Value of serum sTREM-1,D-lactate and I-FABP in diagnosis and prognosis evaluation in liver cirrhotic patients with sepsis


 BackgroundSepsis is can cause serious negative effects among patients with cirrhosis.Finding reliable biomarker in early diagnose and prognosis in septic patients with liver cirrhosis is very important.We assessed value of serumsoluble triggering receptor expressed on myeloid cells-1 ( sTREM-1),D-lactate and Intestinal fatty acid binding protein(I-FABP) in diagnosis and prognosis evaluation in liver cirrhotic patients with sepsis.MethodsA prospective observational study of 158 cirrhosis patients was conducted. A total of 158cirrhosis patients were enrolled, 83 of whom were diagnosed with sepsis and 75 patients without sepsis during hospitalization.We evaluated value of early diagnostic and prognosis of cells-1(sTREM-1),D-lactate and I-FABP in cirrhotic patients with sepsis.Receiver operating characteristic curves (ROCs) were used to assess the ability of tested biomarkers to diagnose and prognosis cirrhotic patients with sepsis.ResultsOur results showed the CRP，PCT，sTREM-1, D-lactate and I-FABP were significantly higher in the sepsis group than in the no-sepsis group.,but only levels of serum sTREM-1 showed increasing trend with the severity of the cirrhotic patients with sepsis. The cutoff value of sTREM-1was 123.62 pg/mL and showed 84.3% sensitivity and 92.0% specificity and high accuracy in the early diagnosis for cirrhotic patients with sepsis.The AUC of sTREM-1 was 0.959(95%CI: 0.934–0. 985) and higher than that of CRP,PCT,D-lactate and I-FABP( P <0.001, P =0.002, P <0.001, P <0.001 respectively).Total 31(37.35%) cirrhotic patients with sepsis cases died during 90-day follow-up.Only sTREM-1 level had significance between survivors and non-survivors(263.99±213.79 vs 672.24±191.44,P=0.031).The AUCs of sTREM-1 for predicting 90-day mortality was 0.904( 95% CI:0.836-.971 ) and higher than that of CRP,PCT,D-lactate (P=0.001,p=0.003, P <0.001 respectively) The AUCs of I-FABP is only 0.482 and had no value of predicting prognosis.ConclusionSerum sTREM-1 is a valuable new biomarker for early diagnosis and prognosis in cirrhosis with sepsis, proving more accurate than CRP,PCT, D - lactate and I - FABP.Serum D-lactate and I-FABP levels can identify early cirrhosis with sepsis, but they is little relationship with the prognosis of patients.


Conclusion
Serum sTREM-1 is a valuable new biomarker for early diagnosis and prognosis in cirrhosis with sepsis, proving more accurate than CRP,PCT, D -lactate and I -FABP.Serum D-lactate and I-FABP levels can identify early cirrhosis with sepsis, but they is little relationship with the prognosis of patients.

Background
Patients with liver cirrhosis have portal hypertension,increased intestinal permeability,the imbalance of intestinal ora and the decreased immunity, which altered defense against bacteri.This reduced bacterial clearance in cirrhotic patients can contribute to a high risk of systemic infection and progress easily to sepsis [1].The latter is a common and serious burden among cirrhotic patients because it can cause liver function to deteriorate further, and has serious negative effects on the disease course [2,3]. Some evidence suggests that systemic in ammation in cirrhotic patients maximizes the risk of complications (e.g. portal hypertensive bleeding, hepatic encephalopathy, acute-on-chronic liver failure), acute renal insults [4],which may easily cause multiple organ dysfunction syndrome (MODS), leading to high mortality.About 30% of cirrhotic patients die within 1 month of infection and another 30% die within 1 year. We have reported that 90-day mortality in cirrhotic patients with systemic in ammatory response syndrome(SIRS)can be up to 38% [5].Early diagnosis,recognition of severity and prognostic evaluation are helpful to improve prognosis and reduce mortality for cirrhotic patients with sepsis.At present, the diagnosis of sepsis mainly depends on nonspeci c physiological conditions and culture based on pathogen detection, which leads to the uncertainty of diagnosis and the delay of treatment.The diagnostic value of traditional detection indexes including Creactive protein (CRP and procalcitonin(PCT) is controversial.Silvestre JP et al [6] found CRP levels were signi cantly decreased in patients with sepsis and fulminant hepatic failure.Therefore, nding reliable biomarker in early diagnose and prognosis in septic patients with liver cirrhosis is very important.
Triggering receptor expressed on myeloid cells-1 (TREM-1) is a member of the immunoglobulin superfamilyexpressedon blood neutrophils andmature monocytes/macrophages [7].TREM-1 is a cell-surface receptor, playing a key-role in the septic process [6,7]. In vitroand invivostudies have shown that TREM-1 expression is upregulated strongly by extracellular bacteria (particularly their cell-wall components) and fungi, although not by mycobacteria [8,9]. sTREM-1 is a soluble form of TREM and is released from activated phagocytes [10].Many studied have indicated that sTREM-1 could be a reliable parameter in differentiating patients with sepsis from those with SIRS [11,12]. Hence, sTREM-1 has attracted attention as a diagnostic/prognostic biomarker for various infectious diseases and sepsis [13,14]. However, value of sTREM-1 for the early diagnosis and prognosis evaluation in liver cirrhotic patients with sepsis has not been done.
In liver cirrhosis, disruption of the intestinal mucosal barrier function and increased mucosal intestinal permeability lead to bacterial translocation and endotoxemia,which increase incidence of systemic in ammatory response syndrome(SIRS) and infection [15,16].Intestinal injury can cause sepsis to develop into septic shock,even multiple organ dysfunction( MDRO) [16]. Intestional tract is often the earliest site of sepsis complicated with organ dysfunction .With the increase of intestinal permeability, bacteria and their metabolites can enter the blood circulation including D -lactate which could evaluate intestinal injury and endoxemia following severe injuries [17,18] .I-FABP is small-cytosolic watersoluble proteins in mature enterocytes.The level of I-FABP can re ected the physiological turnover rate of enterocyte,with elevated levels indicating intestinal epithelial cell damage [19].The level of D-lactate and I-FABP were considered to be a useful marker for the early detection of signi cant intestinal injury [20,21].Nevertheless,little is known about the value of D-lactate and I-FABP in the early diagnosis and prognosis of cirrhosis with sepsis.
In the present study, we aimed to assess (1) performance of serum sTREM-1, D-lactate and I-FABP levels in the early diagnosis and prognosis ofcirrhotic patients with sepsis in comparison with routinely used traditional parameters such as CRP and PCT;(2)whether serum sTREM-1, D-lactate and I-FABP levels are able to identify severity and provide progress information during sepsis episodes in cirrhosis .  [22,23]. The nal diagnosis was based on other clinical data, including diagnostic imaging and microbiological results.

Follow-up
Data of discharged patients were obtained by telephone contact to patients or their family. All patients in the cohort were studied for 90 days or until death.

Laboratory examination
Routine microbiology examination for sepsis patients involved more than one pair of blood cultures.
Analyses of urine, sputum, bronchoalveolar lavage uid, cerebrospinal uid, abscesses and closed wounds were undertaken. Blood was drawn immediately after presentation to the Emergency Department of Zhejiang Provincial People's Hospital and analyzed in the laboratory within 12 h for measurement of WBC counts, neutrophil percentage, C-reactive protein(CRP) , procalcitonin(PCT) , and blood chemistry.A particle-enhanced immunoturbidimetric methodforquantitative determination of whole-range CRP was used employing a kit from Shanghai Upper Biotech Pharma (Shanghai, China). Procalcitonin levels were measured using a Cobas E601 chemiluminescence analyzer (8000 series; Roche, Basel, Switzerland).Additional serum samples were collected, aliquots were prepared, and samples frozen at −80°C until analyses of sTREM-1 . sTREM-1 levels were determined in duplicate using a double-antibody sandwich enzyme-linked immunosorbent assay (ELISA) (Quantikine ® Human TREM-1).D-lactate levels were measured by abcam,ab83429,D-lactate colorimetric Assay Kit.Serum I-FABP concentration was assayed by ELIAS ( Human FABP2 ELISA Kit,abcam).

Statistical analyses
Descriptive data are the mean ± standard deviation or number (percentage). Comparison of continuous variables was carried out using the Student's t-test. Skewness distribution data are the median with range (25-75 interquartile range) and were analyzed using the Mann-Whitney U-test. Categorical variables were analyzed using Fisher's exact test or Pearson's χ 2 -test. Evaluation of the early diagnostic performance and 90-days mortality of CRP,PCT,sTREM-1 , D-lactate and I-FABP was done using receiver operating characteristic curves (ROCs). The latter were compared using a non-parametric method. The cutoff value, which was the maximum area under the ROC curve (AUC), and accuracies were calculated with 95% con dence intervals.) AUCs of the different indices were compared. P < 0.05 was considered signi cant. had signi cantly higher total bilirubin (P < 0.001), prothrombin time (P < 0.001), creatinine (P < 0.001), proportion of Child-Pugh class-C (P < 0.001), Child-Pugh score (P < 0.00), as well as a lower albumin level (P < 0.001), platelet count (P < 0.001), and proportion of Child-Pugh class-A disease (P < 0.001),compared with patients without sepsis group. Cirrhotic patients in sepsis group has more higher 90-day mortality (P < 0.001). (Table 1). Comparison of tested biomarkers between cirrhotic patients with sepsis according to the severity According to the severity of disease, the cirrhotic patients with sepsis were divided into general sepsis group(group 2,n = 33), severe sepsis group (group 2,n = 28), and septic shock group (group 3,n = 22).no-sepsis group (group 1,n = 39).As exhibited in Table 1,only the levels of serum sTREM-1 showed increasing trend with the severity of the cirrhotic patients ( Group 3 vs Group 4,P = 0.017). Changes of serum CRP a,PCT,D-lactate and I-FABP levels had no signi cance with the severity of the cirrhotic patients, even though they was signi cantly higher in the sepsis group than that in non-sepsis group( Table 2). Discriminative power of test biomarkers for the early diagnosis of cirrhotic patients with sepsis upon hospital admission Figure 1 shows the ROC curves for the CRP, PCT,sTREM-1,D-lactateas well as I-FABP, for the early diagnosis ofcirrhotic patients with sepsis upon hospital admission. Values for the AUC, cutoff value, sensitivity, speci city, PPV and NPV are listed in Table 3 and Fig. 1.We compared the pairwise AUCs of test biomarkers.

Discussion
Sepsis is a clinical process in which bacterial infection triggers excessive in ammation,and weakened bers in the immune system.From mild to severe ,sepsis can be divided into three stages:sepsis,severe sepsis and septic shock. [24]Infection and sepsis can occur in patients with liver cirrhosis ofvarious etiology( Table 1).The incidence of sepsis signi cantly promoted the progression of decompensated cirrhosis patients to Acute-onchronic Liver failure(ACLF) [3,25].Organ failure caused by sepsis and sepsis related death are greatly increased in patients with cirrhosis[26]. As we found signi cant differences in levels of ALB, TB,PT,Cr, PLT count,Child-Pugh class C ratio, Child score, and 90-daymortality between patients with and without sepsis Table 1).Cirrhotic patients with sepsiswere in critical conditionand with a poor in-hospital outcome, especially in patientswith Child-Pugh class C and their lower immunity may be responsible for the high incidence of sepsis [ 27,28 ].The 90-day mortality rate was as high as 37.35%( 31/83) in cirrhotic patients with sepsis group. (Table 1). Even if the total number of WBCs and proportion of NEs in sepsis group was higher than that in non-sepsis group,WBC count in two group were within normal range ,which may be related to the presence of hyperslenism and decompensated liver cirrhosis.Partial cirrhotic patients haven't any symptoms at the initial stage of infection.It is not easy to attract the attention of clinicians.Some new biomarkers of sepsis would be particularly useful for the early diagnosis in this population.
Unfortunately, early recognition and prognosis prediction of sepsis in patients with cirrhosis is di cult.
Serum levels of CRP and PCT have obtained special interest in identifying diagnosis and prediction of bacterial infection in liver cirrhosis [29,30]. Procalcitonin(PCT) is a sensitive and speci c serologic marker, and its level increases at the early stage of liver cirrhosiswith serious bacterial infection, and so has value for early diagnosis [31]. CRP is an acute phase I protein synthesized by hepatocytes when the body is stimulated by microbial invasion or tissue damage. PCT is released by the solid tissue cells of liver and peripheral blood mononuclear cellsduring microbial infection.But Liver failure may lead to the formation ofa network of monocytes to reduce theconcentration of acute-phase proteins (APPs) [32]. The more severe the underlying liver dysfunction, the lower the CRP response to bacteremia [33].JJanos Szederjesi et al reported there was no signi cant correlation between the levels of the PCT and CRP biomarkers and severe sepsis .Quantitive procalcitonin and C-reactive protein analysis was not shown to be useful in diagnosing severe sepsis [34].Our data showed that levels of CRP and PCT in liver cirrhosis patients with sepsis did not increase signi cantly ( Table 2), even though they were signi cantly different between the non-sepsis group and sepsis group.
levels of CRP and PCT were no meaningful between the non-survival group and in the survival group.
TREM-1 expression can be induced by bacterial products such as lipopolysaccharides and lipoteichoic acid.TREM-1 is expressed on neutrophils and monocytes. It is implicated in the development and ampli cation of the early in ammatory response to infection and injury. sTREM-1 is the soluble form of this receptor and is released into body uids when TREM-1 expression is upregulated [35]. As reported by Jedynak and colleagues, the serum level of sTREM-1 measured within the rst 24 h of treatment in the intensive care unit is a useful prognostic biomarker for patients with sepsis, severe sepsis or septic shock [36]. Various studies have suggested that the sTREM-1 concentration in different biologic uids is signi cantly higher in patients with bacterial infection than in those with a non-microbial in ammatory process [37][38].sTREM-1 isa reliable biomarker evaluated for the diagnosis and prognosis of sepsis to date [39,40]. Furthermore, the serum level of sTREM-1 re ects the severity of sepsis more accurately than that of CRP ,PCT,is more sensitive for dynamic evaluation of the sepsis prognosis [40,41]. The serum concentration of sTREM-1can early predict the 28-day sepsis mortality [42,43].According to our ndings, sTREM-1 was able to predict the development of cirrhotic patients with sepsis .We revealed that cirrhotic patients with sepsis had a signi cantly increasedserum level of sTREM-1, the highest AUC (0.959), sensitivity of 0.843, and speci city of 0.920 at a cutoff value of 123.620 ng/ml. Serum levels of sTREM-1 could be used to predict the early diagnosis of cirrhotic patients with sepsis compared with the levels of CRP, PCT,D-lactate and I-FABP (Table 3, Fig. 1).The AUC for predicting the 90-day mortality was 0.904 for the sTREM-1 level,and sTREM-1 levelwas superior to PCT ,CRP,D-lactate and I-FABP in predicting90-day mortality in cirrhotic patients with sepsis.
D-lactate is produced by some bacteria including Klabsilla, Escherichia coli,Lactobacilus species and Bacteroides species. Normally,serum level of D-lactate in mammals are quite low.Serum D-lactate concentration would increase critically ill patients with proven intestinal ischemia [44].D-lactate is a sensitive marker for early diagnosis of gut failure and endotoxemia in patients with cirrhosis.During the event that an ischemia/reperfusion insults,the mucosa is injured and intestinal permeability is increased,leading to an e ux of bacteria and the products of metabolism,including D-lactate into the circulation.The bacterial proliferation would be expected to cause an increased bacterial matabolism with increased production of Dlactate.Serum levels of I-FABP can be used to evaluate intestinal injury and is speci cally released by ischemic mucosa.In the past decades studied reported that D-lactate and I-FABP may the best candidate markers for early diagnosis of acute mesenteric ischemia [45]. Our study showed levels of D-lactate and I-FABP were meaningful in the sepsis group than in the non-sepsis group.However there was no signi cant difference between patients with severe sepsis and septic shock,the same result was found in survivors group and non-survivors group.It was obvioused sepsis could further increased mucosal intestinal permeability in patients with cirrhosis,but the level of serum D-lactate and I-FABP was not related to the severity of sepsis and couldn't predict 90-day mortality in cirrhotic patients with sepsis.
Our study showed thatthe serum level of sTREM-1 wasa suitable biomarker for identi cation of cirrhotic patientswith septic. Moreover, the serum level ofsTREM-1 may be an outcome predictor in cirrhotic patientswith sepsis, and could act as an indicator for starting early therapeutic interventions.Recently some data demonstrate that the TREM-1 pathway on Kupffer cells plays an essential role in hepatic in ammation and brogenesis in a mouse model of brosis. TREM-1 controls the mobilization of in ammatory cells in response to injury and consequently enhances liver damage [46]. Maybe we will study the relationship between TREM-1 and severity of liver in ammation in the future.
Our study had two main limitations. First, in this study ,only serum biomarker levels were detected within 12 hours of hospitalization and there is a lack of continuous dynamic change trend.Second,our study cohort was relatively small. Further validation of a larger dataset is required. Third, cirrhotic patients with sepsis' condition was complex,and the intervention measures before hospitalization were different.Some critically ill patients with severe liver function damage may have metabolic disorders and affect the results.

Conclusions
Page 12/19 The CRP,PCT, sTREM-1, D-lactate and I -FABP were helpful for the early diagnosis in liver cirrhotic patients with sepsis.Serum sTREM-1 is a valuable new biomarker for early diagnosis and prognosis in cirrhosis with sepsis, proving more accurate than CRP,PCT, D -lactate and I -FABP.Serum D -lactate and I -FABP levels can identify early cirrhosis withsepsis, but they is little relationship with the prognosis of patients. Availability of data and materials All data generated or analysed during this study are included in this published article.

Con icts of interests
The authors have declared that no competing interests exist.  The ROC curves for the test biomarkers on admission for the identi cation of cirrhotic patients with sepsis.sTREM-1 had the highest AUC and accuracy, followed by PCT .

Figure 2
The 90-day mortality of cirrhotic patients showed increasing trend with the severity of the sepsis