Surgical margins in a single modality transoral robotic surgery—A conundrum

Abstract

We seek this opportunity to congratulate Holcomb et al. for studying one of the important issues of margins in transoral robotic surgery (TORS), regarding which no consensus exists as of now. The current scenario in western countries has changed with rising incidence of human papilloma virus (HPV)-related cancers in younger males without significant history of tobacco and alcohol consumption. The purpose behind performing TORS mainly for HPVrelated oropharyngeal malignancies is to reduce morbidity secondary to open approaches and/or chemoradiation, thus achieving a better quality of life. Surgical margin cutoff for TORS varies across studies with some suggesting margins of 2–5 mm as clear margin. Majority of ongoing de-escalation trials like ECOG3311, ORATOR, and so forth have incorporated 2–3 mm margin as a clear margin, and any margin less than that is considered as close margin (Table 1). The authors in that study had closest margins of less than 3 mm in up to 54.6% patients. Only eight patients had margins between 3 and 4 mm, and five patients had margins between 4 and 5 mm. Such low number of patients in individual subgroups along with a median follow-up of less than 2 years (21 months) reduces the statistical reliability of the margin wise survival analysis. Any statistical conclusion drawn from such a comparison may not be truly valid or representative. The authors make a very valid point about mean muscle thickness of 2.4 mm for superior constrictor at its thinnest point leading to base or deep margins being close in palatine tonsil tumor resections as against mucosal margins for base of tongue (BOT) tumor resection. In about 58% of cases in the study, margins were processed from defect rather than from specimen, thus leading to potential error in studying margins. Reviewing literature for margin assessment favors specimen-driven approach rather than defect-driven approach for margin assessment. Secondary re-resections considered for 13 patients may not be really helpful, as tissue distortion due to previous surgery makes it difficult to make out which part to revise and such a revision may not be really advantageous for survival. Forty-two patients had BOT primary; however, only 14 patients underwent bilateral neck dissections. The rate of contralateral neck node metastasis in HPV-positive BOT tumors is as high as 21%, making it imperative to address contralateral neck either with radiation or neck dissection. This is true even for lateralized BOT tumors. The sample size of 99 patients may be far less a number to help arrive at a conclusion especially when rate of HPV-related oropharyngeal cancers is increasing at 5% per year. Due to this, histopathological factors that have been shown to have impact on prognosis like perineural invasion and lymphovascular invasion have not been adequately represented (3% and 17% in the current cohort), and therefore have not been found to impact prognosis. Quality of life has been the most crucial factor behind TORS procedure gaining importance. Chemoradiation has been associated with dysphagia in up to 50% patients and late grade III and IV laryngopharyngeal toxicity in at least 35% of the patients. These have come down with intensity modulated radiotherapy (IMRT). As per a systematic review, TORS appeared to achieve better functional outcomes. However, the recently concluded ORATOR trial showed that patients treated with radiotherapy versus TORS with neck dissection (with or without chemoradiation or radiation) had superior swallowing quality of life scores at 1 year. Thus, evaluation of these parameters is crucial. Somehow, this aspect has been completely missed by the authors. The authors found encouraging results in terms of no difference in local control or disease-free survival for close margin versus clear margin or even for “ink surface away from tumor” for stage I disease which is similar to results from other investigators who found TORS having good local control for stage I versus stage II and III disease. However, the study criteria of the other studies in terms of margins and other tumor-related factors have been distinct from the current study. As discussed previously, the survival-related conclusions of the present study may not be truly valid. Retrospective studies come with unavoidable shortcomings like selection bias and misclassification bias. Selection bias in terms of exclusion of those who received Received: 10 May 2021 Accepted: 26 July 2021