The coexistence of heart and kidney disease represents an evolving epidemic in an ageing population, with chronic kidney disease (CKD) incidence rates increasing annually.1 The effect of dysfunction of one kidney mediates profound effects on the other, recently described and defined as five distinctive cardiorenal subtypes by the Acute Dialysis Quality Initiative consensus group.2 Much work has been done focusing on the possible mechanisms of this bidirectional cardiorenal axis, aided in part by the discovery and utilisation of several novel biomarkers, representing early injury to different parts of the nephron. Furthermore, new and invaluable insights have also been gained into the further roles novel biomarkers, such as NGAL, may play in controlling iron trafficking and renal tubular epithelial apoptosis during ischaemic and nephrotoxic insults, thus potentially providing new therapeutic options in addition to diagnostic and prognostic utility.
Acute and chronic kidney diseases are associated with multiple effects on the cardiovascular system, including activation of the renin– angiotensin–aldosterone system (RAAS) with volume overload, sympathetic over activation, and systolic and diastolic heart failure (HF) in association with adverse cardiomyocyte remodelling. CKD also mediates systemic dyslipidaemic effects, including increased in low-density lipoprotein (LDL) cholesterol and reduced high-density lipoprotein (HDL) cholesterol, resulting in progressive atherosclerosis and vascular calcification, particularly in patients with end-stage renal disease (ESRD). CKD is thus a strong independent predictor of increased morbidity and adverse clinical outcomes, particularly in the setting of cardiorenal patients with acute coronary syndromes, undergoing cardiac surgery, or with congestive heart failure. For patients with (ESRD), it is estimated that the risk of cardiovascular mortality is 500-fold greater than age-matched controls with preserved renal function.3 Freeman et al.4 showed that patients with renal insufficiency and acute coronary syndromes were less likely to undergo both diagnostic angiography, and percutaneous intervention (p<0.001). This subset of patients also had longer hospital stays and higher rates of in-hospital mortality (8.1 % versus 2.6 %, p<0.001),4 and received less glycoprotein IIb/IIIa receptor antagonists. Similar findings were reported by Beattie et al.5 in patients with ST-elevation myocardial infarctions, who found that as renal function decreased, there was a graded increase risk for occurrence of pulmonary oedema, arrhythmias, and cardiogenic shock, which also represented the mechanisms by which mortality hazard was increased. However, despite their higher risk, the use of beta-blockers, primary angioplasty, and thrombolysis was lower in those with decreased estimated glomerular filtration rates (eGFR), perhaps in part explaining their poor outcomes.
Similarly, in acute and chronic heart failure patients, subgroup analysis of the Acute decompensated heart failure national (ADHERE) registry6 has identified cohorts of patients with chronically impaired renal function (with an incidence of 30 % of the population studied) who received less treatment with angiotensin-converting enzyme (ACE) inhibitors, beta-blockers, and aspirin, highlighting the management challenges faced with these complex cardiorenal patients.
Other authors have also reported adverse outcomes using any measure (hospitalisation, dialysis, readmission and mortality), in patients with AKI and acute HF, traditionally defined as “worsening renal function” (a serum increase ≥0.3 mg/dL during hospitalisation), but also with serum creatinine increases from as little as ≥0.1 mg/dL.7–9 The early diagnosis of AKI in multiple clinical settings (critical care, cardiac surgery, radiocontrast-requiring procedures, acute HF) has been impaired by the inherent limitations of serum creatinine trends in the timely detection and quantification of renal tubular injury, providing one explanation for the failure to develop effective therapies to prevent or ameliorate AKI. In summary, because the diagnosis and thus the treatment is delayed until renal tubular injury is severe enough to cause loss of renal function (and serum creatinine elevation), the window of opportunity for effective AKI therapy has been missed; akin to initiating therapy for acute myocardial infarction when cardiogenic shock has developed. Newer AKI classification systems, such as the Risk, injury, failure, loss, end-stage renal disease (RIFLE) criteria have made improvements towards earlier, standardised diagnosis and staging of AKI. Promising biomarkers of renal injury with improved sensitivity and specificity provide improved tools for the early diagnosis and assessment of AKI in combination with validated AKI classification systems. NGAL is a 25 kDa glycoprotein with bacteriostatic siderophore-binding properties, first discovered in the supernatant of activated neutrophils.10 It is involved in the regulation of cellular apoptosis, transport of iron across cell membranes, and also forms complexes with MMP-9 to attenuate its action in response to inflammatory and interleukin stimuli.11 It has been isolated from renal tubular cells, cardiomyocytes, epithelial and endothelial cells, as well as hepatocytes. NGAL is known to have multiple ligands, the principal ones being hepatocyte growth factor (HGF) and MMP-9, with its specific receptor 24p3R.10 NGAL has shown enormous promise as an early biomarker of renal tubular damage, with superior diagnostic and predictive abilities when compared to current standard markers of renal function (SCr, eGFR, Cystatin C).12Table 1 summarises some of the key cardiorenal studies of the predictive value of NGAL in AKI.
Mishra et al. studied 71 children undergoing cardiopulmonary bypass (where the timing of the event causing renal injury was known), examining the performance of both serum (sNGAL) and urine NGAL (uNGAL) to diagnose acute ischaemic renal injury. In this unique population, with no evidence of CKD, atherosclerosis, diabetes, or exposure to potentially nephrotoxic drugs, uNGAL (using a cutoff value of 50 μg/L) at two hours post-cardiopulmonary bypass was the strongest predictor of AKI by multivariate analysis (area under receiver operating characteristic curve [AUC] 0.99, sensitivity 1.0, specificity 0.98).13 Using urinary NGAL, the diagnosis of AKI was made up to two days before significant changes in serum creatinine occurred. NGAL has also shown to be predictive of AKI in a variety of other clinical settings, ranging from the intensive care unit (ICU) to renal transplant patients, as well as having prognostic value for clinical outcomes such as mortality or initiation of renal replacement therapy.14
Neutrophil Gelatinase-associated Lipocalin Expression in Pre-renal Azotemia
Available data suggest that renal tubular expression of NGAL and circulating and urinary NGAL levels do not increase in patients with pre-renal azotemia (defined as AKI caused by reversible renal hypoperfusion, without tubular injury) in experimental models15 and hospitalised patients.16,17 NGAL appears to be a sensitive distal nephron injury marker (expression increases in the loop of Henle and collecting ducts), but urinary levels also probably reflect proximal tubular injury (because there is constitutive proximal tubular reabsorption of filtered NGAL).18 Defining mild pre-renal azotemia as a change in Scr of 0.3 mg/dL, Paragas et al. found NGAL not to be expressed in any organ of NGAL reporter mice, particularly in the kidney, highlighting the potential differences in NGAL expression in AKI and volume depletion.15 In a diverse group of patients hospitalised following presentation to an urban emergency department, Nickolas and colleagues found that urine NGAL levels distinguished accurately between adjudicated cases of CKD, prerenal-azotemia, AKI with renal tubular injury (acute tubular necrosis-ATN) and normal renal function, better than presenting serum creatinine values. Thus, NGAL may have an important role in helping clinicians treating HF to differentiate between hypovolemia or CHF causing pre-renal azotemia, versus ATN with tubular injury, or CKD.
Conversely, in patients admitted with acute HF, sNGAL levels are associated with an increasing likelihood of developing subsequent worsening renal function (SCr increase from baseline ≥0.3 mg/dL) during the first five days of admission, particularly in patients with normal renal function at baseline.12 In the study by Damman et al., where chronic systolic HF patients had a controlled temporary withdrawal of diuretic therapy, the authors found that uNGAL or sNGAL levels did not rise significantly, whereas urinary KIM-1, N-acetyl-beta- D-glucosaminidase (NAG), and brain natriuretic peptide (BNP) did.17 It is unclear what the significance of these biomarker responses to changes in diuretic therapy and volume status is; an ongoing prospective study of serial NGAL measurements in patients hospitalised to receive IV diuretics with AHF will provide important insights in this syndrome Acute kidney injury neutrophil gelatinase-associated lipocalin (N-GAL) evaluation of symptomatic heart failure study (AKINESIS) ClinicalTrials.gov (NCT 01291836).
Acute Kidney Injury in Patients with Cardiorenal Syndromes
AKI occurs in up to 40 % of patients hospitalised for acute decompensated heart failure (type 1 cardiorenal syndrome), and has a major impact on prognosis, in-hospital outcomes, and morbidity.19,20 Using the definition of worsening renal function (WRF; increase in serum creatinine >26.5 μmol/L from baseline), incidence estimates range from 9 to 19 % for acute decompensated heart failure (ADHF) occurring due to acute coronary syndromes (ACS).21 Renal injury occurring during cardiac surgery with cardiopulmonary bypass (CPB) can complicate up to 30–50 % of cases, thus significantly contributing to increased adverse outcomes,22 including requirements for renal replacement therapy (RRT) and worsening adjusted mortality.
Approximately 1–2 % of patients undergoing conventional coronary bypass grafting require RRT after surgery,23 with this incidence increasing for patients with underlying impaired preoperative eGFR, or for those undergoing heart transplantation or left ventricular assist device implantation. Despite the use of off-pump coronary artery bypass graft surgery and advances in perfusion techniques, such as minimised extracorporeal circulation systems, rates of AKI have remained relatively unchanged. The mechanisms for AKI in this setting are complex and multifactorial, including renal ischaemia/reperfusion, oxidative stress, complement activation, systemic inflammation, and haemolysis with haemoglobinuria.24
Free, labile, iron release catalyses the production of toxic hydroxyl radicals, via the Haber–Weiss, and Fenton reactions, which mediate renal tubular injury. In this environment, the role of NGAL as a siderophore-binding lipocalin, both in its bound and unbound form, may be vitally important to prevent free iron catalyzing generation of reactive oxygen species, serving a protective role in addition to being a marker of tubular injury caused by free iron-mediated oxidant stress.
For patients receiving iodinated contrast media, the development of contrast-induced AKI (CI-AKI) is associated with adverse in-hospital outcomes and increased risk of death.25 In 439 patients with CKD (SCr ≥158 μmol/L) undergoing percutaneous coronary revascularization, 37 % of patients developed CI-AKI, as defined by an increase in SCr ≥25 %. The in-hospital mortality rate for the AKI group was three times higher than those without AKI (14.9 % versus 4.9 %), with an almost twofold increase in one-year mortality.26 Contrast-induced AKI has been associated with longer hospital stays,27 and requirement for renal replacement therapy in 1–4 % of patients, depending on the presence or absence of underlying renal impairment, or the type of contrast used (high or low osmolar versus isosmolar contrast media).28 In a retrospective study of 7,586 patients undergoing percutaneous coronary intervention (PCI), contrast-induced AKI was associated with baseline renal impairment, the presence of acute myocardial infarction, haemodynamic instability, and the volume of contrast administered.23
The proposed consensus cardiorenal syndrome (CRS) definitions2 characterise conditions leading to AKI, such as CPB or CI-AKI, as a type of acute reno-cardiac syndrome – type 3 CRS. In this setting, the development of renal impairment may result in neurohormonal activation, sodium retention, and volume overload, as well as an increased risk of myocardial ischaemia and arrhythmias. In the study by Rihal et al., myocardial infarction was the ultimate cause of death in 69.6 % of patients having suffered contrast-induced AKI, versus 59.8 % in those with no AKI.23
Type 4 CRS defines the impact CKD has on the heart-kidney interaction, best highlighted by those patients with CKD (stages 1–5) and acute coronary syndromes (ACS) or congestive cardiac failure. Across the spectrum of prospective and retrospective studies of both CKD and ESRD, the incidence of ischaemic heart disease is high, while also associating a poor prognosis for those with CKD who have acute ischaemic events.
The reasons for this are multifactorial, relating not only to dysregulated vascular calcification, endothelial dysfunction, and circulating inflammatory mediators, but also to more conservative therapeutic strategies compared to those with no evidence of CKD or AKI. In a prospective series of 4,758 patients with diabetes mellitus and ACS, outcomes were assessed and compared with the subgroup of patients with CKD (34.8 %). Striking differences were apparent between the groups, with only 15.7–35.7 % of the CKD group receiving beta-adrenergic antagonists, and 21.1–27.7 % receiving angiotensinconverting inhibitors. For those with severe CKD, 82 % of patients were treated with medical therapy alone, with 14 % undergoing PCI, and only 3.9 % proceeding to surgical revascularisation. Within the CKD group, those who received PCI had significantly better survival rates (59.0 versus 44.3 months, p<0.0001).29
Simultaneous cardiorenal dysfunction due to systemic illness, such as sepsis or diabetes, comprises type 5 CRS. In the Sepsis occurrence in acutely ill patients (SOAP) study by Vincent et al., examining critically ill patients in European intensive care units, positive fluid balance in patients with sepsis was one of the main predictors of mortality.30 In patients with severe sepsis, cardiovascular organ failure occurred in 62.6 % of patients, with 42.3 % mortality. The role of specific biomarkers in this setting is less well defined, with cardiac troponins and BNP showing some association with adverse outcomes in acute ‘secondary’ cardiorenal syndrome.
Neutrophil Gelatinase-associated Lipocalin as a Biomarker for Progression of Chronic Kidney Disease
The role of NGAL as a risk marker for progression of CKD has been demonstrated in several recent studies, showing inverse and independent relationships to eGFR, as well as showing an association with adverse outcomes. Furthermore, this relationship exists regardless of the primary cause of CKD, underscoring the increasingly appreciated important role that injured renal tubules play in the final common pathway of progression towards ESRD. In 96 patients with CKD of mixed aetiology (such as diabetic nephropathy, biopsy-confirmed glomerulonephritis, and autosomal polycystic kidney disease), significant inverse correlations were seen with both sNGAL and uNGAL, when compared to eGFR (uNGAL R= –0.41, p<0.0001, sNGAL R= –0.44, p<0.0001), as seen in Figure 1.15 In direct comparison for predicting progression of CKD, uNGAL was superior to eGFR (AUC 0.78 versus 0.64). Using multivariate analysis, both sNGAL and uNGAL also predicted increased risk of CKD progression, independent of both eGFR and advancing age.31 In patients with autosomal dominant polycystic kidney disease, sNGAL and uNGAL similarly correlated with severity of CKD.32
Ding et al.33 examined a cohort of 70 patients with IgA nephropathy (IGAN), showing a clear correlation between uNGAL levels and the degree of tubulointerstitial disease, as demonstrated by histopathologic findings from renal biopsies (tubular atrophy, interstitial fibrosis, and inflammatory cell infiltrates). Importantly, uNGAL was also raised up to three times higher than in the controls, in patients with earlier stages of disease (Lee Grade II IGAN), where more traditional markers such as urinary N-acetyl-glucosaminidase (NAG) were still within normal levels. Thus, there is increasing evidence that NGAL is a potentially useful marker of CKD, reflecting ongoing tubular inflammation and injury.
Neutrophil Gelatinase-associated Lipocalin in Atherosclerosis
Beyond the renal tubular epithelium and the neutrophil, NGAL has also been shown to play a role in modulating matrix metalloproteinase-9 (MMP-9), itself associated with both large and small vessel atherosclerosis and plaque stability.34,35 Galis et al.36 showed that MMP-9, a member of the endopeptidase family, plays a key role in proteolytic activity in the plaque, thus contributing to collagen degradation, plaque instability and risk of rupture. Furthermore, high levels of serum MMP-9 are independently associated with adverse cardiovascular outcomes.37 Using a rat carotid injury model, Bu et al. were able to demonstrate induction of intimal smooth muscle cell NGAL expression by an acute vascular inflammatory regulator (NF-κB) in response to angioplastic injury, as well as providing data to suggest that NGAL is also up-regulated in the chronic inflammation associated with atherosclerosis.38 Using an NGAL reporter mouse model to replicate in vivo cellular injury, the use of novel NF-κB inhibitors was also able to inhibit NGAL activity, highlighting the importance of this NF-κB pathway for NGAL regulation.15 The findings by Bu et al. may provide some insight into the outcomes of a human risk factor study carried out a decade earlier. In a series of 156 middle aged patients with early atherosclerosis, plasma NGAL levels (as well as other markers of systemic leucocyte activation-TNF-α, sTNFR-1) correlated with diastolic blood pressure (r=0.22, p<0.005), as well as age and gender, suggesting a possible link between levels of systemic leucocyte activation and risk factors for atherosclerosis.39 Another study by Malysko, in patients with ischaemic heart disease, also found correlations with serum NGAL and leucocyte count, as well as that with NGAL and advancing age in healthy controls.
In keeping with the above findings, there is evidence that in stable CAD patients compared to controls, interleukin-8 (IL-8) levels are increased in response to skin blister models of localized inflammation (designed to simulate atherosclerotic plaques). With raised serum IL-8 levels, circulating neutrophils also produce increased levels of MMP-9/NGAL complexes at sites of intermediate inflammation, compared to controls with no CAD, suggesting that migrating neutrophils have altered activation patterns in patients with chronic CAD,34 with NGAL playing a key role.
Intraluminal thrombus isolated from abdominal aortic aneurysms (AAA) also produces Il-8 via activated neutrophils contained within the luminal layer of the thrombus, where MMP-9/NGAL complexes, along with other chemotactic factors, are also found.35 Thus NGAL, in association with its MMP-9 ligand, also plays an important role in vascular remodeling, proteolytic enzymal activity, and possible aneurysmal progression, in patients with AAAs.
Neutrophil Gelatinase-associated Lipocalin and the Myocardium
In the setting of the failing myocardium, there is increasing evidence that activation of the immune system, particularly involving neutrophils, cardiomyocytes, and fibroblasts, is associated with increased myocardial NGAL mRNA expression and protein concentrations. Detectable sNGAL has been associated with increased mortality and higher NYHA classes in several studies of chronic heart failure patients (Figure 3),40–42 and may signify a cardiomyocyte injury-induced mechanism for NGAL release, as well as a more dynamic cardiorenal signaling interaction than previously appreciated. In a small preliminary study, 46 chronic HF patients showed elevated sNGAL levels compared to healthy age-matched controls (458.5 [62.5–1212.4] versus 37.8 [15.9–46.5] ng/mL; p=0.0001), with NGAL levels >783 ng/mL being associated with increased two-year mortality (HR 4.08 95 % CI 1.29–12.96).42 In support of these hypotheses, NGAL release was also associated with raised levels of cardiac stress biomarkers (NT-proBNP, BNP)40 and cardiomyocyte troponin release, when myocardial perfusion defects were induced using nuclear perfusion stress testing, particularly when patients had evidence of baseline left ventricular dysfunction.43
Further HF studies have focused specifically on the correlation between NGAL, as a marker of tubular damage, and markers of glomerular filtration rate (eGFR), urinary albumin excretion (UAE), and SCr. Damman et al.41 showed uNGAL levels were raised compared to controls, with significant relationships between uNGAL and eGFR (r=–0.29, p=0.002), SCr (r=0.26, p=0.006), UAE (r=0.33, p=0.001), and also NT-proBNP levels (Figure 2). Moreover, raised sNGAL levels in patients with differing HF aetiologies (ischaemic, cardiomyopathic) do not appear to vary significantly.40,41
Examining models of inflammatory heart failure due to myocarditis and ischaemia, in both rats and humans, lipocalin2/NGAL is also strongly expressed in cardiomyocytes and hepatocytes, and is induced by interleukin-1 (IL-1).40,44 Immunostaining analyses in the study by Ding et al. have demonstrated evidence of NGAL-specific receptor (24p3R) expressing properties on cardiomyocytes,44 with corresponding increases in sNGAL. Further to these findings, NGAL immunostaining in the myocardium of humans and rats post-myocardial infarction, is especially pronounced at the ischaemic/non ischaemic border zones, with NGAL activity also seen in endothelial and vascular smooth muscle cells.40 Thus, under certain conditions of cardiac injury, current findings support the role of NGAL in apoptosis and extracellular iron regulation in the heart, by reducing oxidative stress and extracellular free radical formation. As our knowledge of iron homeostasis in the failing myocardium improves, it is clear that the role of NGAL will evolve as both an indicator of cardiomyocyte injury, and a possible therapeutic target.
Summary and Conclusions
The role of NGAL in AKI and renal disease is rapidly evolving, as both an early diagnostic tool for AKI, as well as a prognostic marker for adverse outcomes. NGAL may also have utility for the prediction and assessment of progression of CKD. Multiple studies in common cardiorenal syndromes have demonstrated the diagnostic utility for NGAL in the setting of acute heart failure, cardiac catheterisation/radiocontrast administration, as well as acute coronary syndromes and cardiac surgery. Ongoing research highlights our increasing understanding of the role of neutrophil-mediated immune activation and chemotaxis in the vascular system and myocardium, and how NGAL may play a role in these cardiovascular pathological processes. Using a clinical framework for defining cardiorenal syndromes, translation of ongoing research will permit us to use this novel biomarker to improve cardiovascular and cardiorenal risk stratification and management in a growing population of complex, high-risk patients.