Outcomes of retropharyngeal lymph node dissection in HPV-associated oropharyngeal squamous cell carcinoma patients treated with trans-oral robotic surgery

Presentation: A114
Topic: Oropharynx / HPV Related Disease
Type: Poster
Date:
Session:
Authors: Austin C Cao, BA1; Garrett Largoza, BS2; Jason G Newman, MD1; Karthik Rajasekaran, MD1; Steven B Cannady, MD1; Devraj Basu, MD, PhD1; Ara A Chalian, MD1; Gregory S Weinstein, MD1; Bert W O'Malley, MD1; Robert M Brody, MD1; Christopher H Rassekh, MD1
Institution(s): 1University of Pennsylvania Department of Otorhinolaryngology- Head and Neck Surgery; 2Thomas Jefferson University, Sidney Kimmel Medical College


Objectives:
(1) To describe outcomes of retropharyngeal lymph nodes (RPLN) dissection in HPV+ oropharyngeal squamous cell carcinomas (OPSCCs), and 2) to compare the radiographic and pathological findings of RPLNs.

Study Design: Retrospective chart review.

Methods: Chart review was conducted on patients with HPV + OPSCC receiving TORS at an academic tertiary care center from March 2007 to March 2021. Data collected included patient demographics, treatment information, and imaging and pathology findings for RPLNs. 

Results: From a total of 837 patients who underwent primary TORS for HPV-associated OPSCC, 53 patients (6.3%) had ipsilateral RPLNs resected as indicated by radiographic assessment [median age 58 (IQR = 51-64), males 44 (83%)]. Only 18/53 (34%) had prior imaging with positive or atypical RPLN. No contralateral dissection of RPLNs were performed. The vast majority (92%) of patients who received RPLN dissection were tonsillar primary tumors, with the following tumor staging: 38 (72%) early T-stage tumors (cT1/T2), 9 (17%) advanced T-stage tumors (cT3/T4), and 5 (9.4%) unable to be assessed. Metastases to RPLNs were found on surgical pathology in 15 (28%) patients, with 5 found to have extranodal extension. Pathologic 7th edition N-stage 2b or greater was associated with pathologic RPLN metastases (OR 12.63, p = 0.002), but was not associated with radiographic RPLN findings. In particular, all non-tonsillar primaries (3 base of tongue, 1 glossotonsillar sulcus) that received RPLN dissection were found to have pathologic RPLN and pN2b or greater. Recurrence in an RPLN occurred in 3 patients with prior RPLN dissection, none of which had positive RPLN at time of initial surgery. Positive or atypical radiographic RPLN during pre-treatment period was predictive of positive pathologic RPLN (sensitivity 0.73, specificity 0.99, LR+ 101, LR- 0.27).

Conclusion: High rates of metastatic disease are found on RPLN dissection for HPV+ OPSCC. The presence of non-tonsillar primaries with positive RPLN suggests that atypical lymphatic spread may be occurring from extensive cervical lymph nodes metastases. Positive or atypical RPLN on diagnostic imaging is a strong predictor of positive pathologic RPLN, but some metastases may be radiographically silent.