ABO subgroup incompatibility with severe hemolysis after consecutive allogeneic stem cell transplantations

Abstract Allogeneic hematopoietic stem cell transplantations (HSCTs) represent a curative strategy for treating hematologic malignancies yet bear dangerous and frequently life‐threatening complications including the development of graft‐versus‐host disease. Here, we present a case of a patient that suffered from relapsed/refractory multiple myeloma, a hematologic neoplasm characterized by clonal proliferation of malignant plasma cells in the bone marrow. During the course of his disease, the patient underwent consecutive allogeneic HSCTs, during which he developed a clinical meaningful and hitherto unreported ABO subgroup incompatibility, leading to persistent hemolysis. Testing for ABO subgroups during donor selection, especially after consecutive allogeneic HSCTs, may therefore aid to prevent these complications.


A. Disease characteristics
Disease stage   (Table 1A). Moreover, the malignant plasma cells had expanded in the bone marrow and harbored multiple new genetic aberrations, evolving from the known singular trisomy 17 to a complex aberrant karyotype (51, XY, +3, +9, +11, +15, +17). Subsequently, treatment according to the VRD regimen (bortezomib, lenalidomide, dexamethasone) was initiated [1]. Unfortunately, the patient progressed after one cycle of VRD. We therefore escalated the treatment by administering four cycles of the VCD regimen con- In light of the high-risk disease profile, the patient's young age, and good clinical condition (ECOG 0), a second allogeneic HSCT was considered and after conditioning with fludarabine, treosulfan, and ATG ultimately performed in October 2018 (Table 1B). This resulted in a complete remission (CR) and minimal residual disease negativity (Figure 1A). Administering cyclosporine and mycophenolate mofetil as graft-versus-host disease (GVHD) prophylaxis, the patient had no signs or symptoms of acute GVHD (Table 1B).
This second donor was an HLA-matched unrelated donor bearing blood group A Rh -( U/l) ( Figure 1B). We next sought to identify the causative alloreactive antibody using a series of detection methods. First, we performed a standard antibody screening assay (gel column agglutination test with three test cells) and secondly with a more elaborate screen test panel (further 27 test cells). Then, we scrutinized our assay for possible interference by daratumumab by performing these assays with and without DTT. However, elution of the antibodies bound to the patient's erythrocytes yielded no specificity in any of the used test erythrocytes at our disposal. We therefore concluded that, with the last dosing of daratumumab 6 months ago, daratumumab was no longer interfering in our assay [2].
We first started to treat the hemolysis with corticosteroids (initial dose of prednisone of 2 mg/kg/day) representing the first-line treatment for autoimmune hemolytic anemia. Since this did not alter the course of the hemolysis ( Figure 1B), we sought to target the antibodyproducing B-cell populations by weekly administering rituximab (375 mg/m 2 ) throughout 4 weeks [3,4]. After these measures did not yield a sufficient improvement of the hemolysis, we aimed to deplete the antibody reservoir by initiating a 3-weekly plasmapheresis [5].
Moreover, since all these measures did not bear any sign of success, an off-label treatment with daratumumab (16 mg/kg, weekly for 3 weeks) was started to specifically deplete the plasma cell pool responsible for production of the alloreactive antibodies [6]. Still, crossmatching of A Rh -PRBCs with the patient's plasma was positive in most, but not all PRBCs tested. Surprisingly, those PRBCs demonstrating a negative crossmatch were all blood group A 2 Rhbased on a capture solid phase technology. At the same time, we identified an anti-A antibody in the serum of the patient. Subsequently, we switched to transfusing 0 Rh -PRBCs, which was well tolerated and resulted in a swift return to normal levels of LDH and haptoglobin ( Figure 1B). Further investigation into the first donor's blood group revealed that he was A 1 positive (the most prevalent subgroup in Germany) whereas the second stem cell donor was A 2 positive (Table 1B) [7].
Immune-mediated hemolytic anemia is frequently observed and a common complication after allogeneic HSCT [8], especially in the context of passenger lymphocyte syndrome (PLS) comprising hemolysis associated with minor ABO incompatibility between the donor and the recipient (most common A + recipient, O + donor) [9]. PLS usually occurs immediately (between 5 and 15 days) post-transplant and is generally moderate but has also been described to cause fatal hemolysis and potentially death [9,10].

DISCUSSION
In conclusion, we note that the patient developed a clinically meaningful Coombs positive hemolysis due to the generation of an alloreactive antibody against the A 1 antigen following a second allogeneic HSCT. Erythrocytes in antibody screen assays are usually O Rh + , distracting from the possibility of an alloantibody directed against A 1 .
Based upon our case, we speculate that mismatches within the ABO subgroups may play a more significant role in the future with increasing numbers of consecutive allogeneic HSCTs and broadening pools of potential donors [11]. Implementing rigorous blood group testing dur-ing the process of donor selection may therefore prevent complications arising from for ABO subgroup incompatibilities, especially after consecutive allogeneic HSCTs.

METHODS
Blood group determination and serological testing were performed using automated platforms using the solid-phase technology (Immu-corNeo Galileo analyzer, ImmucorNeo, Dreieich, Germany) and stan-