Realizing the Promise of Digital Biomarkers

The industry is ready to leverage all types of patient captured data and information in order to support better understanding of a drug's safety and efficacy. Digital biomarkers have the potential to provide unique, insightful, longitudinal and high quality data for clinical trials. How can we move the needle in leveraging them?

The digital biomarker (DBM) landscape

Digital biomarkers or DBMs have been around for almost a decade. The earliest DBMs were focused on fine motor movements, specifically in Parkinson's. The main driver had to do with challenges in subjectively assessing the level and progression of tremor, which traditionally was done by finger taps -number, speed and rhythm. Very challenging for a human to rate!

Over time the advent of apps and wearables, e.g. the FitBit watch or Apple Fitness app, have exploded into the consumer and even clinical trial space. Additionally, the development of user-friendly centralized repositories, mainly HealthKit and Fitbit, have democratized health data for the masses.

Smartphones have also evolved to become very powerful computers and all-around sensory machines (camera, video, audio, vibration, gyroscope). While there are 100's of thousands of digital health apps, (naturally) within clinical research the number is much smaller. The scientific rigor needed to ensure a DBM is ready for patient research is a critical element and no small feat.

Even though the technology is there, a majority of Sponsors are still shying away from the use of digital biomarkers or even digital health technologies (to cast a wider net). Below is a chart from a 2022 analysis, showing that smartphones, sensors and apps were being used in less 10% of studies. I am sure the % has gone up in the last 2+ years, but probably still not to the degree that we would expect in the modern age of apps and wearables.

It is important to think about the value and practicality of deploying a digital biomarker, including scientific justification, validation and the ability to operationalize:

Scientific justification

There is an element of the "shiny bauble" when it comes to digital biomarkers. While the technology may look innovative, even cool, it is vital to maintain focus on the scientific value for the study and treatment. Even if the DBM is exploratory, as they often are, it is important to sit with the scientific team and put together a proposal and plan around the value and impact of the digital biomarker.

Remember, anything additional for the trial is adding time, effort, cost and burden (especially for sites and participants), which can potentially detract or distract from your primary endpoints.

Evaluation and Validation

The question of whether an instrument is appropriate for it's application is an important one to answer. If I'm using a smartphone app that was built by an AI bot overnight and "thrown" on the app store, it may not actually work the way I need it to or even the way it claims to. Thoroughly assessing the DBM is crucial:

• Scientific validity: Unless it is a new-to-market DBM, it is important to assess the use of the DBM in prior trials, any published results as well as potentially performing a "scientific audit" -requesting information like the algorithm validation methods and results, number of trials supported (phases, therapeutic areas) and scientific references from other Sponsors.

  
  

• Qualification: It is important to audit the vendor's processes around software release, testing and ongoing support/maintenance. This includes the software and hardware as well as the services surrounding them.

  
  

• Fit for purpose assessment: The DBM should be specific to the trial needs and not capture or require much "ancillary" data or requirements, e.g. registration of patients as "end users" along with personally identifying information (PII). Additionally, the DBM should be able to align with the ideal protocol requirements vs. having to design around the product or solution to make it work -though some give and take may be needed.

  
  

• Beta or pilot testing: If possible, it is helpful to perform a beta or pilot test of the DBM before committing to a clinical trial. This could be as rudimentary as running a hands-on demo of the technology, or as involved as designing and implementing a full study focused around it.

  
  

• Configuration testing: Unless the DBM is a fully off-the-shelf solution with no configuration or customization, it is critical to test out the system once configured for the trial. This should be risk-based, considering anything that was modified or implemented specific to the study and should also cover as many scenarios as possible.

Operationalization

It is also critical to establish the right methods and measures of operational success. Digital biomarkers may hold a lot of promise in delivering high value data for a given trial, but they still need to function, still need to be used (by patients) and still need to deliver the data for analysis.

Here are some factors that need to be considered when deploying digital biomarkers:

• Security and compliance: Is the DBM adherent to HIPAA, GDPR, 21CRF part 11? What are the risks around patient privacy and data security?

  
  

• Usability: Is the technology intuitive? Is it usable considering the challenges of your patients' disease (fine motor requirements, cognitive burden...)?

  
  

• Logistics: How involved is setup as well as certain administrative tasks, and can that be performed by site staff? Is a Wi-Fi connection or cell signal needed for use? What are the charging requirements?

  
  

• Data quality: Is there straightforward tracking or monitoring of data quality? How can data issues, e.g. lost in transmission or missing due to non-use, be addressed?

  
  

• Data delivery: Does the vendor have a strong data delivery process with the ability to output formatted datasets on a required schedule?

Current FDA thoughts

One goal of the FDA is to define "digital biomarkers" correctly. Terminology matters, so it is important to clarify the difference between digital health technology (DHT) vs. digital biomarker (DBM) vs. electronic clinical outcome assessment (eCOA). As a subset of digital health technologies, digital biomarkers are considered by the FDA to be "a defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes, or biological responses to an exposure or intervention, including therapeutic interventions”. A biological measurement and not a patient reported data point, if you will.

The FDA is a proponent of using digital health technologies in clinical trials to reduce patient burden as well as improve convenience and access. However, the encouragement of using DBMs seems less of a focus and still requires a number of hurdles to jump through.

The implementation of the MDDT program is meant to support faster, easier approval of DBMs in support of clinical trials. Vendors will need take on the responsibility of providing evidence of the tool's claims, which may also require investing in studies that show such evidence. Even so, it could pay dividends if the DBM becomes an FDA approved way to measure a certain disease. Perhaps in the vein of Pharma-Biotech partnerships for drug development, a Pharma-vendor partnership model for DBMs would also help move the needle faster. Of course the details like royalties or other incentives would need to be worked out.

As for FDA approved DBMs, the approval of Apple watch's AFIB software as a biomarker for use in clinical trials is a big step forward. Although, it appears to be 1 of 2 approvals at this point.

Summary

Digital biomarkers are exciting, and have a lot of promise. It is important to ensure they have scientific value, are validated from a software and hardware standpoint, fit for purpose, and operationally sound. And while the FDA may find ways to fast track DBM approvals, there is a not an insignificant burden of evidence and approval that need to be met.

As a parting thought: The more Sponsors utilize DBMs in their trials, even as exploratory, the more data and evidence can be gathered in support of the DBM, as well as more exposure for the use of DBMs in general. If we increase the use of DBMs, the FDA will increase focus on supporting them!

Back to table of contents.