Frequency and spectrum of clinically significant molecular alterations detected by tumor gene panel testing for pediatric cancer patients in the Texas KidsCanSeq study.

Abstract

10018 Background: Tumor sequencing has the potential to inform diagnosis, prognosis, and treatment of childhood cancer patients. Prospective data are needed to evaluate the clinical utility of these tests. Methods: Patients less than 18 years of age with extracranial solid tumors (ST), central nervous system tumors (CNS), lymphomas and histiocytic disorders (L/H) were enrolled in the Texas KidsCanSeq clinical genomics study at six Texas institutions. Available tumor samples from a subset of patients designated as high risk (based on poor prognosis at diagnosis or tumor recurrence) were sequenced using pediatric cancer-focused targeted DNA (initially 124 genes; expanded to 169 genes) and RNA (81 genes) panels developed at Texas Children’s Hospital. DNA variants (single nucleotide variants, insertion/deletions, splice site variants, copy number alterations) and fusion genes detected by RNA testing were analyzed. Tumor alterations were classified by molecular pathologists and oncologists as carrying strong (Tier 1) or potential (Tier 2) clinical significance following consensus guidelines from the Association for Molecular Pathology, American Society of Clinical Oncology, and College of American Pathologists. Results: A total of 626 patients were enrolled between 2018 and 2021, of whom 380 were designated as high risk. Tumor samples adequate for molecular testing were available from 342 of the 380 (90%) high risk patients (196 ST [57%], 132 CNS [39%] and 14 L/H [4%]). Clinically significant alterations were identified in 246 of 342 (72%) tumors: 165 (48%) from DNA testing only, 34 (10%) from RNA testing only, and 47 (14%) from both modalities. The frequency of clinically significant alterations was similar across tumor cohorts (74% ST, 68% CNS, 71% L/H). Alterations relevant to therapy-choice were reported in 127 of 342 (37%) tumors (35% ST, 26% CNS, 57% L/H); 87 of these 127 (69%) were not known from prior clinical testing. The most frequently altered therapeutic gene alterations (Table) involved BRAF (n=25), CDKN2A/B (n=20), SMARCB1 (n=17), FGFR1/2/3/4 (n=14), NF1 (n=13), PIK3CA (n=11), ALK (n=10), CDK4 (n=9), and RET (n=7). One hypermutated tumor was identified: a high-grade glioma (tumor mutation burden >190 mut/Mb) in a patient with Lynch syndrome due to a germline MSH6 variant. Conclusions: Tumor alterations of clinical significance according to consensus guidelines were reported in a large proportion of children with diverse high risk or recurrent cancer diagnoses, the majority of which were not detected by other standard pathologic analysis. These data further support broader adoption of tumor molecular profiling for these patients. [Table: see text]