Introduction
Physiological and pathological states leading to volume expansion and/or increased tension on left ventricular myocytes engender the release of cardiac natriuretic peptides.Reference Hall1 The most well studied of these, brain natriuretic peptide (BNP), is a 32 amino acid peptide hormone. Biologically, active BNP is the remaining part of a genetically transcribed prohormone that is cleaved to form BNP and N-terminal (NT)-pro-BNP (76 amino acids), both of which can be measured in the circulation by immunoassay.Reference Hall2 Cardiac myocytes constitute the major source of production of these peptides and the main stimulus for peptide synthesis and secretion appears to be myocyte stretch. In contrast to the closely related atrial natriuretic peptides (ANP/NT-pro-ANP), which originate mainly from atrial tissue, BNP and NT-pro-BNP are mainly produced by ventricular myocytes.Reference Hall1
Of clinical interest, ventricular NT-pro-BNP production is up-regulated in cardiac failure and following a myocardial infarction, probably due to increased mechanical stretch.Reference Hall1, Reference Hall2 After secretion, BNP binds to the natriuretic peptide receptor type A, causing increased intracellular cGMP production, thus inducing diuresis, vasodilatation and inhibition of renin and aldosterone production, as well as cardiac and vascular myocyte growth and leading to decreased cardiac preload.Reference Hall1, Reference Hall2 In mice, knockout of the BNP gene leads to cardiac fibrosis, while BNP over-expression leads to hypotension and bone malformation.Reference Kuhn3 BNP is cleared from plasma through binding to the natriuretic peptide clearance receptor type C and appears relatively resistant to proteolysis by neutral endopeptidases.Reference Martinez-Rumayor, Richards, Burnett and Januzzi4 In contrast, NT-pro-BNP is thought to be principally cleared by renal excretion, but these mechanisms await further study.Reference Martinez-Rumayor, Richards, Burnett and Januzzi4, Reference Khalifeh, Haider and Horl5 Consequently, the half-life of BNP is 20 minutes, whereas the half-life of NT-pro-BNP is 120 minutes,Reference Maisel, Bhalla and Braunwald6 leading to approximately 6-fold higher concentrations of NT-pro-BNP than BNP despite their equivalent secretion.
Chronic kidney disease (CKD) and congestive heart failure (CHF) are both diseases primarily of elderly people. While the prevalence of CKD (defined as decreased glomerular filtration rate or persistent albuminuria) is estimated at 37.8% of the Western population above 70 years old,Reference Coresh, Selvin and Stevens7 CHF is thought to be present in more than 5% of the same age group.Reference Cowie, Mosterd and Wood8 Meanwhile, more than 70% of CKD patients starting dialysis are thought to suffer from CHF.Reference Zolty, Hynes and Vittorio9 As there are significant interactions between the heart and kidneys in both health and disease, this significant co-morbidity deserves special attention. An especially important aspect is the marked dysregulation of multiple central physiological axes that occur with CKD and that have important implications for both the detection and risk of developing CHF, as well as its therapy. We will here attempt to review one such aspect, namely the importance of knowing kidney function when interpreting an elevated NT-pro-BNP measurement obtained in order to diagnose or stage CHF.
Search methods
Relevant material was identified using PubMed searches for: N-terminal pro-brain natriuretic peptide (NT-pro-BNP), older age, mortality, prognosis, congestive heart failure, chronic kidney disease, and dialysis. Based on this material, selected references were chosen.
Age and the heart
The structure and function of arteries change throughout life.Reference Yildiz10 Age is thus an important risk factor for hypertension, coronary heart disease, CHF and stroke.Reference Denys, Cankurtaran, Janssens and Petrovic11 Meanwhile, cardiovascular disease (CVD) is the leading cause of morbidity and mortality both in the general population and in chronic CKD patients.Reference Ikizler12 Indeed, the majority of patients with CKD will ultimately die from a CVD event before their CKD progressed to end-stage renal disease (ESRD).Reference Berl and Henrich13 Advanced age itself significantly increases cardiovascular morbidity by promoting the development of atherosclerosis and by impairing normal cellular function.Reference Csiszar, Toth, Peti-Peterdi and Ungvari14 Stiffening of the large arteries is a common feature of ageing and is exacerbated by a number of disorders such as hypertension, diabetes and renal disease.Reference Wang, Keith, Struthers and Feuerstein15 One of the most prominent organs affected by ageing is the kidney. There is evidence that age-associated phenotypic changes may be an important cause of renal failure.Reference Csiszar, Toth, Peti-Peterdi and Ungvari14
Arterial stiffening is recognized as an important and independent risk factor for cardiovascular events. A number of factors are associated with arterial stiffness and may even contribute to it, including endothelial dysfunction, altered vascular smooth muscle cell (SMC) function, vascular inflammation, and genetic determinants, which overlap to a large degree with atherosclerosis.Reference Wang, Keith, Struthers and Feuerstein15
Age and the kidney
A major confounder when evaluating CVD and NT-pro-BNP in older people is kidney function. Indeed, when CoreshReference Coresh, Selvin and Stevens7 examined the risk of CVD and all-cause mortality in an elderly cohort of 3112 individuals between the ages of 68 and 102 years, the prevalence of elevated urinary albumin excretion was 14.3% for ages 68–74 years, 17.1% for ages 75–84 years, and 26.9% among ages 85–102 years. Furthermore, the incidence of CVD and all-cause mortality was twice as high in individuals with microalbuminuria and 3-fold higher in individuals with macroalbuminuria, compared with matched individuals without kidney damage. Importantly, CKD is a major cause of CVD in the elderly population. Alarmingly, the incidence of kidney disease is increasing in older individuals, with an estimated 37.8% of the Western population above 70 years old having CKD.Reference Coresh, Selvin and Stevens7 This age group also represents the most rapidly increasing segment of the ESRD population. Indeed, while CKD is more common in individuals with prevalent risk factors such as diabetes or hypertension, an estimated 11% of those aged over 65 years without any risk factor have stage 3 (glomerular filtration rate) (GFR <30 ml/min) or higher CKD.Reference Coresh, Selvin and Stevens7
The greatly increased risk of CVD observed with declining GFR is well demonstrated in the elderly population. Indeed, while the 3 years cardiovascular event risk in an individual with a glomerular filtration rate (GFR) of 90 ml/min is 15%, it rose to 40% in those with a GFR of 30 ml/min.Reference Manjunath, Tighiouart and Coresh16 Also, declining GFR and increasing albuminuria independently predict the amount of calcified atherosclerotic plaques in the coronary arteries,Reference Freedman, Langefeld and Lohman17 as well as the risk of clinical CVD events.Reference Acton, Go and Roseman18 However, it should be noted that the reverse association is also true; prevalent CVD is associated with a higher incidence of CKD. About one-quarter of patients with coronary artery disease, one-third of patients with acute myocardial infarction and almost half of all patients with congestive heart failure have an estimated GFR of less than 60 ml/min.Reference Anavekar and Pfeffer19–Reference Ix, Shlipak, Liu, Schiller and Whooley21
NT-pro-BNP and cardiac dysfunction
Clinically, circulating BNP and NT-pro-BNP are routinely used to diagnose heart failure, as well as for prognosticating heart failure and acute coronary syndromes.Reference Christenson22 Both BNP and NT-pro-BNP have been shown to be strong predictors of morbidity, mortality and recurrent cardiovascular events independent of other risk factors, including conventional factorsReference Vickery, Price and John23 and biomarkers representing the acute-phase reaction and endothelial activation and even from left ventricular (LV) systolic function estimated by clinical methods (i.e. by echocardiography or contrast ventriculography). Moreover, NT-pro-BNP appears to have a better prognostic ability than BNP.Reference Tagore, Ling, Yang, Daw, Chan and Sethi24 For example, in a substudy of COPERNICUS,Reference Hartmann, Packer and Coats25 consisting of 1011 patients with symptomatic chronic CHF and ejection fraction (EF) <25%, subjects with a NT-pro-BNP above the median (1767 pg/ml) showed a 2.7-fold increased risk of death compared with those below the median.Reference Mayer, Simon, Plaskova, Cifkova and Trefil26, Reference Masson, Latini and Anand27 Also, Masson et al.Reference Masson, Latini and Anand27 report in a direct comparison of BNP and NT-pro-BNP in 3916 patients with chronic and symptomatic heart failure, that NT-pro-BNP was superior to BNP as a predictor of both morbidity and mortality, as well as a better marker of the risk of hospitalizations for heart failure. Interestingly, an elevated NT-pro-BNP is a strong predictor of mortality even in stable patients at least 6 months after an acute coronary event, even in the absence of any clinically apparent heart failure.Reference Kragelund, Gronning, Kober, Hildebrandt and Steffensen28 Population-based studies also suggest that plasma levels of BNP and NT-pro-BNP are useful screening tests for heart failureReference Lamba and Abraham29, Reference Wu30 and asymptomatic LV dysfunction.Reference Khalifeh, Haider and Horl5
While BNP and NT-pro-BNP are thus useful markers of prognosis in CHF, the exact reasons are unclear. First, while myocyte stretch may be the major driver of BNP and NT-pro-BNP production, many other factors also contribute.Reference Green, Green and Januzzi31 Thus elevated BNP and NT-pro-BNP concentrations occur with the presence of LV dysfunction,Reference Januzzi, van Kimmenade and Lainchbury32 abnormalities in heart diastolic function,Reference O'Donoghue, Chen and Baggish33 right ventricular dysfunction,Reference Chen, Wood and Krauser34 valvular heart disease,Reference Weber, Arnold and Rau35 abnormalities in heart rhythm,Reference Morello, Lloyd-Jones and Chae36 an elevated pulmonary artery pressureReference Weber, Arnold and Rau35 and the presence and severity of ischaemic heart disease.Reference Foote, Pearlman, Siegel and Yeo37
NT-pro-BNP and the kidney
BNP and NT-pro-BNP are highly elevated in patients with CKD.Reference Vickery, Price and John23, Reference Khan, Fink, Nass, Chen, Christenson and deFilippi38, Reference Takami, Horio and Iwashima39 Thus in a survey of asymptomatic patients with CKD who did not yet require dialysis, more than half had NT-pro-BNP levels above normal.Reference Tagore, Ling, Yang, Daw, Chan and Sethi24 In patients who had ESRD and received haemodialysis (HD) or peritoneal dialysis (PD), BNP and NT-pro-BNP levels were uniformly increased compared with normal values.Reference Madsen, Ladefoged, Corell, Schou, Hildebrandt and Atar40–Reference Sommerer, Beimler and Schwenger42 The exact reasons for this elevation remain to be elucidated, but one of the major contributing factors is likely to be the very high prevalence of LV structural and functional abnormalities. However, as NT-pro-BNP is known to be cleared by the kidneys,Reference Palmer and Richards43 a gradual increase in circulating levels due to decreased renal metabolism may be expected with progressive loss of kidney function. Furthermore, the kidney is the major organ for excretion of sodium ions and water, both very much implicated in circulating fluid volume and thus cardiac load. It is thus likely that part of the elevation in NT-pro-BNP observed in CKD patients is driven by actual cardiac stress caused by hypervolaemia and impaired fluid balance. As their name implies, the natriuretic peptides strive to increase the excretion of sodium in the kidneys, leading to a greater loss of water due to tubular secretion and subsequently to an amelioration of hypervolaemia.Reference Palmer and Richards43, Reference DeFilippi, van Kimmenade and Pinto44
Clinically, both BNP and NT-pro-BNP levels are strongly correlated with LV hypertrophy and systolic dysfunction in patients with CKD,Reference Zoccali, Mallamaci and Benedetto45 suggesting a physiological link between hypervolaemia, cardiac hypertrophy and the release of natriuretic peptides. For example, in a recent study by Takami et al.,Reference Takami, Horio and Iwashima39 plasma BNP was a reliable marker of LV overload and a powerful predictor of heart failure in non-dialysed patients with CKD.
The elevation of BNP and NT-pro-BNP has also been shown to reflect the presence of myocardial ischaemia in asymptomatic patients with CKD.Reference Khan, Fink, Nass, Chen, Christenson and deFilippi38 Indeed, when a recent study by Satyan et al.Reference Satyan, Light and Agarwal41 compared the prognostic value of NT-pro-BNP with cardiac troponin T (cTnT) in asymptomatic haemodialysis patients, NT-pro-BNP was more strongly correlated with LV systolic dysfunction and both all-cause and cardiovascular mortality. Also, in another study of PD patients, NT-pro-BNP emerged as a more powerful predictor of mortality, cardiovascular death and events, as well as CHF, than high-sensitivity C-reactive protein (hs-CRP). Finally, Zoccali et al.Reference Zoccali, Mallamaci and Benedetto45 found that BNP, but not ANP, was an independent predictor of mortality in CKD patients also after adjusting for LV mass and ejection fraction.
Despite the previously described differences in clearance, no studies have compared the prognostic value of BNP and NT-pro-BNP in the dialysed ESRD population. A head-to-head comparison in non-dialysis patients with CKD somewhat surprisingly showed similar correlations between BNP and NT-pro-BNP with renal dysfunction, LV hypertrophy and coronary artery disease.Reference Wang and Lai46
NT-pro-BNP in elderly people
When interpreting NT-pro-BNP values, age is paramount. A striking gradual increase in mean and upper percentile values for NT-pro-BNP with increased age has been reported in healthy volunteers.Reference Wang, Larson and Levy47 Indeed, after multivariate adjustment, a 10-year increase in age was associated with a 1.4-fold increase in NT-pro-BNP concentration.Reference Wang, Larson and Levy47 The reasons for this increase are manifold and include a decreasing kidney function, but at least in animal studies, advanced age has been shown to lead to a direct increase in natriuretic peptide gene transcription.Reference Raizada, Thakore, Luo and McGuire48 Thus interpreting a measured NT-pro-BNP value in an elderly patient is complex and requires knowledge of a patient's symptoms as well as their renal function. However, luckily it appears that other co-morbidities, such as diabetes and hypertension, have little impact on circulating NT-pro-BNP.Reference Taylor, Christenson, Rao, Jorge and Gottlieb49
As may be expected due to the complexities involved in interpreting clinical measurements, few studies are available on the specific prognostic value of NT-pro-BNP in elderly patients. Drewniak et al.,Reference Davidoff and Powe50 in a small study specifically investigating older people (>65 years), included 161 patients with acute myocardial infarction. NT-pro-BNP concentration was measured within 48 hours of admission to a hospital, together with LV ejection fraction. The authors found a strong relation between NT-pro-BNP concentrations and clinical outcomes at 6 months.Reference Drewniak, Snopek, Zarukiewicz, Borys and Dabrowski51 Also, in multivariate analysis, age and NT-pro-BNP were both independent predictors of 6-month mortality. However, it should be noted that confounding factors, including kidney function, cardiac ischaemia and fluid balance, were not investigated. Clearly studies in large populations of elderly people are needed.
Evaluating NT-pro-BNP in the elderly patient
When presented with an elevated NT-pro-BNP in an elderly (>65 years of age) patient, several factors must be considered. Patient history is of the utmost importance, as it may reveal previously diagnosed kidney or heart disease. Declining kidney function with age is likely to contribute to an elevation in NT-pro-BNP in the elderly patient and recent studies have also reported a decrease in the clearance of natriuretic peptides from plasma even in the absence of renal dysfunction.Reference Clark, Elahi, Shannon, Wei and Epstein52 Also, impairment of non-renal clearance mechanisms, such as clearance through natriuretic peptide receptors in human platelets, probably contributes to elevated concentrations of NT-pro-BNP in older people.Reference Giannessi, Andreassi, Del Ry, Clerico, Colombo and Dini53 Clearly, age-specific cut-offs are needed to correctly interpret an elevated level of NT-pro-BNP or BNP. These cut-offs should probably also take into account the estimated GFR of the patient. The cut-off values for the measurement of NT-pro-BNP by using an electrochemiluminescence assay (Elecsys proBNP; Roche Diagnostics Corp., Indianapolis, IN, USA) to identify patients with normal and high levels of NT-pro-BNP and to rule out structural heart disease are: 125 pg/ml for patients younger than 75 years and 450 pg/ml for patients 75 years and older. Indeed, given the clear correlation between kidney function and NT-pro-BNP levels in the circulation,Reference Das, Abdullah and Leonard54 as well as the deteriorating kidney function with age,Reference Das, Abdullah and Leonard54 it appears obvious that both age and GFR must be taken into account when evaluating measured NT-pro-BNP levels in the circulation to determine the degree of LV stress and fluid overload.
In summary, when utilizing NT-pro-BNP measurements as a marker of cardiovascular disease and volume status and CHF in elderly subjects, kidney function is an important confounder and must be taken into account when interpreting results.
Conflicts of interest
Bengt Lindholm is employed by Baxter Healthcare. Baxter Novum is an initiative funded by an unconditional grant from Baxter Healthcare to Karolinska Institutet.
