Autoantibody to High Prevalence Kel System Antigen

Background/Case Studies:

A 79-year-old male was admitted to our institution with anemia, with a hemoglobin nadir of 5.5g/dL, and no recent transfusion history. Given a remote history of occult colonic bleeding, the working diagnosis was GI bleeding. The referring blood bank had reported “possible antibody to high prevalence antigens of the Kell system, suspected McLeod.” The American Rare Donor Program had been contacted and confirmed no Ku or KX negative blood was available in the United States.

Study Design/Methods:

Initial testing by our immunohematology reference lab showed that the patient's plasma reacted with all cells tested, including the patient's own red blood cells (RBCs). The patient had a positive DAT (W+ poly, W+ IgG and negative C3d). Further testing showed the plasma and eluate prepared from the patient's RBCs did not react with cells treated with dithiothreitol (DTT) and did not react with an RBC designated as K₀. Red cell morphology and creatinine kinase measurements were not consistent with the McLeod phenotype. The antibody was interpreted as having anti-Ku specificity. A serologic phenotype was performed on the patient's RBCs following treatment with chloroquine diphosphate and showed the following results: K-k-, Kp(b-), Js(b-) and Ku-. A molecular phenotype using Immucor Precise type gave the following results: K-k+, Kp(a-b+), Js(a-b+).

Results/Findings:

A sample was sent to the American Red Cross National Molecular Laboratory for KEL-cDNA analysis and genotype for KX blood group system. The probable genotype was KEL*02/KEL*02 with a predicted phenotype of K-k+, Kp(a-b+), Js(a-b+). The probable genotype was XK*01 (homozygous or hemizygous) with a predicted phenotype of Kx+. Given the positive DAT and the absence of an identifiable source of GI bleeding, a diagnosis of autoimmune hemolytic anemia with suppression of Kell system antigens was made. The patient was initially treated with 100mg oral prednisone daily for two weeks with subsequent weekly taper, four doses of 1g/kg IVIg daily, four doses of rituximab weekly, and three doses of daratumumab on weeks 1, 2 and 4. During the initial work-up and treatment, the patient did not receive a transfusion.

Testing was repeated 3 months later: the DAT remained W+ and the antibody reactivity now demonstrated apparent anti-Js^b specificity. However, serologic antigen typing using chloroquine diphosphate now showed the following results: k+, Kp(b+), Js(b+) and Ku+. At this time, the patient's hemoglobin increased to a maximum of 11.6g/dL.

Unfortunately, the patient relapsed about 4.5 months after his initial presentation. He was restarted on 100mg oral prednisone daily for two weeks followed by weekly taper. This led to only a modest improvement in his hemoglobin concentration. The DAT and phenotyping results remained unchanged but the antibody identification now demonstrated apparent anti-k (Cellano) specificity during this period. As the patient continued to have symptomatic anemia, in the setting of massive splenomegaly, the was decision made to proceed to splenectomy. During the operation, the patient received two K- units that were otherwise not typed for additional Kell system antigens. In the four weeks post-transfusion, there was no evidence of an acute worsening of his hemolytic anemia.

Conclusions:

This is a case of an autoantibody with apparent specificity to a high prevalence KEL antigen(s). The antibody was originally thought to be an alloantibody, requiring Ku- or Kx- RBCs. The KEL system antigens were depressed due to the patient's disease state and were re-expressed following treatment with multimodal immunomodulatory treatment. Specificity of the antibody appeared to change during therapy reacting more strongly with different high incidence Kell system antigens. When the patient was transfused with K- RBCs later in his disease course, there was no acute worsening of hemolysis. This finding is consistent with an autoantibody rather than an alloantibody to KEL antigen(s). This case shows the importance of incorporating molecular testing along with serology.

No relevant conflicts of interest to declare.