Assessment of Acute Cervical Skin Toxicity Due to Fractionated Proton Beam Therapy for Head-and-Neck Cancer

Abstract

By our actual clinical experience, moderate to severe cervical radiation dermatitis occurs in the majority of patients treated with proton beam therapy (PBT) for head and neck cancer (HNC) patients. The aim of this study is to assess the factors affecting the onset of acute cervical dermatitis in HNC patients with nodal metastasis treated with PBT. We retrospectively investigated the clinical and dosimetric predictors of planning system for skin toxicity. Twenty-seven patients with HNC treated with fractionated PBT to nodal metastasis with or without chemotherapy were reviewed. Acute cervical skin toxicity of irradiated area was estimated based on the CTCAE version 4.0 grading. Clinical factors thought to be influencing skin toxicity, such as the prescribed dose and number of fractions, physical beam characteristics, size and location of the target volume, concomitant chemotherapy and DVH parameters of skin surface were analyzed for their relation with risk of acute dermatitis. In this study, the skin surface region was defined the area contoured divided into the thickness in every 2mm structure inside away from the skin surface at the irradiated portions such as surface-2mm, 2-4mm, 4-6mm, 6-8mm, 8-10mm. Typical dosimetric parameters, including the size of the irradiated volume, value of V40-80GyE, and the mean and maximal point dose delivered to skin surface region were obtained from DVH calculation. To evaluate steepness of the entrance dose to skin surface, gradient dose index (surface-2mm volume-dose to 8-10mm volume-dose) was defined. Median biological effective dose (BED) at an α/β ratio of 10 as the prescribed dose was 84.5GyE10 (range, 60-96.8GyE10). Eleven (40%) patients developed acute Grade 1, 12 (45%) patients Grade 2, and 4(15%) patients Grade 3 acute dermatitis. In uni-variate analysis, factors associated with onset of Grade 2 or higher acute dermatitis included BED (GyE10) as prescription dose (p = 0.030), skin dose parameter (V45-70GyE10) (p = 0.044, 0.039, 0.031, 0.015, 0.010, and 0.022, respectively), and gradient dose index (GDI65GyE and 70GyE) (p = 0.031, and 0.048, respectively). When plotting the p value of the logistic regression analysis between the variables, V65 GyE10 of skin surface area show the strongest statistical significance (p ≦ 0.01). High incidence of severe (Grade 2 or higher) radiation dermatitis (60% of entire cohort) was observed during clinical use of PBT in HNC patients. The entrance portal area to skin irradiated volume with 45-70GyE10 was possible predictor for acute skin toxicity and higher dose volume of superficial regions over 65GyE10 should be considered in the treatment planning as sensitive organ volume in PBT.