Evans syndrome in pregnancy and the postpartum period: A retrospective cohort study

Pregnancy triggers changes in immune function, balancing maintaining maternal–fetal tolerance with preserving antigenic response to avoid infection.1, 2 Immune modulations during pregnancy may contribute to the development or exacerbation of autoimmune cytopenias (AIC), including immune thrombocytopenia (ITP) and warm autoimmune haemolytic anaemia (wAIHA).3-5 Approximately 50% of wAIHA patients experience disease exacerbation during the perinatal period.3 While patients with ITP may not have an increased risk of severe exacerbation, around 40% receive therapy in preparation for delivery.5 Most cases of pregnancy-associated ITP and wAIHA can be managed with corticosteroids and/or intravenous immunoglobulin (IVIg) and maternal and fetal outcomes are generally favourable.3-5 Evans syndrome (ES) is a rare autoimmune disorder defined by the concurrent or sequential development of at least two of three AIC: AIHA, ITP and autoimmune neutropenia (AIN).6 Outside of pregnancy, ES features higher mortality and disease-related complications compared to isolated ITP or AIHA.7 Information on pregnancy and ES is limited to case reports.3, 8 Whether ES patients face increased disease relapse and complications in pregnancy remains unknown. This single-centre, retrospective study evaluated clinical characteristics, management and pregnancy and postpartum outcomes of 11 ES patients identified by search of the Mass General Brigham system,9 who experienced pregnancy between 1995 and 2024 (details in Data S1). Among 30 pregnancies, mean age at conception was 30.1 ± 5.3 years. All 11 patients had ITP and wAIHA; 2 (18.2%) also had AIN (Table S1). Of eight with secondary ES (72.7%), six had systemic lupus erythematosus (SLE), including two with antiphospholipid syndrome (APS), one had CLCN7-related osteopetrosis, an inborn error of immunity and one had congenital human immunodeficiency virus (HIV). In addition to the two APS patients, five patients had a positive antiphospholipid antibody (aPL) test without a diagnosis (Table S2). Four pregnancies were conceived on immunomodulatory therapy (Table S3). One continued low-dose prednisone, azathioprine and hydroxychloroquine for her first pregnancy and had a second with hydroxychloroquine monotherapy. One stopped mycophenolate mofetil once pregnancy was confirmed. The last patient took azathioprine throughout pregnancy for wAIHA. Four patients, including both APS patients, used thromboprophylaxis for five pregnancies with aspirin (n = 1), unfractionated or low molecular weight heparin (n = 3) or both (n = 1). Among 30 pregnancies, there were 19 live births (63.3%), two ectopic pregnancies, two elective terminations and seven miscarriages (23.3%), two of which (28.6%) occurred after 12 weeks gestation (Figure 1). Miscarriages were more common among primary ES patients (5 vs. 2), but three of the five primary ES miscarriages were early losses after in vitro fertilization. There was no difference in miscarriage rates for pregnancies before or after ES diagnosis, after age 35, or with a history of aPL positivity, use of thromboprophylaxis or use of immunomodulatory therapy at conception. One pregnancy was terminated given advanced maternal age (38 years) and underlying SLE, APS and ES. One other termination was excluded from further analysis due to a lack of information. Among 29 pregnancies with available data, 9 (31%) were complicated by active AIC: five wAIHA and ITP, three ITP and one wAIHA (Figure 1). These cases required initiation of therapy to address AIC; none were treated to target a desired count for delivery. Of 20 pregnancies that occurred after the onset of ES, seven developed active disease (35% relapse rate). Four relapses occurred in the second trimester. One patient had three pregnancies with relapse of wAIHA and ITP within 2 months postpartum. Perinatal patients who relapsed had a shorter duration of ES (3.4 ± 3.8 vs. 13.1 ± 6.5 years, p = 0.006). There was no difference between patients with primary versus secondary disease. Two patients presented with de novo AIC during pregnancy. The first had SLE, presented with ITP at 16 weeks of gestation and developed wAIHA 13 years later. The other presented with concurrent primary ITP and wAIHA at 25 weeks of gestation without recurrence of ES with 4.3 years of follow-up. Eight of nine (88.9%) perinatal cases were managed with prednisone; details were available for seven. All started with 60 mg/day (0.75–1 mg/kg/day) and all demonstrated complete response (Table S3) with an average time to response of 18.1 ± 14.8 days for ITP and 31.2 ± 14.8 days for wAIHA. Average duration on prednisone was 27.6 ± 8.8 weeks. The patient not receiving prednisone took azathioprine throughout pregnancy for active wAIHA; she was hospitalized for exacerbation of wAIHA triggered by viral respiratory infection at 13 weeks of gestation. She received a partial dose of IVIg, which was halted for allergic reaction. Corticosteroids were avoided given prior mood disturbance, so she received two red blood cell (RBC) transfusions over a week and continued azathioprine with stable disease. The patient diagnosed with de novo ITP and wAIHA during pregnancy was hospitalized and required 17 units of RBC transfusion during initial presentation until attaining response to corticosteroids and IVIg. The remaining seven cases of perinatal ES did not require transfusion or hospitalization. Among six pregnancies complicated by antepartum AIC, four resolved at delivery and one within 2 months postpartum. The patient on azathioprine throughout pregnancy had ongoing haemolysis beyond the postpartum period. One patient with three postpartum ES relapses required 7–8 months of prednisone and could not continue breastfeeding due to persistent bleeding with each relapse. Of 19 pregnancies carried to delivery, 5 (26.3%) delivered before 37 weeks of gestation (Table 1). The majority delivered by caesarean section (C-section; 11/19, 57.9%) and most received spinal analgesia (16/19, 84.2%). Maternal complications (other than C-section) occurred in 12/19 pregnancies (63.2%), all of which occurred after ES onset. More than half of ES pregnancies were complicated by hypertensive disorders (gestational hypertension, pre-eclampsia or eclampsia) and one-third by postpartum haemorrhage (PPH), including two requiring RBC transfusions. Maternal complication rates were not associated with active AIC, pre-existing ES, comorbid SLE, aPL positivity, use of thromboprophylaxis or maternal age over 35. Two patients developed postpartum infection, including one Grade 4 endometritis with febrile neutropenia (absolute neutrophil count 0) managed with intravenous antibiotics without granulocyte colony stimulating factor. This patient was not receiving immunomodulatory therapy during pregnancy. No thrombotic complications occurred. Both patients with APS and patients with ongoing wAIHA on azathioprine received heparin thromboprophylaxis ± aspirin throughout pregnancy. One 28-week neonate died at day 2 of life from congenital anomalies. His twin had respiratory failure and a spontaneous intraventricular haemorrhage at 30 days (platelet count over 300 × 109/L). Blood counts were available for 15 neonates (Table S4): Four had mild neutropenia; none had anaemia or thrombocytopenia. Three of six neonates (50%) had a positive DAT at birth (specificity not recorded). In patients with underlying autoimmune disorders, perinatal alterations in immune function (particularly T-cell-mediated response) may propagate the risk of disease exacerbation, miscarriage and maternal complications like pre-eclampsia and placental abruption.1, 2 Relative to isolated AIC, paediatric and young adult ES is characterized by more significant T-cell dysregulation even among those without an underlying inborn error of immunity10 and non-pregnant patients with ES have a higher risk of mortality and adverse outcomes.7 In ES pregnancies, maternal complications arose in two-thirds carried to delivery. The most common were hypertensive disorders of pregnancy, PPH and preterm delivery. All maternal complications excluding C-section occurred in pregnancies after the onset of ES, while acknowledging the small sample size of the cohort. The observed frequency of composite maternal complications in ES is about two to three times higher than in isolated ITP and AIHA, where the observed rate is approximately 20%–25%.3, 5 Compared to the general population, pregnant SLE patients have higher rates of maternal mortality, preterm labour and hypertensive disorders.11 The high prevalence of underlying SLE in this cohort may have contributed to adverse pregnancy outcomes, although a significant association between SLE status and ES relapse or adverse maternal outcomes was not seen. Relapse occurred in 35% of pregnancies. Reassuringly, perinatal ES was consistently responsive to prednisone and no patient required a second-line agent. The observed time to response was longer than the ~14 days reported in non-pregnant ITP12 and pregnancy-associated3 and non-pregnancy-associated13 AIHA. Larger numbers are required to determine whether these findings are representative of perinatal ES. Approximately one quarter of pregnancies resulted in miscarriage, slightly higher than the estimated 15% in the general population and similar to SLE, where rates of loss are estimated at 15%–20%14 and up to 30% for those with concurrent aPL positivity.15 Prior studies in isolated AIHA and ITP reported miscarriage rates of 13%3 and 5%5 respectively. This is the first observational study of pregnancy outcomes in ES patients. Prior experience is primarily limited to case reports,8 which may be subject to reporting bias. The major limitations are the small sample size and retrospective design. Relapse and complication rates may be overestimated as patients with milder disease may not have been identified by database search. Without a matched cohort of non-pregnant ES patients, we could not assess whether exacerbations were more likely during pregnancy or simply consistent with the nature of the disease. Likewise, without a matched cohort of pregnant patients with isolated AIC or no medical comorbidities, we could not draw direct comparisons of relevant clinical outcomes. Although ES can present as AIN combined with either AIHA or ITP, no patients with AIN with ITP or AIN with AIHA had a documented pregnancy in our database. AIN remains challenging to diagnose, leading to under-recognition of this subset of ES patients,6 who may experience distinct perinatal complications such as higher rates of postpartum infection. In summary, in 30 pregnancies in 11 ES patients, we observed an approximately two- to threefold higher rate of maternal complications (63%) and miscarriages (23%) compared to isolated AIHA or ITP.3, 5 One-third had disease relapse, less common than in AIHA alone, where the rate is reported at 50%.3 Similar to both pregnancy-associated ITP and AIHA, perinatal ES responded to corticosteroids. All but one patient had a successful pregnancy with a live birth, and neonatal complications were rare. Perinatal ES patients should be closely monitored by a multidisciplinary team to provide prompt recognition and management of maternal complications and AIC relapses. DJ performed the data extraction and analysis and drafted the manuscript. DJ, AEF and DJK made intellectual contributions, critical revisions of text and final approval of the manuscript. The authors thank James B. Bussel, MD, for his thoughtful and generous editorial feedback. DJ is supported by the National Bleeding Disorders Foundation Takeda Clinical Fellowship. DJ and AEF declare no conflicts of interest. DJK receives research support from Biocryst, Hutchmed, Novartis, Principia, Sanofi, and has been involved consulting services with Alexion, Alnylam, Alpine, Amgen, Apellis, Argenx, BioCryst, Bristol Myers Squibb, Caremark, Cellularity, Cellphire, Chugai, Hengrui, Hutchmed, Immunovant, Medscape, Merck Sharp Dohme, Novartis, PER, Pfizer, Platelet Disorder Support Association, Principia, Regeneron, Rigel, Sanofi, Seismic, Sobi, Takeda, Up-To-Date, Verve. Data supporting the findings of this study are available within the article and its supplementary materials. Additional data are available from the authors upon reasonable request. Data S1. Tables S1–S4. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.