From the Journals

Center’s experience casts doubt on clinical utility of NGS



Next-generation sequencing (NGS) of tumor samples seldom changes patient management, and even when it does prompt off-label therapy, outcomes are usually poor, one center’s experience suggests.

“NGS has allowed more personalized medicine in oncology. It is well established that treatment of certain actionable mutations improves outcomes in many cancer types,” wrote Gregory J. Kubicek, MD, of MD Anderson Cancer Center at Cooper in Camden, N.J., and colleagues.

However, evidence of its utility to date has been mixed, and key trials – NCI-MPACT (National Cancer Institute Molecular Profiling–Based Assignment of Cancer Therapy) and NCI-MATCH (National Cancer Institute Molecular Analysis for Therapy Choice)—are still ongoing. “In the interim, the oncologist must make clinical decisions with limited empiric data but an exponentially increasing number of options,” they noted.

The investigators studied outcomes of the first 305 consecutive patients at their institution for whom tissue samples were sent to FoundationOne for NGS testing between March 2014 and April 2017. On average, the patients had received two lines of therapy, and the test was ordered 1.1 years from diagnosis of metastatic disease.

Study findings reported in the Journal of Oncology Practice showed that 116 of the tests were unusable because they did not yield a report (most often as a result of insufficient tissue) or yielded a report that could not be acted on owing to follow-up issues (patient loss of contact, transfer to hospice, or death).

Of the 189 potentially usable tests, 40.2% and 66.7% showed an aberration targetable by on-label therapies and off-label therapies, respectively. And fully 89.9% had actionable aberrations via all potential avenues, including clinical trials.

However, only 11.1% of the 189 potentially usable tests (and merely 8.3% of 253 completed tests and 6.9% of all 305 ordered tests) yielded a change in management, including use of on-label or off-label therapies, enrollment in clinical trials, or discontinuation of medications with a predicted poor response.

Of the six patients who were started on an off-label therapy, the median duration of treatment was 46 days, with half of these patients each stopping therapy because of death or because of progression.

“A vast majority of NGS assay results were not actively incorporated into clinical decision making, despite many assays indicating potential on- or off-label therapies,” Dr. Kubicek and coinvestigators wrote. “Given the escalating cost of medical care and scrutiny thereof, it is important to analyze whether tests are changing management and order tests appropriately.”

Several factors may explain the observed low use of NGS test results, they noted. For example, many patients were heavily pretreated, so some NGS-detected mutations would have already been known. Also, clinicians at the center had little experience with NGS testing.

“A variety of factors make precisely defining the utility of these assays in clinical decision making difficult, but we can certainly conclude that we have observed substantial costs with few discernible benefits,” the investigators stated. “It is possible that there will be greater use in the future as familiarity with these assays increases. Similarly, although we found poor outcomes with NGS-directed off-label therapies, we will eagerly await the results of NCI-MPACT and NCI-MATCH.”

Dr. Kubicek disclosed no relevant conflicts of interest. The study did not receive any specific funding.

SOURCE: Davis W et al. J Oncol Pract. 2019 Aug 2. doi: 10.1200/JOP.19.00269.

Next Article: