High Resolution Myocardial Perfusion Imaging at 3 Tesla: Comparison to standard 1.5 Tesla perfusion studies and diagnostic accuracy in patients with suspected CAD

Purpose: Recently, myocardial rest perfusion in healthy volunteers at 3.0T has been shown to significantly increase signal-to-noise-ratio (SNR) in comparison to 1.5T. However, the benefits of 3.0T with respect to image quality in comparison to 1.5T as well as diagnostic accuracy for the detection of coronary artery disease (CAD) have not been evaluated, yet.

The purpose of this study was to evaluate 1) high resolution myocardial rest perfusion in healthy volunteers (group 1) at 3.0T in comparison to a standard approach at 1.5T, as well as 2) the technical feasibility and diagnostic accuracy of 3.0T high resolution myocardial stress perfusion in patients with suspected CAD (group 2).

Methods: All perfusion studies were performed using a T1-weighted saturation-recovery k-space segmented gradient-echo sequence combined with parallel imaging (Gad-DTPA 0.05 mmol/kg BW).

Study group 1 consisted of 17 healthy volunteers (9 males, 8 females; mean age: 25.6 +/- 4.1 years) with low likelihood of CAD (<5% Diamond & Forrester) receiving 1) a high resolution rest perfusion scan at 3.0T (in-plane resolution 1.8×2.1 mm2, pixel size 3.78 mm2, slice thickness 8mm, TE/TR/α 3.7/1.8 ms/15°, SENSE factor 3; Achieva 3.0T, Philips Medical Systems) and 2) a standard perfusion approach at 1.5T (in-plane resolution 2.9×3.4 mm2, pixel size 9.86 mm2, slice thickness 8mm, TE/TR/α 3.7/1.8 ms/ 20°, SENSE factor 2; Intera 1.5T, Philips Medical Systems). Both rest studies were performed in random order on two separate days (interval: 24 to 72 hours). Myocardial signal intensity increase [(peak signal intensity/baseline signal intensity) x 100] was assessed for high resolution and standard scans. Image quality was evaluated by two observers on a four point grading scale (4: excellent, 1: non-diagnostic) with respect to homogeneity of myocardial enhancement, blurring and presence of artifacts.

Study group 2 consisted of 19 patients (12 males, 7 females; mean age: 59.3 +/- 7.2 years) with suspected CAD scheduled for X-ray coronary angiography. Patients underwent a high resolution stress (adenosine 140µg/kg BW over 6min) and rest perfusion study at 3.0T (interval: 30min) with equivalent Gd-DTPA concentration and scan parameters as for healthy volunteers at 3.0T. Perfusion studies were qualitatively assessed for stress-induced hypoperfusion by two experienced observers.

Hemodynamically significant coronary artery disease was defined as luminal diameter stenosis >70% in conventional coronary angiography.

Results: Study group 1: Myocardial signal intensity increase in high resolution rest perfusion studies at 3.0T was not significantly different to the standard approach at 1.5T (207% +/- 43 vs. 199% +/- 52, p>0.05). Image quality was significantly improved in high resolution perfusion imaging at 3.0T in comparison to the standard approach at 1.5T (3.1 +/- 0.7 vs. 2.2 +/- 0.4, p<0.01).

Study group 2: Mean score of image quality was 2.7 +/- 0.7. The mean myocardial signal intensity increase was 170% +/- 30. Sensitivity and specificity of high resolution perfusion studies at 3.0T for the detection of significant coronary artery disease was 90% (n=9/10) and 78% (n=7/9), respectively.

Conclusion: 1) MR myocardial perfusion imaging at 3.0T permits significantly improved spatial resolution allowing for a 60% reduction of pixel size. 2) High resolution perfusion at 3.0T provides significantly improved image quality in comparison to a standard approach at 1.5T. 3) 3.0T perfusion studies are also feasible in patients with suspected CAD in a clinical setting and yield promising results for the detection of significant coronary artery stenosis.