[On the relevance of histopathology results in oropharyngeal cancer with mandibular involvement and the necessary imaging].

Abstract

Background: Planning of surgical procedures in patients suffering from oropharyngeal cancer requires appropriate imaging, particularly in consideration of the spatial relationship to the mandible. Resection of portions of the mandible (box, marginal, or segmental resection) is often necessary, while simultaneously avoiding overtreatment. Typically, a computed tomography (CT) scan is initially performed. However, the question arises of whether CT alone is adequate for reliable assessment of mandibular involvement.

Materials and methods: A patient cohort of 25 individuals aged 44-79 years (mean 62 years) undergoing partial oropharyngeal resection with segmental mandibular resection was examined. The indication for segmental resection was based on the close relationship of the tumor to the mandible observed in imaging. Reconstruction was consistently carried out with a reconstruction plate and free or pedicled flaps. Preoperative radiological findings were compared with histopathology results after decalcification of the mandibular bone.

Results: Cortical tumor infiltration was observed in the mandible in 9 patients (36%). Preoperatively, clear bone infiltration had been identified in CT in only 2 of these 9 patients; in the remaining 7, only direct tumor contact with the mandible was evident. Magnetic resonance imaging (MRI) revealed bone infiltration in 1 of these 7 patients.

Discussion: Adhering to safety margins requires a radical surgical approach and reliable intraoperative frozen section analysis. However, at the bone level, intraoperative frozen section analysis is technically impractical due to the required decalcification. Only after decalcification can the entire specimen be assessed for tumor infiltration. In our cohort, a significant discrepancy in terms of bone infiltration was noted between preoperative imaging assessments and postoperative histopathology. CT preoperatively identified bone infiltration in only 8% of our patients. In two thirds of the specimens, no bone infiltration was evident after decalcification and histopathological processing, retrospectively indicating overtreatment by segmental resection; box or marginal resections may have been sufficient in these cases. Therefore, when tumors are adjacent to the bone in CT, MRI should also be performed preoperatively to more reliably detect bone infiltration.