http://orcid.org/0000-0002-6973-5908
Figures
Abstract
Background
Although urinary neutrophil gelatinase-associated lipocalin (NGAL) has emerged as a promising biomarker for the early detection of kidney injury, previous studies of adult patients who underwent cardiac surgery have reported only moderate discrimination. The age, creatinine, and ejection fraction (ACEF) score is a preoperative validated risk model with satisfactory accuracy for predicting AKI following cardiac surgery. It remains unknown whether combining preoperative risk assessment through ACEF scores followed by urinary NGAL test in a population of high-risk individuals is an optimal approach with improved predictive performance.
Material and methods
A total of 177 consecutive patients who underwent cardiac surgery were enrolled. Clinical characteristics, prognostic model scores, and outcomes were assessed. Urinary NGAL were examined within 6 hours after cardiac surgery. Patients were stratified according to preoperative ACEF scores, and comparisons were made using the area under the receiver operator characteristic curve (AUROC) for the prediction of AKI.
Results
A total of 45.8% (81/177) of the patients had AKI. As expected, patients with ACEF scores ≥ 1.1 were older and more likely to have class III or IV heart failure. They were also more likely to have diabetes mellitus, myocardial infarction, and peripheral arterial disease. Urinary NGAL alone moderately predicted AKI, with an AUROC of 0.732. Risk stratification by ACEF scores ≥ 1.1 substantially improved the AUROC of urinary NGAL to 0.873 (95% confidence interval, 0.784–0.961; P < .001).
Conclusions
Risk stratification by preoperative ACEF scores ≥ 1.1, followed by postoperative urinary NGAL, provides more satisfactory risk discrimination than does urinary NGAL alone for the early detection of AKI after cardiac surgery. Future studies should investigate whether this strategy could improve the outcomes and cost-effectiveness of care in patients undergoing cardiac surgery.
Citation: Lee C-C, Chang C-H, Chen S-W, Fan P-C, Chang S-W, Chen Y-T, et al. (2018) Preoperative risk assessment improves biomarker detection for predicting acute kidney injury after cardiac surgery. PLoS ONE 13(9): e0203447. https://doi.org/10.1371/journal.pone.0203447
Editor: Emmanuel A. Burdmann, University of Sao Paulo Medical School, BRAZIL
Received: June 10, 2018; Accepted: August 21, 2018; Published: September 4, 2018
Copyright: © 2018 Lee et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper.
Funding: This work was supported by a grant from Chang Gung Memorial Hospital, Taiwan CMRPG3E0111 (CHC) and CORPG3C0182 (FCT). The research was also supported in part by the Ministry of Science and Technology grant MOST 103-2314-B-182A-018-MY3 (CHC).
Competing interests: The authors have declared that no competing interests exist.
Introduction
Acute kidney injury (AKI) occurs in 10%–50% of patients undergoing cardiac surgery.[1–5] This statistical variation is largely due to heterogeneous study populations and the use of different criteria to define AKI. Notably, a recent meta-analysis reported that irrespective of the different guideline-based AKI definitions, the presence of AKI in patients undergoing cardiac surgery is consistently associated with a 2- to 4-fold increase in the risk of early mortality.[6] Furthermore, a clear dose–response relationship exists between the severity of AKI and poor outcomes.[2, 6, 7]
Despite decades of research, the clinical management of AKI mainly focused on hemodynamic manipulations, attenuating ischemia and supportive care.[8] A major challenge to developing effective interventions for AKI has been the limited ability to detect this condition in its early stages. Hence, several proteins have been identified as sensitive and specific biomarkers for detecting AKI.[9–11] A frequently studied biomarker is urinary neutrophil gelatinase-associated lipocalin (NGAL), which has demonstrated promise in detecting postoperative AKI earlier than serum creatinine can.[11–13] However, compared with the excellent discrimination reported in pediatric patients, a recent meta-analysis of studies of urinary NGAL in adults reported only moderate discrimination, with an area under the receiver operating characteristic curve (AUROC) of 0.72 (95% confidence interval, 0.66–0.79).[14] Its comparatively high costs render it less useful at present than serum creatinine as a routine clinical tool. Therefore, we examined whether a diagnostic flow chart with stratification by clinical risk assessment as the first line, followed by the urinary NGAL test in a population of high-risk individuals, can improve overall performance.
Numerous clinically validated risk models for cardiac surgery are currently in use. The age, creatinine, and ejection fraction (ACEF) score was initially developed to simplify bedside evaluation by using only 3 critical variables for predicting mortality in patients undergoing coronary artery bypass grafting (CABG).[15] According to the “law of parsimony” for excluding excessive variables that might result in overfitting and the risk of multicollinearity, the ACEF score exhibits similar or superior accuracy to that of a previously established complex model for stratifying mortality risk.[15–17] Our recent studies have demonstrated that ACEF scores can also accurately identify the risk of postoperative AKI in patients undergoing either valve surgery [4] or CABG.[5] We concluded that this extremely user-friendly ACEF score might be an optimal risk assessment tool. The aim of this investigation was to evaluate the discrimination ability of urinary NGAL after risk stratification by ACEF scores.
Materials and methods
Study participants and design
This observational investigation was performed in the cardiac surgery intensive care unit (ICU) of a tertiary care referral center in Taiwan between July 2014 and February 2015. The study protocol was approved by the institutional review board of the Chang Gung Memorial Hospital (No. 103-1993B). Consecutive patients who provided informed consent and who were admitted to the ICU immediately after cardiac surgery were enrolled. Patients who had estimated glomerular filtration rate (eGFR) < 30 mL/min/1.73m2, were receiving dialysis, were aged <20 years, had confirmed AKI before surgery, or reported any prior organ transplantation were excluded.
Data collection and definition
Demographic data, clinical characteristics, echocardiographic data, and routine biochemistry test results were collected in a prospective database and retrospectively evaluated. Comorbidities were evaluated according to the following definitions: myocardial infarction was defined according to the 2007 Expert Consensus Document of Circulation, European Heart Journal; Peripheral arterial disease (PAD) was defined as a confirmed ankle–brachial index < 0.9, diagnosis established through duplex ultrasonography, and need for medication; and chronic obstructive pulmonary disease was confirmed through a spirometer test or a medical history and need for treatment.
Baseline serum creatinine was determined as the concentration measured within one week before surgery. If multiple values were available during this week, the creatinine level was defined as the lowest value. If unavailable, the value measured at the steady state within 2 months before admission was used as a surrogate baseline. After cardiac surgery, we measured serum creatinine levels every morning for up to 1 week. The ACEF scores were calculated as age (years)/ejection fraction (EF)% + 1 (if creatinine > 2.0 mg/dL).[15] The EF was measured through 2-dimensional echocardiography before surgery.
Outcome definition
To determine the predictive value of biomarkers for AKI, the primary outcome was the development of AKI within 7 days after cardiac surgery. Based on the Kidney Disease Improving Global Outcomes (KDIGO) Clinical Practice Guidelines for Acute Kidney Injury, AKI was confirmed under any of the following conditions: serum creatinine levels ≥ 0.3 mg/dL within 48 hours or a ≥1.5-fold increase in serum creatinine levels from baseline within 7 days.[18] The cutoff value of preoperative ACEF scores for predicting AKI in our earlier studies was 1.2 for CABG and 1.1 for valve surgery, according to the optimal Youden index and overall accuracy.[4, 5] To increase the sensitivity of this ACEF score for screening high-risk group for AKI, we applied a cutoff value of 1.1. Other outcome measures included in-hospital mortality, renal replacement therapy, duration of ventilation, and length of ICU stay.
Measurement of NGAL levels
Fresh urine samples were collected within 6 hours postoperatively in nonheparinized tubes via an indwelling Foley catheter and were then centrifuged at 5000 × g for 30 minutes at 4°C to remove cells and debris. The clarified supernatants were stored at −80°C before measurement. Urinary NGAL levels were measured in duplicate by using commercially available enzyme-linked immunosorbent assay kits, according to the manufacturer’s instructions (R&D Systems, DLCN20; Mpls, USA). The intra-assay coefficient of variability for urine NGAL was 8.37%.
Statistical analysis
We compared the continuous variables between groups (i.e., AKI vs. non-AKI or ACEF ≥ 1.1 vs. ACEF < 1.1) using the student t-test. Categorical data between groups were compared using the Fisher exact test. The performance of NGAL in discriminating AKI was assessed using the AUROC. We then performed the ROC analysis stratified by ACEF levels and applied a nonparametric approach[19] to compare the AUROCs of NGAL at different ACEF levels. To determine the optimal cut-off value for NGAL in discriminating AKI, we estimated Youden index, which is defined as (sensitivity + specificity) - 1. The optimal cut-off value for NGAL is that its corresponding Youden index achieves the maximum because it is the point that the highest sensitivity and specificity values occur at the same time. Finally, after incorporating the 40% AKI incidence for individuals who underwent cardiac surgery in Taiwan,[4,5,20] positive and negative predictive values were calculated, given the optimal cutoff value of NGAL. A two-sided P value < 0.05 was considered to be statistically significant. No adjustment of multiple testing (multiplicity) was made in this study. The statistical analyses were conducted using SPSS 22 (IBM SPSS, Armonk, NY: IBM Corp).
Results
Study population characteristics: Non-AKI versus AKI groups
A total of 177 consecutive patients who underwent cardiac surgery between July 2014 and February 2015 were enrolled in the study. Postoperative AKI occurred in 81 (45.8%) patients. Among the study patients, 73 (41.2%) had preoperative ACEF scores ≥ 1.1. Clinical characteristics and laboratory data of all the study patients and those with ACEF scores ≥ 1.1, stratified by the presence of AKI, are summarized in Table 1. There was no significant difference in age, sex, presence of comorbidities, and preoperative cardiac function, either in the whole study population or in the subgroup of patients with ACEF scores ≥ 1.1. The type of cardiac surgery and the proportion of urgent operation were similar between the 2 groups. In both the total patient group and the subgroup of patients with ACEF scores ≥ 1.1, those with AKI had a longer bypass time.
Compared with the patients without AKI, those with AKI had higher postoperative urinary NGAL levels in both the total patient group (P < .001) and the subgroup of patients with ACEF scores ≥ 1.1 (P < .001). Eleven patients died within 30 days, with a significantly higher mortality rate in the AKI group than in the non-AKI group (11.1% vs. 2.1%; P = .024). Furthermore, the AKI group had longer durations of ventilator use and ICU stay.
Patient characteristics across ACEF scores
Among the 177 enrolled patients, 73 (41.2%) had ACEF scores ≥ 1.1 and 104 (58.8%) had ACEF scores < 1.1. The baseline characteristics of the 2 groups are presented in Table 2. As expected, compared with patients with ACEF scores < 1.1, those with ACEF scores ≥ 1.1 were significantly older and had higher serum creatinine and lower EF levels. Patients with ACEF scores ≥ 1.1 were more likely to have diabetes mellitus (DM), myocardial infarction within 6 months, PAD, and class III or IV heart failure. Patients with ACEF scores ≥ 1.1 also exhibited higher mean values in Society of Thoracic Surgeons (STS) and Euro scores. No significant differences were observed in the bypass and clamp times.
Postoperative urinary NGAL values did not differ between the 2 groups (P = .104). The incidence of AKI was significantly higher among the patients with ACEF scores ≥ 1.1 than among those with ACEF scores < 1.1 (57.5% vs. 37.5%, P = .010). A trend of a higher incidence of dialysis-requiring AKI was observed in the patients with ACEF scores ≥ 1.1 (11.0% vs. 3.8%, P = .075).
Performance of NGAL in discriminating AKI
We then examined the performance of the urinary NGAL test for diagnosing AKI. As shown in Table 3, urinary NGAL exhibited satisfactory discrimination ability in the whole study population, with an AUROC of 0.732 (95% confidence interval [CI], 0.656–0.808). The discrimination ability of urinary NGAL for predicting AKI in individuals with ACEF scores ≥ 1.1 (AUROC, 0.873; 95% CI, 0.784–0.961) outperformed that in individuals with ACEF scores < 1.1 (AUROC, 0.606; 95% CI, 0.492–0.720), with a significant difference in the AUROC (P < .001; Fig 1). The optimal cutoff value of urinary NGAL in the total cohort was 82.6 ng/mL, with a sensitivity of 60.5% and a specificity of 80.2%. By contrast, urinary NGAL exhibited a higher sensitivity of 83.3% and specificity of 87.2% for a cutoff value of 77.5 ng/mL when applied in patients with ACEF scores ≥ 1.1. Noticeably, when the population incidence of AKI was set at 40% [4,5,20], the obtained positive predictive value was 81.2% and negative predictive value was 88.7%, given the cutoff value of 77.5 ng/mL in patients with ACEF scores ≥ 1.1.
AUROCs were 0.606 (0.492–0.720) for ACEF scores < 1.1 and 0.873 (0.784–0.961) for ACEF scores ≥ 1.1, with a significant difference in the values (P < .001).
Fig 2 depicts the incidence of AKI across the tertiles of urinary NGAL, stratified by the ACEF score. Notably, the incidence of AKI in patients with ACEF scores ≥ 1.1 stratified by urinary NGAL tertiles increased from 16.7% to 66.7% to 88.0% in the first, second, and third tertiles, respectively, and a significant increasing trend was observed. The odds ratios for the development of AKI in patients with ACEF scores ≥ 1.1 were 10.0 and 36.7 for the second and third tertiles of urinary NGAL compared with those for the lowest tertile. By contrast, the risk for the occurrence of AKI was not increased across urinary NGAL tertiles in patients with ACEF scores < 1.1.
Discussion
In this study, we demonstrated the performance characteristics of a two-step diagnostic framework, combined risk assessment through ACEF scores first and then urinary NGAL test, a urinary biomarker of tubular injury measured within 6 hours after cardiac surgery, for detecting subsequent AKI. We discovered that simple preoperative risk stratification by ACEF scores ≥ 1.1, followed by urinary NGAL levels, offered significant complementary risk stratification with superior discrimination.
AKI remains a major complication in patients undergoing cardiac surgery. AKI incurs significant costs and is independently associated with adverse clinical outcomes.[5, 6, 21] Accumulating evidence has revealed that the distinctive characteristics of cardiac surgery, including cardiopulmonary bypass and aortic clamping, which compromise renal perfusion and induce systemic inflammatory response, contribute substantially to the development of AKI.[8, 22–24] Our data are consistent with these findings. Our study demonstrated that the patients in the AKI group had a higher mortality rate than did those in the non-AKI group (11.1% vs. 2.1%; P = .024). The patients in the AKI group had a longer bypass time than did those in the non-AKI group (187.3 ± 91.5 minutes vs. 159.1 ± 58.3 minutes; P = .026). Because AKI is a well-established risk factor for mortality in patients undergoing cardiac surgery, earlier identification of patients at risk for AKI would allow physicians to initiate interventions to impede disease progression. The usual 24–36-hour delay in serum creatinine rise after a definite renal insult implies that serum creatinine is not a sufficiently sensitive parameter for the measurement of kidney function in AKI.[12, 25, 26]
Our study demonstrated that increases in urinary NGAL levels after cardiac surgery anticipated the subsequent development of AKI (as defined by the KDIGO criteria), with an AUROC of 0.732 (95% CI, 0.656–0.808). This finding is consistent with previous results from the Translational Research Investigating Biomarker End-Points in Acute Kidney Injury (TRIBE-AKI) study[27] and other investigations,[11, 12, 28–30] which have demonstrated that postoperative urinary NGAL is an effective biomarker for early identification of AKI. The TRIBE-AKI study tested the ability of several biomarkers, including urinary NGAL, to detect AKI in 1219 adults who underwent cardiac surgery. The AUROCs of urinary NGAL in this cohort at 0–6 and 6–12 hours were 0.67 and 0.70, respectively.[27] Similarly, in a prospective study of 85 adult patients who underwent cardiac surgery, Matsui et al. stated that the AUROC was 0.74 for predicting the development of AKI according to urinary NGAL levels at 6 hours postoperatively.[29] NGAL is a 25-kDa protein originally identified from human neutrophils as a shuttle for iron transport,[31] and its upregulation has been demonstrated in the proximal tubule shortly after exposure to harmful insults such as ischemia or nephrotoxins.[32, 33] The resultant increased NGAL synthesis is released by injured tubular epithelial cells within hours after AKI. Crucially, both the expression of NGAL in the injured renal tubule and the concentration of NGAL in the urine increase in proportion to the severity of kidney injury,[12, 34, 35] thus providing biological plausibility for its use as an AKI biomarker.
Current evidence does not support routine clinical use because of its moderate discrimination ability and variable thresholds (cutoffs).[14] In addition, unrestrictedly screening every case is not financially feasible. Therefore, risk stratification by robust clinically validated models, followed by biomarkers, should be considered a more optimal and cost-efficient strategy in daily care. Our previous studies have revealed that ACEF scores are a suitable tool for predicting AKI in patients undergoing either CABG[5] or mitral valve surgery[4], with discrimination ability comparable to that of other scoring systems such as STS and Euro scores. Both STS and Euro scores consist of complicated mathematical calculations and require a web-based application to enhance implementation. By contrast, ACEF scores are easy to apply, with a balance between predictability and simplicity. In the current study, patients with ACEF scores ≥ 1.1 exhibited higher mean STS and Euro scores and demonstrated a higher incidence of AKI. These findings contribute to the evidence on the prognostic utility of ACEF scores, and we report another critical finding: risk stratification by ACEF scores raised the AUROC of urinary NGAL from 0.732 to 0.873 for predicting AKI. When the cutoff value of urinary NGAL was set at 77.5 ng/mL, the sensitivity was 83.3%, specificity was 87.2%, and negative predictive value was 88.7%. This superior discrimination ability of urinary NGAL after risk stratification by ACEF scores has 2 possible explanations. First, clinical assessment through ACEF scores enriched the population for the development of AKI and thus largely improved the sensitivity (from 60.5% to 83.3%) and overall performance of urinary NGAL. Indeed, our study revealed that patients with ACEF scores ≥ 1.1 were older and more likely to have DM, myocardial infarction, PAD, and severe heart failure, which are well-established risk factors for AKI.[7, 8, 23] Another explanation is that patients with ACEF scores < 1.1 have probably received damage to their kidneys (as demonstrated by the similar levels of urinary NGAL to that in patients with ACEF scores ≥ 1.1), but not of sufficient severity to cause any subsequent rise in serum creatinine. The exclusion of these populations helps to improve or retain the specificity (from 80.2% to 87.2%) of urinary NGAL for predicting AKI.
Some studies have identified baseline renal function as a possible modifier of the discrimination ability of urinary NGAL as an AKI biomarker, with greater predictive ability in patients with eGFR ≥ 60 mL/min than in those with eGFR < 60 mL/min.[36, 37] This may be due to the less-functional nephron mass with a reduced ability to produce NGAL during renal insult. However, our study demonstrated that patients with ACEF scores ≥ 1.1 had higher serum creatinine values, namely lower eGFRs, than did those with ACEF scores < 1.1. Thus, baseline renal function was unlikely to have contributed appreciably to our favorable discriminatory power of urinary NGAL among patients with ACEF scores ≥ 1.1 in the present series.
We are not aware of previous studies investigating the performance of AKI biomarkers with a two-step diagnostic framework combining risk stratification by a clinically validated model as the first line. However, a few reports have incorporated multiple variables, including one biomarker, for predicting AKI after cardiac surgery. The TRIBE-AKI study combined a biomarker and clinical model, and found that adding clinical model significantly but only modestly raised the AUROC of urine IL-18 and plasma NGAL from 0.74 to 0.76 (P = .03) and 0.70 to 0.75 (P = .01), respectively. Combining urinary NGAL and their clinical model failed to significantly improve model discrimination (P = .12).[27] The difference may be attributable to a different diagnostic framework for the study. Another possible explanation is that different clinical models were used. Our study used a simple validated risk score comprising only 3 preoperative variables, namely age, preoperative creatinine, and ejection fraction. However, the clinical model adopted in the TRIBE-AKI study comprised 8 self-constructed variables, including 7 preoperative variables (age, preoperative eGFR, race, sex, DM, hypertension, and urgent surgery) and one intraoperative variable (bypass time > 120 minutes). Certain factors, such as urgent surgery and bypass time > 120 minutes, could be associated with both increased postoperative urinary NGAL levels and increased risk of AKI, leaving little room for additional improvement. Additional studies are warranted to investigate this possibility.
Although many studies investigating urinary NGAL utility in predicting the development of AKI have normalized NGAL against creatinine to control changes in urine flow rate, the underlying assumption of constant creatinine excretion may be flawed.[38, 39] Lower creatinine excretion in the setting of acute GFR loss during early stage of AKI may amplify the NGAL performance irrespective of whether NGAL synthesis is increased by injury or not. However, we analyzed our data using normalized values and the results are consistent with those using absolute NGAL values. Risk stratification by ACEF scores ≥ 1.1, followed by normalized NGAL, still provides more satisfactory risk discrimination (AUROC, 0.876; 95% CI, 0.800–0.953) than does normalized NGAL alone (AUROC, 0.754; 95% CI, 0.681–0.826) for the early detection of AKI.
This study has several limitations. First, despite how the time course of change depicted that NGAL peaked at approximately 4 hours after cardiopulmonary bypass,[12, 36] repeated measurements might further improve the discrimination. Second, in this study, AKI was identified within 7 days, and postoperative care, postoperative bleeding, and infection also contributed to the AKI; therefore, caution is warranted when expanding the application of our findings to AKI occurring at 7 days after surgery. Third, we did not include urine output criteria in the AKI definition, which might lead to underestimate AKI incidence. Four, the etiologies of AKI might vary in different types of cardiac surgery. The relatively small sample size of this study prevented an adequate subgroup analysis, and thus, the applicability of our data to a specific type of cardiac surgery is limited.
In conclusion, the results of the current study indicated that ACEF scores can help to identify patients with a higher risk of AKI after cardiac surgery. A two-step diagnostic framework combining preoperative risk assessment through ACEF scores first, followed by postoperative urinary NGAL evaluation, provide more accurate risk discrimination than does urinary NGAL alone for the early detection of AKI after cardiac surgery. Additional studies are required to investigate whether this strategy could improve outcomes and reduce the costs of care in patients undergoing cardiac surgery.
References
- 1. Rosner MH, Okusa MD. Acute kidney injury associated with cardiac surgery. Clin J Am Soc Nephrol. 2006;1(1):19–32. pmid:17699187.
- 2. Liotta M, Olsson D, Sartipy U, Holzmann MJ. Minimal changes in postoperative creatinine values and early and late mortality and cardiovascular events after coronary artery bypass grafting. Am J Cardiol. 2014;113(1):70–5. pmid:24176074.
- 3. Ryden L, Ahnve S, Bell M, Hammar N, Ivert T, Sartipy U, et al. Acute kidney injury after coronary artery bypass grafting and long-term risk of myocardial infarction and death. Int J Cardiol. 2014;172(1):190–5. pmid:24502882.
- 4. Chang CH, Lee CC, Chen SW, Fan PC, Chen YC, Chang SW, et al. Predicting Acute Kidney Injury Following Mitral Valve Repair. Int J Med Sci. 2016;13(1):19–24. pmid:26816491; PubMed Central PMCID: PMCPMC4716816.
- 5. Chen SW, Chang CH, Fan PC, Chen YC, Chu PH, Chen TH, et al. Comparison of contemporary preoperative risk models at predicting acute kidney injury after isolated coronary artery bypass grafting: a retrospective cohort study. BMJ Open. 2016;6(6):e010176. pmid:27354068; PubMed Central PMCID: PMCPMC4932284.
- 6. Pickering JW, James MT, Palmer SC. Acute kidney injury and prognosis after cardiopulmonary bypass: a meta-analysis of cohort studies. Am J Kidney Dis. 2015;65(2):283–93. pmid:25445101.
- 7. Lopez-Delgado JC, Esteve F, Torrado H, Rodriguez-Castro D, Carrio ML, Farrero E, et al. Influence of acute kidney injury on short- and long-term outcomes in patients undergoing cardiac surgery: risk factors and prognostic value of a modified RIFLE classification. Crit Care. 2013;17(6):R293. pmid:24330769; PubMed Central PMCID: PMCPMC4056889.
- 8. O'Neal JB, Shaw AD, Billings FTt. Acute kidney injury following cardiac surgery: current understanding and future directions. Crit Care. 2016;20(1):187. pmid:27373799; PubMed Central PMCID: PMCPMC4931708.
- 9. Vaidya VS, Ferguson MA, Bonventre JV. Biomarkers of acute kidney injury. Annu Rev Pharmacol Toxicol. 2008;48:463–93. pmid:17937594; PubMed Central PMCID: PMCPMC2742480.
- 10. Waanders F, Navis G, van Goor H. Urinary tubular biomarkers of kidney damage: potential value in clinical practice. Am J Kidney Dis. 2010;55(5):813–6. pmid:20438985.
- 11. Han WK, Wagener G, Zhu Y, Wang S, Lee HT. Urinary biomarkers in the early detection of acute kidney injury after cardiac surgery. Clin J Am Soc Nephrol. 2009;4(5):873–82. pmid:19406962; PubMed Central PMCID: PMCPMC2676184.
- 12. Bennett M, Dent CL, Ma Q, Dastrala S, Grenier F, Workman R, et al. Urine NGAL predicts severity of acute kidney injury after cardiac surgery: a prospective study. Clin J Am Soc Nephrol. 2008;3(3):665–73. pmid:18337554; PubMed Central PMCID: PMCPMC2386703.
- 13. Koyner JL, Bennett MR, Worcester EM, Ma Q, Raman J, Jeevanandam V, et al. Urinary cystatin C as an early biomarker of acute kidney injury following adult cardiothoracic surgery. Kidney Int. 2008;74(8):1059–69. pmid:18650797; PubMed Central PMCID: PMCPMC2745082.
- 14. Ho J, Tangri N, Komenda P, Kaushal A, Sood M, Brar R, et al. Urinary, Plasma, and Serum Biomarkers' Utility for Predicting Acute Kidney Injury Associated With Cardiac Surgery in Adults: A Meta-analysis. Am J Kidney Dis. 2015;66(6):993–1005. pmid:26253993.
- 15. Ranucci M, Castelvecchio S, Menicanti L, Frigiola A, Pelissero G. Risk of assessing mortality risk in elective cardiac operations: age, creatinine, ejection fraction, and the law of parsimony. Circulation. 2009;119(24):3053–61. pmid:19506110.
- 16. Laurent M, Fournet M, Feit B, Oger E, Donal E, Thebault C, et al. Simple bedside clinical evaluation versus established scores in the estimation of operative risk in valve replacement for severe aortic stenosis. Arch Cardiovasc Dis. 2013;106(12):651–60. pmid:24231053.
- 17. Barili F, Pacini D, Rosato F, Roberto M, Battisti A, Grossi C, et al. In-hospital mortality risk assessment in elective and non-elective cardiac surgery: a comparison between EuroSCORE II and age, creatinine, ejection fraction score. Eur J Cardiothorac Surg. 2014;46(1):44–8. pmid:24401691.
- 18. Section 2: AKI Definition. Kidney Int Suppl. 2012;2(1):19–36. pmid:25018918; PubMed Central PMCID: PMCPMC4089595.
- 19. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44(3):837–45. pmid:3203132.
- 20. Tsai HS, Tsai FC, Chen YC, Wu LS, Chen SW, Chu JJ, et al. Impact of acute kidney injury on one-year survival after surgery for aortic dissection. Ann Thorac Surg. 2012;94(5):1407–12. pmid:22939248.
- 21. Dasta JF, Kane-Gill SL, Durtschi AJ, Pathak DS, Kellum JA. Costs and outcomes of acute kidney injury (AKI) following cardiac surgery. Nephrol Dial Transplant. 2008;23(6):1970–4. pmid:18178605.
- 22. Murphy GJ, Angelini GD. Side effects of cardiopulmonary bypass: what is the reality? J Card Surg. 2004;19(6):481–8. pmid:15548178.
- 23. Parolari A, Pesce LL, Pacini D, Mazzanti V, Salis S, Sciacovelli C, et al. Risk factors for perioperative acute kidney injury after adult cardiac surgery: role of perioperative management. Ann Thorac Surg. 2012;93(2):584–91. pmid:22269725.
- 24. Salameh A, Dhein S. Strategies for Pharmacological Organoprotection during Extracorporeal Circulation Targeting Ischemia-Reperfusion Injury. Front Pharmacol. 2015;6:296. pmid:26733868; PubMed Central PMCID: PMCPMC4686733.
- 25. Ostermann M, Joannidis M. Acute kidney injury 2016: diagnosis and diagnostic workup. Crit Care. 2016;20(1):299. pmid:27670788; PubMed Central PMCID: PMCPMC5037640.
- 26. Waikar SS, Bonventre JV. Creatinine kinetics and the definition of acute kidney injury. J Am Soc Nephrol. 2009;20(3):672–9. pmid:19244578; PubMed Central PMCID: PMCPMC2653692.
- 27. Parikh CR, Coca SG, Thiessen-Philbrook H, Shlipak MG, Koyner JL, Wang Z, et al. Postoperative biomarkers predict acute kidney injury and poor outcomes after adult cardiac surgery. J Am Soc Nephrol. 2011;22(9):1748–57. pmid:21836143; PubMed Central PMCID: PMCPMC3171945.
- 28. Koyner JL, Vaidya VS, Bennett MR, Ma Q, Worcester E, Akhter SA, et al. Urinary biomarkers in the clinical prognosis and early detection of acute kidney injury. Clin J Am Soc Nephrol. 2010;5(12):2154–65. pmid:20798258; PubMed Central PMCID: PMCPMC2994075.
- 29. Matsui K, Kamijo-Ikemori A, Sugaya T, Yasuda T, Kimura K. Usefulness of urinary biomarkers in early detection of acute kidney injury after cardiac surgery in adults. Circ J. 2012;76(1):213–20. pmid:22094907.
- 30. Liebetrau C, Dorr O, Baumgarten H, Gaede L, Szardien S, Blumenstein J, et al. Neutrophil gelatinase-associated lipocalin (NGAL) for the early detection of cardiac surgery associated acute kidney injury. Scand J Clin Lab Invest. 2013;73(5):392–9. pmid:23668886.
- 31. Yang J, Goetz D, Li JY, Wang W, Mori K, Setlik D, et al. An iron delivery pathway mediated by a lipocalin. Mol Cell. 2002;10(5):1045–56. pmid:12453413.
- 32. Mishra J, Ma Q, Prada A, Mitsnefes M, Zahedi K, Yang J, et al. Identification of neutrophil gelatinase-associated lipocalin as a novel early urinary biomarker for ischemic renal injury. J Am Soc Nephrol. 2003;14(10):2534–43. pmid:14514731.
- 33. Mori K, Lee HT, Rapoport D, Drexler IR, Foster K, Yang J, et al. Endocytic delivery of lipocalin-siderophore-iron complex rescues the kidney from ischemia-reperfusion injury. J Clin Invest. 2005;115(3):610–21. pmid:15711640; PubMed Central PMCID: PMCPMC548316.
- 34. Paragas N, Qiu A, Zhang Q, Samstein B, Deng SX, Schmidt-Ott KM, et al. The Ngal reporter mouse detects the response of the kidney to injury in real time. Nat Med. 2011;17(2):216–22. pmid:21240264; PubMed Central PMCID: PMCPMC3059503.
- 35. de Geus HR, Bakker J, Lesaffre EM, le Noble JL. Neutrophil gelatinase-associated lipocalin at ICU admission predicts for acute kidney injury in adult patients. Am J Respir Crit Care Med. 2011;183(7):907–14. pmid:20935115.
- 36. McIlroy DR, Wagener G, Lee HT. Neutrophil gelatinase-associated lipocalin and acute kidney injury after cardiac surgery: the effect of baseline renal function on diagnostic performance. Clin J Am Soc Nephrol. 2010;5(2):211–9. pmid:20056755; PubMed Central PMCID: PMCPMC2827603.
- 37. Endre ZH, Pickering JW, Walker RJ, Devarajan P, Edelstein CL, Bonventre JV, et al. Improved performance of urinary biomarkers of acute kidney injury in the critically ill by stratification for injury duration and baseline renal function. Kidney Int. 2011;79(10):1119–30. pmid:21307838; PubMed Central PMCID: PMCPMC3884688.
- 38. Waikar SS, Sabbisetti VS, Bonventre JV. Normalization of urinary biomarkers to creatinine during changes in glomerular filtration rate. Kidney Int. 2010;78(5):486–94. Epub 2010/06/18. pmid:20555318; PubMed Central PMCID: PMCPMC3025699.
- 39. Ralib AM, Pickering JW, Shaw GM, Devarajan P, Edelstein CL, Bonventre JV, et al. Test characteristics of urinary biomarkers depend on quantitation method in acute kidney injury. J Am Soc Nephrol. 2012;23(2):322–33. Epub 2011/11/19. pmid:22095948; PubMed Central PMCID: PMCPMC3269182.