Objective: To compare the predictability of orthopantomograms (OPG), bone scans, computerised tomography (CT), magnetic resonance imaging (MRI) and periosteal stripping with direct inspection in predicting both the presence and extent of tumor invasion of the mandible.
Design: Prospective study.
Setting: Queen Elizabeth Hospital, Birmingham; Wordsley Hospital, Stourbridge; North Staffordshire Royal Infirmary, Stroke-on-Trent.
Subjects: 35 consecutive patients requiring a mandibular resection as part of their treatment for oral squamous cell carcinoma.
Main outcome measures: The prediction of the extent of bone invasion for each of the investigations and direct inspection after periosteal stripping. The actual extent of invasion of the mandible from a detailed histological assessment.
Results: The OPG underpredicted the width and depth of invasion by on average 13 mm and 2 mm. There were 5 false negative reports. Bone scans overpredicted the width and depth by 14 mm and 15 mm with 1 false negative. CT scans underpredicted the width of invasion by 5 mm and overpredicted depth by 3 mm, but 7 false negatives were returned. MRI scans overpredicted width and depth of invasion by 19 mm and 10 mm with 1 false negative. Direct inspection after periosteal stripping underpredicted the width and depth of invasion by 5 mm and 3 mm with 1 false negative.
Conclusion: OPG's and bone scans are useful for the initial assessment of all tumours in the region of the mandible. MRI is a more useful investigation than CT in the assessment of mandibular invasion by oral squamous cell carcinomas. Exploratory periosteal stripping at the time of resection can accurately predict the presence of tumour invading the mandible.