Background: Measuring residual disease (RD) by multidimensional flow cytometry (MDF) predicts patient outcome following the first induction cycle and is a prognostic indicator in AML, particularly those patients lacking prognostic genetic lesions (i.e. standard risk; SR).Objective: To assess prevalence of RD after induction 1 (EOI1) and induction 2 (EOI2) and correlate presence of RD with clinical outcome.Patients and Methods: COG AML trials AAML0531 and AAML1031 collectively enrolled 2119 patients of whom 1929 met eligibility requirements (AAML0531, N=844, AAML1031, N=1085) and 1642 submitted a specimen for RD at the end of initial induction therapy (EOI1) and continued to subsequent therapy, while 1299 patients submitted a specimen at the end of second induction (EOI2) and continued on therapy. Detection of RD by MDF was determined with a standard panel of reagents applying a “difference from normal” algorithm (1). The clinical cutoff of the assay was 0.1% of total non-erythroid cells. The lower level of detection of the assay was 0.02% of non-erythroid cells. RD was not reported for specimens that were inadequate based on cellularity, level of hemodilution, and constitution of the myeloid compartment. COG AAML0531 and AAML1031 both utilized an MRC backbone but differed in total number of chemotherapy courses, SCT allocation and preparative regimen. Since AAML1031 includes a sorafenib treatment arm that is ongoing, all patients with FLT3/ITD-High allelic ratio were excluded from this analysis.Results: In the combined cohort of 1642 patients with validated RD data, RD was detected in 420 patients at EOI1 (25.6%) and in 110/1299 patients (8.5%) at EOI2. Overall survival from EOI1 (Fig. 1A) and EOI2 (Fig. 1B) is presented below. Actuarial 3 year OS from EOI1 was 47.8 ± 5.2% and 75.1 ± 2.6% for those with and without RD (p<0.001). Corresponding OS from EOI2 as 53.5 ± 10.6% and 74.3 ± 2.7% for those with and without RD (p<0.001). We further inquired whether clearance of RD from EOI1 to EOI2 provides clinical benefit and improved outcome. In the combined cohort, of the 248 RDpos patients at EOI1, 155 patients (62.5%) did not have detectable disease (RDpos/neg) whereas 93 patients had persistent RD (RDpos/pos; 37.5%). Correlation of clinical outcome based on clearance of RD from EOI1 to EOI2 showed that those with RD at EOI1 had similar outcomes regardless of resolution of RD at EOI2 with actuarial overall survival at 3 years from EOI2 of 54.7 ± 11.8% for the RDpos/pos vs. that of 56.4 ± 8.4% in RDpos/neg patients (p=0.755; Fig. 1C). Given that AAML0531 and AAML1031 had different second induction course of therapy, we also compared the rate of clearance of RD between the two study populations and the resultant clinical outcome. In similarly matched standard risk patients with EOI1 and EOI2 RD data, RD conversion rate (RDpos to RDneg) was 13.3% in AAML0531 vs. 17.2% in AAML1031(p=0.262). In AAML0531, those with RD clearance (RDpos/RDneg) had 3 year OS from EOI2 of 46.9 ± 17.6%, nearly identical OS of 42.4 ± 17.2% compared to those with persistent RD (RDpos/RDpos), p=0.485. Similarly, in AAML1031, those with and without RD clearance had similar OS (41.5 ± 13.4% vs. 45.6 ± 17.2%, p=0.144).Conclusions: Persistence of RD detected by MDF after one course of therapy is highly associated with adverse outcome in AML. We further demonstrate that clearance of RD from EOI1 to EOI2 is not associated with improve outcome on two sequential COG trials with similar but not identical therapy. Therefore, reducing leukemia burden after EOI1 may not translate into better survival.Reference:1) Loken MR et. al. Residual disease detected by multidimensional flow cytometry signifies high relapse risk in patients with de novo acute myeloid leukemia: a report from Children's Oncology Group. Blood. 2012 Aug 23;120(8):1581-8. [Display omitted] DisclosuresBrodersen:Hematologics Inc: Employment. Loken:Hematologics Inc: Employment, Equity Ownership.
Read full abstract