Intensity-Modulated Proton Therapy with Dose Painting based on Hypoxia Imaging for Nasopharyngeal Cancer

Abstract

In precision medicine, reducing recurrence rates and toxicities is necessary for nasopharyngeal cancer (NPC) patients. Hypoxia is a well-known risk factor for radioresistance. Radiation dose escalation (DE) to hypoxic regions is one strategy, especially for locally advanced cases in which tumor control may be unsuccessful. Radiation dose reduction is also a valuable approach. A clinical trial of dose de-escalation (DDE) for oropharyngeal cancer based on hypoxia status was recently reported. DDE may also be applicable for NPC patients with good prognostic characteristics. The purpose of this study was to investigate the feasibility of hypoxia-based dose painting for intensity-modulated proton therapy (IMPT) and intensity-modulated X-ray therapy (IMXT) with constraints on doses to organs at risk (OARs) and normal tissue complication probabilities (NTCPs) in DE and DDE scenarios. DE and DDE treatment plans for IMPT and IMXT were generated for ten NPC patients who underwent 18F-fluoromisonidazole (FMISO)-PET before definitive radiotherapy. In the DE scenario, 70 Gy was prescribed to the tumor in a standard manner, with DE to 84 Gy to FMISO-PET-positive (hypoxic) regions. In the DDE scenario, 70 Gy was prescribed to the hypoxic region, with DDE to 60 Gy to non-hypoxic regions. Prescribed doses to high- and low-risk areas were 63 and 56 Gy, respectively, in the DE scenario and 54 and 48 Gy, respectively, in the DDE scenario. The constraints for doses to OARs were a parotid gland mean dose (Dmean) <26 Gy and an oral cavity Dmean <39 Gy. The criterion for NTCP was probability ≤50% for all five toxicities: salivary flow reduction, xerostomia, dysphagia, tube feeding dependence and dysgeusia. The determined doses to hypoxic and non-hypoxic target volumes were prescribed in all IMPT and IMXT plans in the DE and DDE scenarios. The numbers of patients who satisfied the parotid gland, oral cavity and NTCP criteria were 7, 10 and 10, respectively, for IMPT and 0, 1 and 0, respectively, for IMXT in the DE scenario and 10, 10 and 10, respectively, for IMPT and 1, 2 and 5, respectively, for IMXT in the DDE scenario. Dmean for the contralateral parotid gland was 25.2±3.0 for IMPT and 36.3±3.4 for IMXT (p=0.0003) in the DE scenario and 21.4±2.6 for IMPT and 30.9±3.0 for IMXT (p=0.0003) in the DDE scenario. Dmean for the oral cavity was 28.8±7.3 for IMPT and 49.3±4.9 for IMXT (p=0.0003) in the DE scenario and 24.5±6.2 for IMPT and 41.3±4.3 for IMXT (p=0.0003) in the DDE scenario. For all toxicities, NTCPs were lower for IMPT than for IMXT for all patients. ΔNTCP, calculated as NTCPIMXT-NTCPIMPT, was more than 10% for 3 or more toxicities for 9 patients in the DE scenario and 8 patients in the DDE scenario. Compared with IMXT, IMPT satisfied parotid, oral cavity and NTCP criteria in more patients and reduced doses to OARs and NTCPs in both the DE and DDE scenarios. IMPT could be more useful than IMXT when conducting hypoxia-based dose painting while maintaining low toxicities.