Comparison of transfemoral versus upper extremity access to antegrade branches in branched endovascular aortic repair

Objective: We studied the outcomes of transfemoral access (TFA) vs upper extremity access (UEA) for branched endovascular aortic repair (BEVAR). Methods: From January 2016 to October 2019, 152 consecutive patients underwent BEVAR under general anesthesia at a single institution. In 2018, an alternative approach to the antegrade branches using TFA compared with conventional UEA was introduced. The cohort was divided into TFA and UEA groups according to the access approach. The end points were technical success, adverse events (including perioperative stroke/transient ischemic attack), access complications, operation time, and radiation exposure. Results: The TFA group included 60 patients (63% male; median age, 71 years; interquartile range [IQR], 65-76 years). The UEA group included 92 patients (67% male; median age, 73 years; IQR, 66-78 years). The number of target vessels (TVs) was similar in both groups (median, 4.0 TVs per procedure; range, 1-7 TVs for both). Technical success was greater in the TFA group (60 of 60 patients; 209 of 209 TVs) than in the UEA group (87 of 92 patients; 334 of 346 TVs; P <.01). The fluoroscopy time (median, 69 minutes; IQR, 48-87 minutes; vs 88 minutes; IQR, 65-104 minutes; P =.39) and contrast agent volume (median, 141 mL; IQR, 123-165 mL; vs median, 130 mL; IQR, 101-157 mL; P =.34) were similar in both groups. The radiation exposure (221 Gy × cm²; IQR, 138-406 Gy × cm²; vs median, 255 Gy × cm²; IQR, 148-425 Gy × cm²; P =.05) was lower and the operation time (median, 300 minutes; IQR, 240-356 minutes; vs median, 364 minutes; IQR, 290-475 minutes; P =.01) was shorter in the TFA group. Brachial access complications (0 of 60 vs 3 of 92 patients) and perioperative strokes/transient ischemic attacks (0 of 60 vs 8 of 92 patients) only occurred in the UEA group (P =.018). Conclusions: The use of TFA to catheterize antegrade branches was associated with a lower rate of complications in the present study and has become our preferred approach for BEVAR.