Abstract
451
Objectives Extracapsular spread in lymph node metastasis is an important negative prognostic factor in patients with head and neck cancer. The purpose of this study was to evaluate the role of F-18 fluorodeoxyglucose positron emission tomography (FDG PET) in predicting extracapsular spread (ECS) in cervical lymph node metastasis, and to compare its diagnostic accuracy to magnetic resonance (MR) imaging.
Methods Head and neck cancer patients with metastatic lymph nodes confirmed by surgical pathology who underwent pretreatment FDG PET/CT scans were retrospectively identified. Maximum standardized uptake value (SUVmax) was measured for lymph node groups and metabolic tumor volume was measured at a threshold of SUV 2.5 (MTV2.5), 50% of SUVmax (MTV50%), and background-level adjusted for liver (MTVliver). The association of pathologically confirmed ECS status with the SUVmax and various volumetric parameters of the primary tumor and lymph nodes from the FDG PET was evaluated. These findings were then compared to the current noninvasive gold standard, MR imaging.
Results Of the 110 metastatic cervical lymph node groups from 63 head and neck cancer patients included in this study, 60 (55%) were positive for ECS. ECS positive lymph nodes were associated with higher SUVmax (5.175±3.4530 vs. 5.067±3.9032, p=0.417), MTV2.5 (5.231±15.0988 vs. 2.513±4.7454, p=0.04), MTVliver (5.983±16.7905 vs. 2.623±4.7971) and MTV50 (3.028±6.3303 vs. 1.489±1.9306). On logistic regression, however, these parameters along with MR imaging were not statistically significant indicators for ECS.
Conclusions In this retrospective study, high MTV correlated with the presence of ECS, but was not predictive. Both metabolic parameters measured from FDG PET and MR imaging lack diagnostic accuracy, highlighting the need for more research on noninvasive methods to discriminate lymph nodes which are ECS positive in head and neck cancer patients