ESTRO 37 Abstract book

ESTRO 37

S599

1 Radboud university medical center, Radiation Oncology, Nijmegen, The Netherlands

the left and from 1.2 to 8.0 Gy(RBE) for the right gland. The estimated NTCP values ranged from 20 % to 48 % (Table II) with corresponding differences in estimated NTCP values ranging between 1 % and 3 %.

Purpose or Objective MR simulation can facilitate target volume delineation for prostate irradiation. Registration of this MR to the planning CT is hampered by variations in endorectal balloon (ERB) positioning. In this study, the ERB position on MR was compared to planning CT and CBCT. Material and Methods ERB positioning was evaluated for 21 patients that underwent primary prostate irradiation in 28 fractions. Treatment simulation was done using CT and 3T MR imaging. The ERB was filled with 100 cc air during CT and treatment, and 100 cc water on MR. The MR protocol contained a 3D T1 VIBE sequence to visualize gold markers. For 20 patients, this scan was matched to the planning CT using a rigid marker match in Mirada Medical (Oxford, UK). For the remaining patient, this was impossible due to a missing marker. Most patients underwent online EPID during treatment. However, 5 patients underwent online CBCT. For these patients, the first CBCT of every week was collected, assembling a total of 35 CBCT sets. The target volume and organs at risk (including ERB) were manually delineated on the planning CT by a radiation oncologist. For this study, the ERB was also delineated on MR (manually) and CBCT images (region growing). All images were registered to the planning CT using the clinical match (rigid marker match for MR, bone plus mask match in XVI for CBCT). The center-of-mass of each ERB was determined and compared to the treatment isocenter. Results On MR, the mean deviations in ERB position w ith respect to CT were -0.3 ± 2.1 mm in left-right (LR), 0.0 ± 3.5 mm in anteroposterior (AP), and 8.9 ± 7.4 mm in superoinferior (SI) direction. For CBCT, the mean deviations were 0.1 ± 1.9, 0.1 ± 2.7, and -1.7 ± 5.3 mm, respectively. Box plots are shown in Figure 1.

Figure 1: Box plots showing median deviations and interquartile ranges for the ERB position.

Figure 2 shows MR and CT images for the patient with the largest deviation (22 mm) in balloon insertion depth (SI direction). Prostate, bladder and rectal wall were delineated on the matched MR as well. As can be seen from the dose overlay on CT, the prostate (CTV) was still covered with the ERB positioned as on MR. Minimum, mean and max dose for prostate were identical on CT and MR, as was the max bladder dose. Rectal max dose was 0.2 Gy higher for the MR delineation.

Conclusion Inter-observer delineation uncertainty resulted in substantial variations in mean parotid gland doses; however, the influence on the estimated NTCP values for xerostomia was modest. PO-1067 Endorectal balloon positioning on MR simulation for prostate cancer radiotherapy E.J.L. Brunenberg 1 , T.J. Scheenstra 1 , L.C.W. Bouwmans 1 , R.J. Smeenk 1