Results of risk-adapted therapy in acute myeloid leukaemia. A long-term population-based follow-up study.

In 1997-2003, a protocol for treatment of acute myeloid leukaemia (AML) (except promyelocytic leukaemia) was activated in four Swedish health care regions covering 50% of the national population. Based on cytogenetics and clinical findings, patients aged 18-60 yr were assigned to one of three risk groups. In this report we account for the long-term clinical outcome of enrolled patients. Patients received idarubicin and cytarabine in standard doses as induction therapy and consolidation courses included high-dose cytarabine. Allogeneic stem cell transplantation (allo-SCT) from an human leucocyte antigen-identical sibling was recommended in standard and poor-risk patients, whereas unrelated donor transplant was reserved for poor-risk patients. Autologous (auto-SCT) was optional for standard or poor risk patients not eligible for allo-SCT. Two hundred seventy-nine patients with de novo or secondary (9%) AML, median age 51 (18-60) yr, corresponding to 77% of all patients in the population, were included. Twenty (7%) patients were assigned to the good risk group, whereas 150 (54%) and 109 patients (39%) were assigned to standard- and poor-risk groups, respectively. Induction failures accounted for 55 patients; 16 early deaths eight of whom had white blood cell (WBC) >100 at diagnosis, and 39 refractory disease. Thus, complete remission (CR) rate was 80%. At study closure, the median follow-up time of living patients was 90 months. Median survival time from diagnosis in the whole group was 27 months and 4-yr overall survival (OS) rate was 44%. In good, standard, and poor risk groups, 4-yr OS rates were 60, 57 and 24%, respectively. Median relapse-free survival (RFS) time in CR1 was 25 months and RFS at 4 yr was 44%. Four-year RFS rates were significantly (P < 0.001) different between the three risk groups; 64% in good risk, 51% in standard risk and 27% in poor risk patients. One hundred-ten transplantations were performed in CR1; 74 allo-SCT (50 sibling, 24 unrelated donor), and 36 auto-SCT. Non-relapse mortality was 16% for allo-SCT patients. Outcome after relapse was poor with median time to death 163 d and 4-yr survival rate 17%. Three conclusions were: (i) these data reflect treatment results in a minimally selected population-based cohort of adult AML patients <60 yr old; (ii) a risk-adapted therapy aiming at early allogeneic SCT in patients with a high risk of relapse is hampered by induction deaths, refractory disease, and early relapses; and (iii) high WBC count at diagnosis is confirmed as a strong risk factor for early death but not for relapse.