1357743 - Real-world assessment and validation of high-resolution myocardial perfusion imaging with fully integrated quantitative mapping within routine clinical assessment.

Whilst visual assessment of myocardial ischemia remains the clinical standard within stress perfusion cardiac MRI (CMR), quantitative myocardial perfusion mapping is increasingly used in the diagnosis of both epicardial and microvascular coronary artery disease (CAD). Automated inline techniques allow clinicians to accurately assess myocardial blood flow (MBF) values during CMR acquisition [1]. This study assesses the use of a prototype high-resolution cardiac perfusion sequence combined with integrated quantitative MBF mapping in routine clinical use.

Methods:

331 consecutive patients underwent stress perfusion CMR (adenosine stress 279 [84.5%]) at 3T (MAGNETOM Vida, Siemens Healthcare, Erlangen, Germany) at our tertiary center between January and August 2022 [mean age 61.9 ±12.0; mean LVEF (%) 53.4 ±12.4]. All patients were scanned using a prototype high-resolution (1.3-1.5 mm2) perfusion sequence with temporal regulation and motion-correction integrated into an iterative reconstruction [2,3]. Automated pixel-wise myocardial perfusion mapping was fully implemented within the vendor reconstruction software on the scanner system (Figure 1). Further image assessments and map segmentation were conducted using clinical reporting software (Circle, cvi42).

Results: A total of 129 patients [39.0%] were diagnosed with inducible ischemia on visual assessment. Of those, 95 [73.6%] were thought to have epicardial CAD. Only 11 of these patients [11.6%] received coronary intervention following detection of epicardial ischemia (9 percutaneous coronary intervention; 2 coronary artery bypass surgery). 9 patients diagnosed with epicardial ischemia received invasive fractional flow reserve (FFR) assessment of one or more coronary artery. 14 patients without ischemia on CMR had FFR assessment within 6 months of their scan.
 
Inline sequence motion-correction was successful in 314 patients [94.9%], with automated perfusion maps being generated in 309 patients [93.4%]. In patients with no visual evidence of ischemia and normal FFR ( >0.80), MBF values were within accepted normal ranges [global stress MBF 3.36 ±0.69 ml/g/min; global rest MBF 1.53 ±0.31 ml/g/min; global myocardial perfusion reserve (MPR) 2.23 ±0.43]. Coronary territorial segmented stress MBF/MPR values were significantly lower in corresponding FFR-positive (≤0.80) areas [mean MBF 2.49 ±0.69 ml/g/min; mean MPR 1.57 ±0.71] compared with FFR-negative coronary territories [mean MBF 3.14 ±0.71 ml/g/min, p = 0.027; mean MPR 2.19 ±0.44, p = 0.006]. Figure 2 demonstrates the correlation between FFR and MBF/MPR of the corresponding coronary territory.

Conclusion:

High-resolution myocardial perfusion imaging with scanner-integrated quantitative mapping can provide inline assessment of diagnostically significant coronary ischemia with high reporter confidence. Quantitative perfusion values can be easily measured and segmented during routine clinical reporting, augmenting visual interpretation of inducible ischemia.