Percutaneous angioplasty of aorto-coronary venous bypass grafts is associated with a higher incidence of adverse events compared to interventions in native vessels.¹ Reasons for this well-known phenomen are emboli of pieces of friable lipid-rich plaque² and thrombus,³ resulting in the no-reflow phenomen and myocardial ischemia after angioplasty.

Using glycoprotein IIb/IIIa receptor antagonists as adjunctive treatment during intervention on bypass grafts does not improve clinical endpoints after 30 days and 6 months.¹ Several studies^4,5,6,7 demonstrated the usefulness and efficacy of protection devices for graft interventions. The SAFER (Saphenous vein graft Angioplasty Free of Emboli Randomized) trial,⁴ using stenting plus GuardWire® (Medtronic Vascular, Minneapolis, Minnesota) versus standard procedure, showed significant improvement in clinical outcome using the protection device, with a significant reduction in the composite endpoint including death, myocardial infarction, and revascularization.

The GuardWire has some practical limitations due to its long occlusion times (mean time = 388 seconds⁴) and to the difficulty in controlling the intervention because of remaining contrast medium inside the occluded graft where there is no flow. The use of this system is typically time-consuming and complex.

The aim of the present study was to study the efficacy and safety of a new filter-based device with a simpler approach. Its use was restricted for carotid angioplasty until now.

Methods
Device descprition. The Spider™ filter (ev3, Inc., Plymouth, Minnesota) is a heparin-coated nitinol mesh filter mounted on a 0.14 inch coated guidewire. Five nitinol filter sizes are available with a radiopaque proximal loop (3–7 mm) and a nitinol tip with a stainless steel hypotube connector. A radiopaque gold proximal loop and two marker bands provide visualization of vessel apposition. Special filter coating allows patency for up to 60 minutes (Figure 1).

The guidewire supports the deployment of the filter and at 175 cm, the guidewire converts from an over-the-wire system to a monorail system. First passage with the delivery catheter is possible with another guidewire of the operator’s choice, which may allow better steerability. The device was recently approved by the U.S. Food and Drug Administration for use in coronary arteries and vein grafts.

Study design. We included 40 consecutive patients with 50 lesions in which stenosis of the venous bypass graft was > 50%, and who were experiencing angina or had objective signs of ischemia. All consecutive patients with graft lesions were included. The mean age of these patients was 67.6 years (77.5% men), and the mean bypass age was 8.0 years (Table 1).

All patients were given intravenous aspirin (500 mg), 10,000 IU bolus of heparin, and clopidogrel 300 mg as a loading dose before elective angioplasty. Glycoprotein IIb/IIIa antagonists were administered to 10 patients (25%) at the discretion of the operator.

The following data were prospectively collected by an independent operator: TIMI flow before, during placement of the Spider system, and as final result after the procedure.

Thrombus was noted when an intraluminal filling defect separated from the adjactent vessel wall was seen. Embolic events were defined as a new filling defect, slow flow, or no-reflow. Pre-procedure QCA was performed using Quantcor QCA software (version V2.0, Pie Medical Imaging, Maastricht, The Netherlands). Percent diameter stenosis was calculated using the “worst view” method, and lesion length was added in cases where there was more than one lesion.

“Technical success” was defined as successful placement of the device distal to the target lesion. The procedure was considered angiographically successful if post-interventional stenosis was < 50% by visual estimation without device-related or other complications. After removal of the device, careful identification was made of macroscopic debris inside the filter (classified as “yes” or “no”). Following the procedure, after 24 hours, an ECG recording was performed, and creatine kinase (CK) and troponin were determined. Myocardial infarction was defined as a new three-fold rise in CK.

Follow-up was performed according to clinical record forms and after 30 days by phone call. MACE was defined as death, myocardial infarction, CABG, or new intervention. Minor complications such as aneurysm during hospital stay were recorded.

Statistics. Comparisons for continuous variables were performed using the Student’s t-test. A probability value of p < 0.05 was considered statistically significant.

Results
We included 40 consecutive patients with 50 lesions (see above). Baseline parameters are shown in Table 1. Six patients (15%) had acute coronary syndrome defined here as an elevation of CK and/or troponin before starting treatment.

Procedural parameters. Technical success following our defintion was achieved in 37 patients (92.5%). Primary stenting was possible in 80%. Mean lesion length was 8.9 mm and the mean degree of stenosis was 65.1%. Nineteen patients (47.5%) had a type A stenosis following the AHA/ACC criteria; 21 patients (52.5%) had type B stenosis (Table 2).

Acute results. There were no peripheral macroembolism events according to our definition, and macroscopically identified debris were caught with the filter in 21 patients (52.5%). Final TIMI flow 2 was achieved in 3 patients (7.5%), and TIMI flow 3 was achieved in 36 patients (90.0%; Table 3).

Following the AHA/ACC definition of stenosis morphology, 19 (47.5%) patients were found to have type A stenosis, and 21 patients (52.5%) had type B 1/2 stenosis. A trend toward a new rise in ischemic markers and more frequent debris in the filter in type B stenoses was detected, but this trend was not significant (Table 4). No patient had residual stenosis > 50% after dilatation and/or stenting as a final result.

Post-procedural results. Elevation of CK and/or troponin was noted in 11 patients (27.5%), and no patient had new Q-waves at discharge. MACE events occured in-hospital and after 30 days, according to our definition, in 4 patients (10%): One patient died due to cardiogenic shock before starting treatment, while another developed cardiogenic shock during angioplasty and died after multiorgan failure. Both patients had severly reduced ejection fractions before beginning treatment (< 20%), but this was not attributable to the protection device. No device-related failures were noted. Two other patients had peri-procedural myocardial infarctions with more than a three-fold rise in creatine kinase (Table 5). Figure 2 shows an example of a thrombus that was removed.

Discussion
Percutanous intervention in venous aorto-coronary grafts carries an elevated risk of peri-interventional complications. A pooled analysis from five randomized studies revealed a two-fold rise in mortality of 2.1% related to PCI in native vessels.¹ This high complication rate was due to microvascular obstruction resulting from micro- and macroembolism. Slow flow or no-reflow phenomen is a marker highly suggestive of ongoing distal obstruction during angioplasty. Angiographic risk factors predicting embolism are: diffusely diseased vein grafts, eccentric lesions, presence of thrombus, and a large plaque volume.8 However, recent protection devices have shown a highly significant reduction in major adverse events compared to conventional percutaneous interventions,⁴ though the placement of different devices can often be time-consuming.

The Spider system allows the use of a guidewire of the operator’s choice for an easy, single passage of the stenosis, and can quickly be changed after placement from over-the-wire to monorail use, allowing a conventional procedure to be performed. Technical success defined as placement of the device distal to stenosis was possible in 92.5% of our patients, which takes into account the learning curve required. This figure is similar to that from other studies.^4,5,7 Primary stenting was performed in 80% of our patients, and final TIMI flow 3 was achieved in 90%, with no incidence of visible peripheral embolism.

MACE events in our study were low and comparable to those of other series.^4,6,7 A new rise in CK or troponin was observed in 27.5% of patients after treatment, but 8/11 patients (72.7%) experienced only a peri-procedural rise in troponin, with no CK elevation. To our knowledge, no other study has systematically used the troponin marker for angioplasty patients in whom a protection device was used.

Post-procedure CK elevation was observed in 3 of our patients (7.5%), as opposed to 16.3% in the SAFER trial⁴ using the GuardWire. Macroscopically entrapped debris was found in 52.5% of our patients. However, the MACE rate in our patients was 10%, and troponin elevation is a marker of peri-procedural microembolism. The reasons for this phenomen may be due to the porous filter wire structure and microembolism during placement of the device. Using the morphologic criterion of stenosis (type A or B) as a predictor of micro- or macroembolism was not successful. Thus, we recorded a trend toward more frequent entrapped debris and troponin elevation after angioplasty in patients with type B stenosis, but found no significant levels. With this in mind, it appears that all lesion types (except, perhaps, in-stent stenosis^9,10) require a protection device until larger studies can define more exact criteria for its use.

Study limitations. Ours was not a randomized trial, and the results were derived from a small number of patients from a single, high-volume center. The relatively high incidence of microembolism with elevation of ischemic markers in relation to other studies may be due to the fact that our study followed a post-procedure troponin rise, and not just a rise in CK.

Conclusion
The Spider embolic protection device is user-friendly and offers a high technical success rate for angioplasty in aorto-coronary venous bypass grafts. Furthermore, it offers an excellent final TIMI flow, with no macroembolism and a low MACE rate. These results are comparable to those of other published studies using embolic protection devices.

h.korn.kar@zentralk-linik-bad-berka.de