Background Preservation of ovarian perfusion is a critical consideration during uterine artery embolization (UAE), as inadvertent disruption can reduce ovarian reserve. Given the complex and variable anatomy of ovarian blood supply, this study aimed to characterize its anatomical origins and variations using digital subtraction angiography (DSA) in patients undergoing UAE for cesarean scar pregnancy (CSP). Methods This retrospective study included patients with type II or III CSP who underwent first-time UAE at Heibei Maternity Hospital, China, between August 2020 and October 2024. Pre-embolization DSA, including abdominal aortography and selective uterine arteriography, was analyzed to assess visualization of left and right ovarian arteries, uterine artery ovarian branches, and ovarian parenchymal staining. Statistical comparisons between bilateral vessels were performed using chi-square tests. Results Among 243 patients, the left ovarian artery was visualized in 42.8% and the right in 44.86%. Bilateral ovarian arteries were visualized in 18.52%, and neither artery was visualized in 20.58%. Uterine artery ovarian branches were visualized in 39.51% ( n = 96, left) and 38.27% ( n = 93, right), with parenchymal staining observed in 70.83% ( n = 68) and 67.74% ( n = 63) of these, respectively. No statistically significant differences were found between sides for ovarian arteries or uterine branches ( P 0.05). Complete visualization of all four potential vascular sources occurred in 8.23% of cases, while complete absence was seen in 13.58%. Conclusion Ovarian blood supply demonstrates substantial interindividual variation, with uterine artery ovarian branches contributing significantly to perfusion. No significant bilateral differences were identified, indicating that variability arises at the individual rather than side-specific level. Comprehensive bilateral vascular assessment with DSA remains essential to guide embolization strategies that minimize the risk of ovarian dysfunction.
Zheng et al. (Thu,) studied this question.