Severe sepsis has a high risk of mortality and is characterized by features of systemic inflammation and organ dysfunction. Inflammation, hypotension, and hypoxia are pivotal in the development of tissue injury and organ dysfunction. In sepsis a wide range of vasoactive agents, including eicosanoids, endothelins, and nitric oxide, is released, resulting in a disturbance of the regulation of vascular tone and vascular permeability (1, 2). Leakage of plasma into the interstitial compartment causes hypotension and generalized edema. The development of edema negatively affects the respiratory system and tissue oxygenation.
Vascular endothelial growth factor (VEGF) was first identified as a tumor-produced vascular permeability factor (3). VEGF is a key molecule in the control of vascular permeability via interactions with the VEGF-receptor-2 on the endothelial cell. In addition, VEGF is essential for angiogenesis and plays a crucial role in wound healing. The significance of the VEGF pathway is illustrated by the fact that VEGF or VEGF-receptor knockout mice demonstrate embryonic lethality from defective vascular and hematopoietic development. A wide variety of cells produce VEGF, including peripheral blood monocytes, neutrophils, and platelets (4-6). In response to several agents often associated with sepsis such as gram-negative bacterial lipopolysaccharide, gram-positive bacterial components, and tumor necrosis factor-α, VEGF production or secretion of the intracellular pool of VEGF is increased (7-9). In addition, hypoxia increases VEGF production via hypoxia-inducible factor-1a transcription factor (3).
Several authors reporting the identification of new antagonists of VEGF have suggested that VEGF may be involved in the development of capillary leakage and associated hypotension during sepsis; however, whether VEGF levels are indeed increased in patients with sepsis has not been documented in these studies (10, 11).
The first aim of our study was to determine whether plasma VEGF levels are elevated in severe sepsis. The second aim was to determine whether plasma VEGF is associated with severity of organ dysfunction and mortality.
MATERIALS AND METHODS
We investigated plasma VEGF levels in 18 patients fulfilling criteria of severe sepsis (1992 ACCP/SCCM definition) (12). Blood samples from 40 healthy hospital employees served as controls for the VEGF levels. The study protocol was approved by the medical ethics committee of the University Medical Center Utrecht. All patients were enrolled in a clinical trial of adjuvant antithrombin III therapy in severe sepsis and were admitted to the intensive care unit of the University Medical Center Utrecht between November 1997 and January 2000 (13). Informed consent was obtained from the patients or their legal representatives.
Relevant clinical and laboratory characteristics were collected prospectively, including serum albumin levels and plasma lactate levels as a tissue hypoxia marker. Severity of disease was assessed by the Acute Physiology And Chronic Health Evaluation version II (APACHE II) scores and the multiple organ dysfunction score (MODS) (14, 15). Survival was followed for 30 days.
Blood sampling and VEGF measurement
Serial blood samples were collected in heparin-anticoagulated tubes on day 0 (at the time all criteria for severe sepsis were met) and days 1, 2, 3, 7, and 28 following diagnosis. Plasma VEGF levels in healthy volunteers were measured for control values. Blood was centrifuged twice at 1000 g for 10 min at 4°C. Plasma was stored at −80°C until VEGF measurement by ELISA (Quantikine; R&D systems). This assay detects free biologically active VEGF121 and VEGF165. In our hands, the lower limit of detection was 10 pg/mL, the intra-assay precision was 6%, and the interassay precision 9%.
Data are expressed as means ± SE of the mean unless indicated otherwise. Statistical analysis was performed using Student’s t test, two-way ANOVA with Bonferroni post-test at different time points, and Pearson correlation coefficients where appropriate. P values < 0.05 were considered statistically significant.
Eighteen patients with severe sepsis were included in the study, and 40 healthy controls were used for comparison purposes. Demographic and clinical characteristics are summarized in Table 1. In these patients underlying diseases were respiratory tract infection (n = 12), intra-abdominal infection (n = 1), genitourinary tract infection (n = 1), Fournier’s gangrene (n = 1), near drowning complicated by infection (n = 1), meningococcal infection (n = 1), and catheter-related infection (n = 1). Six patients (33%) had a positive blood culture at the time of study enrollment (gram positive n = 2; gram negative n = 4). Patients showed edema, and serum albumin values, used as an indirect measure of capillary leak syndrome, were decreased (albumin nadir 17 ± 1.8 g/L; normal range 35-50 g/L). Plasma lactate values on admission were increased in 6/12 (50%) patients (3.6 ± 0.9 mmol/L [range 0.9-11.7]; normal range 0.5-2.2 mmol/L), indicating tissue hypoxia in at least some of the patients studied. During the study period the individual MODS values further increased (maximum individual MODS 10 ± 1). The mortality in the study group was 10/18 (56%).
Plasma VEGF levels
At study entry plasma VEGF levels were elevated in patients with severe sepsis versus controls (134 ± 20 pg/mL vs. 55 ± 7; P < 0.001; Fig. 1). Maximum plasma VEGF levels of patients receiving antithrombin III (n = 10) or placebo (n = 8) were similar (230 ± 48 pg/mL vs. 251 ± 59). Maximum VEGF levels during the course of disease were higher in nonsurvivors than in survivors (313 ± 53 pg/mL vs. 147 ± 21; P = 0.018).
At day 28, the last sample point of our study, VEGF levels were normalized in seven of eight survivors. VEGF levels at study entry correlated to the maximum MOD score during the course of disease (Pearson correlation coefficient r = 0.75; P = 0.001; Fig. 2). The lag time to the development of maximum MOD scores was 12.1 ± 1 days. VEGF levels at study entry also correlated significantly to the MOD scores at study entry when the sample was taken (Pearson correlation coefficient r = 0.43; P = 0.037).
VEGF levels showed no significant correlation with the other parameters examined, including age, C-reactive protein levels, serum albumin values, plasma lactate levels, white blood cell counts, and platelet counts.
The major finding of the present study is that plasma VEGF levels from patients with severe sepsis are elevated and associated with disease severity.
Experimental studies in human volunteers have demonstrated that intravenous infusion of lipopolysaccharide causes a marked increase in plasma VEGF concentrations as a result of leukocyte activation (7). Our study is the first to report on plasma VEGF levels in clinical sepsis patients and to correlate these data to disease severity. Recently, elevated plasma VEGF values were reported in neonates with postoperative capillary leak syndrome following cardiopulmonary bypass, and elevated serum plasma levels were reported in patients with acute respiratory distress syndrome, suggesting that an increase in plasma VEGF levels is a more generally occurring phenomenon associated with systemic inflammatory response syndrome (16, 17). Moreover, we have reported previously an increase in cerebrospinal fluid VEGF levels in a series of bacterial meningitis patients (18). The increase in cerebrospinal fluid VEGF most likely resulted from VEGF secretion by infiltrating polymorphonuclear leukocytes.
It is likely that many factors contributed to the observed increase in VEGF levels during severe sepsis in the present study. Leukocytes stimulated with bacterial components or proinflammatory mediators, platelets activated by the procoagulant state of the endothelium, and vascular smooth muscle cells stimulated by hypoxia may have contributed to increased VEGF release (8, 19). The circulating half-life in humans has been reported to be 33.7 ± 13 minutes, thus implying sustained VEGF production in patients with severe sepsis (20). However, decreased clearance of plasma VEGF in septic patients may also have contributed to the increased plasma VEGF levels. In the present study, patients were treated with antithrombin III. Because the anticoagulant properties of antithrombin III may have decreased VEGF release from platelets, one might expect even higher plasma VEGF levels in septic patients receiving no anticoagulant therapy. Therefore antithrombin III treatment is unlikely to have confounded the findings in this study. This is in line with the findings by others (7). In our study the controls were of younger age than the patients with severe sepsis. However, we believe this does not compromise the validity of our results, as studies into the effect of age on normal values of circulating VEGF found no difference between healthy adults and elderly (21).
Increased circulating VEGF levels may have several effects. As mentioned previously VEGF is a potent vascular permeability factor. Animal experiments have demonstrated that VEGF is a key molecule in the control of vascular permeability via interactions with the VEGF-receptor-2 on the endothelial cell. VEGF-induced vascular permeability depends the activation state of the endothelial cells such as the presence of transmembrane-TNF-α. Via continuous autocrine stimulation of the endothelium transmembrane-TNF-α has a permissive role for VEGF-induced permeability (22). In vitro anti-TNF antibodies ablate the VEGF-induced permeability of cultivated human endothelial cells. Similarly conditions of hypoxic stress change the endothelial activation state and promote VEGF-induced endothelial permeability (23). This principle has been demonstrated clinically in patients undergoing VEGF gene therapy for lower-extremity ischemia: adenoviral VEGF vector treatment was complicated by augmented lower-extremity edema of the ischemic legs. VEGF gene therapy increased the permeability of the locally activated endothelium of the ischemic legs but did not cause generalized edema (24).
Because circulating inflammatory mediators in sepsis cause generalized stimulation of the endothelium, it may be envisioned that elevated plasma VEGF levels in patients with severe sepsis contributes to systemic vascular leak, hypotension, and finally generalized edema. Of course several other mediators released during systemic inflammation, such as nitric oxide and TNF-α, are also known to affect endothelial permeability. The septic patients in our study population showed signs of generalized vascular leakage, such as decreased serum albumin values. We did not detect a significant correlation between VEGF levels and serum albumin levels. It is possible that the power of this study was too low to detect such a relationship.
Interestingly, plasma VEGF levels did correlate to MODS values in our study; thus, VEGF levels are associated with disease severity. In addition, the maximum VEGF levels during the course of disease in individual patients were slightly higher in nonsurvivors than in survivors. These findings clearly indicate that VEGF levels are negatively associated with outcome.
However, the results of this study do not demonstrate a causal relationship between elevated VEGF levels and dismal outcome. Elevated VEGF levels may also be a sign of activation of endogenous tissue repair mechanisms. VEGF may influence tissue repair and wound healing in septic or other critically ill patients. Elevated serum VEGF levels have been reported in both burn and trauma patients, which may indeed relate to wound healing and tissue repair (25). Contrary to the effect on circulating VEGF concentrations, lipopolysaccharide may suppress the expression of VEGF receptors on endothelial cells, thereby counteracting VEGF related responses (26).
To date no effective therapy for systemic capillary leak is clinically available. However, recent animal studies demonstrated protection of adult vasculature against plasma leakage induced by VEGF and inflammatory agents by the functional VEGF antagonist angiopoietin-1 (11). In addition, other endogenous mediators that counteract the permeability-enhancing effect of VEGF, such as pigment epithelium-derived factor, have been discovered recently (10). Direct targeting of VEGF or its receptors may present a potential alternative therapeutic approach. The demonstrated effects of VEGF on endothelial permeability, the elevated VEGF levels found in patients with sepsis and capillary leak, the association with disease severity, all support a role for VEGF in the pathophysiologic events leading to increased endothelial permeability. Alternatively, VEGF may be involved in tissue repair processes and wound healing. Because our studies included only a small number of patients, our results will have to be confirmed in larger studies. Because in the present study we could detect a statistical association only between plasma VEGF level and the presence and severity of sepsis, additional studies are warranted to establish the functional role of VEGF in sepsis. We conclude that VEGF or its receptors may be a potential therapeutic target to reduce sepsis-related capillary leakage and to improve the prognosis of septic patients.
1. Sriskandan S, Cohen J: The pathogenesis of septic shock. J Infect
2. Marx G, Vangerow B, Burczyk C, Gratz KF, Maassen N, Cobas MM, Leuwer M, Kuse E, Rueckholdt H: Evaluation of noninvasive determinants for capillary leakage syndrome in septic shock patients. Intensive Care Med
3. Ferrara N, Gerber HP, LeCouter J: The biology of VEGF and its receptors. Nat Med
4. Bottomley MJ, Webb NJ, Watson CJ, Holt PJ, Freemont AJ, Brenchley PE: Peripheral blood mononuclear cells from patients with rheumatoid arthritis spontaneously secrete vascular endothelial growth factor (VEGF): specific up-regulation by tumour necrosis factor-alpha (TNF-alpha) in synovial fluid. Clin Exp Immunol
5. Gaudry M, Bregerie O, Andrieu V, El Benna J, Pocidalo MA, Hakim J: Intracellular pool of vascular endothelial growth factor in human neutrophils. Blood
6. Verheul HM, Hoekman K, Luykx-de Bakker S, Eekman CA, Folman CC, Broxterman HJ, Pinedo HM: Platelet: transporter of vascular endothelial growth factor. Clin Cancer Res
7. Mittermayer F, Pleiner J, Schaller G, Weltermann A, Kapiotis S, Jilma B, Wolzt M: Marked increase in vascular endothelial growth factor concentrations during Escherichia coli
endotoxin-induced acute inflammation in humans. Eur J Clin Invest
8. van der Flier M, Coenjaerts F, Kimpen JL, Hoepelman AM, Geelen SP: Streptococcus pneumoniae
induces secretion of vascular endothelial growth factor by human neutrophils. Infect Immun
9. McCourt M, Wang JH, Sookhai S, Redmond HP: Proinflammatory mediators stimulate neutrophil-directed angiogenesis. Arch Surg
10. Liu H, Ren JG, Cooper WL, Hawkins CE, Cowan MR, Tong PY: Identification of the antivasopermeability effect of pigment epithelium-derived factor and its active site. Proc Natl Acad Sci USA
11. Thurston G, Rudge JS, Ioffe E, Zhou H, Ross L, Croll SD, Glazer N, Holash J, McDonald DM, Yancopoulos GD: Angiopoietin-1 protects the adult vasculature against plasma leakage. Nat Med
12. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference: Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med
13. Warren BL, Eid A, Singer P, Pillay SS, Carl P, Novak I, Chalupa P, Atherstone A, Penzes I, Kubler A, Knaub S, Keinecke HO, Heinrichs H, Schindel F, Juers M, Bone RC, Opal SM: Caring for the critically ill patient. High-dose antithrombin III in severe sepsis: a randomized controlled trial. JAMA
14. Knaus WA, Draper EA, Wagner DP, Zimmerman JE: APACHE II: a severity of disease classification system. Crit Care Med
15. Marshall JC, Cook DJ, Christou NV, Bernard GR, Sprung CL, Sibbald WJ: Multiple organ dysfunction score: a reliable descriptor of a complex clinical outcome
. Crit Care Med
16. Abrahamov D, Erez E, Tamariz M, Dagan O, Pearl E, Abrahamov Y, Gendel B, Desai N, Kats J, Vidne B, Barak V: Plasma vascular endothelial growth factor level is a predictor of the severity of postoperative capillary leak
syndrome in neonates undergoing cardiopulmonary bypass. Pediatr Surg Int
17. Thickett DR, Armstrong L, Christie SJ, Millar AB: Vascular endothelial growth factor may contribute to increased vascular permeability in acute respiratory distress syndrome. Am J Respir Crit Care Med
18. van der Flier M, Stockhammer G, Vonk GJ, Nikkels PG, Diemen-Steenvoorde RA, Der Vlist GJ, Rupert SW, Schmutzhard E, Gunsilius E, Gastl G, Hoepelman AI, Kimpen JL, Geelen SP: Vascular endothelial growth factor in bacterial meningitis: detection in cerebrospinal fluid and localization in postmortem brain. J Infect Dis
19. Yancopoulos GD, Davis S, Gale NW, Rudge JS, Wiegand SJ, Holash J: Vascular-specific growth factors and blood vessel formation. Nature
20. Eppler SM, Combs DL, Henry TD, Lopez JJ, Ellis SG, Yi JH, Annex BH, McCluskey ER, Zioncheck TF: A target-mediated model to describe the pharmacokinetics and hemodynamic effects of recombinant human vascular endothelial growth factor in humans. Clin Pharmacol Ther
21. Larsson A, Skoldenberg E, Ericson H: Serum and plasma levels of FGF-2 and VEGF in healthy blood donors. Angiogenesis
22. Clauss M, Sunderkotter C, Sveinbjornsson B, Hippenstiel S, Willuweit A, Marino M, Haas E, Seljelid R, Scheurich P, Suttorp N, Grell M, Risau W: A permissive role for tumor necrosis factor in vascular endothelial growth factor-induced vascular permeability. Blood
23. Fischer S, Clauss M, Wiesnet M, Renz D, Schaper W, Karliczek GF: Hypoxia induces permeability in brain microvessel endothelial cells via VEGF and NO. Am J Physiol
24. Baumgartner I, Rauh G, Pieczek A, Wuensch D, Magner M, Kearney M, Schainfeld R, Isner JM: Lower-extremity edema
associated with gene transfer of naked DNA encoding vascular endothelial growth factor. Ann Intern Med
25. Grad S, Ertel W, Keel M, Infanger M, Vonderschmitt DJ, Maly FE: Strongly enhanced serum levels of vascular endothelial growth factor (VEGF) after polytrauma and burn. Clin Chem Lab Med
26. Power C, Wang JH, Sookhai S, Street JT, Redmond HP: Bacterial wall products induce downregulation of vascular endothelial growth factor receptors on endothelial cells via a CD14-dependent mechanism: implications for surgical wound healing. J Surg Res