Foundation Medicine Big Barrier to Cancer Genomic Sequencing

[gpawelski]

Foundation Medicine and the Big Barrier to Cancer Genomic Sequencing

According to Dr. Eric Topol, Director, Scripps Translational Science Institute, recent studies have highlighted the potential value of whole genome or exome sequencing to precisely guide therapy for patients with cancer. However, almost all samples today go into formalin-fixed, paraffin embedded (FFPE) blocks, which alters the DNA and makes sequencing quite compromised and difficult.

He told Medscape Connect that the company, Foundation Medicine, which works with formalin-fixed paraffin-embedded (FFPE) blocks, and gets about 250-300 genes, the exons or coding elements in those genes, and reads out any potential links to drugs. But the rate-liimiting step appears to be getting something beyond these paraffin blocks. This is, we could do better if we could use either fresh formalin-fixed or frozen tissue samples from a biopsy or surgical specimen.

Topol says the problem is that pathologists are seemingly quite ritualistic. They don't want to go to frozen samples, which would be the best for whole genome sequencing. We're just at the cusp of getting started with this type of limited, not even full exome sequencing, just a few hundred genes, but that isn't enough.

Rencent papers in multiple journals in Nature, Science, Nature Genetics and Cell have shown that with hundreds of tumor samples fully sequenced, no two cancers are the same and a lot of the action is not in the coding elements of the genes per se. Whole genome sequencing certainly appears to be an ideal path to pursue, but we can't do it with the fixed problems that we have with the way samples are handled today.

Topol thinks that maybe we could get fresh formalin-fixed samples, as those appear to be well-suited to whole genome sequencing, although this is still a somewhat bootstrapped situation, like the paraffin-embedded samples. It appears that the long those samples are embedded, the harder it is to get a reasonable sequence beyond very targeted regions.

There are no two cancer tissues that are the same on a molecular basis. There's quite a bit of heterogeneity within the samples and multiple sequencing could account for that. And we also want to anticipate recurrence, match up the right driver mutations and the backseat passenger mutations, whether or not there's needed immunotherapy; all those things that could be done if we could get the right information from the get go.

So Dr. Topol asks this: How are we going to move to a world with a clinic of the future where patients with cancer can get whole genome sequencing rapidly? That is, to have annotation and interpretation of the genome with a day, and have your therapy precisely guided genomically?

[gpawelski]

Researchers have realized that cancer biology is driven by signaling pathways. Cells speak to each other and the messages they send are interpreted via intracellular pathways known as signal transduction. Many of these pathways are activated or deactivated by phosphorylations on select cellular proteins.

Sequencing the genome of cancer cells is explicitly based upon the assumption that the pathways - network of genes - of tumor cells can be known in sufficient detail to control cancer. Each cancer cell can be different and the cancer cells that are present change and evolve with time.

Although the theory behind inhibitor targeted therapy is appealing, the reality is more complex. Cancer cells often have many mutations in many different pathways, so even if one route is shut down by a targted treatment, the cancer cell may be able to use other routes.

In other words, cancer cells have "backup systems" that allow them to survive. The result is that the drug does not affect the tumor as expected. The cancer state is typically characterized by a signaling process that is unregulated and in a continuous state of activation.

In chemotherapy selection, molecular profiling examines a single process within the cell or a relatively small number of processes. All a gene mutation study can tell is whether or not the cells are potentially susceptible to a mechanism of attack. The aim is to tell if there is a theoretical predisposition to drug response.

It doesn't tell you the effectiveness of one drug (or combination) or any other drug which may target this in the individual. There are many pathways to altered cellular function. Functional Profiling measures the end result of pathway activation or deactivation to predict whether patients will actually respond (clinical responders).

It measures what happens at the end, rather than the status of the individual pathway, by assessing the activity of a drug (or combinations) upon combined effect of all cellular processes, using combined metabolic and morphologic endpoints, at the cell population level, measuring the interaction of the entire genome.

Translational science: past, present, and future

http://www.biotechniques.com/multimedia ... _3671a.pdf

Note: Foundation Medicine is not any different than Caris Diagnostics in Phoenix (now Miraca Life Sciences), beyond testing for standard pathology "targets" such as ER, PR, Her2, EGFR mutations, KRAS, BRAF. They aren't worth much for the sorts of chemotherapy which is used in 95% of all cancers and useless with respect to drug combinations. While fresh tissue is very dear and hard to come by, function trumps structure, in terms of potency and robustness of information provided than using archival paraffin blocks.

[gpawelski]

Robert A. Nagourney, M.D.

During the 1960s, 70s and into the 90s, a field of investigation arose that examined buyer’s practices when it came to the consumption of goods and services. Algorithms were developed to interrogate consumer choice. One such treatise was reported in 1994 (Carson, RT et al, Experimental Analysis of Choice, Marketing Letters 1994). What these researchers explored were the motivations and forces that drove consumption. When choices are offered, decisions are driven by such factors as complexity and utility. Complexity demands personal expertise or failing that, input from experts, while utility places a value on the good or service.

A recent report from a small biotechnology company called Foundation Medicine has brought this field of endeavor to mind. It seems that this group will be offering DNA sequencing to select chemotherapy drugs. This service, currently priced at $5,800, will focus upon a small cassette of genes that they described as “key” in tumor growth. Based on their technology they have already raised $33.5 million from the likes of Third Rock, Google and Kleiner Perkins Caulfield & Byers, venture capital sources. The CEO of Foundation substantiates the approach by pointing out that fully 150 people have already used their services. One hundred and fifty!

It seems from this report that our colleagues in the field of molecular profiling have studied the dictates of “Experimental Analysis of Choice” to a “T.” What we have is the perfect storm of medical marketing.

First, the technology is so complex as to be beyond the ken of both patients and physicians alike. Thus, expertise is required and that expertise is provided by those engaged in the field. Second, the utility of drug selection is beyond reproach. Who in their right mind wouldn’t want to receive a drug with a higher likelihood of a response when we consider the toxicities and costs, as well as the consequences of the wrong treatment? Dazzled by the prospect of curative outcomes, patients will, no doubt, be lining up around the block.

But, let’s deconstruct what this report is actually telling us. First, a scientifically interesting technology has been brought to the market. Second, it exists to meet an unmet need. So far, so good. What is lacking, however, is evidence. Not necessarily evidence in the rarefied Cochrane sense of idealized survival curves, nor even Level II evidence, but any evidence at all. Like whirling dervishes, patients and their physicians are drawn into a trancelike state, when terms like NextGen sequencing, SNP analysis and splice variants are bandied about.

Despite the enthusiastic reception by investors, I fear a lack of competent due diligence. To wit, a recent article in Biotechniques, “Will the Real Cancer Cell Please Stand Up,” comes to mind. It seems that cancer cells are not individual entities but networks. A harmonic oscillation develops between tumor, stroma, vasculature and cytokines. In this mix, the cancer cell is but one piece of the puzzle.

Indeed, according to recent work from Baylor, some of the tumor promotion signals in the form of small interfering RNAs, may arise not from the cancer cells, but instead from the surrounding stroma. How then, will even the most punctiliously perfect genomic analyses of a cancer cells play out in the real world of human tumor biology and clinical response prediction? Not very well I fear. But then again such a discussion would require data on the predictive validity of the method, something that appears to be sorely lacking.

Will today’s gene profile companies prove to be the biotech Facebook IPOs of tomorrow?

[gpawelski]

There is the issue about cell-lines vs fresh cells. Cell-lines have always played, and continue to play, an important role in drug screening and drug development.

The problem is that cell-lines do not predict for disease or patient specific drug effects. If you can kill cancer cell-lines with a given drug, it doesn’t tell you anything about how the drug will work in real world, clinical cancer (real-world conditions). But you can learn certain things about general drug biology through the study of cell-lines.

As a general rule, studies from established cell-lines (tumor cells that are cultured and maniplated so that they continue to divide) have proved worthless as models to predict the activity of drugs in cancer. They are more misleading than helpful. An established cell-line is not reflective of the behavior of the fresh tumor samples (live samples derived from tumors) in primary culture, much less in the patient.

Established cell-lines have been a huge disappointment over the decades, with respect to their ability to correctly model the disease-specific activity of new drugs. What works in cell-lines do not often translate into human beings. You get different results when you test passaged cells compared to primary, fresh tumor.

Research on cell-lines is cheap compared to clinical trials on humans. One gets more accurate information when using intact RNA isolated from “fresh” tissue than from using degraded RNA, which is present in paraffin-fixed tissue.

My question would be, do you want to utilize your tissue specimen for “drug selection” against “your” individual cancer cells or for mutation identification, to see if you are “potentially” susceptible to a certain mechanism of attack?

Cell Lines vs Fresh Cells

http://cancerfocus.org/forum/showthread.php?t=3702

Simply finding a mutation

http://cancerfocus.org/forum/showthread.php?t=3829

[gpawelski]

Next-generation sequencing (NGS) technologies have come a long way since 1977 when Frederick Sanger developed chain-termination sequencing, but are they ready for prime time in drug selection?

Researchers have realized that cancer biology is driven by signaling pathways. Cells speak to each other and the messages they send are interpreted via intracellular pathways known as signal transduction. Many of these pathways are activated or deactivated by phosphorylations on select cellular proteins.

Sequencing the genome of cancer cells is explicitly based upon the assumption that the pathways - network of genes - of tumor cells can be known in sufficient detail to control cancer. Each cancer cell can be different and the cancer cells that are present change and evolve with time.

Although the theory behind inhibitor targeted therapy is appealing, the reality is more complex. Cancer cells often have many mutations in many different pathways, so even if one route is shut down by a targted treatment, the cancer cell may be able to use other routes.

In other words, cancer cells have "backup systems" that allow them to survive. The result is that the drug does not affect the tumor as expected. The cancer state is typically characterized by a signaling process that is unregulated and in a continuous state of activation.

In chemotherapy selection, genotype analysis (genomic profiling) examines a single process within the cell or a relatively small number of processes. All a gene mutation study can tell is whether or not the cells are potentially susceptible to a mechanism of attack. The aim is to tell if there is a theoretical predisposition to drug response.

It doesn't tell you the effectiveness of one drug (or combination) or any other drug which may target this in the individual. There are many pathways to altered cellular function. Phenotype analysis (functional profiling) measures the end result of pathway activation or deactivation to predict whether patients will actually respond (clinical responders).

It measures what happens at the end, rather than the status of the individual pathway, by assessing the activity of a drug (or combinations) upon combined effect of all cellular processes, using combined metabolic and morphologic endpoints, at the cell population level, measuring the interaction of the entire genome.

Should oncologists begin using deep genome sequencing in their clinical practice? At the annual meeting of the European Society for Medical Oncology, two key opinion leaders battled it out over this topic in a debate.

[gpawelski]

At ESMO, experts assess the clinical use of genome sequencing

Vienna—Should oncologists begin using deep genome sequencing in their clinical practice? Next-generation sequencing (NGS) technologies have come a long way since 1977 when Frederick Sanger developed chain-termination sequencing, but are they ready for prime time? At the annual meeting of the European Society for Medical Oncology, two key opinion leaders battled it out over this topic in a debate.

The Argument for Deep Genome Sequencing

Arguing the pro position, Fabrice Andre, MD, PhD, of the Institut Gustave Roussy in Villejuif, France, said that embracing deep sequencing in daily clinical practice is not only the right thing to do, it is a necessity. The number of genetic biomarkers known to influence patient outcomes and care has risen dramatically in recent years and is only expected to grow, he said.

“The current system is not sustainable for hospitals and academic centers,” said Dr. Andre. “It’s not possible for [them] to run more than 10 bioassays per patient. We need to move to multiplex technology.”

For breast cancer, he said, clinicians can run tests for ER/HER2, TOP2A, FGR1, IGFR1R, EGFR, PAK1, BRCA1, CYP2D6, PTEN and PI3KCA among others. With whole genome sequencing, “you can assess all the genes that you want,” said Dr. Andre. “When you do one test for each biomarker, each biomarker has a cost. Keep in mind that three FISH [fluorescence in situ hybridization] is equal to the same cost of one whole genome CGH [comparative genomic hybridization] array.”

Whole genome sequencing also offers a number of other potential advantages. High throughput approaches can identify a large number of rare targetable gene alterations. This is increasingly important as researchers find genetic alterations that exist in 1% or 2% of patients. The technology also can capture minority clones that may be hard to identify when there is a low percentage of tumor cells in a sample. The next-generation sequencers have been proven to be accurate and they do not need large samples of tissue. Dr. Andre pointed out that some protein-based assays, which are used because they are less expensive than FISH, are not reliable. One study found that the immunohistochemistry test for the HER2 protein was accurate only 81.6% of the time (J Clin Oncol 2006;24:3032-3038, PMID: 16809727).

The “robust” deep sequencing technology is already being used for patient care at academic centers. One such example is the MOSCATO trial, which began in fall 2011. This trial enrolled 120 patients with difficult-to-treat cancers and is using whole genome sequencing to identify potential therapeutic targets. Once a target has been identified, patients receive targeted therapy in a clinical trial if one is available. The turnover time for sequencing is 15 days and the total cost is 1,500 euros, or roughly $2,000 per patient.

The cost of technology is expected to decrease dramatically in the next few years. By the end of 2012, Oxford Nanopore Technology is expected to launch a technology that is the size of a USB drive and will offer whole genome sequencing in 15 minutes for less than $1,000. Dr. Andre argued that deep sequencing will be less expensive than a multiplicity of tests.

He pointed to a case study recently described in the Journal of Thoracic Oncology as an example of a success story (2012;7:e14-e16). In the case report, a 43-year-old never smoker with lung cancer had tested negative for EML4-ALK on the approved companion genetic test for crizotinib (Xalkori, Pfizer). Sensing that an oncogenic genetic driver was spurring the patient’s cancer, clinicians ordered deep sequencing and identified a novel ALK fusion. The patient was treated with crizotinib and was recently reported to have had a complete response.

“In the context of prospective cohorts, but not clinical trials, I think we need to deliver NGS in order to detect a high number of rare, relevant genomic alterations and then treatment can be done in the context of Phase I trials or drug access programs,” said Dr. Andre.

The Argument Against Deep Genome Sequencing

According to Kenneth O’Byrne, MD, a consultant medical oncologist at St. James Hospital and Trinity College Dublin, Ireland, Dr. Andre is jumping the gun. “He makes the fundamental error that all people who are enthusiastic about new technologies always make and that is the non-application of evidence-based medicine,” Dr. O’Byrne said. “Deep sequencing is a fantastic tool, but it is a research toy and an expensive toy at the moment. For day-to-day practical medicine, we have to go by evidence base.”

Dr. O’Byrne cast doubt on Dr. Andre’s success story example. “They treated the patient with crizotinib and made the false conclusion that the ALK rearrangement they detected was responsible for the response. Do we know if that patient expressed MET? Is there any other reason [he] may have responded to crizotinib?” Dr. O’Byrne said.

He agreed that the cost of the sequencing technology was decreasing, but argued that analysis would remain expensive. He argued that the clinical benefit of identifying genetic drivers is still uncertain.

“I would argue that in lung cancer, and indeed in almost every other tumor, there are only a few proven genetic alterations that can be identified that actually affect the way we treat our patients in clinic,” Dr. O’Byrne said. “EGFR [epidermal growth factor receptor] mutations and ALK rearrangements are the only validated predictive biomarkers in NSCLC [non-small cell lung cancer].” He pointed out that these affect only 15% of lung cancer patients, and although there are targeted agents available, the jury is still out on whether the drugs that target these mutations improve survival.

As an example of this, he pointed out that an interim analysis of the PROFILE 007 trial presented at the ESMO meeting (abstract LBA1) showed that although crizotinib increased progression-free survival by 4.7 months compared with chemotherapy, there was no difference in overall survival. “If you look at all of the EGFR TKI [tyrosine kinase inhibitor] randomized controlled trials versus cytotoxic chemotherapy in EGFR mutation–positive disease, there has yet to be a proven [overall] survival benefit, despite obvious clinical benefits,” Dr. O’Byrne said. Researchers say the lack of overall survival advantage in many of these trials can be blamed on the large numbers of patients who cross over to the experimental therapy. “The argument is crossover, but we don’t know that yet,” he said.

Dr. O’Byrne urged caution, as several years ago, it was thought that tumor angiogenesis inhibitors would be the salvation of lung cancer patients and that did not happen. There was clear evidence that new blood tumor vessels were associated with poor outcome, but when researchers tested a slew of antiangiogenic TKIs in patients with lung cancer, none of them worked. These included apatinib, axitinib, cedarinib, motesanib, pazopanib, sorafenib, sunitinib and vandetanib. “There is still some promise that some of these might break through,” Dr. O’Byrne said, pointing to Boehringer Ingelheim’s BIBF1120. “But to date, we’ve spent billions of euros proving that many of these are of no value.

“In my view, and I feel this quite strongly, predictive biomarker tests must undergo validation and quality assurance before they are used rou- tinely in clinical practice,” Dr. O’Byrne said. “Deep DNA sequencing holds huge promise … but it is a research tool, and I do genuinely believe that a lot of clinically irrelevant data is generated that actually confuses the clinician and the patient.”

Clinical Oncology News Issue: December 2012 | Volume: 07:12