Abstract
Internal tandem duplication of the FLT3 gene (FLT3/ITD) has been associated with high risk of relapse in acute myeloid leukemia (AML) yet nearly 25–30% of the patients with FLT3/ITD have long-term disease free survival with conventional chemotherapy. We hypothesized that FLT3/ITD AML patients with poor clinical outcome may have disease that involves less mature hematopoietic precursors than patients with favorable outcome. To test this hypothesis, we isolated less mature, CD34+CD33− and more mature, CD34+CD33+ precursor cells from 24 pediatric AML patients enrolled on Children’s Cancer Group clinical trials CCG-2891 and 2961 previously identified as having a FLT3/ITD. Granulocyte/monocyte colonies (CFU-GM) were grown in methylcellulose, harvested, and analyzed for the presence of FLT3/ITD after 14 days of growth. Twenty patients yielded sufficient cells and growth of CFU-GM colonies for analysis. FLT3/ITD was detected in CFU-GM colonies derived from CD34+CD33+ cells in all patient samples (median 80% of colonies tested per patient, range 6–100%). In contrast, FLT3/ITD was detected in CFU-GM colonies derived from CD34+/CD33− cells in only 11 of the 20 patient samples (median 46% of colonies tested per patient, range 6–100%). Of the 9 patient samples without FLT3/ITD involvement of CD34+CD33− colonies, 8 achieved a CR, 6 of whom are long-term survivors, and one patient died of non-leukemic causes. In contrast, of the 11 patients with CD34+CD33− cell involvement, 9 either failed to achieve CR or relapsed after achieving CR, and 2 died of non-leukemic causes. Actuarial progression-free survival at 4 years from diagnosis for the patients with and without FLT3/ITD in the CD34+CD33− population was 0% vs. 68% respectively (p=0.017). As allelic ratio of the FLT3/ITD has been used to define high-risk patients within the FLT3/ITD cohort, we determined the FLT3/ITD allelic ratio in our study population and correlated it with the presence of FLT3/ITD in the CD34+CD33− population. Ten of the 11 (91%) of the patient samples with FLT3/ITD involvement of the progenitor cells had high allelic ratio compared to 5 of 9 (56%) of the patients without early cell involvement. Together these data suggest that clonal dominance of FLT3/ITD containing leukemia cells at the CD34+CD33− stage of hematopoietic development is correlated with a high risk of relapse. Further studies are required to determine whether clonal dominance at this hematopoietic stage is a variable that, independent of high allelic ratio, accounts for the poor clinical outcome seen in a subset of FLT3/ITD positive AML patients.
Author notes
Corresponding author