How Real World Evidence Will Create Better Trials
By Jeremy Gilbert, PatientsLikeMe
A large pharmaceutical company was recently set to run a major Phase 3 trial for an underserved and complex condition with few effective treatments. So they took the common road used throughout the industry: they talked to key opinion leaders and investigator experts who were treating the condition and asked which symptoms should be included in the trial to judge if patients were getting better or worse. They received good feedback, but were hesitant to proceed. What gave them pause was a fundamental question: shouldn’t what they asked and measured in the trial be based less on what experts treating the condition thought and more on input from the patients actually living with the condition?
The company decided to look inside our phenomic registry to see what patients with the condition were talking about, and which symptoms they chose to report. What we found was astounding. Patients who live with the condition weren’t talking about some of the established endpoints. But they were talking about a critical issue that the company didn’t consider important enough to include. Since the issue wasn’t represented in the literature, we decided to conduct a survey to see if it was just an anomaly or truly important. In six short weeks we developed our hypothesis based on the retrospective analysis and confirmed it with a prospective study. The results were clear: the symptom absolutely had to be included in the trial.
The company listened to the real world experience of patients, thought about what would be most valuable to them, and changed the Phase 3 design in time. But this scenario is still more the exception than the rule. The path to bring a drug to market is well worn: test an indication in the perfect and pristine world of a clinical trial using measures that were developed, usually decades ago, by researchers. This path rarely acknowledges a medicine’s real world impact on real patients in real care settings. As a result, our ideas about disease as researchers may be very different from patients’ own.
Patients’ experiences outside of the four walls of a clinic will guide future trials in two fundamental ways. First, patients will increasingly use technology to record their day-to-day experiences, and donate their own data to supplement clinical data. This happens today on websites like patientslikeme.com, where patients living with chronic conditions can share health data with other patients and with researchers. Devices like Fitbits and implanted glucose meters are also helping to track how patients are doing, and keeping them on track. But these devices force the patient to learn how to use a new tool. Patients also have to take specific steps to upload information. In the end, their data isn’t synced with other sources to give a whole view of what’s happening.
The good news is we’re headed in that direction. In the future, patients will use the devices they already know, such as their mobile phones, to automatically record answers to clinical trial questions or to collect health statistics. Their data will be continuously uploaded to a centralized, hybrid registry where it can be linked with EMR and clinical data. That’s when we’ll start to see a truly holistic view of the patient in a trial protocol or along a care continuum. In doing so, we will be able to understand where there are treatment gaps and what we need to study next.
Second, patients will become much more involved in the design, planning, implementation and aftermath of a study. They’ve already told us that they want to be involved from the start. In a March 2014 survey of more than 1,600 PatientsLikeMe members, 93 percent of respondents said they would be willing to help researchers with design. But they also want to continue to give and get feedback throughout the trial process. When you think about it, they really should. A patient who has just experienced a drug, who was willing to donate their time and lives to an investigation, is already fully engaged in the process, and as a result is even more valuable to research.
But if you want to engage them even further, do something radical: give them the results of the trial. Time and time again we’ve heard patients express deep frustration that they don’t see how they did during the trial, and, if they did improve, find out how they can stay on the drug while it’s getting approved. That frustration drove our open Lithium study in 2011, the results of which were shared in Nature Biotechnology. This patient-initiated observational study refuted a 2008 published study that claimed lithium carbonate could slow the progression of the neurodegenerative disease amyotrophic lateral sclerosis (ALS). After the original claim was published by the Proceedings of the National Academy of Sciences, 348 ALS patient members reported their off-label use of lithium on PatientsLikeMe. Just nine months later, we presented preliminary results that lithium was not having an observable effect on the disease progression of these patients. The results were revealed before any of the formal-follow up trials enrolled patients and represented the first time a social network had been used to evaluate a treatment in a patient population in real time.
Our more recent Fitbit study with Biogen may also show how to best engage and share data with patients. The goal was to see how people living with multiple sclerosis (MS) can use wearable activity tracking devices to collect and share their mobility data. Nearly 250 PatientsLikeMe members were provided with Fitbit One activity trackers. Study participants were able to track their physical activity levels on the PatientsLikeMe website and connect with each other in the MS discussion forum to talk about changing symptoms, benefits, and issues. Even after the study finished, a large percentage of the participants continued to track their activity and connect with each other. This experience, and others like it, has shown us that patients can participate in interventional research in an open way that benefits patients and researchers alike.
By engaging patients before, during, and after trials, we’ll see a real transformation in the way clinical research is done. Endpoints will be shaped by patient voices, rather than textbooks. When questions about burden and logistics arise during planning, patient voices will guide clinical teams. Beyond pivotal trials, we expect that observational research in clinical settings will be augmented, and in some cases replaced, by virtual trials where any patient can monitor the results and interaction. Virtual trials are a natural and necessary evolution to how we study. They can improve the process for patients because they don’t require them to leave their home, go to a new doctor, or upend their lives to travel to a remote site for simple tests. This isn’t as futuristic as it sounds. We’re on the verge of starting a virtual trial in a specific condition, and we expect to conduct even more in the future. We expect that the learnings we see in such observational studies will start to trickle down to Phase 3 and Phase 2 studies, just as any new technology or process does when it has demonstrated its track record and usefulness. One thing is certain. It will be patients and industry who will create this future, together.
About Jeremy Gilbert
Jeremy Gilbert leads product strategy and development efforts at PatientsLikeMe, a patient network and real time research platform. Prior to joining PatientsLikeMe, he was an Engagement Manager at McKinsey & Company’s health care practice, where he led strategy and execution projects for Fortune 500 clients. Jeremy has co-founded four technology startups and has invented and launched notable products in e-commerce, mobile, and discovery bioinformatics.