SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References
  9. Supporting Information

ACADEMIC EMERGENCY MEDICINE 2012; 19:18–23 © 2012 by the Society for Academic Emergency Medicine

Abstract

Objectives:  Worsening renal function in patients admitted with heart failure is associated with increased morbidity. These changes are not usually apparent initially and often take up to 48 hours to be detected. Using the novel technique of metabolomic analysis, this study aims to determine if markers of renal injury are identifiable at presentation that are associated with the development of worsening renal function in high-risk patients with heart failure.

Methods:  A prospective exploratory study enrolled a convenience sample of patients with suspected heart failure. Eligible patients had to be older than 18 years, have a B-type natriuretic peptide (BNP) level over 100 pg/mL, have a history of diabetes or hypertension, meet Boston criteria for heart failure (>8), and require hospital admission as judged by the treating physician. Patients receiving no more than one dose of diuretic prior to enrollment were excluded. Urine was collected during the emergency department (ED) stay. Initial creatinine and the peak value between 24 to 48 hours were used to determine worsening renal function as defined by a change of >0.3 mg/dL or absolute 25% increase. Urine samples underwent gas chromatography/mass spectrometry (GC/MS) profiling. Peak metabolite values were measured and data were log-transformed. Partial least squares-discriminant analysis (PLS-DA) was used to identify metabolites associated with worsening renal function. Specific urinary metabolites were ranked based on their regression coefficients.

Results:  The 24 enrolled subjects had a median age of 58 years (interquartile range [IQR] = 49.5 to 67.5 years) with 58% being male. Worsening renal function occurred in 10 subjects (41.7%). A total of 156 metabolites were identified. The optimal number of metabolites for class discrimination as determined by PLS-DA was three, with a classification accuracy of 78%. These metabolites were taurine, sulfuric acid, and talose.

Conclusions:  Urinary metabolites found at the time of presentation may be markers of early renal injury. It is therefore possible that the process of renal injury is initiated prior to ED arrival in patients with suspected heart failure, and these may be used to identify a high-risk patient population.

Worsening renal function commonly occurs during the inpatient course of patients admitted to the hospital with acute heart failure. Elevations of creatinine at admission, and worsening creatinine during hospitalization, have been associated with increased rates of readmission, prolonged hospitalization, and death.1,2 Increases in serum creatinine of ≥0.3 mg/dL have been associated with increased mortality and prolonged hospitalization.3

The pathophysiology of worsening renal function in patients with heart failure is complex. Traditionally it has been suggested that renal function declines as a result of inadequate renal perfusion, due to reduced cardiac output. This low-flow state results in an increase of renin release and renin-angiotensin-adolsterone system (RAAS) activation. Elevated central venous pressure and venous congestion has been postulated as another cause of worsening renal function. Heart failure is associated with a marked elevation in renal venous pressure that can result in reduced renal blood flow and urine formation.4 Sympathetic stimulation also occurs in heart failure and may cause vasoconstriction to renal blood vessels.5 This multifactorial pathogenesis leads to decreasing renal perfusion and subsequent decrease in renal function.

The common pathway of renal impairment occurs though activation of the RAAS and the occurrence of oxidative stress, often leading to renal injury. An increase in RAAS results in increased levels of angiotensin II. This leads to an increase in cell growth, inflammation, and fibrotic damage to the renal parenchyma.6,7 Oxidative injury has also been shown to be a common link between heart failure and worsening renal function. These mechanisms appear to alter all three components of cardiac energy metabolism: free fatty acid metabolism, oxidative phosphorylation, and phosphate metabolism.

In this study, we aimed to perform an exploratory investigation utilizing metabolomics to determine if metabolites related to oxidative stress and fatty acid metabolism differ in patients presenting with heart failure who progress to develop worsening renal function, compared to those who do not progress. Metabolic profiling refers to the measurement of a group of small molecule metabolites that reflect cellular responses to organ injury due to a specific cause. In particular, using high-performance liquid chromatography (HPLC) and gas chromatography/ mass spectrometry (GC/MS) allows the quantification and identification of these metabolites. We hypothesize that patients who are prone to developing worsening renal function will have a metabolic change that predisposes them to this outcome. If this is the case, measuring metabolomics is an ideal mechanism for the early identification of patients at risk for developing worsening renal function, as it allows the identification of multiple markers of injury. These markers can be used either singly or in concert.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References
  9. Supporting Information

Study Design

This was a pilot study of prospectively identified patients identified by the treating physician as having heart failure. The study was approved by the institutional review board and was funded by the Department of Emergency Medicine at the University of California, Davis.

Study Setting and Population

A convenience sample of patients was enrolled at an urban tertiary center emergency department (ED). Subjects eligible for enrollment were adults age 18 years or older presenting with symptoms suggestive of acute decompensated heart failure. Eligible subjects had to have a B-type natriuretic peptide (BNP) level over 100 pg/mL, have a Boston criteria for diagnosing heart failure score of ≥8,8,9 and require admission to the hospital as determined by the treating physician for the management of acute heart failure irrespective of etiology or ejection fraction. In addition, to identify a high-risk patient population for the development of worsening renal function, the patient had to either have hypertension or diabetes.

Patients were excluded from the study if they had a hematocrit of <30%; they had a history of organ transplantation; they were currently on immunosuppressive medications; they were septic or on antibiotic therapy; they had a history of, or were currently receiving any form of, dialysis; they had received more than 2 hours of treatment with a diuretic or vasodilator in the ED; the treating physician did not feel that they would be admitted to the hospital at the time of screening; or they had received more than one dose of diuretic therapy. Additionally, patients were excluded if they were hemodynamically unstable or suffering a myocardial infarction. The exclusion of patients with prolonged duration of treatment prior to enrollment was introduced to minimize the potential effect of ED interventions on excretion of metabolites.

Study Protocol

After a patient was identified as fulfilling inclusion and exclusion criteria, and informed consent was obtained, ED data were collected prospectively. This included race, demographics, dietary history, medical history, physical examination, and electrocardiogram findings, as documented by the treating emergency physician. Medications administered in the ED and before arrival were also recorded. Final ED diagnosis was based on the treating physician’s impression. No laboratory tests were mandated as part of the trial study, although eligible patients had to have a BNP, complete blood count, and chest radiograph ordered as a component of standard care. Laboratory tests, chest radiography findings as documented by a board-certified radiologist, and echocardiography reports documented by a board-certified cardiologist were also obtained from the medical record.

Urine samples were collected as a midstream sample or via a Foley bag if one was already placed in the ED at the time of study enrollment. Samples were aliquoted into 2-mL samples and frozen at −80°C. Repeat urine samples were collected at 48 hours if patients were still in the hospital. All patients were followed by chart review throughout their index stay to document in-hospital events. Serum creatinine levels were recorded at presentation and 24 and 48 hours if the patient was still in the hospital.

Outcome Measures

The outcome measure of worsening renal function was defined as an increase in serum creatinine from admission of >0.3 mg or an increase of 25% over the first 24 hours. This outcome measure was based on prior studies in heart failure, ED-based studies examining acute kidney injury, and consensus guidelines.10–12

GC/MS Analysis

Urine samples require no preparation prior to freezing and were lyophilized without further pretreatment. To the dried samples, 20 μL of 40 mg/mL methoxylamine hydrochloride in pyridine was added, and samples were agitated at 30°C for 30 minutes. Subsequently, 180 μL of trimethylsilylating agent N-methyl-N-trimethylsilyltrifluoroacetamide was added, and samples were agitated at 37°C for 30 minutes. GC/MS analysis was performed using an Agilent 6890 N gas chromatograph (Atlanta, GA) interfaced to a time-of-flight (TOF) Pegasus III mass spectrometer (Leco, St. Joseph, MI).13–15 Automated injections were performed with a programmable robotic Gerstel MPS2 multipurpose sampler (Mülheim an der Ruhr, Germany). Initial peak detection and mass spectrum deconvolution was performed with ChromaTOF software (version 2.25, Leco), and later samples were exported to the netCDF format for further data evaluation with MZmine (UC Davis, Davis, CA) and XCMS (Bioconductor, http://bioconductor.org). Metabolites were identified with >50% certainty. If the metabolite could not be matched with this degree of certainty they were titled (var__) followed by a number.

Data Analysis

Continuous data are presented as medians and interquartile ranges (IQRs). The statistical analysis was performed on (natural) log-transformed data to account for increases in the data variance that can occur. Log-transformed data were plotted to assess for normality. A t-test was performed to determine differences in the metabolites, with a p-value threshold of <0.05 mg. Partial least squares-discriminate analysis (PLS-DA) was used to identify metabolites associated with worsening renal function (class difference). The features were ranked based on their regression coefficients. To determine if a past medical history of either diabetes or hypertension was an effect modifier for any difference noted in metabolites, a stratified analysis using a t-test was done on the metabolites identified by PLS by a history of diabetes and hypertension. Statistical analysis was performed using MetaboAnalyst.16 We also used MetPA (Metabolomic Pathway Analysis) to identify the most relevant pathways involved in the change in renal function.17 This program will identify metabolism pathways that contain metabolites most likely to be associated with the outcome measure specified.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References
  9. Supporting Information

Of the 24 subjects who were enrolled, the median age was 58 years (IQR = 49.5 to 67.5 years) and (58%) were male. The median Boston criteria score was 10 (IQR = 8 to 10). Worsening renal function occurred in 10 patients (41.7%). A total of 156 metabolites were identified. Demographics of the group are presented in Table 1. Univariate analysis between groups revealed significant differences in six metabolites. These metabolites are presented by pathway in Table 2.

Table 1.    Clinical Features
 Worsening Renal Function (n = 10)No Change in Renal Function (n = 14)
  1. All values reported as median (IQR) or n (%).

  2. ACE/ARB = angiotensin-converting enzyme inhibitors/angiotensin receptor blockers; BNP = B-type natriuretic peptide; BUN = blood urea nitrogen; COPD = chronic obstructive lung disease; CXR = chest radiograph; HF = heart failure; IQR = interquartile range.

  3. *CXR signs of congestion: presence of cephalization, interstitial edema, or pleural effusion.

Time from urine collection to last meal, hours14.3 (8.6–22)8.4 (3.6–16.5)
Age, yr66 (53–72)55 (49–62)
Male7 (70)7 (50)
History of coronary artery disease6 (66)5 (38)
History of diabetes7 (70)6 (43)
History of hypertension7 (42)14 (100)
History of chronic renal insufficiency8 (80)3 (21)
Medications
 Home diuretics7 (70)9 (64)
 Home ACEI/ARB7 (70)4 (29)
 Home beta-blocker8 (80)8 (57)
Symptoms
 Jugular venous distention3 (30)2 (14)
 Rales4 (40)6 (42)
 Lower extremity edema6 (60)12 (85)
Diagnostic findings
 Systolic blood pressure (mm Hg) 146 (119–158)139 (132–183)
 Pulse (beats/min) 83 (70–107)94 (84–110)
 Respiratory rate (breath/min) 21 (20–24)21 (19–30)
 Oxygen saturation 98 (97–100)98 (97–100)
 Creatinine level    2.0 (1.2–2.9)  1.1 (1.0–1.4)
 BUN level 30 (20–47)18 (11–25)
 Hemoglobin     11.2 (10.7–12.8)  11.8 (10.9–13.1)
 BNP (pg/mL)1,154 (704–2,560)894 (349–2,650)
Prehospital and ED treatments
 Prehospital diuretic0 (0)0 (0)
 Diuretic10 (100)11 (79)
 Nitroglycerin68 (33.3)102 (47.7)
 Final ED diagnosis of HF9 (90)13 (95)
Table 2.    Metabolites With Significant Difference Based on Paired t-test of Log-transformed Data
CompoundsCompound ClassPre-CT peak (mz/rt), median (IQR)Post-CT peak (mz/rt), median (IQR)p-value–log10(p)
  1. IQR = interquartile range; mz/rt = specific mass-charge ratios (mz)/LC column retention times (rt).

Var59 3,315 (1,702–4,795)19,997 (8,522–29,864)0.016491.78286
Glucose-1-phospateSugar15,632 (6,135–35,651)462,637 (13,335–1,310,983)0.036971.4321
TaloseSugar1,548 (1,181–7,074)835,170 (2,437–1,941,418)0.044121.35537
Inositol-alloFatty acid metabolism450 (221–1,021)4,151 (934–7,934)0.047861.32003
FuroylglycineSugar3,148 (742–16,006)8,701 (5,894–120,371)0.049661.30401
TaurineAmino acid785 (5–4,085)10 (0–26)0.049741.30325

The optimal number of metabolites for class discrimination as determined by PLS-DA was three, with a classification accuracy of 78%. The three significant variables ranked by coefficient were taurine, sulfuric acid, and talose (Table 3). Differences between total quantity of taurine, sulfuric acid, and talose are presented in Data Supplement S1 (available as supporting information in the online version of this paper).

Table 3.    PLS-DA of Important Metabolites Ranked Based on Their Regression Coefficients
CompoundsCoefficient Score
  1. PLS-DA = partial least squares-discriminant analysis.

Taurine0.0104
Sulfuric acid0.0091
Talose0.0081

We performed a sensitivity analysis to determine if there was a difference in the significant variables noted based on the prior history of diabetes or hypertension. No difference in the log-transformed mean peak was noted in taurine, sulfuric acid, or talose when stratified by diabetes or hypertension. When reviewing the pathway analysis, the two most relevant pathways associated with worsening renal function were fructose and mannose metabolism with an impact value of 0.04 and amino sugar and nucleotide sugar metabolism with an impact value of 0.09.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References
  9. Supporting Information

Early identification of patients who will develop worsening renal function may allow therapeutic interventions that can alter the progression of disease. In this study we were able to show an association between metabolites identified in the urine at the time of presentation in patients with heart failure and the development of worsening renal function. This knowledge, if validated in a larger study, may affect the treating physician’s decisions regarding pharmaceutical therapy or regarding further studies with known risks of renal dysfunction, such as additional computed tomography scans with contrast.

In review of the specific metabolites, taurine was the most significant variable identified by unsupervised learning using PLS. There has been some evidence that taurine is involved in neuronal modulation, osmoregulation, and protection against oxidative stress.18 Plasma levels are maintained through protein intake and de novo synthesis through the activity of hepatic cysteine sulfinic acid decarboxylase. Alterations in taurine levels may be a result of a decline in overall body pool, poor intestinal absorption, or renal wasting. Studies have reported that taurine has low interindividual variation and low intraindividual variation when measured in the urine of healthy male college students.19 However, in diabetic patients with poor blood glucose control, there is increased renal loss of taurine as a result of hyperglycemia and ketoacidosis. In contrast to the previous findings, a study of type 1 diabetics, type 2 diabetics, and healthy controls showed no difference in baseline taurine values.20 We also noted no difference in taurine values in diabetics. Taurine is absorbed in the renal tubules, and this may be an area of focused damage in heart failure. Hypertaurinurina has been associated with cardiomyopathy and has been attributed to levels of myocardial taurine, although no causal relationship has been determined.21 In patients developing worsening renal function we noted a decrease in taurine excretion. This may have occurred as a result of decreased renal flow or decreased total body pool.

Sulfuric acid is produced as a result of oxidation of organic compounds and is eliminated by renal hydrogen ion excretion. This elimination occurs in the distal nephron. In the setting of heart failure, there is a decrease in fatty acid metabolism as the source of cardiac energy and an increase in exogenous myocardial glucose uptake. Similarly, in renal injury, there is a reduction in the normal oxidative phosphorylation process, and therefore a potential decline in the production of sulfuric acid production. However, our results demonstrated an increase in renal excretion of sulfuric acid in the setting of heart failure. One potential mechanism to explain our findings of increased sulfuric acid may be the oxidative stress that is particularly seen in the renal tubules with heart failure patients. Since sulfuric acid is a common by-product of the oxidation of organic compounds, increasing amounts may be generated in the setting of acute oxidative stress as seen in heart failure. Further investigations are needed to better our understanding regarding sulfuric acid generation, and validation of our laboratory findings is needed as well. We also noted an increase in the breakdown products of carbohydrate metabolism in the urine. This may indicate a change in energy source utilization in the setting of acute heart failure exacerbation.

The use of metabolomics to identify metabolites associated with end-organ damage is a novel concept. Current technology does not allow for real-time processing of the samples, and therefore these metabolites associated with renal injury currently cannot be used as diagnostic tests. However, it is possible that by understanding the metabolite pathways associated with renal injury, therapeutic options may be identified and a better understanding of the pathophysiology of the disease process can occur. Taurine is currently being evaluated as a preventive treatment for renal injury in various disease states.22–24 Therefore, if we can confirm our findings in a larger trial, these preliminary data suggest taurine should be evaluated as a potential treatment to protect against renal injury in acute heart failure patients.

Limitations

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References
  9. Supporting Information

Due to the sample size we were unable to adjust for covariates such as past medical history or medications that may affect our results. In addition, we used the change in creatinine at 24 hours as our measure of worsening renal function. This may under- or overestimate the true occurrence of worsening renal function at 48 hours or longer. Patients were allowed to receive a single dose of diuretics prior to enrollment in the trial. This may also alter the urinary metabolites excreted. Last, we studied a highly specific population, enrolling only patients with elevated BNP levels and hypertension or diabetes. This limits the generalizability of these preliminary findings.

Conclusions

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References
  9. Supporting Information

Metabolites found in the urine at the time of ED presentation may be markers of early renal injury and help identify those patients with heart failure who may have increased risk of worsening renal function during hospitalization. In this study we identified several urine metabolites that had different peak concentrations between those patients who did and did not develop worsening renal function. These findings, if supported by further studies, may support that the process of renal injury is initiated prior to ED arrival in patients with suspected heart failure.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Limitations
  7. Conclusions
  8. References
  9. Supporting Information

Data Supplement S1. Comparison of talose, taurine, and sulfuric acid levels in patients with and without worsening renal function (WRF).

Please note: Wiley Periodicals Inc. are not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
ACEM_1239_sm_DataSupplementS1.doc30KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.