Share this article on:

Neurotech: Virtual Clinical Trials What Investigators Are Learning about the Challenge and Opportunities

Owens, Sarah

doi: 10.1097/01.NT.0000524841.05646.4c
Back to Top | Article Outline




Virtual visits using remote technologies and devices can help improve access, enrollment, diversity, and even the quality of data gathered in clinical trials, researchers told Neurology Today. But issues of feasibility, reliability, and data correlation remain, they said.

Virtual technology, which bridges the gap between doctors and specialists or patients by enabling real-time consultation, is making its way to large multicenter clinical trials. The ability to gather data through open websites, smartphones, or tablets is enabling investigators to pull data from more geographically, ethnically, and socioeconomically diverse sources, experts in the field told Neurology Today, and patients are spared the inconvenience and potential safety risks of traveling great distances to academic centers where trials are conducted.

“I think a fundamental flaw in the way we conduct clinical trials is that we ask people to participate in these trials on our terms,” said Ray Dorsey, MD, MBA, the David M. Levy Professor of Neurology and director of the Center for Health and Technology at the University of Rochester Medical Center, who is spearheading research on the benefits of virtual visits in neurologic disease trials.

“Virtual trials enable people to participate on their terms. The trial comes to them. Instead of a site-centric approach, all activities are centered on the participants,” Dr. Dorsey said. “We can conduct studies that span borders and long time periods. This is especially important for rare neurological disorders. It enables people to participate in studies from parts of world where there may not be a specialist in their condition.”

Back to Top | Article Outline


The question, some researchers say, is whether virtual trials can replace standard in-person clinical observation and scales reliably. Christopher Tarolli, MD, senior instructor in the department of neurology at the University of Rochester, for example, is examining the correlation between virtual and in-person assessments using the United Parkinson's Disease Rating Scale (UPDRS) in the context of a clinical trial.

Dr. Tarolli, Dr. Dorsey, and their colleagues at the University of Rochester conducted an add-on study to the phase 3, double-blind, placebo-controlled trial STEADY-PD III study, which measures the efficacy of isradipine in early PD. For that study, all participants were assessed in the clinic, three times a year; for the add-on study, a sub-cohort of 40 participants also completed a virtual research visit comprising the same motor and non-motor (questionnaire) assessments.

At each virtual visit, “Participants were provided with a large smartphone or small tablet that provides a video link to myself or other investigators,” Dr. Tarolli said. “We have them set the phone up so we can see most of their body, and they perform [all] but two portions of the UPDRS — rigidity and postural instability.”

In an interim analysis of their findings at this year's AAN Annual Meeting, Dr. Tarolli presented data showing a high degree of correlation between virtual and in-person non-motor assessments, as measured by parts IA and IB of the Movement Disorder Society-UPDRS (0.61 and 0.82) and the Montreal Cognitive Assessment (0.78). The correlation between motor scores, however, was lower (0.37, as measured by the UPDRS part III).

That lower correlation “is more likely an issue of inter-relater reliability rather than inter-method reliability,” Dr. Tarolli said. “In the long-term application of virtual visits, that's likely to be less of an issue,” he said, but that remains to be seen.

Back to Top | Article Outline


Richard Bedlack, MD, PhD, associate professor of neurology at Duke University and director of the Duke ALS Clinic, had a patient with amyotrophic lateral sclerosis (ALS) who had regained significant motor function after taking the dietary supplement, lunasin, a peptide found in soy and cereal grains, and he wanted to see if the same treatment might work for other patients, too.

For the Lunasin Virtual Trial, he and his collaborators partnered with PatientsLikeMe, an online patient network and research platform, to enroll patients with ALS who replicated exactly the course of therapy — at least six months of lunasin — of that patient. “We decided to make the study very open in terms of inclusion criteria. We published the protocol online. Anyone with ALS could get into this study, as long as they could use a computer.”

Progression was measured using the ALS Functional Rating Scale by a remote assessor and also was self-administered by participants and their caretakers, who then recorded their results in the PatientsLikeMe database.

Unfortunately, Dr. Bedlack said, no ALS reversals were observed. But the trial enrolled 50 people in five months — which is highly unusual for ALS trials — had a low dropout rate, and had very high compliance. Moreover, Dr. Bedlack and his colleagues found there was “excellent agreement between the self-assessments and the investigator assessments.”

Dr. Bedlack and his colleagues used historical controls to compare outcomes. While he acknowledged the study design is “not perfect,” he added that it offers “a quick way to find a big signal.”

Back to Top | Article Outline


Important metrics of disability are missed in clinical trials of multiple sclerosis (MS), said Jeffrey Gelfand, MD, FAAN, assistant professor of clinical neurology at the University of California, San Francisco (UCSF). In particular, he told Neurology Today, “There can be difficulties measuring real-world change in physical activity and ambulatory function.”

Dr. Gelfand, Valerie Block, PT, DPTSc, Bruce Cree, MD, PhD, and their colleagues, in collaboration with UCSF Health eHeart program, are conducting a cohort study, FITRiMS, to test the ability of an accelerometer, the Fitbit Flex, to capture step count in people with MS — a metric they believe can better measure physical function in the natural environment. Participants with MS wear the device on their wrist and “go on with their usual daily activities,” Dr. Gelfand said.

Findings from the remote assessment are being compared with results from in-clinic assessments using the gold-standard Expanded Disability Status Scale (EDSS), timed walk, and related metrics along with patient-reported outcomes.

So far, the findings show that remote step-count monitoring correlates well with standard functional measures in MS but can also identify important variability in walking otherwise masked by traditional scales, Dr. Gelfand said. In a paper published in the February issue of the Journal of Neurology, he and his colleagues reported that over four weeks of continuous monitoring, lower average daily step-count was significantly associated with greater disability on the EDSS (p<0.001) and that daily steps were lower in progressive versus relapsing MS (p< 0.01). Additionally, the study showed very high retention (97 percent) and high continuous use of the device (97 percent) over four weeks. Longitudinal assessment over one year is ongoing.

“In our research, we found important variability in average daily step-count even within EDSS groups,” Dr. Gelfand said. “We also observed important changes in average daily step count even when the EDSS stays the same. We believe that step count may be an important new outcome measure to accelerate research in MS, particularly for progressive MS and studies of reparative therapies.”

“We may be able to measure changes in real-world function that current scales don't have the responsiveness or granularity to measure,” Dr. Gelfand said. “For example, if someone with MS uses a cane to walk, by current definitions, they would have to no longer need a cane to register improvement on the EDSS, a common benchmark in clinical trials. But if a new treatment or intervention could meaningfully increase step count, even if in the context of still using a cane, that could provide an important — and relatively rapid — signal of benefit to support further research and attention.”



Data captured through virtual visits may also be more true to life, researchers agreed.

“Currently, we rely on data that are detected at arbitrary times in artificial environments,” Dr. Dorsey said. “For example, for pain, we collect data at the beginning of a study visit and at the end of the visit. Now, that might have been influenced by tons of factors — the patient's drive in, their experience in the waiting room, their anxiety, etc.” With virtual visits, he said, “you'll be able to measure pain more frequently.”

Back to Top | Article Outline


Still, experts acknowledged that “virtualization” of clinical trials has limitations. “There are things we currently can't do remotely, like sophisticated imaging,” Dr. Dorsey said. However, he added, for some measures, there may be workarounds — participants can visit a local clinic instead, for example, or “we can also send researchers to participants.”

Many of the devices used in virtual visits, including smartphones, tablets, and sensors, may be too complicated for some participants in clinical trials to use, particularly among older patients with progressive disability, experts agreed.

“We have to consider the concept of the digital divide — between the people who have access and those who don't,” Dr. Tarolli said. “In reality, the folks we really want to reach out to in order to improve diversity may be those who are on the other side of that divide.”

And virtual visits may lose a human component that makes visits meaningful and less isolating.

“When it comes to the technology, we saw that people felt that they didn't like having a screen between themselves and the doctor, or felt the examination wasn't quite as thorough,” even though they were covering the same ground, Dr. Tarolli said.

As a result of these limitations, clinical trials may ultimately include “a combination of in-person, investigator-rated assessments plus remote monitoring or assessments,” Dr. Dorsey said. “If trials are not done entirely remotely, maybe the first visit will be done in person but follow-ups are done remotely.”

“Having studies that are conducted entirely remotely may not be the goal,” he said. “The goal is to reduce the burden on participants and to create clinical trials focused on their needs.”

Back to Top | Article Outline


•. AAN Annual Meeting Abstract P4.005: Tarolli C, Andrzejewski K, Bull M, et al. Virtual research visits in individuals with Parkinson disease enrolled in a clinical trial: REACT-PD Study Interim Analysis.
    •. PatientsLikeMe: The Lunasin Duke Virtual Trial.
      •. Block VJ, Crabtree-Hartman E, Bevan CJ, et al Continuous daily assessment of multiple sclerosis disability using remote step count monitoring http:// J Neurol (2017) 264: 316.
        © 2017 American Academy of Neurology