Article ID: CJ-19-0333
Background: Intravenous vasodilators are commonly used in patients with hypertensive acute decompensated heart failure (ADHF), but little is known about their optimal use in blood pressure (BP) management to avoid acute kidney injury (AKI). The purpose of this study was to investigate the association between systolic BP (SBP) changes and the incidence of AKI in patients with hypertensive ADHF.
Methods and Results: Post-hoc analysis was performed on a prospectively enrolled cohort. We investigated 245 patients with ADHF and SBP >140 mmHg on arrival (mean age, 76 years; 40% female). We defined “SBP-fall” as the maximum percent reduction in SBP 6 h after intravenous treatment. AKI was defined as serum creatinine (SCr) ≥0.3 mg/dL, or urine output <0.5 mL/kg/h (n=66) at 48 h. Mean SBP and SCr levels on arrival were 180 mmHg and 1.21 mg/dL, respectively. Patients with AKI had significantly larger SBP-fall than the others (36.7±15.3% vs. 27.2±15.3%, P<0.0001). Logistic regression analysis showed an odds ratio per 10% SBP-fall for AKI of 1.49 (95% confidence interval 1.29–1.90, P=0.001). SBP-fall was significantly associated with the number of concomitant used intravenous vasodilators (P=0.001). The administration of carperitide was also independently associated with increased incidence of AKI.
Conclusions: Larger SBP-fall from excessive vasodilator use is associated with increased incidence of AKI in patients with hypertensive ADHF.
Acute decompensated heart failure (ADHF) is a life-threating condition that requires urgent and adequate treatment.1,2 Prompt early intervention is necessary to improve the outcome of hospitalized patients with ADHF.3,4 Patients with ADHF are usually classified by systolic blood pressure (SBP) level at admission; this is important not only for prognosis, but also for developing the early-phase treatment strategy.5
Previous studies have shown that more than half of ADHF patients have SBP >140 mmHg on admission.6 For early intervention in hypertensive ADHF, intravenous vasodilators are the primary recommended agent for relieving symptoms.7 They work by reducing cardiac preload and afterload through decreased vessel tone, which consequently improves both congestion and systemic perfusion.8 The current guidelines of the European Society of Cardiology recommend aggressive BP reduction in the range of approximately 25%.7 However, there is no established evidence regarding the optimal degree of BP reduction, particularly in patients with hypertensive ADHF.
Clinicians face the dilemma of insufficient BP reduction delaying improvement of heart failure symptoms, but excessive BP reduction inducing tissue hypoperfusion. Acute kidney injury (AKI), a result of renal hypoperfusion, is a serious complication during ADHF treatment, because AKI is known to negatively effect the outcome of patients with ADHF.9–11
The purpose of the present study was to investigate the association between BP reduction caused by administration of intravenous vasodilators and the incidence of subsequent AKI in patients with hypertensive ADHF.
A retrospective analysis of a prospectively collected cohort was performed. A total of 569 patients who were hospitalized with ADHF between January 2012 and January 2018 were enrolled. The diagnosis of ADHF was based on the Framingham criteria.12 Of all patients, we excluded 2 with insufficient data, 278 with SBP ≤140 mmHg, 15 with endstage renal disease defined as eGFR <15 mL/min/1.73 m2,2,13 3 with acute myocardial infarction, 1 with requirement for mechanical circulatory support at admission, and 1 with isolated right ventricular failure. We also excluded 24 patients who did not receive intravenous agents within 6 h of arrival, because this study focused on the association between BP reduction caused by medical intervention, in particular, intravenous vasodilators, and the incidence of early-phase AKI. Finally, we analyzed 245 patients with ADHF. The study protocol was approved by the institutional ethical review board (approval no. 20111323). All participants provided written informed consent.
In this study, we calculated the incidence of AKI 48 h after the patient’s arrival. According to the Kidney Disease Improving Global Outcome (KDIGO) definitions,14 we defined the development of AKI as an increased level of serum creatinine (SCr) ≥0.3 mg/dL after arrival or urine volume <0.5 mL/kg/h for 6 h. All 245 patients were divided into 2 groups according to incidence of AKI: AKI group (n=66) or Non-AKI group (n=179). Blood samples for laboratory measurement of SCr were taken on arrival and at 24 h, and 48 h. We also recorded in-hospital deaths and duration of hospital stay as the in-hospital outcome. In addition, 1-year outcomes, including all-cause death, cardiovascular death, and unplanned hospitalization because of decompensated HF, were also obtained.
BP data were collected on arrival (0 h), and at 1.5 h, 6 h, 12 h, 24 h, and 48 h. In the present study, we defined “SBP-fall” as the percent maximal reduction in SBP within 6 h after arrival, which was calculated as follows: SBP-fall (%)=[(SBP at 0 h)−(the lowest SBP within 6 h)]/(SBP at 0 h)×100.
We conducted a focused investigation on frequently used intravenous vasodilators, including nitrates (nitroglycerine and isosorbide dinitrate), carperitide, and calcium-channel blockers (CCBs: nicardipine and diltiazem). Carperitide is a synthetic α-human A-type natriuretic peptide that is commonly used to improve ADHF symptoms in Japan.15
Continuous variables are presented as mean±standard deviation or median and interquartile range, and the distributions of these variables between the 2 groups were compared with 2-sample t-test or Wilcoxon rank-sum test. Categorical variables are presented as number and percentage, and the distributions of these variables between groups were compared with the χ2 test. The changes in SBP at each of 1.5, 6, 24, and 48 h were compared between the AKI (n=54) and Non-AKI (n=139) groups by 2-sample t-test. The change ratio in SBP at X h was calculated as follows: [(SBP at X h)−(SBP at 0 h)]/(SBP at 0 h)×100.
The Bonferroni correction was applied to correct the multiple comparisons, and the Dunnett’s t-test was used to compare patients treated with 1, 2, and 3 vasodilators with those treated with loop diuretics alone. We performed univariable logistic analyses to evaluate the odds ratio (OR) and 95% confidence interval (95% CI) and thus assess the influence of each variable on the development of AKI. In the multivariable logistic analysis, variables with P<0.05 in the univariable analyses were selected. The restricted cubic spline was used for visualization of the continuous relationship between SBP-fall and adjusted OR for AKI by multivariate logistic regression model with the adjustment for chronic obstructive pulmonary disease, heart rate, and carperitide. Knots were placed at the 10th, 50th, and 90th percentiles (8.6%, 28.7%, 52.3%).
In the present study, we converted loop diuretic dose to furosemide equivalents with 20 mg oral furosemide, 5 mg oral torasemide, and 30 mg oral azosemide.16
The logistic regression analysis with restricted cubic spline analysis was performed on rms of the R package. Other statistical analyses were performed with JMP pro version 13.0 (SAS Institute, Cary, NC, USA). A value of P<0.05 was considered to indicate statistical significance.
The patients’ characteristics on arrival at hospital are shown in Table 1. There were no significant differences in age, NYHA functional class, SBP, and SCr on admission between the 2 groups. Of all subjects, 41% were prescribed oral loop diuretics, although there were no significant differences in furosemide equivalent dosage (AKI vs. Non-AKI group, 14±27 mg vs. 13±26 mg, P=0.45).
ACEI/ARB, angiotensin-converting enzyme inhibitor/angiotensin receptor antagonist; AKI, acute kidney injury; AF/AFL, atrial fibrillation/atrial flutter; Alb, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; BNP, B-type natriuretic peptide; BUN, blood urea nitrogen; CCB, calcium-channel blocker; CHF, chronic heart failure; COPD, chronic obstructive pulmonary disease; DBP, diastolic blood pressure; DL, dyslipidemia; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; Hb, hemoglobin; HTN, hypertension; LVEF, left ventricular ejection fraction; MRA, mineralocorticoid receptor antagonist; NYHA, New York Heart Association; SBP, systolic blood pressure; SCr, serum creatinine.
The agents administered intravenously within the first 6 h are shown in Table 2. The median interval from arrival to first administration of intravenous agent was 60 (30–122) min, and there was no significant difference between the 2 groups (AKI group, 59 [23–103] min; Non-AKI group, 65 [33–128] min; P=0.21). There was also no significant difference between groups in the total dose of intravenous furosemide during the first 48 h (AKI group 29±52 mg; Non-AKI group, 22±21 mg; P=0.29). CCBs used were 76% nicardipine and 24% diltiazem. Small numbers of patients were treated with digitalis (3%), noradrenaline (2%), dobutamine (4%), and landiolol (2%).
AKI, acute kidney injury; CCB, calcium-channel blocker; i.v., intravenous; ISDN, isosorbide dinitrate; NTG, nitroglycerin.
Noninvasive positive pressure ventilation (NPPV) and tracheal intubation were required in 39 (16%) and 10 (4%) patients, respectively, although there were no significant differences between the 2 groups (AKI group vs. Non-AKI group; NPPV, 17% vs. 16%; P=0.85; tracheal intubation, 6% vs. 3%; P=0.47, respectively).
The SBP continuously and significantly decreased during the first 24 h. The AKI group’s SBP was significantly lower than the Non-AKI group’s at 6, 24 and 48 h (Figure 1).
Percent change in mean systolic blood pressure (SBP) from arrival (0 h) to 48 h in AKI (red line) and Non-AKI (blue line) groups. The error bars indicate standard error. *P<0.05 in the comparison of the 2 groups. AKI, acute kidney injury.
Logistic analyses were performed to detect variables that were predictive of AKI (Table 3). We found that SBP-fall was an independent predictor of AKI. In the comparisons of quartile groups of SBP-fall, a larger SBP-fall was associated with higher AKI incidence (Figure 2). There were no significant differences in the in-hospital deaths or the duration of in-hospital stay. When the incidence of AKI with no SBP reduction (SBP-fall=0%) was defined as the reference, there was a positive association between SBP-fall and the OR for AKI (Figure 3). There were significant differences in SCr at discharge between the AKI and Non-AKI groups (1.41±0.65 mg/dL vs. 1.23±0.51, P=0.021), and according to the 1-year outcome after hospital discharge, the AKI group tended to have higher rates of cardiovascular death and heart failure rehospitalization (Table 4).
AUC, area under the curve; CI, confidence interval; OR, odds ratio. Other abbreviations as in Tables 1 and 2.
Odds ratio (OR) for acute kidney injury (AKI) and in-hospital outcome by the interquartile range of SBP-fall. The error bars indicate 95% confidence interval. Data presented as n (%), median and interquartile range. SBP, systolic blood pressure.
Continuous relationship of SBP-fall and odds ratio (OR) for acute kidney injury (AKI) by SBP-fall. The OR for AKI was adjusted by chronic obstructive pulmonary disease, heart rate, and carperitide (n=244, lack of heart rate data=1). CI, confidence interval; SBP, systolic blood pressure.
AKI, acute kidney injury; CV death, cardiovascular death; HF, heart failure.
The associations between the number of administrated vasodilators and SBP-fall and OR for AKI are shown in Figure 4. The incidence of AKI in patients treated with diuretics alone was defined as the reference (OR=1). Compared with the reference, an increased number of administered vasodilators induced a larger SBP-fall, and accordingly increased the risk of AKI. SBP-fall was 27.5%, even in patients not administered any vasodilators. Regarding the details of the vasodilators, carperitide was a significant predictor of AKI, whereas NTG/ISDN or CCB had no predictive value. In addition, the predictive value of carperitide was independent of SBP-fall (Table 3). In fact, in patients who treated with 1 vasodilator, the SBP-fall with carperitide (n=95) was smaller than that with other agents (NTG/ISDN or CCB) (n=61) (23.6% vs. 38.5%, P≤0.001).
SBP-fall and odds ratio (OR) for acute kidney injury (AKI) by the number of administered vasodilators. The error bars indicate standard error. A loop diuretic alone was deemed as the reference for Dunnett’s test. SBP, systolic blood pressure.
In the present study, we investigated early-phase BP changes in patients with hypertensive ADHF, and report 3 main findings. First, SBP-fall within the first 6 h predicted the incidence of AKI at 48 h. Second, the SBP-fall was associated with the number of vasodilators administered. Third, a larger SBP-fall did not improve the duration of hospital stay or in-hospital mortality.
In early-phase treatment of hypertensive ADHF, vasodilators are commonly used to reduce cardiac afterload and ameliorate the symptoms of ADHF. In contrast, excessive use of vasodilators can induce renal hypoperfusion, which leads to AKI.17 Indeed, several previous studies have shown that BP reduction is associated with increased incidence of AKI in patients with ADHF.3,18,19 However, those studies included patients with mean SBP of 130–140 mmHg at baseline. We guessed that, in patients with normotensive or hypotensive ADHF, BP reduction during the early phase would indicate arterial underfilling rather than optimization of cardiac afterload. Conversely, we expected that patients with hypertensive ADHF would possibly benefit from BP reduction in the early phase. However, unexpectedly, we showed that SBP-fall was associated with the risk of AKI, even in patients with SBP >140 mmHg. In addition, the continuous relationship between SBP-fall and the OR for AKI showed a positive correlation (Figure 3). We considered it significant that there was not a U-shaped relationship between SBP-fall and the incidence of AKI. These results indicated that, consistent with a previous study,20 there is no SBP threshold for the incidence of AKI in patients with hypertensive ADHF. When we performed the receiver-operator characteristic analysis, the best cutoff value of SBP-fall was 30.9%, with 66.7% sensitivity and 60.2% specificity (c-statistic 0.676, P<0.001) (Supplementary Figure). However we thought that this cutoff value would be inappropriate as a target SBP level in the early-phase management of hypertensive ADHF, because in the clinical setting, it is important to minimize the risk of AKI rather than improve the accuracy of prediction of AKI. In terms of the target level of SBP-fall, the current ESC guideline recommends within 25% reduction of SBP, which is not based on established evidence. In our cohort, 25% SBP-fall had high sensitivity (77.2%) with low specificity (43.6%) for the incidence of AKI. Therefore, we consider that in the range of 25% SBP reduction would be reasonable even in patients with hypertensive ADHF.
The progression of left ventricular hypertrophy and increased vascular stiffness are considered important inducible factors of hypertensive ADHF.21 It is known that these structural and functional cardiovascular changes reflect impairment of the renal autoregulatory function.22 The susceptibility of intraglomerular pressure to impairment of renal autoregulation explains why AKI can result from BP reduction even in patients with hypertension.
We also clarified that the number of administered vasodilators was significantly associated with SBP-fall and the risk of AKI (Figure 4). This result suggested that combined use of intravenous vasodilators is associated with SBP-fall and risk of AKI. Some clinicians assert that aggressive BP reduction is reasonable for prompt improvement of ADHF symptoms; however, we should note that, with a higher incidence of AKI, a larger SBP-fall did not reduce in-hospital deaths or shorten the duration of in-hospital stay (Figure 2).
Of course, we do not deny the effectiveness of vasodilator use in patients with ADHF.23 Rapid fluid redistribution caused by venous vasocontraction is considered a mechanism responsible for ADHF. Vasodilators work by reducing vessel tone not only in arteries, but also in veins. Therefore, it is important to counteract unnecessary venous vasoconstriction to decrease central venous pressure (CVP) by using vasodilators. Mullens et al reported that higher CVP is associated with worsening renal function in patients with decompensated HF.24 Although the present study had no data on CVP or serial changes in jugular venous distention, we speculate that the ideal ADHF treatment may be to use vasodilators to reduce CVP with a minimal SBP-fall. Clinicians should minimize the SBP-fall and usage of intravenous vasodilators as long as HF symptoms are improved. Conversely, it would be inappropriate to use vasodilators aggressively to decrease the SBP to an arbitrary target level, particularly within the first 6 h after hospital arrival. In addition, the patients who were administered intravenous loop diuretics alone presented with a certain degree of reduction in SBP (Figure 4). Of course, other interventions such as oxygen administration or artificial ventilation would have a substantial effect on the reduction in SBP. From these perspectives, for management of hypertensive ADHF, clinicians should avoid abrupt BP reduction within the first 6 h resulting from combined use of vasodilators.
We also clarified that the administration of carperitide was also independently associated with increased incidence of AKI (Table 3). Matsue et al reported that carperitide was significantly associated with increased in-hospital deaths of patients with acute HF.25 We thought that the concomitant use of intravenous furosemide and carperitide might be a risk for excessive reduction of effective plasma volume, which contributes to AKI. A previous study showed that nesiritide, a recombinant B-type natriuretic peptide, was not associated with worsened renal function, but it was associated with an increase in the rate of hypotension.26 Interestingly, the predictive value of carperitide for the incidence of AKI was independent of SBP-fall in the present study (Table 3). Although there might be different properties between carperitide and nesiritide, we should take care about excessive reduction of effective plasma volume when using carperitide, even if patients do not present an excessive SBP-fall.
We showed that the SBP on arrival had no association with the incidence of AKI, although previous studies reported an association between higher SBP and risk of AKI.27 We were able to explain this result by the homogeneity of our population; we included only patients with SBP >140 mmHg. In addition, the association between higher SBP and incidence of AKI might derive from larger SBP reductions in the early phase.
Recent previous studies reported that the prognostic ability of AKI on admission may be superior to AKI occurring within the first 5 days,28 and a transient elevation of SCr level does not predict poor outcome because it would reflect the process of decongestion.29,30 However, it is also well known that the complication of AKI predicts poor outcome in patients with ADHF.31 Actually, in the clinical setting, it is not easy to distinguish true AKI from pseudo AKI when the SCr increases. In the present study, 68% patients in the AKI group experienced decreased urine output. Low urine output or poor diuretic response inhibits decongestion. Therefore, clinicians should take care with excessive SBP reductions that would cause AKI.
There were several limitations to the present study. First, we could not determine whether smaller SBP-fall directed management would decrease the risk of AKI, because this study was performed in a post-hoc manner. Second, we could not investigate the dosages and duration of administration of intravenous vasodilators; it was difficult to define their equivalent doses. Third, we could not take into consideration changes in effective fluid volume, which is an important determinant of the incidence of AKI.
In the first 6 h of management of hypertensive ADHF patients, aggressive SBP reduction by combination use of vasodilator agents predicted the incidence of AKI. In addition, a larger SBP reduction was not associated with the in-hospital outcome. The risks and benefits of BP reduction should be considered in patients with hypertensive ADHF.
Please find supplementary file(s);