Abstract
Purpose: Primary radiation therapy with or without chemotherapy (RT/CRT) is the most common treatment for oropharyngeal squamous cell carcinomas (OPSCC), but there has been an increase in transoral surgery (TOS) for T1-2 tumors. Because only a subset of T1-2 tumors are TOS-favorable, nonrandomized comparisons between RT/CRT and TOS could be confounded by indication. We aimed to compare outcomes of potential TOS-candidates versus non-TOS candidates, among patients who underwent RT/CRT for early T-stage OPSCC.
Methods and Materials: For patients treated with RT/CRT for early-stage human papilloma virus positive OPSCC between 2014 and 2018, pretreatment imaging was reviewed by 3 head and neck surgeons, blinded to outcomes, to assess primary-site appropriateness for TOS, and extracapsular extension (ECE) was scored by a head and neck neuroradiologist. We compared outcomes based on surgical favorability pertaining to (1) the primary site tumor alone and (2) the primary site and an absence of ECE. Kaplan-Meier estimates for overall survival (OS), disease-specific survival (DSS), and recurrence-free survival (RFS) were compared using the log-rank test, with Cox regression used for multivariable modeling.
Results: One hundred and forty-three patients were evaluated, of which 121 were male (84.6%), the median age was 59.4 years, and all of them were p16 positive (100%). The primary site was TOS-favorable in 115 of 143 (80.4%). Patients with TOS-favorable primary site experienced superior 5-year OS (89.8% vs 71.2%, P = .017), DSS (90.4% vs 63.4%, P = .022), and RFS (83% vs 49.4%, P = .04) compared with TOS-unfavorable patients. Similarly, patients with a TOS-favorable primary site and no ECE on imaging 101 of 143 (70.6%), had improved OS, DSS, and RFS (P < .05) compared with TOS-unfavorable patients.
Conclusions: In this first study to assess surgical favorability as a prognostic factor among patients with T1/2 p16+ OPSCC, patients with TOS-favorable early-stage OPSCC have better outcomes than TOS-unfavorable patients. This provides valuable prognostic information for patients, and also suggests the risk of confounding by indication in nonrandomized comparisons of treatment modalities. (C) 2022 Elsevier Inc. All rights reserved.