How I treat sickle cell disease in pregnancy

Although sickle cell disease (SCD) is associated with major morbidity, currently in the United States >95% of children survive into adulthood.^1,2 In the early 1970s, the median life expectancy of a person with SCD in the United States was 14 years; now, 50 years later, the median life expectancy is between 40 and 60 years. People live into their reproductive years and beyond.³ Although 50 years ago, people with SCD were discouraged from becoming pregnant, now, most should be supported if they choose to pursue a pregnancy, but they and their providers should be aware of the unique maternal and fetal challenges.

SCD is a vascular disease and, starting with conception, the evolving embryo in a woman with SCD is at greater risk from inadequate vascularization than it would be in a mother who did not have SCD. Fetal growth restriction, preeclampsia, and other consequences of utero–placental insufficiency are more likely to occur in mothers with SCD.^4-8 In 1989, Roberts et al⁹ proposed that a poorly perfused placenta releases factors into the maternal circulation that injure endothelial cells, resulting in the clinical manifestations of preeclampsia.⁴ Furthermore, any maternal endothelial damage from a poorly perfused placenta has the potential of being superimposed on the endothelial damage of SCD. A systematic review and meta-analysis found that women with SCD have a twofold to threefold increased risk of preeclampsia, a threefold to fourfold increased risk of fetal growth restriction, and a fourfold increased risk of stillbirth.¹⁰ 

Mechanical and immune changes during pregnancy contribute to an increased susceptibility to certain infections.¹¹ The immune system during pregnancy is altered to accommodate a fetus whose genetic make-up is half paternal.¹² There is reduced activity of natural killer cells, inflammatory macrophages, and helper T-cell type 1 cells, and reduced production of inflammatory cytokines.¹² Infections that are usually controlled by inflammatory responses are more severe during pregnancy.¹² Pregnant people are more severely affected by influenza virus,¹¹ hepatitis E virus,¹¹ herpes simplex virus,¹¹ malaria parasites,¹¹ and COVID-19,¹³ and possibly coccidioidomycosis, measles, smallpox, and varicella.¹¹ In pregnant people with SCD, this increased susceptibility to certain infections is superimposed on an already increased severity of respiratory and urinary tract infections.¹⁴ Compared with people without SCD, pregnant people with SCD have a 2- to 13-fold increased risk of serious infectious morbidity, including sepsis.^5,15 

Although there is increased susceptibility to certain infections during pregnancy, there is simultaneously increased activation of the complement system, as evidenced by higher concentrations of C3a, C4a and C5a,¹⁶ and increased neutrophils.¹⁷ Neutrophils contribute to the multicellular aggregation and vascular endothelial adherence of red blood cells and platelets that result in vaso-occlusion.¹⁸ The increased complement activation of pregnancy potentially provides a compensatory mechanism that enhances maternal host responses against infection but when combined with the complement activation that is common in SCD, as evidenced by elevated levels of soluble C5b-9,^19,20 has the potential to result in overactivation and tissue damage, including hemolysis. The downstream effects of the release of free heme include inflammation, endothelial dysfunction, microvascular occlusion, and further tissue damage.”²⁰ 

During pregnancy, red cell mass should increase by 400 to 450 mL, or by ∼30%²¹ to support the expected 40% to 50% increase in blood volume. In people with SCD, this increase in red cell mass is often not achievable without transfusion. Even when transfusion is administered selectively as opposed to universally, 30% to 60% of women still receive transfusion.^22-24 

By 24 weeks gestation, cardiac output increases by 45% in singleton pregnancies and by another 15% in twin pregnancies.²⁵ Pregnant people are normally able to accommodate the increased cardiopulmonary demands of pregnancy unless there is underlying cardiopulmonary disease. Although the percentage of SCD pregnancies affected by pulmonary hypertension is not known, pulmonary hypertension, especially morbid in pregnancy, affects an estimated 6% to 11% of patients with SCD.^26,27 Maternal mortality from pulmonary hypertension among all pregnancies was previously reported to be between 30% and 50%, and although the prognosis has improved in recent years, maternal mortality remains in the range of 9% to 28%.²⁸ 

Anatomic and physiologic changes during pregnancy result in dilatation of the renal collecting system, decreased systemic and renal vascular resistance, an increased glomerular filtration rate, and modifications in tubular function.²⁹ Individuals with SCD are already at an increased risk of both acute and chronic kidney disease because of injury to the glomeruli, tubules, and renal vasculature.³⁰ These injuries may manifest as urine concentrating defects, albuminuria, renal tubular acidosis with hyperkalemia, or hypertension; may complicate the diagnosis of preeclampsia; and may increase the susceptibility to acute kidney injury.⁶ A recent analysis using data from the Nationwide Inpatient Sample found a ninefold increased odds of acute kidney failure at the time of delivery in patients with SCD compared with in patients without SCD.¹⁵ 

Pregnancy is a hypercoagulable state with an increase in concentrations of various clotting factors, a decrease in the natural anticoagulant protein S, and decreased fibrinolytic activity because of increased levels of plasminogen activator inhibitor type 1 and the presence of plasminogen activator inhibitor type 2, a product of the placenta.^31,32 Consequently, people are 4 times more likely to experience venous thromboembolism (VTE) when pregnant than when not.³³ Immediately after delivery, the risk is even 5 times higher than during pregnancy.³⁴ SCD is also a hypercoagulable state that carries an increased risk of VTE. By age 40 years, the cumulative incidence of VTE in individuals with SCD is ∼2%.³⁵ In people with SCD, compared with those without SCD, a 2- to 20-fold increased incidence of pregnancy-related thromboembolic events has been reported,^5-8,15,36-38 with a 10- to 22-fold increased odds of stroke.^15,39 

Maternal complications can arise from chronic underlying organ dysfunction such as renal disease or pulmonary hypertension; from acute complications of SCD such as severe anemia, vaso-occlusive crises, and acute chest syndrome; and from pregnancy-related complications. Although causality is not established except for maternal transfusion, anemia, in and of itself, is associated with adverse pregnancy outcomes.^40,41 During pregnancy 50% to 70% of people with SCD require at least 1 hospital admission^22,37,42 and 30% to 40% require transfusion.^22,23 In 1 cohort, women with SCD were hospitalized an average of 6 days during pregnancy.³⁷ In a study of 126 pregnant women with sickle SS disease in Jamaica, there were 135 predelivery admissions among 81 patients compared with 33 predelivery admissions among 31 controls.⁴² The most common reason for predelivery admission among the patients with SCD was vaso-occlusive crisis (34%) followed by acute chest syndrome (8%), urinary tract infection (7%), and preeclampsia (6%). Pregnant patients with SCD not only require transfusion but have frequently received transfusion previously and, according to published reports, 20% to 50% of patients with SCD are alloimmunized.⁴³ Patients who are alloimmunized have limited blood products available to them when transfusion is necessary, but the most serious consequence of alloimmunization at the time of transfusion is the risk of developing a life-threatening delayed hemolytic transfusion reaction. Although alloimmunization is reduced by contemporary extended matching practices, it is not available universally and does not completely eliminate the risk alloimmunization.⁴⁴ Perhaps because of underlying renal disease, perhaps because of possible underlying hypertension, and perhaps because of placental ischemia, patients with SCD are more likely to experience preeclampsia and eclampsia.^5-8,15,45 As mentioned earlier, they are also more likely to experience VTE,^5-8,15,36-39 severe infections,^5,8,15,38 acute renal failure,^6,15 other severe maternal morbidity,⁴⁶ and death. In the United States, the maternal mortality rate is ∼10 times higher than it is for patients without SCD.^5,7 Similarly increased rates are seen in both low and high income countries.^10,47 

Fetal complications during pregnancy are most often the consequence of utero–placental insufficiency but are also due to infections, alloimmunization, spontaneous preterm labor and, less often, opioid exposure. Studies from multiple cohorts, including large population studies, have documented the increased risk of fetal growth restriction,^{5,7,15,38,48,49} preterm delivery,^{5,6,8,36,37,45,50-52} and stillbirth.^6,15,53 In a study using serial ultrasounds, fetal growth was essentially the same between fetuses of mothers with SCD and fetuses of age-matched controls at 25 weeks gestation but lagged by the next ultrasound at 35 weeks gestation.⁴⁹ When fetal growth restriction occurs, it is usually asymmetric,⁴² suggesting that the underlying cause is utero–placental insufficiency. In women with SCD, the increased rate of preterm delivery is predominantly because of medical indications for delivery, not spontaneous preterm labor, which is the opposite of what is true for the general population.⁵² There are fetal as well as maternal consequences of red cell alloimmunization. If patients have formed antibodies associated with hemolytic disease of the newborn, these patients are at risk of having a fetus who is severely anemic or possibly even stillborn.⁵⁴ Although there are only limited prospective data on postnatal opioid withdrawal or neonatal abstinence syndrome specifically for infants born to mothers with SCD, in a retrospective cohort study of 131 infants born to mothers with SCD at Tampa General Hospital, only 4% were diagnosed with neonatal abstinence syndrome.⁵⁵ 

A 26-year-old nulliparous woman with sickle SS disease recently become engaged. She and her fiancé desired to start a family. Although she had not been hospitalized in the last year, she had a history of frequent hospitalizations before starting hydroxyurea (HU). Her fiancé believed he had been tested for the sickle cell trait and was negative.

Individuals who are contemplating a pregnancy and their partners should be aware of the increased risks to an individual’s health during pregnancy. Ideally, they will have the opportunity to speak with a specialist in maternal-fetal medicine. Some studies have found poorer outcomes in patients with sickle cell anemia as opposed to other SCD genotypes,⁴⁵ but other studies have found few, if any, differences.^56,57 A history of frequent hospitalizations and/or episodes of acute chest syndrome and a history of frequent transfusions do correlate with increased risks during pregnancy. Blood pressure should be checked and hypertension treated to lower systolic blood pressure to ≤140 and diastolic blood pressure to ≤90.⁵⁸ A complete blood count (CBC), reticulocyte count, and ferritin level should be obtained, as well as a typing and screening to obtain extended red blood cell phenotype or genotype (if not obtained previously) and to check for red blood cell antibodies. An assessment of renal function and urinary protein, and referral for retinal examination should be performed if not performed within the last year.⁵⁸ We screen for urinary protein with a microalbumin-to-creatinine ratio on a spot urine. If the ratio is >30 μg/mg creatinine, further evaluation is pursued. The American Thoracic Society recommends that individuals with SCD be screened for pulmonary hypertension with Doppler echocardiography every 1 to 3 years.²⁷ These recommendations remain controversial, but, at a minimum, any woman with a history of disordered breathing or symptoms of pulmonary embolism should undergo Doppler echocardiography.^58-60 All individuals with SCD should receive immunizations according to the US Center for Disease Control and Prevention Advisory Committee on Immunization Practices but especially pneumococcal, Haemophilus influenzae type b, and meningococcal vaccinations because of functional asplenia. The latest recommendations can be found on the CDC website. If the ferritin level is elevated, a prenatal vitamin without iron should be prescribed. Prenatal vitamin tablets almost always contain iron, but prenatal gummy vitamins do not. The American College of Obstetricians and Gynecologists recommends that pregnant patients with SCD take 4 mg of folic acid per day.⁶¹ 

The potential fetal risks during a future pregnancy should be considered as well. If a woman has red cell alloantibodies known to cause hemolytic disease of the newborn, the father of the pregnancy should also be tested for the corresponding red cell antigen(s). If he tests positive, the woman and her partner should be counseled about the possibility of hemolytic disease of the newborn and how it is detected. Patients and their partners should also be counseled about the potential fetal and neonatal effects of maternal medications, including opioids. Angiotensin-converting enzyme (ACE) inhibitors should be discontinued. HU, associated with birth defects in experimental animals given high doses, has not been associated with an increased risk of birth defects among infants whose mothers received it during pregnancy.⁶² Recent studies using data from the Sickle Cell Disease Implementation Consortium found that a history of HU use was no more prevalent among women who reported infertility,⁶³ and that HU use at the time of conception alone did not increase the risk of miscarriage or birth defects.⁶⁴ We recommend that patients who are taking HU continue it (or their other disease-modifying therapy) while trying to conceive.

Mothers with SCD are at risk of having a child affected with SCD if the father of the child has SCD, hemoglobin S (HbS) trait, β-thalassemia trait, or is a carrier of another abnormal Hb such as HbC or HbE. If his sickle cell or thalassemia status is unknown, he should be offered testing for hemoglobinopathies (typically Hb electrophoresis and a CBC). If he has 1 of these traits, he and the mother would be at risk of having a child who is affected and should be offered genetic counseling to review their options for preimplantation genetic testing, which does require in vitro fertilization, or prenatal diagnosis. After spontaneous conception, prenatal diagnosis of SCD is possible by cell-free fetal DNA testing (when commercially available), by chorionic villus sampling in the first trimester, or by amniocentesis in the second trimester of gestation.

A 34-year-old woman at 12 weeks gestation with sickle SS disease and a history of 4 previous preterm cesarean deliveries between 24 and 35 weeks gestation was referred because of her history of delayed hemolytic transfusion reaction and a Hb of 6. She intermittently tested positive for the anti-S and anti-c antibodies.

During pregnancy, patients with SCD should be evaluated and treated with respect to the unique maternal and fetal issues associated with SCD. Usually this evaluation will include consultation with a specialist in maternal-fetal medicine who may or may not provide ongoing obstetric care. Whether or not the patient has been under the care of a gynecologist in her community, she should be referred for obstetric care at the institution where her sickle cell providers are based, and where there is a sickle cell expert or team of experts,⁶⁵ maternal-fetal medicine, obstetric anesthesia, a robust blood bank, and newborn intensive care. Pertinent history that should be available to the patient’s providers includes her history of transfusions, hospitalizations, episodes of acute chest syndrome, vaso-occlusive crises, stroke, VTE, cholecystectomy, status of the spleen, Doppler echocardiography, any recent renal evaluation, and immunizations. Issues not addressed during the preconception period (see the preceding case) should be addressed early in pregnancy. Routine prenatal laboratory tests may be able to be obtained at the time of SCD laboratory tests to reduce blood draws. HU, any chelation agents, and ACE inhibitors should be discontinued, if they have not been already, although only ACE inhibitors have been associated with birth defects in humans. Data regarding other disease-modifying therapies are limited. Whether and when to restart HU requires careful consideration and discussion with the patient, given the increased risks of stopping disease-modifying therapy for SCD during pregnancy and the uncertainty about fetal effects; see discussion hereafter. With respect to analgesics, nonsteroidal anti-inflammatory drugs can have adverse fetal effects when used for >48 hours, especially after 30 weeks gestation, but acetaminophen and opioids are considered safe or low risk during pregnancy. Each patient should have an individualized pain management plan updated for pregnancy and available to all clinicians who may be responsible for her care. Although live vaccines are not safe during pregnancy (measles, mumps, rubella, and chickenpox), all the other routine immunizations are safe in any trimester. Penicillin prophylaxis has been shown to reduce the risk of life-threatening Streptococcus pneumonia infections in children aged <5 years⁶⁶ but not in children aged >5 years⁶⁷ or in adults. In the United Kingdom, however, penicillin prophylaxis is recommended for adults and pregnant women.⁶⁸ It is not the standard of care in the United States, nor is it our practice. Low-dose aspirin reduces the risk of preeclampsia and is recommended for all patients at high risk,⁶⁹ such as patients with SCD who have a reported incidence of 10% to 33% based on meta-analyses.^47,70 We recommend low-dose aspirin starting at 12 weeks gestation.⁵⁹ 

HU is the most important disease-modifying therapy available for SCD and has been proven to reduce the incidence of acute chest syndrome, vaso-occlusive crises, and other complications; and to improve survival in people who are not pregnant. Our hematologists and maternal-fetal medicine specialists developed a policy regarding HU. Ideally, HU should not be discontinued until confirmation of pregnancy by either a positive pregnancy test or early ultrasound. If patients desire to continue HU during the first and early second trimesters, they should be counseled about the potential risks (unproven adverse fetal effects) and benefits (likely decreased risks of acute chest syndrome, vaso-occlusive crises, and other complications). Patients who are candidates for HU in pregnancy have almost always been taking HU before pregnancy. We would very seldom consider starting HU for the first time during pregnancy. Candidates should wait, if possible, until after the second trimester anatomy ultrasound has been performed to restart HU. In the absence of contraindications, however, prophylactic transfusion would be the preferred first-line treatment. Although prophylactic transfusion has not been clearly shown to improve pregnancy outcomes,^71,72 a meta-analysis of medical literature databases did find evidence, albeit from a relatively small number of studies with methodological limitations, that there is maternal and fetal benefit from prophylactic transfusions.⁷³ In patients for whom transfusion is contraindicated, HU may be started as long as the patient’s sickle cell provider or providers are in agreement and can provide follow-up. In the case of this patient, we recommended she start on HU because of her severe anemia and her greatly increased risk of life-threatening complications with transfusion.

Ongoing care during pregnancy includes both maternal and fetal surveillance. Blood pressure measurements are obtained at each visit, although normal blood pressure ranges in pregnancy and thresholds for the diagnosis of preeclampsia have not been established for SCD.⁷⁴ Point-of-care testing for urine protein by dipstick is performed at each visit with a concern if the reading is ≥2+. A urine culture is repeated every 1 to 3 months. Because of the increased risk of pyelonephritis and preterm labor with asymptomatic bacteriuria during pregnancy, asymptomatic bacteriuria should be treated. CBCs should be obtained monthly to monitor for severe anemia. In addition to seeing their obstetric provider, patients typically see their sickle cell provider monthly. Fetal surveillance should include serial ultrasonography for fetal growth every 3 to 4 weeks starting at 28 weeks gestation, with antepartum fetal heart rate testing starting ∼32 weeks gestation or sooner if indicated. The risk of stillbirth can be reduced but not eliminated with fetal surveillance. Most adverse fetal outcomes can be attributed to utero–placental insufficiency, but sudden and unpredictable stillbirths can still occur. Because of the ongoing risks to both mothers and their fetuses, some experts, including ourselves, recommend planned delivery at 37 weeks gestation for patients with sickle cell anemia (HbSS and sickle β-null thalassemia), or other severe phenotype.³⁷ Recommendations regarding delivery mode should be according to obstetric indications, recognizing that the rate of cesarean delivery is higher in patients with SCD.^5,6,36,37,51 During any third-trimester hospitalization, the anesthesiologist should be consulted. With 4 previous cesarean deliveries, the aforementioned patient was not a candidate for a vaginal delivery.

It is not clear what the minimum acceptable Hb level is before delivery in persons with SCD. Preoperative transfusion therapy to increase Hb levels to 10 g/dL is strongly recommended.⁵⁸ Because patients with SCD are at high risk for cesarean delivery (43% at our institution,⁵² 45% in the United Kingdom,⁴⁵ and 38% in the United States overall¹⁵), it may be reasonable to target a Hb level of 10 g/dL at the end of pregnancy, although we settle for 8 g/dL. In the case of patients, such as the aforementioned patient, we would settle for even lower level and aim for 7 to 8 g/dL in anticipation of delivery. We work closely with the blood bank who identifies potentially suitable red cell units for transfusion. When transfusion is absolutely necessary and suitable red cell units are not available, we administer immunomodulatory therapies with corticosteroids (typically methylprednisolone, 125 mg daily) and intravenous immunoglobulin (0.5 g/kg daily), both for 4 days along with prophylactic anticomplement therapy (eculizumab^75,76) in patients at very high risk of delayed hemolytic transfusion reaction. Eculizumab is considered safe in pregnancy and during lactation.⁷⁶ 

Patients with SCD have a risk of thrombosis comparable with that of a patient with a high-risk thrombophilia. Patients with SCD should have pneumatic compression devices during antepartum hospitalizations, during labor, and at the time of cesarean delivery. For patients with a history of VTE, we prescribe full-dose anticoagulation with low–molecular weight heparin (enoxaparin, 1 mg per kg every 12 hours) and, unless a patient is on long-term anticoagulation and will continue indefinitely, we continue the enoxaparin for 6 weeks postpartum. For patients with a history of stroke, we prescribe low-dose anticoagulation in addition to low-dose aspirin as suggested by the Canadian Heart Association for prevention of recurrent stroke.⁷⁷ For patients without a history of VTE, we prescribe low-dose anticoagulation with low–molecular weight heparin (enoxaparin, 40 mg daily) during hospitalizations for vaso-occlusive crises and, if hospitalized, continue throughout the remainder of the pregnancy. We also prescribe low-dose anticoagulation (enoxaparin, 40 mg daily) starting 12 hours after a vaginal delivery or 24 hours after a cesarean delivery and continue for 6 weeks postpartum.

A 19-year-old nulliparous patient was referred to our sickle cell team by maternal-fetal medicine for ongoing sickle cell care. She had been maintained on HU since early childhood but had discontinued HU when she realized she was pregnant. She was referred because of repeated admissions for vaso-occlusive crises attributed to pregnancy and to stopping HU.

Individualized pain management plans should provide patients with SCD with a sequence of medications, including opioid doses and other interventions to try at home, and an algorithm for when to present to the day hospital, emergency department, or obstetric unit. When a pregnant woman with SCD does present with pain, she should be evaluated for other complications, but pain should be treated promptly and aggressively.⁵⁸ There are no randomized trials on the safety and efficacy of interventions (transfusion, oxygen therapy, fluid replacement analgesia, and/or steroids) for treating vaso-occlusive crises during pregnancy,⁷⁸ but treatment is essentially the same as for patients who are not pregnant. Nonsteroidal anti-inflammatory drugs, however, except for short courses with <48 hours duration before 30 weeks gestation, are contraindicated, therefore opioids and acetaminophen should be used instead. In general, transfusion should be administered only if there are other indications.⁵⁸ Incentive spirometry should be initiated for patients who are hospitalized.⁵⁸ Oxygen should be given if the oxygen saturation is <95% by pulse oximetry.^58,61 We have not prescribed home oxygen therapy, but some individuals with SCD will have unrecognized nocturnal hypoxemia and the administration of home nocturnal oxygen, including after discharge, may reduce complications of SCD.⁷⁹ 

After careful discussion and consideration, this patient was restarted on HU at 21 weeks gestation. She had no further admissions for vaso-occlusive crises, but at 26 weeks gestation, she developed severe preeclampsia and required emergent delivery.

During pregnancy, SCD poses unique problems to both mother and fetus. Individuals who are contemplating a pregnancy and their partners should be aware of the increased risks to the individual’s health during pregnancy and the increased risks of adverse pregnancy outcome. During pregnancy, patients should be cared for by a multidisciplinary team and monitored closely. Recommendations from this article are summarized in Table 1. We recognize that the approach to care may need to be modified in resource-poor settings in which a safe blood supply may be nonexistent and the risk-to-benefit ratio of various treatments may be different. We also recognize that there is a need for more data to inform the care of pregnant patients with SCD.

The authors thank Joacy G. Mathias for her assistance with the visual abstract.

Contribution: A.H.J. wrote the manuscript; and J.J.S. contributed essential content and edited the manuscript.

Conflict-of-interest statement: A.H.J. reports consultancy for Cerus; received royalties from UpToDate; reports research funding (paid to institution) from Coagulant Therapeutics; and serves on the board of directors of Foundation for Women and Girls with Blood Disorders. J.J.S. is the site principal investigator for Agios clinical trials for mitapivat for thalassemia (ENERGIZE and ENERGIZE-T) and for the Takeda phase 1 study of recombinant human ADAMTS-13 for sickle cell disease (SCD) (inactive); and serves on the scientific advisory board for Disc Medicine, Inc for a phase 1b study of bitopertin in SCD. J.J.S. receives funding from the Human Resources and Services Administration (U1EMC42461-01-00).

Correspondence: Andra H. James, Division of Maternal-Fetal Medicine, Department of Obstetrics & Gynecology Duke University, DUMC 3967, Durham, NC 27710; email: andra.james@duke.edu.