Bolus-dependent dosimetric effect of positioning errors for tangential scalp radiotherapy with helical tomotherapy

The dosimetric effect of errors in patient position is studied on-phantom as a function of simulated bolus thickness to assess the need for bolus utilization in scalp radiotherapy with tomotherapy. A treatment plan is generated on a cylindrical phantom, mimicking a radiotherapy technique for the scalp utilizing primarily tangential beamlets. A planning target volume with embedded scalplike clinical target volumes (CTVs) is planned to a uniform dose of 200cGy. Translational errors in phantom position are introduced in 1-mm increments and dose is recomputed from the original sinogram. For each error the maximum dose, minimum dose, clinical target dose homogeneity index (HI), and dose-volume histogram (DVH) are presented for simulated bolus thicknesses from 0 to 10mm. Baseline HI values for all bolus thicknesses were in the 5.5 to 7.0 range, increasing to a maximum of 18.0 to 30.5 for the largest positioning errors when 0 to 2mm of bolus is used. Utilizing 5mm of bolus resulted in a maximum HI value of 9.5 for the largest positioning errors. Using 0 to 2mm of bolus resulted in minimum and maximum dose values of 85% to 94% and 118% to 125% of the prescription dose, respectively. When using 5mm of bolus these values were 98.5% and 109.5%. DVHs showed minimal changes in CTV dose coverage when using 5mm of bolus, even for the largest positioning errors. CTV dose homogeneity becomes increasingly sensitive to errors in patient position as bolus thickness decreases when treating the scalp with primarily tangential beamlets. Performing a radial expansion of the scalp CTV into 5mm of bolus material minimizes dosimetric sensitivity to errors in patient position as large as 5mm and is therefore recommended.