CADTH Reimbursement Review

Empagliflozin (Jardiance)

Sponsor: Boehringer Ingelheim Canada Ltd.

Therapeutic area: Chronic heart failure

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Stakeholder Input

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Table 1: Submitted for Review

NOC = Notice of Compliance; NYHA = New York Heart Association.

^aThe recommended dosage of empagliflozin for the treatment of chronic heart failure is 10 mg once daily.

Introduction

Heart failure (HF) is a clinical condition whereby the heart is unable to adequately pump blood throughout the body to maintain the metabolic needs of tissues and organs. HF results from structural or functional impairment of ventricular filling or ejection of blood.^1,2 There are an estimated 669,000 people in Canada older than 40 years with HF, with an age-standardized prevalence of 3.5%.³ Between 2001 and 2013, the age-standardized incidence rate of HF in Canada has declined, as has the age-standardized all-cause mortality rate among people living with HF.³ However, people in Canada older than 40 years with HF are 6 times more likely to die than those without an HF diagnosis.³ HF with preserved ejection fraction (HFpEF) accounts for at least 50% of the population with HF, and its prevalence is increasing.⁴ Results from the study by Kalogeropoulos et al.⁵ showed that the mortality and morbidity related to HFpEF were similar or comparable to that of patients with HF with reduced ejection fraction (HFrEF). Common symptoms of HF include dyspnea (breathlessness) and fatigue, exercise intolerance, and fluid buildup, which can lead to pulmonary congestion and peripheral edema (mainly feet, ankles, or legs), which significantly affects patients’ quality of life.¹ The current pharmacological management of HFrEF includes diuretics, beta blockers, angiotensin-converting enzyme inhibitors (ACEIs), or angiotensin receptor blockers (ARBs), as well as mineralocorticoid receptor antagonists (MRAs), sacubitril-valsartan, ivabradine, and dapagliflozin.² According to the expert consulted by CADTH, current strategies for the treatment of HFpEF are limited to supportive therapies, such as ARBs, MRAs, and sacubitril-valsartan, that focus on symptom control rather than morbidity or mortality benefits.

Empagliflozin is a sodium-glucose cotransporter-2 (SGLT2) inhibitor. By inhibiting SGLT2, empagliflozin reduces renal reabsorption of filtered glucose and lowers the renal threshold for glucose, thereby increasing urinary glucose excretion. Empagliflozin is approved by Health Canada for use in adults as an adjunct to standard of care (SOC) therapy for the treatment of chronic HF.⁶ Empagliflozin is available as a 10 mg or 25 mg tablet. The recommended dosage of empagliflozin for the treatment of chronic HF is 10 mg once daily.⁶

The objective of this report is to perform a systematic review of the beneficial and harmful effects of empagliflozin at a dose of 10 mg as an adjunct to SOC therapy for the treatment of chronic HF in adults.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups that responded to CADTH’s call for patient input and from the clinical experts consulted by CADTH for the purpose of this review.

Patient Input

The patient and caregiver input received for this review was collected by the HeartLife Foundation, which is a national charity that through its extensive network, engages patients and their caregivers to provide education, support, and access to treatments and research. Information for this review was gathered through in-person interviews with 3 patients and 1 caregiver, an online survey of 12 respondents held in April 2022, a closed virtual support group of 11 respondents, and literature searches from peer-reviewed publications.

Patients highlighted the common symptoms of HF, such as shortness of breath, extreme fatigue, low blood pressure, dizziness, edema, and bloating. In their input, patients acknowledged that HF has no cure and, if left untreated, will become progressively worse over time. Patients expressed an unmet need for new innovative therapies to improve patient outcomes in terms of both quantity and quality of life because many patients are intolerant to beta blockers and, in some cases, to ACEIs. Respondents expressed a desire to have greater access to proven therapies and improved functional capacity and quality of life, as they would like to spend time with loved ones, be able to work on a regular basis, pursue outdoor activities, and be able to travel. Sixteen respondents with experience using empagliflozin reported the drug was effective in terms of improving ejection fraction and energy level and reducing shortness of breath. According to the HeartLife Foundation survey (N = 12), approximately 33.3% of respondents felt better after taking empagliflozin, while 8.3% reported they felt worse. About 33.3% of respondents described their side effects as manageable, whereas 25% said they were not manageable. The most frequently reported side effects were fatigue and urinary tract infections.

Clinician Input

Input From the Clinical Experts Consulted by CADTH

According to the clinical experts consulted by CADTH, many patients with HFrEF are not being assessed by specialists in Canada, and assistance from other specialists is needed, given the growing number of patients. The clinical experts further noted that the use of goal-directed guideline-recommended pharmacological therapy and medical devices in patients with HFrEF remains suboptimal. The clinical experts highlighted that current treatment strategies in HFpEF are limited to supportive therapies focusing on symptom control rather than morbidity or mortality benefit, including ARBs, MRAs, and ARNIs, while data on SGLT2 inhibitors show clear benefit in this population. The clinical experts indicated that empagliflozin can be used as an alternative to dapagliflozin in combination with other goal-directed guideline-recommended pharmacotherapy in patients with HFrEF, and it is likely to be a first-line therapy for patients with HFpEF, given the ease of use, strength of evidence, safety profile, and familiarity with the use of empagliflozin in patients with type 2 diabetes mellitus. The population least likely to benefit from empagliflozin treatment are patients with low N-terminal pro-brain natriuretic peptide (NT-proBNP) levels and those with NYHA classes I and IV due to limited clinical evidence. The clinical experts indicated that the response to therapy in clinical practice is assessed based on the frequency of hospitalizations for HF, which in turn may lead to a reduction in mortality, improved quality of life, and a slower decline in kidney function. The clinical experts further noted that admission for HF is a major cost burden in the health care system. The clinical experts identified the following factors to consider when deciding to discontinue treatment with empagliflozin:

• the development of severe kidney dysfunction (estimated glomerular filtration rate [eGFR] < 15 mL/min/1.73 m²) and euglycemic diabetic ketoacidosis in patients with diabetes as important AEs

• NYHA functional class IV.

The clinical experts highlighted that empagliflozin is already widely used by primary care providers and endocrinologists for the management of diabetes, by nephrologists to reduce decline in kidney function, and by cardiologists.

Clinician Group Input

No clinician group input was received for this review.

Drug Program Input

Input was obtained from the drug programs that participate in the CADTH reimbursement review process. Key issues raised by the drug plans included concerns over the relevant comparator for empagliflozin, evidence to support the combination use of empagliflozin with sacubitril-valsartan and/or ivabradine, and the potential for additional indications for Forxiga (dapagliflozin) and Jardiance to influence future price negotiations. The clinical experts consulted by CADTH indicated there is no clear evidence to support the benefit of Forxiga over Jardiance in patients with HF, as no head-to-head trials are available yet; however, both drugs showed similar benefits in patients with HFrEF. The clinical experts further noted that empagliflozin would be an addition to the current therapy in patients with HFpEF while, in patients with HFrEF, this would be an alternative to dapagliflozin. The clinical experts do not foresee the combination use of empagliflozin and sacubitril-valsartan and/or ivabradine as an issue. The clinical experts agreed that additional indications would have an impact on future negotiations and acknowledge that the availability of empagliflozin may potentially benefit the payers in terms of price negotiations.

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

Two phase III, double-blind, placebo-controlled randomized controlled trials (EMPEROR-Reduced and EMPEROR-Preserved) were pivotal trials and included in the systematic review. Both trials were multinational and multicentre and included Canadian sites. The EMPEROR-Reduced trial (N = 3,730) was designed to assess the superiority of empagliflozin at 10 mg compared with matched placebo as an adjunct to SOC treatment in patients with HFrEF (LVEF ≤ 40%). In EMPEROR-Reduced, patients had a mean age of 66.8 years (standard deviation [SD] = 11.0 years), 76.1% were male, and the mean LVEF was 27.5% (SD = 6.0%), and most patients (75.1%) had an NYHA functional class of II. The EMPEROR-Preserved trial (N = 5,988) was designed to assess the superiority of empagliflozin at 10 mg compared with matched placebo as an adjunct to SOC treatment in patients with HFpEF (LVEF > 40%). In EMPEROR-Preserved, patients had a mean age of 71.9 years (SD = 9.4 years), 55.3% were male, the mean LVEF was 54.3% (SD = 8.8%), and most patients (81.5%) had an NYHA functional class of II.

In both EMPEROR trials, the primary efficacy end point was the time to first event of adjudicated cardiovascular (CV) death or hospitalization for heart failure (HHF). The key secondary end points were occurrence of adjudicated HHF (first and recurrent) and eGFR (calculated using Chronic Kidney Disease Epidemiology Collaboration creatinine [CKD-EPIcr] equation) slope of change from baseline. Other secondary and further exploratory outcomes in either trial that were important to the CADTH review included other hospitalization-related and mortality outcomes, as well as patient-reported outcomes such as health-related quality of life (HRQoL) and HF symptoms assessed by the Kansas City Cardiomyopathy Questionnaire (KCCQ) and 5-Level EQ-5D (EQ-5D-5L) questionnaires, and functional ability. Harms and notable harms were assessed.

Efficacy Results

A summary of the results for the main efficacy and safety outcomes of both EMPEROR trials is presented in Table 2. Statistical testing in both pivotal trials was conducted based on a hierarchical testing procedure. The following outcomes were controlled for multiplicity in both EMPEROR trials: time to first event of adjudicated CV death or HHF, occurrence of HHF (fist and recurrent), and eGFR (CKD-EPIcr equation) equation slope of change from baseline. The clinical experts consulted by CADTH for this review indicated that both HHF and CV death are the most important outcomes to assess the treatment response in patients with HF, while change in eGFR is not commonly used in clinical practice. Other secondary and further end points were tested in a non-hierarchical fashion without adjustments for multiplicity.

Time to First Event of Adjudicated CV Death or HHF

In EMPEROR-Reduced, a composite of time to first event of adjudicated CV death or HHF occurred in 361 patients (19.4%) in the empagliflozin group and 462 patients (24.7%) in the placebo group. The hazard ratio (HR) for time to first event of adjudicated CV death or HHF was 0.75 (95% confidence interval [CI], 0.65 to 0.86; P < 0.0001) in favour of the empagliflozin group. Although individual components of the composite primary end point were not formally tested for significance, the proportion of HHF was lower in the empagliflozin group (13.2%) compared with placebo (18.3%), while the total proportion of CV deaths was similar across the treatment groups (10.0% versus 10.8%, respectively).

In EMPEROR-Preserved, a composite of time to first event of adjudicated CV death or HHF occurred in 415 patients (13.8%) in the empagliflozin group and 511 patients (17.1%) in the placebo group. The HR for time to first event of adjudicated CV death or HHF was 0.79 (95% CI, 0.69 to 0.90; P = 0.0003) in favour of the empagliflozin group. The proportion of HHF was lower in the empagliflozin group (8.6%) relative to placebo (11.8%), while the total proportion of CV deaths was similar across the treatment groups (7.3% versus 8.2% in the empagliflozin and placebo groups, respectively).

Occurrence of HHF (First and Recurrent)

In EMPEROR-Reduced, the total number of HHF events (first and recurrent) was lower in patients who received empagliflozin compared with those who received placebo (388 versus 553, respectively). The hazard rate of recurrent HHF was significantly reduced in the empagliflozin group compared with placebo, with an HR of 0.70 (95% CI, 0.58 to 0.85; P = 0.0003).

In EMPEROR-Preserved, the total number of HHF events was lower in patients who received empagliflozin compared with those who received placebo (407 versus 541, respectively). The hazard rate of recurrent HHF was significantly reduced in the empagliflozin group compared with placebo, with an HR of 0.73 (95% CI, 0.61 to 0.88; P = 0.0009).

eGFR Slope of Change From Baseline

In EMPEROR-Reduced, over the double-blind treatment period, the rate of decline in the eGFR (CKD-EPIcr equation) per year was slower in the empagliflozin group (−0.55 mL/min/1.73 m² per year; 95% CI, −0.99 to −0.10) than in the placebo group (−2.28 mL/min/1.73 m² per year; 95% CI, −2.73 to −1.83), with a between-group difference in slope of 1.73 per year (95% CI, 1.10 to 2.37; P < 0.0001).

In EMPEROR-Preserved, over the double-blind treatment period, the rate of decline in the eGFR (CKD-EPIcr equation) per year was slower in the empagliflozin group (−1.25 mL/min/1.73 m² per year; ||| ||| ||||| || |||||) than in the placebo group (−2.62 mL/min/1.73 m² per year; ||| ||| ||||| || |||||), with a between-group difference in slope of 1.36 per year (95% CI, 1.06 to 1.66; P < 0.001).

Health-Related Quality of Life and HF Symptoms

Both patients and clinical experts highlighted patient-reported end points as important outcomes and important treatment goals for patients. However, the interpretation of the results must be made with caution, as multiplicity was not controlled for in the analysis of the KCCQ scores.

KCCQ Clinical Summary Score

In EMPEROR-Reduced, the analysis based on the randomized set (RS) showed a smaller decline from baseline of −1.30 points (standard error [SE] = 0.69) in the empagliflozin group than in the placebo group (−3.36 points; SE = 0.69) in the KCCQ clinical summary score (KCCQ-CSS) at week 52, with an adjusted mean difference of 2.06 (95% CI, 0.16 to 3.96) favouring empagliflozin. A responder analysis showed that at week 52, 40.0% of patients in the empagliflozin group reported at least a 5-point increase in KCCQ-CSS, compared with placebo (35.9%) (odds ratio [OR] = 1.23; 95% CI, 1.05 to 1.45).

|| |||||||||||||||||| ||| |||||||| ||||| || ||| || |||||| | |||||| |||||||| || |||| |||||| ||| | ||||| || ||| ||||||||||||| ||||| |||||||| || ||||||| |||||| ||||||| || | ||||| || ||| |||| |||||||| ||||||| ||||| |||| |||||||| || |||| ||| |||| || |||||||| |||| |||||||||| || |||| |||| ||| |||| || ||||| |||||||| |||||||||||||| ||||||||| |||||||| |||||| |||| || |||| ||| ||||| || |||||||| || ||| ||||||||||||| ||||| |||||||| || ||||| | ||||||| |||||||| || |||| |||||||| ||||||| |||||| |||||||| || ||||||| ||||||| ||| | ||||| ||| ||| |||| || ||||||

KCCQ Total Symptom Score

|| |||||||||||||||| ||| |||||||| ||||| || ||| || |||||| | ||||||| ||||||| |||| |||||||| || | ||||| |||||| ||| | ||||| || ||| ||||||||||||| ||||| |||||||| || ||||||| |||||| ||||||| || | ||||| || ||| |||| ||||| ||||||| ||||| || |||| ||| |||| || |||||||| |||| |||||||||| || |||| |||| ||| |||| || ||||| |||||||| |||||||||||||| ||||||||| |||||||| |||||| |||| || |||| ||| ||||| || |||||||| || ||| ||||||||||||| ||||| |||||||| || ||||| | ||||||| |||||||| || |||| |||||||| ||||||| |||||| |||||||| || ||||||| ||||||| ||| | ||||| ||| ||| |||| || ||||||

|| |||||||||||||||||| ||| |||||||| ||||| || ||| || |||||| || |||||||| |||| |||||||| || |||| |||||| ||| | ||||| || ||| ||||||||||||| ||||| |||||||| || ||||||| |||||| ||||||| || | ||||| || ||| |||| ||||| ||||||| ||||| || |||| ||| |||| || |||||||| |||| |||||||||| || |||| |||| ||| |||| || ||||| |||||||| |||||||||||||| ||||||||| |||||||| |||||| |||| || |||| ||| ||||| || |||||||| || ||| ||||||||||||| ||||| |||||||| || ||||| | ||||||| |||||||| || |||| ||||| ||||||| |||||| |||||||| || ||||||| ||||||| ||| | ||||| ||| ||| |||| || ||||||

Harms Results

Adverse Events

In EMPEROR-Reduced, 1,420 (76.2%) patients in the empagliflozin group and 1,463 (78.5%) patients in the placebo group experienced at least 1 adverse event (AE). Patients in the empagliflozin and placebo groups experienced treatment-emergent AEs (TEAEs) at a similar frequency (15.2% and 12.2%, respectively). The most common TEAEs occurring in at least 0.5% of patients in the empagliflozin and placebo groups were hypotension (2.3% versus 1.8%, respectively), renal impairment (1.4% and 1.1%, respectively), urinary tract infection (1.4% in each group), and ||||||||||||| ||||| ||| ||||| |||||||||||||.

In EMPEROR-Preserved, 2,574 patients (85.9%) in the empagliflozin group and 2,585 patients (86.5%) in the placebo group experienced at least 1 AE. |||||||| || ||| ||||||||||||| ||| ||||||| |||||| ||||||||||| ||||| || | ||||||| ||||||||| |||||| ||| |||||| |||||||||||||| ||| |||| |||||| ||||| ||||||||| || || ||||| |||| || |||||||| || ||| ||||||||||||| || ||||||| |||||| |||| ||||||| ||||| ||||||||| ||||| ||| ||||| |||||||||||||| ||||||||||| ||||| ||| ||||| |||||||||||||| ||||| |||||||||| ||||| ||| ||||| |||||||||||||| ||| ||||||||||||| ||||| || |||| ||||||.

In EMPEROR-Reduced, 772 patients (41.4%) in the empagliflozin group and 896 patients (48.1%) in the placebo group experienced 1 or more serious AEs (SAEs). In EMPEROR-Preserved, 1,436 patients (47.9%) in the empagliflozin group and 1,543 patients (51.6%) in the placebo group experienced 1 or more SAEs.

Withdrawals Due to Adverse Events

In EMPEROR-Reduced, the overall frequency of AEs leading to treatment discontinuation was similar between the treatment groups in both pivotal trials (17.3% and 17.6% in the empagliflozin and placebo groups in EMPEROR-Reduced, ||| ||||| ||| ||||| || |||||||||||||||||| |||||||||||||. The most frequently reported types of withdrawals due to AEs in both trials were cardiac failure, death, acute myocardial infarction, renal impairment, and urinary tract infection.

Mortality

|| ||| ||||||||||||||| |||||| ||| |||||||||| || ||||| ||| |||||| ||| |||||||||||| ||||||||| ||||| ||| ||||||| |||||| ||| ||||||||| |||||| ||||||| || |||||||||||||||||| |||| || ||| || ||| ||||||||||||| ||||| ||| |||| || ||| ||||||| |||||| |||||||| || ||||||

Notable Harms

The frequency of notable harms identified in the protocol were comparable between the treatment groups.

In both EMPEROR trials, acute renal failure was the most commonly reported notable AE (9.4% versus 10.3%, and 12.1% versus 12.8% in the empagliflozin and placebo groups in EMPEROR-Reduced and EMPEROR-Preserved, respectively), followed by hypotension (9.4% versus 8.7%, and 10.4% versus 8.6% in the empagliflozin and placebo groups in EMPEROR-Reduced and EMPEROR-Preserved, respectively), urinary tract infection (4.9% versus 4.5%, and 9.9 versus 8.1% in the empagliflozin and placebo groups in EMPEROR-Reduced and EMPEROR-Preserved, respectively), and bone fracture (2.4% versus 2.3%, and 4.5% versus 4.2% in the empagliflozin and placebo groups in EMPEROR-Reduced and EMPEROR-Preserved, respectively). No new safety concerns were identified.

Table 2: Summary of Key Results From Pivotal and Protocol-Selected Studies

AE = adverse event; CI = confidence interval; CKD-EPIcr = Chronic Kidney Disease Epidemiology Collaboration creatinine; Cr = creatinine; CV = cardiovascular; eGFR = estimated glomerular filtration rate; HHF = hospitalization for heart failure; HR = hazard ratio; KCCQ = Kansas City Cardiomyopathy Questionnaire; LLA = lower-limb amputation; NYHA = New York Heart Association; OR = odds ratio; RS = randomized set; SAE = serious adverse event; SE = standard error; TEAE = treatment-emergent adverse event; TS = treated set; WDAE = withdrawals due to adverse event.

^aAn independent group of medical experts performed a central, blinded adjudication of the outcomes.

^bIncidence rate was calculated as the number of patients with events per 100 person-years at risk.

^cCox proportional hazards model included the following factors: treatment, age, geographical region, diabetes status, sex, LVEF, and baseline eGFR (CKD-EPIcr equation).

^dBased on the reduced 2-sided significance level of 0.0496 resulting from the interim analysis.

^eJoint frailty model included the following factors: age, baseline eGFR (CKD-EPIcr equation), geographical region, baseline diabetes status, sex, baseline LVEF, and treatment.

^fPositive correlation between recurrent (HHF) and terminal events (CV death) if alpha > 0.

^gModel included the following factors: age, baseline eGFR, region, baseline diabetes status, sex, baseline LVEF, baseline eGFR-by-time interaction, treatment-by-time interaction, and treatment. Intercept and slope were allowed to vary randomly between patients.

^hBased on RS, including both on- and off-treatment values. For patients who died, the worst score (score of 0) was imputed at all subsequent scheduled visits after the date of death.

ⁱMixed model for repeated measures includes age, baseline eGFR (CKD−EPIcr) as linear covariates, region, baseline diabetes status, sex, baseline LVEF, week reachable, treatment-by-visit interaction, and baseline KCCQ score by visit interaction as fixed effects.

^jLogistic regression includes baseline KCCQ score, baseline eGFR (CKD−EPIcr), treatment, region, diabetes at baseline, sex, and baseline LVEF. Patients who were lost to follow-up, withdrew consent, or died before planned week 52 visit were considered as having deterioration.

^k95% CI was not adjusted for multiple comparisons.

^lDefined by a narrow standardized Medical Dictionary for Regulatory Activities (MedDRA) query (SMQ).

^mDefined by a Boehringer Ingelheim customized MedDRA query (BIcMQ).

ⁿHypoglycemic AEs with a plasma glucose value of ≤ 70 mg/dL or where assistance was required.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Critical Appraisal

Internal Validity

Both the EMPEROR-Reduced and EMPEROR-Preserved trials appeared to have used accepted methods for blinding, allocation concealment, and randomization with stratification. For both EMPEROR trials, a computer-generated block randomization scheme was used, and randomization with stratifications was performed centrally, which typically has a low risk of bias. The demographic and baseline patient characteristics appeared to be generally balanced between the treatment groups in both trials, so randomization was maintained. Both EMPEROR trials included only patients with elevated NT-proBNP, as high concentrations of NT-proBNP can confirm HF in patients who present with dyspnea when the clinical diagnosis remains uncertain.² However, the clinical experts consulted by CADTH highlighted that physicians only need to perform NT-proBNP tests in 10% to 20% of cases, when they are unsure of the diagnosis of HF. A relatively high proportion of patients prematurely discontinued the trial medication (26.7% and 31.5% in EMPEROR-Reduced and EMPEROR-Preserved, respectively, including fatal events), while the cause of discontinuations occurred at a similar frequency between the treatment groups. The clinical experts noted that a high proportion of AEs leading to treatment discontinuation were fatal, which reflects the natural history of the HF more than intolerance to the drug under review. An independent blinded committee of clinical experts performed a central adjudication of the primary and key secondary outcomes based on criteria defined a priori. The clinical experts consulted indicated that CV death and HHF are the main outcomes used in clinical practice to assess the response to HF treatment. While improvement in HRQoL, HF symptoms, and functional ability were of primary importance to patients with HF according to the patient group input, these were exploratory outcomes and were outside the statistical testing hierarchy; thus, the results should be viewed as supportive evidence for the overall effect of empagliflozin. The symptoms associated with HF and HRQoL were assessed using KCCQ and EQ-5D-5L instruments. The clinical experts indicated that these tools are not used in clinical practice but are used in multiple studies, allowing comparisons between different treatments. Since treatment discontinuation rates were relatively high across both treatment groups, and many patients did not complete the KCCQ or EQ-5D-5L at baseline or follow-up, there is a high risk of bias, as the patients who completed the questionnaires may be fundamentally different from those who did not complete (e.g., differences in treatment response, AEs). Assessment of functional ability was based on the change in NYHA functional class from baseline at week 52 using descriptive statistics. The evidence of empagliflozin in patients with chronic HF was limited by 2 placebo-controlled pivotal trials, and no head-to-head evidence for empagliflozin compared against other comparators, including dapagliflozin or sacubitril-valsartan in the HFrEF population, were available for this review.

External Validity

In general, the clinical experts consulted by CADTH for this review confirmed that the populations of both the EMPEROR-Reduced and EMPEROR-Preserved trials were similar to the patients seen in Canadian clinics, and the study results would be generalizable to patients with HF in Canada, with some limitations. While empagliflozin has been approved by Health Canada for use as an adjunct to SOC therapy in patients with chronic HF regardless of NYHA class, CADTH was unable to draw conclusions related to patients with NYHA functional classes I and IV, since both trials excluded patients who had NYHA class I, and there was a very small number of patients who had NYHA class IV. One of the clinical experts consulted highlighted that the benefit of empagliflozin in patients with NYHA class IV is unclear due to limited clinical data and high mortality, while another clinical expert indicated that he would prescribe empagliflozin to patients with NYHA class IV. In addition, the clinical experts indicated they would not prescribe empagliflozin to patients with chronic HF with NYHA class I, as they are asymptomatic, which is consistent with the reimbursement request. About 48% of patients in both trials did not pass the screening, most commonly because of NT-proBNP levels below the pre-specified thresholds at screening, which further reduces the generalizability of the results. The clinical experts consulted indicated that NT-proBNP testing is not widely available in Canada, as some jurisdictions have limited access to it; thus, this patient selection criterion would be difficult to implement in clinical practice. The clinical experts further noted that this inclusion criterion likely created an enriched patient population in both trials; the patients with elevated NT-proBNP appeared to be sicker and could benefit more from treatment with empagliflozin than the population in the real-world setting. In the EMPEROR-Preserved trial, about 33% of patients had mid-range LVEF (41% to 49%); however, the clinical experts do not expect this to be a major issue with the generalizability of the trial results, as the LVEF definition is arbitrary, and estimates of LVEF may vary depending on the patient or technical factors as well as on clinical deterioration. The clinical experts consulted noted the patients included in both EMPEROR trials were younger, as the median age of the population with HF in the real-world setting is approximately 75 years. The generalizability of the EMPEROR-Reduced trial results may be compromised by the high proportion of males (more than 75%) who were enrolled, as half of the population with HFrEF in Canada is female. Nonetheless, the clinical experts consulted noted that they would treat both male and female patients with chronic HF with empagliflozin. The majority of patients in both EMPEROR trials were receiving guideline-recommended treatment of HF; thus, they represented patients who were optimally managed, while the clinical experts noted that a goal-directed treatment of HF is suboptimal in clinical settings. Lastly, although the recommended dose of empagliflozin for the treatment of HF is 10 mg, the clinical experts indicated that both the 10 mg and 25 mg doses of empagliflozin are used in clinical practice.

Indirect Comparisons

Description of Studies

In the absence of direct comparative evidence from trials, the aim of the indirect treatment comparison (ITC) conducted according to the methodology described by Bucher et al. (1997)⁹ was to compare the efficacy of empagliflozin plus SOC versus dapagliflozin plus SOC in patients with HFrEF. ||| ||||||| ||||| || |||||||| ||| ||| || ||| ||| ||||||| ||||| | |||||| ||| |||| ||||| ||||| ||||||||||| ||| ||||||| |||| |||| ||||||||| ||| ||| |||||||| || |||||||||| ||||||||| || |||| || ||||| ||| ||| ||||||| |||| ||| ||||||| |||||| |||| ||| |||| ||||| ||| ||||||||||||||| ||| ||||||| ||||||| |||| ||||||||||||||| ||| ||||||| |||| ||||| || |||||| |||||| |||||||||||| ||||||| ||||||||| ||| || |||||| ||||||||||||| || |||||||||||||| ||||||||||||| ||| |||| |||||||| || ||||||||| ||| |||||||| |||||| || ||||||| ||||| |||||||| |||||||| || ||||||||||||||| ||| |||||||| ||| ||||||||| |||||||| |||| ||||||| |||| ||| ||||||||| || |||||||||||||||| ||||| ||||||| ||| |||||| |||||||| ||||||||| ||| |||||| |||| |||||||||| ||||| ||||||||| || ||| ||||| ||||||||| ||||||||| ||| ||||||| |||||||| || ||||||||||||||| ||| ||||||||||||| ||||||||||||||| ||| ||||| ||||||| || || |||||| ||||||| || |||||||| |||| ||||||||| |||||||| |||| |||||||| |||||| ||||| ||||||| ||||||| ||| ||| |||||||| || ||| ||||||||| ||||| ||||||||| |||| |||||||||| ||||||||||| |||| ||||||| |||| |||||||| || ||| |||||||| ||||||||| || ||| |||||||| |||| |||| || |||| |||||||||| || |||| ||| |||| || |||||

Efficacy Results

||| |||||| ||| |||||||| ||||||||||||| ||||||| ||||||||||||| || |||||||| |||| ||||| || |||| || ||||| ||||| || ||||||||||| || ||||| || ||||||||||| |||| |||| || ||||| ||||||||||| |||| |||| || ||||||||||| || |||||| |||| || ||||||||| |||||||||| ||| |||| || ||||||||| ||||| ||||||||| ||||| |||| || ||||||||| |||| |||||||| || | || ||||||||| | ||| ||| |||||||||| || ||||||||||||| |||||| |||||||||||||| ||| ||||||| |||||||| |||| |||| |||||||| |||| || ||||| ||||| || ||||||||||| || ||||| || ||||||||||| |||| |||||| || |||||||||| ||||||| ||||||||||||| ||| ||||||||||||| |||| | || |||| ||| || |||| ||||| || ||||||

Critical Appraisal

The sponsor conducted a Bucher ITC comparing empagliflozin against dapagliflozin in patients with HFrEF. Studies were identified from a systematic review; however, the included studies from the systematic review were further refined on an ad hoc basis to arrive at the 2 pivotal trials for each drug to be analyzed in the ITC, potentially introducing selection bias. The Bucher methodology for ITC assumes all differences in patient characteristics or study design have no impact on treatment effects, estimating relative treatment effects using the common comparator arm of 2 treatments that have not been investigated in a head-to-head study. Important differences between the EMPEROR-Reduced and DAPA-HF trials included the broader primary composite end point in DAPA-HF (the impact of which is uncertain), baseline characteristics indicating sicker patients in EMPEROR-Reduced, potentially biasing the results in favour of empagliflozin, and the more effective basket of background SOC therapies used in EMPEROR-Reduced, potentially biasing results against empagliflozin | ||||||| ||| ||||||||| ||||||||||| ||||||| ||| | |||||||| ||||| |||| ||| |||||| || || |||||||| ||| | |||||||||| || ||||||||| ||||||| ||||||| ||||||||||||| ||| |||||||||||||| |||||||| |||| ||| ||||||| || |||||||| ||||||| ||||||||||. Two additional ITCs were identified from the literature search conducted by CADTH. Given the lack of details provided, the results were highly uncertain; however, the results indicating no difference between empagliflozin and dapagliflozin were consistent with the ||||||| ||||||||| ||| ||| the opinion of the clinical experts consulted.

Other Relevant Evidence

In addition to the pivotal trials, EMPEROR-Reduced and EMPEROR-Preserved, the CADTH review team identified 2 phase III, multi-centre, randomized, double-blind, placebo-controlled trials that met systematic review inclusion criteria and that were considered relevant for this report: EMPERIAL-Reduced and EMPERIAL-Preserved. However, the CADTH review team did not include the EMPERIAL-Reduced and EMPERIAL-Preserved studies because 1 of the outcomes of interest, KCCQ, was considered exploratory, as the primary end point was not met in the 2 trials. Therefore, although the EMPERIAL-Reduced and EMPERIAL-Preserved studies were not included in the main report, the CADTH review team summarized and appraised the studies to provide additional supportive evidence for KCCQ and safety.

Description of Studies

EMPERIAL-Reduced

The EMPERIAL-Reduced (effect of empagliflozin on exercise ability and HF symptoms in patients with chronic HFrEF) trial was a phase III, multicentre, randomized, double-blind, placebo-controlled study that aimed to evaluate the effect of empagliflozin (10 mg once daily) on exercise capacity and patient-reported outcomes compared with placebo in patients with HFrEF (defined as LVEF < 40%) with or without type 2 diabetes mellitus. A total of 312 patients were enrolled across 109 sites in 11 countries (Australia, Canada, Germany, Greece, Italy, Norway, Poland, Portugal, Spain, Sweden, and the US). Patients were randomized in a 1:1 ratio to receive either empagliflozin at a dosage of 10 mg once daily (n = 156) or matching placebo (n = 156) in a double-blind manner. Among these 312 patients, the mean age was 69.0 years (SD = 10.2 years) and the majority of patients were male (74.4%) and White (84.3%). The cause of HF was ischemic in 50.6% (n = 158) of participants, the mean LVEF was 30.3% (SD = 6.7%), and diabetes was present in 59.9% (n = 187) of patients. The study was funded by Boehringer Ingelheim.^10,11

EMPERIAL-Preserved

The EMPERIAL-Preserved (effect of empagliflozin on exercise ability and HF symptoms in patients with chronic HFpEF) trial was a phase III, multicentre, randomized, double-blind, placebo-controlled study trial that aimed to evaluate the effect of empagliflozin (10 mg once daily) on exercise capacity and patient-reported outcomes as compared with placebo in patients with HFpEF (defined as LVEF > 40%), with or without type 2 diabetes mellitus. A total of 315 patients were enrolled across 108 sites in 11 countries (Australia, Canada, Germany, Greece, Italy, Norway, Poland, Portugal, Spain, Sweden, and the US). Patients were randomized in a 1:1 ratio to receive either empagliflozin at a dose of 10 mg once daily (n = 157) or matching placebo (n = 158) in a double-blind manner. Among these 315 patients, the mean age was 73.5 years (SD = 8.8 years), and the majority of patients were male (56.8%) and White (87.3%). The cause of HF was ischemic in 50.6% (n = 158) of participants, the mean LVEF was 53.1% (SD = 8.0%), and diabetes was present in 51.1% (n = 161) of patients. The study was funded by Boehringer Ingelheim.^10,11

Efficacy Results

The primary end point was change from baseline in 6-minute walk test distance (6MWTD) at week 12. Key secondary end points were change from baseline in KCCQ total symptom score (KCCQ-TSS) and Chronic Heart Failure Questionnaire Self-Administered Standardized Format (CHQ-SAS) dyspnea score at week 12. Results for the KCCQ-TSS and CHQ-SAS dyspnea score are presented in accordance with the protocol for the CADTH review. The median difference from baseline to week 12, empagliflozin versus placebo, in KCCQ-TSS was 3.13 (95% CI, 0.00 to 7.29) and 2.08 (95% CI, −2.08 to 6.25) in EMPERIAL-Reduced and EMPERIAL-Preserved, respectively. The median difference, empagliflozin versus placebo, in CHQ-SAS dyspnea score was 0.10 (95% CI, −0.20 to 0.40) and −0.07 (95% CI, −0.35 to 0.20) in EMPERIAL-Reduced and EMPERIAL-Preserved, respectively.¹¹ The results for other symptom outcomes are presented in the Other Relevant Evidence section.

Harms Results

There was no notable difference for empagliflozin versus placebo regarding the overall frequencies of any AE or any AE leading to treatment discontinuation in both trials. SAEs were reported less frequently with empagliflozin than with placebo in EMPERIAL-Reduced (12.7% for empagliflozin versus 18.4% for placebo) and EMPERIAL-Preserved (13.5% for empagliflozin versus 17.3% for placebo). Decreased kidney function was reported with similar frequencies in both groups. No ketoacidosis or confirmed hypoglycemic events occurred in participants without type 2 diabetes. No new safety concerns were identified.¹¹

Critical Appraisal

The following limitations were identified:

• HF is a chronic condition, which means the progression of HF is generally slow, thus the assessment of change in outcomes may require a long-term follow-up period.

• The follow-up period for the EMPERIAL-Reduced and EMPERIAL-Preserved trials was 12 weeks, which may not be sufficient to assess meaningful changes in the outcome measures.

• The EMPERIAL trials were powered to detect an improvement of 30 m in 6MWTD; however, the study sample size may not be sufficient to detect any between-group changes of less than 30 m.

• As the primary end point (change from baseline in the 6MWTD at week 12) was not met, the analyses of all secondary outcomes, such as the KCCQ-TSS and CHQ-SAS dyspnea score, were considered exploratory.

• While the changes in the KCCQ-TSS and CHQ-SAS dyspnea score may suggest a possible favourable effect of empagliflozin in patients with HFrEF, these results are considered exploratory.

• The baseline demographic and baseline characteristics (sex and 6MWTD) were suggestive of an over-representation of male patients with lower functioning status, which may compromise the representativeness of the study sample compared with the general population of adult patients with HF.

Although the EMPERIAL studies provide additional data on the effectiveness and safety of empagliflozin in patients with HF, the limitations identified introduce uncertainty.

Conclusions

Overall, the efficacy of empagliflozin for use in adults as an adjunct to SOC therapy for the treatment of chronic HF has been demonstrated. Based on the EMPEROR-Reduced and EMPEROR-Preserved trials, empagliflozin is significantly more efficacious than placebo in reducing the hazard rate of the first event of adjudicated CV death or HHF, as well as the occurrence of adjudicated first and recurrent HHF. The annual rate of decline in the eGFR was slower in the empagliflozin group than in the placebo group in both pivotal trials. The benefit of empagliflozin on patient-valued outcomes such as HRQoL, functional ability, and symptoms associated with HF should be viewed as supportive evidence only for the overall effect of empagliflozin. The evidence of empagliflozin in patients with chronic HF was limited by 2 placebo-controlled pivotal trials, and no head-to-head evidence of empagliflozin compared against other relevant comparators, including dapagliflozin, sacubitril-valsartan, and ivabradine in the HFrEF population, were available for this review. The median duration of EMPEROR-Reduced and EMPEROR-Preserved was 1.31 years and 2.15 years, respectively. Thus, long-term efficacy and safety in patients with chronic HF is uncertain. Although empagliflozin has been approved by Health Canada for use as an adjunct to SOC therapy in patients with chronic HF regardless of NYHA class, CADTH was unable to draw conclusions related to patients with NYHA functional classes I and IV because both pivotal trials excluded patients who had NYHA class I, and there was a very small proportion of patients who had NYHA class IV. No new safety signals were identified in patients with HF with reduced and preserved ejection fractions. ||| |||||||| ||||||||| |||| ||||||| |||||||||||| ||||||||| |||| ||||| || || |||||||||| ||||||| ||||||||||||| ||| ||||||||||||| || ||| ||||| ||||||||||| ||||| ||| |||||||||| |||| ||| ||||||| || |||||||| ||||||| ||||||||||

Introduction

Disease Background

HF, sometimes referred to as congestive HF, is a clinical condition whereby the heart is unable to adequately pump blood throughout the body to maintain the metabolic needs of tissues and organs. HF results from structural or functional impairment of ventricular filling or ejection of blood.^1,2 HF is classified based on the percentage of blood that is being pumped out of the left ventricle otherwise known as LVEF.² HFrEF is defined as HF with an LVEF of 40% or less, whereas having an LVEF of 50% or greater is termed HFpEF. HF with an LVEF in the range of 40% to 49% is defined as HF with mid-range LVEF, which may represent a variety of phenotypes, including patients transitioning to and from HFpEF.² There is uncertainty regarding management strategies, including surveillance, treatment, and prognosis, for patients with HF with mid-range ejection fractions.² Additionally, HF with recovered ejection fraction is defined as an LVEF of more than 40% but with a previously documented LVEF of 40% or less.⁵ According to the clinical experts, consulted by CADTH, assessment of LVEF is a routine part of the diagnosis and management of HF and can be carried out using a variety of techniques, the most common being via 2D echocardiography, as well as nuclear medicine, angiography, and MRI. Clinical experts also noted that the classification of LVEF is arbitrary, and LVEF estimates may vary depending on the patient or technical factors as well as clinical deterioration. Another common classification system is the NYHA functional classification, which is based on HF symptoms and patients’ ability to perform physical activities. Patients in NYHA class I have no symptoms (asymptomatic) and those in class IV have symptoms at rest or with any minimal activity.²

Common symptoms of HF include dyspnea (breathlessness) and fatigue, exercise intolerance and fluid buildup, which in turn may lead to pulmonary congestion and peripheral edema (mainly feet, ankles, or legs) that is significantly affecting their quality of life.¹ Other possible symptoms include a rapid heartbeat, frequent urination at night, difficulties concentrating, weight gain, and a dry cough, although these symptoms are present in other conditions, making it difficult to distinguish HF from other medical conditions, particularly during early stages. Depending on symptom severity, HF may go unnoticed, only causing minor symptoms, but patients with advanced HF may find it difficult to carry out normal everyday activities.¹² HF leads to a progressive decline in cardiac function over time, with persistent signs and symptoms interspersed with acute episodes of decompensation needing hospital care.

There are an estimated 669,000 people in Canada older than 40 years with HF, with an age-standardized prevalence of 3.5%.³ Between 2001 and 2013, the age-standardized incidence rate of HF in Canada has declined, as has the age-standardized all-cause mortality rate among people living with HF.³ However, people in Canada older than 40 years with HF are 6 times more likely to die than those without an HF diagnosis.³ HFpEF accounts for at least 50% of the population with HF, and its prevalence is increasing.⁴ Patients with HF, especially those with HFpEF, are often afflicted with multiple comorbid conditions, such as hypertension, atrial fibrillation, renal disease, and diabetes mellitus, contributing to increased morbidity and mortality and impaired quality of life.⁴ Evidence shows that the mortality and morbidity for HFpEF is similar or comparable to those in HF with reduced ejection fraction.⁵ The economic burden due to HF is substantial, with costs associated with health care services, medications, and lost productivity. Hospitalizations due to HF are frequent, with 83% of patients hospitalized at least once, and 43% of patients are hospitalized 4 or more times after a diagnosis of HF.¹³ Approximately half of those with HF have a reduced ejection fraction; it is in this population that the evidence base regarding treatment is more well established.¹³

Standards of Therapy

The current foundational pharmaceutical management of HFrEF encompasses triple therapy, including beta blockers, ACEIs or ARBs or neprilysin inhibitors, and MRAs (e.g., spironolactone, eplerenone).² These drug classes, individually and together, have shown improvement in clinical outcomes including worsening, re-hospitalization, and mortality in patients with HFrEF. More recently, new therapies have emerged to be taken either in addition to, or in lieu of, the triple-therapy regimen.² Specifically, sacubitril-valsartan (an angiotensin receptor-neprilysin inhibitor [ARNI]) has been recommended as a replacement for ACEI or ARB therapy. Current Canadian and international guidelines recommend switching from ACEI or ARB to sacubitril-valsartan in patients with symptomatic HF.² Drugs such as SGLT2 inhibitors, initially developed to treat diabetes, and the soluble guanylate cyclase stimulator vericiguat, have shown marked benefit in HFrEF on top of foundational therapies. Specifically, they have further reduced hospitalizations for HF as well as mortality. Other potential therapies for HFrEF, depending on the situation, have included diuretics, digoxin, temporary inotropic therapy, implantable cardioverter-defibrillator therapy, and cardiac resynchronization therapy.²

There is no clear evidence that pharmacologic therapy, diet, or other therapies reduce the risk of mortality in patients with HFpEF.² According to the experts consulted by CADTH, current treatment strategies in HFpEF are limited to supportive therapies focusing on symptom control rather than morbidity or mortality benefit including ARB, MRA, and ARNI.

Other non-pharmacological measures for both HFrEF and HFpEF include lifestyle recommendations such as fluid restriction, avoiding salt and alcohol, and regular exercise.² According to the clinical experts consulted by CADTH, the goal of HF therapy, primarily for both HFrEF and HFpEF, is to prevent HHF, delay death, and improve quality of life. However, there are differences in how these goals are achieved in patients with HFrEF and HFpEF. In general, patients with HFrEF and HFpEF benefit from lifestyle changes, cardiac rehabilitation attendance, coordinated management with a multidisciplinary team, and pharmacotherapy. The use of mechanical cardiac resynchronization therapy and implantable cardioverter-defibrillators is mostly appropriate for patients with HFrEF. In addition, the clinical experts indicated that diuretics, such as loop diuretics, are used in patients with chronic HF to reduce congestion and improve well-being. Organ transplantation is rarely used for chronic HF and is mainly reserved for young patients with NYHA functional class IV despite optimal therapy.

Drug

Empagliflozin is an SGLT2 inhibitor. By inhibiting SGLT2, empagliflozin reduces renal reabsorption of filtered glucose and lowers the renal threshold for glucose, and thereby increases urinary glucose excretion. Empagliflozin reduces sodium reabsorption and increases sodium delivery to the distal tubules, resulting in a reduction in intraglomerular pressure, and cardiac pre- and afterload, as well as an improvement in diastolic function and cardiac remodelling. Inhibition of glucose and sodium cotransport by empagliflozin is also associated with moderate diuresis and transilient natriuresis. A secondary effect of empagliflozin is an increase in hematocrit.

Empagliflozin is approved by Health Canada for use in adults as an adjunct to SOC therapy for the treatment of chronic HF.⁶ The reimbursement criteria requested by the sponsor are narrower: for the treatment of adults with HF (NYHA functional class II, III, or IV) as an adjunct to SOC therapy.

Empagliflozin has been previously approved by Health Canada and reviewed by CADTH for use as an adjunct to diet and exercise to improve glycemic control in adult patients with type 2 diabetes mellitus for whom metformin is inappropriate due to contraindications or intolerance, and as an add-on combination when metformin used alone does not provide adequate glycemic control, in combination with:⁶

• metformin

• metformin and a sulfonylurea

• pioglitazone (alone or with metformin)

• linagliptin and metformin

• basal or prandial insulin (alone or with metformin).

Empagliflozin has also been approved by Health Canada for use as an adjunct to diet, exercise, and SOC therapy to reduce the incidence of CV death in patients with type 2 diabetes mellitus and established CV disease and has been previously reviewed by CADTH for this indication.⁶

Empagliflozin is available as a 10 mg or 25 mg tablet. The recommended dosage of empagliflozin for the treatment of chronic HF is 10 mg once daily.⁶

Key characteristics of commonly used medical treatments for HF are presented in Table 3.

Table 3: Key Characteristics of Pharmacotherapies for Heart Failure (by Drug Class)

ACEI = angiotensin-converting enzyme inhibitor; ARB = angiotensin receptor blocker; ARNI = angiotensin receptor-neprilysin inhibitor; AT1 = angiotensin type 1; bpm = beats per minute; CV = cardiovascular; HCN = hyperpolarization and cyclic nucleotide; HF = heart failure; HFrEF = heart failure with reduced ejection fraction; I_f = pacemaker current; NA = not applicable; NYHA = New York Heart Association; RAAS = renin-angiotensin-aldosterone system; SGLT2 = sodium-glucose cotransporter-2.

^aHealth Canada–approved indication.

Source: Product monographs for Jardiance (Boehringer Ingelheim Ltd.),⁶ Forxiga (Novartis),¹⁴ Entresto,¹⁵ and Lancora¹⁶ and the Canadian Pharmacists Association.^17,18

Stakeholder Perspectives

Patient Group Input

This section was prepared by CADTH staff based on the input provided by patient groups.

The patient and caregiver input received for this review was collected by the HeartLife Foundation, which is a national charity that, through its extensive network, engages patients and their caregivers to provide education, support, and access to treatments and research. Information for this review was gathered through in-person interviews with 3 patients and 1 caregiver, an online survey of 12 respondents held in April 2022, a closed virtual support group of 11 respondents, and literature searches from peer-reviewed publications.

Heart failure (HF) is a condition that requires daily monitoring, adherence, and vigilance on the part of the patient to control the delicate balance of symptoms. Respondents indicated shortness of breath, extreme fatigue, low blood pressure, dizziness, edema, and bloating as symptoms of HF. Many patients also mentioned having palpitations and arrhythmia because of the underlying cause of their HF. In their input, patients acknowledged that HF has no cure and, if left untreated, will become progressively worse over time. Patients indicated that the current standard “triple therapy” for HF, including ACEIs or ARBs, beta blockers, and MRAs, had shown effectiveness in managing their conditions with respect to reducing mortality and hospitalizations. However, there is a significant unmet need for new innovative therapies to improve outcomes in terms of quantity and quality of life, as many patients are intolerant to beta blockers and, in some cases, to ACEIs. Respondents expressed a desire to have greater access to proven therapies and improved functional capacity and quality of life, as they would like to spend time with loved ones, be able to work on a regular basis, pursue outdoor activities, and be able to travel.

A total of 16 respondents with experience using empagliflozin reported the drug was effective in terms of improving ejection fraction and energy level and reducing shortness of breath. According to the HeartLife Foundation survey (n = 12), about 33.3% of respondents felt better after taking empagliflozin, while 8.3% reported that they felt worse. Based on the survey results, about 33.3% of respondents described their side effects as manageable, whereas 25% said they were not manageable. Only 2 of the 16 patients who had experience with empagliflozin reported side effects, including fatigue and urinary tract infections. One patient reported multiple side effects, including constant yeast infections, back pain, sciatica, runny nose, joint pain, and occasional diarrhea. After 6 months of empagliflozin treatment, the same patient experienced additional side effects of volume depletion, hypotension, urgent urination, lower back pain, and headaches.

Clinician Input

Input From the Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise in the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, and providing guidance on the potential place in therapy). The following input was provided by 2 clinical specialists with expertise in the diagnosis and management of chronic HF.

Unmet Needs

According to the clinical experts consulted by CADTH, many patients with HFrEF are not being assessed by a specialist in Canada, and assistance from other specialists and primary care providers will be required, given the growing number of patients. The clinical experts further noted that the use of goal-directed guideline-recommended drug therapy and medical devices in patients with HFrEF remains suboptimal. The clinical experts consulted highlighted that current treatment strategies in HFpEF are limited to supportive therapies focusing on symptom control rather than morbidity or mortality benefit, including ARBs, MRAs, and ARNIs, while data on SGLT2 inhibitors show clear benefit in this population.

Place in Therapy

The clinical experts indicated that empagliflozin can be used as an alternative to dapagliflozin in combination with other foundational guideline-recommended pharmacotherapy, including beta blockers, ACEIs and ARBs, MRAs, and ARNIs in patients with HFrEF, and it is likely to be a first-line therapy for patients with HFpEF, given the ease of use, strength of evidence, safety profile, and familiarity with the use of empagliflozin in patients with diabetes. In addition, the clinical experts believe that empagliflozin is likely to be better tolerated than other classes of medications.

Patient Population

The clinical reviewers indicated there is no evidence of benefit for empagliflozin in patients with HF with low NT-proBNP levels, as the existing trials were limited to including patients with elevated NT-proBNP levels. The clinical experts also highlighted that NT-proBNP testing is not widely available in Canada, as some jurisdictions have limited access to it; thus, this patient selection criterion would be difficult to implement in clinical practice. The clinical experts consulted by CADTH highlighted that the benefit of empagliflozin in patients with NYHA classes I and IV is unclear due to limited clinical data and high mortality rate in patients with NYHA class IV.

Assessing Response to Treatment

The clinical experts indicated that the response to therapy in clinical practice is assessed based on the frequency of hospitalizations for HF, which in turn may lead to a reduction in mortality, improved quality of life, and a slower decline in kidney function.

Discontinuing Treatment

The clinical experts identified the following factors to consider when deciding to discontinue treatment with empagliflozin:

• the development of severe kidney dysfunction (eGFR < 15 mL/min/1.73 m²), and euglycemic diabetic ketoacidosis in patients with diabetes as important AEs

• NYHA functional class IV.

Prescribing Conditions

The clinical experts indicated that empagliflozin is already widely used by primary care providers and endocrinologists for the management of diabetes, by nephrologists to reduce decline in kidney function, and by cardiologists.

Clinician Group Input

No clinician group input was received for this review.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may impact their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 4.

Table 4: Summary of Drug Plan Input and Clinical Expert Response

CAF = Canadian Armed Forces; CDEC = CADTH Canadian Drug Expert Committee; CKD = chronic kidney disease; CSD = Canadian Space Division; CV = cardiovascular; HF = heart failure; HFpEF = heart failure with preserved ejection fraction; HFrEF = Heart failure with reduced ejection fraction; HHF = hospitalization for heart failure; ITC = indirect treatment comparison; LVEF = left ventricular ejection fraction; NIHB = Non-Insured Health Benefits; NT = Northwest Territories; NYHA = New York Heart Association; ON = Ontario; SGLT2 = sodium-glucose cotransporter-2; T2DM = type 2 diabetes mellitus; YK = Yukon.

Clinical Evidence

The clinical evidence included in the review of empagliflozin is presented in 3 sections. The first section, the systematic review, includes pivotal studies provided in the sponsor’s submission to CADTH and Health Canada as well as those studies that were selected according to an a priori protocol. The second section includes indirect evidence from the sponsor and indirect evidence selected from the literature that met the selection criteria specified in the review. The third section includes sponsor-submitted long-term extension studies and additional relevant studies that were considered to address important gaps in the evidence included in the systematic review.

Systematic Review (Pivotal and Protocol-Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of empagliflozin 10 mg as an adjunct to SOC therapy for the treatment of chronic HF in adults.

Methods

Studies selected for inclusion in the systematic review will include pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those meeting the selection criteria presented in Table 5. Outcomes included in the CADTH review protocol reflect outcomes considered to be important to patients, clinicians, and drug plans.

Table 5: Inclusion Criteria for the Systematic Review

ACEI = angiotensin-converting enzyme inhibitor; AE = adverse event; ARB = angiotensin receptor blocker; HF = heart failure; HRQoL = health-related quality of life; NYHA = New York Heart Association; RCT = randomized controlled trial; SAE = serious adverse event; WDAE = withdrawal due to adverse event.

^aThese outcomes were identified as being of particular importance to patients in the input received by CADTH from patient groups.

The literature search was performed by an information specialist using a peer-reviewed search strategy.

Two CADTH clinical reviewers independently selected studies for inclusion in the review based on titles and abstracts, according to the predetermined protocol. Full-text articles of all citations considered potentially relevant by at least 1 reviewer were acquired. Reviewers independently made the final selection of studies to be included in the review, and differences were resolved through discussion.

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946—) through Ovid and Embase (1974—) through Ovid. All Ovid searches were run simultaneously as a multi-file search. Duplicates were removed using Ovid deduplication for multi-file searches, followed by manual deduplication in Endnote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were Jardiance (empagliflozin) and HF. Clinical trials registries were searched: the US National Institutes of Health’s clinicaltrials.gov, WHO’s International Clinical Trials Registry Platform (ICTRP) search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register.

CADTH-developed search filters were applied to limit retrieval to randomized controlled trials or controlled clinical trials. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. Refer to Appendix 1 for the detailed search strategies.

The initial search was completed on May 4, 2022. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee (CDEC) on August 24, 2022.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool for Searching Health-Related Grey Literature checklist. Included in this search were the websites of regulatory agencies (FDA and European Medicines Agency). Google was used to search for additional internet-based materials. Refer to Appendix 1 for more information on the grey literature search strategy. These searches were supplemented by reviewing bibliographies of key papers and through contacts with appropriate experts. In addition, the manufacturer of the drug was contacted for information regarding unpublished studies.

Findings From the Literature

A total of 7 reports of 2 unique studies^19-25 were identified from the literature for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 6. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Table 6: Details of Included Studies

ACEI = angiotensin-converting enzyme inhibitor; AE = adverse event; AF = atrial fibrillation; ARB = angiotensin receptor blocker; ARNI = angiotensin receptor-neprilysin inhibitor; BMI = body mass index; CKD-EPIcr = Chronic Kidney Disease Epidemiology Collaboration creatinine; CRT = cardiac resynchronization therapy; CV = cardiovascular; DB = double blind; DBP = diastolic blood pressure; DM = diabetes mellitus; ECG = electrocardiogram; eGFR = estimated glomerular filtration rate; HF = heart failure; HHF = hospitalization for heart failure; HFpEF = heart failure with reduced preserved fraction; ICD = implantable cardioverter-defibrillator; KCCQ = Kansas City Cardiomyopathy Questionnaire; LVEF = left ventricular ejection fraction; MI = myocardial infarction; MRA = mineralocorticoid receptor antagonist; NT-proBNP = N-terminal pro-brain natriuretic peptide; NYHA = New York Heart Association; SAE = serious adverse event; SBP = systolic blood pressure; SGLT1 = sodium-glucose cotransporter-1; SGLT2 = sodium-glucose cotransporter-2; TEAE = treatment-emergent adverse event; UACR = urine albumin creatinine ratio.

Note: LVEF was obtained through electrocardiography, radionuclide ventriculography, invasive angiography, MRI, or CT. No prior measurement of LVEF < 40% was performed under stable conditions in HFpEF patients.

^aAn independent group of medical experts performed a central, blinded adjudication of the outcomes.

^bChronic dialysis was defined as dialysis with a frequency of twice per week or more for at least 90 days.

^cSustained was determined by 2 or more consecutive post-baseline central laboratory measurements separated by at least 30 days (the first to last of the consecutive eGFR values).

^dDiabetes was defined as hemoglobin A1C ≥ 6.5% or as diagnosed by the investigator.

^ePre-DM was defined as no history of DM and no hemoglobin A1C ≥ 6.5% before treatment, and a pre-treatment hemoglobin A1C value of ≥ 5.7% and < 6.5%.

^fKCCQ clinical summary score measures HF symptoms (frequency and burden) and physical limitations.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Description of Studies

Two sponsor-conducted trials — EMPEROR-Reduced⁷ and EMPEROR-Preserved⁸ — which met the CADTH review protocol criteria were included in this systematic review.

EMPEROR-Reduced

The EMPEROR-Reduced trial was a phase III, randomized, double-blind, multinational, parallel-group trial that aimed to assess the superiority of empagliflozin 10 mg once daily compared with matched placebo as an adjunct to SOC treatment in patients with chronic HFrEF (LVEF ≤ 40%). A total of 3,730 patients were enrolled across 520 sites from 20 countries in North America (including 35 sites in Canada [136 patients]), Europe, Asia Pacific, South America, South Africa, and other. After a screening period of 4 to 28 days, patients were randomized at visit 2 in a 1:1 ratio to receive either empagliflozin at a dose of 10 mg once daily (N = 1,863) or matching placebo (N = 1,867) in a double-blind manner.

EMPEROR-Preserved

The EMPEROR-Preserved trial was a phase III, randomized, double-blind, multinational, parallel-group trial that aimed to assess the superiority of empagliflozin 10 mg once daily compared with matched placebo as an adjunct to SOC treatment in patients with chronic HFpEF (LVEF > 40%). A total of 5,988 patients were enrolled across 622 sites from 23 countries in North America (including 42 sites in Canada [199 patients]), Europe, Asia Pacific, South America, South Africa, and other. After a screening period of 4 to 28 days, patients were randomized at visit 2 in a 1:1 ratio to receive either empagliflozin at a dose of 10 mg once daily (N = 2,997) or matching placebo (N = 2,991) in a double-blind manner.

In both EMPEROR-Reduced and EMPEROR-Preserved trials, randomization in blocks was conducted centrally via interactive response technology. In EMPEROR-Reduced, randomization was stratified by geographical region (North America, Lain America, Europe, Asia, and other), history of diabetes (diabetes, pre-diabetes, and no diabetes), and eGFR (CKD-EPIcr equation) at screening (< 60 mL/min/1.73 m², and ≥ 60 mL/min/1.73 m²). In EMPEROR-Preserved, randomization was stratified by geographical region (North America, Lain America, Europe, Asia, and other), history of diabetes (diabetes, pre-diabetes, and no diabetes), eGFR (CKD-EPIcr equation) at screening (< 60 mL/min/1.73 m², and ≥ 60 mL/min/1.73 m²), and LVEF (< 50%, and ≥ 50%). Treatment allocation was determined by a computer-generated random sequence. Given that empagliflozin is a diabetes drug, there was a possibility that more patients with diabetes would be recruited in these trials. Therefore, interactive response technology was used to ensure similar proportions of patients with diabetes, pre-diabetes, and no diabetes at the regional level in both trials.

Both EMPEROR-Reduced and EMPEROR-Preserved trials were event-driven trials and were to stop once 841 adjudicated primary end point events (CV death or HHF) were reached. Thus, the duration of double-blind treatment was different for each patient. In both EMPEROR-Reduced and EMPEROR-Preserved trials, the number of confirmed primary end points was continuously monitored in a blinded manner. In both trials, onsite visits were scheduled at 4, 12, 32, and 52 weeks after randomization during the first year, and then every 24 weeks throughout the trial. During the onsite visits, safety and efficacy end points, treatment compliance, and concomitant treatment or intervention were assessed. In addition, follow-up phone calls were scheduled 10 to 12 weeks after each onsite visit, starting at visit 4 and continuing throughout the trial. End of treatment in both trials was defined as reaching the required number of primary end points (841 events), or when the patient permanently discontinued study medication. All patients were required to complete a follow-up visit within 30 days of the regular or premature termination of the treatment period. A schematic of the EMPEROR-Reduced and EMPEROR-Preserved trials is presented in Figure 2.

An interim analysis was performed by the independent data-monitoring committee after 544 and 494 primary end point events (about 60% of information) in the EMPEROR-Reduced and EMPEROR-Preserved trials, respectively, after which it was recommended to continue the trials as planned. The database lock was performed after the completion of the blinded treatment period in the EMPEROR-Reduced and EMPEROR-Preserved trials. In EMPEROR-Reduced, an interim database lock was executed on October 11, 2019, and the final database lock was executed on July 14, 2020. In EMPEROR-Preserved, an interim database lock was executed on January 27, 2020, and the final database lock was executed on June 1, 2021.

In both EMPEROR trials, the primary efficacy end point was the time to first event of adjudicated CV death or adjudicated HHF, and the key secondary end points were occurrence of adjudicated HHF (first and recurrent), and eGFR (CKD-EPIcr equation) slope of change from baseline. HRQoL was assessed using the KCCQ and EQ-5D instruments. Overall, baseline characteristics were well balanced between treatment groups in both pivotal trials.

Figure 2: Study Schema for the EMPEROR-Reduced and EMPEROR-Preserved Studies

EOT = end of treatment; FU = follow-up.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Populations

Inclusion and Exclusion Criteria

The key inclusion and exclusion criteria applied to the EMPEROR-Reduced and EMPEROR-Preserved trials are summarized in Table 6. Briefly, patients eligible for enrolment in the EMPEROR-Reduced trial were adults with chronic HF diagnosed for at least 3 months with NYHA functional class II to IV, reduced ejection fraction (LVEF ≤ 40%), and elevated NT-proBNP (i.e., > 2,500 pg/mL for patients without atrial fibrillation). Patients were also required to have received appropriate doses of HF therapy (i.e., ACEI, ARB, beta blocker, oral diuretic, MRA, ARNI, ivabradine), and appropriate use of medical devices. Patients eligible for enrolment in the EMPEROR-Preserved trial were adults with chronic HF diagnosed for at least 3 months with NYHA functional class II to IV, LVEF greater than 40%, elevated NT-proBNP (i.e., > 300 pg/mL without atrial fibrillation), and evidence of structural heart disease or an HHF within 12 months before the trial. Patients were excluded from both EMPEROR trials if they had a diagnosis of myocardial infarction, stroke, transient ischemic attack, acute decompensated HF, major CV surgery, or any major surgery within the last 90 days. Patients with a history of renal impairment or ketoacidosis, as well as patients with an implanted cardioverter-defibrillator or implanted cardiac resynchronization therapy within the last 3 months, those with current or prior use of an SGLT2 inhibitor, or use of combined sodium-glucose cotransporter-1 (SGLT1) and SGLT2 inhibitors within the last 12 weeks were also excluded.

Baseline Characteristics

EMPEROR-Reduced Study

A summary of baseline characteristics is presented in Table 7. Baseline characteristics were well balanced between the treatment arms. The mean age of all randomized patients in EMPEROR-Reduced was 66.8 years (SD = 11.0 years) and most patients were male (76.1%) and White (70.5%). More patients (62.1%) were aged 65 years and older compared with those under the age of 65 years (37.9%). The mean LVEF was 27.5% (SD = 6.0), and more patients in the placebo group had LVEF < 20% than in the empagliflozin group (9.9% versus 7.3%, respectively). Almost half of the patients were diagnosed with diabetes mellitus (49.8%) at baseline, about 38.6% had a history of atrial fibrillation or flutter, and most patients had NYHA functional class II (75.1%) at baseline. The median NT-proBNP was 1,910 (Q1 [25th percentile] to Q3 [75th percentile], 1,115 to 3,481), and the mean eGFR (CKD-EPIcr equation) was 62.0 mL/min/1.73 m² (SD = 21.6). The mean time of HF diagnosis to enrolment was 6.1 years (SD = 6.3 years), and about 30.8% of patients had a prior HHF. Patients received the following previous HF medications at the start of the EMPEROR-Reduced trial: ACE inhibitors or ARBs (69.7%), ARNI (19.5%), beta blockers (94.7%), MRAs (71.3%), ||| |||||||||| ||||||.

EMPEROR-Preserved Study

Baseline characteristics were well balanced between the treatment arms. The mean age of all randomized patients in the EMPEROR-Preserved study was 71.9 years (SD = 9.4 years), nearly half of the patients were male (55.3%), and most patients were White (75.9%). More patients were aged 70 years and older (64.1%) compared with those under the age of 70 years (35.9%). The mean LVEF was 54.3% (SD = 8.8%), with 33.1% of patients having an LVEF of less than 50%. Most patients had NYHA functional class II (81.5%) at baseline. Almost half of the patients were diagnosed with diabetes mellitus (49.1%) at baseline, and about 52.4% had a history of atrial fibrillation or flutter. The median NT-proBNP was 974 pg/mL (Q1 to Q3, 499 pg/mL to 1,731 pg/mL), while the mean eGFR (CKD-EPIcr equation) was 60.6 mL/min/1.73 m² (SD = 19.8). The mean time of HF diagnosis to enrolment was 4.4 years (SD = 5.1 years), and about 22.9% of patients had a prior HHF. Patients received the following previous HF medications at the start of the EMPEROR-Preserved trial: ACE inhibitors or ARBs (78.6%), ARNIs (2.2%), beta blockers (86.3%), MRAs (37.5%), ||| |||||||||| ||||||.

Table 7: Summary of Baseline Characteristics — EMPEROR-Reduced and EMPEROR-Preserved, RS

ACEI = angiotensin-converting enzyme inhibitor; ARB = angiotensin receptor blocker; ARNI = angiotensin receptor-neprilysin inhibitor; BMI = body mass index; CKD-EPIcr = Chronic Kidney Disease Epidemiology Collaboration creatinine; CRT = cardiac resynchronization therapy; CRT-D = cardiac resynchronization therapy defibrillator; CRT-P = cardiac resynchronization therapy pacemaker; DBP = diastolic blood pressure; eGFR = estimated glomerular filtration rate; HF = heart failure; ICD = implantable cardioverter-defibrillator; IRT = interactive response technology; LVEF = left ventricular ejection fraction; MRA = mineralocorticoid receptor antagonist; NA = not applicable; NT-proBNP = N-terminal prohormone of brain natriuretic peptide; NYHA = New York Heart Association; Q1 = first quartile (25th percentile); Q3 = third quartile (75th percentile); RS = randomized set; SBP = systolic blood pressure; SD = standard deviation; T1DM = type 1 diabetes mellitus; T2DM = type 2 diabetes mellitus.

Note: Patients with missing information are not shown.

^aIncluding patients with no investigator-reported medical history of diabetes and pre-treatment hemoglobin A1C ≥ 5.7% and < 6.5%, or patients stratified to the group of pre-diabetes via IRT and pre-treatment hemoglobin A1C < 6.5% (if available), or patients stratified to the group of no diabetes via IRT and pre-treatment hemoglobin A1C ≥ 5.7% and < 6.5%.

^bPatients without T1DM and with investigator-reported medical history of diabetes, or patients with previously undiagnosed diabetes (pre-treatment hemoglobin A1C ≥ 6.5%), or (in case the information noted previously was missing) patients stratified to the group of diabetes via IRT.

^cPatients with investigator-reported medical history of diabetes and the type was T1DM.

^dInvestigator-reported medical history or baseline electrocardiogram finding.

^eReported either on HF history and diagnosis or health care resource utilization form.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Interventions

In the EMPEROR-Reduced and EMPEROR-Preserved trials, during visit 2 followed by the screening period, all eligible patients were randomized in a 1:1 ratio to receive 1 of 2 interventions: empagliflozin at a dose of 10 mg or matching placebo in a double-blind and single-dummy manner. The drugs were administered orally once daily, with or without food. The empagliflozin and placebo tablets were identical in packaging and labelling. Results from the previous EMPA-REG-OUTCOME trial²⁶ showed that both doses of empagliflozin, 10 mg and 25 mg, are equally effective in reducing CV death, HHF, and the composite of CV death and HHF in patients with HF at baseline. Therefore, given the lower exposure with empagliflozin at 10 mg, empagliflozin at 10 mg once daily was selected in both EMPEROR trials. To ensure a dose interval of about 24 hours, the drug was to be taken in the morning at approximately the same time every day. In both EMPEROR trials, all patients, investigators, and staff involved in conducting and reviewing the trials remained blinded with regard to the randomized treatment assignment until after database lock.

All concomitant medications or other therapies were recorded consistently during the EMPEROR trials. Concomitant antidiabetic medications were adjusted according to the clinical indications of the patient's treating physician. The investigators constantly monitored for symptoms that could be indicative of hypoglycemia in patients without a diagnosis of diabetes. Empagliflozin should be used with caution in patients at a higher risk of ketoacidosis. Patients were assessed and treated for ketoacidosis immediately according to local clinical guidelines. In clinical situations known to predispose to ketoacidosis, the investigators should consider monitoring for ketoacidosis and temporarily discontinuing trial medication. Patients were receiving appropriate care as defined by their physician or practitioner for all CV conditions according to the prevailing guidelines, including Aspirin, statins, diuretics, ACE inhibitors, beta blockers, MRAs, and implantable devices. The use of any SGLT2 inhibitors or combined sodium-glucose cotransporter-1 (SGLT1) and SGLT2 inhibitors was prohibited during the treatment period, except for a 30-day follow-up period.

Outcomes

A list of efficacy outcomes identified in the CADTH review protocol that were assessed in the clinical trials included in this review is provided in Table 8. These outcomes are further summarized subsequently. A detailed discussion and critical appraisal of the outcome measures is provided in Appendix 4.

Table 8: Summary of Outcomes of Interest Identified in the CADTH Review Protocol — EMPEROR-Reduced and EMPEROR-Preserved

CKD-EPIcr = Chronic Kidney Disease Epidemiology Collaboration creatinine; CV = cardiovascular; eGFR = estimated glomerular filtration rate; HHF = hospitalization for heart failure; HRQoL = health-related quality of life; KCCQ = Kansas City Cardiomyopathy Questionnaire; NYHA = New York Heart Association.

^aAn independent group of medical experts performed a central, blinded adjudication of the outcomes.

^bChronic dialysis was defined as dialysis with a frequency of twice per week or more for at least 90 days.

^cSustained was determined by 2 or more consecutive post-baseline central laboratory measurements separated by at least 30 days (the first to last of the consecutive eGFR values). Reduction in eGFR (CKD-EPIcr equation) was defined as reduction in eGFR from baseline of ≥ 40%, eGFR < 15 mL/min/1.73 m² for patients with baseline eGFR ≥ 30 mL/min/1.73 m², or eGFR < 10 mL/min/1.73 m² for patients with baseline eGFR < 30 mL/min/1.73 m².

^dKCCQ total symptom score measures heart failure symptoms (frequency and burden) and physical limitations. KCCQ overall summary score measures the physical limitation, total symptom, social limitation, and health-related quality of life. KCCQ clinical summary score measures physical limitation, symptom frequency, and symptom severity.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Efficacy Outcomes

In both EMPEROR trials, an independent group of clinical experts performed a central adjudication of the following outcomes occurring after randomization in a consistent and blinded fashion:

• all fatal events

• HF hospitalization

• myocardial infarction

• stroke and transient ischemic attack

• ketoacidosis

• hepatic events.

The primary composite end point for both the EMPEROR-Reduced and EMPEROR-Preserved trials was the time to first event of adjudicated CV death or adjudicated HHF.

CV death included death due to:

• acute myocardial infarction, which refers to a death from any CV mechanism that occurs within 30 days after an MI

• sudden cardiac death, which refers to a death that occurred unexpectedly and not following an acute MI

• other causes:

  • HF, which refers to a death in association with clinically worsening symptoms and/or signs of HF regardless of HF etiology

  • stroke, which refers to a death after a stroke that is either a direct consequence or a complication of the stroke

  • CV procedures, which refers to a death caused by the immediate complication of a cardiac procedure

  • CV hemorrhage, which refers to a death related to a hemorrhage such as non-stroke intracranial hemorrhage, non-procedural or non-traumatic vascular rupture, or hemorrhage causing cardiac tamponade

  • other CV causes, which refers to a CV death not included in the previous categories, but with a specific, known cause (i.e., pulmonary embolism or peripheral arterial disease).

Non-CV death was defined as any death with a specific cause that is not thought to be CV in nature.

HHF was defined as an event that met all the following criteria:

• The adjudicated primary diagnosis is admission to hospital for HF.

• The patient’s length of stay in hospital extends for at least 12 hours.

• The patient exhibits documented new or worsening symptoms due to HF on presentation (i.e., dyspnea, fatigue).

• The patient has objective evidence of new or worsening HF, consisting of at least 2 physical examination findings or 1 physical examination finding and at least 1 laboratory criterion, including increased brain natriuretic peptide (BNP) and NT-proBNP levels, radiological evidence of pulmonary congestion, or non-invasive diagnostic evidence of clinically significant elevated left- or right-sided ventricular filling pressure.

• The patient receives initiation or intensification of treatment specifically for HF, including at least 1 of the following: augmentation in oral diuretic therapy, IV diuretic or vasoactive drug, or mechanical or surgical intervention.

• The documented changes in physical signs or laboratory tests, whenever available, were considered to be supportive.

The key secondary end points for both EMPEROR trials included:

• occurrence of adjudicated HHF (first and recurrent)

• eGFR (CKD-EPIcr equation) slope of change from baseline.

In addition to the clinical end points, 2 patient-valued outcomes, such as HRQoL and symptoms associated with HF, were measured in the EMPEROR-Reduced and EMPEROR-Preserved trials using the KCCQ and the EQ-5D instrument.

KCCQ Questionnaire

The KCCQ is a self-administered, 23-item, disease-specific HRQoL questionnaire that was originally developed in 2000 to measure the patient’s perception of their health status within a 2-week recall period.^27-29 The items of the KCCQ can be categorized into the following domains: physical limitation, symptoms (frequency, severity, and recent change over time), social limitation, self-efficacy, and HRQoL. All items are measured using a Likert scale with 5 to 7 response options. Responses are scored using ordinal values, beginning with 1 for the response that implies the lowest level of functioning. Domain scores are transformed to a 0 to 100 range by subtracting the lowest possible scale score, dividing by the range of the scale, and multiplying by 100. Various combinations of the KCCQ domains create 3 KCCQ summary scores including the KCCQ-TSS, the KCCQ-CSS, and the KCCQ overall summary score (KCCQ-OSS). The KCCS-TSS combines the symptom burden and symptom frequency domains and evaluates patient-reported swelling in feet, ankles, or legs, fatigue, shortness of breath, and disturbed sleep.³⁰ The KCCQ-CSS includes the physical limitation and total symptom domains, and the KCCQ-OSS combines the physical limitation, total symptom, social limitation, and HRQoL domains into a single score. Summary scores are then transformed to a 0 to 100 range, where larger scores represent a better outcome: 0 to 24: very poor to poor; 25 to 49: poor to fair; 50 to 74: fair to good; and 75 to 100: good to excellent.^27,29

The KCCQ questionnaire is a generally valid, reliable, and responsive instrument for CV diseases, including HF.^27,30-36 Convergent validity was demonstrated through moderate to strong correlations between the KCCQ-OSS and the KCCQ domain scores with a variety of external indicators of clinical status (r = 0.32 to 0.64).^31-33,35 Internal consistency reliability was demonstrated in a number of studies, where the KCCQ summary and domain scores had Cronbach alpha values greater than 0.7.^27,31,32 Test-retest reliability has been demonstrated (intraclass correlation coefficient [ICC] > 0.7) for the KCCQ symptom, social, and limitation domains.^27,32 High responsiveness of the KCCQ domains, the KCCQ clinical summary and overall summary scores was found when the external indicators of clinical status were NYHA class, the Short Form (36) Health Survey, and the 6MWD.²⁷ The estimated minimal important differences (MIDs) were evaluated using 2 anchor-based methods in patients with HF; they were approximately 5 points for the KCCQ overall summary and total symptom scores, and 6 points for the KCCQ clinical summary scores.³⁷

EQ-5D Instrument

The EQ-5D-5L is a generic self-reported HRQoL outcome measure that may be applied to a variety of health conditions and treatments. The EQ-5D-5L was developed by the EuroQol Group as an improvement to the 3-level EQ-5D (EQ-5D-3L) to measure small and medium health changes and reduce ceiling effects.^38,39 The instrument comprises 5 dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension is rated on 5 levels: level 1 “no problems,” level 2 “slight problems,” level 3 “moderate problems,” level 4 “severe problems,” and level 5 “extreme problems” or “unable to perform.”^38,39 A total of 3,125 unique health states are possible, with 55555 representing the worst health state and 11111 representing the best state. The corresponding scoring of EQ-5D-5L health states is based on a scoring algorithm that is derived from preference data obtained from interviews using choice-based techniques (e.g., time trade-off) and discrete choice experiment tasks.^38,39 The lowest and highest score vary depending on the scoring algorithm used. Scores less than 0 represent health states that are valued by society as being worse than dead, while scores of 0 and 1.00 are assigned to the health states “dead” and “perfect health,” respectively. As an example, a Canadian scoring algorithm results in a score of −0.148 for health state 55555 (worst health state) and a score of 0.949 for health state 11111 (best health state).^38,39 Another component of the EQ-5D-5L is the EQ-5D Visual Analogue Scale (EQ VAS), which asks respondents to rate their health on a visual scale from 0 (worst health imaginable) to 100 (best health imaginable).^38,39 The literature search completed by CADTH did not find any evidence on the validity, reliability, responsiveness, and MID of the EQ-5D-5L questionnaire in patients with HF.

Harms

The primary safety outcomes assessed in the EMPEROR-Reduced and EMPEROR-Preserved trials were:

• AEs

• TEAEs

• SAEs

• AEs leading to the discontinuation of the investigational drug

• AEs of special interests, including hepatic injury, decreased renal function, ketoacidosis, events leading to lower-limb amputation, hypoglycemia

• worsening of underlying condition

• changes from baseline for clinical laboratory measure and vital signs.

An independent group of medical experts performed a central, blinded adjudication of ketoacidosis and certain hepatic events.

Measures Taken During the COVID-19 Pandemic

In both pivotal trials, investigators and coordinators were informed that study procedures should be followed in accordance with the protocol, whenever possible and appropriate. A remote visit (phone contact) was performed when the planned visit to the clinic was not feasible.

Statistical Analysis

The statistical analysis of efficacy end points conducted in both EMPEROR trials is summarized in Table 10.

Sample Size Determination

In the EMPEROR-REDUCED trial, at least 841 primary end point events in the intention-to-treat population were required to achieve 90% power at a 2-sided significance level of 0.05. A true HR of 0.8 between empagliflozin and placebo was chosen based on the HF outcomes in the EMPA-REG-OUTCOME trial.⁴⁰ The sponsor estimated that at least 2,825 patients needed to be enrolled to achieve the required number of events, assuming a yearly event rate in the placebo group of 15%,^41-44 an accrual period of 18 months, and an average follow-up period of 20 months. The yearly dropout rate was assumed to be below 1% and was not considered for the sample size determination. There was 1 planned interim analysis for the primary outcome, which was conducted when 544 (approximately 60%) of the events had occurred. The EMPEROR-Reduced trial used the Hwang, Shin, and De Cani alpha-spending function to control for type I error, yielding a 2-sided significance level of 0.0496 for the final primary end point analysis. Following the interim analysis, the study was recommended to continue as planned. During the trial, based on the actual accrual over time of the primary outcome events, the number of randomized patients was adjusted to 3,600.

In the EMPEROR-Preserved trial, at least 841 primary end point events in the intention-to-treat population were required to achieve a 90% power at a 2-sided significance level of 0.05. A true HR of 0.8 between empagliflozin and placebo was chosen based on the HF outcomes in the EMPA-REG-OUTCOME study.⁴⁰ Based on the previously mentioned assumption, the sponsor estimated that at least 4,126 patients were needed, assuming a yearly event rate in the placebo group of 10%,^45,46 an accrual period of 18 months, and an average follow-up period of 20 months. The yearly dropout rate was assumed to be below 1% and was not considered for the sample size determination. There was 1 planned interim analysis for the primary outcome, which was conducted when 494 (approximately 60%) of the events had occurred. The EMPEROR-Preserved trial used the Hwang, Shin, and De Cani alpha-spending function to control the type I error, yielding a 2-sided significance level of 0.0497 for the final primary end point analysis. Following the interim analysis, the study was recommended to continue as planned. During the trial, based on the actual accrual over time of the primary outcome events, the number of randomized patients was adjusted to 5,750.

Primary Efficacy Analysis

In both EMPEROR trials, the composite end point of time to first event of adjudicated CV death or adjudicated HHF was analyzed using a Cox proportional hazards model adjusted for age, geographical region (Asia, Europe, Latin America, North America, and other), diabetes status (diabetes, pre-diabetes, and no diabetes), sex, LVEF, and eGFR (CKD-EPIcr equation). The proportional hazards assumption was evaluated by plotting Schoenfeld residuals for each covariate and treatment against time and log (time). Each component of the composite primary end point was also summarized separately, but was not formally tested for significance. The primary end point was displayed using a cumulative incidence function, considering non-CV death as competing risk and expressed using HR and 95% CI. Incidence rate was calculated as the number of patients with events per 100 person-years at risk. The time to event was derived from the date of randomization. This analysis was based on the RS (described subsequently), using all data available until completion of the planned treatment phase, including the data after the end of treatment for patients who did not complete the treatment phase as planned. A patient with at least 1 event (CV death or HHF) was considered to have an event and the date of the first event was used for the composite end point analysis. Only the adjudicated and confirmed events were included in the primary analysis. Patients without a specific end point event were censored at the last date the patient was known to be free of the event at the end of the planned treatment period, whichever was earliest.

The following sensitivity analyses (exploratory) were conducted for the primary outcome:

• a Cox proportional hazards model including only treatment, not adjusted for any other factors

• a Cox proportional hazards model with the same covariates as per the primary analysis performed on the treated set (TS), with observation period up to 30 days after treatment discontinuation

• a Cox proportional hazards model with multiple imputations for patients without primary end point events and lost to follow-up before the trial completion

• analysis of investigator-defined events

• a competing risk model by Fine-Gray, including the same set of covariates as in the primary analysis

• a Cox proportional hazards model of all events in the trial (including the follow-up period)

• a Cox proportional hazards model similar to the primary analysis but excluding those missing a physical sign or laboratory test, or both

• a Cox proportional hazards model similar to the primary analysis but including the following additional prognostic covariates: Log(NT-proBNP) and HHF in the last 12 months

• a Cox proportional hazards model similar to the primary analysis but including only events up to cut-off dates before a COVID-19 outbreak.

Sensitivity analyses were performed based on the TS (described subsequently).

Of the subgroups listed in the CADTH review protocol, the following subgroups were pre-specified in the EMPEROR-Reduced trial:

• HF physiology (LVEF ≤ 30% and NT-proBNP < median, LVEF ≤ 30%, NT-proBNP ≥ median, or LVEF > 30%)

• NYHA class (classes II or III/IV)

• history of diabetes (yes or no)

• renal function (eGFR [CKD-EPIcr equation]) < 60 or ≥ 60)

• prior use of ARNI (yes or no)

• prior use of MRA (yes or no).

The following subgroups were pre-specified in the EMPEROR-Preserved trial:

• LVEF (< 50%, 50% to 59%, or ≥ 60%)

• NYHA class (classes II or III/IV)

• History of diabetes (yes or no)

• Renal function (eGFR [CKD-EPIcr equation]) < 60 or ≥ 60)

• History of atrial fibrillation or atrial flutter (yes or no)

• Prior use of ACE inhibitor, ARB, or ARNI (yes or no)

• Prior use of MRA (yes or no).

In EMPEROR-Reduced, the randomization of patients was stratified by geographical region, history of diabetes, and eGFR (CKD-EPIcr equation) at screening. In EMPEROR-Preserved, the randomization of patients was stratified by geographical region, history of diabetes, LVEF, and eGFR (CKD-EPIcr equation) at screening. The subgroup analyses were performed using a Cox proportional hazards model as per the primary end point analysis. There were no adjustments made for multiplicity; thus, all subgroup analyses are exploratory in nature. The between-group treatment effect with a nominal 95% CI for these end points was estimated within each category. Forest plots were created, including interaction P values for treatment by subgroup interactions.

Key Secondary Efficacy Analysis

EMPEROR-Reduced and EMPEROR-Preserved Trials

In both EMPEROR trials, the key secondary end points, including occurrence of adjudicated HHF (first or recurrent) and the eGFR (CKD-EPIcr equation) slope of change from baseline, were tested in order using a hierarchical testing procedure to control the overall type I error rate for multiple end points (Table 9). If the primary end point was statistically significant, the overall type I error was preserved for the test in the next step. This testing procedure continued through each of the key secondary end points until the end point failed to reach statistical significance, after which subsequent key secondary end points were considered exploratory.

In both EMPEROR trials, the occurrence of HHF (first and recurrent) was analyzed using a joint frailty model that accounts for the dependence between recurrent HHF and CV death, with factors of treatment (empagliflozin, placebo), geographical region, baseline status of diabetes, age, sex, LVEF, and eGFR (CKD-EPIcr equation) at baseline as covariates. All data from all randomized patients until the end of the planned treatment period were used. The number of HHF events per patient was summarized descriptively. The number of HHF events was analyzed descriptively. Negative binomial models were additionally fitted for recurrent HHF events. The mean cumulative incidence was displayed for adjudicated first and recurrent HHF. Subgroups analyses were carried out; however, there were no adjustments made for multiplicity.

The following sensitivity analyses of adjudicated HHF (first and recurrent) were conducted:

• a Cox model, including only treatment as covariate, not adjusting for any other factors

• a Cox regression with the same covariates as the primary analysis performed on the TS, including only events up to 30 days after treatment discontinuation

• a Cox regression with multiple imputations for patients without primary end point events and lost to follow-up before trial completion

• analysis of investigator-defined events

• a Fine-Gray competing risk model (considering non-CV death as a competing risk)

• a Cox regression of all events in the trial (including the follow-up period)

• a Cox regression similar to the primary analysis but excluding events without documented physical signs or symptoms

• a Cox regression similar to the primary analysis but including the following additional prognostic covariates: Log(NT-proBNP) and HHF in the last 12 months

• a Cox regression similar to the primary analysis but including only events that were thought to occur before the COVID-19 pandemic.

In both EMPEROR trials, the slope in change from baseline of eGFR (CKD-EPIcr equation) was analyzed using a random coefficient model allowing for random intercept and random slope per patient. The model was adjusted for sex, geographical region, status of diabetes as fixed effects, and eGFR (CKD-EPIcr equation) at baseline, LVEF, age, time, and interaction of treatment by time and interaction of eGFR at baseline by time as linear covariates. Only data from treated patients (based on TS, i.e., measurement up to 1 day after the last intake of study medication) were used. Subgroup analyses were carried out; however, there were no adjustments made for multiplicity.

Table 9: Summary of Hierarchical Testing — EMPEROR-Reduced and EMPEROR-Preserved

CKD-EPIcr = Chronic Kidney Disease Epidemiology Collaboration creatinine; eGFR = estimated glomerular filtration rate; CV = cardiovascular; eGFR = estimated glomerular filtration rate; HHF = hospitalization for heart failure.

Note: If an end point was not successful, all subsequent end points would be evaluated in an exploratory manner.

^aAn independent group of medical experts performed a central, blinded adjudication of the outcomes.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Other Secondary Efficacy Analysis

In both pivotal trials, end points listed as other secondary end points were tested in a non-hierarchical fashion without adjustments for multiplicity. The following time-to-event end points were analyzed using a Cox proportional hazards model similar to the primary outcome analysis: time to the first event in the composite renal end point (chronic dialysis, renal transplant, or sustained reduction in eGFR), time to first adjudicated HHF, time to adjudicated CV death, and time to all-cause mortality. The models were adjusted for sex, geographical region, baseline status of diabetes, eGFR (CKD-EPIcr equation) at baseline, LVEF, and age. If the end point did not include any cause of death, a cumulative incidence function curve was displayed with all-cause of death as the competing risk. A Kaplan-Meier survival curve was displayed for all-cause mortality.

In both pivotal trials, change from baseline KCCQ clinical summary, total symptom, and overall summary scores, and KCCQ domain scores at week 52 were analyzed using a mixed-model for repeated measures analysis, including age, eGFR (CKD-EPIcr equation) as linear covariates, and baseline score by visit, visit by treatment, sex, geographical region, LVEF, and baseline diabetes status as fixed effects. The analysis was carried out for both the RS, and the TS, which included all observed cases with on-treatment data up to week 52. Responder analyses using logistic regression were conducted to evaluate the improvement and deterioration of the KCCQ clinical summary, total symptom, and overall summary scores at week 52.

Further Efficacy End Points

In both pivotal trials, further end points were tested in a non-hierarchical fashion without adjustments for multiplicity. The following end points were analyzed using a mixed-model for repeated measures: change from baseline in KCCQ overall summary score and clinical summary score, and total symptom score at week 52. Furthermore, responders for clinically meaningful improvement (an increase in score of at least 5 points at week 52 from baseline) or deterioration (a reduction of at least 5 points) were analyzed using logistic regression. The following further efficacy end points were analyzed using a Cox proportional hazards model:

• composite of time to first event of all-cause mortality and all-cause hospitalization

• time from first to second adjudicated HHF

• time to first all-cause hospitalization

• time to all-cause hospitalization or all-cause mortality

• time to first investigator-defined CV hospitalization.

Change in NYHA class at week 52 and EQ-5D-5L scores were analyzed using descriptive statistics.

Missing Data

In both pivotal trials, the analyses of primary and key secondary end points were performed based on all available data in the RS (description follows). No data were imputed for time-to-event or safety end points. All efforts were made to follow all patients until the end of the trial for their survival status and for any other end points, including the primary and key secondary end points. With regard to KCCQ scores, for patients who died, the worst score was assigned to all scores scheduled to be assessed after the date of death.

Harms

In both pivotal trials, all safety end points were reported using descriptive statistics and were carried out on the safety population (TS). Separate summaries were provided for most notable safety end points: decreased renal function, ketoacidosis, events leading to lower-limb amputation, hypoglycemic events, urinary tract and genital infections, hypotension, and bone fracture events. The incidence of these end points was analyzed by treatment as well as by subgroups. Safety analyses were based on the TS and included patients who had received at least 1 dose of the trial medication.

Table 10: Statistical Analysis of Efficacy End Points — EMPEROR-Reduced and EMPEROR-Preserved

CV = cardiovascular; CKD-EPIcr = Chronic Kidney Disease Epidemiology Collaboration creatinine; eGFR = estimated glomerular filtration rate; HF = heart failure; HHF = hospitalization for heart failure; KCCQ = Kansas City Cardiomyopathy Questionnaire; LVEF = left ventricular ejection fraction; NT-proBNP = N-terminal prohormone of brain natriuretic peptide; NYHA = New York Heart Association; SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2; TS = treated set.

^aAn independent group of medical experts performed a central, blinded adjudication of the outcomes.

^bChronic dialysis was defined as dialysis with a frequency of twice per week or more for at least 90 days.

^cSustained was determined by 2 or more consecutive post-baseline central laboratory measurements separated by at least 30 days (the first to last of the consecutive eGFR values).

^dReduction in eGFR (CKD-EPIcr equation) was defined as reduction in eGFR from baseline of ≥ 40%, eGFR < 15 mL/min/1.73 m² for patients with baseline eGFR ≥ 30 mL/min/1.73 m², or eGFR < 10 mL/min/1.73 m² for patients with baseline eGFR < 30 mL/min/1.73 m².

^eKCCQ clinical summary score measures HF symptoms (frequency and burden) and physical limitations.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Analysis Populations

All patient populations were defined and documented before database lock. The following analysis populations were used in the statistical analysis: RS, TS, and TS with follow-up (TSFU).

The randomized set, also known as the full analysis set (N = 3,730 and N = 5,988 in EMPEROR-Reduced and EMPEROR-Preserved, respectively), consisted of all randomized patients. Patients were analyzed according to their randomized group. Unless otherwise specified, all efficacy end points were summarized and analyzed using the randomized set.

The TS, also known as the safety set (N = 3,726 and N = 5,985 in EMPEROR-Reduced and EMPEROR-Preserved, respectively) consisted of all randomized patients who received at least 1 dose of the study drug. The summary for the safety analysis set was based on patients “as treated.”

The TS with follow-up, also known as the per-protocol set (N = 1,062 and N = 3,269 in EMPEROR-Reduced and EMPEROR-Preserved, respectively) consisted of all randomized patients who received at least 1 dose of the study drug and who performed the follow-up visit.

Results

Patient Disposition

Details of patient disposition in both pivotal trials are summarized in Table 11.

EMPEROR-Reduced Trial

In EMPEROR-Reduced, 7,220 individuals were screened, of whom 3,490 (48.3%) did not pass screening. The main reasons were not meeting eligibility criteria (95.0%), most commonly because the patients’ NT-proBNP levels were below the pre-specified thresholds at screening, and consent withdrawal (2.3%). In total, 3,730 patients were randomized in the treatment period. Overall, 3,688 patients (98.9%) completed the study or died, with similar completion rates across treatment groups. Vital status was known for 1,857 patients (99.5%) in the empagliflozin group and 1,852 patients (99.4%) in the placebo group. Of the 3,726 patients treated with the study medication, 25.9% of patients in the empagliflozin group and 27.4% of patients in the placebo group discontinued treatment. The most frequently reported reasons for discontinuation were AEs (18.3%), including 8.7% with non-fatal events and 9.5% with fatal events, and patient choice (5.8%). ||| || |||||||| ||||||||| ||||| ||| || ||| ||| || ||||||||| |||||| ||| ||||| ||| || ||||||||| |||||| |||| ||||||||| || ||||| ||||||| ||||||| |||||||||| |||||| |||||||||||

EMPEROR-Preserved Trial

In EMPEROR-Preserved, 11,583 individuals were screened, of whom 5,595 (48.3%) did not pass screening. The main reasons for this were not meeting eligibility criteria (95.4%), most commonly because the patients’ NT-proBNP levels were below the pre-specified thresholds at screening, and consent withdrawal (3.2%). In total, 5,988 patients were randomized in the treatment period. Overall, 5,816 patients (97.1%) completed the study or died, with similar completion rates across treatment groups. Vital status was known for 2,980 patients (99.4%) in the empagliflozin group and 2,972 patients (99.4%) in the placebo group. Of the 5,985 patients treated with the study medication, 31.5% of patients in both the empagliflozin and placebo groups discontinued treatment. The most frequently reported reasons for discontinuation were AEs (18.8%), including 10.6% of non-fatal events and 8.2% of fatal events, and patient choice (9.8%). ||| || |||||||| ||||||||| ||||| ||||| ||| || ||||| ||| ||||| |||||||| || ||||||||| ||| |||||||| || ||||| |||||||| ||||||| ||| || ||||| ||| ||||| ||||||||| || ||||| ||||||| ||||||| |||||||||| |||||| |||||||||| ||| |||| ||| || ||||| ||| ||||| ||||||||| ||||||| ||| |||||||||||||||| ||||||| | ||||| || |||| || |||||||| ||| |||||||||||| || ||||| |||||||||| ||| |||| |||| | |||| ||| || |||||||| ||| |||| ||| ||||||||| ||||||||| ||||||||||||||| ||| || |||||||| ||||||||||

Table 11: Patient Disposition: EMPEROR-Reduced and EMPEROR-Preserved

AE = adverse event; HHF = hospitalization for heart failure; HF = heart failure; ITT = intention to treat; NA = not applicable; PP = per protocol; RS = randomized set; TS = treated set; TS-FU = treated set with follow-up.

^aDefined as all patients with a primary event (cardiovascular death or HHF) or follow-up for the primary end point until study end/death.

^bPatients who discontinued all trial activities but did not withdraw consent to vital status collection at treatment termination.

^cOther patients with incomplete follow-up for the primary end point.

^dIncluding patients from |||| ||| ||||||| and closed sites (who did not complete the trial or died and did not withdraw consent).

^e||| || ||||||||| ||||| ||| |||||| || ||||||||| |||||| |||| ||||||||| || ||||||||| ||||||| ||||||| |||||||||| |||||| ||||||||||| ||||||| || | ||||||| |||||| || |||||||| || ||| ||||||.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Exposure to Study Treatments

| ||||||| || ||| ||||||||||| ||||||| || |||| ||||||| |||||| || |||||||||| || ||||||||||.

In EMPEROR-Reduced, the median observation time up to the end of the planned treatment period was about 1.31 years for the overall study population. ||| |||||||| || |||||||| |||||| |||| |||| |||||||| ||| || ||||| | ||||| ||| |||||| ||||||||||||| |||| ||| ||||||| |||||| ||| ||||||||| ||||||| In EMPEROR-Preserved, the median observation time up to the end of the planned treatment period was about 2.15 years for the overall study population. |||| |||||||| ||||| |||| |||||||| ||| || ||||| | ||||| ||| |||||| ||||||||||||| |||| ||| ||||||| |||||| ||| ||||||||| ||||||||||| || ||||| |||||||||| || |||| ||||||| |||||| || |||||||||| || |||||||| |||

|| |||||||||||||||| ||| |||||| |||||||| || ||||||||| ||| ||||| |||| ||||| |||| ||| |||| || ||||| || ||| ||||||| ||||| ||||||||||| ||| |||||||| || |||||||| |||||| |||| |||| ||||||| ||| || ||||| | ||||| ||| |||||| |||||||| || |||||||| ||| |||||||| ||| ||||||| |||||| ||| ||||||||| ||||||| || |||||||||||||||||| ||| |||||| |||||||| || ||||||||| ||| ||||| |||| ||||| |||| ||| |||| || ||||| || ||| ||||||| ||||| ||||||||||| |||| |||||||| ||||| |||| ||||||| ||| || ||||| | ||||| ||| |||||| |||||||| || |||||||| ||| |||||||| ||| ||||||| |||||| ||| ||||||||| |||||||

Table 12: Redacted

Note: This table has been redacted as per sponsor’s request.

Table 13: Redacted

Note: This table has been redacted as per sponsor’s request.

||||||||| |||||||||| ||| |||||||| ||||| || ||| |||||| || ||||| |||| |||||||| || ||| ||| || ||| ||||||||| ||||||| |||| |||||||| |||||| ||| ||| ||| || |||||||| || ||||||||||||||| ||| |||||||||||||||||| ||||||||||||| |||| ||||||| ||| ||| |||| ||||||||| |||| ||| ||||| |||||||||| |||||| ||| |||||| |||| ||||||| |||||||||| ||||| ||||||||| ||||||

All patients received HF medications during the trials, including ACE inhibitors or ARBs (73.1% and 82.6% in EMPEROR-Reduced and EMPEROR-Preserved, respectively), ARNI (26.2% and 4.4% in EMPEROR-Reduced and EMPEROR-Preserved, respectively), beta blockers (96.4% and 89.3% in EMPEROR-Reduced and EMPEROR-Preserved, respectively), MRAs (77.0% and 46.0% in EMPEROR-Reduced and EMPEROR-Preserved, respectively), and ivabradine (8.3% and 1.7% in EMPEROR-Reduced and EMPEROR-Preserved, respectively).

|||||||| ||||||||||||||||||| |||||||||| || |||| ||||||| |||||| ||| |||||||||| || ||||||||||||| |||||||||||||||| | ||||| || |||| || ||| |||||||||| |||||||| ||| || ||||| | ||||||||| |||||||| |||||||||| |||| ||||||| ||||||||||| |||||| ||| ||||||||| ||||||| ||| |||| |||||| |||||||| |||||||||| |||||||| ||| ||||||| ||| ||||||||| |||||||| || ||| ||||| ||||||| ||| ||| || |||||||||| |||||||||| |||||| ||||||||| |||||| ||||||| ||| |||||||||||||||||| | ||||| || |||| || ||| |||||||||| |||||||| ||| || ||||| | ||||||||| |||||||| |||||||||| |||| ||||||| ||||||||||| |||||| ||| ||||||||| |||||| ||||| ||| |||| || ||| ||||||||||||| ||| ||||||| ||||||| |||||||||||||| ||| |||| |||||| |||||||| |||||||||| |||||||| ||| ||||||| ||| ||||||||| |||||||| || ||| ||||| ||||||| ||| ||| || |||||||||| |||||||||| |||||| ||||||||| |||||| |||||||

Table 14: Redacted

Note: This table has been redacted as per sponsor’s request.

Efficacy

Only those efficacy outcomes and analyses of subgroups identified in the review protocol are reported subsequently. Refer to Appendix 3 for detailed efficacy data.

Time to First Event of Adjudicated CV Death or HHF

The results of the primary efficacy end point for both pivotal trials are presented in Table 15, Figure 3, and Figure 4.

EMPEROR-Reduced Trial

A composite of time to first event of adjudicated CV death or HHF occurred in 361 patients (19.4%) in the empagliflozin group and 462 patients (24.7%) in the placebo group. The incidence rate was lower in the empagliflozin group (15.77 per 100 person-years at risk) compared with the placebo group (21.0 per 100 person-years at risk). The HR for time to first event of adjudicated CV death or HHF was 0.75 (95% CI, 0.65 to 0.86; P < 0.0001) in favour of the empagliflozin group. Although individual components of the composite primary end point were not formally tested for significance, the proportion of HHF was lower in the empagliflozin group (13.2%) compared with placebo (18.3%), while the total proportion of CV deaths was similar across the treatment groups (10.0% versus 10.8%, respectively).

EMPEROR-Preserved Trial

A composite of time to first event of adjudicated CV death or HHF occurred in 415 patients (13.8%) in the empagliflozin group and 511 patients (17.1%) in the placebo group. The incidence rate was lower in the empagliflozin group compared with placebo (6.86 and 8.67 per 100 person-years at risk, respectively). The HR for time to first event of adjudicated CV death or HHF was 0.79 (95% CI, 0.69 to 0.90; P = 0.0003) in favour of the empagliflozin group. The proportion of HHF was lower in the empagliflozin group (8.6%) relative to placebo (11.8%), while the total proportion of CV deaths was similar across the treatment groups (7.3% versus 8.2% in the empagliflozin and placebo groups, respectively).

Subgroup Analysis

The primary end point subgroup analysis in both pivotal trials is presented in (Appendix 3, Table 36). The analyses may not have been powered to detect a treatment difference and there were no adjustments made for multiplicity. As such, all subgroup analyses are exploratory in nature.

In EMPEROR-Reduced, while the effect of empagliflozin on the primary end point events was generally consistent across pre-specified subgroups, potential differences were noted depending on LVEF (P value for interaction < 0.05). In EMPEROR-Preserved, the effect of empagliflozin on the primary end point events was generally consistent across pre-specified subgroups.

In EMPEROR-Reduced, the sensitivity analyses (Appendix 3, Figure 24, and Figure 25) for the primary end point, including those assessing missing data, were exploratory and were generally consistent with the primary analysis. In EMPEROR-Preserved, the sensitivity analyses for the primary end point, including those assessing missing data, were generally consistent with the primary analysis, aside from the treatment effect associated with the COVID-19 outbreak.

Table 15: Time to First Event of Adjudicated CV Death or HHF^a — EMPEROR-Reduced and EMPEROR-Preserved, RS

CI = confidence interval; CV = cardiovascular; HHF = hospitalization for heart failure; HR = hazard ratio; RS = randomized set.

^aAn independent group of medical experts performed a central, blinded adjudication of the outcomes.

^bIncidence rate was calculated as the number of patients with events per 100 person-years at risk.

^cCox proportional hazards model included the following factors: treatment, age, geographical region, diabetes status, sex, left ventricular ejection fraction, and baseline estimated glomerular filtration rate.

^dBased on the reduced 2-sided significance level of 0.0496 resulting from the interim analysis.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Figure 3: Redacted

Note: Confidential figure redacted at the request of the sponsor.

||||| |||||||||||||| ||| | ||||||||| ||||||||| ||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||| ||||| || ||||||||||| || || |||| ||||||||||| |||||| ||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 4: Redacted

Note: Confidential figure redacted at the request of the sponsor.

||||| |||||||||||||| ||| | |||||||||| ||||||||| ||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||| ||||| || ||||||||||| || || |||| ||||||||||| |||||| ||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| |||||||

Occurrence of Adjudicated HHF (First and Recurrent)

The results of the first key secondary outcome for both pivotal trials are presented in Table 16, Figure 5, and Figure 6.

EMPEROR-Reduced

The number of patients with adjudicated HHF was 259 (13.2%) in the empagliflozin group and 324 (18.3%) in the placebo group. The total number of HHF events (first and recurrent) was lower in patients who received empagliflozin compared with those who received placebo (388 versus 553, respectively). The total proportion of CV deaths was similar across the treatment groups (10.0% and 10.8% in the empagliflozin and placebo groups, respectively), with an HR of 0.90 (95% CI, 0.70 to 1.15). The hazard rate of recurrent HHF was significantly reduced in the empagliflozin group compared with placebo, with an HR of 0.70 (95% CI, 0.58 to 0.85; P = 0.0003).

EMPEROR-Preserved

The number of patients with adjudicated HHF was 246 (8.6%) in the empagliflozin group and 352 (11.8%) in the placebo group. The total number of HHF events was lower in patients who received empagliflozin compared with those who received placebo (407 versus 541, respectively). The total proportion of CV deaths was similar across the treatment groups (7.3% and 8.2% in the empagliflozin and placebo groups, respectively), with an HR of 0.89 (95% CI, 0.71 to 1.12). The hazard rate of recurrent HHF was significantly reduced in the empagliflozin group compared with placebo, with an HR of 0.73 (95% CI, 0.61 to 0.88; P = 0.0009).

A secondary end point subgroup analysis for both pivotal trials is presented in Appendix 3 (Table 38). The analyses may not have been powered to detect a treatment difference and there were no adjustments made for multiplicity. As such, all subgroup analyses are exploratory in nature. In EMPEROR-Preserved, while the effect of empagliflozin on the occurrence of adjudicated HHF was generally consistent across pre-specified subgroups, potential differences were noted among LVEF and prior MRA use subgroups (P value for interaction < 0.05). In EMPEROR-Reduced, the effect of empagliflozin on the occurrence of adjudicated HHF was consistent across pre-specified subgroups.

The results of the sensitivity analyses, including those assessing missing data, were consistent with the results of the primary analysis for the occurrence of adjudicated first and recurrent HHF (Appendix 3, Figure 30 and Figure 31).

Table 16: Occurrence of HHF (First and Recurrent) — EMPEROR-Reduced and EMPEROR-Preserved, RS

CI = confidence interval; CV = cardiovascular; HHF = hospitalization for heart failure; HR = hazard ratio.

^aJoint frailty model included factors for age, baseline estimated glomerular filtration rate (CKD-EPIcr equation), geographical region, baseline diabetes status, sex, baseline left ventricular ejection fraction, and treatment.

^bBased on the reduced 2-sided significance level of 0.0496 resulting from the interim analysis.

^cPositive correlation between recurrent (HHF) and terminal (CV death) events if alpha > 0.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Figure 5: Redacted

Note: Confidential figure redacted at the request of the sponsor.

||||| |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| |||||||||||||||| ||||||||| |||||||| ||| ||||||||||| ||||| ||| ||||||||| |||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 6: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| |||||||||||||||| ||||||||| |||||||| ||| ||||||||||| ||||| ||| ||||||||| |||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| ||||||||

eGFR (CKD-EPIcr Equation) Slope of Change From Baseline

The results of the second key secondary outcome for both pivotal trials are presented in Table 17, Figure 7, and Figure 8. The analysis was performed based on the TS and using observed “on treatment” data.

In EMPEROR-Reduced, over the double-blind treatment period, the rate of decline in the eGFR (CKD-EPIcr equation) per year was slower in the empagliflozin group (−0.55 mL/min/1.73 m² per year; 95% CI, −0.99 to −0.10) than in the placebo group (−2.28 mL/min/1.73 m² per year; 95% CI, −2.73 to −1.83), with a between-group difference in slope of 1.73 per year (95% CI, 1.10 to 2.37).

In EMPEROR-Preserved, over the double-blind treatment period, the rate of decline in the eGFR (CKD-EPIcr equation) per year was slower in the empagliflozin group (−1.25 mL/min/1.73 m² per year; ||| ||| ||||| || |||||) than in the placebo group (−2.62 mL/min/1.73 m² per year; ||| ||| ||||| || |||||), with a between-group difference in slope of 1.36 per year (95% CI, 1.06 to 1.66).

A secondary end point subgroup analysis for both pivotal trials is presented in Appendix 3 (Table 39). The analyses may not have been powered to detect a treatment difference and there were no adjustments made for multiplicity. As such, all subgroup analyses are exploratory in nature | || |||||||||||||||| ||||| ||| |||||| || ||||||||||||| || ||| |||| ||||| || |||||| |||| |||||||| ||| ||||||||| |||||||||| |||||| ||||||||||||| |||||||||| ||||||||| ||||||||||| |||| ||||| ||||||||| || |||||||| |||| ||||||||| || ||| ||||||||||| | ||||||. In EMPEROR-Preserved, potential differences in the benefit of empagliflozin on the eGFR slope of change from baseline were noted, depending on baseline diabetes status.

Table 17: eGFR (CKD-EPIcr Equation) Slope of Change From Baseline — EMPEROR-Reduced and EMPEROR-Preserved, TS

CI = confidence interval; CV = cardiovascular; eGFR = estimated glomerular filtration rate; CKD-EPIcr = Chronic Kidney Disease Epidemiology Collaboration creatinine; HHF = hospitalization for heart failure; RS = randomized set.

Note: Model included factors for age, baseline estimated glomerular filtration rate (CKD-EPI cr), region, baseline diabetes status, sex, baseline left ventricular ejection fraction, baseline eGFR-by-time interaction, treatment-by-time interaction, and treatment. Intercept and slope were allowed to vary randomly between patients.

^aBased on a 2-sided significant level of 0.001.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Figure 7: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| || | ||||||| |||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 8: Redacted

|||| | |||||||||||||| || | ||||||| |||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| |||||||

Other Secondary and Further End Points (Exploratory)

Other Mortality Outcomes

A summary of other mortality-related outcomes for both EMPEROR trials is presented in Table 18. These mortality outcomes were tested in a non-hierarchical sequence without adjustments for multiplicity and were exploratory in nature. In EMPEROR-Reduced, a total of 249 patients (13.4%) in the empagliflozin group and 266 patients (14.2%) in the placebo group died from any cause (HR = 0.92; 95% CI, 0.77 to 1.10). In EMPEROR-Preserved, a total of 422 patients (14.1%) in the empagliflozin group and 427 patients (14.3%) in the placebo group died from any cause (HR = 1.00; 95% CI, 0.87 to 1.15) (Figure 9 and Figure 10). The majority of deaths (75.5%) in the EMPEROR-Reduced trial and nearly half of deaths (54.5%) in the EMPEROR-Preserved trial were due to CV causes (Table 37). The analyses showed no differences between treatment groups in the time to adjudicated CV death (HR = 0.92; 95% CI, 0.75 to 1.12; and HR = 0.91; 95% CI, 0.76 to 1.09, in EMPEROR-Reduced and EMPEROR-Preserved, respectively) (Figure 11 and Figure 12) ||| |||| || ||||||||||| |||||| ||||| ||| | ||||| ||| ||| |||| || ||||| ||| || | ||||| |||||| |||| || ||||| || ||||||||||||||| ||| |||||||||||||||||| |||||||||||||| ||||| || |||||||||| |||| ||| ||||||| |||||||| |||||||||

Table 18: Time to All-Cause Mortality, Time to Adjudicated CV Death, and Time to Adjudicated Non-CV Death — EMPEROR-Reduced and EMPEROR-Preserved, RS

CI = confidence interval; CV = cardiovascular; HR = hazard ratio.

^aAn independent group of medical experts performed a central, blinded adjudication of the outcomes.

^bIncidence rate was calculated as the number of patients with events per 100 person-years at risk.

^cCox proportional hazards model included the following factors: treatment, age, geographical region, diabetes status, sex, left ventricular ejection fraction, and baseline estimated glomerular filtration rate (CKD-EPIcr equation).

^dP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Figure 9: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| |||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 10: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| |||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| |||||||

Figure 11: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||| ||||| || ||||||||||| ||| ||||||||||| |||||| ||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 12: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||| ||||| || ||||||||||| ||| ||||||||||| |||||| ||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| |||||||

Other Hospitalization-Related Outcomes

The results of other hospitalization-related outcomes for both pivotal trials are presented in Table 19. These hospitalization-related outcomes were tested in a non-hierarchical sequence without adjustments for multiplicity and were exploratory in nature.

The proportion of patients with 1 or more adjudicated HHFs was lower in the empagliflozin group compared with placebo in both EMPEROR trials. A Cox proportional hazards regression analysis was used to assess the time to first adjudicated HHF (Figure 13 and Figure 14). The hazard of first adjudicated HHF was lower in the empagliflozin group compared with placebo, with an HR of 0.69 (95% CI, 0.59 to 0.81) in EMPEROR-Reduced and 0.71 (95% CI, 0.60 to 0.83) in EMPEROR-Preserved. A joint frailty model was used to examine the occurrence of all-cause hospitalization (first and recurrent) that accounts for the dependence between recurrent all-cause hospitalization and all-cause mortality in both EMPEROR trials. The total number of all-cause hospitalizations was lower in the empagliflozin group compared with placebo (1,364 versus 1,570 in EMPEROR-Reduced, and 2,566 versus 2,769 in EMPEROR-Preserved, respectively). The hazard of recurrent all-cause hospitalization was lower in the empagliflozin group compared with placebo (HR = 0.85; 95% CI, 0.75 to 0.95, and HR = 0.93; 95% CI, 0.85 to 1.01, in EMPEROR-Reduced and EMPEROR-Preserved, respectively), and it was positively correlated with all-cause mortality in both trials. A Cox proportional hazards regression analysis was used to assess the time to all-cause hospitalization or all-cause mortality, and time to investigator-defined CV hospitalization. The incidence rate of first all-cause hospitalization or all-cause mortality was lower in the empagliflozin group compared with placebo in both trials (35.58 versus 43.82 per 100 person-years at risk, and 26.85 versus 29.18 per 100 person-years at risk in EMPEROR-Reduced and EMPEROR-Preserved, respectively), with an HR of 0.81 (95% CI, 0.74 to 0.90), and 0.92 (95% CI, 0.85 to 0.99) in the EMPEROR-Reduced and EMPEROR-Preserved trials, respectively (Figure 15 and Figure 16). The hazard rate of investigator-defined CV hospitalization was lower in the empagliflozin group compared with placebo in both trials, with an HR of 0.75 (95% CI, 0.67 to 0.85) and 0.85 (95% CI, 0.77 to 0.94) in the EMPEROR-Reduced and EMPEROR-Preserved trials, respectively. Since the analysis of the time from the first to the second adjudicated HHF included only patients with 1 or more HHF events, it was not a randomized treatment comparison and is therefore affected by selection bias.

Table 19: Time to First Adjudicated HHF, Time From First to Second HHF, Time to First All-Cause Hospitalization, and Occurrence of All-Cause Hospitalization — EMPEROR-Reduced and EMPEROR-Preserved, RS

CI = confidence interval; CV = cardiovascular death; HHF = hospitalization for heart failure; HR = hazard ratio; RS = randomized set.

^aCox proportional hazards model included the following factors: treatment, age, geographical region, diabetes status, sex, LVEF, and baseline estimated glomerular filtration rate (eGFR [CKD-EPIcr equation]).

^bAn independent group of medical experts performed a central, blinded adjudication of the outcomes.

^cIncidence rate was calculated as the number of patients with events per 100 person-years at risk.

^dP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

^eJoint frailty model included the following factors: age, baseline eGFR (CKD-EPIcr equation), geographical region, baseline diabetes status, sex, baseline left ventricular ejection fraction, treatment.

^fPositive correlation between recurrent (HHF) and terminal events (CV) events if alpha > 0.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Figure 13: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||| ||||||||||| |||| ||||||||||| ||||||||| ||||||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 14: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||| ||||||||||| |||| ||||||||||| ||||||||| ||||||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| |||||||

Figure 15: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||||||| |||||||||||||||| ||||||||||| ||||||||| ||||||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 16: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||||||| |||||||||||||||| ||||||||||| ||||||||| ||||||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| |||||||

Composite Renal Outcome (Exploratory)

The composite renal end point included the following events: chronic dialysis, renal transplant, or sustained reduction in eGFR (CKD-EPIcr equation) (Figure 17 and Figure 18). Sustained reduction of eGFR was determined by 2 or more consecutive post-baseline central laboratory measurements separated by at least 30 days. In EMPEROR-Reduced, the composite renal end point occurred in 30 patients (1.6%) in the empagliflozin group and 58 patients (3.1%) in the placebo group. The incidence rate was 1.56 per 100 person-years at risk in the empagliflozin group versus 3.07 per 100 person-years at risk in the placebo, with an HR of 0.50 (95% CI, 0.32 to 0.77). In EMPEROR-Preserved, the composite renal end point occurred in 108 patients (3.6%) in the empagliflozin group and 112 (3.7%) in the placebo group. The incidence rate was similar between the empagliflozin and placebo groups (2.13 and 2.23 per 100 person-years at risk, respectively), with an HR of 0.95 (95% CI, 0.73 to 1.24).

Table 20: Time to the First Event in the Composite Renal End Point — EMPEROR-Reduced and EMPEROR-Preserved, RS

CI = confidence interval; eGFR = estimated glomerular filtration rate; HR = hazard ratio; NR = not reported; RS = randomized set.

Notes: Composite renal end point = chronic dialysis (with a frequency of twice per week or more for at least 90 days), renal transplant, sustained reduction from baseline in eGFR of ≥ 40%, sustained eGFR < 15 mL/min/1.73 m² for patients with baseline eGFR ≥ 30 mL/min/1.73 m², or sustained eGFR < 10 mL/min/1.73 m² for patients with baseline eGFR < 30 mL/min/1.73 m². Sustained was determined by 2 or more consecutive post-baseline central laboratory measurements separated by at least 30 days (the first to last of the consecutive eGFR values).

The Cox proportional hazards model included the following factors: treatment, age, geographical region, diabetes status, sex, left ventricular ejection fraction, and baseline eGFR (CKD-EPIcr equation).

^aIncidence rate was calculated as the number of patients with events per 100 person-years at risk.

^bP value has not been adjusted for multiple testing (i.e., the type I error rate has not been controlled).

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Figure 17: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||| ||||| || ||| ||||||||| ||||| |||||||| ||||||||||| ||||||||| ||||||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 18: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| ||| || |||| | |||||||| || ||||| || | |||||||||| ||||||||||||||| |||||||||| ||||||||| |||||||| ||| |||| || ||||| ||||| || ||| ||||||||| ||||| |||||||| ||||||||||| ||||||||| ||||||||| || ||||||||| ||||||||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| |||||||

HRQoL, Symptoms of HF, and Functional Status (Exploratory)

KCCQ Scores

||| ||| |||||||| || ||| |||| ||||||| |||| || |||| |||| ||||||| |||| ||||| ||| ||||| || |||||||| |||| ||| ||||||||||||| ||| ||||||| ||||||| |||||||||||| || ||| ||||||||||||||| |||||| ||| |||| |||| ||| ||||| || ||||||||| |||||||||||| || ||| ||||||||||||||||| ||||||

KCCQ Clinical Summary Score

The KCCQ-CSS incorporates the physical limitation and total symptom domain into a single score that is transformed into a range of 0 to 100 (higher scores represent better outcomes).

In EMPEROR-Reduced, the analysis based on the RS, including both on- and off-treatment values (Table 21 and Figure 19), showed a smaller decline from baseline of −1.30 points (SE = 0.69) in the empagliflozin group than in the placebo group (−3.36 points; SE = 0.69) in the KCCQ-CSS at week 52, with an adjusted mean difference of 2.06 (95% CI, 0.16 to 3.96) favouring empagliflozin. For patients who died, the worst score (score of 0) was imputed for all subsequent scheduled visits after the date of death. Responder analyses for an improvement (an increase in score of at least 5 points at week 52 from baseline) or a deterioration (a decrease in at least 5 points) were also conducted, although these outcomes were not part of the statistical testing hierarchy. At week 52, 40.0% of patients in the empagliflozin group reported at least a 5-point increase in the KCCQ-CSS compared with 35.9% of patients in the placebo group (OR = 1.23; 95% CI, 1.05 to 1.45).

|| |||||||||||||||||| ||| |||||||| ||||| || ||| ||| ||||||||| |||| ||| ||| ||||||||||||| |||||| |||||||||||| ||| |||||||||||| |||||| | |||||| |||||||| || ||| |||| |||||||| ||||||| ||||| |||| |||||||| || |||| |||||| ||| | ||||| || ||| ||||||||||||| ||||| |||||||| || ||||||| |||||| ||||||| || | ||||| || |||| ||| |||| || |||||||| |||| |||||||||| || |||| |||| ||| |||| || ||||| |||||||| |||||||||||||| ||||||||| |||||||| |||||| |||| || |||| ||| ||||| || |||||||| || ||| ||||||||||||| ||||| |||||||| || ||||| | ||||||| |||||||| || |||| |||||||| ||||||| |||||| |||||||| |||| ||||| || |||||||| || ||| ||||||| |||||| |||| |||| ||||| || |||| |||| ||| |||| || ||||||

Table 21: Change From Baseline in KCCQ Clinical Summary Score — EMPEROR-Reduced and EMPEROR-Preserved, RS

CI = confidence interval; KCCQ = Kansas City Cardiomyopathy Questionnaire; OR = odds ratio; RS = randomized set; SE = standard error.

Notes: Mixed-model repeated measures included age, baseline estimated glomerular filtration rate (eGFR [CKD−EPIcr]) as linear covariates, region, baseline diabetes status, sex, baseline left ventricular ejection fraction (LVEF), week reachable, treatment-by-visit interaction, and baseline KCCQ clinical summary score by visit interaction as fixed effects.

The 95% CI was not adjusted for multiple comparisons.

^aBased on the randomized set, including both on- and off-treatment values for patients who died, the worst score (score of 0) was imputed at all subsequent scheduled visits after the date of death.

^bLogistic regression includes terms for baseline KCCQ total symptom score, age, baseline eGFR (CKD−EPIcr), treatment, region, diabetes at baseline, sex, and baseline LVEF. Patients who are lost to follow-up, withdrew consent, or died before the planned week 52 visit are considered as having deterioration.

Source: Clinical Study Reports for EMPEROR-Reduced⁷ and EMPEROR-Preserved trials.⁸

Figure 19: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| || | |||||||||| |||| || | |||||||| |||||||||||| || ||| ||||||||| |||| ||| ||| ||||||||||||| ||||||| ||| |||||||| ||| ||||| | ||||| ||||| |||||| || || ||| ||||||| || ||| |||||||||| ||||||||| |||||| ||||| ||| |||| || |||||| |||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 20: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| || | |||||||||| |||| || | |||||||| |||||||||||| || ||| ||||||||| |||| ||| ||| ||||||||||||| ||||||| ||| |||||||| ||| ||||| | ||||| ||||| |||||| || || ||| ||||||| || ||| |||||||||| ||||||||| |||||| ||||| ||| |||| || |||||| |||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| |||||||

KCCQ Total Symptom Score

Symptoms were measured using the KCCQ-TSS, which incorporates symptom burden and frequency into a single score that is transformed into a range of 0 to 100 (higher scores represent better outcomes).

|| |||||||||||||||| ||| |||||||| ||||| || ||| ||| ||||||||| |||| ||| ||| ||||||||||||| |||||| ||||||||||| ||| ||||||||||| |||||| | ||||||| ||||||| |||| |||||||| || ||||| |||||| ||| | ||||| || ||| ||||||||||||| |||| || ||| ||||||| ||||| |||||| ||||||| || | ||||| || ||| |||| ||||| ||||||| ||||| || |||| ||| |||| || |||||||| |||| |||||||||| || |||| |||| ||| |||| || ||||| |||||||| |||||||||||||| ||||||||| |||||||| |||||| |||| || |||| ||| ||||| || |||||||| || ||| ||||||||||||| ||||| |||||||| || ||||| | ||||||| |||||||| || |||| |||||||| ||||||| |||||| |||||||| |||| ||||| || |||||||| || ||| ||||||| |||||| |||| || || || |||| |||| ||| |||| || ||||||

|| |||||||||||||||||| ||| |||||||| ||||| || ||| ||| ||||||||| |||| ||| ||| ||||||||||||| |||||| |||||||||||| ||| |||||||||||| |||||| || |||||||| |||| |||||||| || |||| |||||| ||| | ||||| || ||| ||||||||||||| |||||||| || ||||||| |||||| ||||||| || | ||||| || ||| |||| ||||| ||||||| ||||| || |||| ||| |||| || |||||||| |||| |||||||||| || |||| |||| ||| |||| || ||||| |||||||| |||||||||||||| ||||||||| |||||||| |||||| |||| || |||| ||| ||||| || |||||||| || ||| ||||||||||||| ||||| |||||||| || ||||| | ||||||| |||||||| || |||| ||||| ||||||| |||||| |||||||| |||| ||||| || |||||||| || ||| ||||||| |||||| |||| || || || |||| |||| ||| |||| || ||||||

Table 22: Redacted

Note: Table redacted as per sponsor’s request.

Figure 21: Redacted

Note: Confidential figure redacted at the request of the sponsor.

|||| | |||||||||||||| || | |||||||||| |||| || | |||||||| |||||||||||| || ||| ||||||||| |||| ||| ||| ||||||||||||| ||||||| ||| |||||||| ||| ||||| | ||||| ||||| |||||| || || ||| ||||||| || ||| |||||||||| ||||||||| |||||| ||||| ||| |||| || |||||| |||||||| |||||||| ||||| ||||||| ||| ||||||||||||||| |||||||

Figure 22: Redacted

|||| | |||||||||||||| || | |||||||||| |||| || | |||||||| |||||||||||| || ||| ||||||||| |||| ||| ||| ||||||||||||| ||||||| ||| |||||||| ||| ||||| | ||||| ||||| |||||| || || ||| ||||||| || ||| |||||||||| ||||||||| |||||| ||||| ||| |||| || |||||| |||||||| |||||||| ||||| ||||||| ||| ||||||||||||||||| ||||||

KCCQ Overall Summary Score

The KCCQ-OSS incorporates the physical limitation, total symptom, social limitation, and HRQoL into a single score that is transformed into a range of 0 to 100 (higher scores represent better outcomes).

|| |||||||||||||||| ||| |||||||| ||||| || ||| ||| ||||||||| |||| ||| ||| ||||||||||||| |||||| |||||||||| |||||| | ||||||| ||||||| |||| |||||||| || ||||| |||||| ||| | ||||| || ||| ||||||||||||| ||||| |||| || ||| ||||||| ||||| |||||| ||||||| || | ||||| || ||| |||| ||||||| ||||||| ||||| || |||| ||| |||| || |||||||| |||| |||||||||| || |||| |||| ||| |||| || ||||| |||||||| ||||||||||||||||| |||||||||||||||||| ||| |||||||| ||||| || ||| ||| ||||||||| |||| ||| ||| ||||||||||||| |||||| ||||||||||| |||||| | |||||| ||||||||||| |||| |||||||| || |||| |||||| ||| | ||||| || ||| ||||||||||||| ||||| |||||||| || ||| ||||||| ||||| |||||| ||||||| || | ||||| || ||| |||| ||||||| ||||||| ||||| || |||| ||| |||| || |||||||| |||| |||||||||| || |||| ||||| || ||||| |||||||| ||||||||||||||

Table 23: Redacted

||| |||||||||| ||||||||| || | |||||||| |||||| || | |||||||||| |||||||||| || ||| ||||||||| |||| ||| ||| ||||||||||||| ||||||| ||| |||||||| ||| ||||| | ||||| ||||| |||||| || || ||| ||||||| || ||| |||||||||| ||||||||| |||||| ||||| ||| |||| || ||||||

Notes: Table redacted as per sponsor’s request.

Other KCCQ data reported included the change from baseline to week 52 in the KCCQ quality of life and patient-preferred outcomes, as well as the results of the analyses based on the TS, including only observed on-treatment data (Appendix 3).

|||||||||||||| ||||||| || |||| |||||||| || |||||||||||| ||| |||||||| || ||| ||||| |||||||||| ||||| || ||| || ||| ||| ||||||||||||||| |||||| |||| || |||| |||| ||||||| ||| ||||| ||| ||||| || |||||||| || ||| ||||||||||||| ||| ||||||| ||||||| ||||||||||||| ||||| |||| ||| |||| |||| ||||||| ||| ||||| || |||||||| || |||| |||||| || ||| ||||||||||||||||| |||||| |||| || |||| |||| ||||||| ||| ||||| ||| ||||| || |||||||| || ||| ||||||||||||| ||| ||||||| ||||||| ||||||||||||| ||||| |||| ||| |||| |||| ||||||| ||| ||||| ||| ||||| || ||||||||| ||||||||||||| ||| |||| ||||||| ||||||| ||| |||||| |||| |||||||| || ||| |||||||| ||| |||||||| ||||| ||||||||||| ||||||||||| ||| ||||| |||| || ||||||||||| |||||| ||| ||||||||| |||||| |||||||||||| ||||||||||| |||| ||| ||| |||||| |||||| |||| | |||||| |||||||||| || |||||||| |||||||| ||||||||||| || ||| |||| |||||| || ||||| || ||| ||| || ||| ||||||||||||| ||||| |||||||| || ||||||| || ||| ||||||||||||||| ||| ||||||||||||||||| |||||| ||||||||||||

Table 24: Redacted