Utility of fiducial markers for target positioning in proton radiotherapy of oesophageal carcinoma

Background and purpose Oesophageal mobility relative to bony anatomy is a major source of geometrical uncertainty in proton radiotherapy of oesophageal carcinoma. To mitigate this uncertainty we investigated the use of implanted fiducial markers for direct target verification in terms of safety, visibility, and stability. Materials and methods A total of 19 helical gold markers were endoscopically implanted in ten patients. Their placement at the proximal and distal tumour borders was compared to tumour demarcations derived from [18F]Fluorodeoxyglucose positron emission tomography, their visibility quantified via the contrast-to-noise ratio on daily orthogonal X-ray imaging, and their mobility relative to bony anatomy analysed by means of retrospective triangulation. Results Marker implantation proceeded without complications, but the distal tumour border could not be reached in two patients. Marker locations corresponded reasonably well with metabolic tumour edges (mean: 5.4 mm more distally). Marker visibility was limited but mostly sufficient (mean contrast-to-noise ratio: 1.5), and sixteen markers (84%) remained in situ until the end of treatment. Overall, marker excursions from their planned position were larger than 5(10) mm in 59(17)% of all analysed fractions. On one occasion severe target displacement was only identified via markers and was corrected before treatment delivery. Conclusion Implanted helical gold fiducial markers are a safe and reliable method of providing target-centric positioning verification in proton beam therapy of oesophageal carcinoma.