Introduction
Materials and methods
Data sources
Study selection methods
Study data extraction and analysis
Results
The Angiopoietin system
Study | Year | N | Population | Standard Criteria for SIRS/Sepsis | Association with Sepsis | Other Outcomes |
---|---|---|---|---|---|---|
Parikh et al., [43] | 2006 | 51 | ICU patients with sepsis (22) and hospitalized controls (29) | 2003 ACCP/SCCM [2] | Ang-2 higher in patients severe sepsis than patients with sepsis and controls (23.2 vs. 4.8 and 3.5 ng/mL respectively; P < 0.01) | |
Van der Heijden et al., [45] | 2008 | 112 | Mechanically ventilated patients, with sepsis (24) and without (88) | 1992 ACCP/SCCM [1] | Ang-2 higher in patients with sepsis than without sepsis (4.1 vs. 0.4 ng/mL; P < 0.01) | Higher Ang-2 associated with ALI/ARDS (P < 0.001) and higher in ARDS than in ALI (P > 0.001); Independently associated with the severity of pulmonary leak (r = 0.41; P = 0.014). |
Orfanos et al., [38] | 2007 | 61 | ICU patients | 1992 ACCP/SCCM [1] | Ang-2 higher in severe sepsis compared to patients without SIRS or sepsis (P < 0.05 by analysis of variance) | Ang-2 levels correlated with levels of circulating TNF (P < 0.05) |
Giamarellos-Bourboulis et al., [40] | 2008 | 60 | Trauma patients admitted to ICU (54) and healthy controls (6) | 2003 ACCP/SCCM [2] | Ang-2 higher in sepsis and severe sepsis than in healthy controls, or trauma patients with sterile SIRS (P < 0.05); Predictive of sepsis/severe sepsis (P = 0.017, 0.002 respectively); Increases in serial Ang-2 predicted development of sepsis (P < 0.05) | Ang-2 correlated with 28-day survival (P = 0.015) |
Kumpers et al., [42] | 2008 | 72 | Patients admitted to medical ICU (43) and healthy controls (29) | 2003 ACCP/SCCM [2] | Ang-2 higher in septic patients than in patients without sepsis (16.5 vs. 2.8 ng/mL; P < 0.001); Not correlated with severity of sepsis (median Ang-2 16.5 vs. 28.1 ng/mL; P = 0.12); | Ang-2 correlated with mortality (P = 0.001) |
Davis et al., [44] | 2010 | 124 | Patients admitted to a mixed ICU | 1992 ACCP/SCCM [1] | Ang-2 higher in patients with severe sepsis compared to patients with sepsis without organ failure and non-septic controls (12.4 vs. 6.1 and 2.7 ng/mL, respectively; P < 0.0001). | Ang-2 not predictive of 28-day mortality (P = 0.32) |
Siner et al., [39] | 2009 | 66 | Patients admitted to ICU | 1992 ACCP/SCCM [1] | Ang-2 not correlated with severity of sepsis | Ang-2 correlated with mortality (P = 0.02) |
Ricciuto et al. [48] | 2011 | 70 | Patients with severe sepsis | 1992 ACCP/SCCM [1] | Admission levels of Ang-2 and Ang-2/Ang-1 ratio were not associated with 28-day mortality Serially measured Ang-2 levels correlated directly with the MOD score (P = .003) | |
Ebihara et al. [49] | 2011 | 25 | 25 patients treated with Polymyxin-B column hemoperfusion 11 developed ALI | 1992 ACCP/SCCM [1] | Positive correlation between Ang-1 and PaO2/FiO2 ratio (r = 0.427; P < 0.001) Inverse correlation between Ang-2 and PaO2/FiO2 ratio (r = 0.302; P = 0.003) | |
Page et al., [50] | 2011 | 37 | 16 invasive streptococcal infection and toxic shock 21 invasive steptococcal infection alone | S. pyogenes isolated from normally sterile site and 2009 Consensus definition of streptococcal toxic shock | Ang-2:Ang-1 ratio increased in Streptococcal Toxic Shock Syndrome compared to those with uncomplicated invasive streptococcal infection (P < 0.05) | |
Kranidoti et al., [41] | 2009 | 107 | ICU patients with Ventilator Associated pneumonia (90) and healthy controls (17) | 2003 ACCP/SCCM [1] | Ang-2 higher in septic patients compared to healthy controls. (P < 0.001) | Ang-2 correlated with mortality (P < 0.05); Ang-2 levels greater than 9.7 ng/mL associated with sepsis-related mortality (OR = 3.3; P = 0.033) |
Association with sepsis
Association with clinical outcome
The leukocyte adhesion pathway
Study | Year | N | Population | Standard Criteria for SIRS/Sepsis | Associations with sepsis | Other outcomes |
---|---|---|---|---|---|---|
Shapiro et al., [51] | 2010 | 221 | ED patients with sepsis without organ dysfunction (71), severe sepsis without shock (66), septic shock (71), and non-infected controls (13) | 1992 ACCP/SCCM [1] | sICAM-1 elevated in septic shock compared with non-infected controls (P < 0.05); | sICAM-1 associated with increasing sepsis severity P < 0.05; modest correlation with SOFA and APACHE-II; predicts mortality and severe sepsis (AUC of 0.72 (95% CI 0.57 to 0.87), 0.61 (95% CI 0.53 to 0.69)) |
Schuetz et al., [52] | 2011 | 161 | Patients with hypotension: 69 sepsis 35 cardiac 12 hemorrhagic 12 unknown | Clinical classification based on clinical and microbiological data | ICAM-1 higher in sepsis compared to non-sepsis (P < 0.05) | |
Hofer et al., [55] | 2009 | 147 | Surgical ICU patients with severe sepsis (101) and major abdominal surgery (28), and healthy controls (18) | 2003 ACCP/SCCM [2] | sICAM-1 higher in the septic group than postoperative and volunteer groups at diagnosis (444.7 ng/ml vs 213.7 ng/ml and 219.6 ng/ml, respectively; P < 0.001) | Not predictive of mortality at the time of diagnosis of sepsis, but non-survivors had trend to higher sICAM-1 levels at 48 h and 120 h (683.2 vs 434.1 ng/ml, P = 0.067; 360.2 vs 467.8 ng/ml, P = 0.083, respectively) compared to survivors |
Stief et al., [54] | 2007 | 86 | ICU patients with Sepsis (62), healthy controls (24) | Clinical definition of sepsis | Higher in sepsis than controls (2.56 ug/ml vs 0.19 ug/ml; P < 0.05) | |
Scherpereel et al., [53] | 2006 | 90 | ICU patients with sepsis (63), SIRS (7), healthy controls (20) | 1992 ACCP/SCCM [1] | sICAM-1 higher in sepsis compared to SIRS P < 0.02 | sICAM-1 not predictive of mortality or severity of sepsis |
Kinoshita et al., [56] | 2002 | 56 | Gram negative sepsis from intra-abdominal infection admitted to surgical ICU (47), healthy controls (9) | 1992 ACCP/SCCM [1] | sICAM-1 higher in sepsis than healthy controls | Not correlated with mortality in those with ARDS; Higher in those with ARDS than those without P < 0.05 |
Paterson et al., [57] | 2000 | 16 | ICU patients with SIRS (10), healthy controls (6) | 1992 ACCP/SCCM [1] | sICAM-1 not reported in healthy controls | Not correlated with mortality |
Weigand et al., [58] | 1999 | 21 | Surgical ICU patients with septic shock (14), healthy controls (7) | 1992 ACCP/SCCM [1] | sICAM-1 significantly higher in sepsis than controls (P < 0.05) | sICAM-1 significantly higher in nonsurvivors than survivors, sensitivity and specificity for cutoff of 800 ng/ml was 74.1% |
Froon et al., [73] | 1998 | 42 | ICU patients with sepsis and VAP | 1992 ACCP/SCCM [1] | sICAM-1 higher in VAP patients complicated by severe sepsis or septic shock than other VAP patients, but statistical significance not achieved | Not predictive of mortality, and correlates poorly with SAPS-II (r = 0.16, P = 0.30) |
Kayal et al., [59] | 1998 | 41 | ICU patients with severe sepsis or septic shock (25), ICU controls (7), healthy controls (9) | 1992 ACCP/SCCM [1] | sICAM-1 higher in septic patients than in noninfected ICU controls and healthy volunteers (P < 0.0001); higher in septic shock than those without septic shock (P < 0.05) | sICAM-1 correlated with mortality; correlated with SAPS and MOF score (r = 0.53, P < 0.01 for MOF) |
Boldt et al., [60] | 1997 | 30 | Surgical ICU patients with post-operative sepsis (30), healthy controls (not stated) | 1992 ACCP/SCCM1 | sICAM-1 higher in septic patients than healthy controls | Higher in older than younger patients P < 0.05, and tends to increase in older patients and decrease in younger patients over time |
Egerer et al., [61] | 1997 | 24 | ICU patients with infection (8), severe sepsis (16) | 1992 ACCP/SCCM [1] | sICAM-1 higher in severe sepsis compared with patients with infection (P > 0.05) | Not correlated with mortality in patients with severe sepsis |
Takakuwa et al., [62] | 1997 | 34 | ICU admissions with sepsis (20), trauma (14) | Clinical definition of SIRS and sepsis | sICAM-1 level higher in septic than trauma patients (987.7 vs 472.1 ng/ml; P = 0.0002) | sICAM-1 correlated with endotoxin, TNF-α, IL-6, IL-8, Type II PLA2 (Type II phospholiaps A2), NO (P < 0.05 for all) |
Boldt et al., [63] | 1996 | 30 | Surgical ICU patients with postoperative sepsis (15), trauma (15) | 1992 ACCP/SCCM [1] | sICAM-1 higher in sepsis than trauma (1,266 vs 444 ng/ml; P < 0.01) | |
Endo et al., [64] | 1996 | 28 | ICU patients with sepsis with MOF (8), sepsis without MOF (15), MOF without sepsis (5) | Clinical diagnosis of sepsis | sICAM-1 higher in septic patients with or without MOF than patients with MOF but no infection (1103.3 vs 356.0 ng/ml, and 862.5 vs 356.0 ng/ml, respectively, P < 0.05)) | sICAM-1 level higher in septic patients with MOF than those without (P = 0.0401) |
Moss et al., [66] | 1996 | 55 | ICU patients with sepsis (19), trauma (36) controls (5) | Clinical diagnosis of sepsis | sICAM-1 higher in septic patients than trauma and controls (573 vs 148 and 235 ng/ml, respectively, P < 0.001) | |
Nakae et al., [67] | 1996 | 34 | ICU patients with sepsis (21), trauma (13) | 1992 ACCP/SCCM [1] | sICAM-1 higher in septic patients than in trauma patients (987 vs 472 pg/ml; P = 0.0002) | sICAM-1 correlated with endotoxin, TNF-alpha and IL-8 (P < 0.05 for all) |
Sessler et al., [68] | 1995 | 66 | ICU patients with sepsis (25), SIRS (25), ICU controls (4), healthy volunteers (12) | 1992 ACCP/SCCM [1] | sICAM-1 higher in sepsis than ICU controls and healthy controls (1,259 vs 585 ng/ml, P < 0.001; 1,259 vs 355 ng/ml, P < 0.0001); sICAM-1 is higher in SIRS than ICU controls and healthy controls (937 vs 585 ng/ml, P < 0.05; 937 vs 355 ng/ml, P < 0.001); higher in sepsis vs SIRS (1,259 vs 937 ng/ml; P = 0.12) | sICAM-1 elevated with increasing severity of illness: septic shock, severe sepsis and sepsis (1,551, 796, and 542 ng/ml, respectively, ANOVA P = 0.017); correlated with cumulative MOF score, shock severity score (r = 0.46, P = 0.021; r = 0.50, P < 0.009); higher in nonsurvivors vs survivors (1,697 vs 854 ng/ml; P = 0.0096) |
Cowley et al., [65] | 1994 | 125 | ICU patients with sepsis (21), severe sepsis (14), ICU controls (5), healthy controls (85) | Clinical definition of SIRS and sepsis | sICAM-1 higher in severe sepsis, uncomplicated sepsis, and ICU controls than healthy controls P < 0.05. | sICAM-1 with no significant difference between severe sepsis, uncomplicated sepsis and ICU controls. Not correlated with mortality |
Study | Year | N | Population | Standard Criteria for SIRS/Sepsis | Association with sepsis | Other outcomes |
---|---|---|---|---|---|---|
Shapiro et al., [51] | 2010 | 221 | ED patients with sepsis without organ dysfunction (71), severe sepsis without shock (66), septic shock (71), and non-infected controls (13) | 1992 ACCP/SCCM [1] | sVCAM-1 elevated in septic shock compared with non-infected controls (P < 0.05) | sVCAM-1 associated with sepsis severity P < 0.04; predicts mortality and severe sepsis (AUC of 0.57 (95% CI 0.35 to 0.79), 0.60 (95% CI 0.52 to 0.69)) |
Hofer et al., [55] | 2009 | 147 | Surgical ICU patients with severe sepsis (101), major abdominal surgery (28), healthy controls (18) | 2003 ACCP/SCCM [2] | sVCAM-1 did not differentiate between septic, postoperative and healthy controls | sVCAM-1 not predictive of mortality at the time of diagnosis of sepsis, but nonsurvivors had elevated sVCAM-1 at 48 h and 120 h compared to survivors(1,275.1 vs 882.0 ng/ml, P = 0.027; 1,685.5 vs 748.5 ng/ml; P = 0.021, respectively) |
Kinoshita et al., [56] | 2002 | 56 | Gram negative sepsis from intra-abdominal infection admitted to surgical ICU (47), healthy controls (9) | 1992 ACCP/SCCM [1] | sVCAM-1 higher in patients than healthy controls | sVCAM-1 did not differentiate those with ARDS from those without; not predictive of mortality in those with ARDS |
Presterl et al., [69] | 1999 | 40 | ICU patients with Candida (20) and bacterial sepsis (20) | 1992 ACCP/SCCM [1] | At all times (days 1, 7, 14) sVCAM-1 levels higher in Candida sepsis than bacterial sepsis (P < 0.05) | sVCAM-1 not correlated with mortality |
Knapp et al., [78] | 1998 | 54 | Patients with sepsis (28 gram positive, 11 gram negative), 15 healthy controls | 1992 ACCP/SCCM [1] | sVCAM-1 elevated in sepsis compared with healthy controls (P < 0.05) | sVCAM-1 does not correlate with mortality in gram positive sepsis; does not distinguish between gram positive and gram negative sepsis |
Boldt et al., [60] | 1997 | 30 | Surgical ICU patients with post-operative sepsis (30), healthy controls (not stated) | 1992 ACCP/SCCM [1] | sVCAM-1 higher in septic patients than healthy controls | Higher in older than younger patients P < 0.05, and tends to increase in older patients and decrease in younger patients over time |
Takakuwa et al., [62] | 1997 | 34 | ICU admissions with sepsis (20), trauma (14) | Clinical definition of SIRS and sepsis | sVCAM-1 higher in septic than trauma patients (2,536 vs 1,019.0 ng/ml; P = 0.0004) | sVCAM-1 level correlated with TNF-α, IL-6, IL-8, NO, sE-selectin-1 ((P < 0.05 for all) |
Boldt et al., [63] | 1996 | 30 | Surgical ICU patients with postoperative sepsis (15), trauma (15) | 1992 ACCP/SCCM [1] | sVCAM-1 is higher in sepsis than trauma (1,042 vs 689 ng/ml; P < 0.05) | |
Endo et al., [64] | 1996 | 28 | ICU patients with sepsis with MOF (8), sepsis without MOF (15), MOF without sepsis (5) | Clinical diagnosis of sepsis | sVCAM-1 higher in septic patients with or without MOF than patients with MOF but no infection (2,654.9 vs 945.0 ng/ml, P = 0.0295; 2,045.0 vs 945.0 ng/ml, P = 0.0037) | sVCAM-1 did not differ between septic patients with and without MOF (2,654.9 vs 2,045.0 ng/ml; P = 0.1315) |
Furian et al., [76] | 2011 | 45 | Patients admitted to ICU with severe sepsis or septic shock | 1992 ACCP/SCCM [1] | sVCAM-1 not associated with left ventricular function or size. | |
Schuetz et al., [52] | 2011 | 161 | Patients with hypotension: 69 sepsis, 35 cardiac, 12 hemorrhagic, 12 unknown | Clinical classification based on clinical and microbiological data | VCAM-1 higher in sepsis compared to non-sepsis (P < 0.05) | |
Cowley et al., [65] | 1994 | 125 | ICU patients with sepsis (21), severe sepsis (14), ICU controls (5), healthy controls (85) | Clinical definition of SIRS and sepsis | sVCAM-1 is higher in sepsis than controls | sVCAM-1 higher in severe sepsis than uncomplicated sepsis at baseline (P = 0.06) and peak concentrations P < 0.01. Not correlated with mortality |
Study | Year | N | Population | Standard Criteria for SIRS/Sepsis | Association with sepsis | Other outcomes |
---|---|---|---|---|---|---|
Schuetz et al., [52] | 2011 | 161 | Patients with hypotension: 69 sepsis, 35 cardiac, 12 hemorrhagic, 12 unknown | Clinical classification based on clinical and microbiological data | E-selectin higher in sepsis compared to non-sepsis (P < 0.05) E-selectin independently associated with sepsis after adjustment for age, sex, blood pressure and mortality (P = 0.001) with AUC of 0.74 for discrimination of sepsis and non-sepsis | |
Shapiro et al., [51] | 2010 | 221 | ED patients with sepsis without organ dysfunction (71), severe sepsis without shock (66), septic shock (71), and non-infected controls (13) | 1992 ACCP/SCCM [1] | sE-selectin-1 levels elevated in septic shock compared with non-infected controls | sE-selectin-1 associated with sepsis severity P < 0.001; predicts mortality and severe sepsis (AUC of 0.65 (95% CI 0.49 to 0.82) and 0.71 (95% CI 0.64 to 0.78) respectively) |
Stief et al., [54] | 2007 | 86 | ICU patients with Sepsis (62), healthy controls (24) | Clinically diagnosed sepsis | sE-selectin-1 elevated in sepsis compared to reference value (190 ng/ml vs 55 ng/ml; P < 0.05)) | |
Kinoshita et al., [56] | 2002 | 56 | Gram negative sepsis from intra-abdominal infection admitted to surgical ICU (47), healthy controls (9) | 1992 ACCP/SCCM [1] | sE-selectin-1 does not differentiate between ARDS from non ARDS | Not predictive of mortality in those with ARDS |
Geppert et al., [74] | 2000 | 32 | ICU patients on day two post successfulCPR (25), non-critically ill controls (7) | 1992 ACCP/SCCM [1] | sE-selectin-1 higher in SIRS compared to controls (96.2 ng/ml vs 42.8 ng/ml; P = 0.23), but does not differentiate patients with SIRS vs patients without SIRS | Higher in non-survivors than survivors (114.2 ng/ml vs 85.7 ng/ml; P = 0.025) |
Osmanovic et al., [72] | 2000 | 27 | ICU patients with sepsis with MOF (9), healthy controls (18) | Clinical definition of sepsis | sE-selectin-1 higher in sepsis compared to healthy controls (118 vs 28.5 ng/ml; P < 0.001) | |
Hynninen et al., [70] | 1999 | 20 | ICU patients with severe sepsis (11), severe acute pancreatitis (9) | 1992 ACCP/SCCM [1] | sE-selectin does not differentiation between those with severe acute pancreatitis and severe sepsis | Higher in those with higher SOFA scores (SOFA ≥ 10, P = 0.043), but not correlated with mortality |
Presterl et al., [69] | 1999 | 40 | ICU patients with candida (20) and bacterial sepsis (20) | 1992 ACCP/SCCM [1] | sE-selectin-1 lower in patients with Candida sepsis than bacterial sepsis (P < 0.05) on Day 1 | Higher in non-survivors |
Takala et al., [71] | 1999 | 76 | Hospitalized patients with sepsis with organ failure (8) and without organ failure (12); healthy controls (56) | 1992 ACCP/SCCM [1] | sE-selectin-1 level elevated in septic patients compared to healthy adults P < 0.001 | Not correlated with organ failure |
Weigand et al., [58] | 1999 | 21 | Surgical ICU patients with septic shock (14), healthy controls (7) | 1992 ACCP/SCCM [1] | sE-selectin-1 higher in sepsis than healthy controls (P < 0.05) | Not predictive of mortality or severity of disease |
Froon et al., [73] | 1998 | 42 | ICU patients with sepsis and VAP | 1992 ACCP/SCCM [1] | sE-selectin-1 higher in patients with severe sepsis or septic shock than other VAP patients, but statistical significance not achieved | Day 2 sE-selectin-1 higher in nonsurvivors than survivors (114.3 vs 67.0 ng/ml; P = 0.04); Correlates poorly with SAPSII (r = 0.18, P = 0.25) |
Kayal et al., [59] | 1998 | 41 | ICU patients with severe sepsis or septic shock (25), ICU controls (7), healthy controls (9) | 1992 ACCP/SCCM [1] | sE-selectin-1 higher in septic patients than noninfected ICU controls and healthy volunteers (p < 0.0001); higher in those with septic shock than those without (p < 0.05) | sE-selectin-1 higher in nonsurvivors than survivors on day 0 (286 vs 195 ng/ml; P < 0.05), but decreases after Day 3 of sepsis to reach a level similar to that of survivors Day 14; correlated with SAPS and MOF score (r = 0.45, P < 0.05 for MOF) |
Knapp et al., [78] | 1998 | 54 | Patients with sepsis (28 gram positive, 11 gram negative), 15 healthy controls | 1992 ACCP/SCCM [1] | sE-selectin-1 higher in septic patients than controls p < 0.05 | sE-selectin-1 higher in nonsurvivors than survivors of gram positive sepsis on day 0, 4 and 7 (175 vs 85 ng/ml, P < 0.01; 155.7 vs 78.8 ng/ml, P < 0.05; 180 vs 76.1 ng/ml, P < 0.001, respectively); does not differentiate gram positive from gram negative infections. |
Boldt et al., [60] | 1997 | 30 | Surgical ICU patients with post-operative sepsis (30), healthy controls (not stated) | 1992 ACCP/SCCM [1] | sE-selectin-1 higher in septic patients than healthy controls | Higher in older than younger patients P < 0.05, and tends to increase in older patients and decrease in younger patients over time |
Cummings et al., [79] | 1997 | 119 | ICU patients with sepsis (67), SIRS (44), ICU controls (8) | 1992 ACCP/SCCM [1] | sE-selectin-1 higher in culture positive sepsis than culture negative sepsis, SIRS and ICU controls (15.39 vs 4.87, 2.33, and 1.97 ng/ml, respectively; P < 0.0001) | Day 1 levels higher for nonsurvivors than survivors (10.61 vs 4.35 ng/ml of log transformed mean sE-selectin-1; P < 0.05); sE-selectin-1 correlates strongly to the degree of hemodynamic compromise (P < 0.0001), and moderately with the peak MOF score (r = 0.30, P = 0.001) |
Egerer et al., [61] | 1997 | 24 | ICU patients with infection (8), severe sepsis (16) | 1992 ACCP/SCCM [1] | Higher in patients with severe sepsis and MOF than those with infection alone (P < 0.05) | Higher in nonsurvivors than survivors on Day 7-8, P < 0.05 |
Takakuwa et al., [62] | 1997 | 34 | ICU admissions with sepsis (20), trauma (14) | No Standard Definition | sE-selectin-1 higher in sepsis than trauma (287.9 vs 195.0 ng/ml; P = 0.0055) | sE-selectin-1 level correlated with TNF-α, IL-8, Type II PLA2, sICAM-1 (P < 0.005 for all) |
Boldt et al., [63] | 1996 | 30 | Surgical ICU patients with postoperative sepsis (15), trauma (15) | 1992 ACCP/SCCM [1] | sE-selectin-1 higher in sepsis than trauma (340 vs 57.9 ng/ml; P < 0.05) | |
Endo et al., [64] | 1996 | 28 | ICU patients with sepsis with MOF (8), sepsis without MOF (15), MOF without sepsis (5) | Clinical diagnosis of sepsis | sE-selectin-1 higher in septic patients with or without MOF than patients with MOF but no infection (345.2 vs 121.8 ng/ml, P = 0.0016; 266.2 vs 121.8 ng/ml, P = 0.0054) | sE-selectin-1 did not differ significantly between septic patients with and without MOF (345.2 vs 266.2 ng/ml; P = 0.2939) |
Moss et al., [66] | 1996 | 55 | ICU patients with sepsis (19), trauma (36) controls (5) | Clinical diagnosis of sepsis | Higher in sepsis than trauma and controls (573 vs 148 and 235 ng/ml, respectively, P < 0.001) | |
Simons et al., [75] | 1996 | 50 | Multiple trauma patients, infectious complications in 14 | Not specified | sE-selectin-1 higher in patients who subsequently developed infection, organ dysfunction, or both, by 36 h. P = 0.08 | sE-selectin-1 higher in non-survivors than survivors (P = 0.0018) |
Cowley et al., [65] | 1994 | 125 | ICU patients with sepsis (21), severe sepsis (14), ICU controls (5), healthy controls (85) | Clinical definition of SIRS and sepsis | sE-selectin higher in sepsis than controls (P < 0.01). | sE-selectin-1 higher in severe sepsis than uncomplicated sepsis on presentation (P < 0.01) and more pronounced with peak values (P < 0.001). Not correlated with mortality |
Newman et al., [80] | 1993 | 88 | ICU patients with sepsis with positive blood cultures (17), healthy controls (71) | Clinical definition of sepsis | Higher in septic shock than controls (23.3 vs 0.92 ng/ml; P < 0.05); not elevated in uncomplicated sepsis compared to controls |
Study | Year | N | Population | Standard Criteria for SIRS/Sepsis | Association with sepsis | Other outcomes |
---|---|---|---|---|---|---|
Scherpereel et al., [53] | 2006 | 90 | ICU patients with sepsis (63), SIRS (7), healthy controls (20) | 1992 ACCP/SCCM [1] | Higher in sepsis than SIRS or healthy controls (2.71 vs 0.77 and 0.68 ng/ml; P < 0.001);higher in septic shock than severe sepsis and uncomplicated sepsis (6.11 vs 1.97 and 1.95 ng/ml; P < 0.05, P < 0.02) | Endocan on ICU admission higher in nonsurvivors than patients still alive after 10 days (6.98 vs 2.54 ng/ml; P < 0.01), using a cutoff of 6.2 ng/ml, sensitivity and specificity are 75% and 84% respectively. |
Bechard et al., [23] | 2000 | 28 | Patients with septic shock (8), healthy controls (20) | 1992 ACCP/SCCM [1] | Higher in septic shock than healthy controls (7.815 vs 1.081 ng/ml; P = 0.0173) |
Soluble ICAM-1
Association with sepsis
Association with clinical outcome
Soluble VCAM-1 (sVCAM-1)
Association with sepsis
Association with clinical outcome
Soluble E-selectin
Association with sepsis
Association with clinical outcome
Endocan
Association with sepsis
Association with clinical outcome
Mediators of permeability and vasomotor tone
Study | Year | N | Population | Standard criteria for SIRS/sepsis | Association with sepsis | Other outcomes |
---|---|---|---|---|---|---|
Shapiro et al., [77] | 2008 | 83 | ED patients with septic shock (17), suspected infection without shock (66), and non-infected controls | Suspected infection based on treating clinician | VEGF levels higher in septic shock and infected patients without shock compared with non-infected controls (P < 0.01) | VEGF correlated with APACHE-II score at presentation (P = 0.01) |
Karlsson et al., [82] | 2008 | 280 | Septic ICU patients (250) and healthy controls (30) | 1992 ACCP/SCCM [1] | VEGF levels higher in severe sepsis compared with healthy controls at 0 and 72 h (P = 0.029, 0.003, respectively) | VEGF lower in non-survivors at 0 and 72 h (P = 0.012, 0.009, respectively), no correlation with SOFA scores |
Kumpers et al., [42] | 2008 | 72 | Medical ICU (43) and healthy controls (29) | 2003 ACCP/SCCM [2] | VEGF levels lower in non-septic and septic patients compared with healthy controls (P < 0.0001) | No association with severity of sepsis |
Van der Heijden et al., [45] | 2008 | 112 | Mechanically ventilated patients with sepsis (24) and without (88) | 1992 ACCP/SCCM [1] | VEGF levels higher in patients with sepsis than without sepsis (63.6 vs 20.7 pg/ml, P = 0.012) | VEGF trended higher in patients compared with controls (P = 0.268); No association with incidence of ALI/ARDS |
Van der Flier et al., [83] | 2005 | 58 | Severe sepsis (18) and healthy controls (40) | 1992 ACCP/SCCM [1] | VEGF levels elevated in sepsis compared with healthy controls (134 vs 55 pg/ml, P < 0.001) | VEGF correlated with mortality (P = 0.018) |
Yang et al., [101] | 2011 | 101 | 81 pneumonia and septic shock 20 pneumonia without organ dysfunction | 1992 ACCP/SCCM [1] | VEGF levels lower in septic shock vs. pneumonia (P = 0.005) | Day 1 VEGF did not discriminate survivors from non-survivors (P = 0.46) |
Rafat et al., [84] | 2007 | 62 | Medical ICU with sepsis (32), without sepsis (15), and healthy controls (15) | 1992 ACCP/SCCM [1] | VEGF levels elevated in septic compared with non-septic patients (1,351 vs 477 pg/ml, P < 0.01) | VEGF not correlated with mortality (P < 0.48) |
Study | Year | N | Population | Standard criteria for SIRS/sepsis | Association with sepsis | Other outcomes |
---|---|---|---|---|---|---|
Schuetz et al., [52] | 2011 | 161 | Patients with hypotension: 69 sepsis, 35 cardiac, 12 hemorrhagic, 12 unknown | Clinical classification based on clinical and microbiological data | sFlt-1 higher in sepsis compared to non-sepsis (P < 0.05) SFlt-1 independently associated with sepsis after adjustment for age, sex, blood pressure and mortality (P = 0.03) with AUC 0.70 for discrimination of sepsis from non-sepsis | |
Shapiro et al., [77] | 2008 | 83 | ED patients with septic shock (17), suspected infection without shock (66), and non-infected controls | Suspected infection based on treating clinician | sFLT levels elevated with worsening disease: non-infected, suspected infection without shock, septic shock (159, 386 and 551 ng/dL, respectively, P < 0.01) | sFLT correlated with APACHE-II, SOFA scores upon presentation and at 24 h (P < 0.05 for all) |
Shapiro et al., [51] | 2010 | 221 | ED patients with sepsis without organ dysfunction (71), severe sepsis without shock (66), septic shock (71), and non-infected controls (13) | 1992 ACCP/SCCM [1] | sFLT levels elevated in septic shock compared with non-infected controls (243 vs 41 ng/ml, P < 0.001) | sFLT correlated with SOFA, APACHE-II, lactate; Predicted severe sepsis and mortality (AUC of 0.82 (95% CI 0.76 to 0.88), 0.91 (95% CI 0.87 to 0.95)) |
Study | Year | N | Population | Standard criteria for SIRS/sepsis | Association with sepsis | Other Outcomes |
---|---|---|---|---|---|---|
Schuetz et al., [81] | 2007 | 95 | Consecutive ICU admissions with SIRS, sepsis, septic shock | 1992 ACCP/SCCM [1] | Endothelin-1 rises with sepsis, septic shock, compared with SIRS (64.3, 131.6, 23.1 pmol/L, respectively; P < 0.01 between SIRS and sepsis, P < 0.05 between sepsis and septic shock) | Endothelin-1 not correlated with mortality (p = 0.87) |
Piechota et al., [85] | 2007 | 20 | Medical ICU patients with sepsis | 1992 ACCP/SCCM [1], severity graded by procalcitonin and C-reactive protein level | Endothelin-1 correlates with CRP and PCT levels as estimates of level of sepsis severity (P < 0.05 for both) | Endothelin-1 correlates with SOFA score (p < 0.001) |
Weitzberg et al., [86] | 1991 | 16 | Sepsis (6) and healthy controls (10) | Bone et al., [102] | Endothelin-1 elevated in sepsis compared with healthy controls (11.3 vs. 2.4 pmol/l, P < 0.01) | n/a |
Furian et al. [76] | 2011 | 45 | Patients admitted to ICU with severe sepsis or septic shock | 1992 ACCP/SCCM [1] | Endothelin-1 levels associated with left ventricular and right ventricular function. (p = 0.002) | |
Pittet et al., [87] | 1991 | 40 | Sepsis (14), post-operative cardiac surgery (15) and healthy controls (11) | Bone et al., [102] | Endothelin-1 elevated in septic patients compared with healthy controls (19.9 vs 6.1 pg/ml, P < 0.0001) | Endothelin-1 inversely correlated with cardiac index (p < 0.005); correlated with APACHE-II scores (p < 0.01) |
Soluble VEGF
Soluble FLT (sFlt)
Endothelin-1
Mediators of coagulation
Study | Year | N | Population | Standard criteria for SIRS/sepsis | Association with sepsis | Other outcomes |
---|---|---|---|---|---|---|
Claus et al., [89] | 2009 | 63 | ICU patients with severe sepsis (11), non-elective cardiac surgery (22), elective cardiac surgery as ICU controls (24), and post-exercise as healthy controls (6) | 1992 ACCP/SCCM [1] | VWF:Ag higher in patients with sepsis and post non-elective cardiac surgery than ICU controls (P < 0.05) | VWF:Ag shows tendency to discriminate survivors from nonsurivors |
Bockmeyer et al., [90] | 2008 | 57 | ICU patients with sepsis (11), non-elective cardiac surgery (22), and elective cardiac surgery as ICU controls (24) | Not specified | VWF:Ag higher in sepsis and in non-elective cardiac surgery than ICU controls (both P < 0.001) | VWF:Ag correlated with mortality (P < 0.05) |
van der Heijden et al., [45] | 2008 | 112 | Mechanically ventilated patients, with sepsis (24) and without (88) | 1992 ACCP/SCCM [1] | VWF higher in patients with sepsis than without sepsis (P < 0.001) | VWF correlated with mortality (P = 0.006); VWF higher in those with ALI/ARDS than those without (P < 0.001) |
Hovinga et al., [95] | 2007 | 80 | Medical and surgical ICU patients with severe sepsis or septic shock (40), and healthy controls (40) | 1992 ACCP/SCCM [1] | VWF:Ag and VWF:RCO higher in sepsis than controls (P < 0.001) | VWF not correlated with disease severity, organ dysfunction, or mortality |
Martin et al., [91] | 2007 | 89 | ICU patients with severe sepsis (30), sepsis-unrelated organ failure (29), and healthy controls (30) | 1992 ACCP/SCCM [1] | VWF:Ag tends to differentiate severe sepsis from sepsis-unrelated organ failure (P > 0.05) | VWF:Ag not correlated with mortality |
Scherpereel et al., [53] | 2006 | 90 | ICU patients with sepsis (63), SIRS (7), and healthy controls (20) | 1992 ACCP/SCCM [1] | VWF higher in sepsis than SIRS (P < 0.02) | VWF correlated with mortality (P = 0.039) |
Ware et al., [94] | 2001 | 51 | ICU patients with ALI, ARDS (45% due to sepsis) | Temperature > 38° or < 35°C, systolic blood pressure < 90 mmHg (or a drop of 20 mm Hg or more in the systolic blood pressure from baseline), both present for at least 2 h; AND a clinically identifiable source of infection [103] | VWF:Ag higher in patients with sepsis than those without (P < 0.05) | VWF correlated with mortality (P < 0.005); higher in those with longer duration of ventilation P < 0.005; not correlated with illness severity scores (SAPSII, Lung Injury Score) |
Garcia-Fernandez et al., [92] | 2000 | 80 | ICU patients with SIRS and acute renal failure (40), and healthy controls (40) | 1992 ACCP/SCCM [1] | VWF higher in SIRS than controls (P < 0.001) | |
Bajaj et al., [97] | 1999 | 60 | Ward and ICU patients with ARDS (18), at risk of ARDS (15), and healthy controls (27) | Clinical diagnosis of sepsis | VWF does not differentiate patients with ARDS due to sepsis from other etiologies | VWF higher in ARDS (P < 0.001) and at risk ARDS (P < 0.01) compared to controls but did not differ significantly between these two groups |
Kayal et al., [59] | 1998 | 41 | ICU patients with severe sepsis or septic shock (25), ICU controls (7), healthy controls (9) | 1992 ACCP/SCCM [1] | VWF:Ag higher in sepsis than noninfected ICU controls and healthy controls (P < 0.0001); higher in septic shock than those without septic shock (P < 0.01) | VWF:Ag correlated with mortality (P < 0.01); correlated with SAPS and MOF score (r = 0.57, P < 0.01 for MOF) |
Moss et al., [66] | 1996 | 66 | ICU patients with sepsis (19), trauma (36), healthy controls (11) | Clinical diagnosis of sepsis | VWF:Ag higher in septic patients than trauma patients and controls (both P < 0.001) | |
Moss et al., [98] | 1995 | 96 | Hospitalized patients at risk of ARDS, including sepsis (30) | Clinical diagnosis of sepsis | VWF:Ag not predictive of the development of ARDS | |
Lorente et al., [93] | 1993 | 48 | ICU patients with septic shock | 1992 ACCP/SCCM [1] | VWF:Ag not predictive of mortality | |
Rubin et al., [96] | 1990 | 45 | ICU patients with nonpulmonary sepsis | Clinical diagnosis of sepsis | VWF:Ag correlated with mortality (P < 0.005) and ALI (P < 0.01) |
Study | Year | N | Population | Standard criteria for SIRS/sepsis | Association with sepsis | Other outcomes |
---|---|---|---|---|---|---|
Claus et al., [89] | 2009 | 63 | ICU patients with severe sepsis (11), non-elective cardiac surgery (22), elective cardiac surgery as ICU controls (24), and post-exercise as healthy controls (6) | 1992 ACCP/SCCM [1] | ADAMTS13 activity lower in sepsis than ICU reference group (P < 0.001) | ADAMTS13 activity correlated with mortality (P < 0.05) |
Bockmeyer et al., [90] | 2008 | 57 | ICU patients with sepsis (11), non-elective cardiac surgery (22), and elective cardiac surgery as ICU controls (24) | Not specified | ADAMTS13 activity lower in sepsis than ICU controls (P < 0.01) | ADAMTS13 activity correlated with mortality (P < 0.01) |
Hovinga et al., [95] | 2007 | 80 | Medical and surgical ICU patients with severe sepsis or septic shock (40), and healthy controls (40) | 1992 ACCP/SCCM [1] | ADAMTS13 activity lower in sepsis than healthy controls (P < 0.001) | ADAMTS13 activity not correlated with disease severity, organ dysfunction, or mortality |
Martin et al., [91] | 2007 | 89 | ICU patients with severe sepsis (30), sepsis-unrelated organ failure (29), and healthy controls (30) | 1992 ACCP/SCCM [1] | ADAMTS13 activity lower in severe sepsis than sepsis-unrelated organ failure (P < 0.05) and healthy controls (P < 0.05) | ADAMTS13 activity correlated with APACHE II (r = -0.66, P < 0.0001), number of organ failures (r = -0.66, P < 0.0001), and mortality (P = 0.02 by log rank test) |
Von Willebrand factor (vWF)
Association with sepsis
Association with clinical outcome
ADAMTS13
Discussion
Conclusions
Key messages
-
Multiple molecules reflecting endothelial activation are correlated with the presence of sepsis in humans and other clinically important outcomes.
-
The clinical utility or application of these molecules as biomarkers in sepsis; however, is limited by a lack of standardization in analytical assays, a lack of data regarding receiver operating characteristics and a lack of validation.
-
The consistent association with sepsis, demonstrable dose-response relationship, and temporal progression in patients who develop sepsis make Angiopoietin-2 an attractive potential biomarker in sepsis.
-
Future research should focus on standardization of assays and identification of cut points or thresholds that make biomarkers clinically useful in the diagnosis or stratification of patients presenting with presumed sepsis.
-
Evaluation of multiplexed panels with biomarkers of differential response characteristics may prove useful as a diagnostic strategy.