Magnetic resonance imaging and aneurysm clips A review

The problem of implanted metals causing tissue damage by movement in patients exposed to MRI fields has produced a confusing welter of erroneous, pseudoscientific publications about magnetics, metals, medical equipment, and tissue compatibility. Quite simply, among the devices made for implantation, only those fabricated of stainless steel have the ferromagnetic properties capable of causing such accidents. The author, who introduced the basic design of the modern aneurysm clip in the late 1960s and then a cobalt nickel alloy as an improvement over steel, while chairing the neurosurgical committee assigned to the task of establishing neurosurgical standards at American Society for Testing and Materials, exposes this flawed information and offers clear guidelines for avoiding trouble.

Amid continued furor over lurking danger, it is indeed astonishing to find, in the annals of radiology and neurosurgery, only one reported fatality, this being also the only adverse incident caused by an implanted aneurysm clip moving in the magnetic field of an MRI scanner. One preventable death is, of course, one death too many, and no reasonable safety measure should be spared, nor any effort to eliminate the confusion, where confusion has become rife in a surfeit of flawed information. To this end, the tragedy of this fatality—and the noise behind its singularity—call for another look at the basics of clip design and the involved metallurgy as both relate to MRI and to tissue compatibility in the developing field of microvascular neurosurgery.

Clip Evolvement: Materials

The evolution of materials took a more circuitous route than did the improvements in structure. Early in the era, all spring clips were made of stainless steel. No other available metal at the time afforded both the necessary spring action and acceptable tissue compatibility. In 1956 Mayfield made the first move away from martensitic stainless steel, but some manufacturers continued to use it for a few years because of its stronger spring action. Corrosion problems and clip failure due to faulty design contributed to its abandonment. But many ferromagnetic clips were implanted during the 1960s and 1970s.

The silver solder imbroglio uncovered a disturbing paucity of metallurgy information in neurosurgical annals. The resulting publication26 stimulated the AANS board of directors to send a representative in 1969 to the ASTM to establish standards in a burgeoning neurosurgical technology. Under the aegis of this organization now known as ASTM International, major industry (for example, General Electric, Kodak, General Motors, and medical device companies) work to establish necessary consensus standards. On the scene at the time, other surgical specialties and dentistry were already at work.3 Here, inside open doorways, industry and academia commingle freely to discuss, in roundtable fashion, existing problems. In one of these committee sessions dealing with clip properties, a casual observer from MIT suggested the following: replace steel with MP35N (molybdenum, cobalt, chromium, and nickel), an alloy developed by Dupont for vat storage of acids, and incidentally found to be superior to stainless steels in both spring action and tissue compatibility. The committee approved the idea, thus introducing the cobalt alloys to neurosurgery. Using this metal, MP35N, to make the McFadden Vari-Angle Clip, George Kees took the first step away from using stainless steel in and around the human nervous system. Essentially MP35N is nonferromagnetic, its use by chance predating MRI by some 12 years. Titanium came under scrutiny but was considered at the time not to be the best metal for strong and reliable spring action; L. Steiner introduced it later to diminish clip artifacts in CT and MRI.43 Several companies now make two lines of clips, those using titanium or its alloy and those using Phynox and the similar or identical Elgiloy, a wire alloy devised by the Elgin watch company for watch springs; the metal resembles MP35N with a 17% iron content.


The silver solder misadventure led to other progress and improvements. Activities of the AANS at ASTM attracted attention. The CNS soon joined the efforts. Drugs were added to the tasks, a name change was enacted, and the Committee on Drugs and Devices has since dealt with clips, shunts, other devices, and various drugs, representing the AANS and CNS at ASTM and ISO (International Standards Organization).3 Committee members testified before Congress when laws were being written to regulate devices and drugs, and they later served on the FDA advisory panel for neurology and neurosurgery as regulations were being formulated. In both places they buffeted the too restrictive and too punitive measures under consideration. Among other accomplishments, the committee, writing its consensus standards for aneurysm clips, improved materials and established closing pressures and other physical standards; in the realm of malleable clips, the committee removed silver (actually sterling silver [7.5% copper]) from the list of approved metals because of its extreme corrosive and inflammatory action in body tissue.30 The committee is still active, now under its fifth chairman. Since its inception 42 years ago, the committee has been dedicated to improvement of the metallurgy and the mechanics of aneurysm clips.

Stainless Steel

Steel by definition is predominantly iron (50% or more [personal communication, JR Gladden, 2011]).25,33 The alpha iron in ferromagnetic stainless steel changes to the gamma (nonferromagnetic) phase when heated to the required temperature and, if then alloyed effectively, retains most of the gamma phase at environmental temperatures. These, the austenites, are regarded in neurosurgery as less likely than other steels to move upon exposure to a magnetic field. Thomas W. Eager (personal communication, 2011), Professor of Materials Engineering at MIT, states in his critique of the present article, “…the austenitic steels in their annealed state are non-ferromagnetic, but with mechanical working (such as forming into the strength and shape of a clip) will transform to the ferromagnetic martensite phase. This could have disastrous results. These types of steels are called TRIP steels (Transformation Induced Plasticity) …Types 301 and 302 are some of the most susceptible steels to TRIP.”

The steels used in the Mayfield clip and its descendants belong to this category (Types 301 and 302). These steels, sought after in industry for their unique physical characteristics, form into a satisfactory clip mechanically, but may prove to be treacherous in the MRI scanners in clinical medicine. Higher in the austenitic range (316, 316L), ferromagnetic activity is less likely, but the metal has less spring strength. Despite the ferromagnetic propensities of the austenitic stainless steel used to make the early clips, no adverse incident has been reported from exposure to MRI. In the 1960s and early 1970s, the Mayfield, the McFadden-Mayfield, and the Yaşargil (version of the Mayfield) clips were made of austenitic stainless steel, Types 301–304. The very first models of the helical coiled-spring crossed-action clip (McFadden Vari-Angle) in the late 1960s were fashioned from austenitic stainless steel and then, very soon, changed to the alloy MP35N. The Yaşargil was changed to Phynox.

No one should be making intracranial aneurysm clips of stainless steel; vastly superior metals are now available. If a surgeon has implanted a McFadden Vari-Angle, Yaşargil, Sundt Slim-Line, Spetzler, or Sugita clip made since 1980, his patient should not harbor a stainless-steel clip. Of the 40 or more vascular clip types in existence,11 only the self-closing crossed-action coiled-spring clips are addressed in this statement.


Diamagnetic and paramagnetic materials become magnetic only when exposed to an external magnetic field (personal communication, CL Chien, 2011),2,5,16,44 the first weakly repelled and the other weakly attracted. Both revert when removed from the exposure. The three commonplace ferromagnetic elements—cobalt, nickel, and iron—are strongly attracted to a magnetic field and retain magnetic properties when removed (personal communication, JR Gladden, 2011).26 Quoting Professor Eager again (personal communication, 2011), “…diamagnetism and paramagnetism are properties inherent in the electronic structure of the atom, ferromagnetism is not. Ferromagnetism is created by cooperative action of multitudes of atoms forming crystal domains on a vastly larger scale than the atomic.”

MP35N contains cobalt and nickel. Phynox or Elgiloy contains the same two metals plus 17% iron. Thus alloyed, these metals do not display ferromagnetic behavior when exposed to the scanners currently in use. Professor Eager writes, this from the critique25 of an article by Dejovny et al:6 “…clips…made from MP35N, Elgiloy, Phynox, titanium and its alloy are permanently immune to ferromagnetism, and the cobalt-based alloys (MP35N, Elgiloy, and Phynox) have twice the elastic modulus of titanium and hence greater spring or clamping for equivalent designs.”

Clips tested in an 8.0-T MRI system have shown what Kangarlu and Shellock17 consider to be acceptable reactions to MP35N and to titanium and titanium alloy. Movement of austenitic stainless steel and of Phynox raised concern. Elgiloy, essentially the same alloy as Phynox, raised no concerns in the test, but this result does raise questions about the testing methods and interpretation of the results. Biological effects in the ascending Tesla scale may limit the range of field strengths safe for clinical use. Small laboratory animals and frogs levitate at a 15-T exposure16 through the repulsion of diamagnetic water content of living tissue.


Failed implants in a patient’s skull cavity do not produce a characteristic clinical response as they do in dental (violent tooth ache) and orthopedic patients (fractured prosthesis and collapsed weight-bearing parts, purulent rejection of foreign material through soft-tissue surroundings, and so on). For this reason, early on, neurosurgery was slower to recognize the salient principles of metals and other materials. Beginning in 1969, the AANS addressed the problem and rapidly improved the practices, as described in the standards section above.3,26 Industry was striving for a more powerful spring when making clips out of martensitic stainless steel before the MRI era.

A clip formed from a single piece of corrosion-resistant metal will have less electrochemical activity in tissue than a device formed of several parts. The actuating spring of alligator clips is an appurtenance, and in stainless steel, it may be at a distance in the galvanic series from the other parts, thus contributing to corrosion, and failure in certain models. The alligator clip with its many limitations has been the major offender in clip failures. More acceptable materials are now being used to manufacture clips of the type.

Juxtaposed clips of widely different composition such as stainless steel and sterling silver will react rapidly.26 Lesser differences in kindred alloys do not always cause trouble. For instance, Hamby’s patient13 in whom both Mayfield and Schwartz clips were implanted in 1958 to treat multiple intracranial aneurysms had shown no evidence of complications as of the late 1960s (personal communication at the time). These two clips together in a saline environment would undergo galvanic action.21,26

Spring clip technology, now approaching its 60th year, like other disciplines and other technologies, has progressively evolved and produced improved devices. Beginning with the Schwartz and the Mayfield clips, neurosurgeons used what the industry could produce, and most of the devices were based on ideas and feedback from surgeons. The significance of refinements in corrosion prevention was overshadowed by surgical contingencies early in the era and by relative brief periods of observation. The AANS, later joined by CNS, began in 1969 its contributions to the science and the art. Most of the mechanical and materials deficiencies contributing to clip failure have been eliminated. No doubt better materials and better designs will appear in the future. The muddle in radiology over MRI should not be blamed on the neurosurgical technology in use many years before a patient appears on site to undergo MRI. It’s the wrong place for a witch hunt, and the right place for rationality.

Specious Work

Despite abundant correct information in the annals, the most misleading sophistry has crept into published articles pertaining to clips. In metallurgy, for instance, various authors have referred to Phynox (Elgiloy) (17% iron) and MP35N (no iron) as stainless steel.6,7,9,32,34 Neither material can accurately be labeled stainless steel and to do so confuses the issue. Metallurgical guidelines cannot make sense if the nomenclature is wrong, specifically because stainless steel is the only offender. As a result, confusion and frustration have inhibited the proper handling of patients needing MRI.

Erroneous statements about stainless steel add to the confusion. For instance, Romner et al.34 have written, “Clips…of martensitic stainless steel such as the Mayfield and…” This statement is wrong and serves to confuse any effort to deal with an implanted clip of the Mayfield configuration, a very damaging error, indeed. Mayfield’s clip was always austenitic stainless steel. He pioneered its use in the evolving clip technology.20

Careless language has added to the identification problems. The term “Vari-Angle,” for instance, causes confusion when used without the proper eponym. Thus, Codman has sold several different Vari-Angle products, one being the Sundt-Kees Vari-Angle Clip (the culprit in the fatal case, see The Patient]). Use of the term Vari alone could mean anything, from this involved outdated martensitic stainless-steel clip to the latest product with proper engineering and up-to-date metallurgy. Careless reportage of history has obscured the facts of clip evolution14,19 and created a false impression of neurosurgical irresponsibility. The article written by Louw et al.,19 in particular, is an example of such reportage, reverting to deceptive sensationalism and erroneousness when the authors write, “…and a patient in whom an incompatible clip had been placed died during an MR imaging session. In response the American Society for Testing and Materials Committee developed…” In fact, this unfounded and erroneous claim appeared in print some 30 years after the committee of neurosurgeons and neurosurgical device manufacturers representing the AANS had begun work at ASTM on the mechanics and materials of intracranial aneurysm clips. The committee never approved the use of martensitic stainless steel for implantation, and none of their work was done in response to disasters. Furthermore, the clip in the fatal case was not incompatible. It had been there roughly 14 years, obliterating the aneurysm and doing no harm.

Shellock and Crues37 condemned certain clips with incomplete and inaccurate identification by saying, “… these aneurysm clips (Drake, Mayfield, McFadden, Sundt-Kees) had sufficient ferromagnetism to warrant the exclusion of patients with these implants…,” a statement made in 1988, some 18–20 years after neurosurgery had established materials standards for aneurysm clips. A review of the known characteristics of these clips, one at a time, reveals the faults of the article in question: 1) The early Drake clip, the fenestrated model of the Mayfield, or the McFadden-Mayfield clip, was made of austenitic stainless steel. Beginning in the 1970s, the design was incorporated in all four lines of the helical coiled spring clip, the McFadden Vari-Angle, the Yaşargil, the Sugita, and finally the Spetzler clip; not one by then was made of stainless steel. 2) The Mayfield clip was made of austenitic stainless steel 301–304, and the McFadden-Mayfield clip was composed of austenitic stainless steel. 3) The Mc-Fadden clip, or the McFadden Vari-Angle clip, after the initial austenitic models of 1969, was made of MP35N. 4) The Sundt-Kees Vari-Angle clip was made of martensitic stainless steel, a frankly ferromagnetic stainless steel, in the early models; the later Sundt-Kees Slim-Line clip42 was and still is made of MP35N. So which McFadden, which Sundt, and which Drake model did the authors intend to implicate? This careless use of language, with its eponymous slights, only creates confusion and gives no one any useful information. In another article, Shellock and Crues38 stated, “…the few fatalities that have occurred were the result of failure to follow guidelines…” Nine pages later in the same article,38 the authors wrote, “To our knowledge, only one ferromagnetic aneurysm clip-related fatality has been reported in the peer-reviewed literature.” They offered no reference and no proof for the first claim, and without the evidence this amounts to irresponsible and gratuitous sensationalism.

In an article by Dujovny and colleagues,6 basic metallurgical errors found in tables cast doubt on the accuracy of the text. Consequently, opinions were sought from reliable experts in materials science: Thomas W. Eager, Professor of Materials Engineering at the MIT, and James C. Ho, Trustees Distinguished Professor of Physics and Chemistry at Wichita State University. Both authorities confirmed errors in the tables and the text, and the details were published.25 The same scrutiny should be turned to other articles.

ClipFinder (, with authoritative posturing, sets a flippant tone by classifying aneurysm clips as good guys (nonferromagnetic) and bad guys (ferromagnetic); the article then goes on to introduce mistakes: “It may be an obsolete dangerous ferromagnetic Drake clip…” As stated above, the Drake clip was made of austenitic stainless steel by Kees and Codman but never of a frankly ferromagnetic material such as martensitic stainless steel. A few paragraphs later, addressing materials, the authors remarked, “For the uncommon deformable clips… inert types of steel like tantal [sic] and titanium are used.” Tantal, as they use the word, presumably means tantalum. Neither metal is in any way related to stainless steel. Still further in the document, “Cobaltalloys like the MP 35 N…contain very low steel and much nickel…” Here, again, is an example of the confusing misrepresentation and misunderstanding of steel. In no way is MP35N even remotely related to steel, stainless or otherwise.

The monograph The Sugita Clip, Innovations in Neurosurgery, written by a physicist and translated by John Junkerman, strives to credit Sugita with inventing and developing the helical coiled-spring crossed-action clip, now the dominant type.12 The book is fraught with errors, whether intentional or otherwise, and largely ignores the previous existence of this clip, its composition, and its standards, all developed in the Western world before Sugita appeared on the scene. The absence of a bibliography, references, and precise dates of claimed events points to deception. The erroneous representations are too numerous to be addressed here. In accurate historical significance the book is far worse than inconsequential.


  1. Black SP, German WJ: A clamp for temporarily occluding small blood vessels. J Neurosurg 11:514515, 1954 Abstract, Medline
  2. Burke HE: Handbook of Magnetic Phenomena New York, Van Nostrand Reinhold, 1986. 423 CrossRef
  3. Burton CV, McFadden JT: Neurosurgical materials and devices. Report on regulatory agencies and advisory groups. J Neurosurg 45:251258, 1976 Abstract, Medline
  4. Carvi y Nievas MN, Höllerhage HG: Risk of intraoperative aneurysm clip slippage: a new experience with titanium clips. J Neurosurg 92:478480, 2000 Abstract, Medline
  5. Della Torre E: Magnetic Hysteresis Piscataway, NJ, IEEE Press, 1999. 203
  6. Dujovny M, Alp MS, Dujovny N, Zhao YJ, Gundamraj NR, Misra M, et al.: Aneurysm clips: magnetic quantification and magnetic resonance imaging safety. Technical note. J Neurosurg 87:788794, 1997 Abstract, Medline
  7. Dujovny M, Kossovsky N, Kossowsky R, Perlin A, Segal R, Diaz FG, et al.: Intracranial clips: an examination of the devices used for aneurysm surgery. Neurosurgery 14:257267, 1984 CrossRef, Medline
  8. Dujovny M, Kossovsky N, Laha RK, Leff LL, Wackenhut N, Perlin A: Temporary microvascular clips. Neurosurgery 5:456463, 1979 CrossRef, Medline
  9. Dujovny M, Kossowsky R, Kossovsky N, Diaz FG, Ausman JI: Corrosion of aneurysm clips: evaluation and clinical implications. Part II: Individual performance. Acta Neurochir (Wien) 72:257269, 1984 CrossRef, Medline
  10. Food and Drug Administration: FDA Safety Alert: MRI related death of patient with aneurysm clip ( [Accessed January 24, 2012]
  11. Fox JL: Vascular clips for the microsurgical treatment of stroke. Stroke 7:489500, 1976 CrossRef, Medline
  12. Fukuyama T: The Sugita Clip—Innovations in Neurosurgery Tokyo, Medical Culture Research Institute, 2006
  13. Hamby WB: Multiple intracranial aneurysms: aspects of treatment. J Neurosurg 16:558563, 1959 Abstract, Medline
  14. Heros RC, Morcos JJ: Cerebrovascular surgery: past, present, and future. Neurosurgery 47:10071033, 2000 CrossRef, Medline
  15. Hirashima Y, Kurimoto M, Kubo M, Endo S: Blade crossing of a pure titanium clip applied to a cerebral aneurysm—case report. Neurol Med Chir (Tokyo) 42:123124, 2002 CrossRef
  16. Kaku M: Physics of the Impossible: A Scientific Exploration Into the World of Phasers, Force Fields, Teleportation, and Time Travel New York, Anchor Books, 2008. 316
  17. Kangarlu A, Shellock FG: Aneurysm clips: evaluation of magnetic field interactions with an 8.0 T MR system. J Magn Reson Imaging 12:107111, 2000 CrossRef, Medline
  18. Klucznik RP, Carrier DA, Pyka R, Haid RW: Placement of a ferromagnetic intracerebral aneurysm clip in a magnetic field with a fatal outcome. Radiology 187:855856, 1993 Medline
  19. Louw DF, Asfora WT, Sutherland GR: A brief history of aneurysm clips. Neurosurg Focus 11:2E4, 2001 Abstract
  20. Mayfield FH, Kees G Jr: A brief history of the development of the Mayfield clip. Technical note. J Neurosurg 35:97100, 1971 Abstract, Medline
  21. McFadden JT: Aneurysm clips. J Neurosurg 46:129, 1977. (Letter) Medline
  22. McFadden JT: Aneurysm clips. Neurosurgery 14:521, 1984. (Letter)
  23. McFadden JT: Cerebrovascular surgery: past, present, and future. Neurosurgery 49:231233, 2001. (Letter) Medline
  24. McFadden JT: Evolution of the crossed-action intracranial aneurysm clip. Technical note. J Neurosurg 71:293296, 1989 Abstract, Medline
  25. McFadden JT: Magnetic quantification. J Neurosurg 91:716719, 1999. (Letter) Medline
  26. McFadden JT: Metallurgical principles in neurosurgery. J Neurosurg 31:373385, 1969 Abstract, Medline
  27. McFadden JT: Modifications of crossed-action intracranial clips. Technical note. J Neurosurg 32:116118, 1970 Abstract, Medline
  28. McFadden JT: New aneurysm clip and applier for narrow spaces: technical note. Neurosurgery 46:15331534, 2000 CrossRef
  29. McFadden JT: The origin and evolutionary principals of spring forceps. Surg Gynecol Obstet 130:356368, 1970 Medline
  30. McFadden JT: Tissue reactions to standard neurosurgical metallic implants. J Neurosurg 36:598603, 1972 Abstract, Medline
  31. Miller D: Being an absolute skeptic. Science 284:16251626, 1999 CrossRef
  32. New PF, Rosen BR, Brady TJ, Buonanno FS, Kistler JP, Burt CT, et al.: Potential hazards and artifacts of ferromagnetic and nonferromagnetic surgical and dental materials and devices in nuclear magnetic resonance imaging. Radiology 147:139148, 1983 Medline
  33. Parr JG, Hanson A: An Introduction to Stainless Steel Metals Park, OH, American Society for Metals, 1965. 147
  34. Romner B, Olsson M, Ljunggren B, Holtås S, Säveland H, Brandt L, et al.: Magnetic resonance imaging and aneurysm clips. Magnetic properties and image artifacts. J Neurosurg 70:426431, 1989 Abstract, Medline
  35. Schöller K, Morhard D, Zausinger S, Steiger HJ, Schmid-Elsaesser R: Introducing a freely accessible internet database for identification of cerebral aneurysm clips to determine magnetic resonance imaging compatability. Neurosurgery 56:118123, 2005
  36. Scoville WB: Miniature torsion bar spring aneurysm clip. J Neurosurg 25:97, 1966
  37. Shellock FG, Crues JV: High-field-strength MR imaging and metallic biomedical implants: an ex vivo evaluation of deflection forces. AJR Am J Roentgenol 151:389392, 1988
  38. Shellock FG, Crues JV: MR procedures: biologic effects, safety, and patient care. Radiology 232:635652, 2004 CrossRef
  39. Sokal A, Brickmont J: Fashionable Nonsense: Postmodern Intellectuals’ Abuse of Science New York, Picador, 1998. 294
  40. Sokal AD: Transgressing the boundaries: toward a transformative hermeneutics of quantum gravity. Social Text 46/47:217252, 1996 CrossRef
  41. Sugita K, Hirota T, Iguchi I, Mizutani T: Comparative study of the pressure of various aneurysm clips. J Neurosurg 44:723727, 1976 Abstract, Medline
  42. Sundt TM Jr: Sundt-Kees slim-line high tension aneurysm clips. Neuro News IX (April) 1986
  43. von Holst H, Bergström M, Möller A, Steiner L, Ribbe T: Titanium clips in neurosurgery for elimination of artefacts in computer tomography (ct) a technical note. Acta Neurochir (Wien) 38:101109, 1977 CrossRef, Medline
  44. Vonsovskiĭ SV: Magnetism 1:Toronto, John Wiley & Sons, 1971. 461
  45. Yaşargil MG, Vise WM, Bader DC: Technical adjuncts in neurosurgery. Surg Neurol 8:331336, 1977 Medline