Author + information
- Received December 16, 2004
- Revision received February 12, 2005
- Accepted February 22, 2005
- Published online June 7, 2005.
- Ralf Zahn, MD, FESC⁎,⁎ (, )
- Thomas Ischinger, MD, FESC†,
- Bernd Mark, MD⁎,
- Sabine Gass, MD‡,
- Uwe Zeymer, MD, FESC⁎,
- Wolfgang Schmalz, MD§,
- Klaus Haerten, MD∥,
- Karl Eugen Hauptmann, MD¶,
- Enz-Rüdiger von Leitner, MD#,
- Wolfgang Kasper, MD⁎⁎,
- Ulrich Tebbe, MD, FESC††,
- Jochen Senges, MD, FACC, FESC⁎,
- Arbeitsgemeinschaft Leitende Kardiologische Krankenhausärzte (ALKK)
- ↵⁎Reprint requests and correspondence:
Dr. Ralf Zahn, Herzzentrum Ludwigshafen, Dept. of Cardiology, Bremserstrasse 79, D-67063 Ludwigshafen, Germany.
Objectives We sought to compare the efficacy of a filter embolic protection device (F-EPD) and a distal occlusive embolic protection device (DO-EPD) in patients undergoing carotid artery stenting (CAS).
Background The embolic protection device (EPD) may lower the periprocedural rate of cerebral ischemic events during CAS. However, it is unclear whether there is a difference in effectiveness between the different types of EPD.
Methods We analyzed data from the Carotid Artery Stent (CAS) Registry.
Results From July 1996 to July 2003, 1,734 patients were included in the prospective CAS Registry. Of these patients, 729 patients were treated with an EPD, 553 (75.9%) with F-EPD, and 176 (24.1%) with DO-EPD. Patients treated with DO-EPD were more likely to be treated for symptomatic stenosis (64.5% vs. 53.4%, p = 0.011). The carotid lesions in patients treated under DO-EPD seemed to be more complicated, as expressed by a higher proportion of ulcers (p = 0.035), severe calcification (p = 0.039), a longer lesion length (p = 0.025), and a higher pre-interventional grade of stenosis (p < 0.001). The median duration of the CAS intervention was 30 min in the DO-EPD group, compared with 48 min in the filter group (p < 0.001). No differences in clinical events rate between the two groups of protection devices were observed. Multivariate analysis on the occurrence of the combined end point of in-hospital death or stroke found no difference between filter- and DO-EPD (4 of 176 [2.3%] for DO-EPD vs. 10 of 551 [1.8%] for F-EPD; adjusted odds ratio = 1.04, 95% confidence interval 0.24 to 4.44; p = 0.958).
Conclusions Filter EPD is the currently preferred method of EPD in clinical practice. Both F-EPD and DO-EPD seem to be equally effective during CAS.
Carotid artery stenting (CAS) increasingly emerges as an alternative to carotid endarterectomy (CEA) for the treatment of carotid stenoses (1–3). Both CEA and CAS carry the burden of a small but clinically important rate of periprocedural complications—mainly, cerebral ischemic events.
Embolic protection devices (EPDs) may lower the rate of ipsilateral ischemic events during CAS. This hypothesis is supported by data from registries (4–6), as well as an analysis from the Endarterectomy Versus Angioplasty in Patients With Symptomatic Severe Carotid Stenosis (EVA-3S) trial (7). However, there are no data from randomized, controlled clinical trials on this issue.
Embolic protection can be achieved by different types of devices, such as filter embolic protection devices (F-EPD), distal occlusion embolic protection devices (DO-EPD), and reverse flow devices (8–10). Experimental studies have shown the ability of the different protection types to reduce the amount of embolized material (11–14). However, until now, there are only data from randomized trials from the coronary circulation for the treatment of saphenous vein bypass grafts with percutaneous coronary interventions, showing that F-EPD is as effective as DO-EPD in reducing ischemic events in a clinical scenario (15,16).
To determine the use and compare the effectiveness of F-EPD and DO-EPD during CAS, we analyzed the data of the prospective Carotid Artery Stent (CAS) Registry of the German Arbeitsgemeinschaft Leitende Kardiologische Krankenhausärzte (ALKK).
The CAS Registry of the ALKK is an ongoing registry that was initiated in 1996; it has been described in detail previously (4). In brief, all interventions were prospectively enrolled in the registry. All complications occurring in the catheter room and during the in-hospital phase were documented. All data from the initiation in 1996 until the end of July 2003 were analyzed for this report.
All patients undergoing CAS for a carotid stenosis at the participating hospitals were considered for this analysis. Patients with symptomatic carotid stenoses had a history of an ipsilateral stroke, at least one transient ischemic attack (TIA), or an episode of amaurosis fugax. In these patients, an angiographically documented stenosis ≥70%, according to the diagnostic criteria of the North American Symptomatic Carotid Endarterectomy Trial (NASCET) (17), was suggested to be an indication for an intervention. In the case of asymptomatic patients, a stenosis ≥80% was suggested as an indication for an intervention. The decision to treat a given patient was left to the treating physician.
Amaurosis fugaxwas defined as retinal ischemia with transient monocular blindness, and a TIA as a focal neurologic deficit that resolved spontaneously within 24 h. Strokewas defined as a loss of neurologic function caused by an ischemic or hemorrhagic event, with residual symptoms at least 24 h after onset. A minor strokewas diagnosed if symptoms disappeared within one week after onset, and a major strokeif symptoms persisted for at least more than one week after onset.
No differentiation between ischemic, hemorrhagic, and strokes with unknown etiology and no quantification of the severity of stroke (such as the National Institutes of Health Stroke Scale (18), besides its reversibility, were made.
End points were evaluated by a neurologist either immediately at the occurrence of symptoms or at the end of the hospital stay. We did not separately evaluate 30-day follow-up end points. The combined clinical end point of all deaths and all strokes was prospectively defined as the primary end point.
The CAS procedure
The CAS procedure was performed according to the standard protocol of each participating center.
Every participating center was committed by written consent to include each patient during the study period. The patients gave informed consent for the processing of their anonymous data.
Absolute numbers and percentages are computed to describe the patient population. Median values and quartiles are computed as appropriate. Categorical values were compared by the chi-square test, and crude odds ratios (ORs) with 95% confidence intervals (CIs) were computed. Continuous variables were compared by the two-tailed Wilcoxon rank-sum test. Logistic regression analysis was used to analyze selection of the use of F-EPD instead of DO-EPD. The following variables were examined: age (dichotomized to age ≥70 and <70 years), gender, presence of thrombus, symptomatic carotid stenosis, previous carotid intervention (CAS or surgery), caseload of the hospitals for CAS, and year of inclusion. Logistic regression analyses were also used to adjust for factors influencing the combined clinical end point of all deaths and all strokes. The following variables were examined: age (dichotomized to age ≥70 and <70 years), gender, presence of coronary heart disease, presence of symptomatic carotid stenosis, previous carotid intervention (CAS or CEA), presence of thrombus, presence of calcified lesions, presence of carotid ulcer, caseload of the hospitals (dichotomized to ≤35 and >35 reported CAS/hospital), and type of EPD. A second regression analysis on this end point was performed with age and caseload as continuous variables. We used the cstatistic to assess the association between predicted probabilities and observed response of each logistic regression model. A p value <0.05 was considered significant. All p values are results of two-tailed tests. The calculations were performed using SAS, version 8.02 (Cary, North Carolina).
From July 1996 to July 2003, 1,734 patients from 26 hospitals were included in the prospective CAS Registry. Of 729 patients treated with distal protection, with information on the type of protection device used, 553 (75.9%) were treated with F-EPD and 176 (24.1%) with DO-EPD. The use of EPD grew rapidly over time. Initially, the DO-EPD was used more frequently, but with the introduction of F-EPD, it became the predominantly used device (Fig. 1).Although the GuardWire (Medtronic, Santa Rosa, California) was the only DO-EPD used in our registry, different types of filters were used, as shown in Figure 2.
Patients treated with DO-EPD were more likely to be treated for symptomatic stenosis (64.5% vs. 53.4%, p = 0.011). However, they had a lower prevalence of previous carotid artery dilation (0% vs. 4.3%, p = 0.005). The prevalence of concomitant diseases in both groups is also given in Table 1.
Lesion and interventional characteristics
Lesions in patients treated with DO-EPD seemed to be more complicated, as demonstrated by a higher proportion of ulcers (p = 0.035), severe calcification (p = 0.039), a longer lesion length (p = 0.025), and a higher pre-interventional grade of stenosis (p < 0.001). Pre-dilation without embolic protection was more often performed in the DO-EPD group (53% vs. 26.5%, p < 0.001). The median duration of the CAS intervention was 18 min longer in the F-EPD group (p < 0.001) (Table 2).
Selection of type of EPD
A logistic regression model analysis on the factors being independently associated with the use of a filter instead of DO-EPD revealed the year of implantation (the most current years associated with a more frequent use of filters) to be the most powerful predictor (Table 3).
Concomitant medication and clinical events
No difference in the use of aspirin, ticlopidine or clopidogrel, phenprocoumon, and statins was found between the two groups. However, angiotensin-converting enzyme inhibitors (p = 0.002) and beta-blockers (p = 0.006) were more often prescribed in patients treated with DO-EPD during CAS (Table 4).
No differences in clinical event rates between the two groups of EPD were observed (Table 4). Logistic regression analysis on the occurrence of in-hospital death or stroke, adjusting for possible confounding parameters, found no difference between DO-EPD and F-EPD (2.3% vs. 1.8%, respectively; multivariable OR = 1.04, 95% CI 0.24 to 4.44; p = 0.958; c= 0.688). This result did not change if age and caseload were included as a continuous variable in the multivariate regression model. In patients with symptomatic stenoses, the hospital rates of death or any stroke were 3.6% in the DO-EPD group and 2.6% in the F-EPD group (p = 0.74). In asymptomatic patients, the corresponding rates were 0% versus 1.3% (p = 1.0).
Major findings of current analysis
Our study in a large series of patients undergoing CAS with different EPD shows that currently F-EPD is by far the most preferred EPD. The type of protection device was not associated with differences in acute clinical outcome during CAS.
Selection of type of protection device and interventional characteristics
The DO-EPD was the first EPD that became commercially available. With the introduction of filter devices, it rapidly conquered the market, being used in >90% of protected CAS in 2003 in our registry. We found no other major important factor besides the year of recruitment that was associated with the use of F-EPD instead of DO-EPD. This predilection for F-EPD must have been driven by practical or marketing reasons, for there are no data on the superiority of filters over DO-EPD during CAS. One factor in favor of filters is the avoidance of complete interruption of the blood stream via the internal carotid artery caused by DO-EPD. Such an interruption of blood flow is associated with temporal neurologic deficits in ∼5% of cases (19). Furthermore, the interruption of blood flow also makes an angiographic guidance of the stent implantation more difficult. One advantage of DO-EPD may be that even very small particles can be captured, whereas in the case of filter devices, small particles can pass the filter membrane. However, the clinical relevance of this difference has not been determined yet. A third type of EPD—the reverse flow system (12,20)—was not used in our registry. Therefore, we cannot comment on it.
The median duration of the CAS intervention was 30 min in the DO-EPD group, compared with 48 min in the filter group (p < 0.001). The reason for this may be that the temporal occlusion of the internal carotid artery by DO-EPD puts the investigator under time pressure to quickly terminate the intervention in order to avoid cerebral ischemic symptoms of the patient.
Clinical events with filter versus DO-EPD
Although there are only data from randomized clinical trials on the use of EPD in saphenous vein graft interventions (15,16), the principle of embolic protection during CAS seems so convincing that these devices have already conquered the market. As shown previously (4), our data suggest a small but significant reduction of embolic events with the use of these devices during CAS: all nonfatal strokes and all deaths with (2.1%) versus without (4.9%) protection devices (p = 0.004). This finding is confirmed by other registry data (5,6) and the clinical alert from the EVA-3S trial (7).
However, there are no data comparing the clinical effectiveness of F-EPD versus DO-EPD during CAS. In our study, the combined clinical end point of death or stroke was 2.3% in patients treated with DO-EPD versus 1.8% in CAS patients treated with F-EPD (p = 0.700). After adjusting for confounding parameters, the multivariable OR was 1.04 (95% CI 0.24 to 4.44; p = 0.958). There were also no significant differences in the occurrence of the different types of ipsilateral ischemic events.
Experimental data showed that different types of EPD can reduce the rate of emboli (11,12). An experimental study by Müller-Hülsbek et al. (21) comparing F-EPD with DO-EPD showed no differences in the rate of embolization in the internal carotid artery, but there was a difference in the embolization rates in the external carotid artery. A further investigation of the same group, including different types of filters, showed significant differences concerning the embolization rates between the investigated filters (13). An in vivo study of Rogers et al. (22) demonstrated that the investigational Medtronic AVE filter EPD could retrieve particulate during saphenous vein graft stenting in a similar amount and character as the PercuSurge Medtronic GuardWire.
The randomized, controlled clinical trial that established the effectiveness of EPD was done using a DO-EPD for the percutaneous treatment of coronary saphenous venous bypass grafts (15). Further trials comparing different EPDs revealed conflicting results: some showed a noninferiority of the investigated EPD (16,23), and another failed to demonstrate noninferiority (24), as compared with the reference DO-EPD (i.e., the GuardWire by Medtronic). These trials show that the different types of EPD (F-EPD, DO-EPD, or reverse flow devices) have about the same potential in reducing the amount and characteristics of embolizing particulate. However, due to differences in the design of these devices, each one has to demonstrate its own efficacy.
Because the CAS Registry of the ALKK is an observational study, it is not possible to control totally for the selection of patients to be treated with one of the two types of EPD. In particular, the later introduction of F-EPD into clinical practice may have interfered with the growing experience of the investigators, a factor that cannot be controlled for. However, given the low complication rates in both groups (F-EPD vs. DO-EPD), it is very unlikely that there are clinically relevant differences between the two types of EPD.
Also, different types of F-EPD, as well as different stages of their development, were combined and could not be evaluated separately due to low numbers. Therefore, no conclusions on equivalence or superiority between the different F-EPD devices can be drawn.
We obtained only data until hospital discharge. Therefore, it may be difficult to compare our data to 30-day outcome data of other studies. However, clinical events are extremely low after the first days of stent implantation (25,26).
Institutions and people who participated in the ALKK CAS Registry are listed elsewhere (4).
- Abbreviations and Acronyms
- Arbeitsgemeinschaft Leitende Kardiologische Krankenhausärzte
- carotid artery stenting
- carotid endarterectomy
- confidence interval
- distal occlusive embolic protection device
- embolic protection device
- filter embolic protection device
- odds ratio
- Received December 16, 2004.
- Revision received February 12, 2005.
- Accepted February 22, 2005.
- American College of Cardiology Foundation
- Roubin G.S.,
- New G.,
- Iyer S.S.,
- et al.
- Zahn R.,
- Mark B.,
- Niedermaier N.,
- et al.
- Kastrup A.,
- Groschel K.,
- Krapf H.,
- et al.
- The EVA-3S Investigators
- Martin J.B.,
- Pache J.C.,
- Treggiari-Venzi M.,
- et al.
- Müller-Hülsbeck S.,
- Jahnke T.,
- Liess C.,
- et al.
- Baim D.S.,
- Wahr D.,
- George B.,
- et al.
- Stone G.W.,
- Rogers C.,
- Hermiller J.,
- et al.
- Brott T.,
- Adams H.P. Jr.,
- Olinger C.P.,
- et al.
- Whitlow P.L.,
- Lylyk P.,
- Londero H.,
- et al.
- Rogers C.,
- Huynh R.,
- Seifert P.A.,
- et al.
- Carrozza J.P.J.
- Holmes D.R.
- Cremonesi A.,
- Manetti R.,
- Setacci F.,
- et al.