We are living through a time of rapid change in the clinical genetics laboratory, though at times it may appear that change doesn’t occur fast enough given the challenges within the existing healthcare system. At the recent Molecular Medicine Tri-Conference held in San Francisco March 7 through 11 2016, here are a few summary highlights of the conference.
Seated fourth from the left is SeraCare's Chief Scientific Officer Russell Garlick.
Dr. Elaine Lyon is the Medical Director, Molecular Genetics at ARUP (Salt Lake City, Utah) as well as a past-president of the Association for Molecular Pathology (AMP), and currently serves on their Professional Relations Committee. She presented a keynote entitled “The Diagnostic Dilemma in Inherited and De Novo Disease”, and pointed out the challenges of interpreting gene testing when the technology has moved ahead (multiple gene-panel tests, and whole-exome sequencing in some cases) but the clinical utility (and subsequent payment / reimbursement) still needs to be demonstrated.
The interpretation of genetic (inherited) disease is the challenge, and she asked the question how will the Food and Drug Administration (FDA) regulate this interpretation without regulating the practice of medicine? This is a key point, as AMP has proposed a modernization of the CLIA regulations as an alternative. (For additional background on the FDA’s current stand on Laboratory Developed Procedures, we’ve written a few perspectives called “LDT Oversight: Why the FDA Makes a Point” and “LDT Oversight Counterpoint: Tempering FDA Arguments”.)
On the topic of data sharing, she said ‘no one laboratory has all the information’ needed for the complexities of interpretation. She mentioned a difference in the ClinVar database over the different and conflicting data around a MSH2 gene, which a 20-minute conference call between experts ultimately resolved. There are ‘many types of data that show pathogenicity - segregation, de-novo mutations, bioinformatics approaches’ - and thus molecular testing has many useful attributes.
She also put into perspective that there are some 4,600 disorders currently considered Mendelian, and 100% of individual drug metabolism response is pharmacogenetic; thus randomized prospective studies are needed to test symptomatic and pre-symptomatic individuals to evaluate the clinical utility of genetic testing.
Dr. Helen Fernandes is Associate Professor of Pathology and Laboratory Medicine at Weill Cornell Medical College (New York City, New York) as well as Chief of the Solid Tumor Section Laboratory and showed photographs to compare a surgical resection biopsy slide, a core needle biopsy, and a hematology smear, and asked the question how many cells are needed for 5ng of input material. (The answer is 400 to 600 cells.) She reported failure rates on their core specimens on the order of 20%, with a mean yield of 3.6 ng/sample, and the assay failure is attributed to lack of sufficient DNA for analysis.
Moving to resection samples, there may be as much material as needed, however with an abundance of necrosis the DNA quality varies. She pointed out the C->T and G->A sequencing artifacts induced by FFPE treatment, and referred to this 2015 paper that evaluates QIAGEN’s PAXgene-fixed as an alternative with favorable comparison to fresh-frozen material.
In addition, she showed DNA yield data of a manual-method QIAGEN QIAamp™ DNA FFPE Tissue Kit purification method against an automated Promega Maxwell™ 16 FFPE Tissue LEV DNA Purification Kit and showed favorable comparison both with the variant allele frequency (VAF) and overall variant coverage.
Dr. Derek Raghavan MD, PhD (President of Levine Cancer Institute) shared an interesting perspective he entitled “Integrated Cancer Care: Electronic Standards and Measurement of Impact of Value versus Volume Algorithms”, where he introduced the Levine Cancer Institute forming in 2011 as a network of 40 hospitals throughout the Carolinas serving about 1 million in population, seeing about 15,000 patients per year.
With a new network opening up, they had the challenge of “implementing Obamacare in a red state” (which means it was something not politically popular); they built a system where the electronic infrastructure was created with a single cancer Institutional Review Board, an electronic (i.e. virtual) tumor board created, physicians ‘can regulate themselves’ with electronic patient entry, registration, electronic medical records and tumor registry, and emphasis on evidence-based, cost-effective pathways for state-of-the-art cancer care.
The results speak for themselves - their clinical trial participation rate by minorities is much higher than the national average (typically 3% to 5%, compared to 23% at their center); accrual rates in clinical trials have tripled; consent-sheets are consistent and available across their system; and they even have a side-effect reporter smartphone ‘app’ that results in reduced travel and expense for patients, and improved standardization of care.
Dr. Hoon is the Director of the Molecular Oncology Program at the California-based John Wayne Cancer Institute and also serves on the Advisory Board of Guardant Health, one of the first companies to commercialize a circulating free DNA-based test. (He wrote up a review of circulating tumor DNA here.) One of the interesting things he mentioned about Guardant Health’s current Guardant360 panel was that they tested over 5,000 patient samples in 2015 with over 70% of these being reimbursed.
Their panel covers SNVs, amplifications, insertion-deletion mutations (“indels”) and gene fusions, with sensitivity down to 0.1% and a two-week turnaround time. Their analytical and clinical validation work was published last Fall in PLoS One.
Dr. Hoon went on to review circulating tumor DNA outcomes in melanoma cases, where patients were sentinel lymph-node positive; the John Wayne Cancer Institute pioneered the technique to examine draining lymph nodes during melanoma resection surgery, and was able to show several examples across 2,200 melanoma patients across 15 years with yearly follow-up. These individuals, when going from Stage III pre- to post-operative, their ctDNA showed a drop in all 54 genes, and a spike in ctDNA during Stage IV disease recurrence.
After surgical resection, monitoring is especially important, as once it recurs, 75% of patients do not survive longer than 5 years. He showed convincing data comparing ctDNA monitoring favorably against the existing blood-based biomarker for metastatic Stage IV melanoma, Lactate Dehydrogenase (LDH).
Are you using FFPE samples for your NGS-based oncology assays? You may find our Seraseq Solid Tumor Mutation Mix-I of interest. Are you developing circulating tumor DNA assays? We have you covered as well.