- Email: [email protected]
In concomitant coronary and peripheral arterial disease, inflammation of the affected limbs predicts coronary artery endothelial dysfunction
Atherosclerosis 201 (2008) 440–446
In concomitant coronary and peripheral arterial disease, inflammation of the affected limbs predicts coronary artery endothelial dysfunction Gregorio Brevetti a,∗,1 , Federico Piscione a,1 , Plinio Cirillo a , Gennaro Galasso a , Vittorio Schiano a , Emanuele Barbato a , Francesco Scopacasa b , Massimo Chiariello a a b
Department of Clinical Medicine and Cardiovascular and Immunological Sciences, University of Naples “Federico II”, Naples, Italy Department of Laboratory Medicine, University of Naples “Federico II”, Naples, Italy
Received 29 August 2007; received in revised form 8 December 2007; accepted 31 January 2008 Available online 15 February 2008
Abstract Background: In coronary artery disease (CAD), concomitant peripheral arterial disease (PAD) entails more severe coronary atherosclerosis. We investigated whether the inflammatory status of affected limbs impairs coronary artery endothelial function (CAEF). Methods: We measured the neutrophil myeloperoxidase content (NMPOxC) and plasma levels of interleukin-6 and C-reactive protein in the aorta, femoral vein, and coronary sinus of 22 CAD + PAD and 18 CAD-alone patients. CAEF was assessed by the cold pressure test. Human coronary artery endothelial cells (HCAECs) were incubated with serum from the femoral vein and aorta of CAD + PAD patients to determine whether blood leaving the affected limb activates HCAECs. Results: In CAD + PAD patients, NMPOxC was higher across the femoral circulation than across the coronary circulation (p < 0.01); it was also higher than across healthy femoral circulation of CAD patients (p < 0.01). These findings apply also to interleukin-6, but not to C-reactive protein. The transfemoral gradient of NMPOxC and interleukin-6 significantly correlated with CAEF. The NMPOxC/CAEF relationship was much greater after exercise (R = 0.79, p < 0.001), which increased neutrophil activation across the affected circulation. The post-exercise association remained significant after adjustment for potential confounders (p < 0.01). Serum from the affected limb of CAD + PAD patients induced, in vitro, a significant release of MCP-1 from HCAECs versus serum from the aorta of the same patients (630 [550–740] vs. 547 [490–620]; p < 0.05). Conclusions: In CAD + PAD, triggers from the affected circulation may activate the endothelium at distant sites. Thus, PAD, besides being a marker of cardiovascular risk, could exert a mechanistic function in CAD progression. © 2008 Elsevier Ireland Ltd. All rights reserved. Keywords: Peripheral arterial disease; Coronary artery disease; Inflammation; Coronary artery endothelial function
1. Introduction In patients with coronary artery disease (CAD), the coexistence of peripheral arterial disease (PAD) entails more severe coronary atherosclerosis [1–3]. However, the greater atherosclerotic coronary artery burden is only in part due to ∗ Corresponding author at: Via G. Iannelli 45/A, I-80131 Napoli, Italy. Tel.: +39 081 7462240; fax: +39 081 7462240. E-mail address: [email protected] (G. Brevetti). 1 These authors contributed equally to the paper.
0021-9150/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2008.01.014
classic risk factors [1,2], which suggests that other mechanisms are involved in this process. Inflammation and impaired endothelial function play a prominent role in CAD [4]. Indeed, in stable angina, increased levels of markers of inflammation and endothelial activation seem to be closely correlated with rapid progression of CAD [5]. In this context, it is noteworthy that CAD + PAD patients have a greater inflammatory profile and more severe coronary atherosclerosis than CAD-alone patients [1,3,6,7]. It remains to be determined whether and to what extent the increased levels of inflammatory molecules
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
441
Fig. 1. Study protocol to assess the relationship between the inflammatory status of the leg and the coronary artery endothelial function. BS is simultaneous blood sampling from the femoral vein, aorta, coronary sinus and from an antecubital vein; QCA is quantitative coronary angiography; CPT is cold pressor test; NTG is intra-coronary bolus of nitroglycerin.
result from a primary “extravascular” activation of the acute phase response, or whether they originate from the site of active plaques. The two mechanisms are not mutually exclusive. However, because the peripheral vascular bed provides a large surface for the release of inflammatory molecules, it is intriguing to speculate that an inflammatory response generated in the affected limb of CAD + PAD patients could affect coronary artery endothelial function (CAEF), thus representing a possible mechanistic link between PAD and CAD severity.
2. Methods 2.1. Patients We studied CAD + PAD patients and CAD-alone patients (control group) admitted to our department for elective coronary angiography. CAD + PAD patients had an ankle-brachial index (ABI) < 0.90 and referred a history of intermittent claudication lasting at least one year. None had critical limb ischemia. PAD was excluded in controls on the basis of echo-color-Doppler examination and ABI > 0.90. Exclusion criteria were a history of trauma, surgery or myocardial infarction in the previous 3 months, unstable angina, malignant disease and any other acute or chronic condition associated with inflammation. Twenty-two of 49 consecutive CAD + PAD patients, and 18 of 35 CAD-alone patients gave their informed written consent to the study. No patient was taking anti-inflammatory drugs other than aspirin as antiplatelet agent. All drugs were discontinued for 18 h or longer before the study, which was performed in the morning after an overnight fast. The study was approved by the Ethics Committee of our Institution.
(IL-6) and C-reactive protein (CRP). After a baseline left coronary angiogram, the patient immersed the right hand in iced water for 90 s and a second coronary angiogram was immediately recorded to evaluate the coronary artery endothelial vasoreactivity to the cold pressure test (CPT). Because in claudicants, acute exercise evokes a marked inflammatory response [8], and stimulates neutrophil activation across the circulation of the affected limb [9], we speculated that an exercise-induced increase in the transfemoral gradient of inflammatory markers could be associated with a further exercise-induced deterioration in coronary artery vasoreactivity. Should this be the case, it would provide more direct evidence of the relationship between the inflammatory reaction in the affected limb and coronary artery endothelial function. Accordingly, when the coronary artery diameter returned to baseline, patients were invited to repeatedly flex and extend the foot using the device shown in Fig. 2. CAD + PAD patients exercised until pain in the claudicant limb became unbearable. Controls were stopped when they reached the number of flexion-extensions done by the PAD patient to whom he/she was matched. Five minutes after exercise, blood samples were drawn for measurements of inflammatory parameters, and coronary artery endothelial vasoreactivity was re-assessed. Finally, nitroglycerin (300 g) was injected into the left coronary artery for assessment of endothelium-independent vasodilator capacity. Coronary vasomotion was measured as follows. In each
2.2. Protocol The protocol of the study is shown in Fig. 1. All subjects underwent, before angiography, simultaneous blood sampling from the femoral vein (in the case of CAD + PAD, the femoral vein was that of the claudicant limb), aorta and coronary sinus for the measurement of neutrophil myeloperoxidase (MPOx) content, and plasma levels of interleukin-6
Fig. 2. Device used by the patients to exercise.
442
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
patient, an average of 3.2 segments were selected in one projection on the baseline angiogram. To obtain as many coronary segments as possible, the left anterior descending coronary artery was used as experimental artery. Three coronary segments were considered: proximal (from the ostium to first septal branch), mid (from the first to second septal branch), and distal (after the second septal branch). The presence of coronary stenosis was defined as % diameter stenosis above 50. In stenotic coronary arteries, besides the stenosis, whenever possible, at least one segment proximal and one segment distal to the stenosis were obtained. In case of ostial stenosis, two or more segments distal to the stenosis were obtained. For each segment, the luminal diameter (LD) was measured at end diastole by quantitative coronary angiography using the catheter as a scaling device. In case of stenotic coronary arteries, changes in LD of the stenosis itself were analyzed separately from the changes of the adjacent reference segments. The data from the stenotic and adjacent reference segments were pooled because there was a uniform response to the CPT. Data are presented as average of % LD changes of coronary segments after CPT compared to baseline. To ensure proper filling of the coronaries with contrast medium, even during high-flow situations, an angioplasty guide catheter was used in each case and an automated contrast delivery system was used to inject the same contrast volume at the various steps of the experimental protocol (Acist Medical Systems). The identical projection was used at the different stages of the protocol. Based on the emergence of side branches, exactly the same segments were analyzed at the different stages of the protocol. Angiograms were recorded at 25 frames/s. Heart rate and blood pressures were digitally recorded during the entire study protocol. The contrast medium used for all patients was the nonionic monomer, hypo-osmolar (Iomeron 400, Bracco, Italy). The neutrophil MPOx index of the mean neutrophil MPOx content was determined by the Bayer H*1 analyzer as described elsewhere [10]. Positive values represent MPOxrich neutrophils, and negative values represent neutrophils depleted of MPOx consequent to neutrophil activation. Thus, a lower MPOx index in blood from the femoral vein, as compared with the aorta, represented an index of neutrophil activation through the femoral vascular bed. Plasma levels of IL-6 and CRP were measured by high-sensitive (hs) ELISA methods (Dade Behring Diagnostics). 2.3. In vitro study Experiments were performed with human coronary artery endothelial cells (HCAECs) (Cambrex Bio Science) grown in EGM 2 medium supplemented with 10% FCS. Cells were used at passages 2–5. In a preliminary pilot experiment using the DNA microarray technique, we found that levels of monocyte chemoattractant protein-1 (MCP-1) were significantly higher in HCAECs stimulated with pooled serum from the femoral vein of the affected limb of CAD + PAD patients than
in both HCAECs stimulated with serum from the aorta of the same patients and HCAECs stimulated with serum from the femoral vein of CAD-alone patients (data not shown). Thus, in another set of experiments, HCAECs were incubated for 3 h with medium supplemented with serum from: (1) the femoral vein of the affected limb of CAD + PAD patients; (2) the aorta of CAD + PAD patients; (3) the femoral vein of the healthy limb of CAD-alone patients; and (4) the aorta of CAD-alone patients. The cells were then washed with PBS and placed in fresh medium. Twelve hours later, the cell medium was collected and MCP-1 levels in the four groups were measured by ELISA (R&D Systems). 2.4. Statistical analysis Due to their skewed distribution, the inflammatory parameters were expressed as median and interquartile range and analyzed by non-parametric tests. The Friedman test and the Wilcoxon test were used for intra-group comparisons, and the Kruskal–Wallis and the Mann–Whitney tests for inter-group comparisons. Correlations were tested with the Spearman method. We used linear logistic regression analyses to assess the association of the extent of inflammation in the affected limb with coronary artery endothelial function, as dependent variable.
3. Results The characteristics of the study population are shown in Table 1. Cardiovascular treatments were similar in the two groups. 3.1. Peripheral vascular inflammation Neutrophil activation occurred across the circulation of the affected limb of CAD + PAD patients. As Fig. 3 shows, the median venous–arterial (V–A) difference in MPOx content across the femoral circulation (−0.8 [−2.6; −0.1]) was Table 1 Patients’ characteristics at baseline CAD + PAD (n = 22)
CAD (n = 18)
p
Age (year) Gender (males) Hypertension (%) Dyslipidemia (%) Diabetes mellitus (%) Smoking (%) ABI
61 ± 9 18 (82) 14 (64) 15 (68) 6 (27) 16 (72) 0.62 ± 0.10
64 ± 5 13 (72) 13 (72) 13 (72) 4 (22) 14 (77) 1.12 ± 0.10
0.20 0.47 0.56 0.78 0.71 0.71 <0.01
Treatments Statins (%) ACE-inhibitors (%) Beta-blockers (%) Nitrates (%)
14 (64) 12 (54) 4 (18) 4 (18)
11 (61) 9 (50) 5 (23) 4 (22)
0.87 0.77 0.47 0.75
CAD: coronary artery disease; PAD: peripheral arterial disease; ABI: anklebrachial index; ACE: angiotensin converting enzyme.
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
443
Fig. 3. In CAD + PAD patients, the MPOx and IL-6 venous–arterial (V–A) difference across the femoral circulation was significantly greater than that across the coronary circulation and that across the femoral circulation of the healthy legs of CAD-alone patients. Conversely, in the CAD-only patients, the transfemoral MPOx and IL-6 content was similar to that across the coronary circulation. Contrary to the MPOx and IL-6 data, the venous–arterial difference of hsCRP across the affected limb resembled that in the other vascular districts.
significantly greater than that across the coronary circulation (0.3 [−2.3; 1.2]; p = 0.01) and that in the transfemoral circulation of the legs of CAD-alone patients (0.1 [−0.1; 0.4]; p < 0.01). Conversely, in the CAD-only patients, the transfemoral MPOx content was similar to that across the coronary circulation, namely, 0.1 (0.0; 0.4). Neutrophil counts in the aorta, femoral vein and coronary sinus did not differ within the two groups of patients, but were significantly higher in CAD + PAD than in CAD-alone patients, in each of the vascular sites (data not shown). The IL-6 results confirmed the inflammatory reaction in the affected limb of CAD + PAD patients. As Fig. 3 shows, the IL-6 V–A difference across the affected femoral circulation (0.7 [0.1; 1.1] pg/ml) was greater than that across the coronary circulation (0.1 [−0.1; 0.4] pg/ml; p < 0.05), and that across the femoral circulation of the healthy legs of CADalone patients (0.0 [−0.2; 0.1] pg/ml; p < 0.01). Conversely, in the CAD-alone patients, the transfemoral IL-6 content was similar to that across the coronary circulation, namely, −0.1 (−0.2; 0.1). Contrary to MPOx and IL-6 data, the V–A difference of hsCRP across the affected limb resembled that in the other vascular districts (Fig. 3). In CAD + PAD patients there was no relationship between transfemoral V–A difference of the three inflammatory parameters and ABI, which reflects the severity of ischemia in the disease leg. Furthermore, the ABI did not correlate with CAD severity evaluated by number of coronary stenoses >50% and type C coronary lesions. 3.2. Peripheral vascular inflammation and coronary artery endothelial function at rest Coronary artery endothelial-mediated vasoreactivity was 0.0 (−2.1; 2.4)% in CAD + PAD and 4.0 (1.5; 10.0)%
(p < 0.05) in CAD-alone patients. CAEF was more impaired in CAD + PAD patients with a transfemoral gradient of MPOx > median (−1.4 [−13.8; 2.4]%) than in CAD + PAD patients with a transfemoral gradient of MPOx < median (1.5 [0.0; 4.2]%; p < 0.05) and CAD-alone patients (4.0 [1.5; 10.0]%; p = 0.01) (Fig. 4). Also CAD + PAD patients with a transfemoral gradient of IL-6 > median had greater impairment of CAEF (−1.8 [−18.5; 2.0]%) versus those with less peripheral inflammation (1.3 [−0.6; 4.8]%; p < 0.05) and versus CAD-alone patients (4.0 [1.5; 10]%; p < 0.01). Conversely, CAEF was similar in CAD + PAD patients categorized according to the median value of transfemoral gradient of CRP and in CAD-alone patients (Fig. 4). In CAD + PAD, under resting conditions, endotheliumdependent coronary artery vasoreactivity correlated with the transfemoral V–A difference of leukocyte MPOx content (ρ = 0.53, p < 0.05) and IL-6 plasma levels (ρ = 0.45, p < 0.05), but not with that of hsCRP. Even more important, in the CAD + PAD group, linear regression analyses showed that the association between CAEF and the transfemoral gradient of both MPOx and IL-6 remained significant after adjustment for CAD severity (evaluated by number of coronary stenoses >50%, and number of type C coronary lesions) (β-coefficient −0.55; 95% confidence interval [CI] −16.47 to −1.26, p < 0.05 for MPOx, and −0.68; 95% CI −18.87 to −3.14, p < 0.01 for IL-6). Even when treatments were included in the multivariate analysis, in addition to CAD severity, the only variables that remained associated with CAEF were MPOx (β-coefficient −0.72; 95% CI −22.24 to −0.80, p < 0.05) and IL-6 (β-coefficient −0.60; 95% CI −18.51 to −1.06, p < 0.05). There was no relationship between CAEF and the transcoronary gradient of MPOx, IL-6 and hsCRP.
444
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
Fig. 4. CAD + PAD patients with a transfemoral gradient of MPOx and IL-6 > median had greater impairment of coronary artery endothelial function versus those with less peripheral inflammation and versus CAD-alone patients. Conversely, coronary artery endothelial function was similar in CAD + PAD patients categorized according to the median value of the transfemoral gradient of CRP and in CAD-alone patients.
3.3. Peripheral vascular inflammation and coronary artery endothelial function after exercise Exercise did not further impair endothelial function in 7 of the 22 CAD + PAD patients. Importantly, in these patients exercise did not induce significant changes in the transfemoral gradient of the inflammatory parameters (data not shown). In the remaining 15 patients, exercise either amplified the vasoconstrictive response of coronary arteries to CPT or reversed the response from vasodilatation to vasoconstriction. Changes between rest and post-exercise vasoreactivity of coronary arteries to CPT highly correlated with exercise-induced changes in transfemoral neutrophil MPOx content (ρ = 0.79, p < 0.01). This association remained after adjustment for CAD severity (evaluated by number of coronary stenoses >50%, number of type C lesions and baseline endothelial function) and treatments (β-coefficient 0.69; 95% CI [0.52; 0.28], p < 0.01). Conversely, exerciseinduced changes in transfemoral concentrations of IL-6 and hsCRP were not associated with changes between rest and post-exercise coronary vasoreactivity to CPT. This may be explained by the finding that, whereas exercise increased the transfemoral neutrophil MPOx content from the resting value of −0.9 (−2.6; −0.1) to −2.0 (−2.8; −0.5) (p < 0.05), it did not alter post-exercise transfemoral plasma levels of IL-6 or hsCRP. Noteworthy, exercise did not modify the transcoronary gradient of the inflammatory markers. Therefore, the deterioration in CAEF did not result from an inflammatory response occurring in the coronary district. Similarly, in the CAD-alone group, post-exercise changes in coronary artery endothelium-mediated vasoreactivity did not correlate with transfemoral changes in inflammatory parameters. No patient of either group had angina, arrhythmias or ECG evidence of myocardial ischemia during exercise. The percent increase in luminal diameter after nitroglycerine was similar in the two groups.
Fig. 5. When co-incubated with serum from the affected limb, HCAECs released more MCP-1 than when they were incubated with serum from the aorta of the same patients. Notably, this difference disappeared when HCAECs were incubated with serum from healthy legs or aorta of CAD patients without PAD. FV: femoral vein.
3.4. In vitro study As Fig. 5 shows, serum from the affected limb of CAD + PAD patients induced a significantly higher release of MCP-1 from HCAECs than serum from the aorta of the same patients (630 [550–740] vs. 547 [490–620], p < 0.05). The difference disappeared when HCAECs were incubated with serum from CAD-alone patients.
4. Discussion Atherosclerosis is considered a dynamic, progressive disease arising from the combination of inflammation and endothelial dysfunction [4]. Here we report the novel finding that in CAD + PAD patients, increased gradients of neutrophil MPOx content and IL-6 across the femoral circulation of the
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
affected limb are associated with impaired CAEF. Given the atherogenic potential of endothelial dysfunction, our findings suggest that the association of PAD with more severe CAD [1–3] is not simply the expression of a more extensive disease involving diverse vascular districts, but is probably related also to peripheral inflammation. Interestingly, the transfemoral V–A difference of the inflammatory markers was unrelated to PAD severity measured as ABI [11]. Therefore, the local inflammatory response might be influenced by the quality rather than the quantity of peripheral atherosclerotic lesions. In other words, neutrophils may be activated by interacting with inflamed plaques, which are common in femoral arteries [12,13]. Furthermore, active plaques express IL-6 [14], which may be released into the circulation. Both phenomena may exert systemic effects. In particular, MPOx release from leukocytes, which in our study is reflected by reduced neutrophil content of the enzyme, results in an array of diffusible oxidants [15] that harm the endothelium [16]. In addition, MPOx uses nitric oxide as a physiologic substrate, thereby reducing nitric oxide bioavailability [17] and contributing to endothelial dysfunction [18]. Because also IL-6 may affect the endothelium [19], it is not surprising that, in CAD + PAD patients under resting conditions, the greater the transfemoral V–A difference in IL-6 and in neutrophil MPOx content (and, thus, presumably the greater the release of MPOx from activated neutrophils), the more severe the endothelial dysfunction in the coronary bed. Although correlation coefficients provide no information on cause and effect, numerous findings support the hypothesis that peripheral inflammation contributes to coronary artery endothelial dysfunction in these patients. First, linear logistic regression analysis showed that the MPOx gradient across the affected femoral circulation remained associated with worse coronary endothelial function, after adjustment for CAD severity and cardiovascular treatments. Second, coronary vasoreactivity was not related to the transcoronary gradient of inflammatory markers, which was close to zero, thus excluding that intra-coronary vascular inflammation affected CAEF. Third, we found no relationship between PAD severity measured as ABI, number of coronary stenoses >50% and type C coronary lesions. This, in addition to the fact that ABI did not correlate with peripheral inflammation excludes that PAD severity influences CAEF. Fourth, serum from the claudicant limb of CAD + PAD patients induced a proinflammatory state in HCAECs in vitro. HCAECs released more MCP-1 when exposed to serum from the affected limb than when exposed to serum from the aorta of the same patients. This difference disappeared when HCAECs were incubated with serum from healthy legs or the aorta of CAD-alone patients. Therefore, it is reasonable to assume that the blood leaving the affected limb of PAD patients contains substances that affect the endothelial cells of other vascular districts. Lastly, our post-exercise data support the concept of a mechanistic link between peripheral vascular inflammation and CAEF. In fact, the MPOx/coronary dysfunction association was greater
445
after exercise, and thus CAD + PAD patients who with exercise had the greatest transfemoral reduction in neutrophil MPOx content had the greatest exercise-induced endothelial dysfunction. Not less important is the observation that this post-exercise association was not observed for IL-6, the transfemoral levels of which did not increase in the exercising affected limb. This suggests that MPOx and IL-6 play different roles in the inflammatory process of the claudicant limb. The same applies to CRP whose concentration in blood leaving the affected limb neared that in aortic blood, thus excluding transfemoral release of this molecule. This finding, which coincides with a recent study of coronary circulation [20], suggests a systemic origin of CRP. Notably, the transfemoral concentration of inflammatory markers was unrelated to CAEF in CAD-alone patients, who did not have any inflammatory reaction in the exercising healthy leg.
5. Study limitations Due to the complexity of the protocol, many patients refused to participate in the study, and consequently the CAD + PAD and CAD-alone groups are relatively small. Therefore, the results of this hypothesis-generating study need to be confirmed by larger trials. Secondly, although treatments were discontinued at least 18 h before the study, we cannot exclude that some drugs may have reduced the inflammatory response in the affected limb and/or beneficially influenced CAEF. Furthermore, because of the broad spectrum of drugs and doses administrated, we were unable to examine the interaction between treatment status, inflammation and CAEF. However, multivariate analysis showed that treatments did not affect the significant association between inflammatory status of the claudicant limb and CAEF.
6. Conclusions This study points to a causal link between peripheral inflammation in the affected limb of CAD + PAD patients and coronary artery endothelial dysfunction. Our findings that the exercise-induced increase in the inflammatory response of the claudicant limb caused a further deterioration of CAEF, and that serum from the affected limb of PAD patients exerted proinflammatory effects on HCAECs (a phenomenon not observed with serum from the healthy leg of CAD-alone patients) suggest the existence of circulatory triggers in PAD that could activate the endothelium at distant sites. Should this be the case, PAD, besides being a marker of cardiovascular risk, could exert a mechanistic function in the progression of atherosclerosis in coronary arteries. Should larger, prospective studies confirm this concept by demonstrating that peripheral vascular inflammation portends a higher risk of coronary events, affected individuals would be candidates for more aggressive standard therapy or for emerging therapies aimed at reducing local intra-arterial inflammation.
446
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
References [1] Brevetti G, Piscione F, Silvestro A, et al. Increased inflammatory status and higher prevalence of three-vessel coronary artery disease in patients with concomitant coronary and peripheral atherosclerosis. Thromb Haemost 2003;89:1058–63. [2] Papamichael CM, Lekakis JP, Stamatelopoulos KS, et al. Anklebrachial index as a predictor of the extent of coronary atherosclerosis and cardiovascular events in patients with coronary artery disease. Am J Cardiol 2000;86:615–8. [3] Erren M, Reinecke H, Junker R, et al. Systemic inflammatory parameters in patients with atherosclerosis of the coronary and peripheral arteries. Arterioscler Thromb Vasc Biol 1999;19:2355–63. [4] Libby P, Ridker PM, Maseri A. Inflammation and atherosclerosis. Circulation 2002;105:1135–43. [5] Zouridakis E, Avanzas P, Arroyo-Espliguero R, Fredericks S, Kaski JC. Markers of inflammation and rapid coronary artery disease progression in patients with stable angina pectoris. Circulation 2004;110: 1747–53. [6] Tataru MC, Heinrich J, Junker R, et al. C-reactive protein and the severity of atherosclerosis in myocardial infarction patients with stable angina pectoris. Eur Heart J 2000;21:1000–8. [7] Narins CR, Zareba W, Moss AJ, et al. Relationship between intermittent claudication, inflammation, thrombosis, and recurrent cardiac events among survivors of myocardial infarction. Arch Intern Med 2004;164:440–6. [8] Edwards AT, Blann AD, Suarez-Mendez VJ, Lardi AM, McCollum CN. Systemic responses in patients with intermittent claudication after treadmill exercise. Br J Surg 1994;81:1738–41. [9] Neumann FJ, Waas W, Diehm C, et al. Activation and decreased deformability of neutrophils after intermittent claudication. Circulation 1990;82:922–9.
[10] Buffon A, Biasucci LM, Liuzzo G, et al. Widespread coronary inflammation in unstable angina. N Engl J Med 2002;347:5–12. [11] Fowkes FGR, Allan PL, Tsampoulas C, Smith FB, Donnan PT. Validity of duplex scanning in the d´etection of peripheral arterial disease in the general population. Eur J Vasc Surg 1992;6:31–5. [12] Pasterkamp G, Schoneveld AH, van der Wal AC, et al. Inflammation of the atherosclerotic cap and shoulder of the plaque is a common and locally observed feature in unruptured plaques of femoral and coronary arteries. Arterioscler Thromb Vasc Biol 1999;19:54–8. [13] Vink A, Schoneveld AH, Richard W, et al. Plaque burden, arterial remodeling and plaque vulnerability: determined by systemic factors? J Am Coll Cardiol 2001;38:718–23. [14] Schieffer B, Schieffer E, Hilfiker-Kleiner D, et al. Expression of angiotensin II and interleukin 6 in human coronary atherosclerotic plaques: potential implications for inflammation and plaque instability. Circulation 2000;101:1372–8. [15] Podrez EA, Abu-Soud HM, Hazen SL. Myeloperoxidase-generated oxidants and atherosclerosis. Free Radic Biol Med 2000;28:1717–25. [16] Muller G, Goettsch C, Morawietz H. Oxidative stress and endothelial dysfunction. Hamostaseologie 2007;27:5–12. [17] Eiserich JP, Baldus S, Brennan ML, et al. Myeloperoxidase, a leukocyte-derived vascular NO oxidase. Science 2002;296:2391–4. [18] Vita JA, Brennan ML, Gokce N, et al. Serum myeloperoxidase levels independently predict endothelial dysfunction in humans. Circulation 2004;110:1134–9. [19] Wassmann S, Stumpf M, Strehlow K, et al. Interleukin-6 induces oxidative stress and endothelial dysfunction by overexpression of the angiotensin II type 1 receptor. Circ Res 2004;94:534–41. [20] Maier W, Altwegg LA, Corti R, et al. Inflammatory markers at the site of ruptured plaque in acute myocardial infarction: locally increased interleukin-6 and serum amyloid A but decreased C-reactive protein. Circulation 2005;111:1355–61.
In concomitant coronary and peripheral arterial disease, inflammation of the affected limbs predicts coronary artery endothelial dysfunction Gregorio Brevetti a,∗,1 , Federico Piscione a,1 , Plinio Cirillo a , Gennaro Galasso a , Vittorio Schiano a , Emanuele Barbato a , Francesco Scopacasa b , Massimo Chiariello a a b
Department of Clinical Medicine and Cardiovascular and Immunological Sciences, University of Naples “Federico II”, Naples, Italy Department of Laboratory Medicine, University of Naples “Federico II”, Naples, Italy
Received 29 August 2007; received in revised form 8 December 2007; accepted 31 January 2008 Available online 15 February 2008
Abstract Background: In coronary artery disease (CAD), concomitant peripheral arterial disease (PAD) entails more severe coronary atherosclerosis. We investigated whether the inflammatory status of affected limbs impairs coronary artery endothelial function (CAEF). Methods: We measured the neutrophil myeloperoxidase content (NMPOxC) and plasma levels of interleukin-6 and C-reactive protein in the aorta, femoral vein, and coronary sinus of 22 CAD + PAD and 18 CAD-alone patients. CAEF was assessed by the cold pressure test. Human coronary artery endothelial cells (HCAECs) were incubated with serum from the femoral vein and aorta of CAD + PAD patients to determine whether blood leaving the affected limb activates HCAECs. Results: In CAD + PAD patients, NMPOxC was higher across the femoral circulation than across the coronary circulation (p < 0.01); it was also higher than across healthy femoral circulation of CAD patients (p < 0.01). These findings apply also to interleukin-6, but not to C-reactive protein. The transfemoral gradient of NMPOxC and interleukin-6 significantly correlated with CAEF. The NMPOxC/CAEF relationship was much greater after exercise (R = 0.79, p < 0.001), which increased neutrophil activation across the affected circulation. The post-exercise association remained significant after adjustment for potential confounders (p < 0.01). Serum from the affected limb of CAD + PAD patients induced, in vitro, a significant release of MCP-1 from HCAECs versus serum from the aorta of the same patients (630 [550–740] vs. 547 [490–620]; p < 0.05). Conclusions: In CAD + PAD, triggers from the affected circulation may activate the endothelium at distant sites. Thus, PAD, besides being a marker of cardiovascular risk, could exert a mechanistic function in CAD progression. © 2008 Elsevier Ireland Ltd. All rights reserved. Keywords: Peripheral arterial disease; Coronary artery disease; Inflammation; Coronary artery endothelial function
1. Introduction In patients with coronary artery disease (CAD), the coexistence of peripheral arterial disease (PAD) entails more severe coronary atherosclerosis [1–3]. However, the greater atherosclerotic coronary artery burden is only in part due to ∗ Corresponding author at: Via G. Iannelli 45/A, I-80131 Napoli, Italy. Tel.: +39 081 7462240; fax: +39 081 7462240. E-mail address: [email protected] (G. Brevetti). 1 These authors contributed equally to the paper.
0021-9150/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2008.01.014
classic risk factors [1,2], which suggests that other mechanisms are involved in this process. Inflammation and impaired endothelial function play a prominent role in CAD [4]. Indeed, in stable angina, increased levels of markers of inflammation and endothelial activation seem to be closely correlated with rapid progression of CAD [5]. In this context, it is noteworthy that CAD + PAD patients have a greater inflammatory profile and more severe coronary atherosclerosis than CAD-alone patients [1,3,6,7]. It remains to be determined whether and to what extent the increased levels of inflammatory molecules
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
441
Fig. 1. Study protocol to assess the relationship between the inflammatory status of the leg and the coronary artery endothelial function. BS is simultaneous blood sampling from the femoral vein, aorta, coronary sinus and from an antecubital vein; QCA is quantitative coronary angiography; CPT is cold pressor test; NTG is intra-coronary bolus of nitroglycerin.
result from a primary “extravascular” activation of the acute phase response, or whether they originate from the site of active plaques. The two mechanisms are not mutually exclusive. However, because the peripheral vascular bed provides a large surface for the release of inflammatory molecules, it is intriguing to speculate that an inflammatory response generated in the affected limb of CAD + PAD patients could affect coronary artery endothelial function (CAEF), thus representing a possible mechanistic link between PAD and CAD severity.
2. Methods 2.1. Patients We studied CAD + PAD patients and CAD-alone patients (control group) admitted to our department for elective coronary angiography. CAD + PAD patients had an ankle-brachial index (ABI) < 0.90 and referred a history of intermittent claudication lasting at least one year. None had critical limb ischemia. PAD was excluded in controls on the basis of echo-color-Doppler examination and ABI > 0.90. Exclusion criteria were a history of trauma, surgery or myocardial infarction in the previous 3 months, unstable angina, malignant disease and any other acute or chronic condition associated with inflammation. Twenty-two of 49 consecutive CAD + PAD patients, and 18 of 35 CAD-alone patients gave their informed written consent to the study. No patient was taking anti-inflammatory drugs other than aspirin as antiplatelet agent. All drugs were discontinued for 18 h or longer before the study, which was performed in the morning after an overnight fast. The study was approved by the Ethics Committee of our Institution.
(IL-6) and C-reactive protein (CRP). After a baseline left coronary angiogram, the patient immersed the right hand in iced water for 90 s and a second coronary angiogram was immediately recorded to evaluate the coronary artery endothelial vasoreactivity to the cold pressure test (CPT). Because in claudicants, acute exercise evokes a marked inflammatory response [8], and stimulates neutrophil activation across the circulation of the affected limb [9], we speculated that an exercise-induced increase in the transfemoral gradient of inflammatory markers could be associated with a further exercise-induced deterioration in coronary artery vasoreactivity. Should this be the case, it would provide more direct evidence of the relationship between the inflammatory reaction in the affected limb and coronary artery endothelial function. Accordingly, when the coronary artery diameter returned to baseline, patients were invited to repeatedly flex and extend the foot using the device shown in Fig. 2. CAD + PAD patients exercised until pain in the claudicant limb became unbearable. Controls were stopped when they reached the number of flexion-extensions done by the PAD patient to whom he/she was matched. Five minutes after exercise, blood samples were drawn for measurements of inflammatory parameters, and coronary artery endothelial vasoreactivity was re-assessed. Finally, nitroglycerin (300 g) was injected into the left coronary artery for assessment of endothelium-independent vasodilator capacity. Coronary vasomotion was measured as follows. In each
2.2. Protocol The protocol of the study is shown in Fig. 1. All subjects underwent, before angiography, simultaneous blood sampling from the femoral vein (in the case of CAD + PAD, the femoral vein was that of the claudicant limb), aorta and coronary sinus for the measurement of neutrophil myeloperoxidase (MPOx) content, and plasma levels of interleukin-6
Fig. 2. Device used by the patients to exercise.
442
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
patient, an average of 3.2 segments were selected in one projection on the baseline angiogram. To obtain as many coronary segments as possible, the left anterior descending coronary artery was used as experimental artery. Three coronary segments were considered: proximal (from the ostium to first septal branch), mid (from the first to second septal branch), and distal (after the second septal branch). The presence of coronary stenosis was defined as % diameter stenosis above 50. In stenotic coronary arteries, besides the stenosis, whenever possible, at least one segment proximal and one segment distal to the stenosis were obtained. In case of ostial stenosis, two or more segments distal to the stenosis were obtained. For each segment, the luminal diameter (LD) was measured at end diastole by quantitative coronary angiography using the catheter as a scaling device. In case of stenotic coronary arteries, changes in LD of the stenosis itself were analyzed separately from the changes of the adjacent reference segments. The data from the stenotic and adjacent reference segments were pooled because there was a uniform response to the CPT. Data are presented as average of % LD changes of coronary segments after CPT compared to baseline. To ensure proper filling of the coronaries with contrast medium, even during high-flow situations, an angioplasty guide catheter was used in each case and an automated contrast delivery system was used to inject the same contrast volume at the various steps of the experimental protocol (Acist Medical Systems). The identical projection was used at the different stages of the protocol. Based on the emergence of side branches, exactly the same segments were analyzed at the different stages of the protocol. Angiograms were recorded at 25 frames/s. Heart rate and blood pressures were digitally recorded during the entire study protocol. The contrast medium used for all patients was the nonionic monomer, hypo-osmolar (Iomeron 400, Bracco, Italy). The neutrophil MPOx index of the mean neutrophil MPOx content was determined by the Bayer H*1 analyzer as described elsewhere [10]. Positive values represent MPOxrich neutrophils, and negative values represent neutrophils depleted of MPOx consequent to neutrophil activation. Thus, a lower MPOx index in blood from the femoral vein, as compared with the aorta, represented an index of neutrophil activation through the femoral vascular bed. Plasma levels of IL-6 and CRP were measured by high-sensitive (hs) ELISA methods (Dade Behring Diagnostics). 2.3. In vitro study Experiments were performed with human coronary artery endothelial cells (HCAECs) (Cambrex Bio Science) grown in EGM 2 medium supplemented with 10% FCS. Cells were used at passages 2–5. In a preliminary pilot experiment using the DNA microarray technique, we found that levels of monocyte chemoattractant protein-1 (MCP-1) were significantly higher in HCAECs stimulated with pooled serum from the femoral vein of the affected limb of CAD + PAD patients than
in both HCAECs stimulated with serum from the aorta of the same patients and HCAECs stimulated with serum from the femoral vein of CAD-alone patients (data not shown). Thus, in another set of experiments, HCAECs were incubated for 3 h with medium supplemented with serum from: (1) the femoral vein of the affected limb of CAD + PAD patients; (2) the aorta of CAD + PAD patients; (3) the femoral vein of the healthy limb of CAD-alone patients; and (4) the aorta of CAD-alone patients. The cells were then washed with PBS and placed in fresh medium. Twelve hours later, the cell medium was collected and MCP-1 levels in the four groups were measured by ELISA (R&D Systems). 2.4. Statistical analysis Due to their skewed distribution, the inflammatory parameters were expressed as median and interquartile range and analyzed by non-parametric tests. The Friedman test and the Wilcoxon test were used for intra-group comparisons, and the Kruskal–Wallis and the Mann–Whitney tests for inter-group comparisons. Correlations were tested with the Spearman method. We used linear logistic regression analyses to assess the association of the extent of inflammation in the affected limb with coronary artery endothelial function, as dependent variable.
3. Results The characteristics of the study population are shown in Table 1. Cardiovascular treatments were similar in the two groups. 3.1. Peripheral vascular inflammation Neutrophil activation occurred across the circulation of the affected limb of CAD + PAD patients. As Fig. 3 shows, the median venous–arterial (V–A) difference in MPOx content across the femoral circulation (−0.8 [−2.6; −0.1]) was Table 1 Patients’ characteristics at baseline CAD + PAD (n = 22)
CAD (n = 18)
p
Age (year) Gender (males) Hypertension (%) Dyslipidemia (%) Diabetes mellitus (%) Smoking (%) ABI
61 ± 9 18 (82) 14 (64) 15 (68) 6 (27) 16 (72) 0.62 ± 0.10
64 ± 5 13 (72) 13 (72) 13 (72) 4 (22) 14 (77) 1.12 ± 0.10
0.20 0.47 0.56 0.78 0.71 0.71 <0.01
Treatments Statins (%) ACE-inhibitors (%) Beta-blockers (%) Nitrates (%)
14 (64) 12 (54) 4 (18) 4 (18)
11 (61) 9 (50) 5 (23) 4 (22)
0.87 0.77 0.47 0.75
CAD: coronary artery disease; PAD: peripheral arterial disease; ABI: anklebrachial index; ACE: angiotensin converting enzyme.
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
443
Fig. 3. In CAD + PAD patients, the MPOx and IL-6 venous–arterial (V–A) difference across the femoral circulation was significantly greater than that across the coronary circulation and that across the femoral circulation of the healthy legs of CAD-alone patients. Conversely, in the CAD-only patients, the transfemoral MPOx and IL-6 content was similar to that across the coronary circulation. Contrary to the MPOx and IL-6 data, the venous–arterial difference of hsCRP across the affected limb resembled that in the other vascular districts.
significantly greater than that across the coronary circulation (0.3 [−2.3; 1.2]; p = 0.01) and that in the transfemoral circulation of the legs of CAD-alone patients (0.1 [−0.1; 0.4]; p < 0.01). Conversely, in the CAD-only patients, the transfemoral MPOx content was similar to that across the coronary circulation, namely, 0.1 (0.0; 0.4). Neutrophil counts in the aorta, femoral vein and coronary sinus did not differ within the two groups of patients, but were significantly higher in CAD + PAD than in CAD-alone patients, in each of the vascular sites (data not shown). The IL-6 results confirmed the inflammatory reaction in the affected limb of CAD + PAD patients. As Fig. 3 shows, the IL-6 V–A difference across the affected femoral circulation (0.7 [0.1; 1.1] pg/ml) was greater than that across the coronary circulation (0.1 [−0.1; 0.4] pg/ml; p < 0.05), and that across the femoral circulation of the healthy legs of CADalone patients (0.0 [−0.2; 0.1] pg/ml; p < 0.01). Conversely, in the CAD-alone patients, the transfemoral IL-6 content was similar to that across the coronary circulation, namely, −0.1 (−0.2; 0.1). Contrary to MPOx and IL-6 data, the V–A difference of hsCRP across the affected limb resembled that in the other vascular districts (Fig. 3). In CAD + PAD patients there was no relationship between transfemoral V–A difference of the three inflammatory parameters and ABI, which reflects the severity of ischemia in the disease leg. Furthermore, the ABI did not correlate with CAD severity evaluated by number of coronary stenoses >50% and type C coronary lesions. 3.2. Peripheral vascular inflammation and coronary artery endothelial function at rest Coronary artery endothelial-mediated vasoreactivity was 0.0 (−2.1; 2.4)% in CAD + PAD and 4.0 (1.5; 10.0)%
(p < 0.05) in CAD-alone patients. CAEF was more impaired in CAD + PAD patients with a transfemoral gradient of MPOx > median (−1.4 [−13.8; 2.4]%) than in CAD + PAD patients with a transfemoral gradient of MPOx < median (1.5 [0.0; 4.2]%; p < 0.05) and CAD-alone patients (4.0 [1.5; 10.0]%; p = 0.01) (Fig. 4). Also CAD + PAD patients with a transfemoral gradient of IL-6 > median had greater impairment of CAEF (−1.8 [−18.5; 2.0]%) versus those with less peripheral inflammation (1.3 [−0.6; 4.8]%; p < 0.05) and versus CAD-alone patients (4.0 [1.5; 10]%; p < 0.01). Conversely, CAEF was similar in CAD + PAD patients categorized according to the median value of transfemoral gradient of CRP and in CAD-alone patients (Fig. 4). In CAD + PAD, under resting conditions, endotheliumdependent coronary artery vasoreactivity correlated with the transfemoral V–A difference of leukocyte MPOx content (ρ = 0.53, p < 0.05) and IL-6 plasma levels (ρ = 0.45, p < 0.05), but not with that of hsCRP. Even more important, in the CAD + PAD group, linear regression analyses showed that the association between CAEF and the transfemoral gradient of both MPOx and IL-6 remained significant after adjustment for CAD severity (evaluated by number of coronary stenoses >50%, and number of type C coronary lesions) (β-coefficient −0.55; 95% confidence interval [CI] −16.47 to −1.26, p < 0.05 for MPOx, and −0.68; 95% CI −18.87 to −3.14, p < 0.01 for IL-6). Even when treatments were included in the multivariate analysis, in addition to CAD severity, the only variables that remained associated with CAEF were MPOx (β-coefficient −0.72; 95% CI −22.24 to −0.80, p < 0.05) and IL-6 (β-coefficient −0.60; 95% CI −18.51 to −1.06, p < 0.05). There was no relationship between CAEF and the transcoronary gradient of MPOx, IL-6 and hsCRP.
444
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
Fig. 4. CAD + PAD patients with a transfemoral gradient of MPOx and IL-6 > median had greater impairment of coronary artery endothelial function versus those with less peripheral inflammation and versus CAD-alone patients. Conversely, coronary artery endothelial function was similar in CAD + PAD patients categorized according to the median value of the transfemoral gradient of CRP and in CAD-alone patients.
3.3. Peripheral vascular inflammation and coronary artery endothelial function after exercise Exercise did not further impair endothelial function in 7 of the 22 CAD + PAD patients. Importantly, in these patients exercise did not induce significant changes in the transfemoral gradient of the inflammatory parameters (data not shown). In the remaining 15 patients, exercise either amplified the vasoconstrictive response of coronary arteries to CPT or reversed the response from vasodilatation to vasoconstriction. Changes between rest and post-exercise vasoreactivity of coronary arteries to CPT highly correlated with exercise-induced changes in transfemoral neutrophil MPOx content (ρ = 0.79, p < 0.01). This association remained after adjustment for CAD severity (evaluated by number of coronary stenoses >50%, number of type C lesions and baseline endothelial function) and treatments (β-coefficient 0.69; 95% CI [0.52; 0.28], p < 0.01). Conversely, exerciseinduced changes in transfemoral concentrations of IL-6 and hsCRP were not associated with changes between rest and post-exercise coronary vasoreactivity to CPT. This may be explained by the finding that, whereas exercise increased the transfemoral neutrophil MPOx content from the resting value of −0.9 (−2.6; −0.1) to −2.0 (−2.8; −0.5) (p < 0.05), it did not alter post-exercise transfemoral plasma levels of IL-6 or hsCRP. Noteworthy, exercise did not modify the transcoronary gradient of the inflammatory markers. Therefore, the deterioration in CAEF did not result from an inflammatory response occurring in the coronary district. Similarly, in the CAD-alone group, post-exercise changes in coronary artery endothelium-mediated vasoreactivity did not correlate with transfemoral changes in inflammatory parameters. No patient of either group had angina, arrhythmias or ECG evidence of myocardial ischemia during exercise. The percent increase in luminal diameter after nitroglycerine was similar in the two groups.
Fig. 5. When co-incubated with serum from the affected limb, HCAECs released more MCP-1 than when they were incubated with serum from the aorta of the same patients. Notably, this difference disappeared when HCAECs were incubated with serum from healthy legs or aorta of CAD patients without PAD. FV: femoral vein.
3.4. In vitro study As Fig. 5 shows, serum from the affected limb of CAD + PAD patients induced a significantly higher release of MCP-1 from HCAECs than serum from the aorta of the same patients (630 [550–740] vs. 547 [490–620], p < 0.05). The difference disappeared when HCAECs were incubated with serum from CAD-alone patients.
4. Discussion Atherosclerosis is considered a dynamic, progressive disease arising from the combination of inflammation and endothelial dysfunction [4]. Here we report the novel finding that in CAD + PAD patients, increased gradients of neutrophil MPOx content and IL-6 across the femoral circulation of the
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
affected limb are associated with impaired CAEF. Given the atherogenic potential of endothelial dysfunction, our findings suggest that the association of PAD with more severe CAD [1–3] is not simply the expression of a more extensive disease involving diverse vascular districts, but is probably related also to peripheral inflammation. Interestingly, the transfemoral V–A difference of the inflammatory markers was unrelated to PAD severity measured as ABI [11]. Therefore, the local inflammatory response might be influenced by the quality rather than the quantity of peripheral atherosclerotic lesions. In other words, neutrophils may be activated by interacting with inflamed plaques, which are common in femoral arteries [12,13]. Furthermore, active plaques express IL-6 [14], which may be released into the circulation. Both phenomena may exert systemic effects. In particular, MPOx release from leukocytes, which in our study is reflected by reduced neutrophil content of the enzyme, results in an array of diffusible oxidants [15] that harm the endothelium [16]. In addition, MPOx uses nitric oxide as a physiologic substrate, thereby reducing nitric oxide bioavailability [17] and contributing to endothelial dysfunction [18]. Because also IL-6 may affect the endothelium [19], it is not surprising that, in CAD + PAD patients under resting conditions, the greater the transfemoral V–A difference in IL-6 and in neutrophil MPOx content (and, thus, presumably the greater the release of MPOx from activated neutrophils), the more severe the endothelial dysfunction in the coronary bed. Although correlation coefficients provide no information on cause and effect, numerous findings support the hypothesis that peripheral inflammation contributes to coronary artery endothelial dysfunction in these patients. First, linear logistic regression analysis showed that the MPOx gradient across the affected femoral circulation remained associated with worse coronary endothelial function, after adjustment for CAD severity and cardiovascular treatments. Second, coronary vasoreactivity was not related to the transcoronary gradient of inflammatory markers, which was close to zero, thus excluding that intra-coronary vascular inflammation affected CAEF. Third, we found no relationship between PAD severity measured as ABI, number of coronary stenoses >50% and type C coronary lesions. This, in addition to the fact that ABI did not correlate with peripheral inflammation excludes that PAD severity influences CAEF. Fourth, serum from the claudicant limb of CAD + PAD patients induced a proinflammatory state in HCAECs in vitro. HCAECs released more MCP-1 when exposed to serum from the affected limb than when exposed to serum from the aorta of the same patients. This difference disappeared when HCAECs were incubated with serum from healthy legs or the aorta of CAD-alone patients. Therefore, it is reasonable to assume that the blood leaving the affected limb of PAD patients contains substances that affect the endothelial cells of other vascular districts. Lastly, our post-exercise data support the concept of a mechanistic link between peripheral vascular inflammation and CAEF. In fact, the MPOx/coronary dysfunction association was greater
445
after exercise, and thus CAD + PAD patients who with exercise had the greatest transfemoral reduction in neutrophil MPOx content had the greatest exercise-induced endothelial dysfunction. Not less important is the observation that this post-exercise association was not observed for IL-6, the transfemoral levels of which did not increase in the exercising affected limb. This suggests that MPOx and IL-6 play different roles in the inflammatory process of the claudicant limb. The same applies to CRP whose concentration in blood leaving the affected limb neared that in aortic blood, thus excluding transfemoral release of this molecule. This finding, which coincides with a recent study of coronary circulation [20], suggests a systemic origin of CRP. Notably, the transfemoral concentration of inflammatory markers was unrelated to CAEF in CAD-alone patients, who did not have any inflammatory reaction in the exercising healthy leg.
5. Study limitations Due to the complexity of the protocol, many patients refused to participate in the study, and consequently the CAD + PAD and CAD-alone groups are relatively small. Therefore, the results of this hypothesis-generating study need to be confirmed by larger trials. Secondly, although treatments were discontinued at least 18 h before the study, we cannot exclude that some drugs may have reduced the inflammatory response in the affected limb and/or beneficially influenced CAEF. Furthermore, because of the broad spectrum of drugs and doses administrated, we were unable to examine the interaction between treatment status, inflammation and CAEF. However, multivariate analysis showed that treatments did not affect the significant association between inflammatory status of the claudicant limb and CAEF.
6. Conclusions This study points to a causal link between peripheral inflammation in the affected limb of CAD + PAD patients and coronary artery endothelial dysfunction. Our findings that the exercise-induced increase in the inflammatory response of the claudicant limb caused a further deterioration of CAEF, and that serum from the affected limb of PAD patients exerted proinflammatory effects on HCAECs (a phenomenon not observed with serum from the healthy leg of CAD-alone patients) suggest the existence of circulatory triggers in PAD that could activate the endothelium at distant sites. Should this be the case, PAD, besides being a marker of cardiovascular risk, could exert a mechanistic function in the progression of atherosclerosis in coronary arteries. Should larger, prospective studies confirm this concept by demonstrating that peripheral vascular inflammation portends a higher risk of coronary events, affected individuals would be candidates for more aggressive standard therapy or for emerging therapies aimed at reducing local intra-arterial inflammation.
446
G. Brevetti et al. / Atherosclerosis 201 (2008) 440–446
References [1] Brevetti G, Piscione F, Silvestro A, et al. Increased inflammatory status and higher prevalence of three-vessel coronary artery disease in patients with concomitant coronary and peripheral atherosclerosis. Thromb Haemost 2003;89:1058–63. [2] Papamichael CM, Lekakis JP, Stamatelopoulos KS, et al. Anklebrachial index as a predictor of the extent of coronary atherosclerosis and cardiovascular events in patients with coronary artery disease. Am J Cardiol 2000;86:615–8. [3] Erren M, Reinecke H, Junker R, et al. Systemic inflammatory parameters in patients with atherosclerosis of the coronary and peripheral arteries. Arterioscler Thromb Vasc Biol 1999;19:2355–63. [4] Libby P, Ridker PM, Maseri A. Inflammation and atherosclerosis. Circulation 2002;105:1135–43. [5] Zouridakis E, Avanzas P, Arroyo-Espliguero R, Fredericks S, Kaski JC. Markers of inflammation and rapid coronary artery disease progression in patients with stable angina pectoris. Circulation 2004;110: 1747–53. [6] Tataru MC, Heinrich J, Junker R, et al. C-reactive protein and the severity of atherosclerosis in myocardial infarction patients with stable angina pectoris. Eur Heart J 2000;21:1000–8. [7] Narins CR, Zareba W, Moss AJ, et al. Relationship between intermittent claudication, inflammation, thrombosis, and recurrent cardiac events among survivors of myocardial infarction. Arch Intern Med 2004;164:440–6. [8] Edwards AT, Blann AD, Suarez-Mendez VJ, Lardi AM, McCollum CN. Systemic responses in patients with intermittent claudication after treadmill exercise. Br J Surg 1994;81:1738–41. [9] Neumann FJ, Waas W, Diehm C, et al. Activation and decreased deformability of neutrophils after intermittent claudication. Circulation 1990;82:922–9.
[10] Buffon A, Biasucci LM, Liuzzo G, et al. Widespread coronary inflammation in unstable angina. N Engl J Med 2002;347:5–12. [11] Fowkes FGR, Allan PL, Tsampoulas C, Smith FB, Donnan PT. Validity of duplex scanning in the d´etection of peripheral arterial disease in the general population. Eur J Vasc Surg 1992;6:31–5. [12] Pasterkamp G, Schoneveld AH, van der Wal AC, et al. Inflammation of the atherosclerotic cap and shoulder of the plaque is a common and locally observed feature in unruptured plaques of femoral and coronary arteries. Arterioscler Thromb Vasc Biol 1999;19:54–8. [13] Vink A, Schoneveld AH, Richard W, et al. Plaque burden, arterial remodeling and plaque vulnerability: determined by systemic factors? J Am Coll Cardiol 2001;38:718–23. [14] Schieffer B, Schieffer E, Hilfiker-Kleiner D, et al. Expression of angiotensin II and interleukin 6 in human coronary atherosclerotic plaques: potential implications for inflammation and plaque instability. Circulation 2000;101:1372–8. [15] Podrez EA, Abu-Soud HM, Hazen SL. Myeloperoxidase-generated oxidants and atherosclerosis. Free Radic Biol Med 2000;28:1717–25. [16] Muller G, Goettsch C, Morawietz H. Oxidative stress and endothelial dysfunction. Hamostaseologie 2007;27:5–12. [17] Eiserich JP, Baldus S, Brennan ML, et al. Myeloperoxidase, a leukocyte-derived vascular NO oxidase. Science 2002;296:2391–4. [18] Vita JA, Brennan ML, Gokce N, et al. Serum myeloperoxidase levels independently predict endothelial dysfunction in humans. Circulation 2004;110:1134–9. [19] Wassmann S, Stumpf M, Strehlow K, et al. Interleukin-6 induces oxidative stress and endothelial dysfunction by overexpression of the angiotensin II type 1 receptor. Circ Res 2004;94:534–41. [20] Maier W, Altwegg LA, Corti R, et al. Inflammatory markers at the site of ruptured plaque in acute myocardial infarction: locally increased interleukin-6 and serum amyloid A but decreased C-reactive protein. Circulation 2005;111:1355–61.