Prognostic implications of dipyridamole cardiac MR imaging: a prospective multicenter registry.

PURPOSE To evaluate dipyridamole cardiac magnetic resonance (MR) imaging in the prediction of major events (MEs) in patients with ischemic chest pain in a large multicenter registry. MATERIALS AND METHODS Institutional ethics committee approval and written informed consent were obtained. A total of 1722 patients who were undergoing cardiac MR imaging for chest pain were included. Wall motion abnormalities (WMAs) at rest, hyperemia perfusion defect (PD), late gadolinium enhancement (LGE), and inducible WMA were analyzed (abnormal if more than one abnormal segment was seen) with the 17-segment model. A cardiac MR categorization was created: category 1, no PD, LGE, or inducible WMA; category 2, PD without LGE and inducible WMA; category 3, LGE without inducible WMA; and category 4, inducible WMA. The association with ME was analyzed by using Cox proportional hazard regression multivariate models. RESULTS During a median follow-up period of 308 days, 61 MEs (4%) occurred (36 cardiac deaths, 25 nonfatal myocardial infarctions). MEs were associated with a greater extent of WMA, PD, LGE, and inducible WMA (P ≤ .001 for all analyses). In multivariable analyses, PD (P = .002) and inducible WMA (P = .0001) were the only cardiac MR predictors. ME rate in categories 1, 2, 3, and 4 was 2% (14 of 901 patients), 3% (six of 219 patients), 4% (15 of 409 patients), and 14% (26 of 193 patients), respectively (category 4 vs category 1, adjusted P < .001). Cardiac MR-directed revascularization was performed in 242 patients (14%) and reduced the risk of ME in only category 4 (7% [six of 92 patients] vs 26% [26 of 101 patients], P = .0004). CONCLUSION Dipyridamole cardiac MR imaging can be used to predict MEs in patients with ischemic chest pain. Patients with inducible WMA are at the highest risk for MEs and benefit the most from revascularization.