Background: High-dose dobutamine stress magnetic resonance (DSMR) is a well-established imaging technique for the detection of coronary artery disease (CAD).
Purpose: To investigate the value of DSMR for the detection of in-stent restenoses (ISR) in patients with prior coronary stenting, using invasive coronary angiography (ICA) as the standard of reference.
Material and Methods: 50 patients with 74 stents and without wall motion abnormalities at rest were examined on a 1.5T MR scanner and underwent ICA for clinical reasons within 14 days after DSMR examination. A dobutamine/atropine stress protocol was employed until age-predicted heart rate was achieved, and imaging was performed in at least three long- and three short-axis views using a segmented steady-state free precession sequence (repetition/echo time [TR/TE] 3/1.5 ms, flip angle 60°). All examinations were read by an experienced cardiologist and radiologist in consensus, with myocardial ischemia being defined as a new stress-induced wall motion abnormality in at least one myocardial segment. Statistical analysis was performed on a per-vessel (left circumflex artery [LCX], left anterior descending artery [LAD], and right coronary artery [RCA]) basis and with regard to the number of affected vessels (one-, two- or three-vessel disease).
Results: ICA yielded seven ISR, of which one was missed by DSMR (sensitivity 86%, 95% confidence interval [CI] 0.42–0.99). Sixty-seven coronary arteries showed no ISR in ICA; however, due to new wall motion abnormalities, seven ISR were suspected in DSMR (2×RCA, 3×LCX, and 2×LAD; sensitivity 86%, specificity 90%, positive predictive value 46%, negative predictive value 98%, and diagnostic accuracy 89%). The per-vessel analysis of the three main coronary arteries revealed highest sensitivity (100%), specificity (93%), and diagnostic accuracy (94%) for the LAD.
Conclusion: High-dose DSMR is an accurate, noninvasive technique for the detection of ISR and reliably allows identification of patients who need to undergo control ICA.
