14.14 Autoimmune rheumatic disorders and vasculiti

14.14 Autoimmune rheumatic disorders and vasculitis in pregnancy 2655

ESSENTIALS Autoimmune diseases affect 5–​7% of people, are more common in women of childbearing age, and are frequently encountered in preg- nancy. They may remit or improve during pregnancy, but can flare or present in pregnancy or post partum. Systemic lupus erythematosus The mother—​pregnancy probably exacerbates systemic lupus erythematosus and increases the likelihood of a flare, which can be difficult to diagnose since many clinical features also occur in normal pregnancy. Differentiation of active renal lupus from pre-​eclampsia is notoriously difficult: renal flares are more common if disease is active within six months of conception, in particular in women with hypertension, heavy proteinuria, or high baseline serum creatinine. There is an increased risk of maternal thrombosis and premature atherosclerosis. The fetus—​systemic lupus erythematosus is associated with in- creased risks of adverse pregnancy outcome including fetal death and intrauterine growth restriction. Most fetal losses occur in associ- ation with secondary antiphospholipid syndrome or active disease, particularly renal. For women with systemic lupus erythematosus in remission and without hypertension, renal involvement, or the antiphospholipid syndrome, the risk of problems in pregnancy is similar to that of the general population. Management—​flares of systemic lupus erythematosus must be ac- tively managed, pre-​pregnancy counselling should be encouraged with treatment depending on both organ involvement and severity. Mild cases can be managed with analgesics alone (paracetamol); rash and arthritis will usually respond to non​steroidal anti-​inflammatory drugs, low-​dose prednisolone and/​or hydroxycholorquine; more se- vere disease may require introduction of a disease-​modifying agent (e.g. azathioprine or higher steroid dose). Steroids remain first-​line treatment for severe lupus flares in pregnancy (and treatment of other autoimmune conditions). The baby—​neonatal lupus syndromes are caused by transplacental passage of autoantibodies directed against cytoplasmic ribonucleoproteins Ro and La. Cutaneous neonatal lupus is the most common manifestation (5%) and congenital heart block the most serious (20% mortality). Antiphospholipid syndrome Clinical features—​antiphospholipid antibodies include anticardiolipin antibodies (IgG and/​or IgM), lupus anticoagulant, and anti-​ β2-​glycoprotein-​I antibody. Antiphospholipid syndrome is the combination of any of these with one or both of the following clinical features:  (1) thrombosis—​arterial, venous, or small vessel;
(2) specific pregnancy morbidity. Women with isolated but per- sistent antiphospholipid antibodies without clinical features of antiphospholipid syndrome have obstetric outcomes similar to the general population. Management—​aim is to improve pregnancy outcome and pre- vent maternal thrombosis. This requires low-​dose aspirin from early pregnancy for prevention of pre-​eclampsia ± low-​molecular-​ weight heparin. Close fetal and maternal surveillance are required. Rheumatoid arthritis Rheumatoid arthritis improves with pregnancy in some women. Methotrexate is teratogenic, but with the advent of treatment with biologics, more women with rheumatoid arthritis are now attempting pregnancy. Birth outcomes of women on anti-​TNFα agents are re- assuring and there is no link with teratogenicity, but concern about immunosuppression of the fetus through transplacental transfer means treatment with some agents should be discontinued in later pregnancy. Overall the risk of adverse obstetric outcomes remains minimal and similar to the normal population. Post-​partum flares of rheumatoid arthritis are common. Introduction Autoimmune diseases affect 5–​7% of the population and are more common in women of childbearing age. Many women with auto- immune rheumatic diseases have been advised against pregnancy in the past, but this is no longer appropriate with a new generation of pregnancy-​friendly disease-​modifying antirheumatic drugs and biological agents that afford excellent disease control without com- promising fertility. Nevertheless, many women with autoimmune rheumatic diseases are older and have more comorbidities (i.e. hypertension, obesity, diabetes, cardiovascular disease, and so on) when they do attempt pregnancy. 14.14 Autoimmune rheumatic disorders
and vasculitis in pregnancy May Ching Soh and Catherine Nelson-​Piercy

Section 14  Medical disorders in pregnancy 2656 The importance of planned pregnancies with good preconception advice from clinicians knowledgeable in both the disease process and its effects on pregnancy (and lactation), and vice versa, cannot be overemphasized. Pregnancy is associated with suppressed cell-​mediated immunity (Th1) and enhanced humoral immunity (Th2), but these changes revert post-​partum accompanied by rapid reductions of oestrogen, progesterone, and cortisol levels. The post-​partum period is there- fore a time of susceptibility to autoimmune disorders; and women who already have an autoimmune disorder may suffer disease ex- acerbation following pregnancy. Conversely, some autoimmune diseases may remit or improve during pregnancy, but this is not a universal rule. Flares in pregnancy are often accompanied by adverse obstetric outcomes for both mother and fetus. This chapter considers the relationship between pregnancy and systemic lupus erythematosus, antiphospholipid syndrome, rheumatoid arthritis, vasculitides, and scleroderma, and how preg- nancy affects treatment of these conditions. Systemic lupus erythematosus Systemic lupus erythematosus is much more common in women than men (ratio 9:1), with peak onset during the childbearing years. A  recent extensive review of published epidemiological studies demonstrated that the prevalence ranges from 0.07/​1000 in Caucasian Americans to 1.59/​1000 in a British Afro-​Caribbean population. Effect of pregnancy on systemic lupus erythematosus Flares of systemic lupus erythematosus may be difficult to diagnose during pregnancy since many features such as hair loss, oedema, fa- cial erythema, fatigue, musculoskeletal pain, anaemia, and raised erythrocyte sedimentation rate (ESR) also occur in normal preg- nancy. Several disease activity scores are being validated for use in pregnancy. Active systemic lupus erythematosus in the four to six months preceding conception increases the likelihood of a flare during pregnancy. Several case–​control studies have addressed this issue, but differ in patient ethnicity, criteria for flare and systemic lupus erythematosus activity scales employed. The type of flare usually fol- lows previous disease pattern. Hydroxychloroquine has been linked with better disease control and improved obstetric outcomes, and therefore should be continued in pregnancy. Conversely, steroids do not prevent flares, hence it is not appropriate to prescribe or increase the dose of steroids prophylactically during pregnancy or during the post-​partum period. Ideally, pregnancy should be planned when systemic lupus erythematosus is in remission, while on drugs that are suitable to be continued in pregnancy. Renal flares are more common if disease is active within six months of conception. There is a risk of deterioration of renal function in pregnancy, particularly if the patients are hyperten- sive, have pre-​existing heavy proteinuria, or a high baseline serum creatinine. A recent meta-​analysis reported that the in- cidence of renal lupus flares during pregnancy was 11–​69% and renal impairment occurred in 3–​27%, which was irreversible in up to 10%. Tacrolimus is being increasingly used as a treatment for lupus nephritis. De novo presentations of lupus nephritis during pregnancy are not uncommon, particularly in those who are not treated. Women with systemic lupus erythematosus and secondary antiphospholipid syndrome have an increased risk of maternal thrombosis, especially in the puerperium. There should be a low threshold for empiric treatment with low-​molecular weight heparin pending appropriate diagnostic imaging. Young women with systemic lupus erythematosus are also at risk of premature atherosclerosis, even in the relative paucity of cardio- vascular risk factors. It is possible the pregnancy and its associated complications accelerate this process. Hence, myocardial infarction should be considered as a differential if she presents with chest pain or shortness of breath. Pulmonary hypertension present in up to 14% of patients. It carries a significant risk of maternal death. Idiopathic pulmonary hypertension is associated with an up to 25% risk of maternal mortality, and this risk is even higher in women with underlying connective tissue disorders. Women who have pulmonary hyper- tension should be advised against pregnancy and when pregnant, offered the option of termination on the basis of life-​threatening maternal disease. Nevertheless, the most common cause of ma- ternal death in women with systemic lupus erythematosus is infection. Effect of systemic lupus erythematosus on pregnancy Women with systemic lupus erythematosus remain fertile, except during severe flares or after exposure to prolonged high doses of cyclophosphamide, which results in premature ovarian failure. However, systemic lupus erythematosus is associated with in- creased risk of early pregnancy losses and the later adverse preg- nancy outcome as a result of placental insufficiency, and manifest as pre-​eclampsia, fetal growth restriction, or small for gestational age infants, placental abruption, and stillbirth—​collectively known as maternal-​placental syndrome. These complications often lead to preterm deliveries. The main factors influencing pregnancy out- comes in women with systemic lupus erythematosus are disease activity (especially at time of conception), hypertension, renal in- volvement, secondary antiphospholipid syndrome, and anti-​Ro/​La antibodies (see next). Pregnancy outcomes in women with systemic lupus erythematosus have greatly improved in recent years due to continuation of hydroxychloroquine in pregnancy, fetal surveillance in women who are Ro and La positive, improved disease control even in the non-​ pregnant cohort, likely milder disease course as diagnostic latency is shortened and clinicians are now more comfortable with the man- agement of systemic lupus erythematosus in pregnancy thereby re- ducing the rate of iatrogenic preterm delivery. Presence of renal disease is closely associated with pregnancy outcomes in women with systemic lupus erythematosus. In a meta-​ analysis active nephritis at conception was associated with 25–​50% rate of fetal loss as compared to 8–​12% if the disease was inactive. One case–​control study showed that 28% of patients with class III or IV lupus nephritis developed pre-​eclampsia, 35% had a pre- term delivery, and a significantly lower birth weight compared to the women with systemic lupus erythematosus without nephritis. A higher baseline creatinine is also associated with poor pregnancy outcomes. Women with systemic lupus erythematosus in clinical remission but with persistently low complement levels or positive anti-​double

14.14  Autoimmune rheumatic disorders and vasculitis in pregnancy 2657 stranded DNA (anti ds-​DNA) in the second trimester have worse pregnancy outcomes than seen in women with normal complement or negative anti ds-​DNA. However, it is women with both serological and clinically active disease who have the highest rates of pregnancy loss and preterm delivery. Positive anti ds-​DNA was more closely associated with high clinical systemic lupus erythematosus activity in pregnancy, but low complement alone was not. Management of systemic lupus erythematosus in pregnancy Preconception counselling When possible, management should begin with preconception counselling. Baseline renal function, blood pressure, and the anti­ phospholipid and anti-​Ro/​La antibody status allow prediction of the risks to the woman and her baby (see later). Appropriate baseline investigations are listed in Box 14.14.1. A decision should be made as to whether to start aspirin and/​or low-​molecular weight heparin if the woman is at risk of thromboses. The outlook is better if conception occurs during remission, and women should be advised to avoid pregnancy for at least 6 months post-​flare, especially if there is renal involvement. Her medications should be reviewed and changed to pregnancy-​friendly drugs (if possible). It is important to ensure clinical stability on this regime for at least six months preconception. Maternal surveillance Pregnancy care is best undertaken in multidisciplinary, combined clinics where physicians and obstetricians can monitor disease ac- tivity, fetal growth, and uterine and umbilical artery Doppler blood flow. Women will need to be regularly reviewed for any signs of pre-​ eclampsia or a flare. Frequency of visits depends on gestation and maternal disease activity. As a rule of thumb, the visits increase in frequency towards and during the third trimester—​when superim- posed pre-​eclampsia is commonest. Diagnosis of flare Differentiation of active renal lupus from pre-​eclampsia is notoriously difficult, and the two conditions may co-​exist. Table 14.14.1 lists fea- tures to help to distinguish them. The gold standard is a renal biopsy to differentiate a renal lupus flare from pre-​eclampsia. Renal biopsies can be performed in pregnancy, but there is an increased risk of bleeding and haematoma formation due to increased vascular perfusion to kid- neys during pregnancy. Renal biopsies are performed rarely in preg- nancy and only before 24 weeks’ gestation. Other useful markers for a flare of lupus nephritis include a raised ds-​DNA and active urinary sediment or the presence of casts in the urine. Management of flare Disease flares must be actively managed. Delivery is seldom the sole management of choice as post-​partum flares are common and severe. Treatment will depend on the organ involvement and severity (see the later section in this chapter, ‘The use of antirheumatic drugs, immuno- suppressive agents, and biologics in pregnancy’ for a summary of the therapeutic options available). Most drugs can be used in pregnancy. Teratogenic agents such as cyclophosphamide should be avoided in the first 12 weeks during which organogenesis occurs. There are long-​term safety data on the use of cyclophosphamide later in preg- nancy with a cohort of children exposed in utero showing no signs of neurodevelopmental delay or other adverse effects when followed up to adolescence. Due to the prolonged action of rituximab, it is best avoided for six months prior to delivery. However, its use in pregnancy (for other indications, e.g. idiopathic thrombocytopenic purpura or as part of a chemotherapy regimen) has not been associated with adverse neonatal effects, and the exposed infants are able to mount an adequate response to vaccination despite the reduction of lymphoid B cells in the reticuloendothelial system seen in the offspring of exposed primates. Corticosteroids are the most widely prescribed drug for a flare in pregnancy as they have a well-​established safety profile. Non-​ fluorinated steroids (e.g. prednisolone), are metabolized by the placenta, and the fetus also inactivates steroids by way of sul- phate conjugation, hence very little (<10%) active drug reaches the fetus. Conversely, fluorinated steroids like dexamethasone and betamethasone cross the placenta more readily and are often used for antenatal fetal lung maturation in women at risk of preterm delivery. A recent population-​based study of over 800 000 births over a 12-​ year period in Denmark did not show an increased risk of orofacial clefts with use of oral corticosteroids. However, there is increasing evidence that repeated courses of fluorinated steroids, particularly Box 14.14.1  Baseline investigations to assess risk of pregnancy in patients with systemic lupus erythematosus • Blood pressure • Urinalysis and quantification of proteinuria (albumin/​protein cre- atinine ratio and/​or 24 h urinary protein) • Full blood count, ESR • Urea, creatinine, and electrolytes • Liver function tests • Uric acid • Serology—​antiphospholipid antibodies (lupus anticoagulant and anticardiolipin antibodies, anti-​β2glycoprotein I antibodies), ENA (esp. Ro and La antibodies), ANA, ds-​DNA • Complements: C3 and C4 Table 14.14.1  Features to help distinguish lupus nephritis flare and pre-​eclampsia Pre-​eclampsia Flare of SLE Hypertension Yes Often present Other organs No Malar rash, photosensitive rash, or arthritis Seizures If eclampsia Only if there is neurological involvement Proteinuria ++ ≥0.3 g/​ day or
PCR ≥30 ++ (in lupus nephritis) Urinary casts Absent Present (if lupus nephritis) RBC in urine Absent Present (if nephritic) Urate Elevated Not elevated unless CKD Albumin Low Very low if nephrotic syndrome LFT May be deranged Rarely deranged in a flare of SLE C3 and C4 Unchanged from baseline in early pregnancy Low compared to baseline in early pregnancy Anti-​ds-​DNA Unchanged Elevated CKD, chronic kidney disease; LFT, liver function test; RBC, red blood count; SLE, systemic lupus erythematosus.

Section 14  Medical disorders in pregnancy 2658 dexamethasone, can adversely affect the child’s later neuropsycho- logical development. Corticosteroid usage in pregnancy does in- crease the maternal risk of gestational diabetes, hypertension, infection, risk of preterm rupture of membranes and osteoporosis. Women on long-​term maintenance steroids (>7.5 mg/​day for

2 weeks) require parenteral steroids to cover the stress of labour and delivery. Prednisolone is safe in breastfeeding mothers since less than 10% of active drug is secreted into breast milk. Notwithstanding the above risks, steroids remain first-​line treatment for severe lupus flares in pregnancy (and treatment of other autoimmune condi- tions), as the benefits of rapid disease control outweigh the risks. Disease-​modifying antirheumatic drugs are discussed later in this chapter, in the section ‘The use of antirheumatic drugs, immunosup- pressive agents, and biologics in pregnancy’. Neonatal lupus syndromes About 30% of patients with systemic lupus erythematosus are anti-​ Ro positive. In such women the risk of transient cutaneous lupus is about 5% and the risk of congenital heart block about 2%, with the two conditions rarely coexisting. Anti-​Ro antibodies are present in 90 to 100% of mothers of affected offspring, and 68 to 91% have anti-​La antibodies. Maternal titres of anti-​Ro antibodies as high as 50 U/​ml or more are more likely to be associated with congenital heart block. There is no relationship between anti-​La titre and neonatal lupus. The risk of neonatal lupus is increased if a previous child has been affected, at 15–​25% if one and 50% if two previous children are affected. Not all anti-​Ro/​La positive mothers of neonates with congenital heart block have systemic lupus erythematosus or Sjögren’s syndrome, some are asymptomatic, but 48% of these de- veloped symptoms of connective tissue disease in a median 3.7 years follow-​up in one study. In mothers who do have systemic lupus erythematosus there is no correlation between the severity of ma- ternal disease and the incidence of neonatal lupus. Cutaneous neonatal lupus usually manifests in the first two weeks of life. The infant develops typical annular skin lesions similar to those of adult subacute cutaneous lupus, usually of the face and scalp, which appear after sun or UV light exposure. The rash disappears spontaneously within six months. Residual hypopigmentation or tel- angiectasia may persist for up to two years, but scarring is unusual. No specific treatment is required, except topical steroids in severe cases. Congenital heart block usually appears in utero, around 18–​ 20 weeks and is associated with a structurally normal heart. The mechanism of damage appears to involve binding of the anti-​Ro/​La antibodies to antigens on the fetal cardiocytes, inducing an inflam- matory process which leads to tissue damage and fibrosis of the con- ducting system. In women known to be anti-​Ro/​La positive, the fetal heart rate should be monitored at each visit (from 16 weeks onwards), and fetal cardiology scans offered at approximately 18–​20 weeks’ ges- tation and again at 28 weeks. Complete heart block causes brady- cardia which can be detected on auscultation, but lesser degrees of heart block require Doppler echocardiography. Overall mortality is around 20%, with deaths usually occurring in utero (after developing hydrops, pleural and pericardial effusions) or the neonatal period, but can occur up to three years of age. Most infants who survive the neonatal period do well, although two-​thirds require pacemakers. There is no treatment that reverses complete heart block. In the past, intravenous immunoglobulins, fluorinated glucocorticosteroids, plasmapheresis, salbutamol, and digoxin have all been tried and found to be ineffective. Studies on congenital heart block have been difficult to conduct due to the rarity of the condition. However, a large retrospective study did find an association between maternal hydroxychloroquine use, and a reduction in heart block in the offspring. Its follow-​on study focusing on subsequent pregnancies of women whose off- spring were affected found that there was a definite reduction in the incidence of heart block in the offspring of the women who were re- ceiving hydroxychloroquine. Hence, hydroxychloroquine should be used in women who are planning pregnancy (or in early pregnancy) and are Ro or La antibody positive, even if they are asymptomatic. Antiphospholipid syndrome and antiphospholipid antibodies Antiphospholipid antibodies include anticardiolipin antibodies (IgG and/​or IgM), lupus anticoagulant, and anti-​β2-​glycoprotein-​I antibody. The combination of any of these with one or more of the characteristic clinical features of thrombosis, recurrent pregnancy loss, or adverse pregnancy outcome (as detailed in Table 14.14.2) is known as the antiphospholipid syndrome. antiphospholipid Table 14.14.2  Revised (2006) classification criteria for the antiphospholipid syndrome Revised classification criteria (Sydney criteria) for antiphospholipid syndrome. Vascular thrombosis: ≥1 clinical episode of arterial, venous, or small vessel thrombosis. Thrombosis must be objectively confirmed. For histopathologic confirmation, thrombosis must be present without inflammation of the vessel wall. Pregnancy morbidity: a. ≥1 unexplained death of a morphologically normal fetus ≥10 weeks of gestation. b. ≥1 premature delivery of a morphologically normal fetus <34 weeks’ gestation because of: (i) severe pre-​eclampsia or eclampsia defined according to standard definitions. (ii) recognized features of placental insufficiency. c. ≥3 unexplained consecutive miscarriages <10 weeks’ gestation, with maternal and paternal factors (anatomic, hormonal, or chromosomal abnormalities) excluded. Laboratory criteria: The presence of antiphospholipid antibodies (antiphospholipid antibodies), on two or more occasions at least 12 weeks apart and no more than five years prior to clinical manifestations, as demonstrated by ≥1 of the following. a. Presence of lupus anticoagulant in plasma b. Medium to high-​titre anticardiolipin antibodies (>40 GPL or MPL, or >99th percentile) of IgG or IgM isoforms c. anti-​ß2 glycoprotein-​I antibody (anti-​ ß2GP I) of IgG or IgM present in plasma. Adapted from Miyakis S, et al. (2006). International consensus statement on an update of the classification criteria for definite antiphopholipid syndrome (APS). J Thromb Haemost, 4, 295–​306.

14.14  Autoimmune rheumatic disorders and vasculitis in pregnancy 2659 antibodies are common and are not always associated with a clinical picture of pregnancy losses, obstetric complications, or thrombosis. The implications of antiphospholipid antibodies
and antiphospholipid syndrome in pregnancy Antiphospholipid syndrome is the most commonly acquired thrombophilia. Antiphospholipid antibodies are prevalent in the general population. In a cross-​sectional study of healthy blood donors, lupus anticoagulant was present in 8% and anticardiolipin antibodies in 10%. However, these antibodies were transient and persistence occurred in less than 2%. Antiphospholipid antibodies can be generated after exposure to certain medications, after infec- tion (e.g. with HIV, HTLV-​1, hepatitis A, B, or C, CMV, EBV, VZV, TB, or poststreptococcal rheumatic fever, syphilis, Klebsiella, mal- aria), following a miscarriage, or in association with certain malig- nancies. However, these antiphospholipid antibodies—​particularly those generated after infection or following a pregnancy loss—​are transient and are unlikely to be of clinical significance, hence the importance of retesting antiphospholipid antibodies after more than 12 weeks. The presence of antiphospholipid antibodies does not equate to clinical disease. Only 8% of those with primary antiphospholipid syndrome devel- oped systemic lupus erythematosus or 5% a ‘lupus-​like’ disease over a nine-​year follow-​up interval. The association with thrombosis is more robust in lupus anticoagulant than with anticardiolipin anti- bodies (OR 11.0 vs. 1.6). For those with antiphospholipid antibodies only, several case–​control studies, and a prospective multicentre study have shown the incidence rate of thrombosis is 1.86% per 100 patients per year. In the obstetric setting, women with antiphospholipid syndrome are at increased risk of recurrent early fetal loss, severe early onset pre-​eclampsia, placental abruption, intrauterine fetal death, or fetal growth restriction without hypertension. Recurrent pregnancy loss in women with antiphospholipid syndrome is typically in the second trimester. Fetal death is typically preceded by fetal growth restric- tion and superimposed pre-​eclampsia. Quantifying the risk is diffi- cult. In a meta-​analysis, lupus anticoagulant in particular was more strongly and consistently associated with fetal loss (OR 7.79; 95% CI:2.3–​26.45). The pathogenesis of fetal loss in these patients is not fully understood, although a variety of mechanisms have been sug- gested and there appear to be both thrombotic and inflammatory components. Women with obstetric antiphospholipid syndrome have a much higher risk of severe early onset (<34 weeks’ gestation) pre-​ eclampsia. One study showed 20.9% of these women were found to have anticardiolipin antibodies compared with 7.5% of controls. A systematic review has shown pre-​eclampsia to be more common in women with antiphospholipid antibodies with an OR:  2.73 (95%CI:1.65–​4.51) for anticardiolipin antibodies, and OR 1.45 (95% CI:0.70–​4.61) for lupus anticoagulant. All women with severe early onset pre-​eclampsia should be screened for antiphospholipid syn- drome. However, the presence of antiphospholipid antibodies does not preclude successful pregnancy. Previous poor obstetric history remains the most important predictor of pregnancy outcome in these women. Studies on obstetric outcome in women known to have antiphospholipid syndrome show differing rates of compli- cations depending on their presentation:  those found to have antiphospholipid syndrome as a result of recurrent early miscarriage have lower rates of complications than those with late fetal losses, thrombosis, or other systemic manifestations. By contrast, women with isolated but persistent antiphospholipid antibodies without clinical features of antiphospholipid syndrome have obstetric out- comes similar to the general population. Management of antiphospholipid syndrome
and antiphospholipid antibodies The management of pregnancy in women with antiphospholipid syndrome, in particular the use of anticoagulation, is controversial and will be further discussed. Maternal and fetal surveillance Pregnancy complicated by antiphospholipid syndrome requires ex- pert care and a team approach by obstetricians, obstetric physicians, and rheumatologists/​haematologists. Close monitoring of both mother and fetus is essential. Pre-​eclampsia is the most common complication and should be screened for at each visit with BP check and urine dipstick for proteinuria. Uterine artery Doppler wave- forms are assessed between 20 and 24 weeks’ gestation, and those pregnancies with evidence of compromised blood flow should be monitored more closely with 4-​weekly growth scans because of the high risk of fetal growth restriction. If there is evidence of fetal growth restriction, then use of umbilical artery Dopplers can be fur- ther quantified. Absent or reversed end-​diastolic flow is indicative of poor fetal outcome and delivery should be expedited. By contrast to the patient with antiphospholipid syndrome, women with antiphospholipid antibodies alone do not require add- itional surveillance in pregnancy. Women with isolated but per- sistent antiphospholipid antibodies have normally grown babies and do not have the same complications of pre-​eclampsia, placental abruption, and stillbirth as those with obstetric antiphospholipid syndrome (Table 14.14.3). They can therefore be treated as ‘normal’, except for the use of low-​dose antenatal aspirin 75 mg daily. Pharmacologic options and low-​molecular weight
heparins for antiphospholipid antibodies and antiphospholipid syndrome The main pharmacologic options are low-​dose aspirin (75–​100 mg/​ day) or low-​molecular-​weight heparin, neither, or both. Low-​ molecular-​weight heparin does not cross the placenta. There are few randomized controlled trials in the area of antiphospholipid syn- drome/​antiphospholipid antibodies to guide management of this condition in pregnancy. Many studies are small and of heteroge- neous cohorts that included women who did not fulfil the classifica- tion criteria for antiphospholipid syndrome. Use of aspirin for risk reduction of pre-​eclampsia is almost uni- versally accepted for all women with either antiphospholipid anti- bodies or antiphospholipid syndrome. Two studies have shown that live birth rates are 70% with low-​dose aspirin alone, although a single placebo-​controlled randomized controlled trial showed no difference in live birth rates between the group on aspirin versus those on the placebo. The use of low-​molecular-​weight heparin for pregnant women with antiphospholipid syndrome and previous thrombosis, late preg- nancy losses, or adverse pregnancy outcomes due to placental insuf- ficiency is undisputed, though few recent studies have demonstrated

Section 14  Medical disorders in pregnancy 2660 any improvement in obstetric outcomes. However, in women with only recurrent, early (≤10 weeks) miscarriage, the evidence is less robust, and therefore low-​molecular-​weight heparin should not be routinely started. Laskin’s study looking at low-​molecular-​weight heparin for recurrent pregnancy loss in women (n  =  88) with thrombophilias (47.7% of them had antiphospholipid antibodies) found that live birth rates were 79.1% on aspirin alone and 77.8% on low-​molecular-​weight heparin + aspirin (p = 071). Farquharson’s study (n = 98) also showed no benefit with low-​molecular-​weight heparin use (i.e. 72% live births on low-​dose aspirin vs. 78% with additional low-​molecular-​weight heparin, p = 0.89). In women with antiphospholipid antibodies who are under- going assisted reproduction, low-​molecular-​weight heparin does not improve implantation rates. However, prophylactic low-​ molecular-​weight heparin should be considered due to the high risk of thrombosis in this cohort, especially if they have additional risk factors (i.e. older than 35 years, obese, smokers, or having ovarian hyperstimulation syndrome). low-​molecular-​weight heparin should be discontinued at 12 weeks’ gestation if there are no additional risk factors for thrombosis. For women with antiphospholipid antibodies alone, there is very little evidence to support the use of low-​molecular-​weight heparin antenatally. Most would elect to use low-​dose aspirin as the risk of toxicity is low and there is a modest beneficial effect for the preven- tion of pre-​eclampsia. Aspirin can be discontinued at delivery. Its use does not increase the risk of post-​partum haemorrhage nor is it a contraindication to regional anaesthesia. Women with previous thrombosis on long-​term warfarin should be swapped to low-​molecular-​weight heparin before six weeks’ ges- tation to avoid warfarin embryopathy. Other agents Treatment with high-​dose corticosteroids (in the absence of ac- tive lupus) to suppress antiphospholipid antibodies has previously been recommended (in combination with aspirin), however, high doses of prednisolone (up to 40 mg) caused considerable maternal morbidity and subsequent studies have failed to demonstrate better fetal outcome compared to aspirin and heparin. Steroids may even worsen outcome because of an increased risk of preterm labour. Lower doses have been used successfully in women with recurrent miscarriage despite the use of aspirin and low-​molecular-​weight heparin. Immunosuppression with intravenous immunoglobulin (IVIg) has been used, particularly for recurrent miscarriage. However, more recent studies have shown no benefit either compared to, or in addition to, aspirin and heparin. In the comparison study in re- current miscarriage there was a higher rate of live births (84%) with aspirin and heparin than with immunoglobulin (57%): the study in which IVIg was added showed no significant difference in live birth rate. Other immunosuppressive agents have also been used, including azathioprine and plasmapheresis, but the numbers treated do not allow firm conclusions regarding efficacy. Post-​partum Women with antiphospholipid syndrome and previous thrombo- embolism are at particularly high risk in the puerperium. If the woman has been on aspirin prior to pregnancy, this should be con- tinued. Warfarin can be restarted in the women who were previously taking it once the risk of bleeding is minimal (usually after 10 days) (Table 14.14.4). Aspirin, low-​molecular-​weight heparin, and warfarin are safe for women who are breastfeeding. The combined oral contraceptive pill should be avoided due to the increased risk of thrombosis. The progesterone-​only pill is safe for women with antiphospholipid syndrome and can be used during breastfeeding. Alternative methods such as intrauterine de- vices or depot progesterone, either intramuscularly or subdermally, can be safely used in women with antiphospholipid syndrome. Women with antiphospholipid syndrome, particularly thrombotic antiphospholipid syndrome on warfarin, should be advised to plan their pregnancies. Rheumatoid arthritis Rheumatoid arthritis is common in women (female to male ratio 3:1), with a prevalence of 1–​2 per 1000 women per year in the Table 14.14.3  Risk of adverse obstetric outcomes in women with antiphospholipid antibodies and obstetric antiphospholipid syndrome compared to the normal population Control (n = 292) Isolated persistent aPL (n = 73) Obstetric APS (n = 73) Birthweight g, median (IQR) 3400 (1760–​4580) 3445 (3110–​3685) 3100 (2710–​3380) Small for gestational age, n (%) 31 (11.0) 4 (5.9) 17 (27.0) Adjustedb OR for SGA (95% CI) 1.0 0.5 (0.2–​1.4) 2.9 (1.5–​5.7) All APS-​type complicationsa, n (%) 31 (10.6) 9 (12.3) 28 (38.4) Adjustedb OR for all APS-​type complications (95% CI) 1.0 1.3 (0.6–​2.9) 5.7 (3.0–​10.9) aPL, antiphospholipid antibodies; APS, antiphospholipid syndrome; SGA, small for gestational age. a APS-​type complications include fetal loss >10 weeks’ gestation, early onset pre-​eclampsia necessitating <34-​week delivery, placental abruption, SGA. b Adjusted for maternal age and medical comorbidities (hypertension, renal disease, and diabetes mellitus). Adapted from Soh MC, Pasupathy D, Gray G, et al. (2013) Persistent antiphospholipid antibodies do not contribute to adverse pregnancy outcomes. Rheumatology (Oxford), Sep; 52(9):1642–​7.

14.14  Autoimmune rheumatic disorders and vasculitis in pregnancy 2661 18–​44-​year age bracket. In the past, pregnancy in young women with rheumatoid arthritis was rare as methotrexate (MTX)—​a known teratogen—​was often the first-​line drug used for it treatment. With the advent of biologics, there has been a surge of women with rheumatoid arthritis now attempting pregnancy. Effect of pregnancy on rheumatoid arthritis Newer prospective multicentre studies have challenged the belief that 75% of women with rheumatoid arthritis go into remission in pregnancy. Studies carried out in the United Kingdom and the Netherlands indicate that only 40–​66% of women experience an im- provement in pregnancy, and less than a quarter of women achieve remission by the third trimester. Women without rheumatoid factor (RF), anticyclic citrullinated peptide (CCP), or with less joint damage were more likely to improve in pregnancy. There have been several hypotheses to explain why pregnancy might, in some cases, lead to improvement in rheumatoid arth- ritis. These include the shift in pregnancy from a predominantly Th1 to Th2 immune response and cytokine repertoire; raised levels of circulating hormones (including cortisol, progesterone, and oestrogen); the maternal immune response to fetal paternally in- herited human leukocyte antigen (HLA) class II gene products, with maternal-​fetal mismatch for certain HLA alleles being beneficial; pregnancy-​specific proteins such as a2-​glycoprotein; removal of im- mune complexes by the placenta, and changes in adipocytokine and PPAR​1 signalling pathways. Anti-​CCP, IgM-​RF, and IgA-​RF titres all decline during pregnancy and post-​partum when disease-​modifying antirheumatic drugs are restarted. In the post-​partum period, however, low antibody levels are not linked to quiescent disease, nor are they useful in predicting risk of post-​partum flares of rheumatoid arthritis, which will occur in a third of women within a month; up to 90% will have experienced a deterioration in symptoms within four months post-​partum. There is an increased incidence of developing rheumatoid arth- ritis during the post-​partum period, with an incidence rate ratio of 1.73 (95% CI 1.11–​2.70) within 24 months of delivery of the first offspring. Studies looking at the effect of breastfeeding on disease have shown conflicting results, with some suggesting it is protective against disease development and others that it increases post-​ partum flare. Overall the evidence of an adverse effect is not strong enough to advise women against breastfeeding. Effect of rheumatoid arthritis on pregnancy Women with rheumatoid arthritis appear to have reduced fertility, with a recent study showing increased use of artificial reproductive techniques. A nationwide prospective study of women with rheuma- toid arthritis (n = 152) has shown that active disease in pregnancy is associated with a lower gestational age at delivery and a lower birthweight. In particular, women on prednisone had a significantly higher rate of preterm delivery (<37 weeks’ gestation) compared to those not taking prednisolone. Data from a nationwide database of obstetric hospitalizations in the United States of America indicate Table 14.14.4  Anticoagulation during pregnancy and puerperium in women who have antiphospholipid syndrome or have persistent antiphospholipid antibodies Antiphospholipid antibodies present in medium to high titre 12 weeks apart Antenatal management Postnatal Previous obstetric adverse events Previous thrombosis Aspirin (75 mg/ day) LMWH LMWH None None Yes, but controversial
(see text) No (unless other risk factors for thrombosis, e.g. family history of thrombosis, raised BMI, and so on) Yes for the usual indications Recurrent 1st-​trimester miscarriage (at least three consecutively) None Yes Controversial (see text). Low prophylactic dose from beginning of pregnancy (e.g. enoxaparin 40 mg once daily), but stop either after 1st trimester, or after 20–​24 weeks if uterine artery Dopplers normal (if no history of thrombosis or late fetal loss) Yes Late fetal loss (2nd or 3rd trimester) or Premature delivery <34 weeks because of severe pre-​eclampsia/
​eclampsia or other features of placental insufficiency None Yes Low prophylactic dose throughout pregnancy Yes Previous unprovoked thrombosis but not on long-​term anticoagulation Yes Yes Commence LMWH at prophylactic doses upon confirmation of intrauterine pregnancy Yes for at least six weeks post-​partum Previous recurrent thrombosis and on long-​term warfarin Yes Yes Swap warfarin to LMWH before 6 weeks’ gestation. High prophylactic dose, e.g. enoxaparin 40 mg twice daily, but may need to be therapeutic doses of LMWH esp. if history of recent thromboses Yes. Swap back to warfarin in the post-​ partum interval Thromboses that are recurrent
in pregnancy Yes Yes Full anticoagulant dose, e.g. enoxaparin 1 mg/​kg twice daily. May also consider re-​warfarinizing between 14 and 34 weeks’ gestation (INR 2–​3) esp. if recurrent symptoms of thromboses on therapeutic doses of LMWH Yes. Swap back to warfarin in the post-​ partum interval BMI, body mass index; CNS, central nervous system; LMWH, low-​molecular-​weight heparin.

Section 14  Medical disorders in pregnancy 2662 that women with rheumatoid arthritis have an increased risk of fetal growth restriction—​3.4% compared to the 1.6% incidence in the normal population. There was also a trend towards an increased risk of developing pre-​eclampsia, but this association was less ro- bust compared to women with systemic lupus erythematosus. The odds ratio in various studies for women with rheumatoid arthritis developing pre-​eclampsia is 1.4–​2.2; small for gestational age infants 1.20–​1.56; fetal growth restriction 2.2. Overall, the absolute risk of adverse obstetric outcomes remains minimal and similar to the normal population. Nevertheless, there is an unusually higher rate of elective cae- sarean section in women with rheumatoid arthritis compared to the normal population. This is most likely iatrogenic because of provider-​initiated clinical decision. Limitation of hip abduction is rarely severe enough to impede vaginal delivery, but particular care is necessary in women with juvenile idiopathic arthritis who have had hip joint replacements. Atlantoaxial subluxation (due to erosive disease affecting the cer- vical spine) is a rare complication of a general anaesthetic for an emergency caesarean section. Hence women should be reviewed by an obstetric anaesthetist to assess cervical spine involvement and de- gree of jaw excursion, and to anticipate any problems if they were to require a general anaesthetic. Management of rheumatoid arthritis in pregnancy All women with rheumatoid arthritis should ideally receive pre- conception counselling and their disease activity and medications reviewed prior to pregnancy. First-​line therapy for rheumatoid arthritis is methotrexate, which is highly teratogenic and requires a three-​month washout period and high-​dose folic acid supplementa- tion (5 mg a day) before conception. The risk of teratogenicity ranges between 5 to 25%, and there is a 23% risk of miscarriage in women taking methotrexate in the first trimester. The effect of methotrexate on organogenesis is proportional to the weekly dose ingested. Leflunomide is similarly contraindicated because of its teratogenic potential as a pyrimidine synthesis inhibitor. However, recent data have emerged showing that if cholestyramine washout (8 g a day for 11 days) of leflunomide occurs in early pregnancy, then the risk of fetal malformations is not increased above the normal population (5.4% vs. 4.2%). Confirmation of successful cholestyramine washout with undetectable drug levels (<0.03 µg/​ml) taken two weeks apart is necessary. If in doubt, ongoing washout with cholestyramine should continue until the woman is past the interval of organogenesis (12 weeks’ gestation). Biologics (and other drugs) in pregnancy Biological agents, especially tumour necrosis factor α-antagonists (anti-​TNFα), are now an established treatment for rheumatoid arth- ritis. Many women on anti-​TNFα who have excellent disease control are contemplating pregnancy. Large biologic databases and observational studies of birth out- comes of women on anti-​TNFα agents are reassuring that there is no link with teratogenicity. Current practice is therefore to encourage all women to continue on their anti-​TNFα therapy while attempting to conceive. For many, cessation of these drugs—​particularly if there is a prolonged interval between discontinuing the drug and suc- cessful conception—​could result in recurrent active disease, which in itself is a barrier to a pregnancy. At present, the biggest concern of anti-​TNFα use in pregnancy relates to the transplacental transfer that occurs predominantly after 20 weeks’ gestation. There are two factors that influence this: (i) the efficiency of the transplacental transfer—​which is dependent on the molecular structure of the Fc portion of the anti-​TNFα molecule; (ii) the half-​life of the drug used. Anti-​TNFα agents are immuno- globulins (Ig), so transplacental transfer increases exponentially as the pregnancy progresses with maximal active transplacental transfer occurring after 28 weeks. Immunosuppression of the neo- nate is a major theoretical concern, but small observational studies have not demonstrated an increased risk of infections. Infliximab, adalimumab, and golimumab are IgG1 monoclonal antibodies, which is the Ig subclass with the most active trans- port across the placenta. In early pregnancy, transfer is limited by a cytotrophoblast. By week 14, Fc receptors begin to develop on the trophoblasts and active transport is increasingly efficient as pregnancy progresses. Thus, at the time of delivery infliximab and adalimumab levels in the neonate (as measured in umbilical cord blood) often exceed maternal levels. Moreover, the very long half-​ lives of adalimumab and infliximab (8–​20 days) have led to very high levels (98–​400% of maternal drug levels) of the active drug de- tected in the cord blood of the neonate. Etanercept has a much shorter half-​life (four days) and with its modified Fc portion, may not bind quite as effectively to the pla- cental Fc receptors. Cord etanercept levels have been demonstrated to be much lower, at 3.6–​7.4% of maternal drug levels. The pegylated anti-​TNFα agent, Certolizumab, lacks an Fc por- tion which prohibits active transplacental transport. Therefore, it is reliant on slow diffusion across the placental barrier to reach the developing fetus. Studies have demonstrated negligible transplacental transfer and this drug therefore has a license for use throughout pregnancy. The decision about whether and when to discontinue anti-​TNFα agents in pregnancy should be individualized to the woman’s risk of a flare, the availability of other agents to manage a flare, and the mo- lecular structure of the drug itself. Longer acting anti-​TNFα agents which are actively transported across the placenta (i.e. adalimumab and infliximab), may have to be discontinued earlier than etanercept and certolizumab. A  blanket statement of discontinuing all anti-​ TNFα agents at 28 weeks may not be appropriate, especially if the woman has a preterm delivery and the infant has had recent exposure to the anti-​TNFα agent. There are suggestions when to contemplate discontinuing anti-​TNFα agents based on available published lit- erature (Table 14.14.5). The 2016 British Society of Rheumatology (BSR) guidelines suggest that if we were to aim for low or undetect- able anti-​TNF levels, then infliximab be discontinued by 16 weeks and etanercept and adalimumab should be discontinued in the late second trimester, but certolizumab could be continued throughout pregnancy. New data on the infant’s immune response to vaccination for those exposed to anti-​TNFα agents in utero in the third trimester show that the responses to Haemophillus influenza type B, Streptococcus pneumoniae, and tetanus were normal, and there were no immune deficits of concern in these infants. There is a single case report in the literature detailing a neonatal death from presumed dissem- inated tuberculosis infection following Bacillus Calmette–​Guérin (BCG) vaccination, although Ziehl–​Neelsen stain for acid-​fast ba- cilli and TB PCR methods were equivocal. The infant was exposed to

14.14  Autoimmune rheumatic disorders and vasculitis in pregnancy 2663 infliximab in utero with the last maternal infusion of infliximab for the treatment of her inflammatory bowel disease occurring just two weeks prior to delivery. Based on these data and several other anecdotal cases, all infants with exposure to anti-​TNFα agents after suggested gestation (with the possible exception of pegylated anti-​TNFα agent certolizumab) should not receive any live vaccines in the first six to eight months of life. The BSR suggests an avoidance of live vaccines in an exposed neonate for the first seven months. There is delayed clearance of these antibodies due to the infant’s immature reticuloendothelial system and supra-​therapeutic drug levels (Table 14.14.5). Common live vaccines during that neonatal period include BCG vaccination and the rotavirus vaccine (an orally administered vaccine). The caregivers of exposed neonates should be vigilant that these vac- cines are not routinely administered. All other vaccines should be given. Additionally, despite the lack of direct correlation between drug exposure and neonatal infection, we would still advise that exposed neonates should be carefully examined, and infection ex- cluded if they become unwell. Clinicians probably should have a lower threshold for screening and treating the exposed neonate for a presumed infection. B-​cell depleting agents such as rituximab are being increasingly used for the treatment of systemic lupus erythematosus, rheumatoid arthritis, and many vasculitides. In 2011 a case series of 253 pregnan- cies from the manufacturer’s global drug safety database (dependent on voluntary self-​reporting) has shown that there was a 60% live birth rate and 21% first-​trimester miscarriage rate. The high rate of early pregnancy losses could be related to the underlying disease for which the woman was being treated with rituximab. Rituximab can be detected in cord blood if given within 12 weeks of delivery, and this corresponds to neonatal B-​cell depletion. Yet there does not ap- pear to be any increased risk of infections. Long-​term outcome studies are awaited to determine if there is a future risk of autoimmune disease or lymphoid-​cell malignancies in offspring exposed in utero to biologics. Most disease-​modifying antirheumatic drugs used in rheumatoid arthritis overlap with those used in systemic lupus erythematosus, but additional agents include D-​penicillamine, sulphasalazine, and gold (Table 14.14.6). All disease-​modifying antirheumatic drugs, other than the teratogenic ones (i.e. methotrexate, leflunomide, cyclo- phosphamide, and mycophenolate mofetil), should be continued in pregnancy to prevent flares. Most women with flares of rheumatoid arthritis can be managed on analgesics and low-​dose prednisolone. Non​steroidal anti-​inflammatory drugs (NSAIDs) are not terato- genic and can be used if needed up to 32 weeks’ gestation, after which there is a risk of premature closure of the patent ductus arteriosus and neonatal pulmonary hypertension. The effect is usually reversible within 48-​hours after discontinuation of the NSAID. NSAIDs also have reversible effects on the fetal kidneys with oligohydramnios. Hence their occasional use in late pregnancy for the treatment of polyhydramnios. Data on selective COX-​2 inhibitors in pregnancy are emerging with a population-​based study indicating that there is no teratogenic effect with first trimester exposure. However, there were five cases of musculoskeletal limb defects seen in the group on selective COX-​2 inhibitors. Inflammatory arthritides were also more common in the group on selective COX-​2 inhibitors. Therefore, it is possible that this group were also exposed to methotrexate—​which is associated with limb defects. The future of selective COX-​2 inhibitor in preg- nancy will lie in its improved safety profile in late pregnancy with a less deleterious effect on the fetal ductus arteriosus and kidneys late pregnancy compared to the non​selective COX-​2 inhibitors (or NSAIDs). Vasculitides Granulomatous polyangiits (previously Wegener’s granulomatosis), polyarteritis nodosa, and eosinophilic granulomatosis with polyangiitis (previously Churg–​Strauss syndrome), occur princi- pally in the post-​childbearing years and more often in men, so preg- nancy is very uncommon in these conditions. Henoch–​Schönlein purpura tends to affect the paediatric population, but there have been a few case reports of presentation in pregnancy, or pregnancy after previous Henoch–​Schönlein purpura. Pregnancy is more likely in Takayasu’s arteritis and Behçet’s disease, as they usually occur in the reproductive years. In general, maternal and fetal outcome are dependent on disease activity and pre-​existing complications. Numbers of reported cases are generally too small to determine definitively if disease onset or flare is more likely during pregnancy or post-​partum. The standard advice to plan pregnancy after the disease is in remission still holds as quiescent disease at conception or early pregnancy is usually linked with stable disease in pregnancy. Flares for most conditions are common in the post-​partum interval with maternal immune reconstitution. Table 14.14.5  Accumulation of commonly used anti-​TNFα agents in the neonate and suggested gestation to discontinue Anti-​TNFα agent Half-​life (days) Cord blood to maternal serum concentrations (%) Suggested gestation to discontinue (weeks) Infliximab 8–​10 83–​400 21–​22 Adalimumab 10–​20 98–​293 26–​28 Etanercept 4 3.6–​7.4 30–​32 Certolizumab 14 0–​24.0 Continue throughout Adapted from Soh MC, Nelson-​Piercy, C. (2015) High-​risk pregnancy and the rheumatologist. Rheumatology (Oxford), 54(4), 572–​87, by permission of Oxford University Press, and Soh MC, Mackillop L. (2016) Biologics in Pregnancy—​for the Obstetrician. The Obstetrician and Gynaecologist, January [in press].

Section 14  Medical disorders in pregnancy 2664 Pregnancy outcomes in granulomatous polyangiits are worse with active disease at conception or if there is disease onset during pregnancy, with 57% resulting in miscarriages. Antineutrophil cyto- plasmic antibody (ANCA) titres do not influence obstetric out- comes. If disease is in remission at the beginning of pregnancy, there is still a significant risk of pre-​eclampsia (28.6%) and median gesta- tion at delivery was 35–​36 weeks. There were two maternal deaths from severe flares of the disease out of 43 pregnancies reported in the literature. Pre-​eclampsia was more common with renal involve- ment. Flares post-​partum are common and severe. Flares of granulomatous polyangiits should be aggressively treated. One paper suggests that the use of azathioprine and pred- nisolone to treat a severe flare in pregnancy may be insufficient. Cyclophosphamide—​a known teratogen, but efficacious against severe flares can be used after the first trimester if there is life-​ threatening maternal disease. However, with the advent of rituximab to treat granulomatous polyangiits, this may soon become the pre- ferred option. Other agents to maintain remission in limited granulomatous polyangiits include cotrimoxazole. Although it is a folate antagonist, Table 14.14.6  Commonly used drugs used for the treatment of rheumatic diseases and their effect on pregnancy and breastfeeding Drug Effects on organogenesis (exposure ≤12 weeks’ gestation) Effects on fetus/​neonate (exposure beyond 12 weeks’ gestation) Breastfeeding Authors’ recommendations on its use in pregnancy NSAIDs None Reversible constriction of ductus arteriosus; oligohydramnios. Avoid after 32 weeks or within 48 hours of delivery ✓ Likely a class effect for all NSAIDs. Use if indicated at lowest dose possible until 32 weeks’ gestation COX-​2 inhibitors No teratogenic effects seen in population-​based study Similar effect, but to a lesser degree, as the NSAIDs on fetal heart and kidneys ✓ In the first trimester, till more data emerge, it may be safer to change to a NSAID; though use later in gestation may be less deleterious than NSAIDs Prednisone None Fetus receives <10% of maternal dose ✓ Use lowest dose possible. Plan to taper. To consider addition of a DMARD or biologic if persistently active disease Hydroxychloroquine None None ✓ Continue in pregnancy and breastfeeding Sulfasalazine None None ✓ Commence folic acid supplementation 5 mg/​day, 3 months prior to pregnancy. In men it may affect spermatogenesis and motility Methotrexate Aminopterin syndrome. 15% rate of congenital anomalies—​likely dose related. High rates of pregnancy loss. Safe to be used in men If no congenital anomalies, long-​term follow-​up of children exposed to MTX did not reveal any problems Case report indicates very low levels excreted Reliable contraception advised. Discontinue at least 3 months prior to pregnancy with daily high-​ dose folic acid supplementation. Exposed fetuses should be scanned at 16/​40 to determine if there are any congenital anomalies to facilitate elective termination if the mother wishes. Minimal amounts excreted in breast milk 24 hours after a sc dose of weekly methotrexate—​based in a case report Leflunomide In animal studies, malformations of the head, rump, vertebral column, and limb defects. Increased rate of miscarriages If pregnancy continues, no major structural anomalies noted esp. after cholestyramine wash out as suggested by manufacturer ✘ Reliable contraception advised. Wash out with cholestyramine 8 g TDS 11 days—​repeat till drug level <0.03 μg/​ml taken two weeks apart, or until past 12 weeks’ gestation Azathioprine None Low birth weight and preterm delivery—​could be secondary to maternal disease ✓ Continue in pregnancy and lactation Ciclosporin None Transient immune alterations in the neonate ✓ Continue in pregnancy; probably safe in breastfeeding though a variety of concentrations excreted in breast milk Cyclophosphamide Cyclophosphamide embryopathy with high rate of miscarriages Transient cytopenias. No long-​term effect on the infant ✘ Use only if there is life-​threatening maternal disease >12 weeks. If maternal disease necessitates cyclophosphamide ≤12 weeks’ gestation–​ discuss termination Mycophenolate mofetil OMENS and congenital cardiac defects Most neonates described in the literature, had also been exposed in the period of organogenesis. Phenotype is not dose dependent ✘ Discontinue for at least 3 months prior to pregnancy DMARD, disease-​modifying antirheumatic drug; OMENS, orbital distortion, mandibular hypoplasia, ear anomaly—​N is for seventh cranial nerve involvement, and soft tissue deficiency. † Azathioprine converts to active metabolite 6-​thioguanine nucleotides in 15 minutes but the half-​life of the active metabolite in erythrocytes is weeks to months. ✓—​safe for breastfeeding; ✘—​unsafe or not recommended for breastfeeding Adapted from Soh MC, Nelson-​Piercy C (2015). High-​risk pregnancy and the rheumatologist. Rheumatology (Oxford), 54(4), 572–​87.

14.14  Autoimmune rheumatic disorders and vasculitis in pregnancy 2665 studies have demonstrated its safety in pregnancy, though folic acid 5 mg daily should be co-​prescribed. Polyarteritis nodosa is a rare necrotizing vasculitis with a yearly incidence of two per million, occurring predominantly in those aged 50–​60 years. It is a great mimic, with presenting features of abdom- inal pain, hypertension, and proteinuria, all of which are common in pregnancy. It carries a high risk of maternal mortality (up to 100%) if diagnosed during pregnancy or immediately post-​partum, whereas the maternal and fetal outcome is much better if it is diagnosed pre-​ pregnancy and disease is in remission. There are case reports of neo- natal cutaneous vasculitis in babies born to mothers with cutaneous polyarteritis nodosa. This is thought to be a result of transplacental transfer of maternal antibodies and the neonatal cutaneous changes remit spontaneously. Eosinophilic granulomatosis with polyangiitis which is asso- ciated with late onset asthma, nasal polyps, pulmonary infiltrates, and asthma, is also rare (two to three per million), with a peak incidence in the fourth decade. It tends to be less aggressive than granulomatous polyangiits or polyarteritis nodosa, especially if women conceive during remission. Many women with eosinophilic granulomatosis with polyangiitis will require ongoing aggressive glucocorticoid therapy to keep their pulmonary symptoms under control. Exacerbations of the disease are often due to the reluctance of pregnant women to take medications, in addition to loss of lung capacity from the growing fetus. Maternal and fetal outcomes are significantly worse if eosinophilic granulomatosis with polyangiitis presents during pregnancy, and cardiac disease (from coronary vas- culitis, myocardial fibrosis, and eventual congestive cardiac failure) is an important cause of maternal death. Of the vasculitides, Takayasu’s, the ‘pulseless disease’, is most common in oriental women of childbearing age. Pregnancy has a positive effect on the disease course with lower C-​reactive pro- teins (CRPs) and improved digital plethysmography up to a year following delivery. Issues with Takayasu’s arteritis include hyper- tension and involvement of the aortic valve and abdominal aorta, and superimposed pre-​eclampsia and fetal growth restriction are common, with up to 62% and 11–​52% of pregnancies affected re- spectively. In the past, pre-​eclampsia affected 75% of all pregnant women with Takayasu’s especially if they had hypertension diag- nosed prior to pregnancy. Aspirin use has helped reduce the in- cidence of pre-​eclampsia in the more recent studies. Pulmonary hypertension should be excluded before pregnancy. Obstetric out- comes are likely dependent on the extent of vessel involvement. Fetal growth restriction is common and likely due to a combination of impaired placental perfusion as a result of vascular narrowing of the abdominal aorta and its branches and hypertension. There is a cor- relation between disease severity and poor obstetric outcomes; those with Ishikawa class IIb (two or more vessels affected) have poorer outcomes than those with milder disease. Disease activity can be monitored with ESR and CRP in pregnancy and active disease usu- ally responds well to low doses of steroids. Maintenance doses of prednisolone and azathioprine should be continued in pregnancy. Recent publications on Behçet’s disease in pregnancy suggest, that pregnancy seems to positively affect disease course with a reduction in flares (threefold lower) in pregnancy. Flares occurred in about 30% of patients and were more common in those off their medications. The flares are usually mucocutaneous in nature and usually in the third tri- mester. Behçet’s does not appear to adversely affect maternal and fetal outcomes in case–​control studies. A history of previous thromboses and vascular complications was associated with poorer outcomes. Acute thromboses, including deep vein thromboses (DVTs), inferior vena cava, and dural sinus thrombosis, have been reported and clin- icians should be vigilant that thrombotic complications are prob- ably more likely to recur in pregnancy—​a naturally pro-​thrombotic state. Thromboprophylaxis should be strongly considered in women with any additional risk factors (i.e. age, grand multiparity, maternal obesity, and so on). Colchicine, which is the mainstay of treatment in mucocutaneous Behçet’s, was not teratogenic in a study involving 238 colchicine-​exposed pregnancies when compared to 964 control preg- nancies. Behçet’s is one of the rare indications for thalidomide, and women of childbearing age who are commenced on it should be using reliable contraception. The high rate of early pregnancy losses seen in Behçet’s could be due to elective terminations. Henoch–​Schönlein purpura is typically a disease of childhood, presenting with colicky abdominal pain and purpuric rash as a result of IgA-​mediated vasculitis. It is usually a relatively benign condi- tion with an indolent course, with favourable maternal and fetal out- comes. The main concern regarding Henoch–​Schönlein purpura in pregnancy is renal involvement, which is usually mild and resolves spontaneously, but may lead to nephrotic syndrome or acute kidney injury and needs to be distinguished from pre-​eclampsia. In general, in view of the significant maternal and fetal morbidity and mortality associated with active vasculitides, women should be advised to delay pregnancy until disease is in remission. In the case of flare or onset of disease in pregnancy, it is important to adopt an aggressive approach to treatment with immunosuppression. Corticosteroids are usually first-​line treatment. Depending on the underlying vasculitis and the severity of the flare, this can be followed by azathioprine. Life-​threatening disease may necessitate the use of pulsed cyclophosphamide (especially in granulomatous polyangiits) or rituximab. There are cases in the literature where intravenous im- munoglobulins (IVIG) or plasmapheresis has been successfully used in refractory disease, thus minimizing the fetal risk. Scleroderma Scleroderma (or systemic sclerosis) is more common in women (female to male ratio 3:1), with peak age of onset 30–​50 years old. Although it is a rare disease, more cases are being reported in preg- nancy as maternal age increases. Effect of pregnancy on scleroderma In general, women with limited scleroderma without organ involve- ment do better than those with diffuse disease. The extent of diffuse disease and systemic involvement (particularly lung, cardiac, and renal) influences prognosis. Those with early (<4 years) or diffuse disease, or with antitopoisomerase (anti-​Scl-​70) antibodies, are at greater risk of having more active aggressive disease than those with long-​standing disease and anticentromere antibodies. Pregnancy does not affect the course of scleroderma, with 60% having a stable disease course, 20% improving, and 20% deteriorating in pregnancy. The rationale for deferring pregnancy until at least four years after diagnosis is to ensure that early disease can be aggressively treated with potent and sometimes teratogenic immunosuppressants; and if renal crises occur, they tend to be in early disease.

Section 14  Medical disorders in pregnancy 2666 Renal crisis is the main cause of death, with 33% dying in the post-​ partum interval. With the introduction of angiotensin converting enzyme (ACE) inhibitors, the numbers of such deaths have fallen significantly, and many women with previous renal crisis have sub- sequently had successful pregnancies. In a woman with a previous renal crisis, ACE inhibitors should not be discontinued in preg- nancy and the woman should be carefully counselled about the fetal effects. Avoidance of high doses of steroids, which may precipitate renal scleroderma, is essential. The benefits of antenatal steroids for fetal lung maturation should be very carefully weighed against the risk of a maternal renal crisis. Women with renal involvement often have accelerated hypertension and acute deterioration of renal function necessitating temporary renal replacement therapy. Progressive cutaneous disease is unusual during pregnancy but may occur in the post-​partum period. Raynaud’s phenomenon usually improves as a result of vasodilation. Gastro-​oesophageal reflux is exacerbated by a reduced lower oesophageal tone. Arthralgias also worsen. There is no evidence that pregnancy worsens cardiac or re- spiratory disease. However, symptomatic shortness of breath from the expanding gravid uterus is common. Women with severe pul- monary fibrosis and pulmonary hypertension are at extremely high risk of post-​partum deterioration, with a 30–​50% chance of ma- ternal mortality, as with pulmonary hypertension from any cause. In a study exclusively of pregnant women with pulmonary hyper- tension as a result of scleroderma (n = 85), the authors found a five- fold risk of maternal hypertensive disease and a twofold risk of fetal growth restriction. Worsening skin thickening has been observed in post-​partum women. It has been suggested that pregnancy, including miscarriage, may have an aetiological role in scleroderma, with some studies showing that persistent fetal microchimerism and HLA-​DR compatibility be- tween mother and fetus are more common in women with sclero- derma than controls. Hypotheses implicating the persistence of these fetal cells in the pathogenesis of scleroderma include the de- velopment of a fetal antimaternal graft-​versus-​host reaction and/​or the maternal response to fetal cells becoming redirected against the mother herself. Effect of scleroderma on pregnancy Fertility is unimpaired in women with scleroderma. However, spon- taneous miscarriage is more common in those with established scleroderma compared to those with early disease. Women with dif- fuse disease have twice the rate of miscarriages compared to those with limited disease (24% vs. 12%). There is a 9% risk of preterm deliveries, with a higher risk in those with diffuse disease. Overall, the live birth rate is 84% in those with limited disease and 77% in those with diffuse disease. The risk of adverse outcome is highest for women with early diffuse disease. A subset of women with scleroderma may have concurrent Ro and La positivity, with ranges of 12–​37% for the former and 4% for the latter. Antenatal monitoring for congenital heart block should be undertaken in those women who have tested positive. Management of scleroderma in pregnancy Women should be assessed before conception for the extent of organ involvement. Those with significant renal impairment, severe restrictive lung disease, pulmonary hypertension, or severe cardio- myopathy should be advised against pregnancy. Any woman with previous pulmonary involvement should have lung function tests done in the first trimester and repeated at the latter stages of preg- nancy if clinically indicated. Those with early diffuse disease should delay pregnancy until the disease stabilizes Pulmonary hypertension is a complication in about 8–​12% of patients with scleroderma and is the main cause of disease-​related mortality. Whether corticosteroids (prednisone >15  mg/​day or equiva- lent) precipitate renal crisis remains contentious. High-​dose fluorinated steroids (i.e. dexamethasone and betamethasone) are used for fetal lung maturation, and there are concerns that use in the third trimester, when preterm delivery of a neonate is immi- nent, could precipitate this condition. However, none of the cases of renal crisis during the peripartum interval in the published literature have been caused by antenatal steroids for fetal lung maturation. Anti-​RNA polymerase III (and antitopoisomerase antibodies—​in certain cohorts) may be useful in prognosticating risk of renal crisis, but the strongest predictor remains new onset (<4  years) diffuse disease with rapidly progressive cutaneous involvement. The features of renal scleroderma crisis are almost exactly the same as pre-​eclampsia or HELLP (haemolysis, elevated liver func- tion tests, and low platelets) syndrome. Renal histological changes may aid in differentiating the two pathologies, but renal biopsy is rarely performed in late pregnancy. Most scleroderma renal crisis develop in the third trimester, therefore the management includes prompt delivery prior to initiation of an ACE inhibitor. ACE in- hibitors are usually contraindicated in pregnancy, but the benefits in treating uncontrolled hypertension and preventing maternal mortality usually far outweigh the fetal risks. Otherwise, the management of scleroderma during pregnancy is largely symptomatic, including calcium antagonists for Raynaud’s phenomenon, and histamine antagonists and proton pump inhibi- tors for reflux. NSAIDs are best avoided. Venepuncture, venous access, oxygen saturation, and blood pres- sure measurements may be difficult because of skin, nail, or blood vessel involvement. General anaesthesia may be complicated by dif- ficult endotracheal intubation from microstomia, risk of aspiration, and trauma to telangectatic areas in the nasopharynx. Regional anaesthesia is usually the preferred option. Early assessment by an obstetric anaesthetist is advisable, and epidural anaesthesia and analgesia are encouraged as vasodilation improves skin perfusion of the extremities. Other measures to reduce problems related to Raynaud’s phenomenon include warming of the delivery room and any intravenous fluids as well as socks and gloves. Close observation must continue in the immediate postnatal period, particularly in those with cardiac, pulmonary, or renal in- volvement. Carers must be vigilant of any pressure areas, which may result in cutaneous necrosis. The use of antirheumatic drugs, immunosuppressive agents, and biologics in pregnancy Active disease in early pregnancy often leads to poor placenta- tion or placental insufficiency that manifests itself clinically as maternal-​placental syndrome (as described earlier) and results in

14.14  Autoimmune rheumatic disorders and vasculitis in pregnancy 2667 a preterm delivery of a growth restricted infant. Moreover, there is increasing evidence that antenatal ill-​health affects long-​term maternal well-​being and survival, this is particularly true in those with systemic lupus erythematosus. One of the greatest challenges for clinicians caring for pregnant women is to provide sufficient reassurance that most medications are safe in pregnancy. Active or uncontrolled disease activity has a more deleterious effect on both the mother and developing fetus, therefore every effort should be made to optimize disease control in pregnancy. Counselling a woman about drugs in pregnancy therefore requires not only careful interpretation of the published literature, but also a ro- bust understanding of the basic pathophysiological process and pharmacologic effects of the drugs. Table 14.14.6 lists the various uses of disease-​modifying drugs in rheumatic disorders in pregnancy and breastfeeding. When pregnancy should be avoided There are some circumstances when the risk to mother or fetus are very high. These include the following: • Pulmonary hypertension—​maternal mortality is up to 40%. Effective and reliable contraception should be discussed at each visit. In the event of an unplanned pregnancy, termination of pregnancy before 12 weeks is advised. • Chronic kidney disease (CKD) stage 4 or 5—​prospective studies involving women with CKD have demonstrated 36–​40% chance of pre-​eclampsia and 54–​80% chance of preterm delivery; threefold increased risk of a small for gestational age infant and fivefold increased risk of perinatal mortality. Women with CKD 4/​5 prior to pregnancy are at greater risk of an acceler- ated decline in renal function with the risk of reaching end stage and needing renal replacement therapy either in pregnancy or shortly after. • Severe maternal rheumatic disease in early pregnancy, or ongoing active or poorly controlled disease—​these are seldom conducive to the development of a healthy fetus. • Antiphospholipid syndrome with recurrent placenta-​mediated adverse pregnancy outcomes despite optimal management and anticoagulation. • Women with recurrent intrauterine death, early onset severe pre-​ eclampsia, HELLP syndrome, and severe intrauterine growth restriction with poor neonatal survival despite treatment with aspirin and low-​molecular-​weight heparin may wish to continue attempting pregnancy, but their chances of successful outcome are low. Future maternal health There is increasing evidence that obstetric complications, particu- larly those related to placental insufficiency, are associated with the future development of vascular disease. Pregnancy is a ‘stress test for life’ and many future morbidities that a woman will have are un- masked by the physiological stresses of pregnancy. In the general population, women who have had maternal-​ placental syndrome in pregnancy are four-​times more likely to develop hypertension and future renal disease and also to suffer various forms of vascular complications. Maternal-​placental syndrome-​type complications and preterm deliveries are more common in women with autoimmune connective tissue diseases. These women, particularly those with systemic lupus erythematosus, are at greater risk of future vascular disease, even without trad- itional cardiovascular risk factors. In large population-​based cohort studies, women with systemic lupus erythematosus who have had maternal-​placental syndrome-​type complications in pregnancy are at risk of future cardiovascular events and have a 2.2-​fold risk of pri- mary cardiovascular death. Clinicians caring for women with systemic lupus erythematosus (and like all autoimmune rheumatic diseases) should consider adverse pregnancy outcomes and preterm deliveries as a ‘red-​ flag’ and may wish to commence primary prevention earlier in this group. Acknowledgements We would like to acknowledge the contributions of Sarah Germain and Kate Bramham to the previous versions of this chapter. FURTHER READING General Chakravarty E, et al. (2014). Family planning and pregnancy issues for women with systemic inflammatory diseases: patient and physician perspectives. BMJ Open, 4, e004081. Østensen M, et al. (2015). State of the art: reproduction and pregnancy in rheumatic diseases. Autoimmun Rev, 14, 376–​86. Østensen M, Forger F, Villiger PM (2006). Cytokines and pregnancy in rheumatic disease. Ann N Y Acad Sci, 1069, 353–​63. Petri M (2007). The Hopkins Lupus Pregnancy Center: ten key issues in management. Rheum Dis Clin N Am, 33, 227–​364. Soh MC, Nelson-​Piercy C (2015). High-​risk pregnancy and the rheumatologist. Rheumatology, 54, 572–​87. Systemic lupus erythematosus Bramham K, Lightstone L (2012). Pre-​pregnancy counseling for women with chronic kidney disease. J Nephrol, 25, 450–​9. Bramham K, Soh MC, Nelson-​Piercy C (2012). Pregnancy and renal outcomes in lupus nephritis: an update and guide to management. Lupus, 21, 1271–​83. Chakravarty EF, Nelson L, Krishnan E (2006). Obstetric hospi- talizations in the United States for women with systemic lupus erythematosus and rheumatoid arthritis. Arthritis Rheum, 54, 899–​907. Clowse ME, Magder LS, Petri M (2011). The clinical utility of measuring complement and anti-​dsDNA antibodies during preg- nancy in patients with systemic lupus erythematosus. J Rheumatol, 38, 1012–​16. Clowse MEB, et al. (2005). The impact of increased lupus activity on obstetric outcomes. Arthritis Rheum, 52, 514–​21. Lisnevskaia L, Murphy G, Isenberg D (2014). Systemic lupus erythematosus. Lancet, 384, 1878–​88. Peart E, Clowse ME (2014). Systemic lupus erythematosus and preg- nancy outcomes: an update and review of the literature. Curr Opin Rheumatol, 26, 118–​23.

Section 14  Medical disorders in pregnancy 2668 Ritchie J, et al. (2012). Maternal deaths in women with lupus neph- ritis: a review of published evidence. Lupus, 21, 534–​41. Salmon JE, et al. (2011). Mutations in complement regulatory proteins predispose to preeclampsia: a genetic analysis of the PROMISSE co- hort. PLoS Med, 8, e1001013. Smyth A, et al. (2010). A systematic review and meta-​analysis of preg- nancy outcomes in patients with systemic lupus erythematosus and lupus nephritis. Clin J Am Soc Nephrol, 5, 2060–​8. Soh MC, et  al. (2015). Accelerated development of cardiovascular events in Swedish women with systemic lupus erythematosus—​does pregnancy and its outcomes play a role? A population-​based retro- spective study [OP0050]. Ann Rheum Dis, 7, 85–​6. Webster P, et al. (2014). Tacrolimus is an effective treatment for lupus nephritis in pregnancy. Lupus, 23, 1192–​6. Neonatal lupus Cimaz R, et  al. (2003). Incidence and spectrum of neonatal lupus erythematosus: a prospective study of infants born to mothers with anti-​Ro autoantibodies. J Pediatr, 142, 678–​83. Clancy RM, et al. (2006). Impaired clearance of apoptotic cardiocytes is linked to anti-​SSA/​Ro and SSB/​La antibodies in the pathogenesis of congenital heart block. J Clin Invest, 116, 2413–​22. Friedman DM, et al. (2010). Evaluation of fetuses in a study of intra- venous immunoglobulin as preventative therapy for congenital heart block: results of a multicenter, prospective, open-​label clinical trial. Arthritis Rheum, 4, 1138–​46. Friedman DM, Rupel A, Buyon JP (2007). Epidemiology, etiology, de- tection and treatment of autoantibody-​associated congenital heart block in neonatal lupus. Curr Rheumatol Rep, 9, 101–​8. Izmirly PM, et al. (2010). Cutaneous manifestations of neonatal lupus and risk of subsequent congenital heart block. Arthritis Rheum, 62, 1153–​7. Izmirly PM, et al. (2012). Maternal use of hydroxychloroquine is as- sociated with a reduced risk of recurrent anti-​SSA/​Ro-​antibody-​ associated cardiac manifestations of neonatal lupus. Circulation, 126, 76–​82. Routsias JG, et al. (2011). Association of idiotypic-​anti-​idiotypic anti- body ratio with therapeutic response of IVIG in the prevention of recurrent autoimmune associated congenital heart block. Arthritis Rheum, 63, 2783–​9. Antiphospholipid syndrome Bramham K., et al. (2010). Pregnancy outcome in different clinical phenotypes of antiphospholipid syndrome. Lupus, 19, 58–​64. Bramham K., et al. (2011). First-​trimester low-​dose prednisolone in refractory antiphospholipid antibody-​related pregnancy loss. Blood, 117, 6948–​51. de Swiet M, Redman CW (1992). Aspirin, extradural anaesthesia and the MRC Collaborative Low-​dose Aspirin Study in Pregnancy (CLASP). Br J Anaesth, 69, 109–​10. Farquharson RG, Quenby S, Greaves M (2002). Antiphospholipid syn- drome in pregnancy: a randomized, controlled trial of treatment. Obstet Gynecol, 100, 408–​13. Gomez-​Puerta JA, et al. (2005). Long-​term follow-​up in 128 patients with primary antiphospholipid syndrome: do they develop lupus? Medicine (Baltimore), 84, 225–​30. Laskin CA, et al. (1997). Prednisone and aspirin in women with auto- antibodies and unexplained recurrent fetal loss. N Engl J Med, 337, 148–​53. Laskin CA et al. (2009). Low molecular weight heparin and aspirin for recurrent pregnancy loss: results from the randomized, controlled HepASA trial. J Rheumatol, 36, 279–​87. Lim W, Crowther MA, Eikelboom JW (2006). Management of antiphospholipid antibody syndrome:  a systemic review. JAMA, 295, 1050–​7. Miyakis S, et al. (2006). International consensus statement on an up- date of the classification criteria for definite antiphopholipid syn- drome (APS). J Thromb Haemost, 4, 295–​306. Ruiz-​Irastorza G, et al. (2010). Antiphospholipid syndrome. Lancet, 376, 1498–​509. Soh MC, Nelson-​Piercy C (2010). Antiphospholipid syndrome in pregnancy. Expert Rev Obstet Gynecol, 5, 741–​61. Soh MC, et al. (2013). Persistent antiphospholipid antibodies do not contribute to adverse pregnancy outcomes. Rheumatology, 52, 1642–​7. Tincani A, et al. (2006). Pregnancy, lupus and antiphospholipid syn- drome (Hughes syndrome). Lupus, 15, 156–​60. Triolo G, et al. (2003). Randomized study of subcutaneous low mo- lecular weight heparin plus aspirin versus intravenous immuno- globulin in the treatment of recurrent fetal loss associated with antiphospholipid antibodies. Arthritis Rheum, 48, 728–​31. Rheumatoid arthritis Bowden AP, et  al. (2001). Women with inflammatory polyarthritis have babies of lower birth weight. J Rheumatol, 28, 355–​9. Chakravarty EF (2011). Rheumatoid arthritis and pregnancy: beyond smaller and preterm babies. Arthritis Rheum, 63, 1469–​71. Chakravarty EF, Nelson L, Krishnan E (2006). Obstetric hospi- talizations in the United States for women with systemic lupus erythematosus and rheumatoid arthritis. Arthritis Rheum, 54, 899–​907. Chambers CD, et al. (2006). Human pregnancy safety for agents used to treat rheumatoid arthritis: adequacy of available information and strat- egies for developing post-​marketing data. Arthritis Res Ther, 8, 215. Cheent K, et al. (2010). Case report: fatal case of disseminated BCG infection in an infant born to a mother taking infliximab for Crohn’s disease. J Crohns Colitis, 4, 603–​5. de Man YA, et  al. (2008). Disease activity of rheumatoid arthritis during pregnancy:  results from a nationwide prospective study. Arthritis Rheum, 59, 1241–​8. de Man YA, et al. (2009). Association of higher rheumatoid arthritis disease activity during pregnancy with lower birth weight: results of a national prospective study. Arthritis Rheum, 60, 3196–​206. de Man YA et  al. (2010). Women with rheumatoid arthritis nega- tive for anti-​cyclic citrullinated peptide and rheumatoid factor are more likely to improve during pregnancy, whereas in autoantibody-​ positive women autoantibody levels are not influenced by preg- nancy. Ann Rheum Dis, 69, 420–​3. Lin HC, et al. (2010). Increased risk of adverse pregnancy outcomes in women with rheumatoid arthritis: a nationwide population-​based study. Ann Rheum Dis, 69, 715–​7. Nørgaard M, et al. (2010). Rheumatoid arthritis and birth outcomes: a Danish and Swedish nationwide prevalence study. J Intern Med, 268, 329–​37. Skomsvoll JF, et  al. (1999). Perinatal outcomes in pregnancies of women with connective tissue disease and inflammatory rheumatic disease in Norway. Scand J Rheumatol, 28, 3526. Skomsvoll JF, et  al. (2000). Pregnancy complications and delivery practice in women with connective tissue diseases and inflamma- tory rheumatic disease in Norway. Acta Obstet Gynaecol Scand, 79, 490–​5. Soh MC, Nelson-​Piercy C. (2012). Update of the management of rheumatoid arthritis in pregnancy. Expert Rev Obstet Gynecol, 7, 77–​96.

14.14  Autoimmune rheumatic disorders and vasculitis in pregnancy 2669 Verstappen SM, et al. (2011). Concise report: anti-​TNFα therapies and pregnancy: outcome of 130 pregnancies in the British Society for Rheumatology Biologics Register. Ann Rheum Dis, 70, 823–​6. Wallenius M, et al. (2011). Pregnancy and delivery in women with chronic inflammatory arthritides with a specific focus on first birth. Arthritis Rheum, 63, 1534–​42. Winger EE, Reed JL (2009). Was risk properly assessed in Carter, et al’s safety assessment of tumor necrosis factor antagonists during preg- nancy? J Rheumatol, 36, 2122; author reply 2123. Wolfberg AJ, et al. (2004). Association of rheumatologic disease with preeclampsia. Obstet Gynecol, 103, 1190–​3. Vasculitides Chakravarty EF, et al. (2011). Pregnancy after maternal exposure to rituximab. Blood, 117, 1499–​506. Diav-​Citrin O, et al. (2010). Pregnancy outcome after in utero exposure to colchicine. J Obstet Gynecol, 203, 144.e1–​6. Gatto M, et al. (2010). Pregnancy and vasculitis: a systemic review of the literature. Autoimmun Rev, 11, A447–​59. Hot A, et al. (2007). Marked improvement of Churg–​Strauss vascu- litis with intravenous gamma globulins during pregnancy. Clin Rheumatol, 26, 2149–​151. Ioscovich A, et al. (2008). Peripartum anesthetic management of pa- tients with Takayasu’s arteritis: case series and review. Int J Obstet Anesth, 17, 358–​64. Ishikawa K, Matsuura S (1982). Occlusive thromboaortopathy (Takayasu’s disease) and pregnancy:  clinical course and man- agement of 33 pregnancies and deliveries. Am J Cardiol, 50, 1293–​300. Iskender C, et  al. (2014). Behçet’s disease and pregnancy:  a retro- spective analysis of course of disease and pregnancy outcome. Obs Gyn Res, 40, 1598–​602. Koizumi M, et al. (2004). Schönlein–​Henoch purpura during preg- nancy: case report and review of the literature. J Obstet Gynecol Res, 30, 37–​41. Lakhi NA, Jones J (2012). Takayasu’s arteritis in pregnancy complicated by peripartum aortic dissection. Arch Gynecol Obstet, 282, 103–​6. Noel N, et al. (2013). Behçet’s disease and pregnancy. Arthritis Rheum, 65, 2450–​6. Seo P (2007). Pregnancy and vasculitis. Rheum Dis Clin North Am, 33, 299–​317. Soh MC, et al. (2009). Pregnancy complicating Wegener’s granulomatosis. Obstet Med,  2, 77–​80. Suri V, et al. (2010). Pregnancy and Takayasu’s: a single centre experi- ence from Northern India. J Obstet Gynaecol Res, 36, 519–​24. Scleroderma Bose N, Chiesa-​Vottero A, Chatterjee S (2015). Scleroderma renal crisis. Semin Arthritis Rheum, 44, 687–​94. Cassina M, et al. (2012). Pregnancy outcome in women exposed to leflunomide before or during pregnancy. Arthritis Rheum, 64, 2085–​94. Chambers CD, et al. (2010). Birth outcomes in women who have taken leflunomide during pregnancy. Arthritis Rheum, 62, 1494–​503. Chakravarty EF, et al. (2011). Pregnancy outcomes after maternal ex- posure to rituximab. Blood, 117, 1499–​506. Chung L, et al. (2006). Outcome of pregnancies complicated by sys- temic sclerosis and mixed connective tissue disease. Lupus, 15, 595–​9. Codullo V, et al. (2009). Serologic profile and mortality rates of sclero- derma renal crisis in Italy. J Rheumatol, 36, 1464–​9. Lidar M, et al. (2012). Pregnancy issues in scleroderma. Autoimmun Rev, 11, A515–​19. Nguyen B, et al. (2010). Association of RNA polymerase III antibodies with scleroderma renal crisis. J Rheumatol, 37, 1068; author reply 9. Steen VD (2007). Pregnancy in scleroderma. Rheum Dis Clin North Am, 33, 345–​58. Taraborelli M, et al. (2012). Brief report: successful pregnancies but a higher risk of preterm births in patients with systemic sclerosis: an Italian multicenter study. Arthritis Rheum, 64, 1970–​7. Antirheumatic drugs and immunosuppressive agents in pregnancy Anderka MT, et al. (2009). Reviewing the evidence for mycophenolate mofetil as a new teratogen: case report and review of the literature. Am J Med Genet A, 149A, 1241–​8. Arsenescu R, et al. (2011). TNF-​alpha and the development of the neo- natal immune system: implications for inhibitor use in pregnancy. Am J Gastroenterol, 106, 559–​62. Ben-​Horin S, et  al. (2010). Adalimumab level in breast milk of a nursing mother. Clin Gastroenterol Hepatol, 8, 475–​6. Ben-​Horin S, et al. (2011). Detection of infliximab in breast milk of nursing mothers with inflammatory bowel disease. J Crohns Colitis, 5, 555–​8. Berthelsen BG, et al. (2010). Etanercept concentrations in maternal serum, umbilical cord serum, breast milk and child serum during breastfeeding. Rheumatology, 49, 2225–​7. Chambers CD, et al. (2006). Human pregnancy safety for agents used to treat rheumatoid arthritis: adequacy of available information and strat- egies for developing post-​marketing data. Arthritis Res Ther, 8, 215–​24. Daniel S, et al. (2012). Major malformations following exposure to nonsteroidal antiinflammatory drugs during the first trimester of pregnancy. J Rheumatol, 39, 2163–​9. Flint J, et al. (2016). BSR and BHPR guideline on prescribing drugs in pregnancy and breastfeeding—​Part I: standard and biologic disease modifying anti-​rheumatic drugs and corticosteroids. Rheumatology (Oxford), 55, 1693–​7. Hviid A, Molgaard-​Nielsen D (2011). Corticosteroid use during preg- nancy and risk of orofacial clefts. CMAJ, 183, 796–​804. Hyrich KL, Verstappen SM (2013). Biologic therapies and preg- nancy: the story so far. Rheumatology, 53, 1377–​85. Kane S, et al. (2009). Absence of infliximab in infants and breast milk from nursing mothers receiving therapy for Crohn’s disease before and after delivery. J Clin Gastroenterol, 43, 613–​6. Koren G, et al. (2006). Nonsteroidal anti-​inflammatory drugs during third trimester and the risk of premature closure of the ductus arteriosus: a meta-​analysis. Ann Pharmacother, 40, 824–​9. Liew Z, et al. (2014). Acetaminophen use during pregnancy, behav- ioral problems, and hyperkinetic disorders. JAMA Pediatrics, 168, 313–​20. Lin AE, et  al. (2011). An additional patient with mycophenolate mofetil embryopathy: cardiac and facial analyses. Am J Med Genet A, 155A, 748–​56. Mahadevan U, et al. (2013). Placental transfer of anti-​tumor necrosis factor agents in pregnant patients with inflammatory bowel disease. Clin Gastroenterol Hepatol, 11, 286–​92. Mahadevan U, Terdiman JP, Church J (2007). Infliximab levels in infants born to women with inflammatory bowel disease. Gastroenterology, 132 (Suppl. 1), A144. Mahadevan U, Martin CF, Sandler RS (2012). PIANO: a 1000 patient prospective registry of pregnancy outcomes in women with IBD ex- posed to immunomodulators and biologic therapy. Gastroenterology, 142(Suppl 1), S149. Martinez Lopez J, et  al. (2009). Systematic review of the safety of methotrexate in rheumatoid arthritis regarding the reproductive

Section 14  Medical disorders in pregnancy 2670 system (fertility, pregnancy and breastfeeding). Clin Exp Rheumatol, 27, 678–​84. Mendonca LLF, et al. (2000). Non-​steroidal anti-​inflammatory drugs as a possible cause for reversible infertility. Rheumatology (Oxford), 39, 880–​2. Murashima A, et al. (2008). Etanercept during pregnancy and lacta- tion in a patient with rheumatoid arthritis: drug levels in maternal serum, cord blood, breast milk and the infant’s serum. Ann Rheum Dis, 68, 1793–​4. Østensen M (2009). Are TNF inhibitors safe in pregnancy? Nat Rev Rheumatol, 5, 184–​5. Østensen M, Eigenmann GO (2004). Etanercept in breast milk. J Rheumatol, 31, 1017–​8 Perez-​Aytes A, et  al. (2008). In utero exposure to mycophenolate mofetil: a characteristic phenotype? Am J Med Genet A, 146A, 1–​7. Soh MC, Mackillop L (2016). Biologics in pregnancy—​for the obstetri- cian. The Obstetrician and Gynaecologist, 18, 25–​32. Soh MC, Nelson-​Piercy C (2012). Update of the management of rheuma- toid arthritis in pregnancy. Expert Rev Obstet Gynecol, 7, 77–​96. Soh MC, Nelson-​Piercy C (2015). Comment on: high-​risk pregnancy and the rheumatologist: reply. Rheumatology, 54, 2293. Thorne JC, et al. (2014). Methotrexate use in a breastfeeding patient with rheumatoid arthritis. J Rheumatol, 41, 2332. Weber-​Schoendorfer C, et al. (2014). Pregnancy outcome after metho- trexate treatment for rheumatic disease prior to or during early preg- nancy: a prospective multicenter cohort study. Arthritis Rheum, 66, 1101–​10. Weber-​Schoendorfer C, et al. (2014). No evidence for an increased risk of adverse pregnancy outcome after paternal low-​dose metho- trexate: an observational cohort study. Rheumatology, 53, 757–​63. Winger EE, et al. (2009). Treatment with adalimumab (Humira) and intravenous immunoglobulin improves pregnancy rates in women undergoing IVF. Am J Reprod Immunol, 61, 113–​20. Winger EE, et al. (2011). Degree of TNF-​alpha/​IL-​10 cytokine eleva- tion correlates with IVF success rates in women undergoing treat- ment with adalimumab (Humira) and IVIG. Am J Reprod Immunol, 65, 610–​8. Wolf D, Mahadevan U (2010). Certolizumab pegol use in preg- nancy:  low levels detected in cord blood. American College of Rheumatology (ACR). Arthritis Rheum. Abstract 718. Zelinkova Z, et al. (2011). High intra-​uterine exposure to infliximab following maternal anti-​TNF treatment during pregnancy. Aliment Pharmacol Ther, 33, 1053–​8. Zelinkova Z, et al. (2013). Effects of discontinuing anti-​tumor necrosis factor therapy during pregnancy on the course of inflammatory bowel disease and neonatal exposure. Clin Gastroenterol Hepatol, 11, 318–​21. Future maternal health Bellamy L, et al. (2007). Pre-​eclampsia and risk of cardiovascular dis- ease and cancer in later life: systematic review and meta-​analysis. BMJ, 335, 974. Magnussen EB, et al. (2007). Prepregnancy cardiovascular risk factors as predictors of pre-​eclampsia: population-​based cohort study. BMJ, 335, 978. Magnussen EB, et al. (2009). Hypertensive disorders in pregnancy and subsequently measured cardiovascular risk factors. Obstet Gynecol, 114, 961–​70. Romundstad PR, et al. (2010). Hypertension in pregnancy and later cardiovascular risk: common antecedents? Circulation, 122, 579–​84. Smith GN, et al. (2012). Ten-​year, thirty-​year, and lifetime cardiovas- cular disease risk estimates following a pregnancy complicated by preeclampsia. J Obstet Gynaecol Can, 34, 830–​5. Smith GC, Pell JP, Walsh D (2001). Pregnancy complications and ma- ternal risk of ischaemic heart disease: a retrospective cohort study of 129,290 births. Lancet, 357, 2002–​6. Soh MC, et  al. (2015). Accelerated development of cardiovascular events in Swedish women with systemic lupus erythematosus—​does pregnancy and its outcomes play a role? A population-​based retro- spective study. Ann Rheum Dis, 7(Suppl 2), 85–​6. Soh MC, et al. (2015). Association between pregnancy outcomes and death from cardiovascular causes in parous women with systemic lupus erythematosus: a study using Swedish population registries. Arthritis Rheum, 67, 2376–​83. Soh MC, et al. (2016). Maternal-​placental syndrome and future risk of accelerated cardiovascular events in parous Swedish women with systemic lupus erythematosus—​population-​based retrospective co- hort study with a time-​to-​event analysis. Rheumatology, 55, 1235–​42. Williams D (2003). Pregnancy: a stress test for life. Curr Opin O&G, 15, 465–​71.