3757 - First Large-Scale Analysis of Transposed Ovary Motion Using High-Resolution FBCT from a Chinese-Developed Integrated CT-Linear Accelerator: PRV Margin Optimization for Ovarian-Sparing Cervical Cancer Radiotherapy

Purpose/Objective(s):

Ovarian function preservation stands as a fundamental element in the quality-of-life care for young cervical cancer patients undergoing radiotherapy. Ovarian transposition (OT) can effectively decrease the radiation exposure. Nevertheless, the positional uncertainty of the transposed ovaries during fractionated treatment remains a crucial and unaddressed issue. Current literature faces triple limitations: 1) reliance on limited CT/MRI scans per patient with insufficient temporal sampling; 2) conventional KVCBCT and MVCT lack soft tissue contrast to clearly visualize transposed ovaries; 3) small sample sizes (n<10) and low imaging frequency (<5 scans/patient) fail to capture full 3D motion dynamics. These combined gaps mean no reliable data exists to guide PRV margin design for ovarian-sparing cervical cancer radiotherapy (OS-CRT).

This study aims to address these critical gaps by establishing the first large-scale, high-resolution dataset characterizing 3D transposed ovary motion using frequent 0.3-0.5mm spatial resolution fan-beam CT(FBCT). Our approach capitalizes on the system's unparalleled soft tissue visualization capability, which is unmatched by KVCBCT/MVCT, and high-frequency scanning to derive evidence-based PRV margins for OS-CRT.

Materials/Methods:

Twenty-five cervical cancer patients who had undergone radical surgery and bilateral OT received pelvic radiotherapy (45-50Gy/25-28F) on a novel radiation system, a China-developed integrated CT-linac system featuring 0.3-0.5mm isotropic spatial resolution FBCT. Unlike KVCBCT/MVCT which struggles with soft tissue visualization, this platform provides clear delineation of transposed ovaries, enabling reliable 3D motion analysis via 639 pre-treatment scans. Ovarian contours were independently reviewed by two senior radiation oncologists, and 3D centroid displacements were calculated via deformable image registration.

Results:

Transposed ovaries exhibited consistent 3D motion, with largest amplitude in the SI direction (3.3±0.9 mm), followed by AP (2.5±1.1 mm) and LR (2.0±1.0 mm). Random errors were 1.3 mm (SI), 1.2 mm (AP), 1.0 mm (LR); systematic errors were 1.8 mm (SI), 1.4 mm (AP), 1.1 mm (LR). Maximum displacements represented a 40%-65% reduction compared to cervical target motion. PRV margins of 5 mm (SI), 4 mm (AP), and 3 mm (LR) ensured >95% treatment coverage while reducing normal tissue exposure by 15%-20% compared to uniform margins. The superior soft tissue contrast of FBCT allowed consistent ovarian identification in all scans, a capability not achievable with conventional KVCBCT/MVCT.

Conclusion:

This study using frequent high-resolution FBCT establishes a comprehensive 3D motion dataset for transposed ovaries, addressing critical limitations of KVCBCT/MVCT and traditional CT/MRI studies. The derived asymmetric PRV margins enable personalized ovarian function preservation for young cervical cancer patients and inform adaptive radiotherapy strategies.