Interaction of Galectin-3 Concentrations with the Treatment Effects of β-Blockers and RAS Blockade in Patients with Systolic Heart Failure: A Derivation-Validation Study from TIME-CHF and GISSI-HF

BACKGROUND
Galectin-3 predicts prognosis in heart failure (HF) and may help to select HF patients in need of intensified therapy.

METHODS
This retrospective post hoc analysis included 219 patients from the Trial of Intensified versus Standard Medical Therapy in Elderly Patients with Congestive Heart Failure (TIME-HF) and 631 patients from Gruppo Italiano per lo Studio della Sopravvivenza nell'Insufficienza Cardiaca (GISSI-HF) with HF who had reduced ejection fraction and available galectin-3 plasma concentrations. The interaction between galectin-3, β-blockers, renin-angiotensin system (RAS) blockade, and spironolactone on outcome was evaluated in TIME-CHF and validated in GISSI-HF. End points were all-cause mortality and the composite of mortality with HF hospitalization or any hospitalization.

RESULTS
High galectin-3 concentrations were associated with adverse outcome in both cohorts and remained significantly associated with death after multivariate adjustment [hazard ratio 2.42 (95% CI 1.17-5.01), P = 0.02, in TIME-CHF; 1.47 (1.02-2.10), P = 0.04, in GISSI-HF). In TIME-CHF, patients with low galectin-3 plasma concentrations had a better prognosis when β-blockers were up-titrated, whereas patients with high galectin-3 plasma concentrations did not (interaction P < 0.05 for mortality and death with or without hospitalization). Opposite trends were seen for RAS blockade but were not statistically significant. Patients with high galectin-3 plasma concentrations had neutral prognosis when receiving spironolactone, whereas patients with low galectin-3 plasma concentrations had worse prognosis when receiving spironolactone (interaction P < 0.10 for death with or without hospitalization). In the GISSI-HF validation cohort, these interactions were confirmed for β-blockers (P < 0.05 for all end points) and consistent for RAS blockade (P < 0.10 for death with or without hospitalization), but inconsistent for spironolactone.