Non-invasive quantitative plaque analysis by coronary CT angiography in predicting coronary hemodynamic significance / 中华放射学杂志

ABSTRACT

Objective To evaluate the diagnostic performance of the automated quantitative analysis by coronary computed tomography angiography (CCTA) for lesion specific hemodynamic significance assessed by fractional flow reserve(FFR). Methods One hundred and fifteen patients with one hundred and fifty?one vessels,who successively underwent invasive coronary angiography with evaluation of FFR(values≤0.8 were defined as lesion specific hemodynamically significant), were analyzed by coronary CT angiography. FFR≤0.80 was found in 54(35.76％) of the 151 vessels, which was divided into two groups (group of hemodynamically significant and group of hemodynamically non-significant). CCTA images were quantitatively analyzed with automated software to obtain the following indexminimal lumen diameter(MLD), maximum diameter stenosis(MDS％), minimal lumen area(MLA), maximum area stenosis(MAS％), lesion length (LL), total plaque volume(TPV), total plaque burden(TPB), calcified plaque volume(CPV), calcified plaque burden (CPB), non-calcified plaque volume(NCPV), non-calcified plaque burden(NCPB), lipid plaque volume(LPV), lipid plaque burden(LPB), fibrous plaque volume(FPV), fibrous plaque burden(FPB), napkin-ring sign(NRS), remodeling index(RI) and eccentric index(EI). Logistic regression and area under the receiver operating characteristics were used for statistical analysis. Results MDS％(65.04％± 8.20％), MAS％(73.91％± 7.58％), TPB(57.96％± 11.17％), CPB[4.32％(0.11％, 5.34％)], LPB[14.89％(9.30％, 19.23％)], CPV[30.68 (0.29, 33.36)mm3], LPV[(81.72(33.92, 94.68)mm3]in the group with hemodynamic significance were larger than those in group with normal hemodynamic status[58.27％± 9.50％, 64.83％± 8.31％, 53.88％± 11.77％, 2.05％(0.00％, 3.42％), 11.83％(6.34％, 16.8％), 12.53(0.00, 13.24)mm3, 60.71(24.1, 75.11)mm3, respectively], which was statistically significant(t=4.41,P<0.01;Z=6.63,P<0.01;t=2.08,P<0.05;Z=-2.47,P<0.01;Z=-2.30,P<0.05;Z=-2.48, P<0.01;Z=-2.55, P<0.01, respectively). MLD[1.24(1.04, 1.46)mm]and MLA[3.61(2.40, 4.80) mm2]in the group with hemodynamic significance were smaller than those in group with normal hemodynamic status[1.53(1.32,1.72)mm, 5.28(4.00,6.40)mm2],which was statistically significant[Z=-4.82,-5.40, respectively;P<0.01].In logistic regression analysis, only MAS％(OR1.08,95％CI1.01-1.15,P=0.02), CPB (OR1.16,95％CI1.02-1.33,P=0.02) and LPB(OR1.10,95％CI1.01-1.19,P=0.02), MLA(OR0.69, 95％CI0.49-0.98,P=0.04)were significant predictors of hemodynamic significance. For predicting lesion specific hemodynamic significance, compared with MLA(0.76), MDS％(0.71), CPB(0.62) and LPB(0.61), except for MLA(Z=0.77, P=0.44), the AUC of MAS％(0.79) was significantly increased(Z=2.54, P=0.01;Z=2.91, P<0.01;Z=2.94, P<0.01, respectively). However, combination of other index to MAS％[MAS％+MLA％(0.81), MAS％+MDS％(0.80), MAS％+TPB(0.80), MAS％+CPB(0.80), MAS％+LPB(0.81)] did not show significantly difference over MAS％(Z=1.10, 0.71, 0.40, 0.54, 1.07, respectively;P>0.05). Conclusion Compared with diameter stenosis, area stenosis substantially improves the prediction of lesion specific hemodynamic significance.