Send to

Choose Destination
See comment in PubMed Commons below
J Endovasc Ther. 2004 Apr;11(2):211-8.

Comparison of 4 cerebral protection filters for carotid angioplasty: an in vitro experiment focusing on carotid anatomy.

Author information

Department of Radiology, University Hospital, Arnold-Heller-Strasse 9, 24105 Kiel, Germany.



To assess the influence of internal carotid artery (ICA) tortuosity on the effectiveness of 4 cerebral protection filters in an in vitro bench-top model.


To simulate the anatomical arterial variants, 3 open flow models were constructed: one with normal carotid anatomy, 1 representing a mildly tortuous ICA, and a third imitating a severely tortuous ICA. Polyvinyl alcohol particles (150-1000 microm) served as the embolic material; the emboli were divided into 3 groups according to size: small (150-250 microm), medium (355-500 microm), and large (710-1000 microm). Five milligrams of each size group were injected separately into the ICA proximal to each of 4 protection filters: AngioGuard, FilterWire EX, TRAP, and NeuroShield. Emboli that were not caught by the protection system or were washed into the external carotid artery (ECA) ran into an effluent filter and were weighed and classified according to size.


In 240 test runs, the FilterWire EX presented the lowest weight of emboli in the ICA effluent under all anatomical conditions: normal anatomy 0.39 mg (2.58%), mild tortuosity 0.45 mg (2.99%), and severe tortuosity 0.50 mg (3.33%) (p>0.05). The Angioguard system showed the worst results: normal 1.21 mg (98.03%), mild tortuosity 2.54 mg (16.84%), and severe tortuosity 3.14 mg (20.91%) (p<0.001) compared to the FilterWire EX and the NeuroShield systems. The only protection device displaying no significant differences in all of the 3 emboli sizes was the FilterWire EX (p>0.05). Relevant differences in effectiveness in mildly and severely tortuous ICAs were apparent among the other devices. For all emboli sizes, the differences of the AngioGuard and TRAP systems were highly significant (p<0.001). There was no significant difference between the FilterWire EX and the NeuroShield.


The only protection device showing no significant decrease in efficacy in the tortuous ICA models was the FilterWire EX. In both tested anatomical variants, the protection systems ranked in the same order of effectiveness. None of the tested devices prevented embolization completely.

[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Atypon
    Loading ...
    Support Center