RESUMEN
OBJECTIVE: The accuracy and sensitivity of commonly used imaging modalities in evaluating oral cavity cancer was evaluated by comparing the preoperative radiologic findings and the postoperative pathology report. STUDY DESIGN: Patients with oral squamous cell carcinoma, who had undergone at least 1 imaging test 2 weeks before surgery were included. Radiologic findings were compared with the dissected neck findings to assess the lymph node status. Sensitivity and specificity of the imaging modalities were calculated by using the χ2 test. RESULTS: Sensitivities for detecting metastatic neck lymph nodes at a threshold of 1 cm were 48% (P = .02) and 43.8% (P = .3) for computed tomography (CT) and magnetic resonance imaging respectively. Specificities were 76.3% and 70%, respectively. As for the 1.5 cm threshold, sensitivities were 36% (P = .002) and 31.3% (P = .5), respectively, and specificities were 91.5% and 76.7%, respectively. PET-CT was the most sensitive modality in the present study, with a P value of .02. CONCLUSIONS: The different studied imaging modalities used for preoperative neck staging are not sensitive enough and would lead to underdiagnoses of a significant proportion of patients. Thus, prophylactic neck dissection for occult neck disease is of extreme importance and remains the gold standard for oral cancer treatment.