Evaluation of a Model-based Dose Calculation Algorithm (Acuros) for Intraosseous High Dose Rate Brachytherapy (HDR) With Intramedullary Nailing for Bone Metastases

Abstract

A standard practice for patients with long bone metastases after intramedullary nailing is treatment with postoperative two-dimensional external beam radiotherapy (2D EBRT), delivered to the length of the implanted hardware. Our goal is to use intraoperative, intraosseous HDR brachytherapy delivered prior to intramedullary nailing to deliver an ablative dose to a limited volume immediately surrounding the tumor. This work evaluates the dosimetric effects resulting from treatment through a metal intramedullary drill bit, which in turn, dramatically reduces the volume potentially harboring subclinical contamination from the drill path and subsequent intramedullary nail placement. A model-based dose calculation engine was commissioned using the IROC data-set. Additionally, the accuracy of dose calculations was verified at several distances from the source in water using a custom 3D-printed phantom with a series of concentric rings holding Gafchromic-EBT3 film. A separate set of films from the same batch was calibrated with a 6MV-LINAC beam. Treatment plans were generated for two different source configurations (a single dwell position and a linear array of dwell position) with a standard 4.7Fr catheter surrounded by the stainless-steel drill bit. Measurements were performed at 1cm, 2cm, 3cm, 4cm, and 5cm from the active source and compared with dose calculations from the TG-43 and Acuros algorithms. Without the drill bit surrounding the catheter, measurements were in good agreement with the dose calculated by the treatment planning system (<1.0% on average). The maximum dose deviation (1.5%) was observed at 1cm from the source, as expected due to the steep dose gradient. With the stainless-steel drill bit surrounding the catheter, the Acuros dose calculations agreed with film measurements (<1.5% on average) with the largest difference at 1cm (<2.5%). Using TG-43 for dose calculation, we observed larger deviations from measurements (average 6.3%). At larger distances, 5cm, the agreement is within <1% of the calculated dose by Acuros. A model-based dose calculation engine was commissioned using a consensus dataset. This dose engine was secondarily verified with film measurements about a linear catheter surrounded by a stainless-steel intramedullary drill bit for the purpose of facilitating accurate dose calculations for intraoperative HDR brachytherapy of long bone metastases.