Main Session
Sep
28
PQA 03 - Digital Health Innovation and Informatics, Patient Safety & Quality, and Radiation and Cancer Biology
2580 - BEACON (IRB 20-330): A Prospective, Institution-Wide Digital Phenotyping Registry to Derive Scalable Physiologic Biomarkers for Clinical and CNS Progression Events
Presenter(s)
Luke Pike, MD, DPhil - Memorial Sloan Kettering Cancer Center, New York, New York
L. R. G. Pike1, C. Pryce1, K. Tsai2, T. Erazo1, L. Nystrom1, E. Carbone1, W. Underwood1, C. B. Jackson3, J. Jee1, L. Jones4, C. Fong1, and J. Scott1; 1Memorial Sloan Kettering Cancer Center, New York, NY, 2Carle Illinois College of Medicine, Urbana, IL, 3Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, 4City of Hope, Duartes, CA
Purpose/Objective(s): Objective, high-resolution measures of functional status and physiologic reserve is lacking in oncologic care. Subtle signs of clinical deterioration—such as changes in mobility, sleep, or autonomic function—may precede radiographic or symptomatic findings, especially in neurologic decline due to CNS progression. Current assessments like ECOG/KPS are episodic and prone to inter-observer variability. BEACON (Biometrics-Enabled Analytics for Cancer Observation and Notification) is a prospective, institution-wide registry that collects longitudinal physiologic data from consumer-grade digital devices, such as the Apple Health App, linked to specific clinical outcomes. Its primary aim is to develop and validate scalable digital biomarkers that act as quantitative proxies for performance status and detect physiologic signatures that precede significant oncologic events, including progression or early diagnosis of brain tumors.
Materials/Methods:
BEACON is a minimal-risk, prospective observational registry of adults with and without cancer at MSK. Participants contribute passively collected physiologic data, which are harmonized within a secure institutional infrastructure and linked to EHR-derived outcomes. Participants with an iPhone consent to longitudinal sharing of Apple HealthKit data, including activity metrics (step count, distance, walking speed/step length when available), sleep parameters, heart rate, and heart rate variability (HRV), contingent upon linked wearable devices. Historical and prospective data are ingested via secure APIs and time-aligned with events. The primary endpoints of the study include the feasibility of large-scale longitudinal physiologic data acquisition in oncology populations, adherence and data completeness (frequency, continuity, and durability of data transmission), and data integrity and analytic usability. Secondary/exploratory endpoints include: development of candidate digital biomarkers of functional status and physiologic reserve, identification of time-dependent physiologic signatures preceding key clinical events (hospitalization, treatment toxicity, systemic or CNS progression), and correlation of digital phenotypes with survival and patient-reported outcomes. Eligible participants are adults aged 18–90 years with a solid malignancy (broad tumor cohorts) or no history of cancer (control cohort). HealthKit participants must possess a compatible iPhone.Results:
232 of 3644 patients are enrolled Conclusion: BEACON establishes scalable digital phenotyping infrastructure within a comprehensive cancer center to enable rigorous development of quantitative physiologic biomarkers. This platform is designed to support future CNS-focused analyses aimed at detecting pre-radiographic or pre-symptomatic signatures of neurologic decline and brain metastasis progression, with potential implications for surveillance strategies and early intervention paradigms.