Cardiac biomarkers are associated with maximum stage of acute kidney injury in critically ill patients: a prospective analysis

Acute kidney injury (AKI) is a serious complication of critical illness that affects more than 50% of patients admitted to the intensive care unit (ICU) [1, 2]. AKI is independently associated with short and long-term complications, mortality and increased healthcare costs [3]. There is an urgent need to identify patients who are at risk of developing AKI in order to implement renoprotective strategies and avoid nephrotoxic exposures. Numerous novel biomarkers for AKI have been tested in critically ill patients. Although several markers have been found to indicate the onset of AKI before serum creatinine increases, none has been integrated into routine clinical practice [4].

A few studies have explored the role of cardiac biomarkers in predicting AKI, the rationale being that cardiac and renal function are closely linked (often referred to as cardio-renal syndrome [5]). For instance, Thiengo et al. [6] studied 29 ICU patients with incident sepsis and concluded that cardiac troponin I (cTnI) on admission predicted the development of AKI and the need for acute renal replacement therapy (RRT). Similarly, a study in 34 consecutive ICU patients showed that patients with AKI on presentation or during the ICU stay had significantly higher levels of the cardiac biomarker brain natriuretic peptide (BNP) relative to patients without AKI [7].

These studies were relatively small, and larger studies are required to confirm the findings. Our aim was to explore the relationship between the cardiac biomarkers cTnI, cardiac troponin T (cTnT) and serum N-Terminal pro-brain natriuretic peptide (NT-proBNP) and the development of AKI and the need for RRT in a larger patient cohort than previously undertaken.

We performed a sub-analysis of data that were collected previously for a prospective observational study aimed at exploring the epidemiology of troponin elevation in critically ill patients admitted to the ICU for non-cardiac reasons [8].

Of 144 patients recruited in the original study [8], six were excluded due to pre-existing ESRF. As a result 138 patients were included in the analysis, of whom 73 (53%) developed AKI. Thirteen patients (9.4%) had maximum AKI stage 1, 12 patients (8.7%) had AKI stage 2 and 48 patients (35%) had AKI stage 3, of whom 40 were treated with acute RRT. The median time between the day of admission to the ICU and the day of the worst AKI stage was 3 days.

Patients who developed AKI were significantly older, had higher APACHE II and SOFA scores on admission to the ICU, were more likely to have sepsis and had higher serum bilirubin and lactate concentrations than those without AKI (Table 1).

The key finding of this study was that analysis were that in critically ill patients admitted for non-cardiac reasons, NT-proBNP on admission to the ICU was independently associated with the maximum stage of AKI, including the need for RRT. The independent association persisted after excluding patients who subsequently developed an acute MI during their stay in the ICU. cTnT was independently associated with AKI but not with need for RRT. cTnI was not an independent risk factor.

NT-proBNP had the strongest association with worst stage of AKI and need for RRT. NT-proBNP is a polypeptide secreted by the ventricles in response to excessive stretching of cardiomyocytes and its role is to promote natriuresis [12]. The link between ventricular dilatation, raised central venous pressure (CVP) and renal dysfunction is well established [13, 14] (Fig. 1). In patients with a raised CVP, the increased backward pressure propagates evenly in all districts of the venous system, including renal veins. As a result, renal congestion sets in and glomerular filtration and sodium excretion decrease [15‐19]. Physiologically, it makes sense that NT-proBNP, a biomarker of ventricular dilatation, is associated with the development of AKI.

In patients with cardiac disease, BNP has been extensively investigated and found to have added predictive value for the development of AKI in patients with an ST-elevation MI or heart failure [20, 21]. BNP has also been described as a marker of the ‘cardio renal burden’ in patients admitted with a cardiac emergency and as a risk factor for the development of AKI after cardiac surgery [22‐24].

To the best of our knowledge, the role of BNP as a biomarker for AKI in critically ill patients presenting without an acute cardiac emergency is limited to two studies. de cal et al. [7] conducted a prospective trial in 34 ICU patients admitted with a non-cardiac diagnosis and concluded that BNP levels predicted the development of AKI. However, there was no adjustment for age and severity of illness. Chou et al. [25] measured BNP on admission to the ICU and 24 hours later in 163 critically ill patients, and showed that changes in BNP between the day of admission and 24 hours later predicted the development of AKI, renal recovery and mortality. Again, there was no adjustment for other important risk factors, including severity of illness and underlying cardiac risk factors.

The role of BNP in sepsis has also been studied [26, 27]. Papanikolaou et al. [26] measured BNP concentrations in patients with sepsis and showed that BNP levels were raised in sepsis and septic shock, presumably as a result of pro-inflammatory cytokine release and biventricular dysfunction. However, the impact on renal function was not assessed.

To our best knowledge this is the largest study which demonstrates an independent association between NT-proBNP concentration on admission to the ICU and worst stage of AKI, including requirement for acute RRT. Our conclusions are strengthened by the fact that the association was maintained after exclusion of patients who developed an acute MI during their stay in ICU and after adjustment for cumulative fluid balance. If confirmed in future studies, NT-proBNP could serve as an easily available ‘alert’ for AKI and RRT in routine clinical practice.

However, it is important to acknowledge some potential limitations. As a single-centre retrospective study, the impact of unrecorded confounding factors cannot be excluded. This is particularly important when interpreting the association between NT-proBNP and acute RRT because there is no consensus about the optimal time of starting RRT and high variation in clinical practice. Second, we did not perform routine echocardiography in all patients to correlate the NT-proBNP results with ventricular dilatation and acknowledge that other mechanisms beyond ventricular stretch may stimulate NT-proBNP release [28]. Third, pre-existing creatinine results were only available for 54 patients (39%). In the remaining 84 patients (61%), baseline renal function was estimated by back-calculation using the MDRD formula. Although this method is supported by the KDIGO working group [9], we recognize that there are limitations, including misclassification of AKI [17]. Fourth, we defined sepsis according to the previous consensus criteria because our study was conducted before the new sepsis criteria were published [29]. Fifth, we calculated cumulative fluid balance in the first 24 hours in the ICU but did not include the fluid balance prior to ICU admission. Finally, the exact aetiology of AKI was not always documented by the treating clinical team. Despite these limitations, we believe that our findings are important and should prompt further research to verify our results.

In patients without a cardiac diagnosis on admission to the ICU, admission NT-proBNP concentrations were independently associated with worst stage of AKI, including need for acute RRT. The cTnT and cTnI results on admission were associated with maximum stage of AKI but not with need for RRT. Larger studies are needed to evaluate the potential of NT-proBNP as a biomarker for AKI and RRT.