Minimal Residual Disease by Multiparametric Flow Cytometry Predicts Relapse Free Survival better than Over-Expression of WT1 and BAALC in Acute Myeloid Leukemia

Background: Improved molecular and immunophenotypic characterization of AML offers the opportunity to define markers in individual patients which can be used to monitor minimal residual disease (MRD). Immunophenotypic analyses are based on the multiparametric flow cytometry (MPFC) detection of leukemia-associated immunophenotypes. The most useful molecular prognostic markers implicated in AML characterization are NPM1 mutations, FLT3 ITD, CEBPA mutations, MLL -PTD, RUNX1 and ASXL1 mutations. Deregulated expression of genes have also been identified as prognostic markers eg BAALC, WT1 and ERG. In this study, gene mutations (NPM1, FLT3, CEBPA, MLL-PTD), gene expression (WT1 and BAALC)and immunophenotyping were prospectively assessed to detect MRD.

Methods: Diagnostic Marrow/peripheral blood samples from 98 AML patients (excluding AML-M3) were included between March 2012 and July 2014. Patients received standard induction chemotherapy with daunorubicin 60 mg/m2 for 3 days and cytosine arabinoside 100 mg/m2 iv for 7 days. Marrow was done 21-28 days after start of induction chemotherapy for assessment of morphology, MPFC and molecular markers. If marrow was in remission (CR), then patients received 1st consolidation therapy with 18 gm/m2 of cytosine arabinoside (HiDAC). Marrow was repeated 21-28 days after 1st consolidation chemotherapy for assessment of same parameters as above. Patients then received 2 more HiDAC or underwent allogeneic transplant according to cytogenetic risk. Cytogenetic analyses were performed using standard techniques of chromosome banding and FISH. All cases were divided into three cytogenetic risk groups (i.e. good, intermediate and poor) based on NCCN guidelines. Immunophenotyping was done using 8 color MPFC. The same panel of antibodies was used to characterize the leukemic cells at diagnosis, post induction and post consolidation. MRD was calculated as a percentage of abnormal leukemic cells per total nucleated cells. cDNA synthesis and quantitative real time PCR (RQ-PCR) assays for WT1 and BAALCwere carried out according to Europe Against Cancer guidelines. Mutation profiling was carried out by capillary electrophoresis and direct DNA sequencing methods as described previously. Chi-square method was used to detect any association between risk groups and gene mutations. Probabilities of relapse-free survival (RFS) were estimated using Kaplan-Meier method. Univariate comparisons of RFS for potential prognostic markers (molecular and MPFC) were made using log-rank test.

Results: Of 98 patients enrolled in this prospective study, 86 patients completed induction and 59 patients 3 courses of HiDAC. The median age was 27 years (range 15-58). Based on cytogenetic risk groups 24.4%, 47% and 28.6% were good, intermediate and poor risk cases respectively. Twelve patients died during induction due to sepsis and 16 patients were refractory to induction chemotherapy. Baseline frequencies of FLT3-ITD, MLL-PTD, NPM1-type A and CEBPA gene mutations were 20.4%, 10.2%, 20.4% and 25.5% respectively. The biallelic (TAD1 and ZIP mutations) and monoalleleic CEBPA mutations were found in 4.1% and 21.4% patients respectively. No significant association was found between different risk groups and gene mutations except NPM1-type A mutations were associated with good risk group (P=0.015). MRD by RQ-PCR for BAALC/WT1 and MPFC was assessed after post induction and post first consolidation therapy. There was one log reduction in mean copy number of WT1 (5312 vs 430) and BAALC (19648 vs 3298) between diagnosis and post consolidation samples. No significant differences were found between high expressors of WT1/BAALC and RFS. MPFC analyses revealed that 48% of post induction samples were MRD positive (range: 0.05-38%, mean 6.6%) Similarly, 53% of post consolidation samples were MRD positive (range: 0.01-85%, mean 10.9%). Univariate analysis showed that MRD positivity by MPFC at post induction (RFS at 1.5 years- 85% vs 30%, P=0.012) and post consolidation (RFS at 1.5 years- 90% vs 42%, P=0.015) was associated with poor RFS. Patients without MLL-PTD mutation showed better RFS (P=0.003) as compared to mutated cases and TAD2positivity at post induction showed a trend to better RFS (P=0.097).

Conclusion: Our results suggest that MRD by MPFC predicts RFS more accurately than over-expression of WI1 and BAALC.