NEW YORK – Even though the COVID-19 pandemic significantly reduced new cancer diagnoses and challenged clinical trial operations globally in 2020, a record number of new precision oncology drugs came to market, the US Food and Drug Administration approved the first next-generation sequencing liquid biopsy tests, and the field became more comfortable with telehealth and decentralized studies.
By Precision Oncology News' count, last year, the FDA approved 27 precision oncology drugs, including 10 new molecular entities (NMEs) and 17 new indications of previously approved drugs (see list here). This marks a 35 percent increase in the number of approvals compared to 2019, when the agency approved 20 NMEs and expanded indications.
Notably, the FDA for the first time approved drugs for non-small cell lung and thyroid cancers with RET alterations, selpercatinib (Eli Lilly's Retevmo) and pralsetinib (Blueprint Medicines/Genentech's Gavreto). A new autologous CAR T-cell therapy, brexucabtagene autoleucel (Gilead/Kite Pharma's Tecartus), came to market for refractory mantle cell lymphoma. PARP inhibitors, such as rucaparib (Clovis Oncology's Rubraca) and olaparib (AstraZeneca's Lynparza), became treatment options for hard-to-treat metastatic, castration-resistant prostate cancer patients with mutations in BRCA1/2 or other homologous recombination repair genes.
Drugmakers continued to make headway researching treatments for cancers driven by KRAS mutations, which for decades was thought of as an "undruggable" target. Amgen pulled ahead of the pack with a regulatory filing with the FDA and European Medicines Agency in December for sotorasib in previously treated, advanced or metastatic, KRAS G12C-mutated NSCLC.
In parallel, there is greater recognition among oncologists of the value of genomic testing cancer patients, according to Razelle Kurzrock, director of the Center for Personalized Cancer Therapy at the University of California, San Diego. Previously, given the rarity of many tumor markers such as NTRK fusions, which occur in less than 1 percent of cancer patients, oncologists would question whether it was worth it to profile patients' tumors using large next-generation sequencing panels that can cost several thousands of dollars.
"That argument is going away … because it's not just NTRK fusions that we have now," she reflected. "We have RET, BRAF, tumor mutational burden, microsatellite instability, EGFR, ROS1, and I could go on and on … Each alteration represents a very small subgroup, but when we put them all together, they become a larger group."
The value of genomic profiling is particularly apparent in tumor types like lung cancer where she estimated that up to 40 percent of patients may have a biomarker that can be targeted by available therapy. But there are still tumor types, Kurzrock recognized, such as glioblastoma and pancreatic cancer, where the benefits of precision oncology haven't been demonstrated and more research is needed to identify targetable biomarkers. But that's giving the field reasons to do more genomic testing, not less, according to Kurzrock and other precision oncology supporters. Although oncologists are still debating whether or not to genomically profile all cancer patients, over the past year, she believes it has become a lighter controversy.
The pharmaceutical industry similarly continues to embrace genomic profiling especially in the development of precision oncology therapies for patients with rare tumor alterations such as NRG1 fusions, which like NTRK fusions also occur in less than 1 percent of solid tumors. In order to identify patients with these alterations, Elevation Oncology, which is developing seribantumab for NRG1 fusion-positive solid tumors, has partnered with multiple genomic profiling companies to identify patients eligible for its trials, including Strata Oncology, Caris Life Sciences, Tempus, and Ashion Analytics. Merus, another company developing zenocutuzumab in this subpopulation, has partnered with Sema4 and Caris.
The emergence of drugs targeting NRG1 fusions "feels like the NTRK story again," said Jordan Clark, chief commercial officer at diagnostics data analytics company Diaceutics, referring to the development of larotrectinib (Bayer's Vitrakvi) and entrectinib (Genentech's Rozlytrek) for patients with refractory solid tumors driven by NTRK fusions. Although the drug trials in NRG1 fusion-positive cancers haven't yet reported out, if they follow a trajectory similar to NTRK inhibitors, then Clark is hopeful that these drugs may yield the same "spectacular" response rates in patients.
Still, Clark cautioned that there are lessons to be learned from the commercialization experience with NTRK inhibitors, which are still not reaching all the patients who might benefit due to the limited uptake of comprehensive genomic profiling across tumor types. While genomic testing is becoming standard practice in certain settings, such as lung cancer, "in other cancers, where we haven't had so much next-generation sequencing being conducted, we see patients still being missed," Clark said.
Meanwhile, there was significant acceleration in the early cancer detection liquid biopsy space in 2020 with Illumina's $8 billion acquisition of Grail, as well as Exact Sciences' $2.15 billion purchase of Thrive Earlier Detection and $410 million buyout of DNA methylation analysis company Base Genomics. "Cancer screening seems to have had a tipping point this [past] year," Clark said. While adoption of blood-based early cancer detection tests can yield big benefits for patients there are still significant questions around reimbursement and implementation policy that have yet to be ironed out, he reflected.
Greater use of these tools stands to transform the oncology therapy market by diagnosing cancers earlier when they are more curable, which in turn can push precision oncology in new directions. Genomically informed cancer drugs, typically reserved for patients who are heavily pretreated and have advanced disease, are becoming increasingly available to patients earlier in their disease journey. In December, AstraZeneca garnered FDA approval for osimertinib (Tagrisso) as the first adjuvant treatment for NSCLC patients with EGFR mutations.
However, with greater use of these screening tests, oncologists and researchers will have to grapple with how to treat patients with pre-cancer or early-stage disease. Will they do well with just surgery? Do they even need treatment? "Many pharmaceutical companies could look at this as a threat, rather than an opportunity," Clark said. "This is going to be a whole new area of precision medicine for the scientific community, for medical practices, and for reimbursement policy."
Lessons from a pandemic
In addition to following advances in precision oncology research, treatment, and testing, as a diagnostics data analytics firm Diaceutics is also tracking the impact of the COVID-19 pandemic on cancer diagnoses and related procedures. The company saw the biggest impact in lung cancer, where the number of newly diagnosed patients from February to March dropped by 31 percent. Molecular testing rates for KRAS, BRAF, and EGFR mutations also decreased in March.
Due to the pandemic, several of Kurzrock's patients had their curative surgeries delayed. "And now they're no longer curable," she said. In Precision Oncology News' annual survey of experts at 21 US cancer centers and hospitals, half the respondents said that diagnosis of new cancer patients dropped by 10 percent to 25 percent at their institutions due to pandemic restrictions.
Diaceutics' data shows that cancer diagnosis rates were still low compare to historical levels as of July and August. "We know that these patients have been lost to initial diagnosis," Diaceutics CEO Peter Keeling said. "Therefore, the question is when and how are they going to be present?" Going into 2021, Keeling sees a "perfect storm" brewing, where labs will struggle to diagnose cancer patients coming in with advanced stages of disease while still trying to ramp up COVID-19 testing. There will be test backlogs and longer wait times for results, and in this environment, it will be much harder to launch new diagnostics, Keeling predicts. (Diaceutics plans to release analysis on this in the coming weeks.)
The pandemic also brought to a halt many cancer clinical trials, which is significant for precision oncology since so many of the treatments patients receive based on genomic testing are investigational. "[Clinical trials] were really slowed down at the beginning" of the pandemic, Kurzrock said. "It's coming back, but it's still not as robust as it used to be."
Meanwhile, COVID-19 studies took off at breakneck speed, she noted, which showed that "when you really want to, it's possible" to launch and enroll clinical trials in a few weeks. "That may be an important lesson for diseases like cancer," Kurzrock said. "Why do we have to take a year to open up a trial?"
Some cancer trials continued despite pandemic restrictions thanks to the FDA and cancer centers quickly pivoting to allow remote and decentralized strategies. In a survey by the American Society of Clinical Oncology of 34 academic and community cancer research programs in late March, more than 90 percent of respondents said they reviewed patients' symptoms using remote strategies and nearly 88 percent said they engaged with patients via telemedicine whenever possible.
The success of these approaches has shored up support for more home-based and telehealth-supported trials even after the pandemic so terminally ill patients don't have to spend their last months at a cancer center, hundreds of miles from friends and family. Sameek Roychowdhury, a medical oncologist at the Ohio State University Comprehensive Cancer Center, is working on launching decentralized precision oncology trials using telehealth at OSU, and has gauged interest from cancer foundations, oncologists, and pharmaceutical companies in this regard. "It turns out we probably have a lot more red tape than we need to," he said. "This is something people have recognized for a while."
For example, it's common for cancer patients to travel hours to a cancer center to discuss clinical trial options, only to make the same trip again a few days later to sign consent forms in person in order to join a study. "This is just creating extra barriers and work for everyone, but doesn't help clinical research be accessible or pragmatic," said Roychowdhury. " A lot of these things telemedicine can overcome."
Kurzrock is principal investigator of the TRACK study, a decentralized trial the TargetCancer Foundation launched last year in collaboration with UCSD, MD Anderson Cancer Center, and Foundation Medicine to increase access to precision oncology in the community setting. The trial aims to enroll 400 patients with rare tumor types, who will have genomic tumor profiling and access to a virtual molecular tumor board that will evaluate the test results and recommend treatment approaches. These patients can also consent remotely and continue to receive care locally from their own physician.
When this study was planned two years ago, it was a novel idea for expanding precision oncology access to the community, but because of the pandemic a lot of institutions have made similar adjustments to research protocols. "There's a movement in the country to see if certain types of clinical trials can be done at home," Kurzrock said. "There will be many other trials that go in this direction."
Testing debates, biomarker cutoff squabbles
Even though the pandemic stalled research and made it difficult for cancer patients to get diagnosed and treated, the cancer community continued to debate aspects of precision oncology as new drugs and tests came to market. "We're past an inflection point this last year, and this debate as to whether precision oncology is worth it or valuable or has a true place in cancer treatment broadly is over," said Lee Schwartzberg, medical director of West Cancer Center and Research Institute and chief medical officer at OneOncology, a network of leading oncology practices in the US. "The remaining questions to be answered is timing of testing and in which patients and how best to effectively use the information … We don't want to just generate information and expense with testing unless it actually gets used."
The biggest controversy among oncologists in 2020 involved the FDA's approval of Merck's second tissue-agnostic indication for pembrolizumab using high tumor mutational burden (TMB) status as a predictive biomarker. Specifically, oncologists debated whether the 10 mutations/megabase cutoff for identifying patients with high TMB was arbitrary and would lead to overuse of the immunotherapy in those unlikely to benefit, or appropriate given the intended population is out of other treatment options.
As far as biomarkers are concerned, Kurzrock holds TMB in high esteem, as the "single most important biomarker" in oncologists' arsenal at the moment for identifying those who have a chance at experiencing an enduring benefit from immunotherapy. She has seen patients headed for hospice who received immunotherapy after testing showed they had high TMB and are now alive and disease free for as long as five years. "They're cured. I use the word cure very cautiously, but I don't know what else to call it," she said. "We never saw that before."
Though there is disagreement over the most appropriate TMB cutoff, she noted there is little dispute in the field that patients with lots of mutations in their tumors generally tend to respond better to immunotherapy than those with little or no mutations. "We're splitting hairs with that controversy," she said.
While Schwartzberg believes FDA was reasonable in approving pembrolizumab's tissue-agnostic indication based on TMB, he said the debate over the right cutoff for what constitutes high TMB isn't yet settled. "We need more research to make sure that the patients who we think might benefit are benefitting," he said. "The jury is still out a little bit there."
In Roychowdhury's view oncologists should think more generally about TMB as a biomarker that gives patients the chance to receive immunotherapy if they have a high mutation burden, but not as a biomarker to withhold treatment. He cited the example of interleukin-2, one of the first immunotherapies that emerged in the 1990s as an option for metastatic melanoma patients who tend to have some of the highest TMB, and for metastatic kidney cancer patients who tend to have very low or no tumor mutations. "That is a very simple example of why you shouldn't use TMB-low as a reason not to give somebody immunotherapy," he said. "But TMB-high can be a reason to make sure someone gets immunotherapy."
Beyond quibbling over cutoffs, Kuzrock acknowledged that immunotherapy is overutilized and said the field needs to do a better job of using biomarkers to give immunotherapy to those most likely to benefit. When Kurzrock and her colleagues reviewed the ongoing immunotherapy trials in 2019, they found that only 10 percent employed a biomarker strategy to identify best responders. "The mistakes of targeted therapy are being repeated with immunotherapy. We're giving immunotherapy indiscriminately to everybody," she said. "We know that just as everybody's [tumor is] different genomically, everybody's immune profile is different."
The other emerging consideration for oncologists is the extent to which germline testing should be expanded among cancer patients. Last year, MSK researchers presented data from a study involving 12,000 patients, in which 600 with recurrent or metastatic disease had actionable germline mutations, and 44 percent received targeted drugs either as part of the standard of care or in a study.
"From the MSK data and from our experience, we've seen enough patients that absolutely do not meet criteria for having a germline alteration, and yet, when they got tested, they had one," Kurzrock said. That information is not only important for managing cancer recurrence and guiding treatment in patients, but also for addressing cancer risk in their family members.
Kurzrock is of the view that every cancer patient should get germline testing. Although she acknowledged this is currently a very controversial view, in a few years she predicted that all patients will be getting germline testing.
In 2020, blood-based monitoring of minimal residual disease also gained more support. Medicare Administrative Contractor Palmetto GBA finalized local coverage for ctDNA tests, such as Natera's Signatera, for detecting minimal residual disease in colon cancer patients and guiding adjuvant treatment decisions after surgery and monitoring recurrence. This type of testing represents a big advance for precision oncology and utilization is poised to "explode in the next few years," Schwartzberg said.
Similarly, the momentum behind liquid biopsy as a companion diagnostic to identify best responders to treatment also increased with the FDA's approval of the first two NGS-based liquid biopsy companion diagnostics from Guardant Health and Foundation Medicine. The faster turnaround time for liquid biopsy tests compared to tissue tests can help overcome one of the main reasons cancer patients often don't receive precision oncology drugs based on genomic profiling.
Roychowdhury cited a study from Japan, in which researchers reported significantly higher clinical trial enrollment rates among advanced gastrointestinal cancer patients who had their tumors profiled by circulating tumor DNA testing, compared to those who had tissue-based NGS analysis. According to the paper published in Nature Medicine in October, the time from enrollment to receiving a sample was four days for patients who had ctDNA analysis compared to 14 days for those getting tissue testing; it took seven days versus 19 days for the results to come back from blood- and tissue-based NGS, respectively.
"The testing itself is really the same, next-generation sequencing, whether you analyze tissue or cell-free DNA," said Roychowdhury. "That's not the problem. It's scheduling the new biopsy of tumor tissue that adds three weeks."
Importantly, nearly 10 percent of patients in the study who received ctDNA testing were able to enroll in a drug trial based on detected genomic markers, compared to 4 percent of patients who had their tumor tissue sequenced. Objective response rates and median progression-free survival was similar between these two groups.
This is a significant difference in terms of patient enrollment, in Roychowdhury's view, since advanced cancer patients often forgo the opportunity to enroll in precision oncology trials because they don't want to go without therapy while they wait for tissue-based NGS test results. Advances in ctDNA testing "could really shake up clinical trial accrual and get patients on novel therapies," he said.
There are also disadvantages to these tests. Roychowdhury, who researches drugs targeting FGFR fusions, noted that liquid biopsy tests aren't as accurate as he would like for patients with low tumor burdens and for detecting genomic fusions. As the adoption of liquid biopsy increases, oncologists need to be aware of these limitations, he said.
Improving access, proving value
While the pandemic may have normalized telehealth and decentralized research protocols, which can in turn ease precision oncology access to cancer patients in the community, it provided no solutions for other enduring challenges, like cost and reimbursement. Although access to genomic profiling is increasing with greater use of commercial NGS providers such as Tempus, Caris, and Foundation Medicine, NGS is still not being done as much as it should be in community practices, observed Diaceutics' Clark. "While you may be going to the big labs to get that [testing], that's only a small fraction of the patients that are seen by community oncologists due to challenges in reimbursement and a lack of awareness about the need for NGS," he said.
Another problem is that treatments identified for patients based on genomic profiling often end up being off-label indications. While giving drugs in settings outside of FDA-approved labeling is not an uncommon practice in cancer care, occurring in around 40 percent of cases, Kurzrock has found little logic behind when an insurer will or won't cover an off-label treatment.
Centers of excellence like Moores Cancer Center at UCSD, with significant clinical trials operations, are successful at getting patients access to off-label treatments. For example, within the I-PREDICT study, UCSD researchers were particularly successful in treating patients based on their genomic profile partly due to support from medication acquisition specialists responsible for procuring the drugs patients matched to. "This doesn't solve the issue for patients that are just anywhere in the country," Kurzrock said. "And I would say that getting access to the drug is probably the single biggest barrier for precision oncology."
Precision oncology drugs and the associated genomic profiling also tend to be expensive, and as such there are still those who say the field has yet to prove the value of these strategies in terms of cost-benefit. The FDA has fielded its share of criticism for approving drugs for rare, molecularly defined patient populations based on small, single-arm studies instead of randomized trials, and endpoints, such as response rates, instead of waiting to see if these therapies are extending patients' lives.
Mark Linthicum, director of scientific communications at the Innovation and Value Initiative, a non-profit that evaluates the methods used to ascribe value to healthcare interventions, said that precision oncology certainly presents opportunities for risk-sharing arrangements between drug developers and payors to demonstrate the value of these therapies that are often approved based on incomplete evidence. Whiles these types of deals have cropped up for new cardiology drugs, they haven't yet occurred in cancer, he suspected, because such arrangements may be complicated by the need to factor in the interests of oncologists, who are reimbursed for infusion drugs. Moreover, since precision oncology drugs tend to be for smaller populations and not as widely used as cardiology drugs, "there may just be less appetite for [risk sharing]," Linthicum said.
The Trump administration introduced plans last year to lower drug pricing with much fanfare, though these rules are unlikely to have much impact, according to market observers. However, given the growing interest in the topic, Linthicum predicted that the incoming Biden administration will at least entertain discussions of the possibility of a centralized value assessment body, like the National Institute for Health and Care Excellence in the UK.
"We're likely to see a pretty significant change in momentum in the next couple of years with the change of administration … [and] the possibility of a centralized health technology assessment will become more a part of the conversation," he said, pointing out that even though precision oncology drugs are expensive, people tend to believe these treatments have value, and limiting access to them based on cost has been controversial even in the UK.
From Roychowdhury's perspective, the growing list of targetable tumor markers and approved therapies are chipping away at arguments that precision oncology lacks value. "Just because there are naysayers doesn't mean there aren't people benefiting," he said. He recently came across an NIH grant application from 2011, in which he proposed to sequence the tumor biopsies of metastatic cancer patients in a study and give them therapies based on the biomarker results. The reviewers denied that grant, reasoning that cancer is too heterogenous and that Roychowdhury's approach would be a waste of time.
"And what are we doing today?" he posited, noting that with the continued adoption of genomic testing and molecularly informed drugs as part of the standard of care, fewer oncologists and researchers entering the field today have a naysayer attitude toward precision oncology. "But in 2011, people were laughing at me for writing grants proposing we do sequencing on metastatic cancer patients to choose therapies."