Characterization of a human wearable ECG device for continuous data collection in freely moving non-rodents

Abstract

Wearable devices are commonly used in clinical diagnostic medicine/personal health, but their use to collect nonclinical endpoints is limited due species-specific design specifications, data accessibility limitations, and availability of nonclinical telemetry technologies. These create a challenge to adopt and deploy clinical innovation in nonclinical studies. To determine the feasibility of using a human ECG wearable device (WRB) in nonclinical studies, we compared heart rate and ECG intervals between the simultaneously collected ECG from the WRB and implanted telemetry (TEL) device in canine (n = 5) and non-human primate (NHP, n = 4). Continuous ECGs were collected simultaneously from both devices for 24 h prior and 24 h after oral administration of vehicle or Dofetilide (canine n = 5; 0.003, 0.010, 0.030 mg/kg and NHP n = 4; 0.03, 0.06, 0.12 mg/kg). Individual animal 15-min means were reviewed descriptively and via Bland-Altman (BA) analysis to determine bias and limits of agreement (LOA) between devices. Further, Dofetilide dependent QT changes were evaluated using time-matched QT interval comparisons and the results of QTci exposure-response modeling. Data were qualitatively consistent in magnitude and profile between the two measurement platforms. BA demonstrated that bias between devices was <2 ms for HR, PR, and QRS for both species, 2 ms for canine QT, and 6.6 ms for NHP QT, with acceptable LOA for all endpoints. Mean Dofetilide-dependent QT changes over time were not different between devices and calculated Δ∆ QTci EC_+10ms were < 2× between devices within species. Overall, these studies demonstrated the feasibility of using alternative devices and increasing options to collect in-cage ECG to collect nonclinical safety pharmacology and toxicology endpoints.

Keywords: ECG intervals; Electrocardiogram; Safety pharmacology; Telemetry, noninvasive; Toxicology.