Although innate immunity is designed to initiate healing from vascular damage, an excess innate immune response is also proven to accelerate atherosclerotic plaque growth [1,2]. More recent work has reportedendothelial dysfunction and inflammatory cell invasion at sites of coronary stent implant, causing increased risk of stent thrombosis [3-7]. Endothelial dysfunction and inflammation are detected with bare metal stents (BMS) and drug eluting stents (DES) with late stent thrombosis. Markers for dysregulation of inflammatory and thrombotic pathways are increased in patients athigher risk of restenosis or thrombotic occlusion [8,9]. DESrelease drugs that block cell proliferation, and potentially restenosis and inflammation, e.g. rapamycin and paclitaxel, but the foreign polymer and the metal in the stent canalso interfere withendothelial regrowth that protects against excess inflammation and thrombosis. This is proposed asa cause for late restenosis and also thrombotic occlusions in DES and BMS. Acute stent thrombosis has high associated morbidity and mortality. While the advent of BMS reduced the incidence of restenosis when compared to balloon angioplasty and coated DES stents further reduced restenosis, both types of stents are associated with persistent risk of restenosis and thrombosis, particularly in the absence of adequate antiplatelet therapy [10].

Monocytes and lymphocytes are attracted to sites of vascular damage by cytokines and then migrate along chemokine gradients formed through attachment of chemokinestoglycosaminoglycans (GAGs) at cellular surfaces and in connective tissue layers [11-13]. Chemokines are small 8-12 kDa chemoattractant proteins, classified as C, CC, CXC and CX3C, based upon C terminal sequences. Increased levels of chemokines are seen at sites of vascular injury and mice deficient in selected chemokines (e.g. monocyte chemoattractant protein, MCP-1) and receptors (CCR2) exhibit reduced plaque growth after injury or in transplants.

Inflammatory cell invasion requires access through tissue connective tissues, necessitating breakdown of collagen and elastin through matrix degrading enzymes (matrix metalloproteinases or MMPs), that are activated by serine proteases. The thrombolytic urokinaseand tissue-type plasminogen activators (uPA and tPA, respectively) as well as plasmin, all coordinately activate pro-MMPs leading to connective tissue breakdown and thus orchestrate cell invasion into damaged vasculature [14-17]. Circulating chemokines, e.g. monocyte chemoattractant protein-1 (MCP-1, MIP1α, CCR2, CX3CR1) and also the serine (uPA, tPA) and the matrix metalloproteases (MMP2, MMP9) and the serpin plasminogen activator inhibitor-1 (PAI-1) [9,10] are increased in unstable angina, non ST elevation MI (NSTEMI) and immediately after stent implant [18,19]. PAI-1 is seen in both increased and decreased restenosis [19,20]. tPA, uPA, uPA receptor (uPAR), and PAI-1 are generally elevated after injury and local overexpression of uPA accelerates plaque growth in rabbit carotids [14-17,19,21]. Mice lacking uPA undergo reduced levels of plaque growth after vascular injury [22].

However, whether the chemokine and serine protease pathways have direct causative impact on stent restenosisis not defined, nor is it known which pathways have a greater role in accelerating inflammatory cell invasion and restenosis. Viruses have developed highly active inhibitors that target both uPA/uPAR and chemokine pathways. These inflammatory moderators function at femptomolar to picomolar concentrations when secreted by the originating viruses [23,24]. Two such proteins, Serp-1 that inhibits uPA/uPAR [23-28] and M-T7 [26,29-31] that blocks the C terminal GAG binding domain of C, CC and CXC class chemokines [26,29-31] have been purified individually and studied as anti-inflammatory reagents in a wide range of animal models with demonstrated reduction in inflammatory cell invasion and plaque growth. One protein, Serp-1 has been successfully tested in clinical trial for treatment of patients with unstable coronary syndromes (Unstable angina and NSTEMI) who have received stent implants [26,32]. Serp-1 infusions, administered once a day for three days starting immediately after stent implant, significantly reduced markers of myocardial damagewith variable reductions in inflammatory and clotting (D dimer) [26]. These studies were, however, not sufficiently powered to evaluate potential effects on long term plaque growth.

We report here a study examining and comparing the effects of anti-inflammatory protein treatments with Serp-1 and M-T7, given by intravenous injection, on restenosis and local stent inflammatory cell invasion in cholesterol fed rabbits.

Rabbit arterial injury and stenting

All surgical procedures were approved by the University of Western Ontario/Robarts Research Inst. lab animal care committee and all procedures conformed to University and national guidelines for laboratory animal surgery and research. One hundred and forty one New Zealand White rabbits (weight: 3.1 ± 0.04 kg; mean ± SEM) were fed a cholesterol rich diet (0.5% cholesterol in 6% peanut oil) for two weeks. Femoral arteriotomy was performed under general anaesthetic (40 mg/kg ketalean, 8 mg/kg xylazene and 0.5 mg/kg acepromazine givenby intramuscular injection; Vetrepharm Canada, Inc., London, ON, Canada) and a 4Fr arterial sheath was placed in the femoral artery (30, ref Lucas Circ 1996 and ChristovKorolPhotPhotbiol). Heparin 400 units were given via the arterial sheath. A 3.0-3.5 mm diameter balloon angioplasty catheter was introduced via the femoral arteriotomy and inflated to nominal pressure (8 atmospheres) in the distal abdominal aorta. Vascular injury was produced by advancing the inflated balloon through the abdominal aorta up to the distal thoracic aorta and withdrawing it three times. Stainless steel stents (AVE (Medtronic, Minneapolis MN), ACS (Abbott, Abbot Park, IL), or NIR Niroyal (Boston Scientific Corporation, Natick, MA) were then placed in the distal abdominal aorta and deployed at 14 atmospheres (Table 1). Catheter and arterial sheath were removed and arteriotomy closed. Rabbits were divided into groups receiving either one day, two days, three days or 11 days of treatment with one of three treatment regimens: saline control 1 ml/day (n=39), Serp-1 at 0.05 μg/kg, 5 μg/kg, up to 50 μg/kg/day (n = 63), and M-T7 at 10 μg/kg/daily (n = 38) for one to three daily injections and alternate daysfor 11 day injections (Table 1). M-T7 supplies were smaller than those available for Serp-1 treatments. Injections were initially given immediately after stent implant by intraarterial injection through the catheter lumen. All subsequent injections were given by intravenous injections via ear vein. All rabbits received Aspirin 40 mg per day orally for 10 days. Cholesterol feeding was continued for another two weeks and rabbits were sacrificed at four weeks following stent implant. Aortas were dissected and specimens obtained for analysis. Sections from each stent specimen were made at two to three levels, fixed and stained with haematoxylin and eosin.

Table 1: Rabbit Abdominal Aortic Stent Implant Model.

Morphometric analysis

Harvested arterial sections were processed and stained with hematoxylin and eosin (H & E) or Masson’s trichrome stains as previously described [25-31,33]. Plaque area, was measured by morphometric analysis using an Olympus CCD color video camera attached to an Olympus microscope, and the ImagePro application program calibrated to the microscope objective [26-31,33]. Both the mean total cross-sectional intimal area and the mean intimal thickness after normalized to the medial thickness were calculated for each arterial section. Specimens isolated from rabbits were cut into three sections and two histology cross sections stained per each of the three artery specimens (6 sections per specimen).

Cross sectional area at each level was measured for three parameters: reference vessel area, measured as area inside of the internal elastic lamina; lumen area, measured as area inside of the intima and plaque area measured as area between the internal elastic lamina and the intima. Additionally, plaque thickness was measured at the point of maximum thickness in each section and a mean value calculated for each animal from the measurements for all available sections for that animal. A separate analysis was performed by two of the investigators blinded to treatment. For each animal, a mean plaque area was calculated for all the sections available for that animal. Mean values for all animals are reported for the proximal segments, at the segment with maximum plaque growth and for the mean plaque area in each animal.

Serp-1 and M-T7 expression and purification

Purified Serp-1 protein was provided by Viron Therapeutics Inc. (London, ON, Canada). In brief, Serp-1 was purified from the supernatant of recombinant Chinese hamster ovary (CHO) cell lineby sequential column chromatographic separation as previously described [27,28,32,33]. Serp-1 purity was over 95% as determined by overloaded Coomassie-stained SDS-PAGE gels and reverse-phase HPLC.

M-T7 was isolated from baby green monkey kidney (BGMK) cells infected with vaccinia vector expressing M-T7. M-T1 from vMyxlac- T1gpt infected BGMK cell cultures as previously described [26,29,31]. Concentrated supernatants were fractionated by Mono QHR5/5 anion exchange chromatography followed by gel filtration chromatography using a HiLoadSuperdex 200 column (AmershamPahrmacia Biotech Inc.).

Statistical methods

Data are presented as mean ± standard error of the mean (SEM), as previously described. Mean measurements for each experimental animal for all morphometric data were assessed for significant changes by analysis of variance (ANOVA) with FLSD (Fisher least squares post hoc) secondary analysis. Mortality was assessed by Chi-square analysis.

Effects of anti-inflammatory treatments on adverse events after balloon angioplasty and stent implant

Although not significant, there was an increase in adverse events in saline treated animals when compared to all M-T7 or Serp-1 treated rabbits after stent implant. For all treated animals the surgical mortality numbers were as follows; for Saline control treatment there were 9 deaths (23.1%) with 2 thromboses, 4 bleeds, 1 peritonitis, 1 traumatic leg fracture and one death for undetermined cause; forSerp-1 treatment there were 5 deaths with 3 thromboses, 1 stent fracture (bend), and 1 paralysis (7.9%); and for M-T7 treatment there was1 death with 1 thrombosis (2.6%). By Chi square analysis for all adverse events there was a significant decrease in overall adverse events for Serp-1 compared to saline (P ≤ 0.0334) and reduced thrombosis events for Serp-1 treatment when compared to Saline treatment (P ≤ 0.0071). There was no hemorrhage detected with Serp-1 treatments. For M-T7 treatments, there was only one adverse event, a thrombotic occlusion of the artery after stent implant with an overall significant reduction in adverse events with M-T7 treatments after angioplasty and stent implant (P ≤ 0.0076 by Chi square analysis). There was no significant difference detected for adverse events when comparing all Serp-1 and M-T7 treated animals (P = 0.276 for all adverse events and P = 0.7194 for combined thrombotic and hemorrhagic events). There was 1 paralysis that occurred due to inappropriate securing of the rabbit during an injection resulting in a spine fracture that was unrelated to surgery and is not included in the statistical analyses.

Inhibition of early inflammatory responses reduces late instent plaque growth- Histological analysis

Both M-T7 and Serp-1 significantly reduced plaque growth both inside the arterial area covered by the stent and outside the arterial areas covered by stents in areas of angioplasty injury alone. Large, lipid-laden plaques were detected at sites of stent implant with large numbers of invading inflammatory mononuclear cells (Figure 1A and B). M-T7 reduced plaque growth at doses of 10 μg/kg/d (~30 μg per 3 Kg rabbit body weight) (Figure 1C and D) given at the time of stent implant. Serp-1 also produced significant reductions in plaque size with daily injections at doses of 50 μg/kg/d (150 μg) and higher (Figure 1E and F). When rabbits were treated with either Serp-1 or M-T7 there was a reduction in the plaque size as well as in the local invading inflammatory cell infiltrates, e.g. reduced foam cell macrophage invasion at sites of stent strut implants embedded in the artery wall (Figure 1B,D and F). Serp-1 was tested for dose response with two differing dose levels for rabbits with two and with three daily injections. A clear dose response is detected for either two or three daily injections indicating that the reduction in plaque area is not a random response to injection of a foreign protein. It should also be noted that in prior studies the use of mutagenized inactive Serp1- protein has not produced significant reductions in plaque growth.

Figure 1: Hematoxylin and eosin stained cross sections of rabbit aorta at sites of stent implant at 4 weeks follow up. Marked inflammatory plaque is detected with saline treatment 4 weeks after balloon angioplasty and stent implant. M-T7 and Serp-1 markedly reduce visible plaque in cross sections at 4 weeks follow up. Saline treated rabbit with 11 daily injections - A - Mag 100X, B- Mag 400X. M-T7 treated rabbits with 11 daily injections - C - Mag 100X, D- Mag 400X. Serp-1 treated rabbits with 11 daily injections - E- Mag 100X, F- Mag 400X.

Quantitative analysis

Treatment with either M-T7 or Serp-1 for two or more days after balloon angioplasty and stent implant demonstrated a significant reduction in mean plaque area (Figure 2B-D) and percentage lumen narrowing (Figure 3B-D) for stented aortic sections. Overall there was greater than 53% reduction for the treatment groups, with greater than two days treatment, when compared to the saline treated control group. Although demonstrating a trend, neither M-T7 (P = 0.112) nor Serp-1 (P = 0.476) treatment given for one day only as a single intravenous bolus infusion after balloon angioplasty and stent implant, did not reduce plaque growth significantly when measured either as plaque area (Figure 2A, P = NS) or as percentage lumen narrowing (Figure 3A).The mean plaque area was similar at the three stent segments in all groups.

Figure 2: Morphometric analysis of mean plaque area at sites of stent implant demonstrated reduced plaque areas with greater than two injections of M-T7 and with greater than three injections of Serp-1 at 4 weeks follow up. A dose response is seen with Serp-1 given for three daily doses at three differing doses of Serp-1. A. Single day injection, B. Two daily injections, C. Three daily injections, D. 11 daily injections.

Figure 3: Morphometric analysis of meanlumen percentage narrowing at sites of stent implant demonstrated reduced plaque areas with greater than two injections of M-T7 and with greater than three injections of Serp-1 at 4 weeks follow up. A dose response is seen with Serp-1 given for three daily doses at three differing doses of Serp-1. A. Single day injection, B. Two daily injections, C. Three daily injections, D. 11 daily injections.

Overall M-T7 and Serp-1 produced comparable reductions in plaque growth after balloon angioplasty and stent implant when given as 2 day (Figure 2B and 3B; P<0.005 and <0.004 for M-T7 and Serp-1 for plaque area and P<0.046 and P<0.011for M-T7 and Serp-1 treatment respectively for percentage lumen narrowing), 3 day (Figure 2C and 3C; P<0.045 and P<0.0003 for M-T7 and Serp-1 for plaque area and P<0.001 and P<0.0002 for M-T7 and Serp-1 treatment respectively for percentage lumen narrowing) or 11 day (Figure 2D and 3D; P<0.0003 and P<0.001 for M-T7 and Serp-1 for plaque area and P<0.0001 and P<0.0003 for M-T7 and Serp-1 treatment, respectively for percentage lumen narrowing) injections starting at the day of stent implant. However with a single injection given at the time of stent implant neither M-T7 nor Serp-1 was able to produce a long term reduction in plaque size although M-T7 treatment showed a trend toward reduced plaque growth (Figure 1A and 2A; P=0.112 and P=0.476, for M-T7 and Serp-1 treatment, respectively for plaque area and P=0.235, P=0.987 for M-T7 and Serp-1 treatment, respectively for percentage lumen narrowing).

Measures of intimal to medial thickness ratios, which are used to reduce variations inherent in differing aortic sizes due to natural biological variability, displayed the same significant changes in plaque size with each anti-inflammatory protein treatment. However, in contrast to other measures of plaque growth M-T7 produced a significant reduction in plaque thickness after a single injection given at the time of stent implant (Figure 4; p<0.0001)). Serp-1 did not reduce intimal to medial thickness ratios with a single bolus injection given at the time of stent implant (Figure 4).

Figure 4: Intimal/medial thickness ratios demonstrated reduced intimal thickness with two or more daily doses of M-T7 and with three or more daily doses of Serp-1 (P < 0.0001).

In summary both Serp-1 and M-T7 treatments given for two or more days were capable of reducing plaque growth in the rabbit aorta at 4 weeks follow up after balloon angioplasty and stent implant. M-T7 by some measures (specifically intimal to medial thickness ratios) reduced plaque size when given at lower doses and with alternate day injections for treatment groups. Thus M-T7 may be more effective at blocking plaque growth long term after stent implant.

Plaque area according to stent length and type

Stents used in these studies were donated by the companies and varied somewhat in length and diameter (Table 1). Thus an attempt was made to ensure that the same stent brands (manufacturing styles of stents) were used for each treatment and dosing groups. A similar range of stent lengths was used for each treatment and dosing groups. The mean reference stent size and length at deployment are summarized in Table 1. There were no statistically significant differences between the three groups in these parameters for the dosing treatment groups.

Plaque area, when plotted against stent length for all protein treatment groups, did not show a significant correlation (R²=0.0715, P=NS). There was a negative correlation in the saline group only (R2 =0.454, P=0.03, Saline group only). There was no significant difference in the mean plaque area between different stent types within any of the groups. However, plaque area was greater for Niroyal stents in the saline treatment group only as compared to the Serp-1 and M-T7 treated rabbits, p=0.046, treatment versus saline.

Mononuclear cellular invasion

Invading inflammatory mononuclear cells (macrophage and T cells) in the intimal, medial, and adventitial layers were counted and cell count compared per treatment group at 4 weeks follow up (Figure 5). As for the comparisons of plaque size, both M-T7 and Serp-1 significantly reduced numbers of invading mononuclear cells in all three layers with greater effects on intimal and adventitial layers. Intimal and adventitial cell counts were significantly reduced for both M-T7 and Serp-1 treatments given for more than three daily treatments (Figure 5). Of interest, M-T7 reduced intimal mononuclear cell numbers after a two daily doses (P<0.029), while Serp-1 (P=0.086) was ineffective with two doses. However both M-T7 (P<0.0001) and Serp-1 (P<0.0001) reduced invading mononuclear cell counts in the adventitial layers.

Figure 5: Mean cell counts for intimal and adventitial layer invading mononuclear cells at sites of stent implant at 4 weeks follow up. Cell counts measured in three high power fields using sites with the largest numbers of invading cells for each count. A. Intimal cell count one day dosing, B. Adventitial cell count one day dosing, C. Intimal cell count two day dosing, D. Adventitial cell count two day dosing, E. Intimal cell count three days dosing, F. Adventitial cell count three days dosing, G. Intimal cell count 11 days dosing, B. Adventitial cell count 11 days dosing.

While advances in medical therapy have been made in the prevention and treatment of native vascular atheroma development and acute thrombosis in unstable angina specifically with lipid lowering agents and anti-platelet oranti-thrombotic agents, treatment for restenosis remains limited. Acute thrombosis is reduced with dual antiplatelet therapy and acute post stent infarction may be reduced with high dose statin therapy [34]. However the most effective treatment, proven to reduce restenosis after angioplasty or stent implant is the use of bare metal stents when compared to balloon angioplasty alone or the use of drug eluting stents when compared to balloon angioplastyor BMS [7-10,35-37]. Multiple studies report correlations between acute and chronic inflammatory cell invasion and activation with restenosis as well as later stent thrombosis [7-10,38,39]. Agents targeting inflammatory reactions have been reported in preclinical animal studies and small clinical trials to alter restenosis after stent implant but this work remains investigational [38-43].

Acute thrombosis of stents remains another significant problem, necessitating prolonged use of dual anti-platelet agents. DES has markedly improved risk of restenosis from20-30% with BMS to less than 10% with DES, but there are many clinical contraindications for prolonged use of dual anti-platelet therapy. Longer term dual antiplatelet treatment is recommended for DES implant, than for BMS, due to the risk of late stent thrombosis. This need for longer termantiplatelet treatments influences clinical choices between the use of BMS or DES when patients have a history of recent severe bleeding or have planned surgical procedures. Stenosis and thrombotic occlusions of vein grafts is also frequently seen, up to 50% of vein grafts in coronary artery bypass graft (CABG) patients are occluded within 10 years and or these occlusions 80% occur in the first two years [40,41]. There is a high risk of recurrent graft occlusion after angioplasty and stent implant. Some patients present technically challenging tortuous arteries that necessitate the use of balloon angioplasty alone, e.g. stent implant is not technically possible. Without stent implant the risk of restenosis for balloon angioplasty alone is 30-50%, much higher than with stent implants. The potential to reduce restenosis through antiinflammatory agent treatment might allow for improved outcomes in patients where balloon angioplasty alone or BMS is necessary. Further the use of anti-inflammatory agents might have the potential to also reduce risk of late stent thrombosis.

A significant reduction in stent thrombosis was demonstrated with treatment with Serp-1 (P ≤ 0.0334 for all adverse events; P ≤ 0.0071 for thrombosis and hemorrhage) and M-T7 (P<0.0076 for all adverse events; P ≤ 00516). There was no difference in adverse events when comparing Serp-1 and M-T7 treatments (P = 0.2746 for all adverse events; P = 0.7194 for thrombotic/ hemorrhagic events).This reduction in adverse events may have been due to the anti-inflammatory activity of each protein and is consistent with prior studies demonstrating reduced adverse events with non-specific inhibitors of inflammation such as dexamethasone coated stent implants, statin coated stents or ghrel in peptide treatment [37,38,42,43]. Serp-1 may also have reduced thrombosis and/or hemorrhage through inhibition of fXa and the plasminogen activators (tPA and uPA), respectively. Excess bleeding was not seen after treatment with either agent. The fact that there was no significant difference in adverse events and specifically thrombotic and hemorrhagic events would support prior studies form other researchers that have linked excess inflammation in stent implant areas with later stent thrombosis [7-10,35-43]. Specific or selective inflammatory cell responses werenot examined here and remain to be determined. Prior work has demonstrated that there is modification of both macrophage and T cell responses in animal models of angioplasty or transplant with either Serp-1 or M-T7 treatments, dependent upon the animal model studied. Examination of selected cell responses after stent implant and treatment with these anti-inflammatory agents will require further study.

With this study we examine the potential for two anti-inflammatory protein therapeutics, that block differing inflammatory response pathways, to reduce plaque growth in cholesterol fed rabbit model with balloon angioplasty and stent implant. Targeting either of the two differing pathways, either chemokines or coagulation pathway serine proteases, significantly reduced stent plaque and inflammatory cell invasion. We conclude that blockade of inflammation after stent implant has the potential to reduce in stent restenosis and potentially stent thrombosis, providing newand untapped therapeutic approaches to the prevention of restenosis and thrombosis after coronary stent implant.