Qualitative and Quantitative Concordance of Minimal Residual Disease Data Assessed by Multicolor Flow Cytometry and PCR of Fusion Gene Transcripts In Childhood B-Cell Precursor Acute Lymphoblastic Leukemia

To evaluate qualitative and quantitative concordance between MRD data assessed by FC and PCR of FGt in children with B-cell precursor ALL (BCP-ALL) during treatment.

Concurrent detection of MRD by multicolor FC, RQ-PCR and RT-PCR was performed in 184 follow-up bone marrow samples from 46 children with BCP-ALL. Among them 21 patients (pts) carried ETV6-RUNX1 FGt, 3 pts – TCF3-PBX, while in other 22 pts various types of MLL-rearrangements were detected. 18 pts had CD10(-) and 28 pts – CD10(+) BCP-ALL. 62 samples were obtained during remission induction while 122 – during consolidation/intensification. 6–8-color FC was used. FGt CN was measured by RQ-PCR according to recommendations with normalization to ABL. MRD value was calculated as previously described.

Sensitivity of FC MRD detection varied from 10^-4 to 10^-5. RQ-PCR sensitivity ranged from 5×10^-5 to 1×10^-5. 65 of 184 samples were MRD-negative by both methods, 13 (7.07%) - were negative by FC but positive by RT-PCR, while in one sample (0.54%) tumor cells were detected by FC but not by PCR. Remaining 105 samples were MRD-positive by both methods. High qualitative concordance (92.39%) between FC and RT-PCR data was found. Similar concordance was observed in MLL-rearranged (91.60%), ETV6-RUNX1-positive (95.45%) and TCF3-PBX-positive (88.89%) cases was observed (p=0.650). Samples with and without normal lymphoid regeneration were analyzed separately, because presence of normal B-cell precursors (BCP) in follow-up samples is known to be an obstacle for FC data analysis. Qualitative concordance in BCP-negative and BCP-positive samples was very similar (92.31% and 92.47% respectively, p=0.814). Concordance of data assessed by two methods in samples obtained during remission induction and during consolidation/intensification was also very close (90.32% and 92.62% respectively, p=0.799). Among remission induction samples high qualitative concordance was observed in both day 15 and day 36 (end of induction therapy) samples (91.30% and 89.19% respectively, p=0.859). These results are in contrast with the previously shown data that at the end of remission induction concordance between FC and molecular techniques is relatively low [G. Gaipa et al, ASH-2008]. As the flow cytometric data analysis in CD10(+) and CD10(-) BCP-ALL patients bases on different approaches, these types of leukemia were also analyzed separately. High qualitative concordance was found in samples from both CD10(+) and CD10(-) BCP-ALL (89.66% and 94.85% respectively, p=0.295). Despite high qualitative concordance between FC and RT-PCR data, low quantitative concordance between FC and RQ-PCR results was found (R²=0.291). Significant quantitative difference in FC and RQ-PCR data could be associated with variability of FG expression during treatment that does not correspond to the cells' number. Moreover, percentage of tumor blasts among all nucleated cells is calculated during FC MRD detection, while MRD value in RQ-PCR of FGt is corresponded to the initial FGt and control gene levels. FC appears to be better for the quantitative MRD assessment however FGt detection by RT-PCR is more appropriate for MRD qualitative detection due to higher sensitivity. Hence, FC is more applicable for MRD monitoring during early treatment phases, when the precise MRD value is essential. In contrast, PCR of FGt is more useful for later time-points where MRD-positivity corresponds to poor outcome.

Tandem application of FC and PCR of FGt seems to be a useful tool for long-term MRD monitoring in childhood BCP-ALL.

No relevant conflicts of interest to declare.