Objectives:

To quantify the dosimetric uncertainties of using linear accelerator (LINAC) with 5 mm width MLC (Multi-leaf collimator) to treat multiple 7mm diameter lesions with single iso volumetric arc therapy (VMAT) stereotactic radiosurgery (SRS), using a custom-tailored SRS beam model for small lesions. MPPG9A specifies an end-to-end dosimetric evaluation using an SRS system to be within +/- 5% of measured versus calculated.

Methods:

A multi-lesion single iso VMAT SRS plan with 16 spherical PTVs with 7mm diameters was designed by positioning lesions throughout a flat phantom (25x25x15 cm) in a treatment planning system. Five film planes were placed in the treatment volume at ISO (0 cm), +/- 3 cm, and +/-5 cm for analysis. The positioning of the PTVs included targets at large off-axis distances to replicate multi lesion SRS cases that have planning target volumes (PTV) far from the isocenter.
The multi-lesion VMAT SRS plan was calculated with a custom-tailored dose calculation SRS beam model. Five radiochromic films (8”x10”) were placed in the designated film planes of the flat phantom, and the VMAT SRS plan was delivered with the 5mm MLC LINAC. The 5 films planes were analyzed with a 3%, 2mm gamma pass rate criteria in a 90% isodose region of interest. Dose difference was also analyzed for each film plane and each PTV target. To better delineate dose calculation accuracy, we attempted to minimize geometric errors caused by the machine during delivery by aligning the dose profiles to each PTV individually. This allowed us to look at the best possible fit for each lesion in each film plane.

Results:

The dose difference observed on ISO, 3 cm anterior, and 5 cm anterior film planes was ~2% on these 3 film planes with a respective gamma pass rates of 94.5%, 98.5%, 100.0%. Larger dose differences were found on the 3cm post film plane at 5.4% with a gamma pass rate of 88.5% and 5 cm post film plane at 3.3% and a gamma pass rate of 92.7%.
When analyzing the individual PTV targets attempting to remove the geometric error, the mean dose difference is less than 3% for PTVs less than or equal to 6 cm from the isocenter. Inside of the PTVs there was a local points error of up to +/- 120-160 cGy or ~8-10% dose difference inside the targets.

Conclusion(s):

The dose differences for PTVs were below MPPG 9.a +/- 5% tolerance. We commissioned treatment of PTV volumes down to 7 mm diameter with a maximum 6 cm from iso distance, for all our campus-wide 5 mm MLC LINAC machines.