Study population and design
Potential patients with AS underwent CMR at the Department of Cardiology, Angiology and Pneumology of the Heidelberg University Hospital between January 2009 and December 2021 and were retrospectively identified from our local clinical database. Patients underwent CMR for exclusion of significant CAD. The study was approved by the local institutional ethics committee in accordance with the Declaration of Helsinki (S-154/2015). All patients had undergone stress CMR with either adenosine or dobutamine and two-dimensional echocardiography Doppler study within 12 months of each other and had evidence of AS by echocardiography with aortic valve area (AVA) ≤1.5 cm2. Patients with prior transcatheter aortic valve replacement (TAVR), prior surgical aortic valve replacement were excluded (Figure 1).
Cardiovascular risk factors (arterial hypertension, hypercholesterolemia, diabetes mellitus, obesity, history of smoking and family history of cardiovascular disease) and comorbidities (history of CAD, prior myocardial infarction, prior percutaneous coronary intervention, prior coronary artery bypass grafting (CABG), prior stroke or transitory ischaemic attack, chronic kidney disease and chronic obstructive pulmonary disease (COPD) were assessed using medical reports.
We recorded major complications such as death, myocardial infarction, severe arrhythmias, unstable angina, stroke and new hospital admission as well as minor complications such as induction of supraventricular tachycardia, drop of blood pressure (<60mmHg) and other non-severe complications that did not require a new hospital admission. We also documented complaints, such as chest pain, nausea and emesis or dyspnea that disappeared during CMR or shortly after administration of beta-blockers or sublingual nitroglycerin.
Selection of controls
Age-, sex-, body-mass-index (BMI) matched individuals without AS (controls), who underwent stress CMR with either adenosine or dobutamine, were selected from our CMR database. The same number of controls undergoing adenosine or dobutamine stress CMR were randomly selected from our CMR database.
Study patients underwent conventional transthoracic echocardiographic studies, digitally stored on PACS (Picture Archiving and Communication System) and offline available on workstations (Centricity, GE Healthcare Vingmed, Trondheim, Norway). Echocardiography examinations were performed on commercially available ultrasound systems (Vivid 7, GE Healthcare Vingmed, Trondheim, Norway; ie33, Philips, Eindhoven, The Netherlands and SSD-5500 PureHD, Hitachi Aloka Medical, Tokyo, Japan) according to the guidelines of the American Society of Echocardiography  including standard 2D echocardiography for anatomic imaging and continuous wave and color flow Doppler for the analysis of the Aortic valve gradients. Aortic valve gradients were calculated using continuous wave Doppler signals that afforded the highest peak velocity across the aortic valve using the continuity equation as previously described in the guidelines from the European Association of Cardiovascular Imaging and the American Society of Echocardiography form 2017 . The examinations were analyzed by experienced readers.
Cardiac Magnetic Resonance Imaging
Standard CMR was performed supine in a 1.5T Achieva™, 1.5T Ingenia™ (1.5T) or 3T Ingenia™ (3T) whole body scanner (Philips Healthcare, Best, The Netherlands), with a commercial cardiac phased array receiver coil as previously described [18, 19]. Cine long axis 2-, 3- and 4-chamber views as well as short axis cine images covering the whole left ventricle (LV) from the annulus of the atrioventricular valves to the apex (8 mm slice thickness) were obtained using a breath-hold, segmented-k-space balanced steady-state free precession sequence (bSSFP) employing retrospective electrocardiogram or pulse oximetric gating with 40 (1.5T Achieva™) or 35 (1.5T and 3T) phases per cardiac cycles. Scan parameters were: repetition time (TR) 2.8ms; echo time (TE) 1.4ms; flip angle (FA) 60° (1.5T Achieva™ and 1.5T), TR 2.9ms, TE 1.4ms, FA 45° (3T); with a breath-hold time of 7–10 s per image and prospective gating. Data were analyzed using commercially available workstations (Viewforum™ and IntelliSpace™ Portal, ISP™, Philips Healthcare, Best, The Netherlands) and a certified software (cmr42 Version 5.6.6, Circle Cardiovascular Imaging Inc., Calgary, Canada) as semi-automatic software for volumetric analysis. Ventricular volumes, ejection fraction of LV and right ventricle (RV) and LV myocardial mass were acquired in short axis stacks by manually tracing epi- and endocardial borders, excluding papillary muscles from the myocardium.
Adenosine Perfusion Stress CMR
A three-slice turbo field gradient echo-echo-planar imaging (GRE-EPI) sequence was used as described previously . Stress perfusion imaging was performed using a continuous intravenous infusion of 140μg/kg body weight/min (optional 210μg/kg body weight/min, in case of an inadequate heart rate response or recent caffeine intake) for three minutes over an antecubital vein. Three heartbeats after initiation of the sequence a bolus of gadolinium diethylenetriamine pentaacetic acid /DTPA (Magnevist™, Schering, Berlin, Germany) 0.2mmol/kg body weight (before February 2016) or Gadobutol (Gadovist™, Bayer HealthCare, Leverkusen, Germany) 0.14mmol/kg body weight (1.5T) or 0.1mmol/kg body weight (3T) (after February 2016) was injected over a separate peripheral venous catheter at a rate of 5 ml/s flushed with 20 ml 0.9% saline solution. Semi-quantification of myocardial perfusion was conducted in three LV short‐axis slices using IntelliSpace™ Portal, ISP™ (Philips Medical Systems, Best, The Netherlands). The adenosine perfusion stress CMR protocol was the same for all three vendors.
Dobutamine stress CMR
Dobutamine stress CMR was performed as previously described [21, 22]. A 4, 2, and 3-chamber and three short-axis views (apical, mid-ventricular, and basal) were used. Dobutamine was infused during 3-min stages at incremental doses of 10, 20, 30, and 40 g/kg of body weight/min until at least 85% of the age-predicted heart rate was reached (220-age in years). Atropine was administered in 0.25mg increments (up to maximal dose of 2.0mg) if the target heart rate was not achieved. Cine loops were viewed online as they were acquired. Stress testing was stopped when the target heart rate was achieved or when one of the following occurred: severe chest pain or dyspnea, decrease in systolic blood pressure of >40mmHg, hypertension of >220/120mmHg, severe arrhythmias, new or worsening wall motion abnormalities (WMA) in at least 1 segment. Failure to attain 85% of age-predicted maximal heart rate was considered as a nondiagnostic result. Electrocardiographic rhythm, symptoms, peripheral blood pressure and oxygen saturation were continuously monitored during the stress studies. Inclusion criteria for using dobutamine stress CMR were the following; patients with 3-vessel CAD or after CABG, which is specific for the study center or if contraindications for the use of adenosine were known (known or suspected bronchoconstrictive or bronchospastic disease) . The dobutamine stress CMR protocol was the same for all three vendors.
Statistical analysis was performed using MedCalc™, version 15.7 (Ostend, Belgium), with p<0.05 taken to indicate statistical significance for all statistical tests. Continuous and normally distributed variables (Kolmogorov-Smirnov test, p≥0.05) were expressed as mean ± standard deviation. Group differences for continuous variables were tested using the independent t-test. Continuous variables without normal distribution were stated as median and interquartile range (IQR), group differences were tested using the nonparametric Mann-Whitney U test. Categorical variables were compared using chi-squared test. Correlation analysis for the occurrence of complications was performed using Spearman's rank correlation. Univariable logistic regression models were used to assess the association between each variable and the occurrence of complications. For multivariable modeling, we included 4 and 5 of the most significant variables. Results are reported as odds ratio with 95% confidence intervals (CI).