Background Stress cardiac magnetic resonance imaging (CMR) is increasingly being used in the assessment of myocardial perfusion at rest and in response to exertion. It provides information on the presence or absence of functionally significant coronary artery disease and allows us to predict the benefit of revascularisation. The aim of this study is to assess the accuracy and prognostic value of stress CMR in our tertiary centre.

Method Retrospective single centre study of every adenosine stress CMR done for patients suspected of having ischaemic heart disease between January 2012 and December 2014 (n=525). Reports were reviewed and patients categorised into “reversible ischaemia” (RI, n=94), “non-reversible ischaemia” (NRI, n=146), and “no ischaemia” (NoI, n=285). Follow up data was collected on all-cause-mortality, hospital admissions with acute coronary syndromes and admissions for revascularisation. A Kaplan-Meier survival analysis was done using the composite end point of admission with acute coronary syndrome or for revascularisation.

Using angiography as the reference standard in those that had concurrent angiograms (n=83) sensitivity and specificity was calculated. Where pressure wire studies (PWS) were undertaken (n=15) results were correlated with CMR reports.

Results 525 patients were included in the final analysis. Median follow-up of 15 months. A total of 46% had a perfusion defect on CMR. During the follow up, the distribution in composite outcomes in those with and those without defect was highly significant (p<0.0001) (see figure 1 and table 1).

83 patients had concurrent angiograms, CMR demonstrated 91.1% sensitivity, 77.8% specificity, 89.5% positive predictive value and 80.8% negative predictive value (see table 2). There was 100% correlation between CMR and PWS in those reported as having RI (n=5) and 80% correlation in those who had NoI (n=5). All cause mortality in the NoI group was 2.97% (8/269), and in the ischaemia groups was 4.66% (11/236), though there was no statistical significance our study was limited by the small number of deaths.

Conclusion There is good correlation between pressure wire studies and stress CMR, but our sample size is small, larger studies (such as MR-INFORM) are needed to accurately inform the role of CMR perfusion in investigating and managing coronary artery disease. Patients with a normal stress CMR scan appear to have excellent prognosis.

• Download figure
• Open in new tab
• Download powerpoint

Abstract 109 Figure 1