AIM: ICDs should no longer be contraindicated for an MRI
MRI is safe in some patients with cardiac pacemakers and defibrillators, according to a study published Oct. 4 in Annals of Internal Medicine. The authors’ approach provides a template for other practitioners, writers concluded in an accompanying editorial, but the protocols should be followed strictly after weighing clinical benefits and risks.

As many as 75 percent of patients who receive implantable cardioverter-defibrillators (ICDs) will develop conditions that may require MRI for diagnosis and treatment, and the aging population is likely to increase that need. But all ICDs and most of implanted pacemakers are considered a contraindication for MRI by the FDA and device manufacturers.

Saman Nazarian, MD, of Johns Hopkins Hospital in Baltimore, and colleagues designed a large prospective study to examine the safety of an MRI protocol for certain patients with pacemakers or ICDs, selecting a protocol based on previous studies and device programming that that would minimize inappropriate activation or inhibition of the devices. They selected 438 patients with devices who underwent 555 MRIs after being referred by a primary or subspecialty physician between February 2003 and April 2010.

Exclusion criteria included patients with leads implanted less than six weeks prior to study enrollment; patients with abandoned or epicardial leads; and pacemaker-dependent patients with ICDs. Only patients with pacemakers manufactured after 1998 and ICD generators manufactured after 2000 were included.

Magnetic strength was 1.5 T and imaging was done in one of two centers: Johns Hopkins Hospital and Rambam Medical Center in Haifa, Israel. A registered nurse experienced in device programming and cardiac life support was present and an electrophysiologist was available for immediate backup.

All 438 patients had baseline and immediate follow-up, with long-term telephone follow-up in January. Three of the patients experienced acute power-on-reset events with no long-term effects. Most MRIs produced clinically useful information, and artifacts were limited to thoracic procedures.

“The large number of patients in our study provided adequate power to detect small changes in device variables,” Nazarian and colleagues wrote. “Of note, no change in individual variables was large enough to require system revision or device programming. … Overall, MRI was performed safely in all patients.”

The researchers noted that 53 pacemaker-dependent patients without an ICD underwent MRI with no adverse events. They emphasized that for these patients programming the device to an asynchronous mode, having qualified personnel as monitors and having an external pacing backup available was critical.

“Pacemaker-dependent patients and ICD recipients with device generator models that seem susceptible to power-on-reset events should not have MRI,” they cautioned. “In addition, our findings should not be extrapolated to device models that were not evaluated in this study and lead configurations other than standard transvenous lead systems.”

They noted several limitations, including a 10 percent loss to follow-up, missing data, lack of a control group and small numbers for individual device models despite the overall large number of devices included in the study.

In an editorial, Matthew R. Reynolds, MD, and Peter J. Zimetbaum, MD, of Beth Israel Deaconess Medical Center and Harvard Medical School in Boston, added that only 15 percent of the patients had repeated scans, making accumulative effects unknown. They added that the findings showed MRI use in an appropriate patient population was not risk-free but rather a small risk. Nonetheless, they concluded that if the use of MRI is justified and if the protocol described in the study is followed, MRI using 1.5 T machines should be an option.

“Their and other recent data suggest that the risks of MRI in the presence of CRMDs [cardiac rhythm management devices], although potentially serious, have probably been overestimated and can be managed effectively in many cases,” they wrote. “In our opinion, the presence of a CRMD should no longer be considered an absolute contraindication to MRI.”