The reduction in plaque volume during stent implantation is associated with the release of particulate debris and plaque-derived soluble substances. We studied the potential release of the proinflammatory cytokine TNF-α into the coronary circulation and whether such release is related to the reduction in plaque volume and, possibly, a predictor for restenosis. In 18 male patients (n = 24 stents) with severe stenosis in a saphenous vein aortocoronary bypass graft (SVG), we used a distal balloon occlusion-aspiration device during stent implantation. The aspirate TNF-α levels were determined before and after stent implantation and related to the angiographic and intravascular ultrasound-assessed severity of stenosis and restenosis. We found that TNF-α is, indeed, released into the aspirate of stented SVG (9 ± 1 and 28 ± 3 pg/ml before and after stent implantation, respectively, P < 0.0001) and that such release is related to the reduction in plaque volume (r = 0.88, P < 0.0001) and associated with restenosis after 5 mo (r = 0.71, P = 0.001). The periprocedural release of plaque-derived TNF-α possibly represents the amount and activity of the atherosclerotic process and might be a predictor for restenosis.
- plaque activity
the proinflammatory cytokine TNF-α has been localized in human atheromatous plaques (1, 24, 28), and it contributes to plaque progression, destabilization, and rupture (25, 27), as well as progression of restenosis (11, 13, 15).
The reduction in plaque volume during coronary stent implantation involves plaque compression, axial redistribution, and/or disruption (19, 26). The reduction in plaque volume is also associated with the release of particulate debris, as well as soluble vasomotor and thrombogenic substances, into the coronary circulation (7, 10). Recently, using a distal balloon protection device, which traps and aspirates particulate debris as well as soluble substances during stent implantation, we identified serotonin and thromboxane A2 as potent vasoconstrictors in the aspirate from atherosclerotic saphenous vein aortocoronary bypass grafts (SVG). Both substances were released in relation to the angiographic severity of the stenosis and to plaque volume as determined by intravascular ultrasound (10).
In the present study, we used the same approach to investigate whether TNF-α is also released into the coronary circulation and whether such release is related to the reduction in plaque volume and might be a predictor for restenosis.
MATERIALS AND METHODS
Eighteen consecutive male patients (recruited between May 2005 and November 2006) with a flow-limiting stenosis in their SVG were studied. Patient characteristics are as follows: 64 ± 2 yr, 27 ± 1 kg body wt/m2 body mass index, 2.2 ± 0.2 (range I–IV) Canadian Cardiovascular Society functional class, and 12 ± 1 (range 1–21) yr graft age.
Patients with age-, sex-, and body surface-corrected glomerular filtration rate <60 ml·min−1·m−2 (2, 11) were excluded from the study. The patients' mean serum creatinine level was 1.2 ± 0.03 mg/dl, and glomerular filtration rate was 70 ± 2 ml·min−1·m−2.
Coronary risk factors (in percentage of patients) were arterial hypertension (100%), hypercholesterolemia (100%), diabetes mellitus (17%), and smoking (39%).
Current medication (in percentage of patients) consisted of β-adrenoceptor antagonists (100%), diuretics (100%), angiotensin-converting enzyme inhibitors (89%), angiotensin II receptor type I antagonists (11%), statins (100%), and antidiabetics (6%).
Patients' individual target grafts and baseline and postinterventional angiographic and intravascular ultrasound (IVUS) data are shown in Table 1. Two consecutive stents were implanted for one lesion in the SVG in patient 13. Two stents were implanted in the SVG in patients 14 and 18. In patients 15 and 17, two independent procedures were performed: first one stent and then two additional stents were implanted in the SVG.
Thirteen patients with a total of 18 stents were angiographically controlled for 5.3 ± 0.4 (range 3–8) mo after stent implantation. Patients' individual angiographic data before, immediately after, and 5 mo after stent implantation are given in Table 2.
The study protocol was approved by the local Institutional Review Board of the University of Essen School of Medicine. Full informed consent was obtained from all patients before participation in the study. The investigation conforms with the principles outlined in the Declaration of Helsinki.
Quantitative coronary angiography.
All patients were on aspirin (100 mg/day) and received 10,000 IU of heparin intravenously during the intervention. Coronary angiography was performed after intracoronary injection of 0.2 mg of nitroglycerin via the femoral approach before and after the intervention, as described previously (6, 10). Minimal lumen diameter (MLD) and reference diameter were determined before, immediately after, and 5 mo after stent implantation, and the percent diameter stenosis and restenosis were calculated from quantitative coronary angiography (6, 10).
All IVUS imaging analyses were performed before and after stent implantation with a commercially available solid-state IVUS catheter (Eagle-Eye, 20 MHz, Volcano Therapeutics, Rancho Cordova, CA) after intracoronary injection of 0.2 mg of nitroglycerin.
The IVUS catheter was withdrawn automatically using a motorized pull-back device at 0.5 mm/s. Catheter withdrawal was started at as distal a location as possible, but ≥10 mm distal to the target lesion. IVUS images of the entire withdrawal were recorded digitally and ECG triggered in the DICOM format for offline analysis. An entire segment analysis was performed offline using a computerized analysis system with automated contour detection and editing (QCU-CMS, Medis, Leiden, The Netherlands). Quantitative analyses were performed according to the American College of Cardiology guidelines (14). The analyzed data were averaged over the individually stented segment. A three-dimensional reconstruction with volume measurements of the treated segment was also performed before and after the intervention (5).
In situ, veins do not have an external elastic membrane. However, arterialized SVG undergo morphological changes (intimal fibrous thickening, medial hypertrophy, and lipid deposition), so that an echo-lucent zone, which has been reported to represent the media, can be identified by IVUS. Therefore, the external elastic membrane area was measured by tracing the outer border of this echo-lucent zone (5, 29). Plaque area (− volume) was calculated as external elastic membrane area − lumen area (− volume). Patients' individual change in plaque volume was calculated as plaque volume before the intervention − plaque volume after the intervention.
Balloon-expandable stents [11 bare metal stents and 11 paclitaxel-eluting TAXUS stents (Boston Scientific, Natick, MA) and 2 sirolimus-eluting Cypher stents (Johnson & Johnson, New Brunswick, NJ)] were implanted via the femoral approach using maximum balloon pressures of ≥14 atm and a balloon-to-vessel ratio of 1.1. To prevent microembolization, a distal balloon occlusion device (TriAktiv, Kensey Nash, Exton, PA) was mounted on a guide wire and inflated at 2 atm with CO2 during stent implantation. Plaque debris was aspirated via the guiding catheter after advancement of a special flushing catheter proximal to the inflated wire balloon. During slow withdrawal of the flushing catheter, the stented area was washed with saline. After the extraction process, the balloon was deflated, and a postinterventional angiography was performed.
Blood samples and determination of plasma TNF-α.
Before intervention, arterial blood was collected distal to the stented lesion via the flushing catheter (20 ml; into potassium-EDTA S-Monovette) and compared with the aspirate that was flushed out after stent implantation, resulting in an approximately twofold dilution with saline, and then filtered through a 40-μm mesh filter. In each instance, visible particulate debris was retained on the filter. Arterial blood and filtered aspirate were immediately centrifuged (600 g for 10 min at 4°C), and the plasma was removed, quickly frozen in liquid nitrogen, and stored at −20°C until further use. Plasma TNF-α was determined using a solid-phase, chemiluminescent immunometric assay (IMMULITE/IMMULITE 1000 TNFα, DPC, Gwynedd, UK). The individual increase in plasma TNF-α (ΔTNF-α in pg/ml) was calculated for each patient.
Values are means ± SE. Statistical differences were assessed with paired (angiographic IVUS and TNF-α levels before vs. after stent implantation) or unpaired (bare metal vs.drug-eluting stent and <50% vs. >50% restenosis) two-sided Student's t-tests. Linear correlations were calculated, and all statistics were done with the Prism program (Graph-Pad Software, San Diego, CA). P < 0.05 was considered significant.
Baseline and postinterventional angiographic and IVUS data.
Before stent implantation, the diameter of the stenosis of the SVG was 63 ± 1% and plaque volume was 223 ± 27 mm3 (n = 24 stents). Immediately after stent implantation, the diameter of the stenosis was reduced to 4 ± 1% and plaque volume to 183 ± 24 mm3 (n = 24 stents, each P < 0.0001 vs. before stent implantation). The reduction in diameter of the stenosis was related to baseline diameter of the stenosis (r = 0.76, P < 0.0001, n = 24 stents), and the reduction in plaque volume was related to baseline plaque volume (r = 0.56, P = 0.0041, n = 24 stents).
Aspirate TNF-α before and after stent implantation and relation to plaque extrusion.
In each patient, TNF-α in the aspirate increased after stent implantation: from 9 ± 1 pg/ml before to 28 ± 3 pg/ml stent implantation (n = 24 stents, P < 0.0001; Fig. 1A). The increase in TNF-α (ΔTNF-α) correlated with baseline plaque volume (r = 0.44, P = 0.0319, n = 24 stents) and with the reduction in plaque volume (Fig. 1B).
We found no relations between ΔTNF-α and reduction in the diameter of the stenosis, stent length, mechanical deformation (i.e., increase in vessel cross-sectional area after stent implantation), or graft age.
Periprocedural increase in aspirate TNF-α in relation to restenosis.
MLD was reduced from 3.58 ± 0.14 mm immediately after to 2.28 ± 0.34 mm at 5 mo after stent implantation (P < 0.0001, n = 18 stents); i.e., restenosis averaged 38 ± 9% (n = 18 stents). The patients' individual restenosis correlated significantly with the periprocedural increase in TNF-α (n = 18 stents; Fig. 1C).
The increase in the aspirate TNF-α immediately after stent implantation was 1.9-fold greater in patients with >50% restenosis (ΔTNF-α = 33 ± 8 pg/ml, n = 6 stents) than in patients with <50% restenosis (ΔTNF-α = 17 ± 3 pg/ml, n = 12 stents, P = 0.0355); however, baseline TNF-α in patients with >50% restenosis (10 ± 2 pg/ml, n = 6 stents) was not different from that in patients with <50% restenosis (9 ± 2 pg/ml, n = 12 stents).
The degree of restenosis was independent of implantation of a drug-eluting stent (36 ± 14% with ΔTNF-α = 19 ± 4 pg/ml, n = 8 stents) or a bare metal stent (39 ± 12% with ΔTNF-α = 24 ± 5 pg/ml, n = 10 stents).
In native coronary vessels, TNF-α contributes to the progression from stable to unstable plaques by augmenting the local inflammatory response via multiple effects (11, 19, 25, 27). TNF-α-induced effects include the release of cell surface adhesion molecules (13), the increase in thrombotic activity (21), and the induction of endothelial and smooth muscle cell apoptosis by increased formation of reactive oxygen species within the plaque (3). TNF-α also causes plaque instability by increasing the level of matrix-degrading enzymes, such as matrix metalloproteinases (4). Increased levels of TNF-α with plaque progression are seen predominantly in the cytoplasm and attached to the plasma membrane of macrophages and endothelial and smooth muscle cells (1, 17). TNF-α contributes to coronary endothelial dysfunction in ischemia-reperfusion (30) and also mediates contractile dysfunction after embolization of plaque debris into the coronary microcirculation (23).
In contrast to native coronary artery lesions, SVG plaques contain more foam and inflammatory cells and a poorly developed or absent fibrous cap with relatively low calcium content. Therefore, SVG plaques are more fragile and prone to rupture (16, 18), but the relation of plaque vulnerability to TNF-α is not clear. In the present study, we have demonstrated that TNF-α is indeed released into the aspirate of stented SVG and that this release is related to the reduction in plaque volume and to restenosis.
The correlation of periprocedural TNF-α release to the reduction in plaque volume suggests that TNF-α simply reflects the amount of atherosclerosis. On the other hand, the relation of periprocedural TNF-α release to restenosis suggests that TNF-α also reflects the activity of the atherosclerotic process. In this respect, the released TNF-α may be a mirror of the TNF-α retained in the vascular wall. Since we found no relation of periprocedural TNF-α release to stent length or increase in vessel cross-sectional area after stent implantation, mechanical injury per se is probably not responsible for TNF-α release (8) and restenosis (20).
These findings confirm and extend the observation that restenosis is not a random event but, rather, affects selectively a subset of patients, possibly those with high intraplaque TNF-α levels before stent implantation. TNF-α has previously been suggested to be a predictor for restenosis (9). However, in this prior study, the preprocedural peripheral blood levels of TNF-α reflecting systemic inflammation in patients with stable angina and a small group of patients with unstable angina predicted the rate of restenosis of native coronary artery lesions. In contrast, in our present study, the periprocedural increase in TNF-α at the site of the stented SVG lesion predicted the degree of that individual stenosis 5 mo later, thus establishing an even firmer cause-effect relation. In those patients, the neutralization of TNF-α with locally applied TNF-α antibodies or the use of TNF-α antibody-eluting stents could be promising (8, 22).
Our data are limited to a small number of patients. Future studies in a larger cohort of patients are needed to confirm the potential prognostic value of periprocedural TNF-α release for restenosis, and threshold values need to be determined by receiver-operator characteristics and then analyzed as predictors in a multivariate approach.
Also, SVG plaques are particularly vulnerable to rupture; therefore, care must be taken in extrapolation of our findings to native coronary plaques and their possibly subclinical rupture in daily life. Nevertheless, aspiration of plaque material and soluble factors during stent implantation in SVG is an ethical approach to study pathogenic factors under controlled conditions.
This work was supported by German Research Foundation Grant HE 1320/14-1.
↵* D. Böse and K. Leineweber contributed equally to this work.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 by the American Physiological Society