The prospective cohort study of 626 patients with non-magnetic resonance imaging (MRI)-conditional implantable cardioverter defibrillators (ICDs) suggests said ICDs remain suitable to treat detected tachyarrhythmias post-1.5-Tesla MRI.
Previous reports of adverse events have typically been noted during or immediately after MRI, and have included “end-of-life” battery status messages and loss of programmability during scanning in ICD models manufactured before 2000. Providing the framework for their study however, Joshua Ra (Johns Hopkins University, Baltimore, USA) et al stated that they are “unaware” of reports of non-MRI-conditional ICDs failure to terminate arrhythmias after MRI, but acknowledged that long-term data are currently limited.
Patients were selected by the researchers at Johns Hopkins Hospital on the basis of previous ICD with a clinical indication for MRI. Though, patients were excluded from the study if they had undergone lead implantation within the previous four weeks due to spontaneous dislodgements or lead parameter adjustments after initial implantation which typically occurs within this time period. Exclusion criteria also applied to patients with permanent surgical epicardial leads or nonfictional leads, had subcutaneous ICD systems or were pacing-dependant with cutaneous ICD systems.
During imaging, symptoms including pain, warmth and palpitations were monitored with an in-room speaker system. Blood pressure was also measured non-invasively every three minutes, alongside continuous electrocardiogram (ECG) monitoring.
A total of 813 MRI examinations were carried out—the researchers noted one examination required early termination due to power-on reset, after which generator function was fully restored. Collecting long-term ICD interrogations after initial MRI, Ra et al examined clinical outcomes including arrhythmia detection and treatment, generator or lead exchanges, adverse events, and death.
At a median follow-up of 2.2 years post-MRI to latest available ICD interrogation before generator or lead exchange in 536 patients, 4,177 arrhythmia episodes were detected, and 97 patients received ICD shocks. Sixty-one patients (10%) had 130 spontaneous ventricular tachycardia or fibrillation events terminated by ICD shocks. A total of 210 patients (33%) are known to have died (median, 1.7 years from MRI to death); three had cardiac arrhythmia deaths where shocks were indicated without direct evidence of device dysfunction.
Ra and colleagues asserted theirs is the first study to provide longer term follow-up data on ICD interrogations in a large cohort, concluding: “Our findings are consistent with the accumulating evidence that 1.5-Tesla MRI scans do not induce clinically significant functional changes in non–MRI-conditional ICDs, which could theoretically be due to heating at the lead-tissue interface, lead or generator dislodgement, or hardware and software damage after MRI”, they commented.
They noted several limitations concerning access to ICD interrogations for a subset of patients, and additionally adjudication of cause of death, which relied solely on death certificate data in a subset. Furthermore, as a single-centre study, they draw attention to issues generalising collected data to other clinical settings and MRI facilities.
Non-MRI-conditional ICD systems have been widely implanted over the last four decades, though MRI-conditional ICDs were introduced in 2015. The authors bring awareness to a “substantial” patient population with the former however, for whom access for clinically indicated MRI examinations is limited, likely from “continued perceived risks of MRI” in patients with these devices, they continued.
“Our study provides additional safety data for this patient population under a standardised imaging protocol”, however Ra and fellow authors conclude that future multicentre studies directly comparing non-MRI-conditional and MRI-conditional ICDs after MRI, may be of “interest to validate our findings”.