Author + information
- Received February 16, 2017
- Revision received May 11, 2017
- Accepted May 26, 2017
- Published online September 13, 2017.
- Brian D. Williamson, MDa,∗ (, )
- Douglas C. Gohn, MDb,
- Brian M. Ramza, MD, PhDc,
- Balbir Singh, MDd,
- Yan Zhong, PhDe,
- Shelby Li, MD, MSce,
- Liesa Shanahane,
- on behalf of the SureScan Post-Approval Study Investigators
- aBeaumont Health System, Troy and Royal Oak, Michigan
- bLancaster General Health, Lancaster, Pennsylvania
- cSaint Luke’s Mid America Heart Institute, Kansas City, Missouri
- dMedanta – The Medicity, Gurgaon, Haryana, India
- eMedtronic PLC, Minneapolis, Minnesota
- ↵∗Address for correspondence:
Dr. Brian D. Williamson, William Beaumont Hospital, 4600 Investment Drive, Suite 200, Troy, Michigan 48098.
Objectives This global, multicenter, prospective study, initiated to meet U.S. Food and Drug Administration condition-of-approval requirements, evaluated the safety and efficacy of the Medtronic magnetic resonance imaging (MRI)–conditional pacing system when used in an MRI environment in routine clinical practice. The primary endpoint was MRI-related complications. The secondary endpoint was the cumulative change in pacing capture threshold (PCT) for patients undergoing multiple MRI scans.
Background Large-scale, real-world evaluation of MRI in patients implanted with an MRI-conditional pacing system remains limited, with few published data for patients who undergo multiple MRI scans.
Methods Patients were enrolled and followed up prospectively from the time of implantation. Evaluation of the pacemaker function was performed before and after MRI. The MRI-related complication-free rate was evaluated. Changes in electrical performance after each scan and cumulative changes over multiple scans were analyzed.
Results In 81 centers, 2,629 patients were implanted with a complete SureScan pacing system (41.8% women, age 70.2 ± 12.5 years). A total of 526 patients (28.5%) received 872 clinically indicated MRI scans, including 58 thoracic scans. No MRI-related complications occurred during or after MRI, meeting the primary objective. Six (1%) MRI-related observations (atrial fibrillation, PCT increase, and chest symptoms) were reported. A total of 171 patients (32.5%) underwent 2 or more scans with no cumulative increase in PCT.
Conclusions This report constitutes the largest longitudinal MRI experience in patients implanted with an MRI-conditional pacing system. Results support the safety profile of the SureScan system and demonstrate for the first time that patients may safely undergo multiple MRI scans. (SureScan Post-Approval Study; NCT01299675)
Safe access to magnetic resonance imaging (MRI) has become a critical need for pacemaker patients. Most patients implanted with cardiac devices are more than 65 years of age and are likely to have 1 or more comorbidities. MRI is a fast-growing imaging modality and has become the standard of care for diagnosis across various medical disciplines (1). It is estimated that 50% to 75% of patients with a cardiac implantable electronic device (CIED) will need an MRI scan after the implantation over their lifetime (2). In the past, there have been sporadic reports of patient harm and MRI-related fatalities (3,4). The MRI-associated static magnetic field, gradient fields, and radiofrequency power can adversely affect the electronic operation of a pacemaker. The pacemaker lead can concentrate the energy during an MRI scan, resulting in heat production, potentially leading to cardiac tissue damage at the lead tip. The induced voltage can even stimulate tachycardia (2,5–10). Therefore, magnetic resonance (MR) scanning in patients with a CIED has been considered relatively contraindicated until recently, which has limited the use of MRI in many situations (11–13).
In 2011, Medtronic developed and released the first SureScan pacing system (EnRhythm MRI or Revo MRI implantable pulse generator and CapSureFix 5086 MRI leads) for safe access to MRI at 1.5-T but with positioning restriction for scans in the chest region. Design modifications were made to mitigate the potential hazards from the MR-induced electrical environment (11,14). Medtronic’s second generation SureScan pacing system (Advisa MRI) was developed to provide safe whole-body MRI without scan positioning restrictions (12,13). Preclinical testing (involving bench and animal investigations), computer modeling, and ultimately premarket randomized trials were conducted to understand the effects of MRI on the Revo MRI and Advisa MRI SureScan pacing systems. Recently, the SureScan feature was added to selected defibrillator and cardiac resynchronization systems.
The present study was conducted to meet the U.S. Food and Drug Administration approval requirements aimed to further confirm the safety and efficacy of the SureScan pacing system in the MRI environment in real-world clinical settings. Patient status, electrical performance, and event assessment were collected at implantation, at routine follow-ups, and immediately before and after MRI scans. A key objective of the post-approval study was to gain a greater understanding of the cumulative effect of multiple MRIs on patient safety, achievable only with longitudinal, prospective follow-up. Importantly, this report provides the first large-scale published experience to date for patients who have undergone multiple MRI scans.
Study design and data collection
This multicenter, longitudinal, observational study was conducted in the United States, Canada, and India. All patients provided written informed consent. Patients were eligible for inclusion if they were implanted with a complete SureScan pacing system. Prospective enrollment was defined as prior to the implantation procedure or up to 30 days post-implantation. Patients could be also retrospectively enrolled at the time of an MRI scan, contributing only to the multiple MRI scan objective. There were no protocol exclusions regarding patient age, sex, reason for MRI, pacemaker dependency, or anatomic region scanned.
After the implantation procedure, patients were followed up according to their clinic’s routine care practice every 6 months for 5 years. Data collection was intended to provide product performance assessment from implantation until exit (i.e., electrical measurements, event assessment, MRI status, device explantation, and patient death).
MRI-related information was collected for patients who underwent an MRI. MRI scans were prescribed on the basis of the clinical indication as determined by the patient’s referring physician. Conditions for safely scanning patients in the approved labeling included a maximum whole-body absorption rate value of 2 W/kg for each sequence, maximum gradient slew rate of 200 T/m/s, and SureScan mode programming enabled before the MR exposure (Figure 1). Pre- and post-MRI electrical measurements, MRI scan parameters, indication for scan, and data on anatomic region, patient monitoring during the scan, and image quality, along with event assessment, were collected. Study investigators were also surveyed for care pathway information to gain understanding on coordination between cardiology and radiology for this patient population.
The study complies with the Declaration of Helsinki for investigation in human beings and was approved by the institutional review board or ethics committee at all participating sites.
Events review and classification
Adverse device effect (ADE) events, defined as any untoward and unintended response to a medical device, were reported by sites on occurrence, regardless of whether an action was taken. Site-reported ADEs were reviewed and classified by an independent clinical events committee (CEC) for their relatedness to the pacing system or the MRI procedure. An MRI-related ADE was defined as an ADE caused by the interaction between the SureScan pacing and the MRI system or by the MRI programming and that occurred during the MRI procedure (includes the time the patient is within the 5-G line of the MRI system) or up to 1 month after the MRI procedure. In addition, ADEs were adjudicated as either complications or observations by the independent CEC. A complication was defined as an ADE that resulted in an invasive intervention, termination of significant device function regardless of other treatments, or permanent disability or death of a patient. Other events that were not complications were categorized as observations for this report.
The primary objective was to demonstrate that the MRI-related complication rate was <2%, that is, that the 1-sided confidence interval upper bound was lower than 2%. The secondary objective was to characterize the cumulative change in PCT for patients who underwent multiple MRI scans by measuring PCT longitudinally.
The MRI-related complication rate was calculated by the binomial method. Study sample size was calculated on the basis of the assumption that a population MRI-related complication rate is <2%. To ensure a reliable estimate of the complication rate, the study intended to collect data on MRI scan acquisition parameters from a minimum of 325 patients who were prospectively enrolled and underwent at least 1 MRI scan.
A minimum of 150 patients who underwent multiple MRI scans were required to characterize cumulative changes in PCT for the secondary analysis. To gain a complete understanding of the cumulative effects on PCT, 2 specific multiple-MRI–related questions were evaluated: 1) is there a significant acute effect on PCT with multiple MRI scans (defined as >0.5 V PCT change pre- to post-MR scan); and 2) are there significant long-term effects on PCT with multiple MRI scans? The longitudinal data were analyzed by use of the generalized estimating equations statistical method to account for within-patient effect. These analyses were performed separately for atrial and ventricular PCT values.
In addition, baseline characteristics of patients who underwent an MRI were compared with those patients who did not undergo an MRI scan during the study. The MRI scan rate was estimated as a function of time by the Kaplan-Meier method.
Study enrollment began after the U.S. Food and Drug Administration approval of the Revo MRI SureScan pacing system, with the first enrollment in April 2011. Over the duration of the study, additional MRI-conditional products, including the Advisa IPG (an implantable pulse generator), became commercially available; patients implanted with approved MRI-conditional systems were eligible to enroll within the respective geographic region where available (Online Appendix). A total of 2,690 patients were enrolled from 81 centers located in the United States (90.5%), Canada (2.5%), and India (7.1%). Of the total enrolled, 2,529 patients (94%) provided consent before or up to 30 days after the pacemaker implantation procedure. Of these 2,529 prospectively enrolled patients, 2,468 (97.6%) were implanted with a complete SureScan pacing system. The other 61 patients who initially provided consent to participate in the study received devices other than a SureScan pacing system and subsequently exited from the study. In addition, 161 patients (6%) were enrolled when they were told an MRI scan was indicated. Overall, 2,629 patients (2,468 + 161) were chronically implanted with a SureScan MR conditional pacing system. A summary of study patient disposition is presented in Figure 2.
Patient baseline characteristics were typical of pacemaker patients. Baseline characteristics of patients who received an MRI or MRIs and those who did not during the study are presented in Table 1. Patients with MRI scans were slightly younger (mean 68.1 years vs. 70.8 years), with a shorter QRS duration (105.9 ms vs. 110.9 ms); fewer had diabetes mellitus (22.4% vs. 26.9%), but a higher proportion reported sleep apnea (16.7% vs. 9.2%), and a much higher proportion reported a history of a computed tomography or MRI scan (69.4% vs. 48.4%) before study enrollment compared with those who did not undergo MRI procedures during this study. The mean follow-up was 26.6 ± 15.2 months.
Among all SureScan patients, 526 had undergone a total of 872 MRI scans as of April 2016. All reported MRI scans were medically indicated procedures. The majority (85.5%) of the MRI scans were conducted at facilities affiliated with the participating study site, although a small proportion of the scans were conducted at a facility not affiliated with the study site. Within the first year after the implantation of the pacing system, 10.7% of the patients had an MR scan. Cumulatively, 28.5% of patients had at least 1 MRI scan within 4 years post-implantation (Figure 3). These results confirmed safe access to MRI is a critical need for pacemaker patients.
Of the 526 patients who underwent 872 MRI procedures, 355 (67.5%) had 1 scan, 97 (18.4%) had 2 scans, and 74 (14.1%) had 3 or more scans. The maximum number of scans for an individual patient reported was 11. The most common reasons for MRI scans were neurological disorders (29%) and spinal disorders (16%), followed by cancer (12%). Pain and joint injuries were also important indications. The majority of scans were ordered by physicians other than a cardiologist or electrophysiologist. The most common anatomic locations scanned were head or neck (including brain, head, face, cervical spine, and carotid) at 52%, followed by abdominal or lumbar at 32%, and scans of body extremities at 13.5% (Table 2). The thoracic spine was also an important area of scanning, accounting for 7% of all scans. During the MRI procedure, patients were commonly monitored through pulse oximetry (71.6%), visual observation (70.4%), verbal communication (63.6%), and ECG (57.2%).
Of the total 872 MRI scan procedures, image quality was reported for 512 MRI scans (58.7%), of which 3.7% (19 scans) were reported associated with pacemaker-specific artifacts. Only 1 cardiac MRI scan was reported by the site as “unable to interpret” because of artifacts. A total of 722 participants (82.8%) responded to study survey questions regarding coordination between cardiology and radiology. Of these, 456 (63.2%) reported such coordination as being “easy,” 232 (32.1%) rated it as “neutral,” and 34 (4.7%) reported it as “difficult.”
Of the 526 patients who underwent at least 1 MRI scan after implantation of a SureScan pacing system, none experienced an MRI-related complication as defined previously. The study primary endpoint was met with a 0% MRI-related complication rate (95% confidence interval: 0% to 0.7%).
Although there were no MRI-related complications, 6 MRI-related events were reported in 6 patients implanted with the Revo MRI pacing system. All 6 events were classified as MRI-related observations. Two patients, who were pacemaker dependent and did not have a prior history of atrial fibrillation (AF), experienced AF during the MRI scan. Both events were self-terminating, and MRI scans were completed. Two patients reported chest pain and chest warmth, which resulted in the MRI scan procedures being stopped. Both patients’ symptoms resolved and did not recur. One of these 2 patients was successfully scanned uneventfully 4 days later. In addition, there was 1 reported event of failure-to-capture and 1 event of sudden rise of pacing threshold. In the failure-to-capture event, the threshold rise occurred in a non–pacemaker-dependent patient who underwent the first MRI for ankle and foot imaging due to joint injury. The scan occurred 43 days after pacemaker implantation. Pre-scan right ventricular (RV) PCT was 1.5 V at 0.4 ms, and post-scan RV PCT rose to 4.0 V at 0.4 ms. Two days post-scan, the threshold peaked at >8.0 V at 0.4 ms; however, capture occurred at 4.0 V at 1.4 ms. Subsequent threshold measured at 48 days post-MRI was 6.0 V at 0.4 ms. There was no evidence of lead dislodgement on a radiograph, and the lead was not repositioned. The device was reprogrammed to a ventricular output of 5.0 V at 1.5 ms. For the sudden elevated threshold event, the patient underwent the first MRI of the brain and neck for spontaneous ear pain. After the MRI, the right atrial (RA) PCT rose from 1.0 V at 0.4 ms to 2.0 V at 0.4 ms, and the RV PCT rose from 1.0 V at 0.4 ms to 2.5 V at 0.4 ms. Subsequent follow-up at 1 week post-MRI showed chronic PCTs returning to 0.5 V at 0.4 ms for both RA and RV leads. No intervention was deemed necessary by the investigators. No power reset or change in battery status was observed. All 6 events were resolved without any invasive clinical actions. Two other patients experienced a rise in PCT of 0.6 V and 0.7 V, respectively, which were not believed to represent an event by the participating center principal investigator.
A total of 224 enrolled patients (8.3%) died during the study. All deaths were reviewed and adjudicated by the CEC, and none were considered to be related to the SureScan pacing system or MRI procedure.
Lead performance with MRI
PCT measurements were completed before and after each MRI scan for both atrial and ventricular leads when the MRI was performed at a facility affiliated with the participating study site (85.5% of all scans). Median PCT change was 0 for both RA and RV chambers, with mean differences of −0.02 ± 0.32 V and −0.06 ± 0.32 V, respectively. A total of 4 patients (0.76%) reported a >0.5-V increase of PCT after the MRI scan, including 1 in the RA and 4 in the RV, with 1 patient reporting an increase in both chambers. Two of 4 patients reported MRI-related events as described previously. Figures 4A and 4B report the distribution of acute PCT changes observed in the study. There were no significant acute changes in impedance values after MRI scans.
RA and RV PCT data were summarized longitudinally for patients who underwent multiple MRI scans. PCTs were stable over time. Generalized estimating equations models were used to test whether multiple MRI scans had chronic effects on the pacemaker system by testing the effect of number of MRI scans (0, 1, or 2 vs. 3 or more scans) on the PCT value over time. Average PCT values over time for RA and RV chambers are displayed in Figures 4C and 4D. There was no statistical difference detected for RA and RV PCT in association with number of MRI scans.
This study represents the largest published real-world experience of MRI-conditional pacemaker patients to date. The primary finding is that no MRI-related complications were noted in the real-world clinical settings in this study. Moreover, this is the first report to demonstrate that multiple MRI scans have no cumulative effects on the pacemaker system, with stable PCT over time. There was a significant need for MRIs in pacemaker patients, with 28.5% of patients receiving at least 1 MRI within 4 years post-implantation.
Although hazardous effects of MRI on traditional pacemakers are well documented (2,5,6), several trials have shown that such imaging can be performed safely with certain protocols (7,15–19). These studies, however, typically involved a single site or a small number of sites, or limited scans to nonthoracic regions only. Additionally, they require more extensive resource utilization, such as cardiac monitoring and the presence of an advanced healthcare provider skilled in device management. Our study focused on MRI-conditional pacing systems and was performed in a large number of centers (81) located in 3 countries with a diverse patient population, using clinically available MRI machines and scan configurations for various anatomic scan positions. In contrast to previously published studies with relatively short follow-up after MRI scanning (which could only assess the early effects on the device function), this study followed patients long-term and was able to collect repeated MRI scans and analyze the cumulative effect on patient safety and PCT change.
Several MRI-related observations were noted in patients with the SureScan pacing systems. The maximum reported PCT increase immediately after the MRI scan was 2.5 V in the ventricle, which rose to 8 V 2 days later in 1 non–pacemaker-dependent patient. The site principal investigator raised the possibility that there was antecedent lead microdislodgement in this patient that was unrecognized before the MRI. This MRI scan was performed on the foot and ankle region. The pacemaker device was unlikely within the MRI bore during the scan, and thus, one would not expect the RV lead to have been exposed to sufficient magnetic field strength to result in dislodgement or significant radiofrequency energy heating at the lead-tissue interface. The actual threshold testing tracings for this patient could not be retrieved to confirm the pre-scan threshold, so we cannot exclude the possibility that a high threshold was present pre-scan and not recognized. Nevertheless, this event was counted as an MRI-related observation. No device revision was performed up to the time when the patient exited the study. A second patient undergoing a brain and neck MRI experienced a smaller rise in atrial and ventricular PCT from 1.0 V at 0.4 ms to 2.0 V at 0.4 ms and from 1.0 V at 0.4 ms to 2.5 V at 0.4 ms, respectively. Threshold measurements taken 4 months before the MRI occurred were also recorded as high, as 2.0 V at 0.4 ms and 2.5 V at 1.5 ms. This patient’s PCT returned to baseline at a 1-week post-MRI follow-up. The PCT change could be due to factors unrelated to the MRI, given the threshold measurement variation observed at other times for this patient (including pre-MRI). Two patients (0.4%) without prior AF developed sustained, self-terminating AF. No further episodes occurred after the MRI SureScan mode was programmed off, and no clinical interventions were needed. Further investigation concluded that the AF in these patients was likely induced by competitive pacing during MRI SureScan operation. In both cases, the asynchronous pacing rate was set to 60 beats/min and not sufficiently rapid to avoid competitive atrial pacing. To avoid competitive pacing, an appropriately faster asynchronous pacing rate for patients who require pacing support must be determined before the MRI scan is performed. Two patients (0.4%) reported chest pain or a feeling of warmth in the chest; these patients recovered after MRI termination with no further treatment. Similar chest symptoms have been reported previously (12,20).
A limited number of studies have reported that “legacy” or non–MRI-conditional CIEDs are safe for MRI (15,17,21); however, scanning of these patients has typically been performed in a carefully monitored setting by an advanced cardiovascular life support–certified healthcare provider skilled in device management in a select few experienced centers to ensure patient safety. Furthermore, although the incidence of adverse events is relatively rare, such events are unpredictable and potentially catastrophic.
Our study reflects a real-world experience in a global, multicenter study in centers without previous extensive experience with MRI of pacemaker patients. The MRIs performed were typically ordered by physicians other than those involved in the cardiac care or pacemaker follow-up. The care pathway used included a defined pre- and post-programming of MRI-conditional pacemakers with a SureScan mode that facilitated tailored pacing settings appropriate during MRI that was found to be practical and safe by the enrolling centers.
The observations in this study are an important reminder that even in devices labeled as MRI conditional, a protocol of uniformly checking device function pre-scan, identifying pacemaker dependency, careful rate programming, monitoring during scanning, and post-MRI device evaluation for events are important to ensure patient safety (22,23).
The study is not a randomized study by design but rather was aimed to confirm real-world experience with SureScan systems of the results from 2 randomized, investigational pre-market studies. Additionally, MRIs were obtained using 1.5-T magnets on the basis of the U.S. Food and Drug Administration labeling during the study, and no MRIs were obtained with 3-T MRI systems. No safety conclusions can be made regarding safety of scans with 3.0-T magnets. Fourteen percent of scans performed in this study occurred at facilities not affiliated with study sites. Finally, this report is limited to the Medtronic SureScan products that were evaluated.
This multicenter prospective study provides the largest data source for the evaluation of MR-compatible pacemakers in a real-world setting. Our results support that patients implanted with a SureScan pacing system can safely undergo 1 or multiple MRI examinations, with no evidence of any complications related to the MRI procedure. In approximately 1% of all patients, findings such as a significant rise in PCT, transient AF, or chest warmth or pain were observed. Although MRIs appear to be safe to perform in patients with these MRI-conditional pacemakers, a protocol of scanning consistent with the manufacturer labeling is recommended. Barriers remain for CIED patients to gain access and optimally benefit from the diagnostic utility of MRI (22). Results from our study could provide the clinical experience needed to reduce those barriers and improve patient care.
COMPETENCY IN MEDICAL KNOWLEDGE: MRI has been infrequently performed in patients with implantable pacemakers because of safety concerns. We report the safety of MRI in patients with implantable pacemakers designed for the MRI environment in the largest longitudinal experience to date.
TRANSLATIONAL OUTLOOK: There is an unmet need for MRI in patients with CIEDs. In this study, 28% of patients with implanted pacemakers had at least 1 MRI within 4 years post-implantation. Additional studies will help to expand safe access to MRI for all CIED patients. Although there were no complications after 872 MRIs, 1% of patients had transient atrial fibrillation, a rise in pacing capture threshold, or discomfort over the pacemaker. Additional studies are needed to further ensure these low-incidence events can be reliably avoided.
The authors wish to acknowledge the SureScan Post-Approval Study site investigators and coordinators who contributed to the study execution and data collection. The authors would also like to acknowledge the tremendous contribution of Erika Pouliot, Koen J.P. Verhees, PhD, and many other employees of Medtronic PLC, for their editorial and logistical support of the study and this manuscript. They especially thank all patients who participated in this study.
For supplemental tables, please see the online version of this paper.
The SureScan Post-Approval Study was sponsored by Medtronic (Minneapolis, Minnesota). Although Medtronic funded the study, and Medtronic employees participated in data collection and analysis, Medtronic did not review or have final approval of the manuscript. Drs. Williamson, Gohn, and Ramza have reported that their institutions received compensation fees from Medtronic for conducting clinical research protocols. Dr. Ramza reports that his institution received fellowship support from Medtronic, St. Jude Medical, and Boston Scientific, and he has been a paid speaker for St. Jude Medical. Dr. Singh is a consultant to and receives clinical trials funds from Medtronic. Dr. Zhong, Ms. Li, and Ms. Shanahan are employees of Medtronic.
- Abbreviations and Acronyms
- adverse device effect
- atrial fibrillation
- clinical events committee
- cardiac implantable electronic device
- magnetic resonance
- magnetic resonance imaging
- pacing capture threshold
- right atrial
- right ventricular
- Received February 16, 2017.
- Revision received May 11, 2017.
- Accepted May 26, 2017.
- 2017 The Authors
- Smith-Bindman R.,
- Miglioretti D.L.,
- Larson E.B.
- Sommer T.,
- Naehle C.P.,
- Yang A.,
- et al.
- Martin E.T.,
- Coman J.A.,
- Shellock F.G.,
- Pulling C.C.,
- Fair R.,
- Jenkins K.
- Schwitter J.,
- Kanal E.,
- Schmitt M.,
- et al.,
- for the Advisa MRI System Study Investigators
- Nazarian S.,
- Roguin A.,
- Zviman M.M.,
- et al.
- Bailey W.M.,
- Mazur A.,
- McCotter C.,
- et al.
- Miller J.D.,
- Nazarian S.,
- Halperin H.R.
- Lowe M.D.,
- Plummer C.J.,
- Manisty C.H.,
- Linker N.J.
- Sabzevari K.,
- Oldman J.,
- Herrey A.S.,
- Moon J.C.,
- Kydd A.C.,
- Manisty C.