15 Andrew McCarthy1 and May Ching Soh1,2,3 1 Queen Charlotte’s and Chelsea Hospital, Imperial College Healthcare NHS Trust, London, UK 2 Women’s Health Academic Centre, King’s College London, London, UK 3 John Radcliffe Hospital, Oxford University Hospitals NHS Foundation Trust, Oxford, UK The 2009–2012 MBRRACE (Mothers and Babies: Reducing Risk through Audits and Confidential Enquiries across the UK) report has shown a 48% decrease in direct maternal deaths, yet there has not been a significant decrease in indirect maternal deaths [1]. The rate of indirect maternal deaths is now twice that of death from direct causes (6.87 per 100 000 vs. 3.25 per 100 000). Two‐thirds of women died from medical or mental health problems in pregnancy, of whom 74% had pre‐existing medical (or mental health) problems prior to pregnancy. With more successsful artificial reproductive therapies, a woman’s childbearing years have been greatly extended; some women are facing motherhood at a much older age, thereby increasing the risk of the pregnancy being complicated by coincidental medical conditions. In the UK the latest figures collected through 2009–2012 reveals there were 3 182 873 deliveries of which 3.9% were to women aged 40 years and older; 27% of women who died were obese (body mass index ≥30 kg/m2) and one‐quarter of them smoked. This reflects a major shift, thereby increasing risk of morbidity from medical disorders, and possibly introducing a contributory factor towards indirect deaths in the maternal mortality enquiry. While many medical conditions in this age group do not result in serious morbidity, others have the potential to do so (e.g. epilepsy, asthma, autoimmune rheumatic diseases and infections). It is important that women receive good advice prior to pregnancy about the potential implications of their medical condition, and enter pregnancy with appropriate confidence about their medication or have specific management plans to alter treatment in early pregnancy. This necessitates that they have ready access to specialist advice once they become pregnant. With more data emerging on the safety of drugs in pregnancy, clinicians no longer have to compromise the effectiveness of medical treatment for long‐term conditions. However, changing the mindset of women and other clinicians who are less familiar with pregnancy remains a challenge. It would be ideal if these issues were addressed prior to pregnancy; examples include the continuation of pregnancy‐friendly immunosuppressants or biologics in pregnancy and the safety profile of most antiepileptic drugs in pregnancy. This is so that women will be appropriately reassured and will not abruptly discontinue their medications in early pregnancy. Hence the importance of pre‐conception counselling cannot be sufficiently emphasized, and was further highlighted in the 2014 report on Saving Lives, Improving Mothers’ Care [1]. Management of women with medical disorders is often best coordinated within clinics, with obstetric, medical and midwifery input available. Such clinics make outpatient management much more convenient for the woman, and facilitate good communication between the relevant medical teams. They also serve as a focal point with which the woman may make contact in early pregnancy when treatment changes may need to take place without delay or in later pregnancy if there are problems. Integrated care plans for women with medical disorders should be made. Within units, consideration needs to be given to how such cross‐specialty communication occurs if there is no formal multidisciplinary meeting where high‐risk cases are discussed. The role of the midwife and support workers cannot be emphasized enough in ensuring that there is a more holistic approach to the care provided. Systemic lupus erythematosus (SLE) is the commonest multisystem autoimmune rheumatic disorder affecting women of childbearing age. Prevalence depends on ethnicity, with more severe phenotypes seen in people of African, Far East (Chinese, Korean) and Hispanic extraction. Approximately 30% will have overlap with antiphospholipid syndrome (APS) and are classified as secondary APS. Fertility is unimpaired in women with SLE unless they have had prior significant cyclophosphamide exposure, secondary APS or have current severe flare. Pregnancy outcomes are worse if conception occurs within 4–6 months of a flare, so planning pregnancy is of paramount importance. Contraception should be discussed with these women. The combined oral contraceptive pill is not associated with an increased risk of flares or SLE disease activity, but is contraindicated in those with secondary APS. Pre‐conception counselling enables any potentially teratogenic drugs (e.g. cyclophosphamide and mycophenolate mofetil) to be switched to more pregnancy‐compatible substitutes while the woman’s disease is in remission. Folic acid 5 mg should be commenced at least 3 months prior, vitamin D sufficiency confirmed, and hydroxychloroquine (HCQ) must be continued as observational studies have shown improved obstetric outcomes with reduction of flares when there is uninterrupted HCQ use in pregnancy. Aspirin 75 mg and calcium carbonate 2.5 g daily could be considered in women with SLE as their risk of pre‐eclampsia is significantly higher than the normal population. Good obstetric outcomes are associated with disease that is in sustained remission for at least 4–6 months prior to conception, ongoing HCQ use in pregnancy and lack of major organ involvement (e.g. lupus nephritis, neuropsychiatric lupus) [2]. Flares of SLE are not more common in pregnancy. The flares that occur in pregnancy generally follow a similar pattern to previously: those with predominantly mucocutaneous disease before pregnancy will flare in a similar fashion, whereas those with lupus nephritis will experience renal involvement. De novo presentation with lupus nephritis can occur and may be mistaken for pre‐eclampsia if it occurs after 20 weeks’ gestation. Women with SLE have an increased risk of adverse pregnancy outcomes from placental insufficiency. Up to one‐quarter of women with SLE develop pre‐eclampsia or eclampsia, and 20–31% have preterm delivery. Rates of fetal growth restriction (FGR) range from 5 to 23% in women with SLE. The risk of deterioration of renal function in pregnancy from a flare of lupus nephritis or pre‐eclampsia is higher if the patient is hypertensive or has pre‐existing heavy proteinuria or an elevated serum creatinine at baseline. A meta‐analysis reported that the incidence of renal lupus flares during pregnancy was 11–69% and renal impairment occurred in 3–27%, which was permanent in up to 10% [3]. One case–control study showed that 28% of patients with class III or IV lupus nephritis developed pre‐eclampsia, of whom 35% had a preterm delivery and a significantly lower birthweight compared to the women with SLE without nephritis [4]. Early studies in pregnant women with SLE have demonstrated high circulating levels of soluble fms‐like tyrosine kinase‐1 (sFlt‐1). Similar to the general population, higher circulating levels of sFlt‐1 and soluble endoglin (sEng), lower placental growth factor (PGF) and a high sFlt‐1/PGF ratio assist in the prediction of pre‐eclampsia in SLE pregnancies. In the multicentre Predictors of Pregnancy Outcome: Biomarkers in Anti‐phospholipid Syndrome and SLE (PROMISSE) study, low sFlt levels (<1872 pg/mL) and elevated PGF (>70.3 pg/mL) have a negative predictive value in women with SLE, with less than 5% of the cohort developing adverse pregnancy outcomes [5]. Markers for SLE disease activity are less useful in pregnancy as levels of complement C3 and C4 and erythrocyte sedimentation rate (ESR) all rise in a normal pregnancy; therefore, a relative drop of C3 and C4 from the baseline established in early pregnancy could be a useful marker if previously quantified. Table 15.1 includes helpful tips to differentiate a flare of SLE (especially lupus nephritis) from pre‐eclampsia. Table 15.1 Useful features for differentiating pre‐eclampsia from a flare of SLE or lupus nephritis. * PGF can be raised in a flare of SLE. +, present; –, absent; ↑, raised; ↓, decreased; ALT, alanine aminotransferase; ESR, erythrocyte sedimentation rate; PGF, placental growth factor. The management of a flare of SLE depends on the organ systems involved. Prevention is best, and therefore HCQ should be continued in pregnancy. Steroids remain the mainstay of treatment of a flare, but should never be empirically started for the prevention of a flare. Infusions of immunoglobulins which ‘mop up’ circulating antibodies can be useful for the treatment of a moderate to severe flare; their use must be accompanied by thromboprophylaxis as they increase the risk of thrombosis. Non‐steroidal anti‐inflammatory drugs (NSAIDs) are safe and non‐teratogenic but should only be used up to 32 weeks’ gestation (and not within 48 hours of imminent delivery) due to a reversible effect on the neonate’s patent ductus arteriosus, resulting in neonatal pulmonary hypertension. Azathioprine has a good safety record in pregnancy. There is increasing evidence for the efficacy of tacrolimus for the treatment of lupus nephritis in pregnancy. On the other hand, mycophenolate mofetil and cyclophosphamide are teratogenic and should never be used in the first 12 weeks of pregnancy. However, if the maternal flare is severe enough to necessitate the use of cytotoxics, cyclophosphamide has been used after the initial period of organogenesis. Long‐term follow‐up studies of the offspring exposed to cyclophosphamide have shown normal growth and neurodevelopment when followed up until their adolescent years [6]. Rituximab (a B‐cell‐depleting agent) is not teratogenic but it has a very long half‐life and will suppress the neonate’s B‐cell production for 6–8 months after exposure. Nevertheless, there have not been any adverse effects seen in the exposed offspring, who were able to mount a normal immunological response to vaccination, nor did they have an increased rate of infections. Nevertheless, the number of children exposed to rituximab in utero remains small, and there are more cumulative long‐term data on cyclophosphamide in pregnancy compared with rituximab. Up to 30% of patients with SLE have Ro antibodies. Ro and La antibodies are part of the extractable nuclear antigens that cross the placenta and potentially damage the neonate’s cardiac conduction system resulting in irreversible fibrosis. Ro antibodies are present in 90–100% of mothers of affected offspring, and 68–91% have La antibodies. The risk of congenital heart block is approximately 2% in Ro‐positive mothers and is strongly correlated to the titre of Ro antibodies present in the mother. If affected, the overall mortality is around 20%, with deaths usually occurring in utero (after developing hydrops and pleural and pericardial effusions). Most infants who survive the neonatal period do well, although two‐thirds require pacemakers. The risk of congenital heart block increases exponentially to 15–50% in subsequent pregnancies if there have been previously affected offspring. Detection of heart block can occur from 18 weeks onwards. However, once in second‐degree heart block, the process is almost invariably irreversible, and the treatment would be delivery – if viable – with the implantation of a permanent pacemaker in the neonate. Heart block rarely develops after 28 weeks’ gestation, though there are rare instances of heart block developing up to 3 years post partum. Hence most screening programmes for mothers who are Ro or La antibody positive start at 18 weeks and if the second echocardiogram (to determine the PR interval of the neonate) is normal at 28–32 weeks, then no further screening is needed until an ECG at delivery. The treatment of heart block remains controversial, since none of the therapies previously tried have been effective in reversing second‐degree heart block. HCQ has been shown to reduce the incidence of heart block and could be considered even in asymptomatic women who have Ro or La antibodies [7]. High‐dose dexamethasone and repeated infusions of immunoglobulins have not been efficacious in reversing established second‐degree heart block. Neonatal cutaneous lupus is an erythematous annular rash, and develops soon after exposure to sun or ultraviolet light. It affects approximately 5% of offspring born to mothers who are Ro or La positive. The rash typically resolves within 6–8 months, with clearance of maternal Ro and La antibodies. Treatment with topical corticosteroids is reserved for severe cases as the rash usually clears spontaneously with only residual hypopigmentation or telangiectasias persisting for up to 2 years. In the normal population, antiphospholipid antibodies are often transient and can be induced by a variety of factors. Medium‐to‐high titres of antiphospholipid antibodies do not correlate with adverse obstetric outcomes in the absence of accompanying clinical features of recurrent (and consecutive) spontaneous miscarriages under 10 weeks, unexplained mid‐trimester losses, preterm delivery before 34 weeks as a result of placental insufficiency or previous thrombotic phenomena [8]. Up to 50% of women with SLE will have circulating antiphospholipid antibodies. Lupus anticoagulant is a much stronger predictor of poor outcomes than anti‐cardiolipin antibodies. Women with secondary APS who have experienced thrombosis, late losses and features of placental insufficiency should be offered prophylactic low‐molecular‐weight heparin (LMWH), which should be continued for at least 6 weeks post partum. The use of LMWH to prevent recurrent early pregnancy losses remains controversial; a meta‐analysis concluded that LMWH has not demonstrated additional benefit compared with low‐dose aspirin [9]. Women with SLE have a much higher risk of premature cardiovascular disease and death compared with the normal population, even in the relative absence of any cardiovascular risk factors. Women with SLE who have had their pregnancies complicated by placental insufficiency are at greater risk of death from cardiovascular causes and accelerated development of cardiovascular disease. Preterm delivery before 34 weeks, which could be a surrogate marker for active SLE in pregnancy, has also been associated with accelerated cardiovascular events in this cohort [10]. The adage that rheumatoid arthritis (RA) improves in pregnancy no longer holds true since prospective population‐based studies have indicated active disease in at least 40% of women. Those with more severe disease (erosive disease or joint deformities) and anti‐cyclic citrullinated peptide (anti‐CCP) antibodies often flare more severely than those with seronegative RA. Seronegative spondyloarthropathies, such as psoriatic arthritis, enteropathic arthritis and ankylosing spondylitis, are less responsive to conventional disease‐modifying antirheumatic drugs (DMARDs, e.g. HCQ, methotrexate and sulfasalazine) or prednisolone to treat flares. Though women with RA are less fertile, they have similar obstetric outcomes to the normal population, unless they experience active disease in pregnancy. A nationwide prospective study in the Netherlands has shown that active disease in pregnancy is associated with a lower gestational age at delivery and a lower birthweight. Studies from the USA indicate that women with RA have an increased risk of FGR, particularly if they have severe erosive disease. There was also a trend towards an increased risk of developing pre‐eclampsia, though the absolute risk remains small when compared to women with SLE. There is a higher rate of elective caesarean section in women with RA compared to the normal population, which is most likely iatrogenic. Limitation in hip abduction is rarely severe enough to impede vaginal delivery. Additional care is necessary in women with juvenile idiopathic arthritis who may have had hip joint replacements. Women with inflammatory arthritides should be reviewed by an obstetric anaesthetist to assess cervical spine involvement and degree of jaw excursion in order to anticipate any problems should they require a general anaesthetic in labour. Atlantoaxial subluxation and subsequent paralysis is a rare complication that can occur if a general anaesthetic is administered without due precautions in a woman with erosive disease affecting her cervical spine. DMARDs are likely to have contributed to the reduced fecundity seen in women with inflammatory arthritides. Methotrexate, previously used as first‐line therapy in all patients with RA, is a well‐known teratogen. It must be discontinued at least 3 months prior to any attempt at conception and high‐dose (5 mg) folic acid should be prescribed. Leflunomide, a pyrimidine synthesis inhibitor, is another commonly used DMARD that has teratogenic potential. Cholestyramine 8 mg for at least 11 days ‘washout’ should be given prior to pregnancy, but this can also be commenced in very early pregnancy if necessary. There should be confirmation of successful cholestyramine washout with undetectable drug levels (<0.03 µg/mL) taken 2 weeks apart. However, if this is not possible, then ongoing washout with cholestyramine should continue until organogenesis is completed. Table 15.2 lists a number of DMARDs and other drugs and their suitability in pregnancy and lactation. Table 15.2 Commonly used drugs used for the treatment of rheumatic diseases and their effect on pregnancy and breastfeeding. Source: adapted from Soh & Nelson‐Piercy [13]. * OMENS: O, orbital distortion; M, mandibular hypoplasia; E, ear anomalies; N, seventh cranial nerve root involvement; S, soft tissue deficiency. † Azathioprine is converted to the active metabolite 6‐thioguanine nucleotides in 15 min but the half‐life of the active metabolite in erythrocytes is weeks to months. ✓, safe for breastfeeding; ×, unsafe or not recommended for breastfeeding. ‘Biologics’ are modified part murine and human monoclonal antibodies that act on targeted receptors to reduce the production of various inflammatory molecules. They are increasingly used in women with RA, seronegative spondyloarthropathies and inflammatory bowel disease. At present, they are reserved for patients who have had refractory disease and ‘failed’ conventional therapy. Therefore, by default, these women have more severe disease. The most commonly used drugs at present are the anti‐tumour necrosis factor (TNF) agents, usually infliximab, etanercept, adalimumab, golimumab and certolizumab, and the B‐cell depleting agents rituximab and belimumab. Multiple registries dedicated to studying the effects of these drugs in pregnancy have not demonstrated an increased teratogenic risk. The cases of fetal malformations are sporadic and not part of a syndrome that would suggest interference with organogenesis. Therefore, it is no longer recommended that anti‐TNF agents be discontinued prior to pregnancy. As monoclonal antibodies, these drugs do not cross the placenta (except by slow diffusion in early pregnancy) until after 16 weeks’ gestation when syncytiotrophoblasts develop receptors that will bind with the Fc portion of these modified monoclonal antibodies. Transplacental transfer of these drugs increases exponentially from around 22 weeks to its maximal efficiency around 28 weeks’ gestation. The efficiency of transfer depends on the half‐life of the drug and its Fc portion. Cord blood levels of biologics with a long half‐life and an Fc receptor that is conducive to transplacental transfer (e.g. adalimumab and infliximab) are often higher than maternal serum levels if the drug is administered towards the end of pregnancy [11]. There is significant debate surrounding the timing of drug cessation to avoid transplacental transfer, with the European League Against Rheumatisim (EULAR) [12] suggesting that infliximab and adalimumab be discontinued at 20 weeks, whereas etanercept (with its modified Fc portion and short half‐life) be discontinued at 30–32 weeks; on the other hand, the British Society of Rheumatology [13] recommends that adalimumab and etanercept can be continued until the end of the second trimester and infliximab stopped at 16 weeks’ gestation. However, experts are unanimous that certolizumab, with its large pegylated molecule, only crosses the placenta via slow diffusion and therefore can be continued throughout pregnancy. The predominant concern with anti‐TNF use in pregnancy beyond the recommended gestation is the risk of significant immunosuppression in the newborn. The immature reticuloendothelial system may take 6–8 months to clear the drug, and is dependent on the gestation at which the anti‐TNF was discontinued. Therefore, neonates who are exposed in utero beyond the recommended gestation should be promptly assessed if there is a suspicion of infection. Live vaccinations (usually rotavirus and BCG) should be avoided for the first 6–7 months of life. The woman, her primary caregiver and paediatric teams should be informed prior to delivery so that the necessary precautions can be undertaken. Rituximab has an exceedingly long half‐life but is not associated with any teratogenesis. However, it is associated with a high rate of miscarriage that is most likely due to underlying maternal disease for which this drug (and likely other drugs as well) is prescribed. EULAR guidelines suggest that, if clinically indicated, it could be continued throughout pregnancy. Transient cytopenias (from B‐cell depletion) may occur in the offspring exposed in utero; however, neonatal infections or inability to mount a response to vaccination has not been a problem. The decision to continue any biologic needs to be individualized to the woman, bearing in mind the refractory nature of the underlying disease that has not been amenable to other therapies. If there is a significant risk of flares, with associated maternal (and fetal) morbidity from a flare, then the option to continue these drugs can be discussed, given the caveats that suspected infections in the newborn are promptly assessed and the avoidance of live vaccinations for the first 6–7 months of life. The obstetrician should be aware that women on biologics are significantly immunosuppressed and they should also be carefully screened for infections at each visit. Infections may present atypically (i.e. absence of fever or tachycardia) as these women are unable to mount an adequate immunological response to pathogens. Atypical infections should also be considered. In the longer term, patients on biologics have an increased risk of malignancy, a differential that should be considered if a woman presents unwell [11].
Autoimmune Rheumatic Diseases and Other Medical Disorders in Pregnancy
Background
Systemic lupus erythematosus
Effect of SLE on pregnancy
Differentiating a flare of SLE from pre‐eclampsia
SLE flare/lupus nephritis
Pre‐eclampsia
Hypertension
+
++
Other manifestations of SLE (e.g. rash or oral ulcers)
+
–
Complement (C3, C4)
↓ from baseline (in early pregnancy)
Unchanged
ESR
↑
Unchanged
Double‐stranded DNA titres
↑
Unchanged
Proteinuria ≥0.3 g per 24 hours
++
+
Casts
May be present
–
Serum urate
May be ↑
↑↑
Liver function tests
Unchanged
ALT may be ↑
sFlt‐1/PGF* ratio
↑
↑
Managing a flare of SLE in pregnancy
Congenital heart block and neonatal lupus syndromes
Secondary antiphospholipid syndrome
Long‐term maternal health and cardiovascular disease
Rheumatoid arthritis and other inflammatory arthritides
Precautions and special considerations in women with inflammatory arthritides
DMARDs and other drugs in pregnancy and breastfeeding
Drug
Effects on organogenesis
Effects on fetus/neonate
Breastfeeding
Authors’ recommendations on use in pregnancy
NSAIDs
None
Reversible constriction of ductus arteriosus after 27/40; oligohydramnios; transient anuria and renal failure if used <48hours prior to delivery
✓
Likely a class effect for all NSAIDs. Use if indicated at lowest dose possible till 32 weeks’ gestation and stop if delivery is imminent (within 48 hours)
Cyclooxygenase (COX)‐2 inhibitors
Likely no increase in anomalies in humans
Oligohydramnios is less profound and preterm closure of the ductus arteriosus is delayed compared with NSAID use
✓ for celecoxib
As more data are available on NSAIDs, it would be safer to change to an NSAID in early pregnancy. However, there could be a role for selective COX‐2 inhibitor use in women requiring NSAIDs >32 weeks
Prednisone/prednisolone
None
More than 80% metabolized by placenta and fetus inactivates steroids by hepatic conjugation, and therefore receives <10% of non‐fluorinated steroid dose. Theoretical effects in doses in excess of 80 mg: possible cataracts, adrenal insufficiency and infection
✓
The effects are predominantly maternal with immunosuppression with ascending infections, preterm rupture of membranes, increasing risk of gestational diabetes and hypertension. Use lowest dose possible; ensure there is a plan to taper the dose. If disease persistently active, then consider addition of DMARD/biologic to ensure woman does not remain on prolonged courses of high‐dose steroids
Hydroxychloroquine
None
None
✓
Continue in pregnancy and breastfeeding
Sulfasalazine (and other 5‐aminosalicylic acid compounds)
None
None
✓ except in very preterm jaundiced neonates
Commence folic acid supplementation 5 mg/day 3 months prior to pregnancy. In men with fertility problems, it may need to be swapped with another agent as it affects spermatogenesis and motility
Methotrexate (MTX)
Aminopterin syndrome; high rate of pregnancy loss; 15% rate of congenital anomalies
If no congenital anomalies, long‐term follow‐up of children exposed to MTX did not reveal any problems
×
Reliable contraception advised. Discontinue ≥3 months prior to pregnancy with daily 5‐mg 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. No effect if paternal MTX use
Leflunomide
In animal studies, malformations of the head, rump, vertebral column and limb defects. Increased rate of miscarriage
If pregnancy continues, no major structural anomalies noted especialy after cholestyramine washout in very early pregnancy
×
Reliable contraception advised. Wash out with cholestyramine 8 g t.d.s. for 11 days; repeat until drug levels <0.03 µg/mL taken 2 weeks apart (or continue until 12 weeks’ gestation). If exposed in early pregnancy, offer washout and reassure woman that, to date, birth outcomes of exposed women no different from disease‐matched controls
Azathioprine*
None
None
✓
Continue in pregnancy and lactation
Cyclosporin
None
Transient immune alterations in the neonate
✓
Continue in pregnancy; probably safe in breastfeeding though wide range of concentrations excreted in breast milk
Tacrolimus
None
None
✓
Continue in pregnancy and safe in breastfeeding
Intravenous immunoglobulin
None
None
✓
Concurrent thromboprophylaxis is advised
Cyclophosphamide
Cyclophosphamide embryopathy with high rate of miscarriage
Transient cytopenias. No long‐term effect on the neonate if survives pregnancy
×
Use only if there is life‐threatening maternal disease after the first trimester. If maternal disease necessitates cyclophosphamide in first trimester, discuss termination
Mycophenolate mofetil
OMENS† and congenital cardiac defects. Phenotype is not dose dependent
Neonates described in the literature, had also been exposed in the period of organogenesis
×
Discontinue for at least 3 months prior to pregnancy
Use of anti‐TNF and other biologics
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
