Elevated plasma t-PA (tissue plasminogen activator) and serum CRP (C-reactive protein) concentrations are associated with an adverse cardiovascular risk. In the present study, we investigated whether acute local inflammation causes vascular dysfunction and influences t-PA release in patients with stable coronary heart disease. Serum CRP, plasma t-PA and PAI-1 (plasminogen activator inhibitor type 1) concentrations were determined in 95 patients with stable coronary heart disease. A representative subpopulation of 12 male patients received an intra-brachial infusion of TNF-α (tumour necrosis factor-α) and saline placebo using a randomized double-blind cross-over study design. Forearm blood flow and plasma fibrinolytic and inflammatory variables were measured. Serum CRP concentrations correlated with plasma t-PA concentrations (r=0.37, P<0.001) and t-PA/PAI-1 ratio (r=−0.21, P<0.05). Intra-arterial TNF-α caused a rise in t-PA concentrations (P<0.001) without affecting blood flow or PAI-1 concentrations. TNF-α pretreatment impaired acetylcholine- and sodium nitroprusside-induced vasodilatation (P<0.001 for both) whilst doubling bradykinin-induced t-PA release (P=0.006). In patients with stable coronary heart disease, plasma fibrinolytic factors correlate with a systemic inflammatory marker and local vascular inflammation directly impairs vasomotor function whilst enhancing endothelial t-PA release. We suggest that the adverse prognosis associated with elevated plasma t-PA concentrations relates to the underlying causative association with vascular inflammation and injury.

INTRODUCTION

In epidemiologic studies of patients with CHD (coronary heart disease) [1] and in prospective studies in healthy populations [2], higher plasma concentrations of the pro-fibrinolytic factor t-PA (tissue plasminogen activator) positively and independently predict future cardiovascular events. It would be anticipated that high t-PA concentrations would protect against subsequent cardiovascular events rather than the reverse. This paradoxical association is, in part, explained by the concomitant elevation of PAI-1 (plasminogen activator inhibitor type 1) which complexes with, and inactivates, t-PA. However, the precise stimulus for this increased t-PA release remains unclear.

Areas of endothelial denudation and thrombus deposition are a common finding on the surface of atheromatous plaques and are usually subclinical. Through t-PA release, endogenous fibrinolysis is usually able to prevent thrombus propagation, although organization of the residual thrombus may lead to plaque growth and expansion [3]. The adverse prognosis conferred by elevated plasma t-PA antigen concentrations may, therefore, reflect the extent of occult atheroma and subclinical plaque rupture stimulating t-PA release.

Markers of systemic inflammation, such as CRP (C-reactive protein) and TNF-α (tumour necrosis factor-α), are elevated in patients with cardiovascular disease [4,5]. Indeed, serum CRP concentrations predict the development of cardiovascular disease independently of other risk factors. Previous studies have indicated a direct relationship between serum CRP and plasma t-PA concentrations [6,7]. This raises the question of whether vascular inflammation is causally related to the elevation in plasma t-PA concentrations or whether CRP and t-PA are independently increased by a common factor related to the atherosclerotic process itself, such as acute plaque rupture.

Abnormalities of endothelial function have been demonstrated in patients with atherosclerosis [8] and vascular inflammation [9,10]. In patients with CHD, restoration of endothelium-dependent vasomotor function occurs when there is normalization of CRP concentrations [11], whereas ongoing chronic inflammation is associated with an impaired fibrinolytic response to venous occlusion [12]. As the endothelium is the major source of plasma t-PA, abnormalities of endothelial function may therefore mediate the potential inflammation-induced elevations in plasma t-PA concentrations.

It therefore remains unclear whether elevated t-PA concentrations are implicated in the mechanisms contributing to, or arise as a consequence of, atherothrombotic events.

The aims of the present study were, in patients with stable CHD, to confirm the previous association between plasma CRP and t-PA concentrations and to determine the effect of acute local vascular inflammation provoked by direct intra-arterial infusion of TNF-α on vasomotor function and endothelial t-PA release.

METHODS

Patients

We recruited patients with CHD confirmed by angiography (defined as >70% luminal stenosis of at least one major epicardial coronary vessel) or a previous history of Q-wave myocardial infarction. All patients had stable anginal symptoms and had not undergone coronary revascularization within the preceding 3 months. Exclusion criteria were significant cardiac failure, renal impairment, SBP [systolic BP (blood pressure)] <100 or >190 mmHg, diabetes mellitus, history or clinical features of recent infective illness and immuno-suppressive or non-steroidal anti-inflammatory medication (excluding 75 mg/day aspirin). All studies were undertaken with the approval of the local Research Ethics Committee and in accordance with the Declaration of Helsinki. Written informed consent was obtained from each subject.

Venous sampling and assays

Plasma t-PA, PAI-1 (Coaliza®; Chromogenix), prothrombin F1+2 (fragment 1 and 2; Enzygnost F1+2; Dade Behring), TNF-α (Quantikine; R&D Systems) and IL-6 (interleukin-6; Dako) concentrations were determined using ELISAs, and t-PA activity using a photometric method (Coatest t-PA; Chomogenix) [1315]. Assays of hs-CRP (highly sensitive CRP) were undertaken using the method of particle-enhanced immunonephelometry (BN II nephelometer; Behring). Venous blood was collected into tubes containing acidified buffered citrate (for t-PA), trisodium citrate (for PAI-1 and prothrombin F1+2), potassium EDTA (for cytokines) and serum gel tubes (for CRP). Platelet-free plasma and serum were stored at −80 °C before assay. Haematocrit and white cell count were determined using an automated Coulter counter. Biochemical assays were undertaken on fasting venous samples by the hospital Clinical Laboratory facility.

Drugs

TNF-α (Knoll Pharmaceuticals), bradykinin (Clinalfa), acetylcholine (Novartis) and SNP (sodium nitroprusside; David Bull Laboratories) were administered following dissolution in 0.9% saline.

Study design

All subjects abstained from alcohol for 24 h and from food, tobacco and caffeine-containing drinks for at least 4 h before each study visit. A venous blood sample was taken from all patients for estimation of serum CRP and plasma t-PA and PAI-1 antigen concentrations. Twelve representative male patients were then recruited into a randomized double-blind placebo-controlled cross-over study comparing the effect of direct intra-brachial infusion of TNF-α and saline placebo.

Forearm study protocol

All studies were carried out at 09:00 hours in a quiet temperature-controlled room maintained at 22–25 °C. Patients rested recumbent and strain gauges and cuffs were applied. A 17-gauge venous cannula was inserted into the antecubital vein of each arm, and the brachial artery of the non-dominant arm was cannulated with a 27-SWG (standard wire gauge) needle (Cooper's Needle Works). FBF (forearm blood flow) was measured in both forearms by venous occlusion plethysmography using mercury in silastic strain gauges as described previously [16]. BP and HR (heart rate) were measured using a semi-automated non-invasive sphygmomanometer (Omron 705 IT).

Subjects (n=12) attended on two occasions at least two weeks apart and received an intra-arterial infusion of either TNF-α (80 ng/min) or saline placebo over 60 min, followed on each occasion by a further 60 min of saline infusion. Thereafter intra-arterial bradykinin (100, 300 and 1000 pmol/min), acetylcholine (5, 10 and 20 μg/min) and SNP (2, 4 and 8 μg/min) were administered at 1 ml/min with a 15 min saline washout period between each agent. The dose of TNF-α was chosen to achieve local cytokine concentrations comparable with healthy volunteer studies [13] and those seen in cardiovascular disease [17].

Venous blood samples for t-PA antigen and activity and PAI-1 antigen were obtained at baseline, after 60 min of TNF-α/placebo infusion, following the 60 min infusion of saline, before and during each dose of bradykinin and 15 min after the end of bradykinin infusion. Plasma cytokines, hs-CRP and prothrombin F1+2 were assessed prior to and following the TNF-α/placebo and saline infusions and at the end of each study.

Statistical analysis

Estimated net release of t-PA was defined previously as the product of the infused forearm plasma flow and the concentration difference between the infused and non-infused forearms [16]. Because basal t-PA concentrations were altered by pretreatment with TNF-α, net release of t-PA during bradykinin infusion was calculated by subtracting the mean t-PA release before and 15 min after cessation of bradykinin infusion [13]. The area under the curve was calculated for the estimated net release of t-PA in response to bradykinin. Data were examined, where appropriate, by ANOVA with repeated measures, followed by post-hoc Student's t tests adjusted with a Bonferroni correction for multiple comparisons. Spearman's correlation was used to compare CRP and plasma levels of fibrinolytic factors. As serum CRP concentrations have a skewed distribution, they were logarithmically transformed.

All statistical calculations were undertaken using GraphPad Prism (GraphPad Software). Results are expressed as means±S.E.M., unless otherwise stated, and statistical significance was assigned at the 5% level.

RESULTS

Patients had a typical cardiovascular risk factor profile in keeping with their diagnosis of CHD and most were prescribed secondary preventative medications (Tables 1 and 2). The majority of the subjects were male, two-thirds had previously undergone coronary revascularization and approx. one-third were habitual smokers.

Table 1
Baseline characteristics of the 95 subjects with stable CHD

Values are means±S.D. or n (%). ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; LDL, low-density lipoprotein. *Angiographic data were unavailable on one subject.

Characteristics
Age (years) 59±7 
Body mass index (kg/m229±5 
Male gender (n79 (83%) 
Previous myocardial infarction (n39 (41%) 
Extent of coronary artery disease (n)*  
 One vessel 42 (44%) 
 Two vessels 24 (25%) 
 Three vessels 28 (29%) 
Previous coronary revascularization (n64 (67%) 
Co-morbidity (n 
 Hypertension 46 (48%) 
 Previous hyperlipidaemia 89 (94%) 
 Family history of premature CHD 32 (34%) 
 Smoker/ex-smoker/non-smoker 30/27/38 (32%/28%/40%) 
Medical therapy (n 
 Aspirin 95 (100%) 
 Anti-anginal 82 (86%) 
 Statin 90 (95%) 
 ACE-inhibitor/ARB 30 (32%) 
Serum urea (mmol/l) 5.6±1.3 
Serum creatinine (μmol/l) 93±13 
Plasma glucose (mmol/l) 5.7±0.9 
Total cholesterol (mmol/l) 4.5±0.9 
LDL-cholesterol (mmol/l) 2.5±0.7 
Triacylglycerols (mmol/l) 1.8±0.9 
Characteristics
Age (years) 59±7 
Body mass index (kg/m229±5 
Male gender (n79 (83%) 
Previous myocardial infarction (n39 (41%) 
Extent of coronary artery disease (n)*  
 One vessel 42 (44%) 
 Two vessels 24 (25%) 
 Three vessels 28 (29%) 
Previous coronary revascularization (n64 (67%) 
Co-morbidity (n 
 Hypertension 46 (48%) 
 Previous hyperlipidaemia 89 (94%) 
 Family history of premature CHD 32 (34%) 
 Smoker/ex-smoker/non-smoker 30/27/38 (32%/28%/40%) 
Medical therapy (n 
 Aspirin 95 (100%) 
 Anti-anginal 82 (86%) 
 Statin 90 (95%) 
 ACE-inhibitor/ARB 30 (32%) 
Serum urea (mmol/l) 5.6±1.3 
Serum creatinine (μmol/l) 93±13 
Plasma glucose (mmol/l) 5.7±0.9 
Total cholesterol (mmol/l) 4.5±0.9 
LDL-cholesterol (mmol/l) 2.5±0.7 
Triacylglycerols (mmol/l) 1.8±0.9 
Table 2
Baseline characteristics of 12 patients receiving TNF-α and saline placebo in a randomized double-blind cross-over study design

Values are means±S.E.M or n (%). ACE, angiotensin-converting enzyme; LDL, low-density lipoprotein. *Angiographic data were unavailable on one subject.

Characteristics
Age (years) 60±2 
Body mass index (kg/m224±1 
Male gender (n12 (100%) 
Previous myocardial infarction (n5 (42%) 
Extent of coronary disease (n)*  
 One vessel 5 (42%) 
 Two vessels 4 (33%) 
 Three vessels 2 (17%) 
Previous coronary revascularization (n9 (75%) 
Co-morbidity (n 
 Hypertension 3 (25%) 
 Family history of premature CHD 1 (8%) 
 Previous hyperlipidaemia 12 (100%) 
 Smoker/ex-smoker/non-smoker 2/7/3 (17%/58%/25%) 
Medical therapy (n 
 Aspirin 12 (100%) 
 Anti-anginal 11 (92%) 
 Statin 12 (100%) 
 ACE-inhibitor 1 (8%) 
Serum urea (mmol/l) 5.9±0.3 
Serum creatinine (μmol/l) 93±2 
Plasma glucose (mmol/l) 5.2±0.1 
Total cholesterol (mmol/l) 3.9±0.2 
LDL-cholesterol (mmol/l) 2.1±0.2 
Triacylglycerols (mmol/l) 1.3±0.2 
Placebo visit  
 HR (beats/min) 56±2 
 SBP (mmHg) 145±6 
 Diastolic BP (mmHg) 83±4 
 FBF (ml·100 ml−1 of tissue·min−1 
  Infused arm 2.1±0.2 
  Non-infused arm 2.0±0.2 
TNF-α visit  
 HR (beats/min) 56±3 
 SBP (mmHg) 142±4 
 Diastolic BP (mmHg) 81±2 
 FBF (ml·100 ml−1 of tissue·min−1 
  Infused arm 2.0±0.2 
  Non-infused arm 2.1±0.2 
Characteristics
Age (years) 60±2 
Body mass index (kg/m224±1 
Male gender (n12 (100%) 
Previous myocardial infarction (n5 (42%) 
Extent of coronary disease (n)*  
 One vessel 5 (42%) 
 Two vessels 4 (33%) 
 Three vessels 2 (17%) 
Previous coronary revascularization (n9 (75%) 
Co-morbidity (n 
 Hypertension 3 (25%) 
 Family history of premature CHD 1 (8%) 
 Previous hyperlipidaemia 12 (100%) 
 Smoker/ex-smoker/non-smoker 2/7/3 (17%/58%/25%) 
Medical therapy (n 
 Aspirin 12 (100%) 
 Anti-anginal 11 (92%) 
 Statin 12 (100%) 
 ACE-inhibitor 1 (8%) 
Serum urea (mmol/l) 5.9±0.3 
Serum creatinine (μmol/l) 93±2 
Plasma glucose (mmol/l) 5.2±0.1 
Total cholesterol (mmol/l) 3.9±0.2 
LDL-cholesterol (mmol/l) 2.1±0.2 
Triacylglycerols (mmol/l) 1.3±0.2 
Placebo visit  
 HR (beats/min) 56±2 
 SBP (mmHg) 145±6 
 Diastolic BP (mmHg) 83±4 
 FBF (ml·100 ml−1 of tissue·min−1 
  Infused arm 2.1±0.2 
  Non-infused arm 2.0±0.2 
TNF-α visit  
 HR (beats/min) 56±3 
 SBP (mmHg) 142±4 
 Diastolic BP (mmHg) 81±2 
 FBF (ml·100 ml−1 of tissue·min−1 
  Infused arm 2.0±0.2 
  Non-infused arm 2.1±0.2 

Plasma fibrinolytic factors and correlation with CRP

In the cohort of 95 patients with stable CHD, serum hs-CRP concentrations correlated with plasma t-PA antigen concentrations (r=0.37, P<0.001) and plasma PAI-1 antigen concentrations (r=0.28, P=0.006) and inversely with the ratio of t-PA/PAI-1 antigen (r=−0.21, P<0.05). As anticipated, plasma t-PA antigen concentrations correlated with plasma PAI-1 antigen concentrations (r=0.49, P<0.001).

Effect of acute inflammation on vascular and fibrinolytic function

Those subjects who received TNF-α and saline placebo had similar baseline characteristics to the main cohort and all were receiving statin and aspirin therapy. There were no differences in resting arterial pressure, HR or FBF between the two study visits (Table 2).

There was no change in HR or BP during infusion of either TNF-α or placebo. Haematocrit, temperature, white cell count and hs-CRP were similar on both study visits [P=ns (not significant); results not shown]. The study protocol was well tolerated with no major adverse effects.

Plasma cytokine concentrations

Intra-arterial infusion of TNF-α increased plasma TNF-α concentrations from 1.4±0.2 to 164.5±26.8 pg/ml in the infused arm and from 1.3±0.1 to 33.3±5.6 pg/ml in the non-infused arm (P<0.001; Figure 1). At 1 h after TNF-α infusion, plasma IL-6 concentrations increased from 4.1±1.9 to 6.4±2.3 pg/ml in the infused arm (P<0.001, as determined by ANOVA), but were unchanged in the non-infused arm (4.0±1.9 to 3.7±1.3 pg/ml). Placebo infusion had no effect on plasma IL-6 concentrations in the infused arm (3.9±1.7 to 4.3±1.3 pg/ml; P=ns). Serum CRP concentrations were unchanged following the 60 min infusion of TNF-α and saline placebo.

Plasma TNF-α concentrations in the infused (solid line) and non-infused arm (dashed/dotted line) after 60 min of intra-arterial TNF-α (●) and saline placebo (○)

Figure 1
Plasma TNF-α concentrations in the infused (solid line) and non-infused arm (dashed/dotted line) after 60 min of intra-arterial TNF-α (●) and saline placebo (○)

P<0.001 for the TNF-α dose–response in the infused arm, as determined by ANOVA; P<0.001 when TNF-α compared with saline placebo in the infused arm, as determined by ANOVA.

Figure 1
Plasma TNF-α concentrations in the infused (solid line) and non-infused arm (dashed/dotted line) after 60 min of intra-arterial TNF-α (●) and saline placebo (○)

P<0.001 for the TNF-α dose–response in the infused arm, as determined by ANOVA; P<0.001 when TNF-α compared with saline placebo in the infused arm, as determined by ANOVA.

Vasomotor and fibrinolytic responses

There was no significant change in resting FBF in the 2 h after the start of either TNF-α or placebo infusion. After 60 min of TNF-α, plasma t-PA antigen and activity concentrations in the infused arm had increased from 9.4±1.1 to 11.3±1.2 ng/ml and 0.3±0.1 to 2.1±0.6 IU (international units)/ml respectively (P<0.001), and these remained elevated in the 2 h following discontinuation of the TNF-α infusion (Figure 2). There was no change in plasma PAI-1 antigen concentrations in the infused arm and no change in either plasma t-PA or PAI-1 concentrations in the non-infused arm during the study (results not shown). Prothrombin F1+2 concentrations in the infused arm were unaltered following either saline placebo (0.9±0.1 to 0.9±0.1 ng/ml; P=ns) or TNF-α infusion (0.9±0.1 to 1.0±0.1 ng/ml; P=ns).

Plasma t-PA activity (upper panel) and antigen (lower panel) concentrations in the infused (●) and non-infused (○) arms after 60 min infusion of intra-arterial TNF-α

Figure 2
Plasma t-PA activity (upper panel) and antigen (lower panel) concentrations in the infused (●) and non-infused (○) arms after 60 min infusion of intra-arterial TNF-α

P<0.001 when infused compared with non-infused arms, as determined by ANOVA. *P<0.05 and ***P<0.001, as determined by post-hoc Student's t test for treatment effect.

Figure 2
Plasma t-PA activity (upper panel) and antigen (lower panel) concentrations in the infused (●) and non-infused (○) arms after 60 min infusion of intra-arterial TNF-α

P<0.001 when infused compared with non-infused arms, as determined by ANOVA. *P<0.05 and ***P<0.001, as determined by post-hoc Student's t test for treatment effect.

There was a dose-dependent increase in FBF during bradykinin, acetylcholine and SNP infusion (P≤0.01, as determined by ANOVA). Compared with the saline placebo, TNF-α pretreatment impaired acetylcholine- and SNP-induced vasodilatation (P<0.001 for both), but did not alter the response to bradykinin (Figure 3).

Infused (solid line) and non-infused (dashed/dotted line) FBF during incremental doses of acetylcholine (left-hand panel), SNP (middle panel) and bradykinin (right-hand panel) following pretreatment with TNF-α (●) or saline placebo (○)

Figure 3
Infused (solid line) and non-infused (dashed/dotted line) FBF during incremental doses of acetylcholine (left-hand panel), SNP (middle panel) and bradykinin (right-hand panel) following pretreatment with TNF-α (●) or saline placebo (○)

P≤0.01 for all infused arm responses, as determined by ANOVA. †P<0.001 when TNF-α compared with saline placebo treatments, as determined by ANOVA. *P<0.05, as determined by post-hoc Student's t test for treatment effect.

Figure 3
Infused (solid line) and non-infused (dashed/dotted line) FBF during incremental doses of acetylcholine (left-hand panel), SNP (middle panel) and bradykinin (right-hand panel) following pretreatment with TNF-α (●) or saline placebo (○)

P≤0.01 for all infused arm responses, as determined by ANOVA. †P<0.001 when TNF-α compared with saline placebo treatments, as determined by ANOVA. *P<0.05, as determined by post-hoc Student's t test for treatment effect.

Plasma t-PA concentrations increased in a dose-dependent manner during bradykinin infusion on both study visits (P<0.001). Pretreatment with TNF-α augmented the bradykinin-induced rise in plasma t-PA antigen and activity concentrations (P<0.001 for both; Figure 4) and significantly increased estimated net release of t-PA antigen (63.7±14.8 compared with 120.6±26.1 ng·100 ml−1 of tissue·min−1 at peak dose; P<0.05) and activity (54.8±14.8 compared with 98.8±21.0 IU·100 ml−1 of tissue·min−1 at peak dose; P<0.05). Over the 30 min period of bradykinin infusion, TNF-α increased the area under the curve for net t-PA antigen and activity release by 120% and 188% respectively (P=0.006).

Plasma concentrations of t-PA antigen (left-hand panel) and t-PA activity (right-hand panel) during bradykinin infusion following pretreatment with TNF-α (●) or saline placebo (○)

Figure 4
Plasma concentrations of t-PA antigen (left-hand panel) and t-PA activity (right-hand panel) during bradykinin infusion following pretreatment with TNF-α (●) or saline placebo (○)

P<0.001 when TNF-α compared with saline placebo treatments, as determined by ANOVA. *P<0.05 and **P<0.01, as determined by post-hoc Student's t test for treatment effect.

Figure 4
Plasma concentrations of t-PA antigen (left-hand panel) and t-PA activity (right-hand panel) during bradykinin infusion following pretreatment with TNF-α (●) or saline placebo (○)

P<0.001 when TNF-α compared with saline placebo treatments, as determined by ANOVA. *P<0.05 and **P<0.01, as determined by post-hoc Student's t test for treatment effect.

Subgroup analysis showed a significant impairment of endothelium-dependent vasodilatation as well as bradykinin-induced t-PA response in cigarette smokers. Qualitatively the effect of intra-arterial TNF-α on the blood flow and fibrinolytic responses was similar in both smokers and non-smokers.

DISCUSSION

In the present study, we have confirmed the direct association between plasma t-PA and serum CRP concentrations in patients with stable CHD. For the first time, we have extended this observation using an acute local vascular inflammatory model and demonstrated that direct intra-arterial infusion of TNF-α causes a slow onset and sustained increase in basal t-PA release. This arterial inflammation was also associated with increased stimulated t-PA release in the presence of impaired vasomotor function. Our findings are consistent with the suggestion that t-PA is released during vascular inflammation and endothelial injury and this may, in part, explain the adverse prognosis associated with increased plasma t-PA concentrations.

The link between markers of inflammation and plasma t-PA concentrations suggests that vascular inflammation and injury may be responsible for endothelial t-PA release. However, this association may arise from common aetiological factors and does not establish a causal relationship. A recent meta-analysis has suggested that increases in plasma t-PA concentrations largely reflect the presence of concomitant cardiovascular risk factors [7]. We therefore sought to induce vascular inflammation in a representative sample of our study population. Inducing systemic inflammation will have many biological actions and could be confounded by indirect or extra-vascular effects. This may explain some of the differences between our present results and those of previous work in healthy volunteers using Salmonella typhi vaccination [9,10]. We chose to employ an acute local inflammatory model [13] to assess the direct effects of intra-arterial TNF-α administration on t-PA release. We were able to induce local vascular inflammation with a rise in local plasma IL-6 concentrations to levels comparable with those seen in patients with unstable angina [17], without evidence of a systemic inflammatory response or change in plasma t-PA or PAI-1 concentrations in the non-infused arm. This local vascular inflammation caused a slow onset and sustained increase in plasma t-PA concentrations that continued for at least 2 h after cessation of the TNF-α infusion. This establishes that vascular inflammation directly causes endothelial t-PA release in humans.

The mechanism of TNF-α-induced t-PA release has not been established. Inflammatory cytokines, such as TNF-α, may cause t-PA release via induction of local thrombus formation, activation of specific cellular receptors or through generation of secondary mediators within the local vasculature. The former seems unlikely given that we observed no increase in prothrombin F1+2, a sensitive marker of in vivo thrombin generation [15].

Although it is likely that plasma t-PA concentrations are increased by inducing endothelial injury, smooth muscle cells, macrophages and monocytes also express t-PA mRNA following stimulation by inflammatory cytokines within atherosclerotic plaques [18] and could theoretically contribute to this fibrinolytic response.

Effects of TNF-α on vasomotor function

Impaired vasodilator responses to acetylcholine [19,20] and SNP [20] in patients with cardiovascular risk factors predict an increased risk of adverse cardiovascular events. Previous studies in healthy volunteers have shown acute systemic inflammation is associated with a transient impairment in vasomotor function [9]. In the present study, we have now shown that, in patients with CHD, acute local vascular inflammation decreases the vasomotor response to both acetylcholine and SNP. Acetylcholine is known to stimulate NO (nitric oxide) production via activation of eNOS (endothelial NO synthase) and, together with the impaired SNP response, these findings suggest that local arterial inflammation can decrease NO bioavailability. Interestingly, impaired vasodilator responses to both acetylcholine and SNP have been correlated with plasma TNF-α concentrations in patients with rheumatoid arthritis, a chronic inflammatory condition which is itself associated with an excess cardiovascular risk [21]. Furthermore, intra-arterial infusion of the free radical scavenger vitamin C restores forearm blood flow responses in patients with CHD and elevated serum levels of CRP [22].

Our results suggest that acute inflammation is associated with impaired NO-dependent smooth muscle relaxation in response to direct NO donors such as SNP or endogenously derived NO following stimulation of eNOS. Several lines of evidence support the contention that cytokines, such as TNF-α, may impair NO-dependent signalling. TNF-α decreases eNOS expression [23] as well as increasing reactive oxygen species, such as superoxide anion, that rapidly inactivate NO and are directly cytotoxic to vascular tissues. TNF-α also increases plasma ADMA (asymmetric dimethylarginine), an endogenous inhibitor of eNOS that inhibits endothelium-dependent vasodilatation [24]. Finally, inflammatory states may increase iNOS (inducible NO synthase) expression which is associated with receptor uncoupling and endothelial dysfunction [25].

Bradykinin-induced vasodilatation was unaltered by pretreatment with TNF-α. Previous work has suggested that NO contributes only a small proportion (approx. 15%) to bradykinin-induced vasodilatation [26,27] and does not contribute to the mechanism of bradykinin-induced t-PA release [28]. Moreover, in patients with vascular dysfunction, there may be an increased contribution of EDHF (endothelium-derived hyperpolarizing factor) to smooth muscle vasorelaxation [29], particularly if NO is consumed by free radicals generated by locally active inflammatory cells. These observations are consistent with our present findings of impaired acetylcholine- and SNP-induced vasodilatation, but preserved bradykinin responses.

Effects of TNF-α on acute endogenous fibrinolysis

Despite the presence of higher baseline plasma t-PA concentrations in patients with atherosclerosis [7], we have shown in the present study that direct intra-arterial TNF-α infusion increases bradykinin-induced t-PA release in patients with stable CHD. Thus, although vasorelaxation was impaired, acute inflammation initiates a sustained increase in both basal and stimulated t-PA release. The mechanism of this effect is unknown, but may involve direct endothelial injury, up-regulation of t-PA synthesis or alterations in bradykinin receptor expression.

Under some circumstances, increases in t-PA may protect against the propagation of intravascular thrombosis and thereby avoid the development of an acute coronary syndrome. However, elevations in plasma t-PA concentrations may reflect more widespread endothelial dysfunction and a dominant pro-inflammatory vascular response that may overwhelm any locally protective pro-fibrinolytic effect. Indeed, the pro-fibrinolytic actions of vascular inflammation may potentiate degradation of extracellular matrix and aggravate plaque instability [30]. The clinical outcome of acute vascular inflammation may, therefore, depend upon the relative balance between the protective antithrombotic actions and potential plaque destabilization associated with increased vascular t-PA release.

In the present study, all patients who received TNF-α and placebo infusion were already receiving secondary preventative therapy, including aspirin and lipid-lowering medications. Although these may have influenced the vascular response to TNF-α, it was considered unethical to withhold these and, in clinical practice, a large proportion of patients presenting with acute coronary syndromes and raised inflammatory markers are already established on such therapies. As our study design was focused on the question of the link between vascular inflammation and t-PA release in patients with CHD, we did not include a control population of healthy subjects. However, we have shown previously [13] that intra-arterial TNF-α enhances endothelium-dependent t-PA release by a similar degree in younger healthy volunteers. Although we have again demonstrated that smokers have impaired endothelial responses, including t-PA release [14], intra-arterial TNF-α increased plasma t-PA concentrations to a similar degree in both smokers and non-smokers with established CHD.

We have found that, although TNF-α adversely affects NO-dependent vasodilatation, it enhances other protective mechanisms, such as the endogenous fibrinolytic capacity. This reflects the complex and pleiotropic nature of TNF-α which functions as part of the normal host surveillance mechanisms and response to tissue injury. Although we only determined the effect of acute vascular inflammation in 12 patients, our results may explain some of the contradictory findings of previous clinical studies. For example, in patients with heart failure, TNF-α antagonism causes marked improvements in endothelium-dependent vasodilatation [31], but has failed to demonstrate clinical benefit in randomized controlled trials [32]. Thus the benefits of restoring endothelium-dependent vasomotor function by TNF-α antagonism may be counterbalanced by inhibiting other potentially beneficial acute effects, such as enhancing endogenous t-PA release.

Conclusions

In the present study, we have shown that, in patients with stable CHD, plasma fibrinolytic factors are correlated with CRP, a sensitive and prognostically relevant marker of vascular inflammation. We have also demonstrated that acute vascular inflammation directly impairs vasomotor function whilst enhancing endothelial t-PA release. We suggest that the adverse prognosis associated with elevated plasma t-PA concentrations reflects a causative association with vascular inflammation and injury, rather than representing a marker of occult plaque rupture.

Abbreviations

     
  • BP

    blood pressure

  •  
  • CHD

    coronary heart disease

  •  
  • CRP

    C-reactive protein

  •  
  • F1+2

    fragment 1 and 2

  •  
  • FBF

    forearm blood flow

  •  
  • HR

    heart rate

  •  
  • hs-CRP

    highly sensitive CRP

  •  
  • IL-6

    interleukin-6

  •  
  • IU

    international units

  •  
  • NO

    nitric oxide

  •  
  • eNOS

    endothelial NO synthase

  •  
  • ns

    not significant

  •  
  • PAI-1

    plasminogen activator inhibitor type 1

  •  
  • SBP

    systolic blood pressure

  •  
  • SNP

    sodium nitroprusside

  •  
  • TNF-α

    tumour necrosis factor-α

  •  
  • t-PA

    tissue plasminogen activator

S.D.R. is the recipient of a British Heart Foundation (BHF) Junior Research Fellowship (FS/2001/047), and this work was supported further by a BHF project grant (PG/2001068). We would like to thank Frances Paterson, the Clinical Research Facility at the Royal Infirmary of Edinburgh and the Wellcome Trust Clinical Research Facility at the Western General Hospital for their assistance with this study.

References

References
1
Thompson
 
S. G.
Kienast
 
J.
Pyke
 
S. D.
Haverkate
 
F.
van de Loo
 
J. C.
 
Hemostatic factors and the risk of myocardial infarction or sudden death in patients with angina pectoris. European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group
N. Engl. J. Med.
1995
, vol. 
332
 (pg. 
635
-
641
)
2
Ridker
 
P. M.
Vaughan
 
D. E.
Stampfer
 
M. J.
Manson
 
J. E.
Hennekens
 
C. H.
 
Endogenous tissue-type plasminogen activator and risk of myocardial infarction
Lancet
1993
, vol. 
341
 (pg. 
1165
-
1168
)
3
Mann
 
J.
Davies
 
M. J.
 
Mechanisms of progression in native coronary artery disease: role of healed plaque disruption
Heart
1999
, vol. 
82
 (pg. 
265
-
268
)
4
Haverkate
 
F.
Thompson
 
S. G.
Pyke
 
S. D.
Gallimore
 
J. R.
Pepys
 
M. B.
 
Production of C-reactive protein and risk of coronary events in stable and unstable angina. European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group
Lancet
1997
, vol. 
349
 (pg. 
462
-
466
)
5
Ridker
 
P. M.
Rifai
 
N.
Pfeffer
 
M.
Sacks
 
F.
Lepage
 
S.
Braunwald
 
E.
 
Elevation of tumor necrosis factor-α and increased risk of recurrent coronary events after myocardial infarction
Circulation
2000
, vol. 
101
 (pg. 
2149
-
2153
)
6
Haverkate
 
F.
Thompson
 
S. G.
Duckert
 
F.
 
Haemostasis factors in angina pectoris; relation to gender, age and acute-phase reaction. Results of the ECAT Angina Pectoris Study Group
Thromb. Haemostasis
1995
, vol. 
73
 (pg. 
561
-
567
)
7
Lowe
 
G. D.
Danesh
 
J.
Lewington
 
S.
, et al 
Tissue plasminogen activator antigen and coronary heart disease: prospective study and meta-analysis
Eur. Heart J.
2004
, vol. 
25
 (pg. 
252
-
259
)
8
Ludmer
 
P. L.
Selwyn
 
A. P.
Shook
 
T. L.
, et al 
Paradoxical vasoconstriction induced by acetylcholine in atherosclerotic coronary arteries
N. Engl. J. Med.
1986
, vol. 
315
 (pg. 
1046
-
1051
)
9
Hingorani
 
A. D.
Cross
 
J.
Kharbanda
 
R. K.
, et al 
Acute systemic inflammation impairs endothelium-dependent dilatation in humans
Circulation
2000
, vol. 
102
 (pg. 
994
-
999
)
10
Chia
 
S.
Ludlam
 
C. A.
Fox
 
K. A.
Newby
 
D. E.
 
Acute systemic inflammation enhances endothelium-dependent tissue plasminogen activator release in men
J. Am. Coll. Cardiol.
2003
, vol. 
41
 (pg. 
333
-
339
)
11
Fichtlscherer
 
S.
Rosenberger
 
G.
Walter
 
D. H.
Breuer
 
S.
Dimmeler
 
S.
Zeiher
 
A. M.
 
Elevated C-reactive protein levels and impaired endothelial vasoreactivity in patients with coronary artery disease
Circulation
2000
, vol. 
102
 (pg. 
1000
-
1006
)
12
Speidl
 
W. S.
Zeiner
 
A.
Nikfardjam
 
M.
, et al 
An increase of C-reactive protein is associated with enhanced activation of endogenous fibrinolysis at baseline but an impaired endothelial fibrinolytic response after venous occlusion
J. Am. Coll. Cardiol.
2005
, vol. 
45
 (pg. 
30
-
34
)
13
Chia
 
S.
Qadan
 
M.
Newton
 
R.
Ludlam
 
C. A.
Fox
 
K. A.
Newby
 
D. E.
 
Intra-arterial tumor necrosis factor-α impairs endothelium-dependent vasodilatation and stimulates local tissue plasminogen activator release in humans
Arterioscler., Thromb., Vasc. Biol.
2003
, vol. 
23
 (pg. 
695
-
701
)
14
Newby
 
D. E.
Wright
 
R. A.
Labinjoh
 
C.
, et al 
Endothelial dysfunction, impaired endogenous fibrinolysis, and cigarette smoking: a mechanism for arterial thrombosis and myocardial infarction
Circulation
1999
, vol. 
99
 (pg. 
1411
-
1415
)
15
Paramo
 
J. A.
Orbe
 
J.
Beloqui
 
O.
, et al 
Prothrombin fragment 1+2 is associated with carotid intima-media thickness in subjects free of clinical cardiovascular disease
Stroke
2004
, vol. 
35
 (pg. 
1085
-
1089
)
16
Newby
 
D. E.
Wright
 
R. A.
Ludlam
 
C. A.
Fox
 
K. A.
Boon
 
N. A.
Webb
 
D. J.
 
An in vivo model for the assessment of acute fibrinolytic capacity of the endothelium
Thromb. Haemostasis
1997
, vol. 
78
 (pg. 
1242
-
1248
)
17
Biasucci
 
L. M.
Vitelli
 
A.
Liuzzo
 
G.
, et al 
Elevated levels of interleukin-6 in unstable angina
Circulation
1996
, vol. 
94
 (pg. 
874
-
877
)
18
Lupu
 
F.
Heim
 
D. A.
Bachmann
 
F.
Hurni
 
M.
Kakkar
 
V. V.
Kruithof
 
E. K. O.
 
Plasminogen activator expression in human atherosclerotic lesions
Arterioscler., Thromb., Vasc. Biol.
1995
, vol. 
15
 (pg. 
1444
-
1455
)
19
Suwaidi
 
J. A.
Hamasaki
 
S.
Higano
 
S. T.
Nishimura
 
R. A.
Holmes
 
D. R.
Lerman
 
A.
 
Long-term follow-up of patients with mild coronary artery disease and endothelial dysfunction
Circulation
2000
, vol. 
101
 (pg. 
948
-
954
)
20
Heitzer
 
T.
Schlinzig
 
T.
Krohn
 
K.
Meinertz
 
T.
Munzel
 
T.
 
Endothelial dysfunction, oxidative stress, and risk of cardiovascular events in patients with coronary artery disease
Circulation
2001
, vol. 
104
 (pg. 
2673
-
2678
)
21
Yki-Jarvinen
 
H.
Bergholm
 
R.
Leirisalo-Repo
 
M.
 
Increased inflammatory activity parallels increased basal nitric oxide production and blunted response to nitric oxide in vivo in rheumatoid arthritis
Ann. Rheum. Dis.
2003
, vol. 
62
 (pg. 
630
-
634
)
22
Fichtlscherer
 
S.
Breuer
 
S.
Schachinger
 
V.
Dimmeler
 
S.
Zeiher
 
A. M.
 
C-reactive protein levels determine systemic nitric oxide bioavailability in patients with coronary artery disease
Eur. Heart J.
2004
, vol. 
25
 (pg. 
1412
-
1418
)
23
Yoshizumi
 
M.
Perrella
 
M. A.
Burnett
 
J. C.
Lee
 
M. E.
 
Tumor necrosis factor downregulates an endothelial nitric oxide synthase mRNA by shortening its half-life
Circ. Res.
1993
, vol. 
73
 (pg. 
205
-
209
)
24
Boger
 
R. H.
Bode-Boger
 
S. M.
Szuba
 
A.
, et al 
Asymmetric Dimethylarginine (ADMA): a novel risk factor for endothelial dysfunction: its role in hypercholesterolemia
Circulation
1998
, vol. 
98
 (pg. 
1842
-
1847
)
25
Funakoshi
 
H.
Kubota
 
T.
Machida
 
Y.
, et al 
Involvement of inducible nitric oxide synthase in cardiac dysfunction with tumor necrosis factor-α
Am. J. Physiol. Heart Circ. Physiol.
2002
, vol. 
282
 (pg. 
H2159
-
H2166
)
26
O'Kane
 
K. P.
Webb
 
D. J.
Collier
 
J. G.
Vallance
 
P. J.
 
Local L-NG-monomethyl-arginine attenuates the vasodilator action of bradykinin in the human forearm
Br. J. Clin. Pharmacol.
1994
, vol. 
38
 (pg. 
311
-
315
)
27
Honing
 
M. L. H.
Smits
 
P.
Morrison
 
P. J.
Rabelink
 
T. J.
 
Bradykinin-induced vasodilation of human forearm resistance vessels is primarily mediated by endothelium-dependent hyperpolarization
Hypertension
2000
, vol. 
35
 (pg. 
1314
-
1318
)
28
Brown
 
N. J.
Gainer
 
J. V.
Murphey
 
L. J.
Vaughan
 
D. E.
 
Bradykinin stimulates tissue plasminogen activator release from human forearm vasculature through B2 receptor-dependent, NO synthase-independent, and cyclooxygenase-independent pathway
Circulation
2000
, vol. 
102
 (pg. 
2190
-
2196
)
29
Panza
 
J. A.
Garcia
 
C. E.
Kilcoyne
 
C. M.
Quyyumi
 
A. A.
Cannon
 
R. O.
 
Impaired endothelium-dependent vasodilation in patients with essential hypertension: evidence that nitric oxide abnormality is not localized to a single signal transduction pathway
Circulation
1995
, vol. 
91
 (pg. 
1732
-
1738
)
30
Steins
 
M. B.
Padro
 
T.
Li
 
C.-X.
, et al 
Overexpression of tissue-type plasminogen activator in atherosclerotic human coronary arteries
Atherosclerosis
1999
, vol. 
145
 (pg. 
173
-
180
)
31
Fichtlscherer
 
S.
Rossig
 
L.
Breuer
 
S.
Vasa
 
M.
Dimmeler
 
S.
Zeiher
 
A. M.
 
Tumor necrosis factor antagonism with etanercept improves systemic endothelial vasoreactivity in patients with advanced heart failure
Circulation
2001
, vol. 
104
 (pg. 
3023
-
3025
)
32
Mann
 
D. L.
McMurray
 
J. J. V.
Packer
 
M.
, et al 
Targeted anticytokine therapy in patients with chronic heart failure: Results of the Randomised Etanercept Worldwide Evaluation (RENEWAL)
Circulation
2004
, vol. 
109
 (pg. 
1594
-
1602
)