Lens dose in MLC-based IMRT treatments of the head and neck

Abstract

Purpose: The objectives of this work are: (1) to determine typical dose to the lens during step-and-shoot intensity modulated radiotherapy (IMRT) treatments of the head and neck, and (2) to report on the dose calculation accuracy of a commercial inverse planning system in predicting lens dose.

Methods and materials: The Corvus inverse treatment planning system (Nomos, Cranberry Township, PA) was used to plan IMRT treatments for patients with head-and-neck cancers in our clinic. Patients were treated on Varian C-series linacs (Varian, Palo Alto, CA) with 4-MV or 6-MV X-rays. A Rando phantom (Alderson Laboratories, Stamford, CT) was specially modified to accommodate 1 x 1 x 1 mm(3) thermoluminescent dosimeters at the position of the lens. The IMRT treatment plans were then delivered to the modified Rando phantom. The thermoluminescent dosimeter measurements were converted to dose and taken as an estimate of the lens dose. A total of 20 cases were used in this study (15 cases with 4 MV and 5 cases with 6 MV).

Results: Expressed as a percentage of the prescription dose, the mean dose to the left and right lens for all 4-MV cases was 9.1% (range, 2.0% to 61.3%). For the 6-MV cases, the mean dose to the left and right lens was 12.8% (range, 3.6% to 41.3%). For both the 4-MV and 6-MV cases, the case of maximum dose occurred when the IMRT treatment target included volumes superior to the level of the lens. The field size and number of monitor units did not correlate with the measured lens dose. The only factor of significance affecting lens dose was the inferior-to-superior distance of the target to the lens. For target-lens distance >/=6 mm, the maximum measured lens doses were 5.9% and 9.0% relative to the prescribed dose for the 4-MV and 6-MV beams, respectively. These data are similar to those observed in conventional head-and-neck treatments. For all cases, the difference between the dose measured and that predicted by Corvus was less than 2% and 4% of the dose prescribed to the gross tumor volume for the 4-MV and 6-MV cases, respectively.

Conclusion: In IMRT, factors such as leaf leakage and number of monitor units play a secondary role and are not more significant than what is observed in conventional head-and-neck treatment when the lens is shielded by the collimator jaws. The target-lens distance is the parameter that affects the lens dose most strongly. For cases where the tumor is at or above the level of the lens, the lens dose can amount to an appreciable fraction of the prescription dose. To keep the lens dose to a minimum, noncoplanar beams that enter or exit into the lens should not be used.