Pfizer Perseveres In Pioneering Virtual Clinical Trials
By Wayne Koberstein, Contributing Editor
Success is sometimes measured more by what you tried than how high you scored. In science, the maxim is selfevident: Falling short of expectations can prove more valuable than attaining them — once you look at the evidence and learn from it. How else to explain the excitement that surrounded Pfizer’s REMOTE (Research on Electronic Monitoring of OAB [overactive bladder] Treatment Experience) trial, the first randomized virtual clinical trial conducted under an IND (investigational new drug) application, even after the trial ended in mid-2012 with less than stellar patient response.
REMOTE recruited and qualified patients online for a Phase 4 study of the approved OAB medicine Detrol LA (tolterodine). It attracted candidates with a cartoon video explaining the trial and how to apply and participate. During the trial, applicants went through an extensive online process of education, qualification, and enrollment, followed by self-reporting of their responses. The study was to operate essentially as a giant central trial site, with investigators overseeing data collection and analysis and primary care physicians helping screen and steer patients along the way. Investigators would ship all study drugs to the patients’ homes, rather than dispensing it at a clinic. But REMOTE fell far short of its goal of recruiting 600 total patients from 10 states in the U.S. What went wrong? Therein lie some of the lessons Pfizer learned from the experiment, lessons it plans to apply while developing future studies for consideration in Europe in 2013 — and lessons that can help other companies follow Pfizer’s pioneering push into virtual trials territory.
Craig Lipset, worldwide head of clinical innovation at Pfizer, discusses those lessons in the following conversation, putting them in the context of what REMOTE was intended to accomplish and what it did accomplish in the form of attained knowledge, despite or even because of its poor recruitment results. His colleague and the team leader for the trial, Clinical Development Senior Director Miguel Orri, M.D., adds details to the context and explains how the internal hurdles his team faced influenced results and steepened the company’s learning curve. (See the sidebar “Virtual Trials Trailblazer.”)
THE NAME OF THE REMOTE VIRTUAL TRIAL, “RESEARCH ON ELECTRONIC MONITORING OF OAB TREATMENT EXPERIENCE,” SUGGESTS IT WAS A SCIENTIFIC EXPERIMENT. SO YOU INTENDED TO LEARN FROM IT FROM THE START — EVEN THOUGH IT MIGHT NOT HAVE GONE THE WAY YOU EXPECTED?
LIPSET: Well, in any research project, if you know how it’s going to end, there’s probably no point in doing the project, right? We were attempting to model a new approach. We chose to use the medicine Detrol for overactive bladder as the vehicle for testing this model, because Detrol has a well-characterized efficacy and safety profile. We constructed the experiment as a series of modules to make sure that we could take advantage of what works and continue trying to refine what didn’t. It’s also important that REMOTE fits within a much larger set of experiments, developments, and efforts we have under way to reform medical product development.
WHAT ARE SOME EXAMPLES OF HOW REMOTE HELPED YOU IN THAT CONTEXT?
When we develop new medicines, we have to understand their efficacy and safety before and after they’re approved and gather continuous data on their value for payers, and that is a challenging proposition today. We need to continue to take advantage of new tools that are available to us to help us develop that data. REMOTE was an attempt to leverage many of those new tools. Clinical research is disruptive to healthcare. It requires the physician or patient to stop what they’re doing and change roles — from physician to investigator or from patient to research subject. We want to make participation in research easier for the patient and the provider.
WHEN THE REMOTE INITIATIVE ORIGINATED IN THE COMPANY, WAS IT SOMETHING THAT BUBBLED UP FROM CLINICAL DEVELOPMENT, OR DID IT COME DOWN FROM TOP MANAGEMENT?
There was probably a convergence of three different efforts. One, some folks at Pfizer were internally exploring new data-capture tools and platforms. Two, in parallel, my team began to develop a discrete concept of a virtual trial that could take advantage of some prior work by Lilly and a virtual-trials start-up, then called 1747, Inc. Three, Dr. Steven Cummings at UCSF, one of the founders of 1747, began to open a dialogue with us at Pfizer. So all of those efforts came together with creative leadership at Pfizer that was willing to embrace the innovative new approaches. It was work proposed upward, but certainly it was critical that we got top management to embrace and support it.
TOP MANAGEMENT MUST HAVE HAD A MATCHING AGENDA AT THAT POINT, AT LEAST IN THE SENSE OF IMPROVING THE COST AND EFFECTIVENESS OF CLINICAL TRIALS.
Absolutely. There is general recognition that current clinical development models are largely unsustainable. There are challenges in recruitment and challenges in making clinical research viable for healthcare providers to participate — key challenges that really jeopardize the ability to continue developing medicine. Everyone knows major improvements are needed, but it’s still challenging for companies to step forward and take that risk, especially in an environment that’s increasingly resource-constrained.
WHAT WERE SOME INTERNAL AND EXTERNAL HURDLES OR ROADBLOCKS YOU FACED, ONCE YOU BEGAN MOVING TOWARD THE REMOTE TRIAL?
Pfizer made a decision early on that the project was going to be run within the R&D organization. There are certain innovative initiatives that it makes sense to isolate or protect from the rest of the organization, and there are theories that you need to protect very delicate new innovations or a large organization can crush it. But Pfizer decided to conduct this project within the organization rather than doing it in a small protective group or by simply writing a grant to an academic researcher and walking away. On one hand, it made things a little more difficult. We had to bring on board very pragmatic and experienced clinical research professionals and convince all of them that this was worth their commitment. On the other hand, it created a sense of ownership in the whole organization. It was not something that was just thrust upon them by senior leaders and supervisors. The clinical operations and regulatory colleagues owned it, and that created an upward wave that helped support the culture of innovation that we were trying to encourage within Pfizer. It showed that people can work on out-of-the-box new projects in parallel to their other work and help contribute to corporate change.
HOW DID THE TRIAL FIT INTO THE LARGER PICTURE OF THE FDA’S CLINICAL TRIALS TRANSFORMATION INITIATIVE (CTTI), AND HOW WAS THE AGENCY HELPFUL?
CTTI is an important initiative that relates specifically to quality and efficiency in clinical trials. And though we believed early on that REMOTE was very compatible with the CTTI, we did not see it necessarily as a CTTI project. Yet it was consistent with the spirit of collaboration and transformation that the FDA and other stakeholders want to achieve. We deliberately structured this project under the IND application for Detrol to ensure that we could have a group discussion and engagement with the FDA. This forum created opportunities to discuss key matters such as complying with requirements around drug distribution.
WHAT ASPECTS OF THE STUDY WORKED, AND WHAT FAILED?
We saw tens of thousands of patients who responded to the call, though we might have made some improvements in outreach. Thousands created accounts, thereby expressing interest not only in general, but in going to the next step of demonstrating interest in participating. We saw that our process could ensure patient consent was properly acquired without ever actually seeing the patient live and in person. We demonstrated our ability to distribute blinded investigational drugs directly to the patients in their homes and then to use creative tools and platforms, mobile and Web-based, to capture data from those patients. We showed that you could use a centralized investigator, just as we’ve used centralized IRBs and centralized labs for monitoring patient safety. Now what we failed to do was to find enough eligible patients. But we were looking for patients with severe disease, as in the earlier trials with Detrol, and it was hard to find a sizable number of women purely online who had very severe OAB to qualify for the study. We also learned that for certain diseases and severity level, offline conventional channels work better than the online channels. That is not a claim of success or a claim of failure; it’s a claim of learning.
DOES THAT SUGGEST THAT VIRTUAL TRIALS MIGHT BE MORE SUCCESSFUL IN LESS SEVERE CONDITIONS?
That may be one of the conclusions as we continue to dig through the data and talk to patients who expressed their interest in participating. But there may still be other areas in the near term where virtual trials make sense, say with rare diseases where patients may be very widely distributed and you can’t set up a single investigator site and expect lots of patients in geographic proximity. We also want to continue to explore hybrid approaches; for instance, if healthcare providers and treating physicians are the right channel to help me find patients with severe disease, how do I make this trial model fit with that knowledge? Rather than relying completely on the Internet and social media to find patients, can we reach them through the treating physicians?
SO YOU ARE ENGAGING THE TRADITIONAL HEALTHCARE SYSTEM TO COMPLEMENT THE VIRTUAL DIRECT-TO-PATIENT APPROACH?
Every healthcare interaction is an opportunity to inform research. How do we make sure the interactions of our participating patients, physicians, and investigators are informing our research process? The primary care doctor today has only two options for research participation. They can become an investigator, but eight times out of ten a novice investigator will lose money and never do another study. Or they can participate by referring patients into someone else’s study. But why would a physician refer a patient given the pay-for-procedure reimbursement of healthcare in the U.S.? So the current system isn’t very friendly for the treating physician to participate. But we are creating ways for physicians to participate more easily, to share data more easily.
HOW WERE YOU ABLE TO ELIMINATE INTERFERING FACTORS IN RECRUITMENT, SUCH AS THE EFFECTIVENESS OF THE WEB ANIMATION OR SIMPLY THE LACK OF HUMAN INTERACTION?
We were able to do it by looking at the patient numbers and drop-off at various steps in the process, rather than just looking at the total number of patients we reached or randomized through a social media campaign or other campaign. By seeing where patients are dropping off, we may conclude, say, that it is not the informed consent process causing the loss but disease severity or maybe it’s a mix of both. We actually reacted to the analytics in real time, making adjustments in our outreach tools based on the patient feedback we were receiving. There were other elements that analytics could not answer, such as patient psychology or attitudes. In those cases, we asked the patients questions about why they did or didn’t participate, just to help us toward the goal of learning.
WHAT ROLE DO YOU SEE VIRTUAL TRIALS PLAYING IN THE CONTEXT OF IMPROVING DRUG DEVELOPMENT IN GENERAL?
As an industry, we will continue to fail to recruit patients in our studies if we cannot create an ecosystem of patients already engaged and aware about research studies and research participation. We need to plant the seed of patient engagement earlier on in the process. As we are rethinking patient engagement, we are looking at all of our touch points with the patient — from informing about research participation, informed consent, study participation, and even opportunities after the study concludes.
WHAT ABOUT A SET OF BEST PRACTICES THAT WOULD APPLY TO ANY VIRTUAL TRIAL?
We don’t try to generate intellectual property for ourselves in this space. We are a medicine development company. We also believe that if we are the only one exclusively using a new method, that doesn’t necessarily serve us well. We need others to adopt the new methods so that regulators, providers, patients, and payers are familiar with and support them. If it’s only a Pfizer way of doing things, that process will be slower. Even so, we will be more competitive, we will have a head start, and we will know how to do it better and faster. But we are transparent about this and even prepared to work together with other companies. The recently announced TransCelerate BioPharma collaboration is indicative of the type of collaboration and sharing we are seeing across pharma sponsors to improve the drug development process.
HOW DO YOU ENVISION THE INCREASING USE OF ELECTRONIC PATIENT MEDICAL DATA AFFECTING VIRTUAL TRIALS AND DRUG DEVELOPMENT IN THE FUTURE?
In the REMOTE model, we were highly dependent on patient self-reporting. In the not too distant future, when all health data is digital, we will no longer have to rely on patients to self-report their data. Patients will have access to their electronic health records and will simply authorize and share that data. Today, we have very inconsistent access to our health data electronically, but financial incentives such as the Center for Medicare and Medicaid EHR Incentive Programs through the HITECH act are driving widespread adoption of electronic health records by providers who satisfy specified “meaningful use” criteria — including a requirement that patients have access to their own electronic health data. And when patients are empowered with such access, it changes the game because they become a truly trusted broker of health information for research.
VIRTUAL TRIALS TRAILBLAZER
On the front lines of Pfizer’s trailblazing REMOTE (Research on Electronic Monitoring of OAB Treatment Experience) trial was United Kingdombased Miguel Orri, M.D., clinical development senior director. Orri’s views of the trial complement those of Craig Lipset, worldwide head of clinical innovation, by adding details of how the REMOTE team overcame internal corporate hurdles to implement the virtual trials model in a reality-based experiment. Orri was featured in Rob Wright’s Editor’s Blog in September 2011 for his presentation on REMOTE at The Conference Forum’s Disruptive Innovation in Clinical Trials event. Ironically, the innovation he described, though disruptive to the conventional way of conducting trials, was guided by the central principle of minimizing the disruptions for physicians and patients typically caused by trial recruitment and monitoring. “Our primary aims were to get primary care physicians involved, make clinical trials more accessible to patients, and ease the burden of recruitment for investigators,” Orri says. “We were trying to make clinical trials more efficient but also more attractive and inclusive for all parties.”
Creating a single central recruiting site for the trial simplified patient recruitment for investigators and allowed patients to participate from their own home. Primary care physicians were able to stay engaged with their patients by helping to screen and care for them during the trial. Patients also received their own data at the end of the trial and could choose to share the data with their physician if they wished. “In many studies, it might be a bit far-fetched to go completely remote. Specifically in earlier phases of drug development, you will need initial contact and an in-person assessment to establish the diagnosis, but then all the follow-up could be done remotely.”
Virtual trials may be done in a modular fashion: “Even if you only replace half of all the visits with a remote visit, it might make a big difference to the patients and still enable us to collect data in a more consistent way.”
Use of patient eDiaries can improve data collection, Orri says. “Patients enter the data without this being filtered and transcribed, the entries are time-stamped, and you can verify that the data was actually filled in when they say it was filled in.” He describes what the company learned about patient qualification in REMOTE:
“In the informed consent process, we asked patients to fill in a questionnaire after they had read the informed consent document and watched a video, so we could assess that the patient understands. They then went through a telephone call with the investigator, and finally they signed electronically that they wanted to participate.”
Patients, physicians, and investigators may welcome such improvements over more cumbersome conventional methods, but internal players were less open to them, at least at first. “Recruitment was a challenge, and that was mainly triggered by a tendency in the team to be overly regulatory-cautious, and thus to ensure that no wrong patients got in because we made it so difficult for anyone to get into the study.”
The REMOTE team got the FDA involved at an early stage, Orri says. “They were very supportive of the project, but on occasions some team members wanted to take a more conservative approach than in conventional trials, which would then typically be escalated in the company hierarchy to a governing committee. The result was general affirmation of the project and its goals but a compromise that favored a more conservative approach.”
Patients had to jump through numerous hoops, including creating an account and reconfirming their email details before they knew if they were eligible for the trial. “We learned from that, and there’s actually a plan at the moment to incorporate improvements into future studies. I think we might be able to run the next project with a bit more confidence. I can remember many times saying, ‘This is much more vigorous than what we do in the conventional setting.’ Our increased confidence should make it easier the next time around.”
Orri believes virtual trials will eventually be seen as part of quality by design applied to clinical development. “We have so many options here as well to make the trial population ‘cleaner.’ In clinical trials, protocol violators often get into the study, and here we actually had an electronic system that wouldn’t allow the investigator to enroll patients that were deemed ineligible. So part of the system selected the patients; it still had to go through the investigator, but when a patient was deemed ineligible based on predefined criteria, then the investigator couldn’t override that, so the patients were automatically disqualified from the study. And that basically gave us zero protocol violations for patients entering the study.”