3698 - Rethinking Parotid Lymph Node Assessment in NPC: Why 10 mm is Not Enough and How Node-RADS Refines the Biopsy Threshold
Presenter(s)
J. Zhu1, Y. Zheng2,3, X. Dai1, B. Yang1, Y. Huang3, Z. Wu3, J. Pan2, S. Lin2, and Q. Guo3; 1Fujian Medical University, Fuzhou, Fujian, China, China, 2Department of Radiation Oncology, Clinical Oncology School of Fujian Medical University, Fujian Cancer Hospital, Fuzhou, Fujian, China, 3Fujian Cancer Hospital, Fuzhou, Fujian, China, China
Purpose/Objective(s): Parotid lymph node (PLN) metastasis has been increasingly recognized as a common site of out-field recurrence after radical IMRT in nasopharyngeal carcinoma (NPC), often originating from subcentimeter nodes missed by conventional size criteria. This study aimed to evaluate and compare the diagnostic performance of conventional size criteria (short-axis diameter [SAD] =10 mm) and the MRI-based Node-RADS system for identifying malignant PLNs at the patient level.
Materials/Methods: Patients with newly diagnosed, non-metastatic NPC treated with radical IMRT between January 2014 and December 2019 were candidates for screening. Additional inclusion criteria required that patients had at least one PLN with a SAD =5 mm. All such nodes were independently reviewed by two head-and-neck radiologists, with documentation of anatomical location (superficial/deep intraparotid, subcutaneous pre-auricular, or subparotid) and assignment of a Node-RADS score (1–5). PLNs were classified as malignant based on either pathological confirmation via fine-needle aspiration cytology (FNAC) or in-situ recurrence in the parotid region during follow-up after parotid-sparing IMRT .
Results: Of the 515 enrolled patients, 47 (9.1%) were identified as having malignant PLNs. Among these malignant nodes, 29 (61.7%) measured <10 mm, underscoring the limitations of conventional size criteria. The malignancy rate increased significantly with short-axis diameter (SAD): 1.2% (5–6 mm), 7.0% (6–8 mm), 47.1% (8–10 mm), and 44.8% (=10 mm) (P < 0.001), identifying 8 mm as a critical risk threshold. All 47 malignant PLNs received Node-RADS scores of 4–5. Among the 70 PLNs with scores 4–5, malignancy rates remained consistently high across size subgroups: 60.0% (5–6 mm), 60.0% (6–8 mm), 69.6% (8–10 mm), and 76.5% (=10 mm). Anatomically, most PLNs were located in the superficial intraparotid region (74.6%), followed by subparotid (10.1%), subcutaneous pre-auricular (8.5%), and deep intraparotid (6.8%) regions. The highest site-specific malignancy rate was observed in the deep intraparotid region (14.3%), while the subcutaneous pre-auricular region had the lowest (2.3%). Notably, malignant nodes <10 mm were identified in all anatomical sites except the subcutaneous pre-auricular region.
Conclusion: Given the diagnostic challenges of parotid lymph node (PLN) involvement, a multimodal approach integrating both an 8-mm short-axis diameter threshold and Node-RADS assessment is warranted to guide biopsy decisions. Nodes with Node-RADS scores of 4–5 should undergo biopsy irrespective of size, given their strong association with malignancy. Future integration of artificial intelligence may further refine this strategy, enhancing diagnostic accuracy and optimizing patient selection for pathological confirmation.