Diagnostic yield of clinical tumor exome sequencing for newly diagnosed pediatric solid tumor patients. D. W. Parsons1,2,3,4, A. Roy1,5, F. A. Monzon2,4,5, D. H. López-Terrada1,4,5, M. M. Chintagumpala1,4, S. L. Berg1,4, S. G. Hilsenbeck4,6, T. Wang4, R. A. Kerstein1, S. Scollon1, K. Bergstrom1, U. Ramamurthy1,7, D. A. Wheeler2,3,4, C. M. Eng2, Y. Yang2, J. G. Reid2,3, D. A. Muzny2,3, R. A. Gibbs2,3, S. E. Plon1,2,3,4 1) Pediatrics, Baylor College of Medicine, Houston, TX; 2) Molecular and Human Genetics, Baylor College of Medicine, Houston, TX; 3) Human Genome Sequencing Center, Baylor College of Medicine, Houston, TX; 4) Dan L. Duncan Cancer Center, Baylor College of Medicine, Houston, TX; 5) Pathology and Immunology, Baylor College of Medicine, Houston, TX; 6) Medicine, Baylor College of Medicine, Houston, TX; 7) Institute for Clinical & Translational Research, Baylor College of Medicine, Houston, TX.

   Although advances in sequencing technologies now allow provision of genome-scale data for cancer patients, experience with systematic application of clinical genomic sequencing remains sparse. The BASIC3 study follows pediatric patients with newly diagnosed solid tumors (targeted enrollment n=280) at Texas Childrens Cancer Center for two years after performing CLIA-certified whole exome sequencing (WES) of blood and tumor. Since the study opened in August 2012, 70/82 (85%) eligible families have consented to participate, allowing an early assessment of the feasibility and utility of performing clinical tumor WES for these patients. Snap-frozen tumor and matched normal blood samples are submitted to the Baylor College of Medicine (BCM) Whole Genome and Cancer Genetics Laboratories for WES using a clinically-validated pipeline, with the resulting tumor and germline WES reports submitted into the electronic medical record. The first 70 patients comprise a diverse representation of diagnoses, including 23 (33%) with CNS tumors and 47 (67%) with non-CNS solid tumors. Despite limited diagnostic biopsies in many patients, snap-frozen tumor samples adequate for WES were available from 61/70 (87%). Tumor WES results have been reported for the first 32 patients, revealing a median of 9 (range 1-35) protein-altering mutations per tumor and alterations of known cancer genes such as ALK, KRAS, NRAS, MET, JAK2, FGFR3, ARID1A, CTNNB1, and TP53. Nine of 32 tumors (28%) contained mutations classified as having proven or potential clinical utility, including 9/24 (38%) non-CNS tumors and 0/8 CNS tumors (Fishers exact test, p=0.07). Notably, only four of these mutations would have been detected on the clinically-available BCM Cancer Gene Mutation Panel (v.2). These results demonstrate the feasibility of routine tumor WES in the pediatric oncology clinic, indicate that potentially clinically-relevant mutations can be identified in a substantial minority of pediatric solid tumor patients, and suggest that the diagnostic yield of WES may be greater for non-CNS tumors as compared to CNS tumors. The diversity of mutations identified suggests that exome or genome-scale diagnostic approaches for pediatric solid tumors may be favored over targeted mutation panels which are often optimized for adult malignancies. Clinical utility of WES data as assessed by the pediatric oncologists will be measured longitudinally for each study patient. Supported by NHGRI/NCI 1U01HG006485.

You may contact the first author (during and after the meeting) at