Multiple regression analyses used a stepwise selection procedure with 10% entry and exit criteria. Univariate analyses of other potential clinical factors that may have influenced CD4⁺CD25^high T-cell numbers included recipient age; time from transplantation; donor source, HLA-identical related donor/VUD; HSC source, bone marrow/PBSCs; conditioning protocol, standard myeloablative/reduced intensity, in vivo T-cell depletion; and prior acute GVHD. Treatment with immunosuppressive medication was not included in the analysis because few patients without cGVHD were on treatment at the time of the study.

— indicates that no statistical correlation was found.