Author + information
- Received December 18, 2000
- Revision received April 11, 2001
- Accepted April 23, 2001
- Published online August 1, 2001.
- Antoni Bayes-Genis, MDa,
- Allan R Camruda,
- Michael Jorgensona,
- Janis Donovana,
- Kristin L Shogrena,
- David R Holmes Jr, MD, FACCa and
- Robert S Schwartz, MD, FACCa,* ()
- ↵*Reprint requests and correspondence:
Dr. Robert S. Schwartz, Division of Cardiovascular Diseases, Mayo Clinic, 200 First Street, SW, Rochester, Minnesota 55905
This study evaluates whether rinsing stents with high pressure immediately before implantation minimizes stent-induced inflammation and neointimal formation.
Several reports indicate that manual stent manipulation before implantation results in foreign body contamination and increased neointimal hyperplasia.
A stent-cleaning chamber was developed to rinse stents at a sustained hydrodynamic pressure of 4 atm for 10 s. Commercial pre-mounted stents were examined with different levels of manipulation: 1) untouched stents: no stent manipulation before implantation; 2) handled stents: manual stent re-crimping on the balloon; 3) rinsed stents: pressure-rinsed with the stent-cleaning chamber. In vitro surface analysis was evaluated by scanning electron microscopy. Neointimal hyperplasia and inflammation around stent struts were also assessed in the pig in-stent restenosis model.
In vitro analysis revealed fewer contaminants on rinsed stents compared with untouched (p = 0.01) and handled stents (p < 0.001). In vivo, neointimal thickness, neointimal area and vessel percent stenosis were significantly reduced in rinsed, compared with not-rinsed, stents (p = 0.002, p = 0.007, p = 0.008 respectively). In addition, a significant reduction in the inflammatory infiltrate around struts was observed in untouched, compared with handled, stents (p = 0.04) and in rinsed, compared with not-rinsed, stents (p < 0.001). Regression analysis accounting for injury and neointimal thickness showed significant differences in slopes between “handled + not-rinsed” and “handled + rinsed” stents (p = 0.004), and between “untouched + not-rinsed” and “untouched + rinsed stents” (p = 0.037).
Rinsing stents under high pressure immediately before coronary implantation results in less inflammation around struts and thinner neointima at 28 days in this pig model.
Coronary restenosis after angioplasty and stent implantation placement remains a substantial problem (1). Although recent studies show that stent implantation reduces restenosis significantly compared to balloon angioplasty (2), stents have not eliminated restenosis, especially in complex lesion subsets such as diffuse disease and small vessels. The observation that in-stent restenosis sometimes is diffuse suggests that a generalized reaction to the stent may be a possible etiology. Such a reaction might be to the metal, or alternatively from residual contaminants on the stent from the manufacturing process (3).
Previous animal studies (4,5)established a significant correlation between the degree of arterial injury caused by metallic wire coils, the resulting neointimal thickness, and lumen stenosis at the stent site. More recently, inflammatory reactions induced by stent struts after stent manipulation were found to be positively associated with neointimal hyperplasia (6). Chronic inflammatory cells around stent struts were also commonly seen in a recent histopathologic report of coronary stenting in humans (7). In light of these findings, reducing stent-induced inflammatory response has the potential to limit excessive neointimal formation within stents.
Several reports indicate that manual stent manipulation before implantation may cause foreign body contamination (8,9)and increased neointimal hyperplasia (6). We hypothesized that rinsing commercially available sterile stents with high pressure immediately before implantation might eliminate residual surface stent contaminants and reduce the inflammatory response elicited by the stent struts. Stents were rinsed in a chamber using a pressurized sterile solution before implantation. We evaluated this stent-cleaning chamber in vitro, examined surface stent contaminants with and without rinsing by electron microscopy, and assessed inflammation and neointimal hyperplasia with and without stent rinsing in a porcine coronary in-stent restenosis model.
The stent-cleaning chamber (Fig. 1)
The stent-cleaning chamber was made with silicone rubber tubing connected between two hemostatic valves. Proper ethylene oxide sterilization was accomplished before each use. Pre-mounted stents were advanced through the hemostatic valve and locked in the chamber closing the proximal and distal valves. An inflow port allowed connection to a standard commercial indeflator containing the rinsing solution, and an outreach port permitted drainage of the solution and stent contaminants. A solution of 40 ml of sterile ultrapure water (Barnstead, Dubuque, Iowa) with 25 U/ml of heparin (Elkins-Sinn, Inc., Cherry Hill, New Jersey) was injected through the stent-cleaning chamber at a sustained pressure of 4 atm during 10 s (4 ml/s). After the process was complete, the output port was closed, the distal valve opened, and the stent advanced to the desired coronary artery.
Stent manipulation in this study was as follows: Untouched stents:stents were implanted without manual stent manipulation. The time elapsed between opening of the sterile package and stent implantation was <30 s, and neither the stent nor the balloon was touched by the operator. Handled stents:stents manually re-crimped on the balloon during 10 s before implantation. Sterile powdered gloves directly from their commercial package were used for stent crimping. Stents were placed on a sterile table for ≤3 min before implantation in a manner similar to clinical stent implantation. Rinsed stents:stents were pressure-rinsed with sterile heparinized ultrapure water in a stent-cleaning chamber immediately before implantation. In vivo, four groups of stents were studied in the pig coronary in-stent restenosis model: “untouched + not-rinsed,” “untouched + rinsed,” “handled + not-rinsed” and “handled + rinsed.”
In vitro surface stent analysis
Six pre-mounted balloon-expandable 16-mm NIR stents (Boston Scientific Scimed, Inc., Maple Grove, Minnesota) were expanded in vitro under sterile conditions as untouched (n = 2), handled (n = 2) and rinsed (n = 2). After expansion, stents were collected in a sterile polypropylene tube for immediate ultrastructural evaluation by a Hitachi S-4700 scanning electron microscope. Digital scanning electron microscopy images were obtained, and surface particles were counted by blinded observers using a digital imaging system (Sigmascan Pro 5.0, SPSS Inc., Chicago, Illinois) on three random samples of each stent at a 1:250 magnification.
Animal study protocol
Animal studies were performed with approval of the Institutional Animal Care and Use Committee of the Mayo Foundation. Eight domestic pigs (Sus scrofa; weight 29 ± 3 kg) underwent oversized coronary stent implantation as described previously (10). General anesthesia was achieved with ketamine (3 mg/kg intramuscular [IM]) and xylazine (30 mg/kg IM). Additional medication at the time of induction included atropine (1 mg IM) and antibiotic (flocillin, 1 g IM). During the angioplasty procedure, an intra-arterial bolus of heparin (10,000 U) was administered. Under sterile conditions, an 8F sheath was inserted into the left carotid artery, and a JL3.5 (Cordis) guide catheter was advanced to the ostium of the desired coronary artery under fluoroscopic guidance. Four stents with different degrees of manipulation were randomly implanted. Pre-mounted balloon expandable16-mm NIR stents (Boston Scientific Scimed, Inc.) were deployed at 8 atm for 30 s to achieve a stent/vessel ratio of 1.2:1.
Repeat angiograms were obtained after stent implantation. All equipment was removed, and the carotid artery was ligated. All animals received 325 mg of aspirin and 75 mg of clopidogrel orally daily until euthanasia. Meticulous attention was undertaken during the procedure to avoid contamination of any transcatheter device. Four weeks later animals were euthanized, and the coronary arteries were perfusion-fixed for histologic analysis.
The hearts were fixed in 10% buffered formalin for 24 h. The treated coronary segments were harvested, dehydrated in ascending alcohols and infiltrated with methylmethacrylate (MMA). Specimens were kept at 4°C during dehydration and infiltration of the tissue. The MMA used in this study was a mixture of uninhibited methylmethacrylate, polyethylene glycol distearate, dibutylphthalate and benzoyl peroxide. After polymerization, which takes place at room temperature in presence of nitrogen, 5-μm-thick sections were cut using a heavy-duty rotary microtome (Leica RM 2165, Minneapolis, Minnesosta) with a D-profile tungsten-carbide knife. Modified van Gieson staining and hematoxylin-eosin staining were then performed using these free-floating sections.
Histomorphometry was performed on elastin sections using a microscope coupled to a digital morphometry system (Diagnostic Instruments, Inc., Sterling Heights, Michigan). Measurements were made on three cross-sections from each stent. Injury score was assessed as previously described by Schwartz et al. (4), and the inflammation score for each individual strut was graded as described by Kornowski et al. (6). Vessel percent stenosis was calculated as (stenotic lumen area/original lumen area) × 100. The area within the external elastic lamina was considered the vessel size. All measurements were evaluated by observers blinded to conditions.
Data are presented as mean ± SEM. A sample size of eight arteries per group was chosen to allow detection of a projected difference in neointimal thickness of 0.1 mm at a power of 0.8. Two-way analysis of variance (ANOVA) was used for comparison between untouched/handled stents and rinsed/not-rinsed stents. In addition, the interaction term among groups was tested and interpreted. Associations among groups were assessed by Spearman rank correlation coefficients. Regression modeling was used to account for injury and the injury-dependent neointimal response (11). Note that the variable labeled Gp establishes the “rinsed” stent groups. The two rinsed stent groups (untouched + rinsed and handled + rinsed) were analyzed separately from their not-rinsed counterparts (untouched + not-rinsed and handled + not-rinsed).
Testing for differing intercepts: neointima = constant + injury score + Gp; testing for differing slopes (allowing an arbitrary intercept): neointima = constant + injury score + Gp + Gp × injury score; testing for differing slopes (forcing a fixed intercept): neointima = constant + injury score + Gp × injury score. Similar regression analyses were performed to account for injury and injury-dependent inflammation.
In vitro evaluation of stent rinsing
Untouched stents had an average of 64 ± 9.2 surface contaminant particles, whereas 164.7 ± 10.3 contaminant particles were detected on handled stents and 25.7 ± 6.9 on rinsed stents. The contaminants on rinsed stents were significantly reduced compared with untouched stents (p = 0.01) and handled stents (p < 0.001). The contaminants were both outside and inside the stent surface (Fig. 2).
In vivo response to stent rinsing
Eight animals underwent successful implantation of four stents with different levels of manipulation. All pigs survived until euthanasia at 28 days. Four groups were evaluated (see Methods for details): untouched + not-rinsed stents (eight arteries), untouched + rinsed stents (eight arteries), handled + not-rinsed stents (eight arteries) and handled + rinsed stents (eight arteries).
Neointimal response to injury
Histologic study revealed neointimal formation and lumen stenosis of varying magnitude within all 32 examined stents. Table 1shows the histomorphometric measurements performed in the four studied groups. No significant differences in vessel injury were observed among groups as assessed by two-way ANOVA (Table 2). Vessel percent stenosis, neointimal thickness and neointimal area were significantly reduced in rinsed, compared with nonrinsed, stents (p = 0.002, p = 0.007 and p = 0.008, respectively). The benefit of stent rinsing on these histomorphometric variables was independent of the level of stent manipulation (untouched/handled) (Table 2). Vessel size and lumen were larger among rinsed stents (p = 0.004 and p < 0.001, respectively), with a significant interaction with stent manipulation (Table 2). Multiple comparison analysis using the Tukey method found that handled + rinsed stents had the largest lumen (p < 0.001).
Handled + not-rinsed stents showed a significant association between arterial injury and neointimal thickness (r = 0.75; p = 0.03), but this association was not found in handled + rinsed stents (r = 0.459; p = 0.25). Representative linear fit curves for these correlations are presented in Figure 3, A. Regression analysis that accounted for injury and the injury-dependent neointimal thickness showed statistically significant differences in slopes, both allowing any intercept (p = 0.004) or forcing a fixed intercept (p = 0.003) (Fig. 3A). Untouched + not-rinsed stents showed a significant association between the degree of arterial injury and neointimal thickness (r = 0.72; p = 0.04), and this association was again not found in untouched + rinsed stents (r = 0.6; p = 0.11). Representative linear fit curves for these correlations are presented in Figure 3, B. Regression analysis that accounted for injury and the injury-dependent neointimal thickness performed between untouched and untouched + rinsed stents showed statistically significant differences in slopes, both allowing any intercept (p = 0.037) or forcing a fixed intercept (p = 0.011) (Fig. 3B).
Figure 4is a histopathologic example of the neointimal response between handled + not-rinsed and handled + rinsed stents with the same degree of arterial injury. The trivial stent handling (handled + not-rinsed) performed in this study increased vessel percent stenosis by 15.3% compared with untouched + not-rinsed stents, and by 35.9% compared with handled + rinsed stents.
Neointimal response to inflammation
A mean inflammation score of 0.69 ± 0.1 was found in untouched + not-rinsed stents, 0.41 ± 0.07 in untouched + rinsed stents, 0.99 ± 0.07 in handled + not-rinsed stents and 0.43 ± 0.09 in handled + rinsed stents. A significant reduction in the inflammatory infiltrate around struts was observed in untouched, compared with handled, stents (p = 0.04) and in rinsed compared with not-rinsed stents (p < 0.001). The benefit of stent rinsing on inflammation was independent of whether the stent was untouched or handled before implantation (interaction term, p = 0.16). Thirty-eight of 74 struts (51.4%) in untouched + rinsed and 51 of 77 struts (66.2%) in handled + rinsed stents had inflammatory scores of 0 and none had a score of 2. Twenty-five of 71 struts (35.2%) of untouched + not-rinsed stents had an inflammatory score of 0. Only six of 73 struts (8.2%) of handled + not-rinsed stents had a score of 0, 59 struts (80.8%) had a score of 1, and eight struts (10.9%) had a score of 2. No struts in any group had an inflammation score of 3.
The inflammatory reaction consisted of groups of mononuclear leukocytes adjacent to the struts. Small round lymphocytes and occasional polymorphonuclear leukocytes were also noted surrounding the struts. Fibrin microthrombi around the struts were also identified. Powder remnants such as birefringent particles were not observed under histopathologic examination.
Inflammatory response to injury
Regression analysis performed between injury and injury-dependent inflammation showed a significant difference in intercepts assuming equal slopes (p = 0.008) and a significant difference in slopes assuming a fixed intercept (p < 0.0001) between handled + not-rinsed and handled + rinsed stents. Regression analysis comparing injury-dependent inflammation between untouched + not-rinsed and untouched + rinsed stents showed a significant difference in intercepts assuming equal slopes (p = 0.018) and a significant difference in slopes assuming a fixed intercept (p = 0.01). Representative linear fit curves for these associations are presented in Figure 5A. Figures 5B, C and Dare histologic examples of the inflammatory response observed around stent struts in untouched + not-rinsed, handled + not-rinsed and handled + rinsed stents.
Restenosis remains an important limitation of percutaneous interventions for coronary artery disease, despite major procedural advances over the past decade. This study shows a significant reduction in inflammatory response surrounding stent struts with the use of a custom-made stent-cleaning chamber, which translates into a reduction in neointimal thickness and vessel percent stenosis.
The stent-cleaning chamber appears effective in removing surface contaminants
Numerous studies suggest that foreign bodies activate macrophages (12,13). Products produced by the corrosion of metal implants (14), and powdered biomaterials (15), may elicit these reactions during stent implantation. Our in vitro analysis showed a substantial number of surface foreign materials in untouched and especially in handled stents. The source and identity of this matter is unknown, but it may be talc or other contaminants from the operators’ gloves, from the cath lab environment or from the stent packaging manufacture processing. Jung et al. (16)reported similar results with pre-mounted and hand-crimped stents using scanning electron microscopy and energy-dispersive elemental analysis. Pressure rinsing of stents with the stent-cleaning chamber successfully removed most surface stent contaminants.
Stent rinsing reduces stent-induced inflammatory responses
Experimental studies and pathology reports suggest important relationships among inflammation, vascular injury, and neointimal growth. Monocytes may contribute to neointimal thickening by differentiating into collagen-secreting myofibroblasts (17), by generating injurious reactive oxygen intermediates (18), through elaboration of growth and chemotactic factors (19)and by matrix metalloproteinase production capable of degrading extracellular constituents, thereby facilitating cell migration (20). In the porcine in-stent restenosis model described in this study, rinsed stents using a custom stent-cleaning device before implantation elicited less inflammation and less neointimal hyperplasia at one month than not-rinsed stents. Similarly, untouched stents elicited less inflammation and less neointimal hyperplasia than handled stents. Kornowski et al. (6), using similar methods, found a slightly higher injury score, along with a higher inflammation score, using manually crimped stents that were likely to have been contaminated. These findings have altered our current practice, and we now avoid manipulation of commercially available precrimped stents, either to check their proper implantation onto the balloon or to give them a special curvature.
Stent rinsing modifies surface stent electrostatic forces
Yutani et al. suggested that chronic inflammation stimulated by ion-coating materials around stent struts might give rise to smooth muscle cell proliferation and growth factor production by platelets in the thrombi (21). Electrostatic forces on the surface of metals of the kind used in stents are a determining factor in the interaction of blood with those surfaces and the vascular wall (22). Simon et al. (23)studied the electrostatic forces residing on the surface of metal intravascular prostheses and found that protein binding was relatively uniform for all metallic surfaces, including stainless steel 316L. In their study metals were rinsed with phosphate buffered saline, and post-elution determinations showed that proteins eluted from metallic surfaces, because albumin was more easily eluted than fibrinogen and fibronectin. In light of these findings, it is reasonable to speculate that rinsing stents with high pressure before implantation may modify surface stent electrostatic forces and reduce stent interaction with circulating proteins.
The use of a stent-cleaning chamber is a step beyond the routine hygienic measures used in the clinical setting in the catheterization laboratory. However, meticulous handling techniques (washing of gloves and hands, and minimal handling of catheters and guide wires), and perhaps enhanced manufacturing techniques (i.e., to remove impurities), may help limit stent-induced inflammatory responses and neointima. Further experiments must confirm these results in a clinical context and establish the practical importance of such questions for different commercially available stents. The stent-cleaning chamber could be developed further for such testing in humans.
☆ Supported in part by a grant from “La Caixa” Foundation (Barcelona, Spain) (A.B.G.). We thank Boston Scientific Scimed, Inc. (Maple Grove, Minnesota) for donation of the stents used in this study.
- analysis of variance
- Received December 18, 2000.
- Revision received April 11, 2001.
- Accepted April 23, 2001.
- American College of Cardiology
- Schwartz R.S.,
- Huber K.C.,
- Murphy J.G.,
- et al.
- Karas S.P.,
- Gravanis M.B.,
- Santoian E.C.,
- et al.
- Kornowski R.,
- Hong M.K.,
- Tio F.O.,
- et al.
- Farb A.,
- Sangiorgi G.,
- Carter A.J.,
- et al.
- Schwartz R.S.,
- Murphy J.G.,
- Edwards W.D.,
- et al.
- McKenna C.J.,
- Burke S.E.,
- Opgenorth T.J.,
- et al.
- Rogers C.,
- Edelman E.R.
- Jung F.,
- Bach R.,
- Franke R.P.
- Bayes-Genis A.,
- Campbell J.H.,
- Gabbiani G.,
- et al.
- Peri G.,
- Chiaffarino F.,
- Bernasconi S.,
- et al.
- Assoian R.K.,
- Fleurdelys B.E.,
- Stevenson H.C.,
- et al.