Objective Patients with head and neck cancer (HNC) receiving radiotherapy (RT) are at high risk of weight loss (WL) due to a variety of nutrition impact symptoms (NIS). This study aimed to describe the NIS through the Head and Neck patient Symptom Checklist and body weight over time and further explore the impact of NIS on WL in patients with HNC undergoing RT.
Methods This was a prospective, longitudinal observational study. NIS and body weight of 117 participants were assessed at baseline, mid-treatment and post-treatment of RT. Generalised estimation equations (GEE) were used to conduct repeated measures analysis of NIS interference score and body weight at each time point and estimate the impact of NIS interference score on WL.
Results All participants experienced a substantial increase in the mean number of NIS during RT, with each patient having eight to nine NIS at mid-treatment and post-treatment. Marked increases were noted in almost each NIS score during RT. Compared with their baseline body weight, 97 (82.9%) and 111 (94.9%) participants experienced WL at mid-treatment and post-treatment, with the mean WL of 2.55±1.70 kg and 5.31±3.18 kg, respectively. NIS of dry mouth (β=−0.681, p=0.002, 95% CI −1.116 to −0.247), difficulty swallowing (β=−0.410, p=0.001, 95% CI −0.651 to −0.169) and taste change (β=−0.447, p=0.000, 95% CI −0.670 to −0.225) impacted WL significantly in GEE multivariate model.
Conclusions Patients with HNC experience a variety of NIS which have significant impact on WL during RT. Assessment of NIS, especially dry mouth, difficulty swallowing and taste change, should be given more considerable attention in the supportive care of patients with HNC.
- head and neck
- supportive care
- symptoms and symptom management
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Head and neck cancer (HNC) is the seventh most common cancer worldwide.1 Radiotherapy (RT) either alone or combined with surgery or chemotherapy is frequent in the treatment of these patients.2 3 Although the proportion of individuals with HNC is not high, the population is distinct with regard to their high risk of involuntary weight loss (WL), which is an essential indicator of nutritional status. WL is frequently observed in patients with HNC and the average WL during RT was about 5%–9% according to previous studies.4–6 WL may result in a higher risk of complications, such as reduced immune function, poor wound healing and infections. WL is also associated with less tolerance and lower response to the treatments including surgery, RT and chemotherapy, leading to treatment delay and even lower quality of life.7–10 Langius et al have demonstrated that WL both before and during RT is an important prognostic indicator for 5-year disease-specific survival in patients with HNC.11
WL in patients with cancer can be attributed to a combined effect of reduced dietary intake and metabolic abnormalities.12 In patients with HNC, reduced dietary intake is quite remarkable due to a variety of symptoms (ie, pain, dysphagia, mucositis, taste change, xerostomia) known as nutrition impact symptoms (NIS).13 14 NIS can be caused by tumour itself as well as the side effects of subsequent treatments, especially RT. Previous studies have demonstrated that NIS had significant impact on dietary intake, which eventually resulted in WL.13 15–20 The most studied NIS is dysphagia, followed by mucositis, which are almost universal in patients with HNC undergoing RT. However, there are few studies which focus on more than one symptom at the same time.16 There is limited information about the prevalence of each NIS over time and the specific contribution of each NIS to WL. It is still unknown which symptoms are most strongly related to WL in patients with HNC during RT.
Currently, there are several instruments available to assess the symptoms of HNC, including Common Terminology Criteria for Adverse Events version 3.0,21 Vanderbilt Head and Neck Symptom Survey version 2.022 and M D Anderson Symptom Inventory-Head and Neck module.23 Although widely used, these instruments only assess the intensity of side effects or symptoms in patients with HNC. Our previous study which explored the relationship between taste alteration and WL has indicated that it was the perceived interference of taste alteration with dietary intake, not the taste alteration, that had a significant impact on WL.24 Head and Neck patient Symptom Checklist (HNSC), which is specifically developed for the assessment of NIS in patients with HNC, assesses both the intensity of each NIS and the degree of each NIS’s interference with dietary intake.4 Therefore, we used HNSC to evaluate NIS among patients with HNC undergoing RT in this study. We aimed to describe the prevalence and severity of each NIS over time and further explore the impact of individual NIS and cumulative NIS burden on WL in a cohort population of patients with HNC undergoing RT.
Participants and data collection
This was a prospective, longitudinal observational study. Adult patients with HNC undergoing RT were recruited consecutively at the department of radiation oncology of a university-affiliated hospital in Beijing between March and November in 2017. The inclusion criteria were as follows: (1) diagnosed as HNC through pathological evaluation; (2) undergoing RT either alone or combined with chemotherapy or surgery; (3) be willing to participate. Patients were excluded if they had other cancers, had tube feeding or total parenteral nutrition, or had cognitive or mental problems.
The researchers assessed each patient at three time points of RT, including baseline (before RT), mid-treatment (in the middle of RT) and post-treatment of RT (on completion of RT). It usually took about 1.5 months to finish a patient’s data collection, depending on the radiation regimen.
Demographic and clinical characteristics
Information on patients’ demographic and clinical characteristics including gender, age, tumour location, tumour stage and treatment was collected at the baseline from the medical records by the researchers.
Head and Neck patient Symptom Checklist
The HNSC is a tool specifically developed for the assessment of NIS which are related to decreased food intake in patients with HNC. It was developed in 2013 and includes 17 symptoms (pain, anxiety, dry mouth, loss of appetite, constipation, feeling full, depression, thick saliva, diarrhoea, sore mouth, lack of energy, nausea, difficulty chewing, altered smell, vomiting, difficulty swallowing and taste changes).25 These items were chosen according to a comprehensive review of the literature and clinical experience. Patients are asked to rate the intensity of each symptom and the degree of each symptom’s interference with dietary intake. Both intensity and interference dimensions of each symptom are assessed using a 5-point Likert scale ranging from ‘1=not at all’ to ‘5=a lot’. A symptom is considered ‘present’ if the intensity score is at least 2. The total score is calculated by adding all 17 items, ranging from 17 to 85 for each dimension. A higher total score indicates more severe intensity or interference of the symptom. The Chinese version of HNSC was established and its psychometric properties were evaluated in our previous study. The Cronbach’s alpha is 0.787 for the intensity dimension and 0.797 for the interference dimension, respectively. The test–retest reliability is 0.845 for the intensity dimension and 0.883 for the interference dimension, respectively. It also demonstrates good criterion validity, convergent validity and discriminant validity.26 The Chinese version of HNSC was assessed at baseline, mid-treatment and post-treatment, respectively.
Body weight was measured at each time point, with patients taking off their shoes and wearing light clothes.
Descriptive statistics was used for analysis of demographic and clinical characteristics. Continuous variables (age, weight, score of NIS) were expressed as means and SD. Categorical variables (gender, tumour location, tumour stage, treatment, prevalence of NIS, degree of WL) were expressed as frequency or percentage. A NIS is considered ‘present’ if the intensity score is at least 2, so the prevalence of each NIS in intensity dimension was calculated as the proportion of participants with the score of this NIS≥2. Similarly, the prevalence of each NIS in interference dimension was also calculated. Generalised estimation equations (GEE) were used to conduct repeated measures analysis of NIS interference score (including single item and the total score) and body weight at each time point to describe the changes over time. GEE was also used to estimate the impact of the total symptom interference score as well as individual symptom interference score on WL. GEE was applied because it accounts for the correlation between repeated observations across time within the same individual.27 GEE modelling approach was used to provide robust parameter estimates and SE with an exchangeable working correlation matrix, for a linear function. Individual symptoms reaching p<0.05 at the univariate level were entered into a multivariate model. All statistical analyses used a significant level of 0.05 (two sided). SPSS statistical software (V.24.0) was used to carry out all statistical analyses.
Participants’ demographic and clinical characteristics
A total of 172 patients with HNC were treated with RT during the research period. Among them, 161 patients met the inclusion and exclusion criteria and were included in the baseline assessment. One hundred and forty-four (89.4%) patients were followed up to the mid-treatment, and finally 117 (72.7%) patients were followed up to the post-treatment (figure 1). The average age of all 117 patients was 54.62±13.12 (range: 18–81) years. All patients were treated with intensity-modulated radiation therapy (46–70 Gy in daily factions of 1.2–2.0 Gy). The duration of RT was about 4–7 weeks. Other demographic and clinical characteristics of the participants were shown in table 1.
Nutrition impact symptoms
Prevalence of each NIS in intensity dimension and interference dimension over time
Table 2 showed the prevalence of each NIS in intensity dimension and interference dimension among participants over time, respectively. For the intensity dimension, participants experienced a substantial increase in the mean number of NIS (baseline: 2.65, mid-treatment: 8.70, post-treatment: 9.56). Forty-one per cent of the participants had three or more NIS at baseline. At mid-treatment and post-treatment, almost all participants had three or more NIS, and the symptoms with the incidence rate of more than 50% were pain, dry mouth, thick saliva, difficulty swallowing, taste change, sore mouth, lack of energy, loss of appetite and difficulty chewing.
However, in terms of the symptom’s interference with dietary intake, there were some differences. There was also a great increase in the mean number of NIS interfering with dietary intake over time (baseline: 1.04, mid-treatment: 5.68, post-treatment: 6.56), but obviously lower than the number of NIS in term of its incidence rate. 17.9% of the participants had three or more NIS interfering with dietary intake at baseline. At post-treatment, 89.7% of the patients had three or more NIS interfering with dietary intake, and the interfering symptoms with the incidence rate of more than 50% were pain, difficulty swallowing, taste change, sore mouth, loss of appetite, dry mouth and difficulty chewing.
Mean interference score of NIS over time
The mean interference scores of each NIS and the total interference score of HNSC over time for all participants were shown in table 3. Analysis of variance GEE showed that marked increases were noted in the total interference score and almost each NIS from baseline to post-treatment except ‘altered smell’, ‘depression’ and ‘diarrhea’. At mid-treatment and post-treatment, pain, difficulty swallowing, sore mouth and loss of appetite had the highest interference scores.
Mean body weight over time
The mean body weights for all participants were 65.36±11.76 kg, 63.38±11.39 kg and 60.37±11.03 kg at baseline, mid-treatment and post-treatment, respectively. Analysis of variance GEE showed that marked decreases were noted in mean body weight from baseline to mid-treatment and from mid-treatment to post-treatment (Wald χ2=257.778, p<0.001).
Prevalence of WL
Compared with their baseline body weight, 97 (82.9%) and 111 (94.9%) participants experienced varying degrees of WL at mid-treatment and post-treatment, with the mean WL of 2.55±1.70 kg and 5.31±3.18 kg, respectively. At mid-treatment and post-treatment, a total of 29 (24.8%) and 84 (71.8%) participants experienced ≥5% WL.
GEE modelling to estimate the impact of NIS upon WL
GEE was used to estimate the impact of the score of HNSC on WL. The total interference score of HNSC had a significant impact on WL (β=−0.231, p<0.001, 95% CI −0.261 to −0.200).
Then GEE was used to estimate the impact of each NIS interference score on WL. The univariate results indicated that depression, altered smell and lack of energy were the only NIS having no effect on WL (table 4). Then 14 NIS were entered into the multivariate model, only the NIS of dry mouth (β=−0.681, p=0.002, 95% CI −1.116 to −0.247), difficulty swallowing (β=−0.410, p=0.001, 95% CI −0.651 to −0.169) and taste change (β=−0.447, p<0.001, 95% CI −0.670 to −0.225) impacted WL (table 5).
This study showed that patients with HNC experienced multiple NIS during RT and the number of NIS continued to increase. At baseline, 41% of patients had three or more NIS. The most common NIS were dry mouth (36.8%), pain (35.0%) and difficulty chewing (32.5%), among which the most influential NIS on dietary intake was difficulty chewing. These results were not exactly the same as those of previous studies. Kubrak et al found that the most common symptoms of the patients with HNC before treatment were pain (33%) and dysphagia (29%).14 Farhangfar et al reported that pain, anxiety and lack of energy were the most common symptoms of patients with HNC before chemoradiotherapy.13 Arribas et al found that dysphagia and dry mouth were the most common symptoms of patients at the time of diagnosis.28 Although the symptoms have varied from study to study, most of them were symptoms that affected eating and usually existed at the same time. We found that the number of NIS increased significantly during RT, and the average number of NIS experienced by each patient was 8–9 at mid-treatment and post-treatment. The NIS with incidence rate ≥50% were pain, dry mouth, thick saliva, difficulty swallowing, taste change, sore mouth, lack of energy, loss of appetite and difficulty chewing, most of which had a great interference with eating. Although the incidence of thick saliva and lack of energy at post-treatment reached 96.6% and 78.6%, only 42.7% and 13.7% of patients considered these two symptoms interfering with dietary intake. This result suggested that thick saliva and lack of energy, though high frequent or severe, do not affect eating to the same extent. Our findings were consistent with previous studies. Farhangfar et al discovered that, although pain, anxiety and lack of energy were the most common symptoms in patients with HNC before chemoradiotherapy, loss of appetite was the most influential factor resulting in reduced intake.13
This study showed that the mean interference score of almost each NIS increased significantly from baseline to mid-treatment and from mid-treatment to post-treatment. At mid-treatment and post-treatment, pain, difficulty swallowing, sore mouth and loss of appetite had the highest interference scores. Similar results were found in previous studies.4 29 Ottosson et al found that the most common symptoms were difficulty swallowing, pain, taste change, loss of taste and xerostomia during RT and at the end of RT.29 Kubrak et al discovered that difficulty swallowing, taste change, loss of appetite and pain had the highest interference scores at post-treatment.4 We also found that the total interference score at mid-treatment was almost twice as high as that at baseline. However, the total interference score at post-treatment was just a little bit higher than that at mid-treatment, suggesting that the increasing extent was far lower at the second half of RT than that at the first half of RT. This indicated that the total interference score was close to the peak at mid-treatment. This suggested an early applying of symptom management. In summary, NIS during RT are in the form of a group, and usually become severe early. Therefore, the management of symptoms during RT should be carried out early and focusing on multiple symptoms. According to the degree of effect of symptoms on eating, it is suggested to give priority to the symptoms that have serious impact on dietary intake.
Our study supported that patients with HNC suffered from severe WL during RT. Similar to previous studies,5 6 20 30 we found that the average weight decreased by 5.31±3.18 kg by the end of RT. The univariate analysis found that the majority of NIS significantly impacted WL. In the multivariate model, the NIS of dry mouth, difficulty swallowing and taste change were significant predictors of WL. For example, for each unit increase in difficulty swallowing, weight decreased by 0.410 kg. Difficulty swallowing is considered to be the characteristic symptom interfering with dietary intake in patients with HNC, and its significant impact on WL was no surprise. Our findings were consistent with previous studies which also demonstrated that difficulty swallowing was significantly associated with nutritional deficiencies.31–33 Taste change is another common symptom in patients with HNC during RT, and it includes hypogeusia, loss of taste and taste disorder.34 McLaughlin found that 92.4% of patients had taste dysfunction, and statistically significant WL was associated with taste disorders.35 Jin et al also demonstrated that the perceived interference of taste alteration with dietary intake had a significant impact on WL.24 Xerostomia or dry mouth is thought to be the most prominent complication after RT in patients with HNC.36 It will contribute to difficulty in chewing, swallowing, speaking and sleeping and eventually lead to WL.4 13 20
However, pain and sore mouth, two of the most frequent and severe symptoms during RT, were not significant predictors of WL in the multivariate model. This might be due to the strong correlation between pain, sore mouth and difficulty swallowing. Mucositis, which is one of the most troublesome side effects in patients with HNC undergoing RT, will cause a wide spectrum of symptoms, such as pain, bleeding, sore mouth, infection and even difficulty swallowing.37 So, only difficulty swallowing existed in the final model with pain and sore mouth excluded.
This study had some limitations. First, all participants were recruited from pre-RT and followed up to the completion of RT. In fact, most of them continued suffering from multiple symptoms and WL after the completion of RT. A longer period of follow-up is necessary in future study. Second, this study only focused on the impact of NIS on WL. Some other factors (ie, tumour site, treatment modality and pretreatment WL) which have been demonstrated to impact WL were not included in the GEE model. Inclusion of these factors should be considered in future study. Third, although efforts were made to decrease the missing rate, only 72.7% of participants were followed up to the completion of RT. The majority of missing participants withdrew from the study due to the unbearable side effects of RT. This might lead to a bias in our results. However, the actual situation might be more severe than that in our study.
Patients with HNC suffered from multiple NIS during RT, with each patient having about eight to nine NIS at mid-treatment and post-treatment. The mean interference score of almost each NIS increased significantly during RT. Pain, difficulty swallowing, sore mouth and loss of appetite had the highest interference scores at mid-treatment and post-treatment. The NIS of dry mouth, difficulty swallowing and taste change impacted WL significantly. This study indicated that the supportive care of symptoms during RT should be applied early and focusing on multiple symptoms. It is also suggested to give priority to the symptoms which have a significant impact on WL.
We are grateful to all the doctors and nurses in the Department of Radiation Oncology, Peking University Cancer Hospital, for their generous help in patient’s arrangement and data collection process in this study.
Contributors SJ, QL and YS conceived and designed the study. SJ, SX and BZ participated in the data collection and statistical analysis. SJ, QL, YS, DP and PY contributed to interpretation of findings. SJ led the writing of the manuscript, with input from QL, YS and DP. All authors reviewed the manuscript and approved the final version.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval This study was approved by the Biomedical Ethics Committee of Peking University (IRB00001052-17002). Written informed consent was obtained from all participants.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement Data are available upon reasonable request. Data generated by our research that supports our article are available upon reasonable request.