111 Redundant tooth loss prior to head and neck radiotherapy 5 Radiation dose calculations All RT dose planning was performed in Eclipse (Aria version 15.5; Varian Medical Systems Inc, Palo Alto, California, United States) [21] in which the targets (gross tumor volume (GTV), clinical target volumes (CTV) and planning target volumes (PTV)) were delineated according to international guidelines [22] and the organs at risk (OAR’s) according to the Brouwer’s Atlas [23]. The radiation dose for each extracted tooth was calculated retrospectively: An experienced RT technologist (MG) delineated the location of the extracted teeth on the planning CT. First, the window level was set to bone density. Second, for each extracted tooth a new structure was created and named according to the Fédération Dentaire Internationale (FDI) World Dental Federation notation [24]. If the maxilla and/or mandible had received a maximum dose (Dmax) of less than 25Gy (defined as: the 25Gy isodose line not touching the bone of the mandible/maxilla), this particular extracted tooth was not delineated, but was recorded as <25Gy. To delineate the location of the extracted tooth, the contouring tool was converted to a highresolution segment and a 6 mm wide brush was selected. For each extracted tooth, the position on each CT slice (3 mm slice thickness) where the bone was visible was delineated (Figure 1). After all locations of the extracted teeth were delineated, the mean dose (Dmean) and the Dmax in these locations were exported. All exported data were converted to ipsilateral or contralateral, according to the laterality of the primary tumor region. RT dose was converted to a binary variable comparing sites that received ≥40Gy with sites that received <40Gy, including sites recorded as <25Gy. To calculate the mean values, standard deviations and ranges of Dmean and Dmax, the sites recorded as <25Gy were not included. Statistical analyses Descriptive statistics were reported as numbers and percentages, means with standard deviations (SDs), medians with interquartile ranges (IQRs), and total radiation dose ranges in Gy. Univariable logistic regression analyses was used to test the association between different demographic and clinical variables with dose ≥40Gy, for both Dmean and Dmax. These factors included: age at first dental assessment, sex, tumor location, tumor (T) and nodal stage (N), overall tumor stage (I, II, III, IV), early vs. advanced tumor stage, and number of teeth extracted. Factors with p<.05 were selected as potentially relevant associative variables and subsequently tested using multivariable logistic regression analyses. Data were analysed using SPSS (IBM version 28 for Windows, Armonk, New York, USA). A p-value of less than .05 was considered statistically significant.
RkJQdWJsaXNoZXIy MTk4NDMw