High Percentage of Evanescent Red Cell Antibodies in Patients with Sickle Cell Disease Highlights the Need for a National Antibody Database

Introduction: From 20 to 50% of patients with sickle cell disease are alloimmunized to red cell antigens from transfusions and/or pregnancies.  Pre-transfusion testing (e.g. antibody screening) is essential to avoid hemolytic transfusion reactions from clinically significant antibodies, such as those to Rh, Kell, Duffy and Kidd antigens, among others.  Unfortunately, alloantibodies may evanesce over time, becoming undetectable or even leading to a negative antibody screen. Furthermore, antibodies to low frequency antigens in the donor pool, mostly from Caucasians, may not be detected because they are not typically expressed on standard reagent red cells used for testing.  Both of these facts contribute to the risk of hemolytic transfusion reactions despite negative pre-transfusion screening.  To mitigate this risk, individual transfusion services maintain antibody histories of all patients indefinitely, and must refer to them prior to issuing every unit for transfusion.  

Patients and Methods: We analyzed data collected from the electronic medical records of every adult (19 and older) patient with sickle cell disease who had one or more positive antibody screen in our institution during a 2-year period from 2013-2015 to determine:  1. antibody specificities; 2. percentage of patients with at least one non-reacting alloantibody; 3. specificities of evanescent antibodies; and 4. number (percentage) of patients with antibodies to low frequency antigens.   

Results: We identified 71 patients, of which 5 were excluded because they only had a cold-reacting anti-M (n=2), anti-Lewis, a warm autoantibody (WAA), or a High Titer Low Avidity (HTLA) antibody (n=1 each).  Thus, 66 patients were included in the analysis (62% females). The age range of the study-group was 19-59 years old (mean ± SD, 33 ± 11 years), similar between males and females (p = 0.43).  Males tended to have an antibody screen ordered more often during the study-period, with a trend toward statistical significance (9.4 versus 5.0 times; p = 0.06).  The total number of clinically significant alloantibodies was 218 with mean and median number per patient of 3.3 and 3.0, respectively.  In addition, 16 patients also had a WAA. Anti-E was the most common alloantibody, followed by anti-C and anti-K; the table shows the antibodies identified at least once in 10% or more of the patients, as well as how often they were not reacting.  Of 9 patients with anti-D, 7 were Rh positive, consistent with the propensity of African-Americans to express a D variant. Of note, 30 patients (45%) had antibodies to one or more low frequency antigen such as Cw, V, Vs, Goa, Jsa, Kpa, Lutheran b, Wra, and Ytb; 6 patients (9%) only had alloantibody(ies) to these antigens. Only 12 patients had a positive antibody screen every time they were tested. However, they had significantly fewer tests (average and median of 2, range of 1-6) compared with 54 patients with at least one antibody screen completely negative (average and median of 5 tests, range of 1-30) (p < 0.0001).

	Rh system					Kell		Duffy		Kidd	MNSs	Other
Antibody specificity	D	C	E	V	Goa	K	Jsa	Fya	Fyb	Jkb	S	AUS
Number of patients	9	30	40	14	7	20	15	18	9	12	17	7
Percentage of total antibodies	4%	14%	18%	6%	3%	9%	7%	8%	4%	5%	8%	3%
Times not reacting	5	16	21	8	3	12	12	10	7	8	12	3
Percent evanescent	56%	53%	53%	57%	43%	60%	80%	56%	78%	67%	71%	43%

AUS-antibody of unknown specificity; Fya-Duffy a; Fyb-Duffy b; Jkb-Kidd b;

Conclusions: Our results demonstrate that 81% of patients with sickle cell disease with a history of red cell alloantibodies had at least one test in which the antibody was not detectable. Delayed hemolytic transfusion reactions are the major risk of evanescent alloantibodies not known to the transfusion service preparing the units. Such reactions may even be fatal, especially if not promptly recognized, since they may be confused with a pain crisis. In addition, we noticed that almost half of the patients had antibodies to low frequency antigens that could have been missed during pre-transfusion testing. Considering that patients with sickle cell disease often suffer from lack of continuity of care, unawareness of their antibody history may lead to life-threatening transfusion reactions. We suggest that a national, or even regional, database of red cell alloimmunization in this patient population is warranted for increased transfusion safety.