Effects of dapagliflozin in DAPA-HF according to background heart failure therapy

Abstract

See page 2393 for the editorial comment on this article (doi: 10.1093/eurheartj/ehaa344)

Introduction

In the placebo-controlled Dapagliflozin And Prevention of Adverse-outcomes in Heart Failure (DAPA-HF) trial, the sodium–glucose cotransporter 2 inhibitor (SGLT2i) dapagliflozin reduced the risk of heart failure (HF) hospitalization and mortality, and improved symptoms, in 4744 patients with HF and reduced ejection fraction (HFrEF).^{1  ,  2} It is clearly important to know whether these effects were truly additive to the benefits obtained from other evidence-based treatments in HFrEF. Therefore, in this post hoc analysis, we examined outcomes in patients randomized to dapagliflozin, vs. placebo, according to background drug and device therapy. Because >90% of patients were receiving each of a renin–angiotensin system (RAS) blocker and beta-blocker (BB) at baseline, we examined the effect of dapagliflozin added to a higher or lower dose of these treatments. For other therapies, including a mineralocorticoid receptor antagonists (MRAs), digoxin and devices, we examined outcomes according to whether the treatment was taken at baseline or not. We also examined the effect of dapagliflozin in patients receiving combination therapy.

Methods

The design and results of DAPA-HF are published.^{1  ,  2} Briefly, DAPA-HF was a prospective, randomized, double-blind, placebo-controlled trial in patients with HFrEF, which evaluated the efficacy and safety of dapagliflozin 10 mg once daily, compared with matching placebo, added to standard care. Ethics committees at each of the 410 participating institutions (in 20 countries) approved the protocol, and all patients gave written informed consent.

Study patients

Men and women aged ≥18 years with HF were eligible if they were in New York Heart Association (NYHA) functional Class II–IV, had a left ventricular ejection fraction (LVEF) ≤40%, and, in the view of the investigator, were optimally treated with pharmacological and device therapy for HFrEF, according to local guidelines. The protocol advised that an angiotensin-converting enzyme (ACE) inhibitor, or angiotensin receptor blocker (ARB) or sacubitril/valsartan and a BB, as well as an MRA, should be used at guideline-recommended doses, unless contraindicated or not tolerated.

Participants were also required to have an N-terminal pro-B-type natriuretic peptide (NT-proBNP) concentration ≥600 pg/mL (≥400 pg/mL if hospitalized for HF within the previous 12 months). Patients with atrial fibrillation or atrial flutter were required to have an NT-proBNP level ≥900 pg/mL, irrespective of history of HF hospitalization. Key exclusion criteria included: symptoms of hypotension or systolic blood pressure <95 mmHg, estimated glomerular filtration rate (eGFR) <30 mL/min/1.73 m² (or rapidly declining renal function), Type 1 diabetes mellitus, and another condition likely to prevent patient participation in the trial or greatly limit life expectancy. A full list of exclusion criteria is provided in the design paper.¹

Trial outcomes

The primary outcome was the composite of an episode of worsening HF or cardiovascular (CV) death, whichever occurred first. An episode of worsening HF was either an unplanned hospitalization or an urgent visit resulting in intravenous therapy for HF. Secondary endpoints included the composite of the occurrence of HF hospitalization or CV death. All-cause mortality was a prespecified secondary endpoint. For the purposes of this analysis, we examined the effect of dapagliflozin, compared to placebo, on the primary composite outcome and the individual component of CV death.

Background treatment subgroups

In this post hoc analysis, we compared the effect of dapagliflozin with placebo in subgroups of patients treated with other background pharmacological and device therapies. Subgroups were limited to those with >200 individuals to minimize the likelihood of a Type 1 error.^{3  ,  4} Therefore, the yes/no groups which were analysed, included diuretic, digitalis glycosides (hereafter referred to as digoxin), MRA, sacubitril/valsartan, ivabradine, defibrillating device [implantable cardioverter-defibrillator or cardiac resynchronization therapy plus defibrillator (CRT-D)] and CRT (CRT-P and CRT-D). Because so few patients not taking a BB or a RAS blocker, we analysed the effect of study drug according to BB and RAS blocker dose at baseline, defined as ≥50% target dose and <50% of target dose. We also examined the treatment effect in patients receiving ≥50 and <50% of target MRA dose. Target daily doses were taken from contemporary guidelines^{5  ,  6} and included: carvedilol 50 mg, bisoprolol 10 mg, metoprolol succinate 200 mg, metoprolol tartrate 200 mg, and nebivolol 10 mg; patients (n = 85) taking other β-blockers were classified as taking <50% target dose. Target daily doses for ACE inhibitors and ARB were captopril 150 mg, enalapril 40 mg, fosinopril 40 mg, lisinopril 35 mg, perindopril 16 mg, quinapril 40 mg, ramipril 10 mg, trandolapril 4 mg, candesartan 32 mg, losartan 150 mg, valsartan 320 mg, and irbesartan 300 mg; patients (n = 315) taking other ACE inhibitors/ARB were classified as taking <50% target dose. The target doses of MRA were defined as either eplerenone or spironolactone 50 mg daily. For the purposes of this analysis, MRAs were not classified as a diuretic and sacubitril/valsartan was not included in the analysis of dose of ACE inhibitor/ARB.

Statistical analysis

Baseline characteristics were compared between groups by using the Kruskal–Wallis test for continuous variables and the χ  ² test for categorical variables. The effect of dapagliflozin compared to placebo on each outcome was examined by means of hazard ratio (HR) and 95% confidence intervals (CIs) derived from Cox proportional-hazards models, stratified according to diabetes status and adjusted for a history of hospitalization for HF and treatment-group assignment. We analysed the effect of dapagliflozin compared to placebo on the proportion of patients (presented as an odds ratio) who reported a clinically significant (≥5 point) improvement or deterioration in the Kansas City Cardiomyopathy Questionnaire (KCCQ) Total Symptom Score (TSS) at 8 months following randomization using previously described methods.⁷ Change in systolic blood pressure and serum creatinine was analysed using a mixed model for repeated measurements (adjusted for baseline values, visit, randomized treatment and interaction of treatment, and visit with a random intercept and slope per patient). The least-squares mean differences between treatment groups are presented by subgroup with 95% CI. Prespecified adverse events of interest (volume depletion and renal adverse events) were analysed in patients who were randomized and received at least one dose of dapagliflozin or placebo. All analyses were performed using Stata version 16 (College Station, TX, USA) and SAS, version 9.4 (SAS institute). A P-value <0.05 was considered statistically significant.

Results

Most patients in DAPA-HF were treated with a diuretic (84%), RAS blocker (94%), and BB (96%). Of patients taking a RAS blocker (n = 3952), 1517 patients (38%) were treated with ≥50% of a guideline-recommended target dose. Of those taking a BB (n = 4558), 2349 patients (52%) were treated with ≥50% of the guideline-recommended dose of an evidence-based drug. Overall, 71% of patients were treated with an MRA, 19% with digoxin, 26% with a defibrillator, and 7% had a CRT device. The majority of patients (96%) were treated with at least two of an ACE inhibitor/ARB/ARNI, a BB and/or an MRA, with 3091 (65%) patients on all three of these classes of drugs (Supplementary material online, Figure S1).

Baseline characteristics according to background therapy

Table 1 shows the background characteristics in all randomized patients and by background therapy and Table 2 shows the characteristics by combinations of background therapy. Compared with the overall DAPA-HF population, patients treated with digoxin had a lower LVEF, and, compared to other treatment subgroups, the worst NYHA functional class distribution, worst symptoms [as measured by the Kansas City Cardiomyopathy Questionnaire (KCCQ) Total Symptom Score (TSS)], highest NT-proBNP level and greatest frequency of prior HF hospitalization and atrial fibrillation. Excepting atrial fibrillation, similar trends were seen for patients treated with diuretics and an MRA, although both the latter groups also had worse renal function than the overall DAPA-HF population (especially patients treated with diuretics). Patients receiving sacubitril/valsartan were more likely to have an implanted device (ICD and/or CRT), and a lower mean LVEF and median NT-proBNP level, than the overall population. Patients receiving ivabradine were less likely to be taking a BB or to have atrial fibrillation than the overall DAPA-HF population.

Patients receiving ≥50% of target RAS blocker dose had a higher systolic blood pressure and lower median NT-proBNP level and KCCQ-TSS than patients in the overall DAPA-HF population. Other than a lower KCCQ-TSS, there were no major differences between patients taking ≥50% of target dose BB than in the overall DAPA-HF population.

Among patients with a device, fewer were women (especially an ICD) and patients with a device were generally older than in the overall DAPA-HF population; they also had a worse KCCQ-TSS and lower LVEF. CRT patients had a higher median NT-proBNP level, worse renal function and were more likely to have been hospitalized for HF previously, compared to the overall population.

Effect of dapagliflozin according to background therapy

Supplementary material online, Figure S2 shows the cumulative incidence of the primary endpoint in the major treatment subgroups of interest. The HRs for the effect of dapagliflozin, compared with placebo, for the primary composite outcome and CV death, according to background therapy are summarized in Figures 1 and 2, respectively. The benefit of dapagliflozin over placebo was consistent across all treatment subgroups, including diuretic, digoxin, MRA, sacubitril/valsartan, ivabradine, and devices (ICD or CRT), without any significant interaction between background therapy and the effect of randomized therapy on the primary composite outcome (Figure 1) or CV death (Figure 2).

Figure 1

Effect of dapagliflozin compared to placebo on the primary composite outcome by background heart failure therapy. The primary outcome was a composite of worsening heart failure (hospitalization or an urgent visit resulting in intravenous therapy for heart failure) or death from cardiovascular causes. Hazard ratios and 95% confidence intervals were estimated with the use of Cox regression models, stratified according to diabetes status, with a history of hospitalization for heart failure, and treatment-group assignment as explanatory variables. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor neprilysin inhibitor; CI, confidence interval; CRT, cardiac resynchronization therapy; ICD, implantable cardioverter-defibrillator; MRA, mineralocorticoid receptor antagonist. ^aICD or cardiac resynchronization therapy with defibrillator. ^bCardiac resynchronization therapy with or without a defibrillator.

Open in new tabDownload slide

Figure 2

Effect of dapagliflozin compared to placebo on cardiovascular death by background heart failure therapy. Hazard ratios and 95% confidence intervals were estimated with the use of Cox regression models, stratified according to diabetes status, with a history of hospitalization for heart failure, and treatment-group assignment as explanatory variables. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor neprilysin inhibitor; CI, confidence interval; CRT, cardiac resynchronization therapy; ICD, implantable cardioverter-defibrillator; MRA, mineralocorticoid receptor antagonist. ^aICD or cardiac resynchronization therapy with defibrillator. ^bCardiac resynchronization therapy with or without a defibrillator.

Open in new tabDownload slide

We also examined the effect of dapagliflozin, compared with placebo, according to background ACE inhibitor/ARB, BB and MRA dose. The effect of dapagliflozin, vs. placebo, was the same in patients treated with ≥50% of target doses of these drugs, compared with those treated with a lower dose (Figures 1 and 2).

Effect of dapagliflozin according to combinations of treatment

Figures 2 and 3, respectively, summarize the HRs for dapagliflozin, compared with placebo, for CV death and the primary composite outcome, according to various combinations of background therapy. The beneficial effect of dapagliflozin was consistent regardless of whether patients were treated with the combination of a RAS-blocker, BB and MRA, or were not on all three of these treatments at baseline. Neither the combination of an ICD and ≥50% of target dose of both an RAS blocker and BB, nor an MRA with ≥50% of target dose of both an RAS blocker and BB, modified the response to dapagliflozin (Figures 2 and 3). A small proportion of patients (7%) were on ‘quadruple-therapy’ at baseline with a combined angiotensin receptor neprilysin inhibitor (sacubitril/valsartan), a BB and an MRA; the benefit of dapagliflozin was consistent in this subgroup with no significant interaction when compared with participants not receiving ‘quadruple-therapy’.

Figure 3

Effect of dapagliflozin compared to placebo on the primary composite outcome by combinations of background heart failure therapy. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor neprilysin inhibitor; CI, confidence interval; ICD, implantable cardioverter-defibrillator; MRA, mineralocorticoid receptor antagonist. ^aICD or cardiac resynchronization therapy with defibrillator.

Open in new tabDownload slide

Effect of dapagliflozin on symptom frequency and severity according to background therapy

A significantly greater proportion of patients randomized to dapagliflozin compared to placebo self-reported a clinically significant improvement (≥5 point increase) in the KCCQ-TSS at 8 months following randomization and significantly fewer dapagliflozin patients reported a meaningful deterioration (≥5 point decrease) compared to placebo (Table 3). This effect was consistent across all subgroups of background therapy and combinations analysed.

Effect of dapagliflozin on change in systolic blood pressure and serum creatinine according to background therapy

Compared with placebo, dapagliflozin reduced blood pressure at 8 months following randomization with no significant treatment effect modification observed by individual background therapy or their combinations (Table 3). The small increase in serum creatinine observed with dapagliflozin compared to placebo at 8 months was similar among the subgroups analysed (Table 3).

Effect of dapagliflozin on volume depletion or renal adverse events according to background therapy

In patients treated with a diuretic, adverse events related to volume depletion occurred more frequently with dapagliflozin compared to placebo, with fewer events seen in those not on a diuretic and randomized to dapagliflozin (Table 4). A similar pattern was observed in those treated with ≥50% of target MRA dose, the combination of an RAS blocker, BB, and MRA, and combinations including ≥50% of both RAS blocker and BB target dose (Table 4). Renal adverse events were less common with dapagliflozin compared to placebo in those not on a diuretic at baseline with no difference in those treated with a diuretic (Table 4). Patients treated with ≥50% of target RAS blocker dose more frequently had a renal adverse event with dapagliflozin compared to placebo than those on <50% target dose with the same pattern in those on combinations of therapy that include ≥50% of target RAS blocker dose.