Gadobutrol-Enhanced Cardiac Magnetic Resonance Imaging for Detection of Coronary Artery Disease

ISS Fro Be An ma un Un Au Gle Au U Un Ge Un by Na Cir BACKGROUND Gadolinium-based contrast agents were not approved in the United States for detecting coronary artery disease (CAD) prior to the current studies. OBJECTIVES The purpose of this study was to determine the sensitivity and specificity of gadobutrol for detection of CAD by assessing myocardial perfusion and late gadolinium enhancement (LGE) imaging. METHODS Two international, single-vendor, phase 3 clinical trials of near identical design, “GadaCAD1” and “GadaCAD2,” were performed. Cardiovascular magnetic resonance (CMR) included gadobutrol-enhanced first-pass vasodilator stress and rest perfusion followed by LGE imaging. CAD was defined by quantitative coronary angiography (QCA) but computed tomography coronary angiography could exclude significant CAD. RESULTS Because the design and results for GadaCAD1 (n 1⁄4 376) and GadaCAD2 (n 1⁄4 388) were very similar, results were summarized as a fixed-effect meta-analysis (n 1⁄4 764). The prevalence of CAD was 27.8% defined by a $70% QCA stenosis. For detection of a $70% QCA stenosis, the sensitivity of CMR was 78.9%, specificity was 86.8%, and area under the curve was 0.871. The sensitivity and specificity for multivessel CAD was 87.4% and 73.0%. For detection of a 50% QCA stenosis, sensitivity was 64.6% and specificity was 86.6%. The optimal threshold for detecting CAD was a $67% QCA stenosis in GadaCAD1 and $63% QCA stenosis in GadaCAD2. CONCLUSIONS Vasodilator stress and rest myocardial perfusion CMR and LGE imaging had high diagnostic accuracy for CAD in 2 phase 3 clinical trials. These findings supported the U.S. Food and Drug Administration approval of gadobutrolenhanced CMR (0.1 mmol/kg) to assess myocardial perfusion and LGE in adult patients with known or suspected CAD. (J Am Coll Cardiol 2020;76:1536–47) Published by Elsevier on behalf of the American College of Cardiology Foundation. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). N 0735-1097 https://doi.org/10.1016/j.jacc.2020.07.060 m the National Heart, Lung, and Blood Institute, National Institutes of Health, DHHS, Bethesda, Maryland; Helios Klinikum rlin Buch Klinik für Kardiologie und Nephrologie Abteilung Kardio-MRT, Berlin, Germany; Cedar-Sinai Medical Center, Los geles, California; Robert-Bosch-Krankenhaus Zentrum für Innere Medizin (ZIM) III Abteilung für Kardiologie, Stuttgart, Gerny; Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania; Herzzentrum Leipzig Abteilung für Diagnostische d Interventionelle Radiologie, Leipzig, Germany; Royal Perth Hospital, Perth, Western Australia, Australia; Washington iversity School of Medicine, St. Louis, Missouri; Flinders University, Flinders Medical Centre, Adelaide, South Australia, stralia; Department of Cardiovascular Sciences University of Leicester and the NIHR Leicester Biomedical Research Centre, nfield Hospital, Leicester, United Kingdom; The Prince Charles Hospital Cardiology Research Centre, Brisbane, Queensland, stralia; Medical University of South Carolina, Charleston, South Carolina; National Heart Centre Singapore, Singapore; niversity of Virginia Health System, Charlottesville, Virginia; Departments of Medicine and Diagnostic Radiology, McGill iversity Health Centre, Montreal, Quebec, Canada; Bayer Pharmaceuticals LLC, Whippany, New Jersey; Bayer AG, Berlin, rmany; Cardiovascular Research Centre and CMR Unit at Royal Brompton and Harefield NHS Foundation Trust, London, ited Kingdom; and the National Heart and Lung Institute, Imperial College, London, United Kingdom. Funding was provided Bayer AG, Siemens Healthineers, and in part by the Division of Intramural Research, National Heart, Lung, and Blood Institute, tional Institutes of Health. Dr. Arai has a Cooperative Research and Development Agreement (CRADA) with Bayer, Siemens, and cle CVI Inc.; and has patents and invention reports related to perfusion quantification and cine MRI. Dr. Schulz-Menger has J A C C V O L . 7 6 , N O . 1 3 , 2 0 2 0 Arai et al. S E P T E M B E R 2 9 , 2 0 2 0 : 1 5 3 6 – 4 7 The GadaCAD1 and GadaCAD2 Clinical Trials 1537 AB BR E V I A T I O N S

C MR is a reference standard for assessing ventricular function (1) and for imaging myocardial infarction (MI) (2). In meta-analysis, stress perfusion CMR performs with high diagnostic accuracy, particularly when compared with invasive fractional flow reserve (FFR) (3). The  sion Assessment in Coronary artery disease Trial II) (7) showed that stress perfusion CMR has good diagnostic performance and is superior or not inferior to single-photon emission computed tomography (SPECT) (5,6,8). Stress perfusion CMR and LGE imaging appear in multiple U.S. and international guidelines (9)(10)(11). Despite over 25  stability and high relaxivity (12,13).
The specific aim of the GadaCAD clinical trials was to assess the diagnostic accuracy of gadobutrolenhanced vasodilator stress perfusion CMR and LGE imaging to detect CAD in 2 nearly identical studies using an independent blinded read. The clinical trials had requirements from the FDA to meet or exceed specific diagnostic accuracy criteria for sensitivity and specificity. Gadobutrol-enhanced perfusion CMR had to have higher sensitivity than unenhanced stress cine CMR wall motion for CAD detection. The standard of reference defining CAD was invasive coronary angiography, but coronary computed tomography angiography (CTA) could be used to exclude CAD. Manuscript received May 20, 2020; revised manuscript received July 20, 2020, accepted July 29, 2020.

CMR Core Laboratory
GadaCAD1 and GadaCAD2 had very similar designs, standards of reference, and core laboratories. There were 3 blinded cardiac magnetic resonance (CMR) readers for each study, and thus a total of 6 readers. CAD ¼ coronary artery disease; GadaCAD ¼ Gadobutrol-enhanced CMR to detect Coronary Artery Disease; LGE ¼ late gadolinium enhancement; QCA ¼ quantitative coronary angiography.
Arai et al. LGE was normal, the perfusion finding was categorized as an artifact and the study as normal.
I m a g i n g a n d v a s o d i l a t o r s t r e s s m e t h o d s . The study-specific procedures are summarized in Figure 1 and     GadaCAD2, respectively, while suitability of coronary angiography or CTA led to exclusion of 28 (6.6%) and 45 (9.4%) subjects, respectively (Supplemental Table 2). The demographic characteristics of the study participants are summarized in Table 2.  The meta-analysis ( Overall, the combined results represented the sensitivity and specificity of the individual readers quite well ( Figure 3). In general, the sensitivity for detecting a $70% QCA stenosis was slightly higher in GadaCAD1 (81.5%) than GadaCAD2 (77.1%) as summarized by the majority read (  Table 4). With significant CAD defined by a 70% QCA stenosis, all 6 readers met every endpoint for sensitivity, specificity, and the comparison with stress cine wall motion (Supplemental Table 4). When defining CAD at a 50% QCA stenosis, 5 of 6 readers did not meet or exceed the lower limit of the sensitivity endpoint but all readers met all other study endpoints ( Table 5).
ADVERSE EVENTS. The great majority of adverse events were stressor-related ( Table 2). There were no deaths. Of the 4 adverse events related to gadobutrol, only 1 was considered serious: an anaphylactic reaction. Values are n, n (%), or n/N (%). *Total number is the number of subjects with $50% QCA stenosis.

FIGURE 2 Example of Image Quality
This patient had multivessel coronary stenoses and no clinically recognized prior myocardial infarction (MI), a 95% right coronary artery stenosis, 70% diagonal stenosis, and a 50% obtuse marginal stenosis. Nonmotion-corrected stress perfusion images (bottom) show an obvious inferior and inferolateral perfusion defect (green arrows) and a second less severe anterolateral perfusion defect (red arrows). The perfusion defects were more extensive than the small subendocardial MI detected with LGE imaging (middle). See Videos 1, 2, 3, 4, 5, and 6, including the cine CMR (top), which showed a subtle inferior wall motion abnormality and perfusion images.
Abbreviations as in Figure 1.   Table 3.

Meta-Analysis of GadaCAD1 & GadaCAD2
Including All 6 Reader Results

1.00
The green line plots the receiver-operator characteristic curve with shading surrounding the line representing the 95% confidence interval (CI). Solid red triangles plot the results for the 3 readers from GadaCAD1 versus a 70% QCA stenosis. Solid blue circles plot the results for the 3 readers from GadaCAD1 versus a 70% QCA stenosis. The open symbols plot results for the 2 clinical trials versus a 50% QCA stenosis using the same color scheme. Abbreviations as in Figure 1.  The main findings in this patient were diffuse and severe perfusion defects during stress that were not present during rest perfusion imaging and the absence of a myocardial infarction. The stress perfusion defect appears as a dark band about 50% to 75% in transmural extent in all segments with a thin band of brighter gray near the epicardium (red arrows). Rest perfusion appears normal as documented by uniformly gray enhancement (green arrows). Both sets of late gadolinium enhancement (LGE) images showed no myocardial infarction as the myocardium appears relatively uniformly black myocardium (cyan and yellow arrows). Cine short axis cardiac magnetic resonance (CMR) in multiple, contiguous short-axis imaging planes provided measurements of left ventricular volumes, ejection fraction, and mass. Lowerresolution real-time cine CMR were obtained before and during vasodilator stress to assess for induced regional wall motion abnormalities (RWMA), which cannot be assessed on static images but were not seen on review of the video versions. Long-axis cine CMR confirmed the global and regional left ventricular function.
Arai et al. The FDA approval of gadobutrol-enhanced CMR has important clinical implications (17,22). Multiple observational registries have demonstrated an association of revascularization with and survival benefit in patients with extensive ischemia SPECT. Although the randomized, controlled ISCHEMIA trial (23) did not find a significant difference in hard outcomes in patients with moderate or severe ischemia assigned to an invasive strategy plus optimal medical therapy (OMT) versus OMT alone, improved health status outcomes in the invasive arm were observed (17,22,24). Stress tests are recognized by all major guidelines, and will continue to be used to diagnose and manage CAD.
The published data on FFR suggest that managing patients based on anatomic coronary artery stenoses severity leads to higher rates of revascularization and no benefit or worse outcomes compared with managing patients based on physiological significance of the stenosis by FFR (25)(26)(27)(28)(29) However, the 70% stenosis threshold is widely used was not designed to compare these agents.
Despite methods aimed to obtain identical slices for stress and rest perfusion, matching was not always perfect (Figure 2), but LGE images can also be used to help interpret stress perfusion images (46).

CONCLUSIONS
Gadobutrol-enhanced CMR has high diagnostic accuracy for detecting CAD and is now FDA approved at a dose of 0.1 mmol/kg to assess myocardial perfusion and LGE in adults with known or suspected CAD.