Introduction
Methods
Results
Biomarker [ref] | No. of patients | Sepsis definition | Study population | Reference group | Sensitivity/specificity (%) | AUC |
---|---|---|---|---|---|---|
Interleukin (IL)-27 [10] | 702 | Positive blood cultures | Pediatric ICU patients with infection | Non-infected critical care patients | 84/63 | 0.75 |
Inter-alpha inhibitor proteins [11] | 573 | Positive blood cultures | Neonates with sepsis | Neonates with risk factors for sepsis | 89/99 | 0.9 |
Group II phospholipase A2 [12] | 525 | ACCP 1992 | ED patients with sepsis | ED patients with suspected infection (with and without SIRS) | NR (logistic regression analysis) | NR |
Bactericidal/permeability increasing protein [12] | 525 | ACCP 1992 | ED with sepsis | ED patients with suspected infection (with and without SIRS) | NR (logistic regression analysis) | NR |
CD64 [13] | 468 | International Sepsis Definitions Conference 2001 | Non-selected ICU population with sepsis | ICU patients admitted without sepsis | 89/87 | 0.94 |
Selenoprotein P [14] | 378 | ACCP 1992 | Non-selected population with sepsis or septic shock | Healthy individuals | NR (no test) | NR |
Lipopolysaccharide-binding protein [15] | 327 | ACCP 1992 | Surgical patients without sepsis at admission | Surgical patients with SIRS without sepsis | 60/62 | 0.66 |
Syndecan-1 [16] | 512 | International Sepsis Definitions Conference 2001 | Trauma patients (4 h after admission) without sepsis | Trauma patients without sepsis | NR (logistic regression analysis) | NR |
Presepsin [17] | 440 | Sepsis-3 | ICU patients with sepsis | ICU patients without sepsis | 89/59 | 0.76 |
IL-6 [18] | 306 | SIRS and organ dysfunction, systolic blood pressure < 90 mmHg, or lactate ≥ 4 mmol/L plus infection | ED patients with suspected sepsis | ED patients with SIRS and organ dysfunction, systolic blood pressure < 90 mmHg, or lactate ≥ 4 mmol/L without infection | NR | 0.86 |
Biomarker [ref] | No. of patients | Sepsis definition | Study population | Main finding | Sensitivity/specificity (%) | AUC |
---|---|---|---|---|---|---|
Urokinase plasminogen activator receptor (uPAR) [19] | 1914 | International Sepsis Definitions Conference 2001 | Critically ill patients and patients hospitalized in internal medicine ward | Levels ≥ 12 ng/mL predicted fatal outcome within 30 days | NR/> 70% | 0.708 0.83 (when combined with APACHE II score) |
Plasminogen activator inhibitor (PAI) 1 [20] | 1790 | ACCP 1992 | Septic patients with disseminated intravascular coagulation (DIC) | Levels > 90 ng/mL predict fatal outcome within 30 days | NR (Kaplan-Meier survival functions) | NR |
Interleukin (IL)-12 [21] | 1444 | Proven peritonitis or mediastinitis and systemic inflammation signs | Surgical patients | Pre-surgery IL-12-synthesizing capability was low in patients who had fatal sepsis after operation | NR | 0.72 |
Thrombomodulin [22] | 1103 | ACCP 1992 | Critically ill patients with sepsis | Levels > 14 ng/mL can predict circulatory failure or death—gray zone between 7 and 14 ng/mL | NR (logistic regression analysis) | NR |
Syndecan-1 [22] | 1103 | ACCP 1992 | Critically ill patients with sepsis | Levels > 240 ng/mL can predict circulatory failure or death—gray zone between 70 and 240 ng/mL | NR (logistic regression analysis) | NR |
Fibrinogen [23] | 1103 | ACCP 1992 | Critically ill patients with sepsis | Levels < 200 mg/dL related to increased risk of fatal outcome | NR (logistic regression analysis) | NR |
Antithrombin activity [23] | 1103 | ACCP 1992 | Critically ill patients with sepsis | Decrease in activity > 50% related to increased risk of fatal outcome | NR (logistic regression analysis) | NR |
Brain natriuretic peptide (BNP) [24] | 1000 | International Sepsis Definition Conference 2001 | ED patients | Levels > 113 pg/mL can predict fatal outcome within 28 days | 86/55 | 0.73 |
Angiopoietin-2 [25] | 931 | NR | Critically ill patients with ARDS | Persistently increased levels related to fatal outcome within 90 days | NR (logistic regression analysis) | NR |
Prothrombin time (PT) [26] | 840 | Suspected infection plus ≥ 3 signs of systematic inflammatory response | Critically ill patients with sepsis | Increase in PT time within first 7 days of sepsis was higher in patients who died within 28 days | NR (no test) | NR |
Adrenomedullin [27] | 837 | International Sepsis Definitions Conference 2001 | ED patients sepsis | Levels < 34.4 ng/L predicted fatal outcome within 30 days | 86/61 | 0.77 0.81 (when combined with Mortality in Emergency Department Sepsis (MEDS) score) |
Pro-adrenomedullin [28] | 896 | Clinical suspicion of infection | ED patients with sepsis | Levels ≥ 1.6 nmol/L predicted fatal outcome within 28 days | 75/92 | 0.89 |
Heparin-binding protein [29] | 759 | Suspected infection and at least one clinical sign of systematic inflammatory response | ED patients with sepsis | Levels > 30 ng/mL predicted any organ dysfunction development within 72 h | 78/76 (cross-tabulation analysis) | NR |
D-dimer [30] | 684 | International Sepsis Definitions Conference 2001 | Emergency department patients with sepsis | Higher in non-survivors than survivors within 28 days | NR | 0.68 |
Troponin [31] | 598 | ACCP 1992 | Critically ill patients | Levels > 0.06 ng/mL independent prognostic marker for 28-day mortality | NR (logistic regression analysis) | NR |
YKL-40 [32] | 502 | ACCP 1992 | Critically ill patients | Levels ≤ 505 ng/mL predicted survival in 90 days | 53/76 | 0.64 |
CD64 [13] | 468 | International Sepsis Definition Conference 2001 | Critically ill patients | Sustained elevated levels were related to non-appropriate antibiotic therapy | 93/48 | 0.74 |
Cell-free DNA [33] | 481 | International Sepsis Definitions Conference 2001 | ED patients | Levels > 1.6 μg/mL predicted short-term fatal outcome | 70/76 | 0.77 |
Biomarker | Study group | Reference group | Comment [refs] |
---|---|---|---|
Diagnostic performance similar to or worse than that of PCT and/or CRP | |||
Cell-free DNA (cfDNA) | ICU patients with sepsis | ICU patients with SIRS | |
Copeptin | ED patients with sepsis | ED patients with SIRS | No better than PCT [38] |
ICAM-1 | Patients with necrotic pancreatitis | Patients with sterile necrosis | No better than PCT [39] |
Lipopolysaccharide-binding protein | ED patients with sepsis | ED patients with infection | No better than PCT [40] |
Non-critically ill patients with sepsis | Non-critically ill patients with infection | No better than PCT [41] | |
Children with neutropenia and clinical sepsis and/or bacteremia | Children with febrile neutropenia without infection | No better than PCT [42] | |
Patients with proven bacterial lower respiratory infection | Patients with proven viral lower respiratory infection | No better than CRP [43] | |
Patients treated in internal medicine ward | Healthy control | No better than PCT [27] | |
Pancreatic stone protein | ED patients with sepsis | ED patients without infection | No better than PCT [44] |
sCD22 | Surgical patients with infection after major operation | Surgical patients with SIRS but without infection | Equal value to PCT [45] |
Interleukin (IL)-2 | ICU patients with sepsis | ICU patients with SIRS without infection | No better than CRP [46] |
IL-1β | Neonates with infection and sepsis | Neonates with infection without sepsis | No better than CRP [47] |
RANTES | Neonates with infection | Healthy neonates | No better than CRP [48] |
Neopterin | ICU patients with sepsis | ICU patients without sepsis | |
Macrophage migration inhibitory factor (MIF) | Patients with infection in medical ward or ED | No bacterial infection | No better than PCT [51] |
Adrenomedullin | Neutropenic patients with sepsis | Neutropenic patients with fever and clinically documented infection | No better than PCT [52] |
Pro-adrenomedullin | Sepsis with organ dysfunction and or shock | Patients admitted to coronary unit without infection | No better than PCT [53] |
High-mobility group-box 1 protein (HMGB1) | Infected patients admitted in the ward | Healthy individuals | No better than CRP or PCT [54] |
IL-8 | Neutropenic children with blood culture positive, and/or fever periods with a documented clinical sepsis and/or local infection | Neutropenic children with fever and no infection | No better than CRP [55] |
IL-10 | Patients with bacteremia and SIRS, | Patients with SIRS without bacteremia | Comparable with PCT [56] |
Endocan | Critically ill patients with sepsis and organ dysfunction | Critically ill patients with infection and SIRS | Comparable with PCT [57] |
Pro-atrial natriuretic peptide (ANP) | Burned patients that received antibiotics and had either microbiological confirmation of infection or antibiotics leaded to an improvement in clinical situation | Burned patients without infection | Comparable with PCT [58] |
Pentraxin 3 | Mechanically ventilated patients with ventilator associated pneumonia | Mechanically ventilated patient > 48 h without VAP | No better than CRP [59] |
Hematological patients with bacteremia and/or septic shock | Hematological patients with fever without infection | No better than CRP [60] | |
Better diagnostic value than PCT and/or CRP | |||
Thromboelastometry lysis index | Patients with severe sepsis | Patients after operation without sepsis | Better than PCT [61] |
Decoy receptor 3 | ICU patients with sepsis | ICU patients with SIRS | Positive when PCT was negative [62] |
Group II phospholipase A2 (PLA2-II) | ED patients with sepsis and organ dysfunction | ED patients with SIRS without infection | Better than CRP [63] |
Hepcidin | Infants with sepsis and or bacteremia | Infants with SIRS and not sepsis | Better than CRP [64] |
sCD163 | Patients with sepsis admitted to ICU | Patients with SIRS without sepsis | Better than PCT [65] |
CD64 | ICU patients with sepsis | ICU patients without sepsis | Better than PCT and CRP [66] |
Patients with ventilator associated pneumonia and sepsis | Patients with ventilator associated pneumonia without sepsis | Better than PCT and CRP [67] | |
Serum amyloid A | Full term infants with sepsis | Full term infants with risk for sepsis but without sepsis | Earlier increase in neonates with early onset sepsis than CRP [68] |
Heparin-binding protein | Patients with sepsis for less than 48 h | Patients with infection without sepsis | Better than CRP and PCT [69] |
Delta-like canonical Notch ligand 1 (DLL1) | Patients with abdominal infection or surgical site associated infection | Surgical patients, trauma patients without infection, and healthy volunteers | Better than CRP and PCT [70] |
Conflicting findings | |||
IL-6 | Critically ill patients with sepsis | Patients with SIRS without infection | IL-6 was not found to have lower diagnostic utility compared to PCT (meta-analysis) [71] |
Cirrhotic patients with infection at admission to ICU | Cirrhotic patients without sepsis | IL-6 was found to increase earlier than PCT in cirrhotic patients [72] | |
sCD25 | ED patients with infection | ED patients with suspected infection but finally infection excluded | Equal diagnostic value to PCT for diagnosis of infection in ED [44] |
Patients admitted in ICU with infection and SIRS | Patients with SIRS without sepsis | Better performance than PCT to identify Sepsis I at ICU admission [73] | |
Calprotectin | ICU patients with infection | ICU patients without sepsis | Better than CRP and PCT [74] |
Patients after major operation who developed sepsis | Patients after major operation who did not develop sepsis | Similar value to PCT [75] | |
IL-27 | Critically ill children with sepsis | Children with SIRS without infection | Better than PCT [76] |
ICU patients with sepsis | ICU patients without sepsis | No better than PCT [77] | |
sTREM | ICU patients with sepsis | ICU patients with SIRS | Better than PCT [78] |
ICU patients with sepsis | ICU patients with SIRS | No better than PCT and CRP [79] | |
Presepsin (CD14) | ED patients with sepsis | ED patients with at least two criteria of SIRS without sepsis | Better than PCT in diagnosis of sepsis in ED [80] |
Critically ill patients with sepsis and organ dysfunction | Critically ill patients without infection | No better than PCT regardless of the presence or not of AKI [17] | |
Neonates with SIRS and positive blood cultures | Neonates with SIRS with negative blood cultures | Better than PCT [81] | |
Better performance when combined with PCT and/or CRP | |||
IL-6 | Neonates with infection within the first week of life | Neonates with suspicion of infection but finally excluded within the first week of sepsis | Combination with CRP in neonates with suspected sepsis [82] |
CD64 | Neonates with sepsis | Healthy controls | Combination with PCT and CRP for diagnosis of neonatal sepsis [83] |
Leptin | Patients with community acquired pneumonia with sepsis or complicated intraabdominal infection | SIRS without infection, healthy controls | Combination with CRP [84] |
Pro-adrenomedullin | Septic patients | Patients with SIRS without sepsis | |
suPAR | Septic patients admitted to ICU | Critically ill patients with SIRS without infection and healthy controls | Combination with PCT for diagnosis of sepsis on day 1 of sepsis [86] |
CD11b | Patients with Gram (+) infection | Patients with Gram (−) infection | Combination with CRP for differentiation from Gram (−) infection [87] |
Fibrinogen | Neutropenic patients with sepsis | Neutropenic patients with fever without infection | Combination with CRP for diagnosis of sepsis [88] |
BNP and antithrombin | Neutropenic patients with fever and bacteremia | Neutropenic patients with fever without infection | Combination with PCT for diagnosis of Gram (−) bacteremia [88] |
IL-27 | Pediatric patients with sepsis | Pediatric patients with SIRS without infection | |
α-2 macroglobulin | Surgical patients with sepsis | Surgical patients with SIRS without sepsis | Combination with PCT to exclude sepsis in surgical patients [90] |
Decoy receptor 3 and uPAR | Patients with sepsis | Patients with SIRS without infection, healthy volunteers | Combination with PCT for diagnosis of sepsis [91] |
Situation | Biomarker |
---|---|
To diagnose infection in patients with a particular pathology/condition | |
After cardiac surgery | |
After major surgery | |
COPD | Pentraxin 3 [98] |
Cirrhosis | Interleukin (IL)-6 [72] |
Trauma | |
Catheter-related infections | Citrulline [102] |
Infants with necrotic enterocolitis | IP-10 [103] |
Neutropenic patients | |
Burns | |
Autoimmune diseases | CD64 [108] |
To diagnose specific types of infection | |
Gram (−) vs. Gram (+) | |
Virus vs. bacterial infection or co-infection | |
VAP | suPAR [112] |
Diagnosis of specific conditions | |
Sepsis vs. graft rejection | Lysozyme [113] |
Diagnosis of ARDS | |
Vascular leakage risk in ARDS | |
Recovery from ARDS—endothelial repair | sRAGE [119] |
Identification of low risk of infection in hematological/oncological patients | |
Identification of infection before clinical symptoms | |
Risk of encephalopathy/delirium | |
Disseminated intravascular coagulation |