This is a retrospective analysis of 1084 patients who received an allogeneic HSCT in 10 Brazilian Centers, from February 1983 to March 2003, aiming to validate the EBMT risk score. Data from transplanted patients and donors regarding patients age, disease stage at transplantation, HLA full-match sibling donor or full-match unrelated donors, donor-recipient gender match and the interval from diagnosis to transplantation, were used as variables to calculate the EBMT risk score. This score was analyzed using Cox Proportional Hazards Model. The OS, DFS, TRM and relapse were analyzed using Kaplan-Meier and cumulative incidence, whenever appropriate. In all there were 647 (60%) males and 437 (40%) females, the median age was 32 years old (range 1 – 59); 898 (83%) were in chronic phase, 146 (13%) were in accelerated phase and 40 (4%) were in blast crisis; 151 (14%) were younger than 20 years old, 620 (57%) were between 20 and 40 and 313 (29%) were older than 40; 1025 (95%) received HLA full-match sibling transplant and only 59 (5%) received an unrelated transplant. Female donor to male recipient occurred in 283 (26%) transplants. The interval from diagnosis to transplantation was less than 12 months in 223 (21%) cases and greater in 861 (79%). The OS, DFS, TRM and, relapse were 49%, 50%, 45%, 25%, respectively. The risk score 0–1 occurred in 179 (17%), score 2 in 397 (37%), score 3 in 345 (32%), score 4 in 135 (12%), and score 5–6 in 28 (2%). The risk scores 0–1 and 2 did not show any difference in terms of OS (58% and 55%, respectively) but they were significantly better than scores 3 or more (score 3 – 44%, 4 – 36 % and, 5-6 - 27%, respectively) (P<0.001). DFS and TRM beyond score 3 were 46%, 49%, respectively and the relapse rate beyond score 5–6 was 77%. Disease status had a negative impact on all outcomes (OS, DFS, TRM, and relapse). OS for female donor - male recipient was 40% compared to 52% for the other patients (P=0.004). DFS and TRM were significant for disease phase and female donor-male recipient (P<0.001 and P<0.003, respectively). In our experience, age and interval from diagnosis and transplant did not show any difference in terms of OS, DFS, TRM, and relapse rate. Our results confirm the usefulness of the EBMT risk score for point-decision in the Imatinib era.

Author notes

Corresponding author

Sign in via your Institution