Randomized Comparison of Primary Allogeneic Stem Cell Transplantation and Best Available Drug Treatment in Chronic Myeloid Leukemia.

Early allogeneic hematopoietic stem cell transplantation (HSCT) has been proposed as primary treatment modality for patients (pts) with chronic myeloid leukemia (CML). This concept has been challenged by persisting transplantation mortality and improved drug therapy. In order to verify retrospective and observational results and to counsel pts and doctors about survival prospects with each treatment strategy, a randomized controlled trial was designed to compare primary HSCT and best available drug treatment in a cohort of 621 newly diagnosed CML pts in chronic phase. Assignment to treatment strategy was by eligibility for HSCT and genetic randomization according to availability of a matched related donor. Evaluation followed the intention to treat principle. 354 pts (62% male; median age 40 years, range 11–59) were eligible and randomized. 135 pts (38 %) had a matched related donor of which 123 (91%) received a transplant within a median of 10 months (range 2–106) from diagnosis. 4 pts died before scheduled transplantation, 8 pts withdrew consent. 219 pts (62%) had no related donor and received best available drug treatment. Of these, 97 pts (44%) received a matched unrelated donor (MUD) transplant in 1st chronic phase and were censored at the time of transplantation. As 1st line treatment after randomization pts received interferon alpha based therapy. In the course of the study a total of 197 pts were switched to imatinib after failure of interferon alpha. Currently 31 (57%) of 54 living pts of the drug treatment group receive imatinib or 2nd generation tyrosine kinase inhibitors (dasatinib n=2, nilotinib n=1). With a median observation time of 8.9 years (range 4.2–11.2) median survival of all 621 pts was 8.1 years. During the first 8 years after diagnosis survival curves of drug treated patients were superior to those of transplanted patients reflecting transplant-related mortality. Beyond 8 years survival curves were no longer distinct. 5 (10) year survival was 62% (53%) for transplanted and 73% (52%) for drug treated pts, in the low risk group 68% (59%) for transplanted and 85% (62%) for drug treated pts, respectively. Survival was superior for drug treated pts up to the cutpoint of survival curves at year 8 (p=0.041) and during the study period up to 11 years from diagnosis (p=0.049), particularly so in low risk pts (p=0.027 to cutpoint, p=0.032 overall). Significantly higher proportions of complete cytogenetic remissions (91% vs 61%, p=0.002) and of major molecular responses (ratio BCR-ABL/ABL <0.1%; 81% vs 45%, p=0.001) were found in the transplant group indicating higher levels of residual disease in the group receiving drug treatment. In summary, the general recommendation of HSCT as 1st line treatment option in chronic phase CML can no longer be maintained. It should be replaced by a trial with modern drug treatment first and a risk adapted strategy according to the individual disease and transplantation risks thereafter. HSCT remains an important treatment option in a risk adapted strategy on the basis of higher cytogenetic and molecular long term remission rates.