576 #{149} Radiology August 1996 Aneurysm Clip Testing for Ferromagnetic Properties: Clip Variability Issues’ Emanuel Kanal, MD Frank C. Shellock, PhD Jonathan S. Lewin, MD To assess ferromagnetic properties of intracranial aneurysm clips reported to be nonferromagnetic, 1,765 Yasargil, 11 Sugita, and 15 Pemeczky aneurysm clips were studied for rotation or trans- lation on plate glass in a 13-T MR im- ager. Sixty-three clips (52 Yasargil, 11 Perneczky) weakly reoriented along the static magnetic resonance (MR) field. These results confirm the need for standardized testing for ferromag- netic properties for implantable metal- lic devices. Index terms: Aneurysm, cerebral, I 7.73 ‘ Head, MR. 10.1214 #{149} Magnetic resonance (MR), physics #{149} Magnetic resonance (MR), safety #{149} Radiology and radiologists, iatrogenic injury Radiology 1996; 200:576-578 UMEROUS articles have appeared at- testing to the magnetic resonance (MR) compatibility and presence or ab- sence of ferromagnetic properties of various types of metallic devices and implants (1-13). For most MR sites, these studies have been routinely used as the mainstay for assessing whether a patient with a given implant would be able to safely undergo an MR examina- tion. Nevertheless, as recently as March 1993 the U.S. Food and Drug Adminis- tration (FDA) suggested that findings in these articles and studies not be re- lied on to attempt to determine the MR compatibility of any device (4,13). The FDA explains that there are no current consensus standards for assessing fer- romagnetic properties or for determin- ing MR compatibility in general. Fur- From the Department of Radiology, Uni- versity of Pittsburgh Medical Center, 200 Lothrop St. Rm D-i32, Pittsburgh, PA 15213- 2582 (E.K.); the Department of Radiology, Uni- versity of Southern California, Los Angeles (F.G.S.); and the Department of Radiology, University Hospitals of Cleveland and Case Western Reserve University, Cleveland, Ohio (J.S.L.). From the 1994 RSNA scientific assem- bly. Received September 8, 1995; revision re- quested October 20; revision received March 27, 1996; accepted April 8. Address reprint re- quests to E.K. #{176}RSNA, 1996 ther, the composition and design of im- plants may change without the need for the manufacturer to notify the FDA (7,13). Finally, there is yet no mandate that statements need to be made in package labeling in regard to MR corn- patibility in general or to ferromagnetic properties in particular for any implant or device. To further assess these statements, we conducted a study to evaluate the ferromagnetic properties of multiple aneurysm clips from various manufac- turers. These clips included models that have been claimed by their manufac- turer to be nonferrornagnetic and MR compatible. As static magnetic fields and their associated static gradients can cause rotational as well as translational motion of ferromagnetic objects, both types of motion were assessed. Materials and Methods A 28 x 36-cm sheet of plate glass was washed and dried. We tested 1,765 Yasargil FE intracranial aneurysm clips (Aesculap, San Francisco, Calm of van- ous sizes, shapes, and overall morphol- ogy that were obtained from the store- rooms of the University of Pittsburgh Medical Center. All of these clips had already been removed from the sterile packaging in which they had been mi- tially distributed by the manufacturer. Many of these clips had been nestenil- ized, perhaps on numerous occasions, in the past. Several clips at a time were placed onto the glass in such a way that no clip touched any other. Curved clips were manually positioned with the convex surface oriented downward on the glass to decrease the frictional coef- ficient with the glass, thus facilitating rotation. The clips were placed in a ran- dom orientation relative to the static magnetic field of the imaging system. Some of the Yasargil FE clips were temporary clips. Although such clips are not designed to be permanently im- planted, the similarity in the manufac- turing and testing processes allowed a greaten number of clips to be tested. Further, rare cases may exist (Yonas H, oral communication, 1995) in which a temporary clip may be left in place de- spite the fact that this is not necorn- mended by the manufacturer. We also tested 54 Yasargil ED clips (Aesculap) which, according to the manufacturer, were all distributed prior to 1983. All of these clips were pre- sumed to have been handled exten- sively and perhaps restenilized on nu- merous occasions, although their pre- cise handling history was not known. We also tested II Sugita aneurysm clips (Mizuho Medical, Tokyo, Japan) in the same manner. All of these clips were freshly distributed from the manufacturer and removed from their sealed packages immediately prior to testing and were not handled in any other way prior to testing. We also tested 15 Perneczky (von Zeppelin Surgical Instruments, Pullach, Germany) aneurysm clips. All 15 were nonsterilized: Nine had been exten- sively handled prior to testing, and six were received from the distributor un- used and not previously handled (Paul J, oral communication, 1995). The plate glass with the clips was manually inserted 30 cm into the bore of an unshielded 1.5-T MR imaging magnet (Signa; GE Medical Systems, Milwaukee, Wis) and kept constantly parallel to the floor. The clips were observed for transla- tional motion (sliding across the glass surface) and rotational motion (swivel- ing to orient along the magnetic lines of force of the static field of the MR im- ager). The plate glass was then turned a full 180#{176} around the vertical axis, both clockwise and counterclockwise, and observed for motion of any of the clips (the “at-rest phase” of the test). The plate glass was then vibrated manually to and fro along the lateral (left-to- right), horizontal (front-to-back), and vertical (up-and-down) axes. Such mo- tion induced the clips to slide a short distance over the glass. Further testing was performed by tapping or drum- ming the fingers on the surface of the glass (“in-motion phase”), which served to decrease the frictional coeffi- cient of the clips against the glass. Dun- ing the in-motion shaking phase, the clips were observed for only rotational and not translational motion. During the in-motion finger-drumming phase, the clips were observed for rotational and/or translational motion. The re- sults were then recorded according to clip make, model, and type. Testing was performed by two mdi- viduals simultaneously, and grossly detectable motion in the magnetic field was decided by consensus. All clips that tested positive for any type of mo- tion or alignment during either phase were re-tested by a third observer who independently confirmed the same re- sults, with the clips all placed initially perpendicular to the orientation of the static magnetic field.