Session Item

Nontarget dose should be minimized as it offers no therapeutic benefit. The currently available treatment planning systems (TPSs) are not specifically designed nor intended for accurate dose calculations in regions outside the treated volume. This work presents the rigorous evaluation of RayStation's dose calculation accuracy and reliability in nontarget low dose regions.

The collapsed cone (CC) (v5.2) algorithm implemented in the RayStation (v6) TPS was evaluated using a PTW microDiamond (mD) detector and an Elekta linear accelerator. Different scenarios were investigated, from simple static fields to intensity modulated radiotherapy (IMRT) and volumetric modulated arc therapy (VMAT) clinical treatment plans. Clinical treatment plans were recalculated on the CT-data set of a nearly water equivalent pelvic phantom using the same beam arrangements and control points. Discrepancies and their confidence limits (CLs) between results of calculations and measurements – applying both local (δ) and global (Δ) normalization – were determined for various points of interest (POIs). CLs were calculated as the sum of the absolute value of the mean discrepancy and the standard deviation of the discrepancies multiplied by a factor of 1.96. Results were based on a single-institution experience for one clinical test case (prostate) and evaluated against internationally accepted criteria. Classification of nontarget dose was based on the definitions proposed by AAPM TG 158.

The mD was shown to be suitable for dosimetry in nontarget dose regions. The δ for simple static fields with collimator at 45° depended upon whether jaws were in place (-21.7% ± 29.4%) or retracted (15.9% ± 35.4%) (p-value < 0.01). POIs receiving < 5.0% of the prescribed dose failed the recommend tolerance level for CLδ (40%) as compared to POIs receiving ≥ 5% (88.6% versus 21.7% for IMRT and 62.6% versus 13.6% for VMAT). In addition, the CLδ for POIs that receive only indirect radiation – and can thus be categorized as truly “out-of-field” – as compared to “in-field” nontarget POIs was also above the tolerance limit (40%) for both IMRT (85.4% versus 13.1%) and VMAT (69.2% versus 23.0%). The CLΔ was below the tolerance limit (4%) for both techniques. Overall, no action levels (7%) were exceeded.

This study provides data about the level of dose calculation accuracy and reliability that can be expected for modern prostate radiotherapy in nontarget regions with the CC dose calculation algorithm implemented in the RayStation TPS. Since the measured and calculated doses differed no more than the action level value, our TPS can be considered acceptable for clinical use. However, POIs located in the < 5% dose region or receiving only indirect doses push the algorithm to its limits. A dose calculation algorithm that can accurately compute nontarget doses could be of great value to more comprehensively understand the clinical impact of these low doses.