Next-generation sequencing (NGS) platforms are becoming more automated, more cost-efficient, and while not quite turnkey, are reaching the point in ease of use that clinical applications of the technology are becoming mainstream. Clinical laboratories’ adoption of the technology is occurring at a rate that surpasses the uptake of other molecular technologies, including polymerase chain reaction (PCR), experts say.
DTET surveyed the NGS landscape to evaluate both the latest advances in the technology and trends in adoption by clinical laboratories.
Technological Evolution Faster and cheaper are the two words scientists eagerly listen for when instrument manufacturers unveil their newest sequencing offerings. So far, 2014 has greeted laboratory scientists with some exciting announcements. The first publicly released data generated from U.K.-based Oxford Nanopore Technology’s much anticipated thumb drive-sized MinION sequencer was a mixed bag. A technical review of the device by collaborator David Jaffe, from the Broad Institute in Cambridge, Mass., who used it to assemble two bacterial genomes, concluded that, as promised, the nanopore-based sequencing machine allowed for much longer reads (an average length of 5.4 kilobases, but up to 10 kilobases), compared to Illumina machines, which deliver fragments hundreds of base pairs long. The technology differs substantially from other NGS platforms as it identifies DNA bases by measuring the changes in electrical conductivity DNA generates as it passes through a biological pore. However, the review, presented by Jaffe at the Advances in Genome Biology & Technology conference (AGBT; Marco Island, Fla.; Feb. 12-15), raised concerns about “systematic errors” that prevented the assemblage of the genomes with just the MinION data. Requiring assistance, some at the conference argued, defeats the point of a handheld sequencer. The technology seemingly stalled for two years following the company’s initial announcement.
Science reports that in the interim, the company had to find a new membrane for the pore, as its original choice could not be manufactured on a large scale. The firm also shifted its focus from a large sequencer to a portable device. But following the silence, last month Oxford Nanopore not only made this initial MinION data public, but it also launched its early-access program to individual researchers interested in testing the device in their labs. Researchers must pay a $1,000 deposit, plus $250 for shipping costs. Illumina (San Diego) also had a dramatic start to 2014 with the unveiling of two new products. The company crossed the long awaited $1,000 genome threshold with its new HiSeq X Ten system. “With the HiSeq X Ten, we’re delivering the $1,000 genome, reshaping the economics and scale of human genome sequencing, and redefining the possibilities for population-level studies in shaping the future of healthcare,” said Jay Flatley, Illumina’s CEO in a statement. “The ability to explore the human genome on this scale will bring the study of cancer and complex diseases to a new level. Breaking the ‘sound barrier’ of human genetics not only pushes us through a psychological milestone, it enables projects of unprecedented scale.” Early adopters of the HiSeq X Ten system, which is expected to ship in the first quarter of 2014, include Macrogen, an NGS service organization in South Korea and Rockville, Md.; the Broad Institute; and the Garvan Institute of Medical Research in Australia. But while the announcement was hailed as an important milestone, this version of the $1,000 genome will remain out of reach for many clinical laboratories. The HiSeq X Ten system is available only as a combination of at least 10 HiSeq X systems, which would cost a total of at least $10 million. Few facilities have the sample volume necessary to make the investment worthwhile. “It’s a good deal if you can play in this game,” Chad Nusbaum, co-director of the sequencing program at the Broad Institute, told Nature. “It’s like the high-stakes poker table: If you’re playing $200 a chip, people who can’t afford those chips don’t care.” A possibly more accessible entry in Illumina’s sequencing portfolio is its NextSeq 500 system, which launched in January. The desktop machine can perform the most popular sequencing applications in less than a day (including one whole human genome, up to 16 exomes, up to 20 noninvasive prenatal testing samples, up to 20 transcriptomes, up to 48 gene expression samples, and up to 96 targeted panels) with a price tag of $250,000.
Clinical Adoption Sequencing throughput once reserved for large genome sequencing centers is now capable of being performed in clinical laboratories, thanks to continued platform advancements. As the technology continues to advance and new, cheaper, and more automated benchtop—and even smaller—platforms enter the marketplace, clinical adoption of NGS is expected to proliferate. Studies of the reliability of NGS-based testing using targeted gene approaches for routine clinical care (particularly for oncology and traditional genetic diseases) are materializing in the literature. Simultaneously, early adopters of whole-genome and whole-exome sequencing are already emerging. Gregory J. Tsongalis, Ph.D., director of molecular pathology at Dartmouth Hitchcock Medical Center (Lebanon, N.H.) points to a number of drivers that are pushing clinical laboratories toward NGS, including:
- The need to consolidate single-gene analysis into a single assay for operational efficiency;
- The cost-effectiveness of NGS compared to traditional PCR-based or other molecular methods; and
- Currently nonactionable data can be mined later as advancements in molecular understanding and therapy warrant in the future.
An additional issue is one of comprehensiveness. Unlike single-gene tests, multiplexed NGS assays ensure that all actionable tumor mutations are identified, even if the mutation occurs rarely. Using an NGS cancer hot spot panel, Tsongalis’s group not only detected 100 percent of the mutations seen with PCR, but the laboratory was also able to detect two additional actionable EGFR mutations not included in their laboratory’s single-gene assay. “The current standard approach focusing on single gene and sometimes single exon analysis detects only the most commonly described mutations. Less common mutations are not tested in the single mutation, single assay model because of design, cost, sample, and time constraints,” explains Tsongalis in a recent article published in the March issue of
Clinical Chemistry and Laboratory Medicine. “There is a disconnect between appropriate personalized or precision medicine and current testing algorithms. NGS promises to bridge this gap by allowing for mutation detection in multiple exons from multiple genes in multiple patient samples, simultaneously. . . .
NGS platforms offer an increased breadth of testing at a lower cost and without compromising assay performance and turn-around times.” In the recent paper, Tsongalis’s group evaluated the Ion Torrent AmpliSeq Cancer Hotspot Panelv2 (CHPv2), which is capable of identifying multiple somatic mutations in 50 genes in a single assay. They show that the panel (performed on the Personal Genome Machine) is suitable for clinical testing. The researchers compared the assay to routinely performed, standalone PCR-based methods for mutations in several genes (KRAS, V600E BRAF mutation, and the two most common EGFR activating mutations). Well-characterized cell lines, genetically engineered cell lines in fixed and embedded in paraffin, and 62 clinical samples (lung, colon, melanoma, rectal, and ovarian) that had been previously tested with the laboratory’s current single-gene methods were used. Normal kidney, tonsil, and colon tissues served as controls. The researchers demonstrated that there was 100 percent concordance in accuracy between previous PCR results and the corresponding variants identified using the Ion Torrent panel, as well as high precision. The limit of detection was 5 percent for single nucleotide variants and 20 percent for insertions and deletions. Specificity studies using normal FFPE tissue previously tested by PCR methods were also 100 percent. Some additional practical findings from Tsongalis’s group: 100-times coverage is needed to identify somatic mutation results with confidence; fine needle aspirates, biopsies, and resected surgical pathology specimens were all equally successfully analyzed; and input DNA concentrations well below those recommended by the manufacturer (as little as 1 ng of DNA isolate from FFPE) were adequate. Furthermore, the researchers could return meaningful results through a data analysis process they called “efficient, user-friendly, and robust.” The post-analytical data analysis, the authors say, is critical to ensure that accurate and actionable variants are returned to the clinicians in a comprehensible way. The variant processing pipeline they developed allows the masking of variants of limited clinical value, which can quickly decrease the number of variants returned to clinicians by an order of magnitude. Additionally, the filtered variants can be separately stored and mined when new molecular understandings become available. Tsongalis tells
DTET that the “take home” from his experience has been that the technology available today may not be turnkey quite yet, but laboratories also don’t need significant molecular expertise to enter the NGS field. His biggest advice, though: “Don’t go at it alone.” A multidisciplinary approach that utilizes laboratory folks, pathologists, oncologists, genetic counselors, and information technology/bioinformatics experts is necessary to build the best system for each individual institution.
Not Just Targeted Clinical Sequencing Even as NGS-based testing is beginning to permeate clinical laboratories, early adopters are employing whole-genome sequencing in clinical care. Children’s Mercy (Kansas City, Mo.) is using several NGS-based tests in clinical practice to identify genetic diseases and for the first time, whole-genome testing is even being used in urgent scenarios to inform diagnosis and therapeutic decisions in acutely ill infants in the neonatal intensive care units (NICUs). In October 2013 Children’s Mercy began employing its 50-hour STAT-Seq whole-genome analysis to test for 3,500 genetic diseases in an immediate care scenario. “We believe that 30 percent of the babies in our NICUs are likely to benefit from next-day genome sequencing,” said Stephen Kingsmore, the director of pediatric genomic medicine at Children’s Mercy. STAT-Seq is being developed by the hospital’s Center for Pediatric Genomic Medicine in collaboration with Illumina (using its HiSeq 2500 system) and PerkinElmer. STAT-Seq additionally uses in-house software that integrates physician-entered clinical features for individual patients and a comprehensive set of relevant diseases. This software “substantially automates identification of the DNA variations” that can explain the child’s condition. Future studies are also expected to show that cutting time to diagnosis also cuts costs. Children’s Mercy received a $5 million National Institutes of Health (NIH) grant in 2013 to generate the data needed to guide the use of rapid genome sequencing in the diagnosis and treatment of acutely ill babies. These efforts will further improve the speed and cost-effectiveness of STAT-Seq, as well as evaluate the benefits, and potential harms, of rapid genome sequencing in newborns. In addition to STAT-Seq, Children’s Mercy clinically employs exome sequencing and its TaGSCAN (Targeted Gene Sequencing and Custom Analysis), a test that screens for more than 750 diseases that are the result of a single-gene defect, including muscular dystrophy, cystic fibrosis, and polycystic kidney disease. TaGSCAN uses NGS technology along with proprietary software that allows a symptom-based analysis to diagnose genetic diseases. Unlike the rapid turnaround for STAT-Seq, TaGSCAN results are delivered in six to eight weeks and the test costs less than $3,200. Children’s Mercy is not alone. Partners Healthcare, Geisinger Health System, Scripps Health, and the Medical College of Wisconsin are all using whole-genome sequencing, while Baylor Scott & White Health, Emory Healthcare, and UCLA Health are providing clinical exome sequencing. Partners HealthCare (Boston) is one of the first hospital systems to offer whole-genome sequencing, analysis, and interpretation in clinical care and expects to sequence the genomes from 50 patients in the next year. Partners is reportedly charging about $9,000 for an individual, including interpretation and analysis, and $18,000 analysis for sequencing for a child and both parents to better understand a child’s genetic disorder. Additionally, Partners is enrolling about 200 patients and their primary-care physicians or cardiologists in an NIH-funded project to study the integration of whole-genome sequencing into clinical medicine. Including doctors in the project allows researchers to evaluate how physicians are using sequencing information in caring for their patients. National laboratory giants Quest Diagnostics and LabCorp have both recently entered into multiyear licensing agreements with Illumina, signaling anticipated expansion of their clinical NGS testing. “Investing in next-generation sequencing, which is increasingly used in several clinical areas as well as clinical trials, is a key element of our strategy,” said Jay Wohlgemuth, M.D., Quest’s senior vice president of science and innovation, in a statement at the time of the announcement. In the agreement announced earlier this year, Quest will have broadened rights to use Illumina’s sequencing and genotyping technology, including the MiSeq platform and related consumables, to develop, validate, and offer molecular laboratory-developed tests for several diseases, including several cancers and neurological and women’s health disorders. LabCorp similarly will have expanded rights to use Illumina’s NGS instruments to develop new diagnostic tests in genetic testing, oncology, transplant medicine, and forensics, in addition to human leukocyte antigen tests already planned to be introduced this year.
Takeaway: Clinical adoption of NGS is expected to proliferate at a pace unseen with even other molecular technologies. NGS appears to have reached that inflection point where targeted sequencing assays will shortly become routine in clinical laboratories while more early adopters move into clinical whole-exome and whole-genome sequencing. Side Box 1: Novel NGS Platforms Oxford Nanopore was not the only novel NGS platform maker garnering attention at AGBT. GenapSys (Redwood City, Calif.) unveiled plans for a DNA sequencer with the footprint of an iPad. The company’s platform uses a semiconductor chip like Life Technologies’ system but differs in how the electric signal of the DNA molecule is measured. The company claims it is very easy to use and incredibly low-cost—possibly as low as a few thousand dollars to purchase—but reportedly won’t be available for another year. The company’s founder, Hesaam Esfandyarpour, Ph.D., tells Forbes the goal is to bring this technology “to the hands of the masses,” with a per-run cost of a couple hundred dollars. The company closed a $37 million series B round of financing back in November 2013 that included capital from billionaire technology investor Yuri Milner.
Side Box 2: NGS Testing in Current Clinical Practice Laboratories currently using clinical whole-genome sequencing:
- Children’s Mercy (Kansas City, Mo.)
- Geisinger Health System (Danville, Pa.)
- Medical College of Wisconsin (Milwaukee)
- Partners HealthCare (Boston)
- TruGenome Clinical Sequencing (offered by Illumina [San Diego])
Laboratories currently using clinical whole-exome sequencing:
- Ambry Genetics (Aliso Viejo, Calif.)
- Baylor College of Medicine (Houston)
- Emory Healthcare (Atlanta)
- University of California, Los Angeles
