Search

Background: Stress cardiovascular magnetic resonance (CMR) is a common non-invasive modality used for the diagnosis of coronary artery disease (CAD). However, its accuracy for diagnosing CAD in patients with moderate to severe left ventricular systolic dysfunction (M-S LVSD) is unclear. Although adenosine is the most commonly used agent in stress CMR, haemodynamic responses to adenosine may be diminished in M-S LVSD, so an alternative strategy utilising inotropic stress may provide superior diagnostic performance. We hypothesised that dobutamine stress perfusion CMR would achieve high diagnostic accuracy, exceeding that of adenosine stress CMR for diagnosing severe CAD in patients with M-S LVSD. Thesis outline and aims: This thesis comprises two studies: (1) a retrospective study (Chapter 3) examining the impact of LVSD on haemodynamic responses to adenosine, and (2) prospective study (Chapters 4-6) comparing the diagnostic accuracy of adenosine and dobutamine stress CMR in M-S LVSD. Methods: In the first retrospective study of 497 patients, haemodynamic responses to adenosine were compared in subjects with normal, mild-moderately impaired and severely impaired LV systolic function. For the second study, 41 subjects (31 with M-S LVSD and 10 matched controls [MCs]) underwent both adenosine and dobutamine stress CMR, with invasive coronary angiography or CT coronary angiography as the reference standard. Results: The retrospective study showed that fewer patients in the severe LVSD group (26%) achieved an adequate heart rate (HR) response with adenosine compared to those with mild-moderate (55%) and normal LV systolic function (68%), and more patients with severe LVSD required a dose increase (41% versus 24% and 19%, respectively, p<0.001). In the prospective DISCORDANCE study, 31 LVSD subjects (age 66.1±8.9 years, LVEF 34.57.2%) and 10 MCs (age 63.5±6.4 years, EF 57.3±5.1%) were studied. In MCs, dobutamine produced a higher hyperaemic response than adenosine (2.1±1.4 ml/min/g versus 1.6±0.2 ml/min/g, respectively (p<0.001)). However, in patients with LVSD, a lower hyperaemic response was achieved with dobutamine than with adenosine (1.2±0.3 ml/min/g versus 1.5±0.5ml/min/g, respectively (p<0.001)). For the diagnosis of severe CAD on a per-patient basis, qualitative assessment of adenosine stress CMR perfusion imaging achieved sensitivity 96.2%, specificity 60% and accuracy 90.3%. The sensitivity of dobutamine stress was 92.6%, specificity 60% and accuracy 87.1% (compared with adenosine stress, p=0.93, 0.12 and 0.19, respectively). For quantitative perfusion imaging, the area under the curve for adenosine stress was 0.71±0.54 (95% confidence interval (CI) 0.61-0.82, p<0.001) and 0.78±0.51 (95% CI 0.68-0.88, p<0.001) for dobutamine stress (p=0.36 for difference). Conclusion: Patients with severe LVSD have a blunted response to adenosine, with more patients requiring higher adenosine doses to achieve a satisfactory haemodynamic response. Comprehensive stress CMR with both adenosine and dobutamine achieve good diagnostic performance for identifying significant CAD in patients with severe LVSD. The diagnostic accuracy for detecting significant CAD appears comparable with both stress modalities in M-S LVSD, though the detection of small differences in diagnostic performance is precluded by the relatively small size of the cohort examined.