Novartis Makes Advancements in Digital Biomarkers for Knee Osteoarthritis

The 7th Annual Digital Biomarkers in Clinical Trials Summit recently featured a compelling presentation by William Johnston, Digital Endpoint Lead at Novartis. Johnston’s talk focused on the innovative strides his team is making in knee osteoarthritis (OA) clinical trials, particularly in evaluating physical function using digital health technologies. His insights provided a comprehensive look at the challenges, innovative approaches, and future directions in this critical area of medical research.

The Challenge of Knee Osteoarthritis

Knee osteoarthritis is a progressive and debilitating condition that significantly impacts the quality of life for older adults. Johnston highlighted the grim reality that there are currently no disease-modifying drugs available, leaving patients to rely on symptom management. This often leads to severe cases where 5-10% of patients require total knee replacements, with a substantial 20% of these surgeries being unsuccessful. Johnston elaborated on the symptoms of knee osteoarthritis, which include pain, swelling, stiffness, and crepitus—a grinding or clicking sensation within the joint. These symptoms severely limit mobility, making it difficult for patients to engage in everyday activities such as walking or physical exercise. He cited a systematic review of 19 qualitative studies and the FDA’s Voice of the Patient work, highlighting mobility impairment’s profound impact on patients’ lives.

To illustrate, Johnston shared a poignant example from the FDA’s Voice of the Patient series. A patient had to retire at 59 and found it impossible to keep up with friends during walks. This individual eventually joined a social group of octogenarians with walking frames, as they were the only group with whom they could keep pace. This story vividly demonstrates how knee osteoarthritis can drastically alter one’s social life, mental health, and overall well-being. The lack of effective treatments that address the root cause of the disease further exacerbates these challenges, highlighting the urgent need for innovative solutions in this field.

Novartis’ Innovative Approach

Novartis is working on a promising compound, LNA043, to regenerate damaged cartilage and halt disease progression. Johnston explained that the goal is to develop a disease-modifying osteoarthritis drug (DMOAD) that can intervene early and effectively in the disease process. LNA043 targets damaged cartilage and modulates several biological pathways to encourage cartilage regeneration. Johnston emphasized that this approach aims to manage symptoms and fundamentally alter the disease’s trajectory. This is a significant departure from current treatments, which primarily focus on alleviating pain and other symptoms without addressing the underlying disease.

Johnston also discussed the broader implications of developing a DMOAD. A successful disease-modifying treatment could reduce the number of patients requiring total knee replacements, decreasing the associated healthcare costs and improving long-term outcomes. He noted that the ongoing Phase IIB trial for LNA043 is a critical step in this direction, with the primary endpoint set at two years to evaluate the compound’s efficacy and safety thoroughly. This trial represents a significant investment in the future of knee osteoarthritis treatment, aiming to provide patients with a more effective and sustainable solution.

The Role of Digital Health Technologies

Johnston emphasized the limitations of traditional clinical trial measures, such as patient-reported outcomes (PROs) and point-in-time performance tests like the six-minute walk test. These methods often fail to capture the longitudinal and nuanced aspects of physical activity and mobility. This is where digital health technologies, particularly wrist accelerometry, come into play. Traditional PROs, like the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), primarily focus on pain during specific functional tasks and do not adequately measure overall physical activity. Similarly, point-in-time tests like the six-minute walk test assess functional performance simultaneously, failing to provide a comprehensive picture of a patient’s daily activity levels.

Johnston argued that digital health technologies offer a more holistic and continuous physical activity assessment. Wrist accelerometry, for example, allows for the longitudinal measurement of mobility and activity levels. This technology can capture daily fluctuations and accurately represent a patient’s physical function over time. He also addressed a common misconception: simply giving patients a device and analyzing the data will yield meaningful insights. Johnston stressed that this approach is far from straightforward. It requires careful consideration of several fundamental questions, such as the minimum amount of data needed for a valid measure and how to account for natural variability in daily activity levels.

Key Findings

• Daily Wear Time: The team found that 90% of patients’ activity occurs between 6:00 AM and 9:00 PM. They determined that a minimum of 12 hours of wear within this window is necessary to obtain a representative measure of daily activity. Johnston provided detailed insights into how they arrived at this threshold. By analyzing the activity distribution throughout the day, they observed that only about 5% of activity occurs before 6:00 AM and 5% after 9:00 PM. This pattern is held on weekdays and weekends, although weekend activity starts slightly later. This consistency allowed them to confidently set the 12-hour minimum wear time as a criterion for valid data.
• Number of Days: By analyzing data over two-week periods, they concluded that six days of data collection provide a stable measure of physical activity with good test-retest reliability. Johnston explained that they used a methodical approach to determine the optimal number of days. They aggregated data from day one and day two in one week and day one and day two in the second week, continuing this process for up to 14 days. They then assessed the test-retest reliability at each step. The results showed that six days of data collection were sufficient to achieve a reliability threshold of 0.75 for most physical activity variables. This finding is consistent with published data in other areas, such as COPD, where a minimum of four days with eight hours of wear time is used.
• Data Variability: Johnston illustrated how applying these criteria significantly reduces within-subject variability, thereby enhancing the reliability of the data. He presented a hypothetical example to demonstrate the impact of these criteria. Initially, the data showed no change in a patient’s average daily step count due to high variability. By applying the 12-hour wear time and six-day data collection criteria, they reduced this variability and revealed meaningful changes in activity levels. This approach not only improves the reliability of the data but also enhances the sensitivity of the measures to detect changes over time.

Future Directions

Johnston outlined several future steps, including sensitivity analyses to assess the impact of these criteria on sensitivity to change, exploring the sufficiency of non-consecutive days, and further establishing the tool’s psychometric properties. He emphasized the importance of conducting sensitivity analyses to understand how these criteria affect the ability to detect changes in physical activity over time. This is particularly relevant for evaluating the efficacy of interventions like LNA043. By examining different scenarios, such as varying the number of consecutive days or including non-consecutive days, they aim to refine their approach and ensure robust data collection.

Johnston also highlighted the need to explore the impact of weekend versus weekday activity. While their initial analysis showed minimal differences, further investigation could reveal important nuances affecting the overall physical activity assessment. Additionally, they plan to establish the tool’s psychometric properties, such as its validity, reliability, and responsiveness, to ensure it meets the rigorous standards required for clinical trials. These future directions aim to enhance the robustness and applicability of digital health technologies in clinical research.

If you want to access the complete Digital Biomarker Summit event recordings, click here.