Anatomical versus functional assessment of coronary artery disease: direct comparison of computed tomography coronary angiography and magnetic resonance myocardial perfusion imaging in patients with intermediate pre-test probability

Computed tomography coronary angiography (CTA) and cardiac magnetic resonance myocardial perfusion imaging (CMR-MPI) are state-of-the-art tools for noninvasive assessment of coronary artery disease (CAD). We aimed to compare the diagnostic accuracy of CTA and CMR-MPI for the detection of functionally relevant CAD, using invasive coronary angiography (XA) with fractional flow reserve (FFR) as a reference standard, and to evaluate the best protocol integrating these techniques for assessment of patients with suspected CAD. 95 patients (68 % men; 62 ± 8.1 years) with intermediate pre-test probability (PTP) of CAD underwent a sequential protocol of CTA, CMR-MPI and XA. Significant CAD was defined as >90 % coronary stenosis, 40–90 % stenosis with FFR ≤ 0.80 or left main stenosis ≥50 %. Prevalence of significant CAD was 43 %. CTA was more sensitive (100 %) but less specific (59 %) than CMR-MPI (88 and 89 %, respectively) for detection of significant CAD, with a strong trend for higher global diagnostic accuracy of CMR-MPI (88 vs. 77 %, p = 0.05). An integrated approach based on an initial CTA and subsequent referral to CMR-MPI of positive/inconclusive results had the best diagnostic performance (AUC 0.91). The direct referral to XA of patients with positive/inconclusive CTA performed worse than a selective approach based on CMR-MPI results (AUC 0.80 vs. 0.91, p = 0.005). In this intermediate PTP population, CMR-MPI showed a strong trend toward better performance compared to CTA for the assessment of functionally significant CAD. A combined protocol integrating coronary anatomy and function seems to be a very effective approach in the accurate diagnosis of CAD.