group O, and Bombay, a rare FUT1 null phenotype lacking H and AB antigens.

3Much to our surprise, we observed delayed PLT engraftment among group O patients, but only at low CD34+ cell doses (less than 3 million/kg). We speculated that a group O phenotype may be associated with delayed terminal differentiation relative to group A and B patients due to prolonged fucose-mediated adhesion to stromal fibroblasts. Our study had several weaknesses including the relatively small number of group O patients (n = 18) and non‐group O patients (n = 36) transplanted at low CD34 per kg cell doses, the lack of an A1/A2 red blood cell (RBC) subtype for group A patients, and the absence of any flow cytometric phenotyping on circulating PLTs before transplant. The last two deficiencies may be of critical importance based on subsequent studies by our laboratory and others. The group A2 phenotype, which constitutes 20 percent of all group A donors, is associated with an absence of A antigen on PLTs. 4 Furthermore, a substantial number of A1 (20%) and most group B individuals (50%-80%) have extremely low ABO expression on PLTs when examined by flow cytometry and solid phase. 4 Because PLTs arise from megakaryocytes, these findings are applicable to ABO expression on megakaryocytes as well. Finally, ABO expression on megakaryocytes exhibits clonal variation even in individual donors, possibly reflecting asynchronous maturation during normal megakaryocyte differentiation. 5,6