Prevalence of erythrocyte alloimmunization in polytransfused patients

Objective: To determine the incidence and the rate of red blood cell alloimmunization in polytransfused patients. Methods: A polytransfused patient was defined as having received at least 6 units of red cell concentrates during a 3-month period. The records of all patients (n = 12,904) who had received red blood cell units were examined retrospectively by searching the computer database at Hospital Israelita Albert Einstein in São Paulo, Brazil, over a 6-year period, between 2003 and 2009. results: During this time, 77,049 red cell concentrate transfusions were performed in 12,904 patients. There were 3,044 polytransfused patients, 227 of whom (7.5%) presented with irregular erythrocyte antibodies. The prevalence of alloantibody specificity was: Anti-E>anti-D>anti-K>anti-C>anti-Dia>anti-c>anti-Jka>anti-S in 227 polytransfused patients. We found combinations of alloantibodies in 79 patients (34.8%), and the most common specificities were against the Rh and/or Kell systems. These antibodies show clinical significance, as they can cause delayed hemolytic transfusion reactions and perinatal hemolytic disease. About 20% of the patients showed an IgG autoantibody isolated or combined with alloantibodies. Interestingly, a high incidence of antibodies against low frequency antigens was detected in this study, mainly anti-Dia. conclusion: Polytransfused patients have a high probability of developing alloantibodies whether alone or combined with autoantibodies and antibodies against low frequency antigens. Transfusion of red blood cells with a phenotype-compatible with RH (C, E, c), K, Fya, and Jka antigens is recommended for polytransfused patients in order to prevent alloimmunization and hemolytic transfusion reactions.

Descritores: Transfusão de sangue/efeitos adversos; Eritroblastose fetal; Eritrócitos/imunologia intrODUctiOn One of the risks of blood transfusion is the formation of antibodies against one or more erythrocyte antigens resulting from genetic disparities between donor and recipient (1) .The risk depends on the recipient exposure to the foreign antigen and its immunogenicity (2) , defined as the ability of a given antigen to stimulate antibody production in a patient lacking the antigen (3) .Whether the recipient immune system will react depends on genetic or acquired factors related to the patient, dose, number, and frequency of transfusions (4)(5)(6) .Clinically significant red blood cell (RBC) alloantibodies are developed in more than 30% of patients receiving multiple transfusions, a situation that can pose major problems in the case of long-term transfusion therapy.Several authors found that erythrocyte alloimmunization occurs mainly after the first transfusions (7,8) .Knowledge of clinical conditions that predispose to alloimmunization is important in two ways: it may influence patient management and may lead to a better understanding of the etiology of the transfusion reaction (9) .
When clinically significant non-ABO antibodies are detected in the plasma of patients requiring RBC transfusions, transfusion services must find and administer RBCs lacking the corresponding antigens.Thus, in transfusion medicine, much time and effort are spent in detecting and identifying blood group antibodies.Besides ABO, the most clinically significant antibodies are those in Rh, Kell, Duffy, and Kidd blood group systems (10) .
When non-ABO blood group antibodies disappear, patients are at risk of unknowingly receiving incompatible RBC transfusions and developing delayed hemolytic transfusion reactions (DHTRs).DHTRs are probably the least recognized and most underreported type of transfusion reactions, partly due to their temporal dissociation from the causative transfusion (11) .
The prevalence of blood group alloantibodies was reported in several study populations, including hospital-based patients, patients with hematological disorders requiring chronic transfusion therapy, and blood donors (9,12) .
Giblett calculated the relative immunogenicity of a number of RBC antigens compared to the K antigen.The author compared the frequency that certain particular antibodies are encountered with the calculated probability of exposure.Based on her calculations, the relative likelihoods of non-D blood group formation are K(0.05)> c(0.0205) > E(0.0169) > Fy a (0.0023) > Jk a (0.0007) (3) .
This retrospective study reports the prevalence and rate of alloimmunization in polytransfused patients at Hospital Israelita Albert Einstein (HIAE), São Paulo, Brazil.

OBJectiVe
To determine the incidence and the rate of red blood cell alloimmunization in polytransfused patients.

MetHODS
The records of all patients (n = 12,904) who received RBC units were examined retrospectively by searching the computer database from HIAE, a 530-bed general hospital in São Paulo (SP), over a 6-year period, between 2003 and 2009.
A polytransfused patient was defined as that receiving at least 6 RBCs units in 3 months.We selected all patients with detected RBC auto-and/or alloantibodies.The data collected included patient age, sex, history of RBC transfusion and pregnancy, RBC extended typing, and RBC antibody detection results.

immunohematology testing
Blood samples from patients submitted to blood transfusions were screened for RBC alloantibodies using a selected three-cell set of reagent RBCs for antibody detection.The technique for antibody detection involved the use of 25 μL of serum and 50 μL of 0.8-percent RBCs in low ionic-strength solution gel tests (DiaMed AG, Cressier Switzerland, and Grifols, Barcelona, Spain.)Antibody identification was accomplished with commercial panels of cells tested by similar methods or additional techniques (e.g., polyethylene glycol and enzyme) whenever needed.If specificity could not be clearly determined, the blood sample was sent to our reference immunohematology laboratory for further analysis.The results of both antibody screening and antibody identification were valid for 72 hours (a transfusion episode).Complete crossmatching, including an indirect antiglobulin phase, was performed.

Statistical analysis
Data were analyzed using descriptive statistical analysis by means of absolute frequencies and percentages, and the results were organized in tables.

Patient characteristics
During the study period, we examined 12,904 patients who received 77,049 RBC units.Of these, 3,044 belonged to the selected polytransfused patient category, 227 (7.5%) of whom developed alloantibodies and/or autoantibodies.We found more alloantibodies in men than in women in that population, including patients who already had a RBC antibody before their first transfusion in this hospital.The risk of alloimmunization, defined as the total number of alloantibodies divided by the total number of units transfused, was 0.3%.Patient age did not influence the alloimmunization rate.
In this retrospective study, we observed 57 (3.1%) antibodies to low-incidence antigens (LIA-Ab) alone or in combination, and the anti-Di a was the most prevalent antibody found (Table 3).The single LIA-Abs found was developed after the patient's first transfusion.Anti-E, -C Antibody specificities combinations found in 28 patients Cruz RO, Mota MA, Conti FM, Pereira RAD, Kutner JM, Aravechia MG, Castilho L lia-ab specificities lia-ab associated antibodies Patients

DiScUSSiOn
This retrospective study was undertaken to determine the incidence of antibody formation after RBC transfusion.We demonstrated 349 unexpected antibodies in 3,044 patients characterized as polytransfused.
No influence of age was demonstrated; however, a greater prevalence of alloimmunization among males was found, with a male:female ratio of 1.5:1.The high prevalence of antibodies within the male population was unexpected given that pregnancy as an alloimmunization stimulus was not a factor (1,3,6) .
As noted above, the majority of antibodies detected among males were preexisting antibodies, and most patients had a prior history of surgery.Of all patients 7.5% formed antibodies after consecutive transfusions and 35% of those showed a combination of two or more antibodies.Antibodies within the Rh system (anti-E, -D, -C) combined with anti-K were the most common specificities found.Anti-E was implicated in 16/20 patients with multiple antibodies.
Kidd antibodies represented 5% of the antibodies (17 patients) found and were more frequent than anti-Fy a (5 patients).Fifteen patients developed anti-Jk a shortly after transfusion, suggesting a rapid primary immune response.In two patients, anti-Jk a was demonstrated in the eluate, indicating that these patients probably experienced hemolysis.Heddle et al. prospectively studied alloimmunization after transfusion in 2,082 patients and found 32 new RH, K, FY, JK, or MNS antibody specificities, 9 (28%) of which were anti-Jk a , detected after a median period of 4 days (range, 1 to 94 days) after transfusion.A positive antiglobulin test was found in 31% of the newly alloimmunized patients.Only three patients showed hemolysis, one of which was an anti-Jk a case (13) .Schonewille et al. found 2.8% of anti-Kidd in 1,795 patients with 2,257 antibodies (14) .
These findings differ from those found by Schonewille et al. regarding alloimmunization to the Fy a antigen.A prevalence of 1% of anti-Fy a was found while these authors reported 7.3% of anti-Duffy in their study (14) .
Twenty patients (8.8%) developed multiple antibodies.Several studies reported on the presence of multiple RBC antibodies in chronically transfused patients, with up to a fourfold increased risk of multiple antibodies as compared to the general risk of antibody formation (15)(16)(17)(18)(19)(20) .
The high incidence of RBC alloimmunization found by different authors has posed the question as to whether patients who are transfusion-dependent over a long period, such as those with sickle cell disease or thalassemia, should receive blood matched for antigens other than ABO and D in an attempt to prevent the formation of RBC alloantibodies (21- 24) .The incidence of alloimmunization in patients receiving extended antigen-matched RBC transfusions was quoted as 0 to 6.4%.Extended antigen matching (e.g., c, E, and K) to prevent the formation of the majority of RBC antibodies in chronically transfused patients has been advocated for selected patient populations.Some reports applying this policy showed a significant drop in

Figure 1 .
Figure 1.Incidence of red cell antibodies in patients with multiple transfusuons in HIAE the period 2003 to 2009

table 1 .
Antibody specificity and frequency in 227 alloimmunized patients Figure 2. Associated antibodies in polytransfused patients

table 3 .
Low incidence antibody (LIA-Ab) specificities found alone or in combination in 44 patients LIA-Ab: low incidence antibody.