Red Blood Cell Depletion for Pediatric Major ABO Mismatch BMT: Evaluation of the Risk of Hemolysis and Comparison of Two Techniques

Introduction: Major ABO mismatch occurs when the recipient has preformed isoagglutinins against the donor. Since unmanipulated bone marrow transplant (BMT) grafts contain donor red blood cells (RBCs), major ABO mismatch BMT infusions can cause hemolytic reactions that lead to renal dysfunction. To prevent this complication, RBC depletion of the graft can be performed by various techniques. Strong evidence-based recommendations regarding RBC depletion are lacking. In particular, it is unclear what residual volume of incompatible RBCs in the graft is safe, especially for pediatric patients.

Methods: We conducted a retrospective cohort study of pediatric patients who had major ABO mismatch BMT at a single center. No patients had pre-BMT plasmapheresis to reduce isoagglutinin titers. RBC depletion was performed on the BMT grafts with either hydroxyethyl starch (HES) sedimentation or Ficoll density gradient separation. All patients received similar supportive care around the BM infusion that included hyperhydration and premedication with acetaminophen, diphenhydramine, and hydrocortisone. Serum creatinine the day after the BM infusion was compared to creatinine the day of the BM infusion. Since 2006 all patients had urine screened for blood post-BMT.

Results: Sixty-three patients were identified who received a major ABO mismatch BMT between 9/2004 and 6/2015 (39 had RBC depletion with HES, 24 with Ficoll). Compared to Ficoll RBC depletion, patients who received BM grafts that had RBC depletion with HES received significantly more incompatible RBCs (Table 1). Hemoglobinuria was significantly more common in HES patients, but evidence of post-BM infusion renal impairment was not (Table 1). All patients had donor engraftment with a similar time to neutrophil engraftment for both groups (Table 1). Considering only the HES patients, 8/8 (100%) patients with >25% rise in creatinine had hemoglobinuria compared to 8/21(38%) patients with ≤25% rise in creatinine (p=0.003). Also among just the HES patients, the median amount of incompatible RBCs infused was not significantly different between patients with (0.74 ml/kg) and without (0.62 ml/kg) hemoglobinuria (p=0.42), or between patients with (0.56 ml/kg) and without (0.62 ml/kg) a >25% rise in creatinine (p=0.86).

Table 1: Comparison of patients who had RBC depletion with HES vs. Ficoll

*considering only patients transplanted after 2006.

Conclusion: Our study is the first to analyze pediatric major ABO mismatch BMT infusion hemolysis after two different RBC depletion techniques: HES and Ficoll. Our results suggest that RBC depletion with Ficoll achieves less residual RBCs in the BM graft and likely less resulting hemoglobinuria. However, after both techniques hemolysis-induced renal impairment is rare. The volume of residual incompatible RBCs in the infused BM graft does not appear to strongly determine if clinical hemolysis occurs. The volume that is safe likely depends on other variables that influence the risk of clinical hemolysis.