Fetal anemia

Fetal anemia is a relatively rare but serious condition. An immune-related cause is most common, involving red-blood-cell (RBC) alloimmunization, followed by non-immune causes, such as parvovirus B19 infection and, more rarely, hemoglobinopathies, fetomaternal hemorrhage (FMH) and monochorionic twin complications, among others. Several advances have been made in the diagnosis and treatment of fetal anemia. High-resolution ultrasound now permits accurate, non-invasive screening by measurement of the middle cerebral artery peak systolic velocity (MCA-PSV)1, replacing serial amniocenteses for fetal anemia screening2. Refinement of fetal blood sampling (FBS) and intrauterine transfusion (IUT) techniques has resulted in close to 90% survival of anemic fetuses3, with very good long-term neurodevelopmental outcomes4. In this review, we discuss the most common etiologies of fetal anemia, current screening and diagnostic tools, management of fetal anemia, including FBS/IUT techniques and associated risks and benefits, as well as short- and long-term outcomes following IUT. As fetal hemoglobin (Hb) values increase gradually during pregnancy, anemia may be classified based on the degree of Hb deviation from the mean for gestational age (GA)5 or on multiples of the median (MoM)1 for GA (Table 1). Hydrops typically does not develop until the Hb deficit is > 70 g/L or the absolute Hb value is < 50 g/L6. Alloimmunization occurs following maternal exposure to paternally derived RBC antigens on fetal cells, resulting in maternal antibody formation and secondary fetal RBC hemolysis, anemia, hydrops and, ultimately, fetal demise1. Although over 50 RBC antigens have been associated with fetal hemolysis, severe fetal anemia results most commonly from Rhesus (Rh) D, Rhc or Kell alloimmunization7. RBC alloimmunization is the leading cause of fetal anemia, despite standardized protocols for RhD immune globulin (RhIG) prophylaxis, most often due to unrecognized FMH events, inadequate dosing or missed prophylaxis for antenatal sensitizing events, poor patient compliance, absence of prophylaxis for other RBC antigens and omission of Kell typing of blood transfusions for women of childbearing age7. Pregnancies at risk are identified on the basis of a previous history of hemolytic disease of the fetus and newborn (HDFN) or when causative antibodies are identified on routine maternal blood-group screening. If paternity is certain, initial management involves determining the paternal RBC antigen status and zygosity. If the father is heterozygous for a particular RBC antigen, the fetus has a 50% risk of inheritance and fetal genotyping can be performed. If the father is homozygous, all fetuses will inherit that antigen and genotyping is unnecessary. Traditionally, fetal genotyping was determined by amniocentesis, with a sensitivity of 98.7% and specificity of 100% for RhD status8. The discovery of cell-free DNA in the maternal plasma has since enabled non-invasive determination of the fetal RhD genotype9. Experience with this approach is extensive, with prediction accuracy and test sensitivity approaching 100%, and very few false-negative results9. While results can be obtained as early as 10 gestational weeks, the International Blood Group Reference Laboratory in Bristol, UK currently performs genotyping for RhD, Rhc, RhC and RhE after 16 weeks and for Kell after 20 weeks' gestation. In sensitized pregnancies, serial titers are performed, typically every 4 weeks until 28 weeks' gestation and every 2 weeks thereafter7. Historically, once titers reached a laboratory-specific ‘critical’ level indicative of high risk of fetal anemia, commonly defined as ≥ 1:32 for anti-RhD and most other antibodies and ≥ 1:8 for anti-Kell10, serial amniocenteses were then performed for bilirubin levels to estimate the severity of hemolysis. Spectrophotometry was used to quantify bilirubin level, which was expressed as the change in optical density (OD) at a wavelength of 450 nm (ΔOD450). These values were plotted on Liley's curve11 or Queenan's curve (< 27 weeks)12 to predict fetal anemia. MCA-PSV testing has now replaced serial invasive testing after a randomized control trial demonstrated that Doppler assessment of the MCA-PSV had higher sensitivity and accuracy for prediction of severe fetal anemia compared with amniotic-fluid ΔOD450 in Rh-alloimmunized pregnancies2. Once critical titers are reached, MCA-PSVs are measured weekly to determine optimal timing of FBS. Nearly 65% of women of childbearing age are immune to PB19, and 1.5% of susceptible women will seroconvert during pregnancy13. Typically, transmission is by respiratory droplets; less commonly, it is transplacental or from blood products14. The overall risk of vertical transmission is approximately 17–33%, with the highest risk occurring prior to the third trimester15, 16. PB19 is a single-stranded DNA virus with a particular affinity for erythroid progenitor cells17, potentially leading to hemolysis, bone-marrow aplasia and, occasionally, thrombocytopenia and neutropenia. Erythema infectiosum, or fifth disease, represents the typical presentation in children, with a low-grade fever, malaise and a characteristic ‘slapped-cheek’ facial rash, followed by a maculopapular truncal and extremity rash. Most adults are asymptomatic, or may experience polyarthralgia. Rare presentations include myocarditis and heart failure, aplastic crisis in the setting of chronic anemia, and persistent infection and anemia in immunocompromised individuals17. Fetal infections are mostly asymptomatic without sequelae15, but may result in miscarriage, severe anemia and non-immune hydrops (NIH) and stillbirth. Marked fetal ascites is pathognomonic of PB19 infection (Figure 1). The risk of fetal loss is estimated at 13% when infection occurs < 20 weeks and 0.5% when it occurs > 20 weeks, and approximately 3% of affected fetuses will develop hydrops14. Diagnosis of maternal infection relies on detection of PB19-specific immunoglobulin (Ig)M and/or IgG antibodies14. Viremia develops approximately 1 week after inoculation. IgM is detectable 7–10 days after infection, peaks at 10–14 days, and remains detectable for up to 6 months. Two weeks after infection, IgG becomes detectable and confers lifelong immunity18. Fetal infection is confirmed by polymerase chain reaction for PB19 DNA in amniotic fluid or fetal blood. NIH typically develops 2–6 weeks after seroconversion19 but may occur up to 10–12 weeks later20. Ultrasound and MCA-PSV measurements should be performed every 1–2 weeks for approximately 10–12 weeks from maternal seroconversion14, with referral to a fetal medicine unit for FBS with or without IUT if there is evidence of fetal anemia or hydrops. Rarely, fetal anemia and hydrops can be caused by other congenital infections, including cytomegalovirus (CMV)21, syphilis22 and toxoplasmosis23. Other ultrasonographic features of congenital infections may include ascites, placentomegaly, hepatosplenomegaly, echogenic bowel, liver or intracranial calcifications, ventriculomegaly and intrauterine growth restriction (IUGR). Alpha (α)-thalassemia is caused by the defective synthesis of α-globin chains and is the most common cause of NIH in Southeast Asia. The α-globin gene cluster consists of four copies of the α-globin gene (αα/αα). α-thalassemia can be classified into four types based on the number of functional α-globin genes: silent carrier state (-α/αα), α- thalassemia trait (--/αα (cis) or -α/-α (trans)), Hb H disease (--/-α) and α-thalassemia major (--/--) (Hb Bart's or homozygous α-thalassemia). If both parents carry the cis deletion (--/αα), the inheritance risk for α-thalassemia major is 25%. Fetuses with homozygous α-thalassemia cannot produce normal fetal Hb (α2γ2), instead producing Hb Bart's (γ4). Clinical features include: cardiomegaly, ascites, hydrops, IUGR, limb defects24, intrauterine fetal death (IUFD) and neonatal death (NND)25-27. Given the severe impairment in tissue oxygen delivery, the sequelae of fetal anemia and hydrops may occur at higher Hb values than seen with RBC alloimmunization, making MCA-PSV Doppler less predictive of disease severity. Maternal complications may include ‘mirror’ syndrome and antepartum hemorrhage25. Prenatal diagnosis is established by DNA analysis via chorionic villus sampling performed after 10 gestational weeks or amniocentesis after 15 weeks. Alternatively, serial sonographic evaluation for cardiomegaly and placental thickness can be used as a screening tool in early pregnancy. The sensitivity and specificity of placental thickness for detection of α-thalassemia major were 72% and 97%, respectively, before 12 weeks and 95% and 97%, respectively, after 12 weeks28. Using a cardiothoracic ratio (CTR) ≥ 0.5 to estimate cardiomegaly, sensitivity and specificity were 100% at 12–13 weeks26. When the predictive value of CTR, MCA-PSV and placental thickness were compared, CTR was found to be superior, and sensitivity at 12–15 weeks was further increased by addition of MCA-PSV measurement27. Historically, fetal α-thalassemia major was considered to be a uniformly lethal condition, but, with the advent of IUT, increasing numbers of live births have been reported24. However, in the absence of curative therapy, prenatal treatment raises ethical dilemmas and requires detailed counseling from a multidisciplinary team. Congenital leukemia and occasionally transient myeloproliferative disorder may present prenatally with fetal anemia, hepatosplenomegaly, polyhydramnios, placentomegaly, hydrops and IUFD29, most commonly associated with trisomy 2129, 30. Placental chorangiomas are the most common benign placental tumor and can be associated with fetal anemia31, 32, secondary to fetal hemorrhage31 or consumption and destruction of fetal RBCs within the narrow, thrombosed vasculature of the chorangioma33. Other complications include NIH, polyhydramnios, preterm labor, IUGR and pre-eclampsia, particularly if tumors are > 4 cm32. Fetal sacroccygeal tumors can also result in fetal anemia, due to hemorrhage within the tumor and RBC consumption, with secondary high-output cardiac failure and NIH34. Traditionally, FMH has been defined as hemorrhage of ≥ 30 mL fetal blood into the maternal circulation, as this is the volume of Rh-positive whole fetal blood that is covered by a standard 300-μg dose of RhIG to prevent alloimmunization35. Large or massive FMH has been defined as blood loss of 80–150 mL36 or > 20 mL/kg37, however, there is no fixed volume above which fetal morbidity or mortality will result unequivocally. Although the majority of FMHs remain unexplained35, risk factors include: external cephalic version38, abdominal trauma39, placental abruption, placenta previa, IUFD, Cesarean section, manual removal of the placenta and amniocentesis35. Massive FMH with fetal hydrops at 26–28 weeks has been managed with serial IUTs; however, outcomes are variable, ranging from resolution of hydrops and live birth40 to IUFD41. Whether to expedite delivery or perform an IUT will depend on GA, antenatal test results and availability of expertise. Twin–twin transfusion syndrome (TTTS) and twin anemia–polycythemia sequence (TAPS) are chronic forms of fetofetal transfusion42. TTTS complicates up to 15% of monochorionic diamniotic (MCDA) twin pregnancies and pathogenesis is related to unbalanced blood flow through placental vascular anastomoses43. A gradual blood-flow shift develops, resulting in oliguric oligohydramnios in the donor and polyuric polyhydramnios and volume overload in the recipient43, potentially leading to hydrops secondary to right ventricular dysfunction44. A large intertwin hemoglobin difference is not generally seen, unless there is coexisting TAPS. TAPS is characterized by a Hb discordance in MCDA twins in the absence of the oligopolyhydramnios sequence seen with TTTS, and occurs when there are only a few, tiny (< 1 mm), unidirectional arteriovenous vascular anastomoses, with sparse or absent compensatory arterioarterial anastomoses45. TAPS may occur spontaneously in 3–5% of MCDA twins, usually presenting after 26 weeks' gestation (Figure 2), or in 2–13% of TTTS cases, following incomplete laser ablation of placental anastomoses45. Cotwin demise may result in an acute hemodynamic imbalance, due to a massive transfer of blood from the survivor to the dead twin via the placental anastomoses in monochorionic placentae, causing anemia, neurological handicap or death of the remaining twin46. Although IUT can correct the anemia, whether or not it impacts on neonatal or neurodevelopmental outcome is unknown. Recommended investigations for evaluation of suspected fetal anemia are listed in Table 3. Referral to a maternal fetal medicine specialist and a tertiary care center with expertise in FBS and IUT is indicated. Fetal blood sample (FBS)* should be sent for blood type, CBC, Hb/Hct, platelet count, direct Coombs, reticulocyte count and total bilirubin, PCR for CMV and PB19 with or without syphilis and toxoplasmosis Non-stress test for sinusoidal fetal heart rate The first manifestation of hydrops secondary to fetal anemia is always ascites, followed by placental thickening and hepatomegaly; associated pleural or pericardial effusions are rare47. Fetal liver length > 90th percentile48, splenomegaly (spleen perimeter > 2SD)49 and Doppler measurements of mean blood velocity in the fetal aorta50 and splenic artery PSV51 were also found to correlate with fetal anemia. When Oepkes et al.52 compared fetal liver length, spleen perimeter, umbilical vein (UV) diameter, placental thickness and UV and aortic flow velocities, only Doppler parameters were predictive of severe anemia in non-hydropic fetuses. Evaluate for secondary causes (i.e. fetal or placental) of anemia MCA-PSV > 1.5 MoM1 Placentomegaly, hydrops, ascites97, enlarged fetal liver48 and spleen49 Placental thickness > 2 SD above GA mean28 or > 18 mm at 12–15 weeks98 or > 30 mm at 18–21 weeks99 Cardiomegaly: CTR > 0.5 before 18 weeks or > 0.52 after 18 weeks98 Ascites Initial studies in fetal lambs, measuring blood flow in relation to hematocrit (Hct), demonstrated compensatory increased flow to the brain, heart and adrenals to maintain stable oxygen delivery across a range of Hct53; this was also replicated in humans54. An inverse relationship was also noted between fetal Hct and MCA-PSV55. Subsequently, Mari et al.1 demonstrated that MCA-PSV > 1.50 MoM in non-hydropic fetuses at risk of RBC alloimmunization could detect all cases of moderate to severe anemia, with a false-positive rate of 12%. Oepkes et al.2 thereafter confirmed the superiority of MCA-PSV over amniotic-fluid ΔOD450, making serial amniocenteses for fetal anemia screening essentially obsolete. An axial section of the brain, including thalami, cavum septi pellucidi and greater wing of sphenoid, with the circle of Willis identified by color Doppler, should be obtained1. The MCA closest to the probe is sampled at or near its origin from the internal carotid artery. The waveform peak is measured, with the angle of insonation as close as possible to 0°, and always < 30° (Figure 3). Higher inter- and intraobserver variability results from angle correction and sampling of more distal regions of the MCA56. The fetus should be quiescent, as heart-rate accelerations and movement can alter measurements57. Other factors that may influence MCA-PSV include: gender, cardiac status, uterine contractions, fetal behavioral state58, advanced GA59 and previous IUTs60. Elevated MCA-PSVs may also be seen in abnormal placentation and IUGR, reflecting cerebral autoregulation in response to hypoxemia and hypercapnia61. These factors, in addition to insufficient attention to proper technique, may in part explain the false-positive rate of MCA-PSV for prediction of fetal anemia, described by Mari et al.1. We perform FBS only when the MCA-PSV is above 1.5 MoM and trending upwards. Investigations at the time of FBS are listed in Table 3. Rare causes of anemia should be evaluated in conjunction with a hematologist. Administration of corticosteroids for fetal lung maturity should be considered prior to FBS after viability. The goal of IUT in HDFN is to replace the fetal blood with Rh-negative donor blood, thus supressing fetal erythropoiesis. IUT is generally indicated for fetal Hct < 30%10 or Hb < 10 g/L64. Fresh, CMV-negative, leukoreduced, irradiated, Type-O, RhD-negative donor RBCs, cross-matched to a maternal sample and packed to a Hct of 75–80%, are used64. Autologous washed maternal blood can also be used for IUT once the maternal Hb level is > 120 g/L, which eliminates the risk of sensitization to new RBC antigens in random donor blood. Platelets should be available if PB19 or CMV are suspected, or in the presence of hydrops or hepatosplenomegaly. Coexistent severe thrombocytopenia has been reported in hydropic alloimmunized pregnancies65 and hydropic fetuses with PB1966. When a previous pregnancy has been complicated by HDFN, titers are less reliable predictors of severe fetal anemia and MCA-PSV should be followed serially instead. Intravenous immune globulin (IVIG)67, 68 and plasmapheresis68 have also been used in some cases with a history of previous early mid-trimester loss due to RBC alloimmunization or markedly elevated titers, with a reported reduction in perinatal mortality and delay of the first IUT by 1.5 weeks67. In contrast, one small randomized controlled trial found no additional benefit of IVIG when combined with IUT, compared with IUT alone, in cases of severe RhD alloimmunization69. An intravascular approach, via the UV at the placental cord insertion (PCI), the intrahepatic vein (IHV) or a free loop of cord, has largely replaced intraperitoneal transfusion (IPT), first described in 196370. IPT may still have a role in very early pregnancy, as higher fetal loss rates are associated with intravascular transfusion at GA < 22 weeks71. Rarely, an intracardiac approach may be used. In anatomically normal fetuses, the loss rate with IUT is approximately this rate may however, with and In the presence of or hydrops, loss rates are and risks include from the and fetal heart-rate which are usually and have been reported in up to and of In one outcomes with IUT, performed mostly a approach, of had complications, including preterm of infection, Cesarean section, and or with an overall loss rate of < weeks was found to occur in of cases and only in than Fetal and were more common when a free loop of cord was compared with or Although and are over a free loop of cord, superiority of or has not been demonstrated the on and of the approach include of and secondary and cord less FMH and rates in the of if intraperitoneal it is usually and as an We perform > of via the and this particularly in pregnancies or at advanced gestational when the placenta is Fetal such as and are particularly if a or is may the risk of cord complications, such as or all of which can result from with fetal transfusion and approximately of the volume should be followed by FBS to determine the volume for IUT. Hb level is g/L, in non-hydropic fetuses. Several studies have demonstrated outcomes following IUT in hydropic compared with non-hydropic due to cardiac and increased to volume In a of fetuses that were hydropic secondary to RBC alloimmunization, which IUT, the survival rate overall was in and in severe of hydrops in 65% of cases and only of hydropic fetuses In we not the more than in hydropic fetuses during a IUT, as this has been found to be a of fetal we will perform a within 1–2 days if to the are no fetal following however, weekly assessment of fetal and MCA-PSV Doppler is of is particularly since the of MCA-PSV to predict fetal anemia becomes less reliable after with false-positive rates of and 90% for the detection of 95% of anemic fetuses following the and third IUT, As the in MCA-PSV in anemic fetuses a due to blood the accuracy of MCA-PSV following IUT may be related to of fetal blood, when with blood cells, which have a due to the of fetal RBCs with RBCs following IUT, the of fetal which has a higher affinity for is thus the total oxygen in the fetal have reported higher MCA-PSV for the diagnosis of severe anemia following IUT, that this may be by a compensatory in cerebral blood related to total oxygen and oxygen tissue delivery of blood and the of However, established for diagnosis of fetal anemia following serial have not been Alternatively, the timing of may be based on the in fetal and Hb for the and third IUT or a of in fetal have at 10 days, 2 weeks and weeks, for third and The treatment of TAPS is laser ablation of the placental vascular However, if this is not is to the anemic donor can result in for the with serial this can be by The goal in pregnancies should be to delivery, close to of a requires transfusion In stable pregnancies that have or more a IUT at weeks, with delivery weeks is unless other for delivery prior to survival in the of IUT, is in hydropic and in non-hydropic The neonatal complications are related to or as well as anemia and In the after which evaluated short- and long-term neurodevelopmental outcome after IUT for alloimmunization, of were < weeks, with most between and weeks. neonatal morbidity respiratory syndrome perinatal and severe cerebral When outcome of ≥ weeks was compared between with and without IUT, was for in had been in however, the transfusion rate was in both and was in of or anemia presenting up to after has been described in with with of a RBC count at is related to the number of transfusions that of is the Nearly of alloimmunized additional antibodies following IUT, despite of and blood outcome following IUT was normal in 95% of at a median age of in the severe delay and were in 3% and of children, associated with neurodevelopmental impairment severe hydrops, number of and severe neonatal in hydropic alloimmunized fetuses following IUT is In the severe fetal hydrops was present in of of neurodevelopmental When et evaluated long-term neurodevelopmental outcome in 18 hydropic fetuses, found that death or major morbidity in 16 hydropic were normal at a mean age of 10 and with sequelae had other factors, particularly outcome in hydropic fetuses secondary to PB19 infection is less with a survival rate of that the virus may et evaluated fetuses that were hydropic due to PB19 and which of 16 one third demonstrated In hydropic fetuses followed for a median of survival was and severe neurodevelopmental impairment was noted in RBC alloimmunization is the most common cause of fetal anemia, followed by PB19 infection and, more rarely, hemoglobinopathies, fetal and placental tumors and complications of monochorionic The non-invasive prediction of fetal anemia by Doppler evaluation of MCA-PSV has the management of fetal replacing amniocentesis for bilirubin and making other ultrasound predictors of anemia. Referral to a medicine specialist with expertise in FBS and IUT is indicated when MCA-PSV measurements are ≥ 1.5 maternal investigations should include blood and testing for and and detailed placental and fetal If fetal blood is it should be for blood and Hb/Hct, platelet and reticulocyte count, and Rare causes may be evaluated in conjunction with and FBS and can be performed by an intraperitoneal or intracardiac The intravascular approach is most common, via the or than a free loop of cord, in to the risk of fetal the of FBS and IUT, fetal survival has been reported at and is approaching for RBC alloimmunization in with long-term neurodevelopmental outcome in the of severe fetal hemolytic