Initial response to induction chemotherapy is a significant predictor of outcome in leukemias, where those with rapid response have an improved outcome compared to the non-responders. Morphologic evaluation of post induction marrow has been the gold standard in assessing remission status in leukemias, where those with <5% blasts are considered to be in complete remission (CR), whereas those with '5% blast are considered induction failures. MRC AML 10 study, using morphologic assessment of remission, delineated a clinically significant threshold of 15% marrow blasts after induction I, where those with 5–15% blasts were shown to have a similar outcome to those in morphologic CR, while those with >15% blasts were considered to be high risk. The Children's Oncology Group (COG) phase III AML protocol AAML0531 which treated 1022 eligible patients without Down Syndrome (DS) on an MRC based chemotherapy backbone, utilized the >15% post induction morphologic blast threshold, as assessed at the institutional level, to allocate patients to specific risk groups based on the observed post induction 1 blast prevalence. In this study, all patients regardless of initial response would receive a second course of similar induction chemotherapy. Those with 5–15% blasts after course 1 with no other risk features who achieved a CR after the second course were considered intermediate risk, but those with >15% post course 1 blasts who achieved a CR after course 2 were considered high risk and allocated to stem cell transplantation from the most suitable donor in first CR.
As part of this trial, multidimensional flow cytometry (MDF) was used to assess marrow response after each course of induction chemotherapy. As morphologic evaluation would not be able to distinguish normal and malignant blasts in the marrow, we inquired whether morphologic CR status correlates with MDF findings. Of the 1022 eligible non-DS patients treated on AAML0531, 784 patients had consented to the correlative biology studies and had available MDF data for correlation with morphology. Of the 784 patients, 185 patients (24%) had failed to achieve a morphologic CR (>5% blast by morphology) after the initial course of chemotherapy, of which 94 were partial remissions (PR, 5–15% blast) and 91 were persistent disease (PD, >15%). Of these 185 patients who failed to achieve a morphologic CR, 67 patients (36%) had no evidence of disease by MDF, and the remaining 64% had MDF detectable disease. Clinical outcome evaluation of these patients who failed induction (>5% blast) at the end of course 1 based on MDF status demonstrated that disease free survival for those with MDF detectable disease was 24% compared to that of 52% in those with induction failure with no MDF detectable disease (p<0.0001). Corresponding Overall survival was 48% and 76%, respectively (p=0.005). As 15% blast threshold at the end of induction 1 was used as a threshold for risk allocation, we specifically evaluated the prevalence and clinical implications of MDF-detectable disease in this cohort of high risk patients. Of the 91 patients with >15% blast after course I, 25 (27%) had no evidence of disease by MDF. Disease-free survival at 3 years from end of induction in this high risk cohort with and without MDF-detectable disease was 20% and 55%, respectively (p<0.0001) with a corresponding OS of 47% and 83% (p=0.014). This study highlights the fact that morphologic evaluation of marrow specimens may not be adequate for post induction disease assessment. Multi-dimensional flow cytometry provides significant information for accurate assessment of response.
Smith:Seattle Genetics:; Eisai:; Archimedes Pharma: Membership on an entity's Board of Directors or advisory committees; Pfizer, Inc.: Membership on an entity's Board of Directors or advisory committees.
Asterisk with author names denotes non-ASH members.