Quantitative coronary CT angiography: absolute lumen sizing rather than %stenosis predicts hemodynamically relevant stenosis

OBJECTIVE: To identify the most accurate quantitative coronary stenosis parameter by CTA for prediction of functional significant coronary stenosis resulting in coronary revascularization. METHODS: 160 consecutive patients were prospectively examined with CTA. Proximal coronary stenosis was quantified by minimal lumen area (MLA) and minimal lumen diameter (MLD), %area and %diameter stenosis. Lesion length (LL) was measured. The reference standard was invasive coronary angiography (ICA) (>70 % stenosis, FFR <0.8). RESULTS: 210 coronary segments were included (59 % positive). MLA of ≤1.8 mm(2) was identified as the optimal cut-off (c = 0.97, p < 0.001; 95 % CI 0.94–0.99) (sensitivity 90.9 %, specificity 89.3 %) for prediction of functional-relevant stenosis (for MLA >2.1 mm(2) sensitivity was 100 %). The optimal cut-off for MLD was 1.2 mm (c = 0.92; p < 0.001; 95 % CI 0.88–95) (sensitivity 90.9, specificity 85.2) while %area and %diameter stenosis were less accurate (c = 0.89; 95 % CI 0.84–93, c = 0.87; 95 % CI 0.82–92, respectively, with thresholds at 73 % and 61 % stenosis). Accuracy for LL was c = 0.74 (95 % CI 0.67–81), and for LL/MLA and LL/MLD ratio c = 0.90 and c = 0.84. CONCLUSIONS: MLA ≤1.8 mm(2) and MLD ≤1.2 mm are the most accurate cut-offs for prediction of haemodynamically significant stenosis by ICA, with a higher accuracy than relative % stenosis. KEY POINTS: • Quantitative coronary CT-angiography is accurate for prediction of functional relevant stenosis. • Absolute lumen area and diameter rather than %stenosis predict functional relevance. • Lumen area <1.8 mm (2) and diameter <1.2 mm are the most accurate cut-offs. • Quantitative parameters are helpful for decision-making in terms of patient management.