Introduction

Coronary artery disease (CAD) has a great impact in morbidity and mortality in the long term [1, 2]. Prompt diagnosis allows adequate management of these patients and improved prognosis. According to the most recent guidelines on chronic coronary syndromes, non-invasive detection of CAD with anatomical or functional testing is recommended for diagnosis and risk stratification in patients in whom clinical evaluation alone cannot rule out CAD [2]. In this context, stress perfusion cardiac magnetic resonance (CMR) has shown superior performance compared to other non-invasive tests [2,3,4,5].

Stress-CMR examinations are preferably performed under vasodilator drugs, such as adenosine or dipyridamole [6], which non-selectively target adenosine receptors A1, A2a, A2b, and A3 and cause adverse effects that may limit their use in patients at risk [7, 8]. Regadenoson is a more selective adenosine receptor agonist that preferentially binds to the A2a receptor, responsible for coronary vasodilation. Several studies have shown similar vasodilator effect to that of adenosine [9,10,11], but with fewer adverse events [12,13,14]. Although there are many data supporting the effectiveness of adenosine and dipyridamole in the context of stress perfusion CMR [6, 15,16,17,18,19,20,21], few studies have evaluated the use of regadenoson.

In this study, we sought to address the safety, feasibility, and hemodynamic response of regadenoson in unselected patients who underwent stress perfusion CMR examinations for clinical indication. We also evaluated the diagnostic accuracy of regadenoson stress perfusion CMR in our patient cohort.

Materials and methods

Study population

Between May 2017 and July 2020, 705 consecutive patients with known or suspected coronary artery disease underwent regadenoson stress perfusion CMR. Hemodynamically unstable individuals and patients with myocardial infarction within 24 h, glomerular filtration rate (GFR) < 30 mL/min/1.73 m2, or contraindications for regadenoson perfusion CMR were excluded. Patients were instructed to avoid methylxanthine containing substances 24 h prior to CMR examination [22, 23]. Baseline clinical characteristics were collected from electronic medical record data of our institution. Signed informed consent was obtained from all patients and the ethics committee for drug research approved the study protocol, which was performed in conformity with Royal decree 957/2020 and Declaration of Helsinki.

CMR protocol

CMR examinations were carried out on a 1.5 Tesla system (Magnetom Aera, Siemens Healthineers, Erlangen, Germany) using a conventional stress/rest perfusion protocol, including long and short axis steady state free precession (SSFP) cines, first-pass perfusion imaging under stress and rest conditions, and late gadolinium enhancement (LGE). First-pass stress myocardial perfusion was performed 70 s after the intravenous administration of regadenoson (Rapiscan, GE Healthcare AS) at a fixed dose of 0.4 mg (5 ml). The vasodilator effect of the drug was reverted with euphylline (200 mg i.v.) in all patients, regardless of the clinical symptoms immediately after first-pass stress myocardial perfusion images were acquired, which was approximately 150 s after the administration of regadenoson. A total dose of 0,15 mmol/Kg of gadobutrol (Gadovist, Bayer AG, Berlin, Germany) was administered at 4 ml/s [24].

CMR image analysis

CMR examinations were analyzed with specific software (cmr 42, Circle Cardiovascular Imaging Inc., Calgary, Canada). Endocardial and epicardial contours were traced in the end-diastolic and end-systolic images to calculate left ventricular volumes, function and mass [25]. The myocardial perfusion was visually assessed. Stress-induced perfusion defects were considered ischemic if the decreased signal intensity involved the subendocardium in a coronary artery territory distribution, the signal intensity was normal during rest perfusion, and the defects did not correspond to myocardial infarction on LGE images. Patients with a positive stress perfusion CMR examination were advised to undergo conventional coronary angiography. The final decision on how to proceed was made individually for each patient by the referring physician.

Assessment of clinical symptoms, adverse events, and hemodynamic response to regadenoson

Throughout the procedure, ECG tracing, blood pressure (BP) and heart rate (HR) were constantly monitored. All patients were systematically questioned about their symptoms before and after the administration of regadenoson and euphylline, and the predominant symptom was registered. Resting symptoms were asked just before regadenoson administration, while possible vasodilator-related symptoms were asked during its administration, immediately before first-pass stress myocardial perfusion imaging, and just before administration of euphylline. Clinical symptoms were also queried five minutes after euphylline administration to confirm that any symptoms caused by the vasodilator were reversed. In addition, adverse effects that could be related to induced stress, such as bronchospasm, arrhythmias, atrioventricular block, ventricular tachycardia, ventricular fibrillation, need for hospital admission, myocardial infarction or death were collected.

Hemodynamic response to regadenoson was determined by measuring changes in BP and HR under stress and rest conditions (HR response= [(stress HR– rest HR)/rest HR]*100; BP response= ([stress BP – rest BP]/rest BP)*100) [26]. Rest HR and BP data were collected before regadenoson administration. During stress, HR and BP data were registered before contrast administration, immediately after perfusion imaging acquisition and before euphylline injection, and 5 min after euphylline administration. Stress HR was defined as the highest HR during stress perfusion, whereas stress BP was defined as BP taken just after the actual perfusion scan and before the administration of euphylline.

Diagnostic performance

To establish the diagnostic performance of stress-CMR, sensitivity, specificity, positive and negative predictive values and accuracy were assessed in those patients who underwent invasive coronary angiography in less than one month since the CMR examination. Significant coronary artery obstruction was considered if the fractional flow reserve (FFR) value was < 0.80 or if direct stenting was performed.

Statistical analysis

Continuous data are described as mean ± standard deviation or as median [interquartile range (IQR)] and compared with the independent sample t-test or using the Mann–Whitney U test, as appropriate. Categorical variables are shown as percentages and compared with the Chi-square test. Sensitivity, specificity, positive and negative predictive values and accuracy of regadenoson stress perfusion CMR with respect to conventional coronary angiography were calculated. The statistical analysis was performed using SPSS (version 23.0 / SPSS Inc., Chicago, IL) and a p value < 0.05 was considered statistically significant.

Results

Study population

Seven of the initially included 705 patients were excluded due to technical problems for stress perfusion in three patients, missing clinical data in three patients, and lower back pain that impeded to complete CMR examination in one patient (Fig. 1).

Fig. 1
figure 1

Flowchart of included patients

Full size image

Therefore, a total of 698 patients were considered for the final analysis. The population consisted mainly of men (75.5%) with a median age of 66 years (IQR: 56–73) and a median body mass index (BMI) of 27.1 Kg/m2 (IQR: 24.5–30). Most individuals were in sinus rhythm (74.5%). Patient demographics, clinical characteristics, and indications for stress-CMR are shown in Table 1.

Table 1 Patient demographics, clinical characteristics, and indications for regadenoson stress perfusion CMR.
Full size table

Clinical symptoms and adverse events

Nearly half of patients (48.5%) remained completely asymptomatic after regadenoson administration. Most common clinical symptoms were dyspnea (19.6%) and chest pain (16.6%). These symptoms were mild, transient, and well tolerated (Fig. 2).

Fig. 2
figure 2

Frequency of symptoms induced by regadenoson

Full size image

Adverse events included transient stress-induced ectopies (1.7%), transient atrioventricular block (0.28%), bigeminy (0.14%), a limited episode of chest pain that required nitroglycerine administration (0.14%), and contrast extravasation (0.14%). Severe adverse events that prevented completion of the exam were rare (0.28%). One patient suffered regadenoson-induced symptomatic hypotension that required intravenous fluid therapy, whereas another patient referred chest pain that was treated conservatively. No cases of regadenoson-induced atrial fibrillation, ventricular tachycardia, ventricular fibrillation, need for hospital admission, myocardial infarction or death were observed (Table 2).

Table 2 Adverse events associated with regadenoson
Full size table

Hemodynamic response to regadenoson

Resting median HR was 63 bpm, (IQR: 37–127), median systolic BP was 152 mmHg (IQR: 135.75-165.25) and median diastolic BP was 76 mmHg (IQR: 68–83 mmHg). During the stress, the median HR was 87 bpm (IQR: 78–99), the median systolic BP was 144 mmHg (IQR: 128–157) and the median diastolic BP was 71 mmHg (IQR: 63–79 mmHg). Regadenoson induced an increase in HR response (median 36.2%, IQR: 22.5–50.9), and a decrease in systolic and diastolic BP (median systolic BP response of -5%, IQR: -11.5-0.6; median diastolic BP response of -6.3 mmHg, IQR: -13.4-0).

Patients with symptoms induced by regadenoson showed higher HR response (median 40.3%, IQR: 26.4–56.1) compared to individuals that remained asymptomatic (median 32.4%, IQR: 19-45.6) (p < 0.001). Conversely, blunted HR response was observed in obese (median 31.7%, IQR: 20.7–46.2 vs. median 37.2%, IQR: 23-53.7 in non-obese, p = 0.005), diabetic (median 29.6%, IQR: 18.4–42 vs. median 38.1%, IQR: 24.2–54.4 in non-diabetic, p < 0.001) and patients aged 70 years or older (median 32.9%, IQR: 22.6–43.1 in elderly vs. median 41.8%,IQR: 30.3–53.2 in non-elderly, p < 0.001). No statistically significant differences were observed in BP response. Patients under chronic treatment with beta-blockers did not show differences in the hemodynamic response compared with those untreated (Table 3).

Table 3 Hemodynamic response to regadenoson across subgroups of patients
Full size table

CMR findings

CMR findings are shown in Table 4. Mean LV ejection fraction was 66.3 ± 12.7%, mean indexed end-diastolic volume was 72.3 ± 23.3 ml/m2, and mean indexed end-systolic volume was 26.3 ± 19.3 ml/m2. More than half of individuals (54.7%) had normal left ventricular morphology. Almost two thirds of patients showed late gadolinium enhancement (30% with an ischemic pattern and 32% with a non-ischemic pattern).

Table 4 CMR results
Full size table

Diagnostic performance

In our cohort, the regadenoson stress perfusion CMR was positive in 199 patients (Fig. 3). Conventional coronary angiography was performed in 124 with a positive stress-CMR and in 24 patients with a negative stress-CMR examination but with persisting symptoms. The median time to coronary angiography from stress perfusion CMR was 2 days (IQR 1–6, 90th percentile 20.8). Sensitivity for stress CMR was 95.65% (IQ 91.49–99.81) and specificity 54.84% (IQ 35.71–73.97). The positive predictive value was 86.99% (IQ 82.74–94.68) whereas the negative predictive value was 77.27% (IQ 57.49–97.06). There were no statistically significant differences between diabetic and non-diabetic patients in terms of positive stress-CMR (87% vs. 82.4% p = 0.33) nor in the prevalence of significant coronary obstructions (86.7% vs. 73.5% p = 0.06). The diagnostic accuracy was similar (p = 0.07).

Fig. 3
figure 3

Stress CMR with regadenoson in a 78 year old male with history ofmultiple risk factors (former smoker, hypertension, diabetes mellitus) and percutaneous revascularization of iliofemoral axis stenosis, who presented episodes of chest pain in context of uncontrolled hypertension. (A) Stress perfusion. (B) Rest perfusion. (C) Coronary angiography. The test showed a perfusion defect in the basal, mid and apical inferoposterior segments (arrows in A), with normal perfusion in this segments at rest (image B). This patient underwent invasive coronary angiography that showed severe stenosis in the proximal segment of right coronary artery, and was treated with the implantation of a drug-eluting stent

Full size image

Discussion

The results of this study demonstrate that regadenoson can be used safely in stress CMR examinations. In a routine clinical setting, regadenoson stress perfusion CMR shows high diagnostic performance, comparable to that obtained with other vasodilators.

Stress CMR has many potential advantages over other non-invasive ischemia detection tests. It has a high sensitivity and specificity for diagnosing CAD [3, 5, 19, 27], the technique is the gold standard for evaluating the morphology and function of the heart, does not require ionizing radiation, and the obtained image quality is not influenced by factors such as poor acoustic window. Stress CMR has traditionally been performed with adenosine. The administration of this drug presents, however, some limitations, including the need of an MRI-compatible infusion pump, patient weight based dosage calculation, and the relative contraindications in certain subgroup of patients, such as those with severe respiratory disorders (asthma, chronic obstructive pulmonary disease). Regadenoson may help overcome most of these limitations [12,13,14, 28, 29].

Several publications have emphasized the safety of regadenoson in nuclear medicine perfusion examinations [8, 30, 31] but studies evaluating the safety profile of regadenoson in CMR are scarce [28]. All the series agree that regadenoson presents fewer complications and better tolerability than adenosine. However, the incidence of symptoms related to the administration of regadenoson varies between the publications. For example, in our study we observed a lower incidence of minor symptoms compared to studies that used nuclear medicine imaging techniques [8, 9] but higher than that reported, for example, by a recent CMR study [29]. Rather than the imaging techniques that were employed, we believe that a more plausible explanation for this finding is the way in which symptoms were reported and collected. We decided to systematically question all patients about any possible regadenoson-induced symptoms at many different points in the study, including before and after euphylline administration, and any side effects related by the patient was thoroughly registered. We also consider that the systematic use of euphylline in all patients may have contributed to better tolerability of the vasodilator. In line with the study by Monmeneu Menadas et al. [29] in our cohort patients with asthma or COPD (n = 96) presented a similar safety profile as the general population, showing no significant adverse events. This observation highlights the safety and tolerability of regadenoson in patients with chronic pulmonary disease. In our cohort, two patients suffered severe events that led to premature test ending. One individual was a 49-year-old male with history of CAD, who referred unbearable chest pain after regadenoson administration. The ECG did not show changes suggesting myocardial ischemia. Prompt euphylline infusion relieved the symptoms. The other patient was an 82-year-old obese male, with systemic arterial hypertension and dyslipidemia under treatment and no history of CAD who was referred for stress CMR for chest pain. Patient’s baseline BP was 133/83 mmHg (HR 55 bpm), and after regadenoson administration it dropped to 65/47 mmHg (HR 98 bpm), presenting as pre-syncope that required rapid euphylline and intravenous fluid administration. No myocardial ischemia was detected in the perfusion exam. In line with other publications, no life-threatening events, hospital admission or death occurred after regadenoson administration.

The significant increase in HR is a distinctive feature that reflects the hemodynamic effect of regadenoson. This vasodilator acts on the sympathetic nervous system through baroreflex-mediated activation and through direct activation of the A2a receptor [28]. In our cohort, we observed blunted HR response in those individuals known to have blunted sympathetic response, including the elderly, obese, and diabetic patients. This finding does not appear to affect test accuracy [26, 30,31,32,33], and has been proven to be an independent predictor for poor outcomes in previous studies [33, 34]. Interestingly, patients on chronic beta-blocker treatment did not show a different hemodynamic response to regadenoson, a fact that reassures the performance of stress CMR in the outpatient setting, where medication restriction may not be easy.

All patients received euphylline after stress perfusion despite their clinical symptoms to minimize drug side effects and to reverse regadenoson-induced hyperemia [35]. Being the half-life of regadenoson relatively long as compared with adenosine, concern about residual myocardial hyperemia during the rest perfusion and its impact on the diagnostic accuracy of stress/rest perfusion CMR protocols has been raised. According to our results, however, this fact does not appear to influence the diagnostic performance of the test. Our diagnostic accuracy values are very similar to those reported for stress perfusion CMR. A meta-analysis that compared different cardiac imaging methods against fractional flow reserve (FFR) as the gold-standard to detect lesion-specific ischemia, showed sensitivity of 90% (95% CI, 75–97) and specificity of 85% (95% CI, 79–89%) for stress perfusion CMR [5].

Our study has several limitations. The sample size is smaller than that included by other groups [28,29,30] and data were retrospectively collected. However, it highlights the routine real-life clinical experience of using regadenoson in unselected patients and adds reassurance to the safety and feasibility of using this vasodilator as stressor in perfusion CMR examinations. The hemodynamic effect of the drug was assessed based on the HR and BP response, without studying hyperemia at the myocardial level, which would have provided a more objective assessment. This requires the use of specific quantitative CMR perfusion sequences that are under active research. In their work, Vasu et al. observed that adenosine and regadenoson have similar vasodilator potency (2.04 ± 0.34 ml/min/g vs. 2.12 ± 0.27 ml/min/g) [10]. Lastly, the number of patients who underwent conventional coronary angiography to confirm CMR findings was low and may limit the interpretation of the diagnostic efficacy of the test. Our results, however, are in line with those reported in the literature.

In conclusion, regadenoson is a safe and well-tolerated vasodilator drug for stress CMR. Adverse reactions are very few and drug-induced clinical symptoms are mild, transient and well tolerated. The hemodynamic response consists of significant HR increase and mild hypotension, which appear to be blunted in those patients with diminished sympathetic response, such as the elderly, obese and diabetic individuals. The diagnostic performance of regadenoson stress perfusion CMR is very similar to that achieved with other vasodilators. Further research is warranted to evaluate the safety and tolerability of regadenoson and its diagnostic performance in specific subgroup of patients.