Innovate to Elevate Cardiac Surveillance in Oncology Care

Abstract

Introduction
Osimertinib and alectinib are tyrosine kinase inhibitors (TKIs) used for EGFR- and ALK-mutated non-small cell lung cancer (NSCLC), respectively. While these agents have substantially improved survival in early-stage and metastatic settings, they are associated with clinically relevant cardiac toxicities, QTc prolongation for osimertinib and sinus bradycardia for alectinib. Current guidelines recommend electrocardiographic (ECG) monitoring, yet conventional 12-lead ECGs are impractical for frequent use in ambulatory oncology care. Medically regulated single-lead wearable ECG devices offer a promising alternative for real-time, patient-friendly rhythm surveillance.
Objectives: This thesis aimed to evaluate the clinical performance, reproducibility, and diagnostic concordance of the single-lead ECG function of the Corsano CardioWatch in NSCLC patients treated with TKIs. Specifically, intra-device reproducibility, intra-device agreement with conventional 12-lead ECGs, and diagnostic classification.

Methods
This study was part of the prospective, non-WMO MOVIS-RESP trial conducted at Erasmus MC, including NSCLC patients treated with osimertinib (n = 39) or alectinib (n = 32). Each patient received three consecutive single-lead ECGs with the Corsano CardioWatch and one conventional 12-lead ECG. Reproducibility of ECG parameters was assessed using coefficients of variation (CV), repeated measures analysis, and intra-class correlation coefficients (ICCs). Agreement between devices was evaluated using Bland–Altman plots, ICCs, and correlation analyses. Diagnostic performance was determined via conventional metrics.

Results
Of 213 Corsano ECGs, 96% met signal quality thresholds. Corsano parameter reproducibility was ‘excellent’ for HR, RR, and QTc (ICCs ≥ 0.91), ‘good’ for QT and PR (ICCs 0.78–0.82), and ‘moderate’ for QRS duration (ICC 0.63). Corsano with EMC ECG agreement was ‘excellent’ for HR and QTc (ICCs ≥ 0.85), ‘good’ for QT and RR (ICCs ~0.80), ‘poor-to-moderate’ for PR and QRS (ICCs ≤ 0.47). Bland–Altman analysis showed ‘low’ bias for HR, RR, and QT intervals and wider limits of agreement for PR and QRS durations. The CardioWatch accurately classified rhythm abnormalities, achieving 97% overall accuracy and κ = 0.93 for detecting total abnormal rhythms. Sinus bradycardia (κ = 0.94) and abnormal sinus rhythm (κ = 0.88) were reliably detected.

Discussion
The Corsano CardioWatch was well tolerated and enabled rapid ECG acquisition in seated patients, which benefits those with mobility impairments. Its integrated digital platform supported stable data transmission and clinician access. While wearable ECGs are economically favourable and patient-preferred, concerns remain regarding potential false positives, digital literacy, and algorithm transparency. Transitioning to home use introduces further considerations, including adherence, remote triage protocols, and device accessibility. Structured onboarding, patient support, and robust telecardiology systems will be essential for safe deployment.

Conclusion
The Corsano CardioWatch showed strong intra-device parameter reproducibility and high diagnostic concordance for common arrhythmias. Agreement with 12-lead ECGs was excellent for HR and QTc but limited for PR and QRS intervals, aligning with known constraints of single-lead ECGs. The device shows promise as a user-friendly, medically regulated tool for remote ECG monitoring in oncology. Future studies should assess long-term performance in unsupervised, home-based contexts and determine clinical impact on patient safety, treatment continuity, and healthcare resources.