Springer 2006
Quality of Life Research (2007) 16:399–411 DOI 10.1007/s11136-006-9127-7
What contributes more strongly to predicting QOL during 1-year recovery from treatment for clinically localized prostate cancer: 4-weeks-post-treatment depressive symptoms or type of treatment? Patrick O. Monahan1,2,3, Victoria Champion2,3,4, Susan Rawl2,3,4, R. Brian Giesler5, Barbara Given3,6, Charles W. Given3,7, Debra Burns2,8, Silvia Bigatti9, Kristina M. Reuille4, Faouzi Azzouz1,2, Jingwei Wu1 & Michael Koch1,2 1 Department of Medicine, Indiana University School of Medicine, 1050 Wishard Blvd, RG 4101, Indianapolis, IN 46202-2872, USA (E-mail:
[email protected]); 2Indiana University Cancer Center, Indianapolis, IN, USA; 3Behavioral Cooperative Oncology Group, Indiana University, Indianapolis, IN, USA; 4Indiana University School of Nursing, Indianapolis, IN, USA; 5Department of Psychology, Butler University, Indianapolis, IN, USA; 6College of Nursing, Michigan State University, Lansing, MI, USA; 7College of Human Medicine, Michigan State University, Lansing, MI, USA; 8Indiana University School of Music, Indianapolis, IN, USA; 9Department of Psychology, Indiana University, Indianapolis, IN, USA Accepted in revised form 3 September 2006
Abstract Purpose: Research on prostate cancer and quality of life (QOL) has focused on the effects of treatment type on subsequent QOL, without considering effects of depressive symptoms. The present purpose is to test the independent contribution of depressive symptoms (measured within 4 weeks after treatment) and treatment type in predicting QOL measured 4, 7, and 12 months following treatment for clinically localized prostate cancer. Methods: The 105 patients (all Stage I–II) were newly treated with radical prostatectomy, external beam radiation (EBR) or brachytherapy. Age ranged from 42 to 80 (mean = 64); 88% Caucasian and 9% African American. Repeated measures mixed linear models were adjusted for age, race, education, and marital status. Results: Depressive symptoms significantly (p < 0.01) predicted 8 of 10 disease-specific and 7 of 7 generic QOL outcomes. Treatment type significantly (p < 0.01) predicted urinary function and bowel bother but no generic QOL outcomes. Conclusions: Depressive symptoms appears to predict a wider range of QOL outcomes (measured 4–12 months after treatment) than treatment type; however, when treatment is significant its effect sizes are slightly larger than depressive symptoms. Health care providers should (1) assess depressive symptoms in prostate cancer patients before and after treatment, and (2) provide psychosocial (e.g., counseling, support groups) and pharmacologic treatment options for improving depressive symptoms. Key words: Depression, Depressive symptoms, Prostate cancer, Quality of life, Treatment
Introduction Approximately 234,460 men will be diagnosed with clinically localized prostate cancer in the Unites States in 2006 [1]. The most common
treatment for prostate cancer is radical prostatectomy which involves removal of the prostate gland and surrounding tissue. External beam radiotherapy (EBR) and brachytherapy (seed implant) are popular forms of treatment options using radiation.
400 Watchful waiting is another option. However, considerable controversy surrounds treatment of clinically localized prostate cancer. There is little evidence one treatment surpasses the other in terms of recurrence or mortality at 15-year followup [2]. The lack of strong evidence for either radiation or surgery over watchful waiting has led many clinicians and researchers to conclude that impact on quality of life (QOL) should play an important role in treatment decisions and assessment of treatment success [2]. QOL is especially important for prostate cancer patients because over the past 20 years the five-year survival rate, for all stages at diagnosis combined, has increased from 67% to 99%, and the relative 10-year survival is 92% [1]. Generic QOL is generally measured in four areas: physical, functional, emotional, and social [3]. It is also important to assess QOL specific to the disease [4].
greater, and both groups demonstrate relatively rapid recovery of bowel function at 1-year posttreatment [2]. Sexual dysfunction is a well known side effect of both prostate cancer and its treatments; sexual function is poor before and after treatment, but is exacerbated by certain treatments; specifically, surgery patients show severe declines shortly after treatment and these declines remain worse than the declines for EBR patients throughout the first year of follow-up, although nerve-sparing surgery reduces the declines [2]. In a prospective study, 96% and 93% of surgery patients (less than half had bilateral nerve sparing surgery) exhibited erectile dysfunction at 3 and 12 months posttreatment, respectively, which was well above their 32% baseline rate; the increase in erectile dysfunction was not as great for EBR patients who showed rates of 58% and 67% at 3 and 12 months compared to 45% at baseline [5].
Disease-specific QOL
Depressive symptoms
The most important disease-specific threats to QOL in this population are side effects of treatment, chief of which are urinary, bowel, and sexual dysfunction [2]. Incontinence and other urinary symptoms vary by treatment type and time since diagnosis. In a prospective study, incontinence after prostatectomy was observed in 24–58% of patients 3 months after surgery and 11–35% of patients 12 months after surgery [5]. Incontinence rates for patients recovering from EBR are considerably lower [2], for example, ranging from 2% to 5% over a 12 month follow-up [5]. Of 600 patients receiving brachytherapy, at 60 days and 6 months post treatment, respectively, 78% and 34% of patients exhibited urinary symptoms such as mild to moderate nocturia, dysuria or bladder spasms [6]. Bowel problems are more prominent in patients who undergo EBR compared to other treatments [2]. For example, in a prospective study, 16% and 12% of EBR patients experienced diarrhea (at least several times per week) at 3 and 12 months post-treatment, respectively, compared to 5% and 6% of surgery patients [5]. Although prostate cancer patients undergoing EBR and surgery show declines in bowel function at 4 months post treatment, the declines for EBR patients are
Given the significant disease-specific side effects of treatment, most QOL research in this population has focused on the impact of treatment. However, researchers are beginning to focus upon other factors that place prostate cancer patients at risk for QOL deficits. Several studies demonstrated links between depressive symptoms and QOL in prostate cancer patients. When 88 men were assessed 12–24 months after treatment for radical prostatectomy, depressive symptoms significantly predicted overall QOL (r = ) 0.58) [7]. Psychiatric co-morbidities including past or current depression was associated with worse generic QOL in men with localized prostate cancer [8]. Among prostate cancer patients treated with radical prostatectomy, greater depressive symptoms were associated with incontinence and worse sexual function [9]. For 1138 men with localized or advanced prostate cancer, those with comorbid depression or anxiety (but not dementia) who were on regular psychiatric medication (n = 59) showed worse functioning in physical, role, cognitive, social, and global health, and worse fatigue, pain, constipation, and sexuality than those with comorbid heart and circulatory disease (n = 498) [10]. Recent estimates showed that 15% of 92 ambulatory
401 prostate cancer patients exceeded the cutoff for probable depression [11], and 13% of 45 men with prostate cancer receiving androgen deprivation therapy were diagnosed with DSM-IV major depressive disorder [12]. Links between depressive symptoms and QOL have also been found in the general cancer population. Distress over side-effects, threat of shortened lifespan, disruption of daily routine and social relationships, and other consequences of being diagnosed with and treated for cancer can significantly increase depressive symptoms [13]. A review paper concluded that 19–43% of cancer patients experienced high levels of depressive symptoms, as indicated by cutoff scores on depression screening tools, and 6–15% of cancer patients met strict criteria for a clinical diagnosis of major depression [14]. In longitudinal studies of cancer patients, depression has predicted greater morbidity and hospital stays, and less medical adherence [12], and even mortality before adjusting for age [15]. Depression may also affect treatment efficacy [16]. Depression in cancer patients has been under-recognized and under-referred to psychiatry [12]. Among the highest unmet needs identified by cancer patients continues to be psychological needs related to emotions and coping [17, 18]. Purpose To the best of our knowledge, no study has determined the independent effects of prostate cancer treatment and depressive symptoms on predicting subsequent QOL. This issue is particularly critical for identifying patients who are at risk for QOL deficits. The purpose of the current study is to determine to what extent depressive symptoms and type of treatment independently predict subsequent disease-specific and generic QOL in newly diagnosed prostate cancer patients who were followed for 12 months after the conclusion of treatment for clinically localized disease. Our hypothesis is: Both 4-weeks-post-treatment depressive symptoms and treatment type will significantly predict diseasespecific and generic QOL measured 4, 7, and 12 months post treatment, after adjusting for other covariates.
Methods Overview Data were collected as part of a multi-site prospective, randomized clinical trial of a tailored nurse-driven intervention designed to provide education and support for improving QOL for patient/spouse dyads during the first year following treatment for clinically localized prostate cancer. All institutional review boards approved the project and the informed consent. Participants were randomly assigned to either standard care or to the intervention. In terms of medical treatment, patients were advised as to their medically reasonable options for management which generally depended on age, stage of cancer and medical conditions [1]. Patients then chose a treatment approach in consultation and concurrence with the treating physician. For clinically localized prostate cancer (Stages I and II), although surgery, EBR and brachytherapy are all potential options [1], life expectancy (i.e., age) was a major factor in treatment preference (i.e., our younger patients were more likely to elect surgery). Within 30 days of completing cancer treatment, participants in the intervention arm began sessions with a nurse specialist who provided tailored counseling on cancer and treatment-related symptoms. Outcome data were collected through telephone-assisted interviews. Data are from interviews that occurred at initial assessment (i.e., within 4 weeks after the end of cancer treatment), and at 4, 7 and 12 months following treatment. Sample Eligible patients were (1) newly treated with radical prostatectomy, EBR or brachytherapy, for prostate cancer confined to the gland (Stage I–II), (2) 18 years or older, (3) able to speak English, and (4) with a partner willing to participate. Of 188 eligible dyads, 122 dyads consented, yielding an accrual rate of 65%. Six dyads consented but withdrew prior to initial assessment. Eleven patients had no follow-up data. The present analysis uses data from the 105 patients who had both an initial assessment and at least one follow-up assessment. Patients were accrued from Indiana University Medical Center in Indianapolis, Michigan State University
402 affiliated agencies in East Lansing, the West Michigan Cancer Center in Kalamazoo, and the Jewish and Veteran’s Administration Hospitals associated with the University of Louisville. Age (at time of cancer treatment) ranged from 42 to 79 for the 105 male patients, with a mean of 64.3 (Table 1). The CES-D depressive symptoms score ranged from 0 to 33 with a mean of 7.1 (Table 1). Of the 105 patients, 14% met the CES-D threshold (‡ 16) for clinical depression, and 16% had a score of 1 which was the most frequently occurring score (mode); 9% scored 0 and 10% scored 2. Most patients (88%) were Caucasian, 95% were married and 57% completed at least some college (Table 1). For treatment, 55% of patients elected to have surgery, 27% opted for EBR, and 18% chose brachytherapy. Other treatment and disease information are shown (Table 1). Regarding health insurance, 100% of patients had private insurance, 15% participated in an HMO, 44% had Medicaid, 4% had Medicare, and 21% had VA medical insurance. Instruments Disease-specific QOL The Prostate Cancer Quality of Life Instrument (PC-QoL) is a psychometrically valid and reliable 52-item, multi-scale instrument targeted at men treated for localized disease [19]. The PC-QoL assesses three domains: physiologic dysfunction, limitations in role activities caused by physiologic dysfunction, and bother (i.e., psychological distress) caused by physiologic dysfunction. Each domain is assessed within each of the three organ systems most likely to be affected by treatment (urinary, sexual and bowel), yielding nine scales. A 10th scale assesses anxiety or worry over spread or recurrence of cancer. Among patients treated for clinically localized prostate cancer, all PC-QoL scales demonstrated good test–retest reliability, internal consistency (Cronbach’s a ranged from 0.70 to 0.90, with most being 0.85 or higher), and discriminant/convergent and criterion validity; the scales also differentiated surgical, EBR, and watchful waiting patients [19]. Generic QOL The Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) was used to assess generic
Table 1. Sample characteristics Total number
105
Age at treatment (years) Mean (SD)
64.3 (8.1)
CES-D (4-week post treatment) Mean (SD) Minimum, maximum First quartile, third quartile
7.1 (7.3) 0, 33 2, 10
Race/Ethnicity Caucasion African American Other Unknown
no. (column %) 92 (88%) 9 (9%) 2 (2%) 2 (2%)
Education
