LAUSR

BACKGROUND Automated centerline (CL) measurements have been conventionally used for stent-graft length estimation during thoracic endovascular aortic repair (TEVAR). The purpose of this study was to assess the accuracy of greater curvature length (GL), semiautomated CL and straightened centerline length (SCL) for preprocedural planning in TEVAR. METHODS Immediate postprocedural CT Angiographies of 30 patients (22 males, age-49.2+/-10.1years) who underwent TEVAR between 2015 to 2017 were retrospectively analyzed. CL, GL, SCL and the straightline length(SL) were measured between proximal and distal ends of the stent-graft and results were compared with the true length of the stent-graft (TL). Tortuosity index (TI=CL/ SL) was calculated. RESULTS GL (17.92+/-4.78cm) was the closest in predicting the TL (17.75+/-4.29cm) (p=0.414) overall, as well as in both dissection and aneurysm subgroups (p=0.9). There was a significant difference between CL (16.67+/-4.07cm) and TL (p<0.0001) as well as between SCL (16.86+/-4.16cm) and TL (p=0.001). These differences were greater in dissection subgroup than in the aneurysm group (p<0.0001 and p=0.03 for TL-CL and TL-SCL, respectively). The extent of mismatch between GL or CL and TL did not correlate with tortuosity, but the difference between TL and SCL had a significant positive correlation with tortuosity (r=0.375, p=0.04). TL-GL had a negative linear correlation with the stent-graft length (TL) in the dissection group (r=0.50, p=0.03). CONCLUSION The greater curvature length predicts the actual total length of the deployed stent-graft more accurately than centerline or straightened centerline lengths. Hence, it should be used in planning for the length of stent-graft required for TEVAR.