|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||
Background We sought to identify determinants of health-related quality of life after primary treatment of prostate cancer and to measure the effects of such determinants on satisfaction with the outcome of treatment in patients and their spouses or partners.
Methods We prospectively measured outcomes reported by 1201 patients and 625 spouses or partners at multiple centers before and after radical prostatectomy, brachytherapy, or external-beam radiotherapy. We evaluated factors that were associated with changes in quality of life within study groups and determined the effects on satisfaction with the treatment outcome.
Results Adjuvant hormone therapy was associated with worse outcomes across multiple quality-of-life domains among patients receiving brachytherapy or radiotherapy. Patients in the brachytherapy group reported having long-lasting urinary irritation, bowel and sexual symptoms, and transient problems with vitality or hormonal function. Adverse effects of prostatectomy on sexual function were mitigated by nerve-sparing procedures. After prostatectomy, urinary incontinence was observed, but urinary irritation and obstruction improved, particularly in patients with large prostates. No treatment-related deaths occurred; serious adverse events were rare. Treatment-related symptoms were exacerbated by obesity, a large prostate size, a high prostate-specific antigen score, and older age. Black patients reported lower satisfaction with the degree of overall treatment outcomes. Changes in quality of life were significantly associated with the degree of outcome satisfaction among patients and their spouses or partners.
Conclusions Each prostate-cancer treatment was associated with a distinct pattern of change in quality-of-life domains related to urinary, sexual, bowel, and hormonal function. These changes influenced satisfaction with treatment outcomes among patients and their spouses or partners.
There is no consensus regarding the best way to assemble the distinct domains of the quality of life of cancer survivors into a unified end point.2 Donabedian12,13 and others14,15 proposed that satisfaction with outcome is the ultimate arbiter, but most measures of satisfaction focus on cancer care, rather than outcome. We sought to characterize the quality of life after contemporary treatments for prostate cancer, to identify factors influencing this outcome, and to determine how the quality of life relates to overall satisfaction with the outcome of treatment from the perspective of the patient and his spouse or partner.
Methods
Patients
Patients were enrolled from March 2003 to March 2006 after approval by the institutional review boards at nine university-affiliated hospitals. Men with previously untreated stage T1 to T2 prostate cancer who had elected prostatectomy, brachytherapy, or external-beam radiotherapy as primary treatment and their spouses (partners) were eligible. All patients and their partners provided written informed consent. The group we studied consisted of 1201 such patients who had completed a pretreatment evaluation and 625 partners.
Data Collection and Follow-up
Patient-reported outcome measures, including the Expanded Prostate Cancer Index Composite (EPIC-26) and Service Satisfaction Scale for Cancer Care (SCA),16,17,18,19 were collected by a third-party phone-survey facility before treatment and at 2, 6, 12, and 24 months after the start of treatment. (For details regarding the interview process, see the Supplementary Appendix, available with the full text of this article at www.nejm.org.) The Service Satisfaction Scale for Cancer Care for Partners (SCA-P) and EPIC–Partner modules were administered to partners at 2, 6, 12, and 24 months after the start of the patient's treatment. At the time of this analysis, patients had been followed for a median of 30 months. Of the 1201 patients, 112 did not complete follow-up for the following reasons: 12 died (from causes unrelated to prostate cancer or its treatment), 84 withdrew from the study, and 16 missed two consecutive interviews.
Treatment
Brachytherapy was performed with the use of a transperineal technique with permanent low-dose-rate isotopes.20 Brachytherapy was administered alone in 271 patients; 35 received brachytherapy in combination with external-beam radiotherapy, androgen-suppression therapy, or both. External-beam radiotherapy was accomplished by intensity-modulated radiotherapy or highly conformal techniques7 by itself in 202 patients and with androgen-suppression therapy in 90 patients. Pretreatment interviews were conducted before the start of hormonal treatment. Prostatectomy was performed with the use of retropubic, laparoscopic, or robot-assisted techniques.21,22,23 A total of 561 patients underwent nerve-sparing procedures, and 41 underwent procedures that were not nerve-sparing; surgical-plan data were incomplete for 1 patient. There were three serious adverse events (two patients with pulmonary emboli and one with a rectourethral fistula). There were no treatment-related deaths. (For details regarding treatment and acute adverse effects, see the Supplementary Appendix.)
Statistical Analysis
We assessed differences in baseline demographic characteristics and variables related to cancer severity according to study group with the use of generalized chi-square and nonparametric Wilcoxon rank-sum tests. Scores for patient-reported outcomes, as measured by EPIC-26 and SCA, were calculated as described previously.16,17,18,19 Resulting domain scores for both instruments are on a 0 to 100 scale, with higher values representing a more favorable health-related quality of life or outcome satisfaction. To account for correlation among repeated measures from the same patient, the longitudinal study was analyzed with the use of linear generalized estimating equations (GEE), a multivariate analogue of linear regression for longitudinal data.24,25
First, to assess which quality-of-life domains were affected over time by a specific type of treatment, we evaluated the time profiles of EPIC quality-of-life scores, stratified according to study group, with the use of models containing indicators for each post-treatment time point to assess whether the average score at each time point differed significantly from that at baseline. Standard errors were calculated with the use of the GEE sandwich method to account for within-patient correlation. A clinically relevant change in the quality of life (or a minimum important change) was defined as a difference from baseline to follow-up that exceeded half a standard deviation of the baseline value.1,26
Second, to assess baseline factors that affected changes in quality of life within study groups, we evaluated the independent effects of factors with respect to patients and treatments on changes from baseline by fitting linear regression GEE models for the repeated-measures data. The baseline quality-of-life score was included in each model as a covariate. (See the Supplementary Appendix for additional details regarding model building and other analyses.)
Third, to determine whether changes in quality of life were distressing to the partner, we evaluated the association between the patient's change in quality of life and distress in his partner due to the patient's symptoms at 1 year with the use of Spearman's rank correlation. (For data on survey items that queried spousal distress, see the Supplementary Appendix.)
Fourth, to assess the relative effect of various domain changes of quality of life on overall outcome, we assessed the relationship between the patient's change in quality of life and satisfaction with treatment outcome at 1 year among patients and partners with the use of Spearman's rank correlation and linear regression models.24,25,27 Finally, to identify baseline factors that affected outcome satisfaction, we assessed the relationship between baseline covariates and satisfaction with treatment outcome with the use of linear regression models. Analyses were performed with the use of SAS software. A two-sided P value of less than 0.05 was considered to indicate statistical significance.
Results
Patients
Table 1 shows the characteristics of the patients and baseline differences among the study groups. The median age of the patients was 63 years (range, 38 to 84). The 625 enrolled partners (99% of whom were female) were, on average, younger than the patients (median age, 59; range, 23 to 89; P<0.001). Nine percent of patients and 7% of their partners were black. The number of coexisting illnesses was greatest among patients in the radiotherapy group, intermediate among those in the brachytherapy group, and least in the prostatectomy group. The severity of patients' cancer was greatest in the radiotherapy group, intermediate in the prostatectomy group, and least in the brachytherapy group. Recurrence of prostate cancer was detected in only 14 patients (1%) during follow-up.
|
|
|
|
|
Large prostate size and hormonal treatment exacerbated urinary irritation after brachytherapy or radiotherapy, whereas large prostate size at baseline was associated with improvement in urinary irritation after prostatectomy (Table 3). At 1 year, 5% of partners reported being bothered by the patient's incontinence after prostatectomy or brachytherapy, whereas 7% of partners in the brachytherapy group and 3% each in the radiotherapy group and the prostatectomy group reported being bothered by the patient's symptoms of urinary obstruction, such as urinary frequency (Table 4 and the Supplementary Appendix).
Brachytherapy and radiotherapy were both associated with a reduced quality of life related to bowel function early after treatment, and the change lasted for a year or more (Figure 1). Rectal urgency, frequency, pain, fecal incontinence, or hematochezia caused distress related to bowel function in 9% of patients a year after radiotherapy or brachytherapy (Table 2). No substantive change in bowel symptoms was detected after prostatectomy (Figure 1). The reported bowel symptoms were associated with spousal distress: 5% in the radiotherapy group and 4% in the brachytherapy group reported being bothered by the patient's bowel symptoms (Table 4).
Vitality and other outcomes related to hormonal therapy (such as fatigue, weight change, gynecomastia, depression, and hot flashes) were worse after radiotherapy or brachytherapy among patients who received hormonal treatment (90 patients in the radiotherapy group and 21 in the brachytherapy group), as compared with baseline (Figure 1 and Table 2). After radiotherapy, these symptoms persisted for up to 2 years (despite <1 year of treatment in 94% of men who received androgen-suppression therapy). Obesity, hormone therapy, and prostate size were each associated with worse quality-of-life scores related to vitality or hormonal function (Table 3). Overall, 10 to 19% of patients or their partners reported being distressed by symptoms attributable to hormonal therapy (Table 2 and Table 4).
We also examined the relationship between quality-of-life changes in patients or their partners and satisfaction with the overall outcome of treatment. Changes in each of the five quality-of-life domains, as measured by EPIC, were associated with overall outcome satisfaction among both patients and their partners in bivariate analyses (Table 5). Multivariable analysis showed that, in descending significance, symptoms related to sexual function, vitality, and urinary function were independently associated with outcome satisfaction among patients. The patient's quality of life related to sexual function was also significantly associated with satisfaction in the partner. Analysis of baseline factors in patients showed that blacks were significantly less satisfied with their overall treatment outcome than were patients of other racial backgrounds (P=0.04).
|
Each of three common primary therapies for prostate cancer showed a unique pattern of changes in quality of life related to urinary symptoms, sexual and bowel function, and vitality or hormonal function. Adjuvant hormone therapy exacerbated the adverse effects of radiotherapy or brachytherapy, whereas nerve-sparing surgical procedures mitigated the adverse effects of prostatectomy. Factors that were associated with worse patient-reported outcomes were obesity, a large prostate size, a high pretreatment PSA score, an older age, and black race. The adverse effect of obesity, which was previously linked to urinary incontinence in a single-institution study,28 also exacerbated symptoms related to vitality and hormonal function after radiotherapy or brachytherapy.
Previous multicenter studies that characterized the quality of life after primary therapy for prostate cancer include the Prostate Cancer Outcomes Study (PCOS),4 which relied on a recall of pretreatment quality of life to study the effects of prostatectomy and conventional external-beam radiotherapy, and longitudinal surveys of a limited subgroup of patients in the Cancer of the Prostate Strategic Urologic Research Endeavor (CaPSURE) study.11 Brachytherapy was not evaluated in PCOS, whereas quality-of-life analyses in CaPSURE evaluated a small subgroup of patients, did not measure symptoms related to urinary irritation or hormonal treatment, and had a suboptimal rate of survey response. Our observation that patients who received brachytherapy had symptoms related to sexual or bowel function and urinary irritation or obstruction corroborates effects that were previously noted in single-institution studies.9,10,29,30 Moreover, we observed an association between brachytherapy and both urinary incontinence and reduced vitality.
Previous studies have shown that long-term, primary hormonal therapy is associated with quality-of-life effects.31 We observed that androgen suppression of limited duration, when administered as an adjunct to radiotherapy or brachytherapy, was also associated with long-lasting symptoms involving sexuality and vitality. Notwithstanding the survival benefit of androgen deprivation in randomized clinical trials involving high-risk patients with prostate cancer,8 our findings mitigate enthusiasm for adjuvant hormonal therapy in the setting of disease of low or intermediate risk.
Urinary irritation and obstruction improved after prostatectomy. This observation validates an effect similar to that reported in one single-institution series.32 A Swedish randomized trial, comparing prostatectomy with watchful waiting, also showed fewer obstructive urinary problems among patients who underwent prostatectomy, though the patients' pretreatment values were not evaluated.33
Benefits of nerve-sparing techniques on sexual recovery after prostatectomy have been reported in several high-volume single-surgeon series.34,35,36 In our multicenter study, we also observed better recovery of sexual function among patients undergoing nerve-sparing procedures than among those who did not undergo nerve-sparing procedures, even after adjustment for pretreatment sexual function. Previous multicenter studies may not have detected benefits of nerve-sparing procedures because of differences in surgical technique, limited sample size, or retrospective design.4,9,37,38,39
Previous studies of satisfaction with prostate-cancer care have focused predominantly on treatment, rather than outcome.1,40,41,42,43 We found that changes in quality of life were significantly related to satisfaction with overall outcome among both patients and their partners. Black patients were less satisfied with the treatment outcome, despite having received care at the same high-volume institutions as did other patients.
Treatment-related changes in quality of life among patients caused distress in their partners. A previous multicenter trial of chemotherapy for recurrent prostate cancer showed that a patient's therapy had an effect on the well-being of the patient's spouse or partner.44 Our study showed that the level of spousal distress arising from a patient's sexual and urinary symptoms after primary prostate-cancer treatment was also associated with the partner's level of satisfaction with the treatment outcome. These findings confirm those of single-institution studies suggesting that patients' urinary or sexual symptoms are problematic for their partners.45,46
Limitations of our study include the lack of randomization to treatment and hence the possibility that unmeasured selection factors may have influenced the outcomes. Therefore, our analyses are appropriate only for expectations with respect to outcome within study groups and not for a comparison among treatments. The follow-up period of 2 years is relatively short, and measures of quality of life may change further with longer follow-up. Although black patients reported less favorable outcomes than did other patients (despite ostensibly similar care settings), we could not determine whether these differences in outcome reflected disparities in the quality of care, in the expectations of patients, or in cancer biology. Further study will be required to answer these questions.
In conclusion, our measurement of satisfaction with the overall outcome of cancer treatment was sensitive to divergent changes in quality of life of patients and their partners and can provide a unified assessment of the quality of care for patients with prostate cancer.
Supported by a grant (R01-CA95662, to Dr. Sanda) from the National Institutes of Health.
Dr. Sanda reports serving on advisory panels for Amgen and GTx; Dr. Michalski, receiving consulting fees from Envisioneering, serving on advisory panels for Envisioneering and Brachyscience, and providing expert opinion on radiation toxicity; Dr. Greenfield, holding a copyright for the Service Satisfaction Scale 30, from which the SCA was derived; Dr. Kibel, receiving consulting fees from Sanofi-Aventis, AstraZeneca, and Envisioneering; Dr. Kuban, serving on an advisory panel for Calypso Medical Technologies and having an equity interest in Immtech International; Dr. Wood, receiving consulting fees from Cell Genesys and lecture fees from Intuitive Surgical; Dr. Shah, receiving consulting fees from Intuitive Surgical and Gerson-Lehman Consulting; and Dr. Wei, receiving consulting fees from Sanofi-Aventis and Gen Probe, lecture fees from AMS, and grant support from Sanofi-Aventis and Envisioneering. No other potential conflict of interest relevant to this article was reported.
We thank Angela Titsworth, Valerie Stevenson, and Marsha Jessup of the Data Coordinating Center and Jill Hardy of the Phone Interview Facility for their expert project management; the many clinical-site coordinators for their exceptional professionalism in study enrollment, data collection, and regulatory support; Ruth Coburn and the many other research administrators at each site for their support of grant management and regulatory concerns; the many urologists and radiation oncologists who facilitated the identification of study candidates from among their patients; and the participating patients and their partners.
Source Information
From the Departments of Surgery (M.G.S.) and Radiation Oncology (I.K.), Beth Israel Deaconess Medical Center and Harvard Medical School, and the Department of Biostatistics (X.L.), School of Public Health, Harvard University, Boston; Biostatistics Core (R.L.D.) and the Departments of Radiation Oncology (H.M.S.) and Urology (D.W., N.S., J.T.W.), School of Medicine, and School of Nursing (L.N.), University of Michigan, Ann Arbor; the Departments of Radiation Oncology (J.M.) and Urology (A.K.), Washington University, St. Louis; the Office for Survey Research, Institute for Public Policy and Social Research, Michigan State University, East Lansing (L.H.); Alcohol Research Group, Public Health Institute, Emeryville, CA (T.K.G.); the Departments of Urology (M.S.L., C.S.S.) and Health Services (M.S.L.), UCLA Center for Health Sciences, Los Angeles; the Departments of Radiation Oncology (A.M., J.C.) and Urology (E.K.), Cleveland Clinic Hospitals, Cleveland; and the Departments of Urology (L.L.P.) and Radiation Oncology (D.K.), M.D. Anderson Cancer Center, Houston.
Address reprint requests to Dr. Sanda at the Division of Urology, Beth Israel Deaconess Medical Center, 330 Brookline Ave., Rabb 440, Boston, MA 02215, or at msanda{at}bidmc.harvard.edu.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Quality of Life and Satisfaction with Outcome among Prostate-Cancer Survivors
Yu J. S., Roach M. III, Arap W., Vordermark D., Sanda M. G., Sandler H. M., Wei J. T.
Extract |
Full Text |
PDF
N Engl J Med 2008;
359:200-202, Jul 10, 2008.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
) designate time points at which differences from baseline were significant but were below the threshold of clinical relevance. Double daggers (
) designate mean scores that showed improvement over pretreatment levels. Error bars represent 95% confidence intervals; confidence intervals were symmetric, but error bars are shown as one-sided to avoid overlap with mean scores. Panel D shows the number of patients for whom sufficient time had elapsed after treatment to enter each follow-up period (including patients who were lost to follow-up) and the corresponding number who completed their prescribed follow-up interview at each time point.