Acute kidney injury (AKI) is one of the most common complications of sepsis and is associated with an increased ICU and hospital mortality (1–7). The benefit of preventive and therapeutic measures for AKI has been difficult to confirm because treatments are often initiated when renal injury is already established (8, 9). Perhaps as a result, the prognosis for sepsis patients with AKI remains poor.
Biomarkers for AKI might allow for earlier initiation or more tailored application of renal protection measures and avoidance of iatrogenic harm (10, 11). Two novel urinary biomarkers, tissue inhibitor of metalloproteinases-2 (TIMP-2), and insulin-like growth factor-binding protein 7 (IGFBP7) have been validated for predicting moderate and severe AKI (classified as AKI stage 2 and 3 according to the Kidney Disease Improving Global Outcome [KDIGO] 2012 classification) (12) in critically ill patients. Both TIMP-2 and IGFBP7 are markers of cellular stress in the early phase of tubular cell injury caused by a wide variety of insults (inflammation, ischemia, oxidative stress, drugs, and toxins) (13–16). Furthermore, both molecules can initiate G1 cell-cycle arrest that prevents cells from dividing when potentially injured (17). Importantly, both biomarkers also act as “alarm” proteins exerting paracrine effects on adjacent cells (18). The product of these two biomarkers ([TIMP-2]·[IGFBP7]) outperformed all other known biomarkers or biomarker combinations for predicting moderate-severe AKI (19), and the test has been validated using a clinical adjudication committee as a gold-standard for AKI (20). Finally, [TIMP-2]·[IGFBP7] measurement proved to be a highly sensitive predictor of AKI in cardiac surgery patients (21).
Importantly, performance of novel biomarkers for AKI can suffer in patients with sepsis presumably because many of the pathologic processes of sepsis can affect biomarkers even without injuring the kidney (22). Furthermore, because the mechanisms of organ injury in sepsis may not be specific for the kidney, nonrenal organ failures could mimic AKI (23). Thus, we sought to evaluate the performance of [TIMP-2]·[IGFBP7] in patients with sepsis, with or without nonrenal organ failures. We assessed this biomarker combination in a subset of patients with early sepsis from the Sapphire (19) and Topaz (20) prospective clinical trials using cutoff values of greater than 0.3, 1.0, and 2.0 (ng/mL)2/1,000 (24). At the time of biomarker measurement and adjusting for common sepsis risk factors, we evaluated test performance in conjunction with bedside clinical parameters.
PATIENTS AND METHODS
We conducted a preplanned subgroup analysis of critically ill patients enrolled in either of our two previously reported studies on the discovery/validation (Sapphire) (19) and subsequent secondary validation (Topaz) (20) of [TIMP-2]·[IGFBP7]. We defined sepsis based on international consensus criteria and the clinical diagnosis assigned by the treating physicians at enrolling sites (25). All patients were considered to be at high risk for AKI, characterized by respiratory or cardiovascular dysfunction as previously reported (Fig. 1) (19, 20). The design, execution, and reporting of this study meet the Strengthening the Reporting of Observational Studies in Epidemiology (26) and the Standards for Reporting of Diagnostic Accuracy criteria (27). Data were collected by the investigators and analyzed by independent statisticians not directly affiliated with the study. Both study protocols were approved by the Western Institutional Review Board (Olympia, Washington, DC) and also by the institutional review board or ethics committee of each study site if required. All patients (or authorized representatives) provided written informed consent. In this article, we present data from the Sapphire study, which defined AKI as KDIGO stage 2–3 (12), and from the Topaz study, which used clinical adjudication for AKI (20), in order to examine the performance of the [TIMP-2]·[IGFBP7] test for risk assessment of AKI in patients with sepsis.
Urine and serum samples for biomarker and creatinine assessment, respectively, were obtained within 24 hours of ICU admission. TIMP-2 and IGFBP7 concentrations were measured by immunoassay with the NephroCheck Test on the Asute140 Meter (Astute Medical, San Diego, CA) by technicians blinded to clinical data. Measurements of TIMP-2 and IGFBP7 were made at Astute Medical for the Sapphire study and in triplicate at three independent laboratories (University of California at San Diego, CA; University of Louisville, KY; and ARUP Laboratories in Salt Lake City, UT) for the Topaz study. The median of the triplicate values from Topaz was used for analysis, and values were reported in units of (ng/mL)2/1,000. Serum creatinine testing was performed as previously described (19, 20).
We assessed severity of illness and organ dysfunction/failure with the Acute Physiology and Chronic Health Evaluation (APACHE) III (28) and Sequential Organ Failure Assessment (SOFA) (29) scores. Nonrenal APACHE III and SOFA scores were calculated by subtracting the renal components from these scores.
To assess the performance of [TIMP-2]·[IGFBP7] in predicting AKI, we calculated area under the receiver operating characteristic curve (AUC), sensitivity, specificity, negative and positive predictive values (NPV and PPV), and relative risk (RR). The Delong method was used to estimate the 95% CI for AUC. Bootstrap method was used to estimate the 95% CI for sensitivity, specificity, NPV, and PPV, except for the cases where there were empty cells and the Clopper-Pearson Exact method was used instead. RR was calculated at each [TIMP-2]·[IGFBP7] stratum relative to the lowest stratum. The 95% CI for RR was calculated using bootstrap method except when there were empty cells, and an exact unconditional method was used instead (30).
To examine the effect of nonrenal organ dysfunction and AKI on the levels of [TIMP-2]·[IGFBP7], we performed a multiple linear regression analysis where the response variable was rank transformed [TIMP2]·[IGFBP7]; the explanatory variables were AKI status, subgroup status according to the median of nonrenal SOFA scores, and the interaction between them. The same analysis was also performed for each individual nonrenal SOFA component.
To assess the added value of [TIMP-2]·[IGFBP7] in predicting AKI over using clinical variables alone, we constructed two multivariable logistic regression models: one with only clinical variables as explanatory variables and the other with the addition of [TIMP-2]·[IGFBP7] as an explanatory variable besides clinical variables. A backward stepwise regression procedure with Bayesian Information Criteria was used to select which clinical variables to be included the final models. Model goodness-of-fit was assessed using the Hosmer-Lemeshow method. To quantify the added predictive ability of [TIMP-2]·[IGFBP7], we calculated integrated discrimination improvement (IDI) and category-free net reclassification improvement (cfNRI) using R package “PredictABEL” (31). In addition, we compared the AUCs from these two models using the Delong method for paired AUCs.
A p value of less than 0.05 was considered to indicate statistical significance. All reported p values are two sided. All analyses were performed using SAS 9.3 (SAS Institute, Cary, NC) and R 3.1.0 (31).
Baseline and Clinical Characteristics
We included the 232 sepsis patients in the analysis, including 40 (17%) who developed AKI within 12 hours of testing. Baseline characteristics of all patients stratified by AKI status are shown in Table 1. Almost half of the patients were referred from the emergency department with relatively short delay between hospital and ICU admission (median [intraquartile range] time from hospital to ICU admission was 2 hr [0–8 hr]). Exposure to radiocontrast agents and nephrotoxic drugs and the presence of risk factors such as diabetes mellitus, coronary artery disease, and hypertension were similar in patients who did or did not develop AKI. However, history of congestive heart failure and liver disease were more common in those patients developing AKI. Patients with AKI also had lower mean arterial pressure and were more often treated with vasopressors. Baseline serum creatinine and severity of illness as judged by baseline nonrenal APACHE III and SOFA scores were also greater for patients developing AKI (Table 1).
The unadjusted RR for AKI by strata of [TIMP-2]·[IGFBP7] defined by three cutoffs (0.3, 1.0, and 2.0) are shown in Figure 2. The absolute risk in the low stratum (≤ 0.3) was 2.7% increasing to 53.3% in the highest stratum (> 2.0) for an RR of 19.7 (95% CI, 4.3–69.4; p < 0.001). Patients with AKI had significantly higher levels of [TIMP-2]·[IGFBP7] than patients without AKI (p < 0.001). This effect of AKI on [TIMP-2]·[IGFBP7] levels was not modified by nonrenal SOFA (p = 0.70). In addition, nonrenal SOFA subgroup did not affect [TIMP-2]·[IGFBP7] values after adjusting for AKI status (p = 0.29).
The overall AUC for [TIMP-2]·[IGFBP7] for predicting AKI in this cohort was 0.84 (0.77–0.90). For patients with a nonrenal SOFA score of greater than 7, the [TIMP-2]·[IGFBP7] AUC was 0.85 (0.76–0.94) and similarly for patients with nonrenal SOFA of less than or equal to 7, the AUC was 0.84 (0.73–0.92) (Fig. 3). For comparison, the AUC for a serum creatinine measured at the same time as urinary [TIMP-2]·[IGFBP7] yielded AUC of 0.73 (0.59–0.85) for patients with a nonrenal SOFA score of greater than 7 and AUC of 0.77 (0.65–0.87) for patients with a nonrenal SOFA score less than or equal to 7. Similarly, estimated glomerular filtration rate yielded AUCs of 0.75 (0.61–0.87) and 0.76 (0.63–0.87).
Of note, we reported previously that for patients with sepsis in the Sapphire study (19), the individual marker performance for TIMP-2 was superior to IGFBP7, whereas the opposite was true for surgical patients. Here, in this combined cohort of sepsis patients, we again saw a better AUC for TIMP-2 0.84 (0.77–0.90) compared with that for IGFBP7 0.79 (0.72–0.86). Interestingly, these results were unchanged across low and high nonrenal SOFA subgroups.
[TIMP-2]·[IGFBP7] remained a strong predictor for AKI after adjustment for clinical variables, including severity of illness (APACHE III), nonrenal organ dysfunction (SOFA), body mass index, fluid output, and serum creatinine concentration. Addition of [TIMP-2]·[IGFBP7] to a clinical model significantly improved its predictive ability from 0.86 to 0.94 (p = 0.015) (Table 2). We also performed cfNRI and IDI. Table S1 (Supplemental Digital Content 1, http://links.lww.com/CCM/B893) shows that [TIMP-2]·[IGFBP7] improves the overall predictive ability of the model (statistically significant AUC increase, IDI, and cfNRI). ICU and hospital outcomes stratified by biomarker results are shown in Table S2 (Supplemental Digital Content 2, http://links.lww.com/CCM/B894).
Finally, operating characteristics (sensitivity, specificity, PPV, and NPV) at all three cut-offs (0.3, 1.0, and 2.0) are shown in Table 3. Sensitivity remains near 95% and specificity near 90% for the 0.3 and 2.0 cut-offs, respectively. Table S3 (Supplemental Digital Content 3, http://links.lww.com/CCM/B895) shows operating characteristics for [TIMP-2]·[IGFBP7] cutoffs ranging from 0.1 to 5.0 at 0.1 intervals.
For patients with sepsis, AKI is strongly associated with both short- and long-term adverse consequences (1, 2). Indeed, in a recent study of patients with septic shock, 60-day hospital mortality was 6.2% for patients without AKI, 16.8% for stage 1, and 27.7% for stage 2–3 (32). Early and adequate treatment of sepsis might prevent sepsis-induced AKI, attenuate AKI severity or might reduce the need for renal replacement therapy (RRT) (33, 34). Rapid identification of septic patients at high risk for developing AKI could substantially improve the therapeutic approach. In this context, specific and sensitive biomarkers of renal cell injury or stress could play an important role. We evaluated the performance of the novel urinary cell-cycle arrest biomarker test [TIMP-2]·[IGFBP7] in septic patients at high risk for developing AKI. We chose to assess risk of moderate to severe AKI because this severity (corresponding to KDIGO stage 2 and 3) has been shown to be associated with a significantly increased prevalence of clinically important outcomes such as receipt of RRT and in hospital death (35), as well as adverse effects on long-term survival specifically in patients with sepsis (1).
Sepsis is a challenging area for AKI biomarkers. NGAL and IL-18 have been examined as potential biomarkers of AKI, but both are strongly influenced by systemic inflammation thus degrading their specificity (36, 37). Although a small study recently reported better performance of NGAL (both plasma and urine) for AKI in patients with sepsis (AUCs, 0.83 and 0.89) (38), most studies have noted modest performance of biomarkers in this population. Recently, a sophisticated machine-learning analysis was performed using candidate biomarkers selected from extensive transcriptomic analysis to predict AKI on day 3 in pediatric patients with sepsis (37). Even after including AKI status on day 1 in the model (positive in half the cases), the AUC only reached 0.83 in the test cohort (37). These challenges are perhaps not surprising because the mechanisms of organ injury in sepsis may not be organ specific (37). However, our results indicate that the performance of [TIMP-2]·[IGFBP7] in patients with sepsis is not significantly confounded by nonrenal SOFA. Thus, clinicians can rely on these biomarkers for predicting AKI even in the presence of nonrenal organ failures. This is a notable advance in comparison with other biomarkers that are available around the world (39–42).
As alluded to in a recent study of remote ischemic preconditioning (43), biomarkers of cell-cycle arrest such as TIMP-2 and IGFBP7 may signal that the renal epithelium has been stressed and has shut down function but may still be able to recover without permanent injury to the organ. Importantly, both TIMP-2 and IGFBP7 appear to be able to signal in autocrine and paracrine fashions, thus spreading the “alarm” from the site of injury (15, 44). In terms of timing, this signal could be ideal as it may be early enough that management strategies can still alter the outcome. This is particularly important in septic AKI where delay and early timing remains a major issue (4, 32, 42).
Although not directly examined in these observational trials, we hypothesize that early use of biomarkers of cell-cycle arrest such as TIMP-2 and IGFBP7 could help clinicians intervene early on and thus improve outcomes for patients at risk of or with early evidence of sepsis-induced AKI (45–47). Early awareness that a patient with sepsis is about to develop a major organ failure could change a number of clinical decisions. For example, selection of antibiotics and dosing/monitoring of nephrotoxic medication (antibiotics or others) would be affected. The decision to give intravenous radiocontrast (especially intra-arterial) could be altered. Even the decision to discharge a patient or commence with a detailed diagnostic work-up would be influenced by the probability that a patient has or will have AKI (48). For example, consider a patient with sepsis admitted from the medical ward and started on broad spectrum antibiotics including empiric coverage with vancomycin. The morning after admission (9 hr later), the blood pressure and heart rate were normalized and cultures were pending. Now imagine that creatinine level of the patient was slightly elevated (1.2 mg/dL from 1.0), but urine output was adequate. If the patient had a urinary [TIMP-2]·[IGFBP7] test result more than 2.0, the risk of developing stage 2–3 AKI would be more than 50% and the patient’s clinical team might well wish to stop or dose-adjust the vancomycin and perhaps keep the patient in the ICU where fluids, urine output, and hemodynamics can be monitored more carefully. Conversely, if the result were less than 0.3, then the risk would be less than 3% and thus the patient could be safely continued on current regimen.
Of course, biomarkers will not take the place of clinical judgment and the question whether they offer information in addition to clinical variables is an important one. Although clinicians are unlikely to use statistical models at the bedside, the model shown in Table 2 represents the limits of information that can be derived from clinical variables. Indeed, the model used here was developed in this dataset and is therefore likely overtrained. As such, it represents and unrealistic “benchmark” to compare the biomarker. Nevertheless, the markers show added value even in this setting. Recent work examining the role of electronic surveillance for AKI has shown that although computer programs can help identification of AKI, this alone may not improve patient outcomes. Indeed, electronic alerts may be “too late” because they are based on creatinine that is too late. Wilson et al (49) examined an electronic alert based on KDIGO stage 1 but could not demonstrate changes in physician practice or patient outcome. Physicians may already be aware of an AKI event and so do not change their management. Importantly, the [TIMP-2]·[IGFBP-7] test was developed to assess risk for stage 2–3 AKI 12 hours prior to its clinical manifestation. This has two important potential benefits over the electronic alert. First, 12 hours is a long time in the ICU and could be a decisive period to discontinue nephrotoxins, investigate the source of sepsis, and improve resuscitation. Second, as many as 70% of AKI alerts will be for patients that never progress beyond stage 1 AKI. Alert fatigue is a large factor leading to clinicians ignoring alerts. The [TIMP-2]·[IGFBP-7] test was developed for stage 2–3 AKI, and thus it does not detect stage 1 conditions that are low risk for progression. Conversely, like all diagnostics, biomarkers for AKI should not be used in patient populations for which they were not developed. Low-risk patients (e.g., stable outpatients) will exhibit far greater false-positive rates compared with critically ill patients and would not be appropriate for the [TIMP-2]·[IGFBP-7] test.
Our study has important strengths. We analyzed data from two large and unrelated cohorts with two different methods of determining the AKI endpoint (KDIGO stage 2–3 and clinical adjudication). However, this could be viewed as limitation, and there are also other important limitations to this work. First, because the Sapphire and Topaz clinical trials were not specifically designed to examine sepsis-induced AKI, we did not collect information on the type of organism or on the timing of infection. Nonetheless, our results are consistent across both multicenter cohort studies. Second, long-term outcomes were not available or was quality of life assessed in the Topaz study. Third, and most importantly, we could not examine whether the availability of biomarker results would have changed patient management or outcomes. This question awaits further study. However, given the limited information in the literature about the performance of biomarkers in sepsis-induced AKI, despite its common occurrence, we believe that our results should inform clinical practice and future research.
The urinary [TIMP-2]·[IGFBP7] test provides accurate prediction of AKI in septic patients, and test performance is not affected by nonrenal organ dysfunction. As such, the test may extend the therapeutic window for renal protection and potentially enhance (future) therapeutic interventions to prevent or attenuate AKI.
We acknowledge the many staff, coordinators, and investigators whose work was essential to completion of this study. A complete list of Sapphire and Topaz investigators and support personnel is available online (http://www.ccm.pitt.edu/sapphire-investigators and http://www.ccm.pitt.edu/topaz-investigators).
1. Murugan R, Karajala-Subramanyam V, Lee M, et al; Genetic and Inflammatory Markers of Sepsis
(GenIMS) Investigators: Acute kidney injury
in non-severe pneumonia is associated with an increased immune response and lower survival. Kidney Int 2010; 77:527–535
2. Bagshaw SM, Uchino S, Bellomo R, et al; Beginning and Ending Supportive Therapy for the Kidney (BEST Kidney) Investigators: Septic acute kidney injury
in critically ill patients: Clinical characteristics and outcomes. Clin J Am Soc Nephrol 2007; 2:431–439
3. Mehta RL, Bouchard J, Soroko SB, et al; Program to Improve Care in Acute Renal Disease (PICARD) Study Group: Sepsis
as a cause and consequence of acute kidney injury
: Program to Improve Care in Acute Renal Disease. Intensive Care Med 2011; 37:241–248
4. Godin M, Murray P, Mehta RL: Clinical approach to the patient with AKI and sepsis
. Semin Nephrol 2015; 35:12–22
5. Schortgen F, Asfar P: Update in sepsis
and acute kidney injury
in 2014. Am J Respir Crit Care Med 2015; 191:1236–1231
6. Alobaidi R, Basu RK, Goldstein SL, et al: Sepsis
-associated acute kidney injury
. Semin Nephrol 2015; 35:2–11
7. Honore PM, Jacobs R, Joannes-Boyau O, et al: Septic AKI in ICU patients. diagnosis, pathophysiology, and treatment type, dosing, and timing: A comprehensive review of recent and future developments. Ann Intensive Care 2011; 1:32
8. Medeiros P, Nga HS, Menezes P, et al: Acute kidney injury
in septic patients admitted to emergency clinical room: Risk factors and outcome. Clin Exp Nephrol 2015; 19:859–866
9. Ahmed W, Memon JI, Rehmani R, et al: Outcome of patients with acute kidney injury
in severe sepsis
and septic shock treated with early goal-directed therapy in an intensive care unit. Saudi J Kidney Dis Transpl 2014; 25:544–551
10. Murray PT, Mehta RL, Shaw A, et al; ADQI 10 Workgroup: Potential use of biomarkers in acute kidney injury
: Report and summary of recommendations from the 10th
Acute Dialysis Quality Initiative consensus conference. Kidney Int 2014; 85:513–521
11. Haase M, Kellum JA, Ronco C: Subclinical AKI–an emerging syndrome with important consequences. Nat Rev Nephrol 2012; 8:735–739
12. KDIGO AKI Work Group: KDIGO clinical practice guideline for acute kidney injury
. Kidney Int Suppl 2012;17:1–138
13. Price PM, Safirstein RL, Megyesi J: The cell cycle and acute kidney injury
. Kidney Int 2009; 76:604–613
14. Boonstra J, Post JA: Molecular events associated with reactive oxygen species and cell cycle progression in mammalian cells. Gene 2004; 337:1–13
15. Seo DW, Li H, Qu CK, et al: Shp-1 mediates the antiproliferative activity of tissue inhibitor of metalloproteinase-2
in human microvascular endothelial cells. J Biol Chem 2006; 281:3711–3721
16. Rodier F, Campisi J, Bhaumik D: Two faces of p53: Aging and tumor suppression. Nucleic Acids Res 2007; 35:7475–7484
17. Yang QH, Liu DW, Long Y, et al: Acute renal failure during sepsis
: Potential role of cell cycle regulation. J Infect 2009; 58:459–464
18. Seo DW, Li H, Guedez L, et al: TIMP-2 mediated inhibition of angiogenesis: An MMP-independent mechanism. Cell 2003; 114:171–180
19. Kashani K, Al-Khafaji A, Ardiles T,, et al: Discovery and validation of cell cycle arrest biomarkers in human acute kidney injury
. Crit Care 2013;6:17
20. Bihorac A, Chawla LS, Shaw AD, et al: Validation of cell-cycle arrest biomarkers for acute kidney injury
using clinical adjudication. Am J Respir Crit Care Med 2014; 189:932–939
21. Meersch M, Schmidt C, Van Aken H, et al: Urinary TIMP-2 and IGFBP7 as early biomarkers of acute kidney injury
and renal recovery following cardiac surgery. PloS One 2014;9:e93460
22. Nisula S, Yang R, Kaukonen KM, et al; FINNAKI Study Group: The urine protein NGAL predicts renal replacement therapy, but not acute kidney injury
or 90-day mortality in critically ill adult patients. Anesth Analg 2014; 119:95–102
23. Tu Y, Wang H, Sun R, et al: Urinary netrin-1 and KIM-1 as early biomarkers for septic acute kidney injury
. Ren Fail 2014; 36:1559–1563
24. Hoste EA, McCullough PA, Kashani K, et al; Sapphire Investigators: Derivation and validation of cutoffs for clinical use of cell cycle arrest biomarkers. Nephrol Dial Transplant 2014; 29:2054–2061
25. Calandra T, Cohen J; International Sepsis
Forum Definition of Infection in the ICU Consensus Conference: The international sepsis
forum consensus conference on definitions of infection in the intensive care unit. Crit Care Med 2005; 33:1538–1548
26. von Elm E, Altman DG, Egger M, et al; STROBE Initiative: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Ann Intern Med 2007; 147:573–577
27. Bossuyt PM, Reitsma JB, Bruns DE, et al; STARD Group: Toward complete and accurate reporting of studies of diagnostic accuracy: The STARD initiative. Acad Radiol 2003; 10:664–669
28. Knaus WA, LaRosa JC, Marks RD, et al: The value of commercial funding in health services research: The case of the APACHE III methodology. Health Serv Res 1994; 28:673–678; discussion 683
29. Vincent JL, Moreno R, Takala J, et al: The SOFA (Sepsis
-related Organ Failure Assessment) score to describe organ dysfunction
/failure. On behalf of the Working Group on Sepsis
-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996; 22:707–710
30. Agresti A, Min Y: On small-sample confidence intervals for parameters in discrete distributions. Biometrics 2001; 57:963–971
31. Core Team R: A language and environment for statistical computing. 2014. Vienna, Austria R Foundation for Statistical Computing. Available at: http://wwwR-project.org/
. Accessed August 10, 2015
32. Kellum JA, Chawla LS, Keener C, et al: The effects of alternative resuscitation strategies on acute kidney injury
in patients with septic shock. Am J Resp Crit Care Med 2015; 193:281–287
33. Peters E, Heemskerk S, Masereeuw R, et al: Alkaline phosphatase: A possible treatment for sepsis
-associated acute kidney injury
in critically ill patients. Am J Kidney Dis 2014; 63:1038–1048
34. Honoré PM, Jacobs R, Boer W, et al: New insights regarding rationale, therapeutic target and dose of hemofiltration and hybrid therapies in septic acute kidney injury
. Blood Purif 2012; 33:44–51
35. Hoste EA, Bagshaw SM, Bellomo R, et al: Epidemiology of acute kidney injury
in critically ill patients: The multinational AKI-EPI study. Intensive Care Med 2015; 41:1411–1423
36. Mårtensson J, Bell M, Oldner A, et al: Neutrophil gelatinase-associated lipocalin in adult septic patients with and without acute kidney injury
. Intensive Care Med 2010; 36:1333–1340
37. Parikh CR, Abraham E, Ancukiewicz M, et al: Urine IL-18 is an early diagnostic marker for acute kidney injury
and predicts mortality in the intensive care unit. J Am Soc Nephrol 2005; 16:3046–3052
38. Dai X, Zeng Z, Fu C, et al: Diagnostic value of neutrophil gelatinase-associated lipocalin, cystatin C, and soluble triggering receptor expressed on myeloid cells-1 in critically ill patients with sepsis
-associated acute kidney injury
. Crit Care 2015; 19:223
39. Wong HR, Cvijanovich NZ, Anas N, et al: A multibiomarker-based model for estimating the risk of septic acute kidney injury
. Crit Care Med 2015; 43:1646–1653
40. Bagshaw SM, Langenberg C, Haase M, et al: Urinary biomarkers in septic acute kidney injury
. Intensive Care Med 2007; 33:1285–1296
41. Honore PM, Jacobs R, Joannes-Boyau O, et al: Biomarkers for early diagnosis of AKI in the ICU: Ready for prime time use at the bedside? Ann Intensive Care 2012; 2:24
42. Honore PM, Joannes-Boyau O, Boer W, et al: Acute kidney injury
in the ICU: Time has come for an early biomarker kit! Acta Clin Belg 2007; 62(Suppl 2):318–321
43. Zarbock A, Schmidt C, Van Aken H, et al; RenalRIPC Investigators: Effect of remote ischemic preconditioning on kidney injury among high-risk patients undergoing cardiac surgery: A randomized clinical trial. JAMA 2015; 313:2133–2141
44. Zuo S, Liu C, Wang J, et al: IGFBP-rP1 induces p21 expression through a p53-independent pathway, leading to cellular senescence of MCF-7 breast cancer cells. J Cancer Res Clin Oncol 2012; 138:1045–1055
45. Wajapeyee N, Serra RW, Zhu X, et al: Oncogenic BRAF induces senescence and apoptosis through pathways mediated by the secreted protein IGFBP7. Cell 2008; 132:363–374
46. Zager RA: ‘Biologic memory’ in response to acute kidney injury
: Cytoresistance, toll-like receptor hyper-responsiveness and the onset of progressive renal disease. Nephrol Dial Transplant 2013; 28:1985–1993
47. Palud A, Parmentier-Decrucq E, Pastre J, et al: Evaluation of endothelial biomarkers as predictors of organ failures in septic shock patients. Cytokine 2015; 73:213–218
48. Kellum JA, Bellomo R, Ronco C. Does this patient have acute kidney injury
? An AKI checklist. Intensive Care Med 2015; 42:96–99
49. Wilson FP, Shashaty M, Testani J, et al: Automated, electronic alerts for acute kidney injury
: A single-blind, parallel-group, randomised controlled trial. Lancet 2015; 385:1966–1974