American Journal of Respiratory and Critical Care Medicine

We have previously reported, in patients with acute respiratory distress syndrome (ARDS), elevated plasma levels of vascular endothelial growth factor (VEGF) that became reduced in those who recovered. To examine the potential effect of VEGF on the epithelial side of the alveolar–capillary membrane, we compared VEGF levels in the epithelial lining fluid (ELF) of the same 40 patients with ARDS, and in 28 patients at risk of ARDS. We measured intrapulmonary VEGF levels in 23 patients on Days 1 and 4 after admission to the intensive therapy unit and related these levels to recovery. ELF from subjects with ARDS contained lower levels of VEGF than did ELF from at-risk subjects (1,076 and 7,674 pg/ml, respectively, p = 0.0004) and increased ELF levels at Day 4 were associated with recovery (p = 0.001). Alveolar macrophages from subjects with ARDS produced significantly less VEGF than those from at-risk subjects (6.3 and 13.0 pg/ml, respectively, p = 0.005). Similarly, alveolar neutrophils from subjects with ARDS produced significantly less VEGF than those at risk (13.9 and 31.5 pg/ml, respectively, p = 0.03). ELF VEGF levels inversely correlated with Lung Injury Score (p = 0.003). These studies suggest that VEGF in the alveolar space may reflect the development of, and recovery from, acute lung injury in a manner opposite to that in plasma.

Pulmonary injury in acute respiratory distress syndrome (ARDS) leads to disruption of both sides of the alveolar–capillary membrane with resultant hyperfiltration, alveolar flooding, and hypoxia. Endotoxin (lipopolysaccharide [LPS]), tumor necrosis factor-α (TNF-α), and interleukin (IL)-8 have all been implicated in the pathogenesis of this condition (1). The pulmonary endothelium has been considered to provide the key barrier to protein and fluid flow into the alveolus. However, some animal models of edema have highlighted the potential role of the alveolar epithelium (2). Despite therapeutic advances such as low tidal volume mechanical ventilation, no specific approaches are available to promote edema clearance or alveolar epithelial repair (3).

Human vascular endothelial growth factor (VEGF) occurs as at least four different proteins, VEGF121, VEGF165, VEGF189, and VEGF206, which may have different functions; VEGF165 and VEGF121 are soluble products, whereas VEGF189 and VEGF206 remain cell associated (4, 5). Studies of VEGF have focused on its ability to induce vascular permeability and act as a mitogen for endothelial cells. The biological activity of VEGF is dependent on its interaction with specific receptors. Three well-defined receptors have been identified: Flt-1 (also occurring as a soluble form, sFlt), Flt-4, and KDR (68). A wide variety of cells express VEGF receptors, including activated macrophages and alveolar type II epithelial cells (911). VEGF involvement in the control of alveolar–capillary membrane function has been demonstrated in animal models (1214). Lung overexpression of adenovirally delivered VEGF leads to pulmonary edema and increased lung vascular permeability (12). By contrast, in an interleukin-13 (IL-13) transgenic mouse model of hyperoxic lung injury and a fetal respiratory distress model, a protective effect is in part due to increased intrapulmonary VEGF (13, 14). We hypothesized that VEGF plays a role in both the development and outcome of human ARDS. We have investigated this hypothesis on both sides of the alveolar–capillary membrane and previously reported our findings concerning the vascular compartment, where we found increased plasma VEGF in patients with ARDS (15). To address the role of VEGF in the alveolar compartment, we measured VEGF levels in subjects with ARDS and at risk of ARDS at different stages of the condition as previously reported (15). In view of the potential interaction between VEGF and IL-13, we also measured IL-13 in epithelial lining fluid (ELF) (13). Our findings concur with the study of Maitre and coworkers, suggesting that in the human lung VEGF levels in ELF were reduced in early ARDS (16). We have extended those data by analysis of VEGF in the ELF of patients with resolving lung injury in whom VEGF levels increase, in contrast to our previously reported findings in plasma (15). We further investigated the potential regulation of VEGF production of alveolar cells (macrophages and neutrophils) by LPS, IL-13, IL-8, and hypoxia, which have been shown to regulate VEGF production in other contexts (17, 18). Alveolar cell VEGF production was also reduced in patients with ARDS in comparison with those at risk. These data indicate that VEGF concentrations in the alveolar space and plasma change in opposite directions with recovery from injury (15).


Patients were studied within 48 hours of admission to the Intensive Therapy Unit of Southmead Hospital (Bristol, UK) and have previously been described (15). The severity of lung injury was assessed according to the American–European consensus statement (19). Patients with bilateral infiltrates on chest X-ray and a PaO2:FiO2 ratio less than 200 mm Hg were defined as having ARDS (n = 40). Patients with predisposing risk factors for ARDS were classified as at risk (n = 28). These patients did not subsequently develop ARDS. Although some at-risk patients had elevated lung injury scores with low PaO2:FiO2 ratios, none had radiological evidence of bilateral alveolar shadowing and did not therefore fall into the “acute lung injury” category as defined by the consensus statement. Resolving lung injury was defined as a reduction in Murray Lung Injury Score (LIS) of more than 1 and in all cases was due to a combination of improvement on chest X-ray and PaO2:FiO2. Repeat bronchoscopy was performed in 23 subjects on Day 4 after admission to the Intensive Therapy Unit. Of the remaining 17, 8 had died, 6 were extubated, and 3 withheld consent. Bronchoscopy was also performed on 14 nonsmoking volunteers as normal control subjects and 13 patients who had been ventilated during abdominal aneurysm repair (no cardiopulmonary bypass) as ventilated control subjects. The etiology and severity scores for the ARDS and at-risk patients are shown in Table 1

TABLE 1. Patient characteristics*


 (n = 40)

At Risk
 (n = 28)
Abdominal sepsis10 (25%)8 (28.5%)
Pneumonia14 (35%)7 (25%)
Gastric aspiration9 (22.5%)3 (11%)
Massive blood transfusion3 (7.5%)4 (14%)
Pancreatitis2 (5%)2 (7%)
Other2 (5%)4 (14%)
APACHE II score28.9 (SE, 2.8)21.23 (SE, 2.3), p = 0.069
APACHE III score60.9 (SE, 19.4)57.7 (SE, 3.6), p = 0.53
SAPS II42.4 (SE, 2.5)42.8 (SE, 2.7), p = 0.91
PaO2:FIO2 ratio127.0280 p = 0.65
% BAL neutrophils71.073.7
Mean age (range)62 (18–93)68 (29–78)
Nonsurvival:survival (% mortality)
19:21 (47.5)
13:15 (46)

*As previously reported (15).

Definition of abbreviations: APACHE = Acute Physiology and Chronic Health Evaluation; ARDS = acute respiratory distress syndrome; BAL = bronchoalveolar lavage; F = female; LIS = Lung Injury Score; M = male; SAPS = Simplified Acute Physiology Score.

. The study was approved by the Southmead National Health Service Trust Local Research Ethics Committee.


Bronchoscopy was performed according to standard protocol. After topical anesthesia with 2% lidocaine, bronchoalveolar lavage (BAL) was performed with four 60-ml aliquots of buffered saline instilled into the (right) middle lobe. The BAL fluid was aspirated into a siliconized glass bottle and stored on ice until processing. Samples were processed within 15 minutes of collection. The chilled BAL fluid was strained through a single layer of coarse gauze to remove clumps of mucus and then spun at 400 × g for 5 minutes to recover cells. BAL fluid supernatant was collected and stored at –80°C until analysis.

Cell Purification

Alveolar neutrophils and macrophages were separated by selection with magnetic beads (Dynal, Wirral, UK) from 21 patients with ARDS and from 12 at-risk patients consecutively recruited. Cells were incubated for 30 minutes at 4°C with mouse anti-human CD66b (Serotec, Oxford, UK), an antibody specific for granulocytes. The cells were then washed for 5 minutes in RPMI medium before incubation at 4°C with rat anti-mouse IgG1-specific magnetic beads (Dynal; 100 μl of beads per 107 cells). Granulocytes were depleted from the cell suspension by exposure to a magnet for 2 minutes. The purified macrophages were then washed once in complete medium for 5 minutes at 500 × g and a Cytospin (Thermo Shandon, Pittsburgh, PA) was taken to assess purity. The resultant alveolar macrophages were more than 95% pure, with more than 90% viability. The cell pellet was resuspended in RPMI 1640 medium (Sigma, Poole, UK) supplemented with penicillin (100 U/ml) and streptomycin (100 μg/ml) (Sigma) and adjusted to 2 × 106 alveolar macrophages per milliliter, as defined by morphology. These were cultured in normoxia, alone or with LPS (10 μg/ml; Sigma) and IL-13 (10 or 100 μg/ml; R&D Systems, Abingdon, UK), or under hypoxic (5% O2) conditions for 24 hours.

Neutrophils adherent to magnetic beads were washed in complete medium and a Cytospin was taken to assess cell purity. The resultant neutrophils were used if purity was greater than 95% with greater than 90% viability. The cell pellet was resuspended in serum-free RPMI 1640 medium and adjusted to 2 × 106 neutrophils per milliliter. Neutrophils were cultured alone, with LPS (10 μg/ml), or with IL-8 (10 ng/ml) (R&D Systems).

VEGF and IL-13 Measurements

BAL fluid VEGF and IL-13 were measured by separate sandwich ELISAs (R&D Systems and PharMingen [San Diego, CA], respectively). A specific monoclonal antibody was precoated onto a microplate. Standards and samples were pipetted into the wells. Subsequently, a polyclonal detection antibody was added. Substrate solution was then added to the wells. The resultant color develops in proportion to the amount of specific cytokine present and was read spectrophotometrically in a plate reader. Recovery of VEGF and IL-13 from spiked BAL fluid was more than 95%. The detection limits of these assays were 3 pg/ml for VEGF and 8 pg/ml for IL-13. ELF levels were extrapolated from the BAL fluid data by comparing plasma and BAL urea levels.

Statistical Analysis

Significant differences within a group and between groups were determined by Mann–Whitney U analysis, using a Minitab for Windows package. The data were nonparametric. A p value less than 0.05 was considered significant.

Epithelial Lining Fluid Levels of VEGF Are Significantly Lower in Patients with ARDS

VEGF was detectable in the BAL fluid of all patients and control subjects. The differences between at-risk patients and the patients with ARDS were maintained when the dilutional effects of BAL on ELF were corrected by urea estimation.

VEGF levels in ELF from subjects with ARDS (median, 1,076 pg/ml) were significantly lower than in at-risk subjects (median, 7,674 pg/ml; confidence interval [CI], 1,442–7,290 pg/ml; p = 0.0004), normal (5,621 pg/ml; CI, 259–5,889 pg/ml; p = 0.02), or ventilated control subjects (median, 5,322 pg/ml; CI, 801–5,840 pg/ml; p = 0.0051) (Figure 1)

. There was a significant but modest correlation between ELF VEGF level and LIS at the time of bronchoscopy (Figure 2) .

In both ARDS patients and at-risk patients there were no significant correlations between ELF VEGF and total BAL cell count, cell count per milliliter of BAL fluid, percent neutrophils, percent macrophages, absolute neutrophil count per milliliter of BAL fluid, or absolute macrophage count per milliliter of BAL fluid.

Epithelial Lining Fluid Levels of IL-13 Do Not Differ between Patients with ARDS and Those at Risk

IL-13 levels were low in samples derived from ARDS, at-risk, and normal subjects (median, 795, 982, and 650 pg/ml, respectively). There were no significant differences between these groups. IL-13 was significantly elevated in ventilated control subjects, however, compared with the other three groups (median, 4,485 pg/ml; p = 0.05).

Serial BAL

BAL was performed serially where possible as described in Methods (23 patients). VEGF levels significantly increased in resolving patients only, from Day 1 to Day 4 (from 1,184 to 8,856 pg/ml; p = 0.012). The levels of VEGF on Day 4 were significantly higher in the resolving group compared with the nonresolving group (8,856 versus 1,598 pg/ml; p = 0.001) (Figure 3)

. IL-13 levels did not change significantly in either group between Day 1 and Day 4 (data not shown).

Alveolar Macrophage Culture

Alveolar macrophages from at-risk patients spontaneously produced significantly more VEGF than did alveolar macrophages from patients with ARDS and normal control subjects (13.0 versus 6.3 and 3.8 pg/ml, respectively; p < 0.005) (Figure 4)

. Stimulation of macrophages with LPS did not significantly alter VEGF production in any subjects. Stimulation of macrophages from patients with ARDS, using IL-13, significantly increased VEGF production, from 6.3 to 13.5 pg/ml (CI, 0.49–10.36 pg/ml; p = 0.05) (Figure 5) . Hypoxic treatment of macrophages from subjects with ARDS and at-risk subjects, however, had no effect on VEGF production (data not shown).

Alveolar Neutrophil Culture

Alveolar neutrophils from patients with or at risk of ARDS were cultured at 2 × 106/ml with IL-8 and LPS (Figure 6)

. Alveolar neutrophils from patients with ARDS produced significantly less VEGF in response to LPS than those from patients at risk of ARDS (13.9 pg/ml compared with 31.5 pg/ml; CI, 3.85–28.5 pg/ml; p = 0.03). Neutrophil VEGF production was significantly increased by IL-8 in both groups (to 37.7 pg/ml; CI, 4.5–32.21 [p = 0.009] and 29.4 pg/ml [p = 0.04], respectively).

The biological properties of VEGF as both a growth factor and a permeability factor have led to interest in its role within the lung in both health and disease. VEGF expression is critical for developmental vasculogenesis as demonstrated by embryonic lethality of heterozygous VEGF null mutants (20). During embryonic lung development there is increased expression of VEGF and Flt-1 mRNA in the late fetal and postnatal period in the mouse. The adult mouse lung expresses more VEGF mRNA than the developing lung, suggesting a persistent or additional and important function (21). However, in both animal lung and the human embryonic lung, mesenchymal and alveolar epithelial cells are the predominant source of VEGF as assessed by in situ hybridization (10, 21, 22). Constitutive VEGF within the normal human ELF has previously been described (23). These data demonstrate potential roles for VEGF on both sides of the alveolar–epithelial membrane.

We have explored the potential role for VEGF in the development of ARDS by assessment of both sides of this membrane. We have previously reported increased levels of plasma VEGF in patients with ARDS compared with those at risk (15). We now report our findings in this same group of patients on the epithelial side of the alveolar–epithelial membrane. In contrast to our original hypothesis and the plasma data, we detected lower levels of VEGF in the lungs of patients with ARDS compared with those from normal, at-risk, or ventilated subjects. In addition, we obtained further ELF samples at 4 days and found that increasing VEGF levels were associated with recovery. What is the purpose of this VEGF in the normal adult lung and what causes the apparent reduction in the alveolar space in ARDS?

Some insight into these paradoxical findings comes from considering animal models. Kaner and coworkers have demonstrated that intrapulmonary overexpression of VEGF results in high-permeability edema in the lungs of mice (12). Inducing the expression of sFlt-1 as a biological inhibitor of VEGF blocked this edema. These data suggest that VEGF may regulate baseline microvascular permeability and suggest that elevated alveolar VEGF levels might be important in determining pulmonary edema in ARDS. However, measurements of mouse ELF or plasma VEGF were not documented. In contrast, using an IL-13 transgenic mouse model of hyperoxia-induced lung injury, Corne and coworkers demonstrated a protective role for intrapulmonary VEGF localized to airway epithelial cells, macrophages, and smooth muscle cells (13). In this model, serum levels of VEGF were unchanged in control and hyperoxic animals, that is, the elevation in VEGF was compartmentalized within the lung. These studies initially appear to conflict in terms of VEGF effect, although they both support the lung as a source of VEGF.

In healthy human subjects, VEGF protein levels are highly compartmentalized, with alveolar levels 500 times higher than plasma levels (23). This suggests the lung may be a physiological reservoir of VEGF, but with potentially devastating effects if the epithelial barrier is breached. Our human data show that VEGF levels are lower in the ELF of patients with ARDS than in patients at risk of ARDS, early in the course of the disease. In ARDS there is a widespread but patchy destruction of the alveolar epithelial membrane (24). We speculate that the breached alveolar epithelium leads to the increased plasma levels we have previously observed due to a transepithelial VEGF gradient.

Data on fetal lung and our preliminary observations (25) suggest that the alveolar epithelium is a potential source of VEGF. The correlation between the severity of lung injury and ELF VEGF levels suggests that changes in VEGF levels may reflect the degree of alveolar epithelial damage. Whether this is due to reduced production of VEGF or increased binding is unknown, although preliminary data suggest there may be increased VEGF receptor expression in ARDS lung tissue (26). Our serial bronchoscopy data have demonstrated that ELF VEGF levels increase significantly over the first few days after the development of ARDS. Inevitably, serial bronchoscopy was performed in only 23 of our original 40 subjects. However, this increase in VEGF was greatest in those in whom lung injury was resolving based on improvement in LIS. In this context the alveolar epithelium is recovering and therefore the barrier to VEGF is restored. We hypothesize that this explains the reduction of VEGF in plasma that we have previously observed. Whether the changes in VEGF in ELF are due to reduced binding by Flt-1 or increased production remains to be determined.

Because lung injury stimulates the migration of both neutrophils and macrophages into the lung, and these cell types both produce VEGF, we also considered the possibility that altered VEGF production by inflammatory cells could account for the observed differences in the ELF VEGF between the ARDS group and the control groups.

Alveolar macrophage production of VEGF was significantly different between the groups studied. Normal alveolar macrophages produce little soluble VEGF and are therefore unlikely to contribute much to intrapulmonary VEGF levels, suggesting that they are not the main source in the normal lung. In contrast, the differences between macrophage production of VEGF from ARDS and at-risk subjects reflected the differences in ELF levels, suggesting that they may be a more important source in the injured lung. Interestingly, IL-13 was able to induce VEGF production in macrophages from subjects with ARDS, although the relatively small numbers studied make these data potentially subject to type 2 error. However, overall the data suggest that IL-13 does not determine ELF VEGF levels in lung injury.

The differences in alveolar neutrophil VEGF production between patients with ARDS and at-risk subjects could be explained by tolerance to LPS. LPS tolerance in neutrophils from subjects with ARDS is well documented and is thought to be due to CD14 receptor downregulation (27). The neutrophils were still responsive to IL-8, however, and stimulation of neutrophils from both patients with ARDS and at-risk subjects resulted in increased VEGF production. Because IL-8 is known to be increased in early ARDS, this may be an important trigger for VEGF-mediated alveolar permeability.

Taken together, the inflammatory cell culture data suggest that some of the differences between VEGF levels in ELF from ARDS and at-risk subjects could be accounted for by differential cellular production of VEGF. Inflammatory cell production alone cannot, however, explain the differences between subjects with ARDS and normal subjects. We do not have serial data from alveolar macrophages and neutrophils to determine whether changes in their production of VEGF reflect that found in ELF.

In summary, this study has shown that increasing intrapulmonary VEGF levels are associated with resolution of lung injury in ARDS. Alveolar macrophages and neutrophils are both potential sources of VEGF. VEGF production is responsive to LPS, IL-13, and IL-8, mediators that have previously been implicated in lung injury and repair. It remains to be established whether increased VEGF levels represent a marker of resolution of lung injury, or whether VEGF is actively involved in promoting repair of the alveolar–capillary membrane. We have previously reported that VEGF in the plasma reduces with recovery from ARDS (15). This study shows that the converse occurs in the alveolar space. We suggest that changes in the integrity of the alveolar epithelial barrier may explain these observations.

The authors thank the staff of the Intensive Therapy Unit at Southmead Hospital for their assistance in patient recruitment and Dr. Noeleen Foley for reviewing the manuscript.

1. Pittet JF, Mackersie RC, Martin TR, Matthay MA. Biological markers of acute lung injury: prognostic and pathogenetic significance. Am J Respir Crit Care Med 1997;155:1187–1205.
2. Dobbs LG, Ganzalez R, Matthay MA, Carter EP, Allen L, Verkman AS. Highly water-permeable Type 1 alveolar epithelial cells confer high water permeability between the airspace and vasculature in rat lung. Proc Natl Acad Sci USA 1998;95:2991–2996.
3. Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute lung injury syndrome. N Engl J Med 2000;342:1301–1308.
4. Brenchley PE. VEGF/VPF: a modulator of microvascular function with potential roles in glomerular pathophysiology. J Nephrol 1996;9:10–17.
5. Houck KA, Leung DW, Rowland AM, Winter J, Ferrara N. Dual regulation of vascular endothelial growth factor bioavailability by genetic and proteolytic mechanisms. J Biol Chem 1992;267:26031–26037.
6. Waltenberger J, Claesson-Welsh L, Siegbahn A, Shibuya M, Heldin C. Different signal transduction properties of KDR and Flt-1, two receptors for VEGF. J Biol Chem 1994;269:26988.
7. Barleon B. Migration of human monocytes in response to vascular endothelial growth factor (VEGF) is mediated in the VEGF receptor Flt1. Blood 1996;87:3336–3343.
8. Kendall RL, Thomas KA. Inhibition of vascular endothelial cell growth factor activity by an endogenously encoded soluble receptor. Proc Natl Acad Sci USA 1993;90:10705–10709.
9. Brown KR, England KM, Goss KL, Snyder JM, Accarugei MJ. VEGF induces airway epithelial cell proliferation in human fetal lung in-vitro. Am J Physiol Lung Cell Mol Physiol 2001;281:L1001–L1010.
10. Tsou R, Isik FF. Integrin activation is required for VEGF and FGF receptor protein presence on human microvascular endothelial cells. Mol Cell Biochem 2001;224:81–89.
11. Fehrenbach H, Haase M, Kasper M, Koslowski R, Schuh D, Muller M. Alterations in the immunohistochemical distribution patterns of vascular endothelial growth factor receptors Flk1 and Flt1 in bleomycin-induced rat lung fibrosis. Virchows Arch 1999;435:20–31.
12. Kaner RJ, Ladetto R, Singh N, Fukuda N, Matthay MA, Crystal RG. Lung over expression of the vascular endothelial growth factor gene induces pulmonary edema. Am J Respir Crit Care Med 2000;22:657–664.
13. Corne J, Chupp G, Lee CG, Homer RJ, Zhu Z, Chen Q, Ma B, Du Y, Roux F, McArdle J, et al. IL-13 stimulates vascular endothelial cell growth factor and protects against hyperoxic acute lung injury. J Clin Invest 2000;106:783–791.
14. Compernolle V, Brusselmans K, Acker T, Hoet P, Tjwa M, Beck H, Plaisance S, Dor Y, Lupu F, Nemery B, et al. Loss of HIF-2α and inhibition of VEGF impair fetal lung maturation, whereas treatment with VEGF prevents fatal respiratory distress in premature mice. Nat Med 2002;7:702–710.
15. 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 2001;164:657–664.
16. Maitre B, Boussat S, Jean D, Gouge M, Brochard L, Housset B, Adnot S, Delclaux C. Vascular endothelial growth factor synthesis in the acute phase of experimental and acute lung injury. Eur Respir J 2001;18:100–106.
17. Li J, Perrella MA, Tsai JC, Yet SF, Hsieh CM, Yoshizumi M, Patterson C, Endege WO, Zhou F, Lee ME. Induction of vascular endothelial growth factor gene expression by interleukin-1β in rat aortic smooth muscle cells. J Biol Chem 1995;270:308–312.
18. Ryuto M, Ono M, Izumi S, Yoshida HA, Weich K, Kohno K, Kuwano M. Induction of vascular endothelial growth factor gene by in tumor necrosis factor α in human glioma cells: possible role of SP-1. J Biol Chem 1996;271:28220–28228.
19. American–European Consensus Committee on ARDS. Report to the American–European Consensus Conference on ARDS: definition, mechanisms, relevant outcome and clinical trial co-ordination. Barcelona, Spain. Am J Respir Crit Care Med 1994;149:818–824.
20. Ferrara N, Carver-Moore K, Chen H, Dowd M, Lu L, Powell-Braxton L, Hillan KJ, Moore MW. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 1996;380:439–442.
21. Carmeliet P, Ferreira V, Breier G, Pollefeyt S, Kieckens L, Gertsenstein M, Fahrig M, Vandenhoeck A, Harpal K, Eberhardt C, et al. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 1996;380:435–439.
22. Shifren JL, Doldi N, Ferrara N, Mesiano S, Jaffe RB. In the human fetus, vascular endothelial growth factor is expressed in epithelial cells and myocytes, but not vascular endothelium: implications for mode of action. J Clin Endocrinol Metab 1994;79:316–322.
23. Kaner RJ, Crystal RG. Compartmentalisation of vascular endothelial cell growth factor to the epithelial surface of the human lung. Mol Med 2001;7:240–246.
24. Lamy M, Fallat RT, Koeniger E, Dietrich HP, Eberhart RC, Tucker HJ, Hill JD. Pathologic features and mechanisms of hypoxemia in adult respiratory distress syndrome. Am Rev Respir Dis 1987;135:482.
25. Armstrong L, Medford ARL, Thorley A, Tetley TD, Millar AB. Primary human alveolar epithelial type II cells constitutively produce vascular endothelial growth factor. Am J Respir Crit Care Med 2002;165:A372.
26. Medford ARL, Kendall H, Armstrong L, Millar AB. Expression of vascular endothelial growth factor and neuropilin receptors in normal lung and acute respiratory distress syndrome. Am J Respir Crit Care Med 2002;165:A474.
27. Parsons PE, Gillespie MM, Moore EE, Moore FA, Worthen GS. Neutrophil response to endotoxin in the adult respiratory distress syndrome: role of CD14. Am J Respir Cell Mol Biol 1995;13:152–160.
Correspondence and requests for reprints should be addressed to A. B. Millar, M.D., Lung Research Group, University of Bristol Medical School Unit, Southmead Hospital, Westbury on Trym, Bristol BS10 5NB, UK. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record