Introduction
(Chemo)radiotherapy (C)RT is the primary treatment modality for many laryngeal and pharyngeal cancers in the UK.Reference Paleri and Roland1 The potential detrimental impact of (C)RT on the key laryngeal functions of respiration, voice and swallowing is well recognised and long-term swallowing dysfunction has been found in up to 71 per cent of head and neck cancer (HNC) survivors.Reference Patterson, McColl, Carding and Wilson2 Speech and language therapy interventions such as swallowing exercises may have limited benefit in the context of long-term dysfunction or late-onset radiation-associated dysphagia.Reference Hutcheson, Lewin, Barringer, Lisec, Brandon and Moore3 Recognition of the devastating impact persistent laryngeal dysfunction can have on the quality of life (QoL) of HNC survivors has prompted researchers to investigate options for treatment de-escalation aiming to improve long-term functional outcome.Reference Owadally, Hurt, Timmins, Parsons, Townsend and Patterson4, Reference Nutting, Finneran, Roe, Sydenham, Beasley and Bhide5 While these studies will hopefully lead to fewer patients developing severe laryngeal dysfunction following curative (C)RT for HNC in the future, the impact of this will not be felt for some time.
At its most severe, laryngeal dysfunction may lead to airway obstruction requiring tracheostomy insertion, chronic aspiration with recurrent chest infections, reliance on enteral nutrition (EN) and loss of functional voice for communication. Patients may be offered functional salvage total laryngectomy (FSTL), with the aim of preventing recurrent aspiration pneumonia, restoring some swallowing and voice function and improving QoL.
Total laryngectomy (TL) involves the permanent separation of the airway and the digestive tract, thereby preventing aspiration. The procedure results in a permanent tracheostoma for respiration and the loss of laryngeal voice. Surgical reconstruction of the neopharynx may be with primary closure of the residual pharyngeal mucosa or with local or free tissue flap reconstruction. While TL removes the risk of aspiration, apart from in the cases of fistulisation or a leaking voice prosthesis, dysphagia post-laryngectomy is increasingly recognised as a significant problem for some patients due to factors including stricture, pseudovalleculae and reduced propulsion through the neopharynx.Reference Coffey and Tolley6 Communication options post-TL include electrolarynx, oesophageal voice, mouthing, writing and use of communication apps; however, surgical voice restoration (SVR) is considered the “gold standard” and should be offered to all eligible patients.Reference Clarke, Radford, Coffey and Stewart7 SVR can be completed at the time of TL or as a secondary procedure, and in the UK is the primary communication method for 71 per cent of all people with TL.Reference Woodman, Govender, Baker, Glaister, Rowe and Dunton8
Salvage total laryngectomy (STL), whereby TL is completed in the context of persistent or recurrent disease after primary (C)RT, is associated with higher rates of surgical complications due to the impact of previous radiation on tissue perfusion and subsequent healing.Reference Sayles and Grant9, Reference Hasan, Dwivedi, Gunaratne, Virk, Palme and Riffat10 SVR is often completed as a secondary procedure in this context to limit risk of surgical complications.Reference Brunet, Rovira, Simo, Remacle and Eckel11 Return to oral intake following STL has been shown to be less likely when there are post-operative complications.Reference Meulemans, Demarsin, Debacker, Batailde, Mennes and Laenen12 FSTL, where radical surgery is proposed for functional benefit and in the absence of malignancy, is therefore not a straightforward choice for patients, and information on expected functional outcome is crucial to support patient selection, counselling and decision-making.
There is limited but emerging evidence regarding outcome following FSTL, with studies to date describing small case series in the USAReference Hutcheson, Alvarez, Barringer, Kupferman, Lapine and Lewin13–Reference Evangelista, Nativ-Zeltzer, Bewley, Birkeland, Abouyared and Kuhn18 and the Netherlands.Reference Theunissen, Timmermans, Zuur, Hamming-Vrieze, de Boer and Hilgers19 These studies report on cohorts of patients with limited numbers (range 15–32) and median time from completion of treatment to surgery ranging from 2.8 to 15.5 years. A recent review found that, although TL as treatment for end-stage laryngeal dysfunction offered benefits with regard to prevention of aspiration and alleviation of nil by mouth (NBM) status, many patients still suffer some chronic functional impairment.Reference Olinde, Evangelista and Bewley20 Reported outcomes are variable, with 5–24 per cent of patients remaining reliant on EN at six months post-surgery, and 35–89 per cent of patients eventually achieving functional communication with SVR.Reference Hutcheson, Alvarez, Barringer, Kupferman, Lapine and Lewin13–Reference Wu, Goldsmith, Holman, Kammer, Parikh and Devore16, Reference Theunissen, Timmermans, Zuur, Hamming-Vrieze, de Boer and Hilgers19 Complication rates and functional outcomes have been found to be similar or better for FSTL in comparison with STL,Reference Coviello, Sheehan, Hernandez, Liou, Sandulache and Haskins17 and a recent study has shown improvements in patient-reported swallow function and QoL rating and a reduction in general distress following FSTL.Reference Evangelista, Nativ-Zeltzer, Bewley, Birkeland, Abouyared and Kuhn18
We aimed to investigate the communication and swallowing outcomes for patients undergoing FSTL at our London centre, to help inform patients and their health care teams when considering treatment with FSTL.
Materials and methods
We completed retrospective chart review of all patients treated with FSTL for laryngeal dysfunction following curative (C)RT for HNC at our tertiary cancer centre from 2009 to 2023. Institutional review board approval was obtained for this service evaluation. Patients undergoing functional laryngectomy with other aetiologies (e.g. intractable aspiration associated with a neurological diagnosis) were not included.
Measures were recorded at pre-surgical baseline, point of discharge from inpatient admission and six and 12 months post-surgery and extracted from the electronic record for analysis. Communication outcome was measured by determining each patient’s primary mode of communication. Swallowing outcome was measured with the Functional Oral Intake Scale (FOIS), a clinician-rated ordinal scale of a patient’s oral intake (Table 1).Reference Crary, Carnaby Mann and Groher21 A FOIS score of 1 to 3 indicates feeding tube dependence with varying amounts of oral intake. Scores of 4 to 7 describe full oral intake with or without modifications or adaptations, with scores 6 and 7 indicating a normal diet with or without some avoidances, respectively.
Functional Oral Intake ScaleReference Crary, Carnaby Mann and Groher21
Additional data collected included patient demographics, previous disease and treatment and time since completion of treatment, pre-operative airway status, duration of dependence on enteral feeding and/or tracheostomy pre-operatively and type of flap reconstruction. Data on pulmonary health including prevalence of aspiration pneumonia were unfortunately not available in the electronic record. Descriptive statistics including frequencies, percentages, medians and ranges were used to report results. Further statistical analysis of findings was not possible due to small sample size.
Results and analysis
Ten patients were identified. Patient characteristics and functional outcomes are summarised in Table 2. The median age at time of surgery was 70 years (range 36–79 years) and 70 per cent were male. Four patients (40 per cent) previously underwent radiotherapy alone and six (60 per cent) received concomitant chemotherapy (one [10 per cent] also received neoadjuvant chemotherapy). The median length of time since completion of (C)RT was three years eight months (range four months - 13 years). Primary tumour sites were oropharynx (60 per cent), larynx (40 per cent) and hypopharynx (10 per cent), and all patients were confirmed to be clear of any locally recurrent disease prior to surgery. Half the sample (50 per cent; n = 5) required tracheostomies pre-operatively due to upper airway obstruction, and had been living with a tracheostomy for a median of eight months (range one month - four years five months). Over half of the patients with tracheostomy (60 per cent; n = 3) required a cuffed tube to manage aspiration of secretions. The majority of all patients had free flap reconstruction with antero-lateral thigh (70 per cent), and two patients experienced significant post-operative complications with flap failure requiring secondary reconstruction with pectoralis major pedicled flap. No other significant surgical complications were documented. Free flap reconstruction was an exclusion criterion for primary SVR. All patients were alive and disease-free at 12 months post-surgery.
Patient characteristics and communication and swallow outcomes at baseline, point of inpatient discharge and six and 12 months post-operatively
Abbreviations: ALT, antero-lateral thigh; FOIS, Functional Oral Intake Score; SVR, surgical voice restoration.
Primary mode of communication (Figure 1)
Pre-FSTL
At baseline, 30 per cent (n = 3) were unable to produce voice due to the presence of a cuffed tracheostomy and were reliant on mouthing or writing, while the remaining 70 per cent (n = 7) were communicating verbally.
Primary mode of communication over time.
Post-FSTL
End of surgical admission: One patient (10 per cent) who did not require free flap reconstruction underwent primary SVR at the time of surgery and was using this functionally by the end of surgical admission. 30 per cent (n = 3) were using an electrolarynx effectively, 60 per cent (n = 6) were reliant on mouthing, writing or communication apps.
Six months post FSTL: Secondary SVR had been completed for two further patients (30 per cent in total, n = 3). Six patients (60 per cent) were using mouthing/writing/apps, and one patient (10 per cent) was using an electrolarynx.
Twelve months post FSTL: A further two patients had undergone SVR (50 per cent, n = 5 in total), though only 40 per cent (n = 4) were using this as primary mode of communication. One patient declined SVR, preferring to continue with mouthing/app rather than taking on the additional care required for a voice prosthesis. The remaining 40 per cent (n = 4) had not been offered SVR by 12 months due to factors including prolonged healing and neopharyngeal stenosis. At 12 months, only one patient (10 per cent) continued to use electrolarynx as their primary mode of communication. Five (50 per cent) were using mouthing/writing/apps. All three patients (30 per cent) who had been unable to voice pre-operatively due to cuffed tracheostomy remained reliant on mouthing/writing/apps at 12 months post-surgery.
Swallowing (Figure 2)
Pre-FSTL
At baseline, all patients were NBM due to severe dysphagia and had been reliant on EN for a median of 10 months (range one month - five years one month).
Functional Oral Intake Score over time.
Post-FSTL
End of surgical admission: 20 per cent (n = 2) were tolerating full oral intake of a normal textured diet (FOIS = 6–7), and 20 per cent (n = 2) required diet modifications but no EN (FOIS = 4–5). Most patients were still reliant on EN with 40 per cent (n = 4) managing some oral intake in addition (FOIS = 2–3) and 20 per cent (n = 2) still NBM (FOIS = 1).
Six months post FSTL: Most patients were on full oral intake, with 60 per cent (n = 6) tolerating normal diet (FOIS = 6–7), and 10 per cent (n = 1) still requiring some diet modifications (FOIS = 4–5). A total of 20 per cent (n = 2) required EN to support oral intake (FOIS = 2–3), and 10 per cent (n = 1) remained NBM (FOIS = 1).
Twelve months post FSTL: The number of patients on full oral intake of normal diet (FOIS = 6–7; 60 per cent; n = 6) and modified diet (FOIS = 4–5; 10 per cent; n = 1) was stable. Three patients (30 per cent) remained reliant on EN due to neopharyngeal dysphagia but all were tolerating some oral intake (FOIS = 2–3). All three of these patients underwent contrast swallow assessment under fluoroscopy within the assessment period with no discrete stricture identified.
Discussion
To our knowledge, this is the first UK-based case series reporting communication and swallowing outcomes for patients undergoing FSTL for non-functioning larynx following curative (C)RT for HNC. In this small cohort where all patients were NBM at baseline, most (n = 7) were able to tolerate full oral intake with some special preparations or compensations by six months post-surgery, and this was maintained at 12 months. All patients were able to tolerate some oral intake by 12 months though a minority remained reliant on EN. Almost half (n = 4) were using functional SVR by 12 months post-surgery. While this demonstrates positive outcome for some patients, there remains a significant proportion with persistent functional impairment at one year post-surgery. Almost a third of our cohort was both reliant on EN and without functional SVR at one year follow-up.
There are likely to be a number of factors associated with the rates of SVR in this study. By 12 months post-surgery, 40 per cent of the cohort had not yet been offered SVR, but it is possible that this may change with longer term follow-up as reasons included prolonged healing. SVR can be delayed for a number of reasons including patient factors (e.g., prolonged healing, stricture) or institutional capacity to complete additional procedures though this was not evident in this cohort. Previous work has shown a significant association between communication and swallow outcomes after FSTL.Reference Hutcheson, Alvarez, Barringer, Kupferman, Lapine and Lewin13 In our study, all those who were using SVR functionally at 12 months were also tolerating full oral intake of normal or soft diet, so there may be potential for the other patients with good swallow outcome to also achieve functional SVR in time. All three patients that were reliant on mouthing/writing pre-operatively remained so at 12 months, which may suggest that the presence of cuffed tracheostomy at baseline has an impact on outcome. The potential for a prolonged period without voice should be discussed with patients considering FSTL.
Our findings are broadly in line with outcomes from previous studies, although the proportion of patients not achieving full oral intake was slightly higher than has been previously reported. It is notable that, while previous case series included 73.9-89 per cent of patients reliant on EN pre-operatively, all patients in the current study were EN-dependent at baseline and had been so for a median of 10 months. This may suggest a correlation between severity of baseline dysfunction and potential outcome. The number of patients requiring tracheostomy for airway management at baseline in the present study (50 per cent) is in line with a range of 39.1–79 per cent previously reported.Reference Hutcheson, Alvarez, Barringer, Kupferman, Lapine and Lewin13–Reference Wu, Goldsmith, Holman, Kammer, Parikh and Devore16, Reference Evangelista, Nativ-Zeltzer, Bewley, Birkeland, Abouyared and Kuhn18, Reference Theunissen, Timmermans, Zuur, Hamming-Vrieze, de Boer and Hilgers19
In our study, the two patients who had been reliant on EN for more than two years pre-operatively were both still reliant on EN at 12 months post-surgery and neither had functional SVR. It is not clear from chart review whether these patients had been offered surgery at an earlier date and previously declined. Other case series have not reported on duration of pre-surgical impairment, so it is unclear whether this could be a potential predictor of functional outcome. A longer period of impairment may correlate with increased fibrosis and impact on subsequent outcome; however, conversely an extended period of EN may result in nutritional optimisation and therefore offer better outcomes. It has been suggested that the decision to proceed with FSTL is often postponed by patients and healthcare teams, and that this decision should perhaps be made earlier when patients are likely to be in better overall condition.Reference Theunissen, Timmermans, Zuur, Hamming-Vrieze, de Boer and Hilgers19 The importance of early identification and referral for potential candidates for FSTL is highlighted in a recent prospective study as two patients died of recurrent aspiration pneumonia before their surgical date.Reference Evangelista, Nativ-Zeltzer, Bewley, Birkeland, Abouyared and Kuhn18 Further work on the relationship between duration of impairment and functional outcome is required to determine whether this could be a predictive factor.
While there may be potential for further change over time for some patients, in our study there was little improvement in outcomes from six to 12 months. This suggests that functional gains can be achieved fairly quickly for many patients but that those with poor function at six months are likely to experience persistent impairment. Indeed, previous work has highlighted the risk of deterioration in function over time, with reduced swallow efficiency following STL and FSTL requiring neopharyngeal dilatation.Reference Meulemans, Demarsin, Debacker, Batailde, Mennes and Laenen12, Reference Hutcheson, Alvarez, Barringer, Kupferman, Lapine and Lewin13 Although no dilatations were documented within the first post-operative year for this cohort, it is possible that some patients went on to undergo dilatation after that time point. In view of this, further longitudinal data are required to describe the longer-term trajectory of function following FSTL.
There are a number of limitations to the present study. As a retrospective review, only data that had been clearly recorded could be extracted for analysis. Frequency and severity of chest infections prior to or post-surgery was not documented in the electronic record, for example, and so we cannot comment on the impact of surgery on this important health outcome. Significantly reduced rates of pneumonia have been reported following FSTLReference Hutcheson, Alvarez, Barringer, Kupferman, Lapine and Lewin13 and the potential to prevent recurrent chest infections associated with aspiration is likely to be a key motivator for proceeding with surgery for many patients. Information on pulmonary health status should be included in comprehensive assessment of patients considering and proceeding with FSTL at baseline and in follow up.
A major limitation is the lack of patient reported outcome measures (PROMs) to demonstrate the impact of FSTL on QoL. While decreased levels of QoL have been reported initially following total laryngectomy,Reference Singer, Danker, Guntinas-Lichius, Oeken, Pabst and Schock22, Reference Wulff, Højager, Wessel, Dalton and Homøe23 there is very little evidence around QoL specifically for patients undergoing FSTL. Since patients electing to proceed with FSTL are likely to do so primarily aiming for improvement in function, it is reasonable to assume that most patients also hope for an improvement in QoL and the limited available evidence suggests this may be achieved.Reference Evangelista, Nativ-Zeltzer, Bewley, Birkeland, Abouyared and Kuhn18 Future work should include the prospective longitudinal collection of PROMs, and could aim to compare QoL outcomes following FSTL with the QoL for those patients who have dysfunctional larynx but do not elect to proceed to FSTL.Reference Theunissen, Timmermans, Zuur, Hamming-Vrieze, de Boer and Hilgers19
The cohort described here is smaller than those reported in previous studies,Reference Hutcheson, Alvarez, Barringer, Kupferman, Lapine and Lewin13–Reference Theunissen, Timmermans, Zuur, Hamming-Vrieze, de Boer and Hilgers19 and due to our small sample size statistical analysis was not possible. A larger, multicentre study pooling data from across the UK would allow for more detailed investigation of the factors which may influence or predict functional outcome following FSTL. A USA study found that continuing to smoke after initial treatment and a history of recurrent or chronic pneumonia were associated with dependency on EN following FSTL, and that extent of surgery and type of flap reconstruction were not significant factors.Reference Hutcheson, Alvarez, Barringer, Kupferman, Lapine and Lewin13 Further data are needed to help identify additional factors that impact on communication and swallow outcomes. This would provide crucial information for patients and their healthcare teams to inform decision making in the management of non-functioning larynx following curative (C)RT for HNC.
• Functional salvage total laryngectomy can be offered to manage the effects of severe laryngeal dysfunction following curative radiotherapy
• Salvage surgery carries risk of complications and poor outcome; therefore, evidence is needed to help inform patients and their healthcare teams
• Current evidence is limited, and prior to this work, no UK outcome data were available
• At 12 months post-surgery, all patients in this group tolerated some oral intake though a minority remained dependent on tube feeding. Fewer than half of the group had achieved functional tracheoesophageal voice
Conclusions
In this UK single-centre retrospective analysis of patients undergoing FSTL for non-functioning larynx following curative (C)RT for HNC, most patients were able to re-establish full oral nutrition with some compensations post-surgery, but a minority required ongoing EN at 12 months. Only 40 per cent of patients were using SVR as their primary mode of communication by 12 months post-surgery, and half of patients relied on a combination of mouthing, writing and communication apps. Evidence on functional outcome and impact on QoL is vital for patients considering proceeding with FSTL, and variability in outcome must be explained to patients who are offered surgical management of non-functioning larynx. Further work with a larger cohort is required to provide more robust data on outcome and to identify factors that influence longer term function.
Acknowledgments
The authors would like to acknowledge the contributions of the SLT department at Guy’s Hospital and from referring hospitals, and Mr Richard Oakley and Mr Aleix Rovira, Consultant Head and Neck Surgeons. Preliminary findings of this work were presented in poster format at UKSRG and BAHNO meetings in 2024. Institutional review board approval was obtained for this service evaluation.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
The authors declare none.
