Minimal Residual Disease Monitoring in Childhood Precursor B Lymphoblastic Leukemia with t(12;21)(p13;q22); ETV6-RUNX1: A Comparison Between Quantitative RT-PCR and Flow Cytometry

The translocation t(12;21)(p13;q22) resulting in the fusion gene ETV6-RUNX1, is the most frequent gene fusion in childhood precursor B lymphoblastic leukemia (pre-B ALL), affecting about one in four children with pre-B ALL. In the NOPHO ALL-2008 treatment protocol, treatment assignment in pre-B ALL is based on clinical parameters, genetic aberrations, and results from analysis of minimal residual disease (MRD) at day 29 and 79 during treatment (where MRD >0.1% leads to upgrading of treatment). For pre-B ALL, in this protocol MRD analysis is performed using flow cytometry as the method of choice. In this study, we also analyzed MRD in t(12;21)(p13;q22) cases with quantitative reverse transcription-polymerase chain reaction (qRT-PCR) for the fusion transcript ETV6-RUNX1 in parallel with routine MRD analysis with flow cytometry, to determine if qRT-PCR of the ETV6-RUNX1 fusion transcript would be a reliable alternative to FACS.

Bone marrow samples were collected at diagnosis and at day 15, 29 and 79 during treatment from 31 children treated according to the NOPHO ALL-2000 (n = 3) and NOPHO ALL-2008 (n = 28) protocols in Gothenburg, Sweden, between 2006 and 2013. Samples were analyzed in parallel with qRT-PCR for ETV6-RUNX1 fusion transcript and with FACS. For qRT-PCR, mRNA was isolated, cDNA synthesized, and qRT-PCR performed with GUSB as reference gene. MRD-qRT-PCR was defined as the ETV6-RUNX1/GUSB ratio at the follow-up time point (day 15/29/79) divided with the ETV6-RUNX1/GUSB ratio at diagnosis (%). MRD analysis with FACS was performed, after lysis of erythrocytes, using antibodies against CD10, CD19, CD20, CD22, CD34, CD38, CD45, CD58, CD66c, CD123, and terminal deoxynucleotidyl transferase, and when applicable also CD13 and CD33. Results of MRD-FACS were expressed as % of all cells.

In total, 83 samples were analyzed with both methods in parallel; 31 from day 15 in treatment, 28 from day 29, and 24 from day 79. Overall, MRD-qRT-PCR showed good correlation with MRD-FACS. In total, 31 samples were positive with qRT-PCR and 24 with FACS, with concordant results (positive with both methods or negative with both methods) in 89% of samples, when the limit of decision (positive/negative MRD) was set to 0.1%. The concordance was especially high at the treatment stratifying time points, i.e. day 29 and 79; 89% and 100%, respectively. No samples at these time points were positive with FACS but negative with qRT-PCR. During the follow-up period (6-81 months), one patient relapsed (with negative MRD with both methods at stratifying time points), and two succumbed from therapy-related causes.

Our results show that there is a significant relationship between the results of MRD analysis using FACS and MRD analysis using qRT-PCR of ETV6-RUNX1 fusion transcript. The high concordance between the methods indicates that negative MRD using qRT-PCR is as reliable as negative MRD using FACS, and that qRT-PCR could therefore be an alternative to FACS in cases where FACS is not achievable. In comparison to quantitative PCR of TCR/Ig gene rearrangements, which is the current backup MRD method for cases with pre-B ALL in NOPHO ALL-2008, qRT-PCR of ETV6-RUNX1 is much less time and labor consuming, making it appealing in a clinical laboratory setting.