Anti-U alloimmunisation in a pregnant woman from Niger

Introduction

“U” was the name that Wiener et al.¹ gave in 1953 to recall the almost universal distribution of a new antigen that they described. The antigen was identified through the serum of an Afro-American patient who had a fatal transfusion reaction: the serum reacted with the red blood cells of 977 of 989 Afro-American patients and all 1,000 Caucasian-American patients tested.

The U antigen belongs to the MNS blood group system, the only, together with the Rh system, to be synthesised starting from two very closely linked loci, GYPA and GYPB, which, together with a third locus, GYPE, form a gene cluster on chromosome 4 (4q28→q31)^2–3.

The U antigen, together with the S and s antigens, is expressed on glycophorin B (GPB), a red blood cell membrane glycoprotein, coded for by the GYPB locus^4–6. The amino acid residues from 33 to 39 are essential for the expression of the U antigen. Its antigenicity is not, however, entirely due to this molecular sequence⁷.

The U antigen is a public antigen, occurring almost universally in Caucasian populations⁸. In populations belonging to ethnic groups originating from Sub-Saharan Africa, there is a S-s−U− phenotype, found with a frequency that varies from 1% among Afro-Americans up to 35% among the Pygmies of Congo⁹; the spread of this phenotype among these populations is related to the fact that it hampers the entrance of malaria plasmodia into red blood cells.

This phenotype is caused by the lack of glycophorin B. U− subjects are always also S- and s−. In contrast, the lack of S and s antigens is not always associated with an absence of the U antigen from the red blood cell membrane. According to some research, about 51% of S-s− subjects are U+¹⁰. In these cases, however, the U antigen is variously altered by rearrangements of the amino acid sequence in the residues from 33 to 39: this is indicated by the notation U+^var which stands for a variant of the U antigen. The notation U+^var does, therefore, represent a heterogeneous group of molecules from the points of view of both amino acid sequence and antigens, but a group united by the fact of having some epitopes specific to the U molecule from which each variant is derived¹¹. Subjects with the S-s−U+^var phenotype produce anti-U antibodies that react with all U+ and with a variable proportion of other U+^var, depending on the epitopes expressed on the variant molecule. S-s−U− subjects, on the other hand, produce antibodies that react with all U+ and U+^var red blood cells and should, more correctly, be called anti-U/GPB¹². While the S-s−U− phenotype is a consequence of a deletion of GYPB, the molecular basis of the S-s−U+^var phenotype is a GYPB-like, hybrid gene.

Anti-U is the cause of haemolytic disease of the foetus and newborn, as well as of haemolytic transfusion reactions¹³. This is a rare immunohaematological problem: 24 cases of anti-U immunisation have been reported in the literature, all of which occurred in pregnant women from ethnic groups originating from Sub-Saharan Africa^14–27.

The severity of the haemolytic disease varies from asymptomatic to fatal, with intrauterine death.

Anti-U immunisation causes important diagnostic and transfusional problems, related both to its rarity and to the specific immunohaematological problem of non-Caucasian ethnic groups, a new problem for Italy.

Here we report a case of immunisation against the U antigen, which came to our attention at the Service of Immunohaematology and Transfusion Medicine, Careggi (Italy) in 2009.

Case report and results

This report concerns A.B., a 27-year old woman from Niger, gravida 2, para 0, pregnant by a partner from the same ethnic group who was also the father of the woman’s previous pregnancy. She had a history of past transfusion therapy with red cell concentrates (RCC) in 2005. The patient, who has sickle cell disease, was admitted to the Department of Health Sciences for Women and Children in Careggi University Hospital in the 30^th week of gestation because of anaemia due to a haemolytic crisis (haemoglobin [Hb] 8.9 g/dL; lacate dehydrogenase 397 U/L). The indirect antiglobulin test (IAT) had always been negative. The patient’s blood group was typed as O+ ccDEe kk.

Initially the patient was managed with pharmacological therapy with folic acid and vitamin B12, which stabilised the anaemia but did not improve it (Hb 8.7 g/dL). Given the persistence of the anaemia and the presence of intrauterine growth retardation, delivery by Caesarean section was planned for 34 weeks +3.

In preparation for the operation, it was advised that the woman be transfused with 1 unit of RCC, to be repeated, if necessary, after the delivery. The day before the planned delivery the woman was transfused, without complications, with 1 unit of O+ ccEe kk RCC, which was compatible according to “type and screen” (T&S) studies.

At birth the neonate was in a good general condition. His direct antiglobulin test (DAT) and the mother’s IAT were negative. The neonate’s blood group was found to be O+ ccDEe kk. After delivery the mother’s Hb was lower (8.3 g/dL) but stable; for this reason a second transfusion was not given, but pharmacological treatment was continued.

Ten days after delivery the woman’s Hb had fallen to 5.7 g/dL making another transfusion of 1 unit of RCC necessary. On this occasion the Transfusion Service found a positive IAT with a 4+ score for each of the three test red cells; the DAT was negative. This profile suggested alloimmunisation, probably as a result of the contemporaneous stimulatory action of the pregnancy and the preceding transfusion therapy. The patient’s blood was no longer compatible with any type of blood tested. While awaiting the results of further investigations, the pharmacological treatment was continued.

The process for identifying the specificity was started by testing the serum with a first screening panel of 11 cells: Ortho Panel C. In the first round of tests an incubation temperature of 37 °C was used in a microcolumn with complete anti-gammaglobulin serum (IgG+C3d). The result was homogeneous pan-reactivity with a score of 4+. The use of the same red cells treated with ficin did not alter the results of the test. Similarly, the test of direct agglutination of the serum, with the same panel of red cells, in a test-tube at 20 °C gave a picture of homogeneous positivity with a score of 2+, with all the red cells of the panel. This suggested the presence of an IgM component which, together with the other evidence, outlined the picture of a primary immune response. Cross-matching with 23 units of RCC showed incompatibility in all cases.

The phenotyping of the red cells from the mother and her newly born son was extended to the MNS, Kidd and Duffy systems. The results for the Fy and MNS systems were as follows: mother Fy(a−b−)S-s−U−, son Fy(a−b−) S-s+U+. The following antisera were used: anti-S, anti-s, anti-M, anti-N, anti-Fya, anti-Fyb, anti-Jka, anti-Jkb - Spectra Biologicals (Pieco s.r.l., Capezzano Pianore, Lucca, Italy); anti-U - Bio-Rad (Bio-Rad Laboratories s.r.l., Segrate, Milan, Italy).

The identification process was extended further using two other panels (Twenty Immucor) each containing 20 test red cells, including test cells with the Fy(a−b−)S+s−, Fy(a−b−)S-s+ and Fy(a−b−)S-s−U−He− phenotypes. The tests were performed both at 37 °C in a microcolumn with LISS+IgG+C3d and at 20 °C in a test-tube for the search for agglutinins. The previously observed homogeneous reactivity remained with both methods for all the test red cells used except S-s−U−He− cells, which were the only cells that were not reactive. The tests were also carried out in solid phase with panels not containing U− test stroma. Once again, the method also showed homogeneous 4+ reactivity, indicating the presence of an IgG antibody component. Collectively, these data led to the definitive identification of the antibody specificity: anti-U, in part IgG reactive at 37 °C, in part IgM, reactive also at 20 °C.

The contemporaneous presence of IgG and IgM, as well as the repeated negativity of the IAT and DAT carried out previously on the patient’s serum and the neonate’s red blood cells, are evidence of a further feature making our case particular: this was a primary reaction.

Discussion

In the clinical case described above, the availability of test red cells and a specific serum enabled the correct diagnosis to be made of an immunohaematological problem typical of populations originating from Sub-Saharan Africa; the diagnosis would not have been possible with routinely used reagents.

Although the diagnostic challenge was solved, the transfusion problem remains, since the patient is immunised against the U antigen which is present in 100% of native Italian blood donors. Given the underlying pathology, there is a problem of obtaining compatible blood in the case that the patient should require further transfusions.

The diagnostic difficulties encountered in the case described and the consequent therapeutic implications provoke some thought on the immunohaematological problems that we are ever increasingly facing with non-Caucasian immigrants and on the adequacy of the instruments at our disposal.

Immigrants have red cell phenotypes that are uncommon in the native Caucasian population and have correlated different risks of immunisation. These differences lie in the presence of a significant number of subjects lacking antigens present in the majority of the population (public antigens) and/or carriers of infrequently occurring antigens (private antigens). The test red blood cells used for the diagnosis of immunisation are selected on the basis of the antigenic profile of the Caucasian population and, therefore, lack private antigens that are common among various non-Caucasian ethnic groups (Table I)²⁸. This leads to an increased possibility of false negative results in the IAT, with the well-recognised clinical consequences. On the other hand, immunisation of non-Caucasian subjects against antigens that are public among Caucasians creates both diagnostic and transfusional problems (Table II)²⁸. Test cells negative for public antigens are rare among the test panels marketed for the diagnosis of immunisation. After all, there is a lack of donors typed for these antigens who could fill this gap. We, therefore, lack not only test red cells to confirm the diagnosis, but also units of red blood cells to be able to give patients risk-free effective transfusion therapy. In other cases the clinical irrelevance of the antibodies involved (Table III and IV)²⁸ means the problems are limited only to the diagnosis.

Table I.

Private antigens present in specific ethnic groups (relevant with regards to haemolytic disease of the foetus and newborn and transfusion reactions).

Ethnic group	Antigen	Frequency
Sub-Saharan Africans (Bantu and Sudanese) and Afro-Americans	Jsa	6%
	STEM	6%
	Rh42	2%
	Rh32	1%
	Goa	2%
Ethnic groups from Asia	Mia	>15% of Chinese and South-East Asians
	Mur	6% of Chinese; 7% of Taiwanese; 9% of Thai
	Hil	6% of Chinese
Amerindians	Dia	5% of Chinese, 12% of Japanese, 36% of South-American Indians

Table II.

Public antigens absent in specific ethnic groups (relevant with regards to haemolytic disease of the foetus and newborn and transfusion reactions).

Ethnic group	Antigen	Frequency
Sub-Saharan Africans (Bantu and Sudanese) and Afro-Americans	U	absent in 2 to 35%
	Jsb	absent in 1%
	Fy3	absent in 68%
	Hy	absent in <1%
	Joa	absent in <1%
	Cra	absent in <1%
	Tc(a−b+c−)	<1%
	Tc(a−b−c+)	<1%
	Ata	<1%
Ethnic groups from Asia	Oh (Bombay)	in some subjects from the area of Bombay
	Para Bombay	Reunion Islands, India
	Yta	absent in 2% of Arabs
	Ge3	absent in 50% of Melanesians
	Inb	absent in 4% of Indians
Amerindians	Dib	absent in 4% of native Americans

Table III.

Private antigens present in specific ethnic groups (not relevant with regards to haemolytic disease of the foetus and newborn and transfusion reactions).

Ethnic group	Antigen	Frequency
Bantu	V	30%
	VS	26–40%
	He	6%

Table IV.

Public antigens absent in specific ethnic groups (not relevant with regards to haemolytic disease of the foetus and newborn and transfusion reactions).

Ethnic group	Antigen	Frequency
Bantu	hrS	absent in 1%
	hrB	absent in 2%

The frequency of these problems is destined to increase with the expansion of immigration and the increase in mixed marriages. It is, therefore, important to know the immunohaematological problems of the immigrant populations and establish co-ordination between various centres in order to deal with them in the best manner.

In order to tackle these problems it is essential to raise awareness among immigrants and encourage them to become blood donors. This would also represent a strong source of social integration.

Such subjects need to be typed for the antigens typical of the ethnic group to which they belong in order to be able to recall them if their red cells become necessary for diagnostic purposes or transfusion use.

Ideally, a large bank of rare serum and red cells should be created in the major Transfusion Services. The various centres could share a database to search for possible reagents for diagnostic use or for units to transfuse held in the other centres dedicated to this activity and could obtain them in exchange.