Session Item

Because the 1.5T MR-linac allows for online plan adaptation to the daily anatomy, the exact positioning of the patient on the couch is less relevant than in conventional non-adaptive radiotherapy. Therefore, skin marks can be omitted to improve patient experience. Because we apply full plan adaptation in SBRT prostate treatments (adapt-to-shape, ATS) we have stopped the use of tattoos in this group. However, in conventionally fractionated treatment for high risk disease, translation only corrections (adapt-to-position, ATP) are applied and so pelvic rotations are more critical. We therefore investigated pelvic rotations with and without tattoos.

Clinical MR scans (T2 3D TSE) of 26 prostate patients treated on a 1.5T MR-linac (Unity system, Elekta AB) were used to retrospectively measure the pelvic bone rotations differences relative to the pretreatment MR scans. Pelvic rotations were measured in 20 patients treated with SBRT (100 fractions, ATS) without tattoos and 6 patients (92 fractions, ATP) with tattoos. The Monaco treatment planning system allows a cranial-caudal (CC) offset of 5 cm and 3 cm left-right and dorsal-ventral between pre-treatment and treatment isocenter position. Patients therefore need to be positioned daily to within this margin on the couch. The palpable most cranial part of the pelvic bone and a fixed knee support (Macromedics Holding B.V.) at a certain index were chosen for longitudinal setup without tattoos. For lateral setup, alignment of body midline (belly button-nose) with laser was chosen, together with the fixed knee support.

The location of the knee support and the origin marking were set during pre-treatment MR simulation (using a MR only preparation).

Pelvic rotations in both patient groups were measured in the offline Monaco planning system, using a volume of interest (VOI) around the pelvic bones (figure 1).

Both patient groups had successful treatments with non-significant patient setup time differences (total setup time around 4 minutes). The involved 28 radiotherapy technicians got used to the new workflow within 10 patients, both during simulation and treatment. Two index points (knee support and pelvic bones) were necessary to reproduce the setup within the 5cm longitudinal setup margin, because patients have difficulty to feel the exact position of the knee fix.

The obtained rotations are summarized in table 1. The X-axis rotation spread is somewhat larger without skin markers (SD = 0.5 vs 1.3 degrees, p < 0.05) but the 95% confidence interval remained within our clinical acceptance range of 3 degrees. The spread for the Y- and Z-axis rotations were larger with tattoos but the difference was not statistically significant.

Even though we found slightly larger X-axis pelvic rotations without skin markers, confidence intervals without tattoos remained within 3 degrees and we consider removing tattoos for all prostate patients treated on the MR-linac.