimerism is involved as discussed on Immune-Mediated Connective-Tissue Diseases (Lambert, 2010). Artlett and coworkers (1998) demonstrated Y-chromosomal DNA in almost half of women with systemic sclerosis compared with only 4 percent of controls. Rak and colleagues (2009b) identified male microchimerism in peripheral blood mononuclear cells more frequently in women with limited versus diffuse scleroderma—20 versus 5 percent.
The hallmark is overproduction of normal collagen. In the more benign form—limited cutaneous systemic sclerosis—progression is slow. With diffuse cutaneous systemic sclerosis, skin thickening progresses rapidly, and skin fibrosis is followed by gastrointestinal tract fibrosis, especially the distal esophagus (Varga, 2012). Pulmonary interstitial fibrosis along with vascular changes may cause pulmonary hypertension. Antinuclear antibodies are found in 95 percent of patients, and immunoincompetence often develops.
Frequent symptoms are Raynaud phenomenon, which includes cold-induced episodic digital ischemia in 95 percent, as well as swelling of the distal extremities and face. Half of patients have symptoms from esophageal involvement, especially fullness and epigastric burning pain. Pulmonary involvement is common and causes dyspnea. The 10-year cumulative survival rate is 70 percent in those with pulmonary fibrosis, and pulmonary arterial hypertension is the main causes of death (Joven, 2010; Ranque, 2010). Women with limited cutaneous disease such as the CREST syndrome—calcinosis, Raynaud phenomenon, esophageal involvement, sclerodactyly, and telangiectasia—have milder disease.
Overlap syndrome refers to the presence of systemic sclerosis with features of other connective-tissue disorders. Mixed-connective-tissue disease is a term used for the syndrome involving features of lupus, systemic sclerosis, polymyositis, rheumatoid arthritis, and high titers of anti-RNP antibodies (see Table 59-1). The disorder is also termed undifferentiated connective-tissue disease (Spinillo, 2008).
Although systemic sclerosis cannot be cured, treatment directed at end-organ involvement can relieve symptoms and improve function (Varga, 2012). Hydroxychloroquine and low-dose corticosteroids are helpful, and intravenous immunoglobulin may be indicated for pericarditis and hemolytic anemia. Renal involvement and hypertension are common. At times, angiotensin-converting enzyme (ACE) inhibitors may be required for blood pressure control despite their known teratogenicity (Buhimschi, 2009). Scleroderma renal crisis develops in up to a fourth of these patients and is characterized by obliterative vasculopathy of the renal cortical arteries. This results in renal failure and malignant hypertension. Interstitial restrictive lung disease is common and frequently becomes life threatening. Associated pulmonary hypertension is usually fatal. Cyclophosphamide provides some improvement (Tashkin, 2006).
Pregnancy and Systemic Sclerosis
The prevalence of scleroderma in pregnancy is estimated from a study of nearly 11.2 million pregnant women registered in the Nationwide Inpatient Sample. In this study, Chakravarty and associates (2008) reported that 504 women suffered from systemic sclerosis—a prevalence of approximately 1 in 22,000 pregnancies. These women usually have stable disease during gestation if their baseline function is good. As perhaps expected, dysphagia and reflux esophagitis are aggravated by pregnancy (Steen, 1999). Dysphagia results from loss of esophageal motility due to neuromuscular dysfunction. A decrease in amplitude or disappearance of peristaltic waves in the lower two thirds of the esophagus is seen using manometry. Symptomatic treatment for reflux is described in Chapter 54 (Management). Women with renal insufficiency and malignant hypertension have an increased incidence of superimposed preeclampsia. In the presence of rapidly worsening renal or cardiac disease, pregnancy termination should be considered. As discussed, renal crisis is life threatening and is treated with ACE inhibitors, but it does not improve with delivery (Gayed, 2007). Pulmonary hypertension usually contraindicates pregnancy (Chap. 49, Prognosis).
Vaginal delivery may be anticipated, unless the soft tissue thickening wrought by scleroderma produces dystocia requiring cesarean delivery. Tracheal intubation for general anesthesia has special concerns because of limited ability of these women to open their mouths widely (Black, 1989). Because of esophageal dysfunction, aspiration is also more likely, and epidural analgesia is preferable.
Maternal and fetal outcomes are related to the severity of underlying disease. Steen and colleagues (1989, 1999) described pregnancies in 214 women with systemic sclerosis—45 percent had diffuse disease. Major complications included renal crisis in three women and increased rates of preterm birth. Chung and coworkers (2006) also reported increased rates of preterm delivery, fetal-growth restriction, and perinatal mortality. A multicenter study of 109 pregnancies from 25 centers reported higher rates of preterm delivery, fetal-growth restriction, and very-low-birthweight infants (Taraboreli, 2012). These are likely related to placental abnormalities that include decidual vasculopathy, acute atherosis, and infarcts. These reduce placental blood flow and are found in many cases (Doss, 1998; Papakonstantinou, 2007).
Scleroderma may be associated with subfertility (Bernatsky, 2008; Lambe, 2004). For women who do not choose pregnancy, several reversible contraceptive methods are acceptable. However, hormonal agents, especially combination oral contraceptives, probably should not be used, especially in women with pulmonary, cardiac, or renal involvement. Due to the often unrelenting progression of systemic sclerosis, permanent sterilization should also be considered.
Inflammation and damage to blood vessels may be primary or due to another disease. Most cases are presumed to be caused by immune-complex deposition (Langford, 2012). Primary causes include polyarteritis nodosa, Wegener granulomatosis, Churg-Strauss syndrome, temporal or giant cell arteritis, Takayasu arteritis, Henoch-Schönlein purpura, Behçet syndrome, and cutaneous or hypersensitivity arteritis (Goodman, 2014). Some of these syndromes have antibodies directed against proteins in the cytoplasmic granules of leukocytes—antineutrophil cytoplasmic antibodies—ANCA. Their role in vasculitis is unclear, and disease activity is not related to increasing titers.
This necrotizing vasculitis of small and medium-sized arteries is characterized clinically by myalgia, neuropathy, gastrointestinal disorders, hypertension, and renal disease (Goodman, 2014). Approximately a third of cases are associated with hepatitis B antigenemia (Langford, 2012).
Symptoms are nonspecific and vague. Fever, weight loss, and malaise are present in more than half of cases. Renal failure, hypertension, and arthralgias are common. Diagnosis is made by biopsy, and treatment consists of high-dose prednisone plus cyclophosphamide. Vasculitis due to hepatitis B antigenemia responds to antivirals, glucocorticoids, and plasma exchange (Chap. 55, Hepatitis B).
There have been only a few reports of documented cases of polyarteritis nodosa associated with pregnancy. Even without definitive data, active arteritis in pregnancy is associated with high mortality rates. Owen and Hauth (1989) reviewed the courses of 12 such pregnant women. In seven, polyarteritis first manifested during pregnancy, and it was rapidly fatal by 6 weeks postpartum. The diagnosis was not made until autopsy in six of the seven women. Four women continued pregnancy, which resulted in one stillborn and three successful outcomes.
This is a necrotizing granulomatous vasculitis of the upper and lower respiratory tract and kidney. Disease frequently includes sinusitis and nasal disease—90 percent; pulmonary infiltrates, cavities, or nodules—85 percent; glomerulonephritis—75 percent; and musculoskeletal lesions—65 percent (Sneller, 1995). At least 90 percent have polyangiitis (Langford, 2012). It is uncommon and usually encountered after 50 years of age. Koukoura and associates (2008) reviewed 36 cases reported in association with pregnancy. In another report, a woman with Wegener disease developed thrombotic microangiopathy associated renal impairment at 32 weeks. A second woman had pneumonitis, but pregnancy did not appear to affect disease activity (Pagnoux, 2011). Because subglottic stenosis is found in up to a fourth, the anesthesia team should be consulted during pregnancy (Engel, 2011; Kayatas, 2012).
Corticosteroids are standard treatment, but azathioprine, cyclosporine, and intravenous immunoglobulin may also be used. For severe disease in the late second or third trimester, cyclophosphamide in combination with prednisolone seems acceptable.
Also called pulseless disease, this syndrome is most prevalent in young women. It is a chronic inflammatory arteritis affecting large vessels (Goodman, 2014). Unlike temporal arteritis, which develops almost exclusively after age 55, the onset of Takayasu arteritis is almost always before age 40. It is associated with abnormal angiography of the upper aorta and its main branches and with upper extremity vascular impairment. Death usually results from congestive heart failure or cerebrovascular events. Computed tomography or magnetic resonance angiography can detect this disorder before the development of severe vascular compromise (Numano, 1999). Takayasu arteritis may respond symptomatically to corticosteroid therapy, however, it is not curative. Selection of surgical bypass or angioplasty has improved survival rates.
Severe renovascular hypertension, cardiac involvement, or pulmonary hypertension worsen pregnancy prognosis. Hypertension is relatively common and should be carefully controlled. Blood pressure is measured in the lower extremity. Overall, the prognosis for pregnancy is good (de Jesús, 2012; Hernández-Pacheco, 2011; Hidaka, 2012; Johnston, 2002; Kraemer, 2008; Mandal, 2012). When the abdominal aorta is involved, however, pregnancy outcome may be disastrous (Sharma, 2000). Vaginal delivery is preferred, and epidural analgesia has been advocated for labor and delivery (Langford, 2002).
Vasculitis caused by Henoch-Schönlein purpura is uncommon after childhood. Tayabali and associates (2012) reviewed 20 pregnancies complicated by this vasculitis and described cutaneous lesions in three fourths. Approximately half had arthralgias. Jadaon and colleagues (2005) described 135 pregnancies in 31 women with Behçet disease compared with matched nonaffected controls. Although there were no deleterious effects of pregnancy on the underlying disease, the miscarriage rate was significantly increased threefold. Hwang and coworkers (2009) described necrotizing villitis and decidual vasculitis in a placenta from a first-trimester pregnancy termination and another from a term delivery. Churg-Strauss vasculitis is rare in pregnancy. Hot and associates (2007) described a pregnant woman who responded to intravenous immune globulin therapy. Corradi and associates (2009) described an affected 35-year-old woman at term whose necrotizing vasculitis involved the heart, and she subsequently underwent cardiac transplantation.