Patient experience, functional and quality of life outcomes in patients receiving surgical and non-surgical treatment for residual, recurrent, or new oropharyngeal cancer in a previously irradiated field: A systematic review

1Introduction

Head and neck cancer (HNC) is the sixth most common cancer in the world and is increasing in incidence (Hardman et al., 2020). Squamous cell carcinomas (SCC) account for the majority of these tumours, with an increasing number of oropharyngeal cancers associated with the human papilloma virus (HPV). Treatments for primary disease include (chemo)radiation and/or surgery, with many patients undergoing multi-modality treatment. The impact of such treatments on function, including swallowing, speech, voice and overall quality of life (QoL), in the primary disease setting have been well documented (Patterson, McColl, Carding, & Wilson, 2018; Roe, Drinnan, Carding, Harrington, & Nutting, 2014).

Despite ongoing advancements in the treatment of primary oropharyngeal disease, rates of residual (diagnosed within 12 months of previous treatment), recurrent (diagnosed between 12 months and 5 years of previous treatment) and second primary disease (diagnosed > 5 years following previous treatment) remain high at approximately 20–30% (Hardman et al., 2020; Leeman et al., 2017; Mandapathil et al., 2014; Warnakulasuriya, 2009).

Residual, recurrent or new primary (ReRuNeR) oropharyngeal cancer (OPC) present some of the greatest challenges in HNC practice (Brady, Hardman, Paleri, Harrington, & Roe, 2020). Not only do these patients often present with functional issues such as difficulties with swallowing, speech and/or voice due to long-term/ late-onset effects from their prior treatment (Patterson et al., 2018), but the newly diagnosed disease and/ or the proposed treatment plan may further cause or compound such difficulties. In addition to disease and treatment-related issues, patients who have undergone treatments for HNC are also known to have a high level of psychological stressors, which can further add to symptom burden and overall QoL (Ringash et al., 2018).

The UK standard of care for ReRuNeR OPC is open surgery, an intervention associated with high morbidity (Mehanna, Kong, & Ahmed, 2016). Transoral robotic surgery (TORS) is now being considered as a minimally invasive option with encouraging oncological control (Paleri, Hardman, Brady, George, & Kerawala, 2020). There may be opportunities for re-irradiation; however, this has the potential for severe treatment toxicity. Non-curative options include palliative systemic treatments or best supportive care. More recently, immunotherapy is being utilised. In the UK, pembrolizumab, with or without platinum-based chemotherapy, is now used for untreatable metastatic or unresectable recurrent head and neck squamous cell carcinoma in patients whose tumours express a specific biomarker (PD-L1 with a combined positive score of 1 or more) (NICE, 2020).

Historically, patients with ReRuNeR OPC were considered to have a very poor prognosis with the majority of patients offered palliative treatments or best supportive care. However, over the past 20 years, there has been a paradigm shift in the treatment of recurrent HNC, with evidence to support increased survival rates for patients treated curatively using surgery, in particular for patients with ReRuNeR OPC (Jayaram et al., 2016). For those patients who are not suitable for surgical resection, the focus in recent literature has been on enhanced life-prolonging treatments such as the use of immunotherapy.

National guidelines suggest that patients diagnosed with ReRuNeR disease should be fully counselled on the likely functional and QoL impact of the available treatments (Mehanna et al., 2016).However, although there has been an apparent paradigm shift in studies investigating enhanced survival for patients with ReRuNeR OPC, there appears to be a lack of literature examining functional, QoL and patient experience outcomes for all treatment modalities in ReRuNeR OPC.

The aim of this systematic review is to identify if functional and QoL outcomes and patient experience data are reported, how they are measured and what are the functional and QoL outcomes and patient experience data reported for patients with ReRuNeR OPC. Working definitions for functional outcomes, QoL and patient experience as used in this review will be provided under methods.

2Methods

This systematic review used a population, intervention, comparison, outcome (PICO) framework (Schardt, Adams, Owens, Keitz, & Fontelo, 2007). This framework was divided as follows to report on the following: for patients with ReRuNeR OPC (Population), undergoing curative or non-curative treatment (Intervention), are functional, QoL and patient experience measures reported, and if so, how are they reported and what are the findings (Outcomes). A comparator was not a requirement for this review.

The Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines were used to carry out the systematic review (Moher, Liberati, Tetzlaff, & Altman, 2010). The protocol was registered on PROSPERO an international prospective register of systematic reviews (CRD42021236540).

The main objectives of the systematic review were:

3Results

3.1Study selection

The literature search findings are summarised in Fig. 1 PRISMA flow diagram. Following the removal of duplicate records, 796 studies were identified and 47 included for full text review with 17 studies included for data extraction. The primary reason for exclusion was that only survival outcomes only reported (n = 16); this included studies focusing on surgery (n = 10), chemotherapy (n = 3), re-irradiation (n = 2) and immunotherapy (n = 1). Other reasons for exclusion included general HNC cohort where functional/ QoL or general toxicity outcomes were reported but with no subgroup analysis for patients with ReRuNeR (n = 10) or mixed primary/ ReRuNeR OPC cohorts (n = 1). One qualitative study was identified; however, it was excluded as it was not specific to ReRuNeR OPC but a more general recurrent HNC caseload.

Fig. 1

PRISMA diagram.

4Discussion

This systematic review aimed to assess if function, QoL and patient experience are measured in patients undergoing treatment for ReRuNeR OPC. Only 17 articles were identified reporting functional and/or QoL outcomes for patients undergoing treatments for ReRuNeR OPC, five reporting functional/QoL as a primary measure. All 17 articles focused on surgical trials. No specific functional, QoL, patient experience of qualitative data for patients with ReRuNeR OPC undergoing non-surgical treatments were identified in this systematic review.

The measurement of HR-QoL in HNC has been recommended in both clinical practice and research over the past number of years (Rogers, Semple, Babb, & Humphris, 2016). The routine use of validated specific functional outcome measures to guide rehabilitation for communication and swallowing in HNC continues to emerge and has only recently been added as a recommendation in national guidelines (Schache et al., 2021). As a result, this review included previously validated measures of HR-QoL and all measures (validated and unvalidated) of communication and swallowing function were included.

Aside from reporting survival data only, many articles were excluded from this review due to the inclusion of mixed cohorts of HNC patients, with either general recurrent HNC cohorts or mixed primary and recurrent disease groups, with no subgroup analysis for ReRuNeR OPC. In comparison to other subtypes of head and neck cancer, the OPC subgroup are considered quite unique, in particular those with HPV related disease, who are often younger (less than 60 years) with low co-morbidity, absent or low smoking history (Ang et al., 2010). The incidence of OPC disease is increasing in the UK (Schache et al., 2016). This may be in part due to the emergence of new aetiological categories such as human papilloma virus (HPV), however, non-HPV-related disease which is associated with traditional causative factors including alcohol and tobacco consumption is also increasing (Schache et al., 2016). In the UK, many of these patients will have undergone organ-sparing treatment regimes with radiation therapy +/- chemotherapy. As a result, these patients will most likely be presenting with treatment side-effects in the form of changes to swallowing function, voice/speech and overall QoL (Patterson, McColl, Carding, & Wilson, 2018; Roe, Drinnan, Carding, Harrington, & Nutting, 2014). In addition, there can be a high level of psychological distress and emotional burden following previous treatment (Ringash et al., 2018). Further treatment for disease recurrence may cause further impairments or compound a unique set of baseline physical and psychological burden found in a potentially unique set of younger patients with OPC. With a growing number of patients with ReRuNeR OPC in the UK, specific functional and QoL data in relation to this unique cohort is required.

It is known from previous literature in the primary oropharyngeal disease that to accurately demonstrate the impact of treatment on function/ QoL, multidimensional assessment at pre-defined timepoints is required (Roe et al., 2010). In this review, 7 of the included studies did not report on baseline data and, a further 8 studies reported post-surgical functional and/or QoL data however did not define the follow up period. Pre-treatment baseline level of function has been cited as the key predictor for longer term swallowing outcomes for patients with HNC (Frowen, Cotton, Corry, & Perry, 2010). Given the risk of baseline functional difficulties in patients with ReRuNeR disease (Hardman et al. 2020), accurate baseline and longitudinal data measurements with pre-defined follow-up timepoints are imperative to accurately measure any potential change. Such methodological issues, in addition to the use of different outcome tools, make interpretation of the collective data difficult and any form of meta-analysis impossible.

This systematic review identified just three studies reporting speech outcomes. The standard of care for curative management of ReRuNeR OPC is at present open surgery. National guidelines highlight the potential functional morbidity including speech (Mehanna et al., 2016), however, based on this review, speech outcomes do not appear to be reported routinely. In the studies reporting speech outcomes, only one study used a previously validated speech outcome measure however, all studies did demonstrate altered speech in the post operative phase.

The majority of studies (n = 14) included data on gastrostomy use post-surgery. Eleven of these studies included gastrostomy tube dependence with or without tracheostomy tube dependence as the only measures of function post-surgery. In our review, pooled data (n = 128) demonstrates a long-term gastrostomy feeding rate of almost a quarter (23.42, 95% CI 10.2 to 36.6). In comparison to a previous systematic reviews reporting outcomes for TORS for recurrent HNC where long-term gastrostomy rate was recorded as 5% (Paleri et al., 2020), this estimate does seem to be higher. This may be due to the previous review including all HNC cohorts (not specific to ReRuNeR OPC) or only one treatment method (TORS rather than all surgery). Also, it is noted by the authors here that the definition of long-term use is not clear in the included studies. The current review defined long-term gastrostomy use as at least 6 months post-surgery. Given such ambiguities, the case for future studies to look at specific subgroups of HNC for more accurate/ clearly defined functional outcomes is strengthened. Gastrostomy tube use is widely used in the literature as a proxy for swallowing outcomes. As a unidimensional measure, it does not give an accurate/holistic measure of swallowing status. Many HNC patients may be having a combination of both oral intake and gastrostomy feeding. Also, without the provision of baseline data it is impossible to ascertain if gastrostomy tube presence is a consequence of primary disease management or because of the treatment of ReRuNeR disease.

There appears to be some variation in outcomes in relation to the studies using patient- and clinician-reported measures of swallowing function. It is known from the literature in relation to primary disease that clinician and patient-reported measures of toxicity are not always congruent (Falchook et al., 2016). Both within and across surgery types (TORS and open), mixed findings are reported using the MDADI. One study reports stable post-TORS findings (Williamson et al., 2021) and another reported a deterioration (Charters et al., 2022), although both of these studies included only 3 patients each. A further study, with a much larger cohort (n = 53) (D’Andréa et al., 2022) reports outcomes for both open and TORS surgery and details an overall mean deterioration in MDADI composite scores but that deterioration did not meet statistical or clinical significance. A previous study investigated the clinically meaningful difference in MDADI scores for patients with HNC using a combination of criterion standards including aspiration scores, feeding tube status and diet scores (Hutcheson et al., 2016). The methodology for this study however excluded patients with recurrent disease. It might thus be questioned if the MDADI and the critical 10-point clinically significant difference is suitable in the recurrent setting when we know that most of these patients will have baseline swallowing difficulties with lower diet scores, higher rates of gastrostomy tube usage and increased aspiration from their previous treatments +/- tumour related complications (Paleri et al., 2020).

With regards to QoL measurement, only 2 trials were identified where validated health-related quality of life tools were used, namely the UW-QoL questionnaire and the EORTC QLQ-C30 used in combination with the EORTC QLQ-H&N35. In an area of HNC practice where many treatment decisions are based on the potential impact of further treatments on QoL, one would expect a much greater focus within the literature. Although stable outcomes were reported by Williamson and colleagues (Williamson et al., 2021) it would appear that some aggregate data is missing in the publication. Scoring of the UW-QoL tool typically involves presenting the questionnaire data from two subscale scores, one for ‘Physical Function’ and another for ‘Social-Emotional Function.’ The Physical subscale score is computed as the simple average of 6 domain scores –those of chewing, swallowing, speech, taste, saliva and appearance. The Social-Emotional subscale score is also computed as the simple average of 6 domain scores - those of anxiety, mood, pain, activity, recreation and shoulder function. Such scores are not provided in this publication thus difficult to interpret. Also, a small sample size is noted in this study with just three patients. In a larger study by D’Andréa et al. (2022) non-significant changes in EORTC-QLQ-C30 scores at baseline and 1 year were reported. However, the EORTC QLQ-H&N35 demonstrated some increased difficulties with sticky saliva, difficulty with social contact, eating and speech. This echoes findings by Williamson and colleagues who, when reporting symptoms using the UW-QoL tool, noted increased symptoms in relation to swallowing, speech, taste, saliva, in addition to shoulder, appearance, recreation and mood. However, as noted in previous literature, the content validity of EORTC QLQ-C30 and QLQ-H&N35 (Degboe et al., 2018) has not been widely documented in relation to patients with disease recurrence. Similarly, much of the literature in relation to the development of the UW-QoL questionnaire has focused on patients treated for primary disease, who are now cancer free (Rogers, Lowe, Yueh, & Weymuller Jr, 2010).

This systematic review is not without its limitations. As noted, in preparing the protocol for this review it was hypothesised that literature regarding function and QoL data was in its infancy in the area of ReRuNeR. For this reason, meta-analysis was neither planned nor subsequently undertaken and instead, a narrative review of the evidence is presented. The data presented focuses solely on (potentially curative) surgical outcome data. A high level of bias was noted using the risk of bias assessment for included studies, which must be considered in the interpretation of the findings of this review. The majority of the studies were observational and retrospective in nature. Only two of the included studies were prospective, only one had a comparator group, and neither reported a sample size calculation.

Given the limited evidence on functional and QoL data in the recurrent HNC population, it may have been sensible to include all recurrent HNC studies rather than those specifically pertaining to OPC. For example, two large randomised controlled trials (Burtness et al., 2019; Cohen et al., 2019) reporting standardised QoL outcome measures for patients with recurrent and metastatic HNC including ReRuNeR OPC were excluded, as a subgroup analysis for ReRuNeR OPC disease was not presented in the published data. Although attempts were made to obtain the OPC data from the authors, this was not possible at the time of writing.

Previous consensus guidelines have been published on the conduct of trials in primary disease management to include swallowing and communication standardised outcome tools focusing on a range of clinician-reported, instrumental evaluation and patient-reported outcome measures (Lefebvre, Ang, & Panel, 2009). It is clear from this systematic review that similar standards are neither available nor applied in trials looking at ReRuNeR disease. Further research is required to identify, develop and validate outcome measures for use in the recurrent setting to guide the development of similar recommendations for the conduct of trials focusing on ReRuNeR OPC.

5Conclusion

This review has highlighted that only a small proportion of studies in ReRuNeR OPC report functional and QoL outcomes. These studies focus on surgical treatments, and functional and QoL outcomes for ReRuNeR OPC are not reported specifically in non-surgical trials. Patient experience or qualitative data have not been reported for patients with ReRuNeR OPC.

Where functional and QoL outcomes are reported in surgical trials, there does not appear to be consensus on the outcome tools which should be used, or the timepoints where data should be collected. Consequently, little or no comparison in data is possible across surgical trials. The reviewed literature does show that changes to function and QoL are likely to occur at least transiently. To accurately describe how ReRuNeR OPC and the available treatments impact on QoL and function, there needs to be further research driven by patient experience data highlighting areas of concern. Further research must focus on findings which could lead to a consensus on which tools, or combination of tools should be validated and used at which timepoints, to accurately provide a holistic and multidimensional profile of function and QoL for patients with ReRuNeR OPC.