Cardiovascular magnetic resonance imaging of myocardial oedema following acute myocardial infarction: Is whole heart coverage necessary?

Background: aar measurement is useful when assessing the efficacy of reperfusion therapy and novel cardioprotective agents after myocardial infarction. multi-slice (typically 10-12) t2-stir has been used widely for its measurement,typically with a short axis stack (sax) covering the entire left ventricle,which can result in long acquisition times and multiple breath holds. this study sought to compare 3-slice t2-short-tau inversion recovery (t2- stir) technique against conventional multi-slice t2-stir technique for the assessment of area at risk (aar). methods: cmr imaging was performed on 167 patients after successful primary percutaneous coronary intervention. 82 patients underwent a novel 3-slice sax protocol and 85 patients underwent standard 10-slice sax protocol. aar was obtained by manual endocardial and epicardial contour mapping followed by a semi- automated selection of normal myocardium; the volume was expressed as mass (%) by two independent observers. results: 85 patients underwent both 10-slice and 3-slice imaging assessment showing a significant and strong correlation (intraclass correlation coefficient = 0.92;p < 0.0001) and a low bland-altman limit (mean difference -0.03 ± 3.21 %,95 % limit of agreement,- 6.3 to 6.3) between the 2 analysis techniques. a further 82 patients underwent 3-slice imaging alone,both the 3-slice and the 10-slice techniques showed statistically significant correlations with angiographic risk scores (3-slice to bari r = 0.36,3-slice to approach r = 0.42,10-slice to bari r = 0.27,10-slice to approach r = 0.46). there was low inter-observer variability demonstrated in the 3-slice technique,which was comparable to the 10-slice method (z = 1.035,p = 0.15). acquisition and analysis times were quicker in the 3-slice compared to the 10-slice method (3-slice median time: 100 seconds (iqr: 65-171 s) vs (10-slice time: 355 seconds (iqr: 275-603 s); p < 0.0001. conclusions: aar measured using 3-slice t2-stir technique correlates well with standard 10-slice techniques,with no significant bias demonstrated in assessing the aar. the 3-slice technique requires less time to perform and analyse and is therefore advantageous for both patients and clinicians. © 2016 hamshere et al.