Can radiological examination of mandibular bone invasion accurately predict the need for mandibular resection in oral squamous cell carcinoma?

Abstract

Bone invasion by oral squamous cell carcinoma necessitates jaw resection, with preoperative imaging ideally able to guide the resection. A retrospective review of 109 patients with oral squamous cell carcinoma who underwent mandibular resection was performed. Eighty-three had preoperative computed tomography (CT) imaging and 72 underwent magnetic resonance imaging (MRI). The presence of bone invasion on imaging was compared to histopathology. Bone invasion was detected in 44 of 109 resection specimens (40.4%) and was identified on CT in 31 of 83 cases (37.4%) and on MRI in 35 of 72 cases (48.6%). The sensitivity and specificity of CT for detecting bone invasion was 69.0% and 79.6%, respectively, while for MRI was 87.1% and 80.5%, respectively. Histological detection of bone invasion was associated with greater disease-specific mortality (P=0.002), as was MRI detection of bone invasion (P=0.027). CT detection was not significant (P=0.240). Negative prediction of bone invasion was 95% accurate for both modalities in clinically non-invaded mandibles. Survival was reduced in patients who underwent marginal mandibular resection when bone invasion was detected histologically (33.3% vs. 70.5%, P=0.277) and with CT, although this was not statistically significant. More data are required to determine whether more aggressive resection is warranted when bone invasion is detected preoperatively.

Keywords: bone invasion; computed tomography; magnetic resonance imaging; oral; squamous cell carcinoma.