Cardio–renal–metabolic syndrome: clinical features and dapagliflozin eligibility in a real‐world heart failure cohort

Abstract

Aims

The Cardiovascular Outcomes Retrospective Data analysIS in Heart Failure (CORDIS‐HF) is a single‐centre retrospective study aimed to (i) clinically characterize a real‐world population with heart failure (HF) with reduced (HFrEF) and mildly reduced ejection fraction (HFmrEF), (ii) evaluate impact of renal–metabolic comorbidities on all‐cause mortality and HF readmissions, and (iii) determine patients' eligibility for sodium–glucose cotransporter 2 inhibitors (SGLT2is).

Methods and results

Using a natural language processing algorithm, clinical data of patients diagnosed with HFrEF or HFmrEF were retrospectively collected from 2014 to 2018. Mortality and HF readmission events were collected during subsequent 1 and 2 year follow‐up periods. The predictive role of patients' baseline characteristics for outcomes of interest was assessed using univariate and multivariate Cox proportional hazard models. Kaplan–Meier analysis was used to determine if type 2 diabetes (T2D) and chronic kidney disease (CKD) impacted mortality and HF readmission rates. The European SGLT2i label criteria were used to assess patients' eligibility. The CORDIS‐HF included 1333 HF patients with left ventricular ejection fraction (LVEF) < 50% (413 HFmrEF and 920 HFrEF), who were predominantly male (69%) with a mean [standard deviation (SD)] age of 74.7 (12.3) years. About one‐half (57%) of patients presented CKD and 37% T2D. The use of guideline‐directed medical therapy (GDMT) was high (76–90%). HFrEF patients presented lower age [mean (SD): 73.8 (12.4) vs. 76.7 (11.6) years, P < 0.05], higher incidence of coronary artery disease (67% vs. 59%, P < 0.05), lower systolic blood pressure [mean (SD): 123 (22.6) vs. 133 (24.0) mmHg, P < 0.05], higher N‐terminal pro‐hormone brain natriuretic peptide (2720 vs. 1920 pg/mL, P < 0.05), and lower estimated glomerular filtration rate [mean (SD): 51.4 (23.3) vs. 54.1 (22.3) mL/min/1.73 m², P < 0.05] than those with HFmrEF. No differences in T2D and CKD were detected. Despite optimal treatment, event rates for the composite endpoint of HF readmission and mortality were 13.7 and 8.4/100 patient years. The presence of T2D and CKD negatively impacted all‐cause mortality [T2D: hazard ratio (HR) = 1.49, P < 0.01; CKD: HR = 2.05, P < 0.001] and hospital readmission events in all patients with HF. Eligibility for SGLT2is dapagliflozin and empagliflozin was 86.5% (n = 1153) and 97.9% (n = 1305) of the study population, respectively.

Conclusions

This study identified high residual risk for all‐cause mortality and hospital readmission in real‐world HF patients with LVEF < 50% despite GDMT. T2D and CKD aggravated the risk for these endpoints, indicating the intertwinement of HF with CKD and T2D. SGLT2i treatment that clinically benefits these different disease conditions can be an important driver to lower mortality and hospitalizations in this HF population.

Introduction

Cardiovascular diseases (CVDs) remain the leading cause of death, globally and in the European Union (EU), with costs exceeding €210 billion per year in the EU alone.¹ More than 6 million new cases and over 1.8 million CVD-related deaths occur in the EU each year.¹ With a prevalence of 1–2% in the general population, rising to over 10% among people aged 70 years and above, heart failure (HF) heavily contributes to the high cardiovascular (CV) mortality and morbidity in developed countries.¹ Considering that more than 10 million people in the EU suffer from HF, HF-related costs contribute to more than 2% of the total health care budget.¹ Despite advances in diagnosis and the introduction of specialized therapies, HF prognosis remains poor and its management challenging.²

The incremental impact and the intertwinement of metabolic diseases have led to the terminology of cardio–renal–metabolic disease.³ This novel paradigm conveys that not only HF, chronic kidney disease (CKD), and type 2 diabetes (T2D) frequently co-exist in patients but also they share common risk factors, such as central obesity, dyslipidaemia, hypertension, smoking, and inflammation, thereby heavily impacting clinical outcome.^3,4 As such, the presence of T2D conveys a more than two-fold risk of developing HF and worsens its clinical outcome, resulting in a higher mortality risk for T2D patients vs. people without T2D.^5,6 Similarly, the presence of CKD, characterized by the presence of albuminuria and/or impaired renal function, is associated with an increased risk of developing CVD and HF, HF hospitalization, and CV death in patients with HF.^4,7 Finally, HF also impacts outcomes in patients with CKD and T2D, and the combination of HF and CKD conveys the highest risk of CV-related and all-cause mortality for patients with T2D, who were initially free of CVD and renal disease.^4,8

Due to this close entanglement of CVD and renal and metabolic diseases, it is key to approach them by a medical treatment that simultaneously targets HF, CKD, and T2D. Currently, new medications tackling these different diseases, such as sodium–glucose cotransporter 2 inhibitors (SGLT2is), are available and enable the health care professionals to take a holistic approach and reduce the risk of complications in these patients.

Of note, several CV outcome trials investigating SGLT2is, such as canagliflozin, dapagliflozin, and empagliflozin, have shown a robustly reduced risk of CV events for patients with T2D.^9–11 Furthermore, dedicated HF and renal outcome trials reported a complementary benefit of SGLT2is in reducing the risk of hospitalization for HF and renal complications.^12–15 Recently, the Dapagliflozin and Prevention of Adverse outcomes (DAPA)-HF and DAPA-CKD trials in patients suffering from HF with reduced ejection fraction (HFrEF) and CKD showed a relative risk reduction for worsening HF or CV-related death of 26% and 29%, respectively.^13,16,17 A reduction of 39% was observed for the composite of a sustained ≥50% decline in the estimated glomerular filtration rate (eGFR), end-stage kidney disease, or death from renal or CV causes, irrespective of the presence of T2D.^13,16,17

Despite promising evidence on the benefit of dapagliflozin and empagliflozin for improved clinical outcomes in patients with (co-existing) T2D, HF, and CKD,^12,13,16,18 clinical trials are unavoidably biased due to strict patient inclusion and exclusion criteria. Therefore, data from hospital HF registries and real-world evidence are needed and are of utmost importance to translate clinical trial results into the real-life setting.¹⁹ The present study aimed to characterize a real-world HF patient population in a tertiary centre and to evaluate the effects of the renal–metabolic comorbidities on clinical outcomes, such as all-cause mortality and HF readmission. Furthermore, the study aimed to estimate the proportion of patients who are eligible for initiation of SGLT2is in a real-world setting.

Methods

Study design

This single-centre study, named Cardiovascular Outcomes Retrospective Data analysIS in HF (CORDIS-HF), was conceived as a retrospective analysis of data collected from patient electronic medical records (EMRs). These data were processed by the natural language processing (NLP) algorithm and the artificial intelligence platform of LynxCare (LynxCare Inc., Leuven, Belgium).

Briefly, patient EMRs were collected between the 1 January 2014 and 1 July 2018. The index event was defined as the HF diagnosis, with its date as the corresponding index date. A lookback period of 1 year was considered, such that data from 1 January 2013 to 31 December 2013 were included for patients with index dates close to the study initiation date. Patient comorbidities and baseline characteristics were recorded at 1 July 2018 and, starting from this time, there was a 2 year follow-up during which hospitalization and mortality rates were collected (Figure 1). These dates ensure minimum impact of the COVID-19 pandemic on the collected data, as the latest follow-up information is from 1 July 2020.

[IMAGE OMITTED. SEE PDF]

Data collection and analysis

EMRs of patients with HF were obtained from a representative tertiary HF referral centre [Cardiovascular Centre, Onze Lieve Vrouw (OLV) Hospital, Aalst, Belgium] and collected during the observation period. After anonymization and aggregation, study data on 147 primary and derived variables, including demography, clinical signs and symptoms, medical conditions, haematology and biochemistry profiles, electrocardiograms, and medical treatment, were extracted from the EMRs by a data mining and NLP software compatible with the general data protection regulation (Care Monitor, LynxCare).²⁰ Briefly, the NLP pipeline automatically processed and extracted clinical data from structured, semi-structured, and unstructured reports and encoded the pre-defined data points. Cross-sectional analyses of EMRs using this NLP algorithm have been described and validated previously in patients undergoing total hip arthroplasty.²⁰ The Care Monitor NLP algorithm was specifically adapted to the hospital context and optimized for the corresponding (sub)-specializations and allowed to detect pre-defined data points using statistically validated text mining techniques and comprehensive clinical ontologies.

Study population

All adult (≥18 years) patients with ejection fraction (EF) < 50% and at least one diagnosis of HF (index event) based on the International Statistical Classification of Diseases and Related Health Problems (ICD)-9 or 10 coding and on hospital-specific labelling of the patients prior to the cohort description date (1 July 2018) were included in the patient cohort. Baseline characteristics (demographics, laboratory data, comorbidities, and medication use) were collected prior to the cohort description date (Figure 1), whereas all-cause mortality and hospital admission events were analysed during the 1 and 2 year periods during the follow-up phase.

The study was conducted in accordance with the Declaration of Helsinki and the good clinical practice guidelines of the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use and received ethical approval from the institutional review board of the study centre, OLV Hospital in Aalst. All participants gave their written informed consent for their data to be used for research purposes.

Variables and outcomes

All variables registered in both structured and unstructured free texts by the treating physicians in the EMRs during routine clinical practice were used for analysis.

For all baseline characteristics, the latest value prior to the cohort description date was considered, with the exception for blood pressure, where the average of the three last values is used. Baseline characteristics were summarized for all HF patients with an index event.

All-cause mortality and HF readmission rates were calculated as event rates per 100 patient years with a 1 and 2 year follow-up, for all HF patients who were still alive on 1 July 2018. The number of total events was calculated as the sum of all events. Both outcomes were also coded as time-to-first-event variables. All-cause mortality rates were retrieved from the Belgium national registry and included hospital mortality, mortality at home, and mortality in retirement homes.

Eligibility of patients to start dapagliflozin treatment was determined taking the European Medicines Agency (EMA) indication into account, that is, patients with symptomatic chronic HFrEF [left ventricular ejection fraction (LVEF) ≤ 40%] and/or T2D and/or CKD, with type 1 diabetes (T1D; as a contraindication).²¹ Eligibility of patients to start empagliflozin treatment was according to the EMA indication: adult patients with symptomatic chronic HF and/or T2D with no diagnosis of T1D.²²

Statistical analysis

The entire population of patients with HF (overall HF) was analysed and stratified according to LVEF into those with HF with mildly reduced EF (HFmrEF) and those with HFrEF.

The Cox proportional hazard model was used to examine the predictive role of patients' baseline characteristics for the outcomes of interest. The following parameters were included in a univariate model: gender, age, body mass index (BMI), T2D, atrial fibrillation (AF) history, coronary artery disease (CAD), former coronary revascularization, valve surgery in the medical history, CKD class ≥ 3, the log10 transformed N-terminal pro-hormone brain natriuretic peptide (NT-proBNP) serum level, presence of chronic obstructive pulmonary disease (COPD), and history of cerebrovascular accident (CVA; stroke). Univariate parameter estimates with a P-value < 0.1 significance level were considered further in a multivariate model. Backward selection procedure was used for building the final multivariate model. Proportional hazard assumption was assessed by Schoenfeld's residuals.

Categorical variables were expressed as numbers and percentages and compared using χ² test with Yates' continuity correction. Continuous variables were expressed as means [± standard deviation (SD), in case of normal distribution] or median (minimum and maximum, in case of skewed distribution). Normality of the distributions was determined by graphical assessment. Continuous variables were compared using the Mann–Whitney U (Wilcoxon's rank-sum) test. To assess the relationship of T2D and CKD with outcomes of interest, cumulative Kaplan–Meier survival curves were generated, and differences between curves were assessed by log-rank test. A composite of the two event types (mortality and HF hospitalization) was created by the sum of all individual events in the given period. The total number of patients at risk in the period of interest was used as the denominator for the calculation of the crude event rates. The crude events rates by patient year were subsequently scaled to 100 patient years. Time-to-event analysis was performed, and event rates per 100 patient years were calculated for all-cause mortality and the composite of mortality and HF readmission.

A two-tailed P-value < 0.05 was considered statistically significant. All statistical analyses were performed in the R environment software Version 3.6.2 (R Foundation for Statistical Computing).

Results

Baseline characteristics of the study population

The CORDIS-HF study included 1333 HF patients with LVEF < 50%. Baseline characteristics of the HF population in terms of demography, clinical characteristics, aetiologies, and comorbidities stratified by LVEF are presented in Table 1. Patients had a mean (SD) age of 74.7 (12.3) years and a median (min, max) BMI of 26 (13, 51) kg/m². Over two-thirds (69%; n = 859) were male, and most patients (81%; n = 559) were classified in New York Heart Association (NYHA) classes II and III with median (min, max) NT-proBNP levels of 2410 (5, 35 000) pg/mL. More than half (57%; n = 756) had CKD classes 3–5, whereas 37% (n = 494) and 24% (n = 321) presented T2D and COPD, respectively.

Table 1 Baseline characteristics of patients with heart failure and left ventricular ejection fraction < 50% stratified by ejection fraction

	Total LVEF < 50%	HFmrEF	HFrEF
Number of patients	1333	413	920
Age in years, mean (SD)	74.7 (12.3)	76.7 (11.6)	73.8 (12.4)*
Male gender (%)	859 (69%)	231 (58%)	628 (75%)*
SBP, mmHg, mean (SD)	126 (23.4)	133 (24.0)	123 (22.6)*
BMI, kg/m2, median [min, max]	26.0 [13.0, 51.0]	26.0 [15.0, 51.0]	26.0 [13.0, 47.0]*
LVEF, %, mean (SD)	35.5 (10.4)	46.9 (3.03)	30.2 (8.13)*
NYHA I	72 (10%)	19 (11%)	53 (10%)
NYHA II	296 (43%)	72 (40%)	224 (44%)
NYHA III	263 (38%)	70 (39%)	193 (38%)
NYHA IV	58 (8%)	19 (11%)	39 (8%)
NT-proBNP, pg/mL, median [min, max]	2410 [5.00, 35 000]	1920 [38.0, 35 000]	2720 [5.00, 35 000]*
CAD, n (%)	857 (64%)	244 (59%)	613 (67%)*
Myocardial infarction, n (%)	584 (44%)	146 (35%)	438 (48%)*
Coronary revascularization (CABG + PCI), n (%)	689 (52%)	197 (48%)	492 (53%)
Unstable angina, n (%)	6 (0%)	3 (1%)	3 (0%)
Angina pectoris, n (%)	436 (33%)	158 (38%)	278 (30%)*
Stroke, n (%)	239 (18%)	68 (16%)	171 (19%)
Atrial fibrillation history, n (%)	967 (73%)	338 (82%)	629 (68%)*
Peripheral artery disease, n (%)	6 (0%)	3 (1%)	3 (0%)
Type 2 diabetes, n (%)	494 (37%)	150 (36%)	344 (37%)
COPD, n (%)	321 (24%)	103 (25%)	218 (24%)
eGFR, mean (SD)	52.2 (23.0)	54.1 (22.3)	51.4 (23.3)*
CKD class (eGFR in mL/min/1.73 m2), n (%)
CKD3 (eGFR 30–59)	513 (38%)	172 (42%)	341 (37%)
CKD4 (eGFR 15–29)	208 (16%)	50 (12%)	158 (17%)
CKD5 (eGFR < 15)	35 (3%)	10 (2%)	25 (3%)
Hypertension, n (%)	775 (58%)	288 (70%)	487 (53%)*
Hypotension, n (%)	518 (39%)	140 (34%)	378 (41%)*
HF drug treatment, n (%)
ACE-I	854 (64%)	245 (59%)	609 (66%)*
ARB	45 (3%)	17 (4%)	28 (3%)
Beta-blocker	1120 (84%)	338 (82%)	782 (85%)
MRA	1203 (90%)	372 (90%)	831 (90%)
ARNi	126 (9%)	11 (3%)	115 (12%)*
SGLT2i	21 (2%)	3 (1%)	18 (2%)
Other (HF) treatments, n (%)
Loop diuretics	772 (58%)	246 (60%)	526 (57%)
Digoxin	343 (26%)	101 (24%)	242 (26%)
Device therapy (ICD + CRT)	568 (43%)	100 (24%)	468 (51%)
Nitrates	42 (3%)	10 (2%)	32 (3%)
Warfarin	365 (27%)	118 (29%)	247 (27%)
NOACs	641 (48%)	220 (53%)	421 (46%)*
Receptor P2Y12 antagonists	376 (28%)	113 (27%)	263 (29%)
Metformin	298 (22%)	96 (23%)	202 (22%)
Iron	32 (2%)	5 (1%)	27 (3%)
Corticosteroids	238 (18%)	83 (20%)	155 (17%)
Insulin	198 (15%)	54 (13%)	144 (16%)
GLP-1 RAs	94 (7%)	26 (6%)	68 (7%)
Statins	603 (45%)	182 (44%)	421 (46%)

A high use of guideline-directed medical therapy (GDMT) was observed. Renin–angiotensin–aldosterone system (RAAS) was blocked in 76% of the total population by angiotensin-converting enzyme inhibitors (ACE-Is) (64%; n = 854), angiotensin II receptor blockers (ARBs) (3%; n = 45), or angiotensin receptor-neprilysin inhibitors (ARNis) (9%; n = 126). Furthermore, beta-blockers and mineralocorticoid receptor antagonists (MRAs) were used in 84% (n = 1120) and 90% (n = 1203) of the patients, respectively. The use of the newer HF medication, that is, ARNis and SGLT2is, was limited to 9% (n = 126) and 2% (n = 21) of the patients, respectively, as data collection (2014–18) was done partially before the reimbursement of ARNis in Belgium (November 2016)²³ and before the European registration of SGLT2is for treating HFrEF (November 2020).^14,15

HF with mildly reduced vs. reduced ejection fraction

According to the European Society of Cardiology (ESC) guidelines,² the 1333 included HF patients (with LVEF < 50%) were divided in HFmrEF (LVEF 40–50%; n = 413) and HFrEF (LVEF < 40%; n = 920). HFrEF patients were slightly younger [mean (SD): 73.8 (12.4) vs. 76.7 (11.6) years, P < 0.05] and included a higher percentage of males (75% vs. 58%, P < 0.05) than those with HFmrEF (Table 1). They also had lower mean systolic blood pressure [SBP; mean (SD): 123 (22.6) vs. 133 (24.0) mmHg, P < 0.05], higher median NT-proBNP levels (2720 vs. 1920 pg/mL, P < 0.05), and lower mean eGFR [mean (SD): 51.4 (23.3) vs. 54.1 (22.3) mL/min/1.73 m², P < 0.05]. Comorbidities were different between HFrEF and HFmrEF patients with higher incidence of CAD (67% vs. 59%, P < 0.05) and myocardial infarction (MI) (48% vs. 35%, P < 0.05). In contrast, lower incidences of AF (68% vs. 82%, P < 0.05) and hypertension (53% vs. 70%, P < 0.05) were observed in HFrEF than in HFmrEF patients. The incidence of T2D, CKD, and COPD was similar in both HF populations, whereas small differences in GDMT were observed with relatively more HFrEF patients treated with ACE-Is (66% vs. 59%, P < 0.05) and ARNis (12% vs. 3%, P < 0.05) than the HFmrEF patients. Administration of other treatments was similar in both populations (Table 1).

Mortality and total HF readmission event rates

Despite optimal GDMT, high event rates of 13.7 and 8.4/100 patient years for the composite endpoint of year mortality and HF readmission were observed at 1 and 2 year follow-up (Figure 2). The excess in mortality contributed more than HF readmissions to this high event rate observed during follow-up. Lower overall event rates (including composite rates) were observed in HFmrEF patients, with mortality as the major driver of the event rates (Figure 2). Similar trends were observed for HFrEF patients. So, irrespective of the LVEF and despite proper treatment and follow-up, HF patients still experience substantial mortality and HF readmission event rates, highlighting the need for new therapeutic strategies.

[IMAGE OMITTED. SEE PDF]

Comorbidities as drivers of mortality and HF readmissions

Using the Cox proportional hazard model for univariate and multivariate analyses, several risk factors were identified as drivers for mortality and HF readmission in HF patients (Table 2). Of note, the univariate analysis identified CKD and T2D, the two other components of the cardio–renal–metabolic syndrome (CRMS) as significant drivers for mortality alone [CKD: hazard ratio (HR) = 2.05, P < 0.001; T2D: HR = 1.49, P < 0.01] and for the composite of mortality and HF readmission (CKD: HR = 1.84, P < 0.001; T2D: HR = 1.67, P > 0.001).

Table 2 Calculated hazard ratios for univariate and multivariate Cox proportional hazard models for events of mortality and the composite of mortality and heart failure readmission

Parameter	Mortality	Composite endpoint of mortality and HF readmission
Univariate	Multivariate	Univariate	Multivariate
Gender male	1.81**	1.83**	1.53**	1.43*
Age	1.04***	1.05***	1.03***	1.03***
BMI	0.95**	0.96*	0.97*	—
Type 2 diabetes	1.49**	1.49*	1.67***	1.46**
AF history	2.03***	—	1.83***	1.60*
CAD	2.00***	1.57*	1.81***	1.74***
Coronary revascularization	1.61**	—	1.55***	—
Valve surgery	1.11	—	1.02	—
CKD (≥CKD3)	2.05***	—	1.84***	—
NT-proBNP (log10)	1.75***	—	1.57***	—
COPD	1.49*	1.42*	1.39*	1.33*
CVA	1.58**	—	1.60**	1.38*

The multivariate analysis revealed male gender (HR mortality = 1.83, P < 0.01; HR composite = 1.43, P < 0.05), higher age (HR mortality = 1.05; HR composite = 1.03, both P < 0.001), and the presence of T2D (HR mortality = 1.49, P < 0.05; HR composite = 1.46, P < 0.01), CAD (HR mortality = 1.57, P < 0.05; HR composite = 1.74, P < 0.001), or COPD (HR mortality = 1.42; HR composite = 1.33, both P < 0.05) as significant drivers for mortality alone and for the composite endpoint of mortality and HF readmissions (Table 2).

To further explore the impact of the renal–metabolic comorbidities on HF, Kaplan–Meier survival curves for HF patients stratified by T2D or CKD were generated (Figure 3). They demonstrated higher mortality rates for HF patients presenting T2D (P < 0.01) or CKD (P < 0.0001) than the patients without T2D or CKD (Figure 3A,C). Similarly, the rates of the composite endpoint of mortality and HF readmission were also significantly higher in HF patients presenting T2D or CKD (both P < 0.0001) vs. non-T2D or non-CKD patients (Figure 3B,D). All curves were diverging fast, with a negative impact of the renal–metabolic comorbidities on HF outcome occurring already after 2–4 months.

[IMAGE OMITTED. SEE PDF]

Eligibility for sodium–glucose cotransporter 2 inhibitors

The eligibility for dapagliflozin or empagliflozin was determined taking the EMA indication label into account (see Methods).²¹ For dapagliflozin eligibility, from 1333 total patients, those with HFrEF were 872, those with T2D (within the HFmrEF group) were 150, and those with CKD (within the HFmrEF group) were 131, resulting in 1153 (86.5%) eligible patients. Regarding empagliflozin, from a total of 1333 patients, those with chronic adult HF were 811 and those with T2D were 494, so that in total, 1305 (97.9%) were eligible.

Discussion

In this real-world population cohort of HFrEF patients, mortality and HF readmissions remained high despite the use of GDMT. Importantly, we have shown that the presence of CKD and T2D each significantly aggravated the mortality and HF readmission event rates, highlighting the important role of CRMS in HF mortality. Therefore, a more holistic/multidisciplinary and comprehensive approach is needed to treat HF, focusing on its comorbidities. Furthermore, we reported that, in this population, up to 86.5% (n = 1153) and 97.9% (n = 1305) qualify for treatment with dapagliflozin and empagliflozin, respectively. In fact, dapagliflozin has already been approved for the three intertwined diseases: HFrEF, T2D, and CKD. Herewith, we demonstrated in a real-world population that SGLT2is might be a promising complementary therapy in the armamentarium of the HF cardiologist, which is in line with the CV outcome trials.

The cardio–renal–metabolic syndrome

HF is associated with a high number of comorbidities that negatively impact the patient's outcome.^2,19,24 In this view, previous registries of patients hospitalized for HF have reported up to 45% and 53% of patients having diabetes or CKD, respectively,¹⁹ whereas up to 16% of patients with HF had both T2D and CKD.²⁵ This is aligned with the present CORDIS-HF registry showing that, in a real-world cohort, 57% and 37% of patients present CKD stages 3–5 and T2D, respectively. This cardio–renal–metabolic link was observed in both HFrEF and HFmrEF patients, underlining the similarities between both HF populations. The interaction between HF, T2D, and CKD remains a matter of debate. Their contributions to the reported excess mortality have been attributed to common CV and metabolic risk factors, such as dyslipidaemia and obesity, which exert profound downstream effects on inflammation, oxidative stress, and neurohumoral pathways and play a role in the development of the CRMS.^3,4

The impact of the cardio–renal–metabolic syndrome on clinical outcomes

Not only the presence but also the type of the underlying comorbidity determines outcomes in HF patients.²⁶ Using a univariate Cox proportional hazard analysis, we identified renal (CKD), metabolic (BMI and T2D), and CV (AF, CAD, and CVA) conditions, as well as COPD, as important predictors for HF hospitalization and mortality. This was reinforced by the Kaplan–Meier survival analysis, demonstrating an increased and additive risk for both HF hospitalization and mortality for HF patients with either T2D or CKD. The observation that patients with CRMS have a higher risk of mortality corroborates previous observations, showing that patients with HFrEF and T2D or CKD have a worse outcome and more severe symptoms than those with low EF and no T2D or CKD.^8,13,16,18 Interestingly, we noticed a slightly higher event rate of the composite endpoint than that of recent randomized control trials (RCTs) on HFrEF, such as the DAPA-HF (dapagliflozin),¹⁷ Prospective Comparison of ARNi with ACE-I to Determine Impact on Global Mortality and Morbidity in Heart Failure (PARADIGM-HF) (sacubitril/valsartan),²⁷ and Empagliflozin Outcome Trial in Patients With Chronic Heart Failure With Reduced Ejection Fraction (EMPEROR-Reduced) (empagliflozin).¹⁵ Several reasons may account for this observation. First, the age of the HF patients in our real-world study was higher than the average age of those included in most RCTs. Second, we noticed higher NT-proBNP levels than the RCTs, indicative of patients in worse condition. Finally, the follow-up and standard-of-care treatments of patients in RCTs might be different; for example, in the PARADIGM-HF trial, a run-in period was defined during which the patients who could not tolerate sacubitril/valsartan were excluded, possibly creating an important selection bias.²⁷ These observations stress the need for clinical trials in elderly and worse-off patients.^27,28 They also demonstrate the important limitations of RCTs and the value of also studying the impact of novel therapies in real-world HF cohorts.

Finally, our data extend previous work⁴ by demonstrating that T2D and CKD are also important predictors of mortality in HFmrEF patients. The negative impact of renal–metabolic comorbidities occurred early (after 2–4 months), which underlines the fast and devastating effect of these comorbidities in patients suffering from HF and urges for therapies with immediate impact. The fact that we observed such an early impact of renal–metabolic comorbidities on survival might partially explain why, in all outcome trials of SGLT2is in HF patients,^9,11,13,17 the beneficial effect of adding an SGLT2i to standard of care occurs very early. This early effect of SGLT2is, reflected in the early deviation of Kaplan–Meier curves, could be explained by their holistic targeting of the complete cardio–renal–metabolic triangle.

High usage of guideline-directed medical therapy, but opportunities to implement newer drug classes

The high mortality and HF events were not caused by underusage of GDMT, because 76% of patients received RAAS inhibition (84% and 90% treated with beta-blockers and MRAs, respectively). Of note, the use of neurohumoral medications was similar to what has been reported in RCTs and HF cohorts, with the exception of MRAs that were used in a higher percentage of patients in our population. This discrepancy in MRA treatment has been reported previously and has been attributed to a lower use of this class of medication in North America than in the other regions of the world.²⁹

Moreover, the recent Safety, Tolerability and Efficacy of Rapid Optimization, Helped by NT-proBNP Testing, of Heart Failure Therapies (STRONG-HF) trial reported a similar high percentage of MRA use (95%) in patients admitted with acute HF.³⁰ These data resonate with our findings reporting a high proportion of patients on MRA, which may be explained by the more intensive treatment strategy deployed in our setting, that is, an advanced HF clinic with more frequent follow-up visits, closer follow-up of GDMT, and use of potassium binders to treat MRA-related hyperkalaemia. The use of SGLT2is was low in this population, which was related to the timing of data collection, as the therapy was not reimbursed. Baseline characteristics were extracted from patient files when SGLT2is were only indicated for T2D treatment, and no scientific evidence existed yet for their use in HFrEF. Since then, RCTs demonstrated that SGLT2is reduce the incidence of CV outcomes in T2D patients, as well as the incidence of CV mortality and hospitalization for HF in HFrEF patients.^9–17,27,31 Moreover, dapagliflozin administration also reduced the risk of CV mortality in HFrEF patients, as well as of renal events and CV-related death in patients with CKD, regardless of their T2D status.^10,13,16,18 Remarkably, SGLT2i is the first drug class to improve CV outcomes in HF with preserved EF (HFpEF). Current treatment strategies in HFpEF are limited to symptom control rather than morbidity or mortality benefit. However, the exciting results from the EMPEROR-Preserved and Dapagliflozin Evaluation to Improve the Lives of Patients with Preserved Ejection Fraction Heart Failure (DELIVER) studies show a significant reduction in CV death or hospitalization for HF.^15,32 These results indicate a promising future for SGLT2is in being the first disease-modifying treatment in HFpEF.

Sodium–glucose cotransporter 2 inhibitors eligibility in the study patient population

In our study, 86.5% (n = 1153) and 97.9% (n = 1305) of patients were candidates for the initiation of dapagliflozin and empagliflozin, respectively, when applying the EMA label criteria.^21,22 The slight difference in eligibility between the two SGLT2is is due to the fact that the positive DELIVER trial results,³² including HF patients with LVEF > 40%, have not yet been included in the label of dapagliflozin by EMA, resulting in a lower eligibility for dapagliflozin in the HFmrEF group. A recent analytical model reported similar percentages of SGLT2i-eligible patients and estimated that its optimal implementation across the United States may prevent or postpone 34 125 deaths yearly.³³ Data from the Get With The Guidelines (GWTG)-HF registry using the Food and Drug Administration (FDA) label (eGFR > 30 mL/min/1.73 m², LVEF ≤ 40%, no dialysis and T1D) suggested an eligibility of 81% of patients being candidates for initiation of dapagliflozin.^34,35 Vaduganathan and colleagues estimated an eligibility of 44.1% in GWTG-HF patients when strict European dapagliflozin criteria were applied.³⁵ Although different percentages are reported, all these studies demonstrate that the dapagliflozin eligibility is much higher than prior estimates of eligibility for ivabradine (14%)³⁶ and sacubitril/valsartan (38%),^37,38 suggesting a broader applicability of these drugs in the real-world HF population. The high percentage of eligible candidates indicates that SGLT2is might be a promising complementary new drug in patients with HFrEF with and without CRMS.

Limitations and strengths

This was a single-centre retrospective study, and the applicability of our results to other HF clinics may not be guaranteed. Regarding CKD, these patients were considered as those with CKD stages 3–5 [according to the Kidney Disease: Improving Global Outcomes (KDIGO) guidelines], because the lower severity of CKD stage 1 or 2 limits the amount of evidence needed to correctly identify these patients. Despite all the listed limitations, the present study offers invaluable real-world evidence on outcomes in patients with HF and LVEF < 50%, including a high number of analysed patients and a detailed analysis of their characteristics. The strength of this study consists of the analysis of a real-world cohort of HF patients. Observational data collected from contemporary, real-world, routine clinical practice settings at all health care levels are of increasing importance, given that HF management is rapidly changing due to paradigm-shifting trials and updated guidelines. Hence, ‘real-time’ understanding of the characteristics of HF patients, as well as their burden and treatment, in routine real-world clinical practice is warranted to understand unmet clinical needs and the current implementation of new guidelines. In this regard, our study displays a truer comorbidity pattern of patients in need of intensified prevention. Furthermore, it illustrates more realistic patterns and event rates resulting from HF than does the clinical trial setting.^39,40 This additional value of using real-world data has been confirmed in a recent paper showing that overly stringent RCT exclusion criteria do not appropriately account for the heterogeneity in real-life populations.⁴⁰ There is a clear need to integrate real-world evidence to supplement the learnings from RCTs to optimize evidence-based medicine assessments in evaluating the effectiveness of treatment decisions.⁴⁰

Conclusions and clinical implications

The high prevalence of T2D and CKD in this real-world HF (LVEF < 50%) patient cohort and the strong correlation with mortality and HF readmission events corroborate the importance of improved treatment of renal–metabolic diseases in HF patients. The substantial mortality and high HF readmission event rates despite GDMT identify a subgroup of HF patients with LVEF < 50% suffering from CRMS who need more intensive treatment options and follow-up. As we have shown a high proportion of HF patients eligible for dapagliflozin and empagliflozin, SGLT2is that already demonstrated added value in T2D, HF, and CKD patients, it suggests that many HF patients with residual risk would benefit from the effective and appropriate implementation of an SGLT2i.

Acknowledgements

The authors further thank Annelies Vankeirsbilck, Eef Vandendriessche, Georges El Azzi, Georges De Feu, Dries Hens, Stef Verbrigghe, Roeland Van Kerckhoven, and Frank Staelens (Director Process and Quality, OLV Aalst, Belgium) for providing the data. Medical writing support, and publication co-ordination and editorial support were performed by Modis (Irena Zurnic Bönisch and Sophie Timmery, respectively) and funded by AstraZeneca BeLux. Further medical writing support from Clara L. Oeste is also acknowledged.

Conflict of interest

W.H. received support from AstraZeneca for congress attendance and participation to data safety monitoring/advisory board, outside of the submitted work. M.V. received support from Pfizer and Boehringer Ingelheim for participation to advisory boards, outside of the submitted work. Z.V., M.M., I.F., and M.D.B. are employees of AstraZeneca BeLux. M.B., I.M., S.V., R.D., and J.B. have nothing to disclose.

Funding

The building up of CORDIS-HF database and this study, as well as the cost related to this publication development including writing support, were funded by AstraZeneca.

References

European Heart Network. Heart failure and cardiovascular diseases—a European Heart Network paper. 2019. Available from: https://ehnheart.org/publications‐and‐papers/publications/1202:heart‐failure‐and‐cardiovascular‐diseases.html. Accessed April 2019

McDonagh TA, Metra M, Adamo M, Gardner RS, Baumbach A, Bohm M, Burri H, Butler J, Celutkiene J, Chioncel O, Cleland JGF, Coats AJS, Crespo‐Leiro MG, Farmakis D, Gilard M, Heymans S, Hoes AW, Jaarsma T, Jankowska EA, Lainscak M, Lam CSP, Lyon AR, McMurray JJV, Mebazaa A, Mindham R, Muneretto C, Francesco Piepoli M, Price S, Rosano GMC, Ruschitzka F, Kathrine Skibelund A, Group ESCSD. 2021 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur Heart J. 2021; 42: 3599–3726.

Whaley‐Connell A, Sowers JR. Basic science: pathophysiology: the cardiorenal metabolic syndrome. J Am Soc Hypertens. 2014; 8: 604–606.

Valensi P, Prévost G, Pinto S, Halimi J‐M, Donal E. The impact of diabetes on heart failure development: the cardio‐renal‐metabolic connection. Diabetes Res Clin Pract. 2021; 175: [eLocator: 108831].

Kenny HC, Abel ED. Heart failure in type 2 diabetes mellitus: impact of glucose‐lowering agents, heart failure therapies, and novel therapeutic strategies. Circ Res. 2019; 124: 121–141.

Seferovic PM, Petrie MC, Filippatos GS, Anker SD, Rosano G, Bauersachs J, Paulus WJ, Komajda M, Cosentino F, de Boer RA, Farmakis D, Doehner W, Lambrinou E, Lopatin Y, Piepoli MF, Theodorakis MJ, Wiggers H, Lekakis J, Mebazaa A, Mamas MA, Tschope C, Hoes AW, Seferovic JP, Logue J, McDonagh T, Riley JP, Milinkovic I, Polovina M, van Veldhuisen DJ, Lainscak M, Maggioni AP, Ruschitzka F, McMurray JJV. Type 2 diabetes mellitus and heart failure: a position statement from the Heart Failure Association of the European Society of Cardiology. Eur J Heart Fail. 2018; 20: 853–872.

Nayor M, Larson MG, Wang N, Santhanakrishnan R, Lee DS, Tsao CW, Cheng S, Benjamin EJ, Vasan RS, Levy D, Fox CS, Ho JE. The association of chronic kidney disease and microalbuminuria with heart failure with preserved vs. reduced ejection fraction. Eur J Heart Fail. 2017; 19: 615–623.

Birkeland KI, Bodegard J, Eriksson JW, Norhammar A, Haller H, Linssen GCM, Banerjee A, Thuresson M, Okami S, Garal‐Pantaler E, Overbeek J, Mamza JB, Zhang R, Yajima T, Komuro I, Kadowaki T. Heart failure and chronic kidney disease manifestation and mortality risk associations in type 2 diabetes: a large multinational cohort study. Diabetes Obes Metab. 2020; 22: 1607–1618.

Neal B, Perkovic V, Mahaffey KW, de Zeeuw D, Fulcher G, Erondu N, Shaw W, Law G, Desai M, Matthews DR, Group CPC. Canagliflozin and cardiovascular and renal events in type 2 diabetes. N Engl J Med. 2017; 377: 644–657.

Wiviott SD, Raz I, Bonaca MP, Mosenzon O, Kato ET, Cahn A, Silverman MG, Zelniker TA, Kuder JF, Murphy SA, Bhatt DL, Leiter LA, McGuire DK, Wilding JPH, Ruff CT, Gause‐Nilsson IAM, Fredriksson M, Johansson PA, Langkilde AM, Sabatine MS, Investigators D‐T. Dapagliflozin and cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2019; 380: 347–357.

Zinman B, Wanner C, Lachin JM, Fitchett D, Bluhmki E, Hantel S, Mattheus M, Devins T, Johansen OE, Woerle HJ, Broedl UC, Inzucchi SE. Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes. N Engl J Med. 2015; 373: 2117–2128.

Anker SD, Butler J, Filippatos G, Khan MS, Marx N, Lam CSP, Schnaidt S, Ofstad AP, Brueckmann M, Jamal W, Bocchi EA, Ponikowski P, Perrone SV, Januzzi JL, Verma S, Bohm M, Ferreira JP, Pocock SJ, Zannad F, Packer M. Effect of empagliflozin on cardiovascular and renal outcomes in patients with heart failure by baseline diabetes status: results from the EMPEROR‐Reduced trial. Circulation. 2021; 143: 337–349.

Jhund PS, Solomon SD, Docherty KF, Heerspink HJL, Anand IS, Böhm M, Chopra V, de Boer RA, Desai AS, Ge J, Kitakaze M, Merkley B, O'Meara E, Shou M, Tereshchenko S, Verma S, Vinh PN, Inzucchi SE, Køber L, Kosiborod MN, Martinez FA, Ponikowski P, Sabatine MS, Bengtsson O, Langkilde AM, Sjöstrand M, McMurray JJV. Efficacy of dapagliflozin on renal function and outcomes in patients with heart failure with reduced ejection fraction: results of DAPA‐HF. Circulation. 2021; 143: 298–309.

McMurray JJV, Solomon SD, Inzucchi SE, Køber L, Kosiborod MN, Martinez FA, Ponikowski P, Sabatine MS, Anand IS, Bělohlávek J, Böhm M, Chiang C‐E, Chopra VK, de Boer RA, Desai AS, Diez M, Drozdz J, Dukát A, Ge J, Howlett JG, Katova T, Kitakaze M, Ljungman CEA, Merkely B, Nicolau JC, O'Meara E, Petrie MC, Vinh PN, Schou M, Tereshchenko S, Verma S, Held C, DeMets DL, Docherty KF, Jhund PS, Bengtsson O, Sjöstrand M, Langkilde A‐M. Dapagliflozin in patients with heart failure and reduced ejection fraction. N Engl J Med. 2019; 381: 1995–2008.

Packer M, Anker SD, Butler J, Filippatos G, Pocock SJ, Carson P, Januzzi J, Verma S, Tsutsui H, Brueckmann M, Jamal W, Kimura K, Schnee J, Zeller C, Cotton D, Bocchi E, Böhm M, Choi D‐J, Chopra V, Chuquiure E, Giannetti N, Janssens S, Zhang J, Gonzalez Juanatey JR, Kaul S, Brunner‐La Rocca H‐P, Merkely B, Nicholls SJ, Perrone S, Pina I, Ponikowski P, Sattar N, Senni M, Seronde M‐F, Spinar J, Squire I, Taddei S, Wanner C, Zannad F. Cardiovascular and renal outcomes with empagliflozin in heart failure. N Engl J Med. 2020; 383: 1413–1424.

Heerspink HJL, Stefansson BV, Correa‐Rotter R, Chertow GM, Greene T, Hou FF, Mann JFE, McMurray JJV, Lindberg M, Rossing P, Sjostrom CD, Toto RD, Langkilde AM, Wheeler DC, Committees D‐CT, Investigators. Dapagliflozin in patients with chronic kidney disease. N Engl J Med. 2020; 383: 1436–1446.

Jhund PS, Ponikowski P, Docherty KF, Gasparyan SB, Bohm M, Chiang CE, Desai AS, Howlett J, Kitakaze M, Petrie MC, Verma S, Bengtsson O, Langkilde AM, Sjostrand M, Inzucchi SE, Kober L, Kosiborod MN, Martinez FA, Sabatine MS, Solomon SD, McMurray JJV. Dapagliflozin and recurrent heart failure hospitalizations in heart failure with reduced ejection fraction: an analysis of DAPA‐HF. Circulation. 2021; 143: 1962–1972.

Petrie MC, Verma S, Docherty KF, Inzucchi SE, Anand I, Bělohlávek J, Böhm M, Chiang C, Chopra VK, de Boer RA, Desai AS, Diez M, Drozdz J, Dukát A, Ge J, Howlett J, Katova T, Kitakaze M, Ljungman CEA, Merkely B, Nicolau JC, O'Meara E, Vinh PN, Schou M, Tereshchenko S, Køber L, Kosiborod MN, Langkilde AM, Martinez FA, Ponikowski P, Sabatine MS, Sjöstrand M, Solomon SD, Johanson P, Greasley PJ, Boulton D, Bengtsson O, Jhund PS, McMurray JJV. Effect of dapagliflozin on worsening heart failure and cardiovascular death in patients with heart failure with and without diabetes. JAMA. 2020; 323: 1353–1368.

Ambrosy AP, Fonarow GC, Butler J, Chioncel O, Greene SJ, Vaduganathan M, Nodari S, Lam CSP, Sato N, Shah AN, Gheorghiade M. The global health and economic burden of hospitalizations for heart failure: lessons learned from hospitalized heart failure registries. J Am Coll Cardiol. 2014; 63: 1123–1133.

Van de Meulebroucke C, Beckers J, Corten K. What can we expect following anterior total hip arthroplasty on a regular operating table? A validation study of an artificial intelligence algorithm to monitor adverse events in a high‐volume, nonacademic setting. J Arthroplasty. 2019; 34: 2260–2266.

European Medicines Agency. Forxiga (dapagliflozin) 2012 [updated 5 January 2022]. Available from: https://www.ema.europa.eu/en/medicines/human/EPAR/forxiga. Accessed March 6, 2023

European Medicines Agency. Jardiance (empagliflozin) 2014 [updated 6 September 2022]. Available from: https://www.ema.europa.eu/en/medicines/human/EPAR/jardiance. Accessed March 30, 2023

Commissie voor de bescherming van de persoonlijke levenssfeer (CBPL). Beraadslaging nr. 18/022 van 20 februari 2018 met betrekking tot de verwerking van gecodeerde persoonsgegevens die de gezondheid betreffen, via het platform healthdata.be, in het kader van een procedure artikel 81 voor de terugbetaling van het geneesmiddel entresto (Novartis pharma). 2018.

van Deursen VM, Urso R, Laroche C, Damman K, Dahlstrom U, Tavazzi L, Maggioni AP, Voors AA. Co‐morbidities in patients with heart failure: an analysis of the European Heart Failure Pilot Survey. Eur J Heart Fail. 2014; 16: 103–111.

Lawson CA, Seidu S, Zaccardi F, McCann G, Kadam UT, Davies MJ, Lam CSP, Heerspink HL, Khunti K. Outcome trends in people with heart failure, type 2 diabetes mellitus and chronic kidney disease in the UK over twenty years. EClinMed. 2021; 32: [eLocator: 100739].

Conrad N, Judge A, Tran J, Mohseni H, Hedgecott D, Crespillo AP, Allison M, Hemingway H, Cleland JG, McMurray JJV, Rahimi K. Temporal trends and patterns in heart failure incidence: a population‐based study of 4 million individuals. Lancet. 2018; 391: 572–580.

McMurray JJV, Packer M, Desai AS, Gong J, Lefkowitz MP, Rizkala AR, Rouleau JL, Shi VC, Solomon SD, Swedberg K, Zile MR, PARADIGM‐HF Investigators and Committees. Angiotensin–neprilysin inhibition versus enalapril in heart failure. N Engl J Med. 2014; 371: 993–1004.

McMurray JJV, DeMets DL, Inzucchi SE, Køber L, Kosiborod MN, Langkilde AM, Martinez FA, Bengtsson O, Ponikowski P, Sabatine MS, Sjöstrand M, Solomon SD, on behalf of the DAPA‐HF Committees and Investigators. The Dapagliflozin and Prevention of Adverse‐outcomes in Heart Failure (DAPA‐HF) trial: baseline characteristics. Eur J Heart Fail. 2019; 21: 1402–1411.

Tromp J, Bamadhaj S, Cleland JGF, Angermann CE, Dahlstrom U, Ouwerkerk W, Tay WT, Dickstein K, Ertl G, Hassanein M, Perrone SV, Ghadanfar M, Schweizer A, Obergfell A, Lam CS, Filippatos G, Collins SP. Correction to: post‐discharge prognosis of patients admitted to hospital for heart failure by world region, and national level of income and income disparity (REPORT‐HF): a cohort study. Lancet Glob Health. 2020; 8: [eLocator: e1001].

Mebazaa A, Davison B, Chioncel O, Cohen‐Solal A, Diaz R, Filippatos G, Metra M, Ponikowski P, Sliwa K, Voors AA, Edwards C, Novosadova M, Takagi K, Damasceno A, Saidu H, Gayat E, Pang PS, Celutkiene J, Cotter G. Safety, tolerability and efficacy of up‐titration of guideline‐directed medical therapies for acute heart failure (STRONG‐HF): a multinational, open‐label, randomised, trial. The Lancet. 2022; 400: 1938–1952.

Zelniker TA, Wiviott SD, Raz I, Im K, Goodrich EL, Bonaca MP, Mosenzon O, Kato ET, Cahn A, Furtado RHM, Bhatt DL, Leiter LA, McGuire DK, Wilding JPH, Sabatine MS. SGLT2 inhibitors for primary and secondary prevention of cardiovascular and renal outcomes in type 2 diabetes: a systematic review and meta‐analysis of cardiovascular outcome trials. Lancet. 2019; 393: 31–39.

Solomon SD, McMurray JJV, Claggett B, de Boer RA, DeMets D, Hernandez AF, Inzucchi SE, Kosiborod MN, Lam CSP, Martinez F, Shah SJ, Desai AS, Jhund PS, Belohlavek J, Chiang C‐E, Borleffs CJW, Comin‐Colet J, Dobreanu D, Drozdz J, Fang JC, Alcocer‐Gamba MA, Al Habeeb W, Han Y, Cabrera Honorio JW, Janssens SP, Katova T, Kitakaze M, Merkely B, O'Meara E, Saraiva JFK, Tereshchenko SN, Thierer J, Vaduganathan M, Vardeny O, Verma S, Pham VN, Wilderäng U, Zaozerska N, Bachus E, Lindholm D, Petersson M, Langkilde AM. Dapagliflozin in heart failure with mildly reduced or preserved ejection fraction. N Engl J Med. 2022; 387: 1089–1098.

Bassi NS, Ziaeian B, Yancy CW, Fonarow GC. Association of optimal implementation of sodium‐glucose cotransporter 2 inhibitor therapy with outcome for patients with heart failure. JAMA Cardiol. 2020; 5: 948–951.

Food and Drug Administration. FORXIGA Full prescribing information: FDA. 2020. Available from: https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/202293s020lbl.pdf. Accessed March 6, 2023

Vaduganathan M, Greene SJ, Zhang S, Grau‐Sepulveda M, DeVore AD, Butler J, Heidenreich PA, Huang JC, Kittleson MM, Joynt Maddox KE, McDermott JJ, Owens AT, Peterson PN, Solomon SD, Vardeny O, Yancy CW, Fonarow GC. Applicability of US Food and Drug Administration labeling for dapagliflozin to patients with heart failure with reduced ejection fraction in US clinical practice: the Get With the Guidelines‐Heart Failure (GWTG‐HF) registry. JAMA Cardiol. 2020; 6: 1–10.

Das D, Savarese G, Dahlstrom U, Fu M, Howlett J, Ezekowitz JA, Lund LH. Ivabradine in heart failure: the representativeness of SHIFT (Systolic Heart Failure Treatment With the IF Inhibitor Ivabradine Trial) in a broad population of patients with chronic heart failure. Circ Heart Fail. 2017; 10: [eLocator: e004112].

Parikh KS, Lippmann SJ, Greiner M, Heidenreich PA, Yancy CW, Fonarow GC, Hernandez AF. Scope of sacubitril/valsartan eligibility after heart failure hospitalization: findings from the GWTG‐HF registry (Get With The Guidelines‐Heart Failure). Circulation. 2017; 135: 2077–2080.

Perez AL, Kittipibul V, Tang WHW, Starling RC. Patients not meeting PARADIGM‐HF enrollment criteria are eligible for sacubitril/valsartan on the basis of FDA approval: the need to close the gap. JACC Heart Fail. 2017; 5: 460–463.

Jang HY, Kim I‐W, Oh JM. Using real‐world data for supporting regulatory decision making: comparison of cardiovascular and safety outcomes of an empagliflozin randomized clinical trial versus real‐world data. Front Pharmacol. 2022; 13.

Tan YY, Papez V, Chang WH, Mueller SH, Denaxas S, Lai AG. Comparing clinical trial population representativeness to real‐world populations: an external validity analysis encompassing 43895 trials and 5685738 individuals across 989 unique drugs and 286 conditions in England. The Lancet Healthy Longevity. 2022; 3: e674–e689.