Long-Term Survival and Late Relapse in 5-Year Survivors of Allogeneic Hematopoietic-Cell Transplantation (HCT) for Chronic Myeloid Leukemia (CML) in First Chronic Phase.

Abstract 3321

Poster Board III-209

Prior to the advent of tyrosine kinase inhibitors, allogeneic HCT was standard therapy for CML. However, very long-term outcomes of allogeneic HCT for CML are not well described. To evaluate the probability and risk factors for late mortality and relapse in this patient population, we conducted a retrospective cohort study that included 2444 patients who received a myeloablative allogeneic HCT for CML in first chronic phase between 1978 and 1998 and had survived in continuous complete remission for at least 5 years. Relapse was considered the earliest reported date of the following: hematologic recurrence, cytogenetic recurrence or initiation of therapy for recurrence. The median followup of our cohort was 11 years (range, 5-25) from HCT; 377 patients had followup >15 years. Donor sources were HLA-matched siblings (MSD) in 1692, unrelated donors (URD) in 639 and other related donors in 113 patients. The median age at HCT was 35 years and patients primarily received bone marrow grafts (96%) and conditioning using either total-body irradiation (TBI, 61%) or BuCy (38%) regimens. Acute and chronic graft-versus-host disease (GVHD) occurred in 43% and 62% of patients, respectively. The probabilities of overall survival at 15 years were 88% (95% CI, 86-90%) for MSD and 87% (83-90%) for URD recipients. Corresponding cumulative incidences of relapse at 15 years were 8% (7-10%) and 2% (1-4%), respectively. The latest reported relapse occurred 18 years post-HCT. In multivariable analyses addressing the importance of patient, disease and transplant related factors for long-term survival, older age at HCT, use of female donor for a male recipient, use of TBI based conditioning, acute GVHD and chronic GVHD all independently increased the risk of late mortality. Chronic GVHD reduced the risk of relapse, but increased the risk of non-relapse mortality. Recipients of MSD and URD had similar risks of long-term mortality, relapse and non-relapse mortality. Compared to an age, gender and race adjusted general population, 5 year survivors of HCT for CML were 2.9 times (95% CI, 1.9-3.9) more likely to die at 6 years and 2.5 times (1.3-3.7) more likely to die at 10 years after HCT. However, by 15 years after HCT, their relatively mortality (2.3 [0-4.9]) was not significantly different than the general population. In summary, recipients of allogeneic HCT for CML in first chronic phase who remain in remission for at least 5 years after HCT have very favorable subsequent long-term survival with mortality rates eventually approaching those of the general population. There is a small but continuing risk of relapse even in these long-term survivors. Chronic GVHD protects against relapse but increases the risks of non-relapse mortality.