Significance of immunohematologic testing in mother and newborn ABO incompatibility

Hemolytic disease of the fetus and newborn (HDFN) is a consequence of the difference between the erythrocyte antigens of the fetus and the ABO antibodies of the mother. Unlike other blood group incompatibilities between mother and child in which the mother is immunized during pregnancy or after receiving blood components, ABO incompatibility does not require prior immunization because the mother already has naturally occurring isohemagglutinins of a certain titer.¹

With the introduction of regular and effective anti-D prophylaxis, incidence of RhD HDFN has fallen below 1 percent. The combination of antenatal and postpartum immunoprophylaxis is more than 99 percent effective at preventing maternal sensitization to D.²

In 15 percent of all pregnancies in mothers of European descent, the mother is group O and her infant is group A or B, but clinically obvious HDFN is relatively rare. There are at least two reasons for clinically moderate manifestation: fetal A and B antigens are not fully developed during pregnancy nor at birth, and A and B substances are not limited to erythrocytes. Therefore, only a fraction of IgG anti-A and anti-B, which can cross the placenta, bind to fetal erythrocytes.³ Also, anti-A and anti-B IgG are predominantly IgG2, a subclass of immunoglobulin that has a lower capacity to actively cross the placental barrier.⁴ Anti-A and anti-B occurring in group B and group A individuals, respectively, are predominantly IgM, but in group O individuals, these naturally occurring antibodies are at least partly IgG.³ In pregnancies between a group O mother and a group A or group B child, HDFN develops as a result of the destruction of fetal erythrocytes caused by IgG class anti-A, anti-B, and/or anti-A,B from maternal serum.

The mother–child serologic combination in which a clinically relevant ABO HDFN develops most readily is a group O mother and a group A newborn in mothers of European descent.^4,5 According to some authors in different ethnic populations, neonatal blood group B is more predisposed to significant hemolysis, especially in individuals of African descent.5–7 The incidence of HDFN due to ABO incompatibility is generally higher in African and Arab populations because of the more frequent expression of A and B genes.^3,4 In individuals of African descent, the B antigen is generally more strongly expressed than in individuals of European descent because they have a higher level of B-specified glycosyltransferase in their serum.^8,9 Anti-A and anti-B titers also tend to be higher in individuals of African descent than in individuals of European descent, especially anti-B, but there has also been one case report with an extremely high anti-A titer.¹⁰ Even though the relative potency of anti-A and anti-B has been suspected of being racially determined in the past, some research indicates that environmental factors may be more important.^11,12

ABO HDFN usually manifests in the neonate as hyperbilirubinemia and jaundice within 12–48 hours of birth, which can be treated with phototherapy (PT). Severe cases requiring exchange transfusion are extremely rare.⁵ Unlike RhD HDFN, ABO HDFN results in a short episode, and anemia is not usually found after the first two weeks of life.³

The aim of this study was to define the incidence and risk factors for jaundice and anemia in newborns with a positive direct antiglobulin test (DAT) or incompatible crossmatch due to ABO incompatibility between mother and newborn. We also wanted to describe the methods of treatment in clinically manifested newborns.

This report showcases a retrospective, cross-sectional, noninterventional descriptive study. Between January 2016 and December 2020, 202 newborns were identified as having additional testing in our laboratory because of ABO incompatibility between mother and newborn. In 199 (98.5%) cases, the mother was group O, and in 3 (1.5%) cases, the mother was group B.

The newborns of group O mothers are not routinely tested for ABO, so the incompatibility was recognized in other ways. Umbilical cord blood samples from newborns of D–mothers were tested during evaluation for maternal Rh immune globulin administration. Peripheral blood samples of other newborns were tested for ABO if they presented with jaundice or if pre-transfusion testing was indicated. In our facility, ABO and D testing for infants less than 4 months of age is always accompanied by a DAT. Newborns having a positive DAT or incompatible crossmatch and who were ABO incompatible with their mother were candidates for inclusion in the study.

All newborns were evaluated for other potential causes of hyperbilirubinemia or anemia. Eighteen newborns were excluded from the study because of prematurity (13), perinatal infection (3), Gilbert syndrome (1), or neonatal hepatitis (1).

Ultimately, 184 ABO-incompatible newborns with immunohematologic testing were included. Testing was performed in the laboratory of the Clinical Department of Transfusion Medicine and Transplantation Biology, University Hospital Centre Zagreb (Croatia’s largest hospital).

ABO typing and DAT of newborn samples were performed using column agglutination technology (Ortho Clinical Diagnostics, Raritan, NJ). In DAT+ cases, further testing included the determination of free maternal ABO antibodies in the newborn’s plasma and an indirect antiglobulin test (IAT) on the mother’s sample to exclude antibodies other than ABO. In the case of a mother with ABO incompatibility with her newborn but with a negative IAT, we assumed that the ABO incompatibility was the cause of the positive DAT.

Medical records of all newborns and their mothers were reviewed from the Hospital Information System (IN2, Zagreb, Croatia) and from Hemotools Transfusion software (ZN INFORMATIKA, Zagreb, Croatia).

Neonatal bilirubin level testing was performed if it was triggered by the presence of jaundice. Criteria for PT was determined by the pediatrician in charge of caring for the individual newborn, according to the American Association for Pediatrics criteria for hyperbilirubinemia in effect at the time of treatment.¹³

The characteristics of the two groups of ABO-incompatible newborns were compared: treated versus untreated. In the group of treated newborns, A and B blood groups were also compared. We found nine pairs and one trio of ABO-incompatible siblings who were born during the study period.

The study was approved by the University Hospital Centre Zagreb Ethics Committee in December 2020 (8.1-20/212-2; 02/21 AG) and conducted according to the principles of the Declaration of Helsinki.

Statistical analysis was performed using the MedCalc program (Ostend, Belgium), version 20.009. Data distribution normality was assessed using the Shapiro-Wilk test. Depending on data normality distribution, results are expressed as median accompanied by interquartile range, or as mean and standard deviation. Statistical analysis of quantitative variables with an abnormal distribution was performed using the Mann-Whitney U test, while the t test was used for results with a normal distribution. Fisher exact test was performed to compare qualitative variables. A p value <0.05 was considered statistically significant.

Table 1 lists the characteristics of the two groups of newborns with ABO incompatibility, namely, treated versus untreated. Despite ABO incompatibility, 112 of 184 (61%) newborns did not have any laboratory or clinically significant symptoms of hyperbilirubinemia or anemia. Most of these newborns (173 of 184; 94%) were DAT+, and the other 11 (6%) were DAT–. Six of the newborns with negative DATs were found to be ABO incompatible with their mothers while performing crossmatch testing before surgery for congenital anomalies; therefore, no treatment was required.

Clinically manifested ABO incompatibility was seen in 72 of 184 (39%) newborns, of which 70 were treated with PT. One newborn was treated with erythrocyte transfusion and PT and one with only erythrocyte transfusion. Two newborns who received PT also received intravenous immunoglobulin.

In the treated group of 72 newborns, 4 (5.6%) tested DAT–but were identified with HDFN by an incompatible crossmatch, and 12 (16.7%) did not have free maternal ABO antibodies in their plasma at the time of testing. Within the group of treated newborns, there is an equal number of firstborn and second-born infants, which is typical for HDFN because of the ever presence of maternal ABO antibodies.

Most of the newborns (64 of 72; 89%) in the treated group were partially or exclusively breastfed; in the group of untreated newborns, this information was unknown for 40 of 112 (36%) newborns. Therefore, the type of feeding did not provide significant data for comparison. In both treated and untreated groups, about one-third of newborns came from the first pregnancy (31% and 30%, respectively) and one-third from the second pregnancy (32% and 37%, respectively).

Table 2 presents the characteristics of the two treated groups of newborns according to A and B blood groups. We did not find statistically significant differences between these two groups for any characteristics of interest, except for two newborns with blood group A who received erythrocyte transfusions.

During these 5 years, several mothers gave birth multiple times; as a result, we have interesting data with nine pairs and one trio of siblings, presented in Table 3. There were no unexpected alloantibodies in the plasma of the mothers of these study newborns.

The incidence of ABO incompatibility cannot be defined precisely because of the lack of systemic testing of all newborns with early jaundice and those born to mothers with blood group O. Similarly, as reported by Rahmati et al.,¹⁴ one-third of DAT+ newborns had HDFN, with ABO incompatibility as the most common cause of DAT in their study. Two-thirds of the examined population in our study had a clinically insignificant condition, but the number of newborns treated with PT because of ABO incompatibility may be even higher if the blood group was determined for each newborn after delivery. In Table 3, there is an example of a newborn (N10-2) who was probably ABO incompatible with her mother and received PT treatment, but the blood group was not determined, so we cannot be sure of the cause of jaundice.

In a study in Italy, the incidence of ABO incompatibility was reported to be 11 percent in a shorter period,¹⁵ and in a study of a smaller cohort from India, ABO incompatibility was reported at 37 percent.¹⁶ During the observed period, 14,582 babies were born in the University Hospital Centre Zagreb alone, 1 percent of whom were ABO incompatible with their mothers according to our study—but not all infants were tested. In our study, the positive predictive value of DAT for identifying hyperbilirubinemia/anemia that would require intervention is 0.39, or 39 percent.

There is a statistically significant difference (p = 0.004) in the newborns’ bilirubin values based on the mode of delivery between the two groups that was not previously described; newborns delivered by vaginal birth had higher bilirubin values compared with those born by caesarean section.^17,18 Perhaps we can interpret this as a coincidence because of a higher number of caesarean sections in the untreated group. We cannot, however, consider this as clinically relevant.

Additionally, there is a statistically significant difference between the groups in relation to the time of positive DAT after delivery (p < 0.001), which is probably due to the D status of the mother. Since there were more D–mothers in the untreated group, the positive DAT would have been detected sooner. This testing resulted in a higher proportion of DAT+ newborns in the group of untreated, ABO-incompatible newborns.

As reported in the Results, there are no statistically significant differences for compared variables in the group of treated newborns compared with the difference in blood group (group A and group B), except for blood transfusion (two newborns with blood group A). In different ethnic populations, HDFN due to ABO incompatibility is differently manifested, and fetuses/newborns with certain blood groups present with more severe disease than those with other blood groups. Similar to our research, studies of two cohorts from Turkey and India showed no differences between newborns with blood group A and blood group B,^19,20 unlike a research study from Tunis, in which there was a higher rate of hemolysis in newborns with blood group B.⁶

Researchers in the neighboring country of Serbia published a 17-year experience study with exchange transfusion. According to their results, neonates with blood group A had a greater risk of hyperbilirubinemia and exchange transfusion.²¹ They analyzed almost 400 newborns who underwent exchange transfusion for hyperbilirubinemia because of different causes, and ABO incompatibility was determined to be the cause in 150 newborns.

Our study identified 10 pairs of siblings with similar outcomes (Table 3). In six pairs, both siblings were treated with PT, and in one pair, neither sibling was treated. In a trio, the elder two received PT, but the youngest did not; and in two additional pairs, the elder received no treatment, but the younger did. A possible explanation could be a higher titer of antibody in the maternal serum during the second pregnancy.

ABO antibody titer is not routinely determined for mothers during pregnancy or after birth, but it would certainly be important to see a correlation to the clinical presentation of ABO incompatibility.¹

In our cohort, only 2 of 72 newborns (2.7%) who clinically manifested ABO HDFN received erythrocyte transfusion, which confirms the fact that only about 1.5–2 percent of cases of HDFN require transfusion therapy.^22,23

ABO incompatibility is currently the most common cause of HDFN in the western world. In the future, however, changes in the ethnic structure of populations in Western and Central Europe could lead to even more cases.4 The incidence of HDFN due to ABO incompatibility is the same for first pregnancies as it is in subsequent pregnancies, so the condition is neither preventable nor predictable. Because ABO incompatibility rarely causes severe HDFN, routine antenatal tests to assess the potency of anti-A and anti-B are not indicated.4 In mothers who have a history of a previous infant affected by ABO HDFN, umbilical cord blood should be taken and tested as soon as possible after delivery.³

A limitation of this study is the small number of newborns with ABO incompatibility; future studies should be prospectively conducted on a larger sample. Another potential limitation of our study is the practice at our institution to not regularly perform an eluate in these cases, which would prove that the cause of DAT positivity was related to ABO incompatibility.

Although DAT as a test is not predictive of disease severity, it can be a useful tool to monitor newborns from ABO-incompatible mothers.¹⁵ It would be useful to perform ABO typing and a DAT on all newborns to determine the exact incidence of ABO incompatibility and assume the possible need for therapy, but according to current guidelines this testing is not recommended.²⁴ Our recommendation is to determine the ABO blood group and DAT on umbilical cord blood samples from newborns whose mothers have unknown ABO group and IAT or a positive IAT. An umbilical cord blood sample after birth does not harm the child in any way and does not put the child at risk of anemia.