Session Item

Evaluate the inter and intrafraction motion of patients with bone oligometastasis from pre- and post-treatment CBCTs, to determine systematic and random sources of error that can be corrected and accounted for.

For oligometastatic bone patients the current IGRT strategy is a pre-treatment CBCT (online bone match), followed by a post-treatment CBCT to check for intrafraction motion. Bony oligometastasis are all treated with 3 fractions, using a 3mm CTV-PTV margin and are immobilised with long vacuum bags. This work was approved as a departmental evaluation to determine intrafraction motion for use in the Marcel Van Herk (MVH) margin calculation for hypo-fractionated data. The interfraction setup was also evaluated.

The uncorrected pre-treatment CBCT shifts were used to calculate the group-wise mean [M] and systematic error [∑] to ensure M was small (< 1/3) compared to ∑, or to indicate if any processes were causing systematic errors. The post-treatment CBCT M, ∑ and random error [σ] were calculated. Again, M and ∑ were compared to ensure M was small (< 1/3) compared to ∑.  σ was recorded for use in margin calculation. Data were stratified to site and time to test for trends.

230 CBCTs (115 pre, 115 post) were retrospectively reviewed from 30 patients (37 treatment plans) treated from 2019-2020. Table 1 summarises the calculated M, ∑ and σ. The post-treatment results have been stratified into sites and image assessment time <9min 59sec and >10minutes. The overall uncorrected interfraction data shows that M is not small compared to ∑ (M= 0.9mm, ∑ 1.5mm) in the longitudinal direction; this will not have impacted the patient as this is corrected. The intrafraction data shows that M is small compared to ∑ and ∑ < 1mm in all directions. Because of this, and small sample size, ∑ is assumed to be zero for margin calculation. Assuming a delineation systematic error of 1mm, and random error as shown in Table 1, the calculated CTV-PTV margin is 2.6mm in all directions, with β = 0.84 for hypofractionation and penumbra 0.3.

Due to the small sample in the stratified data, data should be reviewed with caution, however, the random error indicates it is appropriate to use the same margin within these sites. Data also supports current department practice to re-image if CBCTs take longer than 10 minutes to assess. Longer-term data will be analysed once sample sizes are bigger.  

A CTV-PTV margin of 3mm has been found to be appropriate for this treatment site, and the random error associated with intra-fraction motion is small. As a result of this study, a departmental change in practice has occurred to omit post-treatment imaging for this patient cohort. The evaluation has indicated there may be a potential systematic error in the treatment pathway impacting longitudinal setup which requires investigation.