Randomized Comparison of Distal Protection With a Filter-Based Catheter and a Balloon Occlusion and Aspiration System During Percutaneous Intervention of Diseased Saphenous Vein Aorto-Coronary Bypass Grafts
Background— The high rate of periprocedural complications resulting from atherothrombotic embolization after percutaneous intervention in diseased saphenous vein grafts is reduced by distal microcirculatory protection using a balloon occlusion and aspiration system. Whether filter-based catheters, which offer the inherent advantages of maintained perfusion and ease of use, are as effective for this purpose has not been established.
Methods and Results— A total of 651 patients undergoing percutaneous intervention of 682 saphenous vein graft lesions were prospectively randomized to distal protection with the filter-based FilterWire EX versus the GuardWire balloon occlusion and aspiration system. Device success was 95.5% and 97.2% with the FilterWire EX and GuardWire, respectively (P=0.25). Postprocedural measures of epicardial flow and angiographic complications were similar between the 2 groups, although bailout IIb/IIIa inhibitors were required slightly less frequently in the FilterWire EX group (0% versus 1.5%, P=0.03). The primary end point, the composite incidence of death, myocardial infarction, or target vessel revascularization at 30 days, occurred in 9.9% of FilterWire EX patients and 11.6% of GuardWire patients (difference [95% CI]=−1.7% [−6.4%, 3.1%]; P for superiority=0.53, P for noninferiority=0.0008).
Conclusions— Distal protection with the FilterWire EX may be safely used as an adjunct to percutaneous intervention of diseased saphenous vein grafts and, compared with distal protection with the GuardWire balloon occlusion and aspiration system, results in similar rates of major adverse cardiac events at 30 days.
Received April 9, 2003; revision received April 29, 2003; accepted May 5, 2003.
Embolization of atherothrombotic debris occurs commonly during percutaneous coronary intervention (PCI) of diseased saphenous vein bypass grafts (SVGs), frequently resulting in periprocedural myocardial infarction (MI).1–4 In a multicenter, randomized trial, protection of the distal microcirculation by a balloon occlusion and aspiration system (Medtronic GuardWire) significantly reduced the occurrence of angiographic and clinical complications during stent implantation in SVGs5 and is therefore presently accepted as the standard therapy for this application. Balloon occlusion devices, however, obligate a several-minute duration of ischemia that may result in severe angina and left ventricular dysfunction with hypotension and necessitate a hurried procedure. In contrast, distal protection with catheter-based filters maintains antegrade perfusion6,7 and may thus be preferable in selected patients. However, distal filters are larger profile than the GuardWire and possess their own set of technical and operator-related issues that may adversely impact outcomes.7 We therefore performed a pivotal, multicenter randomized trial to evaluate the safety and efficacy of distal microcirculatory protection with a novel filter-based catheter compared with the GuardWire balloon occlusion and aspiration device during PCI of diseased SVGs.
Patients ≥21 years old undergoing PCI with planned stenting of 1 or more de novo lesions in diseased SVGs were eligible for enrollment. Exclusion criteria included acute or recent MI or current elevation of creatine phosphokinase (CPK)-MB isoenzyme, uncontrollable allergy or contraindication to contrast or any of the study medications, cerebrovascular event within 2 months, serum creatinine ≥2.5 mg/dL, prior target vessel PCI within 30 days or planned multivessel intervention in which not all lesions met study enrollment criteria, SVG age <6 months, true aorto-ostial lesion ≤10 mm in length, reference vessel diameter <3.5 mm or >5.5 mm, TIMI 0 flow, lesion within 2.5 cm of the distal anastomosis or >2 cm of relatively straight vessel distal to the lesion not present, unprotected Y-limb branch vessel proximal to the study device, planned use of laser or atherothrombectomy devices, or left ventricular ejection fraction <25%. The Investigational Review Board at each site approved the protocol, and all patients provided informed written consent.
Randomization and Interventional Protocol
Aspirin 325 mg was administered before the procedure. A 300-mg clopidogrel or 500-mg ticlopidine load was recommended before the procedure but was administered in all cases within 4 hours after the procedure. Glycoprotein IIb/IIIa inhibitor use was left to the discretion of the investigator. Intravenous heparin was administered to maintain the procedural activated clotting time to ≥300 seconds (≥200 seconds if a IIb/IIIa inhibitor was used). Randomization was stratified by intended IIb/IIIa inhibitor administration.
Patients were randomized 1:1 to either distal protection with the FilterWire EX system (Boston Scientific) or the GuardWire (Medtronic). As previously described, the FilterWire EX consists of a distal polyurethane filter with 110-μm diameter pores mounted on a 0.014-in steerable guide wire via a spinner tube, which permits unrestricted wire rotation and steering.7 The filter is attached to an elliptical, eccentric self-expanding radiopaque nitinol loop, deployed after crossing the lesion by retracting a restraining 3.9F delivery sheath. The elliptical loop maximizes the cross-sectional area exposed to the bloodstream and affords apposition of vessels ranging from 3.5 to 5.5 mm in diameter. PCI is then performed over the guide wire, with maintained antegrade perfusion carrying the liberated atherosclerotic and thrombotic debris distally into the filter. The filter is then closed with a retrieval sheath and the retained debris removed.
In contrast, the GuardWire consists of a 0.014-in guide wire incorporating a central inflation lumen to which an elastomeric balloon is attached distally.5,8 The crossing profile is 0.036-in (≈2.8F). Once distal to the lesion, injection of diluted contrast results in balloon inflation (3.0- to 6.0-mm-diameter range) with subsequent cessation of antegrade blood flow. Actuating the MicroSeal adapter closes a proximal valve in the GuardWire, allowing the inflation syringe to be disconnected, after which intervention is performed over the wire. Liberated debris is suspended within a stagnant column of blood and is aspirated through the 5F monorail Export Catheter. The GuardWire balloon is then deflated and flow restored.
Once distal protection was established, stenting was performed using approved devices and standard techniques. The decision of whether to perform balloon angioplasty before or after stent implantation was left to the discretion of the operator. CPK-MB levels were measured at least 3 times after the procedure within the first 24 hours, and a 12-lead ECG was obtained immediately and before discharge. Aspirin was continued indefinitely, and a thienopyridine was continued for 4 weeks if stents were implanted.
Definitions, End Points, and Statistical Analysis
The primary end point of the study was the 30-day composite occurrence of major adverse cardiac events (MACE), including death, MI, or target vessel revascularization. An MI was defined as any postprocedural CK-MB elevation to ≥3 times normal. The development of new, pathological Q waves in ≥2 contiguous leads (assessed by a blinded ECG Core Laboratory) was required to diagnose a Q-wave MI. All primary end point events were adjudicated by an independent committee blinded to treatment allocation. Device success for the FilterWire EX was defined as the ability to deliver, deploy, and retract a device at and from the desired target location. Device success for the GuardWire was defined as the ability to deliver a device to the desired target, obtain distal occlusion, and perform aspiration without loss of occlusion attributable to leak or rupture. Clinical success was defined as device success with a final lesion diameter stenosis <50% without in-hospital MACE. Angiographic films were analyzed in a central core laboratory using previously validated techniques.9
The sample size was chosen to demonstrate that the FilterWire EX was noninferior (ie, equivalent or superior) to the GuardWire.10 Assuming that the 30-day MACE rates for both the control and study device were 10%, randomizing 618 patients would provide 80% power to demonstrate noninferiority with 1-sided α error of 5% and δ of 6%. Categorical variables were compared by the likelihood ratio χ2 test or Fisher’s exact test. Continuous variables are presented as mean±SD and were compared by 1-way ANOVA or unpaired t test. The influence of baseline demographic features, angiographic variables, and treatment assignment on 30-day outcomes was evaluated using logistic regression. All analyses were by intention to treat, and all P values are 2-sided. Statistical significance was determined at the P<0.05 level.
Between October 2001 and September 2002, 651 patients at 66 sites in the United States and Canada undergoing stent implantation in 682 SVG lesions were randomized. As seen in Table 1, the population was elderly and characterized by a high incidence of hypertension, hyperlipidemia, diabetes, and prior MI. Baseline angiographic characteristics were well matched between the 2 groups, except that grafts randomized to the GuardWire had a slightly worse degeneration score (Table 2). Thrombus was present in approximately one quarter of lesions, and TIMI flow was reduced in 20% of vessels.
Device and clinical success rates were similar with both devices (Table 3). IIb/IIIa inhibitors were used in approximately half of patients in both groups before the procedure, although the bailout use of these agents was slightly more frequent in patients assigned to the GuardWire. Stent sizing was slightly greater in the GuardWire arm, although inflation pressures were higher in the FilterWire EX arm. Quantitative and qualitative angiographic results and epicardial coronary flow rates were similar in both arms, except for a slightly lower in-lesion diameter stenosis with the FilterWire EX device (Table 5).
In-Hospital and 30-Day Clinical Outcomes
In-hospital MACE occurred in 9.6% of FilterWire EX patients compared with 10.7% of GuardWire patients (difference [95% confidence interval]=−1.0% [−5.7 to 3.6], P=0.67). As shown in Figure 1, the primary composite 30-day MACE end point occurred in 9.9% of FilterWire EX patients and 11.6% of GuardWire patients (P=0.53). The 1-sided test of noninferiority was met at the P=0.0008 level. With 95% confidence, absolute 30-day MACE rates with the FilterWire EX compared with the GuardWire could range from 6.4% better to 3.1% worse or by relative measures from 45% better to 33% worse.
Compared with the GuardWire, 30-day MACE rates were lower with the FilterWire EX in smaller vessels and eccentric lesions; MACE rates were statistically similar with the 2 protection devices stratified by other angiographic variables (Table 5). By multivariate analysis, however, only the number of stents and total stent length were independently correlated with 30-day MACE in the entire population (Table 6).
In the present study, the first to examine the comparative efficacy of distal microcirculatory protection with 2 competitive technologies, the 30-day outcomes of PCI in diseased SVGs using the FilterWire EX system were noninferior (ie, statistically equivalent or better) to those with the predicate GuardWire system. Postprocedural measures of epicardial flow, angiographic complications, and the extent of myonecrosis were also similar between the 2 groups, although bailout use of glycoprotein IIb/IIIa inhibitors was required slightly more frequently with the GuardWire.
The baseline features of the patients in the present trial were similar to those in the Saphenous vein graft Angioplasty Free of Emboli Randomized (SAFER) trial,5 the pivotal study that established the GuardWire as the standard of care for distal protection during SVG intervention. The 11.6% 30-day MACE rate with the GuardWire in the present trial is also similar to the 9.6% rate with the same device in SAFER. Moreover, major device failure with the GuardWire was less in the present trial (2.3%) than reported in the SAFER trial, in which inability to deliver the device or maintain vessel occlusion occurred in 8.6% of patients. It is therefore unlikely that the relatively favorable results with the FilterWire EX system in the present trial are explained by suboptimal results in the control (GuardWire) arm. Thus, within the framework of the inclusion and exclusion criteria that dictated patient selection in this trial, physicians may be 95% confident that use of the FilterWire EX will result in 30-day MACE rates ranging from 45% better than with the GuardWire (absolute 6.4% less MACE) to 33% worse (absolute 3.1% greater MACE). These data are also consistent with the hypothesis that the FilterWire EX is better than control (putative placebo effect), given the 42% treatment effect versus placebo seen in SAFER, using the same definitions as in the present study.
Subtle differences in outcomes may not have been revealed by this study. For example, it was not possible to measure transient intraprocedural events, because the GuardWire results in total vessel occlusion whereas the FilterWire EX typically preserves antegrade perfusion throughout most or all of the procedure. It is also difficult to quantify the relative benefits of ease of use with filter-based distal protection devices, because the ischemia that results with balloon occlusion and aspiration mandates extremely rapid work, which may paradoxically result in shorter procedural times or reduced contrast use. Whether the more relaxed pace of filter-based distal protection allows more considered and appropriate decision-making cannot be answered by the present study. In this regard it is interesting that slightly larger stents were chosen in the GuardWire arm, but slightly greater stent inflation pressures were used with the FilterWire EX. Whether these parameters were affected by the choice of distal protection device is speculative. The diameter stenosis achieved was also slightly lower with the FilterWire EX, although the clinical significance of this slight improvement is doubtful.
Although exploratory subgroup analyses demonstrated a possible benefit of the FilterWire EX compared with the GuardWire in smaller vessels and eccentric lesions, the mechanistic explanation for these observations is uncertain, and these findings may be attributable to chance given multiple comparisons. The outcomes with both protection devices were statistically indistinct in other lesion and vessel subtypes, although this study was underpowered to detect variables that might favor one device rather than the other. Of note, however, event rates were very low in both groups when only 1 stent was implanted or the total length of stents was limited. Conversely, by multivariate analysis, the longer the lesion and the greater the number and length of stents implanted, the higher the event rate with both protection devices. Periprocedural adverse event rates were also increased in thrombotic lesions and degenerated vein grafts. Clearly, adjunctive therapies in addition to distal protection devices are required to additionally reduce the high adverse event rates after PCI in complex SVG lesions.
Future design enhancements are likely to additionally improve outcomes with both devices. As previously reported, lack of vessel wall apposition with the FilterWire EX loop may occur and allow embolic debris to bypass the filter.7 This phenomenon may be reduced by paying close attention to loop apposition in orthogonal views and has been eliminated with the next generation FilterWire EZ device, in which the collection apparatus moves independently from the guide wire shaft. The EZ device is also lower profile (≈3F) and more flexible, which would be expected to facilitate passage in tortuous vessels and crossing of tight lesions. The GuardWire has also been improved, however, with the crossing profile of the GuardWire Plus reduced to 0.028 inches (≈2.3F) and with a simpler-to-use MicroSeal adapter and smaller Export catheter.
The results of this trial apply only to the application studied, ie, distal protection during PCI of diseased SVGs. Whether filter-based catheters or balloon occlusion and aspiration systems are preferred in other clinical settings or vascular territories, such as during neurovascular intervention11,12 or primary angioplasty for acute MI,13 will require separate randomized trials to determine. Finally, additional comparative investigations will also be required to establish whether other filter-based distal protection catheters perform equally well, better, or worse than the FilterWire EX device.
Distal protection with the FilterWire EX system may be safely used as an adjunct to PCI of diseased SVGs and, compared with distal protection with the GuardWire balloon occlusion and aspiration system, results in similar rates of MACE at 30 days. Periprocedural adverse outcomes still occur in ≈10% of patients, however, emphasizing the need for complementary device-based innovations and new pharmacologic regimens to additionally enhance the safety of intervention in these high-risk patients.
Drs Stone, Hermiller, and Herrmann serve as consultants to Boston Scientific Corp (BSC); Dr Caputo is a member of the BSC advisory board; Dr Cox is an advisor to BSC; Drs Feldman and Ramee each have a small amount of equity in BSC; Drs Masud, Popma, Herrmann, and Cox receive research funding from BSC; Dr Rogers is on the speaker’s bureau for BSC; and Dr Stone serves as a consultant to Medtronic.
A complete list of the FilterWire EX Randomized Evaluation investigators and study organization appears in the online-only Appendix available at http://www.circulationaha.org.
Holmes DR, Topol EJ, Califf RM, et al. A multicenter, randomized trial of coronary angioplasty versus directional atherectomy for patients with saphenous vein bypass graft lesions. Circulation. 1995; 91: 1966–1974.
Hong MK, Mehran R, Dangas G, et al. Creatine kinase-MB enzyme elevation following successful saphenous vein graft intervention is associated with late mortality. Circulation. 1999; 100: 2400–2405.
Grube E, Gerckens U, Yeung AC, et al. Prevention of distal embolization during coronary angioplasty in saphenous vein grafts and native vessels using porous filter protection. Circulation. 2001; 104: 2436–2441.
Lansky A, Popma J. Qualitative and quantitative angiography. In: Topol E, ed. Textbook of Interventional Cardiology. Philadelphia, Pa: WB Saunders; 1999: 725–747.
Al-Mubarak N, Roubin GS, Vitek JV, et al. Effect of the distal-balloon protection system on microembolization during carotid stenting. Circulation. 2001; 104: 1999–2002.