Cardiac Magnetic Resonance Imaging Versus Computed Tomography to Guide Transcatheter Aortic Valve Replacement: A Randomized, Open-Label, Noninferiority Trial

Abstract

BACKGROUND: Computed tomography (CT) is recommended for guiding transcatheter aortic valve replacement (TAVR). However, a sizable proportion of TAVR candidates have chronic kidney disease, in whom the use of iodinated contrast media is a limitation. Cardiac magnetic resonance imaging (CMR) is a promising alternative, but randomized data comparing the effectiveness of CMR-guided versus CT-guided TAVR are lacking. METHODS: An investigator-initiated, prospective, randomized, open-label, noninferiority trial was conducted at 2 Austrian heart centers. Patients evaluated for TAVR according to the inclusion criteria (severe symptomatic aortic stenosis) and exclusion criteria (contraindication to CMR, CT, or TAVR, a life expectancy <1 year, or chronic kidney disease level 4 or 5) were randomized (1:1) to undergo CMR or CT guiding. The primary outcome was defined according to the Valve Academic Research Consortium–2 definition of implantation success at discharge, including absence of procedural mortality, correct positioning of a single prosthetic valve, and proper prosthetic valve performance. Noninferiority was assessed using a hybrid modified intention-to-treat/per-protocol approach on the basis of an absolute risk difference margin of 9%. RESULTS: Between September 11, 2017, and December 16, 2022, 380 candidates for TAVR were randomized to CMR-guided (191 patients) or CT-guided (189 patients) TAVR planning. Of these, 138 patients (72.3%) in the CMR-guided group and 129 patients (68.3%) in the CT-guided group eventually underwent TAVR (modified intention-to-treat cohort). Of these 267, 19 patients had protocol deviations, resulting in a per-protocol cohort of 248 patients (121 CMR-guided, 127 CT-guided). In the modified intention-to-treat cohort, implantation success was achieved in 129 patients (93.5%) in the CMR group and in 117 patients (90.7%) in the CT group (between-group difference, 2.8% [90% CI, −2.7% to 8.2%]; P <0.01 for noninferiority). In the per-protocol cohort (n=248), the between-group difference was 2.0% (90% CI, −3.8% to 7.8%; P <0.01 for noninferiority). CONCLUSIONS: CMR-guided TAVR was noninferior to CT-guided TAVR in terms of device implantation success. CMR can therefore be considered as an alternative for TAVR planning. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT03831087.