Session Item

The Halcyon™ system has been engineered with treatment planning algorithms to simplify and streamline clinical workflow for online image-guided intensity-modulated radiotherapy.

Between June and August 2019, 46 patients with localized prostate cancer were treated in the Halcyon 2.0 linear accelerator with Volumetric Modulated Arc Therapy to a dose between 72-77 Gy (2.4-2.2 Gy/d in 30 or 35 fractions) with daily correction of the target position based on kvCBCT imaging.

The workflow in all patients were an online automatic matching, followed by a visual verification and manual adjustment when required, mainly focused in the proper location of the CTV, surrounded by the anterior wall of the rectum, the base of the bladder, the bulb of the penis and the lateral borders of the prostate.

Two automatic matching protocols were used. In the first 23 patients (group 1), the parameters predefined in the Halcyon™ system were apply for an automatic matching of the images. In the remaining patients (group 2), these parameters were modified, limiting the volume of interest to PTV + 1cm and prioritizing the soft tissue. Independent t-test and Mann-Whitney U test were used for comparing groups.

The association between automatic and manual adjustment matching was determined using linear regression.

A total of 1442 CBCT were analyzed, 715 in the group 1 and 727 in the group 2.

In both groups, the mean difference in the vertical (VRT) setup correction between automatic and manual adjustment was < 1mm, that was statistically significant (Group 1: -0.74mm 95%CI -0.89; -0.59; and Group 2: -0.87mm 95%CI -1.04; -0.71;p<0.001) (Table 1), being their linear correlation in this same axis strong (R²>0.7), higher in group 1 (Group 1 R²=0.840 and Group 2 R²=0.737) (Fig. 1a-b).

There were no differences in other axis.

The values of the corrections required between the automatic comparison and the manual adjustment are not different between the two groups (p=0.65) (Table 1 Fig. 1c).

We have observed that the 17 patients who had a mean difference of more than 2mm between automatic and manual matching (8 in group 1; 9 in group 2) had not undergone good rectal preparation during treatment. In the other 29 patients, random differences greater than 3mm were consistent with a poor rectal conditioning as well.

Due to the differences observed between automatic matching and manual adjustment on the VRT axis, the visual verification of the automatic matching should always be performed pretreatment.

Optimization of the automatic matching with specific guidance parameters doesn''t provide any improvement in the Halcyon system.

Poor rectal preparation of the patient leads to worse automatic matching.