Purpose: Anatomic variation has a significant dosimetric impact in intensity modulated proton therapy. Traditionally weekly or bi-weekly CT scans, called QACTs, are used to monitor the anatomic and the resultant dose change to determine if an adaptive plan is needed. Frequent CT scans result in unwanted patient dose and clinical workload. This work proposes a CBCT-based approach to reduce the frequency of QACTs by analyzing patient setup CBCTs.
Methods: we retrospectively analyzed 20 head and neck patients, associated with 885 CBCTs, 105 QACTs, and 22 adaptive plans. For each CBCT, water equivalent thickness (WET) along the pencil beam path between the target surface and the patient body was calculated. For each treatment plan, the WET of the first day CBCT (1st CBCT) was used as the reference, and the mean WET change in each CBCT from the 1st CBCT was used as the surrogate of proton range deviation. Using CBCTs acquired prior to a QACT, a linear regression model was generated to predict the mean WET change on QACT day. The impact of range deviation on target dose was calculated as (1) the predicted mean WET change weighted by the monitor units of each fields and (2) the ratio of mean WET change in (1) over the effective diameter of the target.
Results: When the estimated mean WET change was larger than 8 mm or 8% of the effective diameter of the target, a QACT was required, and otherwise if the change was less than 8 mm AND 8%, the QACT could be skipped. Overall, 40 QACTs could be reduced from 105 QACTs. At least one QACT could have been omitted in 18 out of the 20 cases.
Conclusion: This study shows that QACT frequency can be greatly reduced by calculating range deviation in patient setup CBCTs.
Not Applicable / None Entered.