This study was supported by a grant from the Netherlands Society for Scientific Research (NWO).
Objectives: To describe prospectively the long-term changes of quality of life and mood in patients with squamous cell carcinoma of the head and neck treated with surgery and/or radiotherapy.
Patients and Methods: One hundred seven patients completed the European Organization for Research and Treatment of Cancer (EORTC) Core Questionnaire, the EORTC Head and Neck Cancer Module, and the Center for Epidemiological Studies Depression Scale before treatment, and 6, 12, 24, and 36 months later.
Results: There was limited deterioration of physical and role functioning and of many head and neck symptoms at 6 months, with improvement thereafter. After 36 months only physical functioning, taste/smell, dry mouth, and sticky saliva were significantly worse, compared with baseline. Female sex, higher cancer stage, and combination treatment were associated with more symptoms and worse functioning. Despite physical deterioration, there was a gradual improvement of depressive symptomatology and global quality of life.
Conclusion: Treatment for head and neck cancer results in short-term morbidity, most of which resolves within 1 year. Despite an initially high level of depressive symptomatology, there is gradual improvement of psychological functioning and global quality of life over the course of 3 years. In this prospective study, the impact of the disease and its treatment in long-term survivors seems to be less severe than it is often assumed to be.
In the past decade there has been a considerable increase of interest in quality of life (QOL) issues in oncology.1,2 In clinical research, QOL is recognized as an important endpoint in addition to the traditional endpoints such as response rate, disease-free survival, and overall survival.3 In clinical practice, changes in treatment policies are aimed not only at increasing the chances of cure, but also at improving or maintaining QOL during and after treatment.4
Quality of life is particularly relevant for patients with head and neck cancer because social interaction and emotional expression depend to a great extent on the structural and functional integrity of the head and neck region.5 In patients with advanced disease, intensive treatment may lead to severe physical and psychosocial sequelae. Depression and other psychosocial morbidity occur frequently in these patients.6-9 Measurement of QOL of head and neck cancer patients may help to make decisions about treatment, identify patients with severe physical, and/or psychosocial morbidity, and plan rehabilitation.10
In the early 1990s, there was a lack of prospective QOL studies in head and neck cancer patients using standardized and validated instruments.11,12 In recent years several head and neck-specific QOL instruments have been developed and validated,13-16 and prospective studies using these instruments have been published.17-26 However, no prospective study has been published with a follow-up of more than 1 year.
We performed a prospective study of QOL in patients with squamous cell cancer of the head and neck, treated with surgery and/or radiotherapy. The aim of the study was to describe prospectively QOL (and its domains) before and after treatment. We already described short-term changes in QOL in two cohorts of this study.24,25 All patients in the study now have a follow-up of 3 years. In the present paper we focus on the long-term changes in QOL.
PATIENTS AND METHODS
Patients were eligible for the study if they had squamous cell carcinoma of the oral cavity, oropharynx, hypopharynx, or larynx treated with surgery and/or radiotherapy with curative intent. Other inclusion criteria were age less than 80 years; no previous or synchronous malignancies; no cognitive impairment; and ability to understand and read Dutch. Patients with a recurrence or a second tumor of the head and neck or lung were excluded from the analysis.
During the inclusion period (May 1994-June 1996) 266 patients met the inclusion criteria. Fifty-eight patients refused to participate in the study (refusal rate, 24%). There were no significant differences with regard to sex, age, site, stage, or treatment between patients who participated in the study and patients who refused to participate. Data of 101 patients were excluded from the analysis during the first 3 years after treatment because of recurrence (n=53), development of a second tumor of the head and neck or lung (n=12), death due to other causes (n=7), noncompliance (n=25), or incomplete data (n=4), resulting in 107 fully evaluable patients, alive without recurrence or a second tumor, with a follow-up of 3 years. Patient characteristics are summarized in Table I. Forty-five patients had laryngeal cancer treated with radiotherapy, and 54 patients had cancer of the oral cavity/oropharynx treated with surgery with or without radiotherapy. Analyses were performed for the whole group and for these two subgroups separately.
Table TABLE I.. Patient Characteristics.
The patients completed a questionnaire before treatment and six, 12, 24, and 36 months later. The questionnaire consisted of the European Organization for Research and Treatment of Cancer (EORTC) Core Questionnaire Version 30(+3) (QLQ-C30[+3]), the EORTC Head and Neck Cancer Module (QLQ-H&N35), and the Center for Epidemiological Studies Depression (CES-D) Scale.
The EORTC QLQ-C30 is a widely used questionnaire incorporating extensive QOL issues relevant to a broad range of cancer patients.27 It has been validated for many types of cancer including head and neck cancer.28 It contains five functional scales (physical, role, cognitive, emotional, and social), three symptom scales (fatigue, pain, and nausea/vomiting), a global QOL scale, and six single items (dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties). Version 30(+3) contains two additional items on role functioning and one additional item on overall health. The EORTC QLQ-C30(+3) is meant to be used in conjunction with a tumor-specific module.
The EORTC QLQ-H&N35 is a module used for assessing QOL in head and neck cancer patients.29 It contains seven symptom scales (pain, swallowing, taste/smell, speech, social eating, social contacts, and sexuality) and six symptom items (teeth problems, trismus, dry mouth, sticky saliva, cough, and feeling ill). It has been validated in a sample of 500 patients with head and neck cancer from Norway, Sweden, and The Netherlands.16
All scales and items of the EORTC QLQ-C30(+3) and QLQ-H&N35 range in score from 0 to 100. A high score for a functional or global QOL scale represents a high level of functioning or global QOL, whereas a high score for a symptom scale or item represents a high level of symptoms or problems.30
The CES-D Scale is an instrument used for measuring depression in the general (nonpsychiatric) population.31 The Dutch translation has also been validated.32 The CES-D has been used in head and neck cancer patients.33-35 It results in a total score ranging from 0 to 60. A high score reflects a high level of depression. A cut-off point of 16 can be used, patients with a score of 16 or more being classified as having significant depressive symptomatology.36
The data were analyzed using the statistical package SPSS for Windows (SPSS Inc., Chicago, IL).37 Multivariate ANOVA for repeated measures was used for the comparison of the means of all scales and single items of the QLQ-C30(+3) and of the QLQ-H&N35, and the total score of the CES-D over the study period of 3 years. The influence of sex, age, subgroup, stage, and treatment was studied by entering these variables as between-subject factors in the MANOVA procedure; for this purpose, age was divided into two groups (<60 y and ≥60 years, 60 years being the median). The Cochran Q test was used to detect changes in the proportion of patients with a CES-D score of ≥16 over time. Student t test was used to compare means at two time points and Mann-Whitney's test was used to compare the means of scales and items of patients who completed the study and patients who refused cooperation during the study.
The mean values of all scales and single items of the questionnaire at all time points are shown in Table II. There were significant changes of several scales and items over time. Some of these are depicted in Figures 1 and 2. Changes were usually for the worse, with a pattern of maximal deterioration at 6 months and improvement thereafter. At 12 months fatigue, physical functioning, taste/smell, dry mouth, and sticky saliva were still significantly worse, compared with baseline. However, another pattern was seen for some scales (pain, insomnia, emotional functioning, total score of the CES-D, and global QOL), the mean values of which improved gradually without any deterioration.
Table TABLE II.. Mean Scores* of Scales and Single Items of Questionnaire.
Physical functioning, taste/smell, dry mouth, and sticky saliva were still significantly worse at 36 months, compared with baseline, whereas pain, insomnia, speech, emotional functioning, the total score of the CES-D and global QOL were significantly better. The percentage of patients with a total score of the CES-D of ≥16 showed a nonsignificant (P = .08) trend for improvement (from 27% to 18%). When comparing the values at 12 months and at 36 months, there was only a significant difference (i.e., improvement) for pain as measured with the QLQ-H&N35 and for the total score of the CES-D.
To study whether patients who refused further cooperation during the study introduced a bias, the data from their first and second (if available) questionnaires (n=25) were compared with the data of the patients who completed the study (n=107). At baseline, dropouts reported significantly worse scores for fatigue, pain, dyspnea, swallowing, and social eating; at 6 months they reported significantly worse scores for global QOL, dyspnea, and speech (data not shown).
The changes over time of the scales and single items by subgroup (patients with oral/oropharyngeal cancer treated with surgery with or without radiotherapy, and patients with laryngeal cancer treated with radiotherapy) are shown in Tables III and IV. Patients with oral/oropharyngeal cancer showed a pronounced deterioration of physical functioning, role functioning, fatigue, speech, and trismus. Cough increased in patients with laryngeal cancer, whereas speech, social functioning, and general QOL improved. Pain (as measured with the QLQ-H&N35), emotional functioning, and the total score of the CES-D improved in both subgroups.
Table TABLE III.. Mean Scores* of Selected Scales and Single Items of Questionnaire for Patients With Cancer of the Oral Cavity or Oropharynx Treated With Surgery With or Without Radiotherapy.
Table TABLE IV.. Mean Scores* of Selected Scales and Single Items of Questionnaire for Patients With Cancer of the Larynx Treated With Radiotherapy.
The influence of sex, age, subgroup, stage, and treatment is shown in Table V. Sex differences were seen for many scales and single items of all questionnaires. If there was a significant difference with regard to sex, women consistently reported more complaints and worse functioning. There was hardly any influence of age; the only significant difference was found for physical functioning, patients of 60 years or older having worse scores. Differences between subgroups were only seen for some scales and items of the QLQ-H&N35: pain, social eating, teeth, and trismus (patients with oral/oropharyngeal cancer having worse scores), and speech (patients with laryngeal cancer reporting more speech problems). Stage and treatment had a significant influence, in particular on the scales and items of the QLQ-H&N35; without exception patients with higher stages and patients receiving combination treatment had worse scores.
Table TABLE V.. Influence of Sex, Age, Subgroup Stage and Treatment on Mean Scores of Scales and Items of Questionnaire.
These are the first long-term results of a prospective study of QOL of patients with head and neck cancer to be reported. In accordance with other prospective studies,18-23 we found a temporary deterioration of several physical symptoms and physical and role functioning, most likely caused by treatment. The values of most of these scales and items had returned to baseline at 12 months and there was little change thereafter. At 36 months, physical functioning, taste/smell, dry mouth, and sticky saliva were still significantly worse, compared with baseline. The magnitude of the differences6-16 indicates clinically relevant changes, which are minor to moderate.38
There were a few exceptions to this pattern. Pain, insomnia, and speech all showed a gradual improvement. The improvement in pain was most pronounced in patients with cancer of the oral cavity/oropharynx (continuing even after 12 months), whereas the improvement in speech occurred only in patients with laryngeal cancer treated with radiotherapy. This is likely because pain and speech problems are symptoms that may be tumor-related and may respond favorably to treatment.
Despite the short-term, and sometimes long-term, deterioration of several physical symptoms and functioning, there was a gradual improvement of emotional functioning and the total score of the CES-D (continuing even after 12 months). Both the mean total score and the percentage of patients with a score of 16 or higher were initially higher than that of the average Dutch population,39 but after 36 months these differences had disappeared.
The high level of depressive symptomatology at baseline is in accordance with the literature.20-22,40-42 There seems to be a discrepancy between the deterioration of physical symptoms on the one hand, and improvement of mood on the other hand, not only in the short term, but also in the long term. None of the patients in our study received specific treatment for depression. Several factors may explain the apparent discrepancy between physical deterioration and improvement of mood over time. At the time of the first measurement, the patients had recently heard the diagnosis and were facing a yet unknown and possibly threatening treatment. At the time of the subsequent measurements, they had finished treatment, were recovering from its aftereffects, and had had time to adapt to the situation. The mechanism of response shift (referring to a change of a person's internal standard for determining his or her level of functioning on a given dimension)43 may also be involved. Moreover, as time passed, fear of recurrence and death is likely to have decreased significantly. If we had included patients with a recurrence in the analysis, psychological functioning might have shown a different course.
The results of this study do not seem to be fully compatible with the literature, which indicates a high level of psychological distress in long-term survivors of head and neck cancer.7,9,44 There are some possible explanations for this discrepancy. First, in our study there was a relatively high proportion of patients with limited disease requiring only radiotherapy or surgery, which may have limited long-term psychosocial morbidity. Second, one retrospective study suggested an increase of anger and anxiety as long as 5 years after diagnosis, possibly as a result of patient burnout and cognitive beliefs.9 Thus the follow-up of our study may have been too short. However, Rapoport et al.,9 observed a deterioration after only 18 months, whereas we found an improvement during this period. Third, all these studies were cross-sectional. One prospective study of patients with head and neck cancer with a follow-up of 1 year has been published recently.42 In this study, approximately one third of the patients scored as a probable or possible case of a major mood disorder (depression and/or anxiety) at each measurement point (before treatment and at 1, 2, 3, 6, and 12 months).
Clinicians caring for head and neck cancer patients should be alert for symptoms of a major mood disorder, in particular in patients receiving extensive surgery in combination with radiotherapy. A short questionnaire (such as the Hospital Anxiety and Depression Scale used in the study by Hammerlid et al.42) might be used as a screening instrument for psychiatric morbidity. Patients with severe and/or persistent symptoms should be referred for psychological support and, if necessary, pharmacological treatment.
It is often assumed that head and neck cancer and its treatment has severe implications for social functioning. Interestingly, we found no significant changes in the mean values of the social functioning and the social contacts scales. Again, the relatively high number of patients with limited disease may have influenced the results. For patients with laryngeal cancer treated with radiotherapy, we actually found an improvement in social functioning. This may be related to the improvement of speech in this subgroup that may have facilitated communication. For the whole group, the mechanism of response shift may also have played a role: as a part of the adaptation process patients may have reset their standards for social functioning.
Finally, it is remarkable that the patients indicated a gradual but significant increase in global QOL. The same mechanisms that are involved in the improvement in psychological functioning may apply.
Thus in our study the long-term impact of head and neck cancer and its treatment seems to be less severe than it is often assumed to be. There is a temporary and limited physical deterioration, but in the long run relatively few symptoms remain. Psychosocial functioning and global QOL show a gradual improvement that continues for at least 3 years.
We performed a preliminary analysis to examine which sociodemographic and medical factors had an influence on QOL and mood. Women were found to report more symptoms and worse functioning over time. The data from studies on the influence of sex on QOL of cancer patients are conflicting. Some show no gender differences,23,45 whereas others report worse QOL for women.44,46,47 Age (at least up to 73 years) had no influence except on physical functioning, as was found in other studies.23,47,48 Subgroup, stage, and treatment had a major influence, in particular on head and neck symptoms. Thus site and treatment (which are often interrelated) seem to be important determinants of physical symptoms but not of psychosocial functioning. A multivariate analysis of these factors has been performed and will be reported elsewhere.49
Some limitations of this study warrant discussion. First, the study was performed in an exclusively white Dutch population. The treatment given, and the Dutch patient's perception of QOL and its domains (physical, psychological, and social) may differ from those of other countries and cultures. Second, we have deliberately excluded patients with a recurrence (even if treatable with curative intent) from our analysis. The results of our study are therefore only applicable to long-term survivors who have not developed a recurrence. Third, the question arises whether our sample is representative for the population of patients with squamous cell carcinoma of the oral cavity, pharynx, or larynx. Twenty-two percent of consecutive patients refused to participate in the study and 25% of the patients who consented to participate dropped out from the study at some point during the 3-year period. The patients who refused to participate did not differ with regard to sociodemographic and medical characteristics, but we have no indication of whether their QOL before and after treatment might differ from that of the patients in the study. For the dropouts, there was a tendency for a worse QOL to be reported for some dimensions on the first and second questionnaires. Thus some kind of positive selection bias is likely to have been introduced as a result of noncompliance during the study. Fourth, as mentioned already, the mechanism of response shift may have occurred, making the interpretation of changes in QOL over time difficult. We have not used any methods (for example, “then”-ratings)43 to assess the occurrence of response shift. However, even if internal standards of patients have shifted during the study period (which is likely to have occurred), it could be argued that we are not interested in absolute changes of QOL but in the experience of the patients at the time of the questionnaire, which is by definition dependent on their actual standard. QOL is sometimes defined as the perceived discrepancy between what one has and what one wants, or expects, or has had (the “gap” theory),50 implicating that QOL should always be related to the actual standard.
This prospective study of patients with head and neck cancer demonstrates short-term deterioration of physical symptoms as a result of treatment and limited long-term physical morbidity. Female sex, higher cancer stage, and combination treatment seem to have had a negative influence. During a period of 3 years, there was a gradual improvement of mood and overall QOL. Social functioning did not change and even improved in the subgroup of patients with laryngeal cancer receiving radiotherapy. Thus the prospects for patients with head and neck cancer do not seem to be as bleak as is sometimes thought, in particular when single modality treatment (surgery or radiotherapy) is possible.