RESUMO
PURPOSE: An understanding of the setup variation of the low neck in relation to the upper neck is necessary to define appropriate planning margins, while treating the full neck with intensity-modulated radiotherapy (IMRT) technique. MATERIALS AND METHODS: The setup of 20 sequential head and neck cancer patients was studied. Daily position verification was performed with a computed tomography (CT) on rails. An upper neck point was defined as the anterior-most portion of the cervical spine on the lowest CT cut on which both styloid processes are visible. A low neck point was defined as the anterior-most portion of the cervical spine on the lowest CT cut on which the thyroid gland was visible bilaterally. This procedure was carried out on the planning CT and on each daily treatment CT. The variation of the low neck was analyzed, assuming perfect alignment of the upper neck anatomy. RESULTS: Daily treatment CT of upper neck anterior cervical spine points were normalized to the planning CT. Relative to this coordinate system, the low neck cervical spine point was displaced an average of 3.08 mm anteriorly, ±0.17 mm. There was no systematic lateral or craniocaudal displacement. Random setup errors resulted in low neck standard deviations of 3.9 mm (anteroposterior), 3.3 mm (lateral), and 2.6 mm (craniocaudal). CONCLUSIONS: Position variation in the low neck varied in excess of the planning margins. There was a systematic anterior displacement. Random setup error was greater than expected. The results suggest that the neck volumes located distant from the region of fusion should be drawn with larger planning margins.