Digital biomarkers are changing the face of clinical research.

Fueled by rich, real-world data, often collected by connected devices, digital biomarker development is revolutionizing our measurements of health and accelerating a move towards a patient-centered, preventive healthcare system.  

In the last 10 years, digital biomarkers have received significant attention, as a non-invasive marker of health and disease. In fact, Christian Gossens, Head of Digital Biomarkers at Roche Pharma Research & Early Development says it best, “digital biomarkers are changing how future medicines will be developed and could lead to more personalized treatments. This will transform the lives of patients.” 

Indeed, the benefits of digital biomarker development do not go unseen. They provide crucial information on disease management and treatment and can quantify the subtle changes in symptoms linked to early progression of a disease. For example, researchers could be able to measure the changes in gait and spatial memory to indicate the onset or worsening of neurodegenerative conditions. Beyond a purely clinical setting, digital biomarker development also accelerates label expansion, captures unmet patient needs for regulatory approval, and facilitates the development of Software as a Medical Device (SaMD) with its predictive and diagnostic capabilities. 

Let’s take Parkinson’s Disease, for example. A neurodegenerative disease known for motor impairments, tremors, dyskinesia, and gait abnormalities, which touches the lives of nearly 10 million people worldwide. There is still no approved disease-modifying drug treatment, and the medications that do exist cost an estimated $2,500 a year. In phase three clinical trials, the development of Parkinson’s Disease drugs are difficult and often end in trial failure due to the subjective, sparse data. Digital biomarker development would improve evaluation of a PD drug’s efficacy or track changes in gait and speech to pinpoint the early onset of the disease. In fact, recent progress has been made with an FDA approval of software that uses the Apple's Movement Disorder API on Apple Watch’s to power a data-driven approach to measuring and recording tremors and dyskinetic symptoms common in patients with Parkinson's disease.

Additionally, diagnostic digital biomarker development allows researchers to confirm the presence of disease or a subtype of disease, such as the identification of atrial fibrillation for cardiovascular disease. Using computational intelligence and incorporating wearables, smartphones, and connected devices into clinical trials, allows researchers to better capture real-world lived experiences longitudinally. Simply put, digital biomarker development is at the frontier of precision medicine. 

However, their development is not a simple, one size fits all solution; the process of transforming data from wearables and other sensory devices into digital biomarkers holds many challenges: lack of regulatory oversight, data ownership issues, limited funding, and insufficient clinical validation and standardization methods. To successfully develop them, researchers must prioritize the collection of high quality, dense, real-world data to analyze and discover novel health indicators.

 

What are the challenges in digital biomarker development?

Digital biomarker development is an expensive endeavor for a company to undertake, despite the obvious opportunity the domain presents. Digital biomarker development is often expensive, requiring the stitching together of a patient-generated and other real-world datasets to receive a strong signal from the data. Further, the very inception of a digital biomarker requires significant exploration into traditional, or “wet” biomarkers to understand the potential for the digitization of biomarkers. Creating protocols that accurately translate the literature into a meaningful data collection process across a diverse patient population is a challenge.

Building a digital biomarker must also take critical market research into account: for example, how does a company balance between building a reliable predictive model of a condition, while also ensuring that they are incorporating data that would be realistic to collect in a patient’s everyday life in a post-approval scenario? While the FDA has shown interest in easing and advancing their perspective on artificial intelligence and approval of Software-as-a-Medical-Device, we are arguably still in the early days of the digital biomarker industry as it were. The FDA has a high threshold for digital biomarkers, which would require significant investment from researchers and institutions to meet.

With richer data and artificial intelligence methods, researchers have better opportunities to develop enhanced algorithms for personalized treatments. We are empowering any researcher, whether from a small startup to a Fortune 500 company, to become self-sufficient, minimize the resources required, and take advantage of health innovation. Learn more about how our platform can help your company meet these challenges at https://smartomix.com/digital-biomarkers.

Submit a comment

You may also like

doc.ai and UCB biopharma team up to develop breakthrough AI models to advance care
doc.ai and UCB biopharma team up to develop breakthrough AI models to advance care
19 November, 2020

Today we’re excited to announce a new collaboration with UCB, a multinational biopharma company headquartered in Brussel...

doc.ai to be acquired by Sharecare
doc.ai to be acquired by Sharecare
28 January, 2021

Today, I’m excited to share that doc.ai has entered into a definitive agreement to be acquired by Sharecare, the digital...

PGHD for Digital Phenotyping and Digital Biomarker Development
PGHD for Digital Phenotyping and Digital Biomarker Development
22 June, 2022

Last month, UCB, held its first Digital Health Roundtable on Leveraging Patient-Generated Health Data for Digital Phenot...