- Email: [email protected]
Quality of life after parotid-sparing IMRT for head-and-neck cancer: A prospective longitudinal study
Int. J. Radiation Oncology Biol. Phys., Vol. 57, No. 1, pp. 61–70, 2003 Copyright © 2003 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/03/$–see front matter
doi:10.1016/S0360-3016(03)00361-4
CLINICAL INVESTIGATION
Head and Neck
QUALITY OF LIFE AFTER PAROTID-SPARING IMRT FOR HEAD-ANDNECK CANCER: A PROSPECTIVE LONGITUDINAL STUDY ALEXANDER LIN, M.D.,* HYUNGJIN M. KIM, SC.D.,† JEFFREY E. TERRELL, M.D.,‡ LAURA A. DAWSON, M.D.,* JONATHAN A. SHIP, D.M.D.,§ AND AVRAHAM EISBRUCH, M.D.* *Department of Radiation Oncology and †Center for Statistical Consultation and Research, Department of Biostatistics, Hospital of the University of Michigan, Ann Arbor, MI, ‡Department of Otolaryngology, Center for Practice Management and Outcomes Research, Ann Arbor Veterans Affairs Hospital, Ann Arbor, MI; §University of Michigan School of Dentistry, Ann Arbor, MI Purpose: Parotid-sparing intensity-modulated radiotherapy (IMRT) for head-and-neck cancer reduces xerostomia compared with standard RT. To assess potential improvements in broader aspects of quality of life (QOL), we initiated a study of patient-reported QOL and its predictors after IMRT. Methods and Materials: This was a prospective longitudinal study of head-and-neck cancer patients receiving multisegmental static IMRT. Patients were given a validated xerostomia questionnaire (XQ), and a validated head-and-neck cancer-related QOL questionnaire consisting of four multi-item domains: Eating, Communication, Pain, and Emotion. The Eating domain contains one question (total of six) asking directly about xerostomia. In both questionnaires, higher scores denote worse symptoms or QOL. The questionnaires and measurements of salivary output from the major glands were completed before RT started (pre-RT) and at 3, 6, and 12 months after RT. The association between the QOL scores and patient-, tumor-, and therapy-related factors was assessed using the random effects model. Results: Thirty-six patients participating in the study completed the questionnaires through 12 months. The XQ scores worsened significantly at 3 months compared with the pre-RT scores, but later they improved gradually through 12 months (p ⴝ 0.003), in parallel with an increase in the salivary output from the spared salivary glands. The QOL summary scores were stable between the baseline (pre-RT) and 3 months after RT scores. Patients receiving postoperative RT (whose pre-RT questionnaires were taken a few weeks after surgery) tended to have improved scores after RT, reflecting the subsidence of acute postoperative sequelae, compared with a tendency toward worsened scores in patients receiving definitive RT. After 3 months, statistically significant improvement was noted in the summary QOL scores for all patients, through 12 months after RT (p ⴝ 0.01). The salivary flow rates, tumor doses, mean oral cavity dose, age, gender, sites or stages of tumor, surgery, and use of chemotherapy were not associated with the QOL scores at any point. The mean dose to the parotid glands correlated with the QOL scores at 3 months (p ⴝ 0.05) but not at other post-RT periods. The XQ and QOL summary scores did not correlate before RT but were significantly correlated at each post-RT point (p < 0.01), with a linear correlation coefficient (r) of 0.59, 0.72, and 0.67 at 3, 6, and 12 months, respectively. At these points, the XQ scores also correlated significantly with the scores of each of the individual QOL domains (p < 0.01), including the domains Pain and Emotion, which did not contain any xerostomia-related question. Conclusion: After parotid-sparing IMRT, a statistically significant correlation was noted between patientreported xerostomia and each of the domains of QOL: Eating, Communication, Pain, and Emotion. Both xerostomia and QOL scores improved significantly over time during the first year after therapy. These results suggest that the efforts to improve xerostomia using IMRT may yield improvements in broad aspects of QOL. © 2003 Elsevier Inc. Head-and-neck cancer, Quality of life, Xerostomia, IMRT, Intensity-modulated radiotherapy.
INTRODUCTION In recent years, efforts have been made to reduce the severity of xerostomia after irradiation (RT) for head-andneck cancer by using three-dimensional radiotherapy and
intensity-modulated radiotherapy (IMRT). The goals of these efforts include sparing the major salivary glands while delivering the fully prescribed doses to the targets (1–10). The sparing of the salivary glands results in preservation of salivary flow rates and reduced observer-assessed and pa-
Reprint requests to: Avraham Eisbruch, M.D., Department of Radiation Oncology, Hospital of the University of Michigan, Ann Arbor, MI 48109-0010. Tel: (734) 936-9337; Fax: (734) 763-7370; E-mail: [email protected] Presented in part at the 44th Annual Meeting of the American Society for Therapeutic Radiology Oncology, New Orleans, 2002.
Supported by NIH Grants CA78165 and CA59827. J. A. Ship is currently at New York University College of Dentistry and the Bluestone Center for Clinical Research, New York, NY. Received Dec 9, 2002, and in revised form Mar 7, 2003. Accepted for publication Mar 18, 2003. 61
62
I. J. Radiation Oncology
● Biology ● Physics
tient-reported xerostomia compared with those expected after standard RT (2– 4, 7, 10). The impact of parotid-sparing IMRT on broader aspects of patients’ health-related quality of life (QOL) is not yet known and was the subject of this study. QOL refers to the individual’s sense of well-being and ability to perform daily tasks, potentially affected by an illness and its treatment. Its assessment typically incorporates several domains of health, including the physical, psychological, and social domains. When QOL is self-assessed by the patient, lower scores and more side effects are revealed compared with when QOL is assessed by clinicians (11), and the results are more reliable as predictors of outcome than clinician-rated performance status measures (12, 13). Several head-and-neck cancerspecific QOL assessment instruments have been developed and validated. The recently revised University of Washington QOL instrument (14), the European Organization for the Research and Treatment of Cancer Head and Neck Cancer– specific module (15), the University of South Florida QOL-RT Instrument (16), and the Functional Assessment of Cancer Therapy–Head and Neck (17) contain a variable number of questions covering various aspects of head-andneck–related QOL. All these instruments are patient rated and use a single cumulative QOL score. The Performance Status Scale for Head and Neck Cancer is a clinician-rated instrument addressing eating and speech but not other domains (18). In this study, we opted to use a head-and-neck cancer-specific QOL instrument (HNQOL) that consists of distinct, individually validated, domains: Pain, Eating, Communication, and Emotion (19). This instrument allowed reliable assessment of the effect of therapy on each of the domains. In addition, this questionnaire contains multiple items per domain compared with the single question per domain used in many other instruments. This has the potential of making the instrument more sensitive in discriminating subtle differences in each domain. The HNQOL instrument was validated in a study of 253 patients with various head-and-neck cancers, in which high internal consistency, construct validity, and test-retest reliability were found (19). In another study of 268 patients with laryngeal cancer randomized to laryngectomy or organ preservation, this instrument demonstrated a high utility in distinguishing between the specific QOL-related issues of patients undergoing primary RT vs. surgery and postoperative RT (20, 21). This was a prospective, longitudinal study of QOL in patients treated with multisegmental IMRT aiming to spare the contralateral parotid gland while treating adequately the targets in both sides of the neck. We have previously reported on the predictors of xerostomia in these patients (3). In the current study, a QOL questionnaire and xerostomia-specific questionnaire (XQ) were given to the patients before RT and periodically after therapy to assess their interrelationships and evolvement over time. At the same intervals, patient-, tumor-, and therapy-related relevant factors were used to assess the predictors of QOL.
Volume 57, Number 1, 2003
Table 1. Xerostomia questionnaire 1. 2. 3. 4. 5. 6. 7. 8.
Rate your difficulty in talking due to dryness Rate your difficulty in chewing due to dryness Rate your difficulty in swallowing solid food due to dryness Rate the frequency of your sleeping problems due to dryness Rate your mouth or throat dryness when eating food Rate your mouth or throat dryness while not eating Rate the frequency of sipping liquids to aid swallowing food Rate the frequency of sipping liquids for oral comfort when not eating
METHODS AND MATERIALS The study population consisted of patients with head-andneck cancer treated at the University of Michigan with comprehensive neck IMRT. All patients provided written informed consent approved by the institutional review board of the University of Michigan. They received primary or postoperative RT planned to achieve major salivary gland sparing. IMRT was delivered with static multisegmental, noncoplanar fields according to previously described methods (1, 22) aiming to spare the contralateral parotid gland predominantly, while treating the primary targets and lymph node neck levels at risk of subclinical disease. In all patients, coverage of the targets, including the primary tumor and lymph node levels at risk, was the primary treatment objective. Our guidelines for the selection and delineation of the targets have been previously reported (23). The prescribed doses were typically 70 Gy to the macroscopic tumor planning target volumes, 56 – 64 Gy to the resected tumor bed planning target volumes, and 50 – 60 Gy to nonresected, subclinical disease planning target volumes, at 2.0 Gy/fraction. The XQ (3, 24) is detailed in Table 1. The XQ has previously been tested for validity and reproducibility. It was found to have high test-retest correlations, high internal consistency, and sensitivity for changes in dryness (3). It consists of eight questions, four of which relate to patient-reported dryness while eating or chewing, and four that relate to dryness while not eating or chewing. The XQ does not include questions related to taste or oral pain. Subjects rated each symptom on an 11-point ordinal Likert scale from “0” to “10,” with higher scores indicating greater dryness or discomfort due to dryness. Each item score was added, and the sum was transformed linearly to produce the final summary score ranging from 0 to 100, with higher scores representing greater levels of xerostomia. The HNQOL is detailed in Table 2. It contains 20 questions grouped in four domains: Eating, Communication, Emotion, and Pain. In addition, a question asking how the patient assessed the disturbance due to cancer and its therapy is included. For each question, patients were asked to specify how much they were bothered by the item, with answers of “not at all,” “slightly,” “moderately,” “a lot,” and “extremely.” The HNQOL instrument and each of its four domains have previously been validated, demonstrat-
QOL after IMRT of head-and-neck cancer
Table 2. Head-and-neck cancer-related quality-oflife questionnaire Domain
Item
Communication
Talk to others Talk on the phone Problems with clarity of voice Problems with volume of voice Problems chewing Dryness while eating Problems with taste Problems swallowing soft foods/solids Problems swallowing liquids Problems opening your mouth Embarrassment about condition Concerns about appearance Emotional problems Financial worries Worry that condition will get worse Frustration about condition Shoulder or neck pain General physical problems Pain in mouth Frequency of pain medication use
Eating
Emotion
Pain
ing high internal consistency, construct validity, and testretest reliability. None of the domains was found to measure related or identical concepts (19 –21). In this study, the scaling of scores of each question was modified to range from 0 to 100, with higher scores corresponding to worse conditions. After scaling, each item score for each domain was added, and the sum was transformed linearly to produce the domain score ranging between 0 and 100. A summary score (which was not included in the previous studies) was also calculated for exploratory reasons, using the same scale. It includes all the items in the questionnaire, transformed linearly to produce the final summary score, with 0 the best and 100 the worst score. A comparison of the two questionnaires revealed that only the domain “Eating” in the HNQOL questionnaire contains a direct xerostomia-related question (“dryness while eating”). A few other questions in the Eating and Communication domains of the HNQOL questionnaire are similar to questions asked in the XQ (problems chewing, problems swallowing solid food, and problems talking). In the HNQOL questionnaire, each of these questions asks about problems in general, while in the XQ, each of these questions specifies that the associated problem is due to dryness. The HNQOL domains Pain and Emotion do not contain any question that may be similar to questions asked in the XQ. The patients answered the XQ and the HNQOL questionnaires at baseline (pre-RT) and at 3, 6, and 12 months after completion of therapy. Patients answered the questionnaires while waiting at the Dental or Radiation Oncology clinics for their scheduled follow-up visits. Mailed questionnaires were not used. The salivary flow rates from each of the major glands were measured at about the time of the questionnaires. The
● A. LIN et al.
63
salivary output was measured from each of the parotid glands using a suction cup attached to the orifice of the parotid duct and from the submandibular/sublingual glands by gentle suction using a micropipette from Wharton’s duct orifices, according to previously described methods (25). Unstimulated saliva was measured first, followed by measurement of salivary flow rates stimulated by the application of 2% citric acid on the dorsum of the tongue. The relationship of patient-, tumor-, and therapy-related variables to QOL was studied. The variables examined were age, gender, site and stage of cancer, chemotherapy status, surgery status, prescribed tumor dose, salivary flow rates, mean dose to each of the major salivary glands, and mean dose to the oral cavity, representing the dose received by the minor salivary glands. Statistical analysis For each measurement time, the correlation between the total and each of the domains of the QOL scores and the XQ scores were examined graphically and also using Pearson’s correlation coefficients (r). In assessing these relationships, if a nonlinear relationship appeared to have a better fit, regression analysis was done using the QOL scores as the dependent variable. In these instances, R2 was used, interpreted as the percentage of variation in the QOL scores explained by the predictor. The association between the QOL scores and other baseline variables were assessed using the t test for binary variables, such as surgery status, and using correlation coefficients for continuous variables, such as dose and salivary flow rate. The changes in the scores between different measurement times were assessed using the Wilcoxon matched-paired signed-rank test. Differences in the scores between subgroups such as the definitive RT or postoperative RT subgroups were tested using the Wilcoxon rank sum test. Statistical significance was set at p ⱕ 0.05. Before statistically modeling how the longitudinal scores of QOL changed over time and corresponded with the XQ scores and other variables, extensive graphic exploration was carried out. To model the relationships statistically, random effects models were fit using all post-RT QOL scores as the response variable. The random-effects model accounts for within-person correlations arising from repeated measures and allows the inclusion of time-varying covariates such as the XQ score or salivary flow rates. The final model included variables that were significantly associated with the QOL scores and also included other variables that were thought to be associated with the QOL scores to control for their effects. RESULTS Between July 1997 and December 2000, 57 consecutive patients scheduled to receive comprehensive neck IMRT were accrued to the protocol. Of these, 36 patients who were disease free and who answered the XQ and QOL questionnaire from before RT through 12 months after RT were the
64
I. J. Radiation Oncology
● Biology ● Physics
Table 3. Patient, tumor, and treatment characteristics Sex (n) Men Women Age (y) Median Range Tumor site (n) Oropharynx Oral cavity Larynx/Hypopharynx Parotid, regionally metastatic Skin, regionally metastatic AJCC tumor stage (n) I II III IV Recurrent T stage (n) 1 2 3 4 Recurrent N stage (n) 0 1 2 Recurrent RT (n) Primary Postoperative Chemotherapy (n) Yes No Prescribed dose to primary planning target volume (Gy) Median Range Contralateral parotid dose (Gy) Mean Range Contralateral submandibular gland dose (Gy) Mean Range Oral cavity dose (Gy) Mean Range
28 8 54 37–76 19 8 3 4 2 1 2 12 18 3 8 14 8 3 3 6 9 18 3 15 21 7 29 65 60–70.2 19 8–30 54 13–69 42 10–62
Abbreviations: AJCC ⫽ American Joint Committee on Cancer; RT ⫽ radiotherapy.
subject of this study. Patient, tumor, and treatment characteristics, including relevant organ doses, are detailed in Table 3. The mean contralateral major salivary gland doses are detailed, as well as the mean oral cavity doses, which were found in a previous study to correlate with the degree of patient-reported xerostomia (3). In patients with lateralized tumors, the mean ipsilateral parotid and submandibular gland doses were close to the prescribed tumor doses. Seven of the patients receiving definitive RT also received chemotherapy consisting of induction cisplatin and 5-fluorouracil and concurrent cisplatin. Severe therapy-related side effects
Volume 57, Number 1, 2003
occurred in 4 patients who received chemotherapy and developed Grade 3-4 acute mucositis, which subsided within several weeks after therapy. No patient required a long-term gastric feeding tube. One patient developed exposed mandibular bone that was treated conservatively. None received salivary stimulating or protecting agents during the study. The salivary flow rates before and periodically after RT are detailed in Table 4. After an initial drop, the flow rates from the contralateral parotid glands recovered at 6 to 12 months and returned, on average, to their pre-RT levels. The flow rates from the ipsilateral parotid and submandibular glands had no significant recovery over time. The XQ scores, QOL summary scores, and the scores of each QOL domain before RT and periodically after therapy are detailed in Table 5 and Fig. 1. Because of the nongaussian distribution of the scores, the median and 25th to 75th percentile range were chosen as the descriptive statistics. The summary QOL scores did not change significantly between the pre-RT and the 3-month post-RT periods, but improved significantly after 6 and 12 months, reaching the best scores at 12 months (12 months vs. 3 months, p ⫽ 0.01). An examination of the scores of each QOL domain revealed that by 3 months, the Eating domain scores had worsened (p ⫽ 0.05) and the other domain scores did not change significantly compared with the pre-RT scores. The scores of the domains Eating and Communication improved significantly with time after RT, reaching their best levels at 12 months. The XQ scores worsened significantly after RT compared with the pre-RT levels (p ⬍ 0.001), as expected. However, with time, these scores improved significantly after RT (3 months vs. 12 months, p ⫽ 0.003). Thus, different trends in the summary QOL and XQ scores were observed from the pre-RT to 3-month periods: stable QOL scores compared with significant worsening of the XQ scores. During the post-RT periods, from 3 through 12 months, significant improvements were noted in both scores. The stable QOL summary scores between the pre-RT and 3-month periods masked opposite trends in the definitive RT and postoperative RT patients, in both the summary and the individual domain scores (Table 6). Because of the small number of patients in each subgroup, the differences were not statistically significant in most instances. The summary QOL scores tended to worsen in the definitive RT group (p ⫽ 0.2) and to improve in the postoperative RT group (p ⫽ 0.18), for whom the pre-RT questionnaires were given a few weeks after surgery. An examination of the trends in the individual domains between the pre-RT and 3-month period revealed significant worsening of the Eating domain scores in the definitive RT group (p ⫽ 0.002) compared with stable scores in the postoperative RT patients and significant improvement in the Pain domain scores in the postoperative group (p ⫽ 0.05), reflecting the subsidence of acute postoperative sequelae. An exploration of the potential factors associated with
QOL after IMRT of head-and-neck cancer
● A. LIN et al.
65
Table 4. Salivary flow rates from the major glands
Unstimulated (mL/min) Ipsilateral parotid Contralateral parotid Submandibular* Stimulated (mL/min) Ipsilateral parotid Contralateral parotid Submandibular*
Pre-RT
3 mo
6 mo
12 mo
0.09 ⫾ 0.1 0.12 ⫾ 0.1 0.05 ⫾ 0.04
0 0.05 ⫾ 0.07 0
0 0.07 ⫾ 0.05 0.01 ⫾ 0.01
0 0.10 ⫾ 0.22 0.04 ⫾ 0.13
0.37 ⫾ 0.30 0.51 ⫾ 0.35 0.18 ⫾ 0.17
0.003 ⫾ 0.01 0.42 ⫾ 0.35 0.04 ⫾ 0.11
0.03 ⫾ 0.07 0.64 ⫾ 0.49 0.04 ⫾ 0.07
0.03 ⫾ 0.06 0.68 ⫾ 0.53 0.04 ⫾ 0.08
Abbreviation: RT ⫽ radiotherapy. Data presented as the mean ⫾ SD. * Combined flow rates of both glands.
the QOL scores revealed that none of the patient-, tumor-, or therapy-related factors examined (salivary flow rates, age, gender, site and stage of cancer, chemotherapy status, surgery status, prescribed tumor dose, mean dose to each of the major salivary glands, and mean dose to the oral cavity) was significantly associated with the QOL scores at any point, except for the mean doses to the contralateral parotid glands. At 3 months after RT, but not at other post-RT points, the mean parotid doses correlated significantly with the QOL scores (p ⫽ 0.048). This finding may have been a result of chance alone, resulting from multiple testing. The examination of the relationships between the XQ and QOL scores revealed no statistically significant correlation before RT (r ⫽ 0.22, p ⫽ 0.32). In contrast, a statistically significant linear correlation was found at each of the post-RT points (p ⬍ 0.01; Figs. 2). The strengths of the relationships appear to be similar in the various post-RT points, with a correlation coefficient (r) of 0.59, 0.72, and 0.67 at 3, 6, and 12 months, respectively. We anticipated a correlation between the XQ scores and the Eating domain scores in the QOL questionnaire. Such a correlation could have driven the correlation between the XQ and global QOL scores. We therefore assessed the correlation between the XQ scores and the scores of each of the domains in the QOL instrument. A statistically significant correlation between the XQ scores and each of the
QOL domains was found at each post-RT point. The relationships for the 12-month post-RT scores are presented in Fig. 3. A statistically significant linear correlation was found between the XQ scores and Eating domain (r ⫽ 0.77, p ⬍ 0.01), Emotion domain (r ⫽ 0.60, p ⬍ 0.01), and the Pain domain (r ⫽ 0.41, p ⫽ 0.01). For the Communication domain, linear and quadratic XQ score terms together were more predictive than a linear term alone (R2 was 0.65 for a quadratic relationship compared with an R2 of 0.22 for a linear relationship), suggesting that the Communication scores are affected by xerostomia only after xerostomia levels reached a certain level. Similar significant relationships between the QOL domain scores and XQ scores were also found in the 3- and 6-month post-RT periods. Table 7 shows the results of a random-effects model describing the QOL summary scores from all post-RT measurement times using XQ scores at the corresponding times. To reduce the baseline interpatient differences, the model was adjusted for the pre-RT scores. The time variable was introduced to adjust for the improving scores over time. The model shows that at each post-RT point, the XQ scores were associated with QOL scores after adjusting for the baseline scores and the time since RT. A 10-point increase in the XQ score would be associated with a 2.6-point increase in the QOL score at 3 months and a 2.1-point increase at 6 or 12 months.
Table 5. QOL and XQ scores before and at various times after RT
XQ scores QOL total scores QOL domain scores Eating Pain Communication Emotion
Pre-RT
3 mo
6 mo
12 mo
7.1 (0–17) 28.6 (10–45)
48.1 (25–56)* 23.8 (14–33)
38.1 (26–53)* 16.1 (11–26)*†
30.6 (22–48)*† 14.3 (9–27)*†
16.7 (0–46) 31.3 (16–58) 12.5 (0–38) 14.6 (8–38)
29.2 (21–42)* 18.8 (13–38) 18.8 (6–34) 12.5 (6–21)
20.8 (17–33) 18.8 (6–31)*† 9.4 (0–25)† 12.5 (4–21)*
20.8 (13–38)† 18.8 (13–28)* 6.3 (0–19)*† 12.5 (4–21)*
Abbreviations: QOL ⫽ quality of life; XQ ⫽ xerostomia questionnaire; RT ⫽ radiotherapy. Data presented as the median, with the 25th to 75th percentile range in parentheses; higher scores denote worse symptoms/reduced QOL. * Significant difference from pre-RT score (p ⬍ 0.05). † Significant difference from 3-mo score (p ⬍ 0.05).
66
I. J. Radiation Oncology
● Biology ● Physics
Fig. 1. Median QOL and XQ scores (and 25th to 75th percentile range) before RT and at various times after RT. Lower scores indicate better patient-reported QOL or xerostomia.
DISCUSSION In many surveys monitoring RT of head-and-neck cancer, patients cited xerostomia as the most prevalent long-term treatment side effect and as a major factor contributing to reduced QOL (11, 26 –28). This study demonstrated statistically significant correlations between patient-reported xerostomia and each of the domains of QOL: Eating, Communication, Pain, and Emotion. These correlations have not been demonstrated before. Although the domains Eating and Communication contain some questions similar to those in the XQ (although not asking specifically about drynessrelated problems, apart for one question in the Eating domain), the domains Pain and Emotion do not contain questions that may relate directly to xerostomia. These findings suggest that there is a major role for the degree of xerostomia in determining the degree of seemingly unrelated domains of QOL. This can occur by the contribution of xerostomia to anxiety/depression, sleep disturbances, and nutritional issues. These findings also suggest that the efforts to reduce xerostomia may improve broad aspects of QOL. Other studies examining determinants of QOL after RT for head-and-neck cancer, using a variety of instruments, have reported conflicting results. Some studies reported no statistically significant correlation between any patient-, tumor-, or therapy-related factor, other than socioeconomic patient status (29, 30) or pretherapy QOL scores (31). Other
Volume 57, Number 1, 2003
studies found a correlation between QOL scores and the tumor stage (20, 26) or target volume (26, 32). The wide range of xerostomia resulting from parotidsparing IMRT may have helped in revealing the correlations between xerostomia and the various QOL domains in our study. A previous, larger, prospective longitudinal study assessed the determinants of the XQ scores in 84 patients similarly treated, using the same XQ used in the current study (3). Statistically significant predictors of the XQ scores were the mean doses to the major salivary glands, mean doses to the oral cavity (representing potential effect of RT on the minor salivary glands), pretherapy XQ scores, and time since RT. Notably, the salivary flow rates correlated only weakly with the degree of xerostomia. The search in the current study for potential determinants of QOL found that only patient-reported xerostomia was a consistently significant determinant throughout the post-RT study period. The doses to the parotid glands were found to be a significant factor at 3 months, but not at 6 and 12 months, after RT. The correlation between the parotid doses and the QOL scores at 3 months, found in this study, is consistent with our previous findings that the salivary gland mean doses correlated with xerostomia (3). However, this may be a chance finding, because many statistical tests were performed. No other patient-, tumor-, or therapy-related factor was found to be a statistically significant determinant of QOL. Thus, despite the correlation between the XQ and QOL scores, and the similar trend of improvement in both scores over time after RT, the predictors of each of the two seem to differ. This may have been a result of the smaller patient numbers in the current study. Another possible explanation is that the QOL scores after IMRT are dependent on additional factors that were not examined in this study. These factors may include the extent of the irradiated mucosa and other nonspecified tissue in the head and neck. The conformal dose distributions produced by IMRT may improve QOL, not only through improving xerostomia by reducing the doses to the salivary glands, but also by reducing the volume of other nontarget tissue receiving a high dose. A recently reported randomized study of pilocarpine vs. placebo concurrent with RT for head-and-neck cancer has
Table 6. QOL scores in definitive and postoperative RT groups before and 3 months after RT Definitive RT (n ⫽ 15)
Summary scores Communication Eating Pain Emotion
Postoperative RT (n ⫽ 21)
Pre-RT
3 mo
Pre-RT
3 mo
19 (8–31) 0 (0–38) 4 (0–15)*† 19 (16–38)† 29 (13–42)
24 (14–33) 19 (0–50) 29 (17–42)* 19 (13–44) 17 (4–25)
29 (10–45) 31 (0–38) 25 (17–54)† 34 (19–58)*† 13 (8–25)
21 (10–33) 16 (6–25) 29 (21–42) 19 (13–38)* 13 (8–21)
Abbreviations as in Table 5. Data presented as the median, with the 25th to 75th percentile range in parentheses; higher scores denote worse symptoms/reduced QOL. * Significant change from pre-RT to 3 mo (p ⬍ 0.05). † Significant difference between definitive and postoperative subgroups (p ⬍ 0.05).
QOL after IMRT of head-and-neck cancer
Fig. 2. (a– c) Correlations between XQ scores and QOL summary scores at 3, 6, and 12 months after RT.
demonstrated increased saliva production in patients receiving pilocarpine (33). However, no advantage of the drug could be demonstrated in the scores of the University of Washington QOL instrument used in the study (33, 34).
● A. LIN et al.
67
Several important differences exist in the outcome of using drugs to stimulate or protect the salivary production vs. the use of IMRT to spare the salivary glands. These differences predict a greater potential of IMRT to affect broad QOL measures. The first is the extent of the salivary flow sparing. Parotid-sparing IMRT preserves both the unstimulated and the stimulated salivary flows, which over time can recover significantly (2, 3), compared with modest sparing by concurrent pilocarpine of only the unstimulated flows (33). The second difference is the sparing of additional tissues achieved by the highly conformal dose distributions of IMRT, which may lead to better QOL compared with pharmacologic interventions that improve only the salivary flows. QOL instruments are useful in deciphering differences in outcome between treatment strategies that achieve similar tumor control and survival. In our study, the differences in QOL between the postoperative and definitive RT subgroups were confined to the pre-RT and early post-RT periods. Notably, the pre-RT scores in the surgery group were dominated by the sequelae of surgery performed a few weeks earlier. These results suggest that the baseline, pretherapy QOL assessment in patients receiving postoperative RT should be taken before surgery, rather than after surgery and before RT. This may pose logistical difficulties, but it would enhance the reliability of the longitudinal assessment of QOL in patients undergoing surgery and postoperative RT. Harrison et al. (28) and the University of Florida group (35) reported QOL and functional results after RT for baseof-tongue cancer that were superior to those expected after surgery. Terrell et al. (20) reported better QOL scores after chemoradiation compared with laryngectomy and postoperative RT in advanced laryngeal cancer patients. These results were related to more freedom from pain and better emotional well-being in the organ preservation group rather than to the preservation of speech function. In our study, the heterogeneity in tumor sites, operations performed, and the small patient numbers in each group precluded a reliable systematic comparison of QOL in patients receiving definitive or postoperative RT. Larger patient numbers will allow a better assessment of QOL after definitive vs. postoperative IMRT in the future. In our study, the summary QOL scores and scores of most domains improved continuously after therapy through 12 months. Other studies assessing longitudinal trends in QOL scores of head-and-neck cancer patients have reported conflicting results. Morton (36) reported little difference in most QOL domains between pretherapy and 12 months after therapy testing. Other authors reported worsening scores from therapy until about 12 months (31, 37, 38). Notwithstanding the heterogeneity in patients, therapy, and QOL instruments, a major difference between the patients in our study and the patients in the QOL studies cited above is the RT technique. After parotid-sparing IMRT in our study, the salivary flow rates and XQ scores improved throughout the posttherapy period, in parallel with an improvement in
68
I. J. Radiation Oncology
● Biology ● Physics
Volume 57, Number 1, 2003
Fig. 3. Correlations between XQ scores and each QOL domain score. (a) Eating domain, (b) Communication domain, (c) Pain domain, and (d) Emotion domain. Fitted values were linear in (a), (c), and (d) and linear and quadratic in (b).
the QOL scores. In contrast, after the standard RT techniques used in other studies of QOL in head-and-neck cancer patients, severe xerostomia was stable or even worsened over time (30, 37, 38). The differences in the extent and kinetics of recovery of xerostomia, related to the different RT techniques, may explain the differences between our study and the previous QOL studies in the longitudinal pattern and the trend over time of the QOL scores. Table 7. Model of post-RT QOL summary scores (at each point, corresponding xerostomia score used in model)
Pre-RT QOL Months since RT 3-mo XQ score 6-mo XQ score 12-mo XQ score Intercept
Coef.
SE
p
95% CI
0.19 ⫺0.30 0.26 0.21 0.21 7.59
0.09 0.55 0.08 0.08 0.09 5.50
0.025 0.589 0.002 0.006 0.015 0.167
0.02, 0.36 ⫺1.38, 0.78 0.10, 0.42 0.06, 0.36 0.04, 0.39 ⫺3.18, 18.37
Abbreviations: Coef. ⫽ coefficient; SE ⫽ standard error; CI ⫽ confidence interval; other abbreviations as in Table 5.
The statistically significant association between patientreported xerostomia and general QOL found in this study at all post-RT periods, as well as the similarity in their longitudinal recovery over time, suggests that the efforts to reduce xerostomia by IMRT may result in even broader benefits in QOL. We regard these results as hypothesis generating, rather than proving a cause– effect relationship. There may still be undetected factors that are related to both the xerostomia and the QOL scores, and other factors that may be independent of the technique of RT. The hypothesis suggested by the data in this study needs to be tested therefore, preferentially in a randomized study of IMRT vs. standard RT. Such a study, in which xerostomia and general QOL will be assessed, will start soon in Europe (40). CONCLUSION This study has demonstrated strong relationships between xerostomia and broad aspects of QOL after IMRT of headand-neck cancer. These findings strengthen the rationale for the efforts to reduce xerostomia by sparing the salivary glands, without compromising target RT.
QOL after IMRT of head-and-neck cancer
● A. LIN et al.
69
REFERENCES 1. Eisbruch A, Marsh LH, Martel MK, et al. Comprehensive irradiation of head and neck cancer using conformal multisegmental fields: Assessment of target coverage and noninvolved tissue sparing. Int J Radiat Oncol Biol Phys 1998; 41:559–568. 2. Eisbruch A, Kim HM, Ten Haken R, et al. Dose, volume and function relationships in parotid glands following conformal and intensity modulated irradiation of head and neck cancer. Int J Radiat Oncol Biol Phys 1999;45:577–587. 3. Eisbruch A, Kim HM, Terrell JE, et al. Xerostomia and its predictors following parotid-sparing irradiation of head and neck cancer. Int J Radiat Oncol Biol Phys 2001;50:695– 704. 4. Chao KSC, Deasy JO, Markman J, et al. A prospective study of salivary function sparing in patients with head and neck cancers receiving intensity-modulated or three-dimensional radiation therapy: Initial results. Int J Radiat Oncol Biol Phys 2001;51:938–946. 5. Hunt MA, Zelefsky MJ, Wolden S, et al. Treatment planning and delivery of intensity-modulated radiation therapy for primary nasopharyngeal cancer. Int J Radiat Oncol Biol Phys 2001;49:623–632. 6. De Neve W, De Gersem W, Derycke S. Clinical delivery of IMRT for relapsed or second-primary head and neck cancer using a multileaf collimator with dynamic control. Radiother Oncol 1999;50:301–314. 7. Lee N, Xia P, Quivey JM, et al. Intensity modulated radiotherapy in the treatment of nasopharyngeal carcinoma: An update of the UCSF experience. Int J Radiat Oncol Biol Phys 2002;53:12–22. 8. Wu Q, Manning M, Schmidt-Ullrich R, et al. The potential for sparing of parotids and escalation of biologically equivalent dose with intensity modulated radiation treatments of head and neck cancers: A treatment design study. Int J Radiat Oncol Biol Phys 2000;46:195–205. 9. Butler EB, Teh BS, Grant WS, et al. SMART (simultaneous modulated accelerated radiation therapy) boost: A new accelerated fractionation schedule for the treatment of head and neck cancer with intensity modulated radiotherapy. Int J Radiat Oncol Biol Phys 1999;45:21–32. 10. Maes A, Weltens C, Flamen P, et al. Preservation of parotid function with uncomplicated conformal radiotherapy. Radiother Oncol 2002;63:203–211. 11. Bjordal K, Kasa S, Matekaasa A. Quality of life in patients treated for head and neck cancer: A follow-up study 7 to 11 years after radiotherapy. Int J Radiat Oncol Biol Phys 1994; 28:847–855. 12. Ganz P, Lee J, Siau J. Quality of life assessment: An independent prognostic variable for survival in lung cancer. Cancer 1991;67:3131–3135. 13. Ruckdeschel J, Piantodosi S, for the Lung Cancer Study Group. Quality of life assessment in lung surgery for bronchogenic carcinoma. J Thorac Surg 1991;6:201–205. 14. Weymuller EA, Alsarraf R, Yueh B, et al. Analysis of performance characteristics of the University of Washington quality of life instrument and its modification. Arch Otolaryngol Head Neck Surg 2001;127:489–493. 15. Bjordal K, Kaas S. Psychometric validation of the EORTC core quality of life questionnaire, 30-item version and a diagnostic-specific module for head and neck cancer patients. Acta Oncol 1992;31:311–321. 16. Trotti A, Johnson DJ, Gwede C, et al. Development of a head and neck companion module for the quality of life-radiation therapy instrument (QOL-RTI). Int J Radiat Oncol Biol Phys 1998;42:257–261.
17. Cella DF, Bonomi AE. Measuring quality of life. Oncology (Huntingt) Suppl 1995;9:47–60. 18. List MA, D’Antonio L, Cella DF, et al. The performance status scale for head and neck cancer patients and the functional assessment of cancer therapy-head and neck cancer scale: A study of utility and validity. Cancer 1996;77:2294– 2301. 19. Terrell JE, Nanavati KA, Esclamado RM, et al. Head and neck cancer-specific quality of life: Instrument validation. Arch Otolaryngol Head Neck Surg 1997;123:1125–1132. 20. Terrell JE, Fisher SG, Wolf GT. Long-term quality of life after treatment of laryngeal cancer. Arch Otolaryngol Head Neck Surg 1998;124:964–971. 21. Terrell JE. Quality of life assessment in head and neck cancer patients. Hematol Oncol Clin North Am 1999;13:849–863. 22. Fraass BA, Kessler ML, McShan DL, et al. Optimization and clinical use of multisegment intensity-modulated radiation therapy for high-dose conformal therapy. Semin Radiat Oncol 1999;9:60–77. 23. Eisbruch A, Foote RL, O’Sullivan B, et al. IMRT for head and neck cancer: Emphasis on the selection and delineation of the targets. Semin Radiat Oncol 2002;12:238–249. 24. Henson BS, Inglehart MR, Eisbruch A, et al. Preserved salivary output and xerostomia-related quality of life in head and neck cancer patients receiving parotid-sparing radiotherapy. Oral Oncol 2001;37:84–93. 25. Ship JA, Fox PC, Baum BJ. Normal salivary gland function: How much saliva is enough? J Am Dent Assoc 1991;122:63– 69. 26. Epstein JB, Emerton S, Kolbinson DS, et al. Quality of life and oral function following radiotherapy for head and neck cancer. Head Neck 1999;21:1–11. 27. Jensen AB, Hanse O, Jorgensen K, et al. Influence of late side-effects upon daily life after radiotherapy for laryngeal and pharyngeal cancer. Acta Oncol 1994;33:487–491. 28. Harrison LB, Zelefski MJ, Pfitzer DG, et al. Detailed quality of life assessment in patients treated with primary radiotherapy for cancer of the base of tongue. Head Neck 1997;19:169– 175. 29. Sehlen S, Hollenhorst H, Lenk M. Only sociodemographic variables predict quality of life after radiotherapy in patients with head and neck cancer. Int J Radiat Oncol Biol Phys 2002;52:779–883. 30. Fang FM, Chiu HC, Kuo WR. Health-related quality of life for nasopharyngeal carcinoma patients with cancer-free survival after treatment. Int J Radiat Oncol Biol Phys 2002;53:959– 968. 31. List MA, Sison A, Haraf D, et al. Quality of life and performance in advanced head and neck cancer patients on concomitant chemoradiotherapy: A prospective examination. J Clin Oncol 1999;17:1020–1028. 32. Huguenin PU, Taussky D, Moe K, et al. Quality of life in patients cured from a carcinoma of the head and neck by radiotherapy: The importance of the target volume. Int J Radiat Oncol Biol Phys 1999;45:47–52. 33. Scarantino CW, Leveque FG, Scott CB, et al. A phase III study of concomitant oral pilocarpine to reduce hyposalivation and mucositis associated with curative radiation therapy in head and neck cancer: RTOG 9709 [Abstract]. Proc ASCO 2001;20:225a. 34. Fisher JJ, Scott CB, Scarantino CW, et al. Phase III quality of life study: Reduction in hyposalivation does not improve QOL for head and neck cancer patients post radiation therapy [Abstract]. Proc ASCO 2002;21:356a. 35. Moore GJ, Parsons JT, Mendenhall WM. Quality of life outcomes after primary radiotherapy for squamous cell carci-
70
I. J. Radiation Oncology
● Biology ● Physics
noma of the base of tongue. Int J Radiat Oncol Biol Phys 1996;36:351–354. 36. Morton RP. Laryngeal cancer: Quality of life and cost effectiveness. Head Neck 1997;19:243–250. 37. Rogers SN, Lowe D, Brown JS, et al. The University of Washington head and neck cancer measure as a predictor of outcome following primary surgery for oral cancer. Head Neck 1999;21:394–401.
Volume 57, Number 1, 2003
38. Hammerlid E, Silander E, Hornestam L. Health related quality of life three years after diagnosis of head and neck cancer—A longitudinal study. Head Neck 2001;23:113– 125. 39. Bundgaard T, Tandrup O, Elbrond O. A functional evaluation of patients treated for oral cancer. Int J Oral Maxillofac Surg 1993;22:28–34. 40. Nutting C. Personal communication, 2002.
doi:10.1016/S0360-3016(03)00361-4
CLINICAL INVESTIGATION
Head and Neck
QUALITY OF LIFE AFTER PAROTID-SPARING IMRT FOR HEAD-ANDNECK CANCER: A PROSPECTIVE LONGITUDINAL STUDY ALEXANDER LIN, M.D.,* HYUNGJIN M. KIM, SC.D.,† JEFFREY E. TERRELL, M.D.,‡ LAURA A. DAWSON, M.D.,* JONATHAN A. SHIP, D.M.D.,§ AND AVRAHAM EISBRUCH, M.D.* *Department of Radiation Oncology and †Center for Statistical Consultation and Research, Department of Biostatistics, Hospital of the University of Michigan, Ann Arbor, MI, ‡Department of Otolaryngology, Center for Practice Management and Outcomes Research, Ann Arbor Veterans Affairs Hospital, Ann Arbor, MI; §University of Michigan School of Dentistry, Ann Arbor, MI Purpose: Parotid-sparing intensity-modulated radiotherapy (IMRT) for head-and-neck cancer reduces xerostomia compared with standard RT. To assess potential improvements in broader aspects of quality of life (QOL), we initiated a study of patient-reported QOL and its predictors after IMRT. Methods and Materials: This was a prospective longitudinal study of head-and-neck cancer patients receiving multisegmental static IMRT. Patients were given a validated xerostomia questionnaire (XQ), and a validated head-and-neck cancer-related QOL questionnaire consisting of four multi-item domains: Eating, Communication, Pain, and Emotion. The Eating domain contains one question (total of six) asking directly about xerostomia. In both questionnaires, higher scores denote worse symptoms or QOL. The questionnaires and measurements of salivary output from the major glands were completed before RT started (pre-RT) and at 3, 6, and 12 months after RT. The association between the QOL scores and patient-, tumor-, and therapy-related factors was assessed using the random effects model. Results: Thirty-six patients participating in the study completed the questionnaires through 12 months. The XQ scores worsened significantly at 3 months compared with the pre-RT scores, but later they improved gradually through 12 months (p ⴝ 0.003), in parallel with an increase in the salivary output from the spared salivary glands. The QOL summary scores were stable between the baseline (pre-RT) and 3 months after RT scores. Patients receiving postoperative RT (whose pre-RT questionnaires were taken a few weeks after surgery) tended to have improved scores after RT, reflecting the subsidence of acute postoperative sequelae, compared with a tendency toward worsened scores in patients receiving definitive RT. After 3 months, statistically significant improvement was noted in the summary QOL scores for all patients, through 12 months after RT (p ⴝ 0.01). The salivary flow rates, tumor doses, mean oral cavity dose, age, gender, sites or stages of tumor, surgery, and use of chemotherapy were not associated with the QOL scores at any point. The mean dose to the parotid glands correlated with the QOL scores at 3 months (p ⴝ 0.05) but not at other post-RT periods. The XQ and QOL summary scores did not correlate before RT but were significantly correlated at each post-RT point (p < 0.01), with a linear correlation coefficient (r) of 0.59, 0.72, and 0.67 at 3, 6, and 12 months, respectively. At these points, the XQ scores also correlated significantly with the scores of each of the individual QOL domains (p < 0.01), including the domains Pain and Emotion, which did not contain any xerostomia-related question. Conclusion: After parotid-sparing IMRT, a statistically significant correlation was noted between patientreported xerostomia and each of the domains of QOL: Eating, Communication, Pain, and Emotion. Both xerostomia and QOL scores improved significantly over time during the first year after therapy. These results suggest that the efforts to improve xerostomia using IMRT may yield improvements in broad aspects of QOL. © 2003 Elsevier Inc. Head-and-neck cancer, Quality of life, Xerostomia, IMRT, Intensity-modulated radiotherapy.
INTRODUCTION In recent years, efforts have been made to reduce the severity of xerostomia after irradiation (RT) for head-andneck cancer by using three-dimensional radiotherapy and
intensity-modulated radiotherapy (IMRT). The goals of these efforts include sparing the major salivary glands while delivering the fully prescribed doses to the targets (1–10). The sparing of the salivary glands results in preservation of salivary flow rates and reduced observer-assessed and pa-
Reprint requests to: Avraham Eisbruch, M.D., Department of Radiation Oncology, Hospital of the University of Michigan, Ann Arbor, MI 48109-0010. Tel: (734) 936-9337; Fax: (734) 763-7370; E-mail: [email protected] Presented in part at the 44th Annual Meeting of the American Society for Therapeutic Radiology Oncology, New Orleans, 2002.
Supported by NIH Grants CA78165 and CA59827. J. A. Ship is currently at New York University College of Dentistry and the Bluestone Center for Clinical Research, New York, NY. Received Dec 9, 2002, and in revised form Mar 7, 2003. Accepted for publication Mar 18, 2003. 61
62
I. J. Radiation Oncology
● Biology ● Physics
tient-reported xerostomia compared with those expected after standard RT (2– 4, 7, 10). The impact of parotid-sparing IMRT on broader aspects of patients’ health-related quality of life (QOL) is not yet known and was the subject of this study. QOL refers to the individual’s sense of well-being and ability to perform daily tasks, potentially affected by an illness and its treatment. Its assessment typically incorporates several domains of health, including the physical, psychological, and social domains. When QOL is self-assessed by the patient, lower scores and more side effects are revealed compared with when QOL is assessed by clinicians (11), and the results are more reliable as predictors of outcome than clinician-rated performance status measures (12, 13). Several head-and-neck cancerspecific QOL assessment instruments have been developed and validated. The recently revised University of Washington QOL instrument (14), the European Organization for the Research and Treatment of Cancer Head and Neck Cancer– specific module (15), the University of South Florida QOL-RT Instrument (16), and the Functional Assessment of Cancer Therapy–Head and Neck (17) contain a variable number of questions covering various aspects of head-andneck–related QOL. All these instruments are patient rated and use a single cumulative QOL score. The Performance Status Scale for Head and Neck Cancer is a clinician-rated instrument addressing eating and speech but not other domains (18). In this study, we opted to use a head-and-neck cancer-specific QOL instrument (HNQOL) that consists of distinct, individually validated, domains: Pain, Eating, Communication, and Emotion (19). This instrument allowed reliable assessment of the effect of therapy on each of the domains. In addition, this questionnaire contains multiple items per domain compared with the single question per domain used in many other instruments. This has the potential of making the instrument more sensitive in discriminating subtle differences in each domain. The HNQOL instrument was validated in a study of 253 patients with various head-and-neck cancers, in which high internal consistency, construct validity, and test-retest reliability were found (19). In another study of 268 patients with laryngeal cancer randomized to laryngectomy or organ preservation, this instrument demonstrated a high utility in distinguishing between the specific QOL-related issues of patients undergoing primary RT vs. surgery and postoperative RT (20, 21). This was a prospective, longitudinal study of QOL in patients treated with multisegmental IMRT aiming to spare the contralateral parotid gland while treating adequately the targets in both sides of the neck. We have previously reported on the predictors of xerostomia in these patients (3). In the current study, a QOL questionnaire and xerostomia-specific questionnaire (XQ) were given to the patients before RT and periodically after therapy to assess their interrelationships and evolvement over time. At the same intervals, patient-, tumor-, and therapy-related relevant factors were used to assess the predictors of QOL.
Volume 57, Number 1, 2003
Table 1. Xerostomia questionnaire 1. 2. 3. 4. 5. 6. 7. 8.
Rate your difficulty in talking due to dryness Rate your difficulty in chewing due to dryness Rate your difficulty in swallowing solid food due to dryness Rate the frequency of your sleeping problems due to dryness Rate your mouth or throat dryness when eating food Rate your mouth or throat dryness while not eating Rate the frequency of sipping liquids to aid swallowing food Rate the frequency of sipping liquids for oral comfort when not eating
METHODS AND MATERIALS The study population consisted of patients with head-andneck cancer treated at the University of Michigan with comprehensive neck IMRT. All patients provided written informed consent approved by the institutional review board of the University of Michigan. They received primary or postoperative RT planned to achieve major salivary gland sparing. IMRT was delivered with static multisegmental, noncoplanar fields according to previously described methods (1, 22) aiming to spare the contralateral parotid gland predominantly, while treating the primary targets and lymph node neck levels at risk of subclinical disease. In all patients, coverage of the targets, including the primary tumor and lymph node levels at risk, was the primary treatment objective. Our guidelines for the selection and delineation of the targets have been previously reported (23). The prescribed doses were typically 70 Gy to the macroscopic tumor planning target volumes, 56 – 64 Gy to the resected tumor bed planning target volumes, and 50 – 60 Gy to nonresected, subclinical disease planning target volumes, at 2.0 Gy/fraction. The XQ (3, 24) is detailed in Table 1. The XQ has previously been tested for validity and reproducibility. It was found to have high test-retest correlations, high internal consistency, and sensitivity for changes in dryness (3). It consists of eight questions, four of which relate to patient-reported dryness while eating or chewing, and four that relate to dryness while not eating or chewing. The XQ does not include questions related to taste or oral pain. Subjects rated each symptom on an 11-point ordinal Likert scale from “0” to “10,” with higher scores indicating greater dryness or discomfort due to dryness. Each item score was added, and the sum was transformed linearly to produce the final summary score ranging from 0 to 100, with higher scores representing greater levels of xerostomia. The HNQOL is detailed in Table 2. It contains 20 questions grouped in four domains: Eating, Communication, Emotion, and Pain. In addition, a question asking how the patient assessed the disturbance due to cancer and its therapy is included. For each question, patients were asked to specify how much they were bothered by the item, with answers of “not at all,” “slightly,” “moderately,” “a lot,” and “extremely.” The HNQOL instrument and each of its four domains have previously been validated, demonstrat-
QOL after IMRT of head-and-neck cancer
Table 2. Head-and-neck cancer-related quality-oflife questionnaire Domain
Item
Communication
Talk to others Talk on the phone Problems with clarity of voice Problems with volume of voice Problems chewing Dryness while eating Problems with taste Problems swallowing soft foods/solids Problems swallowing liquids Problems opening your mouth Embarrassment about condition Concerns about appearance Emotional problems Financial worries Worry that condition will get worse Frustration about condition Shoulder or neck pain General physical problems Pain in mouth Frequency of pain medication use
Eating
Emotion
Pain
ing high internal consistency, construct validity, and testretest reliability. None of the domains was found to measure related or identical concepts (19 –21). In this study, the scaling of scores of each question was modified to range from 0 to 100, with higher scores corresponding to worse conditions. After scaling, each item score for each domain was added, and the sum was transformed linearly to produce the domain score ranging between 0 and 100. A summary score (which was not included in the previous studies) was also calculated for exploratory reasons, using the same scale. It includes all the items in the questionnaire, transformed linearly to produce the final summary score, with 0 the best and 100 the worst score. A comparison of the two questionnaires revealed that only the domain “Eating” in the HNQOL questionnaire contains a direct xerostomia-related question (“dryness while eating”). A few other questions in the Eating and Communication domains of the HNQOL questionnaire are similar to questions asked in the XQ (problems chewing, problems swallowing solid food, and problems talking). In the HNQOL questionnaire, each of these questions asks about problems in general, while in the XQ, each of these questions specifies that the associated problem is due to dryness. The HNQOL domains Pain and Emotion do not contain any question that may be similar to questions asked in the XQ. The patients answered the XQ and the HNQOL questionnaires at baseline (pre-RT) and at 3, 6, and 12 months after completion of therapy. Patients answered the questionnaires while waiting at the Dental or Radiation Oncology clinics for their scheduled follow-up visits. Mailed questionnaires were not used. The salivary flow rates from each of the major glands were measured at about the time of the questionnaires. The
● A. LIN et al.
63
salivary output was measured from each of the parotid glands using a suction cup attached to the orifice of the parotid duct and from the submandibular/sublingual glands by gentle suction using a micropipette from Wharton’s duct orifices, according to previously described methods (25). Unstimulated saliva was measured first, followed by measurement of salivary flow rates stimulated by the application of 2% citric acid on the dorsum of the tongue. The relationship of patient-, tumor-, and therapy-related variables to QOL was studied. The variables examined were age, gender, site and stage of cancer, chemotherapy status, surgery status, prescribed tumor dose, salivary flow rates, mean dose to each of the major salivary glands, and mean dose to the oral cavity, representing the dose received by the minor salivary glands. Statistical analysis For each measurement time, the correlation between the total and each of the domains of the QOL scores and the XQ scores were examined graphically and also using Pearson’s correlation coefficients (r). In assessing these relationships, if a nonlinear relationship appeared to have a better fit, regression analysis was done using the QOL scores as the dependent variable. In these instances, R2 was used, interpreted as the percentage of variation in the QOL scores explained by the predictor. The association between the QOL scores and other baseline variables were assessed using the t test for binary variables, such as surgery status, and using correlation coefficients for continuous variables, such as dose and salivary flow rate. The changes in the scores between different measurement times were assessed using the Wilcoxon matched-paired signed-rank test. Differences in the scores between subgroups such as the definitive RT or postoperative RT subgroups were tested using the Wilcoxon rank sum test. Statistical significance was set at p ⱕ 0.05. Before statistically modeling how the longitudinal scores of QOL changed over time and corresponded with the XQ scores and other variables, extensive graphic exploration was carried out. To model the relationships statistically, random effects models were fit using all post-RT QOL scores as the response variable. The random-effects model accounts for within-person correlations arising from repeated measures and allows the inclusion of time-varying covariates such as the XQ score or salivary flow rates. The final model included variables that were significantly associated with the QOL scores and also included other variables that were thought to be associated with the QOL scores to control for their effects. RESULTS Between July 1997 and December 2000, 57 consecutive patients scheduled to receive comprehensive neck IMRT were accrued to the protocol. Of these, 36 patients who were disease free and who answered the XQ and QOL questionnaire from before RT through 12 months after RT were the
64
I. J. Radiation Oncology
● Biology ● Physics
Table 3. Patient, tumor, and treatment characteristics Sex (n) Men Women Age (y) Median Range Tumor site (n) Oropharynx Oral cavity Larynx/Hypopharynx Parotid, regionally metastatic Skin, regionally metastatic AJCC tumor stage (n) I II III IV Recurrent T stage (n) 1 2 3 4 Recurrent N stage (n) 0 1 2 Recurrent RT (n) Primary Postoperative Chemotherapy (n) Yes No Prescribed dose to primary planning target volume (Gy) Median Range Contralateral parotid dose (Gy) Mean Range Contralateral submandibular gland dose (Gy) Mean Range Oral cavity dose (Gy) Mean Range
28 8 54 37–76 19 8 3 4 2 1 2 12 18 3 8 14 8 3 3 6 9 18 3 15 21 7 29 65 60–70.2 19 8–30 54 13–69 42 10–62
Abbreviations: AJCC ⫽ American Joint Committee on Cancer; RT ⫽ radiotherapy.
subject of this study. Patient, tumor, and treatment characteristics, including relevant organ doses, are detailed in Table 3. The mean contralateral major salivary gland doses are detailed, as well as the mean oral cavity doses, which were found in a previous study to correlate with the degree of patient-reported xerostomia (3). In patients with lateralized tumors, the mean ipsilateral parotid and submandibular gland doses were close to the prescribed tumor doses. Seven of the patients receiving definitive RT also received chemotherapy consisting of induction cisplatin and 5-fluorouracil and concurrent cisplatin. Severe therapy-related side effects
Volume 57, Number 1, 2003
occurred in 4 patients who received chemotherapy and developed Grade 3-4 acute mucositis, which subsided within several weeks after therapy. No patient required a long-term gastric feeding tube. One patient developed exposed mandibular bone that was treated conservatively. None received salivary stimulating or protecting agents during the study. The salivary flow rates before and periodically after RT are detailed in Table 4. After an initial drop, the flow rates from the contralateral parotid glands recovered at 6 to 12 months and returned, on average, to their pre-RT levels. The flow rates from the ipsilateral parotid and submandibular glands had no significant recovery over time. The XQ scores, QOL summary scores, and the scores of each QOL domain before RT and periodically after therapy are detailed in Table 5 and Fig. 1. Because of the nongaussian distribution of the scores, the median and 25th to 75th percentile range were chosen as the descriptive statistics. The summary QOL scores did not change significantly between the pre-RT and the 3-month post-RT periods, but improved significantly after 6 and 12 months, reaching the best scores at 12 months (12 months vs. 3 months, p ⫽ 0.01). An examination of the scores of each QOL domain revealed that by 3 months, the Eating domain scores had worsened (p ⫽ 0.05) and the other domain scores did not change significantly compared with the pre-RT scores. The scores of the domains Eating and Communication improved significantly with time after RT, reaching their best levels at 12 months. The XQ scores worsened significantly after RT compared with the pre-RT levels (p ⬍ 0.001), as expected. However, with time, these scores improved significantly after RT (3 months vs. 12 months, p ⫽ 0.003). Thus, different trends in the summary QOL and XQ scores were observed from the pre-RT to 3-month periods: stable QOL scores compared with significant worsening of the XQ scores. During the post-RT periods, from 3 through 12 months, significant improvements were noted in both scores. The stable QOL summary scores between the pre-RT and 3-month periods masked opposite trends in the definitive RT and postoperative RT patients, in both the summary and the individual domain scores (Table 6). Because of the small number of patients in each subgroup, the differences were not statistically significant in most instances. The summary QOL scores tended to worsen in the definitive RT group (p ⫽ 0.2) and to improve in the postoperative RT group (p ⫽ 0.18), for whom the pre-RT questionnaires were given a few weeks after surgery. An examination of the trends in the individual domains between the pre-RT and 3-month period revealed significant worsening of the Eating domain scores in the definitive RT group (p ⫽ 0.002) compared with stable scores in the postoperative RT patients and significant improvement in the Pain domain scores in the postoperative group (p ⫽ 0.05), reflecting the subsidence of acute postoperative sequelae. An exploration of the potential factors associated with
QOL after IMRT of head-and-neck cancer
● A. LIN et al.
65
Table 4. Salivary flow rates from the major glands
Unstimulated (mL/min) Ipsilateral parotid Contralateral parotid Submandibular* Stimulated (mL/min) Ipsilateral parotid Contralateral parotid Submandibular*
Pre-RT
3 mo
6 mo
12 mo
0.09 ⫾ 0.1 0.12 ⫾ 0.1 0.05 ⫾ 0.04
0 0.05 ⫾ 0.07 0
0 0.07 ⫾ 0.05 0.01 ⫾ 0.01
0 0.10 ⫾ 0.22 0.04 ⫾ 0.13
0.37 ⫾ 0.30 0.51 ⫾ 0.35 0.18 ⫾ 0.17
0.003 ⫾ 0.01 0.42 ⫾ 0.35 0.04 ⫾ 0.11
0.03 ⫾ 0.07 0.64 ⫾ 0.49 0.04 ⫾ 0.07
0.03 ⫾ 0.06 0.68 ⫾ 0.53 0.04 ⫾ 0.08
Abbreviation: RT ⫽ radiotherapy. Data presented as the mean ⫾ SD. * Combined flow rates of both glands.
the QOL scores revealed that none of the patient-, tumor-, or therapy-related factors examined (salivary flow rates, age, gender, site and stage of cancer, chemotherapy status, surgery status, prescribed tumor dose, mean dose to each of the major salivary glands, and mean dose to the oral cavity) was significantly associated with the QOL scores at any point, except for the mean doses to the contralateral parotid glands. At 3 months after RT, but not at other post-RT points, the mean parotid doses correlated significantly with the QOL scores (p ⫽ 0.048). This finding may have been a result of chance alone, resulting from multiple testing. The examination of the relationships between the XQ and QOL scores revealed no statistically significant correlation before RT (r ⫽ 0.22, p ⫽ 0.32). In contrast, a statistically significant linear correlation was found at each of the post-RT points (p ⬍ 0.01; Figs. 2). The strengths of the relationships appear to be similar in the various post-RT points, with a correlation coefficient (r) of 0.59, 0.72, and 0.67 at 3, 6, and 12 months, respectively. We anticipated a correlation between the XQ scores and the Eating domain scores in the QOL questionnaire. Such a correlation could have driven the correlation between the XQ and global QOL scores. We therefore assessed the correlation between the XQ scores and the scores of each of the domains in the QOL instrument. A statistically significant correlation between the XQ scores and each of the
QOL domains was found at each post-RT point. The relationships for the 12-month post-RT scores are presented in Fig. 3. A statistically significant linear correlation was found between the XQ scores and Eating domain (r ⫽ 0.77, p ⬍ 0.01), Emotion domain (r ⫽ 0.60, p ⬍ 0.01), and the Pain domain (r ⫽ 0.41, p ⫽ 0.01). For the Communication domain, linear and quadratic XQ score terms together were more predictive than a linear term alone (R2 was 0.65 for a quadratic relationship compared with an R2 of 0.22 for a linear relationship), suggesting that the Communication scores are affected by xerostomia only after xerostomia levels reached a certain level. Similar significant relationships between the QOL domain scores and XQ scores were also found in the 3- and 6-month post-RT periods. Table 7 shows the results of a random-effects model describing the QOL summary scores from all post-RT measurement times using XQ scores at the corresponding times. To reduce the baseline interpatient differences, the model was adjusted for the pre-RT scores. The time variable was introduced to adjust for the improving scores over time. The model shows that at each post-RT point, the XQ scores were associated with QOL scores after adjusting for the baseline scores and the time since RT. A 10-point increase in the XQ score would be associated with a 2.6-point increase in the QOL score at 3 months and a 2.1-point increase at 6 or 12 months.
Table 5. QOL and XQ scores before and at various times after RT
XQ scores QOL total scores QOL domain scores Eating Pain Communication Emotion
Pre-RT
3 mo
6 mo
12 mo
7.1 (0–17) 28.6 (10–45)
48.1 (25–56)* 23.8 (14–33)
38.1 (26–53)* 16.1 (11–26)*†
30.6 (22–48)*† 14.3 (9–27)*†
16.7 (0–46) 31.3 (16–58) 12.5 (0–38) 14.6 (8–38)
29.2 (21–42)* 18.8 (13–38) 18.8 (6–34) 12.5 (6–21)
20.8 (17–33) 18.8 (6–31)*† 9.4 (0–25)† 12.5 (4–21)*
20.8 (13–38)† 18.8 (13–28)* 6.3 (0–19)*† 12.5 (4–21)*
Abbreviations: QOL ⫽ quality of life; XQ ⫽ xerostomia questionnaire; RT ⫽ radiotherapy. Data presented as the median, with the 25th to 75th percentile range in parentheses; higher scores denote worse symptoms/reduced QOL. * Significant difference from pre-RT score (p ⬍ 0.05). † Significant difference from 3-mo score (p ⬍ 0.05).
66
I. J. Radiation Oncology
● Biology ● Physics
Fig. 1. Median QOL and XQ scores (and 25th to 75th percentile range) before RT and at various times after RT. Lower scores indicate better patient-reported QOL or xerostomia.
DISCUSSION In many surveys monitoring RT of head-and-neck cancer, patients cited xerostomia as the most prevalent long-term treatment side effect and as a major factor contributing to reduced QOL (11, 26 –28). This study demonstrated statistically significant correlations between patient-reported xerostomia and each of the domains of QOL: Eating, Communication, Pain, and Emotion. These correlations have not been demonstrated before. Although the domains Eating and Communication contain some questions similar to those in the XQ (although not asking specifically about drynessrelated problems, apart for one question in the Eating domain), the domains Pain and Emotion do not contain questions that may relate directly to xerostomia. These findings suggest that there is a major role for the degree of xerostomia in determining the degree of seemingly unrelated domains of QOL. This can occur by the contribution of xerostomia to anxiety/depression, sleep disturbances, and nutritional issues. These findings also suggest that the efforts to reduce xerostomia may improve broad aspects of QOL. Other studies examining determinants of QOL after RT for head-and-neck cancer, using a variety of instruments, have reported conflicting results. Some studies reported no statistically significant correlation between any patient-, tumor-, or therapy-related factor, other than socioeconomic patient status (29, 30) or pretherapy QOL scores (31). Other
Volume 57, Number 1, 2003
studies found a correlation between QOL scores and the tumor stage (20, 26) or target volume (26, 32). The wide range of xerostomia resulting from parotidsparing IMRT may have helped in revealing the correlations between xerostomia and the various QOL domains in our study. A previous, larger, prospective longitudinal study assessed the determinants of the XQ scores in 84 patients similarly treated, using the same XQ used in the current study (3). Statistically significant predictors of the XQ scores were the mean doses to the major salivary glands, mean doses to the oral cavity (representing potential effect of RT on the minor salivary glands), pretherapy XQ scores, and time since RT. Notably, the salivary flow rates correlated only weakly with the degree of xerostomia. The search in the current study for potential determinants of QOL found that only patient-reported xerostomia was a consistently significant determinant throughout the post-RT study period. The doses to the parotid glands were found to be a significant factor at 3 months, but not at 6 and 12 months, after RT. The correlation between the parotid doses and the QOL scores at 3 months, found in this study, is consistent with our previous findings that the salivary gland mean doses correlated with xerostomia (3). However, this may be a chance finding, because many statistical tests were performed. No other patient-, tumor-, or therapy-related factor was found to be a statistically significant determinant of QOL. Thus, despite the correlation between the XQ and QOL scores, and the similar trend of improvement in both scores over time after RT, the predictors of each of the two seem to differ. This may have been a result of the smaller patient numbers in the current study. Another possible explanation is that the QOL scores after IMRT are dependent on additional factors that were not examined in this study. These factors may include the extent of the irradiated mucosa and other nonspecified tissue in the head and neck. The conformal dose distributions produced by IMRT may improve QOL, not only through improving xerostomia by reducing the doses to the salivary glands, but also by reducing the volume of other nontarget tissue receiving a high dose. A recently reported randomized study of pilocarpine vs. placebo concurrent with RT for head-and-neck cancer has
Table 6. QOL scores in definitive and postoperative RT groups before and 3 months after RT Definitive RT (n ⫽ 15)
Summary scores Communication Eating Pain Emotion
Postoperative RT (n ⫽ 21)
Pre-RT
3 mo
Pre-RT
3 mo
19 (8–31) 0 (0–38) 4 (0–15)*† 19 (16–38)† 29 (13–42)
24 (14–33) 19 (0–50) 29 (17–42)* 19 (13–44) 17 (4–25)
29 (10–45) 31 (0–38) 25 (17–54)† 34 (19–58)*† 13 (8–25)
21 (10–33) 16 (6–25) 29 (21–42) 19 (13–38)* 13 (8–21)
Abbreviations as in Table 5. Data presented as the median, with the 25th to 75th percentile range in parentheses; higher scores denote worse symptoms/reduced QOL. * Significant change from pre-RT to 3 mo (p ⬍ 0.05). † Significant difference between definitive and postoperative subgroups (p ⬍ 0.05).
QOL after IMRT of head-and-neck cancer
Fig. 2. (a– c) Correlations between XQ scores and QOL summary scores at 3, 6, and 12 months after RT.
demonstrated increased saliva production in patients receiving pilocarpine (33). However, no advantage of the drug could be demonstrated in the scores of the University of Washington QOL instrument used in the study (33, 34).
● A. LIN et al.
67
Several important differences exist in the outcome of using drugs to stimulate or protect the salivary production vs. the use of IMRT to spare the salivary glands. These differences predict a greater potential of IMRT to affect broad QOL measures. The first is the extent of the salivary flow sparing. Parotid-sparing IMRT preserves both the unstimulated and the stimulated salivary flows, which over time can recover significantly (2, 3), compared with modest sparing by concurrent pilocarpine of only the unstimulated flows (33). The second difference is the sparing of additional tissues achieved by the highly conformal dose distributions of IMRT, which may lead to better QOL compared with pharmacologic interventions that improve only the salivary flows. QOL instruments are useful in deciphering differences in outcome between treatment strategies that achieve similar tumor control and survival. In our study, the differences in QOL between the postoperative and definitive RT subgroups were confined to the pre-RT and early post-RT periods. Notably, the pre-RT scores in the surgery group were dominated by the sequelae of surgery performed a few weeks earlier. These results suggest that the baseline, pretherapy QOL assessment in patients receiving postoperative RT should be taken before surgery, rather than after surgery and before RT. This may pose logistical difficulties, but it would enhance the reliability of the longitudinal assessment of QOL in patients undergoing surgery and postoperative RT. Harrison et al. (28) and the University of Florida group (35) reported QOL and functional results after RT for baseof-tongue cancer that were superior to those expected after surgery. Terrell et al. (20) reported better QOL scores after chemoradiation compared with laryngectomy and postoperative RT in advanced laryngeal cancer patients. These results were related to more freedom from pain and better emotional well-being in the organ preservation group rather than to the preservation of speech function. In our study, the heterogeneity in tumor sites, operations performed, and the small patient numbers in each group precluded a reliable systematic comparison of QOL in patients receiving definitive or postoperative RT. Larger patient numbers will allow a better assessment of QOL after definitive vs. postoperative IMRT in the future. In our study, the summary QOL scores and scores of most domains improved continuously after therapy through 12 months. Other studies assessing longitudinal trends in QOL scores of head-and-neck cancer patients have reported conflicting results. Morton (36) reported little difference in most QOL domains between pretherapy and 12 months after therapy testing. Other authors reported worsening scores from therapy until about 12 months (31, 37, 38). Notwithstanding the heterogeneity in patients, therapy, and QOL instruments, a major difference between the patients in our study and the patients in the QOL studies cited above is the RT technique. After parotid-sparing IMRT in our study, the salivary flow rates and XQ scores improved throughout the posttherapy period, in parallel with an improvement in
68
I. J. Radiation Oncology
● Biology ● Physics
Volume 57, Number 1, 2003
Fig. 3. Correlations between XQ scores and each QOL domain score. (a) Eating domain, (b) Communication domain, (c) Pain domain, and (d) Emotion domain. Fitted values were linear in (a), (c), and (d) and linear and quadratic in (b).
the QOL scores. In contrast, after the standard RT techniques used in other studies of QOL in head-and-neck cancer patients, severe xerostomia was stable or even worsened over time (30, 37, 38). The differences in the extent and kinetics of recovery of xerostomia, related to the different RT techniques, may explain the differences between our study and the previous QOL studies in the longitudinal pattern and the trend over time of the QOL scores. Table 7. Model of post-RT QOL summary scores (at each point, corresponding xerostomia score used in model)
Pre-RT QOL Months since RT 3-mo XQ score 6-mo XQ score 12-mo XQ score Intercept
Coef.
SE
p
95% CI
0.19 ⫺0.30 0.26 0.21 0.21 7.59
0.09 0.55 0.08 0.08 0.09 5.50
0.025 0.589 0.002 0.006 0.015 0.167
0.02, 0.36 ⫺1.38, 0.78 0.10, 0.42 0.06, 0.36 0.04, 0.39 ⫺3.18, 18.37
Abbreviations: Coef. ⫽ coefficient; SE ⫽ standard error; CI ⫽ confidence interval; other abbreviations as in Table 5.
The statistically significant association between patientreported xerostomia and general QOL found in this study at all post-RT periods, as well as the similarity in their longitudinal recovery over time, suggests that the efforts to reduce xerostomia by IMRT may result in even broader benefits in QOL. We regard these results as hypothesis generating, rather than proving a cause– effect relationship. There may still be undetected factors that are related to both the xerostomia and the QOL scores, and other factors that may be independent of the technique of RT. The hypothesis suggested by the data in this study needs to be tested therefore, preferentially in a randomized study of IMRT vs. standard RT. Such a study, in which xerostomia and general QOL will be assessed, will start soon in Europe (40). CONCLUSION This study has demonstrated strong relationships between xerostomia and broad aspects of QOL after IMRT of headand-neck cancer. These findings strengthen the rationale for the efforts to reduce xerostomia by sparing the salivary glands, without compromising target RT.
QOL after IMRT of head-and-neck cancer
● A. LIN et al.
69
REFERENCES 1. Eisbruch A, Marsh LH, Martel MK, et al. Comprehensive irradiation of head and neck cancer using conformal multisegmental fields: Assessment of target coverage and noninvolved tissue sparing. Int J Radiat Oncol Biol Phys 1998; 41:559–568. 2. Eisbruch A, Kim HM, Ten Haken R, et al. Dose, volume and function relationships in parotid glands following conformal and intensity modulated irradiation of head and neck cancer. Int J Radiat Oncol Biol Phys 1999;45:577–587. 3. Eisbruch A, Kim HM, Terrell JE, et al. Xerostomia and its predictors following parotid-sparing irradiation of head and neck cancer. Int J Radiat Oncol Biol Phys 2001;50:695– 704. 4. Chao KSC, Deasy JO, Markman J, et al. A prospective study of salivary function sparing in patients with head and neck cancers receiving intensity-modulated or three-dimensional radiation therapy: Initial results. Int J Radiat Oncol Biol Phys 2001;51:938–946. 5. Hunt MA, Zelefsky MJ, Wolden S, et al. Treatment planning and delivery of intensity-modulated radiation therapy for primary nasopharyngeal cancer. Int J Radiat Oncol Biol Phys 2001;49:623–632. 6. De Neve W, De Gersem W, Derycke S. Clinical delivery of IMRT for relapsed or second-primary head and neck cancer using a multileaf collimator with dynamic control. Radiother Oncol 1999;50:301–314. 7. Lee N, Xia P, Quivey JM, et al. Intensity modulated radiotherapy in the treatment of nasopharyngeal carcinoma: An update of the UCSF experience. Int J Radiat Oncol Biol Phys 2002;53:12–22. 8. Wu Q, Manning M, Schmidt-Ullrich R, et al. The potential for sparing of parotids and escalation of biologically equivalent dose with intensity modulated radiation treatments of head and neck cancers: A treatment design study. Int J Radiat Oncol Biol Phys 2000;46:195–205. 9. Butler EB, Teh BS, Grant WS, et al. SMART (simultaneous modulated accelerated radiation therapy) boost: A new accelerated fractionation schedule for the treatment of head and neck cancer with intensity modulated radiotherapy. Int J Radiat Oncol Biol Phys 1999;45:21–32. 10. Maes A, Weltens C, Flamen P, et al. Preservation of parotid function with uncomplicated conformal radiotherapy. Radiother Oncol 2002;63:203–211. 11. Bjordal K, Kasa S, Matekaasa A. Quality of life in patients treated for head and neck cancer: A follow-up study 7 to 11 years after radiotherapy. Int J Radiat Oncol Biol Phys 1994; 28:847–855. 12. Ganz P, Lee J, Siau J. Quality of life assessment: An independent prognostic variable for survival in lung cancer. Cancer 1991;67:3131–3135. 13. Ruckdeschel J, Piantodosi S, for the Lung Cancer Study Group. Quality of life assessment in lung surgery for bronchogenic carcinoma. J Thorac Surg 1991;6:201–205. 14. Weymuller EA, Alsarraf R, Yueh B, et al. Analysis of performance characteristics of the University of Washington quality of life instrument and its modification. Arch Otolaryngol Head Neck Surg 2001;127:489–493. 15. Bjordal K, Kaas S. Psychometric validation of the EORTC core quality of life questionnaire, 30-item version and a diagnostic-specific module for head and neck cancer patients. Acta Oncol 1992;31:311–321. 16. Trotti A, Johnson DJ, Gwede C, et al. Development of a head and neck companion module for the quality of life-radiation therapy instrument (QOL-RTI). Int J Radiat Oncol Biol Phys 1998;42:257–261.
17. Cella DF, Bonomi AE. Measuring quality of life. Oncology (Huntingt) Suppl 1995;9:47–60. 18. List MA, D’Antonio L, Cella DF, et al. The performance status scale for head and neck cancer patients and the functional assessment of cancer therapy-head and neck cancer scale: A study of utility and validity. Cancer 1996;77:2294– 2301. 19. Terrell JE, Nanavati KA, Esclamado RM, et al. Head and neck cancer-specific quality of life: Instrument validation. Arch Otolaryngol Head Neck Surg 1997;123:1125–1132. 20. Terrell JE, Fisher SG, Wolf GT. Long-term quality of life after treatment of laryngeal cancer. Arch Otolaryngol Head Neck Surg 1998;124:964–971. 21. Terrell JE. Quality of life assessment in head and neck cancer patients. Hematol Oncol Clin North Am 1999;13:849–863. 22. Fraass BA, Kessler ML, McShan DL, et al. Optimization and clinical use of multisegment intensity-modulated radiation therapy for high-dose conformal therapy. Semin Radiat Oncol 1999;9:60–77. 23. Eisbruch A, Foote RL, O’Sullivan B, et al. IMRT for head and neck cancer: Emphasis on the selection and delineation of the targets. Semin Radiat Oncol 2002;12:238–249. 24. Henson BS, Inglehart MR, Eisbruch A, et al. Preserved salivary output and xerostomia-related quality of life in head and neck cancer patients receiving parotid-sparing radiotherapy. Oral Oncol 2001;37:84–93. 25. Ship JA, Fox PC, Baum BJ. Normal salivary gland function: How much saliva is enough? J Am Dent Assoc 1991;122:63– 69. 26. Epstein JB, Emerton S, Kolbinson DS, et al. Quality of life and oral function following radiotherapy for head and neck cancer. Head Neck 1999;21:1–11. 27. Jensen AB, Hanse O, Jorgensen K, et al. Influence of late side-effects upon daily life after radiotherapy for laryngeal and pharyngeal cancer. Acta Oncol 1994;33:487–491. 28. Harrison LB, Zelefski MJ, Pfitzer DG, et al. Detailed quality of life assessment in patients treated with primary radiotherapy for cancer of the base of tongue. Head Neck 1997;19:169– 175. 29. Sehlen S, Hollenhorst H, Lenk M. Only sociodemographic variables predict quality of life after radiotherapy in patients with head and neck cancer. Int J Radiat Oncol Biol Phys 2002;52:779–883. 30. Fang FM, Chiu HC, Kuo WR. Health-related quality of life for nasopharyngeal carcinoma patients with cancer-free survival after treatment. Int J Radiat Oncol Biol Phys 2002;53:959– 968. 31. List MA, Sison A, Haraf D, et al. Quality of life and performance in advanced head and neck cancer patients on concomitant chemoradiotherapy: A prospective examination. J Clin Oncol 1999;17:1020–1028. 32. Huguenin PU, Taussky D, Moe K, et al. Quality of life in patients cured from a carcinoma of the head and neck by radiotherapy: The importance of the target volume. Int J Radiat Oncol Biol Phys 1999;45:47–52. 33. Scarantino CW, Leveque FG, Scott CB, et al. A phase III study of concomitant oral pilocarpine to reduce hyposalivation and mucositis associated with curative radiation therapy in head and neck cancer: RTOG 9709 [Abstract]. Proc ASCO 2001;20:225a. 34. Fisher JJ, Scott CB, Scarantino CW, et al. Phase III quality of life study: Reduction in hyposalivation does not improve QOL for head and neck cancer patients post radiation therapy [Abstract]. Proc ASCO 2002;21:356a. 35. Moore GJ, Parsons JT, Mendenhall WM. Quality of life outcomes after primary radiotherapy for squamous cell carci-
70
I. J. Radiation Oncology
● Biology ● Physics
noma of the base of tongue. Int J Radiat Oncol Biol Phys 1996;36:351–354. 36. Morton RP. Laryngeal cancer: Quality of life and cost effectiveness. Head Neck 1997;19:243–250. 37. Rogers SN, Lowe D, Brown JS, et al. The University of Washington head and neck cancer measure as a predictor of outcome following primary surgery for oral cancer. Head Neck 1999;21:394–401.
Volume 57, Number 1, 2003
38. Hammerlid E, Silander E, Hornestam L. Health related quality of life three years after diagnosis of head and neck cancer—A longitudinal study. Head Neck 2001;23:113– 125. 39. Bundgaard T, Tandrup O, Elbrond O. A functional evaluation of patients treated for oral cancer. Int J Oral Maxillofac Surg 1993;22:28–34. 40. Nutting C. Personal communication, 2002.