AFFILIATIONS
1 CHU Sainte-Justine, Montréal, Quebec, Canada, Université de Montréal, Montréal, Quebec, Canada
2 CHU Sainte-Justine, Montréal, Quebec, Canada
3 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada
4 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada
5 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada
6 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada
7 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada
8 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada
9 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada
10 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada
11 Molecular Diagnostic Laboratory, CHU Sainte-Justine, Montréal, Quebec, Canada
12 Molecular Diagnostic Laboratory, CHU Sainte-Justine, Montréal, Quebec, Canada
13 Molecular Diagnostic Laboratory, CHU Sainte-Justine, Montréal, Quebec, Canada
14 CHU Sainte-Justine, Montréal, Quebec, Canada
15 CHU Sainte-Justine, Montréal, Quebec, Canada, Université de Montréal, Montréal, Quebec, Canada
16 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada, CHU Sainte-Justine, Montréal, Quebec, Canada, Université de Montréal, Montréal, Quebec, Canada
17 Division of Hematology-Oncology, Department of Pediatrics, Centre Hospitalier Universitaire de Québec, Université Laval, Québec, Quebec, Canada
18 McGill University, Montréal, Quebec, Canada
19 McGill University, Montréal, Quebec, Canada
20 McGill University, Montréal, Quebec, Canada
21 Research Center Du Chus, Sherbrooke, Quebec, Canada, University of Sherbrooke, Sherbrooke, Quebec, Canada
22 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada, CHU Sainte-Justine, Montréal, Quebec, Canada, Université de Montréal, Montréal, Quebec, Canada
23 Université de Montréal, Montréal, Quebec, Canada
24 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada, CHU Sainte-Justine, Montréal, Quebec, Canada, Université de Montréal, Montréal, Quebec, Canada
25 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada, CHU Sainte-Justine, Montréal, Quebec, Canada, Université de Montréal, Montréal, Quebec, Canada
26 Centre de recherche Azrieli du CHU Sainte-Justine, Montréal, Quebec, Canada, CHU Sainte-Justine, Montréal, Quebec, Canada, Université de Montréal, Montréal, Quebec, Canada
ABSTRACT
Background
Risk-directed treatment of pediatric acute lymphoblastic leukemia (ALL) relies on timely integration of clinical, biological, and molecular features. Current diagnostic workflows require multiple assays, with results becoming available over days to weeks. Oxford Nanopore Technologies (ONT)-based whole-genome sequencing with adaptive sampling (asWGS) is a promising single-assay approach capable of detecting structural variants, copy number alterations (CNAs), small variants, gene fusions, and DNA methylation-based classification. ONT sequencing enables real-time analysis, creating the potential for rapid and comprehensive molecular characterization. However, evidence remains limited to small cohorts, and prospective real-time analysis has not been evaluated to date. We conducted a multicenter study to evaluate the performance of asWGS for molecular risk stratification in pediatric ALL and prospectively assess its turnaround time relative to standard multi-assay diagnostic testing.
Methods
Fifty-four patients with newly diagnosed ALL from the SIGNATURE precision medicine program (all 4 pediatric centers in Quebec, Canada) were included in study. All underwent standard testing including karyotyping, fluorescence in situ hybridization (FISH), comparative genomic hybridization (CGH), whole-exome sequencing (WES), and whole-transcriptome sequencing (WTS). asWGS was performed as previously described (Geoffrion et al., medRxiv 2025). Sequencing data were analyzed using the development branch of OncoSeq (nf-core-oncoseq), enabling real-time analysis and automated integrated reporting, including DNA methylation classification with Marlin. Real-time analyses were triggered after 1, 18, and 72 sequencing hours. Prospective real-time analyses were performed during sequencing for the most recent 28 patients. Of these, 15 patients consented to study enrollment and clinical testing concomitantly, enabling prompt initiation of real-time analyses and head-to-head comparison of asWGS turnaround times with conventional testing (real-time cohort).
Results
Among included patients, 42 were diagnosed with B-ALL (78%) and 12 with T-ALL (22%). Median age was 6 years [3-12] and male:female ratio was 1.30.
DNA methylation classification after 1 sequencing hour accurately identified leukemia lineage in 53/54 cases (98%). Among B-ALL cases represented in Marlin, subtype assignment was above threshold and accurate in 30/37 (81%). All risk-stratifying aneuploidies were identified within the first sequencing hour, including hyperdiploidy (8/8), hypodiploidy (1/1), and iAMP21 (1/1). Likewise, all focal CNAs involving CDKN2A/B (n=20), PAX5 (n=7), IKZF1 (n=4), and ERG (n=1) were detected across 21 patients. asWGS also identified two IKZF1 CNAs missed by CGH, including a subclonal intragenic deletion and an exonic duplication.
All but one subtype-defining fusion were identified by asWGS (31/32, 97%). The only missed fusion, P2RY8::CRLF2, was retrospectively recovered using Stellerator, now integrated into OncoSeq. asWGS also detected a CEBPD::IGH fusion missed by standard testing. All but one Tier 1 SNV/indel were detected, including sequence mutations involving PAX5 (2/3), IKZF1 N159 (n=2), and TP53 (n=1). The sole missed variant was a subclonal PAX5 mutation (VAF 7%), likely reflecting intratumoral heterogeneity (5/140 supporting reads on manual inspection). Overall, asWGS enabled molecular classification concordant with clinical testing in 53/54 patients while refining molecular classification and/or risk stratification in two patients (CEBPD-r and IKZF1 intragenic deletion).
In the real-time cohort (n=15 patients), median time from sample collection to comprehensive molecular classification with asWGS was 3 days (IQR 2.5-4), corresponding to 4 days (3.5-6) from the initial blood sampling. This outperformed FISH (5 days [3-6]), CGH (7.5 days [7-8.8]), karyotype (17 days [15.5-23]), WTS (23 days [21-26.3]), and WES (29.5 days [22.8-34.5]).
Conclusions
asWGS provided accurate molecular classification and risk stratification of pediatric ALL in a single assay with markedly reduced turnaround times compared to current clinical multi-assay testing approaches. This approach has the potential to deliver clinically actionable molecular alterations within days of presentation, streamline contemporary leukemia diagnostics, and represent a future standard-of-care in advanced genomic diagnostics of ALL.