Benchmarking Photon-Counting Computed Tomography Angiography Against Invasive Assessment of Coronary Stenosis: Implications for Severely Calcified Coronaries

Background: Clinical guidelines do not recommend coronary computed tomographic angiography (CTA) in elderly patients or in the presence of heavy coronary calcification. Photon-counting coronary computed tomographic angiography (PCCTA) introduces ultrahigh in-plane resolution and multienergy imaging, but the ability of this technology to overcome these limitations is unclear. Objectives: The authors evaluate the ability of PCCTA to quantitatively assess coronary luminal stenosis in the presence and absence of calcification, comparing both the ultrahigh-resolution (UHR)-PCCTA and the multienergy standard-resolution (SR)-PCCTA with the criterion-standard 3-dimensional invasive quantitative coronary angiography (3D QCA). Methods: The authors included 100 patients who had both PCCTA and invasive coronary angiography (ICA). They comparatively evaluated luminal diameter stenosis with PCCTA and 3D QCA, anatomic disease severity (according to CAD-RADS [Coronary Artery Disease–Reporting and Data System]) and the diagnostic performance of PCCTA in identifying coronary arteries with ≥50% diameter stenosis on 3D QCA requiring invasive hemodynamic severity evaluation and/or revascularization. Results: The authors analyzed 257 vessels and 343 plaques. UHR-PCCTA luminal evaluation relative to 3D QCA was more precise than SR-PCCTA (median difference: 3% [Q1-Q3: 1%-6%] vs 6% [Q1-Q3: 2%-11%]; P < 0.001), particularly in severely calcified arteries (median difference 3% [Q1-Q3: 1%-6%] vs 6% [Q1-Q3: 3%-13%]; P = 0.002). Per-vessel agreement for CAD-RADS between UHR-PCCTA and 3D QCA was near-perfect (κ = 0.90 [Q1-Q3: 0.84-0.95]; P < 0.001), and it was substantial for SR-PCCTA (κ = 0.63 [Q1-Q3: 0.54-0.71]; P < 0.001), especially in severely calcified arteries: κ = 0.90 (Q1-Q3: 0.83-0.97; P < 0.001) and κ = 0.67 (Q1-Q3: 0.56-0.77; P < 0.001), respectively. Per-vessel diagnostic performance of SR- and UHR-PCCTA was excellent: AUC: 0.94 (95% CI: 0.91-0.98; P < 0.001) and 0.99 (95% CI: 0.98-1.00; P < 0.001), respectively. UHR-PCCTA diagnostically outperformed SR-PCCTA: ΔAUC: 0.05 (95% CI: 0.01-0.08; P = 0.01). Conclusions: PCCTA compares favorably with ICA for lumen assessment and anatomic disease severity classification in patients presenting with acute coronary syndrome or patients referred for ICA. UHR-PCCTA luminal evaluation is superior to SR-PCCTA, especially in patients with heavy coronary calcification. UHR-PCCTA has excellent diagnostic performance in identifying coronary arteries with ≥50% luminal stenosis on 3D QCA, outperforming standard-resolution imaging.