Severe sepsis is a major cause of death in critically ill patients. However, diagnosis of sepsis can be challenging because several of its symptoms overlap with those observed in other inflammatory states. Early identification of sepsis and appropriate initial treatment have a major effect on the clinical course and outcome of critically ill patients (1). Critical care physicians use a variety of indicators, such as white blood cell (WBC) count, C-reactive protein (CRP), erythrocyte sedimentation rate, and procalcitonin (PCT) to help discriminate between infectious and noninfectious conditions and also to predict clinical outcomes (2).
Triggering receptor expressed on myeloid cells 1 (TREM-1), a member of the immunoglobulin superfamily expressed on blood neutrophils and a subset of monocytes, is upregulated in various inflammatory diseases (3, 4). In a previous study, a soluble form of TREM-1 (sTREM-1) was detected at significant levels in the sera of patients with sepsis, and sTREM-1 level appeared to be the most helpful parameter for differentiating between patients with sepsis and those with systemic inflammatory response syndrome (5).
A trial of early goal-directed therapy (EGDT) involving emergency department (ED) patients with severe sepsis showed that EGDT significantly reduced mortality (6). Early goal-directed therapy is an algorithmic goal-oriented approach for restoring systemic oxygen delivery by manipulating preload (volume), afterload (blood pressure), and contractility (stroke volume), guided by monitoring central venous pressure (CVP), mean arterial pressure (MAP), and central venous oxygen saturation (ScvO2) to preserve effective tissue perfusion within the first 6 h of disease presentation (6). The primary outcome variable, in-hospital mortality, was 46.5% in the control group versus 30.5% in the EGDT group (6). Early goal-directed therapy modulates inflammation and significantly reduces morbidity, mortality, and healthcare resource consumption. The findings of the EGDT trial have been externally validated and have been consistently shown to be generalizable (7).
As far as we are aware, no previous study has shown prognostic factors for patients who achieved the initial goal of EGDT. In addition, it is unknown whether plasma sTREM-1 level may be useful for predicting the survival of patients with severe sepsis who have completed initial resuscitation via EGDT. In the present study, we conducted a prospective study of EGDT using successful treatment guidelines that involve a multidisciplinary team approach, applied the guidelines to patients presenting to the ED with severe sepsis, and measured plasma sTREM-1 level along with PCT and CRP levels. The goal of this analysis was to determine whether sTREM-1 could be used as a reliable marker for predicting prognosis in patients with severe sepsis who had completely achieved the initial resuscitation goal through EGDT.
MATERIALS AND METHODS
Patients and study design
The prospective study in patients treated with EGDT for severe sepsis was carried out at Severance Hospital, a 2,000-bed university-affiliated tertiary care referral hospital in Seoul, Republic of Korea. The study took place from April 2009 to May 2010, and the protocol was approved by the institutional review board of Severance Hospital. All participants provided written informed consent.
Severe sepsis including septic shock was defined as sepsis with organ dysfunction, which may include persistent hypotension, hypoxemia, oliguria, metabolic acidosis, and an altered level of consciousness. Patients with severe sepsis were assessed for inclusion, which required that patients be older than 18 years with confirmed or presumed infection. Patients were eligible if the initial resuscitation according to EGDT protocol was immediately performed at the ED. We excluded the patients with the following conditions: pregnancy, presence of an acute cerebrovascular attack, acute coronary syndrome, acute pulmonary edema, an absolute contraindication to central venous catheterization, active gastrointestinal bleeding, trauma, drug overdose, a requirement for immediate surgery, and do-not-resuscitate status. Analysis was based on 28-day all-cause mortality except cancer-related death.
A multidisciplinary team of ED, intensive care unit (ICU), and infectious disease specialists collaborated to implement the protocol and methods described by Rivers et al. (6). Our protocol was identical to that, except for one modification: the protocol was initiated in the ED and was subsequently transitioned to the ICU. If a patient completed the initial resuscitation goal within 6 h, the patient was enrolled in the study. Using these criteria, a total of 63 patients were included.
The following data at the time of admission to the ED and pre-EGDT were collected: age, sex, comorbid conditions, duration of hospital stay, source of infection, routine blood tests, lactate level, and severity of illness as calculated by the Acute Physiology and Chronic Health Evaluation (APACHE) II score (8), Simplified Acute Physiology Score (SAPS) II (9), and Sepsis-Related Organ Failure Assessment (SOFA) score (10). In addition, the goal-directed resuscitation protocol required monitoring of CVP, MAP, and ScvO2.
Resuscitation goals and definitions
The patients enrolled in EGDT received the insertion of a central venous catheter capable of measuring ScvO2 (PreSep Oximetry Catheter; Edwards Lifesciences, Irvine, Calif), and the catheter was connected to a computerized spectrophotometer for continuous monitoring. Patients were treated according to a protocol for EGDT for at least 6 h, after which monitoring of ScvO2 was discontinued, if the initial resuscitation goal was achieved. The accomplishment of the initial resuscitation goal according to the EGDT protocol was defined as a CVP of 8 to 12 mmHg, a MAP of 65 to 90 mmHg, and an ScvO2 greater than 70% in the initial 6 h.
Measurements of sTREM-1, PCT, and CRP
Day 0 was defined as the day of admission to the ED and the time of pre-EGDT. Pre-EGDT means the time between the presentation to the ED and the performance of EGDT protocol. After separation, plasma was aliquoted and immediately frozen at −70°C until the measurement. Plasma sTREM-1 level was measured with a commercialized specific enzyme-linked immunosorbent assay kit (R&D Systems, Minneapolis, Minn) according to the manufacturer’s protocol. Intra- and interassay coefficients of variation were 3.6% and 7.1%, respectively. The PCT level was measured using a monoclonal immunoluminometric assay (LUMI test PCT; BRAHMS Diagnostica, Berlin, Germany). The functional assay sensitivity for PCT with a 20% interassay coefficient of variation was 0.05 ng/mL. C-reactive protein concentration was assessed in serum using a nephelometric method (Beckman Coulter, Fullerton, Calif). Repeated determinations of plasma sTREM-1, PCT, and CRP concentrations were performed on days 0, 3, 7, and 14.
All variables are expressed as mean (SD), unless otherwise indicated. Categorical variables were compared by χ2 analysis, and continuous variables with normal distributions were compared using the Student t test. We used the Mann-Whitney U test to compare the continuous variables in skewed distribution. The time courses of plasma sTREM-1, PCT, and CRP levels were assessed by analysis of variance. The values of sTREM-1 were skewed without normal distribution through the one-sample Kolmogorov-Smirnov test. Therefore, we performed the natural logarithm transformation of the sTREM-1 variable for normal distribution. Variables with P < 0.05 in bivariate analysis were included in the logistic regression model for multivariate analysis to estimate the odds ratio (OR) of dying, along with the 95% confidence interval (CI). The predictive power of variables to discriminate patients who died from those who did not was assessed by calculating the area under receiver operating characteristic curve (AUC) (11). The AUC ranged from 0.5 to 1.0, and the greater the AUC, the better the variables. These analyses were done in using SPSS version 12.0 software (SPSS Inc, Chicago, Ill).
A multiple logistic regression analysis using the stepwise method with mortality was performed, including the following parameters: SAPS II, SOFA score, log(sTREM-1), ScvO2, and lactate. This statistical analysis was performed using SAS 9.2 (SAS Institute Inc, Cary, NC).
Statistical significance was set at the level of P < 0.05. To account for the inflation in type I error due to multiple testing, all results were corrected with Holm correction (12). If the significance of a result changed after Holm correction, the change was stated after the result.
A total of 63 patients (33 men) diagnosed with severe sepsis and septic shock were enrolled in the study. The mean age of patients was 63.7 (SD, 12.8) years, and the all-cause 28-day mortality rate except cancer-related death was 25.4% (16/63). Twenty-four patients (38.1%) had a documented bloodstream infection. Age, sex, and underlying diseases did not differ significantly between survivors (n = 47) and nonsurvivors (n = 16). In addition, two groups did not have the statistically significant difference for the occurrence of acute renal failure and the presence of bacteremia (Table 1).
Regarding the severity of illness at the time of admission to the ED and pre-EGDT, the SAPS II (59.25 [SD, 14.15] vs. 43.51 [SD, 14.33], P = 0.001) and SOFA score (9.50 [SD, 3.16] vs. 6.13 [SD, 2.72], P = 0.001) were significantly higher in nonsurvivors than in survivors. Also, nonsurviving patients had significantly higher plasma sTREM-1 levels (514.1 pg/mL [interquartile range, 412.7–1,749.5 pg/mL] vs. 182.4 pg/mL [interquartile range, 54.3–327.0 pg/mL], P = 0.001) and arterial lactate levels (4.7 mmol/L [interquartile range, 0.6–17.7 mmol/L] vs. 1.8 mmol/L [interquartile range, 0.5–11.2 mmol/L], P = 0.016) at the time of admission and pre-EGDT than in surviving patients. However, the plasma PCT level did not show statistical significance between nonsurvivors and survivors (16.7 ng/mL [interquartile range, 0.1–200.0 ng/mL] vs. 5.7 ng/mL [interquartile range, 0.1–200.0 ng/mL], P = 0.439). The CRP levels and WBC count did not differ between nonsurvivors and survivors, whereas the ScvO2 values at the time of admission and pre-EGDT were significantly different between nonsurviving and surviving patients (70% [interquartile range, 53%–92%] vs. 80% [interquartile range, 40%–95%], P = 0.042) (Table 2).
Infections of gastrointestinal (14 [29.8%] in survivors and 7 [43.8%] in nonsurvivors), genitourinary (14 [29.8%] in survivors and 1 [6.3%] in nonsurvivors), and pulmonary (10 [21.3%] in survivors and 5 [31.3%] in nonsurvivors) origin were considered the major cause of severe sepsis in total study participants (Table 3).
Patients with a poor prognosis tended to have higher plasma PCT and sTREM-1 levels at the time of ED admission and pre-EGDT. The difference in PCT levels did not reach statistical significance, whereas sTREM-1 levels of nonsurvivors remained, until death, significantly higher than those of survivors (P* = 0.004 on day 0 and P* = 0.012 on day 3 in sTREM-1) (Fig. 1).
On multivariate analysis, the log(sTREM-1), ScvO2, and SAPS II at the time of admission to the ED and pre-EGDT were independently predictive of 28-day all-cause mortality in patients with severe sepsis receiving EGDT (OR, 2.784; 95% CI, 1.116–6.945 [P = 0.028]; OR, 0.931; 95% CI, 0.875–0.990 [P = 0.022]; OR, 1.055; 95% CI, 1.001–1.112 [P = 0.048], respectively) (Table 4).
We found that log(sTREM-1) level was more accurately correlated with mortality than was CRP or PCT in patients receiving EGDT for severe sepsis and achieved an initial resuscitation goal within 6 h. Triggering receptor expressed on myeloid cells 1, recently identified to be involved in inflammatory responses (3), acts in synergy with Toll-like receptor signaling pathways to amplify the inflammatory response mediated by several microbial components (13). However, TREM-1 is not upregulated in noninfectious inflammatory disorders. In addition to its membranous form, a soluble counterpart of TREM-1 (sTREM-1) exists and is released during various infectious processes.
A few studies have evaluated the clinical value of sTREM-1 in sepsis. Gibot et al. (14) described the changes in systemic sTREM-1, PCT, and CRP levels during sepsis. In their prospective study comprising 63 adults with sepsis, severe sepsis, or septic shock in a medical ICU, this group found that sTREM-1 level was significantly lower at admission in nonsurvivors than in survivors. Whereas systemic PCT and CRP decreased during the 2-week study in both survivors and nonsurvivors, sTREM-1 level remained stable or even increased in nonsurviving patients and decreased in survivors.
In contrast, we found that nonsurviving patients had higher sTREM-1 levels at the time of ED admission and continued to show these elevated sTREM-1 concentrations. Recently, several studies have concurred with our results showing significantly higher levels of sTREM-1 in nonsurvivors at an early stage of sepsis (15, 16). Thus, sTREM-1 measurements may be useful for early diagnosis of severe sepsis and timely intervention.
Sepsis, a syndrome characterized by a dysregulated and destructive response to infection, is a complex disease. Populations of patients with severe sepsis are heterogeneous with respect to the cause of the comorbid conditions. Therefore, a variety of strategies have emerged to manage this state (17). Rivers et al. (6) demonstrated that EGDT, consisting of fluid resuscitation, vasoactive interventions, and transfusions based on hemodynamics and other physiological parameters, results in an absolute mortality reduction of 16%. Since the initial publication of the Surviving Sepsis Campaign Guidelines, centers have reported reductions in morbidity and mortality from severe sepsis when using a standardized approach to care (18). Recently, Rivers et al. (7) reported a significant reduction in mortality rate (mean relative and absolute risk reductions of 0.47% [SD, 0.19%] and 21% [SD, 9.9%], respectively). In the present study, the mortality rate was 25.4% for patients with severe sepsis after EGDT, an improvement over that of historical control subjects (7). Considering the complexity of sepsis and the heterogeneity of its management, EGDT could be standardized as a good clinical practice for sepsis treatment (19). Therefore, we believe that clinical data including the levels of plasma sTREM-1 and other inflammatory markers in patients receiving EGDT are valuable and reliable.
In our study, initial ScvO2 was an independent predictor of mortality in patients with severe sepsis. A low ScvO2 may indicate a decrease in oxygen delivery, an increase in oxygen extraction, or a combination of the two. In the EGDT protocol, a low ScvO2 is addressed through oxygen delivery optimization by improving arterial oxygen saturation, cardiac output, or oxygen-carrying capacity to increase oxygen delivery. Accordingly, one may broadly categorize deficiencies in oxygen exchange into three types of failure: macrocirculatory, microcirculatory, and mitochondrial. Macrocirculatory failure is typically assessed through parameters such as CVP, MAP, cardiac index, and ScvO2 (6, 20). The initial EGDT study targeting an ScvO2 level greater than 70% suggested a mortality benefit to normalizing ScvO2 in septic patients (6). Moreover, another study found that nonsurvivors of sepsis had more episodes of mixed venous oxygen saturation desaturation (<65%) than survivors, suggesting a mismatch of oxygen supply and demand (21). One study also found mixed venous oxygen saturation to be an independent predictor of mortality, but only after 48 h in the ICU (22).
On the other hand, another study found that patients with initial ScvO2 values demonstrating hyperoxia (ScvO2 90%–100%) as well as hypoxia were associated with a worse in-hospital mortality than those with values demonstrating normoxia, and they suggested that, whereas low ScvO2 value may be a marker for macrocirculatory failure, high ScvO2 value may reflect microcirculatory or mitochondrial failure (23). However, we found that the target goal of EGDT including ScvO2 (>70%) was accomplished in the initial 6 h in all eligible patients, except for the 6 patients who were excluded from the study. We found no statistical difference in hyperoxia between survivors and nonsurvivors, although the power of our study to detect such a difference was low because of the small number of patients with the condition (data not shown). Therefore, we believe that future research on methods for normalizing high ScvO2 values with therapies that improve microcirculatory flow or mitochondrial dysfunction may be warranted to confirm the significance of the predictive factors for mortality in patients with severe sepsis.
There are several limitations to the current study. First, the study is limited by its small sample size. However, this prospective study has been carefully conducted and includes a significant number of subjects, given the rarity of the disease when considering the strict definition of severe sepsis. In addition, we found that the predictive power for the key variables was high enough in the final multivariate logistic regression model. Further larger-scale studies of sTREM-1 in a diverse clinical spectrum of patients with sepsis should be carried out. Second, because we did not use control groups receiving standard therapy in the ED, we are not able to draw solid conclusions regarding mortality reduction in EGDT. Finally, the present study was performed on patients with severe sepsis in the ED. Therefore, our results might not be generalizable to patients with mild infections.
In conclusion, in the present study, the most powerful predictor of mortality was the plasma sTREM-1 level at the time of admission in patients with severe sepsis despite complete initial resuscitation. In addition, by measuring serial sTREM-1 level according to clinical course, we were able to confirm that the change in clinical condition reflected the sTREM-1 level. Therefore, plasma sTREM-1 level may be a useful biomarker for identifying severely septic patients with a relatively poor prognosis.
The authors thank Hye Sun Lee, a biostatistician (Department of Biostatistics, Yonsei University College of Medicine), for her support in the execution of this study.
1. Harbarth S, Garbino J, Pugin J, Romand JA, Lew D, Pittet D: Inappropriate initial antimicrobial therapy and its effect on survival in a clinical trial of immunomodulating therapy for severe sepsis
. Am J Med 115: 529–535, 2003.
2. Bates DW, Cook EF, Goldman L, Lee TH: Predicting bacteremia in hospitalized patients. A prospectively validated model. Ann Intern Med 113: 495–500, 1990.
3. Bouchon A, Dietrich J, Colonna M: Cutting edge: inflammatory responses can be triggered by TREM-1, a novel receptor expressed on neutrophils and monocytes. J Immunol 164: 4991–4995, 2000.
4. Bouchon A, Facchetti F, Weigand MA, Colonna M: TREM-1 amplifies inflammation and is a crucial mediator of septic shock. Nature 410: 1103–1107, 2001.
5. Gibot S, Kolopp-Sarda M, Bn MC, Cravoisy A, Levy B, Faure GC, Bollaert P: Plasma level of a triggering receptor expressed on myeloid cells-1: its diagnostic accuracy in patients with suspected sepsis. Ann Intern Med 141: 9–15, 2004.
6. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich M: Early goal-directed therapy in the treatment of severe sepsis
and septic shock. N Engl J Med 345: 1368–1377, 2001.
7. Rivers EP, Coba V, Whitmill M: Early goal-directed therapy in severe sepsis
and septic shock: a contemporary review of the literature. Curr Opin Anaesthesiol 21: 128–140, 2008.
8. Knaus WA, Draper EA, Wagner DP, Zimmerman JE: APACHE II: a severity of disease classification system. Crit Care Med 13: 818–829, 1985.
9. Le Gall JR, Lemeshow S, Saulnier F: A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA (Chicago, Ill) 270: 2957–2963, 1993.
10. Vincent JL, Moreno R, Takala J, Willatts S, De-Mendona A, Bruining H, Reinhart CK, Suter PM, Thijs LG: 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 22: 707–710, 1996.
11. Akobeng AK: Understanding diagnostic tests 3: receiver operating characteristic curves. Acta Paediatr 96: 644–647, 2007.
12. Holm S: A simple sequentially rejective multiple test procedure. Scand J Stat 6: 65–70, 1979.
13. Tessarz AS, Cerwenka A: The TREM-1/DAP12 pathway. Immunol Lett 116: 111–116, 2008.
14. Gibot S, Cravoisy A, Kolopp-Sarda M, Béné M, Faure G, Bollaert P, Levy B: Time-course of sTREM (soluble triggering receptor expressed on myeloid cells)-1, procalcitonin
, and C-reactive protein plasma concentrations during sepsis. Crit Care Med 33: 792–796, 2005.
15. Zhang J, She D, Feng D, Jia Y, Xie L: Dynamic changes of serum soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) reflect sepsis severity and can predict prognosis
: a prospective study. BMC Infect Dis 11: 53, 2011.
16. Latour-Perez J, Alcala-Lopez A, Garcia-Garcia MA, Sanchez-Hernandez JF, Abad Terrado C, Viedma Contreras JA, Masi-Canuto M, Broch Porcar MJ, Arizo-Len D, Gonzalez-Tejera M, et al.: Prognostic value of the sTREM-1 plasma values in patients with sepsis: a cohort study [in Spanish]. Med Intensiva 34: 231–236, 2010.
17. Azevedo LC, Park M, Schettino GP: Novel potential therapies for septic shock. Shock 30 (Suppl 1): 60–66, 2008.
18. Dellinger RP, Levy M, Carlet JM, Bion J, Parker M, Jaeschke R, Reinhart K, Angus DC, Brun Buisson C, Beale R, et al.: Surviving Sepsis Campaign: international guidelines for management of severe sepsis
and septic shock: 2008. Crit Care Med 36: 296–327, 2008.
19. Bozza FA, Carnevale R, Japiassu AM, Castro-Faria-Neto HC, Angus DC, Salluh JI: Early fluid resuscitation in sepsis: evidence and perspectives. Shock 34 (Suppl 1): 40–43, 2010.
20. da Silva Ramos FJ, Azevedo LC: Hemodynamic and perfusion end points for volemic resuscitation in sepsis. Shock 34 (Suppl 1): 34–39, 2010.
21. Krafft P, Steltzer H, Hiesmayr M, Klimscha W, Hammerle AF: Mixed venous oxygen saturation in critically ill septic shock patients. The role of defined events. Chest 103: 900–906, 1993.
22. Varpula M, Tallgren M, Saukkonen K, Voipio Pulkki L, Pettil V: Hemodynamic variables related to outcome in septic shock. Intensive Care Med 31: 1066–1071, 2005.
23. Pope JV, Jones AE, Gaieski DF, Arnold RC, Trzeciak S, Shapiro NI: Multicenter study of central venous oxygen saturation
(ScvO(2)) as a predictor of mortality in patients with sepsis. Ann Emerg Med 55: 40–46.e41, 2010.