Pacemakers (PM) and MRI

On the basis of their carefully performed experimental studies and analysis of post-mortem explanted pacemakers (PMs), Irnich et al.¹ review some of the literature concerning PM and MRI. With excellence, the paper gives insights into the theoretical background of PM behaviour in magnetostatic and gradient fields. The report of fatal cases with post-mortem-investigation of the devices is as valuable and significant as the proposal of the well thought out strategy of performing MRI in PM-patients (ECG-triggered sequences while pacing in VVT mode). Ventricular fibrillation (VF) as a cause of death was supposed in three cases and was described as proven in three other cases. However, there is no evidence, because there was no documented rhythm monitoring during MRI scanning in any of these cases. Therefore, strictly spoken, the cause of death was probably VF in three cases and unknown in three further cases. If in such an uncontrolled setting the cause of death is diagnosed post-mortem as cardiac arrest, VF induced by asynchronous pacing may be one probable trigger. Primarily, non-PM-dependent patients, especially those with SSS, often become PM-dependent because of a prolonged sinus node recovery time. In such a case, cardiac arrest caused by sinus arrest in a setting of an MRI-inhibited PM could also be the trigger, leading to VF following cardiac hypoxia. There is no doubt that a risk of VF during asynchronous pacing, in particular, when intrinsic rhythm and asynchronous pacing compete at comparable rates, is present. Of course, this risk has to be minimized, but the risk per se should not be overestimated, as fast asynchronous pacing during MRI never induced VF under controlled conditions.

The authors state that heating is not a real problem in MRI scanning of PM patients, because it is concentrated around the surface of the electrode. However, lead impedance frequently decreases immediately after MRI (with recovery to baseline values in all cases) but almost never increases. This cannot only be regarded as a trend² but has also been shown to be a significant effect in a larger cohort (Sommer et al., submitted for publication). A decrease in lead impedance may reflect local reversible oedema, infiltrating the unexcitable layer surrounding the electrode surface. In most cases, such reversible effects on conduction properties are not detected by measurements of sensing and stimulation threshold. In contrast, in some cases, an increase in threshold values was found, indicating that heating reached and damaged the excitable layer.

1. In a previous study, early completely reversible increases in threshold values were observed in two PM-patients.²

2. In addition, several intentional MRIs in one ICD-patient showed a marked and persistent increase in sensing and stimulation thresholds.³

3. Temporary deterioration of pacing threshold has also been observed in an ICD-patient who underwent MRI by mistake.⁴

4. In a chronically instrumented canine model, a temporary deterioration of pacing threshold was found in one ICD-animal.⁵

In case of persistent deterioration, programmed output has to be adapted advancing PM replacement surgery and, therefore, influencing quality of life. Increase in temperature of up to 20°C with RF exposure was not only observed in vitro⁶ but also in vivo.⁷ This is in a range that is used for RF catheter ablation of cardiac accessory pathways and is probably capable of inducing tissue injury at the lead tips. As heating also occurred in the presence of blood flow,⁷ protection by the cooling effect of myocardial blood flow in any in vivo or clinical setting is small and must no longer be overestimated. In Luechinger's study, pathology showed no evidence for heating-induced damage, just because cell damage of implantation could not be distinguished from cell damage from heating. With all this knowledge of the effects of MRI on the lead tip properties Irnich's conclusion, that heating of the electrode is not a real problem in MRI scanning, should be carefully reconsidered. The theoretical considerations for estimation of expected heating effects by Irnich et al.,¹ no matter how brilliant they are and how convincing they appear, are only mathematical models. Deductions cannot beat reality, least of all when focusing on safety.

References