Redefining Patients at Risk of Contralateral Neck Disease for HPV-related Oropharyngeal Cancer: A Pathologic Study of Patients with Bilateral Neck Dissection

Abstract

HPV positive oropharynx squamous cell carcinomas (OPSCC) have a better prognosis than HPV negative OPSCC. Standard radiation treatment volumes cover at risk nodal regions that drain the primary tumor, and avoidance of the contralateral neck may improve morbidity. However, optimal treatment volumes are still undetermined, especially for base of tongue (BOT) primary tumors, and few studies to date have used surgical series to risk stratify HPV positive patients. We reviewed all patients who received trans-oral robotic surgery (TORS) of primary OPSCC and dissection of bilateral necks to determine risk of contralateral neck disease. After IRB approval, patients with cT1-T3 SCC of the tonsil or BOT who received primary resection and bilateral neck dissection were identified to have HPV positive disease by PCR between 2010 and 2018. Pre-surgical PET scans, CT scans, and physician notes were reviewed for clinical staging according to AJCC 8th edition, and well-lateralized primary disease was defined as lack of soft palate or midline structure involvement. Fisher’s exact test was used to evaluate association of contralateral nodal disease with pre-surgical treatment clinical information. Univariate and multivariate odds ratios were constructed using logistic regression. Out of 111 cases in our cohort, there were 10 (9%) with positive contralateral lymph nodes on pathology, including 6% (3/50) of tonsillar and 11.5% (7/61) of BOT cases. Neither cT stage nor primary site was predictive of contralateral nodal disease. Patients with a well-lateralized BOT primary and without bilateral clinical nodal disease (cN0/N1) were not likely to have pathologic contralateral disease (OR 0.05; 95% CI:0.008-0.317), and was present in 5% of such patients. Among the whole cohort, presenting without bilateral clinical nodal disease was the strongest predictor of lack of pathologic contralateral disease (adjusted OR 0.03; 95% CI:0.005-0.19). Those presenting with both cN0/N1 disease and with a well-lateralized tumor were not likely to have contralateral nodal disease on pathology (OR 0.06; 95% CI:0.013-0.25). Radiographic extranodal extension, smoking history, and multiple clinically suspicious nodes were not associated with contralateral pathologic disease. HPV-related OPSCC cancers that are clinically and radiographically N0-N1 have exceedingly low rates of contralateral disease on pathology. This is the first pathologically driven study to suggest that well-lateralized HPV positive BOT primaries with limited clinical nodal disease may be able to receive elective nodal irradiation to the ipsilateral neck only. Future prospective trials should determine if such BOT primaries can be treated with unilateral neck irradiation in a manner akin to some tonsil primaries, thereby expanding opportunities for HPV-related treatment de-intensification.