Laurel: So when it comes to the pandemic, it really shows us how important and worrying the race to get new treatments and vaccines to patients is. Can you explain what evidence generation is and how it fits into drug development?
Anub: certainly. So, as a concept, generating evidence in drug development is nothing new. This is the art of bringing together data and analytics to successfully demonstrate the safety, efficacy and value of a product to many different stakeholders, regulators, payers, providers and ultimately, most importantly, patients. To date, I would say that evidence generation includes not only the trial readings themselves, but now pharmaceutical or medical device companies conduct different types of studies, which may be literature reviews or observational data studies or analyses showing disease burden or even treatment patterns. If you look at how most companies design, the clinical development team focuses on designing the protocol, executing the trial, and they are responsible for the successful readout of the trial. Much of the work occurs in clinical development. But as the drug nears the market, health economics, outcomes research, epidemiology teams are helping to paint a picture of what the value is and how can we better understand the disease?
So I think we’re at a very interesting inflection point in the industry right now. Generating evidence is a multi-year activity, both during a trial and, in many cases, long after it. We think this is especially true for vaccine trials, but also for oncology or other therapeutic areas. In the midst of the Covid-19 pandemic, vaccine companies have been putting together their evidence kits in record time, an incredible effort. Now I think what’s happening is that the FDA is finding a tricky balance, and they want to promote the innovations we’re talking about, advances in bringing new treatments to patients. They have established tools to accelerate treatment, such as accelerated approvals, but we need confirmatory trials or long-term follow-up to really understand the evidence and understand the safety and efficacy of these drugs. That’s why this concept we’re talking about today is so important, and how can we do it more quickly?
Laurel: That’s certainly important when you’re talking about life-saving innovation, but as you mentioned earlier, with the rapid pace of technological innovation and the combination of data being generated and reviewed, we’re at a particular inflection point here. So how has the generation of data and evidence evolved over the past few years, and how different might this ability to make a vaccine and all the evidence kits be 5 or 10 years ago?
Anub: It is important here to distinguish between clinical trial data and so-called real-world data. Randomised controlled trials are and remain the gold standard for evidence generation and submission. We know that in clinical trials, we have a very tightly controlled set of parameters and focus on a subset of patients. There is a lot of specificity and granularity in what is captured. There are regular assessment intervals, but we also know that the experimental setting is not necessarily representative of how patients will ultimately perform in the real world. And the word “real world” is a bit of a Wild West of a bunch of different things. It is an insurance company’s claims data or billing record. It’s electronic medical records from providers, hospital systems and laboratories, and even the increasingly newer forms of data you might see from devices and even patient-reported data. RWD, or real world data, is a large and diverse set of disparate sources that capture patient performance as they move in and out of different healthcare systems and environments.
Ten years ago, when I first worked in this field, the term “real world data” didn’t even exist. It’s like a dirty word, basically a word coined by pharmaceutical and regulatory authorities in recent years. So, I think another important part or dimension that we’re seeing now is that regulators have initiated and pushed through very important legislation like the 21st Century Cures Act, how real world data can be used and incorporated to enhance our understanding of treatments and understanding of disease. So, there’s a lot of momentum here. Actual data is used in 85%, 90% of FDA approved new drug applications. So, this is a world we have to navigate.
How do we maintain the rigor of clinical trials and tell the whole story, and then how do we bring in real-world data to complete that picture? This is something we’ve been focusing on for the past two years, and we even built a solution around it called Medidata Link during the COVID-19 pandemic, which actually links patient-level data in clinical trials with all non-trial data exists in the world for individual patients. As you can imagine, this makes a lot of sense during covid, we actually started with a covid vaccine manufacturer so we could look at long-term outcomes so we could relate the trial data to what we’re seeing after trial. Does a vaccine make sense in the long run? is it safe? Does it work? I think it’s something that’s on the horizon in terms of how we collect data and has been a big part of our development over the past few years.
Laurel: The stories of collecting data are certainly part of the challenge of generating this high-quality evidence. What other gaps in the industry do you see?
Anub: I think the big picture in the pharma industry development space is that despite all the advances in data and analytics, the likelihood of technological success or regulatory success as it requires drugs, going forward is still very low. The overall likelihood of Phase 1 approval is consistently below 10% for many different therapeutic areas. It’s under 5% in Cardiovascular and a little over 5% in Oncology and Neurology, and I think the root cause of these failures is the lack of data demonstrating efficacy. This is where many companies submit or include what regulators say are flawed study designs, inappropriate statistical endpoints, or in many cases, trials are underpowered, meaning the sample size is too small to reject the null hypothesis. So that means if you just look at the trials themselves and some of the gaps where the data should be more involved and have a bigger impact on decision making, you’re grappling with a lot of critical decisions.
So when you design a trial, you’re evaluating “What are my primary and secondary endpoints? What inclusion or exclusion criteria do I choose? What are my comparators? What do I use for the biomarkers? Then how do I Understand the outcome? How do I understand the mechanism of action?” It’s an array of countless different choices and different decisions that have to be made in parallel, all with data and information from the real world; we talked about the momentum where the value of electronic health records lies. But the gap here, the question is, how was the data collected? How do you verify where it came from? Can it be trusted?
So while the numbers are good, the gap actually contributes and there is a large opportunity for bias in a variety of different domains. Selection bias, which means there are differences in the types of patients you choose to receive treatment. There are many issues with performance bias, detection, and the data itself. So, I think what we’re trying to navigate here is, how do you do that in a robust way where you put these datasets together and address some of the key questions I mentioned earlier about drug failure? Our personal approach has been to take curated historical clinical trial datasets on our platform and use it to contextualize what we see in the real world and better understand patient responses to treatments. In theory, and what we’ve seen in our work, this should help clinical development teams use a novel approach to using the data to design trial protocols, or to improve some of the statistical analysis work they do.