Cover Story

MRI-Conditional Cardiac Devices: Continued Controversy Over the Future Not Being What It Used to Be

J. Rod Gimbel, MD
University Hospitals Case Medical Center
Cleveland, Ohio

J. Rod Gimbel, MD
University Hospitals Case Medical Center
Cleveland, Ohio

With the recent passing of Lawrence Peter “Yogi” Berra, we were reminded of his many important everyday insights, such as “When you come to the fork in the road, take it,” as well as the importance of avoiding “too many wrong mistakes.”1 Every day, device implanters must make very specific choices regarding the devices they implant, and a fork in the road is faced. One such choice is whether to implant an MRI-conditional device. 

“The towels were so thick there I could hardly close my suitcase.” 

Likening the MagnaSafe Registry to the Magna Carta in the January 2015 issue of EP Lab Digest,2 Dr. Bradley Knight began his monthly editorial noting, “Despite evidence from Johns Hopkins and other centers that an MRI can be safely performed in most patients with cardiac implantable electrical devices (CIEDs) when certain protocols are followed, many hospitals still do not allow patients with a device that is not labeled as MRI-compatible to undergo an MRI, and Medicare reportedly will not reimburse for it.”2 Dr. Knight went on to succinctly summarize the encouraging results of off-label MRI of device patients during the recent presentation of the MagnaSafe Registry at the November 2014 AHA meeting,3 writing that the MagnaSafe “results suggest that MRI scanning in these patients is very safe.”2 

Perhaps feeling any controversy on the matter is now closed, Eyal and Roquin recently wrote “The MagnaSafe analysis included 1500 patients and proved that MRI, at 1.5 Tesla, can be performed for patients with standard devices at no significant clinical risk, while adhering to several important safety rules.”4 In this particularly enthusiastic endorsement of the findings, it might be noted that the authors failed to mention the MagnaSafe cohort specifically excluded pacemaker-dependent ICD patients and patients undergoing thoracic MRIs — both high risk and important populations — surely details worth considering. 

Clearly, the accumulating evidence that safe off-label MRI of device patients can be performed has not gone unnoticed. Recent society position statements document the evolution of off-label MRI of device patients from being an “absolute contraindication” to a “relative contraindication” using a more patient-individuated, risk-based approach.5 

Given the encouraging evolutionary view regarding off-label MRI of devices, why might clinicians choose to consistently implant an MRI-conditional system? Simply put, choosing an MRI-conditional system leads to important incremental safety advantages, diminished burdensome supervisory commitments, reduced administrative headaches, a straightforward payment pathway, and most importantly, improved patient access to MRI. A view of device choice guided not by particular vendor, but rather by the presence of MRI-conditional labeling, is explored here.

“Why buy good luggage? You only use it when you travel.” 

It has been estimated that up to 75% of CIED recipients require an MRI over the lifetime of their device,6 suggesting that device patients travel frequently to the MRI after they receive their implant. Who can say that an 81-year-old who receives a pacemaker today might not fall a year later, needing an MRI to appropriately sort out the orthopedic and neurologic issues that may follow?

The strong static magnetic and powerful dynamic gradient magnetic and radiofrequency fields present in the MRI suite pose a significant challenge to the uninterrupted safe operation of luggage (CIEDs) that was never specifically designed to travel to or within the MRI suite. Even exposure to simply a (comparatively weak) magnetic field may alter the behavior of a pacemaker, causing no pacing output and leading to a potentially life-threatening situation.7

The implication8-10 that gradual improvements in the materials used to construct CIEDs are largely responsible for the improved safety of off-label scanning of “modern” devices are speculative and unsubstantiated. Importantly, no such official statement from any manufacturer has ever been issued substantiating such a claim. Indeed, as recently noted by Mayo Clinic investigators, “abnormalities in pacemaker function during MRI continue to be reported, some of which are potentially concerning, even with the current generation of available devices.”11 

Perhaps forgotten in all the kerfuffle, survey data from multiple sources suggest even inadvertent MRI of devices (no screening, monitoring, or programming strategy) occurs safely far more frequently than medical misadventure,12,13,26 raising the possibility that a substantial element of sheer luck (not safety by design) is largely responsible for the “safe outcome” of off-label MRI. Even luggage not designed for the task might survive the rigors of adventurous travel. 

“If you don’t know where you are going, you might wind up someplace else.” 

Excellent, thorough, and balanced summaries of the many potential complications of off-label MRI have been presented.5,6 Despite the fact that it has been suggested that off-label scanning only be performed in expert centers, the road map outlining the directions for off-label scanning is not particularly complex. Redundant monitoring (EKG and pulse oximeter), simple pre-MRI programming, and MRI power deposition reduction strategies (SAR below 2 W/kg)14 appear to mitigate some (but not all) of the ill effects that might occur during off-label scanning. However, such strategies do not solve the problem entirely or sufficiently as power on reset (PoR), for instance, remains an important issue.

Indeed, tasked as editorialists for the Johns Hopkins’ manuscript presenting their off-label MRI experience, Reynolds and Zimmetbaum tempered their largely and appropriately positive view of the findings, noting during off-label scanning that “another potential device behavior, called ‘power on reset,’ can override even the most effective device reprogramming. In most CRMDs, power on reset, which is prompted by extreme electromagnetic fields, is accompanied by a reversion of the programmed settings; this may include switching back to a synchronous pacing mode, and resumption of tachycardia detection and therapies. Thus, a power-on-reset event can reverse the programming maneuvers available to avoid the important risks for asystole and inappropriate shocks, should oversensing occur.”15

In the MagnaSafe Registry, six PoR events were recorded and, while not related to a PoR event, one ICD undergoing MRI required replacement post MRI after it was found unresponsive.3 This was blamed on “pilot error” and was not a “defect of the airplane” itself, as the clinician involved failed to program the ICD “off” prior to MRI. The suggestion, of course, is that had the clinician properly programmed the device “off” prior to MRI, this event would have never occurred.2 However, clinicians should be aware that despite appropriate pre-MRI programming, any non MRI-conditional ICD undergoing MRI may experience a PoR event, which enables the ICD’s sensing and charging capacity during an MRI. Such PoR events are unpredictable and entirely beyond the control of the clinician, and may occur without the staff or patient being aware that PoR has taken place. 

After a PoR event, and with the device now in a sensing mode, the ICD lead acts as an antennae soaking up the EMI present during the scan, feeding it to the pulse generator, which may interpret the EMI falsely as ventricular tachycardia or ventricular fibrillation.16,17 The ICD may then experience a series of charge and dump sequences, leaving the device dysfunctional at the end of the MRI. In short, the unpredictable nature of PoR undoes the thoughtful pre-MRI programming that may have been performed, leaving the patient with the same programming strategy that takes place when an inadvertent scan occurs — that is to say, effectively no programming strategy at all. Because of the many special design features of MRI-conditional ICDs, and when programmed appropriately to “MRI mode,” such an event is avoidable if one uses an MRI-conditional ICD.

While only a small fraction of scans resulted in PoR events in the MagnaSafe Registry, other investigators using similar protocols have found a 13-16% incidence of PoR during off-label MRI.11,18 That PoR can result in a potentially life-threatening situation during off-label MRI is aptly illustrated by previous reports.13,19,26

Such reports, in which the clinician was intending to go one place but wound up someplace else during off-label scanning because of PoR, are not likely to go unnoticed by regulatory authorities that are tasked with strict rules of evidence when labeling a device “MRI conditional.”

Again, it should be noted that in contrast, because of the many system modifications explicitly designed to permit safe MRI, systems labeled “MRI conditional” provide a safe, predictable road to travel for both the patient and physician wishing to have an uneventful MRI and not end up someplace else, as evidenced by the absence of reported adverse events during MRI when using MRI-conditional devices. Importantly, such predictably safe scans take place irrespective of the patient’s pacemaker dependence or region of the body undergoing MRI.

“We have deep depth.” 

The U.S. Food and Drug Administration approved the first MRI-conditional pacemaker system in February 2011. Criticisms emerged of the pulse generator initially available for use (somewhat limited functionality), and in particular, the lead (less satisfying handling, stiffer and perhaps prone to perforation).20 Subsequent developments rapidly led to the release and/or relabeling of more leads as MRI conditional, giving the practitioner leads with proven, reliable track records. Implanters now have available both familiar full-featured single- and dual-chamber pacemakers and ICDs as well as the flexibility of scanning any part of the body. An expanded range of choices is expected to follow as more manufacturers release MRI-conditional devices.

Perhaps “the problem” of off-label scanning will eventually resolve itself as device companies in the near future will no longer make non-MRI-conditional devices. It may be, however, that market forces driven by relentless pricing pressures, customer segmentation, and tiered pricing for various premium features (MR conditional) will lead to a marketplace where non-MRI-conditional devices remain available as an option for years to come. Thus, the “problem” of which device to choose may not resolve itself so easily. 

Investigators from the Mayo Clinic note that “even if MRI-conditional pacing systems achieve widespread availability, patients with existing pacing systems will continue to need MRI for some time.”11 As such, both the Johns Hopkins data21 and the results of the MagnaSafe Registry3 are quite valuable, as they provide a measure of reassurance that off-label scanning can be done. However, perpetuating the problem associated with off-label MRI by continuing to implant non-MRI-conditional devices, thereby creating a burgeoning future population of compromised off-label CIED patients, may not be a wise choice, particularly when safer MRI-conditional devices exist. Mishaps during off-label scanning may not be an attractive option when the tools exist to avoid them.22 The fact that some or many of the adverse effects reported during off-label MRI scanning are relatively infrequent or, for example, have been reported after inadvertent scanning,13,26 may do little to diminish the concern of the everyday practitioner who simply wants to regularly scan device patients in a hassle-free manner. A simple solution: choose an MRI-conditional device from the now deep depth of offerings.

“I usually take a two-hour nap from one to four.”

While it is no doubt true the vast majority of off-label scans proceed uneventfully, it is also true few things in life are more soporific than watching an uneventful MRI. You meet the patient outside the MRI suite, choose the programming strategy, take time to assure the patient is appropriately monitored, and then watch as 45 minutes or more of unrecoverable (and unrecompensed) time quietly drifts away while the scan is being performed. You then recheck the device post MRI and dictate a note, all the while thinking about your other clinical responsibilities that await your attention. In the sprawl of many modern medical centers, the MRI suite is nowhere near the cardiovascular base of action, requiring a long walk to and from the MRI suite, a pedestrian effort perhaps best rationalized as part of a 10,000 steps per day program. Who has time for this, particularly when 50-75% of device patients may need an MRI? If not you, then someone from your valuable team needs to supervise the off-label event.

Regarding the risks of off-label scanning, Sommer et al note “risks cannot be absolutely eliminated and are not quantifiable in individual cases.”23 Because of these difficult to predict and unquantifiable risks, off-label scanning requires supervision from a member of the cardiology or electrophysiology team, a supervisory burden not imposed during MRI of patients who have MRI-conditional devices. The expensive supervisory burden, and the time sink that is “off-label MRI,” can largely be mitigated by choosing an MRI-conditional device as the risks of MRI become predictably and consistently low.

“A nickel ain’t worth a dime anymore.”

One argument for not using an MRI-conditional system is the added expense. Other than the increased upfront costs of these devices, very little has been written about the comparative differential economics of choosing one type of device over another in total. Carefully crafted proprietary device contracts chock-full of complex volume discounts and tailored rebates vary from hospital system to hospital system, making it difficult to gauge just what the “uplift” (price increase) is for a premium feature like MRI conditional. However, at an uplift of $500 compared to a non-MRI-conditional device, a back-of-the-envelope calculation suggests that, amortized over the life of an eight-year battery, the cost of gaining unfettered, safe MRI access for your patient is 17 cents per day. 

No one should begrudge hospital administrators for trying to manage operational costs — that is part of their job. No one wins if a hospital closes, because its costs exceed its revenues. However, a more nuanced global view of the “total cost of ownership” should be embraced when clinicians and administrators choose which devices to have on contract. No doubt it is difficult to estimate the previously mentioned time value costs of the supervisory burden imposed by choosing a non-MRI-conditional device, but this should be considered.24 Time is valuable.

Delays in diagnosis can occur to a patient who needs an MRI but cannot have one because their device is not MRI conditional. Notably, the care pathways suggested for off-label MRI ask the clinician to consider if an alternative imaging modality is feasible.5,6,11,21,23 Sometimes this decision is easily made and an equivalent imaging modality is possible, but sometimes not.23 Often, one clinician must contact another and bring that clinician up to speed on the patient’s care. This often takes time, and delays in care can take place, with statements such as “I need to pull the patient’s chart and see the patient…see if a CT will work…and tell them I’ll call them back right after my PVI.” This delay can extend an already expensive hospital stay, adding to the cost of care and causing a delay in diagnosis. Perhaps a valid question is, given the options to avoid the problem, why should a cardiologist’s choice of pacemaker or ICD today constrain a neurologist’s, oncologist’s, or orthopedist’s imaging choices tomorrow? 

Clinicians and administrators who decide which devices should be on contract should also be aware of economic offsets to the upfront cost of MRI-conditional devices. The increased costs of the MRI-conditional product are often offset substantially, because now device patients with MRI-conditional systems can obtain clinically important (and reimbursed) MRIs when they could not before. Many patients receive multiple MRIs.

Maybe your hospital does perform off-label MRI. Let us not forget about the patient, who must often pay out-of-pocket for the MRI if they have a non-MRI-conditional device, as insurers usually will not reimburse off-label MRIs. At between $500-2500, an off-label MRI can be an unwelcome budget-busting burden for an elderly, sick patient on a fixed and penurious income. MRIs performed on patients who have MRI-conditional devices are readily approved for reimbursement without additional paperwork petitioning for payment.

Puzzling, perhaps, is why clinicians and administrators who make economic arguments against the use of MRI-conditional devices are the same ones who, when asked what device they would like to have if faced with the need for a device, insist that they or a family member would choose an MRI-conditional device.

“All pitchers are liars or crybabies” and “If the world were perfect, it wouldn’t be.”

Hats off to Dr. Melissa Walton-Shirley for taking the time to opine on the problem of obtaining off-label MRIs.25 Her own frustration is palpable, and this particular piece (and the associated comments) might be read as an impassioned broad overview on the many externalities beyond the dry technical issues that inform the MRI-conditional debate in the traditional medical literature.

The results of the MagnaSafe Registry,3 the Johns Hopkins’ data,21 and other work clearly indicate we may reasonably and safely expand the availability of off-label scanning. Clinicians and institutions wishing to take on the risk and service burdens of off-label scanning should be reimbursed accordingly, as they are providing a valuable service. Feel free to petition CMS and insurers to change their reimbursement policy. 

Unfortunately, the world is far from perfect, as Dr. Walton-Shirley has discovered.25 While infrequent, the naggingly persistent and sometimes serious complications that can inevitably occur during off-label MRI of devices cannot be simply swept away by magical thinking, hyperbole, or polemic. Further, we might temper our criticisms of those who choose not to bear the potentially serious risks, compliance headaches, and administrative burdens that accrue with off-label MRI. 

As noted by Walton-Shirley, “all over the world, there are patients with headaches, paresthesias, and presumptive and unconfirmed diagnoses of MS, brain tumors, stroke, and multi-infarct dementia who still wait for an MRI because they have a device.”25 Carefully explaining to patients who have non-MRI-conditional devices the virtues of the MagnaSafe Registry and the Johns Hopkins’ data does little to provide them access when no nearby hospital will allow them in their MRI scanner. The simple solution going forward is to implant MRI-conditional devices and leads wherever possible. Get the device out of the way of the imaging study that your patient needs.

People, as they say, are entitled to their own opinions; they are not entitled to their own facts. No one can rightly claim that devices never designed for MRI can regularly undergo MRI safely under the wide variety of conditions expected during MRI without appropriate, thorough testing that validates the labeling of “MR conditional”. The MagnaSafe Registry results and the Johns Hopkins’ data3,21 are important, and the authors should be congratulated for their excellent work. However, such findings did not prove that those devices are “safe” to scan or even deserving of the “MR conditional” label. The label itself, “MR conditional,” has strict meaning — a hurdle not overcome by current off-label devices and the data presented to date.
Perpetuating the problem by continuing to implant non-MRI-conditional devices and leads ignores the important incremental safety benefits, diminished supervisory and administrative burdens, and improved MRI access that accrues with choosing an MR-conditional product. Choose wisely, and when you come to the fork in the road, take it.

Disclosure: The author has no conflicts of interest to report regarding the content herein. Outside the submitted work, Dr. Gimbel reports personal fees as a consultant and speaker for Medtronic and BIOTRONIK.  


  1. Scott N. The 50 greatest Yogi Berra quotes. Published September 23, 2015. Available online at Accessed Dec 15, 2015.
  2. Knight BP. MagnaSafe: The Magna Carta for Safely Performing MRIs in Patients with Pacemakers and Defibrillators. EP Lab Digest. 2015;15(1):4. 
  3. Russo RJ and The MagnaSafe Investigators. Determining the Risks of Magnetic Resonance Imaging at 1.5 Tesla for Patients with Non-MRI Conditional Pacemakers and Implantable Cardioverter Defibrillators: Final Results of the MagnaSafe Registry. Late-Breaking Clinical Trials, AHA Nov 2014.
  4. Eyal A, Roguin A. Cardiology Patient Page. Magnetic Resonance Imaging in Patients With Cardiac Implantable Electronic Devices. Circulation. 2015;132(14):e176-178.
  5. Verma A, Ha AC, Dennie C, et al. Canadian Heart Rhythm Society and Canadian Association of Radiologists consensus statement on magnetic resonance imaging with cardiac implantable electronic devices. Can Assoc Radiol J. 2014;65(4):290-300.
  6. Ipek EG, Nazarian S. Safety of implanted cardiac devices in an MRI environment. Curr Cardiol Rep. 2015;17(7):605.
  7. Kolb C, Luik A, Hessling G, Zrenner B. Magnetic catheter navigation system interference with a dual-chamber pacemaker. J Cardiovasc Electrophysiol. 2007;18(8):892-893.
  8. Roguin A, Zviman MM, Meininger GR, et al. Modern pacemaker and implantable cardioverter/defibrillator systems can be magnetic resonance imaging safe: in vitro and in vivo assessment of safety and function at 1.5 T. Circulation. 2004;110:475-482.
  9. Shellock FG, Fieno DS, Thomson LJ, et al. Cardiac pacemaker: in vitro assessment at 1.5 T. Am Heart J. 2006;151:436-443.
  10. Martin ET, Sandler DA. MRI in patients with cardiac devices. Curr Cardiol Rep. 2007;9(1):63-71. 
  11. Boilson BA, Wokhlu A, Acker NG, et al. Safety of magnetic resonance imaging in patients with permanent pacemakers: a collaborative clinical approach. J Interv Card Electrophysiol. 2012;33(1):59-67.
  12. Gimbel JR, Lorig RJ, Wilkoff BL. Safe Magnetic Resonance Imaging of Pacemaker Patients (Abstract). J Am Coll Cardiol. 1995. doi:10.1016/0735-1097(95)91501-N 
  13. Ferris NJ, Kavnoudias H, Thiel C, Stuckey S. The 2005 Australian MRI safety survey. AJR Am J Roentgenol. 2007;188(5):1388-1394.
  14. Gimbel JR. The safety of MRI scanning of pacemakers and ICDs: What are the critical elements of safe scanning? Ask me again at 10,000. Europace. 2010;12(7):915-917.
  15. Reynolds MR, Zimetbaum P. Magnetic resonance imaging and cardiac devices: how safe is safe enough? Ann Intern Med. 2011;155:470-471.
  16. Roguin A, Schwitter J, Vahlhaus C, et al. Magnetic resonance imaging in individuals with cardiovascular implantable electronic devices. Europace. 2008;10(3):336-346.
  17. Anfinsen OG, Berntsen RF, Aass H, et al. Implantable cardioverter defibrillator dysfunction during and after magnetic resonance imaging. Pacing Clin Electrophysiol. 2002;25(9):1400-1402.
  18. Naehle CP, Carsten M, Thomas D, et al. Safety of brain 3-T MR imaging with transmit-receive head coil in patients with cardiac pacemakers: pilot prospective study with 51 examinations. Radiology. 2008;249:991-1001.
  19. Gimbel JR. Unexpected asystole during 3T magnetic resonance imaging of a pacemaker-dependent patient with a “modern” pacemaker. Europace. 2009;11:1241-1242.
  20. Elmouchi DA, Rosema S, Vanoosterhout SM, et al. Cardiac perforation and lead dislodgement after implantation of a MR-conditional pacing lead: a single-center experience. Pacing Clin Electrophysiol. 2014;37(1):4-10.
  21. Nazarian S, Hansford R, Roguin A, et al. A prospective evaluation of a protocol for magnetic resonance imaging of patients with implanted cardiac devices. Ann Intern Med. 2011;155:415-424.
  22. Kamerow D. No accidents. BMJ. 2004;328(7453):E300. 
  23. Sommer T, Luechinger R, Barkhausen J, et al. German Roentgen Society Statement on MR Imaging of Patients with Cardiac Pacemakers. Working Group on Cardiovascular Imaging, German Roentgen Society. Rofo. 2015;187(9):777-787.
  24. Wehrwein P. MRIs may be safe for people with pacemakers and ICDs. Harvard Health Blog. Available online at Accessed Oct 5, 2011.
  25. Walton-Shirley M. Cardiology’s biggest lie: No MRI for your device patient. Medscape. Published July 29, 2014. Available at: /viewarticle/829046. Accessed July 30, 2014.
  26. Al-Sabagh KH, Christensen BE, Thøgersen AM, et al. [Safety of magnetic resonance imaging in patients with pacemaker and implantable defibrillator]. Ugeskr Laeger. 2010;172:1740-1744.