DDW sits down with Mark Matson, Senior Solution Specialist for Sensor Cloud at Medidata, a Dassault Systèmes company, about the use of digital biomarkers in life sciences and how they can help advance our understanding of disease.
How do we define digital biomarkers?
Digital biomarkers are data, or metrics, collected using digital health technologies to help describe a biological process. For example, digital technology can measure a patient’s movement with an accelerometer – this is a digital biomarker. These biomarkers help expand our knowledge of biological processes. To better understand Parkinson’s disease, for example, we can use digital tools to measure bradykinesia, or slowing of gait, as a digital biomarker of the disease, which allows us to understand disease processes and the impact of therapeutic intervention.
When looking at digital biomarkers, we think about the phenomenon we’re trying to measure, how we measure it, how we prove that our measurements are accurate, and how they correlate with the disease process.
Is this a new way of looking at and understanding disease?
We’re just beginning to discover these digital biomarkers because of how sophisticated digital technologies are becoming. 10-15 years ago, there were few validated medical devices for clinical research, and now we have a lot of technology to choose from. Companies are using sensors that we can put on our wrist or chest and capture all types of data. These different sensors are being used in various ways to try to come up with new insights that could lead to the discovery of new digital biomarkers.
Why are digital biomarkers important?
By understanding diseases better, we can treat them faster and more accurately. Most diseases affect the physiology and biomarkers can help to understand the normative ranges and how diseases deviate from these normative ranges. Then we can figure out how to bring people back into the normative ranges.
Technology is changing the way we measure these phenomena. For example, there’s a new product called Biobeat that continuously measures your blood pressure. This device can sample blood pressure much more often compared to measuring your blood pressure in a clinic at one point in time. So we have a much better understanding of the normative ranges and when people deviate from this range and the impact of those changes on outcomes. Having these new data leads to greater understanding and enables precision medicine.
Why are digital biomarkers important in clinical trials?
Collecting objective data with sensors as part of a clinical trial not only helps verify the subjective data collected from patients and investigators, it also provides data gathered more often and in the patient’s natural setting (compared to a clinical setting). This new clinical evidence supports the traditional endpoints and has the potential to replace many of the more subjective endpoints available today, especially when measuring a patient’s functional capacity.
What expertise does Medidata have in the digital biomarker space?
In 2020, Medidata acquired a digital biomarker business MC10 to enhance its remote biometric data capture capabilities. The BioStamp team successfully created and cleared 47 metrics with the FDA to measure vital signs, activity, posture, sleep, and the relationships among them.
Bringing the digital biomarker experts from MC10 together with a team of big data architects at Medidata allowed us to create a scalable sensor platform and speed up the development of digital biomarkers by normalising the data into a common data model where downstream analysis is based on a common data structure. This means we can combine all the data coming in from different medical grade sources and streamline the discovery of new clinical insights.
What do sponsors and CROs need to consider when choosing a partner for their sensor needs?
The foremost consideration when choosing sensor technology is to ensure that it is fit-for-purpose and will provide the clinical evidence required for the study. Beyond that, there are other operational considerations associated with deploying sensors in a global clinical trial, including documentation to support Clinical Trial Applications in various jurisdictions, training of sites and patients, and ongoing support to achieve high compliance and high data yield. Once collected, there are data analysis considerations surrounding the large amount of data and the connection to other data sources, such as EDC and eCOA data. Lastly, ensuring that the data collected can be pooled with results from future research will be important, not only to obtain longitudinal data, but to garner insights from that data to better understand disease processes and long term treatment effects.
What are the challenges using sensors in clinical trials? How does Sensor Cloud address this?
The challenges when deploying sensors into clinical trials fall into three categories: technical, operational, and analytical. The technical challenges emerge from the types, variety, and sheer number of sensors available today, each with their own technical solution, data models, APIs, and roadmaps. The operational challenges surface as clinical teams attempt to distribute devices, train sites, support patients, and troubleshoot inevitable issues. The analytical challenges mount after the study is complete and the terabytes of data collected represent a needle-in-the-haystack issue for the data science teams.
Overcoming these challenges requires a scalable and secure software solution to manage the high-volume and high-velocity data, and a means to normalise the data for interoperability and processing efficiency. Medidata’s Sensor Cloud allows for rapid integration of most sensors and devices, provides standardisation and analysis of medical grade data under one common data model, measures a wide array of patient data, and delivers a unified experience for patients, sites, CROs and sponsors.
What are some disease areas where Medidata is looking at digital biomarkers?
The use of medical devices to characterise physiology is valuable in many therapeutic areas, including most disorders of the central nervous system (CNS). We’re involved with Parkinson’s Disease, Huntington’s Disease (HD), Multiple Sclerosis, Essential Tremor, and Dementia trials. These digital measures add clinical support to standard assessments conducted by clinicians, such as the unified Parkinson’s Disease Rating Scale, and help to remove the subjectivity associated with the eCOA instruments.
HD is a neurodegenerative disease that affects ~30,000 people in the US. The earliest symptoms are often subtle problems with mood or mental ability, and a general lack of coordination, which is followed by an unsteady gait and chorea (the uncontrolled movements that HD patients experience due to progressive brain degeneration). Medidata successfully utilised sensors to characterise gait abnormalities and chorea in a Huntington’s Disease Phase 3 study, providing the sponsor with clinical evidence to support their primary endpoint.
We’re also developing an at-home Six Minute Walk Test, combining the standard distance measure with other vital signs, such as heart rate, heart rate variability, heart rate recovery, and other cardiac response to exercise. The new assessment will be able to determine not just how far a patient walks, but how their body reacts to the exertion, adding a new, important dimension to a well-known and utilised functional assessment.
What does the future hold for digital biomarkers?
Fundamentally, a better understanding of diseases and disease processes. Measuring components of functioning, like postural changes, activity, gait, and sleep, along with vital signs, speech, and other physiology, allows clinical researchers to characterize disease progression and accurately determine treatment effects.
About the author
Mark Matson is a research professional with more than 25 years in drug development, sales leadership, licensing, and program management. Mark’s experience includes an extensive career at Pfizer in R&D, program leadership at MannKind Corporation, and stints in sales and sales leadership at IQVIA, Covance, and ICON. Mark joined Medidata through the acquisition of MC10 and is currently Managing Partner for Medidata’s Patient Cloud.
