Optimized Construction of Multi-Channel Brachytherapy Cylinder Applicators

Abstract

The purpose of this work is to identify an optimal construction for rigid multi-channel brachytherapy cylinder applicators, such that the clinical user can maximize lateral dose coverage while minimizing surface and normal tissue doses. Cylindrical multi-channel applicators are commonly used in HDR gynecological brachytherapy to treat the post-hysterectomy vaginal cuff. Literature suggests that adjuvant vaginal brachytherapy has high efficacy and low morbidity for intermediate-risk endometrial cancer [1]. Commercial vaginal cylinders may not have fully considered dosimetric consequences of the cylinder’s construction. Therefore we seek to provide manufacturers with guidance on the optimal radial offsets for peripheral channels within multi-channel cylinder applicators. Commercial multi-channel applicators typically consist of a cylindrical obturator with various diameter build-up caps. The obturator contains one central channel and two or more peripheral channels positioned at a fixed radial distance from the central channel. The radial distance from the central channel to peripheral channels affects the radioactive loading/weighting and usability of peripheral channels. For example, peripheral channels that are located very close to the patient surface cannot be loaded without quickly exceeding a mucosal surface dose limit (e.g. 200%). An ideal radial offset for peripheral channels can maximize asymmetric peripheral dose coverage of the target while maintaining surface dose <200% of prescription dose. This work simulated a multichannel cylinder consisting of a central channel and two lateral channels. A point source approximation was used. Initially, three Ir-192 VS200 sources were simulated for a vaginal cuff treatment, one source in each channel. Optimal dwell times were determined using the methods of Deufel and Furutani [2] to achieve 100% prescription dose 5 mm apically and 5 mm posteriorly from the cylinder surface, with mucosal surface dose <200%. (See Fig1 (a)) The radial distance of peripheral channels was varied from 0 to Rapplicator. The optimal radial distance, dlat, is the distance where lateral 100% dose coverage is maximum. This method was then extended to a vaginal cuff plus vault treatment for a cylinder with 6.5cm treatment length (Fig1 (b)). Dose optimization points were placed at the apex, along the cylinder surface and posterior 5mm along the treatment length to produce a homogenous dose. Results were obtained for cylinders with 3.0, 3.5, and 4.0 cm diameters. A comparison of point and line source results was also performed. Fig 1(c) demonstrates the lateral coverage as a function of dlat for a 3.5 cm diameter cylinder. The optimal lateral channel position is located at 2.8 mm, 4.1 mm and 5.2 mm away from the central channel for the cylinder in diameters of 3, 3.5 and 4cm and provides maximal lateral coverage of 5.7 mm, 6.3 mm and 7 mm from the applicator surface, respectively. Doses 5mm apically, 5mm posteriorly, and at the lateral surface are shown in Fig 1(d). The optimal radial distance for peripheral channels increases and maximum lateral coverage distance increases as cylinder diameter increases. Figs 1(e) and (f) provides results for cuff plus vault treatment. Compared to three-source loading, cuff plus vault treatment with more inferior loading can push dose more laterally (e.g. 7.4 mm vs. 4.1 mm for the 3.5cm applicator) This work proposes a method for optimized construction geometry of multi-channel vaginal cylinders. Manufacturers may use results to construct an applicator that provides maximum lateral coverage while maintaining normal tissue constraints. Furthermore, this work shows that a library of Atlas plans for vaginal cuff treatment may be efficiently and optimally generated using the methods of Deufel and Furutani.