Objectives: To determine the sensitivity and specificity of magnetic resonance imaging (MRI) for the detection of perineural spread (PNS) along the intratemporal facial nerve (ITFN) in patients with head and neck cancers.
Study design: Case series with chart review.
Setting: Tertiary care center.
Subjects and methods: We included 58 patients with head and neck malignancies who underwent sacrifice of the ITFN between August 1, 2002, and November 30, 2015. Demographics, preoperative facial nerve function, prior oncologic treatment, and timing between MRI and surgery were recorded. Histopathology slides and preoperative MRI were reviewed retrospectively by a neuropathologist and a neuroradiologist, respectively, both blinded to clinical data. The mastoid segment of the facial nerve (referred to as the descending facial nerve [DFN]) and stylomastoid foramen (SMF) were evaluated separately. A grading system was devised when radiographically assessing PNS along the DFN.
Results: Histopathologic evidence of PNS was found in 21 patients (36.2%). The sensitivity and specificity of MRI in detecting PNS to the DFN were 72.7% and 87.8%, respectively. MRI showed higher sensitivity but slightly lower specificity when evaluating the SMF (80% and 82.8%, respectively). Prior oncologic treatment did not affect the false-positive rate ( P = .7084). Sensitivity was 100% when MRI was performed within 2 weeks of surgery and was 62.5% to 73.3% when the interval was greater than 2 weeks. This finding was not statistically significant (SMF, P = .7076; DFN, P = .4143).
Conclusion: MRI shows fair to good sensitivity and good specificity when evaluating PNS to the ITFN.
Keywords: facial nerve; imaging; mastoid; perineural invasion; perineural spread; temporal bone.