ABSTRACT
AIMS: Subclinical left ventricular dysfunction (LVD) is a prelude to symptomatic heart failure (HF). We hypothesised that screening-guided treatment with spironolactone would prevent incident HF in at-risk patients. METHODS AND RESULTS: We randomised asymptomatic, community-dwelling subjects aged ≥65 years old, with at least one non-ischaemic HF risk factor (hypertension, type 2 diabetes mellitus or obesity) to echocardiography-guided therapy or usual care. Spironolactone therapy was triggered by the detection of subclinical LVD (global longitudinal strain [GLS] ≤16%) or diastolic abnormalities (at least one of E/e' >15, E/e' >10 with left atrial enlargement [LAE] or impaired relaxation [E/A < 0.8, IR], LAE with IR), or borderline GLS (17%) with IR or borderline GLS with LAE. The primary outcome was incident HF at 24 months. Secondary outcomes were change in 6-min walk test (6MWT) and change in left ventricular function. LVD was identified in 161 (46%) of 349 participants (age 70 [68-73] years, 201 [58%] women). The trial was stopped because of a 55% rate of spironolactone discontinuation, due primarily to decline in renal function. Incident HF developed in 11 (3.5%) of 310 participants completing follow-up, with no difference between usual care and intervention (4 [2.5%] vs. 7 [4.7%], p = 0.29), decline in 6MWT distance (p = 0.28), persistent or new LVD (p = 0.58), nor change in GLS with intervention (p = 0.15). A per-protocol analysis of 131 patients with baseline LVD and a follow-up echocardiogram, showed resolution of LVD with spironolactone therapy (59% vs. 33%, p = 0.01). CONCLUSION: The study was underpowered to determine whether screening-guided spironolactone therapy reduced incident HF because spironolactone was frequently discontinued due to renal function criteria. However, LVD resolved in more patients treated with spironolactone than in untreated patients. Future trials should use less conservative renal criteria for spironolactone discontinuation.
