Alterations in Kidney Function Associated With Heart Failure

Epidemiology of Chronic Kidney Disease and the Impact on Heart Failure

Kidney dysfunction is commonly seen in both stable patients and those with acute decompensated heart failure (ADHF). Ezekowitz and colleagues demonstrated that almost 40% of outpatients with both coronary artery disease and HF have Stage 3 or higher CKD, defined as an eGFR less than 60 mL/min/1.73m ² . The prevalence of Stage 3 or higher CKD was as high as 64% among patients admitted for ADHF.

The presence of CKD is one of the most important prognostic factors for patients with HF, regardless of whether the ejection fraction (EF) is preserved or reduced ( Fig. 15.1 ). In a prospective cohort of 754 adults referred to an outpatient HF clinic, a 1% increase in 1-year mortality was seen for every 1 mL/min decrease in creatinine clearance. In an observational study of 24,331 adults with HF, Smith and colleagues found a graded association of eGFR with hospitalization for HF, all-cause hospitalization, and mortality. The association was similar between patients with preserved (n = 14,579) and reduced (n = 9762) EF. Importantly, patients with CKD experienced the same reduction in mortality from treatment with angiotensin-converting enzyme inhibitors and beta-blockers as did patients with normal kidney function, although they were less likely to receive the medications.

The presence of CKD is also associated with poor short-term outcomes. Data from the Acute Decompensated HF National Registry (ADHERE) showed that in-hospital mortality increased from 1.9% among patients with normal kidney function to 7.6% and 6.5% among patients with baseline CKD and kidney failure, respectively. While GFR remained a significant predictor of mortality after adjusting for confounders, the blood urea nitrogen (BUN) level remained the single best predictor. Other investigators have confirmed the usefulness of BUN as a marker of increased risk. In the Acute and Chronic Therapeutic Impact of a Vasopressin Antagonist in Chronic Heart Failure (ACTIV in CHF trial), patients with a baseline BUN greater than 40 mg/dL experienced an event rate of 30%, whereas patients whose baseline BUN less than 18 mg/dL had an event rate of 8.6%. It is likely that, in the acute setting, BUN is a marker of neurohormonal activation, which leads to constriction of the afferent arteriole, secretion of vasopressin, and enhanced reabsorption of sodium, water, and urea. These pathways will be discussed in more detail as follows.

Worsening Kidney Function and Prognosis

Worsening kidney function (WKF) among patients admitted with ADHF is often defined as an increase in creatinine ≥0.3 mg/dL. As many as one-third of patients may experience WKF, which typically occurs in the first 3 to 4 days of admission, but has also been shown to occur in the week after hospital discharge. In the Prospective Trial of Intravenous Milrinone for Exacerbations of Chronic Heart Failure (OPTIME-CHF), 12% of participants experienced a greater than 25% decrease in eGFR and 39% experienced a greater than 25% increase in BUN. In a meta-analysis of eight randomized controlled trials of renin-angiotensin-aldosterone system (RAAS) inhibitors in stable outpatients (6 HFrEF, 1 HFpEF and 1 that included both populations), the incidence of WKF was slightly higher in HFrEF compared with HFpEF (12% vs. 7%). Risk factors for WKF include baseline creatinine greater than 1.5 mg/dL, diabetes, and pulmonary edema.

While earlier studies suggested that WKF portended a worse prognosis, more recent analyses suggest that the association of WKF with prognosis is highly dependent on the clinical context. For example, in a study of 599 adults admitted to the hospital with ADHF, the development of WKF was associated with 1-year mortality only in patients who also had signs on physical exam of congestion at the time of hospital discharge. In a study of 1232 adults admitted for ADHF, Salah and colleagues compared the relative contribution of a decrease in N-terminal-pro-B-type natriuretic peptide (NT-pro-BNP) and WKF to prognosis. The authors found that WKF was not associated with 180-day mortality or readmission to the hospital for CVD among patients who had at least a 30% reduction in their NT-pro-BNP. In addition, the change in NT-pro-BNP did not predict development of severe WKF (defined as an absolute increase in serum creatinine ≥0.5 mg/dL), suggesting that severe WKF is multifactorial, and not simply explained by overly aggressive diuresis. Ibrahim and colleagues also evaluated the relative importance of NT-pro-BNP versus WKF in a randomized controlled trial of biomarker-guided HF therapy for patients with class II-IV HF with reduced ejection fraction (HFrEF). The authors found that adults who achieved NT-pro-BNP levels less than 1000 pg/mL were on higher doses of RAAS inhibitors, regardless of WKF. Further, patients with WKF but an NT-pro-BNP less than 1000 pg/mL experienced significantly fewer events at 1 year (worsening HF, hospitalization for HF or CVD death) compared with patients with no WKF but an NT-pro-BNP greater than 1000 pg/mL. Paradoxically, in the Diuretic Optimization Strategies Evaluation (DOSE) trial, a prospective multicenter trial investigating strategies of loop diuretic administration, increases in serum creatinine were associated with a lower risk for death, emergency room visits within 60 days, or rehospitalization (hazard ratio [HR] = 0.81 per 0.3 mg/dL increase, 95% CI 0.67–0.98, P = .026). Moreover, compared with those with stable kidney function, there was a strong association between improved kidney function (n = 28) and the composite endpoint (HR = 2.52, 95% CI = 1.57–4.03, P < .001). Taken together, available data suggest that for both acute and chronic management of HFrEF, it is important to prioritize decongestion and the use of evidence-based therapy, such as RAAS inhibitors, even with modest degrees of WKF.

There are few studies comparing the risk of WKF in patients with HFpEF versus HFrEF. In a meta-analysis of eight RCTs of RAAS inhibitors versus placebo (6 HFrEF, 1 HFpEF, 1 both), adults with HFrEF who were randomized to RAAS inhibitors and had WKF experienced higher rates of HF hospitalization, compared with patients who experienced no WKF (RR, 1.19 [1.08–1.31]; P < .001). However, the risk associated with WKF in patients allocated to placebo was larger (RR, 1.48 [1.35–1.62]; P < .001), and significantly different from patients randomized to RAAS inhibitors with WKF ( P for interaction = .005). Adults with HFpEF who were taking RAAS inhibitors also experienced a higher risk of hospitalization if they developed WKF compared with those without WKF (RR 1.78, 95% CI 1.43–2.21). However, unlike adults with HFrEF, WKF for those on placebo did not confer an increased risk of hospitalization. While is it possible that the pathophysiology of WKF differs between HFrEF and HFpEF, it is likely that the use of RAAS inhibitors simply unmasked greater disease severity.

Pathophysiology of Chronic Kidney Disease in Heart Failure

The homeostatic relationship between the heart and kidneys with respect to maintenance of body fluid homeostasis has been appreciated for centuries. Indeed, Sir William Withering noted that patients with “dropsical” conditions, characterized by the swelling of soft tissues due to the accumulation of excess water, often experienced a brisk diuresis following the administration of foxglove. In the early phases of HF, the ratio of the glomerular filtration rate to kidney blood flow (referred to as the filtration fraction) is maintained, and can even be increased. This is largely due to vasoconstrictor peptides, such as arginine vasopressin (AVP), angiotensin II, and norepinephrine (NE), which help maintain GFR by maintaining blood pressure and constricting the efferent arteriole resulting in increased intraglomerular capillary pressure ( Fig. 15.2 ). The remainder of the chapter will review some of the normal physiology that controls salt and water homeostasis, and some of the cardio-kidney pathways activated and contributed to a failing heart in clinical HF.

Peripheral Volume Sensors

Baroreceptors and Mechanoreceptors

Numerous pressure receptors (baroreceptors, mechanoreceptors) are located in critical areas in the circulatory tree—the carotid sinus, aortic arch, afferent glomerular arterioles in the kidney, and the superior and inferior vena cava. The receptors detect alterations in pressure and stretch of the vessel wall. In volume-depleted states, where stretch is decreased, there is a subsequent loss of tonic inhibition by the parasympathetic nervous system, which results in reflex increase sympathetic neural tone ( Fig. 15.3 ). The baroreceptors of the carotid body and aortic arch contribute to the antinatriuretic response observed in HF patients. Studies have demonstrated a reduction in urinary output and sodium excretion following volume expansion in the presence of carotid sinus excitation, despite a constant arterial pressures.

The atria also possess baroreceptors of two types: (1) type A, located primarily at the entrance of the pulmonary veins, which discharge at the onset of systole and are not affected by volume; and (2) type B, which demonstrate increased activity with atrial filling and increase atrial size. The neuronal inputs of the cardiac atria and the carotid body are transmitted to the medulla and hypothalamus through cranial nerves IX and X. Atrial transmural pressure has a significant influence on kidney function. Experiments in dogs show that decreased atrial wall distension culminated in decreased sodium excretion and urine flow rate by the kidney. In addition, the normal diuretic response to volume expansion was attenuated when atrial wall distension was limited. Other studies examine the interaction between the carotid body and neuronal control of sodium homeostasis. For example, infusion of hypertonic saline increased neural activity along the tracts, leading to the hypothalamus and medulla. Conversely, increased left atrial stretch decreased activity of these neural tracts, and with a resultant increase in the urinary flow rate and sodium excretion. Plasma levels of atrial natriuretic peptide (ANP) were increased, while AVP and aldosterone levels were decreased. The atria and ventricles have their own baroreceptors that stimulate secretion of ANP and brain natriuretic peptide (BNP) in response to atrial or ventricular distension. ANP release can stimulate up to a 10-fold increase in sodium excretion. ANP and BNP will be discussed in more detail as follows.