Mean density of computed tomography for predicting rotational atherectomy during percutaneous coronary intervention.

Multi-slice computed tomography (CT) allows noninvasive evaluation of the severity of coronary calcification. However, there has yet to be a definitive parameter based on the cross-sectional CT image for predicting the need for rotational atherectomy (RA). Therefore, we aimed to investigate the mean density of cross-sectional CT images to predict the need for RA during percutaneous coronary intervention (PCI). A total of 154 lesions with moderate to severe calcification detected in coronary angiography were identified in 126 patients who underwent coronary CT prior to PCI for stable angina. PCI with RA was performed for 48 lesions, and the remaining 106 were treated without RA. Multi-slice CT was retrospectively evaluated for its ability to predict the use of RA. We chose the most severely calcified cross-sectional image for each lesion. The mean density within the outer vessel contour, calcium arc quadrant of the cross-sectional CT image, calcium length, calcification remodeling index, and per-lesion coronary artery calcium score was studied. Receiver-operator characteristic curve analysis revealed 637 Hounsfield units (HU) (area under the curve ​= ​0.98, 95% confidence interval: 0.97–1.00, p ​< ​0.001) as the best mean density cutoff value for predicting RA. Multivariate logistic regression analysis showed that a mean calcium level >637 HU was a strong independent predictor (odds ratio: 32.8, 95% confidence interval: 7.0–153, p ​< ​0.001) for using RA. The mean density of the cross-sectional CT image, a simple quantitative parameter, was the strongest predictor of the need for RA during PCI. TOC Summary : We investigated the mean density of cross-sectional computed tomography (CT) images to predict the need for rotational atherectomy (RA) during the percutaneous coronary intervention (PCI). A total of 154 lesions (48 with RA and 106 without RA) of stable angina were enrolled. Receiver operator characteristic curve analysis revealed that the area under the curve of mean density for predicting RA was 0.98 (95% confidence interval:0.97–1.00). A mean density >637 Hounsfield units was the strongest independent predictor for RA, among other potential predictors: calcium arc quadrant, calcium length, calcification remodeling index, and per-lesion coronary artery calcium score. [ABSTRACT FROM AUTHOR]