Background and objective: To analyze the results of unrelated bone marrow transplantation (UDBMT) as treatment for chronic myeloid leukemia (CML) in Spain.
Design and methods: Eighty-seven consecutive UDBMT performed in 9 centers between October 1989 and February 1998 were evaluated. This represents more than 95% of UDBMT for CML performed in adult transplant centers in Spain during this period. The patients' median age was 31.5 years (range, 12-49). The median interval from CML diagnosis to UDBMT was 30 months (range, 3-160). Seventy-nine percent of transplants were performed during the first chronic phase (1CP).
Results: Actuarial probability of survival and disease-free survival at 4 years for the whole series was 24% (95% confidence interval [CI]: 14%-34%) and 20% (CI: 10%-30%), respectively. The cumulative incidence of relapse and transplant-related mortality (TRM) was 7% (CI: 4%-10%) and 71% (CI: 60%-82%), respectively. The main causes of death were graft failure (n=7), infection (n=23), and graft-versus-host disease (GvHD) (n=25). The actuarial probability of acute GvHD grade II-IV and grade III-IV was 56% (CI:46%-66%) and 36% (CI: 26%-36%), respectively. The cumulative incidence of extensive chronic GvHD was 18% (CI: 9%-27%). Univariate analyses showed that the pre-transplant factor with the highest influence on survival was disease status at transplant (30% in 1CP vs. 0% in advanced phases; p=0.0001). Other pre-transplant factors influencing survival among patients in 1CP were: patient's age (older than 30 years 11% vs. 48%), interval diagnosis-transplantation (longer than 2 years 17% vs. 55%), donor type (HLA, B, DRB1 identical 32% vs. 25%), CMV serologic status (donor and recipient negative 63% vs. 24%), year of transplantation (before 1995 19% vs. 40%), and conditioning regimen (cyclophosphamide plus total body radiation 40% vs. 16%). The main risk factors had a cumulative effect on survival. Thus, probability of survival ranged from 66% (CI: 39%-93%) in patients in 1CP, under 40 years of age, transplanted from an HLA, A, B, DRB1 identical donor during the first two years after diagnosis, to 0% in those with three or more risk factors.
Interpretation and conclusions: This experience shows that UDBMT used to have a high TRM that has progressively decreased along the years. At the present time, the results are encouraging, particularly when UDBMT is performed under favorable conditions.