Purpose: To determine how to use node response on computed tomography (CT) to indicate the need for neck dissection.
Patients and methods: Five hundred fifty patients with lymph node-positive head and neck cancer were treated between 1990 and 2002 with radiotherapy (RT) at a median dose of 74.4 Gy; 24% of these patients (n = 133) were treated with chemotherapy. Three hundred forty-one patients (62%) underwent planned post-RT neck dissection. Physical examination and contrast-enhanced CT were performed 30 days after completion of RT. CT images were reviewed in 211 patients for lymph node size (largest axial dimension) and presence of a focal abnormality (lucency, enhancement, or calcification). By correlating post-RT CT to neck dissection pathology, criteria associated with a low likelihood of residual disease were identified. A subset of patients who fit these criteria of radiographic response who did not undergo post-RT neck dissection was observed for recurrence.
Results: Radiographic complete response (rCR) was defined as the absence of any large (> 1.5 cm) or focally abnormal lymph node. Correlation of response with neck dissection pathology indicated a negative predictive value of 77% for complete clinical response and 94% for rCR. In 32 patients (median follow-up time, 3.2 years) with rCR who did not undergo post-RT neck dissection, the 5-year ultimate neck control rate (100%) and cause-specific survival rate (72%) were not significantly different from the rates of patients with a negative post-RT neck dissection.
Conclusion: Patients with rCR 4 weeks after RT can be spared from a post-RT neck dissection regardless of initial node stage.