How can biopharma use oncology real-world data to support clinical development? Q&A with Viraj Narayanan, VP of Life Sciences, COTA
In oncology, real-world data (RWD) has emerged as an important tool to support clinical research and decision-making around treatment approaches, particularly in rare tumor types or instances where traditional clinical trials are neither ethical nor feasible.
COTA Healthcare, Inc., a new Aetion collaborator, offers an RWD set with deep clinical oncology data to help life sciences partners understand how their treatments perform in the real world. Founded by oncologists, COTA also works with clinicians to ensure its data, which captures elements such as biomarkers, performance status, and unstructured data—from clinician notes, for example—is regulatory-grade and equipped to answer questions about treatment safety and effectiveness.
Together, the collaboration aims to enhance the transformation of RWD into real-world evidence (RWE). It enables streamlined contracting for users of the Aetion Evidence Platform® (AEP) and quicker access to the rich oncology data sets for RWE generation. COTA’s research-grade oncology RWD will be also available for demonstration on the AEP.
We sat down with Viraj Narayanan, M.B.A., Vice President of Life Sciences at COTA, to discuss how this data can support clinical development work in oncology, and the role external comparators can play to offer a virtual alternative to clinical trials. Viraj, a decision scientist by training, has worked in biotechnology research and advisory roles to life sciences and providers, and with his team works to leverage RWD and analytics to accelerate drug development.
Responses have been edited for clarity and length.
Q: How is COTA’s data set well positioned to enable observational research?
A: Our real-world data set includes data from a mix of academic and community sites, including a heavy concentration in tertiary referral sites. This mix is particularly important because patient profiles and treatment patterns can vary substantially across care settings, so when sponsors consider real-world data sets, it’s important that there is representation from both.
In addition to that, our data sets include not only structured data from electronic medical records, but unstructured data as well, which is where most of the clinical richness exists. We pull out relevant information from the doctors’ notes, lab reports, pathology reports, and other unstructured data, which is particularly useful for sponsors as they look to answer clinical questions about their products. Another differentiator is that our data is deep both clinically and longitudinally, because we continue to look at the patient journey over time.
Our focus area is the clinical development, pre-launch space. COTA’s data is well suited for those types of analyses because we have an adaptable data model, and we bring the oncologist’s perspective to the conversation. That is particularly important as we partner with sponsors on external comparators to a single-arm trial, for example. If the sponsor needs to identify specific data elements in the data set to match the clinical trial inclusion list, our adaptable data model allows us to find those additional elements and provide them back to the sponsor.
Providing the oncologist’s point of view is a key element, too. Our Chief Medical Officer is involved in data reviews and data walkthroughs with our sponsors to help answer any questions that may arise in the data. If sponsors stress test the data and find things that seem clinically improbable—patients receiving an unusual sequencing of drugs, for example—you need the clinician’s point of view to explain what happened.
Q: How have you seen biopharma leverage RWE for clinical development?
A: We’ve seen the most RWE adoption and interest from sponsors around virtual control arms and external comparators, which are most typically found in phase II trials in oncology with the potential for accelerated approvals. The sponsor will decide that, because it will be difficult to enroll patients in a traditional control arm for a rare condition, and the product is showing efficacy, a virtual approach makes the most sense.
We’ve also seen some interest in looking at off-label use in real-world data to explore label expansion, as well as, in the post-launch setting, using real-world data to inform label updates. And because the FDA is so forward thinking on leveraging alternatives to clinical trials, they’re assessing ways to integrate real-world evidence into labels, too.
Q: What are some challenges you’ve seen sponsors face as they adopt RWE, and how do you help solve for them?
A: I’d say there are three layers of challenges we see sponsors face: challenges with data, the organizational or cultural appetite for RWE, and the process of establishing an overarching strategy for using real-world evidence.
Within the data component, we see challenges with sample size (“we don’t have enough data to match the control arm”), data harmonization (“how do we know that we’re measuring endpoints in the same way across data sets?”), and assessing data fitness for purpose (“is this the right fit-for-purpose data we need to answer our question?”). In clinical development, COTA is especially equipped to provide fit-for-purpose data because of our deep clinical data—claims data is less useful in this case.
When it comes to cultural appetite, there’s certainly an adoption curve, and some organizations are more innovative than others. Today, we see a lot of disparity in organizations that are ready to adopt real-world evidence in the clinical development setting. The main challenge they face is around positioning RWE as a different way of meeting a goal to help augment and accelerate insights, rather than as a threat to the status quo. On top of that, there’s the associated challenge of developing the set of capabilities needed for RWE—a data science capability, an epidemiology capability, a medical capability, for example.
And then from a strategy perspective, there are important questions to answer: When do you use real-world evidence, and at what phase of development? Do you have advocacy at the right levels of the organization? There has to be an overarching strategy and someone responsible for it.
Our most successful partnerships with life sciences companies are open and collaborative. We typically have senior sponsorship at a very high level in the organization, and we learn together and transparently. These challenges are difficult to solve, and neither of us will be able to solve them alone, so we must share back and forth. That’s the model we try to build towards in our biopharma partnerships.
Q: How have you seen COVID-19 impact data collection and research initiatives?
A: Everything is impacted by COVID-19, unfortunately. A lot of the sponsors that we’ve talked to are having earlier stage trials slowed down or shut down to ensure patient safety. The implication of this is that program timelines will be extended.
As we slow down trials, we could speed things up with real-world data—it’s digital, and doesn’t require someone going to an office or a trial site. How do you use that digital patient experience to ensure that you’ve got the right patient population, study design, and external comparator? It could be a pivotal moment for sponsors to leverage real-world evidence to speed up development timelines when the system returns to normal.
Q: What are the most exciting opportunities for RWE today? And how do you see that impacting the future of clinical development and regulatory applications?
A: As we discussed, we think external comparators are the biggest area of opportunity today. In terms of the future, there are two areas where real-world data is underutilized.
One is in phase I, where sponsors can use real-world evidence to inform population selection and trial design, including what hazard ratios you should consider for your trial. Historically, program teams have used historical clinical trial data to make these assessments, but often that data is too old. With real-world evidence, they’d be able to assess how similar patients to their target population perform today, which can help validate their hypothesis on the patient population of interest. It can also help them understand how much better their drug has to perform than the previous benchmark.
The second area is the post-approval space. Precision medicine is no longer a buzzword; it’s happening, and in cancer it’s even further accentuated—one-third of the drug pipeline is aimed at targeted mutations. These patient populations are super small, and approvals are going to be based on small populations. But then the regulators, payers, and patients have to ask an important question: Are the patients in the real world truly like the ones in the trial? Or, do we need to use real-world data to generalize a population that is more representative of actual patient populations? Once we’ve approved a drug based on 100 patients, what will performance look like in 1,000 patients? What’s their safety and efficacy to that treatment?
I think payers are going to be asking those questions because of the cost of cancer drugs. I think regulators are going to be asking those questions because they want to make sure the drug they approved is actually working. And I think patients are going to be asking those questions as they assess, for themselves and their loved ones, whether to take a chance on a new product. Real-world evidence has a huge opportunity to prove value to each of those groups.