Background

Currently, non-invasive cardiac imaging tests are first-line diagnostic tools in symptomatic patients with known or suspected ischaemic heart disease [1]. Vasodilator stress cardiovascular magnetic resonance (CMR) has shown high diagnostic accuracy for detection of myocardial ischaemia and in the prognostic evaluation of these patients [2,3,4,5,6]. Vasodilators such as adenosine, which has a direct action on adenosine receptors, and dipyridamole, an inhibitor of the cellular reuptake of adenosine, have been used. More recently, regadenoson has been approved for this purpose [7]. Regadenoson has several advantages over other vasodilators, it is administered intravenously as a single dose, with no need for dose adjustment with respect to weight, administration is done in 10 s, thus saving time compared to other stressors, and it is a selective agonist of A2A receptors. This is particularly important in patients with asthma or chronic obstructive pulmonary disease (COPD), in which adenosine and dipyridamole are contraindicated or must be used with caution [8, 9].

Regadenoson safety has been evaluated in nuclear stress studies, with favorable results compared to adenosine [10,11,12], and also in stress-CMR [13,14,15]. Thus, the aim of this study was to provide additional insight into the feasibility and safety of regadenoson stress-CMR in comparison with adenosine in a large referral population.

Methods

Study population

Between January 2016 and July 2019, a total of 6184 consecutive patients from 13 hospitals and 7 outpatient clinics within the Valencian Community area in Spain, were scanned with CMR. Of them, 2970 subjects were referred to stress-CMR with a clinical indication. Regadenoson was initially used only in patients with COPD or asthma of any degree of severity in which adenosine was contraindicated. Later, in several hospitals, its use was approved in all patients as a first-line vasodilator. Thus, finally 2285 subjects were scheduled for regadenoson and 685 for adenosine stress-CMR. Interestingly, 99.5% of stress-CMRs were performed in a hospital environment (four University Hospitals and three Community Hospitals) and only 0.5% in an outpatient CMR unit. All scans were performed on 1.5 T scanners (Sonata, Avanto, Symphony and Aera, Siemens, Erlangen, Germany; Signa HDxt, GE Healthcare, Waukesha, Wisconsin, US). All the CMR scans were performed and interpreted by five cardiologists with average 11 years of experience in stress-CMR. All the CMR units had adequate equipment, drugs, including bronchodilator therapy, and resuscitative measures to treat any potential stress-related cardiovascular complications.

Study protocol

The indications, contraindications, diagnostic criteria regarding test results, and reasons for termination of the tests were in accordance with the current recommendations [16]. All patients were required to be clinically stable for the scan to be performed. Absolute contraindications for regadenoson and adenosine stress testing included: (1) second or third-degree atrioventricular block or sinus node dysfunction without a functioning pacemaker, (2) known hypersensitivity (allergy) to adenosine or regadenoson, and (3) systolic blood pressure less than 90 mmHg. Importantly, there was no restriction for regadenoson stress in patients with moderate or severe bronchoconstrictive lung disease such as asthma or COPD.

Patients were previously instructed to avoid caffeine, theophylline and dipyridamole at least 24 h before the CMR scan. A resting 12 lead electrocardiogram was performed outside the scan room immediately before the CMR study in all cases. Heart rate (HR) and blood pressure (BP) were continuously monitored and a patient intercommunication system was available throughout the scan.

The CMR protocol included: (a) Rest study with acquisition of black blood HASTE sequences and steady-state free precession (SSFP) cine sequences in the usual views for quantification of ventricular dimensions and function. (b) Stress study which included administration of intravenous regadenoson (Rapiscan®, GE Healthcare Bio-Sciences, S.A, Madrid, Spain) at a fixed dose of 400 μg in 10 s, or adenosine (Adenocor®, Sanofi-Aventis S.A, Barcelona, Spain) at a dose of 140 μg/kg body weight/min, (with increase up to 210 μg/kg body weight/min if after 2–3 min HR does not increase by 10 bpm and or BP does not drop by > 10 mmHg) followed 60–90 s later by infusion of 0.15 mmol/kg gadoteric acid (Dotarem®, Guerbet, France) at a flow of 5 mL/s during simultaneous acquisition of first pass saturation-recovery sequences in one long-axis and three ventricular short-axis views. Images were sampled every RR in most cases and only if the fast HR avoids it, they were sampled every second RR. In all the sites the myocardial perfusion sequence was similar, with TR = 135–160 ms, TE = 1 ms, flip angle = 12 degree; acceleration factor = 2; voxel size ≤ 2.8 × 2.2 mm; slice thickness 8 mm and receiver bandwidth 790 Hz/pixel. This was followed by acquisition of stress cine sequences in two long-axis and three ventricular short-axis views. Finally, aminophylline 200 mg was administered intravenously if regadenoson was used. We did this to accelerate the disappearance of symptoms and promote rapid recovery of heart rate before acquiring late gadolinium enhancement (LGE) images and rest perfusion if necessary. (c) Segmented inversion recovery SSFP sequences were acquired at least 10 min after gadolinium injection, for detection of myocardial LGE. (d) Whenever considered necessary, rest perfusion sequences were acquired to rule out artifacts in stress perfusion images. Reasons for termination of the stress test were: presence of severe symptoms or side effects, technical reasons and patient refusal. Subsequently, the images were visualized and studies with artifacts or poor image quality that prevented adequate image analysis were excluded from the analysis.

Ventricular dimensions and function were eventually quantified with the Simpson´s method (Medis, The Netherlands). Stress myocardial perfusion, regional contractility and LGE were assessed by visual analysis.

Study variables

Relevant clinical and demographic data were prospectively collected in a database along with previous stress tests undergone, need for pharmacologic sedation, history of kidney failure. Variables obtained from the CMR scan included cardiac rhythm, left (LV) and right ventricular (RV) dimensions and function, presence and extent of myocardial ischemia, necrosis, and focal fibrosis, type of contrast used and contrast dose. A perfusion defect was considered ischaemic if ≥ 2 consecutive segments in at least three consecutive phases were affected, as previously described [17]. Inducible wall motion abnormalities are considered as a criterion of severity in cases with perfusion defects, not for diagnosis of ischaemia. HR and BP were measured at baseline, peak of stress, and then monitored every 2 min until the end of the test [18]. The increase in HR during stress was calculated as a percentage ([Peak stress HR−Resting HR/Resting HR] × 100). The change in BP was calculated accordingly.

In order to classify complications, we have established the severity criteria based on the absence of improvement in the CMR unit and the need for hospitalization. Thus, we considered severe complications death, myocardial infarction, unstable angina, severe or persistent arrhythmias or arrhythmias requiring continuous monitoring, second degree atrioventricular block, significant dyspnoea or severe chest pain, severe bronchospasm requiring bronchodilators and hospital admission, and new hospital admissions by any other cause. Non-severe complications were defined as those moderate in intensity, responding to aminophylline or specific management in the CMR unit and not requiring hospital admission. We also recorded minor symptoms as those mild in intensity, that disappeared immediately or shortly after administration of aminophylline.

After completing the study, we performed a subjective assessment of the overall quality of the study. Studies of optimal quality were those without artefacts of any type and in which contractility, first-pass perfusion, and LGE could be analyzed reliably. Studies with some artefact that did not prevent a correct interpretation of images were considered of suboptimal quality. Finally, those with artefacts that prevented reliable interpretation of images were considered of poor quality. Likewise, stress test was considered non diagnostic only when the image quality was poor.

Statistical analysis

Statistical analyses were carried out with SPSS 17 for windows (SPSS Inc., Chicago, IL, USA). Quantitative variables were presented as absolute values and percentages. Continuous variables were expressed as mean ± standard deviation and categorical variables as proportions. Two-tailed unpaired Student’s T-test was used for intergroup comparisons and χ2 test for comparisons of proportions. A p value < 0.05 was considered statistically significant.

Results

Feasibility and baseline characteristics

Between January 2016 and July 2019, consecutive patients were selected to perform stress-CMR (Fig. 1). Finally, regadenoson stress-CMR was performed in 2253 patients (98.6% of requested) and adenosine stress-CMR in 655 patients (97.5% of requested). Importantly, 1948 patients (86.5%) underwent regadenoson stress as first line vasodilator according to their hospital policy. The reasons have been mainly based on its ease of use and time savings. In the remaining 305 patients (13.5%) regadenoson was used since contraindication to adenosine was present due to COPD or asthma by clinical criteria. None of the patients referred to our unit reported a history of hypersensitivity to adenosine or regadenoson.

Fig. 1
figure 1

Flow diagram showing patient selection for stress-CMR in this study. After excluding patients with general contraindications to CMR and with contraindications to pharmacologic stress we performed stress-CMR in 2908 patients

Full size image

The clinical characteristics of the study population are shown in Table 1. It was a population with an intermediate prevalence of risk factors, in which approximately a quarter or a fifth of patients had a known coronary artery disease (CAD). There were no significant differences in the main clinical variables between the regadenoson and adenosine subgroups, except in the prevalence of COPD or asthma (logically). In 24.8% of patients, a previous exercise tolerance test was available that was non-conclusive in 465 (64%). Sinus rhythm was reported in 89.8% of patients. There were 31 patients (1.1%) who had one or more previous stress-CMR. Also, 32% of subjects were inpatients at the time of the study. Mild sedation was administered with midazolam (1–3 mg intravenous) in 123 patients (4.2%).

Table 1 Clinical characteristics of the study population
Full size table

Study quality was optimal in 2612 cases (90.0%), suboptimal in 276 cases (9.5%), and poor in 20 cases (0.5%). Suboptimal or poor-quality studies were associated with older age of the patients (72 ± 10 vs. 67 ± 12 years, p < 0.0001), history of COPD or asthma (23.2 vs. 12.3%, p < 0.0001), history of heart failure (13.7 vs. 7.5%, P = 0.008), in-patients (39.1 vs. 28.8%, p = 0.025), presence of rhythm other than sinus (49.8 vs. 7.1%, P < 0.0001), lower LV ejection fraction (57 ± 15 vs. 63 ± 14%, p < 0.0001), lower RV ejection fraction (58 ± 12 vs. 63 ± 8%, p < 0.0001), and administration of midazolam (9.6 vs. 3.9%, p = 0.002). All these factors can be associated with each other and influence a worse performance of apneas by patients, e.g. patients with right ventricular systolic dysfunction were associated with a higher presence of COPD or asthma and worse LVEF. We must emphasize that, although in patients with non-sinus rhythm (most of them in atrial fibrillation) the quality of the studies was significantly worse, only in one case was not possible to obtain diagnostic information about the function, presence of ischaemia or necrosis. There were no differences with respect to gender, vasodilator used, CMR unit in which the test was done or other stress-CMR parameters.

Stress-CMR data

Stress was performed in 2908 patients (97.9% of those requested). In a sample of 20 studies with each drug the average duration of the tests was 34 min with regadenoson compared to 42 min with adenosine. The stress tests provide diagnostic information in 2875 patients (98.9% of the tests performed). Table 2 shows the main findings of stress-CMR. Resting wall motion abnormalities were detected in 28.9% of the patients, inducible ischaemia was detected in 17.7% of cases and 25.4% had myocardial necrosis on LGE sequences. The subgroup of patients evaluated with adenosine had a slightly higher LV ejection fraction and a slightly higher percentage of ischaemia than the subgroup with regadenoson.

Table 2 CMR characteristics of the study population (n = 2908)
Full size table

The haemodynamic response to regadenoson and adenosine was assessed (Fig. 2a and b). Resting HR and BP were similar between subgroups. Peak stress HR was higher with regadenoson than with adenosine (90 ± 15 vs. 86 ± 15, p < 0.0001). The subgroup of patients with atrial fibrillation (n = 175) had a higher baseline heart rate (77 vs. 66 bpm, p < 0.0001) and a lower systolic blood pressure at rest (137 vs. 141 mmHg, p < 0.05) as well as a lower HR increase than patients in sinus rhythm (19 vs. 23 bpm, p < 0.0001). These differences were similar in the group treated with regadenoson and in the adenosine group. In 37 patients there was a decrease in HR, less frequent with regadenoson than with adenosine (22 [1.1%] vs. 15 [2.3%], p = 0.03), but in no case was severe, with the maximum HR always being ≥ 44 bpm.

Fig. 2
figure 2

Haemodynamic response with stress-CMR. a Absolute values of baseline heart rate (HR), systolic blood pressure (BP) and diastolic BP. b Percent changes in HR, systolic and diastolic BP with stress. ∆ heart rate (%) = [(Peak stress heart rate−Resting heart rate)/Resting heart rate] × 100. ∆ systolic blood pressure (%) = [(Peak stress systolic blood pressure−Resting systolic blood pressure)/Resting systolic blood pressure] × 100. ∆ diastolic blood pressure (%) = [(Peak stress diastolic blood pressure−Resting diastolic blood pressure)/Resting diastolic blood pressure] × 100. Values reported are means. Error bars represent standard deviation

Full size image

Safety

No patient died during the performance of the stress-CMR. Complications and minor symptoms after stress.CMR are shown in Fig. 3a and b. Two patients without previously known COPD or asthma had severe bronchospasm after adenosine (0.3%) while no severe immediate complications occurred after regadenoson, p = 0.05 (Table 3). Only 11 patients (0.5%) in the regadenoson subgroup and 5 (0.8%) in the adenosine subgroup had non-severe complications, p = n.s. All cases were resolved in the CMR unit with no need for further specific care. Tables 4 and 5 depicts the characteristics of the patients who had complications. Given the low number of complications, only a univariate analysis was performed, which showed that patients who presented complications had more prevalence of known COPD or asthma and ischaemia during the test (Fig. 4a) compared to the subgroup with no complications. There were only two cases of non-severe bronchospasm (0.088%) following the use of regadenoson, both with known COPD or asthma, which resolved in the CMR unit after administering inhaled beta-agonists and intravenous corticosteroids. We have analyzed separately, based on clinical history, the patients who received regadenoson with more severe reactive airway disease for being taking inhaled or oral corticosteroids or being on active treatment with bronchodilators in the last weeks (131/305–43%). There were not significant differences in the number of complications between patients with severe in comparison with less severe reactive airway disease (3/131 vs 3/174, p = ns). No patient presented high-grade atrioventricular block when regadenoson was used, compared to two patients who presented paroxysmal atrioventricular block after use of adenosine. The presence of inducible wall motion abnormalities in the patients with ischaemia were not associated with a higher rate of immediate complications. The resting HR, BP, haemodynamic response or presence of atrial fibrillation were not associated to presence of complications in the complete group nor according to the drug used.

Fig. 3
figure 3

Complications (a) and minor symptoms (b) during stress-CMR

Full size image
Table 3 Complications After Stress-CMR (n = 2908)
Full size table
Table 4 Complications After Regadenoson-CMR (n = 11/2253)
Full size table
Table 5 Complications After Adenosine-CMR (n = 7/655)
Full size table
Fig. 4
figure 4

Factors associated with the presence of complications (a) and minor symptoms (b) during stress-CMR. Bars represent the percentage of cases that each factor presents in the different categories

Full size image

Obese patients (body mass index, BMI > 30 kg/m2, n = 1051) were shown to have no more complications than non-obese patients (BMI < 30 kg/m2, n = 1803) (0.7% vs. 0.6%, p = ns). No other demographic, clinical, hemodynamic, and CMR-derived parameters including the presence of scar in LGE images were associated with the occurrence of immediate complications. Neither were any of the factors regarding CMR unit in which the study was performed, referral department, performing cardiologist, or patient type (inpatient vs. outpatient).

Chronic renal failure was present in 71 patients (2.4%), with glomerular filtration rate (GFR) figures between 30 and 60 mL/min. No complications occurred in these patients.

Gadolinium related non-severe adverse effects occurred in only 14 patients (0.5%), mostly nausea and vomiting and only three cutaneous allergic reactions (mild urticaria) that resolved using standardized treatment for allergic reactions.

Minor symptoms occurred in 847 patients (29.1%) and resolved in all cases after administration of aminophylline, self-limited dyspnoea being the most frequent with regadenoson and mild chest pain with adenosine (Table 6). Minor symptoms were more frequent with the use of adenosine and were associated with female sex, myocardial ischaemia in the stress-CMR and previous history of CAD (Fig. 4b). Although the presence of ischaemia was associated with a higher percentage of complications, most of the patients who had ischaemia did not present any complications. Figures 5 and 6 show two clinical examples of stress-CMR with adenosine and regadenoson.

Table 6 Minor Symptoms After Stress-CMR
Full size table
Fig. 5
figure 5

Example of Adenosine stress first pass perfusion in a 69-year-old woman diagnosed with elevated blood pressure, hypercholesterolemia, diabetes, ex-smoker and multivessel coronary disease, with no history of reactive airway disease. Adenosine is administered at 140 mcg/kg/min, achieving an adequate vasodilator response at 4 min. Images are ordered from base to apex (a to d). A perfusion defect is evident at the inferior, inferoseptal and anterior segments (arrowheads). The patient has no angina during the administration of adenosine but increasing dyspnoea occurs with low O2 saturation (85%) that requires the administration of oxygen therapy, methylprednisolone and hydrocortisone e.v. This patient presents a severe bronchospasm following the administration of adenosine

Full size image
Fig. 6
figure 6

Stress-CMR performed with Regadenoson 400 mcg in a 66-year-old woman diagnosed with elevated BP and moderate COPD. Images are ordered from base to apex (a to d). A perfusion deficit can be seen in the anteroseptal, anterior and apical wall (arrowheads). No symptoms or immediate complications arise following the administration of regadenoson

Full size image

Discussion

This multicentre prospective registry demonstrates the feasibility and safety of regadenoson in comparison to adenosine stress-CMR in a large referral population. This study confirms the safety of regadenoson in patients with reactive airway disease. Stress-CMR with regadenoson is easier and faster to perform than with other vasodilators or dobutamine since the dose is fixed and it does not require dose adjustment to patient´s weight, it is administered intravenously in 10 s, and there is no HR or BP target. Also, regadenoson is usually better tolerated by patients, allowing completion of the study with images of diagnostic quality in most patients. Actually, the percentage of diagnostic studies is higher with regadenoson than with other drugs [19, 20].

Vasodilator stress-CMR has been traditionally done with dipyridamole, which inhibits the reuptake of adenosine, and with adenosine, which binds non-selectively to receptors A1, A2A, A2B and A3. A2A receptors are responsible for the vasodilation required in the stress study, while A1, A2B and A3 receptors may be responsible for serious side effects such as high-grade atrioventricular block (A1) or bronchospasm (A2B and A3), which prevent adenosine and dipyridamole to be used in patients with COPD or asthma [8, 9]. Adenosine stress requires two intravenous lines, often in different arms, one for adenosine and the other for contrast administration. In recent years, regadenoson has been approved for use in myocardial stress testing [7]. Regadenoson is a modification of the adenosine molecule that selectively stimulates A2A receptors with a minimal effect on the other receptors, thus avoiding some side effects derived from the use of adenosine [11, 21]. This drug has the advantage of being administered in a single intravenous bolus of 400 μg without the need for adjustment by weight or renal function of the patient. The peak plasma concentration of regadenoson is achieved within 1 to 4 min after injection and parallels the onset of the pharmacodynamic response. The half-life of this initial phase is approximately 2 to 4 min. An intermediate phase follows, if aminophylline is not given, with an average half-life of 30 min coinciding with loss of the pharmacodynamic effect. The last phase consists of a decline in plasma concentration with a half-life of approximately 2 h. The vasodilator efficacy of regadenoson is similar to that of adenosine and superior to dipyridamole, as demonstrated by quantitative perfusion CMR [22]. Also, the fractional flow reserve (FFR) is similar between intravenous administration of regadenoson or adenosine [23, 24]. Regarding the hemodynamic response, regadenoson produces a significant increase in HR, greater than the one produced by adenosine, without significantly affecting BP [11, 12, 22]. The mechanism seems to be a direct sympathetic stimulation rather than a pure baroreflex response [25]. Cases of severe bradycardia and asystole attributable to the intense vagal stimulation caused by the common route of A2A stimulation of the nucleus tractus solitary or hypothalamus have been described [26]. However, in our study, no patient presented this side effect.

Safety

The safety profile of stress-CMR has been demonstrated, with published data regarding adenosine [27], dipyridamole or dobutamine stress-CMR [19, 28]. Safety of regadenoson has been evaluated mainly in nuclear studies, with favorable results compared to adenosine [10,11,12, 29], though reports on regadenoson stress-CMR are scarce [15]. Thus, in a study of regadenoson CMR in 728 patients, a very low incidence of adverse effects with no deaths, one hospitalization for decompensation of heart failure and a single case of bronchospasm were reported, although there were only ten patients with COPD or asthma in the study population. In particular, the safety of this drug has been evaluated in patients with COPD or mild or moderate asthma in randomized, double-blind, placebo-controlled studies [30,31,32]. Likewise, a very low profile of complications has been demonstrated in patients with severe lung disease [33]. In our study, which included a significant proportion of patients (305, 13.5%) with COPD or asthma in the regadenoson subgroup, there were only two cases of non-severe bronchospasm registered, this supports the safety of regadenoson stress-CMR in patients with reactive airway disease. This is, in our opinion, an important aspect since this population has been routinely studied using dobutamine, which is significantly less clinically tolerated, as we have previously demonstrated [20]. The detection of ischaemia is associated with a greater number of complications, although this factor is of little use since we did not know it before performing the stress test. The presence of induced wall motion abnormalities is related to a worse prognosis and need to revascularization, as we previously demonstrated [5]. Although the presence of necrosis does not condition a greater number of immediate complications, it is an important factor since it has recently demonstrated its influence on prognosis, even in cases of unrecognized necrosis [34]. No differences were found in the presence of complications related to the presence of atrial fibrillation, which supports the safe use of regadenoson and adenosine in this patient population. Despite the limitation of ECG monitoring to detect advanced AV block in setting of CMR, we were able to obtain an adequate ECG recording during the test in the vast majority of patients. Thus, no patient in the regadenoson subgroup presented high-grade AV block (second or third degree), which shows that this complication is less frequent with regadenoson than with non-selective adenosine receptor agonists [35]. One of the concerns with the use of regadenoson is the intermediate phase of the stress agent, so we think that the use of aminophylline in all patients may have contributed to better tolerability, as described by Rangel et al. [36]. Likewise, no increased appearance of side effects has been demonstrated in patients with chronic kidney disease, as already described in nuclear tests [37], although this drug is eliminated in 57% unchanged by the renal route. Despite the elimination half-life of regadenoson is prolonged in chronic kidney disease, its maximal plasma concentration, severity and number of adverse side effects are not affected significantly, and its use can be safe even in patients with end-stage renal disease, as has already been demonstrated [38,39,40]. On the other hand, the use of certain gadolinium-based contrast agents can lead to nephrogenic systemic fibrosis (NSF). In spite of the risk factors for developing NSF include estimated glomerular filtration rate (eGFR) less than 30 mL/min/1.73 m2, the use of low-risk contrast agents, as gadoteric acid, remains approved in Europe and is fairly reasonable, even in patients with severely reduced renal function [41, 42]. Recently, a higher rate of acute adverse events of gadolinium-based contrast agents was found among patients receiving pharmacological stressors compared to non-stress imaging in a large registry [43]. Patients receiving regadenoson had higher event rates when compared to adenosine or dobutamine, but none of the events were severe. Moreover, adverse event rates were balanced between adenosine and regadenoson when gadoteric acid was used (as in our study).

Although it has been postulated that women had a significantly higher rate of any side-effect [38, 44], we have only found this relation in the rate of appearance of minor symptoms. Recently, results from the European Society of Cardiovascular Radiology (ESCR) MRCT Registry have been published including 1151 patients with regadenoson stress-CMR reporting a percentage of complications (2.95%) which could be higher than with other drugs [39]. But in this study, with a significantly lower sample size than our work, most of the complications were moderate, mainly dyspnoea, and there were no severe complications, which is in concordance to our results. In addition, the authors did not mention the percentage of concomitant lung disease or the use of aminophylline after stress. Importantly, the authors suggest that there is a multiplicative association of adverse effects between stress drugs and gadolinium compounds, which would be significant with gadobutrol but not with gadoteric acid, the gadolinium compound we used in our study.

The incidence of minor symptoms in our study was low compared to other series [11, 12, 45, 46]. This may be caused by differences in the subjective assessment of the importance of symptoms. The rate was also lower with regadenoson than with adenosine. Moreover, we found less symptoms with regadenoson stress-CMR than we have previously reported using dipyridamole or dobutamine [20].

Study limitations

We do not have long term prognostic data from the study cohort. We have not made a complete clinical long-term follow-up of patients for nephrogenic systemic fibrosis as a late complication. However, as far as we are aware, we have not had any documented cases.

All patients with COPD or asthma have been selected for regadenoson use without excluding any, so we suggest that the results can be applied to a non-selected population. However, we do not have information about respiratory tests that some patients had performed. In any case, the only pilot studies that were conducted to investigate the safety of regadenoson in COPD and asthma patients [30, 31] found no significant differences in bronchoconstrictive reactions, mean maximum decline in FEV1 or other parameters of pulmonary function tests between study and placebo groups. Furthermore, Husain et al. [47] found 0% incidence of clinical exacerbation of COPD or asthma after regadenoson MPI in 228 unselected patients with underlying lung disease. Therefore, we cannot analyze the relation between the severity of the reactive airway disease and the incidence of complications after regadenoson stress-CMR.

Importantly, all tests were carried out and interpreted by cardiologists with a high level of experience in performing stress-CMR, so the results in terms of image quality, diagnostic information, and presence of complications should not be extrapolated to other units with a different internal organization.

Although it has not been the goal of the study, we must note that the use of regadenoson and aminophylline represents a higher additional economic cost than adenosine, approximately 42 euros for each scan, although cost-effectiveness studies would be necessary for a complete analysis of this issue.

Conclusion

Stress-CMR with regadenoson and adenosine was evaluated in a large multicentre prospective registry of more than 2900 referral patients with known or suspected CAD. Stress-CMR with regadenoson is feasible, and in the vast majority of patients provides optimal image quality and useful diagnostic information for their diagnostic workup. Regadenoson stress-CMR showed an excellent safety profile with no serious immediate complications and a low incidence of non-severe complications and minor symptoms in comparison with adenosine.

Only inducible ischaemia and history of COPD or asthma were associated with complications. Nevertheless, the use of regadenoson stress-CMR is safe in this type of patients.

The registry was approved by the institutional review board and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. Written informed consent was obtained from all the subjects included.