Adaptive Sampling enables rapid and complete profiling of pediatric cancers | LC26

Abstract
Timely and comprehensive molecular classification is essential for therapeutic decision-making in pediatric oncology, yet current diagnostic workflows rely on multi-step, time- and resource-intensive testing. We present a standardized Oxford Nanopore Technologies whole-genome sequencing approach using Adaptive Sampling (AS-WGS), optimized for pediatric cancers, that enables rapid and comprehensive molecular profiling in a single assay. AS-WGS was uniformly applied to 63 pediatric cancer samples (42 leukemias, 21 solid tumors) and compared with extensive clinical testing including cytogenetics, exome and RNA sequencing. AS-WGS achieved somatic-grade coverage (mean 167x) across 380 clinically relevant loci while maintaining sufficient pan-genomic coverage (18x) for genome-wide analyses. It reliably detected all classes of clinically actionable alterations, including copy-number changes, gene fusions, and single-nucleotide variants, capturing sub-clonal variants down to 5% allelic frequency, comparable to clinical sequencing. Four methylation classifiers (Malin, Alma, CrossNN, and Sturgeon) were also successfully applied, enabling cancer subtype confirmation and resolution of complex cases. Most clonal alterations were confidently identified within 24 hours of sequencing, with some, such as methylation-based classification and large copy number variations, detected within the first hour, providing same-day insights. We developed an open-source bioinformatic pipeline, nf-core-oncoseq, to enable on-the-fly analysis and automated reporting during sequencing. Together, this integrated platform consolidates complex molecular diagnostics into a single, rapid assay, establishing AS-WGS as a transformative approach for pediatric cancer diagnostics with the potential to accelerate molecular classification and expand timely access to precision therapies. Prospective validation on additional participants, including time to reporting and performance metrics, will be presented at the meeting.

Biography
Vincent-Philippe Lavallée, MD PhD, is an associate professor of pediatrics at University of Montreal and a physician scientist in pediatric oncology. He directs the Integrated Genomic Platform at CHU Sainte-Justine, where he leads the implementation of advanced molecular diagnostics. His work integrates exome and transcriptome sequencing with third generation long-read technologies and circulating DNA analysis to accelerate precision medicine and improve outcomes for children with cancer.