Objective: To assess the prognostic value of early treatment response in children with acute myeloid leukemia (AML).
Methods: Sixty-one children with AML, 37 male and 24 female, aged 96 months (6 - 154 months), underwent treatment according to the protocol AML-XH-99 with a total treatment course of 15 months and were followed up for 12 months (1 - 74 months). Bone marrow smear was made 48 hours after the end of the first course of induction treatment. Then the children were divided into 2 groups according to the number of bone marrow blast cells: group with the number of blast cells > or = 0.15 and group with the number of blast cells < 0.15. Second bone marrow smear was made when complete remission was achieved after the end of the treatment course. Then the children were divided into 2 groups according to the number of bone marrow blast cells: group with the number of blast cells of 0.00 and group with the number of blast cells between 0.00 and 0.05. The probability of event-free survival (EFS) was estimated by Kaplan-Meier analysis. Log-rank test was used to compare the 5-year EFS (pEFS) of different groups. The differences in the biological features were compared by Chi-square analysis or Fisher exact test.
Results: The pEFS of the group with the number of blast cells > or = 0.15 was 18% +/- 15%, significantly shorter than that of the group with the number of blast cells < 0.15 (49% +/- 11%, P = 0.079). The 3 patients without morphologically identifiable blast all survived 5 years after complete remission had been achieved, and the pEFS of the 39 patients with the number of blast cells between 0.00 and 0.05 was 53% +/- 10%. The pEFS of the patients among which complete remission was achieved after the first course of treatment (n = 39) was 54% +/- 10%, significantly higher than that of the patients without complete remission after the first course of treatment (10% +/- 9%, P = 0.0002). Multiple factor analysis showed that achievement of complete remission after the first course of treatment and existence of central nervous system leukemia were both independent prognostic factors with the hazard ratios of 4.007 and 7.050 respectively and the 95% confidential intervals of 1.019 to 6.163 and 0.018 to 0.547 respectively (P = 0.045 and P = 0.008). The number of blasts 48 hours after the end of the first course of induction treatment was highly correlated with the rate of complete remission after the first treatment course (P = 0.000 028 8).
Conclusion: With important prognostic significance, early treatment response, such as the number of blasts 48 hours after the end of the first course of induction treatment can predict whether complete remission can be achieved.