Session Item

Pretreatment dose verification of stereotactic body radiotherapy (SBRT) is generally considered necessary because of its high accuracy requirements. These measurements come at the cost of extra work that must be balanced with their capability of detecting critical plans. Also secondary dose calculations (SDCs) can be used for finding (large) errors in the dose calculation of a plan, whereas the extra workload for using these SDC systems is limited.

In our clinic the question arose whether the number of pretreatment measurements for lung SBRT can be reduced. Whereas pretreatment measurements for lung SBRT are evaluated at g criteria of 3%/3mm, the 3D SDC is generally done at 5%/3 mm. The aim of the study was to develop a method to recalculate a large number of lung SBRT plans with the 3D SDC at the stricter 3%/3mm criteria and to judge if pretreatment measurements can be minimized.

Plans of patients treated with lung SBRT from January 2020 to August 2021 were taken from the clinical database. No other selection criteria were applied. The plans were made with Pinnacle TPS (Philips, USA) using VMAT arcs of 10 MV photons. All plans were verified pretreatment on a Delta4 phantom (ScandiDos, Sweden) and fulfilled our clinical g criteria of 3%/3mm with a passing rate ≥95% at a 40% dose threshold. Additionally, all plans passed our clinical 3D SDC (Mobius3D, Varian Medical Systems, USA) using our clinical g criteria of 5%/3mm with a passing rate ≥95% at a 40% dose threshold.

The selected plans were recalculated with the 3D SDC using criteria of 3%/3mm. In order not to interfere with the standard workflow, a dicom tag was modified during transfer from our radiotherapy PACS to the 3D SDC and that triggered the use of the tighter settings. Results were automatically stored.

136 unique lung SBRT plans, including patients with multiple lesions, were automatically recalculated at 3%/3 mm with the 3D SDC. 129 of the plans (95%) had a g passing rate ≥95%. Although the treatments were heterogenous in prescription and fractionation, the histogram of (monitor units)/(prescribed dose) had a relatively small distribution (see fig.). The plans that failed the criteria are indicated in the figure; 4 of the 7 failed plans (57%) were “in the tails” of the distribution.

A large number of lung SBRT plans was re-evaluated using our 3D SDC at the same criteria as for pretreatment measurements. The majority of plans (95%) passed these criteria. It is often recommended to do extra QA for plans that are more complex than the “standard” plans. However, by using (monitor units)/(prescribed dose) as plan complexity measure, 3 plans would have escaped extra QA.

By using tighter 3D SDC criteria potentially critical plans can be identified. We recommend a risk-based selection for pretreatment QA by measuring those plans that do not meet the pretreatment criteria in the 3D SDC and to combine this with sampling the rest of the plans. This will result in a vast reduction of work load.