Mandibular invasion in oral squamous cell carcinoma: investigation by clinical examination and orthopantomogram.

Assessing the relationship of oral squamous carcinoma with the mandible prior to definitive therapy poses a perplexing problem for the head and neck oncologist. We carried out a prospective open study of 51 (21 female and 30 male; mean age of 53.4 years) patients undergoing mandibular resections for oral squamous cell carcinoma to examine the incidence of mandibular bone invasion and to assess the predictive capabilities of clinical and radiological examination in detecting bone involvement. A detailed clinical examination was followed by radiographic evaluation of mandible for bone invasion. After resection, the mandible was sectioned serially at every cm to find the pathological bone involvement. Sensitivity, specificity, and positive and negative predictive values of clinical and radiological findings were calculated. Specimens from 25 patients (49%) (4 segmental 21 hemi) demonstrated tumour invasion on histological examination. Clinical impression of mandibular invasion showed a sensitivity of 96% and specificity of 65%, whereas radiological examination had a sensitivity of 92% and specificity of 88%. When considered together, clinical and radiological examinations were able to detect all the cases of bone invasion, but specificity was only 58%. This study advocates careful correlation of clinical and radiological findings prior to definitive therapy, as clinical examination tends to over diagnose bone invasion in tumours adjacent to the mandible. The specificity of imaging was also found to be low pointing towards the need for more specific diagnostic tools in doubtful cases. Aggressive surgical therapy, namely segmental or hemi resection of mandible is warranted in case of tumours of the lower alveolus with definite bone invasion. In case of carcinomas of the buccal mucosa and tongue the mandibular resection can be limited to that required for clearance of margins provided the radiology is negative.