Dosimetric comparison of multiple SBRT delivery platforms for pancreatic cancer

Abstract

Background Stereotactic body radiation therapy (SBRT) has been widely used for pancreatic cancer. However, there is still a lack of studies comparing the latest SBRT techniques in terms of clinical efficacy and safety. Objectives: This study aims to evaluate three latest SBRT delivery platforms: CyberKnife (CK), Tomography Radixact (TOMO), and Halcyon volume rotation intensity modulation therapy (VMAT) for the treatment of pancreatic cancer. Methods Sixteen patients with pancreatic cancer treated with CK were retrospectively analyzed. SBRT plans were designed using Precision and Eclipse software. CK plans were optimized in two forms: fixed collimator (CK-Fixed) and multi-leaf grating collimator (CK-MLC). TOMO plans were designed with 2.5 cm Fixed Jaw, pitch 0.123–0.43 and 4.0 modulation factors in precision system. In Eclipse 15.6 system, photon optimizer (OP) algorithm was used to design the coplanar two-arc Halcyon VMAT. The median radiation dose was 40 Gy (35–45 Gy) in 5 fractions. The effectiveness of clinical treatment was evaluated by comparing the homogeneity index ( HI ), conformity index ( CI ), coverage of the planning target volume (PTV) and dose distribution parameters of organs at risk (OAR). Results All plans met the limits of clinical target dose and OAR. CK-MLC plans had the lowest maximum dose of 2 cm normal tissue from PTV margin (D 2cm ), indicating a low risk of peripheral radiation damage. Additionally, the CK-MLC plans had the lowest dose parameters and provided the best protection for the kidney, spinal cord, small intestine, and duodenum, with a paired t-test p < 0.05, indicating a statistical difference. Conclusion High conformity and adjustability of CK-MLC allowed for precise complex target localization and conformal dose distribution, benefiting tumor treatment while maximally reducing damage to OAR. This study provides valuable dosimetric evidence for SBRT technique selection for pancreatic cancer.