Efficacy and safety of dapagliflozin in patients with acute heart failure

Unlike diuretics or other GDMT, early dapagliflozin initiation could achieve both AHF goals.OBJECTIVESThe authors aimed to assess the diuretic efficacy and safety of early dapagliflozin ini

Efficacy and Safety of Dapagliflozin in Patients With Acute Heart Failure

Zachary L Cox, PHARMD,a,bSean P Collins, MD, MSC,c,dGabriel A Hernandez, MD,eA Thomas McRae III, MD,f

Beth T Davidson, DNP,fKirkwood Adams, MD,gMark Aaron, MD,hLuke Cunningham, MD,iCathy A Jenkins, MS,j

Christopher J Lindsell, PHD,jFrank E Harrell, Jr, PHD,jChristina Kampe, MACC,cKaren F Miller, RN, MPA,c

William B Stubblefield, MD, MPH,cJoAnn Lindenfeld, MDk

BACKGROUNDThe primary goals during acute heart failure (AHF) hospitalization are decongestion and guideline-directed medical therapy (GDMT) optimization Unlike diuretics or other GDMT, early dapagliflozin initiation could achieve both AHF goals.

OBJECTIVESThe authors aimed to assess the diuretic efficacy and safety of early dapagliflozin initiation in AHF.

METHODSIn a multicenter, open-label study, 240 patients were randomized within 24 hours of hospital presentation for hypervolemic AHF to dapagliflozin 10 mg once daily or structured usual care with protocolized diuretic titration until day 5 or hospital discharge The primary outcome, diuretic efficiency expressed as cumulative weight change per cu-mulative loop diuretic dose, was compared across treatment assignment using a proportional odds model adjusted for baseline weight Secondary and safety outcomes were adjudicated by a blinded committee.

RESULTSFor diuretic efficiency, there was no difference between dapagliflozin and usual care (OR: 0.65; 95% CI: 0.41-1.02; P¼ 0.06) Dapagliflozin was associated with reduced loop diuretic doses (560 mg [Q1-Q3: 260-1,150 mg] vs 800 mg [Q1-Q3: 380-1,715 mg]; P¼ 0.006) and fewer intravenous diuretic up-titrations (P # 0.05) to achieve equiv-alent weight loss as usual care Early dapagliflozin initiation did not increase diabetic, renal, or cardiovascular safety events Dapagliflozin was associated with improved median 24-hour natriuresis (P ¼ 0.03) and urine output (P ¼ 0.005), expediting hospital discharge over the study period.

CONCLUSIONSEarly dapagliflozin during AHF hospitalization is safe and fulfills a component of GDMT optimization Dapagliflozin was not associated with a statistically significant reduction in weight-based diuretic efficiency but was associated with evidence for enhanced diuresis among patients with AHF (Efficacy and Safety of Dapagliflozin in Acute Heart Failure [DICTATE-AHF];NCT04298229) (J Am Coll Cardiol 2024;83:1295–1306) © 2024 by the American College of Cardiology Foundation.

From theaDepartment of Pharmacy Practice, Lipscomb University College of Pharmacy, Nashville, Tennessee, USA;bDepartmentof Pharmacy, Vanderbilt University Medical Center, Nashville, Tennessee, USA;cDepartment of Emergency Medicine, VanderbiltUniversity Medical Center, Nashville, Tennessee, USA;dVeterans Affairs Tennessee Valley Healthcare System, Geriatric Research,Education and Clinical Center, Nashville, Tennessee, USA;eDivision of Cardiology, University of Mississippi Medical Center,Jackson, Mississippi, USA;fTriStar Centennial Medical Center, Nashville, Tennessee, USA;gDepartment of Medicine, University ofNorth Carolina School of Medicine, Chapel Hill, North Carolina, USA;hDepartment of Cardiac Sciences, Saint Thomas WestHospital, Nashville, Tennessee, USA;iDepartment of Cardiology, INTEGRIS Baptist Medical Center, Oklahoma City, Oklahoma,USA;jDepartment of Biostatistics, Vanderbilt University Medical Center, Nashville, Tennessee, USA; and thekDivision of Cardi-ology, Vanderbilt University Medical Center, Nashville, Tennessee, USA.

The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’institutions and Food and Drug Administration guidelines, including patient consent where appropriate For more information,visit theAuthor Center.

Manuscript received December 12, 2023; revised manuscript received February 6, 2024, accepted February 7, 2024.

Listen to this manuscript’saudio summary byEditor-in-ChiefDr Valentin Fuster onwww.jacc.org/journal/jacc.

The 2 primary therapeutic goals during acute heart failure (AHF) hos-pitalization are complete deconges-tion and optimizadeconges-tion of guideline-directed medical therapy (GDMT) before discharge.1-3

Studies of loop diuretics combined with either acetazolamide or hydrochlorothiazide demonstrated improved decongestion, but these diuretic combinations do not inher-ently optimize GDMT and failed to improve postdischarge outcomes.4,5

Early addition of sodium glucose co-transporter 2 inhibitor (SGLT2i) is a poten-tial strategy to achieve both primary AHF therapeutic goals, but the acute efficacy and safety of this strategy are unknown.1,2 Larger SGLT2i trials randomized patients at the end of hospitalization and evaluated safety and efficacy across intermediate to long-term postdischarge outcomes but not in-hospital safety outcomes, natriuresis, or decongestion.6,7 Early, smaller studies of SGLT2i in AHF reported increased urine output but lacked protocolized loop diuretic regimens between treatment arms or larger populations for safety assessment.8,9Thus, concerns persist regarding the safety and immediate benefits of SGLT2i initiation before patient stabilization during AHF hospitalization Endocrinology guidelines and international diabetes experts recommend against the routine use of SGLT2i during hospitalization, cit-ing concerns about increased risks of genitourinary infections, ketoacidosis, and acute kidney injury.10-12

These concerns may explain, at least in part, why only 20% of patients hospitalized for AHF without con-traindications are prescribed SGLT2i despite efforts to improve GDMT optimization during AHF hospitali-zation.13We evaluated dapagliflozin initiated within 24 hours of hospital presentation for hypervolemic AHF to assess its safety and additive diuretic benefits to protocolized loop diuretic therapy in DICTATE-AHF (Efficacy and Safety of Dapagliflozin in Acute Heart Failure).

TRIAL DESIGN We conducted an investigator-initiated, multicenter, prospective, randomized, open-label study at 6 hospitals in the United States

assessment of safety outcomes Details regarding the trial design and rationale have been previously pub-lished.14 Although dapagliflozin has not increased hypoglycemic events in outpatients, we adopted an open-label design to facilitate appropriate use of

inpatient insulin therapies with dapagliflozin given the well-established elevated hypoglycemia risk during hospitalization.10-12 The trial protocol and

Supplemental Appendix The study protocol was approved by each Institutional Review Board All patients provided written informed consent before any study procedures An independent Data and

protocol and interim safety analysis A blinded

(Supplemental Appendix).

PATIENTS Beginning in April 2020, adult patients with type 2 diabetes and an estimated glomerular filtration rate (eGFR) of at least 30 mL/min/1.73 m2 admitted to the hospital with AHF and current or planned treatment with intravenous (IV) loop di-uretics were eligible for participation Additional in-clusion criteria included at least 2 objective measures of hypervolemia and randomization within 24 hours of presentation to the emergency department or direct admission to the hospital In September 2021, the protocol was amended to allow enrollment of patients with or without type 2 diabetes and decrease the eGFR inclusion criterion to 25 mL/min/1.73 m2 because of the publication of new safety data.15The main exclusion criteria were type 1 diabetes, systolic

of <80 mg/dL, or a history of diabetic ketoacidosis Full inclusion and exclusion criteria are shown in Supplemental Table 1.

RANDOMIZATION AND STUDY PROCEDURES.Within 24

randomly assigned to either dapagliflozin 10 mg once daily or structured usual care The Research Elec-tronic Data Capture randomization module was used to randomize patients using permuted blocks of 4 stratified by site.16

At randomization, natriuretic peptide concentra-tion, standing weight, and congestion assessment using a validated orthoedema scale assessed by a study physician blinded to treatment assignment were performed.17Daily weight measurements used the same standing scale to minimize measurement variation Patients received a sodium-restricted diet (2 g/d) and a 2,000-mL/dfluid restriction.

An IV loop diuretic titration protocol targeting a urine output of 3 to 5 L/d was used for both study arms to standardize diuretic therapy and prevent empiric IV diuretic changes given the open-label design (Supplemental Table 2) If no IV diuretics were administered before randomization, the initial

SEE PAGE 1307

A B B R E V I A T I O N SA N D A C R O N Y M S

AHF= acute heart failure

CEC= Clinical Endpoint

NT-proBNP= N-terminal pro–B-type natriuretic peptide

SGLT2i= sodium glucose co-transporter 2 inhibitor

IV loop regimen was 2 to 2.5 times the home diuretic divided into twice daily administration If IV diuretics were administered before randomization, the treating provider could use the previous diuretic response to guide the initial loop diuretic dose selection Subse-quently throughout the study, loop diuretic doses were titrated every 12 to 24 hours to a dose of at least 960 mg/d of IV furosemide equivalents before a thiazide diuretic was added Use of nonstudy diuretic

allowed), nonstudy thiazides (excluding

considered a protocol deviation For patients with type 2 diabetes, endocrinology was consulted, and a standardized protocol was used for in-hospital gly-cemic control throughout the study (Supplemental Table 3) All other medication changes were at the discretion of the treating team.

A spot urine sample was collected after the initial IV loop diuretic dose, but before dapagliflozin administration, to measure baseline diuretic-induced urine chemistries On study day 2, a spot urine sample was collected after a loop diuretic dose and after dapagliflozin administration in the intervention arm to measure urine chemistries reflecting combination

completed on day 2 for measuring 24-hour natriuresis and urine volume.

At day 5 or discharge, whichever camefirst, natri-uretic peptide concentration, final standing weight, and a congestion assessment via an orthoedema scale by a study physician blinded to treatment assignment were performed After hospital discharge, patients were contacted via telephone to assess 30-day post-discharge outcomes.

diuretic efficiency calculated as the cumulative change in weight per cumulative loop diuretic dose (IV and oral) from enrollment to day 5 or discharge, if

kg/40 mg IV furosemide equivalents.18,19Diuretic ef-ficiency is independently associated with prognosis and quantifies diuretic response relative to the loop diuretic requirement to account for titration over the study.5,8,9,18,19

Secondary outcomes, adjudicated by a blinded CEC, included inpatient worsening heart failure (HF) and 30-day hospital readmission for HF or

Supplemental Appendix Safety outcomes included change in eGFR from baseline and the following events adjudicated by a blinded CEC: prolonged

hospitalization for hypotension or hypoglycemia, ketoacidosis, and inpatient mortality Exploratory endpoints included changes in natriuretic peptides,

diuresis, blinded physician congestion score, and hospital length of stay We evaluated the cumula-tive incidence of hospital discharge and duration of IV diuresis during the study period, treating death as a competing risk, as post hoc analyses to un-derstand how dapagliflozin affected these compo-nents of length of stay.

Adverse events were collected through 30 days postdischarge Mild hypoglycemia was defined as serum glucose of 54 to 69 mg/dL and severe hypo-glycemia as a serum glucose of <54 mg/dL.10 Hypo-tension was defined as either: 1) systolic blood pressure of<90 mm Hg sustained for at least 30 mi-nutes; or 2) symptomatic hypotension necessitating IV fluid or vasopressor therapies Hypovolemic hy-potension was defined as a systolic blood pressure of <90 mm Hg requiring IV fluid administration for hypovolemia.

STATISTICAL ANALYSIS The prespecified statistical

Appendix, and details of the study design have been

distributed standard deviation within each arm of 0.25 kg/40 mg IV furosemide,8,14we calculated that a sample size of 120 experimental subjects and 120 control subjects would have 87% power to detect a probability of 0.611 that an observation in the dapa-gliflozin arm is less than usual care with a 0.05 2-sided significance level The primary, secondary, and safety outcomes were analyzed according to the intent-to-treat principle, including all randomized patients who either received dapagliflozin or had baseline weight measured in usual care.P values and 95% CIs were not adjusted for multiplicity Partici-pant characteristics were summarized using median (Q1-Q3) or frequency (percentage), as appropriate The primary outcome was compared across treatment assignment using a proportional odds regression model adjusted for baseline weight Baseline weight wasfit with restricted cubic splines (3 knots) to relax the linearity assumptions ORs and estimated means were reported Measures of diuresis and natriuresis were compared across treatment assignment using Wilcoxon rank sum tests The association of N-ter-minal pro-B-type natriuretic peptide (NT-proBNP) at discharge with treatment arm was modeled using a proportional odds regression model adjusted for baseline NT-proBNPfit with restricted cubic splines (3 knots) Similarly, the association of discharge eGFR

with treatment arm was modeled using a proportional odds regression model adjusted for baseline eGFRfit with restricted cubic splines (3 knots) Missing data

matching with 20 imputation replications Statistical analyses were performed using R statistical software (R Core Team).20All tests were 2-sided with an alpha level of 0.05.

Between April 2020 and May 2023, 3,672 patients were screened, and 240 were randomized to either dapagliflozin or structured usual care at 6 sites Two patients were withdrawn (usual care arm: n¼ 1, un-able to stand for weights; dapagliflozin arm: n ¼ 1, treating physician’s decision to avoid SGLT2i ther-apy), leaving 238 patients eligible for analysis (Supplemental Figure 1) Baseline characteristics were balanced between treatment arms (Table 1, Supplemental Table 4) Type 2 diabetes was present in 71% of patients, and 52% had a left ventricular ejection fraction of#40% Baseline and daily IV loop diuretic doses with daily IV diuretic titrations are presented inSupplemental Table 5.

PRIMARY OUTCOME After adjustment for baseline weight, there was no difference in the OR for improved diuretic efficiency between dapagliflozin and usual care (OR: 0.65; 95% CI: 0.41-1.02;P ¼ 0.06) (Table 2) Model-based estimates of the mean diuretic efficiency were –0.42 kg/40 mg IV furosemide (95% CI:–0.52 to –0.33 kg/40 mg IV furosemide) with dapagliflozin and –0.31 kg/40 mg IV furosemide (95% CI:–0.39 to –0.23 kg/40 mg IV furosemide) with usual care If this difference was significant, a dif-ference of 0.11 kg/40 mg IV furosemide would trans-late to an additional weight loss of 1.5 kg for the median furosemide cumulative dose (560 mg) over the study period Distribution of the primary outcome is shown inSupplemental Figure 2 For the compo-nents of the primary outcome, the addition of dapa-gliflozin to IV loop diuretics resulted in equivalent cumulative weight change (–4.0 kg [Q1-Q3: –6.9 to –1.5 kg] vs –4.2 kg [Q1-Q3: –6.5 to –1.8 kg]; P ¼ 1.00) as usual care, but the dapagliflozin arm required signif-icantly less loop diuretic (560 mg [Q1-Q3: 260-1,150 mg] vs 800 mg [Q1-Q3: 380-1,715 mg];P ¼ 0.006) to achieve this weight loss (Figure 1) In compliance with the diuresis-guided IV loop diuretic titration protocol, providers did not empirically down-titrate the IV loop diuretic regimen when dapagliflozin was initiated, but rather, dapagliflozin significantly reduced the need for daily IV loop diuretic up-titrations for low urine output per the diuretic pro-tocol (Supplemental Table 5) The effect of dapagli-flozin on the primary outcome was generally consistent across prespecified subgroups (Figure 2) We did not observe a statistically significant interac-tion between baseline eGFR and the primary outcome between treatment arms (P for interaction ¼ 0.08) MEASURES OF DIURESIS AND NATRIURESIS.Among the 176 patients with a complete 24-hour urine

TABLE 1Baseline Characteristics

Atrialfibrillation or flutter50 (42)49 (41)Heart failure characteristics

Systolic blood pressure, mm Hg121 (112-136)120 (110-136)Serum laboratory valuesIV loop diuretic bolus dose before randomization116 (97)113 (95)Prerandomization IV loop diuretic dose in FE (mg)80 (40-160)80 (80-120)IV loop diuretic continuous infusion before

Values are median (Q1-Q3) or n (%).

ACEI¼ angiotensin-converting enzyme inhibitor; ARB ¼ angiotensin receptor blocker; ARNI ¼ angiotensinreceptor neprilysin inhibitor; BUN¼ blood urea nitrogen; eGFR ¼ estimated glomerular filtration rate;FE¼ furosemide equivalents in milligrams; IV ¼ intravenous; LVEF ¼ left ventricular ejection fraction;NT-proBNP¼ N-terminal pro–B-type natriuretic peptide; T2DM ¼ type 2 diabetes mellitus.

collection on day 2, patients excreted a median of 262 mmoL of sodium (Q1-Q3: 156-378 mmoL) and 2,952 mL of urine (Q1-Q3: 2,094-3,974 mL) Dapagli-flozin increased 24-hour measured natriuresis compared with usual care (median: 50 mmoL sodium/ 40 mg IV furosemide [Q1-Q3: 24-102 mmoL sodium/ 40 mg IV furosemide] vs 35 mmoL sodium/40 mg IV furosemide [Q1-Q3: 19-63 mmoL sodium/40 mg IV furosemide]; P ¼ 0.025) (Figure 3A) Stated differ-ently, the usual care arm required an additional 80 mg of IV furosemide to achieve equivalent 24-hour natriuresis as the dapagliflozin arm (Supplemental Table 6) Dapagliflozin significantly increased 24-hour urine output (median: 634 mL/40 mg IV furo-semide [Q1-Q3: 333-1,275 mL/40 mg IV furofuro-semide vs 403 mL/40 mg IV furosemide [Q1-Q3: 249-750 mL/ 40 mg IV furosemide];P ¼ 0.005) (Figure 3B) Urine sodium concentration and fractional excretion of

sodium per 40 mg IV furosemide in spot urine

(Supplemental Table 6).

exploratory analyses treating death as a competing risk, dapagliflozin reduced both the time to complete IV diuretic therapy (P ¼ 0.003) (Supplemental Figure 3) and the time to hospital discharge (52% vs

(Figure 3C) Despite an observed difference during the study period, total length of stay including days after the end of the study was not statistically different (Table 2) Although the usual care arm received IV diuretics for more time, the dapagliflozin arm had a nonsignificant improvement in orthoedema score at the end of the study (OR: 0.66; 95% CI: 0.38-1.13) (Supplemental Figure 4) NT-proBNP decreased sub-stantially in both the dapagliflozin (median: 47%

TABLE 2Primary, Secondary, Safety, and Exploratory Outcomes and Adverse Events

Change in weight (kg)/40 mg IV furosemide,adjusted for baseline weight

–0.42 (–0.52 to –0.33)–0.31 (–0.39 to –0.23)0.65 (0.41-1.02)0.06Change in weight (kg)/40 mg IV furosemide, unadjusted–0.24 (–0.53 to –0.08)–0.16 (–0.35 to –0.07)0.64 (0.41-1.00)0.05Secondary outcomes

30-day hospital readmission for AHF or

Select adverse events during treatment

Values are median (Q1-Q3) or n, unless otherwise indicated Treatment effect is presented as OR All P values, except the primary endpoint, are descriptive because no corrections for multipletesting were prespecified.

AHF¼ acute heart failure; HF ¼ heart failure; other abbreviations as inTable 1.

reduction) and usual care (median: 43% reduction) arms by the end of the study (Table 2) Patients ran-domized to early dapagliflozin were more likely to be on an SGLT2i at hospital discharge (P ¼ 0.006) (Supplemental Table 7).

ADVERSE EVENTS Early dapagliflozin initiation did not increase the risk of in-hospital worsening HF or 30-day readmissions for either HF or diabetes-related reasons compared to usual care (Table 2) No patient experienced ketoacidosis The change in eGFR from

FIGURE 1Primary Outcome and Components

(A) The odds of a favorable primary outcome expressed as weight change per 40 mg of IV furosemide adjusted for baseline weight failed was neutral The componentsof the primary outcome—(B) cumulative weight change and (C) cumulative loop diuretic dose—are presented as median (IQR), with whiskers representing 5th to 95thpercentiles Dapagliflozin significantly reduced the cumulative loop diuretic dose required to achieve an equivalent weight change compared with structured usual care.IV¼ intravenous.

FIGURE 2Heterogenicity of Treatment Effect Forest Plot

Treatment effect across prespecified subgroups based on the median value at randomization BMI ¼ body mass index; eGFR ¼ estimated glomerular filtration rate;FENa¼ fractional excretion of sodium; NT-proBNP ¼ N-terminal pro–B-type natriuretic peptide.

baseline to the end of the study was no different with dapagliflozin than usual care (–2.0 mL/min/1.73 m2vs –3.2 mL/min/1.73 m2;P ¼ 0.79) (Figure 3D).

The incidence of adverse events was similar be-tween treatment arms (Table 2) Adverse events

Supplemental Table 8, and acute kidney injury events are further described in Supplemental Table 9.

Changes in serum chemistries did not differ between treatment arms (Supplemental Table 10).

Dapagliflozin did not increase total or severe

Table 11) Mean blood glucose was not different with dapagliflozin relative to usual care for all patients (122 mg/dL [Q1-Q3: 106-147 mg/dL] vs 127 mg/dL [Q1-Q3:

FIGURE 3Safety and Exploratory Outcomes

Dapagliflozin significantly increased (A) 24-hour natriuresis and (B) 24-hour urine output per 40 mg IV furosemide in the n ¼ 176 patients with a complete24-hour urine collection on study day 2 Data are presented as median (IQR) with whiskers representing 5th to 95th percentiles (C) The dapagliflozin armhad a greater probability of hospital discharge during the study period, treating death as a competing risk (D) The mean SD change in estimatedglomerularfiltration rate did not differ between treatment arms P values are descriptive because no corrections for multiple testing were prespecified.IV¼ intravenous.

diabetes (139 mg/dL [Q1-Q3: 116-158 mg/dL] vs 134 mg/dL [Q1-Q3: 120-165 mg/dL];P ¼ 0.71) Among pa-tients with type 2 diabetes, the median daily insulin dose did not differ between dapagliflozin and usual care (9.6 U [Q1-Q3: 2.0-33.5 U] vs 9.3 U [Q1-Q3: 1.2-26.7 U];P ¼ 0.72).

In this multicenter, randomized trial with high-dose protocolized IV diuretic therapy titration, we investi-gated the efficacy and safety of dapagliflozin initiation within 24 hours of hospital presentation for AHF The centralfindings are as follows: 1) dapagliflozin ach-ieved equivalent weight reduction with significantly lower loop diuretic cumulative doses and fewer IV diuretic up-titrations required than usual care, resulting in a nonsignificant treatment effect for the primary outcome; 2) early dapagliflozin initiation was safe across all diabetic and cardiorenal in-hospital

increased natriuresis and urine output per 40 mg of IV furosemide The primary goals of AHF are deconges-tion and GDMT optimizadeconges-tion The totality of data supports the early initiation of dapagliflozin in AHF to facilitate both these goals (Central Illustration).

These observations have important implications for SGLT2i use across the spectrum of HF First, dapagliflozin demonstrated safety in the present study, which, compared to previous studies, had a larger multicenter population with blinded, stan-dardized adverse event adjudication Dapagliflozin can be safety started on thefirst day of AHF hospi-talization before patient stabilization, inherently facilitating rapid GDMT optimization Second, dapa-gliflozin provided acute diuretic benefits against a

diuretics Dapagliflozin improved multiple measures

weight-based diuretic efficiency favored dapagliflozin but failed to meet statistical significance Unlike hard cardiovascular outcomes, diuretic response lacks any one definitive metric; thus, we evaluated several outcomes reflecting diuretic efficiency to ensure a consistent treatment effect.2,21,22 Dapagliflozin

diuretic efficiency, reduced IV loop diuretic up-titration for poor diuretic response, and shortened the duration of IV diuresis.

Numerous AHF therapies have been investigated with the hope of improving acute decongestion and postdischarge outcomes Neither dopamine,

increased diuresis or outcomes in AHF.1,2Recently,

empiric additions of acetazolamide, thiazides, and SGLT2i improved diuresis Both acetazolamide’s and SGLT2i’s diuretic effects occur largely by inhibiting renal sodium hydrogen exchanger 3–mediated sodium transport.4,23-27 Given the potential mechanistic duplication, acetazolamide’s additive diuretic effect and clinical utility in patients receiving SGLT2i are unknown.4,28 Although loop-thiazide combinations increase natriuresis more than SGLT2i-loop combina-tions, thiazides increase the risk of worsening renal function and severe electrolyte depletion.5,29 Early addition of SGLT2i in AHF may decrease the need for thiazide addition and avoid these risks Poor diuretic efficiency is independently associated with several adverse HF outcomes including mortality across mul-tiple AHF cohorts.18,19Whether this risk can be miti-gated by improving diuretic efficacy is unclear Increasing decongestion and diuretic efficiency with acetazolamide or thiazides did not improve post-discharge outcomes.4,5 SGLT2i have convincingly reduced HF hospitalizations and mortality in prior

mechanisms.1,2,6,7,26Therefore, DICTATE-AHF was not designed to evaluate improvement in postdischarge outcomes For these reasons, early SGLT2i initiation should be prioritized over other diuretic combinations Our observations build on EMPA-RESPONSE-AHF (n¼ 80) and EMPAG-HF (Early Empagliflozin Initia-tion on Diuresis and Kidney FuncInitia-tion in Patients With

reporting increased diuresis with the addition of SGLT2i to IV loop diuretics in a larger multicenter population.8,9 The use of a urine output–based IV diuretic titration protocol with a standard net nega-tive goal of 2 to 3 L/d in DICTATE-AHF resulted in a higher median cumulative loop diuretic dose in the comparator group (800 mg IV furosemide) than in EMPA-RESPONSE-AHF (300 mg IV furosemide) and EMPA-AHF (351 mg IV furosemide) Thus, SGLT2i en-hances diuresis even when diuretic goals are

resistance exists Collectively, these studies indicate that early SGLT2i initiation is safe and provides modest diuretic benefit in AHF.

Safety concerns related to hypoglycemia, ketoaci-dosis, and changes in renal function have limited SGLT2i use during hospitalization Endocrinology so-cieties advocate for trials of noninsulin therapies with prespecified glycemic measures in hospitalized pa-tients with diabetes and cardiovascular disease.10-12

We designed DICTATE-AHF to evaluate in-hospital and postdischarge diabetes-related outcomes Dapa-gliflozin did not increase the risk of hypoglycemia, and no patient experienced ketoacidosis during

hospitalization eGFR commonly worsens during aggressive diuresis, and SGLT2i decreases eGFR within thefirst weeks of therapy.30,31However, prior studies of SGLT2i initiation during AHF observed conflicting changes in eGFR.8,9,32,33 Despite aggres-sive IV diuresis in DICTATE-AHF, dapagliflozin did not worsen eGFR or cause significant acute kidney injury events (Supplemental Table 8) compared to usual care Importantly, this was not the result of empiric IV loop diuretic down-titration in the dapa-gliflozin arm It may be that the acute, mild changes in eGFR with SGLT2i are obscured by the common un-derlyingfluctuations in eGFR increases and decreases during decongestion in AHF.30,33,34

Despite the well-established benefits of SGLT2i in HF and kidney disease, significant implementation gaps exist Only 20% of SGLT2i candidates hospital-ized for AHF are prescribed SGLT2i, and only 12% of

patients with chronic kidney disease receive SGLT2i therapy.13,35GDMT prescription at hospital discharge significantly increases the likelihood of chronic medication optimization and adherence.36-39 Our safety findings should encourage early in-hospital use, which can translate into improved chronic

STUDY LIMITATIONS The open-label design may have introduced unmeasured bias despite the use of a standardized diuretic titration protocol, objective outcomes, and blinded adjudicators for prespecified subjective assessments Given the limited experience with in-hospital SGLT2i at the time of study design, this compromise was acceptable to ensure that the hypoglycemia risk was mitigated Dapagliflozin had a

Early dapagliflozin initiation before patient stabilization on the first day of acute heart failure (AHF) hospitalization was safe, allowing rapid guideline-directed medicaltherapy (GDMT) optimization After adjustment for baseline weight, the OR for the primary outcome of diuretic efficiency was neutral for dapagliflozin (OR: 0.65;95% CI: 0.41-1.02; P¼ 0.06) compared to usual care However, dapagliflozin improved multiple other measures of diuretic efficacy, with the totality of data sug-gesting a modest diuretic benefit AE ¼ adverse event; DE ¼ diuretic efficiency; IV ¼ intravenous; SGLT2i ¼ sodium glucose cotransporter 2 inhibitor; UOP ¼ urineoutput.