Joy, Saju D. MD*; Rossi, Karen Q. RN*; Krugh, Dave†; O'Shaughnessy, Richard W. MD*
Red blood cell alloimmunization is a well-known cause of hemolytic disease of the fetus or newborn. Of the 270 red blood cell antigens with the potential to cause hemolytic disease of the fetus or newborn, Rh(D) antigen has been the most studied.1 However, given the widespread use of Rh(D) immune globulin, there has been a relative increase in the importance of non-Rh(D) alloimmunization as a cause of hemolytic disease of the fetus or newborn.2–5 Of the remaining 43 antigens in the Rh system, the other frequently observed antigens include C, c, E, and e.
The obstetrician encounters a dilemma upon demonstration of anti-E during routine antenatal screening for red blood cell antibodies. Although it is established that the E antigen can cause alloimmunization, the effect in causing clinically significant hemolytic disease of the fetus or newborn continues to be debated.6 Pregnancies complicated by alloimmunization due to any of the atypical antibodies are generally managed using clinical strategies established for anti-D alloimmunization.1
At The Ohio State University, the Fetal Therapy Program has maintained a database of pregnancies affected by alloimmunization since June 1959. Each case was reviewed and followed up contemporaneously by our institutional isoimmunization committee. The objective of this study was to retrospectively review cases of anti-E alloimmunization to determine whether prenatal management similar to that for anti-D is appropriate for anti-E.
MATERIALS AND METHODS
A computerized database containing the records of all women with alloimmunized pregnancies who had a consultation for a positive antibody screen at The Ohio State University Medical Center from June 1959 to April 2004 was used to identify all pregnant women affected by anti-E. Before conducting this study, permission to retain and evaluate these patients’ data were obtained from The Ohio State University Institutional Review Board. Data were obtained for the computerized database from hospital charts and transfusion service and physician records. This group was composed of patients from our institution, as well as referrals from central and southeastern Ohio and neighboring regions. Maternal data included pregnancy and transfusion history, indirect and direct antiglobulin tests, antibody titer, results of ΔOD450, middle cerebral artery peak systolic velocity, and fetal hemoglobin and antigen testing. When available, paternal antigen typing was included. Neonatal data included gestational age at delivery, hemoglobin and hematocrit, cord blood direct antiglobulin test result, newborn antigen typing, and neonatal morbidity and treatments for hemolytic disease of the fetus or newborn. Pregnancies affected by anti-E alloimmunization confirmed by a positive direct antiglobulin test due to anti-E or positive E antigen typing in the fetus or newborn were included. Pregnancies affected by multiple antibodies or with a positive direct antiglobulin test due to ABO incompatibility were excluded, as were patients with incomplete data.
To eliminate any interlaboratory variation, all serum titers were analyzed at The Ohio State University Medical Center Prenatal Reference Laboratory. This laboratory follows published guidelines endorsed by the American Association of Blood Banks.7 An isoimmunization committee evaluated all laboratory reports and recommended a care plan for each of these patients as described below. The isoimmunization committee is made up of obstetricians, pediatricians, nurses, and transfusion medicine staff.
Affected pregnancies with anti-E alloimmunization were monitored using the same criteria as for anti-D alloimmunization. Titers were measured at 4-week intervals or less, depending on the initial level and trends in the titers. In cases where invasive procedures were necessary, fetal E antigen phenotype or genotype or both was determined using the fetal red blood cells obtained by cordocentesis or by polymerase chain reaction testing of amniocytes obtained by amniocentesis.
We have established 1:32 as the critical titer at our institution. If the anti-E titer rose to greater than or equal to 1:32, or at lower titer levels when there was a history of a prior affected child, an amniocentesis was performed for ΔOD450 evaluation.1 The ΔOD450 results were plotted on a modified Liley graph (O’Shaughnessy R. Amniotic fluid spectrophotometry is useful after 20 weeks gestation in the care of pregnancies complicated by red blood cell isoimmunization [abstract]. Am J Obstet Gynecol 1991;164:317).8,9 Amniocentesis was begun as early as 20 weeks of gestation. Subsequent amniocenteses were repeated at intervals determined by the ΔOD450 values. Middle cerebral artery peak systolic velocity has recently been added as an adjunct test to evaluate for fetal anemia10 and was obtained in patients with titers greater than or equal to 1:32 beginning in 2001.
The clinical use of maternal serologic titers and amniotic fluid spectrophotometry remained largely unchanged over the study period. Cordocentesis for monitoring and intravascular transfusion of the fetus was implemented in our program in 1987. In some of the cases presented in this article, cordocentesis was used rather liberally compared with today's standards. This reflected the period of the early 1990s when some authorities favored the use of cordocentesis as a primary tool for fetal blood typing and evaluation of fetal anemia.11,12 We now use cordocentesis for direct assessment of fetal hematologic characteristics when amniotic fluid ΔOD450 levels are in zone III or rising or plateauing in zone IIB.
Before 1987, intraperitoneal intrauterine transfusions were performed at our institution when either hydrops fetalis or amniotic fluid ΔOD450 in zone III was identified. Beginning in 1987, fetal anemia was confirmed by cordocentesis, and fetal transfusion was considered when the hemoglobin was less than 10 g/dL or hematocrit less than 30%.11–13 In patients who required intrauterine transfusion therapy, only data obtained before the first transfusion is reported.
A literature review using PubMed and MEDLINE was performed using the keywords “anti-E,” “alloimmunization,” erythroblastosis fetalis,” and “hemolytic disease of the newborn.” Articles available in the English language were reviewed.
This review identified anti-E in 283 pregnancies from September 1966 to April 2004. Of these, there were 32 pregnancies in 27 women with only anti-E antibodies, confirmed fetal or neonatal risk for hemolytic disease of the fetus or newborn, and complete data. The average age of the patients was 29 years with a range from 18–44 years. Age was not recorded in 5 pregnancies. Of the 32 pregnancies, 25 (78%) were managed in the 16 years between 1987–2004. Anti-E antibody cases referred to our program increased in frequency after 1981.
Sixteen pregnancies (50%) had titers less than 1:32. Amniocentesis was performed in 1 of these cases for a history of a previously affected child, and the ΔOD450 value was in zone I of the modified Liley graph. Neonatal hemoglobin was recorded in 7 of these 16 newborns, with all of the results more than 13 g/dL. All the newborns in this group were delivered at term and had a normal, uncomplicated neonatal course. In this group with titers less than 1:32, there were no cases of hydrops fetalis or fetal demise.
The other 16 pregnancies (50%) had titers of 1:32 or greater (Table 1). The critical titer of 1:32 predicted all cases of anemic fetuses and newborns. Amniocenteses were performed for ΔOD450 in 15 of these pregnancies. One patient did not have an amniocentesis due to noncompliance. Delivery and newborn care in that case were unremarkable. Overall, 85 amniocenteses were performed based on the established criteria defined above. When the ΔOD450 of the 15 patients were plotted, ΔOD450 values in zone IIB or zone III identified all pregnancies with significant anemia (hemoglobin < 10 g/dL or hydrops fetalis) before transfusion or at delivery. There were 38 cordocenteses performed in 4 pregnancies and a total of 11 intravascular intrauterine transfusions performed in 3 pregnancies. Fetal hemoglobins before the first intrauterine transfusions ranged from 8.1 to 9.5 g/dL. There was 1 intraperitoneal transfusion performed in 1970 for hydrops fetalis. There were no emergent deliveries resulting from complications of these invasive procedures. Five of the 16 newborns from pregnancies with titers of 1:32 or greater required red blood cell transfusion for hemolytic disease of the fetus or newborn after birth.
There were 5 pregnancies (15%) identified with a fetal or neonatal hemoglobin less than 10 g/dL and 1 pregnancy complicated with hydrops fetalis due to anti-E alloimmunization (Table 1, pregnancies G, J, K-6, L-7, L-5, M). Figure 1 shows ΔOD450 values for these 6 pregnancies. In 4 of these 6 pregnancies, cordocenteses were performed, with 3 receiving intravascular intrauterine transfusions. There were 2 perinatal deaths. One pregnancy (Table 1, pregnancy M) underwent an intraperitoneal intrauterine transfusion for hydrops fetalis, with subsequent intrauterine fetal death. In 1 patient (Table 1, pregnancy L-5), a sudden rise in the ΔOD450 value in association with fetal ascites prompted treatment with an intrauterine transfusion. Emergent delivery ensued, with resultant neonatal death due to strangulation and subsequent perforation of the intestine from congenital malrotation.
Since 2001, our institution has also included middle cerebral artery peak systolic velocity in the management of hemolytic disease of the fetus or newborn. During this period, there were 2 cases of anti-E with titers of 1:32 or greater with normal middle cerebral artery peak systolic velocity (< 1.5 multiples of the mean) measurements. The ΔOD450 values were in zone I of the modified Liley graph, and both pregnancies delivered at term without complications or hemolytic disease of the fetus or newborn.
With the institution of Rh(D) immune globulin prophylaxis beginning in 1968, there has been a decline in anti-D alloimmunization and a relative increase in alloimmunization associated with other red blood cell antigens.5 Anti-E is frequently encountered, often second or third in frequency to anti-Kell and anti-D.4,5 Anti-E can occur as a natural immunoglobulin M antibody without immune stimulation or an immunoglobulin G antibody in those with a history of a transfusion or prior pregnancy.14 Most often anti-E alloimmunization is associated with mild to moderate hemolytic disease of the fetus or newborn.2–5,15,16
The literature includes some case reports17,18 and several case studies2,4,5,15,16,19 of isoimmunization due to anti-E alone. Patient G in this study has been previously reported.17 The study by Moran et al19 includes 62 infants with anti-E who had a positive direct antiglobulin test. They show that anti-E can cause clinically important hemolytic disease of the fetus or newborn, but they found no correlation between disease severity and antibody titer. Pepperell et al15 included 44 patients with anti-E with information on newborn Coombs’ status, need for infant exchange transfusion, and stillbirth (1 case). However, there was no information regarding serologic titers or amniocentesis for this group. Kornstad,4 Jovanovic-Srzentic et al,5 and Bowell et al2 identified 61, 67, and 90 cases, respectively, of anti-E, but did not provide past medical history or any information regarding hemolytic disease of the fetus or newborn, serologic titers, or other indices. Finally, Wu et al16 reported 6 cases of anti-E from 1991–2000 among Taiwanese women, with 1 case of hydrops fetalis.
Our findings confirm that anti-E alloimmunization can cause significant hemolytic disease of the fetus or newborn requiring prenatal intervention. In contrast to the conclusions by Moran et al,19 our data indicate that antibody critical titer is useful. In our study population, a titer of 1:32 or greater identified all of the anemic fetuses. We believe that in the absence of a prior affected infant this is an appropriate critical titer. In 50% of cases reported here, only maternal serologic titers were necessary to monitor the fetus in utero.
Some authorities have questioned the continued usefulness of amniocentesis, an indirect index of fetal hemolysis, when more direct analysis of fetal hematologic characteristics is available with cordocentesis.20 In our data, amniotic fluid ΔOD450 patterns detected all of the significantly anemic fetuses. Amniocentesis, while an invasive test, is associated with less risk to the fetus than cordocentesis. The recent development of noninvasive testing for fetal anemia with middle cerebral artery peak systolic velocity holds great promise. In our study, 3 patients (Table 1, pregnancies G, J, L-5) demonstrated a significant increase in ΔOD450 after multiple amniocenteses with or without cordocenteses. We acknowledge that although amniocentesis and cordocentesis provide valuable data regarding fetal condition, these invasive tests pose the risk of aggravating the disease process.1,21
Limitations of our study include its retrospective nature and selection bias based on patients referred to our tertiary level institution. Our study population therefore is not intended to estimate an incidence for anti-E alloimmunization or for severity of disease. In addition, we use a modified Liley graph that allows for evaluating ΔOD450 values from 20–40 weeks of gestation.8–10 Others may use different modifications of the Liley graph.22
Fetal hemolytic disease of the fetus or newborn due to anti-E alloimmunization can be monitored in most cases using maternal serologic analysis supplemented by amniotic fluid spectrophotometry and cordocentesis when necessary. The same criteria used to follow Rh(D) alloimmunization are appropriate in patients with E alloimmunization. In some cases, fetal transfusion may be necessary. The use of middle cerebral artery peak systolic velocity is not clarified by these data, but middle cerebral artery peak systolic velocity holds promise as a useful noninvasive tool to monitor the severity of fetal anemia. Our data show that a critical serologic titer of 1:32 in the absence of a previously affected fetus warrants further evaluation with amniocentesis, cordocentesis, and possible treatment with intrauterine transfusion.
1. American College of Obstetricians and Gynecologists. Management of isoimmunization in pregnancy. ACOG Educational Bulletin 227. Washington, DC: ACOG; 1996.
2. Bowell PJ, Allen DL, Entwistle CC. Blood group antibody screening tests during pregnancy. Br J Obstet Gynaecol 1986;93:1038–43.
3. Lee CK, Ma ES, Tang M, Lam CC, Lin CK, Chan LC. Prevalence and specificity of clinically significant red cell alloantibodies in Chinese women during pregnancy—a review of cases from 1997 to 2001. Transfus Med 2003;13:227–31.
4. Kornstad L. New cases of irregular blood group antibodies other than anti-D in pregnancy. Acta Obstet Gynecol Scand 1983;62:431–6.
5. Jovanovic-Srzentic S, Djokic M, Tijanic N, Djordjevic R, Rizvan N, Plecas D, et al. Antibodies detected in samples from 21,730 pregnant women. Immunohematology 2003;19:89–92.
6. van Dijk BA, Dooren MC, Overbeeke MA. Red cell antibodies in pregnancy: there is no “critical titre”. Transfus Med 1995;5:199–202.
7. American Association of Blood Banks. Technical manual of the American Association of Blood Banks. 14th ed. Arlington (VA): American Association of Blood Banks; 2002.
8. Liley AW. Liquor amnii analysis in the management of the pregnancy complicated by rhesus sensitization. Am J Obstet Gynecol 1961;82:1359–70.
9. O'Shaughnessy R, Kennedy M. Isoimmunization. In: Ling FW, Duff P, editors. Obstetrics and gynecology: principles for practice. New York (NY): McGraw Hill; 2001. p. 308–26.
10. Mari G. Noninvasive diagnosis by doppler ultrasonography of fetal anemia due to maternal red cell alloimmunization. N Engl J Med 2000;342:9–14.
11. Weiner CP, Williamson RA, Wenstrom KD, Sipes SL, Grant SS, Widness JA. Management of fetal hemolytic disease by cordocentesis. I. Prediction of fetal anemia. Am J Obstet Gynecol 1991;165:546–53.
12. Weiner CP, Williamson RA, Wenstrom KD, Sipes SL, Widness JA, Grant SS, et al. Management of fetal hemolytic disease by cordocentesis. II. Outcome of treatment. Am J Obstet Gynecol 1991;165:1302–7.
13. Moise KJ Jr. Management of rhesus alloimmunization in pregnancy. Obstet Gynecol 2002;100:600–11.
14. Harrison J. The “naturally occurring” anti-E. Vox Sang 1970;19:123–31.
15. Pepperell RJ, Barrie JU, Fliegner JR. Significance of red-cell irregular antibodies in the obstetric patient. Med J Aust 1977;2:453–6.
16. Wu KH, Chu SL, Chang JG, Shih MC, Peng CT. Haemolytic disease of the newborn due to maternal irregular antibodies in the Chinese population of Taiwan. Transfus Med 2003;13:311–4.
17. Strohm PL, Iams JD, Kennedy MS. Hemolytic disease of the newborn from anti-E. J Reprod Med 1988;33:404–6.
18. To WW, Ho SN, Mok KM. Anti-E alloimmunization in pregnancy: management dilemmas. J Obstet Gynaecol Res 2003;29:45–8.
19. Moran P, Robson SC, Reid MM. Anti-E in pregnancy. BJOG 2000;107:1436–8.
20. Nicolaides KH, Rodeck CH, Mibashan RS, Kemp JR. Have Liley charts outlived their usefulness? Am J Obstet Gynecol 1986;155:90–4.
21. MacGregor SN, Silver RK, Sholl JS. Enhanced sensitization after cordocentesis in a rhesus-isoimmunized pregnancy. Am J Obstet Gynecol 1991;165:382–3.
22. Queenen JT, Tomai TP, Ural SH, King JC. Deviation in amniotic fluid optical density at a wavelength of 450 nm in Rh-immunized pregnancies from 14 to 40 weeks gestation: a proposal for clinical management. Am J Obstet Gynecol 1993;168:1370–6.