STEP-HFpEF: Semaglutide in patients with HF & ejection fraction >=45% & obesity, with or without diabetes

Pooled analysis of STEP-HFpEF & STEP-HFpEF DM. Lancet 2024

STEP-HFpEF main paper. NEJM 2023;389:1069-84.

STEP-HFpEF DM main paper. NEJM 2024;390:1394-407.

Bottom line: In patients with symptomatic HF, ejection fraction >=45% & BMI >=30, semaglutide improved quality of life & reduced weight (average ~11% weight loss at 1 year), but increased the risk of discontinuation due to GI intolerance. For every 100 patients treated for 1 year, 12 patients will get a noticeable improvement in their quality of life because of semaglutide, & 8 patients will discontinue it due to intolerable side-effects (mostly gastrointestinal).

Patients (n=1145 randomized)

  • 13 countries, 2021-2022

  • Included:

    • Age >=18 years

    • Symptomatic (NYHA 2-4) HF with left ventricular ejection fraction >=45%

    • At least one of the following:

      • Elevated LV filling pressures;

      • Elevated BNP/NTproBNP plus echocardiographic abnormalities;

      • HF hospitalization in the last 12 months, plus ongoing treatment with diuretics OR echo abnormalities

    • BMI >=30

    • Kansas City Cardiomyopathy Questionnaire (KCCQ)-Clinical Summary Score (CSS) <90/100

    • 6-minute walk distance (6MWD) >=100 meters

    • STEP-HFpEF excluded patients with diabetes (prior dx, A1c >=6.5% at screening), whereas STEP-HFpEF DM required types 2 diabetes & A1c <=10%

  • Key exclusions:

    • ESRD or dialysis dependence

  • Baseline:

    • Age 70 y, male ~50%

    • White 90%, Black 4%

    • LVEF median 57% (45-49% in 16%)

    • NYHA 2 (69%), 3 (31%)

    • KCCQ-CSS median 59/100, 6MWD median 295 meters

    • HF hospitalization in last year 17%

    • Weight median 104 kg, BMI median 37 (66% BMI >=35)

    • Comorbidities: AF 45%, CAD 21%, HTN 84%

    • Meds: Loop diuretic 62%, ACE/ARB/ARNI 79%, beta-blocker 81%, MRA 34%, SGLT2i 20% (4% in those without T2DM)

Intervention: Semaglutide subcutaneous once weekly at “weight loss doses”

  • Starting dose: 0.25 mg q1w

  • Titration: Uptitrated every 4 weeks (to 0.5 -> 1.0 -> 1.7 -> 2.4 mg q1w) as tolerated

  • Target dose: 2.4 mg q1w (reached after 16 weeks)

    • 84% of those still taking the drug at 1 year received the target dose

Comparator: Matching placebo

Outcomes @ median 1.1 year

Co-primary outcomes: Mean change from baseline to week 52:

  • KCCQ-CSS mean difference +7.5

    • Clinically-important improvement (>=5-point improvement) in KCCQ-CSS: 74% vs 57% (+17%)

      • >=10-point improvement: 61% vs 43% (+18%)

    • Consistent mean improvement over placebo across KCCQ-Overall Summary Score (+7.4) & across all sub-scores

  • Weight mean difference -8.4% or -8.9 kg (greater weight loss in non-diabetic patients)

    • >=20% reduction: 12.2% vs 1.2% (NNT ~9)

Key secondary outcomes

  • 6MWD: mean +17 meters with semaglutide vs placebo

  • Exploratory composite (time to first HF hospitalization, urgent visit, or CV death): 2% vs 6% (hazard ration 0.31, 95% confidence interval 0.15-0.62)

Safety

  • Serious adverse events: Semaglutide 28.7 vs 52.7 %/y

  • Discontinued due to GI adverse events: 10.7 vs 3.3 %/y

Internal validity = low risk of bias

  • Computer-generated random sequence generation

  • Allocation concealment by centralized interactive web-based response system

  • Blinding by matching placebo & titration schedule

  • Intention-to-treat analysis

  • Loss to follow-up (LTFU): KCCQ data missing for 8% on semaglutide & 11% on placebo at 1 year

Generalizability & other considerations

  • Similar improvement (no significant treatment-subgroup interaction) in QoL with semaglutide across studied LVEF in mildly-reduced/preserved range (45% to >=60%)

  • In STEP-HFpEF, similar improvement in QoL with semaglutide regardless of BMI (but all >=30), but KCCQ-CSS improvement in the semaglutide group was associated with weight loss >=5%

    • Impossible to say whether lesser KCCQ improvement in patients who lost <5% of their body weight due to an actual cause-effect relationship between weight loss & QoL improvement, or whether this is confounded by some other factor (e.g. lower adherence to semaglutide could explain lack of both weight loss & QoL improvement)

  • Individual-patient-level meta-analysis of patients with HFmrEF/HFpEF in STEP-HFpEF, STEP-HFpEF-DM, SELECT, and FLOW trials

    • Reduction in HF composite (time to first worsening HF or CV death): HR 0.69 (0.53-0.89); absolute risk reduction ~0.9%/y

    • But no significant reduction in CV death (HR 0.82, 0.57-1.16)

TOPCAT & FINEARTS-HF: Spironolactone & finerenone in heart failure with ejection fraction >40%

TOPCAT (spironolactone) & FINEARTS-HF (finerenone)

Bottom line:

  • In patients with HF with EF >40% (i.e., HF with mildly-reduced or preserved EF), MRAs reduced the risk of HF hospitalization vs placebo, did not reduce the risk of dying, and increased the risk of hypotension, hyperkalemia, and eGFR reduction.

  • For every 1000 patients treated with an MRA per year, 15 HF hospitalizations would be avoided, but there would be 20-30 more people with hypotension (with or without symptoms), 70-130 more patients with eGFR reduced by >30%, and 70-80 more patients with K >5.5 mmol/L.

  • There is no direct head-to-head comparison of finerenone vs spironolactone. Efficacy & safety seem to be similar with these 2 drugs; this likely represents a class effect of MRAs in HFmrEF/HFpEF.

TOPCAT 

TOPCAT: Spironolactone for heart failure with preserved ejection fraction. NEJM 2014;370:1383-92

TOPCAT-Americas: Circulation 2015;131:34-42

Patients (n=3445)

  • Inclusion

    • Age 50+

    • Clinical HF

    • LVEF >45% (measured within 6 months prior to randomization & after any prior ACS)

    • Controlled SBP (<140, or <160 if receiving 3+ antihypertensives)

    • Either,

      • 1+ HF-related hospitalization in prior 12 months, or

      • BNP >100 pg/mL (or N-terminal pro-BNP >360 pg/mL) within 30 days, not explained by another disease entity

  • Exclusion

    • Pericardial constriction

    • Hemodynamically significant uncorrected primary valvular heart disease

    • Infiltrative or hypertrophic obstructive cardiomyopathy (HoCM)

    • Stroke, MI or CABG in past 90 days

    • AF with resting HR >90 bpm

    • Heart transplant recipient or currently implanted LVAD

    • Chronic pulmonary disease: Requiring home O2 or PO steroids, hospitalization for exacerbation within 12 months, or significant in the opinion of the investigator

  • "Average" patient

    • Age 69 y (age 75+ 27%), female 52%, white 89%

    • HF characteristics

      • Hospitalization in last 12 months 71%

      • NYHA functional class 1 (3%), 2 (64%), 3 (33%), 4 (<1%)

      • BNP 236

      • LVEF 56%

    • BP 130/80 mm Hg

    • Meds: Diuretic 80%, ACEI/ARB 85%, beta-blocker 75%

Intervention: Spironolactone

  • Initial dose: 15 mg PO daily x4 weeks

    • If initial dose tolerated, increased to 30 mg daily

    • If HF still symptomatic at month 4, increased to 45 mg daily

  • Monitoring: Measurement of SCr & serum K required <1 week after start or change of study drug dose

Comparator: Matching placebo

Outcomes @ mean 3.3 y

Efficacy

  • Death: Spironolactone 14.6% versus placebo 15.9%, hazard ratio (HR) 0.91 (95% confidence interval 0.77-1.08)

  • Hospitalization for any cause: 44.5% vs 46%, HR 0.94 (0.85-1.04)

  • Primary outcome (CV death, aborted cardiac arrest, HF hospitalization): 18.6% vs 20.4%, HR 0.89 (0.77-1.04)

  • Quality of life

    • Kansas City Cardiomyopathy Questionnaire (KCCQ; 100-point scale, minimal clinically important difference -5): Versus placebo, +1.5 at 4 months & +1.9 at 36 months (p=0.02)

Safety

  • Discontinuation of study drug: 34.3% vs 31.4%

  • Doubling of SCr: 10.2% vs 7% (NNH 32)

  • Hyperkalemia (serum K >5.5 mmol/L): 18.7% vs 9.1% (NNH 11)

Internal validity

FINEARTS-HF

FINEARTS-HF: Finerenone in heart failure with mildly reduced or preserved ejection fraction. NEJM 2024

Patients (n=6001)

  • 7463 screened -> 6016 randomized -> 6001 analyzed

  • Inclusion:

    • Age >=40

    • HF with NYHA 2-4

    • Outpatient or hospitalized for heart failure (hemodynamically stable off inotropes)

    • Receiving diuretic >=30 days

    • LVEF >=40% measured in last year (could be <40% before)

    • Structural heart abnormality (LA enlargement, LVH)

    • NTproBNP >=300 pg/mL in SR or >=900 pg/mL in AF

  • Key exclusions:

    • SBP <90 mm Hg

    • eGFR <25

    • K >5.0 mmol/L; prior hyperkalemia with MRA

    • Untreated HTN; alternative causes for HF symptoms

  • “Average” trial patient:

    • 72 y/o, 46% female, white 79%/Asian 17%

    • PMHx: HTN 88%, T2DM 42%, AF 38%, prior LVEF <40% 5%

    • HF characteristics

      • Prior HF hospitalization 60%

      • Time since HF event: <=1 week 20%, 7-90 days 34%, >3 months/none 46%

      • NYHA 2 (69%), 3 (30%), 4 (<1%)

      • LVEF 52% (36% <50%)

      • NTproBNP ~1000

    • SBP 129 mm Hg, K 4.4 mmol/L, eGFR 62

    • Meds: Loop diuretic 87%, beta-blocker 85%, ACEI/ARB ~70%, ARNI 8-9%, SGLT2i 13-14%, GLP1-RA 2-3%

Intervention: Finerenone

  • If eGFR >60:

    • Starting dose: Finerenone 20 mg PO once daily

    • Increase to 40 mg PO once daily after 4 weeks (target dose)

    • Dose range 10-40 mg daily

  • If eGFR 25-60:

    • Starting dose: Finerenone 10 mg PO once daily

    • Increase to 20 mg PO once daily after 4 weeks (target dose)

    • Dose range: 10-20 mg daily

Comparator: Matching placebo

Outcomes @ median 2.7 y

Efficacy

  • Death: 16.4% vs 17.4% (HR 0.93, 95% CI 0.83-1.06)

  • Primary outcome (composite of cardiovascular death or worsening HF event [first or recurrent hospitalization or urgent visit for HF]):

    • Finerenone 14.9 vs placebo 17.7 per 100 patient-years

    • Rate ratio 0.84, 95% confidence interval (CI) 0.74-0.95

    • Similar result using traditional “time to first event”: 20.8% vs 24.0% (hazard ratio [HR] 0.84, 95% CI 0.76-0.94)

    • Results consistent across subgroups

  • Kidney composite outcome (sustained eGFR decrease ≥50%, sustained eGFR to <15, or initiation of long-term dialysis or kidney transplantation): 2.5% vs 1.8% (HR 1.33, 95% CI 0.94-1.89)

  • NYHA improvement at 12 months: ~19% in both groups (odds ratio 1.01, 95% CI 0.88-1.15)

  • Change from baseline in Kansas City Cardiomyopathy Questionnaire (KCCQ) total symptom score: +1.6 (95% CI +0.8 to +2.3) out of 100

Safety

  • Discontinued study drug: 20% in both groups

  • SBP <100 mm Hg (symptoms not specified): 18.5% vs 12.4%

  • Hyperkalemia (K >5.5 mmol/L): 14.3% vs 6.9%

    • >6: 3% vs 1.4%

    • “Investigator-reported hyperkalemia”: 9.7% vs 4.2%

    • Death from hyperkalemia: 0% in both groups

  • Serum creatinine increase to >221 umol/L: 2% vs 1.2%

Internal validity

  • Low risk of bias

    • Allocation bias: Randomized, allocation concealed

    • Performance & detection bias: Patients, clinicians, outcome assessors unaware of treatment assignment

    • Attrition bias: Intention to treat analysis; low loss to follow-up (~0.1%)

Meta-analysis

Mineralocorticoid receptor antagonists in heart failure: an individual patient level meta-analysis. Lancet 2024

Pooled results of TOPCAT + FINEARTS-HF:

  • Composite of cardiovascular death or HF hospitalization pool: HR 0.87 (95% CI 0.79-0.95), 1.5% absolute risk reduction

  • Death: HR 0.94 (0.85-1.03)

Safety: Similar absolute increase (vs placebo) with both spironolactone & finerenone

  • SBP <90 mm Hg:

    • TOPCAT: Spironolactone 4% vs placebo 2%

    • FINEARTS-HF: Finerenone 5% vs placebo 3%

  • eGFR reduction >30%:

    • TOPCAT: 31% vs 24%

    • FINEARTS-HF: 35% vs 22%

  • K >5.5 mmol/:

    • TOPCAT: 12% vs 5%

    • FINEARTS-HF: 15% vs 7%

SODIUM-HF: Sodium restriction in patients with heart failure

By Hans Haag with editorial support from Ricky Turgeon

SODIUM-HF. Lancet 2022;399:1391-400.

Bottom line: In patients with heart failure (HF), achieved dietary sodium restriction ~1600 mg/day did not reduce the risk of death or cardiovascular hospitalizations/ED visits compared with usual care achieving sodium about 2000 mg/day.

Patients (n=841 randomized, 806 analyzed)

  • Included:

    • Adults with clinical HF

    • NYHA class 2-3

    • Receiving optimal guideline-directed medical therapy

  • Key exclusions:

    • Dietary sodium intake <1500 mg/day

    • Serum sodium <130 mmol/L

    • eGFR <20 mL/min/1.73 m^2 or requiring hemodialysis

    • Within 1 month of cardiovascular hospitalization

  • Baseline

    • Age 67, 33% female, 58% Canadian

    • NYHA 2 (71.4%), NYHA 3 (26.9%), LVEF median 36%, NT-proBNP ~800, HF hospitalization in last 12 months ~33%

    • Comorbidities: HTN 62%, CAD 47%, AF/AFL 39%, diabetes 33%

    • SBP 118, eGFR 61

    • Meds: ACEI/ARB/ARNI 81%, beta-blocker 87%, MRA 57%

      • Diuretic use not reported

    • Baseline median sodium intake ~2200 mg/day (self-report)

Intervention: Dietary sodium intake <1500 mg/d

  • Country-specific meal plans & menus

  • Achieved ~1600 mg/d at 12 months (approximately 400 mg/d lower than comparator group)

Comparator: Usual care

  • General advice to restrict dietary sodium (as provided during routine clinical practice)

  • Achieved ~2000 mg/d at 12 months

Outcomes @ 12 months

  • No difference for any of the clinical outcomes

  • Improvements in some subjective secondary outcomes with the intervention

    • QoL: Mean difference in Kansas City Cardiomyopathy Questionnaire +3.4/100 vs placebo

    • NYHA class 1: 16% vs 11% (odds ratio 1.69 [95% CI 1.16-2.50] for improvement ≥1 NYHA class)

    • 6min walk distance (6MWD): +6.6 meters (95% CI -9 to +22)

Internal validity

  • Low risk of bias for clinical outcomes (death, hospitalizations)

    • Low risk of selection bias: Independent statistician generated randomization list; allocation concealment by automated web-based system. Block randomization according to the study site.

    • Low risk of performance & detection bias: Patients and clinicians unblinded to treatment allocation, but relatively objective outcomes & blinded endpoint adjudication

    • Attrition bias: ITT analysis, LTFU <1%

  • High risk of bias for subjective outcomes (QoL, NYHA, & to lesser extent 6MWD)

    • Low risk of selection bias as above

    • But high risk of performance & detection bias due to subjective nature of these outcomes & extra healthcare visits with in-person dietician visits at months 3 & 9 (possibility to identify & mitigate worsening HF or optimize pharmacotherapy)

  • Trial stopped early due to operational feasibility issues resulting from the COVID-19 pandemic

Other considerations

  • Generalizability: Limited to individuals already fairly restrictive in dietary sodium

    • Patients/comparator:

      • All patients were restricting sodium intake at baseline (~2.2 g/d) compared to the general population (average 4 g/day worldwide). The value of this restriction & ideal targets (e.g. 2-2.5 g/d vs 3-4 g/d) remains unknown.

      • Notably, these results do not apply to patients who are acutely hospitalized with HF, require escalating diuretic doses, or those with advanced chronic kidney disease

    • Intervention: Highly transportable/scalable due to use of meal plans & menus rather than prepared meals

  • Proportion of patients on diuretics used, doses used, & potential changes during the trial have not (yet?) been reported

    • These results could shed light on the secondary outcomes (e.g. could sodium restriction allow for lower diuretic doses -> less urinary frequency -> improve QoL?)

    • Similarly, information on changes to HF pharmacotherapy during the trial could offer further insights into these results

Context

  • Sodium restriction has historically been a cornerstone of HF management for decades due to the role of sodium in contributing to sodium retention and congestive symptoms of HF

    • However, evidence for sodium retention (& specific targets) has been limited to observational studies and small, inconclusive randomized controlled trials.

  • The latest Canadian HF guideline recommendations on dietary sodium intake (2017) acknowledged this uncertainty and recommended personalizing daily targets while still recommending 2000-3000 mg of sodium per day (consistent with the usual care group in SODIUM-HF)

PARAGON-HF: Sacubitril-valsartan in heart failure with ejection fraction >=45%

By Hans Haag with editorial support from Ricky Turgeon

PARAGON-HF. NEJM 2019;381:1609-20

Bottom line:

  • In patients with heart failure (HF) with left ventricular ejection fraction (LVEF) ≥45%, sacubitril-valsartan did not reduce the composite total HF hospitalizations/cardiovascular (CV) death or death from any cause over ~3 years.

  • Sacubitril-valsartan increased the risk of hypotension (+5%) and angioedema (+0.4%) compared with valsartan.

  • In subgroup analyses (which often lead us astray) sacubitril-valsartan reduced HF hospitalizations in females by ~1.5%/year (but not males) with LVEF in the “lower” range (45-60%).

Patients (n=4822 randomized)

  • Screened 10,359 -> 5746 entered valsartan run-in (~9% discontinued) -> 5205 entered sacubitril-valsartan run-in (7% discontinued) -> 4822 randomized

  • Included:

    • Age 50+

    • NYHA 2-4

    • LVEF ≥45% measured in the last 6 months

    • At least one of the following:

      • HF hospitalization within 9 months prior + NT-proBNP >200 pg/mL (>600 pg/mL if AF)

      • NT-pro-BNP >300 pg/mL (>900 pg/mL if AF)

    • Echo evidence of LA enlargement (e.g. LAV ≥55 mL or LAVi ≥29 mL/m^2) or LVH (septal thickness ≥1.1 cm)

    • Symptomatic HF for at least 30 days prior to screening visit

    • Structural heart disease (LA enlargement or LVH)

  • Key exclusions:

    • Any prior LVEF <40%, history of dilated cardiomyopathy, hemodynamically significant valvular heart disease

    • Uncontrolled/life-threatening dysrhythmia, including AF-RVR

    • History of angioedema

    • Alternate diagnosis to explain HF symptoms (e.g. anemia with Hb <100 g/L, severe COPD)

    • Uncontrolled hypertension

    • SBP <100 mm Hg or symptomatic hypotension

    • eGFR <30 mL/min/1.73m^2 or a reduction of >35% after run-in period

    • K >5.2 mmol/L

  • Baseline:

    • Age 73, 52% female, 82% white, 12% Asian, 2% Black

    • NYHA 2 (77%) & 3 (20%), LVEF median 57%, NT-proBNP ~900 pg/mL, HF hospitalization in last 12 months ~23%

    • Comorbidities: HTN 96%, diabetes 43%, AF/AFlutter 33%

    • Meds: Diuretic 96%, ACEI/ARB 86%, beta-blocker 80%, MRA ~25%

    • SBP 130, eGFR 63

Interventions: Sacubitril-valsartan 97/103 mg BID vs valsartan 160 mg BID

  • Intervention: Sacubitril-valsartan 97/103 mg BID (82% on target dose)

  • Comparator: Valsartan 160 mg BID (85% on target dose)

  • Co-intervention: MRA permitted, all other non-study RAAS inhibitors stopped

Outcomes @ median 35 months (2.9 years)

paragon.png

Internal validity: Low risk of bias selection, performance, detection & attrition bias

  • Selection bias: Computer-generated random sequence; allocation concealment by interactive web-response system

  • Performance & detection bias: Patients, clinicians & investigators blinded to study allocation via double-dummy placebo

  • Attrition bias: ITT analysis, LTFU <0.1%

Other considerations

  • Generalizability:

    • 16% excluded in single-blind run-in

      • Consisted of (1) Valsartan 40-80 mg BID x 1-2 weeks, then (2) sacubitril-valsartan 49/51 mg BID x2-4 weeks, then randomized

    • PARAGON-HF defined “preserved” ejection fraction as >45%, which differs from the (subsequent) 2021 universal definition and classification of HF’s classification of HF with preserved LVEF as ≥50% and HF with LVEF 41-49% as mildly-reduced

  • Pre-specified subgroup analysis of the primary outcome based on baseline LVEF suggests greater benefit with ARNI in patients with LVEF below the study median (≤ 57%) and in females (but no reduction in death in any subgroup)

  • Are the results clinically important?

    • Yes for: Patients who place higher value on reducing HF hospitalizations and improving quality of life than they do the increase in hypotension, angioedema (and higher cost)

    • Not for:

      • Males

      • Female patients with LVEF ≥60%

      • Female patients who do not care about the benefits noted above, or who are more concerned about costs, pill burden, and the adverse effects noted

    • However, given the more robust evidence and greater certainty for SGLT2 inhibitors and MRAs in this setting (noted below), ARNI should generally be offered after exhausting those other options.

Context

EMPEROR-Preserved: Empagliflozin in heart failure with preserved or mildly-reduced ejection fraction

References:

Bottom Line: In patients with symptomatic heart failure with preserved (≥50%; HFpEF) or mildly-reduced (41-49%; HFmrEF) ejection fraction, empagliflozin reduced the risk of HF hospitalization vs placebo (-3.2%) and increased the probability of a clinically important improvement in quality of life (+3.8%), but did not reduce deaths or total hospitalizations at 2.2 years. Empagliflozin increased the risk of symptomatic hypotension (+1.4%), genital fungal infections (+1.5), and UTIs (+1.8%).

Patients (n=5988)

  • 11,583 screened -> 5988 randomized

    • Most common reasons for exclusion:

      • 78% NT-proBNP below inclusion criteria threshold

      • 5% LVEF <40%

      • 4% exclusion criteria based on safety

  • Enrolment 2017-April 2020

  • 23 countries

  • Included: Symptomatic chronic HFpEF or HFmrEF with elevated natriuretic peptide

    • Chronic HF (≥3 months)

    • NYHA 2-4

    • LVEF >40% without any prior LVEF ≤40%

    • NT-proBNP

      • >300 pg/mL if in sinus rhythm

      • >900 pg/mL if in atrial fibrillation

    • Either

      • Structural heart disease (LAE &/or LVH) documented on echo

      • HF hospitalization within past 12 months

  • Key exclusions:

    • SBP <100 mm Hg

    • eGFR <20

    • BMI ≥45

    • SGLT2i contraindication (history of ketoacidosis, allergy/hypersensitivity)

    • “Cardiovascular (CV) disease/treatment that increase the unpredictability of or change the patients’ clinical course independent of HF” (e.g. MI/stroke/TIA/CV surgery in past 90 days; infiltrative cardiomyopathy; heart transplant recipient/wait list; severe valvular disease)

    • “Untreated CV condition that might influence the course of HF/study drug tolerability” (e.g. AF with uncontrolled HR, SBP ≥180 mm Hg)

    • “Significant comorbidity that might influence clinical course” (e.g. pulmonary disease requiring O2, PO steroids or requiring hospitalization; acute/chronic liver disease)

  • Baseline characteristics:

    • Age 72, 45% female, 76% White/14% Asian

    • NYHA 2 (81%), 3 (18%)

    • Mean LVEF 54% (~1/3 each in categories 41-49%, 50-59%, ≥60%)

    • Median NT-proBNP ~950-1000 pg/mL

    • HF hospitalization in last 12 months ~23%

    • Comorbidities: HTN 90-91%, AF 51%, eGFR <60 50%, diabetes 49%

    • Meds: Beta-blocker 86%, ACEI/ARB 79%, ARNI ~2%, MRA 37-38%, digitalis 9-10%

    • SBP 132, HR 70

Intervention: Empagliflozin 10 mg qAM

Comparator: Matching placebo

Outcomes at median 26.2 months (2.2 years)

Efficacy outcomes

Effect on quality of life (using Kansas City Cardiomyopathy Questionnaire [KCCQ]; range 0 [worst] to 100 [best]):

  • More likely to have a clinically-important (≥5/100) improvement in quality of life with empagliflozin vs placebo

    • KCCQ-overall summary score at 1 year: Empagliflozin 49.6% vs placebo 45.8% (+3.8%)

    • Similar effect over time (e.g. difference +4.7% at 3 months vs 3.8% at 12 months)

    • Similar difference if considering clinically-important decline (-4.8% at 1 year) or different cutoffs for improvement (+2.3% for ≥10-point improvement & +3.6% for ≥15-point improvement)

    • Similar difference if considering KCCQ subscores (e.g. +4.6% for KCCQ-total symptoms score [HF symptom burden + frequency] at 1 year)

Cumulative incidence curve for the primary composite outcome showing immediate separation of empagliflozin and placebo curves (suggesting early benefit)

Cumulative incidence curve for the primary composite outcome showing immediate separation of empagliflozin and placebo curves (suggesting early benefit)

Safety outcomes

Effect on biometrics & biomarkers (difference vs placebo):

  • Body weight: -1.3 kg

  • SBP -1.2 mm Hg

  • A1c: -0.2%

  • NT-proBNP: -20 pg/mL

Internal validity: Low risk of bias

  • Computer-generated random sequence using permuted blocks

    • Stratified by geographic region, diabetes status, eGFR <60 or ≥60, & LVEF <50% or ≥50%

  • Allocation concealment by central randomization via interactive response technology

  • Blinding of participants and treating clinicians with matching placebo

  • Blinded outcome adjudication

  • Intention-to-treat analysis

  • 3% loss-to-follow-up for primary outcome, 0.6% for death

Other considerations

Are the results clinically important?

  • Maybe; this will very much depend on individual patient/clinician preferences

    • Overall, likely net clinical benefit based on composite of % who died or had a hospitalization due to any cause

      • HF hospitalizations only accounted for 18% of total hospitalization outcomes in this trial, and therefore the 3.2% absolute reduction in the risk of a first HF hospitalization is diluted in total hospitalizations

      • Neutral effect on all-cause death & inconclusive effect on CV deaths

        • CV death accounted for 55% of deaths (sudden death > HF > other), & non-CV deaths accounted for 45% (infection > malignancy > other)

    • QoL improvement with empagliflozin consistent with results of the PRESERVED-HF trial & effects of SGLT2i on QoL in HFrEF trials

      • Brief summary of PRESERVED-HF:

        • P: 324 patients with NYHA 2-4 HF & LVEF >=45% (mixed HFpEF/HFmrEF) + elevated NT-proBNP/BNP + receiving a diuretic + additional enrichment criteria + eGFR >=20 + SBP >=100

        • I: Dapagliflozin 10 mg daily

        • C: Placebo

        • O: KCCQ-23 @ 3 months

          • Mean +4.5/100 in overall-summary score with dapa

          • Clinically-important improvement: Dapa 45.4% vs placebo 34.9% (+10.5%) at 3 months

How do we apply these results to patient care (generalizability)?

  • Although the study defined “preserved” ejection fraction as >40%, the 2021 universal definition and classification of HF further sub-classify HF as HFmrEF if 41-49% (~1/3 of the study population) & HFpEF if ≥50%

    • Subgroup analysis of the primary outcome comparison based on baseline LVEF suggested attenuation of efficacy with increasing LVEF, with uncertain efficacy with LVEF ≥60%

      • Hazard ratio progressively attenuated from LVEF 41-49% (0.71, 95% CI 0.57-0.88), 50-59% (0.80, 95% CI 0.64-0.99), ≥60% (0.87, 0.69-1.10)

      • Risk of the primary outcome increased with lower LVEF, leading to a greater absolute risk reduction in those with lower baseline LVEF (even if we assume constant 21% relative risk reduction regardless of LVEF)

        • LVEF 41-49%: Risk in placebo group 19.5%, absolute risk reduction 4.1%

        • LVEF 50-59%: Risk in placebo group 16.8%, absolute risk reduction 3.5%

        • LVEF ≥60%: Risk in placebo group 14.9%, absolute risk reduction 3.1%

  • Efficacy on primary outcome (in terms of relative effect) similar in females/males, diabetes/no diabetes, AF/no AF, eGFR <60/≥60, & regardless of race/ethnicity

More to come…