Affective Symptoms and Oropharyngeal Dysphagia in Head-and-Neck Cancer Patients: A Systematic Review

Swallowing is a complex neurocognitive process. It relies on accurate coordination of a variety of muscle and nerve groups aiming at efficient bolus preparation and safe and efficient bolus transfer from the oral cavity and pharynx to the esophagus [1]. Damage of upper aerodigestive tract tissue due to a head-and-neck malignancy or its oncological treatment may cause oropharyngeal dysphagia (OD) [2, 3]. The incidence of head-and-neck cancer (HNC) is rising, partly due to increasing numbers of human papilloma virus (HPV)-related oropharyngeal cancer resulting in a growing population of patients with a need for long-term healthcare [4]. OD is a high impact morbidity in head-and-neck cancer (HNC) patients with a reported prevalence of 45% [2, 5]. OD can be accompanied by severe complications such as aspiration pneumonia, sepsis, or malnutrition [3, 6]. Therefore, early screening, diagnosis, and treatment are essential to minimize the consequences of OD.

Besides affecting overall health, HNC may also affect mental health, social functioning, and employment [2]. This range of issues may have major effects on social and psychological well-being. The diagnosis and treatment of HNC itself may result in a significant burden on the patients’ psychological state because patients often find themselves in an ‘existential crisis situation’ [7]. Moreover, OD is often accompanied by anxiety, depression, reduced self-esteem, and social isolation, further amplifying the HNC-related suffering [6]. The recognition and treatment of the psychosocial burden in patients with HNC is important as distress may interfere with the ability to cope with the disease, its oncological treatment, and rehabilitation. The increasing incidence of HNC, combined with a high prevalence of psychological comorbidity in HNC patients [8], emphasizes the importance of an interdisciplinary approach including mental health care.

A wide variety of instruments are developed to screen for affective (anxiety and depression) symptoms and swallowing dysfunction. Screening tools are used for early identification of individuals at potentially high risk for a specific disorder. A screening tool such as the hospital anxiety and depression scale (HADS) is useful to identify clinically relevant anxiety and depressive symptoms [9]. For further identification of the nature and severity of a psychological disorder, a neuropsychological diagnostic workup is required. Regarding swallowing function, a screening tool such as the water swallow test (WST) is a quick and non-invasive method to identify patients at risk for unsafe swallowing [10]. After positive screening for OD, a clinical examination by a speech and language pathologist and/or instrumental swallowing assessment are recommended. In the literature, fiberoptic endoscopic evaluation of swallowing (FEES) and videofluoroscopic swallow study (VFSS) are considered the golden standard examinations to assess the swallowing function [11]. Besides clinician-reported outcome (CRO), self-evaluation is covered by patient-reported outcome (PRO) questionnaires. These questionnaires can roughly be divided into two different concepts: health-related quality of life (HRQoL) versus functional health status (FHS) questionnaires. FHS is often defined as one’s ability to perform daily activities required to meet basic needs, fulfill usual roles, and maintain their health and well-being [12]. HRQoL is a multi-dimensional concept that includes domains related to physical, mental, emotional, and social functioning [13]. HRQoL focuses on the impact of health status on quality of life [11, 14]. The concepts of HRQoL and FHS are often mixed, making it difficult to distinguish between tools that measure disease-related-QoL and functioning.

The aim of the present study is to systematically review and appraise the literature to obtain insight into the prevalence, strength, and causal direction of the relationship between OD and clinically relevant affective symptoms in HNC patients. A better understanding of this relationship will contribute to a better interdisciplinary approach to both problems (OD and affective symptoms) that can adversely affect each other during oncological treatment and rehabilitation.

The systematic searches across all databases yielded a total of 139 abstracts after duplicate references were removed. Of these, 86 abstracts were removed based on the exclusion criteria. The full-text of the remaining 53 abstracts was reviewed. Two articles were identified after hand searching the reference lists of the included studies. Finally, 15 articles were included in this systematic review. The PRISMA flow diagram illustrates the search process (Fig. 1).

The two reviewers had 78% agreement (κ = 0.55) on the selection based on title and abstract screening. Articles that were selected by just one of the reviewers were subsequently screened using the full-text. At the full-text review stage, the two reviewers had 84% agreement (κ = 0.63) on their ratings. The two reviewers had 100% agreement on the ABC-rating scale and 93.3% on the QualSyst ratings. Disagreements were discussed and resolved in consensus.

The initial purpose of the present study was to systematically review and appraise the literature on the prevalence of affective symptoms and to identify the strength and direction of the relationship between OD and clinically relevant affective symptoms in HNC patients. A better understanding of this relationship will contribute to a better interdisciplinary approach to both problems (OD and affective symptoms) that can adversely affect each other during oncological treatment and rehabilitation. In general, OD and affective symptoms were related to each other as described below. However, the results of this systematic review have not been able to adequately answer the question on the strength and direction of this relationship.

In total, 15 articles were included. The methodological quality of the included studies ranged from adequate (0.86) to strong (0.96). Due to the heterogeneity of study designs, terminology, outcomes, and outcome measures used, a meta-analysis could not be conducted.

A positive relationship between OD and affective symptoms or emotional status was described in the majority of the studies [20, 22, 24‐28, 30‐32, 51]. Nguyen et al. reported that HNC patients experience anxiety and depression related to their OD severity which can be explained by the functional impairment and disfigurement resulting from HNC and its treatment [30]. Eating and drinking is an important part of social interaction, but dysphagic HNC patients often experience eating difficulties in public and home environment [23]. In some cases, this may lead to exclusion of invitations or to patients declining to eat out. Therefore, patients may become socially isolated leading to symptoms of anxiety and depression. Maclean et al. and Zhang et al. described that an improvement in swallowing function may reduce the severity of affective symptoms in HNC patients [28, 30]. On the other hand, affective symptoms can cause physical complaints such as dry mouth which may enhance swallowing impairment [45]. Furthermore, affective symptoms can negatively affect motivation and consequently compliance during cancer rehabilitation, resulting in a poor functional outcome [50]. Bozec et al. described that depressive symptom scores are independent predictors of poorer swallowing function in HNC patients, highlighting the fact that swallowing function is highly dependent on psychological, emotional, and social conditions in addition to tumor or treatment characteristics [26]. In addition, the patient’s psychological baseline should be taken into account as affective symptoms may already be present prior to the HNC diagnosis [26].

Although a positive relationship between OD and affective symptoms has been described in the majority of the included studies, other studies found a non-significant or negative relationship between swallowing function versus anxiety, depression, or emotional status [21, 23, 29, 33]. The reasons for these divergent findings can be multiple. For example, the timing of assessment of OD and affective symptoms varies widely between these studies. The moment of measurement plays an important role in the outcome of swallowing-, physical-, and emotional functioning. The longitudinal study of Hammerlid et al. reported that patients may develop coping skills or undergo changes in the experience of the disease and in their expectations of health over time resulting in improved symptom scores [51]. These findings justify the recommendation to systematically screen for affective symptoms and swallowing disorders at baseline (before oncological treatment) and during the oncological follow-up.

A variety of tools measuring swallowing function and affective symptoms were used in the included studies. These different tools vary in use and purpose (screening versus diagnostic) and the interpretation and clinical relevance of the outcome measures should be taken into account. Multiple PRO HRQoL questionnaires were used to measure patients’ perception of the swallowing function based on the multi-dimensional concept of HRQoL including domains related to physical, mental, emotional, and social functioning. However, only one study included PRO FHS questionnaires to determine the impact of OD on the ability to perform daily activities [29]. In a cross-sectional study, Campbell et al. aimed to determine associations between instrumental assessment (VFSS), PRO HRQoL measurement (UW-QOL), and FHS measurements (FACT-G, FACT-H&N and PSS-HN) in HNC survivors, 5 years post-treatment [29]. Patients presenting aspiration of the bolus into the airway scored significantly lower on the UW-QOL swallowing domain, FACT-H&N additional concerns, and PSS-HN ‘normalcy of diet’ domain, compared to non-aspirators. Aspiration was not associated with PSS-HN ‘willingness to eat in public’ domain nor with any of the FACT-G well-being scales. So, despite unsafe swallowing (VFSS) and poor swallowing-related HRQoL (UW-QOL), aspiration of the bolus into the airway does not seem to impact everyday activities and fulfilling usual roles.

It is not uncommon to find a discrepancy between the results of PRO and CRO measures [52]. Airoldi et al. reported on a discrepancy between a self-reported prevalence of depressive symptoms of 30% measured using the HADS versus a clinician-reported prevalence of 44.4% measured by the MADRS in a population of patients following treatment for oral cancer [27]. The authors concluded that this discrepancy might be related to an inadequate self-awareness of HNC patients concerning illness-related psychological distress. Furthermore, according to the authors, a higher prevalence of having a vulnerable socioeconomic status, addictive behavior, and anosognosia can play a role in limited self-awareness in this patient population [27]. In addition, Florie et al. reported on the very few statistically significant mean differences of MDADI subscale scores between the ordinal scale levels of several FEES variables in a heterogenous population of HNC patients following cancer treatment. This study also described a weak relationship between the severity of OD and PRO OD-specific HRQoL. The authors concluded that adaptive changes in swallowing function, radiation neuropathy with a decreased oropharyngeal sensibility, and the patients’ noncomplaining nature and lack of initiative may affect the perception (underestimation) of their swallowing difficulties. That perception or underestimation, in turn, may determine their score on the MDADI questionnaire [21]. The MDADI was developed to measure the influence of OD on the patients’ HRQoL. Nevertheless, it still remains unclear if the MDADI can be used as an indicator for the severity of OD [21]. Although these PRO and CRO measures do not correlate well, it remains important to realize the existence of these different dimensions of OD as well as their application and relevance in both scientific research and daily clinical practice.

Non-validated measurement tools were used in eleven of the included studies. The use of high-quality measurement tools based on robust psychometric properties such as validity and reliability is strongly recommended and essential to accurately estimate the prevalence of affective symptoms and OD [53].

Although no meta-analytic conclusions can be drawn from the included articles, OD and affective symptoms often appear to be coincidental in HNC patients. The included studies reported a prevalence of OD ranging from 16 to 100% and a prevalence of clinically relevant affective symptoms ranging from 12 to 54%. This wide variation in the prevalence of OD and clinically relevant affective symptoms in the different studies may be due to several reasons: different study designs with various HNC patient samples (variation in age, tumor location, tumor stage, oncological treatment, etc.), the use of different and/or non-validated measuring instruments for OD and affective symptoms in HNC patients, different timing of measurement, etc.

Seven studies reported the effects of tumor location, tumor stage, and/or oncological treatment modality on swallowing, and affective symptoms [23, 26, 28‐32, 51]. Depending on the tumor location, swallowing function may be affected in different ways. For instance, in the literature, a higher incidence of aspiration before the start of the oncological treatment is reported in patients with laryngeal and hypopharyngeal cancer [54]. However, the included studies of the present review did not report on the relationship between tumor location and OD severity nor on the relationship between tumor location and clinically relevant affective symptoms. A significant relationship between tumor stage versus OD and depressive symptoms was reported [26, 29, 30]. An advanced tumor stage can cause more severe OD and functional impairment due to a greater extent of damage to essential structures of the upper aerodigestive tract. However, this relationship was established after oncological treatment, so most likely the type and number of oncological treatment modalities play a role in this relationship. Advanced primary site disease often necessitates aggressive multimodality treatment, which puts patients at greater risk of long-term disability as a result of surgical and adjuvant (chemo)radiotherapy induced functional loss [30]. (Chemo)radiotherapy may induce mucositis, stomatitis, hyposalivation, trismus, soft tissue necrosis, fibrosis, and osteoradionecrosis of the mandible [27]. The effect of surgery on swallowing function and affective symptoms should also be considered as OD, and affective symptoms seem to be related to the location and extent of the resection. A significant association between the extent of tongue(base) resection versus OD and depressive symptoms was reported [26, 31]. However, six studies did not find any effect of tumor location, tumor stage, and type of oncological treatment modality on the prevalence of affective symptoms and OD showing that this relationship is not yet well understood [23, 25, 28, 32, 33, 51].

All included studies applied patient-reported questionnaires on OD and affective symptoms. However, only few studies screened or assessed the level of cognition of the included patients [21‐23]. When using PRO tools, it is necessary to screen or estimate the level of cognition prior to completing a self-report questionnaire to guarantee that patients are able to understand and answer the questions accordingly. Cognitive impairment may complicate recall-based assessment with questionnaires resulting in recall bias. Besides cognition, alcohol consumption should also be reported when evaluating affective symptoms. Prolonged alcohol use is often seen in HNC patients and known to cause structural changes in the brain as well as cognitive deficits [55]. Four articles reported on a significant association between alcohol abuse and long-term psychological distress, resulting in a vicious circle in which these phenomena reinforce each other [27, 28, 30, 33]. This highlights the importance to support HNC patients with an active addiction. Finally, although the use of psychotropic drugs is likely to influence the severity of affective symptoms, only two studies reported on the use of psychotropic drugs [22, 23].

This systematic review has some limitations. The systematic search generated a low number of articles on affective symptoms and OD in HNC patients. Reasons for this low number may be related to the inconsistent terminology used in this research topic. The lack of randomized controlled trials and pre- and post-treatment data studies may limit the strength of the findings. All studies had methodological limitations (e.g., lack of details provided regarding selection criteria; small sample sizes; limited information on methods, incomplete information about measurement tools and procedures; no information about test result interpretation). The heterogeneity of the included studies is likely to have contributed to the overall variation in reported frequencies of OD and affective symptoms and precluded eligibility to pool data across studies.

This study shows that screening for affective symptoms in dysphagic HNC patients should be considered as affective symptoms and OD seems to be prevalent and coincident in this population. The strength and direction of the relationship between affective symptoms and OD still remain unclear. Considering the impact of affective symptoms and OD on patients’ daily life, early detection and an integrated interdisciplinary approach are recommended. Future studies should use validated measurement tools, bigger sample sizes, and study designs that lead to high-quality evidence.