Background: The propensity for early lymphatic metastases of the oropharynx necessitates management of the neck even in the absence of clinical and radiographic disease. The likelihood of contralateral neck metastases in the absence of radiographic disease is not well understood for tonsil and tongue base cancers. Although some centers advocate unilateral neck treatment for lateralized tonsil cancers, the true rates of occult contralateral metastases for tonsil and especially tongue base cancers is not known. Knowledge of occult contralateral disease would help inform practitioners as to who may be candidate for unilateral neck management.
Objective: The objective of this study was to determine the rate of pathologic contralateral positive nodes in patients without suspicious preoperative imaging in a multi-institutional population of OPSCC patients treated with TORS and bilateral neck dissections.
Methods: A retrospective review of medical records was performed at the participating institutions: Princess Margaret Cancer Center, Toronto; Icahn School of Medicine at Mount Sinai, New York City;, and Montefiore Medical Center, New York City. Patients with OPSCC undergoing TORS and bilateral neck dissections were identified for analysis. Preoperative imaging, as reviewed by experienced neuroradiologists blinded to pathologic status, was compared to postoperative pathology reports. The AJCC 7th Edition was applied for staging.
Results: Thirty-three patients were identified who underwent bilateral neck dissections with TORS for OPSCC as initial therapy. There were 28 men and 5 women, and the median age of the population was 63 (SD = 10.5) years. Thirteen patients (39%) had tonsil primary, 19 (58%) patients had a base of tongue primary, and 1 (3%) patient had a pharyngeal wall primary. Twenty-five (74%) of patients were known to be p16+. Preoperative radiographic stages were as follows: 2 (6%) N0; 5 (15%) N1; 5 (15%) N2a; 16 (48%) N2b; 5 (15%) N2c; and 1 (3%) N3. Twenty-nine patients (88%) were radiographically negative in the contralateral neck preoperatively. Of the 5 patients with radiographically suspicious contralateral nodes, only 1 (20%) was positive on final pathology. Three patients were ultimately found to have pathologically positive contralateral nodes, and of these three, 2 were not anticipated on preoperative imaging. The occult contralateral nodal disease rate was 6.9%. One occult node was radiographically T1N1 p16+ tonsil primary 7 mm from midline, and the other was radiographically T4 (deep tongue muscle invasion) N2b p16- base of tongue primary at the midline. The sensitivity, specificity, positive predictive value, and negative predictive value of suspicious contralateral nodes on preoperative imaging were 33%, 84%, 20% and 93% respectively.
Conclusion: OPSCC patients undergoing TORS and elective contralateral neck dissection will have a very low rate of unanticipated contralateral nodal positivity. Many patients with suspected contralateral disease may in fact be down-staged postoperatively. Prediction of pathologic staging remains a challenge with current preoperative imaging modalities.