NGS-Based MRD Quantitation: An Alternative to qPCR Validated on a Large Consecutive Cohort of Children with ALL

Together with multicolor flow cytometry, quantitation of clonal immunoglobulin (IG) and T-cell receptor (TR) gene rearrangements represents the current standard for the detection of minimal / measurable residual disease (MRD) in treatment protocols for pediatric acute lymphoblastic leukemia (ALL) patients. Despite the adoption of next generation sequencing (NGS) in the routine identification of clonal IG/TR gene rearrangements as markers for MRD detection, real-time quantitative (q)PCR is still the standard for MRD quantitation in follow-up samples. So far, no large-scale direct comparison of qPCR- and NGS-based MRD quantitation has been performed.

We compared qPCR- and NGS-MRD evaluation in a cohort of children with B-cell precursor (BCP) ALL treated on the AIEOP-BFM ALL 2009 protocol and assessed the feasibility and relevance of this method for the stratification at day 33 (EOI).

In total, 459 patients were diagnosed with BCP-ALL from 2010 to 2018, and 437 of them were included in our study based on the availability of residual DNA material isolated from day 33 bone marrow aspirates and having at least one IG/TR MRD marker detectable by standard qPCR with protocol-required sensitivity of 10 ^-4. Sequencing libraries were prepared according to the EuroClonality-NGS group SOP (Brüggemann et al, Leukemia 2019) with the total DNA input normalized to the equivalent of 150,000 nucleated cells to reach MRD sensitivity of 10 ^-5 and sequenced on Illumina NovaSeq and MiSeq instruments. In total of 780 IG/TR markers evaluated by both NGS and qPCR. Sequencing data were analyzed using the ARResT/Interrogate (Bystry et at, Bioinformatics 2017) pipeline and a custom bioinformatic analysis process and the NGS-MRD results were normalized to the EuroClonality-NGS central in-tube quality/quantification control (cIT-QC; Knecht et al, Leukemia 2019).

From the total 780 IG/TR MRD markers evaluated by both methods, 629 (80.6%) were concordant with 242 markers being MRD positive and 387 negative. From 82 markers that were only positive by qPCR and not by NGS, 76 were positive below the quantitative range (positive non-quantifiable). Specificity analysis was performed for each marker by searching for the junction sequence across the dataset of all patients' NGS results. Based on these results, 22 out of 82 markers positive only by qPCR were classified as potentially unspecific (false positive) and similarly 32 unspecific markers were identified among the 69 positive only by NGS. This was also supported by unspecific amplification of the polyclonal control in 27 out of these 32 corresponding qPCR systems, in some cases leading to qPCR negative classification determined by the EuroMRD guidelines.

Overall stratification of patients based only on day 33 MRD by qPCR or NGS was concordant in 76% of patients by both methods, while in 19% of patients, NGS-MRD quantitation led to the assignment to a lower-risk group, mainly due to the elimination of false-positive results. Furthermore, analysis of all positive markers across all patients' NGS libraries showed, that one out of 10 markers (mainly in the IGK, TRG and TRD loci) used for qPCR-MRD stratification did not provide satisfactory specificity, although they fully met EuroMRD criteria during the optimization of qPCR patient-specific assays.

Our results show that NGS-MRD is highly concordant with traditional qPCR-based strategy and has comparable sensitivity and clinical value in the setting of a BFM-based clinical protocol, while being less laborious and providing significantly more specific results and additional information on the IG/TR repertoire (Kotrova et al, Blood 2016). Our study also emphasizes the importance of selecting MRD markers of adequate specificity at diagnosis. Currently, this selection can be assisted by these broad sequencing data on IG/TR repertoire of large number of patients. Based on these results, we propose that frontline NGS-MRD evaluation developed by the EuroClonality-NGS working group can be used as an alternative to traditional qPCR-based MRD quantitation in future MRD-based treatment protocols.

Supported by grants NU20-03-00284 and NU20-07-00322 from the Czech Health Research Council and 534120 from Charles University. All methods were established through collaboration within the EuroClonality-NGS and EuroMRD groups.