Skip to main content
Premium Trial:

Request an Annual Quote

Real-World Data from OneOncology, Roche Shed Light on NSCLC NGS Trends in Community Practices


NEW YORK – Biomarker testing rates have improved to some extent in recent years for non-small cell lung cancer patients treated in community cancer centers, but disparities in access and test turnaround time persist, according to new data from OneOncology and Genentech.

In two presentations during the American Society of Clinical Oncology (ASCO) Quality Care Symposium, Ari VanderWalde of Tennessee-based West Cancer Center and Clinic shared data from real-world evidence studies jointly sponsored by Genentech and OneOncology, a national network of 10 independent community oncology practices comprising some 500 providers across 175 care sites. The studies demonstrated an uptick in biomarker testing rates among advanced NSCLC patients treated at OneOncology-affiliated community cancer centers over recent years.

According to Lee Schwartzberg, OneOncology's chief medical officer and medical director of West Cancer Center and Research Clinic, OneOncology conducted the study to "get a quantitative assessment with real numbers" for a trend that has historically been tough to measure. While large, academic cancer centers may estimate testing rates within their own institutional networks, it can be difficult to pin down testing rates across disparate community care settings, where the majority of cancer patients in the US receive their care.

By partnering with Roche subsidiaries Genentech, Foundation Medicine, and Flatiron Health, OneOncology was able to identify testing trends within real-world data collected from its network practices.

In one study presented at the meeting, researchers looked at testing rates at four OneOncology sites — West Cancer Center, Tennessee Oncology, Eastern Connecticut Hematology and Oncology, and New York Cancer and Blood Specialists — and found that for all six therapeutically actionable NSCLC biomarkers of interest (EGFR, ALK, ROS1, PD-L1, KRAS, and BRAF) testing increased from 13 percent in 2015 to 57 percent in 2020.

In all, 83 percent of 3,860 advanced NSCLC patients treated at one of these OneOncology community cancer centers from 2015 to 2020, received testing for at least one of these biomarkers versus 17 percent who received no biomarker tests at all. When researchers considered the type of testing these patients received, they found 63 percent of patients had next-generation sequencing tests versus 37 percent who were tested for biomarkers in single genes or used older testing technologies, such as fluorescence in situ hybridization, immunohistochemistry, or polymerase chain reaction.  

The findings may not capture biomarker testing patterns in community oncology beyond OneOncology practices, since this is a group that makes a point of keeping network providers up to date on the latest biomarker testing guidelines and precision cancer drugs and holds molecular tumor boards to help providers interpret molecular test results.

That said, the second OneOncology-Genentech study presented during the ASCO Quality Care Symposium also included data from a larger cohort of 29,572 advanced NSCLC patients treated within the broader Flatiron Health network. Among these patients, who were treated at roughly 800 care sites across the country, mostly community practices, the NGS testing rates increased from 22 percent in 2015 to 60 percent in 2020, as in the OneOncology cohort.  

"These rates probably represent better than the average community oncology practice," said Schwartzberg. "It was very gratifying to see that the numbers went up over time, as one might expect, with the advent of additional biomarkers being identified that could be acted on in non-small cell lung cancer."  

Beyond the scope of OneOncology's study, past studies, including one conducted by researchers in the Molecularly Informed Lung Cancer Treatment in a Community Cancer Network, or MYLUNG, have indeed pointed to NSCLC biomarker testing rates in community settings that lag behind the OneOncology improvement. According to this MyLUNG data presented in June, only 37 percent of 3,500 metastatic NSCLC patients who received first-line treatment at community practices from 2018 to 2020 underwent NGS.  

Strategies for improved access

These two studies suggest that doctors at OneOncology practices are adopting NGS testing for NSCLC patients at a higher pace than physicians in the broader real-world dataset of 800 practices, Schwartzberg pointed out. The reason, he said, is that OneOncology has resources that support its physicians' ability to identify precision treatment opportunities for their patients.

For example, OneOncology has a precision medicine committee that includes at least one member from each practice. The committee meets on a regular basis to discuss operational and research issues surrounding personalized medicine and the ways that OneOncology can use its aggregated comprehensive genomic profiling data to answer questions pertaining to targeted drug use and biomarker actionability.

Additionally, OneOncology providers can tap into multiple databases of deidentified genomic profiling test results to answer research questions, including Foundation Medicine's database as well as its own large clinico-genomic database. The OneOncology clinico-genomic database is still in its early stages, but rapidly growing, according to Schwartzberg.

Perhaps most important, in Schwartzberg's eyes, are the educational resources shared throughout the OneOncology network. For example, OneOncology providers can attend tumor boards that meet to discuss the genomic and clinical features of several patients' cancers per month.

The network also holds what Schwartzberg called a "process tumor board," which meets on a weekly basis to review cases from across the network and triage the most pressing or most ambiguous cases. These tumor boards also bring in third-party experts to weigh in and notify providers if their patients have alterations that could make them eligible for a targeted drug's clinical trial.

Turnaround time falls short

Although Schwartzberg was proud to note an uptick in the percentage of NSCLC patients receiving biomarker tests, he noted there is still room for improvement with respect to the time it takes to identify patients for biomarker testing, collect tissue samples, send them out for analysis, and get the results back. According to the OneOncology-Genentech study, the median turnaround time from advanced NSCLC diagnosis to biomarker test result over the last five years was 20 days, and the median time from tissue specimen collection to receiving biomarker test results was 13 days.

Especially for patients diagnosed at an advanced stage, waiting multiple weeks to begin treatment can be detrimental. This could explain why many patients begin systemic treatment before learning if they have any actionable biomarkers.

In 2015, the percentage of patients starting therapy before getting biomarker test results was 29 percent, and while it dropped to 16 percent in 2017, that figure has not improved since 2017. The analysis showed chemotherapy was the most common treatment prescribed to patients who started treatment before receiving their biomarker testing results. Even after these patients received results and found out they had actionable biomarkers in ALK, ROS1, PD-L1, KRAS, or BRAF, fewer than half switched over to receiving a matched targeted treatment.

For patients who tested positive for EGFR mutations, 61 percent switched to biomarker-targeted treatment. "Further studies are warranted to better understand the reasons that patients received treatments that were not specific to their actionable mutations," VanderWalde said.

Test turnaround time "is an issue, and it's exaggerated in non-small cell lung cancer because many of these patients are pretty sick," Schwartzberg said. "There's enormous pressure on the physician, the patients, and their families to initiate therapy … and this is a bit harder in the community setting than it might be in an academic setting … where people tend to travel [for treatment] and are usually a little healthier."

This observation — that biomarker testing timeliness may pose a greater challenge in the community oncology setting than in the academic setting — mirrors a discrepancy highlighted in a recent LUNGevity and Association for Community Cancer Centers (ACCC) survey. Among community oncologists, the survey found 52 percent of respondents said they would order biomarker testing at the time of a patient's initial biopsy, versus 76 percent of academic oncologists who said they would do the same.

Treating patients with chemotherapy or chemotherapy and immunotherapy combinations before learning if they have actionable mutations that might make them eligible for first-line targeted therapy, in many cases, may affect the degree of benefit patients derive from those targeted therapies. For example, some targeted agents are approved specifically for patients who have not received prior treatment and certain clinical trials for biomarker-defined cancer treatments limit eligibility criteria to treatment-naïve patients. Even in cases where biomarker-targeted cancer treatments are indicated for later-line settings as well, the degree of benefit tends to narrow with each subsequent line of therapy.

"I personally don't like the idea of giving a cycle of platinum-based chemotherapy … or just giving everyone anti-PD-1 agents, then stopping after a couple of cycles and switching to a targeted therapy," Schwartzberg said. "That's not good medicine. I would like to move beyond the empiric treatment argument to personalized treatment for all patients."

While there isn't a clear solution to improving the turnaround time for biomarker test results, some precision medicine proponents have encouraged physicians to order testing when patients are first diagnosed with advanced disease and decreasing the percentage of patients who receive systemic therapy before getting their biomarker test results. Schwartzberg said that he and others in the broader OneOncology network have been discussing strategies, such as homing in on the best way to sequence liquid biopsies — which only require a blood draw — and tissue biopsies, and to make this sequence "work most efficiently without costing more."

"I don't think that the strategy should be that everybody should get a tissue and a liquid [test] at the same time … most payors won't pay for that, and it's not the most efficient and thoughtful way to use resources," he said. "But we do know that with sending out samples … the [process] is not always as smooth as it could be, so we've worked a lot on making those linkages better and emphasizing the need for speed."

Addressing insurance, socioeconomic status

Beyond the need to address test turnaround time, Schwartzberg highlighted an urgent need to address access disparities related to patients' insurance and socioeconomic status.

Within the OneOncology-Genetech studies presented at the symposium, VanderWalde said there was a trend in the data suggesting patients were more likely to receive NGS testing if they had commercial insurance. Although insurance data were missing for roughly a quarter of patients treated in the broader Flatiron Health network, he noted that in both the OneOncology and Flatiron cohorts a greater percentage of the patients who received NGS were commercially insured than were the percentage of patients receiving non-NGS biomarker tests.

"Is there still an inequity in offering un- or underinsured patients these tests?" Schwartzberg asked hypothetically. "There are so many more hoops to jump through to get these tests covered [for these patients], so that may be a part of it."

This data point regarding insurance status is particularly troubling when considered alongside another finding from the LUNGevity-ACCC survey. Among surveyed oncologists from both academic and community settings, a total of 57 percent of respondents said that they would be "extremely likely" to order biomarker testing for an advanced NSCLC patient with high socioeconomic status, versus only 32 percent who said they would do the same for patients with the same disease characteristics but with low socioeconomic status.

"There is a growing inequity problem with personalized medicine for people who are of lower socioeconomic status or from particular groups that are underserved," Schwartzberg said, sharing that he hopes to conduct additional research about these disparities using OneOncology's burgeoning clinico-genomic database while continuing to work on narrowing these gaps throughout the network.