Galectin-3-enhanced prognostic models for major adverse cardiovascular events in Vietnamese patients with acute coronary syndrome: A prospective cohort study.

ObjectiveWe evaluated whether admission Galectin-3 (Gal-3) was associated with higher apparent discrimination and reclassification when added to the TIMI, HEART and GRACE scores for 12-month major adverse cardiovascular events (MACE) in Vietnamese patients with acute coronary syndrome (ACS).MethodsThis prospective observational cohort included 103 adults with ACS diagnosed according to the 2023 European Society of Cardiology criteria between August 2024 and April 2025. Patients were followed for 12 months, with final follow-up completed in April 2026. Serum Gal-3 was measured at admission. The primary endpoint was MACE, comprising all-cause death, nonfatal myocardial infarction, nonfatal stroke, unplanned hospitalization for heart failure or unplanned coronary revascularization. Discrimination of the TIMI, HEART and GRACE scores, with and without Gal-3, was assessed by area under the receiver operating characteristic curve (AUC). Reclassification was evaluated using continuous net reclassification improvement (NRI) and integrated discrimination improvement (IDI), with 95% confidence intervals estimated from 1,000 bootstrap resamples.ResultsDuring follow-up, 16 patients (15.5%) experienced MACE. Admission Gal-3 was higher in patients with MACE than in those without MACE (median 28.7 vs 15.4 ng/mL) and the adjusted median difference remained significant (p=0.034). Adding Gal-3 yielded higher apparent AUCs for TIMI (0.584 vs 0.804), HEART (0.580 vs 0.802) and GRACE (0.620 vs 0.797), with optimism-corrected AUCs of 0.785, 0.789 and 0.783, respectively. Given the limited number of outcome events, these estimates remain exploratory and potentially optimistic.ConclusionsHigher admission Gal-3 levels were associated with 12-month MACE. Adding Gal-3 to established risk scores yielded higher apparent discrimination and secondary exploratory reclassification metrics; however, the magnitude of these differences remains uncertain and may still be optimistic despite bootstrap correction. These analyses should be regarded as hypothesis-generating rather than as the development of candidate clinical prediction models, and external validation in larger multicenter cohorts is required.