PO08

Abstract

Purpose Brachytherapy (BT) is an essential pillar in the treatment of cervical cancer. One method of gynecological (GYN) BT uses a transperineal catheter-guiding template in combination with an intrauterine (IU) tandem. Typically, catheter locations are decided using pre-BT imaging, and radiation treatment plans are then determined based on imaging taken after catheter insertion. Due to changes in patient anatomy once the IU tandem is inserted, this approach can lead to suboptimal tumor coverage, and often unused catheters, which contribute to increased implantation time, unnecessary tissue injury, and increased risk of bleeds. Images taken after insertion of the Template and APPlicator (post-TAPP) used in conjunction with a simple geometric catheter placement optimization algorithm may result in fewer unused catheters with better or equivalent dosimetry. In previous studies on prostate BT, the use of a Centroidal Voronoi Tessellation (CVT) algorithm for catheter optimization led to equivalent or superior treatment plans using fewer IS catheters compared to clinical cases. This work aims to verify these findings for cervical cancer BT. Materials and Methods Cases of locally advanced cervical cancer treated from 2016 to 2020 using IS BT with a Syed Neblett template were selected (N=12). Post-insertion imaging with target, organ-at-risk (OAR), and clinical catheter delineations were retrieved from the first BT fraction for each patient. CVT was used to simulate post-TAPP optimized catheters for each case by uniformly distributing catheters throughout a 2D projection of the target volume. The number of catheters in the CVT arrangement was equal to the number of catheters in the clinical implant, or the number of available template positions in the CVT planning region (defined as the high risk clinical target volume [HR-CTV] minus OARs projected onto the template plane), whichever number was lower. Treatment plans were then manually generated by a single expert for both the CVT and clinical catheter arrangements. Plan acceptability was evaluated via compliance with the EMBRACE-II dose-volume histogram limits in equieffective dose in 2 Gy per fraction (EQD2). Measures of dose inhomogeneity (%V150HR-CTV and %V200HR-CTV) were also recorded. Results The mean time for catheter optimization using CVT was 11.49 s. In all cases but 1, the CVT arrangements led to improved or EMBRACE-II compliant treatment plans with as many or fewer inserted catheters compared to the clinical cases (Figure 1). An increase in mean D90HR-CTV, D98HR-CTV, and D98IR-CTV was observed in the CVT group compared to the clinical group, as well as a statistically significant 2.2 Gy increase in mean D98GTVres (p < 0.05). A 2.2% increase in mean %V150HR-CTV was observed in the CVT group (p < 0.05). The mean number of unused catheters decreased from 4 in the clinical plans to 0 in the CVT plans (p < 0.001), and the acceptability rate increased from 50% with clinical catheters to 66.67% with CVT catheters. Conclusions This retrospective work supports the hypothesis that CVT with post-TAPP imaging can achieve clinically acceptable dosimetry while limiting insertion of extraneous catheters for cervix cases using the IU and Syed Neblett template. Future work will increase the cohort size and further quantify the effect of lowering the number of CVT catheters. Brachytherapy (BT) is an essential pillar in the treatment of cervical cancer. One method of gynecological (GYN) BT uses a transperineal catheter-guiding template in combination with an intrauterine (IU) tandem. Typically, catheter locations are decided using pre-BT imaging, and radiation treatment plans are then determined based on imaging taken after catheter insertion. Due to changes in patient anatomy once the IU tandem is inserted, this approach can lead to suboptimal tumor coverage, and often unused catheters, which contribute to increased implantation time, unnecessary tissue injury, and increased risk of bleeds. Images taken after insertion of the Template and APPlicator (post-TAPP) used in conjunction with a simple geometric catheter placement optimization algorithm may result in fewer unused catheters with better or equivalent dosimetry. In previous studies on prostate BT, the use of a Centroidal Voronoi Tessellation (CVT) algorithm for catheter optimization led to equivalent or superior treatment plans using fewer IS catheters compared to clinical cases. This work aims to verify these findings for cervical cancer BT. Cases of locally advanced cervical cancer treated from 2016 to 2020 using IS BT with a Syed Neblett template were selected (N=12). Post-insertion imaging with target, organ-at-risk (OAR), and clinical catheter delineations were retrieved from the first BT fraction for each patient. CVT was used to simulate post-TAPP optimized catheters for each case by uniformly distributing catheters throughout a 2D projection of the target volume. The number of catheters in the CVT arrangement was equal to the number of catheters in the clinical implant, or the number of available template positions in the CVT planning region (defined as the high risk clinical target volume [HR-CTV] minus OARs projected onto the template plane), whichever number was lower. Treatment plans were then manually generated by a single expert for both the CVT and clinical catheter arrangements. Plan acceptability was evaluated via compliance with the EMBRACE-II dose-volume histogram limits in equieffective dose in 2 Gy per fraction (EQD2). Measures of dose inhomogeneity (%V150HR-CTV and %V200HR-CTV) were also recorded. The mean time for catheter optimization using CVT was 11.49 s. In all cases but 1, the CVT arrangements led to improved or EMBRACE-II compliant treatment plans with as many or fewer inserted catheters compared to the clinical cases (Figure 1). An increase in mean D90HR-CTV, D98HR-CTV, and D98IR-CTV was observed in the CVT group compared to the clinical group, as well as a statistically significant 2.2 Gy increase in mean D98GTVres (p < 0.05). A 2.2% increase in mean %V150HR-CTV was observed in the CVT group (p < 0.05). The mean number of unused catheters decreased from 4 in the clinical plans to 0 in the CVT plans (p < 0.001), and the acceptability rate increased from 50% with clinical catheters to 66.67% with CVT catheters. This retrospective work supports the hypothesis that CVT with post-TAPP imaging can achieve clinically acceptable dosimetry while limiting insertion of extraneous catheters for cervix cases using the IU and Syed Neblett template. Future work will increase the cohort size and further quantify the effect of lowering the number of CVT catheters.