Key points are not available for this paper at this time.
We have analyzed factors associated with acute graft-versus-host disease following allogeneic bone marrow transplantation in 469 patients with histocompatible sibling donors between 1979 and 1987. Overall, 46 +/- 5% (95% confidence interval) developed clinical grade II-IV acute GVHD following transplantation. In univariate analysis, patient or donor age greater than or equal to 18 years was significantly associated with increased GVHD risks (greater than or equal to 18, 63 +/- 6% grade II-IV GVHD vs. less than 18, 27 +/- 6%, P less than .0001), without incremental risk in older adults. Univariate analysis showed that donor:recipient sex match and female:female transplants were associated with less-frequent GVHD. More frequent GVHD was associated with chronic myelogenous leukemia, cytomegalovirus seropositivity, and prior donor alloimmunity (pregnancy or transfusion). Additionally, the allele HLA-A26 was associated with increased risk of GVHD (72%, P = .005) while HLA-DR3 was associated with less GVHD (31%, P = .03). Stepwise multivariate analysis confirmed the increased GVHD risks associated with older recipient age, HLA-A26 and donor:recipient gender (not female:female) and the protective effect of HLA-DR3. Similar results were found using the different analytic technique of recursive partition analysis, which identified within the adult population the lowest GVHD risk in female recipients with nonalloimmunized female donors (20%), while other gender combinations had 68% acute GVHD, regardless of donor alloimmunity. In children (less than 18 years), lower GVHD risk accompanied donor:recipient sex-matched (18%) versus mismatched (33%) BMT. Clinical trials undertaken to lessen the hazards of GVHD must be designed with appropriate attention to these reproducibly identified clinical variables associated with different GVHD risks.
Weisdorf et al. (Sat,) studied this question.