A systematic review on Transoral robotic surgery (TORS) for carcinoma of unknown primary origin: Has tongue base mucosectomy become indispensable?

Abstract Background Transoral robotic surgery (TORS) is increasingly used in head and neck surgery and in carcinoma of unknown primary (CUP) origin specifically. Due to the rising incidence of human papillomavirus (HPV)‐related oropharyngeal squamous cell carcinoma (OPSCC), there is a rationale for finding ways to de‐escalate treatment strategies. This review aims to test the hypothesis that TORS is a meaningful adjunct in the diagnostic (and therapeutic) pathway in CUP in head and neck. Methods A structured search of the literature was performed with the search terms ‘TORS’ and ‘Carcinoma of Unknown Primary’. Results Two hundred and seventy four cases of CUP in which TORS was used were identified for further analysis. Workup for CUP was comparable in all series with regard to physical examination, fine and/or gross needle examination of cervical nodes, fibre optic endoscopy, imaging and robot assisted mucosectomy of the base of tongue (BOT). Identification rate of the primary tumour was 72% on average (range 17%‐ 90%), and 55%‐ 96% were HPV positive. Clear margins were achieved in 60% (range 0%‐85%) of resected occult tumours. Complication rate of TORS BOT mucosectomy was low with mainly grade I‐III sequelae according to Clavien–Dindo. Conclusions Transoral robotic surgery seems to be a useful and safe adjunct in the diagnostic and therapeutic pathway in case of CUP in an era of increasing incidence of HPV‐positive OPSCC.


| INTRODUC TI ON
Carcinoma of unknown primary (CUP) metastatic to cervical lymph nodes represents 1%-5% of all cases of head and neck malignancies, and consists mainly of squamous cell carcinoma (SCC) (50%-70%). 1,2 Once identified, the majority of initially occult oropharyngeal cancers turn out to be high-risk human papillomavirus (hrHPV) related.
As a result of the increase in HPV-related oropharyngeal squamous cell carcinoma (OPSCC), the incidence of CUP is increasing. 3,4 The reported HPV prevalence for CUP in head and neck ranges between 22%-91% worldwide. p16 protein expression is considered to be a good surrogate marker of HPV-status (although 15% of p16 positive tumours is HPV DNA negative) which is currently the most important independent prognostic factor in OPSCC. 5 There is no general consensus on preferred diagnostic investigations in CUP, although physical examination including (office-based) endoscopy and diagnostic imaging such as computed tomography (CT), magnetic resonance (MR) imaging and positron-emission tomography (PET) is routinely performed in most institutions. The improvement of imaging techniques, high-definition endoscopic instruments and the introduction of narrow-band imaging (NBI) has significantly improved detection rates of the primary head and neck tumour over the years. 6 Despite these efforts, approximately 50% of primary tumours remain undetected. 7,8 Identification of the primary tumour site is important for optimal treatment. The current standard treatment for (occult) OPSCC is based on either surgery and/ or radiotherapy, both associated with comparable, high tumour control rates but with different side effects profiles and technical constraints.
In order to decrease the potential morbidity of open surgery, transoral approaches have been developed within the last decades, including transoral robotic surgery (TORS).
There is no general consensus in the various national guidelines such as the United States National Comprehensive Cancer Network (NCCN) and the British National Institute for Health and Care Excellence (NICE) on the role of TORS in CUP diagnostics. [7][8][9][10] The present study aims to determine the true benefit of TORS in detecting unknown primary tumours by conducting a systemic review of the literature.

| Ethical considerations
Institutional ethical approval needed not to be obtained for this systematic review for which publicly accessible data were used. None of the used data are individually traceable.
A MEDLINE (PubMed) search was performed with the search terms 'TORS' AND 'Carcinoma of Unknown Primary' using a combination of MeSH headings and keywords. The study was not designed to identify all studies on TORS in head and neck cancer, but to analyse those that focused on TORS for CUP.
The search was limited to humans, clinical trials, randomised controlled trials (RCTs), case reports and English language articles from January 2013 to September 2018. Full articles of all citations resulting from this search were obtained. We scrutinised all articles for details of the methodology used to obtain an unknown primary tumour. Two reviewers independently screened all identified studies by title and abstract for further full-text review and then independently reviewed these studies for eligibility. When multiple studies were published by a single institution, only the most recent study was included to avoid inclusion of the same patients more than once in this review.
Disagreements were resolved by consensus. Data from the included studies were extracted and entered onto a Excel spreadsheet for collation and analysis.

| RE SULTS
A total of 274 cases of CUP were included from 12 case series  Table 1.
The average identification rate of the primary oropharyngeal tumour using TORS was 72% (range 17%-90%).
In 142 cases, the primary tumour was identified in the BOT. Fifty four (54) cases involved the palatine tonsil. Five studies report on mucosectomy of the base of tongue (BOT) only (without palatine tonsillectomy) in case of CUP. 14,16,18,20,21 The studies reviewed reported a range of 55%-96% positivity of HPV/p16 in CUP. In 60% (range 0%-85%) of all detected CUP, negative margins were observed after TORS resection.( Table 2).
The complications reported were re-classified according to the Clavien-Dindo classification of surgical complications and are presented in Table 3. 11-23 • TORS is a meaningful adjunct in the diagnostic and therapeutic pathway in CUP and should be considered to be structurally implemented in the guidelines of CUP.

| D ISCUSS I ON
an improved disease-specific survival and overall survival compared to non-HPV-related OPSCC's. Since occult tumours often prove to be located in the oropharynx, a relatively large percentage of HPVpositive OPSCC's present a CUP. 24 In the non-surgical workup, detection methods for CUP in the head and neck region changed significantly over the years, with the introduction of CT, MR, PET, NBI and, more recently, p16 and HPV DNA testing. 6,24 (18)F-FDG-PET/CT was used in the majority of cases reviewed. It should however be noted that the sensitivity and specificity for BOT lesions on (18)F-FDG-PET/CT is moderate due to physiological isotope uptake in the lymphoid tissue of the lingual tonsils, possibly leading to false-positive results. 25 Pattani et al 26    An important footnote concerning these numbers is the relatively small number of studies on CUP and TLM (n = 3). 33 One of the reasons of the robotic's system superior detection rate of mainly occult BOT tumours is the camera of the Da Vinci ®  The results of the current study could be influenced by confounding factors such as the differences in diagnostic workup as reported in Table 2.
The emphasis in most reports on TORS for CUP is on the BOT based on the fact that an occult primary of the palatine tonsil would be identified by classic dissection tonsillectomy as well.
Tonsillectomy combined with panendoscopy is not routinely performed in all reports analysed. Palatine tonsillectomy has shown to provide cancer detection rates superior to biopsy of tonsillar tissue. 35 Differences in the surgical technique-described in five papers-of BOT mucosectomy might influence results in terms of identification rates. 11,13,17,19,21 Paleri et al 36 recommend to use a midline incision for two separate BOT specimens for proper orientation and reduction of specimen trauma (see Figure 2A,  the cost-effectiveness of TORS in case of CUP. They also advocated a sequential strategy of primary EUA with tonsillectomy followed by BOT mucosectomy when necessary in a second procedure. In their series, this seemed more cost-effective due to shorter admission time due to less postoperative pain. 16

| CON CLUS IONS
This systematic review supports the added value of TORS for the identification of primary HNSCC of unknown origin in an era of increasing incidence of HPV-positive OPSCC. The vast majority of primary (mainly HPV positive) tumours is found through TORS, and the complication rate is relatively low.
The BOT harbours the majority of occult tumours which is emphasised by the identifications rates of BOT mucosectomy.
Transoral robotic surgery for CUP may lead to de-intensification of treatment by refraining from pharyngeal radiation and/or dose de-escalation in select cases but results of forenamed de-intensification trials need to be awaited. Prerequisites for TORS in CUP of the head and neck are well defined and uniform surgical and histopathological protocols.

CO N FLI C T O F I NTE R E S T
No conflict of interest to declare for all authors.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are openly available https://www.ncbi.nlm.nih.gov/pubmed.