Introduction: Patients with sickle cell anemia require transfusions for severe complications of their underlying disease. Both allo- and autoantibodies to red blood cells are reported in patients receiving transfusions of packed red blood cells (PRBCs). Since 1978, we have provided PRBCs with extended antigen matching for patients requiring intermittent or chronic transfusions. Extended matching reduced the rate of alloimmunization but patients continued to develop autoantibodies.

Methods: Records of patients with sickle hemoglobinopathies enrolled in the Colorado Sickle Cell Treatment and Research Center between December 31, 1993 and January 1, 2006 were reviewed under a protocol approved by the COMIRB at UCDHSC. At enrollment, serologic testing was completed on patients for the following blood group antigens: ABO; Rhesus (C,c,D,E,e); Kell (K,k); Duffy (Fya,Fyb); Kidd (Jka,Jkb); Lewis (Lea,Leb); and MNS (M,N,S,s). Donors were typed for the same antigens. For all transfusions, a perfect match was sought. However, when exact matches could not be found, mismatches were allowed when necessary for Fyb and MNS because of lower risk of sensitization and for Le because of infrequent hemolytic transfusion reactions. Antibody screens were completed as part of crossmatch technique at the time of each transfusion by standard methods. When an antibody screen was positive, standard antiglobulin and enzyme techniques and cell panels were used to identify the antibody.

Results: Over the 13 years, 104 patients received 6,978 PRBC units under the matching protocol when transfusion was required. The average was 68 units per patient (range 1–519). Of this group, 62 (60%) received intermittent transfusions and 42 (40%) chronic transfusions (repeated transfusions every 3–6 weeks for at least 6 months). Eleven patients had complications requiring both schedules of transfusions. Seven patients (6.7%) developed alloantibodies; 3 anti-D (mosaic), 1 anti-Lea, 1 anti-Kpa and 2 anti-M. Since the three cases of anti-D would have been typed as Rh(D) positive and would have occurred with any program for antigen matching by serologic techniques, the rate of alloimmunization was reduced from 3.5 antibodies/100 transfusions (historic control before antigen matching) to 0.06 per 100 transfusions with extended matching. Although there was a reduction in alloantibodies, 12 patients had positive DAT associated with an autoantibody. Only one patient with an autoantibody had an associated alloantibody. Of eleven patients with autoantibodies only, eight developed warm autoantibodies; 4 with anti-e specificity, 1 with anti-E specificity and 3 panagglutinins. Six antibodies were cold, 3 with I specificity and 3 non-specific. Nine patients had 1 antibody and 3 patients had two antibodies.

Conclusion: Although extended matching reduced the rate of alloimmunization, autoantibody formation was not altered. Patients on both chronic and intermittent transfusions develop both warm and cold autoantibodies to red cells.

Author notes

Disclosure: No relevant conflicts of interest to declare.