Methods – 206 patients (58.1±15.5 years; 64% male, 16% non-conditional) underwent 235 consecutive CMR studies at 1.5 Tesla including (amongst others) cine- and late gadolinium enhancement (LGE)-imaging as well as first-pass myocardial perfusion imaging (MPI) at rest and during peak hyperemia (after 400µg regadenoson i.v.). Assessment of image quality was performed with a four-category grading scale (1=without limitation, 2=minor limitation with overall good assessability, 3=major limitation with poor assessability, 4=no assessability at all) for cine-, LGE- and MPI.

Results – There were no relevant procedure-associated technical or clinical safety issues. However, in 11 out of 34 patients with S-ICD (32%), we observed an inactivation of the alarm. A change of > 0.3V/0.4ms in pacing threshold or a change of > 2mV in sensing was detected in 15 devices after CMR (overall 6%; 6 out of 102 PM (6%), 7 out of 79 ICD (9%), 2 out of 15 CRT-ICD (13%)). These changes in electrode parameters were within normal limits and there was no relationship to device type, location, MR conditionality and required pacing during CMR in a binary logistic regression model. Ordinal logistic regression analysis demonstrated that CMR image quality in cine-, LGE- and MPI was significantly associated with a) device type and b) location (cine: p<0.001, R²=0.49; LGE: p<0.001, R²=0.54; MPI: p<0.001, R²=0.50) - but not with device conditionality or required pacing during CMR. Compared to PM, a significantly reduced image quality was observed for ICD, CRT-ICD and S-ICD regarding cine-, LGE- and MPI. Moreover, S-ICD showed a reduced image quality compared to transvenous ICD regarding cine-, LGE- and MPI. Right pectoral location showed a substantially better image quality compared to left pectoral location in cine- and LGE-images as well as MPI (p<0.001).

Conclusions – CMR studies can be safely performed in patients with CIEDs. Both device type and location are significantly associated with CMR image quality. In general, CMR image quality is not affected in case of PM and right pectoral located devices. However, CMR image quality can be reduced in the anterior LV segments in ICD and CRT-ICD patients, and in LV lateral segments in S-ICD patients. These potential limitations in CMR image quality should be considered in CIED patients when making the indication and during post-procedural image analysis.