B-Type Natriuretic Peptide in the Critically Ill with Acute Kidney Injury

Introduction. Acute kidney injury (AKI) is common in the intensive care unit (ICU) and associated with poor outcome. Plasma B-type natriuretic peptide (BNP) is a biomarker related to myocardial overload, and is elevated in some ICU patients. There is a high prevalence of both cardiac and renal dysfunction in ICU patients. Aims. To investigate whether plasma BNP levels in the first 48 hours were associated with AKI in ICU patients. Methods. We studied a cohort of 34 consecutive ICU patients. Primary outcome was presence of AKI on presentation, or during ICU stay. Results. For patients with AKI on presentation, BNP was statistically higher at 24 and 48 hours than No-AKI patients (865 versus 148 pg/mL; 1380 versus 131 pg/mL). For patients developing AKI during 48 hours, BNP was statistically higher at 0, 24 and 48 hours than No-AKI patients (510 versus 197 pg/mL; 552 versus 124 pg/mL; 949 versus 104 pg/mL). Conclusion. Critically ill patients with AKI on presentation or during ICU stay have higher levels of the cardiac biomarker BNP relative to No-AKI patients. Elevated levels of plasma BNP may help identify patients with elevated risk of AKI in the ICU setting. The mechanism for this cardiorenal connection requires further investigation.


Introduction
Acute kidney injury (AKI) is a common clinical problem in intensive care unit (ICU) patients and independently predicts poor outcome [1][2][3][4]. In the ICU setting, the overall incidence of AKI is approximately 36% [5,6], and an increasing trend has been reported [7,8]. Cardiac dysfunction is also common in patients with AKI in the ICU, and increasing interest exists in how the interaction of these two systems affects clinical outcomes in this group of patients.
However, it remains unclear whether plasma levels of BNP are useful in predicting AKI in critically ill patients. Therefore, our study aimed to investigate whether BNP levels in the first 48 hours may be useful in diagnosis of established AKI.

Patients and Study Protocol.
We studied a cohort of 34 consecutive patients admitted to the ICU of "San Bortolo" Hospital, Vicenza, Italy, between December 2007 and April 2008. Patients requiring mechanical ventilation and admitted on any day from Monday to Wednesday were included, and we excluded patients with acute coronary syndrome or acute myocardial infarction. The study was approved by the Institutional Review Board. The primary outcome was presence of AKI during admission or development of AKI during ICU stay. Patients were classified as having AKI if any time during the first 48 hours after enrollment they had (1) an increment of serum creatinine (SCr) of 0.3 mg/dL or more or an increase of at least 50% from baseline and/or, (2) an episode of urine output less than 0.5 mL/kg/hr for more than six hours despite fluid challenge of 500 mL or more.
AKI was classified according to the RIFLE (Risk, Injury, Failure, Loss, and End-stage kidney disease) criteria [21]. The Sequential Organ Failure Assessment (SOFA) scores were calculated using standard methods [22]. Blood samples for plasma BNP and renal function were taken within 6 hours from admission and 24 and 48 hours later, to investigate association of BNP levels with clinical and laboratory parameters and SOFA score. Normally or near normally distributed variables were presented as means ± standard deviations (SD); non-normally distributed continuous data were presented as medians and interquartile ranges (IQR). Differences between groups were analyzed using Student's t and Mann-Whitney tests as appropriate. Differences between repeated measures within a group were analyzed using Friedman test. Statistical analysis was performed using the SPSS 15.0 (SPSS Inc, Chicago, Ill, USA). A P value < .05 was considered significant.

Results
During the study, a total of 34 patients were admitted to the ICU, and, of these, 26 met the inclusion criteria and had sufficient data for analysis. Furthermore, 9 (34.6%) fulfilled criteria for AKI, 5 (19.2%) had AKI on admission, and 4 (15.4%) more developed AKI during 48 hours. Characteristics of the patients are shown in Tables 1 and 2. Given the differences between SCr, age, and BNP at baseline, we did examine for correlations between these variables. While baseline SCr and BNP were not significantly correlated (r = 0.27, P = . 19), there was a weak correlation between  age and SCr (r = 0.39, P = .048) and a stronger correlation between age and baseline BNP (r = .46, P = .02).
An additional 4 patients developed AKI after admission. These patients had on admission higher SCr (1.14 versus 0.82 mg/dL) and BNP (338 versus 197 pg/mL) levels compared to no-AKI patients.
We also analyzed levels of SCr and BNP in all patients developing AKI at any point during 48 hours (n = 9). For these patients, the difference in BNP versus no-AKI patients at admission was even more pronounced (510, IQR 232-832, versus 197, IQR 36-353 pg/mL; P = .038). Also, for the 9 patients developing AKI at any time during ICU stay, SCr and BNP levels at baseline and at 24 and 48 hours were significantly higher compared to no-AKI patients (Table 4).

Discussion
A large proportion of patients admitted to hospital, especially in the critical care setting, have various degrees of heart and kidney dysfunction [23]. Primary disorders of one of      these two organs often result in secondary dysfunction or injury to the other [24]. Such pathophysiological interactions represent the basis for a clinical entity often referred to as the cardiorenal syndrome (CRS) [25]. Limited data are available regarding the diagnostic and prognostic utility of BNP in patients with AKI in intensive care unit. In a recent study, Park et al. [26] demonstrated that BNP levels have the diagnostic and prognostic capability for CRS type 4 in ICU patients, according to the novel classification of CRS [27,28]. In our study, BNP was able to predict the presence of AKI on admission or development of AKI during ICU stay with a ROC-AUC 0.830 (Figure 1). No previous studies have focused on the significance of BNP in patients with AKI admitted to the ICU.
In ICU setting, emerging cardiac and renal impairment are strongly connected on neurohormonal basis via reninangiotensin-aldosterone system BNP and nitric oxide, the sympathetic nervous system and other pathways such as coagulation and inflammation. Burchill et al. have shown in experimental animal models that the acute effects of AKI on the heart occur as early as few hours after kidney injury, and that changes in cardiac structure are associated with increased cardiac BNP [29].
It could also be speculated that the changes in BNP observed in our study may partly reflect the pathophysiology between kidney and heart in AKI, the so-called CRS type 3 or acute renocardiac syndrome. In this category, AKI is believed to be the primary inciting factor, and cardiac failure is a common and in often times a fatal complication of AKI [27].
In our study, we demonstrate a dynamic interaction between AKI and plasma BNP levels in a cohort of mechanically ventilated ICU patients who were admitted primarily for noncardiac diagnosis. This calls attention to the possible utility of this marker in detecting AKI. Further, the results support the need for additional study of the potential value of plasma BNP levels in discrimination between AKI and no-AKI in critically ill patients.
To our knowledge, this is the first investigation of the association between plasma BNP levels and AKI in critically ill patients. We acknowledge some limitations in this study. Extensive information regarding patient comorbidities was not available and could not be added to our analysis. Plasma BNP levels can be affected by other variables such as age, and in our study AKI patients were significantly older than no-AKI patients, and age was correlated with baseline BNP, hence some of the association between BNP and AKI may have reflected the risk of AKI related to age. Additionally, while patients with acute myocardial infarction and acute coronary syndromes were excluded, massive information about previous or current cardiac dysfunction was not collected and could have influenced plasma BNP levels.
Furthermore, we did not perform objective assessment of cardiac function to document that increased BNP in this setting would be due to myocardial dysfunction. Finally, this study was not designed to look at the prognostic value of plasma BNP levels in critically ill patients with AKI. We hope that these preliminary results will encourage further study of these important questions.

Conclusion
In this pilot study, we have demonstrated for the first time an association between plasma BNP levels and AKI in critically ill patients. Patients with AKI have higher levels of BNP compared to no-AKI patients, and in AKI patients BNP levels continue to increase during the subsequent 48 hours. Our results suggest that plasma BNP may distinguish the occurrence of AKI. Additional studies are necessary to confirm our findings and to further shed light on the pathophysiologic interaction between kidney and heart during AKI.