Association of galectin-3 with echocardiographic parameters in acute COVID-19 heart failure phenotypes and a post-acute comparator cohort

Background: Galectin-3 is involved in inflammatory and fibrotic pathways and may reflect myocardial remodeling in COVID-19. Its relationship with echocardiographic findings across heart failure phenotypes during acute disease and the post-acute period remains incompletely defined. This study aimed to examine associations between circulating galectin-3 and echocardiographic parameters in acute COVID-19 heart failure phenotypes and in a post-acute comparator group.

Methods: This single-center retrospective observational study included 86 participants: 61 adults assessed during hospitalization for acute COVID-19 and 25 adults assessed approximately 2 months after infection. The acute-phase participants were classified as HFrEF (n=6), HFmrEF (n=20), or HFpEF (n=35); the post-acute participants formed a separate comparator group (n=25). Between-group comparisons were recalculated from reported group sample sizes, means, and standard deviations using Welch analysis of variance with Games-Howell post hoc tests. Associations were summarized using reported Spearman correlation coefficients.

Results: Galectin-3 differed among the four analytical groups (Welch p<0.001). Games-Howell pairwise differences were observed between HFrEF and HFpEF (p=0.027), HFrEF and the post-acute comparator (p=0.009), HFmrEF and HFpEF (p=0.019), and HFmrEF and the post-acute comparator (p<0.001). In HFpEF, galectin-3 correlated with left atrial volume (rho=0.538, p<0.001) and left atrial volume index (rho=0.532, p=0.001). In the post-acute comparator, galectin-3 correlated inversely with LVEF (rho=-0.501, p=0.011) and positively with E/e’ (rho=0.525, p=0.007) and LVESV (rho=0.535, p=0.006). Large correlations in HFrEF were based on only six participants and were considered unstable.

Conclusions: Galectin-3 was associated with selected echocardiographic measures in this exploratory cohort. The small, uneven groups, absence of a robust adjusted model, and lack of longitudinal outcomes preclude conclusions regarding independent, prognostic, or phenotype-specific clinical utility.