Better Utilization of Broad Panel NGS Possible in 2018

Stuart Goldberg, MD

A recent study of oncologists’ use of next-generation sequencing (NGS) to test for actionable genetic mutations reveals a stark gap between the appropriateness and the actual utilization of these tests. That situation can be expected to change dramatically in the year ahead, owing in part to growing recognition of the value of NGS in guiding the use of precision medicine, according to Stuart Goldberg, MD, chief scientific officer of Cota, who spoke with OncologyLive^® about how clinical practice is expected to catch up with scientific advances in 2018.

The results of a study of genomic profiling of patients with advanced non—small cell lung cancer (NSCLC) in 15 community practice settings in New Jersey and Maryland revealed significant underuse of genomic testing (TABLE).¹ The patients underwent treatment from January 2013 to December 2015, and although national guidelines call for routine testing for EGFR and ALK mutations for these patients (N = 814), only two-thirds underwent testing for both aberrations. Among those who were not tested for these biomarkers, 52% received chemotherapy without documented reasons for not testing. Potential barriers to testing included lack of sufficient biospecimens, failure to integrate biomarker testing into routine pathology workflow, slow lab turnaround times, and uncertainty of payment.

Table. Genomic Testing Patterns in Advanced NSCLC¹

The report also found a correlation between patient survival rates and failure to conduct NGS testing. For the 131 patients who received a targeted therapy during their treatment, the median overall survival (OS) was 31.8 months. In contrast, 482 patients received cytotoxic chemotherapy, and their median OS was 12.7 months. For 17 patients with EGFR or ALK mutations who did not receive targeted therapy but instead received chemotherapy, OS was 15.5 months, investigators reported.

Goldberg, who is one of the report’s authors, said the evaluation of the 89 physicians’ testing habits showed that two-thirds of the time, patients underwent immunohistochemistry or fluorescence in situ hybridization tests, possibly because these simple tests could readily be performed in the lab. For the more comprehensive tests, such as the FoundationOne CDx (F1CDx) test for 324 genes and 2 genomic signatures, the utilization rate fell to under 20%.

Cota, based in New York, New York, catalogs and analyzes patient medical data to find value in healthcare. The study was supported by Guardant Health of Redwood City, California, a manufacturer of liquid biopsy tests using NGS. The company also is working with Foundation Medicine of Cambridge, Massachusetts, on a clinical study to measure changes in survival benefit and cost savings among patients with untreated metastatic NSCLC who undergo comprehensive genomic profiling with F1CDx.

Goldberg said the results of the New Jersey— Maryland study suggest that physicians were willing to test for suspected genetic alterations for which FDA-approved drugs were available, but less likely to test for a broader slate of mutations recommended by the National Comprehensive Cancer Network. The panel recommends testing for EGFR, ALK, BRAF, ERBB2 (HER2), MET, RET, and ROS1. Just 8% of patients in the 814-patient cohort were tested for those mutations. The report stated, “Despite the guidelines specifically recommending testing without regard to the clinical characteristics, the physicians appeared to withhold testing for active tobacco users. Smokers as a group have a lower frequency of EGFR and ALK genomic alterations, but BRAF and MET mutations occur more frequently in smokers.” A broader application of NGS testing “may have identified targets that were amenable to experimental drugs,” Goldberg said.

A 2014 report from Montefiore Medical Center in the Bronx, New York, had similar findings of a sizable underuse of genomic testing of patients with NSCLC. From 2009-2013, of 1910 patients with NSCLC (874 with nonsquamous histologic types), the testing rates for EGFR and ALK were 62% and 23%, respectively.²

The pendulum may swing in the opposite direction given the November 2017 FDA approval for the F1CDx test, Goldberg said. Since 2014, Foundation Medicine has struggled to obtain payer coverage approvals for the broad-based panel test, although the FDA had previously approved diagnostics that are specific to a single drug. The F1CDx approval came with a simultaneous preliminary national coverage determination from CMS. Assuming the administrative review process goes smoothly, a payment code would be issued during the first quarter of 2018, and private payers would be expected to follow suit, providing coverage for the test, Goldberg said.

The FDA and CMS decisions regarding F1CDx signify that “the era of precision medicine is officially arriving in 2018,” he said. “We will start to see even more acceptance and more utilization of genomic profiling, now that it’s likely to be covered by Medicare, to help direct treatment decisions and treatment planning.” Foundation Medicine has obtained payment as high as $3416 for its F1CDx test, but in the third quarter of 2017, the average was $2600 for each test.

Although cost may be a factor in physicians not ordering the tests, the weight of clinical evidence that genomic testing can point out successful modes of treatment will inevitably uproot that objection, Goldberg predicted, noting the ongoing MATCH and TAPUR studies of the value of genomic profiling in matching patients with advanced therapies that are targeted to genomic disorders. “We’re going to see a lot more broad applications of genomic profiling across cancers, and that—combined with some of the advances that, hopefully, will come from the therapeutic side— will move us to a world of precision medicine,” he said.

With NGS, diagnosing and prescribing will become much more accurate, according to Goldberg. “The advantage of these tests is you don’t have to pick and choose which mutation to look for,” he said. “They give you very broad coverage. You may see genes that you didn’t think about, that might have experimental drugs or drugs that are commercially available that you might consider using off-label.”

It’s also likely that new tests and procedures will emerge to compensate for the shortfalls of current NGS testing, such as physicians’ difficulty in obtaining sufficient biospecimens. “Liquid solutions,” such as plasma-based cell-free circulating tumor DNA testing, will find their place in the new landscape, Goldberg said. Indeed, Goldberg and colleagues stressed the importance of such tests “to obviate the need for repeat invasive tissue biopsies when the initial samples have been exhausted by pathologic examination or to avoid repeating biopsies at progression.”

The large data sets emerging from broad-scale use of NGS are likely to spur biopharmaceutical companies to dig more deeply for new targeted therapies, Goldberg added.

Whereas clinical evidence and FDA acceptance will help encourage use of genomic testing, the public’s growing desire to be informed about healthcare issues will add to the pressure to adopt, Goldberg said. Case in point: the genomics and biotechnology company 23andMe of Mountainview, California. For a fee, the direct-to-consumer company will analyze DNA samples and inform customers about their predispositions to any of 10 diseases or conditions, from Parkinson disease to late-onset Alzheimer disease.

In April 2017 the company became the first direct-to-consumer company to offer genetic testing services. This followed an initial FDA moratorium on sales by 23andMe, during which the company was asked to prove that its tests were accurate, which it did. The company has tested more than 2 million saliva samples since 2007, and public demand for such services is so strong that it’s almost inevitable that consumer testing will expand to include multiple other risk factors, Goldberg predicts. It’s becoming clear that some genes are smoking guns for disease, and consumers will want to test themselves for those risk factors, he said. “The ones 23andMe [have already] included allow a consumer to look at the results and bring it up with a doctor and not be panicked, but I wouldn’t be surprised if in the foreseeable future some of those genes that are currently restricted to tests that doctors order, such as the BRCA breast cancer gene, would become part of a consumer test,” he said.

“The medical concern right now is, would a consumer, without any doctor available to talk about the findings, understand the results?” Goldberg asks. “That’s why they’re going slow with this, but that’s going to happen. We’re moving toward a society, at least in the United States, where consumers are very conscious of their health and want detailed information about their medical status. Patients are able to access their health information, they’re able to read doctor notes—what I’m writing about them—and read their lab results. I can’t believe it’s going to be very long before consumers are able to order their own genetic profiling. And 23andMe has started the process.”

References

1. Gutierrez ME, Choi K, Lanman RB, et al. Genomic profiling of advanced non—small cell lung cancer in community settings: gaps and opportunities. Clin Lung Cancer. 2017;18(6):651-659. doi: 10.1016/j.cllc.2017.04.004.
2. Inal C, Yilmaz E, Cheng H, et al. Effect of reflex testing by pathologists on molecular testing rates in lung cancer patients: experience from a community based academic center. Presented at: 2014 Annual Meeting of the American Society of Clinical Oncology, May 31, 2014; Chicago, IL. Abstract 8098.