Efficacy of dapagliflozin according to geographic regions in patients with heart failure: a patient-level pooled analysis of DAPA-HF and DELIVER

Abstract

Abstract Background Clinical characteristics and prognosis vary by geographic region in patients with heart failure (HF). Some previous reports suggested the efficacy of certain treatments for HF may be modified by region. However, most analyses based on regions in individual trials are hampered by modest subgroup sizes and small numbers of events and can give unreliable results. Purpose To evaluate the difference in outcomes between regions, and to examine whether the efficacy and safety of dapagliflozin in HF were consistent across regions, using the patient-level pooled data of DAPA-HF and DELIVER. Methods The DAPA-HF (n = 4744) and DELIVER (n = 6263) trials enrolled patients with HF, functional limitation, and elevated natriuretic peptide. The main difference between the trials was that patients with a left ventricular ejection fraction (LVEF) ≤40% were randomized in DAPA-HF and those with a LVEF >40% in DELIVER. In both trials, patients were randomized to dapagliflozin 10 mg daily, or placebo, in addition to standard care. The primary outcome of both trials was the composite of worsening HF or cardiovascular death. Results Of the total 11,007 patients, 5159 patients (46.9%) were enrolled in Europe; 1528 (13.9%) in North America; 1998 (18.2%) in South America; and 2322 (21.1%) in Asia. Patients were younger in Asia and South America and were more often female in South America and Europe. The mean baseline LVEF and median natriuretic peptide level did not vary significantly across regions. Mineralocorticoid receptor antagonists were prescribed less frequently in patients with HF and reduced ejection fraction in North America (46.4%) compared to other regions (70-80%). The rate of the primary outcome (per 100 person-years) was higher in North America [13.9 (95%CI 12.5-15.4)] than in other regions: 10.8 (10.1-11.5) in Europe, 10.0 (9.0-11.1) in South America, and 10.5 (9.5-11.5) in Asia. The rate of the primary outcome increased with decreasing LVEF similarly across all regions studied (Figure 1). The benefit of dapagliflozin on the primary outcome was not modified by region: hazard ratio for dapagliflozin versus placebo 0.85 (95%CI 0.75-0.96) in Europe, 0.75 (0.61-0.93) in North America, 0.72 (0.58-0.89) in South America, and 0.74 (0.61-0.91) in Asia (p-interaction = 0.40). This was the same when evaluated separately for patients with LVEF ≤40% (p-interaction = 0.39) and for those with LVEF >40% (p-interaction = 0.84). In each region, the effect of dapagliflozin was consistent regardless of LVEF (Figure 2). Patients in North America discontinued randomized treatment more frequently than anywhere else (in the placebo group, 21.8% in North America versus 6.4% in South America), but the discontinuation rates did not differ between placebo and dapagliflozin by region. Conclusions The efficacy and safety of dapagliflozin were consistent across global regions despite geographic differences in patient characteristics, background treatment, and event rates.Incidence rate according to regionDapagliflozin effect according to region