NEW YORK (GenomeWeb) – When the choice is whether to prescribe patients experimental, targeted drugs based on genetic markers driving their tumors, or recommend immunotherapies, doctors are often going with the latter, a recent survey found.
A sampling of more than 100 physicians showed that more than two-thirds preferred prescribing patients a type of immunotherapy, called a checkpoint inhibitor, rather than direct them to an experimental targeted drug based on genomic testing.
OmniSeq, a company that a year ago launched a 144-gene next-generation sequencing cancer profiling assay, conducted the survey. Based on the findings and one-on-one conversations with oncologists in the community, executives at OmniSeq believe that doctors may be shifting away from targeted drugs using a precision testing approach in favor of recently approved checkpoint inhibitors.
There may be a number of reasons for this, including implementation challenges that keep genomic test results from being added to patients' medical charts. But OmniSeq Chief Commercial Officer Matt Klusas suspects that the firm's survey reflects the sobering reality that though some of these targeted drugs initially melt away tumors, patients' cancers almost always recur.
With immunotherapies, between 20 percent and 30 percent of cancer patients respond, and there are risks that a proportion of them will experience severe side effects. Still, there is the chance of durable responses and a shot at living longer with immunotherapies, Klusas pointed out.
Oncologists working at academic centers and healthcare systems that have invested heavily in NGS testing and precision oncology programs aren't surprised to hear that their colleagues around the country are eager to prescribe patients immunotherapies. "Patients are always asking me when I give them cytotoxic chemotherapy about what that will do to their immune system," said Terence Rhodes, an oncologist at Intermountain Healthcare’s Southwest Cancer Center in St. George, Utah. "When I tell them a particular therapy will work with their immune system to target and kill cancer cells, their eyes light up."
Intermountain Healthcare was an early adopter of NGS testing in cancer care and recently spun out a subsidiary, Navican, in an effort to bring genomic testing services to community oncologists who are having difficulty making sense of lengthy test reports and complex mutational profiles. NGS test results can take weeks to come back to the doctor, which can also hinder adoption of precision medicine strategies.
"If I say to a patient, 'I can give you this test,' even though it has a two week turnaround, or immunotherapy, for patients who already view immunotherapy as a more holistic approach, it just adds to the idea that that would be the best way to go," Rhodes reasoned.
In a recently published study, Intermountain researchers compared 36 patients who got targeted treatments within its precision oncology program and 36 historical controls who received standard chemotherapy or best supportive care, and reported that those who received targeted drugs guided by genomic test results had average progression-free survival of 22.9 weeks compared to 12 weeks in the control cohort.
Though the study suggested a survival advantage for those undergoing a precision oncology approach, skeptics have balked at drawing definitive conclusions from the analysis since it involved a small number of patients and wasn't a randomized-controlled trial, making it more prone to bias. But Rhodes said that while some oncologists may be biased against precision therapy based on studies that show that genomic testing benefits a minority of patients, "our bias is that it works based on our own studies."
Benefitting a few
A Medscape survey of 132 oncologists recently found (see related story) that the top two reasons they were ordering NGS testing were to guide treatment decisions (66 percent), and to direct patients to clinical trials (16 percent). Though 71 percent thought NGS testing was an important advance in cancer, 55 percent felt it was over-promoted since test results are currently helping only a small portion of their patients. Around 60 percent of respondents indicated that genomic testing benefits fewer than 25 percent of patients, and approximately half said that NGS testing resulted in fewer than 10 percent of patients being enrolled in a clinical trial.
"The results of the survey really portray the ambivalence about precision medicine," said Jack West, medical director of the thoracic oncology program at the Swedish Cancer Institute, who authored the survey in partnership with Medscape.
OmniSeq's survey similarly signals growing reticence on the part of oncologists toward investigational genomically guided targeted treatments. At first blush, when OmniSeq asked doctors if they altered therapeutic decisions after ordering its NGS test, 60 percent said they did. "But then, when we peeled back the onion further, we were really shocked and amazed," said OmniSeq CEO Mark Gardner.
Of those that changed their treatment decisions, 27 percent said they didn't choose a targeted drug for their patients (meaning they likely chose chemotherapy or the standard of care), 13 percent prescribed immunotherapy, 9 percent prescribed a targeted drug, and 8 percent directed patients to a precision medicine clinical trial.
The fact that 40 percent of docs decided on a strategy other than a targeted drug or a trial based on actionable markers identified by OmniSeq's NGS test came as a surprise to the company. OmniSeq claims that around 90 genes on its NGS panel can help guide patients to available clinical trials, but the test doesn't report predictive markers that would help identify best responders to checkpoint inhibitors, such as microsatellite instability, PD-L1 expression, or mutational burden.
In general, the surveyed doctors — employed at academic centers, community hospitals, or in private and group practices around the country — tended to choose a targeted drug if NGS testing found that patients fit the FDA-approved, molecularly defined indication for the treatment. But when they didn't, then doctors tended to keep patients on standard of care or choose immunotherapy. Many of the surveyed doctors indicated they were using NGS test results as a way to rule out targeted therapy due to contraindications.
When asked for their reasons for preferring immunotherapy, 85 percent cited prior success with checkpoint inhibitors both on and off label, while 70 percent blamed the lack of an available clinical trial, and just under 50 percent noted the time and effort it takes to match patients to a trial investigating a genomically guided approach.
"They're moving on," Gardner said, noting that doctors don't always have time to manage all the considerations for matching patients to a genotype-guided trial. Klusas worried that if physicians turn to alternatives like immunotherapy instead of directing patients to clinical trials, it will be even more difficult for companies like OmniSeq to generate the evidence showing that NGS testing is clinically useful for patients with rare cancer markers.
In the Medscape survey, doctors indicated they were mostly ordering genomic testing when standard treatments stopped working, and in the metastatic and research setting. However, they said that 73 percent of patients aren't willing to pay out of pocket for their testing, though 78 percent of surveyed oncologists felt that insurers should pay for this testing based on currently available evidence. But insurers are already unwilling to pay for many of the markers on multi-gene NGS panels without evidence from clinical studies, Klusas noted.
Although there are so-called basket studies attempting to speed up evaluation of drugs in rare, genomically defined cancer subpopulations, some of these trials, such as NCI-MATCH, have faced enrollment challenges. Community oncologists may not have ready access to these studies, and often lacking genomics expertise and resources, "may have greater difficulty drawing useful conclusions from the reports that come back with many pages of relatively obtuse results," West said. Similarly, patients may not have the resources or be in good enough health to travel long distances to partake in these trials.
At a community hospital-based cancer program in Rhode Island, for example, Kalyan Mantripragada and colleagues reported in the Journal of Clinical Oncology last year that only 9.2 percent of 164 patients who had sufficient samples for NGS testing had enrolled in a genotype-directed clinical trial. Among the reasons for the low accrual rate was that about a fifth of the patients had seen their health deteriorate between initiating NGS testing and receiving the results, and around half of the patients didn't match to a locally available trial.
Amid these challenges hindering enrollment of genotype-directed studies in cancer, there is growing concern among some experts that the hype around precision medicine is resulting in premature adoption of NGS. Major cancer centers in the US have all invested in NGS and are promoting their focus on precision medicine.
An advertisement about advanced genomic testing by Cancer Treatment Centers of America two years ago started by asking: "Shouldn't what makes each of us unique, also make our treatment unique?"
"There are no standard treatments, there is only the treatment that works for you," explains an ad from the Inova Center for Personalized Health. And a Memorial Sloan Kettering Cancer Center promotional video promises "more precise, personal, innovative care" that will change the way people think about the disease.
West has been warning his colleagues to check the promotional enthusiasm around precision oncology against data from well-designed studies. "I would be the first to shout about the success of precision medicine and broad genomic testing from the highest rooftops if we get evidence showing that it's beneficial to test a broad population of patients as the standard of care," West said. "There is a lot of concern that the hope exceeds the current reality."
He acknowledged that immunotherapy is also overpromoted. While there is a real basis for the optimism for it, "we're not at a point where the current utility meets the rabid level of hype," he reflected.
"It's important to recognize that the current standards of care, even if it is something as unglamorous as conventional chemotherapy, has become a standard of care because it improves survival in a way that's been established in hundreds to thousands of patients over time," he added.
The hype cycle
David Carbone, director of the James Thoracic Center at Ohio State University, has seen this kind of hype cycle with every new type of cancer therapy, particularly when the treatments are being as heavily advertised as immunotherapies are. Television ads for checkpoint inhibitors tell viewers that with these drugs they may have a chance to live longer. "Patients see this and think that's what they want," he said. "In spite of what the doctor recommends, that's the only thing they want."
He recently saw a patient with early-stage lung cancer who wanted immunotherapy, but for whom he recommended chemo and radiation. "I argued with her. I told her immunotherapy is not approved for early-stage lung cancer, and that there is no evidence that it has a cure rate that exceeds that of what I'm recommending," said Carbone. "She refused and just stormed out. The only thing she'd accept was immunotherapy."
Although this example didn't involve a choice between immunotherapy and a targeted drug, Carbone, a lung cancer specialist, said he is usually able to convince patients to go with his recommendation for a precision medicine approach, because there are a number of FDA-approved options in this setting. The National Comprehensive Cancer Network "strongly advises broader molecular profiling" to ensure that non-small cell lung cancer patients with rare driver mutations receive appropriate treatments and have the chance to enroll in studies.
OSU conducts NGS testing and can direct lung cancer patients into 80 locally available clinical trials — luxuries that many community oncologists don't have. Comparatively, patients with other types of cancers have fewer established and experimental options.
"Even when it works, it doesn't work well enough," Carbone said of genomically guided treatment approaches. He gets excited about the fact that lung cancer patients with ALK rearranged tumors can now realistically hope to live for two to four years without their cancer progressing, when a few years ago the average patient with metastatic disease lived for six months. But the reality is that these ALK-positive lung cancer patients tend to be in their 50s and cancer is still cutting their lives prematurely short.
"Compared to historical controls, it's great, but the fact is, it still stinks," Carbone said. "We have a lot of work to do to convert those responses into cures, convert months into years and into decades."
Immunotherapy, meanwhile, also benefits a minority of cancer patients, but some have experienced durable responses. For example, updated data from a study called KEYNOTE-010 showed that 90 percent of 47 patients who had received the checkpoint inhibitor Keytruda (pembrolizumab) for 24 months saw their tumors shrink, and 100 percent had a clinical benefit.
That chance to be one of those patients, however small, is what makes them eager to get on an immunotherapy. Comparatively, the option to get genetic testing and maybe get on a trial if one is available, sounds less hopeful.
"The final nail in the coffin for targeted therapies is that we know you're positively selecting for resistant genotypes," OmniSeq's Gardner said. "Because of tumor heterogeneity and positive selection, the overall survival for patients isn't substantially higher, though progression-free survival is. Because of that, there is a lot of willingness to move on to the next thing."
In Carbone's view, however, the answer isn't to move away from precision oncology strategies and genomic testing. To do so would risk missing that rare lung cancer patient with an alteration in ALK, ROS1, or another seldom seen marker, who could get on a targeted drug and gain a few more months or years with family and enjoy a better quality of life. As research advances, he believes the field will have a better understanding of which patients respond best to targeted agents and immunotherpies, and be able to more effectively combine agents.
Many experts in oncology feel there currently aren't good predictive markers for personalizing immunotherapies. In the OmniSeq survey, there was a 50/50 split between doctors prescribing checkpoint inhibitors on-label and off-label, and the majority were unaware of predictive testing to individualize checkpoint inhibitors.
The FDA has approved checkpoint inhibitors in certain cancer indications with companion and complementary diagnostics that assess PD-L1 expression. Experts disagree on the utility of PD-L1 expression to identify best responders to immunotherapies, since it appears that the marker cannot clearly differentiate responders from non-responders, but can identify who derives the most benefit from these agents.
Even though PD-L1 isn't predictive of treatment response in the same way an EGFR or ALK mutation is, precision medicine approaches are proving useful in research and patient care. Carbone routinely tests patients for PD-L1 expression and uses that information along with other data and NGS testing to prioritize treatment options for them. But he is looking forward to publishing data on newer predictive markers for immunotherapies, such as tumor mutational burden, that appear to be an improvement over PD-L1.
Foundation Medicine, which provides an NGS service that assesses more than 300 cancer genes, last year began quantifying tumor mutational burden as a way to predict which patients will respond to checkpoint inhibitors. The company recently presented initial data on this at a major cancer conference.
Meanwhile, OmniSeq is developing an assay that it says will provide a patient's immune profile using five different modalities: RNA-Seq, PD-L1 IHC testing, FISH to assess copy number variations of PD-L1/2, mutational burden, and microsatellite instability. "This assay will help clinicians identify specific immunotherapy strategies by interrogating T-cell receptor signaling, tumor infiltrating lymphocytes, and mutational burden together," said Klusas.
Gardner also highlighted that Roswell Park Cancer Institute, which OmniSeq spun out of in 2015, is studying the checkpoint inhibitor Opdivo (nivolumab) in combination with CIMAvax-EGF, an immunotherapy that blocks the epidermal growth factor protein from attaching to the receptor on cancer cells. In this study, researchers will track outcomes based on EGFR and PD-1 expression, and tumor mutations. As such, "the mission of right drug for right patient will still be with us," Gardner said.
"It's nearly a uniform belief even among those who are wary" that NGS testing "falls short of expectations now, that we're going to be living in a much more molecularly driven oncology world in the next several years," West agreed.
From Carbone's perspective, immunotherapy and genomically guided cancer care are going through the cycle of hype, disappointment, and realistic implementation that's all too common in oncology.
"There is an incredible enthusiasm and everybody thinks it works in everyone," he said. "Then, there is a phase where everyone sees the results and it doesn't work in everyone and they become incredibly disappointed with it. The third phase is where we figure out how to use it and in whom to use it, and it becomes a valuable cornerstone of therapy."