Search

The Clinical Genome Resource (ClinGen) Somatic Cancer Variant Interpretation Committee (CVI) provides oversight and recommendations in the expert panel process for ClinGen Somatic Cancer Variant Curation Expert Panels (SC-VCEPs). SC-VCEP use of the Oncogenicity SOP (PMID: 35101336) provides critical clarity where modifications (SOP changes), recommendations (CVI guidance), or specifications (SC-VCEP gene/cancer-specific use) are needed. We summarize SC-VCEP Oncogenicity SOP use informing guideline development. For example, Histone H3 SC-VCEP highlighted the need to incorporate classic phenotypes by creating a supporting evidence code (OP5). The CVI requested testing phenotype specificity within OP2, leading to developing recommendations on OP2 extended use. Conversely, if SC-VCEPs specify thresholds or alternative databases for cancer hotspots, these specifications are within the SC-VCEP scope and are not general CVI recommendations or modifications to the Oncogenicity SOP. Additional work includes CVI developing guidance for using in-silico predictors (OP1). We are investigating two commonly used predictors - REVEL and FATHMM - that operate by distinct random forest ensemble prediction algorithms and the Hidden Markov model, respectively. On a small variant set from the Oncogenicity SOP manuscript, manually curated 'true' oncogenic GOF and LOF variants displayed modest but significant differences in REVEL scores and larger differences in FATHMM, with potential discriminatory applications. We will report on scaling this comparison as well as general-use considerations. In addition, we share updates and seek community feedback on incorporating detailed functional data (OS2) in the Oncogenicity SOP as well as considering structured text for the curation of resistance variants. [ABSTRACT FROM AUTHOR]

Roughly half of the genomic alterations in pediatric cancers are not present in common adult cancers and are underrepresented in oncology knowledgebases. As many of these are fusions or large genomic rearrangements, they may be missed by standard analytic pipelines. Additionally, gene panels (biased towards adult cancers) may not include critical genes recurrently mutated in pediatric cancers. Therefore, it is important for sequencing centers processing pediatric tumor samples to have a comprehensive list of variants that have clinical impact in pediatric cancers. To facilitate and improve tumor profiling for pediatric patients, the Pediatric Taskforce of the ClinGen Somatic Cancer Clinical Domain Working Group has used the Pediatric Tumours volume of the WHO Classification of Tumours Online to extract recurrent genomic alterations in non-benign diagnoses with diagnostic, prognostic, or therapeutic significance. Diagnostic variants were divided into 2 categories, 'defining' and 'supportive'. Defining variants, such as MYOD1 L122R (spindle cell/sclerosing rhabdomyosarcoma), provide a strong indication of the diagnosis in the absence of any additional tumor information. Supportive diagnostic variants, like SMARCB1 loss in malignant rhabdoid tumors, may aid in diagnosis when coupled with additional information such as tumor histology or location. Due to the importance of recognizing these alterations in pediatric tumors, we have prioritized 99 defining variants and 71 supportive variants considered essential for diagnosis in 105 pediatric malignancies for curation in CIViC (www.civicdb.org). This includes 134 fusions/rearrangements and 36 SNVs/CNVs that will be discussed in this presentation. [ABSTRACT FROM AUTHOR]