Effect of dapagliflozin on left ventricular mass in patients with kidney failure receiving maintenance hemodialysis: Primary results from the DAPA-HD trial

Background Sodium–glucose co-transporter 2 inhibitors (SGLT2i) reduce cardiovascular events across a broad spectrum of kidney function, including advanced chronic kidney disease. However, whether their cardiovascular benefits persist in patients receiving dialysis and whether they depend on residual kidney function remain unknown. Purpose The Dapagliflozin in Hemodialysis (DAPA-HD) trial was designed to evaluate the effect of dapagliflozin on cardiac structure and function in patients with kidney failure receiving maintenance hemodialysis, with a prespecified interaction analysis stratified by residual urine output as a pragmatic marker of residual kidney function. Methods This academic, multicenter, randomized, double-blind, placebo-controlled trial enrolled patients with kidney failure on maintenance hemodialysis and left ventricular hypertrophy. Participants were randomized 1:1 to receive dapagliflozin 10 mg once daily or placebo for 6 months, stratified by residual urine volume (>200 mL/24h vs. ≤200 mL/24h). The primary endpoint was the change in left ventricular mass index (LVMI) adjusted for body surface area, measured by echocardiography performed immediately after dialysis and analyzed using ANCOVA adjusted for baseline LVMI. A prespecified interaction term tested effect modification by residual urine output. Results In total, 220 patients were randomized (65 yrs, 26% female, 42% T2DM, LVEF 57%, 28% anuric). Baseline echocardiography suitable for LV mass assessment was available in 210 patients, and 142 had evaluable 6-month echocardiographic data. Baseline LVMI was similar between groups (dapagliflozin: 136 ±35 g/m²; placebo: 140 ±38 g/m²). At 6 months, LVMI increased by 4.5 ±25.0 g/m² in the dapagliflozin group and by 8.0 ±20.6 g/m² in the placebo group. After adjusting for baseline LVMI, no significant difference was observed between groups (adjusted mean difference: -3.7 g/m²; 95% CI -11.2 to 3.8, P=0.33). However, in patients with urine output >200 ml, dapagliflozin was associated with lower LVMI at 6 months compared with placebo (adjusted mean difference −9.2 g/m², P=0.035), whereas no significant difference was observed in anuric patients (adjusted mean difference 9.0 g/m², P=0.19; P-value for interaction 0.025). During follow-up, death occurred in 2 placebo-treated patients and in 1 dapagliflozin-treated patient. Conclusions In patients with kidney failure receiving maintenance hemodialysis, dapagliflozin did not reduce LV mass overall. However, a prespecified interaction indicated that residual urine output significantly modified treatment response, with reverse remodeling observed only in patients with preserved urine output. These findings suggest that residual kidney function may be required for cardiac structural benefits of SGLT2i in dialysis and provide mechanistic insight into SGLT2i biology.