Plasma neutrophil gelatinase-associated lipocalin and factors related to acute kidney injury and mortality in critically ill cancer patients

Bertha M Córdova-Sánchez¹, Erika B Ruiz-García², Alicia López-Yañez², Mireya Barragan-Dessavre¹, Andoreni R Bautista-Ocampo¹, Abelardo Meneses-García¹, Angel Herrera-Gómez¹ and Silvio A Ñamendys-Silva^1,3,4

¹Department of Critical Care Medicine, Instituto Nacional de Cancerología, Mexico City 14080, Mexico

²Translational Medicine Research Laboratory, Instituto Nacional de Cancerologia, Ciudad de Mexico 14080, Mexico

³Department of Critical Care Medicine, Fundación Clínica Médica Sur, Mexico City 14050, Mexico

⁴Department of Critical Care Medicine, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán, Mexico City 14000, Mexico

Correspondence to: Professor Silvio A Ñamendys-Silva. Email: snamendyss@medicasur.org.mx; snamendys@incan.edu.mx

Abstract

Rationale: Acute kidney injury (AKI) is a frequent complication in critically ill cancer patients.

Objectives: To assess plasma neutrophil gelatinase-associated lipocalin (NGAL) levels and risks factors associated with AKI and mortality.

Methods: We recruited 96 critically ill cancer patients and followed them prospectively. Plasma NGAL levels were determined at intensive care unit (ICU) admission and at 48 hours. We generated receiver operating characteristic curves to assess the ability of NGAL to predict AKI. Logistic regression analysis was performed to determine risks factors associated with AKI. Cox-regression analysis was performed to evaluate 6-month mortality.

Measurements and main results: From 96 patients, 60 (63%) developed AKI and 33 (55%) were classified as stages 2 and 3. In patients without AKI at admission, plasma NGAL levels revealed an area under the curve (AUC) = 0.522 for all AKI stages and an AUC = 0.573 for stages 2 and 3 AKI (85% sensitivity and 67% specificity for a 50.66 ng/mL cutoff). We identified sequential organ failure assessment (SOFA) score (without renal parameters) at admission as an independent factor for developing stages 2 and 3 AKI, and haemoglobin as a protective factor. We observed that metastatic disease, dobutamine use and stage 3 AKI were independent factors associated with 6-month mortality.

Conclusions: In our cohort of critically ill cancer patients, NGAL did not predict AKI. SOFA score was a risk factor for developing AKI, and haemoglobin level was a protective factor for developing AKI. The independent factors associated with 6-month mortality included metastatic disease, dobutamine use, lactate and stage 3 AKI.

Keywords: plasma neutrophil gelatinase-associated lipocalin, acute kidney injury, critically ill cancer patients, intensive care, critical care

Copyright: © the authors; licensee ecancermedicalscience. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Published: 14/02/2019; Received: 09/07/2018

Introduction

Acute kidney injury (AKI) is a common complication in cancer patients due to exposure to contrast media, cytotoxic agents, surgical treatment, antibiotics and nonsteroidal anti-inflammatory drugs. In addition, critically ill cancer patients are at increased risk for AKI associated with volume depletion and sepsis [1].

Studies including critically ill patients with hematologic and solid neoplasms have identified an incidence of AKI ranging from 22 to 100%, using the criteria proposed by the Acute Kidney Injury Network and the organisation Kidney disease: Improving Global Outcomes (KDIGO), additionally the presence of AKI is associated with higher mortality [2–4].

Serum creatinine (SCr) and urinary output (UO) are traditional markers of AKI; however, decreased muscle mass, inflammation, volume expansion or medications can alter SCr production, limiting its sensitivity. Therefore, biomarkers that are more sensitive are under investigation [1].

Plasmatic neutrophil gelatinase-associated lipocalin (NGAL) has demonstrated good predictive performance in children [5]. Nevertheless, in adult critically ill patients, the results are heterogeneous, with an area under the curve (AUC) of less than 0.7 in recent studies [6, 7].

Most studies investigating AKI biomarkers in cancer patients focus on early detection of nephrotoxicity mediated by anticancer drugs [8, 9]. However, in critically ill cancer patients, inflammatory biomarkers, such as NGAL, may be influenced by numerous sources of inflammation and do not reflect tubular damage. Therefore, we aim to explore the predisposing factors of AKI and the performance of NGAL in a population of critically ill cancer patients.

Methods

We performed a prospective observational study from April 2014 to July 2015, at the Instituto Nacional de Cancerología, a tertiary care cancer centre. The institutional ethics committee approved this study (Prot. No. 014/011/SCI/CEI/860). Written informed consent was obtained from all participants or primary caregivers. We included 96 critically ill cancer patients at the time of admission to the ICU. We excluded patients who had an ICU stay of ≤24 hours or patients under 18 years, with known chronic kidney disease (CKD), long-term or acute renal replacement therapy (RRT), previous kidney transplantation and renal or urologic cancer (Figure 1). ICU readmissions were not considered. We estimated glomerular filtration rate (GFR) using the CKD Epidemiology Collaboration equation, CKD was defined as GFR < 60 mL/min/1.73m² for at least 3 months prior to admission [10].

Researchers also recorded demographic characteristics, comorbidities, cancer status, sepsis, mechanical ventilation, acute physiology and chronic health evaluation (APACHE II) [11], sequential organ failure assessment (SOFA) [12] score and laboratory values. In addition, registered days with mechanical ventilation, days with norepinephrine ICU stay and death at 180 days were recorded.

Acute kidney injury criteria

We defined AKI according to KDIGO Clinical Practice Guidelines. Stage 1 AKI was defined as 0.3 mg/dL SCr elevation above baseline within 48 hours, and/or UO less than 0.5 mL/kg/h for 6–12 hours. Stage 2 AKI was defined as SCr with a 2–2.9-fold increase compared with the baseline value, and/or UO less than 0.5 mL/kg/h for 12–24 hours. Stage 3 AKI was defined as SCr greater than three-fold the baseline value, UO less than 0.3 mL/kg/h for greater than 24 hours, anuria for greater than 12 hours or RRT requirement [13].

We considered baseline creatinine as the lowest value in the 3 months preceding hospitalisation. In patients without historical values, the baseline was calculated by estimating the eGFR at 75 mL/min/1.73 m² [13].

Figure 1. Cohort selection flow diagram.

Plasma collection and processing

Plasma concentrations of NGAL were measured at 24 hours and 48 hours during the stay in the ICU. Plasma samples were freshly collected in tubes with heparin, placed on ice and centrifuged for 15 minutes. The supernatant was stored at −20°C until assayed. Plasma samples were diluted 20 times, with a calibrator included in the commercial kit Human Lipocalin kit-1/NGAL (R&D Systems). The conjugated sample remained stored at 2°C−8°C. The other reagents were left at room temperature before use. The washing buffer was incubated at room temperature and gently stirred until the crystals dissolved. We diluted 20 mL with deionised water to prepare 500 mL of washing buffer. Twenty millilitres of the calibrator RD5-24 were diluted in 80 mL of deionised water to prepare 100 mL solution. The coloured reagents A and B were mixed in equal volumes 15 minutes before use and protected from light. The lipocalin-2 standard was reconstituted with 1 mL of deionised water and incubated for 15 minutes at room temperature. Subsequently, 100 μL of the standard and 900 μL were used for the standard curve. Samples and standards were performed in triplicate.

We used the software elisaanalysis.com to perform the analysis and curve adjustment with four logistic parameters. W obtained r² = 0.9967093.

Statistical methods

Continuous variables are expressed as the mean (± standard deviation) or median (interquartile range 25%–75%). Categorical variables are expressed as proportions. To compare patients with and without AKI, we performed Student’s t-test or Mann-Whitney U test for continuous variables, as appropriate, and X² test was employed for categorical variables. We constructed receiver operating characteristic (ROC) curves to estimate the sensitivity and specificity of NGAL to predict AKI within 72 hours. We performed a logistic regression analysis to determine risks factors associated with AKI, as defined by KDIGO during hospitalisation (expressed as odds ratio and 95% confidence interval: OR, 95% CI). We constructed a univariate Cox-regression analysis to assess factors related to time to death (expressed as hazard ratio: HR). Variables with p values <0.05 were included in multivariate analysis.

Results

Acute kidney injury

We included 96 critically ill cancer patients. Specifically, 60 patients (63%) developed AKI and 33 (55%) were classified as stages 2 and 3. None of these patients required RRT. Comparison of general characteristics between patients with AKI and patients without AKI is presented in Table 1.

In total, 28 patients already had AKI at ICU admission and 32 developed AKI during their ICU stay. From the 68 patients without AKI at admission, 32 (47%) developed AKI: 19 (28%) were stage 1 patients, 10 (15%) were stage 2 patients and 3 (4%) were stage 3 patients.

In 68 patients without AKI at admission, plasma NGAL levels exhibited an AUC = 0.522 (95% CI: 0.383–0.661, p = 0.759) for all AKI stages, with an 81.3% sensitivity and 69.5% specificity for a 45.4 ng/mL cutoff. We observed an AUC = 0.573 (95% CI: 0.416–0.731, p = 0.413) for stages 2 and 3 AKI with an 84.6% sensitivity and 67.3% specificity for a 50.66 ng/mL cutoff (Figure 2).

The OR of AKI decreases depending on the increase in haemoglobin level, therefore, we constructed a ROC curve to identify the haemoglobin threshold related to a lesser AKI risk, resulting in ≥ 8.8 g/dL. Then, we categorised patients into two groups and observed that patients with haemoglobin level ≥ 8.8 g/dL had decreased risk of AKI (OR 0.138, 95% CI: 0.036–0.521, p = 0.004).

We identified the SOFA score (without renal parameters) at admission as an independent factor (OR 1.30, 95% CI: 1.00–1.68, p = 0.049) for developing stages 2 and 3 AKI. Male gender exhibited a trend towards significance (OR 4.6, 95% CI: 0.94–22.12, p = 0.06), and haemoglobin appeared as a protective factor (OR 0.76, 95% CI: 0.42–0.94, p = 0.026) (Table 2).

We assessed variables for collinearity and interactions and constructed a multivariate model. The AUC of the model was 0.892 (95% CI: 0.807–0.978, p = 0.000) for predict stages 2 and 3 AKI and the p-value for Hosmer-Lemeshow test was 0.822.

Factors associated with 6-month mortality

We included 96 patients in mortality analysis (Table 3). We founded that age (HR 1.03, 95% CI: 1.00–1.05, p = 0.043), metastatic disease (HR 4.04, 95% CI: 1.45–11.29, p = 0.008), dobutamine use (HR 21.92, 95% CI: 3.68–130.5, p = 0.001), lactate at admission (HR 1.03, 95% CI: 1.01–1.05, p = 0.003) and stage 3 AKI (HR 2.83, 95% CI: 1.07–7.48, p = 0.036) were independent factors for 6-month mortality. In multivariate analysis, we observed that NGAL measured at 48 hours of admission had an HR = 1.047 (95% CI: 0.993–1.104, p = 0.087) for each 10 ng/mL that was not statistically significant to statistical significance to predict 6-month mortality.

The value of haemoglobin at the time of ICU admission demonstrated statistical significance as a protective factor for mortality (HR 0.81, 95% CI: 0.66–0.99, p = 0.047).

Discussion

During the recruitment period, we observed an AKI prevalence of 63%, corresponding with other studies in critically ill oncology patients using definitions such as AKIN and KDIGO [2, 3].

In our group of cancer patients without AKI at the time of admission to the ICU, plasma NGAL displayed a low diagnostic performance with an AUC of 0.522 for all AKI stages. Plasma NGAL also predicted moderate to severe AKI with an AUC of 0.573 for stages 2 and 3. This finding is consistent with various previous studies in critically ill adult patients, in whom the diagnostic performance of NGAL is more variable compared with paediatric patients [5, 14] and adult patients exposed to contrast media or cardiac surgery [15].

Studies designed to assess the diagnostic performance of NGAL to predict AKI in critically ill patients have revealed a low AUC ranging from 0.44 to 0.7 for plasma NGAL [6, 7, 16] and values from 0.5 to 0.8 for urinary NGAL [7, 16–18].

Generally, these studies compared the diagnostic accuracy of biomarkers with SCr level and did not employ precise methods, such as renal scintigraphy, to verify the reduction in glomerular filtration [15]. The results can be conflicting in ICU patients, given that hypercatabolism, malnutrition, fluid overload and drugs that affect tubular excretion can modify SCr [19, 20].

On the other hand, biomarkers utility depends on sample timing after renal insult. However, in ICU patients, AKI is frequently multifactorial and it is not possible to determine the exact moment of the renal insult [21].

Table 1. Clinical characteristics of patients on admission to the ICU.

Figure 2. (A) Plasmatic NGAL levels in patients at any stage based on KDIGO criteria. (B) Patients meeting the KDIGO criteria for stages 2 and 3.

Table 2. Factors associated with AKI.

Moreover, the burden of comorbidities, such as cardiovascular and renal disease may be the cause of the reported difference in NGAL performance between adults and children [18]. Additionally, oncology patients possess factors that could alter NGAL expressions, such as exposure to chronic inflammation and anti-neoplastic drugs that accumulate in the renal cortex and predispose to kidney damage before acute insult [22, 23].

Table 3. Cox proportional hazards analyses of factors associated with increased 6-month mortality rate.

Recently, novel biomarkers of kidney injury have shown promising results to predict AKI but these biomarkers are prone to fail in a heterogeneous population of adult patients. Therefore, it is necessary to use predictive models that consider epidemiological and clinical factors [24, 25]. We found that SOFA score at admission may be useful for predicting the development of AKI and that haemoglobin level is a protective factor. This finding is consistent with previous studies demonstrating that AKI in critically ill cancer patients is associated with multiple organ failures [26], and anaemia is a risk factor for AKI in general ICU patients [24].

Han et al [27] previously demonstrated that anaemia defined by haemoglobin < 10.5 g/dL is an independent risk factor for AKI and mortality in a large cohort of critically ill patients, including less than 18% with neoplastic disease. In our population, we observed a mean haemoglobin level of 9.6 g/dL and a threshold of 8.8 g/dL as a risk factor for AKI. This lower haemoglobin level could be related to cytopenia in oncology patients.

Critically ill patients are exposed to hypoperfusion, and the presence of anaemia decreases the delivery of oxygen to tissues, including renal parenchyma, which normally receives 20%–25% of cardiac output. Additionally, we observed that a higher level of haemoglobin relates to lower mortality rates. The relation between anaemia and mortality occurs because anaemia can aggravate organic dysfunctions [27].

In a six-month mortality analysis, we identified independent factors, such as age, metastatic disease, and severe AKI, previously described in this group of patients [28, 29]. We also identified that the use of dobutamine was statistically significant. One possible explanation is that almost one-third of our patients were septic, and actual guidelines for sepsis management suggest the use of dobutamine in the presence of myocardial dysfunction manifested by elevated cardiac filling pressures, low cardiac output, and signs of hypoperfusion, which may reflect illness severity. However, dobutamine increases cardiac output and splanchnic blood flow. In addition, dobutamine has not demonstrated a benefit in mortality and could even decrease survival [30, 31].

Based on the inclusion of SOFA score and haemoglobin in a risk model without considering renal biomarkers, we identified an AUC = 0.892. Although we obtained this result in a small cohort of patients, these findings can be useful in the construction of future models in critically ill cancer patients, for further validation in larger cohorts. We observed that NGAL measured at 48 hours of admission had an HR = 1.047 (95% CI: 0.993−1.104 for each 10 ng/mL, p-value = 0.087) that was not statistically significant to predict 6-month mortality. Increased NGAL levels at 48 hours could be a manifestation of the persistence of inflammation independent of kidney function [32–34].

Our study has several limitations, such as the number of patients, the heterogeneity of the population and the inability to determine the precise moment at which the renal insult occurred in a cohort of medical and surgical critically ill patients. However, most studies about renal biomarkers in cancer patients evaluate their ability to predict AKI in haemodynamically stable patients who receive nephrotoxic antineoplastic drugs. Currently, cancer patients are more frequently admitted to ICUs; therefore, exploration of the diagnostic performance of new instruments is required to improve patient care.

Conclusion

In our cohort of critically ill cancer patients, NGAL did not predict AKI. SOFA score was a risk factor for developing AKI, and haemoglobin level was a protective factor for developing AKI. The independent factors associated with 6-month mortality included metastatic disease, dobutamine use, lactate and stage 3 AKI.

Author contributions

Bertha M. Córdova-Sánchez, Silvio A. Ñamendys-Silva had full access to all data used in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Data responsible: Erika B. Ruiz-García, Alicia López-Yañez, Mireya Barragan-Dessavre, Andoreni R. Bautista-Ocampo, Abelardo Meneses-García, Angel Herrera-Gómez. Study concept and design: Silvio A. Ñamendys-Silva. Statistical analysis: Bertha M. Córdova-Sánchez, and Silvio A. Ñamendys-Silva. Interpretation of data: All authors. Drafting of the manuscript: Bertha M. Córdova-Sánchez, and Silvio A. Ñamendys-Silva. Critical revision of the manuscript for important intellectual content: All authors. Obtained funding: Abelardo Meneses-García, Angel Herrera-Gómez, Erika B. Ruiz-García, and Silvio A. Ñamendys-Silva. Study supervision: Silvio A. Ñamendys-Silva.

Scientific knowledge on the subject

AKI is a frequent complication in ICU patients. Studies designed to analyze the ability of NGAL to predict AKI in ICU have shown variable results. Oncological patients exhibit factors that can alter the expression of NGAL. However, studies of biomarkers in oncological patients mostly include haemodynamically stable patients exposed to cytotoxic agents.

What this study adds to the field

This study demonstrates that in critically ill cancer patients the plasma NGAL was not a good predictor of AKI. We identified SOFA score at admission as a risk factor and haemoglobin level as a protective factor for developing AKI.

Conflicts of interest

None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.

Funding

None.

Notes

This study was presented in part at the The American Thoracic Society 2018 International Conference, 18-23 May 2018 in San Diego, CA, USA.

References

1. Lameire N, Vanholder R, and Van Biesen W, et al (2016) Acute kidney injury in critically ill cancer patients: an update Crit Care 20(1) 209 https://doi.org/10.1186/s13054-016-1382-6 PMID: 27480256 PMCID: 4969681

2. Samuels J, Ng CS, and Nates J, et al (2011) Small increases in serum creatinine are associated with prolonged ICU stay and increased hospital mortality in critically ill patients with cancer Support Care Cancer 19 1527–1532 https://doi.org/10.1007/s00520-010-0978-7 PMCID: 3037438

3. Darmon M, Vincent F, and Canet E, et al (2015) Acute kidney injury in critically ill patients with haematological malignancies: results of a multicenter cohort study from the Groupe de Recherché and Reanimation Respiratoire en Onco-Hematologie Nephrol Dial Transplant 30 2006–2013 https://doi.org/10.1093/ndt/gfv372 PMID: 26597921 PMCID: 4832999

4. Maccariello E, Valente C, and Nogueira L, et al (2011) Outcomes of cancer and non-cancer patients with acute kidney injury and need of renal replacement therapy admitted to general intensive care units Nephrol Dial Transplant 26 537–543 https://doi.org/10.1093/ndt/gfq441

5. Mishra J, Dent C, and Tarabishi R, et al (2005) Neutrophil gelatinase-associated lipocalin as a biomarker for acute renal injury after cardiac surgery Lancet 365 1231–1238 https://doi.org/10.1016/S0140-6736(05)74811-X PMID: 15811456

6. Kashani K, Al-Khafaji A, and Ardiles T, et al (2013) Discovery and validation of cell cycle arrest biomarkers in human acute kidney injury Crit Care 17 R25 https://doi.org/10.1186/cc12503 PMID: 23388612 PMCID: 4057242

7. Aydogdy M, Gürsel G, and Sancak B, et al (2013) The use of plasma and urine neutrophil gelatinase-associated lipocalin (NGAL) and cystatin C in early diagnosis of septic acute kidney injury in critically ill patients Dis Markers 34(4) 237–246 https://doi.org/10.1155/2013/740351

8. Lin HY, Lee SC, and Lin SF, et al (2013) Urinary neutrophil gelatinase-associated lipocalin levels predict cisplatin-induced acute kidney injury better than albuminuria or urinary cystatin C levels Kaohsiung J Med Sci 29 304–311 https://doi.org/10.1016/j.kjms.2012.10.004 PMID: 23684135

9. Shahbazi F, Sadighi S, and Dashti-Khavidaki S, et al (2015) Urine ration of neutrophil gelatinase-associated lipocalin to creatinine as a marker for early detection of cisplatin-associated nephrotoxicity Iran J Kidney Dis 9 306–310 PMID: 26174458

10. Levey AS, Stevens A, and Schmid CH, et al (2009) A new equation to estimate glomerular filtration rate Ann Intern Med 150(9) 604–612 https://doi.org/10.7326/0003-4819-150-9-200905050-00006 PMID: 19414839 PMCID: 2763564

11. Knaus WA, Draper EA, and Wagner DP, et al (1985) APACHE II: a severity of disease classification system Crit Care Med 13(10) 818–829 https://doi.org/10.1097/00003246-198510000-00009 PMID: 3928249

12. Vincent JL, Moreno R, and Takala J, et al (1996) The SOFA (sepsis-related organ failure assessment) score to describe organ dysfunction/failure. On behalf of the working group on sepsis-renal problems of the European society of intensive care medicine Intensive Care Med 22(7) 707–710 https://doi.org/10.1007/BF01709751 PMID: 8844239

13. KDIGO AKI workgroup (2012) Kidney disease: improving global outcomes (KDIGO) clinical practice guideline for acute kidney injury Kidney Int 2(1) S1–S138

14. Dent CL, Ma Q, and Dastrala S, et al (2007) Plasma neutrophil gelatinase-associated lipocalin predicts acute kidney injury, morbidity and mortality after pediatric cardiac surgery: a prospective uncontrolled cohort study Crit Care 11 R127 https://doi.org/10.1186/cc6192 PMID: 18070344 PMCID: 2246223

15. Haase M, Bellomo R, and Devarajan P, et al (2009) Accuracy of neutrophil gelatinase-associated lipocalin (NGAL) in diagnosis and prognosis in acute kidney injury: a systematic review and meta-analysis Am J Kidney Dis 54(6) 1012–1024 https://doi.org/10.1053/j.ajkd.2009.07.020 PMID: 19850388

16. De Geus HR, Bakker J, and Lesaffre EM, et al (2011) Neutrophil gelatinase-associated lipocalin at ICU admission predicts for acute kidney injury in adult patients Am J Respir Crit Care Med 183(7) 907–914 https://doi.org/10.1164/rccm.200908-1214OC

17. De Geus HR, Fortrie G, and Betjes MG, et al (2013) Time of injury affects urinary biomarker predictive values for acute kidney injury in critically ill, non-septic patients BMC Nephrol 14 273 https://doi.org/10.1186/1471-2369-14-273 PMID: 24321290 PMCID: 3878913

18. Liangos O, Tighiouart H, and Perianayagam MC, et al (2009) Comparative analysis of urinary biomarkers for early detection of acute kidney injury following cardiopulmonary bypass Biomarkers 14(6) 423–431 https://doi.org/10.1080/13547500903067744 PMID: 19572801 PMCID: 2743298

19. Putucheary ZA, Rawal J, and McPhail M, et al (2013) Acute skeletal muscle wasting in critical illness JAMA 310 1591–1600 https://doi.org/10.1001/jama.2013.278481

20. Macedo E, Bouchard J, and Soroko SH (2010) Program to improve care in acute renal disease study. Fluid accumulation recognition and staging of acute kidney injury in critically-ill patients Crit Care 14 R82 https://doi.org/10.1186/cc9004

21. Ralib AM, Pickering JW, and Shaw GM, et al (2014) The clinical utility window for acute kidney injury biomarkers in the critically ill Crit Care 18(6) 601 https://doi.org/10.1186/s13054-014-0601-2 PMID: 25366893 PMCID: 4255650

22. Luke DR, Vadiei K, and López-Berenstein G (1992) Role of vascular congestion in cisplatin-induced acute renal failure in the rat Nephrol Dial Transplant 7(1) 1 PMID: 1316576

23. Hartmann JR, Kollmannsberger C, and Kanz L, et al (1999) Platinum organ toxicity and possible prevention in patients with testicular cancer Int J Cancer 83(6) 866–869 PMID: 10597214

24. Malhotra R, Kashani KB, and Macedo E, et al (2017) A risk prediction score for acute kidney injury in the intensive care unit Nephrol Dial Transplant 32(5) 814–822 https://doi.org/10.1093/ndt/gfx026 PMID: 28402551

25. Cruz DN, Ferrer-Nadal A, and Piccinni P, et al (2014) Utilization of small changes in serum creatinine with clinical risk factors to assess the risk of AKI in critically ill adults Clin J Am Soc Nephrol 9 663–672 https://doi.org/10.2215/CJN.05190513 PMID: 24677553 PMCID: 3974351

26. Soares M, Salluh JI, and Carvalho MS, et al (2006) Prognosis of critically ill patients with cancer and acute renal dysfunction J Clin Oncol 24(24) 4003–4010 https://doi.org/10.1200/JCO.2006.05.7869 PMID: 16921054

27. Han SS, Baek SH, and Ahn SY, et al (2015) Anemia is a risk factor for acute kidney injury and long-term mortality in critically ill patients Tohoku J Exp Med 237(4) 287–295 https://doi.org/10.1620/tjem.237.287 PMID: 26607258

28. Ostermann M, Ferrando-Vivas P, and Gore C, et al (2017) Characteristics and outcome of cancer patients admitted to the ICU in England, Wales, and Northern Ireland and National Trends between 1997 and 2013 Crit Care Med 45(10) 1668–1676 https://doi.org/10.1097/CCM.0000000000002589 PMID: 28682838

29. Córdova-Sánchez BM, Herrera-Gómez A, and Ñamendys-Silva SA (2016) Acute kidney injury classified by serum creatinine and urine output in critically ill cancer patients Biomed Res Int 2016 6805169 https://doi.org/10.1155/2016/6805169 PMID: 27803928 PMCID: 5075588

30. Wilkman E, Kaukonen KM, and Pettilä V, et al (2013) Association between inotrope treatment and 90-day mortality in patients with septic shock Acta Anaesthesiol Scand 57(4) 431–442 https://doi.org/10.1111/aas.12056 PMID: 23298252

31. Sato R and Nasu M (2017) Time to re-think the use of dobutamine in sepsis J Intensive Care 5 65 https://doi.org/10.1186/s40560-017-0264-6 PMID: 29201378 PMCID: 5699177

32. Cowland JB and Borregaard N (1997) Molecular characterization and pattern of tissue expression of the gene for neutrophil gelatinase-associated lipocalin from humans Genomics 45 17–23 https://doi.org/10.1006/geno.1997.4896 PMID: 9339356

33. Cai L, Rubin J, and Han W, et al (2010) The origin of multiple molecular forms in urine of HNL/NGAL Clin J Am Soc Nephrol 5 2229–2235 https://doi.org/10.2215/CJN.00980110 PMID: 20829422 PMCID: 2994084

34. Martensson J and Bellomo R (2014) The rise and fall of NGAL in acute kidney injury Blood Purif 37 304-10 https://doi.org/10.1159/000364937 PMCID: 25170751