`© 2005 by the American College of Cardiology Foundation
`Published by Elsevier Inc.
`
`STATE-OF-THE-ART PAPERS
`
`Vol. 46, No. 8, 2005
`ISSN 0735-1097/05/$30.00
`doi:10.1016/j.jacc.2005.04.061
`
`Distal Myocardial Protection During
`Percutaneous Coronary Intervention
`When and Where?
`Diana A. Gorog, MD, PHD, MRCP, Rodney A. Foale, MD, FRCP, Iqbal Malik, PHD, MRCP
`London, United Kingdom
`
`The discrepancy between angiographic success and microvascular perfusion has been
`recognized for some time. In the face of an open artery, the degree of microvascular perfusion
`determines post-infarct prognosis. Despite successful epicardial recanalization, tissue perfu-
`sion may be absent in up to 25% patients with acute myocardial infarction. Historically
`associated with saphenous vein graft intervention, embolization is increasingly recognized in
`native coronary arteries, particularly in patients undergoing primary percutaneous coronary
`intervention (PCI). With more than two million PCI procedures performed worldwide each
`year,
`there is enormous interest
`in protecting the left ventricular myocardium from
`embolization during PCI. This article reviews the evidence for distal myocardial protection
`and discusses the relative merits of the different available techniques.
`(J Am Coll Cardiol
`2005;46:1434 – 45) © 2005 by the American College of Cardiology Foundation
`
`Distal protection devices (DPDs) were first introduced for
`cerebral protection during carotid artery stenting (CAS) (1).
`In this setting, registries have demonstrated that use of
`DPDs may halve the combined end point of stroke or death
`(2,3).
`Although angiographic evidence represents only the tip
`of the embolization iceberg during percutaneous coronary
`intervention (PCI), even this occurs in up to 15% patients
`undergoing primary PCI (4). Angiographic indicators of
`embolization such as corrected Thrombolysis In Myocardial
`Infarction (TIMI) frame count and myocardial blush grade
`(MBG), as well as rapidity of ST-segment resolution, are
`highly predictive of clinical and functional outcome (5,6).
`That these phenomena are a manifestation of emboliza-
`tion, rather than de novo thrombus formation, is borne out
`by histological data showing, during elective PCI, emboli
`comprised of mucopolysaccharide components and necrotic
`cores (6 –11). Vulnerable plaque morphology, namely dis-
`ruption or thinning of the fibrous cap, overlying thrombus,
`and increased lipid content are associated with complica-
`tions from endovascular procedures (12,13). High plaque
`macrophage content and plasma matrix metalloproteinase 9
`(MMP9) levels may predict embolization during PCI,
`possibly due to thinning of the fibrous cap by MMP9
`secreted by plaque macrophages (14). In addition to me-
`chanical obstruction, the local response to embolization may
`contribute to myonecrosis (15–19).
`It is thus not surprising that emboli are resistant to
`antiplatelet medication. Although use of glycoprotein (GP)
`IIb/IIIa inhibitors has contributed to improved success rates
`
`From the Waller Cardiac Department, St. Mary’s Hospital, London, United
`Kingdom.
`Manuscript received February 5, 2005; revised manuscript received April 12, 2005,
`accepted April 15, 2005.
`
`with PCI, intervention in saphenous vein grafts (SVG) and
`in native vessels with high intraluminal thrombus burden
`continues to be hampered by thromboembolic events.
`
`TYPES OF DEVICES
`
`Available devices fall into four categories (Table 1):
`Distal filtration devices
`Distal occlusion devices
`Proximal occlusion devices
`Thrombus extraction devices
`The devices with widest evidence base in each category
`are the EZ-FilterWire (Boston Scientific, Natick, Massa-
`chusetts), the GuardWire (Medtronic AVE, Santa Rosa,
`California), Proxis (Velocimed, Maple Grove, Minnesota),
`and X-Sizer (EndiCOR Medical, San Clemente, Califor-
`nia) systems, respectively. This review will therefore con-
`centrate predominantly on these systems.
`The FilterWire system (Fig. 1) incorporates a nonocclu-
`sive filter (pore size 110 m) in the shape of a windsock,
`mounted on a nitinol loop, and fixed on its own guidewire,
`which is deployed through a 3.2-F delivery sheath. The
`nitinol loop self-expands to fit vessels 3.5 to 5.5 mm in
`diameter, and intervention performed over the wire. Finally,
`the device is captured using a 4-F retrieval sheath. Other
`filtration devices work similarly.
`The GuardWire temporary occlusion-aspiration system
`(Fig. 2) consists of a guidewire incorporating a central
`inflation lumen, to which an elastomeric balloon is attached.
`This has a 2.8-F crossing profile, and injection of diluted
`contrast results in balloon inflation (2.5- to 5.0-mm or 3.0-
`to 6.0-mm diameter), arresting anterograde flow. Interven-
`tion is performed over the wire, and liberated debris trapped
`proximal to the balloon aspirated through a 5-F monorail
`
`Page 1
`
`
`
`IPR2020-00126/-127/-128/-129/-130/-132/-134/-135/-136/-137/-138
`
`Medtronic Ex-1811
`Medtronic v. Teleflex
`
`
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`Gorog et al.
`Embolic Protection During PCI
`
`1435
`
`CAS
`DPD
`GP
`MACE
`MBG
`MMP9
`PCI
`PRIDE
`
`SAFE
`
`SAFER
`
`Abbreviations and Acronyms
`⫽ acute myocardial infarction
`AMI
`⫽ acute coronary syndromes
`ACS
`CAPTIVE ⫽ CardioShield Application Protects During
`Transluminal Intervention of Vein Grafts
`by Reducing Emboli
`⫽ carotid artery stenting
`⫽ distal protection device
`⫽ glycoprotein
`⫽ major adverse cardiac event
`⫽ myocardial blush grade
`⫽ matrix metalloproteinase 9
`⫽ percutaneous coronary intervention
`⫽ PRotection During Saphenous Vein Graft
`Intervention to Prevent Distal
`Embolization
`⫽ Saphenous Vein Graft Angioplasty Free of
`Emboli trial
`⫽ Saphenous Vein Graft Angioplasty Free of
`Emboli Randomized trial
`⫽ saphenous vein graft
`
`SVG
`
`Export Aspiration catheter. The balloon is then deflated
`and flow restored.
`The Proxis system incorporates a sealing balloon that is
`deployed upstream of the stenosis to create a stagnant
`
`column of blood in which intervention is performed
`(Fig. 3). Protection is thus in place before any device crosses
`the lesion. The device is 7- or 8-F guide compatible, and
`protects vessels 2.5 to 5 mm in diameter. The stent is
`delivered through the Proxis system, and flow is reversed,
`aspirating debris, before the sealing balloon is deflated,
`restoring flow.
`The X-Sizer system (Fig. 4) consists of a 1.5- or 2.0-mm
`stainless steel helical cutter in a protective housing con-
`nected to a 4.5- or 5.5-F dual-bore catheter shaft containing
`the guidewire and vacuum/extraction lumens. The catheter
`shaft is linked to a handheld control module and vacuum
`bottle in which debris is collected. Activating the control
`unit simultaneously activates the helical cutter, which ex-
`tends 1 mm beyond the protective housing, rotating at
`⬃2,100 rpm and initiates the vacuum, resulting in tissue
`maceration, excision, and aspiration.
`
`USE IN ACUTE CORONARY SYNDROMES (ACS)
`
`Table 2 shows the trials employing DPD in ACS (trial
`acronyms explained in Table 3). There have been concerns
`that DPD use during primary PCI might delay reperfusion.
`The first study using the FilterWire in primary PCI showed
`that successful DPD positioning was achieved in 89% of
`
`Table 1. Embolic Protection and Thrombectomy Devices Available or in Development
`Distal Occlusion-
`Aspiration Systems
`
`Filter-Based Systems
`
`Thrombectomy Devices
`
`Proximal Balloon Occlusion-
`Flow Reversal Systems
`
`EZ-FilterWire (EPI, Boston Scientific)
`-fixed to its own guidewire
`-3.2-F crossing profile
`-110 m pores
`-6-F guide-compatible
`
`Spider and Microvena Trap
`(eV3, Minneapolis, Minnesota)
`-monorail system
`-Heprotec coating prevents thrombin
`build-up
`-3.2- and 2.9-F crossing profiles
`-6F guide-compatible
`
`Angioguard (Cordis)
`-100 m pores
`-7-F guide-compatible
`
`Cardioshield and Neuroshield
`(Mednova/Abbott, Galway, Ireland)
`
`Rubicon (Rubicon Medical Corp.)
`-100 m pores
`-fixed to its own guidewire
`-2-F crossing profile
`-6-F guide-compatible
`
`Interceptor (Medtronic AVE)
`-100 m pores
`-6-F guide-compatible
`
`GuardWire (Medtronic)
`-2.8-F crossing profile
`-7-F guide-compatible
`
`X-Sizer (Endicor)
`-4.5- or 5.5-F crossing profile
`-6- or 8-F guide-compatible
`-compatible with any guidewire
`
`Proxis (Velocimed)
`-7/8-F guide-compatible
`-compatible with
`any guidewire
`
`TriActiv system (Kensey Nash)
`
`AngioJet (Possis)
`
`Kerberos Rinspirator/Protection
`device (Kerberos)
`
`Hydrolyzer (Cordis)
`
`Rescue (Boston Scientific)
`-4.5-F crossing profile
`-7-F guide-compatible
`
`Parodi Anti-embolization
`device (ArteriA)
`
`MO.MA occlusion system
`(Invatec, Brescia, Italy)
`
`Pronto (Vascular Solutions)
`-7-F guide-compatible
`
`Diver (Invatec, Brescia, Italy)
`-3.8-F crossing profile
`-6-F guide-compatible
`
`Page 2
`
`
`
`1436
`
`Gorog etal.
`Embolic Protection During PCI
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`Figure 1. The FilterWire system. (Panel 1) The polyurethane porous membrane filter attached to a nitinol loop. (Panel 2) The filter is deployed
`distal to the lesion, and the nitinol loop self-expands to fit the vessel upon retraction of the delivery sheath. (Panel 3) Saphenous vein graft containing
`thrombus, seen as intraluminal filling defect (A), treated with percutaneous coronary intervention using FilterWire protection (B), achieving a good result after
`stenting (C).
`
`patients in ⱕ10 min (6). Compared to historical controls,
`FilterWire use improved final angiographic characteristics
`and left ventricular performance. Among the several limi-
`tations of this study, final TIMI flow grade ⬍3 in 85% of
`patients in the control group may have lead to overestima-
`tion of the benefit of the DPD.
`A small study using the GuardWire in patients with
`angiographic “high-burden thrombus” undergoing primary
`PCI showed improved flow and MBG, but this did not
`translate into a reduction in 30-day major adverse cardiac
`events (MACE) (20). The EMERALD study was the first
`large randomized trial to evaluate the GuardWire in pa-
`tients with acute myocardial infarction (AMI). Although
`the device markedly reduced the incidence of angiographic
`slow/no-reflow, there was no significant overall effect on
`ST-segment resolution or infarct size. Preliminary results
`
`from the first 188 patients undergoing PCI using the
`GuardWire in the RUBY registry have suggested that direct
`device delivery was possible in 87% cases, with favorable
`angiographic and electrocardiogram characteristics and low
`clinical event rates. In the PROMISE study, use of the
`FilterWire-EX in patients undergoing primary PCI did not
`improve reperfusion and did not reduce infarct size com-
`pared with usual care.
`There are no data on the safety or efficacy of proximal
`occlusion systems in AMI.
`X-Sizer thrombectomy was first assessed in a small
`randomized study of patients with suspected intracoronary
`thrombus (21). Although the study failed to show a benefit
`on final angiographic characteristics, creatine kinase-MB, or
`30-day MACE, X-sizer pretreatment was associated with
`more rapid normalization of epicardial flow and, in patients
`
`Page 3
`
`
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`Gorog et al.
`Embolic Protection During PCI
`
`1437
`
`Figure 2. The GuardWire system. Upper panel corresponds to lower panel: the GuardWire, used to cross the lesion, is inserted into the MicroSeal Adapter
`(A), connected to the EZ Flator, which is inflated to occlude the vessel (B). Debris is aspirated using the Export Aspiration catheter (C). SVG ⫽ saphenous
`vein graft.
`
`infarction, more
`with ST-segment elevation myocardial
`rapid ST-segment resolution. The study was underpowered
`to detect a benefit in clinical parameters. In AMI patients
`with angiographic evidence of thrombus (22), thrombec-
`
`Figure 3. The Proxis system is delivered through a guiding catheter, and
`the sealing balloon (A) is inflated proximal to the stenosis, arresting flow,
`and debris aspirated through the Proxis system (B).
`
`tomy significantly improved pre-PCI flow, post-procedural
`MBG, and ST-segment resolution, but this was not re-
`flected in hard clinical end points, and the device failed to
`traverse the lesion in 9% of patients.
`In the VeGAS-2 study, the AngioJet thrombectomy
`device (Possis, Minneapolis, Minnesota) was compared
`with intracoronary urokinase infusion (23). Although
`thrombectomy was technically successful and reduced in-
`hospital MACE, the results were clouded by subsequent
`studies demonstrating worse outcomes with urokinase than
`with placebo in patients with thrombotic lesions undergoing
`PCI (24). The disappointing results of the AIMI study,
`presented at Transcatheter Cardiovascular Therapeutics
`(TCT) 2004, showed that AngioJet use paradoxically in-
`creased infarct size.
`There seem to be no data to suggest that routine use of
`any DPD system is beneficial in patients with ACS under-
`going PCI. However, it may be hard to show the benefit of
`protection against embolization that undoubtedly happens
`during angioplasty for AMI. In a prothrombotic milieu,
`thrombi may form on the downstream (low pressure) side of
`the protection device and embolize. Furthermore, fragmen-
`
`Page 4
`
`
`
`1438
`
`Gorog etal.
`Embolic Protection During PCI
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`Figure 4. The X-Sizer thrombectomy system comes ready to assemble in a tray (panel 1). Schematic of mechanism of action (panel 2). Panel 3 shows
`angiogram of right coronary artery proximally occluded by thrombus (A), X-sizer thrombectomy device in situ (B), and angiographic appearance after
`thrombectomy (C).
`
`tation of large thrombi by GP IIb/IIIa inhibitors or throm-
`bolysis may result in small particles that pass through the
`filter. Lastly, in contrast to the cholesterol emboli released in
`SVG intervention, the consequence of embolization during
`AMI is not a fait accompli, because thrombotic platelet
`emboli may subsequently be lysed in the distal myocardial
`bed, without sequelae. Thus, the significance of emboliza-
`tion during AMI may depend on the nature of the embolic
`material and integrity of endogenous thrombolytic response.
`
`USE IN SVG
`
`Pathophysiology. Saphenous vein graft interventions carry
`a 20% risk of MACE, predominantly AMI, and significant
`risk of no-reflow (25). The protection offered by GP
`IIb/IIIa inhibitors during native vessel PCI has not been
`mirrored in SVG intervention (26,27), reflecting the differ-
`ing composition of plaque in these settings; SVG plaques
`tend to be cholesterol-rich, with relatively low calcium
`
`Page 5
`
`
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`Gorog et al.
`Embolic Protection During PCI
`
`1439
`
`content and less intimal proliferation than plaques in native
`coronaries (28). Accordingly, debris embolized from SVG
`largely consists of fibrin and necrotic core (29).
`
`device, failed to demonstrate both superiority to no embolic
`protection, and noninferiority to the GuardWire.
`
`CLINICAL STUDIES
`
`The SAFE registry was the first to report on use of DPD in
`SVG intervention (Table 4) (29). Patients had relatively
`low-grade thrombus burden and good pre-procedural flow
`(TIMI flow grade 3, 84%). GuardWire balloon inflation
`(mean 5.4 min) was well-tolerated. Final angiographic
`appearances were excellent, and post-procedural creatine
`kinase-MB release compared favorably with historical con-
`trols (30). These results spawned the larger SAFER trial,
`which confirmed the favorable effects on angiographic/
`myonecrotic markers and clinical end points (31). Throm-
`bus burden was higher than in SAFE.
`Use of the FilterWire was first reported in a registry
`which,
`like SAFE and SAFER, excluded patients with
`AMI or severe left ventricular failure (32). Pre-procedure
`TIMI flow grade 3 was present in ⬃85% cases, with a
`higher thrombus burden (65%) than in the GuardWire
`trials. Analysis of phase I results identified several correct-
`able factors that led to improved results in phase 2 (see
`technical concerns in the following text). The trial high-
`lighted the difficulty in predicting embolic risk during SVG
`intervention and the need for DPD deployment from the
`outset of intervention.
`A comparison of the GuardWire and FilterWire (FIRE)
`in patients with low thrombus burden and good pre-
`procedural flow revealed similar rates of successful device
`deployment, angiographic and myonecrotic end points, and
`30-day MACE (33). Subgroup analysis suggested an ad-
`vantage of the FilterWire over the GuardWire in smaller
`vessels and eccentric lesions.
`In an early safety and feasibility study in 40 patients
`undergoing mainly SVG PCI (FASTER) (34), the Proxis
`system was successfully deployed in 95% of cases and
`appeared safe (MACE 5%). The first clinical trial with this
`device (PROXIMAL) is now under way (Table 5).
`In the X-TRACT trial (35), use of the X-Sizer system
`(70% SVG) was not associated with reductions in MACE.
`Although the overall
`incidence of procedural AMI was
`unchanged, the incidence of large infarcts was significantly
`reduced in the X-Sizer arm.
`The results of two noninferiority studies, PRIDE (Tri-
`Activ, Kensey Nash, Exton, Pennsylvania) and CAPTIVE
`(CardioShield, MedNova, Galway, Ireland), using new
`DPD in SVGs were presented at TCT 2004. The TriActiv
`system has three components: a distal protection balloon, a
`3-F flushing catheter, and a peristaltic pump extraction
`system, which allow constant flushing and aspiration of
`debris. The PRIDE trial demonstrated noninferiority to
`established DPD with respect to 30-day MACE. The
`CAPTIVE trial, which assessed the CardioShield filter
`
`TECHNICAL CONCERNS
`
`Failure to cross the lesion. The relatively large crossing
`profile and lack of torquability may make it technically
`challenging to advance these devices beyond a tight stenosis
`or in a tortuous vessel. Some devices come loaded on their
`own delivery wire, the handling characteristics of which will
`clearly be better suited to crossing some vessels than others.
`Pre-dilatation can overcome this problem, at the risk of
`distal embolization before protection is in place. The im-
`portance of such early embolization should not be underes-
`timated; significant embolization has been documented in
`ex vivo studies (36), and in patients undergoing CAS with
`transcranial Doppler monitoring (37,38).
`Positioning. In AMI with TIMI flow grade 0, it may be
`difficult to know where to position the DPD, in order to be
`far enough away from the lesion to allow unimpeded
`stenting, yet
`in a part of
`the artery large enough to
`accommodate the device, and proximal
`to major side
`branches.
`Sizing the device. Some devices, such as the FilterWire,
`expand to fit a range of vessel diameters. However, with
`other systems, the size of the distal vessel may be underes-
`timated if flow is reduced, and device malapposition due to
`undersizing may allow distal embolization.
`Side-branch protection. In Y anastomotic grafts or distal
`graft lesions where the native vessel run-off gives off early
`major side-branches, it may be difficult to position the DPD
`such that side-branches are protected. To overcome this,
`two similar devices may be used in each branch (“kissing”
`filters), assuming that the caliber of both branches is
`commensurate with use of a DPD. Alternatively, a filter-
`type device may be deployed in one branch and an
`occlusion-aspiration or thrombectomy device used in the
`other. Importantly, these techniques will be feasible for
`balloon angioplasty but may make stenting challenging, in
`order to avoid jailing the wires of either DPD during stent
`deployment.
`Persistent embolization. Using balloon occlusion systems,
`embolization may occur due to gradual balloon deflation
`during the procedure. Minor deflation may be difficult to
`detect without frequent injections of contrast to ensure a
`tight seal, but this itself may cause embolization with an
`inadequate seal. Furthermore, debris may fail to be aspirated
`by the suction catheter, either because of resistance to
`aspiration or because the catheter cannot approach suffi-
`ciently close, leaving a “suction shadow” of embolic material
`behind (39). With filter-based devices, incomplete apposi-
`tion is a significant concern and has been found to correlate
`with periprocedural AMI (32). This may be easily missed
`without orthogonal views to assess the filter both en face
`and in profile. The most common cause of malapposition is
`lifting of the nitinol frame away from one side of the vessel
`
`Page 6
`
`
`
`1440
`
`Gorog etal.
`Embolic Protection During PCI
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`Results
`(Intervention vs. Control)
`
`89%
`22 vs. 31, p ⫽ 0.005
`63 vs. 36, p ⫽ 0.006
`80% vs. 54%, p ⫽ 0.006
`236 vs. 333, p ⫽ 0.01
`7% vs. 4%, p ⫽ 0.012
`86% vs. 90%
`(GuardWire vs. FilterWire)
`84% vs. 89%
`(GuardWire vs. FilterWire)
`4.7 vs. 5.1
`(GuardWire vs. FilterWire)
`10% vs. 22%
`(GuardWire vs. FilterWire)
`95% vs. 75%, p ⫽ NS
`2.4 vs. 1.6, p ⫽ NS
`4.7% vs. 16.5%, p ⫽ 0.006
`
`p ⫽ NS
`p ⫽ NS
`89% vs. 96% p ⫽ NS
`2.7%
`96%
`50%
`42%
`p ⫽ NS
`p ⫽ NS
`
`90% vs. 84%, p ⫽ NS
`18% vs. 25%, p ⬍ 0.03
`83% vs. 52%, p ⫽ NS
`6% vs. 6%, p ⫽ NS
`142 vs. 171, p ⫽ NS
`11% vs. 11%, p ⫽ NS
`52% vs. 50%, p ⫽ NS
`182% vs. 229%, p ⫽ NS
`94% vs. 96%, p ⫽ NS
`72% vs. 37%, p ⫽ 0.006
`83% vs. 52%, p ⫽ 0.001
`
`2.7% vs. 8.2%, p ⫽ NS
`
`Table 2. Trials Employing Embolic Protection in ACS
`Trial
`Name
`
`Clinical Syndrome
`
`Device
`
`Limbruno
`et al. (6)
`
`Orrego et al.
`Cath
`Cardiovasc
`Int 2003
`
`Primary PCI
`
`FilterWire
`
`Primary PCI
`
`FilterWire vs.
`GuardWire
`
`Yip et al. (20)
`
`Primary PCI (high
`thrombus burden)
`
`GuardWire
`
`EMERALD
`
`AMI, primary or rescue
`PCI
`
`RUBY
`
`ACS, predominantly
`STEMI
`
`GuardWire
`
`GuardWire
`
`PROMISE
`
`AMI (primary PCI)
`
`FilterWire
`
`Beran et al. (21)
`
`ACS, predominantly
`STEMI
`
`X-Sizer
`
`Napodano
`et al. (22)
`
`AMI with angiographic
`thrombus
`
`X-Sizer
`
`No. of
`Patients
`
`53
`
`10
`23
`
`108
`
`500
`
`328 total; results
`of first 188
`available
`
`200
`
`66
`
`92
`
`TOPIT
`
`Thrombus-rich vessels
`
`TEC
`
`251
`
`Prospective randomized trial,
`randomized to TEC vs. no
`thrombectomy pre-PCI
`
`Management Strategy
`
`Other Agents Used
`
`End Point
`
`Registry; matched historical
`controls
`
`Non randomized, first 22
`consecutive patients treated
`with GuardWire, next 10
`with FilterWire
`
`Abciximab
`
`Registry; matched historical
`controls treated with PCI
`with tirofiban, without
`GuardWire
`Prospective, randomized trial,
`GuardWire vs. no DPD
`
`Prospective registry
`
`Tirofiban
`
`GP IIb/IIIa inhibitors in
`83%
`
`GP IIb/IIIa inhibitors in
`35%
`
`Prospective, randomized trial
`
`Abciximab
`
`Prospective, randomized trial,
`X-Sizer vs. no
`thrombectomy
`
`GP IIb/IIIa inhibitors in
`⬃70% of both study
`arms
`
`Prospective randomized trial,
`randomized to X-Sizer vs.
`no thromectomy
`
`GP IIb/IIIa inhibitors in
`43% of both study
`arms
`
`Success of DPD delivery
`Post-procedure cTFC
`Post-procedure MBG 3
`ST-segment resolution
`Peak CKMB
`30-day LVEF
`Success of DPD delivery
`Final TIMI flow grade 3
`ST-segment score
`Distal embolization
`
`Final TIMI flow grade 3
`Final MBG
`30-day MACE
`
`(1°) ST-segment resolution
`(1°) Infarct size on sestamibi
`(2°) TIMI flow grade 3
`(1°) 30-day MACE
`(2°) TIMI flow grade 3
`(2°) Normal MBG
`(2°) ST-segment resolution
`(1°) Max flow velocity in IRA
`(2°) Infarct size (% LV mass on
`MRI)
`TIMI flow grade 3
`cTFC
`ST-segment resolution
`30-day MACE
`Peak CKMB
`30-day MACE
`30-day LVEF
`Peak CKMB
`Post-procedure TIMI flow
`grade 3
`Post-procedure MBG3
`ST-segment resolution
`In-hospital MACE
`
`Page 7
`
`
`
`Gorog et al.
`Embolic Protection During PCI
`
`1441
`
`Table 3. Summary of Trial Acronyms
`AIMI ⫽ AngioJet Rheolytic Thrombectomy in Patients Undergoing
`Primary Angioplasty for AMI
`AMEthyst ⫽ Assessment of the Medtronic AVE Interceptor Saphenous
`Vein Graft Filter System
`CAPTIVE ⫽ Cardio-Shield Application Protects During Transluminal
`Intervention of Vein Grafts by Reducing Emboli
`DEAR-MI ⫽ Dethrombosis to Enhance Acute Reperfusion in
`Myocardial Infarction
`EMERALD ⫽ Enhanced Myocardial Efficacy and Recovery by
`Aspiration of Liberalized Debris
`FASTER ⫽ Evaluation of the Proxis embolic protection system during
`stenting of coronary arteries and degenerated saphenous vein grafts
`FIRE ⫽ FilterWire EX During Transluminal Intervention of
`Saphenous Vein Grafts
`GUARD ⫽ SVG Intervention Using AngioGuard for reduction of
`Distal Embolization
`PRIDE ⫽ PRotection During Saphenous Vein Graft Intervention to
`Prevent Distal Embolization
`PROMISE ⫽ Protection Devices in PCI-Treatment of Myocardial
`Infarction for Salvage of Endangered Myocardium Study
`PROXIMAL ⫽ PROXimal Protection During Saphenous Vein Graft
`Intervention Using the Proxis Embolic Protection System: A
`Randomized, Prospective, Multicenter TriAL
`RUBY ⫽ Revascularization Utilizing Balloon protection in Acute
`CoronarY Ischemic Syndrome
`RULE SVG ⫽ RUbicon FiLtEr SVG
`SAFE ⫽ Saphenous Vein Graft Angioplasty Free of Emboli trial
`SAFER ⫽ Saphenous Vein Graft Angioplasty Free of Emboli
`Randomized trial
`SAPPHIRE ⫽ Stenting and Angioplasty with Protection in Patients at
`High Risk for Endarterectomy
`SPIDER ⫽ Evaluation of the Spider embolic protection system during
`stent treatment of degenerated saphenous vein grafts
`TOPIT ⫽ TEC or PTCA in Thrombus
`VeGAS-2 ⫽ Vein Graft AngioJet Study 2
`X-TRACT ⫽ Prospective, randomized evaluation of thrombectomy
`prior to PCI in diseased SVGs and thrombus-containing coronary
`arteries
`
`For further information, refer to: http://www.tctmd.com.
`
`wall due to wire bias (32), and may be corrected by
`repositioning the filter.
`The benefit of filter devices over balloon occlusion sys-
`tems is the preservation of flow. This is a two-edged sword
`because it also allows the passage of smaller microemboli.
`The SAFE registry revealed that 80% of particles collected
`were smaller than 100 m diameter, although it is impos-
`sible to know how much suction/aspiration contributed to
`particle break up (29).
`Embolization may also occur during device retrieval.
`Filters may become full, and may spill their contents when
`collapsed during retrieval. Newer devices will incorporate a
`built-in shutter mechanism to close the mouth of the filter
`before retrieval.
`Retrieval. The relatively large-profile aspiration catheter or
`the retrieval catheter of the filter devices may become caught
`in stent struts.
`Use in small vessels. Use of either DPD or thrombectomy
`is generally recommended for vessels ⬎3.0 to 3.5 mm in
`diameter. Recently, the first small study employing the
`FilterWire in small native vessels (2.6 ⫾ 0.5 mm diameter)
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`graft.OtherabbreviationsasinTable1.
`cardiacevents;MBG⫽myocardialblushgrade;MRI⫽magneticresonanceimaging;NSTEMI⫽non-ST-segmentelevationmyocardialinfarction;PCI⫽percutaneouscoronaryintervention;STEMI⫽ST-segmentelevationmyocardialinfarction;SVG⫽saphenousvein
`ACS⫽acutecoronarysyndromes;AMI⫽acutemyocardialinfarction;CKMB⫽creatinekinase-MB;CTFC⫽correctedTIMIframecount;DPD⫽distalprotectiondevice;GP⫽glycoprotein;IRA⫽infarct-relatedartery;LV⫽leftventricle;MACE⫽majoradverse
`
`p⫽NS
`p⫽NS
`p⫽NS
`p⫽NS
`12.5%vs.9.8%,p⫽0.02
`
`(2°)FinalMBG
`(2°)FinalcTFC
`
`grade
`
`(2°)post-procedureTIMIflow
`(2°)ST-segmentresolution,
`
`sestamibi
`
`trialpublication
`
`(1°)30-dayinfarctsizeon
`
`Uncertain,pendingfull
`
`29%vs.30%,p⫽NS
`12%vs.25%,p⬍0.05
`
`(1°)30-daycompositeMACE
`NSTEMI
`
`arms
`14%ofbothstudy
`
`GPIIb/IIIainhibitorsin
`
`nothrombectomypre-PCI
`RandomizedtoAngioJetvs.
`
`PCI
`Urokinaseinfusionbefore
`RandomizedtoAngioJetvs.
`
`(Interventionvs.Control)
`
`Results
`
`EndPoint
`
`OtherAgentsUsed
`
`ManagementStrategy
`
`480
`
`346
`
`Patients
`No.of
`
`AngioJet
`
`required)
`thrombusNOT
`AMI(angiographic
`
`nativeorSVG
`
`AIMI
`
`AngioJet
`
`Angiographicthrombusin
`
`VegAS-2(23)
`
`Device
`
`ClinicalSyndrome
`
`Name
`Trial
`
`Table2Continued
`
`Page 8
`
`
`
`1442
`
`Gorog etal.
`Embolic Protection During PCI
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`Table 4. Trials Using Embolic Protection in SVG Intervention
`Trial
`Clinical
`Name
`Syndrome
`
`Device
`
`SAFE
`
`Elective SVG intervention
`(low thrombus burden)
`
`GuardWire
`
`No. of
`Patients
`
`105
`
`Management
`Strategy
`
`Other Agents
`Used
`
`Registry
`
`SAFER
`
`Stone et al.
`(32)
`
`FIRE
`
`Elective SVG intervention
`(moderate thrombus
`burden)
`Elective SVG intervention
`(High thrombus burden)
`
`Elective SVG intervention
`(Low thrombus burden)
`
`GuardWire
`
`801
`
`Randomized to GuardWire
`vs. no DPD
`
`FilterWire
`
`60 (phase 1)
`248 (phase 2)
`
`Registry
`
`FilterWire vs. GuardWire
`
`651
`
`Prospective randomized trial
`
`X-TRACT
`
`SVG (70%) or thrombus-
`rich native vessel (30%)
`
`X-Sizer vs. no DPD
`
`50 (phase 1)
`797 (phase 2)
`
`Prospective randomized trial
`
`GP IIb/IIIa inhibitor
`in ⬃58% lesions in
`both arms
`GP IIb/IIIa inhibitor
`in 30% lesions
`(phase 1), in 52%
`lesions (phase 2)
`GP IIb/IIIa inhibitor
`in ⬃52% lesions in
`both arms
`
`GP IIb/IIIa inhibitor
`in ⬃76% lesions in
`both arms
`
`PRIDE
`
`SVG
`
`TriActive vs.
`FilterWire/GuardWire
`
`CAPTIVE
`
`SVG
`
`CardioShield vs.
`GuardWire
`
`ULN ⫽ upper limits of normal; other abbreviations as in Tables 1 and 2.
`
`631
`
`652
`
`Prospective randomization to
`TriActive system vs.
`established DPD
`(FilterWire or
`GuardWire)
`Prospective randomization to
`CardioShield vs.
`GuardWire
`
`Uncertain, pending
`full trial
`publication
`
`Uncertain, pending
`full trial
`publication
`
`End Point
`
`(1°) In-hospital MACE
`(2°) Final TIMI grade 3
`flow
`(2°) No reflow
`(2°) CKMB 3⫻ ULN
`30-day MACE
`No reflow
`
`30-day MACE
`
`(1°) 30-day MACE successfull
`DPD delivery
`Angiographic and
`myonecrotic end points
`30-day MACE
`1-year MACE
`Periprocedural AMI (any)
`Large AMI (CKMB ⬎8⫻
`ULN)
`30-day MACE
`
`Results
`(Intervention vs. Control)
`
`5%
`99%
`0%
`5%
`
`3% vs. 9%, p ⫽ 0.02
`9.6% vs. 16.5%, p ⫽ 0.004
`
`21% (phase 1)
`11% (phase 2)
`
`10% vs. 12% (FilterWire vs.
`GuardWire)
`p ⫽ NS for superiority
`p ⫽ 0.008 for inferiority
`17% vs. 17%, p ⫽ NS
`31% vs. 28%, p ⫽ NS
`16% vs. 17%, p ⫽ NS
`5% vs. 10%, p ⫽ 0.002
`
`11% vs. 10%, p ⫽ NS
`
`30-day MACE
`
`10% vs. 12%, p ⫽ NS
`
`Page 9
`
`
`
`JACC Vol. 46, No. 8, 2005
`October 18, 2005:1434–45
`
`Gorog et al.
`Embolic Protection During PCI
`
`1443
`
`Table 5. Ongoing Trials Using Embolic Protection in Coronary Intervention
`Type of
`Investigational
`Device
`
`Device
`
`No. of
`Patients
`
`Management Strategy
`
`Trial Name
`
`Clinical
`Syndrome
`
`Primary End Point
`
`30-day MACE
`
`-Myocardial reperfusion
`(ST-segment resolution)
`-Echocardiographic
`recovery of LV
`function
`30-day MACE
`
`30-day MACE
`
`Results
`
`Ongoing
`
`Ongoing
`
`Ongoing
`
`Ongoing
`
`(Protocol not published) Ongoing
`
`30-day MACE
`
`Ongoing
`
`PROXIMAL SVG intervention
`
`PPD
`
`Proxis
`
`DEAR-MI
`
`AMI
`
`Thrombectomy
`
`Pronto
`
`RULE-SVG PCI to SVG
`
`SPIDER
`
`PCI to SVG
`
`GUARD
`
`PCI to SVG
`
`Filter
`
`Filter
`
`Filter
`
`Rubicon
`
`SpideRX
`
`AngioGuard
`
`600
`
`200
`
`60
`
`770
`
`800
`
`AMEthyst
`
`PCI to SVG
`
`Filter
`
`Interceptor
`
`600
`
`Randomized to Proxis or
`other (FDA approved)
`DPD
`Randomized to Pronto
`thrombectomy vs. no
`treatment prior-to
`primary PCI
`
`Randomized to Rubicon
`vs. no protection
`Randomized to SpideRX
`or GuardWire
`Randomized to
`AngioGuard or
`GuardWire
`Randomized to
`Interceptor or
`GuardWire
`
`PPD ⫽ proximal protection device; other abbreviations as in Tables 1 and 2.
`
`with moderate-complex lesions has revealed high rates of
`procedural success, no device-related vessel dissection, and
`distal embolization in only 4% patients (40). Although
`vasospasm (50%) and reduction in coronary flow (45%) were
`frequent, these universally resolved after device retrieval.
`Use in large vessels. Degenerative SVGs can become
`markedly ectatic, and the risk of no reflow during PCI is
`particularly high. Although DPD do not generally expand
`to ⬎6 mm in diameter, a larger range of devices is expected
`to become available soon.
`Uncertain clinical scenarios. Almost all available clinical
`data are in men, with very limited data in diabetics. There
`are no data on the tolerability of the balloon occlusion
`systems in patients with poor left ventricular function or
`AMI and cardiogenic shock. Studies are needed to deter-
`mine whether combined treatment with both thrombec-
`tomy and DPD will confer additional advantages. Whether
`use of GP IIb/IIIa inhibitors or lytic therapy combined with
`DPD will offer additional benefit, or whether it reduces the
`size of embolizing particles such that these pass through
`filters unhindered, remains to be determined. In the X-Sizer
`AMI registry, where 43% of patients received abciximab
`(22), no difference in TIMI flow grade 3 or MBG was
`found according to abciximab use.
`
`FILTER, BALLOON-OCCLUSION,
`AND ASPIRATION, OR THROMBECTOMY?
`
`In deciding on the best protection strategy, it is important to
`consider specific lesion and vessel characteristics, as well as
`thrombus burden. Electron microscopy of material aspirated
`with the GuardWire system showed particles ranging from
`3.6 to 5,262 m in diameter, 50% of which were ⬍100 m
`(29,41). Filter devices have 80- to 150-m pores, and may
`therefore allow significantly more debris to pass through.
`
`Although smaller pore sizes might appear advantageous,
`these also cause more frequent filter thrombosis (42) with a
`size of ⬃100 m representing a reasonable compromise.
`Why then do filters with 100-m pores appear to offer equal
`benefit to balloon occlusion systems when half the particles
`are under that size? Although it is possible that th