Progress of patients hospitalized with acute heart failure treated with empagliflozin

A total of 195 patients were admitted with HF between 1 May 2022 and 31 May 2023. Of these, 83 patients were excluded for various reasons. Therefore, the final study population comprised 112 patients (57.4%) who started treatment with empagliflozin before discharge; of these, 97 completed follow-up (15 patients died) (Figure 1). Since only two patients had started dapagliflozin at the end of the recruitment period, they were not included in the final analysis to avoid inconsistencies in the results.

The baseline characteristics of the population are shown in Table 1. Mean age was 85.2 ± 6.5 years, and 67.9% of patients were women. Most patients were in NYHA functional class III (35.7%) or IV (31.3%), 73.2% had HFpEF, and most cases (85.7%) involved an exacerbation in patients previously diagnosed with HF. As for the presence of other comorbidities, 89.3% had arterial hypertension, 51.8% had atrial fibrillation, 40.1% chronic kidney disease, 38.4% diabetes and 10.7% a history of myocardial infarction.

As for treatment of HF before admission, 80.4% were taking a loop diuretic, mainly furosemide, 70.6% were taking a RAAS inhibitor, 49.1% were taking betablockers, and 25% were taking mineralocorticoid receptor antagonists (Table 1).

Median time to starting empagliflozin during admission was 3 days (IQR, 0–269). At admission, 94.6% were taking furosemide (15.2% at high doses [≥60 mg/24 h]; 36.6% at intermediate doses [>40-<60 mg/24 h]). On initiating empagliflozin, 12.5% were taking high doses of furosemide and 26.8% were taking intermediate doses. The dose of furosemide was reduced in 12.4% of patients (Supplementary Table 1).

As for the events recorded, 26.8% of patients (n = 30) had attended the emergency department and 13.4% (n = 15) had died 3 months after discharge. Patients who had started empagliflozin with diuretics at higher doses than those prescribed at onset of the episode that led them to be hospitalized did not progress more poorly, more frequently die, visit the emergency room more often, or experience worse adverse effects during follow-up (Supplementary Table 2).

Of the patients who remained alive, 93.8% were still taking empagliflozin at 3 months of follow-up. The reasons for withdrawal (n = 22; 19.6%) were death (15 patients [13.4%]), discontinuation of therapy for unknown reasons (4 [3.6%]), intolerance (1 [0.9%]), urinary discomfort (1 [0.9%]) and palliative care (1 [0.9%]). Type of acute HF, obesity, hypertension, Type 2 diabetes and edema were included as potential confounding factors in the logistic regression analysis, although no statistically significant findings were recorded.

This real-world study, which involved more than 100 patients admitted to a Spanish internal medicine department with HF, showed that it was feasible to initiate empagliflozin before discharge from hospital and that treatment was maintained by a large number of participants during follow-up.

In the study sample, mean age was high (85 years), most patients were in NYHA functional class III or IV, and up to 75% had HFpEF. Moreover, patients had numerous comorbid conditions. In the RICA registry, which included patients aged ≥50 years admitted with HF to internal medicine departments in Spain between 2008 and 2019, the mean age was 80 years, up to a third of patients were in functional class III, more than 60% had HFpEF, and comorbidity was very common [17]. Various studies have indicated that in clinical practice, a high number of patients with HF would be eligible for treatment with SGLT2i [18,19]. In fact, in a recent study that analyzed the eligibility of the four pharmacologic pillars for treatment of HFrEF at discharge from hospital, SGLT2i were more frequently prescribed than RAAS inhibitors and mineralocorticoid receptor antagonists, with kidney failure being the main factor limiting prescription of these agents, in contrast with SGLT2i [20]. Furthermore, major differences can be observed in the clinical profile and type of HF depending on the specialist treating the patient (e.g., internal medicine vs cardiology). Older age, more comorbid conditions, and a higher frequency of HFpEF also play a role [21]. Additionally, there are relevant differences with clinical trials. For example, whereas in the EMPA-RESPONSE-HF studies the presence of chronic kidney disease with estimated glomerular filtration rate ≤30 ml/min was an exclusion criterion, and EMPULSE included up to 20 ml/min. In our cohort, 7.1% of patients had a history of G4 chronic kidney disease. Moreover, at the time of admission, 19.6 and 4.5% of patients had an estimated glomerular filtration rate <30 and <20 ml/min, respectively. Patients were older, with more comorbidities and a higher Charlson index. Moreover, the present study showed that 73.2% had HFpEF, whereas in the EMPULSE trial only 28.7% had HF with EF >40%, and in EMPA-RESPONSE-HF, the mean left ventricular ejection fraction was 36 ± 17%, indicating that the phenotype of patients hospitalized in the internal medicine department differs from that shown in clinical trials [12,13]. Consequently, the information provided in the present study is relevant, since the sample is representative of patients with acute HF admitted to internal medicine departments in Spain and treated early with an SGLT2i before discharge to ensure they obtain the maximum benefit from treatment, as recommended in clinical practice guidelines [6,10,11].

The present study showed that before admission, 70.6% of patients were taking a RAAS inhibitor, 49.1% a betablocker and 25.0% a mineralocorticoid receptor antagonist. A recent study performed in Spain and including patients with HFpEF up to 2019 found that 79.8% were taking RAAS inhibitors (12% sacubitril–valsartan), 65.9% were taking betablockers, 27.7% mineralocorticoid receptor antagonists and only 5.1% SGLT2i [1]. A study performed in France showed that 73.7% were taking RAAS inhibitors, 80.3% were taking betablockers, and 21.6% mineralocorticoid receptor antagonists [22]. Despite having improved the percentage of patients taking treatment recommended for HF compared with previous studies, there is still considerable room for improvement in the form of a more favorable prognosis and reduced risk of readmission with HF, especially when treatment is started early, including both before and after discharge, which is the most vulnerable period for affected patients [6,23–25]. Furthermore, the EMPULSE study found that the benefit of adding empagliflozin before discharge was independent of the underlying treatment the patient was taking for HF [26]. In any case, SGLT2i are underused in clinical practice, and clinical findings indicate that these drugs should be prescribed more frequently [27,28].

The EMPULSE study [12] randomized patients during admission when they were hemodynamically stable (mean time from admission to randomization, 3 days). In fact, before randomization, patients had to be treated with a minimum single dose of 40 mg of iv. furosemide (or equivalent iv. loop diuretic, defined as 20 mg of torasemide or 1 mg of bumetanide). The EMPA-RESPONSE-AHF trial [13] was a randomized, placebo-controlled, double-blind, parallel group, multicenter pilot study, in which 80 acute HF patients with and without diabetes were randomized to receive either empagliflozin 10 mg/d or placebo for 30 days. Of note, patients had to be treated with loop diuretics at screening. In the present study, prior to admission, 25% of patients were taking thiazide diuretics and 25% mineralocorticoid receptor antagonists that were maintained during admission. At initiation of empagliflozin, 17.9% were taking high-dose furosemide (≥60 mg), 2.7% were taking hydrochlorothiazide, and 5.4% did not receive any diuretic treatment because the drug was started the day before hospital discharge. The median time until initiation of empagliflozin was 3 days, although patients who started earlier did not experience more complications during follow-up. Similarly, the results of the present study revealed no differences in clinical profile, indicating that empagliflozin could be used when the patient's condition stabilizes during admission. As for the dose of furosemide, while not recommended by guidelines [6,10,11], this study revealed that starting empagliflozin with higher doses of furosemide than at admission did not imply a poorer prognosis, possibly because empagliflozin leads to osmotic diuresis that does not interfere with other treatments or involves more clinical risks than usual in a patient admitted with HF [25,26].

With respect to events at 3 months after discharge, 26.8% of patients had attended the emergency department (at 6 months) and 13.4% had died. In the EMPULSE trial, cardiovascular death or worsening of HF occurred in 34 (12.8%) patients in the empagliflozin group and in 49 (18.5%) in the placebo group (HR: 0.69; 95% CI: 0.45–1.08). In the EMPA-RESPONSE-HF trial, death, re-hospitalization for HF, or worsening HF during hospitalization was reported in four patients (10%) in the empagliflozin group and in 13 patients (33%) in the placebo group (p = 0.014). The findings of the present study can be explained by the profile of patients with acute HF admitted to internal medicine, namely, older age with more comorbidities, a higher Barthel index and more complex management requirements than those recruited in the pivotal clinical trials, thus highlighting the importance of the results reported here.

Finally, most patients continued treatment with empagliflozin during the 3-month follow-up. The frequency of adverse events was similar to that described in clinical trials [8,28,29]. These results could prove even more relevant if we consider the older age, comorbid conditions and frailty that characterize the study patients, since suspending treatment that has proven clinically beneficial is associated with a poorer prognosis and greater risk of complications [30,31].

The present study is subject to a series of limitations. Given its retrospective design, the only data collected were those already included in the patients' electronic clinical history, thus favoring the loss of some variables of clinical interest. However, it is very unlikely that visits to the emergency department, death and changes in treatments, including withdrawal, were not detected, since the health system in Andalusia, Spain is completely computerized and information is readily available, including changes in treatment (drug withdrawal and/or dose modification), which are detected quickly and in real time. Furthermore, the absence of a comparison group limits the generalization of the results obtained. Finally, it would be necessary to replicate these results in other populations with similar characteristics in multicenter studies.

Despite this study has relevant limitations (i.e., small sample size and lack of comparative group), our data suggest that in clinical practice it may be feasible to introduce empagliflozin before discharge in patients admitted to internal medicine with HF. The frequency of complications was as expected owing to the highly complex nature of the patients, and treatment was largely maintained during follow-up.

Funded by the Fundación para la investigación biosanitaria de Andalucía Oriental – Alejandro Otero (FIBAO), though an unrestricted grant from Boehringer Ingelheim (BI) – Lilly Alliance. BI was given the opportunity to review the manuscript for medical and scientific accuracy as it relates to Boehringer Ingelheim substances, as well as intellectual property considerations. M Raya-Cruz has given talks, attended conferences and participated in clinical trials sponsored by AstraZeneca, Boehringer Ingelheim, Eli Lilly and Company, Pfizer, Cantabria Labs and Rovi. J Gascón Jurado has attended lectures sponsored by Pfizer. G Olalla de la Torre Peregrín has attended lectures sponsored by Rovi. N Montúfar has attended conferences sponsored by Rovi, Pfizer and Janssen. F Gómez Delgado has given talks, attended conferences and participated in clinical trials sponsored by Ferrer, Esteve, Boehringer Ingelheim, Eli Lilly and Company, Daiichi Sankyo, Novartis, GSK, Pfizer, Janssen-Cilag, Amgen, Viatrix and Sanofi. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Content Ed Net (Madrid, Spain) provided editorial assistance in the preparation of this manuscript, which was funded by the Fundación para la investigación biosanitaria de Andalucía Oriental – Alejandro Otero (FIBAO).

The authors state that they have obtained appropriate institutional review board approval or have followed the principles outlined in the Declaration of Helsinki for all human or animal experimental investigations.

This work is licensed under the Attribution-NonCommercial-NoDerivatives 4.0 Unported License. To view a copy of this license, visit https://creativecommons.org/licenses/by-nc-nd/4.0/