The objective of this study was to investigate whether it is safe to perform 1.5-Tesla magnetic resonance imaging (MRI) scans in pacemaker (PM) patients without pulse oximetry or electrocardiogram monitoring and with no special specific absorption rate (SAR) or time limits, provided that the PMs are interrogated and programmed to asynchronous mode prior to the scan.
This study reports the outcome of 207 MRI scans on PM patients at Rigshospitalet, Copenhagen University Hospital from June 2010 to September 2013. All MRIs were performed with the PMs in asynchronous mode and without additional monitoring. There were no adverse events registered among the PM patients during the study period. The only statistically significant change after MRI scans was a small, but clinically insignificant increase in atrial sense. No occurrences of reprogramming to power-on-reset were registered.
It is possible to perform MRI scans relatively safely in PM patients without additional monitoring or change in the normal MRI protocol, given that the PM has been assessed and reprogrammed prior to MRI. This is especially important to remember in the acute setting where MRI scans may be delayed when monitoring facilities are unavailable.
Magnetic resonance imaging (MRI) can be performed safely on non-MRI conditional pacemaker (PM) patients without additional monitoring if PMs have been assessed and reprogrammed beforehand.
In the acute setting where exact diagnostics is crucial, MRI scans should not be cancelled or postponed because monitoring facilities are not available.
Introduction
In recent years, the use of both magnetic resonance imaging (MRI)1 and the use of pacemakers (PMs) for treatment of cardiac arrhythmias, have been increasing. In the United States, the overall use of PMs increased with 55.6% from 1993 to 2009, and PM patients have become older.2 It has been estimated that patients with PMs have a 50–75% risk of being referred for an MRI during the lifetime of their device.3
However, implantable electronic devices are subjected to strong electromagnetic fields when placed in an MRI scanner, and although poorly characterized, several fatalities have been documented on MRI procedures in PM patients.4 The effects of MRI can be categorized into three different groups: (i) The force from the static magnetic field, (ii) the energy from the gradient magnetic field, and (iii) the energy from the radiofrequency (RF) field. The potential problems are related to force and torque, induction of voltage, heating, pacing, programming changes, and reed switch malfunction.5,6 As a precaution against these potential issues, it has been advised to keep PM patients closely monitored during MRI scans.6,7 A thorough review of these potential problems reveals that the issues of force and torque, induction of voltage, and heating are insignificant. Other matters such as inhibition of pacing, programming changes, and reed switch malfunctions can be handled with proper preparation and assessment prior to MRI.6,8 While changes in practices of monitoring during MRI scans are emerging,9–11 no studies have been published demonstrating the safety of unmonitored MRI scans following standard clinical protocols in PM patients.
The PM may register electromagnetic interference (EMI) from the MRI scanner as electric activity from the heart. If the PM misinterprets EMI as spontaneous cardiac electric action, it may respond by inhibiting PM output. This can cause bradycardia and/or asystole. The solution to the problem of inhibition has been to reprogramme the PM to asynchronous mode, e.g. VOO/DOO. Thus, the PM will not be inhibited by detection of cardiac signals or EMI.6
Unexpected programming changes, e.g. resetting to default parameters (power-on-reset/electrical reset) have been reported during MRI scans. This is due to battery depletion or EMI interfering with the PM.9 Power-on-reset is resetting, for example to VVI mode, with a frequency of 65 beats per minute (bpm) and a high output. Since EMI during an MRI scan may result in PM output inhibition, resetting to VVI will result in asystole in a fully PM-dependent patient. Power-on-reset mode may also be problematic in patients who are dependent on alternative programming, e.g. small children who need a heart rate above 65 or patients who need a PM output higher than the standardized value. Though power-on-reset during MRI has been reported, with varying frequencies12–15 none of these incidents have had a fatal outcome. Preliminary results from a still ongoing study report 881 MRI scans of PM with 5 (0.6%) episodes of power-on-reset.16
All PMs have a reed switch that functions as the PM's ‘panic switch’. Activation of the reed switch, by placing a strong magnet on the skin close to the PM, will result in temporary reprogramming of the PM to asynchronous mode. It is impossible to predict the position of the reed switch in the MRI scanner.17 Since the PM should already be programmed to an asynchronous mode prior to an MRI, the position of the reed switch is not significant. However, if the reed switch mode has a different heart rate than the pre-programmed asynchronous mode, one would expect to see a heart rate that changes during the scan, depending on the position of the reed switch, as experienced by Kaasalainen et al. (2014).9
In June 2010, the policies of surveillance during MRI scans in PM patients at Rigshospitalet were changed. From June 2010 until September 2013, PM patients were scanned in the same way as all other patients, except that they in collaboration with the PM clinic had their PMs reprogrammed during the scans. This is, to our knowledge, the only study on the risk associated with MRI scanning of PM patients without any extra cardiac monitoring.
Material and methods
We have registered all PM patients having MRI scans performed from June 2010 to September 2013. Only PMs were included, no ICD's and no leadless PM systems were included in this study. All patients went through our pacemaker MRI protocol (see below) before MRI scans. According to this protocol, no additional monitoring or changes in MRI protocols were applied for PM patients compared with other patients. It is clinical standard in our department that patients always have an alarm button for emergencies during the scan. Visual contact is limited during the scan, and only voice contact is available during pauses in the scan due to noise from the scanner during active scanning.
All adverse events during MRI scans were registered in the MRI department. An adverse event was defined as termination of MRI, arrhythmias requiring immediate therapy, programming changes, electrode failure, or device failure. Besides adverse events, we also investigated if changes in PM status could be detected. PM interrogations before and after MRI scans were retrieved retrospectively from the PM clinic and analysed for sensing, threshold, and lead impedance, if available. Since data were retrieved retrospectively, only interrogations that had been printed and saved on file were available. If PM interrogations were unavailable, the patients' records were read through to detect any cases of adverse events. If any of the interrogated values were registered as an interval, the average value was calculated and used. Programming of the devices, type of PM (DDD, VVI, AAI), device manufacturer (Medtronic, Biotronik, St. Jude Medical, Boston, Guidant, Pacesetter, Sorin, Ela Medical), implantation date, and body part scanned (heart, spine, chest, neck, abdomen, brain, upper extremities, pelvic region, lower extremities) were also registered. If more than one body part was scanned, the one closest to the PM was registered. All patients in Denmark with PMs are in a registry (Danish Pacemaker Register) where all procedures, leads, and generators are registered, as are explanted and abandoned leads. This register was checked for abandoned leads, lead type, implantation site, and PM type.
All MRI scans were performed on a 1.5-Tesla Magnetom Vision, Siemens Medical System, Erlangen or a 1.5-Tesla GE Signa scanner. Maximum gradient strength was 25 milliTesla/meter (mT/m). Specific energy absorption rate (SAR) for the entire body did not exceed normal clinical limits. The project was considered a quality control project by the Ethics Committee for the Capital Region of Denmark (protocol number H-4-2015-FSP).
Pacemaker MRI protocol
Just before the MRI scans, PMs were interrogated for status on sensing, pacing threshold, lead impedances, and battery capacity. To avoid programming changes during the MRI scan, battery voltage had to be acceptable (>2.7 V and estimated battery lifetime >6 months). Patients were not excluded from MRI because of recent PM implantation. If PM status was acceptable, the PMs were programmed to AOO, VOO, or DOO mode, to avoid MRI-induced PM inhibition, regardless of the patient being PM dependent or not. Pacing rate was set to 80–110 bpm, 20 bpm above the actual intrinsic heart rate/pace frequency. All PMs were programmed to high output (5 V, 1 ms). Other PM functions such as rate-response function and mode-switch function were inactivated. No additional monitoring was performed on PM patients during the MRI scans, compared with patients without a PM. Immediately after the MRI scans (typically within 1 h), the PMs' sense, threshold, impedance, and battery status were checked, and the PMs were reprogrammed to the original mode and settings. In the hospital, a cardiac rescue team is available within 2 min, and the PM clinic is within reach in 5–10 min.
Our present guideline for MRI scans of PM patients.
Statistics
Sense, threshold, and lead impedance were analysed using paired Student's t-test. Due to a skewed distribution, the variables were logarithmically transformed. Thus, results of the paired t-tests report per cent change of values after the scans compared with before the scans. A P-value of ≤0.05 was considered statistically significant. If patients had undergone several scans, only the first one in the study period was included and subsequent scans were discarded to ensure independency of observations.
Results
In the period between June 2010 and September 2013, a total of 207 MRI scans in 184 different PM patients have been registered. No adverse events were registered in any of the 207 MRI scans during the study. None of the scans performed were deemed inconclusive due to PM artefacts.
Of the 184 patients included, 1 had a cardiac resynchronization therapy (CRT) system, 5 patients had MRI conditional PMs, and the rest had standard PM systems. Two patients had epicardial systems. Of these patients, 134 patients had left-sided systems, 11 right-sided systems, 2 abdominal systems, and 37 did not have an implantation site registered. Four patients had registered abandoned leads: 3 with abandoned endocardial leads and 1 with epicardial leads. There were no reports of complications in these patients either.
If patients had abandoned leads registered, their case was evaluated by a senior cardiologist to investigate whether the leads could be problematic. Chest X-ray was not taken as a routine.
Unfortunately, some PMs were interrogated, but printed data were not saved on file afterwards. Consequently, complete data set was not available for all scanned patients. Device interrogations before and after MRI were available in 137 PM patients since 16 scans were excluded because the patient had a prior scan in the study period and 54 scans were excluded due to inadequate data (i.e. no PM interrogations on file or only PM interrogations from before or after the MRI scan). Ninety-two patients had a dual-chamber PM (DDD) (67.2%), 35 patients a VVI PM (25.5%), and 10 patients an AAI PM (7.3%). Of these 137 patients, 6 were scanned within 42 days of PM implantation (range 9–31 days).
Table 1 shows the distribution of scanned regions. The majority of scans were directed at the spine (42.3%) and brain (37.2%), whereas the rest of the regions accounted for <6% each.
Scanned regions
| Region scanned | Number | % |
|---|---|---|
| Heart | 7 | 5.1 |
| Spine | 58 | 42.3 |
| Chest | 1 | 0.7 |
| Neck | 1 | 0.7 |
| Abdomen | 6 | 4.4 |
| Brain | 51 | 37.2 |
| Upper extremities | 2 | 1.5 |
| Pelvic region | 3 | 2.2 |
| Lower extremities | 8 | 5.8 |
| Total | 137 | 100.0 |
| Region scanned | Number | % |
|---|---|---|
| Heart | 7 | 5.1 |
| Spine | 58 | 42.3 |
| Chest | 1 | 0.7 |
| Neck | 1 | 0.7 |
| Abdomen | 6 | 4.4 |
| Brain | 51 | 37.2 |
| Upper extremities | 2 | 1.5 |
| Pelvic region | 3 | 2.2 |
| Lower extremities | 8 | 5.8 |
| Total | 137 | 100.0 |
Distribution of different regions scanned in the patients.
Scanned regions
| Region scanned | Number | % |
|---|---|---|
| Heart | 7 | 5.1 |
| Spine | 58 | 42.3 |
| Chest | 1 | 0.7 |
| Neck | 1 | 0.7 |
| Abdomen | 6 | 4.4 |
| Brain | 51 | 37.2 |
| Upper extremities | 2 | 1.5 |
| Pelvic region | 3 | 2.2 |
| Lower extremities | 8 | 5.8 |
| Total | 137 | 100.0 |
| Region scanned | Number | % |
|---|---|---|
| Heart | 7 | 5.1 |
| Spine | 58 | 42.3 |
| Chest | 1 | 0.7 |
| Neck | 1 | 0.7 |
| Abdomen | 6 | 4.4 |
| Brain | 51 | 37.2 |
| Upper extremities | 2 | 1.5 |
| Pelvic region | 3 | 2.2 |
| Lower extremities | 8 | 5.8 |
| Total | 137 | 100.0 |
Distribution of different regions scanned in the patients.
Table 2 shows the distribution of PM manufactures in the study. The majority of PMs were produced by Medtronic (52.6%) and St. Jude Medical (22.6%).
Pacemaker manufactures
| PM manufacturer | Number | % | Implantation date (range) | PM age at MRI in days (range) | ||
|---|---|---|---|---|---|---|
| Medtronic | 72 | 52.6 | 4 June 2004 | 5 April 2013 | 14 | 2760 |
| Biotronik | 11 | 8.0 | 31 July 20009 | 22 April 13 | 24 | 1188 |
| St. Jude Medical | 31 | 22.6 | 8 January 2002 | 22 May 2011 | 176 | 3866 |
| Boston | 10 | 7.3 | 5 March 2010 | 26 January 2012 | 110 | 1040 |
| Guidant | 6 | 4.4 | 15 January 2003 | 10 December 2007 | 1341 | 3141 |
| Pacesetter | 2 | 1.5 | 1 July 1997 | 24 September 2012 | 59 | 5008 |
| Sorin | 4 | 2.9 | 6 November 2009 | 26 September 2012 | 9 | 633 |
| ELA Medical | 1 | 0.7 | 15 January 2007 | 15 January 2007 | 1858 | 1858 |
| Total | 137 | 100.0 | 1 July 1997 | 22 April 2013 | 9 | 5008 |
| PM manufacturer | Number | % | Implantation date (range) | PM age at MRI in days (range) | ||
|---|---|---|---|---|---|---|
| Medtronic | 72 | 52.6 | 4 June 2004 | 5 April 2013 | 14 | 2760 |
| Biotronik | 11 | 8.0 | 31 July 20009 | 22 April 13 | 24 | 1188 |
| St. Jude Medical | 31 | 22.6 | 8 January 2002 | 22 May 2011 | 176 | 3866 |
| Boston | 10 | 7.3 | 5 March 2010 | 26 January 2012 | 110 | 1040 |
| Guidant | 6 | 4.4 | 15 January 2003 | 10 December 2007 | 1341 | 3141 |
| Pacesetter | 2 | 1.5 | 1 July 1997 | 24 September 2012 | 59 | 5008 |
| Sorin | 4 | 2.9 | 6 November 2009 | 26 September 2012 | 9 | 633 |
| ELA Medical | 1 | 0.7 | 15 January 2007 | 15 January 2007 | 1858 | 1858 |
| Total | 137 | 100.0 | 1 July 1997 | 22 April 2013 | 9 | 5008 |
Distribution of the different manufactures of the PMs implanted in the patients who underwent MRI scanning and their implantation date range and PM age at MRI.
Pacemaker manufactures
| PM manufacturer | Number | % | Implantation date (range) | PM age at MRI in days (range) | ||
|---|---|---|---|---|---|---|
| Medtronic | 72 | 52.6 | 4 June 2004 | 5 April 2013 | 14 | 2760 |
| Biotronik | 11 | 8.0 | 31 July 20009 | 22 April 13 | 24 | 1188 |
| St. Jude Medical | 31 | 22.6 | 8 January 2002 | 22 May 2011 | 176 | 3866 |
| Boston | 10 | 7.3 | 5 March 2010 | 26 January 2012 | 110 | 1040 |
| Guidant | 6 | 4.4 | 15 January 2003 | 10 December 2007 | 1341 | 3141 |
| Pacesetter | 2 | 1.5 | 1 July 1997 | 24 September 2012 | 59 | 5008 |
| Sorin | 4 | 2.9 | 6 November 2009 | 26 September 2012 | 9 | 633 |
| ELA Medical | 1 | 0.7 | 15 January 2007 | 15 January 2007 | 1858 | 1858 |
| Total | 137 | 100.0 | 1 July 1997 | 22 April 2013 | 9 | 5008 |
| PM manufacturer | Number | % | Implantation date (range) | PM age at MRI in days (range) | ||
|---|---|---|---|---|---|---|
| Medtronic | 72 | 52.6 | 4 June 2004 | 5 April 2013 | 14 | 2760 |
| Biotronik | 11 | 8.0 | 31 July 20009 | 22 April 13 | 24 | 1188 |
| St. Jude Medical | 31 | 22.6 | 8 January 2002 | 22 May 2011 | 176 | 3866 |
| Boston | 10 | 7.3 | 5 March 2010 | 26 January 2012 | 110 | 1040 |
| Guidant | 6 | 4.4 | 15 January 2003 | 10 December 2007 | 1341 | 3141 |
| Pacesetter | 2 | 1.5 | 1 July 1997 | 24 September 2012 | 59 | 5008 |
| Sorin | 4 | 2.9 | 6 November 2009 | 26 September 2012 | 9 | 633 |
| ELA Medical | 1 | 0.7 | 15 January 2007 | 15 January 2007 | 1858 | 1858 |
| Total | 137 | 100.0 | 1 July 1997 | 22 April 2013 | 9 | 5008 |
Distribution of the different manufactures of the PMs implanted in the patients who underwent MRI scanning and their implantation date range and PM age at MRI.
Table 3 shows the registered PM measurements before and after MRI. Both sense and threshold values showed an increasing, but clinically insignificant trend. This applies to both atrial and ventricular electrodes. The change in sense values for the atrial electrodes was statistically significant, but without clinical relevance. The impedance of both atrial and ventricular leads showed a decreasing trend, but neither was statistically significant. There were no observations of threshold increases >0.75 mV.
Pacemaker measurements
| N | Before MRI | After MRI | Mean change in per cent | P-value | |
|---|---|---|---|---|---|
| Atrial electrodes (n = 102) | |||||
| Sense (mV) | 69 | 2.3 (0.9, 6.1) | 2.6 (0.7, 9.2) | 13% (3, 24) | 0.015 |
| Threshold (V) | 51 | 0.6 (0.2, 1.5) | 0.7 (0.3, 1.6) | 12% (0, 26) | 0.06 |
| Impedance (Ω) | 82 | 428 (269, 681) | 427 (272, 671) | 0% (−1, 1) | 0.87 |
| Ventricular electrodes (n = 127) | |||||
| Sense (mV) | 64 | 11.1 (4.2, 29.7) | 11.3 (4.0, 31.7) | 1% (−4, 7) | 0.63 |
| Threshold (V) | 83 | 0.7 (0.3, 1.6) | 0.7 (0.3, 1.6) | 3% (−4, 10) | 0.41 |
| Impedance (Ω) | 109 | 519 (302, 891) | 518 (309, 869) | 0% (−1, 1) | 0.87 |
| N | Before MRI | After MRI | Mean change in per cent | P-value | |
|---|---|---|---|---|---|
| Atrial electrodes (n = 102) | |||||
| Sense (mV) | 69 | 2.3 (0.9, 6.1) | 2.6 (0.7, 9.2) | 13% (3, 24) | 0.015 |
| Threshold (V) | 51 | 0.6 (0.2, 1.5) | 0.7 (0.3, 1.6) | 12% (0, 26) | 0.06 |
| Impedance (Ω) | 82 | 428 (269, 681) | 427 (272, 671) | 0% (−1, 1) | 0.87 |
| Ventricular electrodes (n = 127) | |||||
| Sense (mV) | 64 | 11.1 (4.2, 29.7) | 11.3 (4.0, 31.7) | 1% (−4, 7) | 0.63 |
| Threshold (V) | 83 | 0.7 (0.3, 1.6) | 0.7 (0.3, 1.6) | 3% (−4, 10) | 0.41 |
| Impedance (Ω) | 109 | 519 (302, 891) | 518 (309, 869) | 0% (−1, 1) | 0.87 |
Measurements on atrial and ventricular electrodes before and after MRI scanning.
N, number; MRI, magnetic resonance imaging; mV, millivolt; V, volt; Ω, Ohm.
Pacemaker measurements
| N | Before MRI | After MRI | Mean change in per cent | P-value | |
|---|---|---|---|---|---|
| Atrial electrodes (n = 102) | |||||
| Sense (mV) | 69 | 2.3 (0.9, 6.1) | 2.6 (0.7, 9.2) | 13% (3, 24) | 0.015 |
| Threshold (V) | 51 | 0.6 (0.2, 1.5) | 0.7 (0.3, 1.6) | 12% (0, 26) | 0.06 |
| Impedance (Ω) | 82 | 428 (269, 681) | 427 (272, 671) | 0% (−1, 1) | 0.87 |
| Ventricular electrodes (n = 127) | |||||
| Sense (mV) | 64 | 11.1 (4.2, 29.7) | 11.3 (4.0, 31.7) | 1% (−4, 7) | 0.63 |
| Threshold (V) | 83 | 0.7 (0.3, 1.6) | 0.7 (0.3, 1.6) | 3% (−4, 10) | 0.41 |
| Impedance (Ω) | 109 | 519 (302, 891) | 518 (309, 869) | 0% (−1, 1) | 0.87 |
| N | Before MRI | After MRI | Mean change in per cent | P-value | |
|---|---|---|---|---|---|
| Atrial electrodes (n = 102) | |||||
| Sense (mV) | 69 | 2.3 (0.9, 6.1) | 2.6 (0.7, 9.2) | 13% (3, 24) | 0.015 |
| Threshold (V) | 51 | 0.6 (0.2, 1.5) | 0.7 (0.3, 1.6) | 12% (0, 26) | 0.06 |
| Impedance (Ω) | 82 | 428 (269, 681) | 427 (272, 671) | 0% (−1, 1) | 0.87 |
| Ventricular electrodes (n = 127) | |||||
| Sense (mV) | 64 | 11.1 (4.2, 29.7) | 11.3 (4.0, 31.7) | 1% (−4, 7) | 0.63 |
| Threshold (V) | 83 | 0.7 (0.3, 1.6) | 0.7 (0.3, 1.6) | 3% (−4, 10) | 0.41 |
| Impedance (Ω) | 109 | 519 (302, 891) | 518 (309, 869) | 0% (−1, 1) | 0.87 |
Measurements on atrial and ventricular electrodes before and after MRI scanning.
N, number; MRI, magnetic resonance imaging; mV, millivolt; V, volt; Ω, Ohm.
Statistical power for subgroup analyses was unfortunately insufficient. These analyses included stratifications for high-impact vs. low-impact region scanned (high impact defined as the heart, chest, neck, and abdomen); repetitive scans vs. single scans; PM manufacturers vs. each other; and patients with recent implantation vs. all other scans. All analyses were non-significant.
Discussion
Many studies have investigated the impact of MRI scans on PMs.4,9–14,16,18–22 After 12 years without any unexpected adverse events in PM patients closely monitored (blood pressure, electrocardiogram, and pulse oximetry) by a doctor during MRI scans,23 we chose not to use any extra monitoring of PM patients in the MRI scanner. During these 12 years prior to this study, a single incident of power-on-reset was encountered. This happened in a patient with an 11-year-old PM with low battery capacity and therefore predisposed to resetting.23 The present study is, to our knowledge, the first study published on PM patients with non-MRI conditional PM undergoing standard clinical MRI without close monitoring. Previous studies have been limited by strict safety restrictions. This includes restrictions with respect to SAR limits, scanning time, scanned body region, types of PMs, PM dependency, and the type of monitoring of the patients. It is important to emphasize that we are an experienced centre within the area of cardiac implantable devices and MRIs. Our centre performed more than 800 device procedures and 426 new implantations (PMs, CRTs, and ICDs) in 2014 and performed 85 MRI scans of PM patients in the same year. Our pacemaker MRI protocol did not apply any extra safety restrictions during MRI scans of patients with PMs. Thus, PM patients with indication for a clinical MRI scan were subjected to normal scanning times, normal SAR values, and no extra monitoring compared with other patients undergoing MRI scans, except that MRI scans were only performed on 1.5-Tesla scanners. This resulted in no incidents of programming changes, malignant arrhythmias, electrode failure, or device failure during the study period. All PM patients had their PM interrogated prior to MRI with regard to sense, impedance, threshold, and battery voltage before reprogramming the PMs to asynchronous mode. After MRI, all PMs were assessed immediately for parameter changes and change in programming, and reprogrammed to pre-MRI settings. The analysis of PM status showed a small, but statistically significant increase in the atrial sensing. Though clinically not significant, there was an overall tendency towards increasing values of sensing and threshold for both atrial and ventricular electrodes. We did not have a later follow-up to see whether these changes were transitory. Previous studies on the effect of MRI on PMs have reported small, clinically insignificant changes,4,12,14,16,18–21 whereas other studies have found no changes in PM sensing, threshold, or impedance.13,22 Findings have not been consistent, and changes have varied between increased threshold and impedance,19 decreased sensing and impedance and increased threshold,14 decreased sensing and impedance,18 and increased threshold.12 Hence, there is no clear picture of which changes MRI does induce, if any. In our opinion, the possible event of power-on-reset is the remaining concern when MRI scanning PM patients. As stated by Higgins et al.,15 power-on-reset is an infrequent event, but it can cause deleterious changes in pacing mode and heart rate. The risk of power-on-reset is, however, very low if the patient's PM is relatively new (market released after 2001),14,15 and the battery status of the PM is good.6
We have followed these simple rules and did not experience any problems during 207 scans. The purpose of this paper is not to suggest that there are no risks when performing MRI scans on PM patients. We are aware that the relatively small number of patients included in this study is a limitation. We cannot, based on this study, conclude that it is perfectly safe to perform MRI scans on PM patients. To reach such a conclusion, a large controlled trial with several thousand patients would be necessary. Nevertheless, the risk of experiencing an adverse event with PMs programmed to asynchronous mode in the MRI scanner is very low, and we worry that clinically important scans are being postponed or cancelled because of concerns about clinically irrelevant PM problems. Monitoring is very resource consuming, and we believe that it adds little to patients' safety when our safety protocol is followed.
We would like to emphasize that we after concluding this study have changed our guidelines to include monitoring during scheduled MRI scans of fully PM-dependent patients and patients with older devices. This decision has been based on findings in recent large international studies,14,16 suggesting that power-on-reset happens in 0.6–0.8% of scans. This has been done to detect the extremely rare event of a power-on-reset happening in a fully PM-dependent patient. However, we still advocate that if a patient, who is fully PM dependent, needs an urgent scan, this should not be postponed if cardiology staff is unavailable for monitoring the patient. Since the risk of serious adverse events (power-on-reset) is very small, the benefits of performing the MRI scan will outweigh the risks in any number of cases.
We recommend that resources spent on monitoring be directed towards systematic assessment before MRI scans, especially focusing on battery voltage and age of PM, to avoid power-on-reset. As many studies suggest, after experiencing power-on-reset, many of these cases can be explained by low battery capacity or an old device.14,15,24 We believe that with thorough preparation before MRI scans, it is relatively safe to refrain from additional monitoring in PM patients undergoing MRI. We strongly advise that
- Pacemakers only are scanned if they have a battery voltage of >2.7 V and an estimated battery lifetime of >6 months. If this implies a PM replacement earlier than otherwise planned, it is our opinion that this is preferable to avoid problems during an MRI scan.
- A cardiac rescue team is available shortly for emergencies to introduce Zoll pacing until transvenous temporary pacing can be established.
- Fully PM-dependent patients and patients with older PMs are monitored during the scan.
Limitations
The greatest limitation to this study is its relatively small size. If, as literature suggests, a power-on-reset event is expected for each 200–250 scans, a single dangerous episode could have been missed simply by chance. The present study was performed retrospectively. We were unable to identify full measurements of all patients, and thus the number of values for comparison does not reach the total number of patients included in the study. Since some patients were referred from other hospitals, we may not have registered smaller problems following the scan.
Conclusions
In conclusion, this study supports the assumption that it is relatively safe to perform MRI scans in PM patients without additional monitoring, given that the PM has been reprogrammed to asynchronous mode prior to the MRI and there is an easily accessible in-house cardiac rescue team and PM clinic. However, we recommend that planned MRI scans in fully PM-dependent patients be performed with monitoring.
Funding
The work was funded by internal sources.
Conflict of interest: H.H.P. has received speaker's fee from and is a consultant in Medtronic, St. Jude, Biotronik, Biosense Webster, and Boston Scientific. J.H.S. has received research grant from Medtronic and speaker bureau for Medtronic, Biotronik, and Boehringer-Ingelheim. Remaining authors have no conflicts of interest.
