Long-term survival after chemotherapy for acute myeloid leukemia: the experience of the Southwest Oncology Group.

Background: Reports on outcomes of chemotherapy trials in acute myeloid leukemia (AML) have rarely included results of long-term follow-up beyond 10 years. The authors therefore chose to review long-term follow-up data from 3 studies conducted by the Southwest Oncology Group (SWOG) between 1978 and 1990.
Methods: The analysis included data on 2083 patients enrolled in SWOG studies S7823, S8124, and S8600. The results were based on data available as of November 15, 1996.
Results: The probability of survival 8 years after entry was 9% in Study S7823, 14% in S8124, and 15% in S8600. For patients age < 50 years, the probabilities were 14%, 24%, and 20%, respectively. For patients ages 50-64 years, the probabilities were 7%, 8%, and 8%, respectively. For those age < 50 years who achieved complete remission, the 8-year probability of disease free survival was 17% in Study S7823, 28% in S8124, 17% with standard dose cytarabine in S8600, and 26% with high dose cytarabine in S8600. Relapse was the major reason for failure after complete remission in all three studies. When the results of the 3 studies were combined, most of the 743 relapses had occurred by Year 3 and nearly all the rest by Year 5. Among the prognostic factors universally available for study, three were highly associated with survival in all three studies: age, French-American-British disease classification, and white blood cell count at diagnosis.
Conclusions: In view of the fact that most deaths occurred during the first 3 years, it is appropriate to report the results of clinical trials after patients have been followed for 4 years. Despite modest gains, the results of chemotherapy for AML remain disappointing, especially in the treatment of older patients.