SANTA CLARA, Calif. – Just as it does with treatments, the National Comprehensive Cancer Network (NCCN) offers detailed guidelines on genomic testing by cancer type. These are key in determining what physicians can prescribe routinely and what insurers will cover. But those guidelines aren’t followed regularly outside a major research hospital setting, thereby obviating access to tumor genetic information that could help to better guide treatment.
Even if current guidelines are followed, physicians and patients can get information back from the tests that neither party is prepared to process. Analyzed tumor mutations can be either somatic or germline, with the former being cellular mutations specific to that tumor and the latter being heritable mutations. The latter have implications not only for the risk of cancer recurrence, but also for the family of the patient who run the risk of inheriting or themselves passing on that mutation. These were the conclusions of a standing-room only panel on the topic at the Precision Medicine World Conference.
Without a guideline
The lack of guideline-based genomic testing outside the research hospital means that precision medicine isn’t a reality for certain patients – when it easily could be. Physicians, even oncologists, are inhibited not just by a lack of genetic knowledge – many didn’t receive training about the topic in medical school – but also sometimes by the difficulty of ordering outside tests.
“All of us in NGS labs think that patients are getting completely genotyped. In actuality, day-to-day, they're really not in the community setting,” said Rebecca Nagy, the senior director, medical affairs at Guardant Health Inc. “Patients are not getting guideline, complete testing. And when they don't get guideline, complete testing, they go on the wrong drug at the wrong time.”
She cited a study conducted by Guardant Health at a community health practice with about 75 sites. It evaluated the use of genomic testing that conforms to NCCN guidelines specifically in non-small-cell lung cancer (NSCLC) and colorectal cancer (CRC).
It found that only 8% of NSCLC patients were getting tested for all the guideline biomarkers. Even one of the most established ones with multiple drugs to treat those with it, EGFR, was only tested in about 70% of NSCLC patients in this community hospital system. In CRC, only about 40% were getting testing that conforms to NCCN guidelines. Therefore, a large segment of patients was identified that had not received chemotherapy as they should have or received a contraindicated drug.
In addition, patients who participate in cancer clinical trials have a very specific profile. Only 3% of adult cancer patients take part in a clinical trial, although that ratio is much higher for children. Participation in a clinical trial has been found to be associated with an increased life expectancy.
The typical adult clinical trial participant is a white male with metastatic disease who has private health insurance and is being treated at an academic center. That points to another source of inequity in terms of who has access to the innovative tests and treatments.
Some of the most established cancer mutations, such as BRCA, are at the germline. But the protocol for validating, handling and communicating this information remains far from routine. Testing companies are technically prohibited from reporting results other than those ordered, noted Shari Brown, who is a molecular pathologist at Neogenomics. Germline results occasionally fall in that category, but not reporting them presents an ethical conundrum.
In addition, physicians may not understand the implications of a germline finding, and, therefore, not communicate it correctly to patients. There are times suspicion of a germline finding is raised in a tumor biopsy test, but then it would need to be confirmed via a blood test.
One company with a representative on the panel, Tempus, runs tests simultaneously on blood, normal tissue and tumor tissue. That enables it to distinguish between somatic and germline mutations and to advise physicians accordingly.
“We do a tissue normal germline match. We run the DNA on the tumor sample and a normal tissue sample, then we match the blood,” explained Mark Oldroyd, the senior vice president of commercial markets at Tempus. “So, we're actually trying to capture everything that might be inherited in terms of germline all under one test platform. The advantage obviously is [that] we’re bringing in the blood at the same time [that] we're doing the solid tumor biopsy.”
But many physicians and payers often remain uncomfortable with and uncertain about how to handle or share germline data. It may not technically be required to treat that cancer – a nonrelevant germline mutation may turn up, but it remains unclear if the physician routinely has the knowledge or the disposition to report it to the patient and the family. Payers also are looking for immediately actionable data for a particular patient and may not even insure other family members for whom the data is relevant.
All the panelists underscored the usefulness of every patient working with a trained genetic counselor, as they would be able to explain information on germline mutations and convey the implications for an individual’s family.
These issues are more relevant for certain kinds of cancers vs. others, noted Brown. She added that with carcinomas and melanomas, it’s much more common to find germline mutations. With sarcomas and leukemias, the vast majority of tumors only have somatic mutations.
The panel had a long wish list of advances in terms of analyzing and conveying cancer genomic testing information. In addition to ubiquitous genetic counseling access, another goal was a universal electronic health record (EHR). Brown observed that genomic tests often are repeated unnecessarily, while there are also often discrepancies with different panels measuring distinct mutations. A universal EHR would make it much easier to manage and harmonize the data, as well as help make it more usable to the physician.
The ultimate dream is to match a particular genomic profile with optimized treatment outcomes based on histories of highly similar patients. The CEO of Foundation Medicine, Cindy Perettie, pointed to her company’s collaboration with real-world data analytics player Flatiron Health that dates back to 2014. Both have since been acquired by Roche, with an eye toward retaining an edge in genomic-driven drug development.
“Today, we have about 350,000 patients where we can follow their outcomes and understand based on their genomic profile,” said Perettie. “Our goal is to get this to the point of care. Today, we're piloting this clinical genomic database with a select number of sites, essentially doing molecular profiling and then looking at it together with those sites to understand how those patients can be matched to a therapy.”
She concluded, “But imagine a day where you have a profile for a patient, and you can look at 100 other patients [who] look similar to your patient – and understand what is the best treatment paradigm for them in front-line, second-line and third-line to guarantee that patients are going to have the best outcomes.”