Interscan reproducibility of computed tomography derived coronary plaque volume measurements using a semi-automated analysis software

• Published in: European heart journal Vol. 45; no. Supplement_1
• Main Authors: Iraqi, N, Mortensen, M B, Sand, N P, Busk, M, Grove, E L, Dey, D, Pedersen, K B, Kanstrup, H, Madsen, K T, Parner, E, Jensen, J M, Noergaard, B L
• Format: Journal Article
• Language: English
• Published: 28.10.2024
• ISSN: 0195-668X; 1522-9645
• DOI: 10.1093/eurheartj/ehae666.175

Abstract Background Coronary computed tomography angiography (CCTA) enables detailed quantification and characterization of coronary atherosclerotic plaque, offering diagnostic and prognostic value. Despite the growing demand for accurate plaque analysis, studies on interscan reproducibility of automated plaque quantification are limited in number and size. Moreover, impact of clinical and technical factors on the reproducibility of plaque assessment tools are scarce. Purpose The objective of this study was to assess the interscan reproducibility of CCTA-derived plaque volume measurements and the implications of clinical and technical characteristics on interscan reproducibility. Methods A total of 101 patients with known coronary artery disease underwent two CCTA scans within hour interval using identical CT acquisition protocol. Coronary plaque volume measurements were quantified on a per-patient level using a contemporary semi-automated plaque analysis software. Results There was no statistically significant difference in median plaque volumes between the first and second scan. The direct correlation was excellent for volumes of non-calcified plaque (NCP), calcified plaque (CP), and total plaque (TP) across all analyses (Pearson’s coefficient 0.96 to 0.99), while moderate for low-density non-calcified plaque (LD-NCP) volumes (Pearson’s coefficient 0.77). Bland Altman analyses demonstrated high interscan agreement with narrow absolute and relative limits of agreement (LoA) across plaque subtypes, except for LD-NCP volumes (Table 1). Interscan reproducibility on CP volumes was affected by CT image quality with narrower LoA in scans with the highest image quality score (p = 0.003), or lowest image reconstructive iteration level (p <0.001). Limits of agreement were significantly narrower for NCP, CP, and TP volumes in lesions located in LAD compared to non-LAD lesions (p <0.001). However, no significant difference in LoA was present when comparing patients according to BMI (<27 and ≥27 kg/m2), heart rate (<65 and ≥65 beats per minute), or Agatston score (<193 and ≥193). Conclusions This comprehensive interscan reproducibility study revealed good agreement in the measurement of plaque subtypes. Adding new insight to existing literature, we have demonstrated that factors such as image quality, image reconstructive algorithms, and lesion location affect the reproducibility of plaque volume measurements. CT-derived plaque quantification using a semi-automated plaque analysis software holds potential for monitoring disease progression or regression in patients with CAD, provided high CCTA image quality.