Impact of coronary calcium score on the diagnostic accuracy of ultra-high-resolution CT for detecting obstructive coronary artery disease in the CORE PRECISION multicenter study

Background Dense coronary calcification limits the use of conventional coronary CT angiography (CCTA) due to blooming artifacts that obscure the lumen and reduce its diagnostic accuracy. Ultra–high–resolution CT (UHR-CT), with improved spatial resolution, may mitigate this limitation and allow a more reliable assessment even in patients with high calcium score levels. Purpose To investigate the influence of coronary artery calcium score (CACS) in the diagnostic accuracy of UHR-CT for detecting hemodynamically significant coronary artery disease (CAD) at patient level using quantitative flow ratio (QFR) as the reference standard, and to explore whether a clinically meaningful CACS threshold exists beyond which UHR-CT accuracy declines. Methods In the CORE Precision study, participants with suspected obstructive CAD underwent non-contrast calcium scoring and UHR-CT prior to invasive coronary angiography. UHR-CT images were analyzed and stenosis ≥70% were deemed obstructive. Functional significance (reference standard) was defined by an abnormal QFR. CACS was analyzed both categorically (0–99, 100–399, 400–999, ≥1000) and as a continuous variable using a moving-threshold approach. Diagnostic accuracy was calculated across increasing CACS cutoffs, restricting analyses to thresholds where each subgroup comprised ≥10% of the cohort to ensure stability. Results A total of 118 participants (mean age 67±11 years, 75% male) were included. The mean calcium score was 1,004 (± 1,108). The global diagnostic accuracy of UHR-CT was 80.3%. Diagnostic accuracy declined progressively across increasing CACS categories, measuring 90.9% for CACS 0–99, 88.5% for CACS 100–399, 83.8% for CACS 400–999, and 76.7% for CACS ≥1,000. Performance attenuation at very high CACS values was primarily driven by reduced specificity, while sensitivity remained preserved. Detailed diagnostic performance metrics and confidence intervals are presented in Figure 1A. In the continuous analysis, no abrupt cutoff point was identified. The accuracy remained stable, especially in patients with CACS <1,000, with a gradual decline with increasing calcium burden (Figure 1B). The histograms in Figure 1B illustrate correctly and incorrectly classified cases across CACS ranges. Conclusions UHR-CT maintains robust diagnostic accuracy across a wide range of coronary calcium scores, particularly in patients with CACS <1000, with no evidence of a clear threshold for abrupt decline in diagnostic performance. The high accuracy for detecting obstructive CAD in patients previously considered unsuitable for the test suggests that UHR-CT may represent a noninvasive diagnostic alternative in this group.