Bringing Oncology Treatment Down to Earth
October 18, 2021
Real-world evidence can accelerate access to life-saving medications – but as Dr. Parneet Cheema explains, it has to be the right kind.
As Medical Director of Oncology at William Osler Health System, Dr. Cheema leads the first-of-its-kind immunotherapy program at the institution. A medical oncologist with a worldwide reputation, Dr. Cheema is also an assistant professor at the University of Toronto’s Faculty of Medicine. Dr. Cheema and her team are currently recruiting subjects for the Pan-Canadian Lung Cancer Observational Study (PALEOS), a multicentre observational study that will collect data on patients with specific subtypes of lung cancer in Canada. Here, Dr. Cheema explains how we can use real-world evidence (RWE) to greater advantage.
You do a lot of work in lung cancer. Can you tell us how the medical understanding of this form of cancer has evolved?
We used to think of lung cancer as one disease, but we know about numerous subtypes. So we’re looking at many different diseases, some of them with orphan-type status, all under the non-small-cell lung cancer (NSCLC) umbrella.
What is standing in the way of value-based or outcomes-based agreements (OBAs) for lung cancer medications?
Historically, the evidence collected from databases hasn’t consistently panned out in clinical trials, which has led to a bit of skepticism about RWE. So first and foremost, we need to create the infrastructure to generate high-quality RWE that can supplement clinical trial data rather than just generate hypotheses.
Can you give us an example in which RWE and OBAs might have helped obtain a listing?
There is a targeted therapy for a form of NSCLC called BRAF V600E. It’s a combination of two medications, dabrafenib and trametinib. The pCODR expert review committee (pERC) initially recommended against listing it because of the limited evidence from clinical trials. But it’s impossible to conduct a large trial for such a small slice of the NSCLC pie—there simply aren’t enough patients. As it was, it took 14 months to enroll 59 patients from 9 countries in a phase 2 trial, which did show a benefit. pCODR also maintained there were other treatment options for NSCLC patients, essentially lumping this subtype together with several others. At the time, this decision was a big step back for precision medicine. The drug did eventually get a positive recommendation, but the lag time between NOC and listing exceeded three years. That's a long time for patients to wait.
“The data is with the patient. We need ways to get patients enrolled into a data collection pathway. And we need to keep it simple.”
To flip the question around, has RWE ever actually helped expedite a listing of a lung cancer treatment?
Yes. There is a medication called crizotinib that targets a lung cancer subtype called ROS1-positive NSCLC. It’s a rare subtype, with only 250 cases per year in Canada. A small phase 2 trial showed a clinical benefit. In this case, pERC considered not only the trial results but input from a group of clinicians. I was part of this group, and we submitted our observations that the medication had a durable response and improved patients’ quality of life. pERC went on to recommend a listing. Which begs the question: what constitutes good RWE? What can we actually submit? We still don’t have clarity on these questions.
There are so many avenues to getting RWE. Where do we start?
We should remember that “the data is with the patient.” The patients who come to my clinic, sitting in front of me—those are the patients who can give us good RWE. We also need RWE from patients treated in the community, as patients in academic centres don’t represent the entirety of the affected population.
You’ve explained the “where.” What about the “how”?
We need mechanisms to enroll patients in a data collection pathway. And we need to keep it simple. Right now it’s a bit of a mess for clinicians. We barely have time to write orders for our patients, so how are we going to find the time to collect all this extra data, not to mention data on historical controls? We need help with this, ideally from both industry and government.
So how would you advise industry and government to proceed with respect to RWE?
My plea to government and industry is this: please invest in long-term solutions as opposed to one-off RWE studies, which are costly, hard to run, and subject to bias. Invest in an infrastructure that includes clinical coordinators and mechanisms to collect prospective data, which is critical for clinician buy-in. Invest in databases with the capability to transfer data easily between sites and to integrate AI capabilities.
What can we do to help patients get on board?
We may need campaigns to communicate the value of participating in registries and databases and to address concerns about privacy. We need to make it easy for patients to provide consent and to include patients from all socioeconomic groups. In Ontario, we’re getting a jump on this with PALEOS.
Tell us more about PALEOS. What type of data will the study collect?
As I mentioned earlier, lung cancer has so many subtypes that we can’t get enough patients to conduct clinical trials with sufficient power. PALEOS is designed to fill this gap by generating real-world data on natural history, treatment patterns, and outcomes in relation to lung cancer subtypes, using both retrospective and prospective methods. To reflect the diversity of Canadian patients, we are recruiting from both academic and community cancer settings. We have funding for clinical coordinators and data analysis support. We will provide centralized education to ensure all sites are entering prospective data the same way, so we can generate standardized variables that can be used by health technology assessors.
“We should always keep sight of what we’re trying to accomplish with outcomes-based agreements: facilitating patients’ access to life-saving drugs.”
How can we get moving on using RWE to support OBAs?
We clinicians can’t produce RWE unless our patients can access the medication being evaluated. Industry can help with this, with the understanding that we provide RWE in return. From a regulatory perspective, it would make sense for pharma companies to include an RWE generation plan with their pCODR submission. The plan should address clinical uncertainty about a drug, so the data is strong enough to support an OBA. And we should always keep sight of what we’re trying to accomplish with OBAs: facilitating patients’ access to life-saving drugs.
Next-level testing44
Clinicians need to know the status of several gene mutations to optimally treat NSCLC patients, making timely biomarker testing a necessity. Dr. Cheema was part of an expert consensus group that convened in 2020 to create recommendations for biomarker testing. The panel recommended that all patients with nonsquamous NSCLC, regardless of stage, should undergo comprehensive reflex biomarker testing at diagnosis with targeted next-generation sequencing. “That’s a big jump from just testing for the EGFR mutation, which is what we used to do,” says Dr. Cheema. “Precision medicine keeps on getting more precise.”
References
44. Cheema PK et al. Consensus recommendations for optimizing biomarker testing to identify and treat advanced EGFR-mutated non-small-cell lung cancer. Curr Oncol 2020;27:321.