Empagli ﬂ ozin in Patients With Heart Failure, Reduced Ejection Fraction, and Volume Overload EMPEROR-Reduced Trial

BACKGROUND Investigators have hypothesized that sodium-glucose cotransporter 2 (SGLT2) inhibitors exert diuretic effects that contribute to their ability to reduce serious heart failure events, and this action is particularly important in patients with ﬂ uid retention. OBJECTIVES This study sought to evaluate the effects of the SGLT2 inhibitor empagli ﬂ ozin on symptoms, health status, and major heart failure outcomes in patients with and without recent volume overload. METHODS This double-blind randomized trial compared the effects of empagli ﬂ ozin and placebo in 3,730 patients with heart failure and a reduced ejection fraction, with or without diabetes. Approximately 40% of the patients had volume overload in the 4 weeks before study enrollment. RESULTS Patients with recent volume overload were more likely to have been hospitalized for heart failure and to have received an intravenous diuretic agent in an outpatient setting in the previous 12 months, and to experience a heart failure eventfollowingrandomization,eventhoughtheyweremorelikelytobetreatedwithhighdosesofaloopdiureticagentasan outpatient(allp < 0.001).Whencomparedwithplacebo,empagli ﬂ ozinreducedthecompositeriskofcardiovasculardeathor hospitalization for heart failure, decreased total hospitalizations for heart failure, and improved health status and functional class. Yet despite the predisposition of patients with recent volume overload to ﬂ uid retention, the magnitude of these bene support a dominant role of in mediating the physiological or bene ts of inhibitors on the course of failure in patients with a reduced

T he action of sodium-glucose cotransporter 2 (SGLT2) inhibitors to promote natriuresis and osmotic diuresis (1)(2)(3)(4) has been proposed as a central mechanism by which these drugs reduce heart failure hospitalizations in large-scale trials (5)(6)(7). Sodium retention in patients with heart failure is related to increased sodium reabsorption in the proximal tubule (8), and by attenuating the sodium reabsorption at this site, SGLT2 inhibitors may potentiate the effect of diuretic agents acting at the loop of Henle (2,9,10). The effects of SGLT2 inhibitors to increase hematocrit and decrease body weight and circulating natriuretic peptides (6,7) have been linked by some investigators to an effect of diuresis to contract plasma and extracellular volume (11)(12)(13).
However, the actions of SGLT2 inhibitors on these biomarkers can be explained by mechanisms that are independent of sodium and water excretion (14)(15)(16)(17).
It has been proposed that if a diuretic action of SGLT2 inhibitors leads to a reduction in heart failure events, such a benefit would be particularly manifest in patients with heart failure who have fluid retention as a major feature of their history and clinical course (18,19). Typically, these patients manifest ongoing episodes of edema and other signs of organ congestion and volume overload, experience worse symptoms and clinical instability, and require repeated outpatient and inpatient interventions, especially the intensification of therapy with diuretic agents (20,21). This unfavorable course is often seen despite the use of high maintenance doses of oral diuretic agents, a finding suggesting a state of diuretic resistance (21,22). By acting on sodium reabsorption in the proximal tubule, SGLT2 inhibitors are poised to overcome resistance to loop diuretic agents that act more distally (23).
If the ability of SGLT2 inhibitors to potentiate the effect of loop diuretic agents is important, then patients who experience recent or repeated episodes of volume overload may be particularly likely to show a reduction in heart failure events during long-term treatment with SGLT2 inhibitors. This hypothesis was specifically tested in a secondary analysis of the EMPEROR-Reduced (EMPagliflozin outcomE tRial in Patients With chrOnic heaRt Failure With Reduced Ejection Fraction) trial, which evaluated the efficacy and safety of empagliflozin in patients with heart failure and a reduced ejection fraction.

METHODS
The design of the EMPEROR-Reduced trial has been described previously (8). Ethics approval was obtained at each study site, and all patients provided informed consent; the registration identifier at Clin-icalTrials.gov is NCT03057977.
Participants had New York Heart Association (NYHA) functional class II to IV heart failure and an ejection fraction #40%, and they were receiving all appropriate treatments for heart failure. We preferentially enrolled patients with an ejection fraction of #30% by requiring those patients with higher ejection fractions to have been hospitalized for heart failure within 12 months or to have markedly increased levels of N-terminal pro-B-type natriuretic peptide (NT-proBNP) (i.e., $1,000 pg/ml or $2,500 pg/ ml in patients with an ejection fraction of 31% to 35% or 36% to 40%, respectively); these thresholds were doubled in patients with atrial fibrillation. All patients were recruited as outpatients, but they were excluded if they had acute decompensation or a change in diuretic agents within 1 week before study enrollment.
Before randomization, investigators identified patients who had clinical evidence of volume overload in the 4 weeks before the first study visit and distinguished them from those patients who had been clinically euvolemic or volume depleted.
Because there is no consensus on the definition of euvolemia, no guidance was provided to investigators on how to identify patients with volume overload. We assessed the baseline use of diuretic agents, and patients were categorized as receiving high or low doses of a loop diuretic agent, with the former defined as >40 mg daily of furosemide or its equivalent (24). Patients were randomized and double blinded (in a 1:1 ratio) to receive placebo or empagliflozin 10 mg daily, in addition to their usual therapy for heart failure. Following randomization, all treatments for heart failure or other medical conditions could be initiated or altered at the discretion of the investigator. Patients were periodically assessed for major outcomes and functional capacity related to heart failure, intensification of diuretic therapy, vital signs, heart failure-pertinent biomarkers, and adverse events.
The primary endpoint was the composite of cardiovascular death or hospitalization for heart failure, analyzed as time to first event. The first secondary endpoint was the occurrence of all (first and recurrent) hospitalizations for heart failure. The second secondary endpoint was the slope of the change in estimated glomerular filtration rate (eGFR) during double-blind treatment. Serious adverse renal outcomes included long-term dialysis, renal transplantation, a sustained eGFR reduction of $40%, or a sustained eGFR <15 ml/ min/1.73 m 2 (if baseline eGFR was $30) or a sustained eGFR <10 ml/min/1.73 m 2 (if baseline eGFR was <30).
Additional analyses included the following: 1) time to reported intensification of therapy with diuretic agents; 2) changes in the Kansas City Cardiomyopathy Questionnaire at 12 and 52 weeks; and 3) changes in NYHA functional class, hematocrit, uric acid, NT-proBNP, body weight, systolic blood pressure, serum sodium, and serum albumin at 4 and 52 weeks.

OUTCOME MEASURES AND STATISTICAL ANALYSIS.
For time-to-first-event analyses, between-group differences were assessed using a Cox proportional hazards model, with pre-specified covariates of age, sex, region, diabetes, ejection fraction, and eGFR at baseline. For the analysis of total events, betweengroup differences were assessed using a joint frailty model, with cardiovascular death as a competing risk, and changes in NYHA functional class were evaluated by logistic regression. Both analyses used the same covariates as the time-to-first event analyses and included baseline NYHA functional class for the analysis of NYHA functional class. The eGFR slope analysis was determined on the basis of on-treatment data using a random coefficient model, with age and baseline eGFR as linear covariates and sex, region, ejection fraction, diabetes, baseline eGFR, time, and recent volume overload interaction terms as fixed effects.
For vital signs and laboratory measurements, treatment effects were assessed using a mixed model for repeated measures, with age and baseline eGFR as linear covariates and baseline score by visit, visit by treatment, sex, region, ejection fraction, individual last projected visit, and diabetes as fixed effects. All analyses of changes in NT-proBNP were performed on log-transformed data.
Interaction p values were used to compare the magnitude of the effect of empagliflozin on prespecified outcomes in groups defined by the presence or absence of recent volume overload. To assess differences in the course of the patients with or without recent volume overload at baseline, analyses were performed on placebo recipients only, by using the same covariate adjustments. Values are mean AE SD, n (%), or median (interquartile range). The p values refer to the difference between patients with or without recent volume overload, combining the 2 randomized treatment groups. Patients who self-identified with $1 race or with no race are classified as "other"; data for "other" or missing for both race and region are not shown. *Implantable cardioverter-defibrillator with or without cardiac resynchronization therapy. †Cardiac resynchronization therapy with or without a defibrillator.
Packer et al.  Cumulative incidence plots, with hazard ratio (HR), 95% confidence interval (CI), and p value for the comparison of empagliflozin and placebo. Interaction p value for the difference in the effect of empagliflozin on the left and right is 0.34.  (Table 3). Importantly, changes in body weight were poorly correlated with changes in NT-proBNP or with changes in hematocrit in individual patients treated with empagliflozin (r ¼ 0.12 and r ¼ À0.17, respectively, at 4 weeks; and r ¼ À0.14 and r ¼ 0.045, respectively, at 52 weeks).
Additionally, the frequency of reports of adverse events related to symptomatic hypotension, volume depletion, or worsening renal function was not increased by empagliflozin, either in patients with or without recent volume overload. The incidence of volume depletion was somewhat higher and the incidence of hyperkalemia somewhat lower in empagliflozin-treated patients with recent volume overload (Supplemental Table 1), but the betweengroup differences in the number of events were small.

DISCUSSION
Patients who had volume overload in the 4 weeks before enrollment in the EMPEROR-Reduced trial had a history of clinical instability and remained at elevated risk for heart failure events during followup. Patients with recent volume overload were Packer et al.    Values are hazard ratio (95% confidence interval) or mean AE SD. Because of the exceptional non-normal distribution, changes in NT-proBNP are shown as the ratio of adjusted geometric means and 95% confidence intervals.
Interestingly, in the EMPEROR-Reduced trial, patients with an ejection fraction of 30% or less were particularly responsive to treatment with empagliflozin with respect to the effects on the risk of cardiovascular death and hospitalizations for heart failure (6).
Finally, some investigators have hypothesized that SGLT2 inhibitors act preferentially to reduce the accumulation of fluid in the interstitial space (5), an action that would not be readily discerned by measurements of weight or blood constituents. However, this mechanism has been based on modeling analyses and not on direct measurements of interstitial fluid.  KEY WORDS diuretic agent, heart failure, SGLT2 inhibitors APPENDIX For a supplemental table, please see the online version of this paper.