Impact of amalgam dental filling on radiotherapy of head and neck cancer: In vivo dosimetry and dose calculation using AAA and Acuros algorithms.

INTRODUCTION: Dental restorations using high-density materials can cause inaccuracies in target and organ-at-risk (OAR) delineation and dose calculations during radiotherapy. These materials, such as amalgam, lead to dose scattering, resulting in enhanced mucositis in adjacent tissues. Minimizing the impact of these artifacts is crucial to improve dose calculation accuracy. This study evaluates the effects of amalgam tooth fillings on dose distribution, compares two dose calculation algorithms (AAA-anisotropic analytical algorithm and AXB-Acuros XB), and assesses their impact on mucosal toxicity during head and neck radiotherapy. PATIENTS AND METHODS: Forty-nine patients with one to five dental amalgam fillings treated with intensity-modulated radiotherapy (IMRT) for head and neck cancer between 2016 and 2021 at the Oncotherapy Department of Szeged University were included. Planning CTs with and without metal artifact reduction (MAR) were used to delineate targets and OARs. Treatment plans were optimized using the Eclipse Treatment Planning System with the AAA and AXB algorithms. In vivo dosimetry was performed using Gafchromic EBT3 films embedded in thin Styrofoam slabs during one of the first five treatment sessions. Statistical analyses, including t-tests, ANOVA, paired t-tests, and Kaplan-Meier curves, were conducted to evaluate the influence of clinical and dosimetric factors on dose perturbations and mucositis onset. RESULTS: Metal artifact reduction (MAR) correction improved contouring accuracy. Dose values calculated with AAA were higher than those with AXB for both mean and maximum dose to OARs and mucosa (Dmean: AAA > 10.57%, Dmax: AAA > 6.8% compared to AXB). Measured doses showed better agreement with AAA-calculated Dmean values (p = 0.341) but were significantly underestimated by AXB (p < 0.001). There was no difference in dose perturbation according to tumor localization, gross tumor volume, planning target volume, or the use of a tongue wedge. The number of amalgam-filled teeth correlated significantly with the earlier onset of mucositis, with each additional filling advancing mucositis appearance by 1.7 days. CONCLUSION: High-density dental materials cause significant dose perturbations in the oral cavity during head and neck radiotherapy. In our clinical IMRT setup, the AAA algorithm demonstrated closer agreement with in vivo film measurements compared to AXB, despite the theoretical and literature-reported superiority of AXB in heterogeneous conditions. This discrepancy underscores that algorithm performance depends on clinical technique, dosimetric methodology, and dose reporting mode. Additionally, each dental filling was associated with earlier mucositis onset.