A dosimetric phantom study of dose accuracy and build-up effects using IMRT and RapidArc in stereotactic irradiation of lung tumours

The study was executed with two heterogeneous phantoms with targets of ？1.5 and ？4.0？cm. Dose distributions in the simulated tumours delivered by small sliding window apertures (SWAs), IMRT and RapidArc treatment plans were measured with radiochromic film. Calculation algorithms of pencil beam convolution (PBC) and anisotropic analytic algorithm (AAA) were used to calculate the corresponding dose distributions.Peripheral doses of the tumours were decreased as SWA decreased, which was not modelled by the calculation algorithms. The smallest SWA studied was 2？mm, which reduced the 90% isodose line width by 4.2？mm with the ？4.0？cm tumour as compared to open field irradiation. PBC was not able to predict the dose accurately as the gamma evaluation failed to meet the criteria of ±3%/±1？mm on average in 61% of the defined volume with the smaller tumour. With AAA the corresponding value was 16%. The dosimetric inaccuracy of AAA was within ±3% with the optimized treatment plans of IMRT and RapidArc. The exception was the clinical RapidArc plan with dose overestimation of 4%.Overall, the peripheral doses of the simulated lung tumours were decreased by decreasing the SWA. To achieve adequate surface dose coverage to small lung tumours with a difference less than 1？mm in the isodose line radius between the open and modulated field, a larger than 6？mm SWA should be used in the dose delivery of SLRT.