ESTRO 2024 - Abstract Book

S348

Beachytherapy - Physics

ESTRO 2024

In brachytherapy (BT) treatment planning for cervical cancer it is common to use manual methods. In the guidelines from ESGO/ESTRO/ESP, on the management of patients with cervical cancer, it is stated that it is important to adhere to the pear-shaped dose distribution with a high central dose and to avoid long dwell times in the interstitial part of the implant [1]. There are also recommendations on the distribution of the dwell times between the different parts of the implant, ideally approximately 50% in the tandem, 35-45% in the vaginal part (in this case a ring), and 0-20% in the catheters [2]. With these things in mind, it is also important to achieve a high dose to the target volume and to spare the organs at risk (OAR), according to published dose planning aims [2]. Manual BT treatment planning for cervical cancer is a demanding task, which requires substantial training, and treatment planning can be time-consuming. An automated method using the available optimisation tools in the treatment planning systems (TPS) could ease the implementation of the method and ease the learning curve for new treatment planners. The clinical treatment planning systems do not all include tools to be able to control the distribution of dwell times in the different parts of the implant and do not produce treatment plans that fulfill all the different demands. To be able to control the distribution of the dwell times even without these tools, we introduce an automated treatment planning method using pseudo-structures which enables the optimisation to be done in one step. Treatment data from twelve patients previously treated for cervical cancer was used in the comparison. Six patients with intracavitary (IC) implants and six with combined intracavitary and interstitial implants (IC+IS). Three treatment plans were compared for all patients. The first was the treatment plan used clinically and was achieved by adjusting the dwell times manually. A second treatment plan was obtained using the VEGO optimizer in BrachyVision v16 (Varian medical systems Inc., Palo Alto, USA) optimizing on the clinical structures, the target volume (GTV and high-risk CTV) and the OARs (rectum, bladder, small bowel and sigmoid), called straightforward method. A third treatment plan was performed using VEGO optimizer in BrachyVision using the clinical structures and additional pseudo-structures, called pseudo-structures method. The pseudo-structures are used to control the distribution of the dwell times and are done in a couple of minutes using the operators in the contouring module in BrachyVision. First, a pseudo-structure around the CTV HR was used to decrease the dose to the surrounding tissue and to reduce the contribution from the interstitial part of the implant. Then, a structure around the ring part of the implant (excluding the OARs with a margin) was included to increase the contribution from this part. All the automated treatment plans were done in one single step, not allowing any individual patient tuning. The prioritizing was done to harmonize with the priority order in Tanderup et al. [3]. The treatment plans were compared using dose-volume histogram (DVH) parameters in accordance with the planning aims recommendations from Tanderup et al. [3]. They were also evaluated on the distribution of the dwell times in the different parts of the implant. Material/Methods:

Results:

Figure 1 shows a boxplot with the DVH-parameters for all the treatment plans. The hard constraints for the target volumes are fulfilled in seven, ten and, eleven cases for the manual, pseudo-structures respectively straightforward method. The hard constraints for the OARs are fulfilled in nine, six and, six cases for the manual, pseudo-structures respectively straightforward method.

The distribution of the dwell times in the different parts of the implant is shown in Table 1.