Transfusion considerations concerning patients with anti-hr^B and confounding co-morbid conditions: pregnancy and coagulopathy

Introduction

Alloimmunisation to hematopoietic antigens is generally considered to arise from three sources: pregnancy, blood transfusion, or organ transplantation. The most common red blood cell (RBC) antigens of clinical significance are those of the ABO and Rhesus (Rh) blood group systems. The Rh blood antigens comprise two linked genes showing a high degree of homology. For historical reasons, the genes are named RHD and RHCE. RHD encodes the ubiquitous RhD protein, for which only one common allele exists, whose absence in Rh-negative individuals elicits a strong antigenic response when challenged by exposure to allogeneic RhD-positive RBC. The well-known alloimmunisation reaction can lead to frank complement-mediated haemolysis in individuals with a past history of exposure; in pregnancy, this can lead to haemolytic disease of the foetus and newborn, or in severe cases, erythroblastosis foetalis. By contrast, at the linked RHCE locus, two intragenic antigenic determinants exist and each has a pair of alleles, designated C, c, E, and e^1–4. While conventional variants at these loci do not usually elicit a strong immune response, rare polymorphic alleles at these loci may result in haemolytic disease of the foetus and newborn when allogeneic RBC challenge occurs.

One infrequent allele present in African-Americans of South African or Caribbean descent is designated hr^B or Rh31⁵. The hr^B variant is thought most commonly to result from three amino acid changes in the RHCE allele that appear to confer a dominant allointolerance to common RHCE alleles^6,7. The antigen conferring sensitivity in hr^B-negative individuals has recently been more extensively studied by molecular characterisation of RHCE haplotypes in alloimmune individuals. This molecular analysis has revealed molecular heterogeneity among genetically susceptible individuals carrying the associated (C)ce^S haplotype⁸.

Features of haemolysis in offspring of hr^B mothers are minimal⁵ and hr^B is generally thought not to cause clinical sequelae of haemolytic disease of the foetus and newborn³; nevertheless, avoidance of alloimmunisation is recommended. As hr^B-negative units are extraordinarily rare, family matches or Rh^null units from rare donor pools are often solicited. We report two instances of hr^B alloimmunity, one in a multiparous patient with a twin pregnancy and the other in a male with sickle cell disease and coagulopathy.

Case report 1

A 26-year old African-American female presented to an outside facility with a complaint of abdominal pain. Evaluation revealed that the patient had a twin gestation. This was further complicated by several additional factors: absent prenatal care, one foetus in a breech position, advanced cervical dilatation, and severe anaemia with an admission haemoglobin (Hb) of 5.7 g/dL. The patient was transferred to our facility for further evaluation and management.

The patient had a past medical history of multiple prior RBC transfusions without any significant reactions (medical necessity for previous transfusions not documented). Her obstetric history included five prior pregnancies: two full term, two preterm, and one elective abortion. Her past surgical history included dilatation and curettage for the aforementioned elective abortion. Based on her last menstrual period, the estimated gestational age was 31 weeks and 6 days.

On presentation, the patient’s physical examination was significant for a gravid, non-tender abdomen. A complete blood count showed severe anaemia (Hb 5.1 g/dL), a decreased RBC count, and a mean corpuscular volume of 58.8 fL. In addition, the patient’s RBC distribution width was increased at 20.2%. In accordance with the complete blood count values, a peripheral blood smear showed anisopoikilocytosis with mainly hypochromic, microcytic RBC, and several target cells and ovalocytes. Hb electrophoresis was normal with 87.7% HbA.

Serum iron level, total iron-binding capacity, percentage iron saturation and ferritin level were consistent with iron deficiency. Stool samples were negative for occult blood. Given the patient’s severe anaemia and gravid state, intravenous iron, sucrose and epoetin-α were administered. Prior to transfusion, the patient was phenotyped as A-negative, her antibody screen was positive, and autocontrol was negative. Her complete Rh phenotype was Ccee. Anti-D antibodies were noted but the remainder of the screening cell panel (Panocell-20, Immucor, Norcross, GA, USA) was inconclusive. A blood sample was, therefore, sent to the American Red Cross Blood Services, Southern Region in Atlanta (ARC-ATL) for further characterization.

The ARC-ATL report concluded that a direct antiglobulin test was negative and that the patient had anti-D, anti-E, and anti-hr^B antibodies with titres of 32, <1, and 4, respectively. Specifically, agglutination was observed in the anti-globulin phase of testing using PEG enhancement (Gamma PeG, Immucor, Norcross, GA, USA) and was unchanged by papain or 0.2 M dithiothreitol treatment of the test cells. The patient’s serum was adsorbed once at 37 °C with selected allogeneic red cells following treatment with papain to reveal antibody specificity. The adsorbed serum was then tested with reagent red cells and showed anti-D, as well as weak-D with two of four RhD-negative red cells.

The patient’s gravid state and the possibility of an upcoming Caesarean section given the breech presentation of foetus B necessitated urgent transfusion of packed RBC to the patient. The Transfusion Medicine Service was, therefore, consulted and the nearest compatible packed RBC unit was identified in Arkansas. The Arkansas unit tested type O-negative, E-negative, and hr^B-negative; an hr^B phenotype was confirmed via previously characterised hr^B antisera in the Red Cross archives. The patient received this leucoreduced unit after PEG-enhanced cross-match compatibility had been confirmed with anti-human globulin (Immucor, Norcross, GA, USA). The patient had no complications during the transfusion and no post-transfusion reaction was seen.

Genetically related family members were identified and tested for RBC compatibility. Fortunately, the patient’s mother was found to be compatible and was able to donate, and frozen blood was stored for her daughter. Specifically, the mother tested negative for hr^B with a previously characterised source of anti- hr^B antiserum in the Red Cross archives at low-ionic-strength indirect antiglobulin phase testing.

In addition, another compatible unit was obtained via a rare donor registry and stored frozen in case of future need. During this time, the patient’s preterm labour was successfully abated with magnesium sulphate tocolysis. Three weeks after admission, at an estimated gestational age of 35 weeks, preterm labour ensued. At this time, the woman’s Hb was 10 g/dL, and she delivered viable twins by a primary Caesarean section with two compatible units of packed RBC available if needed.

The neonates were found to be dichorionic, diamniotic and both placentas were submitted to pathology for evaluation. Both placentas were in the 60^th percentile for gestational age by weight. No significant histopathological changes were noted, except a focal area of villous chorangiosis in the placenta of twin B.

The patient had an uneventful post-operative course with a post-operative Hb of 8.4 g/dL and was provided oral iron tablets and discharged home after 5 days. No further transfusions were performed. Both neonates had a protracted post-natal course with initial Hb values of 13.3 g/dL and 14.3 g/dL. They were transferred to the neonatal intensive care unit because of low birth weight and respiratory immaturity, which subsequently improved. Cord RBC from babies A and B were typed as A-positive and O-positive, respectively. Both infants had positive antibody screens and direct antiglobulin testing at 2+ agglutination; this was attributed to passive immune globulin transfer from the mother and not further investigated. Baby A’s bilirubin was 4.3 mg/dL at 16 hours of life and 13.7 mg/dL at day 4 of life; the corresponding values for baby B were 4.5 mg/dL and 14.6 mg/dL, respectively. All these values were below the threshold indicating the need for phototherapy. Both babies were discharged approximately 2 weeks after birth in good health.

Case report 2

A 29-year old African-American male with sickle cell disease presented at the emergency department with a 3-day history of cough and nasal congestion. The symptoms were worsening progressively, and the patient was producing yellow sputum accompanied by persistent sharp chest and leg pain. The patient did not report any dizziness, shortness of breath, abdominal pain, nausea, vomiting or diarrhoea. In the emergency department, the patient was afebrile, but appeared jaundiced. He had a past medical history of bilateral deep vein thromboses and pulmonary embolism all within the preceding 5 months and was receiving warfarin anticoagulation. The patient had a remote history of avascular necrosis of the hip, cholelithiasis and laparoscopic cholecystectomy. He had no known drug allergies, and his medications included folic acid, hydroxyurea, oxycodone/acetaminophen, as well as warfarin.

At physical examination the patient was awake, alert and oriented, and in no apparent respiratory distress with unremarkable vital signs. He was jaundiced and pharyngeal erythema was noted. Auscultation revealed clear lungs, bilaterally, and normocardia in sinus rhythm. His abdomen was soft, non-tender, non-distended, and had positive bowel sounds. By palpation neither the spleen nor liver was enlarged. His extremities appeared jaundiced; however, no other positive findings were discovered.

A complete blood count showed normocytic anaemia and leucocytosis with mild left shift. No infiltrates were seen on a chest X-ray. The patient was given intravenous fluid support, levaquin for an upper airway infection, and low molecular weight heparin for prophylaxis of deep vein thrombosis and pulmonary embolism. The patient’s International Normalised Ratio was considered to be subtherapeutic, and the warfarin dose was increased.

The patient, whose last transfusion was 4 months earlier, was typed as B-positive, and had a positive antibody screen with a negative autocontrol. The RBC panel antibody work-up showed alloimmune anti-E as well as a possible anti-c specificity. The ARC-ATL was consulted and concluded that the blood sample was B-positive with a negative direct antiglobulin test. The patient had the alloimmune profile described above, including anti-E, but a previously identified anti-hr^B -like antibody was not detected. Although the direct antiglobulin test was negative, the immunohaematological work-up revealed the presence of warm autoantibody. The warm auto-reactivity was only 1+ via indirect antiglobulin testing with ficin pre-treatment.

It was decided, if necessary, to transfuse with ABO/Rh compatible units typing negative for E and hr^B antigens. As discussed below, units that are negative for both E and hr^B are exceedingly rare. Fortunately, this patient did not require imminent blood cell support and was discharged home after resolution of his pain crisis the following day.

Further review of the patient’s blood bank record revealed that 12 years previously, another inconclusive antibody screen had also prompted further testing at the same reference laboratory, and at that time, a request for 6 units of cross-matched compatible blood had been made. It was then that this patient’s anti-hr^B - like antibody was first discovered. The transfusion recommendation made by the reference laboratory at that time was to transfuse ABO/Rh compatible e-negative or hr^B-negative blood. It was also commented that: (i) locating possible donors for future transfusions could be very difficult because of the likelihood that this sickle cell patient would develop additional antibodies; (ii) some individuals with partial e-antigen may also produce complex antibodies which react with all red cells with normal Rh antigens; and (iii) since the immune response in anti-hr^B-positive patients is highly variable, re-evaluation of the antibody specificities present in the serum would be required prior to shipping rare compatible units. At the time, the laboratory was able to provide five units of O-positive, e-negative, cross-match-compatible units.

Discussion

Hr^B and hr^B were first described in 1972 by Shapiro et al.⁵ These antigens are in the Rh blood group system and, according to the International Society of Blood Transfusion, hr^B and Hr^B are denoted as Rh31 and Rh34, respectively^1–4. Both of these antigens are described as high frequency variants with hr^B present in 98% of whites and Hr^B in >99.9% of whites⁹.

Anti-Hr^B and anti-hr^B are rare alloantibodies that are typically found in African-American individuals, especially those with sickle cell disease requiring frequent transfusion support¹⁰. Both patients described in this report had a history of multiple transfusions, the obstetric patient secondary to her severe iron deficiency, and the sickle-cell patient due to the natural history of disease treatment.

While anti-Hr^B is reputed to be clinically significant, it has been assumed that only a high-avidity anti-hr^B would result in significant haemolytic sequelae. Anti-hr^B is described as an alloanti-e-like antibody and, therefore, patients with anti-hr^B are typically transfused with e-negative (R²R²) red blood cells¹¹. However, these patients readily develop anti-E^10,11, as did both patients described in this report, which further complicates their subsequent care.

While anti-hr^B is rare and typically not associated with haemolytic complications³, the theoretical risks of intravascular haemolysis and HDFN require that RBC phenotypic compatibility be maintained, if possible. Strategies include the triage of rare blood types such as r’r patients who are hr^B-negative, and canvassing rare donor registries for known hr^B-negative donors. If compatible hr^B-negative RBC units are not available, and time permits, testing of genetically related family members for compatibility should be undertaken, as in the first case. As for any patient with a rare RBC phenotype who is likely to undergo repeated transfusion therapy, precautions should be undertaken during indicated transfusions to avoid the generation of additional alloantibodies.

Gammon and Velasquez proposed an algorithm to locate units of blood for African-American recipients with sickle cell disease, anti-hr^B, and anti-D¹². All identified African-American first time donors underwent Rh phenotyping (D, C, c, E, e), performed by a primary reference laboratory. The specimens from African-American r’r individuals were then sent to a secondary reference laboratory for hr^B phenotyping after each subsequent donation. They found that out of 24 donors who typed r’r, seven were hr^B-negative. It had previously been demonstrated that D-negative, hr^B-negative donors may be found among African-American r’r individuals¹³. On a much larger scale, the American Rare Donor Program consists of more than 35,000 rare donors from the United States, Puerto Rico, and Milan, Italy. This program has been instrumental in identifying units of blood that are negative for high-prevalence Rh phenotypes such as hr^B7,¹⁴.

Based on the considerable molecular heterogeneity of the RCHE locus, recent studies advocate molecular characterisation of selective exons in the RCHE gene among donors identified in such rare donor registries⁸. RHD and RHCE are two highly homologous genes that encode proteins that display D, and C or c and E or e antigens, respectively. At present, more than 120 different RHD variants, and approximately 50 different RHCE variants have been noted⁷. Patients, such as those described here, who are of African descent and have persistent anaemia or sickle cell disease, are often dependent on chronic transfusion therapy, and may have variant antigens that may make them more likely to produce unusual antibody combinations. In these patients, the finding of concurrent anti-E and an anti-hr^B-like antibodies, though unusual, was not, therefore, entirely unexpected. In such patients, identifying additional antibodies and, subsequently, finding compatible blood proved a formidable challenge¹⁰.

These two cases highlight the difficulty entailed in blood bank laboratory testing of blood samples from certain individuals and the complexities involved in obtaining cross-match-compatible units for them. An algorithmic approach to Rh phenotyping of African-American donors provides a viable means to quicker, more efficient access to blood products for this particular population of patients.