Opportunity and cost of clinical whole genome sequencing. F. Dewey1,2, M. Grove1,2, C. Pan2,3, B. Goldstein4, J. Bernstein5, H. Chaib2,3, R. Goldfeder6, K. Ormond3,7, C. Caleshu1,2,5, K. Kingham8, T. Klein3, M. Whirl-Carrillo3, K. Sakamoto3,7, M. Wheeler1,2, A. Butte9, J. Ford8, L. Boxer4, J. Ioannidis4,10,11,12, A. Yeung1, R. Altman3,13, T. Assimes1, M. Snyder2,3, E. Ashley1,2,3, T. Quertermous2,3 1) Stanford Center for Inherited Cardiovascular Disease, Stanford, CA; 2) Stanford Center for Genomics and Personalized Medicine, Stanford, CA; 3) Stanford Department of Genetics, Stanford, CA; 4) Stanford Department of Medicine, Stanford, CA; 5) Stanford Department of Pediatrics, Stanford, CA; 6) Stanford Biomedical Informatics Training Program, Stanford, CA; 7) Stanford Center for Biomedical Ethics, Stanford, CA; 8) Stanford Division of Medical Oncology, Stanford, CA; 9) Stanford Division of Systems Medicine, Stanford, CA; 10) Stanford Department of Statistics, Stanford, CA; 11) Stanford Prevention Research Center, Stanford, CA; 12) Stanford Department of Health Research and Policy, Stanford, CA; 13) Stanford Department of Bioengineering, Stanford, CA.
Background Exome and whole-genome sequencing (WGS) are increasingly applied in clinical genetics and clinical research. WGS is expected to uncover potentially clinically significant findings regardless of the primary indication for sequencing. We investigated the burden of clinically reportable findings in WGS data, resources required for their discovery and interpretation, and costs of associated secondary evaluation. Methods We performed WGS on two platforms and clinical genome interpretation for adult participants in a primary care setting. A multi-disciplinary team reviewed all potentially reportable genetic findings and generated medical genomics reports for discussion with the study participant, genetic counselor, primary physician and medical geneticist. These physicians proposed secondary evaluations based on the report. Results Review of 90-127 Mendelian disease risk candidates in each participant (n=10, median age 53, 6 female) required a median of 54 (range 5-223) minutes per genetic variant, resulted in modestly good inter-rater classification agreement between genomic professionals (Gross kappa 0.61, 95% confidence interval 0.47-0.74), and reclassified 70% of genetic variants cataloged as disease causing in mutation databases to variants of uncertain or lesser significance. A median of 5 (range 2-6) reportable personal disease risk findings were discovered in each individual, including a frameshift deletion in BRCA1 strongly associated with hereditary breast and ovarian cancer that was found in a participant without family or personal history of the disease. Four of ten participants harbored variants cataloged by mutation databases as disease causing in genes recommended by the ACMG for return of incidental findings, and seven harbored apparent expected pathogenic novel variants in these genes. Cross-platform concordance of these latter variants was low (0-10%). Physician review of medical genomics reports prompted consideration of a median of 3 (range 0-10) follow-up diagnostic tests and referrals at an estimated total cost of $594 (range $362-1427) per individual. Conclusions WGS-based genomics involves considerable manual interpretation of potentially clinically significant genetic variants, uncertainty among genomics professionals and knowledge bases about these findings, and a moderate financial cost of secondary clinical evaluation. In select cases WGS will reveal genetic findings warranting early medical intervention.
You may contact the first author (during and after the meeting) at