Experimental evaluation of the impact of different head-and-neck intensity-modulated radiation therapy planning techniques on doses to the skin and shallow targets.

PURPOSE To investigate experimentally the impact of different head-and-neck intensity-modulated radiation therapy (IMRT) planning techniques on doses to the skin and shallow targets. METHODS AND MATERIALS A semicylindrical phantom was constructed with micro-MOSFET dosimeters (Thomson-Nielson, Ottawa, Ontario, Canada) at 0-, 3-, 6-, 9-, and 12-mm depths. The planning target volume (PTV) was pulled back 0, 3, or 5 mm from the body contour. The IMRT plans were created to maximize PTV coverage, with one of the following strategies: (a) aim for a maximum 110% hotspot, with 115% allowed; (b) aims for a maximum 105% hotspot; (c) aims for a maximum 105% hotspot and 50% of skin to get a maximum 70% of the prescribed dose; and (d) aim for 99% of the PTV volume to receive 90-93% of prescribed dose, with a maximum 105% hotspot, and with the dose to the skin structure minimized. Doses delivered using a linear accelerator were measured. Setup uncertainty was simulated by intentionally shifting the phantom in a range of +/-8 mm, and calculating the delivered dose for a range of systematic and random uncertainties. RESULTS From lowest to highest skin dose, the planning strategies were in the order of c, d, b, and a, but c showed a tendency to underdose tissues at depth. Delivered doses varied by 10-20%, depending on planning strategy. For typical setup uncertainties, cumulative dose reduction to a point 6 mm deep was <4%. CONCLUSIONS It is useful to use skin as a sensitive structure, but a minimum dose constraint must be used for the PTV if unwanted reductions in dose to nodes near the body surface are to be avoided. Setup uncertainties are unlikely to give excessive reductions in cumulative dose.