TBM improves the detection rate of occult malignancy and has been shown to half the number of true SCCOP after PET CT and EUA UADT with tonsillectomy [5]. Its inclusion as a diagnostic technique in national guidelines reflects its efficacy and indicates its use will continue to increase [4]. However, TBM has its own challenges and comes with potential complications. The need to identify a primary is driven by the morbidity of radical treatment. Extensive irradiation of both sides of the neck and pharyngeal mucosa with concurrent chemotherapy results in significant xerostomia and dysphagia. The rate of dilatation for oesophageal stricture after TMR is 40% [30]. However, there is limited data on morbidity caused by TBM, which is implemented to target adjuvant radiotherapy . Previous systematic reviews do not explore functional outcomes after TBM [10, 31].

Saliva production is a problematic area for patients post head and neck cancer treatment. Xerostomia is a widely reported side-effect of radiotherapy and it has been shown to be the most important domain to patients in the UW-QOL score, with 33% choosing this side-effect as most problematic [32,33,34]. In our patient cohort saliva production had the lowest median UW-QOL score of 70, a figure that correlates with “less saliva than normal, but enough.”. There were no reports of dysphonia by patients in the UW-QOL score, or by speech and language therapist assessments. Median post-operative pain score was 100 at 24.5 months post-treatment. The lack of pain immediately post-operatively is demonstrated by a return to soft diet within 24 h, optimised by our local protocol which provides patients with regular opioid-based analgesia. Mehta et al. (2013) reported similar findings, with all five of their TBM patients reporting UW-QOL pain scores of 75–100 [19]. Pain management should be a priority in order to optimise swallow rehabilitation. Using topical local anaesthetic and consulting a pain specialist both help, the latter may adjunctively treat with neuropathic agents.

Dysphagia is a common effect of radiotherapy and the median UW-QOL swallowing score after completion of oncological treatment in our case-series was 70. This is lower than reported outcomes for oncological treatment alone which range between 80 and 84 [11, 35]. However, the median normalcy of diet score shows that TBM patients can expect to manage full diet with or without liquid assist after completion of treatment. Follow-up by the speech and language team is essential in rehabilitating swallow and monitoring improvement. The median MDADI emotional score in our series was 67. This is an aspect of swallow function that can often be overlooked and it is therefore important to address this with psychological support if necessary. The functional score reflects ability to perform daily activities and had a median score of 80 while the median physical score of 65 reflects the action of swallowing, weight maintenance and aspiration.

The global MDADI score consists of only one statement: “My swallowing ability limits my day-to-day activities,” which in this series gave a median global score of 80 (“Disagree”). This finding can be used to counsel patients, that change in swallow is unlikely to impact their day-to-day activities after TBM and curative treatment, along with the finding that all patients were gastrostomy free at latest follow-up. Contrary to the differences in UW-QOL swallowing score previously discussed, the global MDADI score compares favourably to chemoradiotherapy and radiotherapy in the literature – which gives a mean global MDADI score at 3 months of 61.3 for chemoradiotherapy, 66.3 for radiotherapy 63 Gy in 30 fractions and 76.8 for radiotherapy 50 Gy in 16 fractions [32]. This further supports the argument for TBM in potentially reducing the volume and therefore the total dose of radiation.

Owen et al. (2017) assessed whether pre-treatment swallowing measures predict swallowing recovery at 6 weeks after trans-oral robotic TBM [16]. The patients had a mean of 14 days of post-operative tube feeding, similar to transoral robotic assisted tongue base tumour resections (19 days). The mean length of tube feeding was likely impacted by the small patient numbers and early experience in the evolution of RTBM. The authors concluded that poor pre-operative swallow correlated with post-operative duration of tube feeding. This emphasises the need for formal swallow assessment and documentation pre-operatively, so patients that are more likely to have poor swallow outcomes post TBM can be identified and start “pre-habilitation”. That said, without extensive neck nodal disease, it is unlikely that those without an identifiable primary will have a poor swallow. Comorbidities should be considered when identifying high risk patients. For instance, chronic obstructive pulmonary disease can increase the likelihood of a gastrostomy tube need, as these patients are more vulnerable to aspiration pneumonia. The effects of age and neuromuscular disorders on swallowing function and recovery is not clearly demonstrated in the literature but should be considered in pre-operative work-up.

Our case series shows that the majority of patients can return to a soft diet within 12 h of TBM, with or without tonsillectomy and neck dissection. Three patients in our case-series had a delay in return to soft diet over 4 days. One did not return to soft diet for 9 days due to post-operative pain and UTI-related delirium. His oral intake was supplemented by nasogastric feed for a further 6 days. The second patient did not return to soft diet until 7 days post-operatively, due to poor initial pain control. The third patient did not return to soft diet before commencing radiotherapy – he had a gastrostomy tube inserted prophylactically, which was removed after completing treatment. His swallow was likely impacted by the need for early re-intubation for tongue swelling and intensive care admission for 4 days. At latest follow-up 41 months post-treatment, he had a functional MDADI score of 68 and normalcy of diet score of 50, correlating with soft chewable foods. An uneventful return to oral intake for the other 85% (17/20) of patients resulted in a median length of nasogastric feeding of 0 days. It has therefore become our institution’s practice to not routinely insert a nasogastric tube after TBM.

As illustrated in Table 1, previous studies have not addressed taste or speech outcomes and only one study has addressed pain outcomes after robotic TBM. Swallowing outcomes were described in 50% (9/18) of TBM studies. Those that addressed swallowing outcomes did not use validated head and neck specific assessments tools of dysphagia, such as MDADI and only one study reported beyond swallow function beyond the immediate post-operative period [28]. Those that reported on swallow outcomes showed rapid return to normal swallow [5, 14, 15, 25], apart from Owen et al. (2017) as previously mentioned [16]. The use of laser or endoscopic monopolar to undertake TBM also reported return to normal swallow [22,23,24, 26]. Hardman et al. (2024) reported on the outcomes of step serial sectioning histopathology from patients included in a prospective multicentre study on TBM. They are gathering MDADI and pain scores on these patients, but the functional outcomes are yet to be reported [29, 36]. The authors hypothesize a return to near normal swallow function occurs 6 weeks post-surgery, and subsequent swallow deterioration to be due to adjuvant radiotherapy [36].

Post-operative haemorrhage rates are similar in most studies to those after tonsillectomy [37]. It results in readmission and in some cases a return to theatre. As a life-threatening event, this should be reported consistently. The random effects pooled bleeding rate is 5% (95% CI 0.03: 0.08, I2 = 0%, = 1.0).

The majority of patients in our study had a bilateral tonsillectomy at the time of TBM with the aim of improving the primary identification rate. In the literature, there is currently no consensus as to whether unilateral or bilateral tonsillectomy is the optimum approach. We identified no malignancy in the contralateral tonsil, which reflects the low rate of contralateral tonsil malignancy in the literature: Farooq et al. (2019) described a contralateral tonsil malignancy rate of 0.9% [10] and Hardman et al. (2024) identified no malignancy in the contralateral tonsil [29]. The most recent United Kingdom guidelines state that at least unilateral tonsillectomy should be undertaken alongside TBM, and that contralateral tonsillectomy should be considered [2]. More data is required to identify if swallow function differs after unilateral or bilateral tonsillectomy alongside TBM and the benefit of a slightly increased detection rate should be weighed against the potential additional swallow morbidity of bilateral tonsillectomy [2].

Swallow, saliva production and pain are the three most important issues reported by head and neck cancer patients in their treatment [38]. The anticipated increase in TBM procedures indicate a need for consistent reporting of morbidity and studies with larger cohorts to quantify these outcomes after diagnostic TBM. Our case series is the first to employ a validated scoring method to assess swallow, pain, speech, taste and saliva production after TBM beyond the immediate post-operative period. Looking forward, a multi-centre database is required to record prospective functional and oncological outcomes of TBM patients.

Limitations of this study include the number of patients in our cohort, although this seems to be similar to other studies. The follow-up is variable in our study and the literature. Our study shows the variation in post-TBM oncological treatment, such as radiotherapy, chemoradiotherapy or nothing, and surgical procedures such as tonsillectomy and neck dissection concurrently or consequently. This makes it difficult to ascertain the impact of each modality on quality-of-life outcomes. The variation or lack of standardized measures used for quality-of-life outcomes precludes comparison of many functional aspects. The focus in the literature has been on oncological outcomes, rather than quality-of-life data. What constitutes a TBM and variation in resection samples may also vary across different centres.