Next Generation Sequencing: The Next Iteration of Personalized Medicine

Next generation sequencing sounds futuristic, maybe because it evokes the television series Star Trek: The Next Generation. “It does seem like science fiction, but next generation sequencing is going to become a science reality in a very big way starting this year,” says David I. Smith, PhD, professor of Laboratory Medicine and Pathology and chairman of the Technology Assessment Group of the Center for Individualized Medicine at Mayo Clinic, in Rochester. “One of the first applications will be in cancer therapy.”

Cancer has long been considered a disease of the genome, so it makes sense that figuring out what’s gone haywire in an individual’s genome can provide major clues about which drugs could block the aberrant pathways that promote cancer and can also help scientists to develop new drugs to hit those targets.

Think of next generation sequencing (NGS) applied to cancer care as personalized oncology. Synonymous with second generation sequencing, NGS refers to those DNA sequencing methods that came after capillary-based Sanger sequencing (first generation) back in 2005. Current next generation DNA and RNA sequencing companies include Illumina (TruSeq, HiSeq), Life Technologies (Ion Torrent, SOLiD), Complete Genomics (DNA nanoball sequencing), 454 Sequencing (pyrosequencing), and Oxford Nanopore Technologies (GridION).

One reason NGS is making its cancer care debut in 2012 is the dramatic advances in genetic sequencing technologies. Faster, cheaper, and more accurate sequencing technologies are leapfrogging each other on their way to the research community and to academic medical centers, carrying an opportunity to transform how cancer is diagnosed and treated, and to venture capitalists, who see an opportunity to make history and a return on their investment.

“Because biopsy samples are becoming smaller and smaller, the concept of a single test and a single drug is completely unsustainable,” says Kevin Krenitsky, MD, Foundation Medicine’s chief operating officer.

Transforming cancer care

“Right now, the most popular thing in NGS is sequencing a small target region or sequencing the full exome of 38 megabases of the 200,000 exons,” says Smith at Mayo Clinic. “Foundation Medicine Inc. is right in the middle of that, and one of the advantages is that they can do greater depths of sequencing. That’s important because cancers aren’t homogeneous.”

For Foundation Medicine Inc. (FMI), based in Cambridge, Mass., the pieces came together in 2010. Founding advisors include genomic all-stars Levi Garraway, MD, PhD, Todd Golub, MD, and Matthew Meyerson, MD, PhD — affiliated with Harvard Medical School, the Broad Institute, and the Dana-Farber Cancer Institute — and Eric Lander, PhD, with the Massachusetts Institute of Technology, Harvard Medical School, and the Broad Institute. FMI has been financed by a consortium of Third Rock Ventures, Kleiner Perkins Caufield & Byers, and Google Ventures to the tune of $40 million with another round of financing in the works.

In 2011, Michael Pellini, MD, left Clarient, GE’s cancer diagnostics and testing company, to join FMI as president and CEO. “The care of oncology patients is on the verge of being individualized,” says Pellini. “The molecular make-up of each tumor is going to drive personalized medicine. Our goal is to work with oncologists and pathologists and do nothing short of transforming cancer care. We want to change the way cancer patients are managed — not just in academic medical centers but in communities across the United States and internationally.”

“We’re looking at hundreds of cancer-related genes and about 21 translocations right now,” says FMI chief operating officer Kevin Krenitsky, MD. “Our cancer genomicists have been studying the field for more than 20 years. We’re confident that we are finding everything that is actionable in clinical care.”

The universe of what’s actionable — i.e., treatable with existing therapies — will expand as more tumors are sequenced and new drugs are developed. FMI has already contributed to the literature in collaboration with the Dana-Farber Cancer Institute in Boston. A paper in Nature Medicine reports that NGS sequencing has revealed genomic alterations directly associated with clinically available therapeutics or a relevant clinical trial of a targeted therapy in 72 percent of 24 non-small cell lung cancer (NSCLC) tumors and in 52.5 percent of 40 colorectal cancer (CRC) tumors.

Two novel gene fusions, KIF5B-RET in NSCLC and C2orf44-ALK in CRC, were among the alterations that might be treated by drugs. The fusion of C2orf44 and ALK produces an overexpression of anaplastic lymphoma kinase (ALK), the target of crizotinib (Xalkori), approved for the treatment of ALK-positive NSCLC, which suggests the possibility that ALK-positive CRC patients may respond to ALK-inhibitor treatment.

Deep coverage

Krenitsky thinks it’s no coincidence that these novel gene fusions were detected by the FMI comprehensive cancer genomic test.

Optimized for clinical-grade analysis of tumor tissues and designed to overcome multiple complexities, such as purity, ploidy, and clonality inherent in tumor genomes, the test will make its debut at the 2012 American Society of Clinical Oncologists annual meeting.

FMI uses the Illumina HiSeq 2000 system and Life Technologies’ Ion Torrent in-house, but Krenitsky does not go into detail about his company’s flagship test, which will list for under $6,000. “We’re always evaluating what’s best for us — and that could change at any time.”

Tumor specimens coming into FMI’s CLIA-certified lab are standard formalin-fixed, paraffin-embedded tissue blocks, but because the actual amounts of biopsied tumor tissue are getting smaller, it’s important to get all the genomic alterations on the first pass. Krenitsky claims the FMI test is “extraordinarily effective” at acquiring data on point mutations, copy number alterations, insertions, deletions, and rearrangements from very small tumor samples.

“If you cannot take small amounts of tissue and run a comprehensive assay that gives you all the relevant markers and then determine, based on the results, which therapies fit the patient, then the model breaks down — not only economically but technically because biopsy samples are becoming smaller and smaller,” Krenitsky explains. Sequencing is the wave of the future, he adds. “The concept of a single test and a single drug is completely unsustainable.”

Tumor tissue is complex. A tissue specimen consists of both normal cells and cancerous cells, and the proportion of cancer cells can be as low as 10 percent. Deep coverage, i.e., sequencing the same nucleotide base multiple times, increases the likelihood that as many cancer cells as possible are sequenced. Coverage depth is expressed as the number of times a base is read followed by an X. FMI routinely sequences at 500X to 1500X, which Krenitsky describes as “extraordinarily deep.”

The tumor or somatic tissue sequence is compared with the individual’s normal or germline sequence and the differences between the tumor and normal sequences indicate where potential treatment targets lie. Databases such as the Catalogue of Somatic Mutations in Cancer (Wellcome Trust/Sanger Institute) and The Cancer Genome Atlas (National Institutes of Health) are consulted. The treating physician gets a report annotated with scientific and medical literature relevant to the patient’s genomic alterations along with information on targeted therapies and clinical trials.

Collaborative research

“I think we’re very early on,” says Paul Billings, MD, PhD, chief medical officer at Life Technologies, in Carlsbad, Calif., which is conducting beta testing of its new Ion Torrent sequencing technology. “Until we accumulate more data and more experience, we could go through a period where there might be a lot of back and forth about how important a particular mutation is in the genome or in the 200 cancer genes that FMI happens to have on their assay, or the 170 that Oxford Biomedical Research Centre is going to have on their assay, or the 46 genes that are being used by the Knight Cancer Center at Oregon Health & Science University.”

Billings notes that Myriad Genetics launched its BRACAnalysis test in 1996 and has refined it based on its understanding of BRCA1 and BRCA2 gene mutations derived from a database of several hundred thousand cases. “We can’t wait around 17 years to have that quality of a database,” he adds. “We need one in a few years.”

“We are looking at a transformation — the technology advances have been tremendous,” says Lynda Chin, MD, at MD Anderson.

Interrogating genes known to be involved in cancers and for which effective treatments are either available or in trials makes sense for clinical purposes now, Billings says. But to build a comprehensive database of correlations between genomic alterations, signaling pathways, and disease will require intensive discovery and research work with whole genome and exome sequences. It will also require computational biologists and bioinformaticians to “sort the nonsense from the sense and to look at interactions in the genome,” as Billings puts it.

At the same time, clinicians will need to document patient clinical states, disease progression, and therapeutic response and nonresponse in searchable electronic medical records so that the information can be correlated with patient genetic data.

Lynda Chin, MD, leads another effort at the MD Anderson Cancer Center, in Houston, as scientific director of the MD Anderson Institute for Applied Cancer Science and as chair of the department of genomic medicine.

“We are looking at a transformation,” says Chin. “The technology has advanced over the last two or three years, and it has gotten ahead of our understanding. Commercial entities providing next generation sequencing can further advance the technical and analytic aspects, but we need insights on what to do with the data.”

The academic cancer research community, says Chin, will drive efforts to make sense of the complex genomic data. Once an insight is provided, the traditional academic model may not be efficient at converting that insight into a practical application, like a drug, a test, or device.

“We built a professional team to do drug discovery biology,” Chin explains. “You list a finite number of key biology questions that need to be answered, and you go after those questions in a targeted fashion. You reward product development instead of publication.”

Sequencing could soon follow the traditional trajectory of lower cost and faster turnaround.

Drivers and passengers

“That’s precisely what we’re doing,” says Christopher L. Corless, MD, PhD, referring to the collaboration between computational biologists and clinicians proposed by Billings and Chin. “That’s Dr. Druker’s vision for the whole Knight Cancer Institute. It’s his goal to provide personalized cancer care to every patient in Oregon.”

Corless is medical director of Knight Diagnostic Laboratories at the Knight Cancer Institute at Oregon Health & Science University in Portland. He works for Brian Druker, MD, director of the Knight Cancer Institute and another alumnus of Dana-Farber Cancer Institute. Druker is best known for developing imatinib mesylate (Gleevec), for which he and collaborators Nicholas Lydon and Charles Sawyers received the 2009 Lasker-DeBakey Clinical Medical Research Award. His laboratory is one of the big beta test sites for the new Life Technologies Ion Torrent technology. The Druker team has already developed its own database of 643 mutations across 53 genes in 4,000 patients using Sequenom microarray technology. Corless acknowledges that NGS is “hot.” He is also careful to note that NGS has two distinct applications — at least for now.

“If you use NGS to sequence the entire genome or the entire exome of a cancer, you’re going to find not only the common mutations but also the rare ones, plus a lot more data than you know what to do with,” he says. “If you use NGS as a very focused, cost-effective way of looking at a select group of genes that you already know are relevant, then NGS is going to give you exactly what you want.”

One such relevant mutated gene in the Philadelphia chromosome is the driver of chronic myeloid leukemia inhibited by Gleevec, “converting a fatal cancer into a manageable chronic condition.” Those last nine words come from the Lasker-DeBakey Clinical Medical Research Award and are also a motto for Blueprint Medicines, in Cambridge, Mass., whose business is developing highly selective drugs that inhibit aberrant and resistance mechanisms in cancer. Founded in 2011 by a team of scientists including Druker and Lydon, Blueprint Medicines scored $40 million in Series A financing led by Third Rock Ventures and made it onto Fierce Biotech’s 2011 “Fierce 15.”

“Our proprietary Insights-to-Validation platform enables us to quickly separate the driver genes and aberrations from passengers within specific genetic contexts,” wrote Christoph Lengauer, PhD, Blueprint Medicines chief scientific officer in an e-mail to Biotechnology Healthcare. According to Chris Varma, PhD, president and CEO of Blueprint Medicines, potential partners include academic medical centers, patient advocacy groups, pharmaceutical companies, and technology developers. Potential customers include oncologists, healthcare systems, payers, and ultimately patients.

Will NGS replace the one drug/one test companion diagnostics that check for known mutations? It depends. For known alterations, it would be more cost-effective to use an NGS gene panel instead of multiple one drug/one test diagnostics. But, such a panel may miss a rare mutation that may be a “driver” instead of a “passenger” in someone’s cancer.

“As we understand these cancers better and have more FDA-approved drugs available, I see these NGS panels making a serious entry into the marketplace, whether it’s a huge one like Foundation Medicine’s or one of our smaller ones,” says Corless. “I think they’ll begin to replace all the single gene assays that we have right now. How long will that take? Several years because oncologists are not very quick to change practice.”

More sequencing to come

“Right now whole genome sequencing as a clinical test doesn’t make any sense,” says Smith at Mayo Clinic. “It’s unaffordable, and what information do you get out of it? But soon, it’s going to be either because you want the genome sequence to figure out what’s going on, or the patient actually shows up with their genome already sequenced.”

For now, whole genome sequencing (WGS) is primarily a research tool for de novo sequencing, the kind of work Leroy Hood, MD, PhD, and colleagues at the Institute for Systems Biology (ISB) and other systems biology researchers are doing. As the technology advances, WGS will follow the familiar trajectory of lower cost and faster turnaround and very quickly may turn out to be even quicker than Smith predicts. Meanwhile, whole exome sequencing, transcriptome sequencing, epigenomic sequencing, and proteome analysis are waiting in the wings.

“Cancers change their mutations over time, so mutations aren’t always stable,” says Billings. “If you want to know whether a mutation has affected the amount of the protein or RNA, you will have to do transcriptomics and maybe even proteomics. Eventually, we’re going to have to incorporate epigenomics to identify the most important parts of the genome that are involved in cancer.”

For an oncologist or a cancer patient, having only part of the picture is not good.

Complete Genomics, in Mountain View, Calif., sequences only whole genomes and is a vendor to systems biology companies like ISB. Both Hood and George M. Church, PhD, are on the Complete Genomics scientific advisory board.

Professor of genetics at Harvard Medical School, director of the Center for Computational Genetics, and co-founder and advisor to Knome, in Cambridge, Mass., Church sees three main applications for NGS in the near future: to discover inherited predispositions for cancer, to compare a cancer genome to a normal genome, and to analyze a cancer for drug-resistant mutations.

Launched in 2007, Knome, which calls itself a genome interpretation company, provides researchers, drug developers, and clinicians in more than two dozen countries with tools and solutions that help determine the genetic basis of human disease and drug response. Knome was the first company to interpret a human whole genome for a commercial client in 2008 and will begin licensing its software to hospitals and clinics this year.

GenomeQuest, in Westborough, Mass., Omicia, in Emeryville, Calif., and Softgenetics, in State College, Pa., are also developing software to help clinicians identify actionable genetic variations.

Demand for NGS is increasing

The escalating demand for NGS services is fueling an increase in strategic partnering and merger and acquisition activities. Based on successful collaborations like the landmark 1,000 genomes asthma study by Johns Hopkins University School of Medicine, Illumina selected Knome as a partner for the Illumina Genome Network earlier this year.

“The so-called actionable genome is rather small today,” says Chin at MD Anderson. “It could be as small as 10 to 20 genes. Maybe we know about the functionality of a hundred genes, but we don’t have a hundred drugs that we can select from. A lot of questions need to be answered before we can make use of this genomic information.”

FIGURE.

Sample cancer genomic test report