Systemic lupus erythematosus.

Systemic lupus erythematosus is an autoimmune disease of unresolved cause afflicting predominantly women in their childbearing years. Speculation exists as to the relative roles of viruses, histocompatibility loci such as HLA-DR3, and estrogenmediated decreases in suppressor cell function. Whatever the cause of familial tendency, the disease results in a pathologic hyperactivity of the B-lymphocyte population and a subsequent production of autoantibodies. This hyperactivity may reflect a primary B-cell defect or a deficiency of regulatory cells.

oregnancy A. Gluco- )rd serum py to pre- e- J Clin I Compli- adelphia: prenatal nfants of immune Gynecol Systemic Lupus Erythematosus Craig H. Syrop, MD, and Michael W. Varner, MD University of Iowa Hospitals and Clinics Iowa City, Iowa Systemic lupus erythematosus (SLE) is an autoimmune disease of unresolved cause afflicting predominantly women in their childbearing years. Speculation exists as to the relative roles of viruses, histocompati­bility loci such as HLA-DR3, and estrogen- mediated decreases in suppressor cell function.1,2 Whatever the cause or familial tendency, the disease results in a pathologic hyperactivity of the B-lymphocyte popula­tion and a subsequent production of auto­antibodies. This hyperactivity may reflect a primary B-cell defect or a deficiency of regulatory cells.2,3 Although the pathogen­esis of the disease has undoubtedly remained unchanged over the past 30 years, the diag­nosis, maternal prognosis, and obstetric implications have changed. In 1955, Turner et al.4 remarked that "Pregnancy is not a good omen for women with extensive lupus erythematosus." They expressed the hope that the introduction of adrenocorticotropic hormone (ACTH) and cortisone could pro­long the lives of women affected by SLE. In 1974, Dubois et al.5 demonstrated improved survival rates in association with high doses of corticosteroids. Although there has been a marked decrease in deaths from central ner­vous system (CNS) manifestations and ure­mia, fungal and opportunistic bacterial in­fections have become an important cause of death paralleling the increased utilization of steroids, dialyses, and transplantation ther­apies. Because of increasing survivorship and the natural tendency for SLE to affect wo­men of childbearing potential, an under­standing of SLE in pregnancy is of growing importance to the obstetrician and the in­ternist. To enhance such understanding, it is necessary for us to examine the interrela­tionships of the disease and pregnancy. Diagnosis and Manifestations The diagnosis of SLE is often inhibited by its protean manifestations and its tendency to periods of exacerbation and remission. Specific preliminary criteria for standard­izing the diagnosis of SLE were proposed by Chen et al.6 in 1971. A proposed 1982 revision of criteria appeared in abstract form from Tan et al.7 Both sets of criteria (Table 1) require the identification of four or more manifestations to establish the diagnosis of SLE with a high degree of sensitivity and specificity against rheumatoid arthritis, scleroderma, or nonrheumatic diseases.6,7 SLE manifestations vary, but arthralgias or arthritis present in roughly 90% of pa- Clinical Obstetrics and Gynecology, Vol. 26, No. 3, September 1983 Copyright © 1983 by Harper ir Row, Publishers, Inc., 0009-9201183/0900/0547$01.35 547 548 SYROP AND VARNER TABLE 1. The Diagnosis of Systemic Lupus Erythematosus* 1971 Criteria2 Proposed 1982 Revision3 1. Facial erythema, butterfly rash 1. Malar rash 2. Discoid lupus 2. Discoid lupus 3. Raynaud's phenomenon 3. Photosensitivity 4. Alopecia 4. Oral ulcers 5. Photosensitivity 5. Arthritis 6. Oral or nasopharyngeal ulceration 6. Antibody to DNA or Sm or LE cells 7. Arthritis without deformity or false-positive STS 8. LE cells 7. Proteinuria > 0.5 g/day or 9. Chronic false-positive STS cellular casts 10. Profuse proteinuria > 3.5 g/24 hr 8. Pleuritis and pericarditis 11. Cellular casts 9. Psychosis or seizures 12. Pleuritis and/or pericarditis 10. Hemolytic anemia or leukopenia or 13. Psychosis and/or convulsions thrombocytopenia 14. Hemolytic anemia or leukopenia or thrombocytopenia •For each set of criteria, four or more are required for diagnosis of SLE. dents. Dermatologic manifestations are seen in 70-80%, and renal disease is present in 46% of patients. Hematologic abnormalities are present in more than 50% of cases, but cardiovascular disease afflicts 30-50%.x Of the hematologic aspects, anemia is the most common (57-78%) but is most often the anemia of chronic disease and not auto­immune in nature.3 A positive direct Coombs test is relatively common, and leukopenia is seen in about 50% of patients. There is a greater absolute decrease in circu­lating granulocytes, in part secondary to marrow inhibition. A decrease in the abso­lute number and total proportion of T and B cells is seen, and the decrease in T cells reportedly parallels disease activity.3 Lupus anticoagulant (IgG), sometimes associated with a chronic false-positive serologic test for syphilis (STS), interferes with activation of prothrombin by the prothrombin activa­tor complex and inhibits clotting factor interaction on phospholipid surfaces.9 The IgG fraction that contains lupus antico­agulant has been shown to reduce the release of prostacyclin and thereby plays a role in arterial thrombosis as well as the obstetric problem of reduced uterine blood flow and repeated intrauterine deaths.9 Some mani­festations of SLE may be altered by preg­nancy. Tozman et al.10 noted significantly less pleuritis and pericarditis and a sugges­tion of less CNS disease when comparing pregnant SLE patients with the total SLE population. In addition, Varner et al.11 found the subset of patients in whom SLE developed in pregnancy had thrombocyto­penia and heavy proteinuria more fre­quently. Indeed, can pregnancy cause an alteration in the onset or course, short-term or long-term, of SLE? The first manifestations of SLE may occur during pregnancy or the puerperium. Both the initial diagnosis of SLE in pregnancy and the differentiation of preeclamptic tox­emia (PET) from SLE nephritis can be extremely difficult.11'16 PET is often associ­ated with renal manifestations and may be present in 18-25% of all pregnant SLE patients.12,14,16 Varner et al.11 found that of the 18% of patients in whom symptoms of preeclampsia developed, 70% had their dis­ease first diagnosed in pregnancy. Some feel that a renal biopsy in the third trimester is relatively contraindicated and the approach should be empiric.14 Antinuclear antibody is not normally present in sera of normal or preeclamptic patients and may therefore aid in diagnosis.15 Decreased gestational age (27-30 weeks), severity of hypertension, severity of renal involvement, anemia, thrombocytopenia, coagulation defects, and circulating antibodies may be atypical fea­tures suggesting a coexisting multisystemSYSTEMIC LUPUS ERYTHEMATOSUS cells enia or >anng 1 SLE al. 11 1 SLE Dcyto- ; fre- se an -term occur Both lancy c tox­in be ssoci- ay be SLE lat of ms of r dis- e feel >ter is oach Ddyis al or efore ional sion, ?mia, ;, and 1 tea­' s tern disease.11 Analysis of several studies2,11'17'18 indicates that of all pregnancies compli­cated by SLE, 9-21 % first had manifestations during pregnancy. Fine et al.19 believe that this is not likely to represent an actual increase in the onset of SLE if compared with any 1 -year period during the childbear­ing years. Effect of Pregnancy on SLE There is no unanimous agreement as to the effect of pregnancy on the frequency of exacerbations or worsening of SLE nephri­tis. It is the current general feeling that a pregnancy conceived during a period of SLE remission is likely to remain quiescent. Conversely, it is less likely to undergo ex­acerbation. 18,20 If exacerbations do occur in such cases, they are usually mild.18 Hayslett and Lynn20 found of those in remission for 6 months before conception, the remission persisted in two-thirds; 32% had exacerba­tions, 10% severe but reversible in nature. If conception occurred during active SLE or lupus renal disease, the course was more severe. The manifestations worsened in roughly 50%. The occurrence of exacerbations in differ­ent trimesters and the postpartum period • j 1 ■ -U !•* . 8,10,11,13,14,17,28 varies widely in the literature.................... Garsenstein et al.25 formulated the relative risk of an exacerbation during either half of pregnancy or postpartum by comparing these periods to 32 weeks before conception and 9-40 weeks postpartum. The number of exacerbations or remissions per 100 weeks at risk was then calculated for each period. The relative risks were roughly 3.3 times those of the control period in the first half of preg­nancy, 1.7 times those of the control period in the second half, and almost 7 times those of the control period in the postpartum period.19 Zulman14 notes that SLE may have a profound short-term effect, noting in a liter­ature review12,14,17,25 that 19 of 20 maternal deaths occurred postpartum. The cause of such flares and remissions remains un­29 known, but they are probably dependent upon the degree of imbalance between trig­ger factors and regulatory factors. A post­partum drop in suppressive factors, a de­crease in suppressor-cell activity or ab­normal helper-cell function (or all three) could predispose a patient to altered disease activity.2 In general, as opposed to pre-1970 reports, later studies show little or no in­crease in postpartum SLE activity. Zurier comments that the characteristic increase in postpartum disease activity is apparently decreasing because more mild cases are being diagnosed. Additional factors include greater experience with steroid therapy and a greater general awareness of SLE.29 Bear30 states that pregnancy may lead to serious and even lethal activation of SLE and/or lupus nephritis. However, most re­cent reviews indicate that pregnancy does not adversely affect the long-term course of SLE.14,20,31,32 More specifically, evidence of disease activity in the 6 months preceding conception predisposed to permanent or transient renal function deterioration dur­ing or after pregnancy.20 Yet in 61% of those same pregnancies antedated by the onset of SLE, the clinical course of the disease was not adversely influenced by pregnancy. This is in agreement with the findings of Zul­man. The long-term course of primary renal disease showed no evidence of preg­nancy- related deterioration or lesion pro­gression beyond that expected in the non­pregnant state.32 In general, SLE patients who become pregnant have a more favorable prognosis than those who do not; however, this may well reflect disease severity as it relates to the capacity to become preg­nant. 14,18 Effects of Pregnancy Termination on SLE Having explored the effects of pregnancy on SLE, we shall consider the effects of pregnancy termination upon SLE. Donald­son et al.12 noted that of 12 patients for whom therapeutic abortions were per-550 SYROP AND VARNER formed for acute SLE, only 1 patient improved, 11 of 12 had no change or experienced disease reactivation, and subse­quently 3 patients died. It is believed that for disease management, induced abortions exert little, if any, positive influence on subsequent clinical course,21 and there is no evidence that therapeutic termination alone is "therapeutic" in a flare.14 It will not prevent maternal morbidity.19 By the same token, elective pregnancy termination should be an option for the patient. Zulman et al.14 reported upon 10 patients who opted for elective abortions, including 1 with active disease, most of whom were given supplemental prednisone for 48 hours be­fore returning to their regular dosage. In these cases, no detrimental short-term or long-term effects were noted. Maternal Effects of SLE SLE exerts influences upon the mother. To the disappointment of physicians caring for patients 20 years ago when SLE was con­sidered a "poor omen," it was noted that steroids did not prevent ovulation, although amenorrhea was sometimes seen.18 Fertility was unimpaired when the disease was under 33 • • 17 26 28 control. Several investigators ' ' noted that sterility and fertility were virtually unchanged by maternal SLE. Fraga et al.17 found that the overall fertility rate (preg­nancies per fertile patient) prior to the clinical onset of disease was 3.4 and after diagnosis, 2.1. When compared with control subjects, however, the fertility of SLE pa­tients was not seen to be significantly altered. Involuntary sterility was noted in 24.5% of patients after the diagnosis and in 25% of the control group. Although fetal effects of SLE are discussed later, it should be noted that SLE exerts an adverse effect on perinatal outcome before as well as after its diagnosis. In one study population, there were 23.1% spontaneous abortions prior to diagnosis and 40.5% following di­agnosis. The control group rate was 12.5%. Maternal steroid therapy did not improve the abortion rate. It has been recommended that preclinical disease be suspected in mothers experiencing recurrent losses, par­ticularly in black women.31 Patients with SLE have increased ma­ternal morbidity and mortality. The risks of exacerbation as well as the increased risk of preeclampsia (18-25%)12'14 were previously discussed. Houser et al.23 found that only those patients in whom severe preeclampsia or nephritis developed exhibited changes in renal function status. Despite this finding, it was their conclusion that although the patient with lupus nephropathy has a high- risk pregnancy, the likelihood of an unfa­vorable outcome in patients without ab­normal renal function or active SLE seems small. When under intensive medical and obstetric management, there was no long­term risk to maternal well-being.23,26 Fine et al.19 found that in 114 pregnancies, only 15 patients suffered permanent deterioration.22 In patients in whom SLE developed in pregnancy, there was no evidence that the subsequent course would be different from that expected in the absence of pregnancy. Earlier reports4,28 stressed increased ma­ternal mortality, particularly in the face of renal or cardiac disease. Subsequently, re­ports have shown that although pregnancy in SLE may exacerbate the disease, there is no proof that it leads to increased maternal mortality.13,25 Furthermore, in patients with renal involvement, there was no significant ^ difference in deaths due to lupus glomeru­lonephritis during pregnancy as compared with the nonpregnant state.25 The subse­quent maternal course is regrettably impos­sible to predict on the basis of the course of the disease during a prior pregnancy.18 Fetoplacental Effects of SLE SLE affects the fetoplacental unit before and after its clinical recognition. Fetal outcome is dependent upon disease activity.20,21,23 Most authors would agree that pregnancy wastage is increased secondary to CT T? 11,15,17,19-21,23-25,27 tj r J- • SLE. Before diagnosis, spon­taneous abortion estimates range up to 23- 30%.11,17 Following diagnosis, spontaneousSYSTEMIC LUPUS ERYTHEMATOSUS i in par­ma­ks of >k of 'Usly only psia es in ling, the igh- infa- ab- sems and ong- ne et ly 15 22 on. d in t the irom ncy. ma­ce of re- ancy ;re is ;rnal with cant leru- iared lbse- pos­se of and abortions occur in 5-40% of pregnan­cies. 11'17,19'27 The normal spontaneous abor­tion rale ranging from about 8% to 12.5%.11,17 Fetal loss is further increased by sequelae of prematurity, which occurs in 16.6-37% of pregnancies following the diagnosis of „T t- 8,1 2,15,18,21,23,25 ^ . , T 18 SLE. Estes and Larson re­ported an increased risk of fetal death late in pregnancy, and estimates of stillbirths range from 12 to 30%,12,22 with fetal wastage of 25-33.3%12,21,22,24 Loss is related to maternal disease activ­ity. 11,19 21,23,25 With respect to lupus renal disease, a maternal serum creatinine of 1.5 mg/dl carried a 50% fetal loss rate,20 and a maternal BUN greater than 50 mg/dl was associated with adverse outcome.19 Houser et al.23 found, in patients with previously documented lupus nephropathy, that all ten pregnancies without clinical renal disease or active SLE at conception delivered at term and experienced a benign neonatal course. In eight pregnancies with clinical renal disease or active SLE at conception, only one resulted in a term delivery, and three were premature. Fine et al.19 found that although there was a minimal effect of pregnancy on renal involvement, fetal survival was greatly compromised. If proteinuria and decreased creatinine clearance were present, preg­nancy wastage was increased to 80%. On the other hand, if patients with lupus renal disease were in remission for 6 months before conception, the rate of successful live births was 92% (comparable to normotensive gravidas with diverse types of renal dis­ease). 20 The rate of prematurity did not exceed the expected rate. Aside from pregnancy loss or prematurity, what other effects does SLE or its therapy exert upon the fetus? No study1 1,13,19,20,23,25 found any increased incidence of congenital anomalies among viable pregnancies. Like­wise, no increased anomaly rates have been related to steroid or immunosuppressive therapy. Concordant growth retardation has been reported in infants of mothers with SLE.19 If one combines two recent series,11'19 13 of 62 (21%) infants were small for gesta­tional age, 46 of 62 (74%) were appropriate for gestational age, and 3 of 62 (5%) were large for gestational age. Even after going through the birth pro­cess, the infant may suffer sequelae related to maternal disease. Neonatal SLE is usually considered to be a distinct entity from "familial" lupus. The onset of familial lupus is usually later in life and, more variable in course, and related family mem­bers have hypergammaglobulinemia.34 The two most common manifestations of neo­natal SLE are dermatologic and cardiac in nature. Cutaneous lesions are erythematous, scaly, and atrophic, usually involving the face and upper thorax, with resolution by 12 months of age. Congenital cardiac involve­ment consists of atrioventricular block. It is unusual for an infant to have both mani­festations. 35 Transplacental humoral factors (ANA) disappear within weeks.15 Hemato­logic abnormalities such as anemia, leuko­penia, and thrombocytopenia can exist con­currently or in the absence of the rash or congenital heart block (CHB).35 CHB has a poorer prognosis than the skin manifestations.35 The underlying pathology is endocardial fibroelastosis, fibrosis of the conducting system, and secondary mitral insufficiency and/or small patent ductus arteriosus.35,36 Fifteen to 20% of these infants have associated congenital heart defects.37 The cause of this condition is speculative- perhaps it occurs because of transplacental passage of immune complexes or maternal antigens that stimulate the fetal immune system to form complexes. Injury may occur early in gestation (4-6 weeks) during a critical period of organogenesis, and the nodal anlage then fails to develop normally. Alternatively, the definitive atrioventricular node develops, is injured, and subsequently undergoes fibrosis.38 These infants may re­quire pacemakers,36 but most who survive the neonatal period do relatively well.39 CHB is a permanent but presumably non­progressive insult.25 In a review of 42 infants with neonatal SLE, only 2 of 23 with the cutaneous form were male, whereas 12 of 30 with the CHB form were male.35 This study contrasts with the work of Oleinick,40 whoT SYROP AND VARNER 552 found that proximal siblings of SLE pa­tients had a decreased sex ratio (fewer males to females), but offspring of SLE patients do not have an altered sex ratio.40,41 Just as immunologic factors may dictate the mani­festations of maternal lupus upon the fetus, the placenta is subject to such influences and may also affect fetal survival.42'45 Bresnihan et al.44 found that the incidence of lymphocytotoxic antibodies in sera of SLE patients was significantly lower during pregnancies ending in live births than those ending in spontaneous abortions. Further­more, these lymphocytotoxic antibodies could be absorbed with trophoblastic anti­gens, perhaps explaining increased fetal loss. The data of Lom-Orta et al.45 is in opposition to this, however. Placental histo­logic and immunofluorescent examinations have shown immune complex deposition along the trophoblast membrane and necro­tizing decidua] vasculopathy with fibrinoid necrosis.42,43 This may represent one of many mechanisms resulting in placental ischemia, impaired placental development and function, and subsequent fetal growth retardation or death.42 Management of Pregnancy and SLE Avoidance of adverse fetal outcome and provision of safe maternal-fetal passage through pregnancy, labor, and delivery are the goals of every obstetrician and internist involved in caring for the pregnant SLE patient. Realizing these goals requires an understanding of the interactions of SLE and pregnancy, utilization of laboratory and clinical assessments, and a working knowl­edge of current drug therapy. Routine obstetric laboratory work may aid in diagnosis or indicate disease activ­ity. As discussed before, 50% of SLE pa­tients have a hematologic abnormality,8 usually anemia. Leukopenia, thrombocyto­penia, and a positive Coombs are not un­common. 3 A chronic false-positive STS serves as a major diagnostic criterion and is noted in some pregnant SLE patients.6,7,11 A routine urinalysis and sediment examina­tion are important tools, because renal dis­ease is present in almost 46% of SLE pa­tients. 8 Other generally available determi­nations such as serum creatinine, blood urea nitrogen (BUN), creatinine clearance, and urinary total protein have been related to fetal outcome.19,20 Several,10,14,21,24 but not ajj 11,20,21 autjlors £eej that serial complement levels are predictive of the course of the mother's disease. Baines et al.46 found that in normal preg­nancies the third component of complement (C3) and total hemolytic complement (CH50) showed gradually increasing activity as pregnancy progressed following a signifi­cant depression in the first trimester. They questioned whether this initial decrease in complement resulted from increased anti- gen-antibody reactions associated with the allograft of pregnancy. In a study of com­plement levels in normal and preeclamptic pregnancies, it was found that complement levels in preeclamptics at delivery, 1 day postpartum, and at 6 weeks postpartum approximated the control levels.47 Zurier et al.24 investigated serum comple­ment levels in pregnancies complicated by SLE. Although they did not specifically correlate abnormal values with exacerba­tions, they noted that the mean complement values in all trimesters were slightly in­creased but within normal limits. Tozmanet al.10 reported that in inactive SLE, C3 and CH50 rose during pregnancy, but C3 re­mained in the normal range. When SLE was active, the patients had significantly lower CH50 levels at the onset and throughout pregnancy. Devoe et al.21 examined the rela­tionship of lupus activity and complement levels prospectively in 13 pregnancies and concluded that exacerbations were signaled by decreased or falling levels of complement. Remission states were reportedly associated with normal or near normal C3 or C4 levels. There was considerable variability in anti- nuclear immunofluorescent studies and anti-DNA (Farr) assays, leading to the con elusion that complement is more useful in assessment and prognosis than antinuclearSYSTEMIC LUPUS ERYTHEMATOSUS amina- ial dis- LE pa- etermi- od urea ce, and a ted to >ut not lemem of the il preg- lement (CH50) i/ity as signifi- They ease in d anti- 'ith the )f com- amptic lement 1 day oartum omple- ated by ifically acerba- lement tly in­man et C3 and C3 re- LE was r lower jghout le rela- lement ies and gnaled ement. ociated i levels, n anti­' s and ie con eful in luclear factors. It is interesting to note that a review of the data of Devoe et al.21 shows that although exacerbations were associated with decreased complement levels in six of seven exacerbations, one patient had an exacerbation with normal C3 and C4. Fur­thermore, 6 of 26 times complement levels were obtained from patients in clinical re­mission, C3 and/or C4 were decreased. In total, 8 of 33 listed complement levels did not correlate with clinical disease activity. Varner et al.11 looked at serial complement levels in 12 pregnancies and commented that C3 did not correlate with the course of the disease in 5 of 7 pregnancies with SLE exacerbations. C4 paralleled C3 in 10 of 12 pregnancies. Hayslett et al.20 believe there is a poor correlation between the results of serologic tests and disease activity. Serial anti-DNA titers are of limited usefulness in signaling exacerbations or predicting the course of the disease.11 It would appear that clinical parameters and renal and hemato­logic studies are most useful in assessing the course of SLE in pregnancy as well as the need for medication alteration.11,20 Clinical obstetric precautions in preg­nancies associated with SLE are similar to those required for competent care of other high-risk pregnancies. As in all preg­nancies, early exhaustive efforts should be made to establish firm pregnancy dating. This is obviously important in view of increased prematurity8,12,15,18,21,23,25 and its role in possible intervention decisions. In­trauterine growth retardation occurs in SLE,11'19 and the early diagnosis and appro­priate management of this hinges upon accurate obstetric dating. Serial ultrasono­graphic examinations may be indicated. In attempts to reduce the high rate of still­birth, 12,18,22 fetal surveillance consisting of electronic fetal heart rate monitoring and estriol determinations should be instituted at a time when reliable interpretation and successful intervention, if indicated, is pos­sible- usually at 30-32 weeks' gestation.11 It is important to remember that steroids may affect fetal estriol production, but subnor­mal estrogen excretion in pregnant women receiving less than 75 mg of cortisol daily is indicative of fetal failure to thrive and sub­sequent growth retardation or stillbirth.48 As in other high-risk pregnancies, careful fetal monitoring should! be employed throughout labor. If a maternal antiplatelet antibody is present (as is also the case in ITP) a fetal scalp platelet count should be performed early in labor.49 Delivery for the patient with SLE should be individualized and in accordance with good obstetric practice. Although the cesar­ean section rate for SLE patients has been reported as high as 59%,10 Varner et al.11 noted a rate of 12.9% (the same as the overall institution rate). Delivery for the majority of pregnancies completing the second tri­mester was accomplished spontaneously or with elective low forceps.11 SLE per se is not an indication for cesarean section. Further­more, there is no evidence that cesarean section is of benefit in cases of CHB.39 Reduction of postpartum exacerbations with intrapartum and postpartum cortico­steroid augmentation has been cham­pioned, 8 and its use has been frequently reported.8,10,11,19,21,25 The usual dosage men­tioned is 100-150 mg of intravenous hydro­cortisone every 8 hours, in addition to the maintenance dosage until after the first postpartum day.8,21 Although some11 would question the effectiveness of augmentation in preventing a postpartum flare, most would favor its prophylactic use.8,10,19,25 It is recommended that immunosuppression not be decreased in the postpartum period.8,11 Although dialysis19 and plasmapheresis50 may play a therapeutic role, immunosup­pressive and anti-inflammatory agents form the mainstay of current SLE therapy. It is clear that since the introduction of cortico­steroid therapy, 5-year survival has im­proved from 70% to 93%,19 although cortico­steroids have had no effect on pregnancy outcome.18 Patients who conceive with active disease should receive steroid and, if necessary, immunosuppressive therapy as in the nonpregnant state.14 It is agreed that there is no justification for alteration of needed therapy because of pregnancy.11,19 OfSYROP AND VARNER note, Varner et al.11 found that of nine pregnancies in which disease worsened, four had been advised to decrease medication prior to the exacerbation. Medical management of SLE can poten­tially produce undesired effects in preg­nancy. Although the benefits outweigh the risks,11'14 physician understanding and sub­sequent communication of the relative risks to patients is important. Aspirin is fre­quently used but may induce bleeding in patients with circulating anticoagulants. Furthermore, in a review, high-dose aspirin consumers had longer gestational length, longer labor, an increased incidence of post­maturity, and increased blood loss.51 Cyto­toxic drugs are sometimes used in preg­nancy. Although the risk of a malformed viable infant following the first-trimester administration of a cytotoxic drug must be considered, no reproducible anomaly has been reported in humans. For drugs admin­istered in the second and third trimester, the risk of a fetal malformation is no greater than normal. Importantly, 40% of babies born after cytotoxic drug administration are low in birth weight, and no follow-up studies of their fertility are available.52 Azathioprine is a purine analog, felt to be a "steroid-sparing" agent with less im­munosuppressive and more antiinflamma­tory activity than cyclophosphamide.51,53 Sudden withdrawal of azathioprine can cause a severe refractory exacerbation.19,51 Follow-up karyotype examinations, intel­ligence testing, and neurologic examina­tions in infants of mothers receiving azathi­oprine have revealed no abnormalities.53 The skeletal system of laboratory animals appears to be the primary target of azathio­prine as a teratogen; yet only one case of pre-axial polydactyly has been reported in association with its use.54 It appears that azathioprine use is relatively safe in preg- 53,54 nancy. Beneficial maternal effects of steroids were previously discussed, but with their use should come the realization that there is a relationship between corticosteroids and infection. Infection is now a leading cause of death among SLE patients.5 Although glu­cocorticoids in combination with immuno­suppressive drugs may cause decreased thymic size and a slight decrease in the neonatal lymphocyte count,19 adverse fetal effects are rare. There have been a few reports of reduced fetal growth, but no report shows an increased human incidence of cleft lip or palate.19 The human placenta converts a large fraction of prednisone and predniso­lone to inactive 11-keto forms, but not dexamethasone or betamethasone.55 The re­sultant concentrations of these steroids dic­tate that prednisone or prednisolone, not betamethasone or dexamethasone, be used to treat maternal diseases.56 Prednisone and other steroids are secreted in breast milk. Although a maternal prednisone dosage of less than 30 mg/day is unlikely to create problems,57 breast-feeding while on gluco­corticoids or azathioprine is to be discour­aged. 19,57 Family Planning Traditionally, the postpartum period has been a time to discuss contraception and family planning. Warnings have been given that the increased exertion associated with child care may tax the mother and thereby increase her risk for increased disease activ­ity. 58 To avoid undesired pregnancy, indi­vidualized careful consideration should be given to safe contraception. The fear has been expressed that the use of intrauterine devices, particularly in immunosuppressed patients, poses undue risk of infection.11,58 Use may also result in increased tissue damage and elevated antibody formation.58 Oral contraceptive pills have been associ­ated with thrombosis, hypertension, and possible disease exacerbation. The use of the traditional condom plus a spermicide or diaphragm has seemed prudent. Recently, support for this view was produced.58 In summary, the use of estrogen-containing compounds, even at low doses (30-50 ^ig), should be avoided in women with SLE. When reversible contraception is chosen because of the possible desire for futureSYSTEMIC LUPUS ERYTHEMATOSUS >ugh glu- immuno- decreased »e in the erse fetal w reports ort shows left lip or on verts a predniso- but not 55 The re- roids dic­lone, not , be used isone and ast milk, dosage of to create >n gluco- ; discour- 'riod has tion and sen given ited with i thereby ase activ- icy, indi- hould be fear has rauterine ippressed ction.11,58 ?d tissue nation.58 n associ- ion, and use of the licide or Recently,. ced.58 In ntaining )-50 ng), ith SLE. s chosen )r future pregnancy, mechanical methods should be tried initially. If this is not possible, pure progestogens may be valuable because of antigonadotrophic and antiestrogenic ef- r . 59 fects. Routine gynecologic care should include regular Papanicolaou (Pap) smears. In a recent series,11 3 of 38 pregnancies showed abnormal results in Pap smears, and all 3 women were using immunosuppressive agents. Immunosuppression is known to predispose the human cervical epithelium to dysplastic change. Conclusions SLE in pregnancy is a high-risk condition for both the mother and the fetus, demand­ing intensive and aggressive obstetric and medical management. Through improve­ments in therapeutic regimens, more fre­quent diagnosis of milder cases, and dili­gent obstetric surveillanceand management, the maternal and fetal prognoses appear to be improving. Ideally, patient education should occur at initial diagnosis; then preg­nancies, if desired, should be planned dur­ing periods of remission. Educated about the maternal and fetal implications of the dis­ease, the patient is better prepared both for the intensive observation needed and pos­sible obstetric or neonatal complications that might occur. Successful management of such preg­nancies requires a team approach linking the obstetrician and the internist. It is rec­ommended that patients with SLE in preg­nancy undergo serial clinical evaluations using ARA criteria and careful scrutiny for development of PET. Maternal monitoring should include routine obstetric laboratory work as well as evaluation of hematologic and renal indices. Complement level de­terminations are favored by some, but the Farr assay appears to have little predictive ability. Careful pregnancy dating should be obtained, and fetal surveillance should be instituted as soon as the potential exists for fetal salvage. The route of delivery should be based upon appropriate obstetric indica­tions. The postpartum period is a high-risk period for exacerbation of the condition, and prophylactic steroids may be beneficial in reducing this risk. Finally, it appears that barrier methods of contraception are the safest reversible forms of family planning for patients with SLE. References 1. Lahita RG, Bradlow L, Fishman J, Kunkel HG. Estrogen metabolism in systemic lupus erythematosus-patients and family mem­bers. Arthritis Rheum 1982;25:843. 2. Smolen JS, Steinberg AD. Systemic lupus erythematosus and pregnancy: clinical, im­munological and theoretical aspects. Reprod Immunol 1981,1:283. 3. Budman DR, Steinberg AD. Hematologic aspects of systemic lupus erythematosus. Ann Intern Med 1977;86:220. 4. Turner SJ, Levine L, Rothman A. Lupus erythematosus and pregnancy. Am J Obstet Gynecol 1955;70:102. 5. Dubois EL, Wierzchowiecki M, Cox MB, Weiner JM. Duration and death in systemic lupus erythematosus-an analysis of 249 cases. JAMA 1974;227:1399. 6. Cohen AS, Reynolds WE, Franklin EC, et al. Preliminary criteria for the classification of systemic lupus erythematosus. Bull Rheum Dis 1971:21:643. 7. Tan EM, Cohen AS, Fries J, et al. Criteria for the classification of systemic lupus erythema­tosus. Arthritis Rheum 1982;25:53. 8. McGee CD, Makowski EL. Systemic lupus erythematosus in pregnancy. Am J Obstet Gynecol 1970; 107:1008. 9. Carreras LO, Machin SJ, Deman R, et al. Arterial thrombosis, intrauterine death and "lupus" anticoagulant: detection of im­munoglobulin interfering with prostacyclin formation. Lancet 1981; 1:244. 10. Tozman ECS, Urowitz MB, Gladman DD. Systemic lupus erythematosus and preg­nancy. J Rheumatol 1980;7:624. 11. Varner MW, Meehan RT, Syrop CH, Strott- man MP, Goplerud CP. Pregnancy in pa­tients with systemic lupus erythematosus. Am J Obstet Gynecol (in press). 12. Donoldson LB, DeAlvarez RR. Further ob­servations on lupus erythematosus associated556 SYROP AND VARNER with pregnancy. Am J Obstet Gynecol 1962; 83:1461. 13. Dziubinski E, Winkelmann RK, Wilson R. Systemic lupus erythematosus and preg­nancy. Am J Obstet Gynecol 1962;84:1873. 14. Zulman JI, Talal N, Hoffman GS, Epstein WV. Problems associated with the manage­ment of pregnancies in patients with sys­temic lupus erythematosus. J Rheumatol 1980;7:37. 15. Kitzmiller JL. Autoimmune disorders: ma­ternal, fetal and neonatal risks. Clin Obstet Gynecol 1978;21:385. 16. Abramoksy C. Lupus erythematosus, the placenta, and pregnancy: a natural experi­ment in immunologically mediated repro­ductive failure. Reprod Immunol 1981; 1:309. 17. Fraga A, Mintz G, Orozco J, Orozco JH. Sterility and fertility rates, fetal wastage and maternal morbidity in systemic lupus ery­thematosus. J Rheumatol 1974; 1:293. 18. Estes D, Larson DL. Systemic lupus erythem­atosus and pregnancy. Clin Obstet Gynecol 1965;8:307. 19. Fine LG, Barnett EV, Danovitch GM, et al. Systemic lupus erythematosus in pregnancy. Ann Intern Med 1981;94:667. 20. Hayslett JP, Lynn RI. Effect of pregnancy in patients with lupus nephropathy. Kidney Int 1980; 18:207. 21. Devoe L, Taylor RL. Systemic lupus ery­thematosus in pregnancy. Am J Obstet Gyne­col 1979;135:473. 22. Mund A, Simson J, Rothfield N. Effect of pregnancy on the course of systemic lupus erythematosus. JAMA 1963; 183:109. 23. Houser MT, Fish AJ, Tagatz GE, Williams PP, Michael A. Pregnancy and systemic lupus erythematosus. Am J Obstet Gynecol 1980,138:409. 24. Zurier RG, Argyros T, Urman J, Warren J, Rothfield NF. Systemic lupus erythemato­sus: management during pregnancy. Obstet Gynecol 1978;51:178. 25. Garsenstein M, Poliak V, Kark R. Systemic lupus erythematosus and pregnancy. N Engl J Med 1962:267:165. 26. Cox BJ. Disseminated lupus erythematosus in pregnancy. Obstet Gynecol 1965;26:511. 27. Madsen JR, Anderson GV. Lupus erythema­tosus and pregnancy. Obstet Gynecol 1961; 18:492. 28. Friedman E, Rutherford JW. Pregnancyand lupus erythematosus. Obstet Gynecol 1956; 8:601. 29. Zurier RB. Editorial-SLE and motherhood. J Rheumatol 1980;7:591. 30. Bear R. Pregnancy and lupus nephritis. Obstet Gynecol 1976;47:715. 31. Grigor RR, Shervington PC, Hughes GRV, Hawkins DF. Outcome of pregnancy in systemic lupus erythematosus. Proc R Soc Med 1977;77:99. 32. Strauch BS, Hayslett JP. Kidney disease and pregnancy. Br Med J 1974;4:578. 33. Ricks P. Pregnancy and lupus erythemato­sus. Obstet Gynecol 1962;20:36. 34. Jackson R, Gulliver M. Neonatal lupus ery­thematosus progressing into systemic lupus erythematosus. Br J Dermatol 1979; 101:81. 35. Draznin TH, Esterly NB, Fureu N, DeBotsky H. Neonatal lupus erythematosus. J Am Acad Dermatol 1979; 1:437. 36. McCue C, Mantakas M, Tingelstad JB, Ruddy S. Congenital heart block in new­borns of mothers with connective tissue dis­ease. Circulation 1977;56:82. 37. Stephensen O, Cleland WP, Hallidie-Smith K. Congenital complete heart block and per­sistent ductus arteriosus associated with ma­ternal systemic lupus erythematosus. Br Heart J 1981;46:104. 38. Chameides L, Truex R, Vetter V, Rashkind W, Galioto FM, Noonan JA. Association of maternal systemic lupus erythematosus with congenital complete heart block. N Engl J Med 1977:297:1204. . 39. Scott JS. Systemic lupus erythematosus and allied disorders in pregnancy. Clin Obstet Gynecol 1979;6:461. 40. Oleinick A. Family studies in systemic lupus erythematosus: I, prenatal factors. Arthritis Rheum 1969; 12:10. 41. Hytten FE. Commentary: boys and girls. Br J Obstet Gynaecol 1982:89:97. 42. Abramowsky CR, Vegas ME, Swinehart G, Gyves MT. Decidual vasculopathy of the placenta in lupus erythematosus. N Engl J Med 1980:303:668. 43. Grennan DM, McCormick JN, Wojtacha D, Carty M, Behan W. Immunological studies of the placenta in systemic lupus erythema­tosus. Ann Rheum Dis 1978;37:129. 44. Bresnihan B, Oliver M, Grigor RR, et al.557 SYSTEMIC LUPUS ERYTHEMATOSUS "lancy and ecol 1956; therhood. nephritis. hes GRV, ;nancy in oc R Soc isease and ythemato- lupus ery- mic lupus 9;101:81. DeBotsky us. J Am ;lstad JB, < in new- tissue dis- die-Smith k and per­l with ma- itosus. Br Immunological mechanism for spontaneous abortion in systemic lupus erythematosus. Lancet 1977;2:1205. 45. Lom-Orta H, Diaz-Jouanen E, Alarcon- Segovia D. Lymphocytoxic antibodies dur­ing pregnancy in systemic lupus erythema­tosus. Lancet 1979; 1:1034. 46. Baines MG, Millar KG, Mills P. Studies of complement levels in normal human preg­nancy. Obstet Gynecol 1974;43:806. 47. Kitzmiller JL, Stoneburner L, Yelenoskey PF, Lucas WE. Serum complement in nor­mal pregnancy and pre-eclampsia. Am J Ob­stet Gynecol 1973;117:312. 48. Oakey RE. The interpretation of urinary oestrogen and pregnanediol excretion in pregnant women receiving corticosteroids. J Obstet Gynaecol Br Commonw 1970;77:922. 49. Pitkin RM. Autoimmune disease in preg­nancy. Semin Perinatol 1977; 1:161. 50. Hubbard HC, Portnoy B. Systemic lupus ery­thematosus in pregnancy treated with plas­mapheresis. Br J Dermatol 1979; 101:87. 51. Bulmash JM. Systemic lupus erythematosus and pregnancy. Obstet Gynecol Annu 1978; 153. 52. Nicholson HO. Cytotoxic drugs in preg­nancy. J Obstet Gynaecol Br Commonw 1968; 75:307. 53. Sharon E, Jones J, Diamond H, Kaplan D. Pregnancy and azathioprine in systemic lupus erythematosus. Am J Obstet Gynecol 1974;118:25. 54. Williamson RA, Karp LE. Azathioprine tera­togenicity: review of the literature and case report. Obstet Gynecol 1981 ;58:247. 55. Blanford AT, Murphy BEP. In vitro metab­olism of prednisolone, dexamethasone, beta­methasone and cortisol by the human pla­centa. Am J Obstet Gynecol 1977;127:264. 56. Davies J. The fetal adrenal. In: D Tulchin- sky, KJ Ryan, eds. Maternal-fetal endocri­nology. Philadelphia: WB Saunders, 1980. 57. Romero R, Berkowitz R. The use of anti­asthmatic drugs in pregnancy. In: Niebyl, JR ed. Drug Use in Pregnancy. Philadelphia: Lea and Febiger, 1982. 58. Zurier RB. Systemic lupus erythematosus and pregnancy. Clin Rheumatol Dis 1975; 1: 613. 59. Jungers P, Dougados M, Relissier C, et al. Influence of oral contraceptive therapy on the activity of systemic lupus erythematosus. Arthritis Rheum 1982;25:618. Rashkind )dation of tosus with N Engl J itosus and lin Obstet :mic lupus . Arthritis d girls. Br inehart G, :hy of the N Engl J ojtacha D, cal studies erythema- !9. RR, et al.