CT angiography in stent-graft sizing: impact of using inner vs. outer wall measurements of aortic neck diameters.

Purpose: To assess the impact of using inner wall vs. outer wall measurements on stent-graft sizing for endovascular repair of abdominal aortic aneurysm (AAA). Methods: Preoperative static and electrocardiographically-gated 64-slice computed tomographic angiography (CTA) datasets were acquired on 40 consecutive AAA patients (29 men; mean age 78.9 +/- 6 years, range 75-89). On static images, the aortic neck diameters were manually measured twice by 2 readers at 3 clinically relevant levels (supra-, juxta-, and infrarenal). The measurements were obtained from intima-to-intima (inner wall) and from adventitia-to-adventitia (outer wall). Dynamically reconstructed scans were also reviewed in each phase of the cardiac cycle to identify inner and outer minimum/maximum diameters. Using inner and outer wall measurements performed on static images, readers selected the size of a stent-graft that required inner neck diameter measurements and then one that required outer wall diameters. To calculate the relative oversizing, each selected stent-graft size was compared to that obtained using dynamic measurements. Oversizing <4% or >30% was considered inadequate. Results: Mean variations for the inner and outer wall diameters of 9.75%+/- 4.01% and 8.66%+/- 3.71%, respectively, were recorded on static CTAs; the absolute changes in diameters were 1.82 +/- 0.63 mm and 1.91 +/- 0.64 mm, respectively. No statistically significant differences were found relative to aortic pulsatility at the 3 levels in the neck for the inner or outer wall diameters. Significant variability was seen between inner (mean 20.8 +/- 3.4 mm) vs. outer (mean 23.7 +/- 4.3 mm; p<0.05) wall diameters. Stent-graft sizes significantly changed on the basis of the measurement method and device; for example, using the outer diameter to size a stent-graft that requires an inner diameter reference changed 36% of the selected stent-graft sizes, with similar to 20% being excessively oversized. Conversely, using the inner diameter to size an outer-diameter-based stent-graft resulted in nearly 40% of the sizes being altered. Based on dynamic measurements, the changes were more dramatic: the oversizing was considered excessive in up to 90% of patients if the measurement method did not match the stent-graft's stipulated reference. Conclusions: These data suggest that stent-graft sizing should follow the manufacturer's recommendations for using inner or outer diameter references based on dynamic patterns (mean value between diastolic and systolic diameters suggested). J Endovasc Ther. 2011; 18: 280-288