Systemic lupus erythematosus (SLE) is a multisystemic chronic inflammatory disease that affects patients in many different ways over a varying course of time. The disease is typically characterized by periods of remission and relapse, although the causes of exacerbation remain uncertain. SLE, like most autoimmune diseases, has a clear predilection for women. Indeed, women are affected seven times more frequently than men. The disorder may be diagnosed between the ages of 15 and 50 years, although it is most often detected in women in their twenties. Therefore, SLE is the most commonly encountered autoimmune disease in pregnancy. Although no specific gene mutation for SLE has been identified, the disease likely has a genetic component.¹ Approximately 10% of affected patients have a relative with SLE and monozygotic twin studies demonstrate that 50% of affected twins are concordant for the disease. It is estimated that about 2% of children born to mothers with SLE will develop the disease themselves.² The symptoms of SLE are extremely heterogeneous which can make the diagnosis difficult. The disease may affect joints, skin, kidneys, lung, nervous system, and other organs. The most common presenting complaints are extreme fatigue, arthralgias, fever, and rash (Table 27-1).

In 1982, the American Rheumatism Association (ARA) revised previously set criteria for the diagnosis of SLE³ (Table 27-2). According to the ARA, a person must have had at least 4 of the 11 specific criteria in order to carry the diagnosis of SLE. However, many patients have less than four clinical or laboratory features of SLE and do not meet strict diagnostic criteria. These patients should not be considered to have SLE, but are often referred to as having lupus-like disease. Such individuals may benefit from therapies for SLE and some will ultimately develop the clinical syndrome.

The Effect of Pregnancy on SLE

Fertility

Typically, patients with SLE do not have impaired infertility. However, patients on high dose steroids may become amenorrheic or anovulatory. Women with end-stage lupus nephritis requiring dialysis also are frequently amenorrheic. In addition, depending on the cumulative dose of medication and the age of the patient, 10% to 60% of patients who have been treated with cyclophosphamide become permanently amenorrheic. Patients with mild-moderate disease have fertility rates comparable to the general population and should be counseled appropriately about contraception unless they desire to become pregnant. Estrogen-containing oral contraceptives (and other forms of contraception) are considered safe to use in women with SLE as long as they do not have comorbidities such as thrombosis or hypertension.

Maternal Complications

A recent study using a database including detailed information regarding about 20% of all (not necessarily pregnancy related) hospitalizations in the United States estimated a 20-fold increased risk in mortality in women with SLE.⁴ There was a 3- to 7-fold increase in the risk for thrombosis, infection, thrombocytopenia, and the need for blood transfusion. Women with SLE also are more likely to have comorbid conditions such as diabetes and hypertension that are associated with adverse maternal (and fetal) outcomes.

Lupus Flares

There is an association between estrogen and SLE, as evidenced by the female predilection for the disorder.⁵ Thus, conditions such as pregnancy that are associated with high estrogen levels have the potential to exacerbate SLE. The reported incidence of flares during pregnancy ranges between 15% and 63%. Several retrospective, uncontrolled studies performed prior to 1985 suggest that pregnancy exacerbates lupus flares. There have been numerous, mostly prospective studies on the frequency of lupus flares during pregnancy.^2,6-14 It is difficult to interpret available data because control groups were often unmatched, and the SLE cohorts among studies vary greatly regarding patient characteristics including race, severity of disease, and the definition of lupus flares. Furthermore, normal physiologic changes of pregnancy such as palmar erythema, facial blushing, proteinuria, and alopecia can be misinterpreted as lupus flares.

Doria and colleagues investigated the relationship of steroid hormone levels in pregnancy to SLE activity.¹⁵ The group prospectively studied 17 women with lupus during pregnancy and matched them to eight healthy pregnant controls. They reported that women with SLE had significantly lower serum levels of estradiol and progesterone than controls. Furthermore, the highest levels of estrogen and progesterone occurred in the third trimester, when patients with SLE had both the lowest disease activity and serum immunoglobulin levels. These data challenge previous work that supports the association between increased levels of steroid hormones and lupus activity, and raise the question of whether or not estrogens and progesterones suppress humoral immune responses and therefore disease activity.

Regardless of whether or not the rate of SLE flares increase during pregnancy, flares are common and may occur in any trimester, or in the postpartum period. In general lupus flares during pregnancy are mild and easily treated.¹⁶ Furthermore, it has been demonstrated that active disease at the time of conception, active nephritis, a systemic lupus erythematosus disease activity index (SLEDAI) score of five or more, and abruptly stopping hydroxychloroquine therapy are significant risk factors for lupus flares. Approximately, 50% of patients with active disease at the time of conception experience flares during pregnancy compared to 20% of patients who are in remission when they conceive. Conversely, patients who have been in remission for 6 to 12 months prior to conception have a lower risk of lupus flares and do better than those with active disease.^17-19

Preexisting Renal Disease

Approximately, 50% of patients with lupus will develop renal disease. Lupus nephritis is a result of immune complex deposition, complement activation, and inflammation in the kidney. Several reports have emphasized the potential for a permanent decrease in renal function after pregnancy in women with lupus nephritis. On the other hand, more recent series indicate excellent outcome for most women with mild renal disease.^20-22 Burkett reviewed several retrospective reports including 242 pregnancies in 156 women with lupus nephritis.²³ He demonstrated that 59% of patients had no change in their renal function, 30% experienced transient renal impairment, and 7% had permanent renal insufficiency. Similar results were noted in several recent cohorts.^17,18,24 Most patients with prior nephritis had successful pregnancies and flares were predicted by renal status (remission decreases the risk) at conception.

It is clear that there is a strong correlation between renal insufficiency prior to conception and the risk of deterioration during and after pregnancy. Women with a serum creatinine level greater than 1.5 mg/dL have a significantly increased risk of deterioration in renal function. Conversely, patients with serum creatinine levels less than 1.5 mg/dL can be reassured that pregnancy will not increase the rate of deterioration of renal function. The specific type of renal disease as demonstrated by histologic studies does not appear to influence pregnancy outcome or postnatal renal function.

Preeclampsia

Preeclampsia is among the most common pregnancy complications in patients with SLE. The incidence ranges between 20% and 35%. The cause of the increased incidence of preeclampsia in women with SLE is not clear, but may be due to unrecognized renal disease that is likely present in many patients with SLE. Renal disease, hypertension, and antiphospholipid syndrome (APS) all increase a patient’s risk for developing preeclampsia. In the prospective study of Lockshin et al.,¹⁰ 8 of 11 (72%) patients with lupus nephritis developed preeclampsia compared to 12 of 53 (22%) women who did not have nephritis. In some cases, it is difficult to distinguish preeclampsia from a lupus flare manifesting as lupus nephritis. Both disorders may be characterized by increased proteinuria, hypertension, and fetal growth restriction (FGR). Table 27-3 lists several features that may aid in the distinction of preeclampsia from nephritis. Despite these parameters, it is often difficult to distinguish between the two, and in some cases a renal biopsy may be required to differentiate between the conditions. For example, confirmation of lupus nephritis may prevent unnecessary iatrogenic preterm birth in an attempt to treat preeclampsia. Although there is a theoretical increased risk of complications from the procedure, it has been performed safely during pregnancy. Preeclampsia and lupus nephrits may also coexist and a definitive diagnosis cannot always be made.

Fetal Complications

Pregnancy Loss

Patients with SLE have an overall increased risk of pregnancy loss. The rate of first trimester spontaneous miscarriage is as high as 35%. The risk of fetal death is also increased and approaches 22% in some series. Several factors have been associated with pregnancy loss in women with SLE including APS (discussed later), renal disease (especially Class III-IV glomerulonephritis), active disease during pregnancy, a history of fetal loss, and African American and Hispanic race/ethnicity.^25-29 However, in the absence of these, SLE patients have pregnancy loss rates that are similar to the general population.

Preterm Delivery

There is a higher incidence of preterm birth in patients with lupus than in healthy women. Preterm delivery less than 37 weeks has been reported in as few as 3% and as many as 73% of SLE pregnancies (median 30%). The variation in preterm birth in these studies may be due, in part, to the tendency of some obstetricians to intentionally deliver patients with SLE in order to avoid fetal morbidity. However, a well-designed cohort study by Johnson et al.³⁰ including careful obstetric detail, reported a 50% rate of preterm birth in patients with SLE. A recent meta-analysis and systematic review noted a 2.05 relative risk (RR) for preterm birth with SLE (95% confidence interval [CI], 1.72-3.32).³¹ The risk was higher (RR 2.98; 95% CI, 2.32-3.38) for those with active disease.^30,31 Preterm delivery typically occurs because of preeclampsia, FGR, abnormal fetal testing, and preterm premature rupture of membranes.³⁰ Increased disease activity, chronic hypertension, and antiphospholipid antibodies (aPL) are all associated with an increased risk for both medically indicated and spontaneous preterm delivery.

Neonatal Lupus Erythematosus

Neonatal lupus erythematosus (NLE) is a rare condition that occurs in approximately 1:20,000 live births. The disease is characterized by neonatal or fetal heart block, skin lesions, or less commonly, anemia, thrombocytopenia, and hepatitis. Approximately, 50% of fetuses with NLE have skin lesions, 50% have heart block, and 10% have both. NLE is an immune-mediated disease and is a result of transplacental passage of maternal autoantibodies. Most cases are associated with antibodies to the cytoplasmic ribonucleoproteins SSA (Ro), more specifically the five anti-SS2-kDa epitope of SSA. SSB (La) antibodies are also detected in 50% to 75% of these women. However, NLE is rarely associated with isolated antibodies to SSB.

Typical skin lesions associated with NLE are erythematous, scaling plaques usually seen on the scalp or face of the infant. The lesions appear within the first weeks after delivery and last only for a few months. Skin biopsies of the lesions show changes typical of cutaneous lupus in adults. The hematologic abnormalities of NLE also resolve within a few months, coinciding with the disappearance of maternal autoantibodies.

Cardiac lesions associated with NLE are heart block and endocardial fibroelastosis.^32,33 The anti-SSA (most specifically anti-SSA-52), binds to myocardial tissue. Histologic analysis of affected fetal hearts demonstrates mononuclear cell infiltration, fibrin deposition, calcification of the conduction system (specifically the AV and SA nodes), and diffuse fibroelastosis throughout the myocardium. It is hypothesized that the earliest effect of the antibody-mediated disease is global pancarditis with subsequent fibrosis of the conduction system. Congenital heart block is usually detected as fetal bradycardia with a rate between 60 and 80 beats/min between 16 and 25 weeks of gestation. Fetal echocardiography demonstrates a structurally normal heart with AV dissociation. In some cases, fetal hydrops develops in utero.

The presence of autoantibodies alone is insufficient to cause NLE. This is demonstrated by the fact that approximately 30% of patients with SLE have anti-SSA antibodies, and 15% to 20% of patients have anti-SSB autoantibodies. However, prospective studies indicate that the incidence of congenital heart block in infants born to women with SLE is only 2%. Also, the recurrence risk ranges between 15% and 20% and there are reports of twins who are discordant for NLE. Thus, anti-SSA alone does not always lead to NLE. It is also important to recognize that maternal SLE is not a prerequisite for NLE. In fact, up to 50% of cases occur in the offspring of healthy women with circulating autoantibodies. Some, but not all of these women eventually develop connective-tissue disorders. Prospective studies of women with anti-SSA or anti-SSB antibodies (regardless of SLE status) have confirmed that about 2% will have a child with NLE.

The clinical course of NLE is highly variable. Cutaneous and hematologic abnormalities resolve by 6 months of age. However, heart block is a permanent condition that is associated with significant morbidity and even mortality. Approximately, 15% to 20% of fetuses affected with heart block die within 3 years of age due to a fatal cardiomyopathy. Up to 60% of neonates require pacing during the neonatal period and most affected children eventually require permanent pacemakers before adulthood. There are few data regarding long-term outcomes. However, preliminary reports suggest that anti-SSA may be associated with an increased risk of dyslexia. Thus, long-term neuropsychological evaluation may be useful.

Whether treatment of heart block detected in utero is beneficial is not clear. Many clinicians advocate the use of flourinated corticosteroids since they cross the placenta. The rationale for steroid treatment is based on the fact that the cardiac histology of fetuses with CHB demonstrates diffuse inflammation, IgG, fibrin, and complement deposition. Initially, improvement in myocardial function was reported by some investigators. However, it is now clear that once heart block is complete it is irreversible even with steroid treatment. Thus, steroids should not be routinely used once heart block is complete. Some authorities advocate their use in cases of myocarditis, heart failure, or mild hydrops but efficacy is uncertain.³³

It is even more controversial as to whether screening for first or second-degree heart block and treating such patients with steroids can reduce progression to complete heart block. A cohort of pregnancies in women with anti-SSA and anti-SSB antibodies was prospectively followed with serial fetal echocardiograms (the PR interval and dexamethasone evaluation [PRIDE] study). The investigators noted rapid progression from normal sinus rhythm to complete heart block without a graded progression through early stage heart block. In some cases steroids appeared to reverse a prolonged PR interval. However, spontaneous reversal may also occur. Thus, the benefit of screening for and/or treating prolonged PR intervals in women with anti-SSA or anti-SSB remains unproven.^34,35

Importantly, there are significant maternal and fetal risks to treatment with high dose, chronic fluorinated corticosteroids including osteoporosis, glucose intolerance, adrenal suppression, FGR, decreased brain growth, learning disabilities, and developmental delay. Therefore, steroid therapy in the treatment of congenital heart block diagnosed in utero should be considered experimental and used with extreme caution.

Intavenous immune globulin also seemed promising as a prophylactic or therapeutic agent for the prevention of congenital heart block in at risk pregnancies. However, two recent, large clinical trials did not show benefit.^36,37 Hydroxychloroquine may reduce the rate of CHB but proof of efficacy is lacking. However, based on a lower recurrence risk in case series in women taking hydroxychloroquine,³⁸ some authorities recommend the drug for all pregnancies in women with anti-SSA and SSB antibodies.³⁹

The Management of Pregnancies Complicated by SLE

Table 27-4 summarizes the management of a patient with SLE during pregnancy. Ideally, women with SLE should have preconceptual counseling to discuss both medical and obstetric risks including lupus flares, preeclampsia, FGR, pregnancy loss, and preterm delivery. Patients should also be made aware of the risk of NLE and the clinical implications of the disease. All patients with SLE should have an assessment of their renal function in the form of a serum creatinine level and a 24-hour urine analysis for protein and creatinine clearance. In addition, a hematocrit and platelet count should be determined to exclude hematologic abnormalities associated with SLE. Finally, all patients should be tested for aPL (see below in lupus and APS). A number of studies have demonstrated that active lupus at the time of conception increases the risk of lupus flares, preeclampsia, and fetal loss. Thus, the optimal timing of conception in SLE patients is after a patient has been in remission for 6 months. In addition, nonsteroidal anti-inflammatory drugs (NSAIDs) and cytotoxic agents should be stopped prior to conception (see below in medications).