Presented by Jeremy Gilbert, PatientsLikeMe
Clinical trials shouldn’t be burdensome for patients. Keep it simple by partnering with patients to ensure trials are designed with their needs in mind. Also covered: the effectiveness of PatientsLikeMe when designing and developing a clinical trial.
What I wanted to share with you is a little bit about what we’ve been doing in kind of three areas. But let me start a little bit, just letting you know about PatientsLikeMe as a company, just so you have a little bit of a sense of where we’ve come from. And I promise I’ll tie all this back to trials and the world of compound development.
So PatientsLikeMe was founded by two brothers who had a third brother diagnosed with ALS. And they were convinced—Jamie and Ben—were convinced that they were going to cure ALS, and they raised about $80 million over five years to start a therapy development institute. Probably one of the first non-profits to actually file an IND. And they didn’t succeed in finding a cure for ALS, but they did learn a lot about the plight of people with chronic degenerative conditions along the way. And PatientsLikeMe came out of that experience. It was basically, what the brothers realized and is something that I think we all know when we work with people with chronic conditions, degenerative conditions is that the patient-to-patient communication is absolutely vital in the management of these diseases. And so what you might learn from your neurologist is, well you know, you’re seeing X, Y, and Z motor neuron deficiency, but what you’re going to learn from other patients is, here’s how I decided to go onto a med, here's what I did when I needed to remodel my home for a wheelchair. And in terms of the ALS is just an extreme example of a very large variety of largely patient-managed conditions, where the patient starts to become a bit of an expert on their own about the condition. And that’s really the type of patient that we build our platform for. And what we do for them is really unique because we’re not—we are a social network, I mean we have social relationships that form in our community, it’s really diverse, people talk about “Dancing With the Stars” and they talk about a new medical treatment—but what we’ve built that is kind of an unusual twist, is a whole journaling tool. So I’ll describe this a little bit. Basically what it does is it’s almost like an eCOA system except it’s all online and it’s all patient driven. And then you can basically populate in your PatientsLikeMe profile, the symptoms that you're experiencing, the treatment start dates, treatment stop dates, basically your whole personal health record can be populated inside of it, and you can share it with other patients.
And why would you do that? Well it turns out there are actually a lot of reasons why patients want to know how other patients are managing their condition. Just a few examples. We've had patients who have never met other patients who have had the condition that they have. And the fact that they can even see what those patients are taking—say it’s an area where there’s no standard of care. Another example is we've had patients who were taking things like baclofen, which is used for stiffness and spasticity in MS, and they learned that they were on the wrong dosage of it, really just because they could see what other patients were taking. And then we’ve also seen cases where literally things are not approved for a particular indication, but people take it anyway. Like for instance, Adderoll is taken very frequently for brain fog. We’ve actually had patients use the accumulated crowdsource data on our platform to go to their insurance provider and say hey actually I think you should cover it, here’s data from 400 patients like me who are also taking Adderoll for brain fog, and they can get their approval letter. And so it’s both a social thing in terms of social support and connectivity, but it’s also about getting a better medical outcome by being a part of a group.
And we’re really principled about it. I mean one of the things—we don't make money off of our patients. We don't advertise to them in any fashion, we don’t take money from them in any way. Our entire model is to do research with patients. And so our entire interaction modality has to be very very patient centric. We’re basically telling patients, we’re going to make it worthwhile for you to be involved in research projects with us. And we’re going to make sure you’re kept in the loop and you’re part of that experience with us and that we’re being transparent and open with you on the way, and that you’re getting enough out of the experience of being in research that it’s worthwhile for you to do this. Because we’re not compensating them for being part of this network, nor are they paying to be part of it.
And so we’ve developed a series of really deep understanding of these types of patients and what they need. We've done hundreds of interviews and thousands of surveys to really get under the skin of about 20-30 different chronic conditions. And that’s really how we designed our system.
So I want to talk about what are some of the implications for this as we think about a world of eCOA and trials and what have we learned as that sort of I think is going to influence ahead of the ticker of where this industry is going. And the first is this area, that doesn’t have a great name but I think we all have a sense of it, and it’s something that we’ve been calling “phenomics.” And phenomics is what you’re measuring with ePRO, right? It’s the thing that’s patient accessible and not physical accessible. It’s the gestalt of a QoL survey or an outcomes survey. But I think one of the things that we’ve realized along the way is that this is really a very undiscovered area. People don’t understand the phenomics of disease nearly as well as they should, and I’ll show you some examples of that. But what we’ve seen, especially in inflammation and mental health and degenerative neurological conditions that are inflammation mediated, is that the environmental aspects of a patient’s condition or whether or not they feel good in the morning or whether or not they slept well at night, have all sorts of wide-ranging secondary effects through the progression of their condition. And so the question is, how do you systematize this in a more useful way rather than just single surveys, how do you think about this as a computable thing in the same way that a genome is computable or a phenome is computable. And that’s really one of the things that we’ve tried to dedicate ourselves to in the last the years.
This is an example of a patient profile in PatientsLikeMe, and you can see that we’re giving—you can actually see that this patient can put her symptom scales on a five-point Likert scale. We’re giving her a quality of life score. And what we're doing on the back end is we’re taking all this data and we’re coding it back to the medical vocabulary that you guys would be familiar with—Meddra, SNOMED, Multum, CPT-10, things like that.
And so what we’re really trying to do is not just deploy single surveys to patients for one research purpose but rather just try to say, how are you doing overall as a patient. And let’s ask you that question today and let’s ask you that question in a month and let’s see what patterns and things that we can learn. And the way we do that is that we actually have an AI system, which basically is taking all of this journal data and mapping it back to the controlled medical vocabularies, and we have a team that’s scanning it for adverse events, and then that’s what we publish on and we build about 60, maybe 65, peer-reviewed publications basically on back of this real-world patient experience data.
And one of the main applications of this has been in patient experience of disease. And I think sleep is a really interesting example of this, where most patients who say that they have insomnia don’t really recognize the medical definition of insomnia, which includes feeling rested in the morning, and not waking up during the night. People think bout insomnia typical as I can’t fall asleep. So what we did is a very large, we engaged 7000 of our patients across a huge number of conditions and we asked about those four different dimensions of insomnia, and we looked—this is a project we did with Merck—we looked at the ways that that experience of insomnia varied between conditions, we looked at the way that the different pieces, the components of insomnia, varied day to day for different patients. And as a result, we were able to actually build a picture for Merck around how insomnia, and the Belsomra drug that they were launching when we did this work, would fit in on a specialty basis.
And then I want to tie this to trials because this is a relatively recent example. We worked with AstraZeneca, I guess it was last year, in lupus. And they were building a secondary symptom scale to use for a very large Phase III trial. And they were curious whether or not the symptoms they were sticking on their symptom scale were actually the symptoms that patients cared about. So I think this is a pretty common question people have. And what we did was we basically said all right, well let’s do this. Let’s look at what people spontaneously report with lupus. So we’ll just go and we have 6000 lupus patients, they’ve already been just characterizing themselves over time in our system, let’s just look and see what symptoms they choose to say that they’re experiencing. And we found some discrepancies. In particular, we found that the patients were much more likely to feel concerned about body fatigue and body pain than you would normally find in the lupus literature. And AstraZeneca, of course, wanted a little more confirmation of this before they changed their trial. So we then did a prospective survey, and that’s actually what’s nice about our network, and this is where I think social media is kind of very key here. These are not just anonymous patients in an EMR database that we downloaded. These are actually patients we have a relationship with, so it was very easy to go back to those patients in a prospective way and say, okay, we’ve looked at your data and we saw this pattern, can you tell us more about it. And we engaged another four or five hundred lupus patients and just said, hey can you actually prospectively comment on your experiences of fatigue and pain in lupus. And this led to a change in the symptom scales being used in AstraZeneca's trial.
So another area that I want to talk about, that I think applies into the trial space, is this concept of reciprocity in patient research. And I think one of the things that we’re very keen about because of our business model is how do you make the reciprocity equation work for patients. And I think that you know when we run a project with patients—and I’ll show you an example of one—we’re very keen on giving patients data along the way. We’re very keen on making sure the patients feel like a partner in that research. And the reason we do that is, part of it is because of our mission, but another reason is that we know that it gets engagement, we know that people are more engaged when they’re a part of something. And I think this is—you know, so when we run a project, we’ll engage, we’ll build a whole consumer communications campaign round it, and then we’ll take the data and the insights from all that participation and we’ll push it back to patients. We’ll send them newsletters, we’ll send them updates, we’ll say hey, we learned the following things, or hey here’s the open access publication you can download based on the data that you generated. We’re basically tapping into the excitement people have to be part of something. My wife and I are part of a Kickstarter to make little dresses that have rocket ships for my daughter because we like the idea of we want her to be into STEM stuff. And you know we look at that, we refresh that blog every week, you know, we want to know where the dresses are because I think it’s kind of cool. That’s the kind of excitement that can, in fact, tap into with patients.
And I’ll just give a very quick example—I won’t belabor it—but this is an example. We’ve been asked to speak on this Biogen Fitbit study quite a lot recently. This is where we gave a lot of patients. I think it was 248 patients were given Fitbits. And we enrolled this basically in less than 24 hours, these patients linked, more than 80% of them linked and authorized their Fitbit and then stayed on throughout the entire process of the study, and this was done 100% virtually. So there was no site investigator, we basically just engaged these patients online. And it was pretty good, considering that there was absolutely no relationship with and investigator or a physician at any point in this process. And the way we did it was, we made it really exciting. We basically unboxed the Fitbit, we threw away all the packaging, because we said no 60-year-old MS patient is going to be able to deal with any of this, we put it in a case, in a box that we thought the patients actually could engage with, we made all the communication come from this woman named Liz Morgan, and she was the single point of contact and so every email message, every text message, even every little thing in the box was Liz speaking to the patient. And that’s how we got a relatively older population to link up Fitbits in a completely virtual study.
And you know, patients loved this, by the way. They said, we had all sorts of things that people said that it helped them manage their condition better. We learned a lot about Biogen about how patients found it kind of useful to track activity but less about tracking sleep. And we actually even were able to correlate measures. So we have a clinical measure of patients’ walking ability that has been clinically validated. And we were actually—we were able to even to see that the Fitbit was picking up fairly good sensitivity on—their step count had a fairly good sensitivity to the walking domain in their multiple sclerosis rating scale. So it was kind of a cool project. But the point of it was that we were using the power of patient enthusiasm and engagement.
And I think that when you look at the way clinical trials work, we’ve sterilized them, and for good reasons, right. But we’ve sterilized them in a way that we try to take away the patient-to-patient interaction, we actually in many ways keep information from the patients that we’re concerned about them knowing. And this barrier, which I know is there for statistical controlling reasons and things like that, but it works against us in terms of engaging patients in an authentic way. And I think we have to be able to balance these two things as an industry. Because I’m here to tell you that when you get it right, patients get really excited about research. And it’s not a question of retention. Your trials will fill so much faster if people got a vibe that they were part of something bigger and you could engage them in this way. And I particularly think that the social connections between patients are one of the most ignored pieces of this equation.
So finally what I want to talk about is this idea of patient input, which is the third application of our technology. And what we’ve been doing in the patient input space is basically trying to take what we do as a company and bring that out to the way that a trial is run. So when we did our Biogen study, the one I just showed you, we engaged patients throughout the whole process. When we were designing the kits, with the packaging and everything, we would just basically just send flash polls out to ten, fifteen patients, just ask them what do you think. We take everything to our patient advisory board. And my question is, does this have any applicability in the trial space. Can you actually systematize and make scientific the collection and utilization of patient input. And this has been a really big expanding area. There is a patients partners conference, there is some really work that has been shared at DIA and at PCCT about this type of patient engagement.
We did a very very early study in this space, this is 2014, we believe this might have been the first time that patient input was ever broadly solicited to design a Phase IIb study, which is kind of crazy if you think about how long it is that we’ve been running Phase IIb studies. But basically this was a case where Genentech was trying to decide if they were going to add a lumbar puncture into an MS study, and the reason they wanted to have these lumbar punctures was obviously for a biomarker validation study they wanted to run as a companion. And they went to their physician ad board, and they said do you think we could add two LPs into our protocol. And the physicians said no way, your protocol is sunk as it is, patients will not go for it. And so they said, well let's actually talk to patients. So we engaged with about 400 MS patients, we segregated—some of them had LPs before, some of them didn’t have that prior experience with LPs. And our questions were not—like, we didn’t ask them do you want to have a lumbar puncture, I mean no patient's going to say yes, right. But what we asked them was, if you were part of a clinical trial, how would it change your decision making around that clinical trial if it had that lumbar puncture. And what are the kinds of things that we could do in that clinical trial to make you more comfortable about having one? And if you haven’t previously had an LP, what are you worried about? what are the things that you’ve heard that could happen that you would want—what are the concerns that you would want us to be able to assuage? So it wasn’t a question about patient preference, it was more about, how can we work with you as a patient if we wanted to add this lumbar puncture in. And what we learned was actually that patients who had not previously had lumbar punctures had a lot of misconceptions. They actually thought that—they were worried about things that they probably didn’t need to be worried about and they weren’t asking questions about other things. So what a lot of this ended up doing was shaping both the decision to add the LP, but also the materials and the communication and the scripting and the things that were put around the study to help communicate this to patients. And Genentech, as part of this work, did actually go ahead and add that into the study.
So how do we systematize this. I’m very interested in making patient input into trials something that we can do on a repeatable and ongoing basis. I don’t want to live in a world where we sort of like—I’ve talked to clinical operations people all the time who say yeah, if only we had asked the patient about whether or not their refrigerator could hold this kit, we could have avoided this whole drug distribution problem. I mean, these examples come up over and over and over again. So we have a framework that we’ve been playing with, where we say, look let’s take the measurable attributes of a successful study, things like protocol compliance, retention, recruitment rates, these are the things that our industry knows how to measure. And then let’s work back in a systematic way and say what are the places specifically where patient input can affect some of those things, right. So we’re not just asking patients to feel good about it, we’re actually asking patients about drivers that affect our business. And we have four categories that we’ve been using around enrolment, protocol, and site experience. And the point is that we try to draw a link.
So if you think that, you know, your measurable outcome here is an accrual rate, what’s going to drive that. Well it’s going to be your mix of broad-based recruiting versus site-based recruiting. It’s going to be the way that you design your study portals and your websites. It’s going to be how you think through the informed consent and decision to enroll. Well those are all things you decide to do based on some understanding that your study team has, right, around well what do patients care about, what do they need to know, what are the kinds of things that we’re going to tell our call centre to answer, what kind of materials are we going to put on the website. That plan comes from some understanding of patients that we can interrogate, right. We can understand what patients, what are you worried about in an asthma trial that involves needles. Or for a small cell lung cancer case we recently did, if you had infusions that lasted longer than six hours, what are the kinds of things you’d want the site to be able to provide. These are interrogable questions.
And so we actually have created standard question sets that address the forms of patient understanding that we know can actually trickle through into a decision around how you enroll in a trial or how you retain in a trial or how you configure the sites in a more patient friendly way. And so the point of this is that we’re trying to advance essentially a science, if you will, of patient input and trying to make the idea that any study that a sponsor does actually is something that patients can weigh in on. And we’ve had enormous success, especially with AstraZeneca, where more often than not they now have as part of their study startup or study protocol design phase, an engagement with patients either at an ad board level or on a survey level to gather information and basically bring the patient voice into the equation.
So I’ll just leave you guys with three thoughts. I recognize this industry runs in a very constrained world. You guys have crazy regulations, it’s not easy to make patients social inside clinical trials. I’m there, I’m with you. But in terms of where trials have to be in order to come into the 21st century, in terms of where this research institution needs to be in order to have the same kind of engagement that people are beginning to expect in every other area of their life that isn’t a clinical trial, I kind of think that there are three things that we need to think about. We really need to get phenomic characterization right, this idea that eventually there is going to be additional signals that will come about from the environment, from a digital biomarker, from something that patients do online. We have to be able to adapt to that and I think we all know that that’s true. But I think it’s coming in a very serious way. And I also think, secondly, that the reciprocity issue in patients really has to get tackled. We have to think about really how do we design a study where more information gets back to patients and we engage them, we make them more human, and we bring the same kind of excitement and engagement and connectivity that many many other areas of human life have, because that’s ultimately—if we can do that in scientifically acceptable ways, we’re going to drive a different form or engagement, a different form of relationship to research, which will benefit the economics of the industry. And then finally I think that this idea of design partnership with patients, which I’m sure, you know, the ePRO designers are doing on a micro level, but this idea broadly of how do we design things thinking about the patient as a stakeholder that can actually provide actual information, I think that’s a key direction we have to go in as an industry.
So anyway, that’s what I had to share. I’m happy to leave some time for questions and dialogue and debate. Thank you very much.
[END at 23:32]