Abstract
Left ventricular ejection fraction (EF) and global longitudinal strain (GLS) are widely used for cardiac surveillance in patients receiving potentially cardiotoxic cancer therapy. While GLS is more sensitive for detecting early myocardial dysfunction, the relationship between EF and GLS in routine clinical practice remains incompletely characterized, and discordance between these parameters may complicate clinical interpretation. This retrospective observational study included 117 oncology patients undergoing serial echocardiographic assessment at baseline, early follow-up (6-12 weeks), and short-term follow-up (3-6 months). EF was measured by biplane Simpson's method, and GLS by speckle-tracking echocardiography as absolute values. EF deterioration was defined as a ≥ 10% absolute reduction or EF < 53%, while GLS deterioration was defined as a ≥ 15% relative reduction in absolute GLS magnitude. Concordance between EF and GLS deterioration was assessed, using Cohen's kappa and McNemar's test. Mean EF and GLS values remained stable (EF p = 0.124; GLS p = 0.803). EF deterioration occurred in 32 patients (27.4%), and GLS deterioration in 25 patients (21.4%). Discordance was common, with isolated GLS deterioration in 17 patients (14.5%), and EF deterioration without GLS decline in 24 patients (20.5%), representing discordance in 41% of patients. Agreement was poor (κ = 0.05), and McNemar's test was not significant (p = 0.27). ΔGLS showed a weak association with ΔEF (r = 0.19, p = 0.036). Although systolic function remained stable overall, clinically relevant deterioration occurred in a subset. Frequent EF-GLS discordance highlights their complementary but non-interchangeable roles in cardio-oncology surveillance.