Michelle Petri, MD, MPH

Systemic lupus erythematosus (SLE) is a multi-system autoimmune disease that affects women of childbearing age. Although most pregnancies are successful with reported live birth rates of 85%,

 SLE pregnancies remain high-risk. The challenges posed by such pregnancies arise from increased maternal risk of flares,

 as well as increased fetal risks of intrauterine growth restriction (IUGR), pregnancy loss6 and preterm birth (PTB).

 In this review, we consider what is known about the immunology of lupus in pregnancy as well as the maternal and fetal risks and outcomes. 

It is amazing how little is still known about the immunology of SLE in pregnancy. Gonadal and adrenal steroid hormones are dysregulated in SLE pregnancies.

 In normal pregnancies, estrogen and progesterone stimulate Th2 and inhibit Th1 cytokines, while dehydroepiandrosterone sulfate (DHEA-S) plays an immunosuppressive role. Levels of these three hormones are significantly lower in SLE pregnancies, particularly during the second and third trimesters, possibly due to placental compromise.

In pregnant SLE patients, prolactin levels are elevated in the second trimester and peak in the third trimester.

 High prolactin level correlates with pregnancy disease activity

Cytokine imbalances have also been found in pregnancies complicated by SLE, regardless of whether the disease is clinically active or inactive. Increases in interleukin (IL)-6, a Th2 cytokine, are lower than expected and IL-10, B-cell growth factor, is persistently elevated.

In the PROMISSE Study, overactivation of the complement pathway, detected by increased Bb and SC5b-9 (products of this pathway), was present early in 487 SLE pregnancies when compared with 204 healthy pregnancies.

 Increased Bb and SC5b-9 correlated with adverse pregnancy outcomes, which included fetal/neonatal death, preterm delivery or preeclampsia and/or IUGR.

 In one SLE pregnancy, mutations in complement system regulatory proteins were found.

These proteins are highly expressed on trophoblast membranes and prevent excessive complement activation in uncomplicated pregnancies.

Angiogenic imbalance is present in SLE pregnancies complicated by preeclampsia. In a case-control study using stored serum samples, soluble fms-like tyrosine kinase-1 (sFlt1), an anti-angiogenic factor, was significantly higher in pregnant patients with SLE who had preeclampsia.

 Elevation of sFlt1, as early as 12 to 15 weeks, correlated strongly with adverse pregnancy outcomes.

Given the number and breadth of the immunologic, endocrine, and angiogenic changes in pregnancies complicated by SLE, it is no wonder that they are associated with multiple adverse outcomes. However, these changes have not been integrated in a way that would be useful clinically at the individual patient level.

Lupus-affected pregnancies are at risk for renal flares,

 preeclampsia, cesarean delivery and a twenty-fold increase in maternal mortality.


	Multiple case- control studies have yielded conflicting results regarding the likelihood of flares during pregnancy. While some centers found no difference in flare rates,

others have found increased risk of flare (Table 1)

 in pregnant compared to non-pregnant women with SLE. The differences are likely due to patient selection rather than study design or definition of flare. In particular, studies that included African-Americans, lupus nephritis and women with unplanned pregnancies have shown higher flare rates.

Patients with organ-specific lupus activity during the 6 months prior to conception were more likely to have persistence of, or an increase in, the same type of activity during pregnancy.

 Most of the flares were mild to moderate in intensity

 and manageable with small increases in prednisone. Severe flares constituted only 2% to 20% of flares.

 There has been no overall pattern of timing of flares, with some studies showing flares early

Patients with SLE who have lupus nephritis constitute a special population that has been studied separately. Case-control studies that evaluated risk of flares in lupus nephritis concluded that pregnancy does not increase risk for renal flares. However, the study population included mostly white women in complete or partial remission at time of conception (Table 2).

 Compared to patients without lupus nephritis, pregnant women who have nephritis are at increased risk of renal flare.

Lupus activity (manifested as mild, moderate or severe disease flares) in the months prior to conception is a major predictor of pregnancy activity. Studies with more than one-third of their patients with active disease at conception reported flare rates ranging from 45% to 70%,

 while those with predominantly inactive or stable disease at conception reported flare rates lower than 20%.

In patients with lupus nephritis, renal disease activity prior to or at the time of conception, and shorter renal disease remission duration were predictors of pregnancy flare.

 Renal flares occurred in 5% of pregnancies in women who were in complete remission prior to conception compared to 40% of pregnancies in those who had active renal disease.

 Renal activity within 4 months prior to conception was the strongest predictor.

 Kidney biopsy early in pregnancy is safe and informative in patients with suspected renal flares. In a case series of 11 patients with SLE who underwent kidney biopsy during pregnancy, all but one underwent a change in management as a result of findings on renal biopsy and none had biopsy-related complications.


	Women who have SLE-affected pregnancies are three times more likely to suffer from preeclampsia.

 Reported rates of preeclampsia range from 13% to 35% in lupus-affected pregnancies.

Presence of hypertension is the main predictor of preeclampsia.

 Hypertensive patients are 40 times more likely to have pregnancies complicated by preeclampsia.

 Lupus nephritis is also a predictor, especially when combined with prolonged disease duration (14% per month), previous renal flares (with a relative risk of 10), and active nephritis.

Major fetal complications of pregnancy affected by SLE include pregnancy loss, preterm birth, IUGR, and neonatal lupus.  


	Fetal loss rates for pregnancy complicated by lupus have decreased over the past 40 years, from a mean of 40% to 17%.

 In patients with SLE, pregnancy loss occurs mostly during the first trimester. Attribution of early losses to SLE is often not possible, as early pregnancy losses are also common in the general population. Second-trimester losses are mostly associated with secondary antiphospholipid syndrome (APS).

We have found that proteinuria
	(> 500 mg in a 24-hour urine collection or a urine protein-to-creatinine ratio
	> 0.5 g; OR = 2.1), secondary APS (APS in the setting of another autoimmune disorder; OR = 3.4), hypertension (blood pressure > 140/90 mm Hg; OR = 4.4) and thrombocytopenia (platelet count < 150,000; OR = 3.0) at the first prenatal visit in patients with SLE is predictive of pregnancy loss. We coined the acronym PATH (proteinuria, antiphospholipid syndrome, thrombocytopenia and hypertension) to remember these risk factors.

 Increased lupus disease activity in the first and second trimester,

 particularly when combined with low complement levels

or second-
	trimester positive anti-dsDNA antibodies,

 is associated with higher risk of pregnancy loss.

In our center, only presence of lupus anticoagulant at the first pregnancy visit (not a past history of positivity) predicted loss of that pregnancy.

 Thus, if lupus anticoagulant is positive, we recommend low-dose aspirin and low-molecular-weight heparin.

 In patients with thrombosis due to secondary APS, full anticoagulation is indicated and should be done with unfractionated or low-molecular-weight heparin.

A history of pregnancy loss in the first pregnancy is not a predictor of poor outcomes in future pregnancies.

 In an Australian cohort, 90% of women with SLE and a history of first pregnancy loss had a live-born infant, with no recurrence of perinatal death.


	PTB is the most frequent adverse pregnancy outcome in SLE,

 with a reported incidence of approximately 50%.

 PTB can be spontaneous, such as preterm labor or premature rupture of membranes (PROM), or indicated for maternal or fetal complications such as preeclampsia, IUGR or fetal distress. 

Maternal hypertension is an important predictor of PTB.

 In the Hopkins study, mean diastolic blood pressure correlated with preterm delivery.

 Use of low-dose aspirin therapy was not predictive of better outcome, but instead was predictive of increased PTB,

 likely due to a clinician sensing that the patient was already at high risk.

Disease activity, particularly in the second trimester and when combined with hypocomplementemia or anti-dsDNA antibodies, is a risk factor for PTB.

 In the Hopkins cohort, 45% of patients who delivered preterm had a high physician global assessment of disease activity and 70% were on 20 mg or more of prednisone.

 Active, but not quiescent lupus nephritis during pregnancy (defined as achievement of 50% reduction in urine protein/creatinine ratio 4 months prior to conception)  is associated with a higher incidence of PTB (46.3% compared to 25.9%, respectively).

Low estradiol, elevated ferritin and elevated uric acid

 levels at mid-gestation and maternal serum alpha-fetoprotein (AFP) levels

 have been found to be potential markers of subsequent PTB.

In the Hopkins cohort, PROM occurred in 40% of preterm SLE pregnancies and was also common in term SLE pregnancies.

 Occurrence of PROM did not correlate with disease activity, prednisone use, or serologic tests. In the same cohort, preeclampsia or pregnancy-induced hypertension was the reason for PTB in 32%, while spontaneous premature labor occurred in only 11%.

IUGR affects 20% to 30% of pregnancies complicated by SLE.

 Mean birth weight was significantly less for infants born to mothers with SLE than mothers without SLE

. This effect became more marked the longer the gestational period, suggesting that IUGR simply becomes worse as the pregnancy progresses.

Despite the fact that infants of mothers with SLE are more likely to be premature and to have IUGR, they tend to do well.

 Fewer than 2% had an Apgar score below 7.

 Risk of neonatal intensive care unit admission, however, was three times higher in infants born to mothers with SLE.


	Neonatal lupus consists of neonatal lupus rash or congenital heart block due to transplacental transfer of anti-Ro and anti-La antibodies. The rash usually resolves by about 6 months. Cardiac manifestations, including heart block and cardiomyopathy, are associated with 20% mortality and 60% rate of permanent cardiac pacing.

 Anti-Ro-positive women without previously affected pregnancies have a 2% estimated risk of  congenital heart block.

 The rate of congenital heart block when a previous pregnancy resulted in cutaneous or cardiac neonatal lupus is 20%.

 Weekly fetal cardiac monitoring (by echocardiography assessing PR intervals, valvulopathy, and myocardial function) between weeks 16 and 24 is recommended. Fluorinated steroids are given-to prevent cardiomyopathy-when a fetal echocardiograph detects heart block.

 To date, hydroxychloroquine is the only drug that has shown to be beneficial in decreasing incidence of neonatal lupus.

Preconception preparation is key. Prior to pregnancy, it is essential to assess disease activity (especially renal) in a patient, know which treatments she is taking for SLE (to switch to medications allowed in pregnancy),  and be aware of whether she is anti-Ro- and La-positive and has APS. 

Hydroxychloroquine should be maintained or initiated in all pregnancies complicated by SLE. We tell patients that it is desirable in pregnancy-for both mother and fetus. The safety of hydroxychloroquine in pregnancy was first demonstrated by Parke et al.

 Other studies have confirmed the drug’s lack of fetal toxicity and congenital abnormalities.

 Maintaining hydroxychloroquine during pregnancy offers maternal benefits. It decreases disease activity and flares,

 particularly joint and constitutional flares.

 In the Hopkins cohort, among patients with low disease activity who discontinued hydroxychloroquine, 30% developed high disease activity during pregnancy compared to only 3% of those taking hydroxychloroquine.

 Hydroxychloroquine has an anti-thrombotic
	effect

 and should be used in pregnancies complicated by SLE in view of the 10-fold increase in thrombosis risk.

 Hydroxychloroquine also reduces neonatal morbidity. Rates of prematurity and IUGR were lower in 41 pregnancies complicated by SLE exposed to hydroxychloroquine compared to
	77 pregnancies not exposed to hydroxychloroquine.

 Neonates born to mothers who continued hydroxychloroquine had higher gestational age, birth weight, and Apgar scores.

Nonsteroidal anti-inflammatory drugs (NSAIDS) should be avoided in the second and third trimesters. They have been shown to increase risk of premature closure of ductus arteriosus and impair fetal renal function.

Increases in prednisone to doses equal or greater than 10 mg should be avoided. Exposure to glucocorticoids did not increase risk of major anomalies, but did increase risk of cleft palate in pregnancies to women who did not have SLE.

 Exposure to prednisone during the first trimester, in the general population, was associated with increased rates of miscarriage, PTB and low birth weight.

 It was not possible to separate out the effects of the underlying conditions from the effects of prednisone. Moreover, prednisone use in pregnancy was shown to be associated with increased risk of gestational diabetes when used in patients with idiopathic thrombocytopenia

A limited number of steroid-sparing agents are available for treatment and maintenance of disease control during pregnancy. Methotrexate, mycophenolate, and cyclophosphamide must be stopped prior to conception. In particular, we recommend that mycophenolate be stopped 3 months prior to conception and azathioprine substituted (after first ensuring safety by thiopurine methyltransferase testing). During that period, monitoring can be done for occurrence of renal flares on azathioprine.

 A low risk of flares was observed in patients with lupus nephritis who were switched from mycophenolate to azathioprine, compared to those who were initially on azathioprine and maintained on it.

 If renal flare occurs, activity can be controlled by returning to mycophenolate prior to pregnancy. Once a patient’s flares are controlled, she can be switched to azathioprine plus tacrolimus (again waiting to make sure the combination does control the renal lupus, prior to conception).

Azathioprine has not been associated with major congenital anomalies or poor pregnancy outcomes at daily doses not exceeding 2 mg/kg.

 Azathioprine safety is explained by absence of the enzyme inosinate pyrophosphorylase in the immature fetal liver. This enzyme is essential to convert azathioprine to its active metabolite mercaptopurine.

Tacrolimus has been used successfully for maintenance and control of flares during pregnancy in nine patients with lupus nephritis, with no congenital abnormalities reported.

 In a retrospective review of 15 pregnancies complicated by SLE, tacrolimus use was not associated with adverse fetal and maternal outcomes.

Most pregnancies exposed to rituximab (monoclonal antibody against CD20 present on B cells) resulted in uncomplicated live births with no pattern of congenital anomalies.

 Transient B-cell depletion in infants has been reported in three pregnancies exposed to rituximab less than 12 weeks before delivery. The recommendation remains to stop rituximab 12 months prior to conception.

Seventy-seven pregnancies exposed to belimumab (monoclonal antibody against B-cell activating factor) resulted in 38 live births and four congenital anomalies. The latter were chromosomal, urogenital, neural tube, and cardiovascular anomalies.

Three issues regarding SLE in pregnancies were highlighted in recent studies. First, during pregnancy, patients with SLE do not visit their rheumatologists as often as they should. This is of concern and could indicate that patients are ignoring their disease while focusing on pregnancy.

Second, immunosuppressant use decreased in women with SLE during pregnancy. This is possibly due to inappropriate concern about fetal safety of these drugs,

 since azathioprine and tacrolimus therapy are permissible in pregnancy. Controlling lupus activity remains key to a successful pregnancy outcome. 

Third, inadequate counselling and use of contraceptives in patients with SLE who are at risk of pregnancies contributes to an increased number of unplanned pregnancies.

The authors report no potential conflicts of interest with regard to this article.

Petri M. Hopkins Lupus Pregnancy Center: 1987 to 1996. 

Zulman JI, Talal N, Hoffman GS, Epstein W V. Problems associated with the management of pregnancies in patients with systemic lupus erythematosus. 

Petri M, Howard D, Repke J. Frequency of lupus flare in pregnancy. The Hopkins Lupus Pregnancy Center experience. 

Ruiz-irastorza G, Lima F, Alves J, et al. Increased rate of lupus flare during pregnancy and the puerperium: a prospective study of 78 pregnancies. 

Clowse MEB, Jamison M, Myers E, James AH. A national study of the complications of lupus in pregnancy. 

Clark CA, Spitzer KA, Laskin CA. Decrease in pregnancy loss rates in patients with systemic lupus erythematosus over a 40-year period. 

Petri M, Howard D, Repke J, Goldman DW. The Hopkins Lupus Pregnancy Center: 1987-1991 update. 

Doria A, Cutolo M, Ghirardello A, et al. Steroid hormones and disease activity during pregnancy in systemic lupus erythematosus. 

Jara LJ, Pacheco-Reyes H, Medina G, Angeles U, Cruz-Cruz P, Saavedra MA. Prolactin levels are associated with lupus activity, lupus anticoagulant, and poor outcome in pregnancy. 

Doria A, Ghirardello A, Iaccarino L, et al. Pregnancy, cytokines, and disease activity in systemic lupus erythematosus. 

Kim MY, Guerra MM, Kaplowitz E, et al. Complement activation predicts adverse pregnancy outcome in patients with systemic lupus erythematosus and/or antiphospholipid antibodies. 

Salmon JE, Heuser C, Triebwasser M, et al. Mutations in complement regulatory proteins predispose to preeclampsia: A genetic analysis of the PROMISSE cohort.

Holmes CH, Simpson KL, Okada H, et al. Complement regulatory proteins at the feto-maternal interface during human placental development: distribution of CD59 by comparison with membrane cofactor protein (CD46) and decay accelerating factor (CD55). 

Qazi U, Lam C, Karumanchi SA, Petri M. Soluble Fms-like tyrosine kinase associated with preeclampsia in pregnancy in systemic lupus erythematosus. [Erratum appears in J Rheumatol. 2008 Apr;35(4):729]. 

Kim MY, Buyon JP, Guerra MM, et al. Angiogenic factor imbalance early in pregnancy predicts adverse outcomes in patients with lupus and antiphospholipid antibodies: Results of the PROMISSE study. 

Koh JH, Ko HS, Lee J, et al. Pregnancy and patients with preexisting lupus nephritis: 15 years of experience at a single center in Korea. 

Lockshin MD, Reinitz E, Druzin ML, Murrman M, Estes D. Lupus pregnancy. Case-control prospective study demonstrating absence of lupus exacerbation during or after pregnancy. 

Meehan RT, Dorsey JK. Pregnancy among patients with systemic lupus erythematosus receiving immunosuppressive therapy. 

Urowitz MB, Gladman DD, Farewell VT, Stewart J, McDonald J. Lupus and pregnancy studies. 

Georgiou PE, Politi EN, Katsimbri P, Sakka V, Drosos AA. Outcome of lupus pregnancy: a controlled study. 

Borella E, Lojacono A, Gatto M, et al. Predictors of maternal and fetal complications in SLE patients: a prospective study. 

Articles

Maternal and fetal outcomes complicated by SLE require that patient, rheumatologist, and ob/gyn work as a team.  

Michelle Petri, MD, MPH

Articles

Lupus in pregnancy