CHU Sainte-Justine – Molecular Diagnostic Lab, Montreal, QC, Canada
Pediatric cancers present a highly heterogeneous and unique genetic landscape. Relevant alterations include fusion genes and other rearrangements, mutations and copy number variants (CNV). Methods currently used in clinical labs (e.g. FISH, PCR) are sensitive but assess one alteration at the time. Here we describe the deployment of transcriptome (RNA-Seq) and exome sequencing into a healthcare system funded clinical lab for molecular characterization of pediatric cancers.
Illumina NovaSeq and Dragen were used for sequencing and bioinformatic processing. Retrospective samples were used for clinical validations. RNA-Seq achieved 97% sensitivity for fusion detection. Exome sequencing identified 99% of SNVs and 95% of indels at a minimum variant allelic frequency (VAF) of 5%. Sensitivity for CNVs was 95% (20% VAF) at ~10Mbp resolution. Implementation into a ISO15189 molecular diagnostic lab was conducted with the provincial universal health care system. Exome and RNA-Seq were performed on tumour-infiltrated material, and another exome on non-tumoral blood. Virtual gene panels, ACMG and AMP-ASCO-CAP guidelines were used to classify variants for final reporting. Most patients (>90%) showed at least one somatic alteration. Germline pathogenic variants were found in <10% of cases. Implementation challenges were encountered: pre-analytical (i.e. multiple sample weeks apart, triage of clinical information), logistical (cost-effectiveness, turnaround time, low volume) and analytical. Finally, the performance of whole genome sequencing (WGS) was evaluated for high-risk or molecularly unresolved cases. WGS was able to retrieve all previously captured alterations.
Integration of genomic analyses into routine clinical care provides a comprehensive and unique molecular profile for pediatric cancers.