Research and Opinion in Anesthesia & Intensive Care

ORIGINAL ARTICLE
Year
: 2018  |  Volume : 5  |  Issue : 1  |  Page : 8--14

Urinary soluble triggering receptor expressed on myeloid cells-1 is an early predictor for sepsis outcome and sepsis-induced acute kidney injury


Ahmed Farghal1, Hossam M Sherif1, Adel Al Sisi1, Sameh Al-Maraghy2,  
1 Critical Care Medicine Department, Faculty of Medicine, Cairo University, Cairo, Egypt
2 Critical Care Medicine Department, Faculty of Medicine, Beni Suef University, Beni Suef, Egypt

Correspondence Address:
Hossam M Sherif
Critical Care Center, Cairo University Hospitals, El Manial, Cairo, 11562
Egypt

Abstract

Background Recent reports had indicated the usefulness of urinary soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) as a prognostic marker for sepsis and sepsis-induced acute kidney injury (AKI). Objective This study aimed to detect the prognostic value of urinary sTREM-1 in sepsis in terms of the clinical course, development of AKI, and the survival rate. Patients and methods Thirty critically ill patients with sepsis were included (57.6±7.5 years, 18 men), in addition to a group of 10 controls (45.6±3.5 years, six men). Urinary sTREM-1 and C-reactive protein serum levels were estimated on admission, and days 3 and 7. The sepsis-related organ failure assessment scoring system was calculated at baseline and daily until discharge, mortality, or up to 28 days. Length of stay in ICU, need for vasopressors, mechanical ventilation or hemodialysis, and development of AKI and the outcomes were recorded. Results Compared with stable patients, patients who required vasopressors (23 patients) or hemodialysis (patient 4) showed significantly higher sTREM-1 values (4.06±1.22 vs. 2.86±0.51 ng/ml, P<0.001, and 5.27±1.2 vs. 3.55±1.05 ng/ml, P<0.05). Fair correlation could be detected between the sepsis-related organ failure assessment scoring system and sTREM-1 on day 1 and 7 (r=0.45, P<0.05 and r=0.47, P<0.05). In patients who developed AKI (12 patients), sTREM-1 showed significantly higher values than those who did not develop AKI (4.37±1.34 vs. 3.39±0.95 ng/ml, P<0.05). Compared with the survivors, the nonsurvivors (14 patients) showed significantly higher sTREM-1 values (4.6±1.14 vs. 2.96±0.52 ng/ml, P<0.001). The area under the curve for sTREM-1 to predict AKI on day 1 was 0.73 (95% confidence interval: 0.53–0.92), with the best cut-off value at 4.02 ng/ml (sensitivity 66.7% and specificity 83.3%). The area under the curve for sTREM-1 to predict the ICU mortality on day 1 was 0.91 (95% confidence interval: 0.81–1.01), with the best cut-off value at 4.02 ng/ml (sensitivity 73.3% and specificity 100%). Conclusion Urinary sTREM-1 can be used as a clinical outcome predictor for the development of AKI and ICU mortality in patients with sepsis.



How to cite this article:
Farghal A, Sherif HM, Al Sisi A, Al-Maraghy S. Urinary soluble triggering receptor expressed on myeloid cells-1 is an early predictor for sepsis outcome and sepsis-induced acute kidney injury.Res Opin Anesth Intensive Care 2018;5:8-14


How to cite this URL:
Farghal A, Sherif HM, Al Sisi A, Al-Maraghy S. Urinary soluble triggering receptor expressed on myeloid cells-1 is an early predictor for sepsis outcome and sepsis-induced acute kidney injury. Res Opin Anesth Intensive Care [serial online] 2018 [cited 2018 Sep 26 ];5:8-14
Available from: http://www.roaic.eg.net/text.asp?2018/5/1/8/223839


Full Text

 Introduction



Sepsis is a main factor for admission to ICU and patient mortality [1]. Because of its rapid progression, the disease might, within a relatively short period, lead to secondary multiple organ dysfunction syndrome (MODS), which could be life threatening [2],[3],[4]. Despite recent advances in the comprehensive management of patients, sepsis is still a life-threatening condition with a poor outcome [1],[4],[5],[6]. Early diagnosis of sepsis plays a significant role; for each hour of delay of appropriate therapy, the mortality increases by 7.6% [5].

However, the accurate and timely detection of sepsis remains a significant challenge because of the various, insidious, and nonspecific clinical manifestations as well as the compound and indeterminate pathophysiologic process [1],[5]. Therefore, neither the clinical nor the most traditional biomarkers can fulfill all the existing needs in the early diagnosis and management of sepsis [5]. Clinical practice has shown that for patients with septic shock, an early and effective intervention can improve the prognosis and reduce the mortality [7].

Acute kidney injury (AKI) is a well-known health problem associated with longer length of stay, morbidity, and mortality in adults [2],[8]. Sepsis-related AKI occurs in about 19% of sepsis patients, which may reach 23% among septic shock patients, showing a mortality rate of 70% [9]. Therefore, it is clinically important to identify indicators that can be used for the early diagnosis and prognosis of sepsis and the AKI induced.

Triggering receptor expressed on myeloid cells-1 (TREM-1) is considered an immunoglobin expressed on the cell membrane of neutrophils, monocytes, and macrophages. TREM-1 belongs to a family related to the natural killer cell receptors [10]. TREM-1 upregulates the expression levels of proinflammatory chemokines and cytokines and amplifies the inflammatory responses mediated by Toll-like receptors. Urinary soluble triggering receptor that is expressed on myeloid cells-1 (sTREM-1) is a form of TREM-1 that may be released into urine upon the upregulated expression of TREM-1; this soluble form can be measured in all biological fluids and may be used as a diagnostic biomarker for evaluating the severity and prognosis of sepsis [7],[10].

In clinical studies involving patients with severe sepsis, sTREM-1 has shown the potential to provide an excellent predictive value for septic shock/death. For instance, the sTREM-1 level was found to be associated significantly with AKI, which may be used as a diagnostic and prognostic biomarker for AKI in critically ill patients with sepsis [11].

The main aim of this study is to evaluate the prognostic role of sTREM-1 as an early predictor for clinical outcome in critically ill patients with sepsis and in the detection of sepsis-related acute kidney injury. Also, our aim is to compare sTREM-1 with one of the traditional common biomarkers in the setting of sepsis [C-reactive protein (CRP)].

 Patients and methods



This prospective cohort study included 30 patients and a group of 10 controls who were all admitted from June 2013 to May 2014 to the Critical Care Department, Cairo University Hospitals. This study was approved by the Ethical Committee Review Board of the Faculty of Medicine, Cairo University. Informed consent was obtained from all patients and controls before enrollment in the study.

Inclusion criteria

All patients included in this study (age ≥18 years with a likely source of infection) fulfilled the criteria of sepsis and severe sepsis/septic shock according to the American College of Chest Physician/Society of Critical Care Medicine Consensus Conference [12] patients with two or more of the following:Temperature more than 38°C (100.4°F) or less than 36°C (96.8°F).Heart rate more than 90/min.Respiratory rate more than 20/min or PaCO2 less than 32 mmHg, white blood cell count more than 12 000/mm3 or less than 4000/mm3 or more than 10% immature neutrophils.

Patients were considered to have sepsis when the systemic inflammatory response syndrome (SIRS) was the result of infection. Severe sepsis was taken into consideration when dysfunction accompanied sepsis in one or more vital organs. Septic shock was taken into account when refractory hypotension accompanied severe sepsis to volume infusion.

The patients were divided as follows:Patients with sepsis.Patients with severe sepsis.Patients with septic shock.To determine the standard reference range of the urinary sTREM-1 concentration, 10 healthy volunteers were included in this study and served as controls.

Exclusion criteria

Any patient who had a history of chronic renal failure or was subjected to renal transplantation was excluded from the study.

Study design

All patients were subjected to daily collection of baseline clinical and routine investigational data, including blood urea and serum creatinine, arterial blood gasses, complete blood count, and urine analysis.sTREM-1 estimation in urine was performed on days 1, 3, and 7 during ICU stay [11],[13],[14].Using the double-antibody sandwich ELISA technique, the urine samples were collected within 24 h after ICU admission (day 1) and recollected on days 3 and 7. The urinary sampling was prepared for EDTA tubes. Urinary sTREM-1 serum concentration was determined by ELISA (Kit Quantikine, Human TREM-1 Immunoassay; R&D Systems, Minneapolis, MN, USA). The microtiter plate was precoated with an antibody specific to sTREM-1 samples that were added to the appropriate microtiter plate wells with a biotin substrate reaction; the reactions were terminated by the addition of a sulfuric acid solution and the color change was measured spectrophotometrically.All the procedures were performed by laboratory technicians who were blinded to the clinical data.CRP estimation in blood was performed on days 1, 3, and 7 during ICU stay.Using the ELISA technique, the blood samples were collected within 24 h after ICU admission and repeated on the morning of days 3 and 7, respectively.All the procedures were performed by laboratory technicians who were blinded to the clinical data.Sepsis-related organ failure assessment scoring system (SOFA-score) [15].The SOFA-score was applied daily for the entire ICU stay. The SOFA-score included the following parameters: respiratory (PaO2/FiO2), coagulation (platelets), liver function tests (total serum bilirubin), cardiovascular (hypotension), central nervous system (Glasgow coma scale), and renal function (serum creatinine/urine output). The lowest score is 6 and the highest score is 24.Acute kidney injury was evaluated according to the RIFLE-criteria (Risk, Injury, Failure, Loss and End-stage Kidney) in oliguric patients ([Table 1]).{Table 1}A standard case report form was used for every patient to record daily progression by reviewing medical records until 28 days during ICU stay or death.All patients were followed up in terms of the following:Length of ICU stay.The need for and duration of mechanical ventilation (MV).The need for and duration of vasopressors (VP).Need for renal replacement therapy (RRT).Outcome and mortality.

Statistical methods

Statistical package for the social sciences (SPSS), version 2.0 (IBM, Armonk, New York, United States) was used for our statistical analysis. Data were summarized as mean±SD, minimum, and maximum values. Comparisons between groups were made using a nonparametrical Kruskal–Wallis test and the Mann–Whitney test. For comparison of serial measurements, the nonparametrical Friedman test and the Wilcoxon test were used. For comparison of categorical data, the χ2-test was performed. We used the analysis of variance test to analyze the differences among the means of multiple groups. Correlations between variables were determined using Spearman’s correlation coefficient. Receiver operator characteristic (ROC) curves were derived and an area under the curve (AUC) analysis was carried out to determine the best cut-off value of urinary sTREM-1 to detect patients with AKI. P values less than 0.05 were considered statistically significant.

 Results



This study was carried out on 30 (57.57±7.51, 19–84 years; 18 men) critically ill sepsis patients. Ten controls were included (45.6±3.5, 25–72 years; six men).

Etiology for ICU admission

Twenty-one cases were admitted to ICU because of medical problems, whereas nine patients were admitted postoperatively.

The three main clinical problems for ICU admission included chest infection (50%), intra-abdominal sepsis (7%), and urinary tract infection (6.7%).

The patient population included seven patients with sepsis, 10 patients with severe sepsis, and 13 patients with septic shock.

Clinical course and final outcome data are shown in [Table 2].{Table 2}

Urinary soluble triggering receptor expressed on myeloid cells-1 data

Compared with the controls, the patient data showed statistically significant increased sTREM-1 values on day 1 (3.78±1.21 vs. 0.78±0.16 ng/ml, P<0.001). However, our results showed comparable sTREM-1 values in patients admitted with medical problems (21 patients) versus patients admitted postoperatively (nine patients) (3.96±1.26 vs. 3.36±0.99 ng/ml).

Comparable sTREM-1 values on day 1 could be detected among patients with sepsis versus patients with severe sepsis, but on comparing either patients with sepsis versus patients with septic shock or patients with severe sepsis versus patients with septic shock, statistically significant increased sTREM values could be detected ([Figure 1]).{Figure 1}

Analysis of differences between groups indicated statistically significantly increased sTREM-1 values on day 1 upon shift from sepsis, severe sepsis, to septic shock groups (P<0.05) ([Figure 1]).

Analysis of differences in all patients indicated statistically significant increased sTREM-1 values from days 1, 3 to day 7 (3.78±1.21, 4.50±1.79, 5.89±2.99 ng/ml, P<0.05), respectively.

Need for mechanical ventilation and vasopressors

Our data showed comparable sTREM-1 values on day 1 between patients who needed MV (24 patients) and patients who did not (six patients) (3.63±1.05 vs. 4.37±1.68 ng/ml).

However, compared with patients (7) who did not need VP, patients (23) who needed VP showed statistically significantly increased sTREM-1 values on day 1 (4.06±1.22 vs. 2.86±0.51, P<0.05).

Need for renal replacement therapy

Compared with patients (26) who did not need RRT, patients (4) who needed RRT showed statistically significantly increased sTREM-1 values on day 1 (5.27±1.2 vs. 3.55±1.05, P<0.05).

Laboratory data

A good correlation could be detected between sTREM-1 and CRP values only on day 7, but no correlations could be detected on day 1 or 3 ([Table 3]).{Table 3}

No correlations could be detected between sTREM-1 values and the total leukocytic count on day 1, 3, or 7 ([Table 4]).{Table 4}

Sepsis-related organ failure assessment clinical score

Good correlations could be detected between sTREM-1 values and the SOFA-score on days 1 and 7, but no correlation could be detected on day 3 ([Table 5]).{Table 5}

Urinary soluble triggering receptor expressed on myeloid cells-1 as an outcome predictor

Compared with sTREM-1 values in patients who did not have AKI (18 patients), sTREM-1 values in patients who had AKI (12 patients) showed statistically significantly higher values on day 1 (4.37±1.34 vs. 3.39±0.95 ng/ml, P<0.05).

For ICU length of stay (16.83±7.33 days), our data showed no correlation with the values of sTREM-1 on day 1 (r=−0.13, P=NS), but compared with patients with ICU stay less than 28 days (six patients 20%), patients with ICU stay more than 28 days (nine patients 30%) showed statistically significantly higher sTREM-1 values on day 1 (3.31±0.49 vs. 2.73±0.42 ng/ml, P<0.05).

The mortality rate in this study was 50% (15 patients); compared with survivors, the nonsurvivors showed statistically significantly higher sTREM-1 values on day 1 (4.6±1.14 vs. 2.96±0.52 ng/ml, P<0.001).

ROC curve was calculated for sTREM-1 values to predict AKI on day 1. The AUC was 0.73, P value less than 0.05 [95% confidence interval (CI): 0.53–0.92], with the best cut-off value for sTREM set at 4.02 ng/ml (sensitivity 66.7% and specificity 83.3%) ([Figure 2]).{Figure 2}

ROC curve was calculated for sTREM-1 values to predict ICU mortality on day 1. AUC was 0.91, P value less than 0.001 (95% CI: 0.81–1.01), with the best cut-off value for sTREM set at 4.02 ng/ml (sensitivity 73.3% and specificity 100%) ([Figure 3]).{Figure 3}

 Discussion



Accurate prediction of prognosis in ICU patients is important to enable appropriate treatment decisions to be made by medical attendants [16]. Currently available tools for the prediction of prognosis in ICU such as the SOFA-score [3],[15], which predicts morbidity and mortality, are based on several physiological indices and chemical analytics.

Over the years, several problems, pitfalls, and limitations of these scoring systems have been identified. Furthermore, they are very cumbersome and time-consuming as they are based on several biochemical measurements and several physiological indices [17].

Several biomarkers have been used to predict morbidity and mortality in critically ill patients, although none have been proven to be entirely useful [1],[5]. However, recently, investigations have been performed on the potential use of urinary sTREM-1 for the clinical diagnosis and prognosis of sepsis syndrome and associated AKI [9],[11],[14]. Accordingly, the aim of our study was to investigate the prognostic value of urinary sTREM-1 in patients with sepsis.

In our study, the urinary sTREM-1 showed statistically significantly higher values in all sepsis patients compared with the healthy controls. The same results were obtained by a study carried out by Palmiere et al. [18], but the study evaluated serum TREM-1 postmortem.

To our knowledge, no investigation similar to ours has been carried out for comparison between sepsis patients and controls in terms of urinary sTREM-1.

In the present study, the elevated levels of urinary sTREM-1 on day 1 increased further with increasing severity of sepsis. Patients with septic shock had statistically significantly higher urinary sTREM-1 values than either those with severe sepsis or those with sepsis. However, no correlation could be detected between urinary sTREM-1 levels in patients with sepsis and severe sepsis. Su et al. [19], studied 104 ICU patients; 16 patients showed SIRS, 35 patients showed sepsis, and 53 patients showed severe sepsis. They reported that patients with severe sepsis and septic shock had significantly higher levels of urinary sTREM-1 at 6 different time points (on days 1, 3, 5, 7, 10, and 14) than those with sepsis.

Our data showed comparable results between urinary sTREM-1 in patients who needed MV and those who did not on day 1. This was in agreement with Su et al. [19], who reported that there was no correlation between the level of urinary sTREM-1 in patients who needed MV and those who did not.

However, our data showed statistically significantly higher urinary sTREM-1 values in patients who needed VP support/RRT versus those who did not on day 1. However, Dai et al. [20], showed an insignificant difference between sepsis patients who needed RRT and urinary sTREM-1 values. This discrepancy can be attributed to the fact that Dai et al. [20] included only patients with sepsis, but we included patients with sepsis, severe sepsis, and septic shock as well.

In our study, a good correlation could be detected between the values of urinary sTREM-1 and serum CRP only on day 7, but no correlation could be detected on day 1 or 3. Also, no correlations could be detected between sTREM-1 values and the total leukocytic count (TLC) on day 1, 3, or 7. Su et al. [19], showed that, on days 1, 3, 5, 7, 10, and 14, urinary sTREM-1, serum CRP, and TLC levels were higher only in the severe sepsis group compared with the sepsis group (P<0.05). Urinary sTREM-1 and serum CRP levels increased continuously among nonsurvivors, whereas TLC and serum CRP concentrations decreased in both groups. Our explanation for this disagreement is that we correlated sTREM-1 with either CRP or TLC in all patients on days 1, 3, and 7 in all patients as our aim from the beginning was to test one prognostic biomarker for all types of sepsis patients irrespective of the severity.

In our study, good correlation could be detected between sTREM-1 values and the SOFA-score on days 1 and 7, but no correlation could be detected on day 3. Accordingly, urinary sTREM-1 can serves as a surrogate marker for the severity of sepsis. This result was in agreement with Su et al. [19]. They showed that there were statistically significantly higher differences in urinary sTREM-1 levels on day 1 and the SOFA-score among the 104 patients admitted to the ICU studied: 16 patients with SIRS, 35 patients with sepsis, and 53 patients with severe sepsis (P<0.001).

On day 1, our data showed that, compared with patients who did not show AKI, in patients who showed AKI, the urinary values of sTREM-1 showed statistically significantly higher values. This was in agreement with Su et al. [19]; they reported that in 17 patients with AKI, the urinary sTREM-1 on day 2 showed higher values than those of non-AKI patients (P<0.05). In another study carried out by Dai et al. [20] compared with the non-AKI sepsis patients, the sepsis AKI patients showed a markedly higher level of serum TREM-1 24 h before AKI diagnosis (P<0.01) and the score was significantly higher in sepsis patients with AKI than in the non-AKI patients (P<0.05).

In our study, we used the urinary sTREM-1 concentration as a predictor of AKI. The ROC-curve showed that AUC was 0.73 (95% CI: 0.53–0.92). The best cut-off value of urinary sTREM-1 on day 1 for the prediction of AKI was 4.02 g/ml with a sensitivity of 66.7% and a specificity of 83.3%. In the Su et al. [19] study, they reported that the AUC was 0.92 (95% CI: 0.85–1), with a best cut-off point at 6.9 ng/ml and a sensitivity of 94% and a specificity of 76%. This disagreement in the best cut-off values for sTREM could have been because we carried out our calculations on day 1, but they carried out their calculations on day 2. Calculation on day 2 led to a higher chance for more patients who developed AKI to be included, which would increase the cut-off values of the urinary sTREM-1.

Our results showed that the ICU mortality rate was 50% (15 patients). The urinary sTREM-1 concentration in nonsurvivors was significantly higher than that in survivors. Our results were in agreement with Su et al. [19], who reported that the urinary sTREM-1 levels were significantly higher in a nonsurvivor group than in the survivors (P<0.001).

In our study, we used the urinary sTREM-1 concentration on day 1 as a predictor of ICU mortality. The ROC-curve showed that AUC was 0.91; (95% CI: 0.74–1.1). The best cut-off value of urinary sTREM-1 on day 1 for the prediction of AKI was 4.02 g/ml with a sensitivity of 73% and a specificity of 100%.

To our knowledge, no study has shown the best cut-off value of urinary sTREM-1 in the prediction of ICU mortality in sepsis or sepsis-induced AKI.

 Limitations



Besides the limited number of patients in our study, we found limited number of published articles discussing our relevant data, but this also can be considered as a strength, but of course, further investigation would be of added value to elucidate the emerge of urinary sTREM-1 as a predictor of sepsis and sepsis-induced AKI.

 Conclusion



Urinary sTREM-1 may be used as a feasible, easy, and reproducible test to perform and interpret the prediction of clinical outcomes of sepsis patients on admission to the ICU and the early prediction of induced AKI.

This early risk stratification is of particular significance in the intensive care environment as it enables clinicians to make more rational therapeutic decisions to ensure that the hospital resources are used efficiently and appropriately.

In terms of practicability, urinary sTREM-1 meets the demands of a readily available biomarker under clinical routine and emergency conditions.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1Bentzer P, Russell JA, Walley KR. Advances in sepsis research. Clin Chest Med 2015; 36:521–530.
2Post EH, Kellum JA, Bellomo R, Vincent JL. Renal perfusion in sepsis: from macro- to microcirculation. Kidney Int 2017; 91:45–60.
3Vincent JL, Martin-Loeches I, Annane D. What patient data should be collected in this randomized controlled trial in sepsis? Intensive Care Med 2016; 42:2011–2013.
4Sakr Y, Rubatto Birri PN, Kotfis K, Nanchal R, Shah B, Kluge S et al. Higher fluid balance increases the risk of death from sepsis. Crit Care Med 2017; 1:386–394.
5Liu X, Ren H, Peng D. Sepsis biomarkers: an omics perspective. Front Med 2014; 8:58–67.
6Dellinger R, Levy M, Rhodes A. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock, 2012. Intensive Care 2013; 41:580–637.
7Otero RM, Nguyen HB, Huang DT, Gaieski DF, Goyal M, Gunnerson KJ et al. Early goal-directed therapy in severe sepsis and septic shock revisited: concepts, controversies, and contemporary findings. Chest 2006; 130:1579–1595.
8Post EH, Su F, Hosokawa K, Taccone FS, Herpain A, Creteur J et al. Changes in kidney perfusion and renal cortex metabolism in septic shock: an experimental study. J Surg Res 2017; 207:145–154.
9Lameire N, van Biesen W, Vanholder R. Acute renal failure. Lancet 2005; 365:417–430.
10Bouchon A, Facchetti F, Weigand MA, Colonna M. TREM-1 amplifies inflammation and is a crucial mediator of septic shock. Nature 2001; 410:1103–1107.
11Liu XR, Xu J, Wang YM, Ji MS, Liu FS. The effects of paeoniflorin injection on soluble triggering receptor expressed on myeloid-1 (sTREM-1) levels in severe septic rats. Korean J Physiol Pharmacol 2016; 20:565.
12Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA et al. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The ACCP/SCCM Consensus Conference Committee. American College of Chest Physicians/Society of Critical Care Medicine. Chest 1992; 101:1644–1655.
13Su L, Xie L, Liu D. Urine sTREM-1 may be a valuable biomarker in diagnosis and prognosis of sepsis-associated acute kidney injury. Crit Care 2015; 19:281.
14Derive M, Gibot S. Urine sTREM-1 assessment in diagnosing sepsis and sepsis-related acute kidney injury. Crit Care 2011; 15:1013.
15Safari S, Shojaee M, Rahmati F, Barartloo A, Hahshemi B, Forouzanfar MM et al. Accuracy of SOFA score in prediction of 30-day outcome of critically ill patients. Turk J Emerg Med 2016; 16:146–150.
16Wijeratne S, Butt A, Burns S, Sherwood K, Boyd O, Swaminathan R. Cell-free plasma DNA as a prognostic marker in intensive treatment unit patients. Ann N Y Acad Sci 2004; 1022:232–238.
17Wu Y, Wang F, Fan X, Bao R, Bo L, Li J et al. Accuracy of plasma sTREM-1 for sepsis diagnosis in systemic inflammatory patients: a systematic review and meta-analysis. Crit Care 2012; 16:R229.
18Palmiere C, Bardy D, Mangin P, Augsburger M. Value of sTREM-1, procalcitonin and CRP as laboratory parameters for postmortem diagnosis of sepsis. J Infect 2013; 67:545–555.
19Su L, Feng L, Zhang J, Xiao Y, Jia Y, Yan P et al. Diagnostic value of urine sTREM-1 for sepsis and relevant acute kidney injuries: a prospective study. Crit Care 2011; 15:R250.
20Dai X, Zeng Z, Fu C, Zhang S, Cai Y, Chen Z. 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.