Volume-Rendered Multidetector CT Angiography: Noninvasive Follow-Up of Patients Treated with Renal Artery Stents

Abstract

OBJECTIVE. The purpose of our study was to evaluate the role of multidetector CT (MDCT) angiography with volume rendering for estimating the patency of renal artery stents. SUBJECTS AND METHODS. In 16 patients, 16 renal artery stents were evaluated with MDCT renal angiography and digital subtraction angiography (DSA). CT data were evaluated using multiplanar volume reformations and the volume-rendering algorithm with three different volume-rendered parameter settings (low-to-high, high-to-low, and high—low—high opacity transfer functions: VR_LH, VR_HL, and VR_VE, respectively). Targeted images of each stent were rendered in paraaxial and paracoronal planes and were interactively interpreted. The overall restenosis severity was measured on postprocessed paraaxial and paracoronal images and compared with that obtained on DSA using linear regression analysis. Image quality and lumen delineation on rendered images were also compared using Wilcoxon's signed rank test. RESULTS. Eight restenoses were identified on DSA. Correlations between restenosis severity measured with DSA and those measured with MDCT were significant (p < 0.001). Volume rendering with VR_HL allowed the best correlation with DSA (reviewer 1, r² = 0.86; reviewer 2, r² = 0.94) and was significantly better than multiplanar volume reformations (p = 0.028). Overall image quality was high with all rendering techniques and with no significant differences (p > 0.59, for all comparisons). Stent lumen was well delineated with volume-rendering modalities; however, VR_HL was significantly better than VR_LH (p = 0.033). CONCLUSION. Volume-rendered MDCT angiography enabled high-quality three-dimensional reproducible evaluation of the patency of implanted renal artery stents. Volume rendering with VR_HL achieved the best performance.