Abstract 4138835: Efficacy and Safety of Sodium-glucose Transporter 2 Inhibitors in Acute Decompensated Heart Failure Patients Concomitant with Physical or Cognitive Frailty: Finding from the WET-HF Registry

Background: Recent evidence suggests the efficacy and safety of sodium-glucose transporter 2 inhibitors (SGLT-2is) in heart failure (HF) patients. However, there still remain concerns in their use among patients with physical or cognitive frailty. Methods: West Tokyo Heart Failure (WET-HF) Registry is an ongoing prospective multicenter registry in Japan enrolling all patients who were hospitalized for acute decompensated heart failure. We retrospectively analyzed the impact of physical or cognitive frailty-related factors (FF) on the association of SGLT-2i prescription with long-term outcomes using the data of WET-HF (2018-23, n=5106). Patients without data of SGLT-2i prescription at discharge were excluded. Primary endpoint (PE) was defined as the composite of cardiac death and HF rehospitalization during 2 years after discharge. Results: At baseline 3208 of 4845 patients (66%) had FF defined by any of following 5 factors: BMI <18.5 kg/m 2 , Barthel index ≤85, clinical frailty scale ≥4, geriatric nutritional risk index (GNRI) <82, and a history of dementia. The study population was divided by prescription of SGLT-2is at discharge. SGLT-2i users (n=737) showed younger age (73 vs 80 years, P<0.001), higher number of male sex (67% vs. 57%, P<0.001), higher prevalence of HF with reduced ejection fraction (55% vs. 40%, P<0.001), and diabetes mellitus (63% vs. 26%, P<0.001) compared to SGLT-2i non-users. This trend was similarly seen in FF subgroup. SGLT-2i users showed significantly lower rate of PE compared to non-users (19% vs 23%, P=0.008, log-rank test). The results were directionally similar in the patients with FF (20% vs. 26%, P=0.006), but in those without FF there was no difference in PE between SGLT-2i users and non-users (17% vs. 16%, P=0.96). These associations persisted even after adjusting for patient background with a propensity score-based inverse probability of treatment weighting for SGLT-2i prescription (FF, HR 0.59 [0.39 – 0.91], P=0.016; No FF, HR 0.75 [0.47 – 1.20], P=0.23, P for interaction 0.48). At 1 year after discharge (follow up rate 61%), there was no difference in SGLT-2i continuation rate between FF and No FF groups (81% vs. 85%, P=0.14). The patients with FF who continued SGLT-2i showed no significant temporal changes in BMI, GNRI, or eGFR, whereas those without SGLT-2i showed significant decline in eGFR (P<0.001). Conclusions: SGLT-2is was safe and efficacious even for HF patients with FF in our retrospective analysis.