Evolution of Cancer-Related Symptoms Over an 18-Month Period

Transcription

1 Vol. 45 No. 6 June 2013 Journal of Pain and Symptom Management 1007 Original Article Evolution of Cancer-Related Symptoms Over an 18-Month Period Claudia Trudel-Fitzgerald, BA, Josee Savard, PhD, and Hans Ivers, PhD School of Psychology, Universite Laval and Laval University Cancer Research Center, Quebec, Quebec, Canada Abstract Context. Previous studies have revealed inconsistent findings about the longitudinal evolution of cancer-related symptoms. In addition, the contribution of medical factors (e.g., cancer site, and treatments) in explaining the changes in these symptoms is yet to be established. Objectives. This prospective study investigated longitudinal changes of five symptoms (i.e., depression, anxiety, insomnia, fatigue, and pain) in patients scheduled to undergo surgery for cancer (N ¼ 828). Methods. The patients completed the Hospital Anxiety and Depression Scale, the Insomnia Severity Index, the Multidimensional Fatigue Inventory, and a pain questionnaire at baseline and after 2, 6, 10, 14 and 18 months. Results. Several time changes were statistically significant but effect sizes only revealed one change of a medium magnitude, that is, a reduction of anxiety from T1 to T2 (d ¼ 0.58). Women with breast or gynecological cancer were the only subgroups to exhibit significant changes (i.e., reduction of a small magnitude of anxiety symptoms from T1 to T2; ds ¼ 0.27 and 0.30, respectively). However, numerous differences were found across adjuvant treatments, including greater variations in depression and insomnia scores in the chemotherapy group (ds ¼ 0.71 to 0.20) and a transient increase in fatigue symptoms in patients receiving all adjuvant treatments (ds ¼ 0.24 to 0.37). Conclusion. The severity of cancer-related symptoms varies during the cancer care trajectory, especially anxiety scores, which importantly decrease during the first few months after the surgery. This study also suggests that treatment regimens better account for individual differences than cancer site in the evolution of symptoms. J Pain Symptom Manage 2013;45:1007e1018. Ó 2013 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Key Words Anxiety, cancer, depression, diagnosis, fatigue, insomnia, longitudinal, pain, prospective study, treatments Address correspondence to: Josee Savard, PhD, Laval University Research Center, 11 C^ote du Palais, Quebec, Quebec G1R 2J6, Canada. josee.savard@ psy.ulaval.ca Ó 2013 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved. Accepted for publication: June 3, /$ - see front matter

2 1008 Trudel-Fitzgerald et al. Vol. 45 No. 6 June 2013 Introduction Mostly because of improvements in screening procedures and cancer treatments, a large proportion of patients now survive many years after the diagnosis of nonmetastatic cancer. 1 However, cancer and its treatment are associated with many negative consequences. Among the most frequently reported symptoms, 2e4 anxiety, depression, insomnia, fatigue, and pain have been associated with reduced quality of life and treatment adherence. 5e8 However, longitudinal studies have revealed inconsistent findings on the evolution of these disturbances during the cancer care trajectory. Many studies have suggested that the level of depressive symptoms significantly increases during the cancer care trajectory. 9e13 Conversely, other studies have shown a significant decrease or stable scores of depressive symptoms over time. 14e16 Mixed results also have been found for anxiety symptoms, with some studies showing reduced anxiety scores, for instance, when comparing pretreatment scores to those obtained during and after treatment. 15,16 However, other studies have reported stable scores throughout the disease trajectory. 9,10,12,13 More consistent results have been observed for other cancer-related symptoms. Most longitudinal studies have revealed stable scores of sleep disturbances across the cancer care trajectory. 17e21 However, these studies often measured sleep with a limited number of items from quality-of-life questionnaires, rather than specific validated scales. Fatigue level has been found to significantly increase during chemotherapy and radiation therapy, and then decrease after their termination. 22e25 Whereas constant high levels of pain have been reported in patients with metastatic cancer, 26 in patients with non-metastatic cancer, the level of pain appears to increase during cancer treatments and decrease afterward. 19,22,27,28 Different time assessments and characteristics of the study sample (e.g., cancer sites and treatment regimens) may explain some of these inconsistencies. There is some evidence suggesting that head and neck cancer patients show higher levels of depressive, fatigue, and pain symptoms and that breast cancer patients have greater levels of depressive and insomnia symptoms, 19,25,29e32 whereas patients treated with chemotherapy and radiation therapy have been found to display distinct patterns of fatigue over time. 25,29 However, other studies reported no significant differences across cancer sites and treatment regimens. 9,26,33,34 These results on the evolution of cancerrelated symptoms need to be interpreted cautiously. Previous prospective studies were often limited by their small sample sizes (between 50 and 200 subjects for most studies), resulting in a lack of statistical power when comparing groups. As many authors have emphasized, there is a need for larger-scale longitudinal studies assessing cancer-related symptoms with several time assessments throughout the disease trajectory. 4,35e37 The goals of this longitudinal study were to describe prospectively, over an 18-month period, the evolution of anxiety, depression, insomnia, fatigue, and pain levels in a large population-based sample of cancer patients. This research also aimed at evaluating potential differences between cancer sites and treatment regimens. The hypotheses were: 1) breast cancer patients will report higher levels of symptoms than patients with cancer at other sites; and 2) more severe symptoms will be obtained in patients receiving adjuvant treatments. Methods Participants Inclusion criteria were 1) confirmed first diagnosis of nonmetastatic cancer; 2) scheduled to receive curative surgery; 3) between 18 and 80 years of age; and 4) able to read and understand French. Exclusion criteria were: 1) administration of neoadjuvant cancer treatment; 2) upcoming surgery as part of brachytherapy for prostate cancer; 3) severe cognitive impairments (e.g., Alzheimer s disease) or severe psychiatric disorder (e.g., psychosis) as noted in the medical chart, observed at recruitment, or reported by the patients; 4) having been diagnosed by a physician or being treated for a sleep disorder other than insomnia (e.g., sleep apnea) as reported by the patient; and 5) severe visual, hearing or language defects impairing the capacity to complete the measures. This study was part of a larger longitudinal study on the epidemiology of insomnia in the context of cancer. 32,38 Potential participants

3 Vol. 45 No. 6 June 2013 Evolution of Cancer-Related Symptoms 1009 were recruited at L H^otel-Dieu de Quebec (Centre hospitalier universitaire de Quebec) and H^opital du St-Sacrement (Centre hospitalier affilie universitaire de Quebec), in Quebec City, Canada. The study was approved by the research ethics committees of the Centre hospitalier universitaire de Quebec, the Centre hospitalier affilie universitaire de Quebec, and the Universite Laval. All patients meeting the initial inclusion criteria were approached by a research assistant on the day of their preoperative visit to explain the study goals and further assess their eligibility. Of the 3196 patients solicited to take part in the larger study, 1677 (52.5%) were eligible and 962 agreed to participate (57.4%). Differences between these patients and the ones who refused to participate are described elsewhere. 32 From the original sample of 962 participants, 86.1% completed at least 50% of this study s variables for at least 50% of the assessment points (N ¼ 828), which formed the sample for the current study. This decision was supported by the evidence that patients who were excluded on this basis (N ¼ 134) had a much larger amount of missing data (77.3%) on the symptom variables, compared with 6.4% only for participants included in the final sample. Patients who were excluded because of this criterion were more likely to be widowed, c 2 (1, N ¼ 134) ¼ 18.16, P < ; to receive chemotherapy, c 2 (1, N ¼ 134) ¼ 39.4, P < ; to have a gynecological, c 2 (1, N ¼ 134) ¼ 3.90, P ¼ 0.05 or a urinary and gastrointestinal (UGI) cancer, c 2 (1, N ¼ 134) ¼ 7.43, P ¼ 0.006, and less likely to be married, c 2 (1, N ¼ 134) ¼ 5.78, P ¼ 0.02; and to receive a combination of chemotherapy, radiation and hormone therapy as adjuvant treatments, c 2 (1, N ¼ 134) ¼ 5.27, P ¼ However, they were not significantly different at baseline on any of the main dependent (symptoms) variables. Procedure The study used a prospective longitudinal design comprising six time assessments: baseline (T1), 2 months (T2), 6 months (T3), 10 months (T4), 14 months (T5), and 18 months (T6). At T1, participants completed a battery of self-report scales within the following week and mailed it back. Then, a phone interview was conducted to complete missing data and review the most challenging items to fill out to increase the reliability of the data. Although patients were recruited before surgery, the majority (79.7%) completed the baseline measures after being operated on because of the small time interval between the preoperative visit and the surgery. The same procedure was used from T2 to T6, except that the questionnaires were mailed. Participants received compensation of $20 Canadian for each time assessment completed. Measures Dependent Variables Hospital Anxiety and Depression Scale. The Hospital Anxiety and Depression Scale (HADS) includes 14 items divided into two subscales: depression (HADS-D: seven items) and anxiety (HADS-A: seven items). 39 The HADS contains no somatic items, which could be confounded with symptoms of the medical condition. The four-point Likert scale ranges from 0 to 3. The total for each subscale ranges from 0 to 21, a score of seven or more suggesting the presence of clinical levels of depression or anxiety. 40,41 The French-Canadian version has psychometric qualities equivalent to those of the original English version. 41 Insomnia Severity Index. The Insomnia Severity Index (ISI) includes seven items that evaluate the perceived severity of 1) difficulties falling asleep, 2) difficulties maintaining sleep, 3) early morning awakenings, 4) the degree of dissatisfaction with current sleep, 5) the degree to which sleep difficulties interfere with daytime functioning, 6) the degree to which others notice the deterioration of functioning related to the sleep problem, and 7) the level of distress or worry caused by the sleep difficulties. 42 The five-point Likert scale ranges from 0 (not at all) to 4 (extremely), for a total score ranging from 0 to 28. A score of eight or more indicates clinical levels of insomnia, whereas a score of 15 or more suggests an insomnia syndrome. 43 The French-Canadian version of the ISI was empirically validated among cancer patients, 44,45 with psychometric properties similar to those found in the general population. 42 Multidimensional Fatigue Inventory. The French- Canadian version used is a short form of the original Multidimensional Fatigue Inventory

4 1010 Trudel-Fitzgerald et al. Vol. 45 No. 6 June 2013 (MFI). 46,47 It contains 15 items divided into four subscales but only the general and physical fatigue subscale was used for this study (MFI- GENPHY; seven items) to avoid the potential overlap between other dimensions of fatigue and depression (e.g., loss of motivation). The items are evaluated on a Likert scale ranging from 0 (not at all) to 4 (extremely). The French-Canadian version of the MFI has excellent psychometric qualities. 46 Physical Symptoms Questionnaire. The Physical Symptoms Questionnaire (PSQ), an adaptation of the Memorial Symptom Assessment Scale, 48 was developed to assess the frequency of 19 somatic symptoms often reported by cancer patients. Each item is scored on a Likert scale, ranging from 0 (never) to 4 (often). Only the pain item was used (PSQ-P) in the current study. Independent Variables Demographic and Cancer Characteristics. Participants age, education level, and marital, employment, and socioeconomic status were collected. Information regarding current and past psychological or medical conditions and medication use also was obtained with this questionnaire. Cancer-related data (e.g., cancer site and adjuvant treatments received) were taken from the patient s medical record. Statistical Analyses Raw data were entered twice by independent assistants and were cross-validated to ensure maximal integrity. The data were examined for missing data and outliers using standard procedures. 49 All analyses were completed using SAS (SAS Institute, Inc., Cary, NC), at a standard alpha level of 5%. As already explained, participants who had significant missing data (n ¼ 134) were excluded from the sample. A single imputation was then performed, using the expectation maximization algorithm, 51 which yielded a final sample of 828 participants with complete data (all five symptoms for all six assessments). Analyses of variance with repeated measures using linear mixed models were computed to investigate overall time effects for each symptom throughout the six time assessments. Afterward, a priori orthogonal contrasts were conducted to determine the presence of significant differences between pairs of consecutive time assessments (T1 vs. T2, T2 vs. T3, etc.). To evaluate time effects by cancer sites and treatment regimens, two series of mixed-model analyses of variance with a factorial design (i.e., time cancer site and time treatment regimen) were computed. Then, simple effects were used to decompose significant interactions. Cancer sites were categorized in five groups: breast, prostate, gynecological, UGI, and others. The five treatment regimens were coded: none (i.e., surgery only; no adjuvant treatment), radiation therapy only, chemotherapy only, hormone therapy only, and all (i.e., chemotherapy, radiation and hormone therapies), to study their unique effect and because every patient had his or her own sequence of treatments (note that the 223 patients receiving other combinations of treatment, therefore, were excluded from these comparisons). Potential covariates (i.e., age, gender, marital status, education, number of children, number and severity of stressful life events, use of anxiolytics and antidepressant medications, use of analgesics, past psychological difficulties, and medical comorbidity) were investigated for possible inclusion in the linear mixed models. Because none of these variables had a correlation $0.30 with the dependent variables, 52 no covariate was included. Because of the large sample size, a substantial proportion of time effects was significant. Effect sizes (Cohen s ds), which provide an index of magnitude of changes independent of the sample size, were computed to identify the largest changes, hence, those that were the most likely to be clinically significant. Effect sizes were categorized based on conventional criteria: ¼ small, 0.50 ¼ medium, and 0.80 ¼ large. Results Demographic and Medical Characteristics Table 1 presents the characteristics of the study sample at baseline (N ¼ 828). The French- Canadian participants were 56.9 years old on average and most of them were women (68.8%). The most frequent cancer sites were breast (49.4%) and prostate (28.3%), and, most commonly, patients had localized disease (i.e., Stage I or II: 73.7%). Surgery only (40.8%) and the combination of chemotherapy, radiation and

5 Vol. 45 No. 6 June 2013 Evolution of Cancer-Related Symptoms 1011 Table 1 Characteristics of Participants at Baseline (N ¼ 828) Characteristics M (SD) n (%) Age (years; range: 23e79) 56.9 (9.8) Gender (women) 567 (68.8) Marital status (n ¼ 824) Married/cohabitating 567 (68.8) Single 87 (10.6) Separated/divorced/ 170 (20.6) widowed Education (n ¼ 816) Primary school or less 54 (6.6) High school 340 (41.7) College 210 (25.7) University degree 212 (26.0) Annual family income in Canadian dollars (n ¼ 700) Less than $20, (16.0) $20,001 to $40, (32.9) $40,001 to $60, (20.6) $60,001 to $80, (15.6) $80,001 and more 105 (15.0) Current occupation (n ¼ 822) Working (full/part time) 325 (39.5) Domestic work 35 (4.3) Sick leave 116 (14.1) Retired 322 (39.2) Unemployed 24 (2.9) Time since initial diagnosis 2.2 (1.9) (months; n ¼ 810; range: 0.1e7.1) Cancer site Breast 409 (49.4) Prostate 234 (28.3) Gynecologic 89 (10.8) Urinary and gastrointestinal 52 (6.3) Other 44 (5.3) Cancer stage 0 37 (4.5) I 296 (35.8) II 314 (37.9) III 143 (17.3) IV a 20 (2.4) Unspecified 18 (2.2) Adjuvant treatments received (n ¼ 826; during the entire study) None (surgery only) 337 (40.8) Radiation therapy 73 (8.8) Chemotherapy 18 (2.2) Hormone therapy 28 (3.4) Radiation therapy and 71 (8.6) chemotherapy combined Radiation and hormone 138 (16.7) therapies combined Chemotherapy and 12 (1.5) hormone therapy combined Radiation therapy, 149 (18.0) chemotherapy and hormone therapy (all) Medical comorbidity b (n ¼ 827) Any 457 (55.3) Cholesterol 178 (21.5) Diabetes 72 (8.7) Thyroid disorders 68 (8.2) Hypertension/heart disease 246 (29.8) (Continued) Table 1 Continued Characteristics M (SD) n (%) Arthritis 97 (11.7) Epilepsy 7 (0.9) Hypoglycemia 2 (0.2) Osteoporosis 44 (5.3) Respiratory difficulties 25 (3.0) Intestinal disorders 10 (1.2) Fibromyalgia 10 (1.2) Other 17 (2.1) a All of the patients with Stage IV cancer were included because they did not have distant metastases. b Some patients did not have a medical comorbidity, whereas others had more than one comorbid condition. hormone therapies ( all group; 18.0%) were the most common treatment regimens received. The majority of chemotherapy treatments were administered between two to six weeks after T1 (median duration of 14 weeks), between seven to 23 weeks after T1 for radiation therapy (median duration of five weeks) and between 18 and 28 weeks after T1 for hormone therapy (median duration of 57 weeks). General Evolution of Symptoms Table 2 presents the mean scores obtained for each scale across the six time assessments. Despite a global time effect found for each symptom (all Ps # ), only the reduction of anxiety from T1 to T2 was of a medium magnitude (d ¼ 0.58), whereas all other changes were less than a small magnitude (ds between and 0.19). Most significant changes detected by orthogonal contrasts were observed in the first 10 months of the study (from T1 to T4; see subscripts in Table 3). Interestingly, mean HADS-A scores exceeded the clinical cutoff score (score $ 7) at T1 only, whereas depression scores never did. ISI scores were in the clinical range (score $ 8) at T1 and T2 only. Evolution of Symptoms by Cancer Sites Main Effects. Significant differences between cancer sites were found for all symptoms. Prostate cancer patients reported lower anxiety (all Ps # 0.05; Fig. 1) and fatigue (all Ps # 0.01) levels than breast, gynecological, and UGI cancer patients throughout the duration of the study. They also showed lower levels of insomnia (all Ps # 0.01) and pain (all Ps # 0.05) compared

6 1012 Trudel-Fitzgerald et al. Vol. 45 No. 6 June 2013 Table 2 Mean Scores, SE, and Time Effects Obtained for Each Symptom Time (months) Time Effects Symptoms T1 (0) T2 (2) T3 (6) T4 (10) T5 (14) T6 (18) F (5, 4135) P Anxiety (HADS-A) (SE ¼ 0.13) 7.62 a 5.50 b 5.46 b 5.16 c 5.08 c 4.90 c < Depression (HADS-D) (SE ¼ 0.10) 3.50 a 3.17 b 2.88 c 2.62 d 2.61 d 2.55 d < Insomnia (ISI) (SE ¼ 0.19) 8.86 a 8.78 a 7.69 b 7.04 c 7.04 c 6.73 c < Fatigue (MFI-GENPHY) 1.53 a 1.68 b 1.65 b 1.57 c 1.57 c 1.57 c < (SE ¼ 0.02) Pain (PSQ-P) (SE ¼ 0.05) 1.22 a 1.14 a 1.29 b 1.30 b 1.38 b 1.53 c < HADS-A ¼ Anxiety subscale of the Hospital Anxiety and Depression Scale; SE ¼ standard error; HADS-D ¼ Depression subscale of the Hospital Anxiety and Depression Scale; ISI ¼ Insomnia Severity Index; MFI-GENPHY ¼ general and physical fatigue subscale of the Multidimensional Fatigue Inventory; PSQ-P ¼ Pain item of the Physical Symptoms Questionnaire. Mean scores with different subscripts are significantly different (Ps # 0.01) according to orthogonal contrasts. with patients with breast, gynecological, and other cancer types. UGI cancer patients had higher levels of depression compared with breast and prostate cancer patients (all Ps # 0.05). Interaction Effects. Table 3 presents the results of interaction (cancer site time) and simple effects, between each time assessment, within each cancer site. A significant interaction effect was found for all symptoms. Despite several significant time effects (25 of 125; subscripts in Table 3), only two changes were at least of a small magnitude. Namely, women with breast or gynecological cancer exhibited a small reduction of anxiety symptoms from T1 to T2 (ds ¼ 0.27 and 0.30, respectively) (Fig. 1 and brackets in Table 3). Evolution of Symptoms by Adjuvant Treatments Main Effects. Over time, patients receiving no adjuvant therapy ( none group) reported significantly lower levels of anxiety and pain than patients treated with radiation therapy and those of the all group (all Ps < 0.05). Patients of the none group also showed significantly lower insomnia scores than patients of the chemotherapy and the all groups (all Ps < 0.05; left part of Fig. 2). Also, patients from the none group displayed lower levels of fatigue, compared with all other groups except the hormone therapy group (all Ps < 0.05; middle part of Fig. 2). No significant differences were found on the severity of depressive symptoms across treatment regimens. Interaction Effects. Table 4 presents the findings of interaction (treatment time) and simple effects, between each time assessment, within each treatment regimen. An interaction effect was found for all symptoms except pain. Among time effects computed, 16.8% (21 of 125) were significant (subscripts in Table 4). Calculation of effect sizes revealed numerous important changes (i.e., d $ 0.20; brackets in Table 4), including some time effects that were not statistically significant. A decline in anxiety symptoms of a medium magnitude was found in all groups between T1 and T2 (ds from 0.45 to 0.66), and of a small magnitude between T5 and T6 for hormone and radiation therapy groups (ds ¼ 0.28 and 0.22, respectively; these last two were not statistically significant). Important changes in depression and insomnia scores were mostly found in the chemotherapy group (depression: increase of d ¼ 0.20 between T2 and T3, increase of d ¼ 0.38 between T4 and T5, and decrease of d ¼ 0.49 between T5 and T6; insomnia: increase of d ¼ 0.31 between T1 and T2, decrease of d ¼ 0.71 between T2 and T3, increase of d ¼ 0.28 between T3 and T4, increase of d ¼ 0.21 between T4 and T5, and decrease of d ¼ 0.56 between T5 and T6). However, it should be noted that only the largest changes (i.e., decrease of d ¼ 0.71 between T2 and T3 and of d ¼ 0.56 between T5 and T6) were statistically significant in this group. Furthermore, patients in the radiation therapy and the all groups showed small decreases in their depressive symptoms at the beginning of the study (d ¼ 0.30 between T1 and T2, and d ¼ 0.20 between T2 and T3, respectively). In addition, patients treated with hormone therapy reported small reductions in their sleep difficulties early in the study, but they were not statistically significant (d ¼ 0.29 between T1 and T2, d ¼ 0.20

7 Vol. 45 No. 6 June 2013 Evolution of Cancer-Related Symptoms 1013 Table 3 Mean Scores Obtained for Each Symptom at Each Time Assessment by Cancer Site Breast (n ¼ 409) Prostate (n ¼ 234) Gynecologic (n ¼ 89) UGI (n ¼ 52) Other (n ¼ 44) Anxiety (M) 0 month (T1) 6.05 a 7.48 a 6.95 a 2 months (T2) 4.47 b 6.00 b 5.70 b 6 months (T3) 5.82 b 4.74 b 5.44 b 5.58 b 5.77 b 10 months (T4) 5.46 c 4.27 c 5.53 b 6.02 b 5.34 b 14 months (T5) 5.25 c 4.50 c 5.18 b 5.77 b 5.55 b 18 months (T6) 5.12 c 4.27 c 5.01 b 5.54 b 5.23 b Interaction F (20,4115) ¼ 2.90, P < Depression (M) 0 month (T1) 3.45 a 3.03 a 4.20 a 4.81 a 3.59 a 2 months (T2) 3.26 a 2.79 a 3.13 b 3.94 b 3.48 a 6 months (T3) 2.88 b 2.68 a 3.13 b 3.29 b 2.98 a 10 months (T4) 2.46 c 2.59 a 2.83 b 3.62 b 2.75 a 14 months (T5) 2.32 c 2.74 a 2.57 b 3.60 b 3.52 b 18 months (T6) 2.27 c 2.71 a 2.54 b 3.52 b 3.18 b Interaction F (20,4115) ¼ 2.83, P < Insomnia (M) 0 month (T1) a 6.53 a 9.57 a 8.23 a 8.50 a 2 months (T2) 9.68 a 7.10 a 9.06 a 8.50 a 9.05 a 6 months (T3) 8.57 b 5.85 b 7.94 b 7.98 a 8.43 a 10 months (T4) 7.76 c 5.48 b 7.65 b 7.06 a 7.34 a 14 months (T5) 7.40 c 6.09 b 6.84 b 7.31 a 8.82 b 18 months (T6) 7.20 c 5.56 b 7.01 b 6.98 a 7.64 b Interaction F (20,4115) ¼ 2.47, P ¼ Fatigue (M) 0 month (T1) 1.38 a 1.24 a 1.57 a 1.60 a 1.46 a 2 months (T2) 1.48 b 1.32 b 1.50 a 1.55 a 1.48 a 6 months (T3) 1.45 b 1.25 c 1.42 a 1.48 a 1.43 a 10 months (T4) 1.34 c 1.22 c 1.39 a 1.45 a 1.41 a 14 months (T5) 1.30 c 1.27 c 1.32 a 1.58 a 1.40 a 18 months (T6) 1.28 c 1.24 c 1.35 a 1.53 a 1.40 a Interaction F (20,4115) ¼ 2.71, P < Pain (M) 0 month (T1) 1.29 a 0.89 a 1.49 a 1.38 a 1.50 a 2 months (T2) 1.37 a 0.81 a 1.04 b 0.71 b 1.52 a 6 months (T3) 1.56 b 0.88 a 1.26 b 0.75 b 1.61 a 10 months (T4) 1.50 b 0.95 a 1.28 b 1.12 b 1.64 a 14 months (T5) 1.54 b 1.04 a 1.41 b 1.27 b 1.73 a 18 months (T6) 1.66 b 1.30 b 1.51 b 1.52 b 1.52 a Interaction F (20,4115) ¼ 1.64, P ¼ M ¼ mean scores; UGI ¼ urinary and gastrointestinal. Note: Mean scores with different subscripts are significantly different (P # 0.05) based on simple effects, between times, within each group (cancer sites). Brackets represent effect sizes: dotted brackets ¼ small changes (d ¼ 0.20). between T2 and T3). For fatigue symptoms, effect sizes only revealed a small increase from T1 to T2 (d ¼ 0.37), followed by a small decrease between T3 and T4 (d ¼ 0.24) in the all group. For pain (right part of Fig. 2), patients from the none group showed a small decrease in their symptoms between T1 and T2 (d ¼ 0.20), and patients treated with chemotherapy reported a small (although not statistically significant) increase for this same period of time (d ¼ 0.27). Patients receiving radiation therapy exhibited a small and transient augmentation in their symptoms of pain between T2 and T3 (d ¼ 0.22; not statistically significant) and between T3 and T4 (d ¼ 0.23; statistically significant), whereas patients treated by hormone therapy exhibited a small increase between T4 and T5 (d ¼ 0.29; not statistically significant). Discussion The goals of this longitudinal study were to assess, over an 18-month period, the evolution of anxiety, depression, insomnia, fatigue, and pain and investigate the differences between cancer sites and treatment regimens. Effect sizes indicated a change of a moderate magnitude only on anxiety symptoms, a reduction,

8 1014 Trudel-Fitzgerald et al. Vol. 45 No. 6 June 2013 Fig. 1. Evolution of anxiety scores by cancer site. UGI ¼ urinary and gastrointestinal. which occurred within the first two months after surgery. All other changes that were statistically significant (probably because of the large sample size and strong statistical power) were below a small magnitude, which questions their clinical significance. Nonetheless, the orthogonal contrasts showed that depression, anxiety, and insomnia symptoms generally decreased in the first 10 months after the perioperative period and then remained stable for the next eight months. All together, these results may indicate a progressive psychological adjustment to the disease. Indeed, the cancer diagnosis and surgery, which occured soon before the study baseline, are recognized as significant stressors. 5,54,55 It is important to remember that, in most patients (79.7%), the measures were completed after the surgery. Hence, because of the general reduction that was observed in the following months, it may be presumed that scores on all symptoms would have been even higher if the measures had been completed at presurgery. This study revealed a few differences in the evolution of symptoms as a function of cancer sites: only a reduction in the anxiety levels between T1 and T2 in breast and gynecological cancer patients were at least of a small magnitude according to effect sizes. Interestingly, their anxiety scores at baseline were higher than those of patients with cancer at other sites, which is consistent with our initial hypothesis postulating higher levels of symptoms in breast cancer patients. The finding that women with gynecological cancer had the same pattern of results raises a possible gender effect. However, given that most investigations have failed to find major differences between men and women on various symptoms, 11,34,56 explanations other than gender should be sought. In our sample, breast and gynecological cancer patients were younger than patients with cancer at other sites. Contextual differences (e.g., impact on life goals) associated with younger age may better explain these differences. This study also showed that prostate cancer patients exhibited lower levels of anxiety, insomnia, fatigue, and pain throughout the duration of the study, which could be explained by the fact that most of them received no adjuvant treatments (88% were treated by radical prostatectomy only) and had a better prognosis in general (i.e., 61% with a Stage II cancer). Regarding the possible influence of treatment regimens, effect sizes obtained revealed changes Fig. 2. Evolution of insomnia, fatigue, and pain scores by treatment regimens. None ¼ no adjuvant treatment, all ¼ combination of chemotherapy, radiation therapy and hormone therapy.

9 Vol. 45 No. 6 June 2013 Evolution of Cancer-Related Symptoms 1015 Table 4 Mean Scores Obtained for Each Symptom at Each Time Assessment by Adjuvant Treatments M ¼ mean scores; None ¼ no adjuvant treatment after surgery; All ¼ combination of chemotherapy, radiation therapy and hormone therapy. Mean scores with different subscripts are significantly different (P # 0.05) based on simple effects, between times, within each group (adjuvant treatments regimens). Brackets represent effect sizes: dotted brackets ¼ small changes (d ¼ 0.20); plain brackets ¼ medium changes (d ¼ 0.50); bold brackets ¼ large changes (d ¼ 0.80). of small to large magnitude in depression and insomnia symptoms in the chemotherapy group during the study. There also was an augmentation of a small magnitude of fatigue levels in the first two months of the study for the all group, which coincided with the administration of at least one adjuvant treatment in 53.6% of the patients. Furthermore, the fatigue scores decreased between the six- and 10-month evaluations in this group, which corresponded to the end of chemotherapy or radiation therapy for many patients in this sample (only 3.4% and 9.8%, respectively, were still receiving these treatments). These results are consistent with our hypothesis that physical symptom levels increase with the administration of adjuvant treatments and fade out thereafter. 22,24,25 Other results indicated that patients treated with surgery only consistently showed lower levels of all symptoms except depression. Likewise, patients who received a combination of chemotherapy, radiation and hormone therapy (i.e., the all group) displayed levels of anxiety, insomnia, fatigue, and pain symptoms that were statistically superior to those displayed by patients treated by surgery only. These differences may be explained by their combined side effects, but also by the additional emotional burden and

10 1016 Trudel-Fitzgerald et al. Vol. 45 No. 6 June 2013 anticipation associated with each new treatment being administered. 55 This study has several strengths. First, the large population-based sample size contributed to maximizing the generalization of the findings, and the use of reliable and valid questionnaires enhanced the study s internal validity. Second, prospective, repeated measures at specific points in time over an 18-month period allowed us to characterize the natural evolution of symptoms during and after treatment. Third, the investigation of symptom trajectories across cancer sites and treatment regimens allowed the identification of some possible risk factors. Nevertheless, there are limitations to this study. First, despite the large sample size, some subgroups remained underrepresented, making it impossible to study them as a separate group (e.g., head and neck cancer). Second, because the time points were fixed in time, independent of when treatments were administered for each patient, it was impossible to assess the temporal coincidence between treatment initiation/ending and symptom aggravation/ improvement. Third, because of the large sample size, numerous significant differences were obtained that may not be clinically significant. However, the calculation of effect sizes to document important changes helped to resolve that issue. Although they need to be replicated, findings of this population-based research add to those that have been found in previous studies and have several implications. From a clinical point of view, these findings suggest critical periods in the disease trajectory when patients are more vulnerable and would need more psychological support to maintain their quality of life. The first few months after surgery appear to be an important window of opportunity for offering an intervention, when clinical levels of psychological symptoms are reported by the highest proportion of cancer patients. Moreover, it would appear that specific sleep interventions should be offered more systematically at the perioperative period, given the high probability that clinical levels of insomnia will persist over time, especially during chemotherapy. Fatigue symptoms also should be closely monitored and managed as soon as adjuvant treatments are started, especially in patients receiving a combination of treatments. The results from this study also call for the development of specific psychological interventions for particular groups of patients. Women with breast and gynecological cancer experience high levels of anxiety at the perioperative period. Although a natural remission of these symptoms is likely, psychoeducational interventions such as promoting the use of active coping strategies could be useful for these women in preventing the persistence of their symptoms. In addition, a systematic screening procedure for depressive symptoms appears to be particularly relevant for UGI cancer patients. From an empirical standpoint, more research is needed on the factors explaining variations in severity levels of cancer symptoms, such as biological, psychological, and environmental factors. Additional research also is needed on symptom clusters, given that many cancer patients experience a combination of symptoms. Disclosures and Acknowledgments This research was supported by a training award held by the first author from the Canadian Institutes of Health Research, a grant from the Canadian Institutes of Health Research (MOP ) and a research scientist award from the Fonds de la recherche en sante du Quebec to the second author. The authors report no conflicts of interest. The authors sincerely thank the patients for their participation and Fred Sengmueller for revising the manuscript. References 1. Canadian Cancer Society. Canadian cancer statistics. Toronto: Canadian Cancer Society, Kim JE, Dodd MJ, Aouizerat BE, Jahan T, Miaskowski C. A review of the prevalence and impact of multiple symptoms in oncology patients. J Pain Symptom Manage 2009;37:715e Honea N, Brant J, Beck SL. Treatment-related symptom clusters. Semin Oncol Nurs 2007;23: 142e National Institutes of Health. NIH State-of-the- Science Statement on symptom management in cancer: pain, depression, and fatigue. NIH Consens State Sci Statements 2002;19:1e Ballenger JC, Davidson JR, Lecrubier Y, et al. Consensus statement on depression, anxiety, and oncology. J Clin Psychiatry 2001;62(Suppl 8):64e67.

11 Vol. 45 No. 6 June 2013 Evolution of Cancer-Related Symptoms Goodell TT, Nail LM. Operationalizing symptom distress in adults with cancer: a literature synthesis. Oncol Nurs Forum 2005;32:E42eE Reich M, Lesur A, Perdrizet-Chevallier C. Depression, quality of life and breast cancer: a review of the literature. Breast Cancer Res Treat 2008;110: 9e Wagner LI, Cella D. Fatigue and cancer: causes, prevalence and treatment approaches. Br J Cancer 2004;91:822e Butt Z, Wagner LI, Beaumont JL, et al. Longitudinal screening and management of fatigue, pain, and emotional distress associated with cancer therapy. Support Care Cancer 2008;16:151e Byar KL, Berger AM, Bakken SL, Cetak MA. Impact of adjuvant breast cancer chemotherapy on fatigue, other symptoms, and quality of life. Oncol Nurs Forum 2006;33:E18eE Chen AM, Jennelle RL, Grady V, et al. Prospective study of psychosocial distress among patients undergoing radiotherapy for head and neck cancer. Int J Radiat Oncol Biol Phys 2008;73:187e Stone P, Richards M, A Hern R, Hardy J. Fatigue in patients with cancers of the breast or prostate undergoing radical radiotherapy. J Pain Symptom Manage 2001;22:1007e Yan H, Sellick K. Symptoms, psychological distress, social support, and quality of life of Chinese patients newly diagnosed with gastrointestinal cancer. Cancer Nurs 2004;27:389e de Graeff A, de Leeuw JR, Ros WJ, et al. Longterm quality of life of patients with head and neck cancer. Laryngoscope 2000;110:98e Geinitz H, Zimmermann FB, Stoll P, et al. Fatigue, serum cytokine levels, and blood cell counts during radiotherapy of patients with breast cancer. Int J Radiat Oncol Biol Phys 2001;51: 691e Vahdaninia M, Omidvari S, Montazeri A. What do predict anxiety and depression in breast cancer patients? A follow-up study. Soc Psychiatry Psychiatr Epidemiol 2010;45:355e Ahlberg K, Ekman T, Gaston-Johansson F. The experience of fatigue, other symptoms and global quality of life during radiotherapy for uterine cancer. Int J Nurs Stud 2005;42:377e Hartl K, Engel J, Herschbach P, et al. Personality traits and psychosocial stress: quality of life over 2 years following breast cancer diagnosis and psychological impact factors. Psychooncology 2010;19: 160e Hickok JT, Morrow GR, Roscoe JA, Mustian K, Okunieff P. Occurrence, severity, and longitudinal course of twelve common symptoms in 1129 consecutive patients during radiotherapy for cancer. J Pain Symptom Manage 2005;30:433e Marchand V, Bourdin S, Charbonnel C, et al. No impairment of quality of life 18 months after high-dose intensity-modulated radiotherapy for localized prostate cancer: a prospective study. Int J Radiat Oncol Biol Phys 2010;77:1053e Palesh OG, Roscoe JA, Mustian KM, et al. Prevalence, demographics, and psychological associations of sleep disruption in patients with cancer: University of Rochester cancer Center-Community clinical oncology Program. J Clin Oncol 2010;28: 292e Barker CL, Routledge JA, Farnell DJ, Swindell R, Davidson SE. The impact of radiotherapy late effects on quality of life in gynaecological cancer patients. Br J Cancer 2009;100:1558e Dhruva A, Dodd M, Paul SM, et al. Trajectories of fatigue in patients with breast cancer before, during, and after radiation therapy. Cancer Nurs 2010; 33:201e Geinitz H, Thamm R, Scholz C, et al. Longitudinal analysis of quality of life in patients receiving conformal radiation therapy for prostate cancer. Strahlenther Onkol 2010;186:46e Schwartz AL, Nail LM, Chen S, et al. Fatigue patterns observed in patients receiving chemotherapy and radiotherapy. Cancer Invest 2000;18:11e van den Beuken-van Everdingen MH, de Rijke JM, Kessels AG, et al. Prevalence of pain in patients with cancer: a systematic review of the past 40 years. Ann Oncol 2007;18:1437e Kenefick AL. Patterns of symptom distress in older women after surgical treatment for breast cancer. Oncol Nurs Forum 2006;33:327e Klee M, Machin D. Health-related quality of life of patients with endometrial cancer who are diseasefree following external irradiation. Acta Oncol 2001;40:816e Hofman M, Ryan JL, Figueroa-Moseley CD, Jean- Pierre P, Morrow GR. Cancer-related fatigue: the scale of the problem. Oncologist 2007;12(Suppl 1): 4e Massie MJ. Prevalence of depression in patients with cancer. J Natl Cancer Inst Monogr 2004;32: 57e Potter J, Higginson IJ. Pain experienced by lung cancer patients: a review of prevalence, causes and pathophysiology. Lung Cancer 2004; 43:247e Savard J, Villa J, Ivers H, Simard S, Morin CM. Prevalence, natural course, and risk factors of insomnia comorbid with cancer over a 2-month period. J Clin Oncol 2009;27:5233e Prue G, Rankin J, Allen J, Gracey J, Cramp F. Cancer-related fatigue: a critical appraisal. Eur J Cancer 2006;42:846e Servaes P, Verhagen C, Bleijenberg G. Fatigue in cancer patients during and after treatment:

12 1018 Trudel-Fitzgerald et al. Vol. 45 No. 6 June 2013 prevalence, correlates and interventions. Eur J Cancer 2002;38:27e Carlson LE, Bultz BD. Cancer distress screening: needs, models, and methods. J Psychosom Res 2003;55:403e Fiorentino L, Ancoli-Israel S. Sleep dysfunction in patients with cancer. Curr Treat Options Neurol 2007;9:337e Zaza C, Baine N. Cancer pain and psychosocial factors: a critical review of the literature. J Pain Symptom Manage 2002;24:526e Savard J, Ivers H, Villa J, Caplette-Gingras A, Morin CM. Natural course of insomnia comorbid with cancer: an 18-month longitudinal study. J Clin Oncol 2011;29:3580e Zigmond AS, Snaith RP. The Hospital Anxiety and Depression scale. Acta Psychiatr Scand 1983; 67:361e Roth AJ, Kornblith AB, Batel-Copel L, et al. Rapid screening for psychologic distress in men with prostate carcinoma: a pilot study. Cancer 1998;82:1904e Savard J, Laberge B, Gauthier JG, Ivers H, Bergeron MG. Evaluating anxiety and depression in HIV-infected patients. J Pers Assess 1998;71: 349e Bastien CH, Vallieres A, Morin CM. Validation of the Insomnia Severity Index as an outcome measure for insomnia research. Sleep Med 2001;2: 297e Blais FC, Gendron L, Mimeault V, Morin CM. Evaluation of insomnia: validation of three questionnaires. [in French]. L Encephale 1997;23:447e Savard MH, Savard J, Simard S, Ivers H. Empirical validation of the Insomnia Severity Index in cancer patients. Psychooncology 2005;14:429e Smith S, Trinder J. Detecting insomnia: comparison of four self-report measures of sleep in a young adult population. J Sleep Res 2001;10:229e Fillion L, Gelinas C, Simard S, Savard J, Gagnon P. Validation evidence for the French Canadian adaptation of the Multidimensional Fatigue Inventory as a measure of cancer-related fatigue. Cancer Nurs 2003;26:143e Smets EM, Garssen B, Bonke B, De Haes JC. The Multidimensional Fatigue Inventory (MFI) psychometric qualities of an instrument to assess fatigue. J Psychosom Res 1995;39:315e Portenoy RK, Thaler HT, Kornblith AB, et al. The Memorial Symptom Assessment Scale: an instrument for the evaluation of symptom prevalence, characteristics and distress. Eur J Cancer 1994;30A: 1326e Tabachnik B, Fidell L. Using multivariate statistics, 5th ed. Boston, MA: Allyn & Bacon, SAS Institute. SAS/STAT 9.1 user s guide. Cary, NC: SAS Institute, Roth PL. Missing data: a conceptual review for applied psychologists. Pers Psychol 1994;47:537e Frigon J-Y, Laurencelle L. Analysis of covariance: a proposed algorithm. Educ Psychol Meas 1993;53:1e Cohen J. Statistical power analysis for the behavioral sciences, 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates, Berger AM, Sankaranarayanan J, Watanabe- Galloway S. Current methodological approaches to the study of sleep disturbances and quality of life in adults with cancer: a systematic review. Psychooncology 2007;16:401e Evans WK, Connor Gorber SK, Spence ST, Will BP. Health state descriptions for Canadians: Cancers, in n MIE Ottawa: Statistics Canada, Pirl WF. Evidence report on the occurrence, assessment, and treatment of depression in cancer patients. J Natl Cancer Inst Monogr 2004;32:32e39.