Dental amalgam artifact: Adverse impact on tumor visualization and proton beam treatment planning in oral and oropharyngeal cancers

Purpose: We evaluated the incidence and impact of dental filling artifacts on the definition of clinical target volume (CTV) for oropharyngeal/oral cavity cancers receiving radiation therapy. We performed phantom proton beam dosimetric analyses using a low-density composite filling to investigate artifact reduction and dose distribution. Methods and materials: We reviewed oral cavity/oropharynx radiation treatment plans between 2010 and 2012. Plans were evaluated for artifacts and impact on CTV visualization. We constructed a head and neck phantom, obtaining planning computed tomography images at baseline (native tooth) and for each filling (composite and metal amalgam) interchanged into a tooth adjacent to the tumor. We performed uniform scanning proton plans with each filling, evaluating for planning target volume (PTV) coverage and overall dose distribution. Results: A total of 110 treatment plans were reviewed (71 oropharynx, 39 oral cavity). Artifacts were identified in 81 plans (73.6%), including 53 oropharynx (74.6%) and 28 oral cavity (71.8%). Artifacts obscured the CTV in 77 cases (95%), including 49 of 53 oropharynx cases (92.5%) and all 28 oral cavity cases. On phantom testing, the metal amalgam obscured the tumor while the composite did not. Hounsfield unit (HU) values (range, mean) for the tumor were: baseline (- 484.0 to 700.0 HU, 104 HU), composite (- 728.5 to 1038.0 HU, 105 HU), metal amalgam (- 1023.0 to 807.0 HU, 90.74 HU). The percent of planning target volume receiving 95% of prescription dose of the PTV was baseline (100%), composite (100%), and metal amalgam (92.3%). PTV dose ranges were baseline (98%-106%), composite (98%-107%), and metal amalgam (66%-111%). PTV coverage and dose distributions of the composite and native tooth plans were identical. Conclusions: A high incidence of artifacts was found on the planning scans of oral/oropharyngeal cancer patients, adversely impacting CTV visualization. In our phantom model, metal amalgam impacted tumor and tissue density. The PTV was underdosed with the metal amalgam compared with the composite filling. A potential solution involves exchanging metal fillings with composite before proton treatment planning for improved tumor visualization and dosimetry.