Effects of adenosine and regadenoson on hemodynamics measured using cardiovascular magnetic resonance imaging

A total of 25 healthy subjects between the ages of 18–45 years with normal electrocardiograms were prospectively enrolled to undergo a 6-min infusion of adenosine followed by regadenoson administration for assessment of changes in ventricular volumes and function by CMR. Patients were excluded if they reported a history of moderate to severe reactive airway disease, CAD, pulmonary hypertension, structural heart disease, theophylline use, contraindications to CMR, or were pregnant. All patient reported data was confirmed by study personnel via review of electronic medical records. Baseline demographic data to include age, gender, height, and weight were also collected. Study patients were instructed to refrain from caffeine consumption 48 h prior to CMR and presented the day of the study in a fasting state. All patients provided written informed consent and the study was approved by the San Antonio Military Medical Center institutional review board. No monetary compensation was provided for participation. DT and KS had full access to all the data in the study and takes responsibility for its integrity and the data analysis.

All study participants underwent CMR on a 1.5 T CMR scanner (Siemens Magnetom Espree, Siemens Healthineers, Erlangen Germany [Tim 76 × 18]) with a 16 channel body coil (Invivo). Initial image sequences included localization and scout acquisitions. Baseline short axis (SAX) cine (steady state free procession) acquisitions (8 mm thickness, 2 mm skip) were acquired from the mid-atrial cavity level through the ventricular apex. This same SAX cine acquisitions were repeated immediately following conclusion of vasodilator stress administration (both adenosine and regadenoson) and at 5, 10, and 15 min after vasodilator administration. Intravenous gadolinium was not given. All CMR acquisition sequences were analyzed on a 3D workstation (QMass, Vital Images, Medis BV, Leiden, the Netherlands) by a cardiac imaging trained radiologist (MM) and Level III fellowship-trained CMR cardiologist (KS). Endocardial borders were defined for both the right ventricle (RV) and left ventricle (LV) using semi-automated software application with manual correction. Values were then indexed to body surface area (BSA) to determine LV end-diastolic volume (LVEDVi), LV end-systolic volume (LVESVi), LV ejection fraction (LVEF), RVEDVi, RVESVi, and RVEF.

The stress protocol is outlined in Fig. 1. Abstinence from caffeine was verified orally on the day of the procedure. Following baseline SAX cine acquisitions, study participants received an infusion of adenosine (140mcg/kg/min IV for 6 min). As above, SAX cine stacks were obtained immediately following vasodilator administration and at 5 min, 10 min, and 15 min. Following the completion of the 15 min adenosine scan an additional washout period of 10 min was taken prior to administration of intravenous regadenoson (0.4 mg over 10 s) followed by the same SAX image acquisitions. Aminophylline was not given following regadenoson in accordance with published Society for Cardiovascular Magnetic Resonance (SCMR) protocols [9]. Blood pressure (BP), heart rate, and pulse oximetry (SpO2) were obtained at baseline, prior to each vasodilator administration and at the same time points as repeat image acquisition (immediately following vasodilator administration, 5 min, 10 min, and 15 min). For the purposes of assessing temporal changes following regadenoson administration, the volumetric and hemodynamic data acquired at the 15 min post-adenosine infusion mark were used as the baseline data for the regadenoson acquisitions. Following completion of image acquisition, study participants were surveyed regarding any side effects experienced with adenosine and regadenoson infusion and vasodilator preference.

The study population consisted of 17 males and 8 females ranging in age from 28 to 42 years (mean 34 ± 5 years) with a mean body mass index (BMI) of 25.6 ± 4.1 kg/m2 (Table 1). No subjects had a history of theophylline use or caffeine consumption within 48 h. The baseline heart rate, BP and SpO2 were 64 ± 8 bpm, 136 ± 16/69 ± 12 mmHg, and 98 ± 1%, respectively. Eight subjects (32%) had initial baseline heart rates <60 bpm.

Table 2 shows heart rate and BP changes with both adenosine and regadenoson. Administration of both vasodilatory agents resulted in an increase from baseline to peak heart rate immediately (Fig. 2) following vasodilator administration (adenosine: 64 ± 8 to 96 ± 13 bpm vs regadenoson: 65 ± 13 to 107 ± 10 bpm, p < 0.0001 for both). Conversely, heart rate returned to baseline by the 5 min mark following adenosine administration (p = 0.11) whereas a persistent, significant elevation in heart rate was observed in patients following 15 min after regadenoson administration (p < 0.0001). No significant durable effects were observed in systolic BP, diastolic BP, or Sp02 with all values returning to baseline by the 15 min mark following both regadenoson and adenosine administration.

Administration of both adenosine and regadenoson resulted in an increase in left ventricular ejection fraction (LVEF) compared with baseline (Table 2). There was no significant difference in the peak increase in LVEF (Fig. 3a), which was observed immediately following administration of both vasodilators (p = 0.68), however LVEF following adenosine returned to baseline by the 10 min mark (baseline: 63.0 ± 4.6 vs 63.3 ± 5.4, p = 0.84). Conversely, LVEF following regadenoson administration remained increased when compared to baseline at the 15 min mark (baseline: 63.0 ± 6.9 vs 67.2 ± 4.9, p = 0.0015). This represents a mean increase in LVEF of 4.2 ± 1.3% baseline following regadenoson administration at the end of the sampling period (15 min mark).

Similar to LVEF, an initial increase was also observed in right ventricular ejection fraction (RVEF) following administration of both adenosine and regadenoson (Fig. 3b). Additionally, RVEF returned to baseline by the 5 min mark following adenosine administration (baseline: 57.0 ± 7.5 vs 59.1 ± 6.1, p = 0.13) while RVEF remained increased, though to a lesser degree when compared to effects on LVEF, following regadenoson administration out to 15 min (baseline: 58.5 ± 8.2 vs. 61.9 ± 7.5, p = 0.016).

Adenosine and regadenoson were observed to have significant effects on LV end-diastolic volume index (EDVi) (Fig. 4a). LVEDVi dropped significantly when compared to baseline starting at the 10 min mark before reaching a nadir at the 15 min time point (net reduction in LVEDVi 4.9 ± 1.6 mL/m2, p = 0.002). LVEDVi trended downward following regadenoson administration, as well, reaching a nadir at the 15 min time point (net reduction in LVEDVi 4.1 ± 1.6 mL/m2, p = 0.010). When comparing the magnitude of LVEDVi reduction following both vasodilator agents, a greater reduction was observed following adenosine administration by the end of the protocol (p = 0.010).

An immediate, significant reduction in LV end-systolic volume index (ESVi) (Fig. 4b) was observed following administration of both adenosine and regadenoson (p = 0.0002 and p < 0.0001 for difference from baseline, respectively). Following adenosine infusion, however, LVESVi returned to baseline by the 10 min mark (baseline: 26.4 ± 7.0 mL/m2 vs 24.3 ± 7.9 mL/m2, p = 0.072) and remained as such through the 15 min period (p = 0.52). At termination of post-adenosine monitoring, the net decrease from baseline in LVESVi was 0.7 ± 1.2 mL/m2. Conversely, LVESVi remained decreased throughout the 15 min following administration of regadenoson (baseline: 25.6 ± 9.7 mL/m2 vs 20.3 ± 5.5 mL/m2, p < 0.0001). This translates to a 5.3 ± 4.2 mL/m2 reduction from baseline in LVESVi following regadenoson administration.

When combining these effects by looking at LV stroke volume indexed for body surface area (LVSVi), LVSVi following regadenoson returned to baseline levels within 5 min (p = 0.19 for net difference from baseline) of administration and remained as such through the 15 min observation period (p = 0.99 for net difference from baseline). This is explained by a fairly similar net reduction in both LVEDVi and LVESVi following regadenoson. Thus, with no change in LVSVi and a significant reduction in LVEDVi, a significant increase in LVEF would be expected and was observed. Conversely, LVSVi following adenosine infusion was trending toward significance (p = 0.058) with a net reduction of 2.9 ± 3.1 mL/m2 at the end of the observation period (15 min mark). This is explained by a more significant reduction in LVEDVi when compared to LVESVi following adenosine. Thus, LVSVi would be reduced to a similar magnitude to LVEDVi resulting in a negligible change in LVEF.

For completeness, temporal changes in RV volumes following adenosine and regadenoson administration are listed in Table 2. In general, the effects on RVEDVi and RVESVi following adenosine and regadenoson roughly mirror those seen in the LV volumes. Of note, LV mass, when indexed to BSA, predictably did not change significantly when assessed at all time points following administration of both adenosine and regadenoson.

Fifteen of the 25 subjects (60%) reported one or more side effects (Fig. 5) with adenosine compared and eleven (44%) with regadenoson (p = 0.40). Following administration of adenosine, 42% of the subjects reported dyspnea, 19% nausea, 4% a headache, 8% abdominal pain, and 27% chest pain. Following regadenoson, 44% of the subjects had dyspnea, 19% nausea, 12% a headache, 19% abdominal pain, and 6% chest pain. Only the incidence of chest pain differed significantly between adenosine and regadenoson (p < 0.049). When the total number of reported side effects were compared, a significant reduction in total side effects was noted with regadenoson (p = 0.015). Additionally, patient surveys revealed that subjects preferred >5:1 (p = 0.001) regadenoson over adenosine with respect to tolerability of side effects.

Our study population was young and healthy therefore hemodynamic responses may not be similar to individuals with cardiovascular disease and/or myopathic processes. In addition, study subjects were instructed to refrain from caffeinated products 48 h prior to participation with oral verification at the time of the procedure and were not verified through serum analysis of caffeine levels. In an effort to limit significant alterations in plasma volume, subjects received two vasodilators during the same CMR session with a 25 min delay between injections of adenosine to the injection of regadenoson. Despite the fact that all easily detectable measures had returned to normal following adenosine administration (i.e. vital signs), not all of the ventricular volumes had returned to baseline, specifically LVEDVi, RVEDVi and RVESVi. This may impact the baseline comparison to post-regadenoson acquisitions. The use of the 15 min post-adenosine time point as the regadenoson baseline in lieu of performing a second, separate baseline data acquisition prior to administration of regadenoson may have resulted in some cross contamination with residual adenosine effects despite its short half-life of less than 10 s [18]. Reported side effects following regadenoson were increased compared to published data and may have been the result of consecutive vasodilator agent applications [19]. Some centers have incorporated the routine use of aminophylline following regadenoson stress to truncate hemodynamic effects as Dandekar et al. demonstrated rapid return to baseline of myocardial blood flow using this strategy [12]. While some CMR centers routinely use aminophylline as an abortive agent, we chose to follow the most current SCMR guidelines due to variations in this practice pattern. Our findings on the persistent augmented ejection fraction induced by regadenoson provides support for the use of an abortive agent if an accurate measurement of the EF is clinically impactful.