Session Item

Re-irradiation of locally recurrent nasopharyngeal carcinoma (rNPC) is always challenging. Target volume coverage is often suboptimal as organs at risk (OARs) usually receive significant dose in the primary course of radiotherapy. While most patients with rNPC are treated with Intensity Modulated Radiation Therapy (IMRT), there is growing interest in the potential dosimetric and clinical advantages of Intensity Modulated Proton Therapy (IMPT) to provide a wider therapeutic window. This study aims to evaluate the dosimetric performance of IMPT compared to IMRT in terms of target volume coverage and sparing of neurological OARs.

Twenty patients with locally rNPC who received second course of IMRT were retrospectively re-planned using IMPT (Varian Eclipse External Beam Planning System: Version 15.6). All of them had been previously treated with IMRT to 70Gy in the primary treatment. Compared with 9-fields sliding window IMRT re-irradiation, 3-4 fields robust-optimized IMPT plan was computed with field-specific target and uncertainties (range +/-3 % and ISO shift 3 mm, total 12 perturbation scenarios). The planning goal was to maximize dose coverage of volume of 60Gy to Clinical Target Volume (CTV) without exceeding the neurological OARs tolerances. The accumulative dose limits of OARs were defined as 130 % of dose constraints of OARs of primary course. The dose received by Gross Tumor Volume (GTV) (V100%), CTV (V100% and V98%), planning neurological OARs volume (PRV)(i.e. brainstem, spinal cord, optic nerves, optic chiasm, and temporal lobes), internal carotid arteries (ICA) and nasopharyngeal (NP) mucosa were compared.

IMPT plans significantly improved the prescribe dose coverage received by GTV V100% (+10.8%), CTV V100% (+10.4%) and CTV V98% (+2.3%) with p-values of <0.05. IMPT plans also achieved much better neurological OARs sparing, which showed significant reduction (p < 0.05) in average maximum dose (D max) to spinal cord (-2.51Gy); brainstem (-3.01Gy); optic chiasm (-3.02Gy); left optic nerve (-1.72Gy); right optic nerve (-2.04Gy); left temporal lode (-3.45Gy) and right temporal lode (-2.41Gy). These superiorities were attributed by the physical property of Bragg peak with proton beam. However, there was a paradoxical increase in the mean hotspot dose with IMPT which was 16. 5% higher than that of IMRT, and this led to significant increase in doses to ICA and NP mucosa.

Target under-dosage is a potential cause of treatment failure in rNPC patients undergoing re-irradiation. IMPT is a promising novel treatment technique that demonstrates superiority in providing more conformal dose coverage compared to IMRT. However, clinicians need to be aware that dose to the ICA and NP mucosa are often exceedingly high, which may result in increased risk of carotid blow-out and massive epistaxis. Careful clinical and dosimetric evaluation is required when deciding on the treatment strategy.