Estimating Proton Stopping Power Ratio of a Novel Iodinated Rectal Hydrogel Spacer Using Dual Energy Computed Tomography

Abstract

A novel iodinated rectal hydrogel spacer (I-RS) has been developed to make the gel radiopaque on CT compared to the standard rectal hydrogel spacers (S-RS). As the iodine in I-RS strongly absorbs kV photons, the accuracy of currently employed CT calibration built into a single energy CT (SECT) becomes inaccurate to determine the proton stopping power ratio (PSR) of the injected gel. Direct measurements of PSR mediated by I-RS with proton beams has been reported; however, those measurements were conducted under ideal conditions without fully accounting for density heterogeneity mediated by air bubbles co-injected with the gel during placement. This study used dual-energy CT (DECT) to estimate I-RS PSR in patients undergoing CT simulation for prostate cancer (PCa) radiotherapy (RT).Sequential scans at 90 and 140 kVp were acquired for 7 intact PCa (I-RS n = 5, S-RS n = 2) on a Philips Brilliance scanner. PSR was reconstructed following the method proposed by Torikoshi and improved by Yang, and compared to similar measurements when I-RS was replaced with S-RS. A pre-calculated table correlating effective atomic number with all possible Hounsfield Unit (HU) combinations from two sequential scans was implemented to accelerate reconstruction. PSRs were compared between DECT and SECT, and between I-RS and S-RS.In gel, HU measurements were heterogeneous (range -372 - +358). The mean PSR on reconstruction when using DECT was 0.966 (0.940-0.980) for I-RS. Replacement of I-RS with S-RS HU density resulted in a mean PSR of 0.937 (0.926-0.941) by DECT. PSR for I-RS on SECT was 1.107, and 1.014 for S-RS. The estimated difference between PSR in I-RS and S-RS on DECT vs. SECT were 14% and 8%, respectively.These data demonstrate that the use of I-RS in proton RT treatment planning could overestimate proton range due to underestimating PSR when CT simulation is performed using standard SECT. Notably, PSR absolute value reported here is different compared to previously reported in vitro data under ideal conditions and may be related to differences in gel heterogeneity in vivo. Heterogeneity is highlighted by the wide range of gel HU measurements. Direct measurements of PSR will be obtained in a planned future experiment. These data suggest that PSR overestimation due to I-RS should be considered when planning PCa treatment using proton therapy for patients who get I-RS to reduce treatment toxicity.