Treatment of Chronic Heart Failure in Advanced Chronic Kidney Disease: The HAKA Multicenter Retrospective Real-World Study Open Access

Patients with chronic kidney disease (CKD) are at high risk for cardiovascular events. The prevalence of heart failure (HF) in patients with CKD reaches up to 40% [1, 2] and the incidence of HF rises even more when estimated glomerular filtration rate (eGFR) decreases [1]. It has been reported that up to two-thirds of patients with an eGFR lower than 30 mL/min/1.73 m² or advanced CKD (aCKD) have HF [3]. Patients with aCKD and HF are at high risk of premature death and hospitalization (HF-specific and all-cause) [1, 4].

Based on left ventricular ejection fraction (EF), the European Society of Cardiology (ESC) guidelines classify HF into three types [5, 6]: HF with reduced EF (HFrEF), HF with mildly reduced EF (HFmrEF), and HF with preserved EF (HFpEF). Well-established ESC guidelines provide treatment recommendations with a very strong level of evidence for HFrEF [5, 6] and weaker for HFmrEF and HFpEF [7].

However, many pivotal clinical trials have excluded HF patients with aCKD, as some HF medications may affect renal hemodynamics and/or be associated with electrolyte disturbances [8]. Therefore, there is a current significant lack of clinical trial-based evidence for the management of HF in patients with aCKD and information derived from real-world evidence is crucial to ensure their appropriate management [8].

In this context, the aim of this study was to investigate the current clinical practice regarding the HF treatment in patients with aCKD who are clinically managed in specialized aCKD units, with particular attention to the implementation of recommendations by the 2016 ESC Guidelines on treatment with foundational drugs (such as renin-angiotensin system inhibitors [RASi], angiotensin receptor blocker/neprilysin inhibitor [ARNI], beta-blockers [BBs], and mineralocorticoid receptor antagonists [MRAs]) in addition to sodium-glucose cotransporter-2 inhibitors [SGLT2i] [5‒7].

The Heart And Kidney Audit (HAKA) is a retrospective observational real-world study, promoted by the Spanish Society of Nephrology (S.E.N.), including outpatients with aCKD and HF cared for in specialized aCKD units between January 1, 2021, and March 31, 2021. We selected these dates to ensure adequate collection and availability of all required information, as well as to assess the adequacy of treatment in accordance with ESC recommendations for managing HFrEF [5, 6]. To have an updated overview of the implementation of the recommended treatment of patients with HFmrEF and HFpEF [7], we also analyzed the basic drugs prescribed to patients with aCKD presenting with these two types of HF.

The main objective was to assess the current treatment of all types of HF in aCKD outpatients, with particular emphasis on treatment with RASi/ARNI, BB, MRA, and SGLT2i. We also evaluated the management of comorbidities in this population.

In December 2022, the S.E.N. invited all centers in Spain with a specialized aCKD unit to participate in this study. Inclusion criteria were an age ≥18 years, having an eGFR <30 mL/min/1.73 m² calculated by the creatinine-based 2009 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation and a diagnosis of chronic HF, defined by clinical symptoms and signs, elevated natriuretic peptides and echocardiographic findings [5]. Exclusion criteria were patients on dialysis or who required dialysis within the prior 6 months, patients under conservative management (noncandidates for kidney replacement therapy), and those with a history of kidney or heart transplantation.

Using electronic medical records, we collected epidemiological, clinical, and analytical data on demographics, CKD stages 4 and 5, diagnosis of HF types, comorbidities, and pharmacological and non-pharmacological treatments.

CKD G4 or G5 were diagnosed when eGFR values were 15–29 and <15 mL/min/1.73 m², respectively. The diagnosis of HF was made by a combination of clinical criteria (signs and/or symptoms), elevated natriuretic peptides, and echocardiographic findings according to current ESC guidelines [5‒7]. Patients with HF were divided into three groups: HFrEF (EF ≤40%), HFmrEF (EF 41–49%), and HFpEF (EF ≥50%) [5].

We collected history of hypertension, dyslipidemia, diabetes, coronary artery disease, hypertensive heart disease, peripheral vascular disease, atrial fibrillation, and stroke. Body mass index (BMI) was calculated to define obesity (body mass index ≥30 kg/m²). Anemia was defined as hemoglobin <13 g/dL in men and Hb <12 g/dL in women [9]. Absolute iron deficiency was defined according to European Renal Best Practice (ERBP) guidelines as transferrin saturation index ≤20% and serum ferritin concentration ≤100 ng/mL [9]. Relative iron deficiency was defined as transferrin saturation index ≤20% and ferritin >500 ng/mL [9]. Hyperkalemia was considered when serum potassium was >5.5 mEq/L.

Beyond RASi/ARNI, BB, MRA, and SGLT2i, other HF therapies analyzed were furosemide, ivabradine, digoxin, chronic resynchronization therapy, and implantable cardioverter defibrillators. Additional comorbidities-related therapies were also assessed, including statins, proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i), erythropoiesis-stimulating agents (ESAs), iron, and potassium binders. Maximum doses were considered according to technical data sheets and the 2016 and 2021 ESC Guidelines for RASi, ARNI, BB, and MRA [5, 6].

Data are expressed using the mean (standard deviation) or median (interquartile range) depending on the distribution of the variable (tested with the Shapiro-Wilk test). Comparisons between categorical variables were done using the Fisher’s exact test; between categorical and continuous variables using the t or Mann-Whitney tests, and between continuous variables and more than two categories with analysis of variance. Logistic regression was performed for assessing adjusted models. Statistical analysis was performed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA). A p value <0.05 was considered statistically significant.

Twenty-nine centers with specialized aCKD units responded to the call made by the S.E.N. and provided the required information. Among the 5,012 included patients with aCKD, 532 had a medical record of HF (median 13% [range between centers 2–44%]) (Fig. 1). According to the KDIGO classification, 410 (77%) patients had CKD G4 and 122 (23%) CKD G5.

The clinical characteristics of patients grouped by CKD stage are shown in Table 1. Patients with CKD G5 were younger, more frequently men and presented higher prevalence of peripheral arterial disease, hypertensive heart disease, valvular heart disease, atrial fibrillation, and a lower E/e’ ratio (14 ± 6 vs. 12 ± 4, p = 0.006). In addition, anemia and the presence of an arteriovenous fistula were significantly more frequent in patients with CKD G5 than in patients with CKD G4. Patients with CKD G5 required significantly more Cardiology and Nephrology outpatient visits per year than patients with CKD G4.

HF was diagnosed based on cardinal HF symptoms and signs (86%), elevated natriuretic peptide values (50%), and the presence of structural abnormalities on echocardiography (75%). There were some differences in these diagnostic criteria according to HF type (Fig. 2a), but not according to CKD category (Fig. 2b).

The distribution of HF types was as follows: 20% HFrEF, 13% HFmrEF, and 67% HFpEF. The clinical characteristics of the patients grouped according to HF types are shown in Table 2. Significant differences across the types of HF were found in sex, the prevalence of hypertension, coronary artery disease, hypertensive heart disease, anemia, EF, and E/e’ ratio. Annual outpatient visits in Cardiology were significantly more frequent for HFrEF patients than for the other two types of HF (Table 2). However, yearly outpatient visits in nephrology were similar across all subtypes of HF.

Table 3 shows the distribution of HF therapies in the entire group of patients with HF and in patients separated by type of HF, in both cases according to CKD category. In the entire HF population, furosemide was administered to the majority of patients. However, BB was the only foundational drug prescribed to more than half of the population. The low prescription of foundational drugs should be highlighted, especially in patients with CKD G5. Very few patients were treated with other HF drugs and devices, regardless of CKD category.

Table 4 shows that treatment with erythropoietin and potassium binders was significantly more frequent in patients with CKD G5 than in patients with CKD G4. There were no other significant differences in therapies between the two groups.

Among patients with HFrEF, those with CKD G5 received ARNI significantly less frequently than patients with HFrEF and CKD G4. The percentages of patients with HFrEF receiving the remaining pharmacological and non-pharmacological therapies were similar in both CKD categories. Of interest, the number of patients with HFrEF receiving triple (RASi/ARNI, BB, and MRA) or quadruple (RASi/ARNI, BB, MRA, and SGLT2i) for HF was low (35%, Fig. 3a). Indeed, no patients with CKD G5 were receiving all foundational HF drugs (Fig. 3c) compared to 11% of patients with CKD G4 (Fig. 3b). Factors associated to receive triple or quadruple therapy in HFrEF were diabetes (p = 0.048), lower GFR (p = 0.003) and higher annual cardiology outpatient visits (p = 0.015). Adjusted logistic regression demonstrated that GFR was the only associated factor to triple or quadruple therapy prescription (HR 1.13, 95% CI: [1.03–1.24], p = 0.010).

Figure 4 shows that the maximum recommended dose of RASi/ARNI, BB, and MRA [5, 6] was not reached in over half of the patients with HFrEF regardless of CKD stage. SGLT2i are not represented because there is no maximum recommended dose for these compounds in patients with HFrEF [5].

No significant differences in the treatment of HFmrEF were observed between CKD stages (Table 4). It should be noted that very few patients with HFmrEF were treated with SGLT2i, especially patients with GKD G5.

As seen in Table 4, patients with HFpEF and CKD G5 significantly received less frequently treatment with RASi than patients with HFpEF and CKD G4. The opposite was true for implantable cardioverter defibrillators. No other significant therapeutic differences were observed between CKD categories in the treatment of HFpEF. It is worth mentioning that very few patients with HFpEF were treated with SGLT2i, especially patients with GKD G5.

No great differences were observed in treating of comorbidities by type of HF (Table 4). Potassium binders were given significantly more in patients with HFrEF and CKD G5 than in patients with HFrEF and CKD G4. In addition, ESAs were significantly more prescribed in patients with HFpEF and CKD G5 than in patients with HFpEF and CKD G4.

Several findings of this study are noteworthy. First, we observed a high degree of heterogeneity in the prevalence of HF across various aCKD units, which may indicate an underdiagnosis in some units. Second, our research revealed suboptimal HF treatment, as evidenced by the limited use of the three foundational drugs recommended for HF therapy in the 2016 ESC HF guidelines. Third, a low prevalence of full dosing of therapeutic agents was observed. Overall, most patients were treated with HF pharmacological regimens recommended in the 20th century rather than with 21st-century regimens (Fig. 5)[5, 6, 7, 10]. Fourth, other aspects of the comorbidity treatment were also suboptimal.

Our data revealed that 13% of aCKD patients attended in specialized aCKD units during the study period of the study had a diagnosis of HF. This prevalence is not far from the prevalence of aCKD in patients with HF found in a recent Spanish study [11]. In that study, performed in 13 HF clinics over 4 months, among 1,107 outpatients with HF diagnosed according to the ESC guidelines [5‒7], 16.2% had aCKD [11]. Both cross-sectional studies are in line with data on the high prevalence of HF in patients with aCKD from long-term longitudinal studies with large population samples [1‒3].

The prevalence of HF was highly variable across participating centers, ranging between 2% and 44%. This variability may reflect important differences in the perception of the relevance of HF and/or in the reliability of criteria for its diagnosis in the real-world of aCKD units. Diagnosis of HF requires the presence of nonspecific typical symptoms that are common in aCKD and, consequently, cannot be used alone to make a diagnosis. Therefore, additional recommended diagnostic criteria include elevated plasma concentrations of natriuretic peptides and demonstration of structural heart disease by echocardiography [5‒7]. The renal clearance of natriuretic peptides decreases when eGFR falls below 30 mL/min/1.73 m², supporting the diagnostic imprecision of the standard cut-off for natriuretic peptides in HF in patients with aCKD [12]. Conversely, structural heart diseases such as left ventricular hypertrophy are prevalent in CKD patients, and this prevalence increases as renal function declines, regardless of other confounding factors. Consequently, it is present in over 80% of aCKD patients [13]. These limitations may explain why the three recommended diagnostic criteria for HF were not applied in all aCKD patients included in this study, especially in those with HFmrEF and HFpEF.

The 2016 and 2021 ESC guidelines highlight the challenges in extrapolating prospective HF randomized clinical trial data to clinical practice when aCKD remains an exclusion criterion in most studies [5, 6]. However, even patients with CKD and HF who have eGFR well above 30 mL/min/1.73 m² are less likely to be prescribed HF-modifying therapies [14]. Therefore, independent groups have proposed that given the clear benefit, the use of RASi/ARNI, BB, MRA, and SGLT2i should be strongly supported by nephrologists in patients with non-dialysis aCKD and HFrEF, with usual cautions and close monitoring of serum creatinine and potassium [8, 15]. In this sense, it is striking that, apart from BB, other foundational drugs for the treatment of HF are being used well below these expectations in patients with aCKD and HFrEF, especially in patients with CKD G5. Moreover, the proportion of patients with HFrEF who receive the full dose of each of these drugs is very low. And last but not least, the percentage of patients with aCKD and HFrEF receiving all four foundational drugs simultaneously is unacceptably low, especially in patients with CKD G5.

The use of RASi/ARNI, BB, and MRA in patients with HFrEF was already recommended with a high level of evidence in the 2016 ESC guidelines for the treatment of HFrEF [5]. Data from the present study indicate that little progress has been made in the treatment of patients with aCKD and HFrEF with these drugs 5 years later.

Although the use of SGLT2i for the treatment of HFrEF was recommended with a high level of evidence in the 2021 ESC guidelines for the first time [6], the low prescription of these drugs in patients with HFrEF is especially astounding. In fact, the beneficial effects of dapagliflozin on cardiovascular death and HF hospitalization in patients with HFrEF and eGFR 25–30 mL/min/1.73 m² were comparable to those seen in patients with early stage CKD [16]. Additionally, empagliflozin has shown to reduce HF hospitalization in patients with HFrEF and with eGFR as low as 20 mL/min/1.73 m² [17]. Our data revel that GFR is independently associated to the prescription of triple or quadruple therapy, showing that renal function is considered as the main limiting factor for an optimal treatment of HFrEF.

In the case of with HFmrEF and HFpEF the 2016 and 2021 ESC Guidelines made weak or no recommendations for the use of disease-modifying therapies that have class I evidence for use in patients with HFrEF [5, 6]. In the 2023 Focused Update of the 2021 ESC Guidelines, RASi, ARNI, BB, and MRA have a recommendation evidence IIb for the treatment of HFmrEF [7]. Findings from the present study show that the prevalence of prescription of foundational drugs does not exceed 50% in aCKD patients with HFmrEF or HFpEF, with the sole exception of BB.

Since 2021, two trials testing the SGLT2i empagliflozin [18] and dapagliflozin [19] in patients with HF and EF >40% support the recent recommendation (with class I evidence) to prescribe SGLT2i to reduce the risk of HF hospitalization or cardiovascular death in patients with HFmrEF and HFpEF7. Of interest, the empagliflozin study included patients with eGFR ≥20 mL/min/1.73 m² and the dapagliflozin study patients with eGFR ≥25 mL/min/1.73 m². Based on these results, SGLT2i is currently recommended in patients with HFmrEF or HFpEF, including those with an eGFR ≥20–25 mL/min/1.73 m², to reduce the risk of HF hospitalization or cardiovascular death. Data from a recent meta-analysis including 13 large trials suggest that this recommendation applies to diabetic patients, but not to nondiabetic patients [20]. The findings from the present study show the paucity of aCKD patients with HFmrEF or HFpEF receiving SGLT2i, especially among patients with CKD G5, and most of them had type 2 diabetes mellitus.

Regarding other treatments, dyslipidemia was treated with statins in 80% of patients but only in 55% at full doses. This is quite surprising, as aCKD is considered a very high-risk comorbidity for CV events and requires treatment targeting a cLDL <55 mg/dL as stated in 2019 ESC/EAS Guidelines [21], which is difficult to achieve with suboptimal dosing of statins [22]. PCSK9i, an effective alternative for patients with uncontrolled LDLc, was prescribed residually in aCKD patients from the present study [22].

Finally, the IROMAN Trial demonstrated that iron administration in HFrEF reduced the risk of hospital admissions for HF and cardiovascular death [23]. In patients with aCKD, treatment of absolute and relative iron deficiency in the context of anemia is associated with lower CV events [24]. However, our study found that 29% of patients with absolute iron deficiency were not receiving iron, while nearly 50% of patients with iron deficiency were receiving ESA (but no iron). Indeed, ESA was prescribed in at least 16% of the non-anemic patients. ESA overshooting of hemoglobin values is associated with an increased risk of CV events, while iron deficiency is associated with higher rates of ESA hypo-responsiveness [25].

This study has some limitations to be acknowledged. First, it carries the inherent biases of a retrospective design. There were few missing data, likely related to the accreditation programs of HF and aCKD units in Spain that started 10 years ago, leading to improved registration of clinical data. Second, albuminuria values are scarce and reported in heterogeneous units, preventing an accurate description and analysis of this important variable. However, our study focused on optimizing HF treatment in aCKD, where albuminuria is not a decisive factor for the prescription of drugs in HF. In any case, we encourage using standardized procedures and units to assess and report albuminuria values in all CKD patients, as the level of albuminuria influences the risk of death. Our study was conducted during the first trimester of 2021 as the study period to avoid the COVID-19 pandemic and also because all evaluated drugs were available and reimbursed. Third, as a transversal study, we do not have information of prognosis. Finally, it may be questionable whether the 2021 ESC Guidelines [6] and the 2023 Focused Update of the 2021 ESC Guidelines [7] used as reference were already implemented during the study period. Therefore, the results should serve as a reference for the evolving implementation of ESC guidelines on the treatment of HF in future analyses.

In conclusion, the suboptimal treatment of HF in patients with aCKD in real-life may underlie the suboptimal HF-specific outcomes in this population. As renal function is a limiting factor for an accurate prescription, cardiorenal units appear to be the solution for an optimal management of HF. Population-based, pragmatic real-world clinical trials are necessary to confirm the disease-modifying beneficial impact of HF therapy optimization in aCKD.

This study protocol was reviewed and approved by the Ethics Committee of Hospital Universitario de la Princesa, Madrid, Spain, approval number 5040, 12/12/22. The study has been granted an exemption from requiring written informed consent by the Ethics Committee of Hospital Universitario de la Princesa, Madrid, Spain.

Borja Quiroga has received honoraria for conferences, consulting fees and advisory boards from Vifor-Pharma, Astellas, Amgen, Bial, Ferrer, Novartis, AstraZeneca, Sandoz, Laboratorios Bial, Esteve, Sanofi-Genzyme, Otsuka. Alberto Ortiz has received grants from Sanofi and consultancy or speaker fees or travel support from Adviccene, Alexion, Astellas, AstraZeneca, Amicus, Amgen, Boehringer Ingelheim, Fresenius Medical Care, GSK, Bayer, Sanofi-Genzyme, Menarini, Mundipharma, Kyowa Kirin, Lilly, Freeline, Idorsia, Chiesi, Otsuka, NovoNordisk, Sysmex and Vifor Fresenius Medical Care Renal Pharma and Spafarma and is Director of the Catedra UAM-AstraZeneca of chronic kidney disease and electrolytes. He has stock in Telara Farma. José Jesús Broseta has received honoraria for conferences, consulting fees, and advisory boards from Novo-Nordisk, Kyowa Kirin, AstraZeneca, Daichii-Sankyo, Esteve, Lilly, CSL Vifor, and Chiesi. Beatriz Escamilla Cabrera has received honoraria for conferences or travel support from Vifor-Pharma, Astellas, AstraZeneca, and NovoNordisk. David Rodriguez Santarelli has received honoraria for Conferences from Lilly, Boehringer Ingelheim, AstraZeneca, for the manuscript writing from Vifor-Pharma, and travel support from Vifor-Pharma, Boehringer Ingelheim, Zambon, Servier, Esteve and Novo-Nordisk. Francisca Lopez Rodriguez has received honoraria for conferences, meetings and travel support from Vifor-Pharma, Abbott, AstraZeneca, Esteve, Otsuka, Astellas. Bárbara Cancho Castellano has received honoraria for conference from Vifor-Pharma. Adriana Puente García has received honoraria for conferences, consulting fees, and travel support from Aztra Zeneca, Boehringer Ingelheim, Faes Farma, Lilly, Menarini, Novo-Nordisk, Sanofi-Genzime, Otsuka, Viatris. Nuria Aresté has received honoraria for conferences, consulting fees, advisory boards, or travel support from Vifor, AstraZeneca, Boehringer Ingelheim, Baxter, Astellas, GSK. Miguel Ángel Rojas has received honoraria for conferences consulting fees, and travel support from: CLS-Vifor-Pharma, Servier, Astellas, Sanofi, AstraZeneca, and Boehringer Ingelheim. R. Haridian Sosa Barrios has received honoraria for conferences, consulting fees, and advisory boards from AstraZeneca. Sara Núñez, Amparo Soldevila Orient, Maria Kislikova, Henar Santana Zapatero, Esperanza Moral Berrio, María Ibáñez Cerezo, Maria Constanza Glucksmann Pizá, Yaiza Rivero Viera, Zoila Stany Albines, Ana Ródenas Gálvez, Belén Campos Gutiérrez, Sagrario Balda Manzanos, Silvia González Sanchidrián, Carmen Patricia Gutierrez Rivas, Laura Salanova Villanueva, Gema Rangel Hidalgo, Sandra Beltrán Catalán, Mayra Ortega Diaz, Lucía Rodríguez Gayo, Ana Maria Martinez Canet, and Javier Díez do not present any conflict of interest that need to be declared.

Alberto Ortiz research is supported by FIS/Fondos FEDER (PI22/00469, PI22/00050, PI21/00251, ERA-PerMed-JTC2022 (SPAREKID AC22/00027), FRIAT, Comunidad de Madrid en Biomedicina P2022/BMD-7223, CIFRA_COR-CM. Instituto de Salud Carlos III (ISCIII) RICORS program to RICORS2040 (RD21/0005/0001) funded by European Union – NextGenerationEU, Mecanismo para la Recuperación y la Resiliencia (MRR) and SPACKDc PMP21/00109, FEDER funds. This publication is based upon work from COST Action PERMEDIK CA21165, supported by COST (European Cooperation in Science and Technology) 2023-2027. PREVENTCKD Consortium Project ID: 101101220 Programme: EU4H DG/Agency: HADEA.

Conceptualization and methodology: Borja Quiroga and Javier Díez. Formal analysis and writing – original draft: Borja Quiroga, Alberto Ortiz, and Javier Díez. Investigation, visualization, and writing (review and editing): Borja Quiroga, Alberto Ortiz, José Jesús Broseta, Beatriz Escamilla Cabrera, David Rodriguez Santarelli, Francisca Lopez Rodriguez, Bárbara Cancho Castellano. Adriana Puente García, Nuria Aresté, Miguel Ángel Rojas. R. Haridian Sosa Barrios, Sara Núñez, Amparo Soldevila Orient, Maria Kislikova, Henar Santana Zapatero, Esperanza Moral Berrio, María Ibáñez Cerezo, Maria Constanza Glucksmann Pizá, Yaiza Rivero Viera, Zoila Stany Albines, Ana Ródenas Gálvez, Belén Campos Gutiérrez, Sagrario Balda Manzanos, Silvia González Sanchidrián, Carmen Patricia Gutierrez Rivas, Laura Salanova Villanueva, Gema Rangel Hidalgo, Sandra Beltrán Catalán, Mayra Ortega Diaz, Lucía Rodríguez Gayo, Ana Maria Martinez Canet, and Javier Díez.

The data underlying this article will be shared on reasonable request to the corresponding author.