Abstract No. 227 Use of dedicated planning and guidance software results in radiation dose reduction in prostate artery embolization

Abstract

Complexity and variability of pelvic arterial anatomy and the need to catheterize tortuous, small vessels often results in high radiation dose to patients and staff during prostate artery embolization (PAE). Cone-beam CT (CBCT) has been shown to provide relevant 3-dimensional information to plan and guide PAE, but this benefit is mitigated by the intraprocedural time necessary for physicians to review the patient’s vascular anatomy. The purpose of this study is to investigate whether the use of a PAE planning and guidance software has an impact on radiation dose. Routine PAE procedures performed by a single physician from December 2019 to September 2020 were retrospectively analyzed. All procedures were performed in angio-interventional suite (Discovery IGS740, GE Healthcare, Chicago, IL), using the same acquisition protocol and dose settings. The new software (EmboASSIST, GE Healthcare) was used for procedure planning and guidance in every PAE since its installation at our institution in June 2020. Using virtual injection technology, the pelvic arterial anatomy was automatically segmented, and prostatic arteries were identified on CBCT and overlaid as a 3D model onto live fluoroscopy. Augmented fluoroscopy was used in all cases to facilitate navigation. The automated radiation reports were retrospectively reviewed and dose area product (DAP), cumulative air kerma (CAK) and fluoroscopy time were analyzed. The number of digital subtraction angiography (DSA) runs and CBCTs were also recorded. Data was compared between cases with and without use of the software using Mann-Whitney test. 20 consecutive PAE cases were included in the study, 10 cases before the software installation and 10 cases after. In terms of patient age (average 69 before vs 73 after), number of DSAs (8.0 and 8.4) and CBCTs (4.5 and 4.7) no significant difference was found. The patient BMI, average of 24 kg/m2 after and 27 kg/m2 before, was significantly different (P = 0.02). The median DAP was significantly decreased at 47 Gy.cm2 with software use versus 97 Gy.cm2 without (P = 0.02), whereas no significant difference was detected in CAK and fluoro time, 0.39 Gy versus 0.656 Gy, 55.3 min versus 55.4 min, respectively. Bilateral success was 90% when using the software versus 70% without. Use of dedicated CBCT-based planning and guidance software during PAE facilitated pelvic arterial assessment and prostatic artery navigation and may have contributed to significant reduction in radiation dose. These preliminary results prompt the need for further investigation into this technology.