This week I am talking to Laura Fernandes, PhD, Senior Statistical Director COTA (@cotahealthcare) who provide real-world data and analytics to help health systems expedite cancer research and standardize cancer patient care, and empowers life sciences companies to accelerate drug development and regulatory approvals for life-saving cancer drugs and treatments.
Laura comes from an international background that influenced her research into statistical methods to improve cancer research. We talk about the work she is doing in developing real world data (RWD) and what that is and how it relates to clinical trials and the gold standard in medicine and science the Randomized Clincal Trial (RCT). Her work has allowed the application of RWD to cancer care and allowed for new uses and updated labelling by the FDA for drugs in oncology care. We discuss the evolution of clinical trials, and insights on how organizations are tackling disparities in cancer care and research and some of the broader health data challenges facing healthcare and medical researchers
We discuss how the Cota data was used to review the effectiveness of Hydroxychloroquine for the treatment of COVID19 and how their data was used to demonstrate clearly that it has no place in for treating COVID19. We discuss labels in disease and in particular cancer and how that might change as our understanding grows
Listen in to hear how this RWD is being used to accelerate our understanding of drugs in pediatric patients where the ability to test drugs is much harder leading to delays in availability of treatments and options for children and infants.
Listen live at 4:00 AM, 12:00 Noon or 8:00 PM ET, Monday through Friday for the next week at HealthcareNOW Radio. After that, you can listen on demand (See podcast information below.) Join the conversation on Twitter at #TheIncrementalist.
Listen along on HealthcareNowRadio or on SoundCloud
Raw Transcript
Today, I’m delighted to be joined by Laura Fernandez. She is the Senior statistical director at Kota. Laura. Thanks for joining me today.
Laura Fernandes
Hi, Nick, thank you so much for having me over. I’m so excited to be here today.
Nick van Terheyden
So as I do with all my guests, tell us a little bit about your background, you You and I share some international heritage with perhaps a little bit of confusion and differences to the average bear. Tell us how you got here. And you know, a little bit about your your past.
Laura Fernandes
Definitely. So once again, thank you for having me over i My I’m currently working as a senior statistical Director for Research at Kota. Prior to that I was at the University of Michigan, where I obtained my PhD in biostatistics in and my thesis topic focused more mainly on clinical trial designs in phase one setting. And so the, the question that I’ve often asked is, so tell me more about your name, right. And so my name is Laura Fernandez. And yes, that is my given name. And that is because I was born and brought up in Goa, which is a small state in India that was colonized by the Portuguese. Unlike the rest of India, that was a British colony, which kind of explains my name, Laura Fernandez, because we had a lot of Jesuit missionaries who came in and did a lot of conversions to Christianity. And that’s how I got my name from my grandmother. And my name is spelled with the f and s, unlike the Hispanic spelling, which is usually H and Z Hernandez, but minus bananas. So like the way I said, I opened my PhD, I came to the US for further studies. And I studied at University of Michigan in Ann Arbor, my thesis topic focused on clinical trial designs in oncology. In the phase one setting, after my graduation, I worked at the FDA for about seven years, in the space of drug approvals in oncology and hematology. And so if you look at my career trajectory, in some sense, it is rich with the training and statistics and in oncology, which are, which are some areas that are very close to my heart. And I, when I was looking at areas to pursue after my work at the FDA, I looked at corta, which is pioneering the work in real world data in oncology. And so that is how I got here, right now,
Nick van Terheyden
I always love when people have these sort of different names that, you know, it makes you memorable. I mean, one of the things that happens as a result of that, and that’s good when you want to be remembered, it was terrible for me when I had problems at school, and I didn’t want to be remembered, I wanted to face into the background. But, you know, it’s delightful to sort of hear some of those stories. So. So you’ve clearly focused on this sort of data element, you had a whole set of research your thesis, you’re now working at Kota, and you bring up a topic that, you know, is prevalent in a lot of the conversation currently, as a result of the pandemic, I think people have become more aware of statistics, data trials, and so forth. And you bring up something that is extremely important, and something that is essential that people understand, but I don’t think they do so tell us what you mean by real world data and how that compares to other types of data that are used in trials and the development of drugs and treatments.
Laura Fernandes
Yes, so you’re you’re absolutely right, due to the pandemic, the use of real world data has become kind of mainstream, you know, everybody talks about real world data. And then the obvious question is, what is the meaning of real world, right? And how is that different? And so basically, the idea of real world data came up to differentiate it from a what we call data that’s collected in a clinical trial. And so clinical trials are in in our basic, in essence, experiments done using human beings, you know, human patients, and in a typical randomized clinical trial, you would have an internet interventional therapy or a new therapy that is compared for safety and effectiveness to a standard of care or your control therapy. So, you have this controlled experiment with human beings and data is collected at predefined time points and so You have this rich source of data. So in contrast, real world data can also be called observational data that is routinely collected, say, for example, in your clinical practice, when you go to see your doctor, the data that your doctor collects using electronic health records, electronic medical records, or when you have your insurance claim, or your prescription is filled out, all this data that is collected in the routine clinical setting is what is called observational data. There are other sources also, for example, you know, specific registry data’s or hospital registries. So what we see is that in clinical trials, it is hard to do to answer some basic questions because of enrollment or because we do not have enough patients sometimes, you know, to answer basic questions. And so then we try to look out for other alternative sources to answer these questions that are becoming harder, especially in the case of oncology, where, you know, we are studying very specific genetic mutations. And it’s because as you can imagine, it’s becoming harder to enroll patients into these large randomized clinical trials. And so we’re trying to look out for alternative data sources, and this is where real world data comes with the picture.
Nick van Terheyden
So extremely helpful, you know, the concept of observational and all of this additional information that’s starting to contribute to our understanding of, you know, what’s happening in our world. And, you know, obviously, we’re gathering more and more of that data. But there’s been a number of examples of that in use that I think will help sort of center an understanding and, you know, some of this real world data has actually been used to assess the effectiveness of treatments or potential treatments. Tell us a little bit about that.
Laura Fernandes
Yes, so So let’s, let’s stick with an example of cancer because that is what we do at at Kota. Basically, what happens is, now, when you study lung cancer, for example, it is not just a lung cancer, right? It is now lung cancer that is met Mattawan with madman mutations, or Ross one mutations or EGFR mutated. And so your cancer subpopulations are becoming smaller and smaller. And because of these smaller, well defined patient populations, it is becoming harder to study a new drug in a randomized clinical trial. And because of that we have we are seeing a lot of single arm clinical trials, you know, and so it’s becoming harder to get this to get this answer of how well does my drug do in comparison to an existing drug? And so to answer such questions, we need what is known as a control arm. And instead of assigning patients, you know, to the same standard of care in a clinical trial, you could recycle the data that was, you know, collected in, in a real world setting to fulfill this requirement. And so this is where the augmenting of data from an observational data setting into a clinical trial works out very well, you know, and so there are, I feel, there are like two or three different examples where real world data has been used by the FDA, which have led to labeling changes, quota has been instrumental in one of those approvals, where quota data was also used in a real world setting, you know, for a clinical trial approval. So, so yes, we are having examples. And the FDA is acceptance to the use available data is also increasing.
Nick van Terheyden
So I just for the benefit of the listeners, you know, let’s briefly talk about labeling. Labeling is the process that says this drug for this purpose, you know, fairly defined consequences. When you talk about changing labeling, it’s, you know, opening it up and I’ll use an example I imagine, most people will understand Botox, Botox is not traditionally labeled for use in reduction of lines, which most people would know it for. It is used for migraines, or migraines. And you know, when people talk about off label, then that’s the change. But this would be data that essentially says, Oh, now we’ve we can show that it worked for plastic surgery. And I’m not saying that’s a real example. I’m just trying to explain it for the listeners. But you actually have been essential in the pandemic, to informing some of the insights that I think were essential very early on, there was a number of therapies that we looked at that, you know, there was Certainly some initial excitement and lots of interest, in fact, a number of trials. Tell us a little bit about that. And hydroxychloroquine? Oh, yes,
Laura Fernandes
definitely. So there have been like several examples. And this is the reason why real world data has become so popular or mainstream during the pandemic, is because we have used real world data to inform our health care policy decisions. And so many of our mandates were dictated by the use of real world data. hydroxychloroquine is a perfect example where we should and in fact, it was done using port as data. So Porter was responsible in curating this large data set that helped answer this question. And again, it was observational data. And it showed that using hydroxychloroquine did not provide any benefit in terms of reducing all case mortality for COVID patients. In the same observational trial, we also looked at the use of Tasi, loser Mab and also the use of azithromycin, you know, whether in or without hydroxychloroquine. And so these are perfect examples of where, where our understanding of what works or does not work, or was guided by the use of real world data.
Nick van Terheyden
So for those of you just joining, I’m Dr. Nick the incrementalist and today I’m talking with Laura Fernandez, She is the Senior statistical director at Kota, we were just talking about real world data, the application of its use in the pandemic, which informed a very clear understanding, let’s you know, I think anybody that follows me online knows that I am categorical about this, there is no value to taking hydroxychloroquine own for the treatment of COVID 19. Based on the data and based on all of the research studies, and Kotor essentially contributed to that. You’ve talked Laura a little bit about this being observational data. And the specific, you know, this is, as we move to the precision medicine, you know, lower numbers of patients, there’s a couple of things that sort of occur, or, you know, strike me about this. The first is, you know, randomized controlled studies are the gold standard for any judgment and assessment of treatments. As we move forward, we’re sort of narrowing the spectrum of these diseases. Do you think the randomized control study is something that we’ll look back on and say, well, it was interesting, it was useful, but it’s no longer applicable? Or is this something that coexists with real world data?
Laura Fernandes
So you are absolutely right in saying that randomized clinical trials are the gold standard and establishing the safety and effectiveness offer of a new therapy in the market? You know, so there is no, no, no arguing with that. But there has, but we all also agree that randomized clinical trials, they do provide internal validity and internal validity implies that we know how well drug works in this well defined patient population. But what they lack is external validity. Wherein when the drug is given to patients who were not necessarily studied in the clinical trial, you know, so when it’s given to the general population, we do not know if those results are if they have general if they’re generalizable. And so this is what we struggle with with randomized clinical trials is that we do not have a certain age demographics that engenders certain race categories, also not studied in clinical trials, which kind of limits our understanding of how the drug will work, when it is given to a patient like me, you know, I was involved in lots of disparities work when I was at the FDA in clinical trials. And that is one of the areas that we struggled is that we had all these different subgroups of patients demographic subgroups, and we had no clue how the drug worked in those particular subgroups, not because they were meant to be studied, but the trial was powered to be for for showing a treatment benefit. But it it was not, that was not the objective to study, you know, in a large in a generalizable population. And that is where I can see a real world data actually helping clinical trials where when you have trials that are easily accessible, where patients can be easily enrolled, the use of decentralized clinical trials I have the there’s been news where you know, certain pharmacy centers like Walgreens And CBSs could be potential clinical trial sites in the future where patients can easily enroll. And this would increase our our confidence in the results that we see in clinical trials.
Nick van Terheyden
So I think it’s an augmentation of the gold standard of randomized controlled studies. It builds on that information, but you bring up another interesting point. And, you know, it’s always hard to know how well understood this is outside of, you know, the folks that really focus on this. And that is the disparity of the source of data that is used. And by source data, as you described it earlier on human trials, we test out drugs on humans, ultimately, that’s the only way that we can go through this process. And, you know, we’re grateful for people that volunteer and, you know, we’ve learned over the course of time to improve the regulatory surroundings. But what we fail to do, certainly earlier on, was to make that a representative population to test out on in the very early days, I don’t know that this is entirely true, but it feels like it was, it was almost all men, if it wasn’t exclusively men. So we excluded 50% of the population, deliberately, and I don’t think it was, you know, to be clear, this was not an intent to say we want to make this wrong or, you know, ineffective, the other reasons for it, we’ve started to include women, but now we’re sort of thinking about ethnicity. There’s all sorts of variations that we understand, you know, and it sounds like Kota and the data, and this real world evidence is going to help both look at that, and then start to address it. Tell us a little bit about that.
Laura Fernandes
Yes, so. So what is unique about Kota is that we, we partner. So our objective is to bring clarity to cancer. And we partner with both academic and community centers, cancer centers, which means that we have a very good representation of patient population in our database, we have over 1.7 million individual patient records. And we have records in like various hematology over 20, different hematology and oncology types. And so using this rich database, maybe I should take a step back. So in addition, what Kota does is, I highlighted earlier what was electronic health records, right, and so, electronic health records are have this structured component, and also this unstructured component. So there is this all these notes that are written by the doctor, you know, in this so that was what we would say, is the unstructured component. What Kota does is converts all the structured and unstructured components into longitudinal data for for an individual patient’s journey. So we could now tell looking at a patient’s individual patient, when he or she was diagnosed with a particular cancer, what were the initial baseline characteristics? How long did it take a particular therapy? You know, what were the outcomes of that first therapy? And then what were the subsequent therapies? And then eventually, is the patient still living or dying, you know, so we get the entire patient journey for a particular patient, or for a particular cancer type. And so there is rich information in this data source, right? And so we can mine it to answer so many questions, we give it back to our academic partners, our academic and community partners, and they use it to answer so many questions about the effectiveness of the therapies that the doctors prescribe to their patients, you know, so you can you can mine it for different to test different hypothesis to answer different questions. And this is one way that Kota is, is helping in the US available data, and also in the US in clinical trials. You know, so we highlighted this earlier, how we could use this real world data to augment clinical trial data. And so that is where quota also plays a role.
Nick van Terheyden
So as you think about the future, clearly, you know, the past has shown us some of the failings of you know, approach to this. You know, we didn’t I don’t think looked at or maybe we did, maybe there was a process, pre any randomized control study, it was all real world data, and then we move to this gold standard and, you know, appropriately so and now. I think there’s a, a move back, or at least an opening to add this in. As you think about it. What does that future look like? What are the key elements that comes to the forefront of research and technology and innovation that is brought to the table by this real world data, what excites you?
Laura Fernandes
So, so there are lots of areas where real world data actually could play a role. You know, I have to say that we are in the early stages of the use of real world data. But I would say there are three main areas that I really look forward to where real world data could play a role. The first would be in efficient clinical trial designs where we just don’t reach where we just don’t assign patients to the same standard of care, you know, in a clinical trial. But we augmented from the real world setting, this would have very good downstream effects, where we would be able to approve effective and safe therapies much faster, you know, so we don’t have to wait for longer to get like these good effective therapies. A second way I would say, where reliable data would definitely play a role is in this personalized medicine or precision medicine. Imagine a patient coming to your clinic and as a as a, as a treating physician, you can actually put in your your characteristics of your patient, you know, you are a particular patient, you are a female in this age category with this race. And you can see what has what have been the outcomes for patients of a similar demographic in the past, and you can assign the best therapy for your patient in real time, you know, at this time, what is the best work in therapy. And in third, I would say the excess of new therapies for really disadvantaged demographic patient populations, there is going to be an increased accessibility. And we would see like fewer disparities in clinical trials and in clinical research because of the use of real world data. So I’m very hopeful about about the future and the use of real world data in this setting.
Nick van Terheyden
So as as you think about the future, does this start to inform our policy, and the way that we approached the development of drugs? And maybe it’s the way that we labeled disease? You know, I’ve had a number of conversations with folks and, you know, labels are important, you know, they contribute to our understanding, you talked about lung cancer, and then you see it, you know, it spawned into four or five versions, and then I’m thinking, I’m willing to bet that that’s not just lung anymore, it’s in other places, you think that’s going to inform the future of healthcare, definitions and labels?
Laura Fernandes
Definitely not. We have already seen this right during the pandemic, where our our use of real world data informed our policies. So the the need for a booster shot for a vaccine was in fact defined or helped by the use of observational clinical trials that were done in Israel and Qatar. So it was observational data that guided our mandates for the use of a booster shot in vaccines. And I can only see this trend improving in the future. In oncology, we did see an initial use of global data. And I can see this being translated to other disease types too, you know, especially rare disease types, it’s so difficult to have a pediatric indication, and pediatric pediatric trials or trials, which are done in like infants, or like, you know, under adolescence. And so it’s harder to enroll patients in this very rare patient to subgroup. Because, because it’s hard not to do clinical trials in this group. And that’s where we can see an improvement by use of real world data. Definitely.
Nick van Terheyden
Fantastic. Well, unfortunately, as usual, we’ve run out of time, just remains for me to thank you for joining us on the show, obviously, exciting opportunities with the expansion of this data, I think, you know, a long challenged area of pediatric enablement of drugs, you know, must resonate with every parent who goes, gosh, you know, if I have a child with a disease, you just think about the, the development of vaccines and how long that took to get the children in, if we could supplement the adult data and provide some faster path to be able to get to the approvals and safer drugs for children and infants even we’ve still got, you know, a whole group that aren’t satisfied. tremendous opportunity. So these are exciting times, you know, delighted to have had you on the show and learn a little bit more. So, thanks for joining me, Laura.
Laura Fernandes
Thank you so much, Nick. It really has been a pleasure. You know, I also wanted to say that during my time at the FDA, I have seen a wide accept You know, by the FDA, it is also paving the way by showing that it is open to using real world data. And so, companies like Kota do help in, in, in, in using this acceptance by the FDA for real world data by providing an avenue you know, for for such use of data in clinical research. I’ve been very happy to be here today on the podcast. It was a very great opportunity to speak with you, and I hope your audience does have a better understanding now available data and clinical trials in general. Thank you
