Jeremy P. Segal, MD, PhD, is the Director of Clinical Development and Policy at NYGC.

Part I of this blog post can be read here

Of all the many topics under discussion at the NCI Clinical NGS for Cancer conference, perhaps the most contentious was the issue of which tests should be performed and how they should be reported. In this instance, the great advantage of the new technology—inexpensive and high-volume sequence data—is also in some ways its Achilles heel.

The problem is that the new systems produce such a large volume of sequence data that they necessitate the application of a different testing paradigm. For example, if a lab were test a patient’s cancer sample for a small handful of individual cancer mutations in the traditional fashion using an advanced NGS platform, the test may occupy just one millionth of the capacity of the instrument. Finding a million samples to run is simply not feasible. So, labs are faced with options: either use smaller machines, or expand the amount of sequence information gathered from each sample.

Today, any lab currently performing clinical cancer testing on NGS machines is examining at least a sizeable panel of genes, and as described above, even a limited panel produces data on more than the small number of genes for which there are validated effective targeted drugs. Though so much data can be generated for little cost, the question remains: what to do with the data? In the end, clinicians will want only data that they can use to make patient care decisions, a class of data that attendees at the meeting termed “actionable.”

Beyond the genes for which we have known targeted compounds shown to be effective in clinical trials, an argument could be made that any number of other genes could be actionable if there exist clinical trials offering experimental compounds to cancer patients with mutations in those genes. Even if no clear treatment could be selected based on the results, the physician could at least intelligently recommend an appropriate trial that might benefit the patient. Others might go further and argue that any gene could be actionable so long as a clinician could make a reasonable medical judgment that a discovered mutation could provide a rationale for treatment with a known compound, even one approved for a different purpose.

Obviously, this last argument raises significant ethical questions related to treating patients based on minimal evidence. Those in favor of this approach at the meeting countered with the fact that medicine is often an art and that many treatment decisions are based on physician instinct. Additionally, when dealing with cancer patients whose diseases have progressed despite multiple attempts at therapy and whose prognoses are dismal, perhaps an educated guess based on genome sequencing is a morally acceptable approach. Counter-arguments were also made. For example, if terminal patients are treated in a piecemeal fashion with a best-guess approach, how does the research and clinical community evaluate that process to determine if it has been successful?

In the end, it was decided that a group should be convened to compile and tend a list of actionable targets, but there was no clear consensus on where the line should be drawn regarding what makes a truly actionable target. It’s a challenging issue and a moving target, one made all the more complex by the prospect of compensation either by insurance companies or patients, neither of whom enjoys paying for services that don’t have easily-defined value. Additionally, regulatory issues come into play, because as a lab begins to cast a wider net beyond the most thoroughly validated targets, the line begins to blur between clear clinical diagnostics and research, which have very different oversight and standards.

Many attendees’ opinions on this topic came down to a matter of personal philosophy regarding the ultimate role of physicians, researchers, and diagnostics in the treatment of cancer. For everyone, the ultimate goal was the same: personalized, in-depth, genomics-based analysis as a means to provide the best possible treatment strategy for each patient. This is the inevitable and promising future of cancer therapy, and now that we have the technology needed to make it a reality, it is just a question of figuring out how to get there.

Fortunately, widespread adoption of next-generation sequencing for personalized assessments of cancer genomes seems likely in the future, and it will undoubtedly accelerate as additional research-validated targets emerge and as sequencing costs continue their rapid decline.

About the New York Genome Center

The New York Genome Center (NYGC) is an independent, non-profit organization that leverages the collaborative resources of leading academic medical centers, research universities, and commercial organizations. Its vision is to transform medical research and clinical care in New York and beyond through the creation of what will be one of the largest genomics and bioinformatics facilities in North America.