Juan Martínez-Milla✉, Marcelino Cortés García, Julia Anna Palfy, Mikel Taibo Urquía,Marta López Castillo, Ana Devesa Arbiol Ana Lucía Rivero Monteagudo,María Luisa Martín Mariscal, Inés Jiménez-Varas, Sem Briongos Figuero,Juan Antonio Franco-Pelaéz, José Tuñón,7,8

1. Department of Cardiology, Hospital Universitario Fundación Jiménez Díaz-Quirónsalud; 2. Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain; 3. Department of Cardiology, Hospital Central de Asturias; 4. Department of Cardiology, Hospital Universitario Ruber-Quirónsalud; 5. Department of Cardiology, Hospital Universitario Infanta Leonor; 6. Department of Endocrinology, Hospital Clínico San Carlos; 7. Universidad Autónoma de Madrid; 8. CIBERCV,Madrid, Spain

ABSTRACT OBJECTIVE To assess the role of beta-blockers (BB) in patients with chronic kidney disease (CKD) aged ≥ 75 years.METHODS AND RESULTS From January 2008 to July 2014, we included 390 consecutive patients ≥ 75 years of age with ejection fraction ≤ 35% and glomerular filtration rate (GFR) ≤ 60 mL/min per 1.73 m2. We analyzed the relationship between treatment with BB and mortality or cardiovascular events. The mean age of our population was 82.6 ± 4.1 years. Mean ejection fraction was 27.9% ± 6.5%. GFR was 60−45 mL/min per 1.73 m2 in 50.3% of patients, 45−30 mL/min per 1.73 m2 in 37.4%, and ＜ 30 mL/min per 1.73 m2 in 12.3%. At the conclusion of follow-up, 67.4% of patients were receiving BB. The median follow-up was 28.04 (IR: 19.41−36.67) months. During the study period, 211 patients (54.1%) died and 257 (65.9%) had a major cardiovascular event (death or hospitalization for heart failure). BB use was significantly associated with a reduced risk of death (HR = 0.51, 95%CI: 0.35−0.74; P ＜ 0.001). Patients receiving BB consistently showed a reduced risk of death across the different stages of CKD:stage IIIa (GFR = 30−45 mL/min per 1.73 m2; HR = 0.47, 95% CI: 0.26−0.86, P ＜ 0.000 1), stage IIIb (GFR 30−45 mL/min per 1.73 m2; HR = 0.55, 95% CI: 0.26−1.06, P = 0.007), and stages IV and V (GFR ＜ 30 mL/min per 1.73 m2; HR = 0.29, 95% CI: 0.11−0.76; P =0.047).CONCLUSIONS The use of BB in elderly patients with HFrEF and renal impairment was associated with a better prognosis.Use of BB should be encouraged when possible.

Heart failure (HF) is one of the most prevalent cardiovascular (CV) disorders worldwide. Approximately half of all patients with HF have reduced or mid-range ejection fraction.[1,2]Due to their negative inotropic action, for many years beta-blockers (BB) were contraindicated in patients with HF. Toward the end of the last century, however, these drugs were shown to have highly positive effects in patients with HF.Since then, they have become a cornerstone in the treatment of HF with systolic dysfunction,[3]and the most recent clinical practice guidelines encourage BB use to reduce mortality and CV events.[4,5]

Despite such advances, data remain scarce on the role of these drugs in elderly patients with chronic kidney disease (CKD). Classically, both elderly and CKD patients have been underrepresented in clinical trials, creating a gap in the evidence base. In addition, CKD is highly prevalent among elderly patients with HF and reduced ejection fraction(HFrEF).[6]As a result, though HF drugs (e.g., Angiotensin-Converting Enzyme inhibitors (ACEi)/Angiotensin Receptor Blockers (ARB), BB, Mineralocorticoid Receptor Antagonist (MRA), Angiotensin Receptor-Neprilysin Inhibitors (ARNI))provide a substantial cardiovascular (CV) benefit,the underrepresentation of elderly patients with CKD in the primary clinical trials and the existence of side effects may limit their use.[6,7]As well as ACEi/ARB and MRA,[8]the elderly may be less likely to receive BB than other populations with HFrEF,particularly in cases with associated CKD, despite the fact that there is no clear reason to avoid this medication.[9]Our study analyzes the role of BB therapy in elderly patients with HFrEF and CKD.

METHODS

Patients

We carried out a single-center, observational cohort study. From January 2008 to July 2014, we consecutively enrolled 802 patients 75 years of age or older with left ventricular ejection fraction (LVEF) ≤35% as measured by 2-dimensional echocardiography. Of the total population, 390 had renal impairment, defined as a glomerular filtration rate(GFR) ＜ 60 mL/min per 1.73 m2.

A specific database compiled in the cardiac imaging department of Hospital Fundación Jiménez Díaz (Madrid, Spain) was used to screen for patients meeting both criteria. All patients underwent regular medical supervision according to their symptoms and the indications of their physician(cardiologists or general practitioners) to optimize treatment.

This investigation was carried out in accordance with the principles outlined in the Declaration of Helsinki.

Outcomes and Follow-up

The outcomes analyzed in our study were the rate of all-cause death and major CV events. Here,CV events included death from any cause or admission due to HF. HF admission was defined as admission to a health-care facility lasting ＞ 24 h due to the worsening of HF symptoms and followed by specific treatment for HF (regardless of the cause of cardiac decompensation). Data on clinical events and death during follow-up were collected from patients’ electronic health records or, if unavailable,from telephone interviews with patients or relatives.

Statistical Analysis

Data were subjected to descriptive statistical analysis via frequency measurements (absolute frequencies and percentages) for qualitative variables and using mean and standard deviation for quantitative variables. The magnitude of the effects of the variables was expressed as hazard ratio (HR) and 95% confidence interval (95% CI). Univariate analysis of the quantitative variables was performed using the Student t test when the variables were normally distributed, and the Mann-Whitney U test when distribution was not normal. Qualitative variables were analyzed using the χ2or the Fisher exact test.

Because observational studies do not allow for randomization, we planned 2 different approaches to avoid potential confounding factors: multivariate Cox proportional hazard and propensity score(PS)-matched analysis. These two analyses were used to determine significant predictors of CV events and mortality. First, we performed a multivariate analysis with Cox (backward stepwise) regression. Of all the baseline variables collected, we selected those with the potential to act as confounding factors. The selection criteria were as follows:first, clinical and biological plausibility and, second,the statistical criterion of Mickey, excluding all those variables that returned a P value ＞ 0.20 on univariate analysis. Second, we performed a PS-matched analysis. The PS was calculated by means of an ordered logistic regression model, taking the BB group as the dependent variables and adopting a parsimonious approach. In a first step, all the following variables were included in the univariate analysis: age, gender, hypertension, diabetes mellitus, obesity, GFR, chronic obstructive pulmonary disease (COPD), peripheral vascular disease, any degree of cognitive impairment, any degree of functional disability, ischemic origin of reduced EF, previous HF admission, sinus rhythm, wide QRS complex, LVEF, and New York Heart Association (NYHA)Class I or II (vs. III, IV, or not available) at initiation of follow-up. All variables with a P-value ＜ 0.2 were entered into a multivariate binary logistic regression model, which served to estimate the PS of every patient. Patient matching was performed at a 1:1 ratio with the nearest neighbor method (caliper =0.2 × SD [logitPs]).

Results are expressed as hazard ratio (HR) and 95% CI. Statistical analyses were performed with SPSS version 22.0 (SPSS, Inc, Chicago IL, USA).

RESULTS

Baseline Characteristics

During the study period, 802 consecutive patients with LVEF ≤ 35% were assessed for eligibility.Of these, 390 patients were included due to associated renal impairment. Table 1 shows the baseline characteristics of our population. In terms of sex,62.3% were male, and the mean age was 82.6 ± 4.1 years. Mean LVEF was 27.9% ± 6.5%. An ischemic etiology was found in 50.6% of cases. GFR was between 60 and 45 mL/min per 1.73 m2in 50.3% of patients, 45−30 mL/min per 1.73 m2in 37.4%, and ＜30 mL/min per 1.73 m2in 12.3%.

At the end of follow-up (32 ± 23 months), 263(67.4%) patients were undergoing treatment with BB. The most commonly used BB by type was bisoprolol once daily in 143 (54.4%) patients followed by carvedilol twice daily in 111 (42.2%) patients,metoprolol twice daily in 6 (2.3%) patients, and nebivolol once daily in 3 (1.1%) patients. Chronic lung disease (32.4%), followed by bradycardia (9.0%), asthenia (4.5%), and deterioration of HF (4.5%) were the most frequent reasons why study subjects did not take BB; however, in 29.7% of these patients no formal contraindication was found. Dose levels of BB used are shown in Table 2 for both the entire population and according to GFR.

2.3 验证预测效果 验证组平均sFlt-1(3 207.79±3 206.82)pg/ml，平均PLGF(729.30±756.96)pg/ml，平均sFlt-1/PLGF(12.87±22.76)，取sFlt-1/PLGF比值12.68作为截断值，统计出验证组的灵敏度为80.0%,特异度为92.6%,符合率为90.6%。

Outcomes

After a median follow-up of 28.04 (IR: 19.41−36.67) months, 211 patients (54.1%) died and 257 patients (65.9%) developed a major CV event (death or hospitalization for HF). Of the patients who died,the cause of death was CV in 56 cases (26.5%), and non-CV causes accounted for 73 deaths (34.6%). We were unable to determine the cause of death in 82 patients (38.9%). Regarding HF hospitalization alone, 146 patients (37.4%) of the total study population were admitted due to HF decompensation.We performed a multivariate analysis (Cox regression) of our study population in order to identify significant predictors of total mortality, following the methodology described above. In similar fashion, we performed another multivariate analysis(Cox regression) to determine significant predictors of CV events. Tables 3 and 4 show the results of univariate and multivariate analyses of overall mortality (Table 3) and CV events (Table 4). A multivariate Cox regression analysis revealed that the use of BB was significantly associated with lower mortality rates (HR = 0.53, 95% CI: 0.37−0.78, P logrank ＜ 0.001), as compared with patients not receiving BB (Figure 1A). However, BB use was not significantly associated with differences in CV events.When we used propensity score matching specifically aimed at analyzing the role of BB in our population, we found that BB had benefited our population, producing a difference that reached statistical significance (HR = 0.45, 95% CI: 0.27−0.75, P =0.002) (Figure 1B). Similarly, we found no relationship between BB and CV events. Finally, a multivariate Cox analysis considering HF hospitalization alone revealed no relation between BB and a reduction in HF admissions; only ACEi/ARBs played a protective role in this regard (HR = 0.467; 95%CI: 0.313−0.696).

When we analyze the role of BB according to eGFR, we see similar results throughout the study population. In the subgroup of patients with stage IIIa CKD (GFR 45−60/min per 1.73 m2), BB significantly reduced mortality (HR = 0.47; 95% CI:0.26−0.86; P log-rank ＜ 0.000 1); the same was true for patients with stage IIIb (GFR = 30−45/min per 1.73 m2; HR = 0.55, 95% CI: 0.26−1.06; P log-rank =0.007) and stages IV and V disease (GFR ＜ 30 mL/min per 1.73 m2; HR = 0.29, 95% CI: 0.11−0.76; P log-rank =0.047) (Figure 2).

When we analyzed the population by BB dose, no differences in the mean dose of bisoprolol and carvedilol were found between the different glomerular filtration groups. A similar analysis was not done for metoprolol and nebivolol because of their low rate of use in our population.

Table 1 Baseline characteristics.

DISCUSSION

Blocking the adrenergic system with BB has proven effectiveness in patients with HFrEF,[3]and this treatment is currently included in clinical guidelines.[4,5]A closer look at the studies reporting evidence in support of using these drugs, however,reveals that the populations studied present a relatively low number of comorbidities, with few patients over 75 years of age (average age commonly under 65 years),[10,11]making these studies unrepresentative of routine clinical practice.[12]Elderly patients make up a substantial portion of the population with severe left ventricular dysfunction,[9]and the rate of renal failure in this cohort is often 3-fold higher than that of the general population.[13]In addition, these patients have a higher proportion of other comorbidities and polypharmacy, and more than 70% of HF patients older than 80 years fulfil frailty criteria.[14,15]These differences are important to bear in mind when interpreting the results of randomized clinical trials on BB, and nowadays, specific data remain limited and controversial.[16]

Table 2 Daily doses of beta-blockers used.

Table 3 Univariate and multivariate analysis of overall mortality.

Although the available evidence on BB therapy in patients ＞ 70 years with HFrEF is limited, recent studies support the use of these drugs. The SENIORS trial compared the use of BB against a placebo in patients over 70 years of age with HF;[17]all patients included in the study had a clinical history of chronic HF with one or both of the following features: documented hospital admission within the previous 12 months with a discharge diagnosis of congestive HF or documented LVEF ≤ 35% within the previous 6 months. The trial demonstrated a correlation between nebivolol use and a significant(14%) reduction of the primary end-point, a composite of all-cause mortality or cardiovascular hospital admission. Though a secondary end-point, no favorable impact of nebivolol on all-cause mortality was demonstrated (HR = 0.88, 95% CI: 0.71−1.08;P=0.21).[18]We believe that the main difference between the SENIORS study and both our study and pivotal clinical trials is the inclusion of patients with preserved LVEF, since 35% of patients were reported to have LVEF ＞ 35%. In fact, no treatment has demonstrated a clear survival benefit among patients with preserved LVEF-HF.[4,5]There is a lack of robust data evidencing decreased mortality associated with BB administration in the elderly population. Hernandezet al., in the OPTIMIZE-HF registry, suggest that BB are beneficial in elderly patients with HFrEF.[16]Few studies have reported clear benefit of BB in elderly population in terms of mortality. Our group conducted a retrospective, observational study in elderly patients (＞ 75 years) with HFrEF, concluding that BB therapy improves survival in patients with LVEF lesser than or equal to 0.35, although this effect seems unrelated to the dose received (P= 0.025).[18]

Table 4 Univariate and multivariate analysis of cardiovascular events.

Figure 1 All-cause mortality in overall population. (A): Kaplan-Meier curve showing all-cause mortality in the overall population,comparing the group under beta-blocker therapy (green) against the group not receiving this treatment (blue); (B): Kaplan-Meier curve showing all-cause mortality, comparing the group under beta-blocker therapy (green) versus the group that did not use beta-blockers(blue), after propensity score matching.

Another issue to take into account when evaluating these patients is CKD status. We know this disorder is more prevalent in patients with HF and has an important influence on prognosis.[19,20]In addition, we know that the presence of CKD in HF patients affects the prescription, dosage, and maintenance of therapies that have demonstrated benefits in HFrEF.[4,5]In addition, advanced-stage CKD was an exclusion criterion in many of the clinical trials analyzing the role of BB therapy in patients with HFrEF.[21]Furthermore, those studies that have examined the role of BB therapy in this group of patients are mostly observational in design, and the endpoints used are less relevant (i.e., other than major factors or nonfatal clinical events).[22,23]Despite this lack in the knowledge base, presence of CKD is one of the primary factors associated with increased mortality.[22]

Most clinical trials carried out to date use exclusion criteria based on glomerular filtration rate(GFR), and as a result patients with Stage I and II kidney disease are well-represented (GFR ＞ 90 mL/min per 1.73 m2and 60−89 mL/min per 1.73 m2, respectively).[24,25]However, this representativeness decreases at lower GFR, and the available data on patients at stage IV and V are scant.

This pattern can be seen in the classical studies investigating the role of BB therapy in HFrEF patients with associated Stage-IIIa, Stage-IIIb, and Stage-IV-V CKD(25). Although there is no strong evidence of the effect of BB in CKD, stage III is better represented in the different clinical trials. In the MERIT-HF trial (metoprolol vs placebo), which included patients with HFrEF, there was a significant relative risk reduction in the composite endpoint of CV hospitalization/all-cause mortality in patients with GFR of 45−60 mL/min per 1.73 m2(HR = 0.68(0.52−0.90)) and even in patients with GFR ＜ 45 mL/min per 1.73 m2.[26]

In the CIBIS-II trial on the effect of bisoprolol in patients with HFrEF, BB significantly reduced the mortality and HF-related hospital stay in the subgroup of patients with GFR ＜60 mL/min per 1.73 m2as well as those with GFR ＜ 45 mL/min per 1.73 m2.[27]As in the SENIORS trial in patients with reduced GFR, the effect of BB was not different from the effect in patients with GFR above 60 mL/min per 1.73 m2.[28]Finally, a meta-analysis of the effect of carvedilol in the COPERNICUS (Carvedilol Prospective Randomized Cumulative Survival) and CAPRICORN (Carvedilol Post Infarct Survival Control in LV Dysfunction) trials showed that this BB significantly improved outcome in patients with eGFR between 45 and 60 mL/min per 1.73 m2.There was no interaction between the effect of carvedilol treatment and eGFR categories (＜ 45vs.45 to 60 mL/min per 1.73 m2).[29]

Stages IV and V are not well-represented, though data from both the MERIT-HF and CIBIS-II trials suggest that BB are effective in patients with CKD stage IIIb-V(26,27). Specially, in the MERIT-HF study, the metoprolol/placebo hazard ratio was 0.41 (95% CI: 0.25 to 0.68) in the 493 patients with eGFR ＜ 45 mL/min per 1.73 m2(12% of the whole study population). This subgroup had a mean eGFR of 36.6 ± 6.8 mL/min per 1.73 m2, which included patients with eGFR ＜ 30 mL/min per 1.73 m2.[26]In the SENIORS study, only 3.1% of patients had stage IV CKD, but no subgroup analysis has been performed on these patients.[28]However, only 8% of all patients in these studies had stage 4 CKD. In a small trial of hemodyalisis patients with HF, carvedilol significantly improved the secondary combined endpoint of all-cause mortality and CV death.[30]

Recently, Kotecha,et al.[31]published the largest meta-analysis including patients with left ventricular dysfunction and CKD. The authors considered 10 double blind placebo-controlled trials that included more than 16,000 patients. They found that BB reduced the relative risk of all-cause mortality by 27% (95% CI: 0.62−0.86) in patients with GFR of 45−60 mL/min per 1.73 m2and 29% (95% CI:0.58−0.87) in those with a GFR of 30−44 mL/min per 1.73 m2. This benefit was only seen in patients in sinus rhythm. In patients with GFR ＜ 30 mL/min per 1.73 m2there were no enough patients to draw conclusions due to the exclusion criteria of the different trials.[31]

Concern for increased toxicity often leads clinicians to undertreat these patients with CKD, causing less therapeutic resources to be devoted to individuals with myocardial infarction and concomitant CKD.[32]However, it has been shown that these therapeutic measures are beneficial in this population.[33]

Our population is particularly elderly (mean age,82.6 ± 4.1 years), and as such is representative of the individuals we treat in our daily practice. There is currently no solid evidence on the role of BB in the elderly population with HF and CKD. Given this lack of data about the role of BB in this common population: elderly with HF and CKD. For this reason, we believe that our findings are relevant for overall practice. We found a significant reduction in all-cause mortality, and this benefit was maintained when separately assessing the role of BB treatment in advanced CKD patients (＜ 45 mL/min per 1.73 m2), as the protective effect of this treatment continues to be statistically significant in terms of all-cause mortality. The effect of treatment with BB is neutral, however, when we isolate the variable of mortality. The high rate of associated cardiovascular co-morbidities may have attenuated the beneficial effect of BB therapy in our study population in terms of CV events.

In sum, according to our data, treatment with BB in elderly patients presenting HFrEF and CKD was associated with a lower rate of all-cause mortality.Our data thus show that BB therapy could improve the prognosis of this selected population when there is no formal contraindication for its use.

STUDY LIMITATIONS

Our study has certain limitations. First, the study population is relatively small, which could influence the statistical results. In addition, it is a retrospective, non-randomized study using a historical cohort from a single center. A third limitation is the relatively short follow-up period, potentially masking a long-term benefit of BB in reduction of CV events. Nevertheless, this last issue is less relevant due to the short life expectancy of elderly patients and the higher number of CV events they present.Lastly, we were unable to discern the cause of death in 82 (38.7%) patients as this information was lacking from their clinical records.

CONCLUSIONS

According to our results, use of BB is significantly associated with a reduction in all-cause mortality in elderly patients with HFrEF and CKD irrespective of GFR. As a result, these drugs may be beneficial for these patients provided there are no formal contraindications. Nevertheless, this is an observational study and that residual confounding may exist.

ACKNOWLEDGMENTS

To Oliver Shaw for editing the manuscript for aspects related to English language usage and style.