Anatomic distribution of local recurrence of nasopharyngeal carcinoma in the Era of IMRT: Suggestion of modification of clinical target volume delineation.

Abstract

e18060 Background: Target volume delineation in nasopharyngeal carcinoma (NPC) remains a challenge because of the narrow therapeutic margin. There are no-well established guidelines for clinical target volume (CTV) delineation with long-term follow-up. Current consensus continues to rely heavily on bony structure and experience from conventional two and three- dimensional radiation. The aim of our study, is to clarify different anatomic distribution of local recurrence of NPC based on various T stages and spread patterns, provide a high-risk map of local recurrence of NPC, and thus provide suggestion for modification of CTV delineation based on T stages and spread patterns. Methods: We retrospectively enrolled 869 patients with newly diagnosed nonmetastatic NPC treated with definitive IMRT in our institution. With a median follow-up of 54.4 months, 72 patients developed local failure. All the cases of local failure were reviewed and those of local residual disease were excluded. After that, 52 cases with tracaeble primary radiotherapy plans and MR imaging of local recurrence were included in this study. Gross tumor volume of local recurrence (GTVr) and normal structures were delineated in each recurrent patient on primary CT of radiotherapy. The overlap of normal structures with GTVr were analyzed, as well as the stepwise patterns of tumor spread in NPC, based on different T stages and relapsed risk (≥5%, ≥10%, ≥20%, etc). Results: The anatomic distributions of local recurrence of various primary T stages were compared between early and late stages of NPC. Of note, these structures were significantly highly involved in T3/4 cases, compared with T1/2 cases, including lateral pterygoid muscle, inferior orbital fissure，posterior ethmoid sinus，cavernous sinus as well as hypoglossal canal. As for tumor spread patterns of local relapse, most cases relapsed in the nasopharynx and retropharyngeal lymph node (38.5%), followed by anterior-lateral tumor extension (via nasal cavity-pterygoid process-pterygopalatine fossa, 17.31%), posterior tumor extension (via prevertebral muscle-clivus, 13.46%), anterior-superior extension (via nasal cavity-ethmoid sinus-sphenoidal sinus,13.46%), superior extension(via foramen lacerum-cavernous sinus, 11.54%), etc. Finally, we depicted a map of high-risk area based on different T stages and relapsed risk (≥5%, ≥10%, ≥20%). Conclusions: Comprehensive analysis of local recurrence distributions and tumor spread patterns in real-world IMRT provides important reference for modification of CTV delineation for primary NPC. Certain structures such as cavernous sinus and lateral pterygoid muscle are not routinely recommended in the delineation of high-risk CTV for T1-2 disease. Future individualized delineations should be based on relapsed risk and tumor spread patterns.