Parotid gland tumors: a comparison of postoperative radiotherapy techniques using three dimensional (3D) dose...

To compare different treatment techniques for unilateral treatment of parotid gland tumors.

The CT-scans of a representative parotid patient were used. The field size was 9 x 11 cm, the separation was 15.5 cm, and the prescription depth was 4.5 cm. Using 3D dose distributions, tissue inhomogeneity corrections, scatter integration (for photons) and pencil beam (for electrons) algorithms and dose-volume histogram (DVH), nine treatment techniques were compared. [1] unilateral 6 MV photons [2] unilateral 12 MeV electrons [3] unilateral 16 MeV electrons [4] an ipsilateral wedge pair technique using 6 MV photons [5] a 3-field AP (wedged), PA (wedged) and lateral portal technique using 6 MV photons [6] a mixed beam technique using 6 MV photons and 12 MeV electrons (1:4 weighting) [7] a mixed beam technique using 6 MV photons and 16 MeV electrons (1:4 weighting) [8] a mixed beam technique using 18 MV photons and 20 MeV electrons (2:3 weighting) [9] a mixed beam technique using 18 MV photons and 20 MeV electrons (1:1 weighting).

Using dose-volume histograms to evaluate the dose to the contralateral parotid gland, the percentage of contralateral parotid volume receiving > or = 30% of the prescribed dose was 100% for techniques [1], [8] and [9], and < 5% for techniques [2] through [7]. Evaluating the "hottest" 5 cc of the ipsilateral mandible and temporal lobes, the hot spots were: 152% and 150% for technique [2], 132% and 130% for technique [6]. Comparing the exit doses, techniques [1], [8] and [9] contributed to > or = 50% of the prescribed dose to the contralateral mandible and the temporal lobes. Only techniques [2] and [6] kept the highest point doses to both the brain stem and the spinal cord below 50% of the prescribed dose.

The single photon lateral field [1] and the mixed electron-photon beams [8] and [9] are not recommended treatment techniques for unilateral parotid irradiation because of high doses delivered to the contralateral parotid gland and high exit doses which are associated with Xerostomia. The en face electron beam technique [2] and the mixed electron-photon beam technique [6] are unacceptable due to the excessive dose heterogeneity to the contiguous normal structures. In spite of optimal dose fall-off achieved using the en face technique [3], most patients cannot tolerate the resulting high skin doses. We conclude that the ipsilateral wedge pair [4], the 3-field [5], and the mixed electron-photon beam [7] techniques are optimal techniques in providing relatively homogeneous dose distributions within the target area and for minimizing dose to the relevant normal structures.