PII S (01) CLINICAL INVESTIGATION

Transcription

1 PII S (01) Int. J. Radiation Oncology Biol. Phys., Vol. 51, No. 3, pp , 2001 Copyright 2001 Elsevier Science Inc. Printed in the USA. All rights reserved /01/$ see front matter CLINICAL INVESTIGATION Prostate A PROSPECTIVE QUALITY-OF-LIFE STUDY IN MEN WITH CLINICALLY LOCALIZED PROSTATE CARCINOMA TREATED WITH RADICAL PROSTATECTOMY, EXTERNAL BEAM RADIOTHERAPY, OR INTERSTITIAL BRACHYTHERAPY W. ROBERT LEE, M.D., M.S., M. CRAIG HALL, M.D., RICHARD P. MCQUELLON, PH.D., L. DOUGLAS CASE, PH.D., AND DAVID L. MCCULLOUGH, M.D. Comprehensive Cancer Center of Wake Forest University School of Medicine, Winston-Salem, NC Purpose: To prospectively assess the health-related quality of life (HRQOL) and changes in HRQOL during the first year after 3 different treatments for clinically localized prostate cancer. Methods and Materials: Ninety men with T1 T2 adenocarcinoma of the prostate were treated with curative intent between May 1998 and June 1999 and completed a quality-of-life Functional Assessment of Cancer Therapy-Prostate (FACT-P) questionnaire before treatment (T0) and 1 month (T1), 3 months (T3), and 12 months (T12) after treatment. Forty-four men were treated with permanent source interstitial brachytherapy (IB), 23 received external beam radiotherapy (EBRT), and 23 men were treated with radical prostatectomy (RP). The mean age of the entire study population was 65.9 years (median 67, range 42 79). The mean pretreatment prostate-specific antigen level of the entire study population was 6.81 ng/ml (median 6.25, range ). The Gleason score was <6 in 65 (72%) of 90. The repeated measures analysis of variance and analysis of covariance were conducted on all quality-of-life and urinary outcome measures. Results: A comparison of the demographic characteristics of the 3 treatment groups demonstrated significant differences. The men treated with RP were significantly younger than the men in either the IB or EBRT group (median age 61.0 RP, 67.1 IB, 68.8 EBRT; p ). The men in the IB group were more likely to have a Gleason score of <6 than the EBRT group (Gleason score <6, 86% IB and 48% EBRT; p 0.015). The mean score (standard deviation) at T0, T1, T3, and T12 for the FACT-P questionnaire for each group was as follows: IB (17.0), (21.7), (18.4), and (14.2); EBRT (12.1), (21.0), (19.2), and (15.6); and RP (14.7), (18.3), (17.8), and (14.9), respectively. Statistically significant differences over time were observed for the FACT-P in the IB and RP groups (p <0.0001), but not for the EBRT group (p 0.08). The examination of the subscales within the FACT-P instrument demonstrated statistically significant changes over time in the IB and RP groups for the following: physical well-being, functional well-being, and prostate cancer symptoms. After adjusting for age, race, T stage, Gleason score, use of hormonal therapy, and baseline FACT-P scores, statistically significant differences in the FACT-P score at T1 according to treatment group were observed. At T12, the FACT-P scores were not significantly different than the baseline FACT-P scores for any group. Conclusions: The results of this analysis suggest that significant decreases in HRQOL, as measured by the FACT-P instrument, are evident in the first month after IB or RP, but not after EBRT. One year after treatment, however, the FACT-P scores were not statistically different from the baseline measures for any group. For all treatment groups, most of the HRQOL decreases were observed in the physical, functional, and prostate cancer-specific domains. These results suggest that the HRQOL changes are likely to be treatment-specific, further emphasizing the importance of a randomized trial comparing the different treatment options in this population of men Elsevier Science Inc. Prostate brachytherapy, Quality of life, FACT-P, Radical prostatectomy, External beam radiotherapy. INTRODUCTION In patients diagnosed with cancer, the length of disease-free or overall survival has traditionally measured the success of any particular treatment. In the past decade, the health care community, prompted by the medical outcomes movement, Reprint requests to: W. Robert Lee, M.D., Department of Radiation Oncology, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC Tel: has placed more emphasis on patient-reported end points such as health-related quality of life (HRQOL) (1). An individual s HRQOL has been defined as an appraisal of and satisfaction with their current level of function as compared to what they perceive to be possible or ideal. (2) Measuring the HRQOL in men with clinically localized ; Fax: ; wrlee@wfubmc.edu Received Feb 20, 2001, and in revised form May 5, Accepted for publication June 4,

2 Quality of life after treatment for prostate cancer W. R. LEE et al. 615 Table 1. Patient demographics and clinical characteristics (n 90) IB EBRT RP Total participants (n) Median age (y) 67.1 (49 79) 68.8 (51 79) 61.0 (42 68) Stage (%) T1 26/44 (59) 12/23 (52) 19/23 (83) T2 18/44 (41) 11/23 (48) 4/23 (17) Gleason score (%) /44 (86) 11/23 (48) 16/23 (70) 7 6/44 (14) 10/23 (43) 5/23 (22) 7 0 2/23 (9) 2/23 (9) Median PSA (m/ml) 6.5 ( ) 8.1 ( ) 6.2 (1.3 12) NAS (%) 11/44 (25) 2/23 (9) 3/23 (13) Abbreviations: IB interstitial brachytherapy; EBRT external beam radiotherapy; RP radical prostatectomy; PSA prostatespecific antigen; NAS neoadjuvant androgen suppression. Numbers in parentheses are the ranges unless otherwise noted. prostate cancer has recently captured the attention of physicians and patients alike. This increased interest is understandable for several reasons. First, adenocarcinoma of the prostate is the most commonly diagnosed solid tumor in American men (3). Second, numerous treatment options, including expectant management, radical prostatectomy (RP), external beam radiotherapy (EBRT), interstitial brachytherapy (IB), and cryotherapy, are available to these men, and no consensus has been reached as to the optimal therapy (4). Third, it is clear that RP, EBRT, and IB can result in sexual, urinary, and bowel dysfunction. Finally, any morbidity related to treatment may persist for several years given the natural history of most prostate cancers. It should not be surprising then that numerous reports have been published in the past decade examining the HRQOL of men with clinically localized prostate cancer (5 22). Most of these reports have been retrospective or cross-sectional, a serious methodologic weakness when considering the effects of treatment on HRQOL. Most men diagnosed with prostate cancer will be 65 years of age, and many men of this age without prostate cancer will have sexual, urinary, and bowel dysfunction that can be measured by sensitive instruments (23). This observation underscores the importance of longitudinal or prospective study designs. Indeed, repeated measurements of HRQOL allow each patient to act as his own control, markedly reducing the sample sizes required to detect meaningful changes in HRQOL. We analyzed a cohort of men with clinically localized prostate cancer treated by a small group of physicians at Wake Forest University during the past 2 years. These men were recruited to participate in a longitudinal, observational study of HRQOL and patient-reported symptoms after treatment with RP, EBRT, or IB. We elected to use the Functional Assessment of Cancer Therapy-Prostate (FACT-P) questionnaire and the International Prostate Symptom Score (IPSS) questionnaire (10, 24, 25). These instruments were chosen because they are robust validated measures, and our previous experience demonstrated that it was possible to integrate these questionnaires into a busy clinical practice (14, 15). METHODS AND MATERIALS Patients Between May 1998 and June 1999, a total of 98 men with localized prostate cancer received treatment with IB, EBRT, or RP at Wake Forest University School of Medicine. All 98 men were asked to participate in this study, and 90 (91%) agreed to enroll and completed the FACT-P (Appendix) and IPSS questionnaires before treatment and 1, 3, and 12 months after treatment. The demographic characteristics of all men according to treatment group are presented in Table 1. All patients were staged according to a directed history and physical examination. The 1992 American Joint Committee on Cancer clinical staging system was used (26). All patients underwent serum prostate-specific antigen (PSA) level determination (Hybritech, San Diego, CA) before prostate biopsy. Quality-of-life questionnaires The preliminary development of the FACT-G and FACT-P instruments has been described previously (10, 25). In brief, the FACT-G is a 29-item self-report questionnaire that measures the quality of life of cancer patients. The FACT-G consists of 5 subscales measuring physical (PWB), functional (FWB), social/family, and emotional (EWB) well-being and satisfaction with the doctor/patient relationship (RWD). The questionnaire can be self-administered or used in an interview format. The prostate cancer subscale (PCS) of the FACT includes 12 items specifically designed to measure the quality of life in prostate cancer patients. The PCS includes questions about pain, urinary symptoms, and bowel and sexual function. The FACT-P is scored by summing the 5 subscales of the FACT-G plus the PCS, yielding a composite quality-of-life score. The Trial Outcome Index (TOI) is the sum of the PWB, FWB, and PCS and is thought to be the most focused and sensitive

3 616 I. J. Radiation Oncology Biology Physics Volume 51, Number 3, 2001 indicator of the physical aspects of HRQOL (10). All men in this analysis completed FACT-P version 3.0 and were instructed to rate themselves on how well they felt that day and during the past 7 days. Higher scores on the overall measure and subscales indicate a better quality of life. The IPSS is an instrument designed to measure lower urinary tract symptoms (LUTS) resulting from benign prostatic hyperplasia in men. It was initially developed in the United States and was recently adopted by the World Health Organization (24). Higher scores indicate more LUTS. It is important to point out that the IPSS questionnaire does not contain questions assessing continence or dysuria. All men completed the questionnaires in the Department of Radiation Oncology or the Department of Urology within the Comprehensive Cancer Center at Wake Forest University. The questionnaires were self-administered after the men gave informed consent. All questionnaires were in English. The procedure for completing the questionnaires was described by a single clinician familiar with the questionnaires. Other health care professionals checked the completed questionnaires and asked the participants to complete any items that had been omitted. All men completed the questionnaires time points. No patients were lost to follow-up. The study received Institutional Review Board approval before activation. Treatment In the IB group (n 44), all men were treated with brachytherapy alone, no EBRT was given. Eleven men (25%) received androgen deprivation therapy to reduce the size of the prostate gland before IB. The specifics of the IB technique have been previously described (27, 28). Pretreatment planning using ultrasound images was routinely performed using commercially available treatment planning software. All men received IB with 125 I alone. The prescription dose was 144 Gy according to the TG T43 formalism (29). All men in the IB group were treated by the same 2 physicians (W.R.L., D.L.M.). In the EBRT group (n 23), all men were treated with EBRT alone, no brachytherapy was given. All men were treated with 10 MV photons. The median dose was 70.2 Gy (range ) prescribed to the 95% volume. CT was performed on all patients to assist in the treatment planning process. The four-field technique (AP:PA:right:left) and Cerrobend blocking were routinely used. The Cerrobend blocking was shaped to conform the radiation dose to the shape of the prostate gland (conformal radiotherapy). The pelvic lymph nodes were not irradiated. All men treated with EBRT were treated by the same physician (W.R.L.). In the RP group (n 23), all men underwent radical RP. A nerve-sparing technique was performed at the discretion of the operating surgeon. Pelvic lymph node dissection was routinely performed. Two different urologists contributed patients to this study (M.C.H., D.L.M.). Statistical analysis On the basis of previously reported information, we estimated that the mean and standard deviation of the FACT-P scores in this population of men would be approximately This study was designed to detect changes in HRQOL over time. With 23 subjects per group, we could detect differences in the HRQOL of 15 units (an 11.5% relative difference), with 90% power at the 5% two-sided level of significance, assuming no correlation between the before and after measures. Univariate repeated measures analyses of variance were conducted on all outcome measures. Pairwise comparisons within a treatment group were made using the paired t test, and Bonferroni s method was used to adjust for multiple comparisons. Analysis of covariance was used to assess treatment group differences in posttreatment HRQOL after adjusting for baseline values and patient demographics. The demographic characteristics between the treatment groups were compared by analysis of variance. All p values reported are two-sided. RESULTS Demographics of treatment groups Table 1 outlines the demographic characteristics of the study population according to treatment group. The mean age of all 90 men was 65.9 years (median 67; range 42 79). The mean pretreatment PSA level for all men was 6.81 ng/ml (median 6.25; range ). Nearly two-thirds of men had no palpable evidence of prostate cancer on digital rectal examination (57 of 90; 63.3%), and most had a Gleason score of 6 (65 of 90; 72.2%). A comparison of the men in each of the 3 treatment groups demonstrated significant differences among the groups at baseline. The men treated with RP were significantly younger than the men in either the IB or EBRT group (median age 61.0 RP, 67.1 IB, 68.8 EBRT; p ). The men in the IB group were more likely to have a Gleason score of 6 compared with the EBRT group (Gleason score 6, 86% IB and 48% EBRT; p 0.015). The pretreatment PSA level was slightly higher in the EBRT group than in the RP or IB group, although this was not statistically significant (median PSA 8.1 ng/ml EBRT, 6.5 ng/ml IB, and 6.2 ng/ml RP; p ). Interstitial brachytherapy The mean and standard deviation for the total FACT-P, total FACT-G, IPSS, all the FACT-P subscales, and the TOI score in all men treated with IB are presented in Table 2. The overall p values reported in Tables 2, 3, and 4 represent the test for any differences across all time points. The p values for pairwise comparisons are also reported in Tables 2 4, assessing the change between each of three posttreatment points relative to the baseline. In Tables 2 4, 30 pairwise comparisons were made, and the level of (0.05 of 30) was chosen to represent the differences that were statistically significant among the 3 time points. The composite FACT-P score decreased significantly 1

4 Quality of life after treatment for prostate cancer W. R. LEE et al. 617 Table 2. Repeated measures analysis of variance of all quality-of-life measures for men treated with IB (n 44) p* HRQOL measure T0 T1 T3 T12 Overall T0 vs. T1 T0 vs. T3 T0 vs. T12 FACT-P (17.0) (21.7) (18.4) (14.2) FACT-G (11.8) 92.5 (14.7) 97.3 (13.0) (9.1) PWB 25.9 (2.8) 21.6 (4.7) 23.8 (4.2) 25.3 (2.6) FWB 23.3 (4.4) 18.9 (6.4) 21.4 (4.6) 24.1 (3.7) EWB 20.8 (3.0) 21.1 (2.9) 21.7 (3.2) 22.3 (2.1) SFWB 23.0 (4.6) 23.1 (4.3) 22.8 (3.9) 22.7 (3.7) RWD 7.7 (0.8) 7.7 (1.0) 7.6 (0.9) 7.8 (0.7) PCS 37.7 (6.6) 28.0 (7.9) 32.6 (7.2) 36.3 (6.1) TOI 86.9 (12.3) 68.6 (17.4) 77.8 (13.9) 85.8 (10.8) IPSS 8.3 (6.0) 20.8 (7.7) 17.2 (7.1) 10.4 (7.3) Abbreviations: HRQOL health-related quality of life; T0, T1, T3, T12 before and 1, 3, and 12 months after treatment; FACT-P Functional Assessment of Cancer Therapy-Prostate; FACT-G FACT-general; PWB physical well-being; FWB functional WB; EWB emotional WB; SFWB social/family WB; RWD doctor/patient relationship; PCS prostate cancer subscale; TOI Trial Outcome Index; IPSS International Prostate Symptom Score. Data presented as the mean, with the standard deviation in parentheses, unless otherwise noted. * Thirty pairwise comparisons were made, and the level of (0.05/30) was chosen to represent statistically significant differences. Statistically significant. month after IB. Close examination of the results listed in Table 2 demonstrates that most of this early decrease occurred in three subscales or domains: the PWB, FWB, and PCS subscales. The TOI combines the three subscales that were most affected, and it is no surprise that this combination of subscales demonstrated a dramatic change in the first month. At 3 months after treatment with IB, the PCS and TOI scores remained less than the baseline levels. The social/family well-being and RWD subscale scores changed very little during the first year, but the EWB subscale score actually increased, although the increase was of only borderline significance. Comparing the HRQOL scores 12 months after IB with the baseline measures revealed th scores, with the exception of the EWB subscale, returned to baseline. The IPSS score is a measure of LUTS, and it is important to note that the IPSS more than doubled in the first month after IB. This increase to an IPSS score of 20 indicates moderate to severe LUTS. Urinary symptoms, as measured by the IPSS, persisted for at least 3 months, with the IPSS score at 3 months more than double the score at baseline. At 12 months, however, the difference in the IPSS score from the baseline score was statistically significant. External beam RT The mean and standard deviation for the total FACT-P, total FACT-G, IPSS, all the FACT-P subscales, and the TOI score in all men treated with EBRT are presented in Table 3. The FACT-P score declined between T0 and T1, but this decrease was not statistically significant. The scores of the PWB, PCS, and TOI subscales decreased but not significantly. Examining the pairwise comparisons demonstrated Table 3. Repeated measures analysis of variance of all quality-of-life measures for men treated with EBRT (n 23) p* HRQOL measure T0 T1 T3 T12 Overall T0 vs. T1 T0 vs. T3 T0 vs. T12 FACT-P (12.1) (21.0) (19.2) (15.6) FACT-G 99.9 (7.9) 96.1 (12.5) 98.1 (13.2) (10.1) PWB 25.2 (2.2) 22.6 (4.7) 23.8 (3.7) 25.1 (4.1) FWB 22.9 (3.2) 21.7 (5.0) 22.2 (5.2) 23.2 (4.8) EWB 20.7 (3.3) 21.5 (3.4) 20.6 (4.2) 21.9 (2.4) SFWB 23.6 (2.8) 22.8 (3.2) 23.8 (2.4) 23.1 (3.1) RWD 7.5 (0.8) 7.5 (0.7) 7.8 (0.5) 7.7 (0.5) PCS 37.2 (6.0) 33.3 (10.0) 36.3 (7.0) 35.8 (6.3) TOI 85.3 (9.1) 77.6 (18.1) 82.2 (14.5) 84.1 (13.7) IPSS 11.9 (6.4) 13.8 (7.5) 9.3 (5.3) 8.5 (5.4) Abbreviations as in Tables 1 and 2. Data presented as the mean, with the standard deviation in parentheses. * Thirty pairwise comparisons were made, and the level of (0.05/30) was chosen to represent statistically significant differences; no p values were statistically significant.

5 618 I. J. Radiation Oncology Biology Physics Volume 51, Number 3, 2001 Table 4. Repeated measures analysis of variance of all quality-of-life measures for men treated with RP (n 23) p* HRQOL measure T0 T1 T3 T12 Overall T0 vs. T1 T0 vs. T3 T0 vs. T12 FACT-P (14.7) (18.3) (17.8) (14.9) FACT-G 99.8 (9.6) 88.9 (13.2) 98.5 (12.0) (11.3) PWB 26.3 (2.3) 20.9 (5.0) 25.4 (3.6) 26.3 (2.5) FWB 23.6 (4.3) 16.5 (5.1) 21.9 (4.6) 23.3 (4.1) EWB 17.9 (4.4) 19.9 (3.8) 21.0 (3.4) 21.7 (3.3) SFWB 24.4 (3.2) 23.7 (4.5) 22.7 (4.1) 22.8 (4.5) RWD 7.7 (0.6) 7.9 (0.3) 7.5 (1.1) 7.7 (0.5) PCS 38.5 (7.6) 28.8 (7.2) 35.9 (6.8) 38.6 (4.6) TOI 88.3 (12.3) 66.2 (13.3) 83.3 (13.3) 88.2 (10.1) IPSS 12.5 (9.3) 17.2 (10.3) 7.4 (4.3) 5.5 (3.7) Abbreviations as in Tables 1 and 2. Data presented as the mean with the standard deviation in parentheses, unless otherwise noted. * Thirty pairwise comparisons were made, and the level of (0.05/30) was chosen to represent statistically significant differences. Statistically significant th HRQOL scores were not significantly different from baseline by 12 months after EBRT. Urinary symptoms, as measured by the IPSS, changed very little 1 month after the completion of EBRT. In fact, 3 months after EBRT, the IPSS score was significantly lower than the baseline IPSS score, indicating an improvement in LUTS. Radical prostatectomy The mean and standard deviation for the total FACT-P, total FACT-G, IPSS, all the FACT-P subscales, and the TOI score in all men treated with RP are presented in Table 4. The FACT-P score decreased significantly in the first month. Similar to the results seen in the IB group, the scores in the PWB, FWB, and PCS subscales and TOI all decreased. The social/family well-being and RWD subscale scores remained stable in the first month, but the EWB subscale score actually increased at T3 compared with T0, indicating an improvement in emotional well-being. This increase was maintained 1 year after RP. Comparing the HRQOL scores, 12 months after RP with the baseline measures revealed that none of the scores was significantly different from baseline, except the EWB subscale mentioned previously. Urinary symptoms, as measured by the IPSS, increased slightly 1 month after the completion of RP. By 3 months after RP, the IPSS score was significantly lower than the baseline IPSS score, indicating an improvement in LUTS. In fact, 1 year after RP, the LUTS, as measured by the IPSS, were significantly lower than at baseline. Comparing HRQOL among groups Because this was a nonrandomized, observational study, we were primarily interested in assessing the HRQOL changes over time within each treatment group. Men who select the various treatments are different, and they may differ in ways that have not been quantified by the measured covariates. We did, however, attempt to compare the HRQOL profiles observed among the treatment groups (Fig. 1). The initial between-group comparisons were performed after adjusting for the baseline FACT-P, and treatment was observed to be an independent predictor of HRQOL at 1 month. Men treated with IB or RP experienced a larger decrease in HRQOL in the first month than men treated with EBRT (IB vs. EBRT, p ; RP vs. EBRT, p ). No significant differences in HRQOL at T1 were observed between the IB and RP groups (p ). The results of subsequent between-group comparisons adjusting for potential confounders, including age, race, Gleason score, T stage, use of hormonal therapy, and baseline HRQOL scores, are listed in Table 5. Three different HRQOL scores are included in Table 5: FACT-P, TOI, and IPSS. After including the previously mentioned variables in the model, the only variables that independently predicted for HRQOL differences at T1 were the HRQOL score at baseline and the treatment the participant received. No treatment group differences were observed in any of the HRQOL scores 12 months after treatment (data not shown). Fig. 1. FACT-P scores by treatment.

6 Quality of life after treatment for prostate cancer W. R. LEE et al. 619 Table 5. Multivariate analysis of various HRQOL scores at 1 month FACT-P Baseline FACT-P score Age Race (white vs. black) T stage (T1 vs. T2) Pretreatment hormonal therapy (yes vs. no) Gleason score (2 6 vs. 7 10) Treatment (IB vs. EBRT vs. RP) TOI Baseline TOI score Age Race (white vs. black) T stage (T1 vs. T2) Pretreatment hormonal therapy (yes vs. no) Gleason score (2 6 vs. 7 10) Treatment (IB vs. EBRT vs. RP) IPSS Baseline IPSS score Age Race (white vs. black) T stage (T1 vs. T2) Pretreatment hormonal therapy (yes vs. no) Gleason score (2 6 vs. 7 10) Treatment (IB vs. EBRT vs. RP) Abbreviations as in Tables 1 and 2. DISCUSSION We used a longitudinal study design to measure patientreported HRQOL before and 12 months after 3 distinct treatments for clinically localized prostate cancer. Using a validated HRQOL instrument (FACT-P), we observed clinically significant decreases in HRQOL after IB and RP. The HRQOL scores also decreased after EBRT but to a smaller degree. The urinary symptoms, as measured by the IPSS, increased dramatically after IB. Much smaller changes in the IPSS were observed after EBRT and RP, and, in fact, the IPSS score decreased 1 year after RP. The baseline scores of FACT-P and FACT-G in our population were similar to those reported by Esper et al. (10) and Litwin et al. (16). We believe, therefore, that our study population is representative of men with clinically localized prostate cancer. It is important to point out that the baseline FACT-P and FACT-G scores were similar among the treatment groups despite the differences in the age distributions of the 3 groups. This observation suggests that the type of treatment received, rather than age, was a greater determinant of the HRQOL changes. This study was primarily designed to measure the changes within the treatment groups, such that each patient would serve as his own control. The changes in HRQOL observed in the IB group were very similar to those reported from an earlier population of men receiving IB at our institution. The FACT-P score decreased significantly during the first month, and the change was most evident in the PWB, FWB, and PCS subscales. Our previous conclusions p appear justified by the present experience. We know of no other report using a validated HRQOL measure in a prospective manner before and after IB. In the group of men treated with EBRT, the FACT-P score declined less markedly (8 points). This magnitude of change (roughly one-half standard deviation) is considered by many HRQOL investigators to be of marginal clinical significance. With 23 participants in the EBRT group, this study had a 90% power to detect changes of 15 points on the cumulative FACT-P score. Others have used a longitudinal design to assess patientreported symptoms and HRQOL before and after EBRT. Talcott et al. (22) examined 279 men receiving EBRT or RP in the Boston metropolitan area from All men completed a symptom questionnaire focusing on sexual, urinary, and bowel complaints. Participants completed the questionnaire before treatment and 3 and 12 months after treatment. Compared with baseline, at 3 months after treatment, men treated with EBRT reported an increase in bowel urgency or tenderness (28% of men) and an increase in urinary frequency (26% of men). Sexual dysfunction 3 months after EBRT did not differ from baseline in their study. No HRQOL instrument per se was analyzed in their report, but these results are consistent with the modest HRQOL decreases we observed in the EBRT group. The HRQOL changes observed in the RP group are consistent with the results of a recent prospective HRQOL analysis in the first year after RP (18). Litwin et al. (18) studied 90 men who were assessed at baseline and at several time points after RP, using the Rand 36-Item Health Survey (SF-36) and the Prostate Cancer Index. This analysis focused on a return to baseline end point. Three months after RP, only 23 38% of the men had returned to baseline functioning in the various domains of the SF-36, and only 7 37% had returned to baseline functioning in the urinary, sexual, and bowel domains of the Prostate Cancer Index. At 6 months, however, most men ( 70%) had returned to baseline in all the domains of the SF-36 and the Prostate Cancer Index, with the exception of the urinary and sexual domains. Our results are similar in that the patients reported a decrease in HRQOL, especially in the physical, functional, and prostate cancer-specific domains, but had returned to baseline by 12 months. It is interesting to observe that the only subscale that was statistically different from the baseline measures was the EWB subscale; in fact, EWB improved during the 12-month period. This improvement in the emotional domain has been reported by Clark et al. (9) in a similar study of men treated with RP or EBRT. A comparison of the HRQOL scores among the groups identified some differences. In the first month after treatment, men treated with IB or RP experienced a decrease in HRQOL that was of a larger magnitude than that observed in the men receiving EBRT (Fig. 1). At 3 months, the HRQOL differences among the groups were less striking, and at 12 months, the HRQOL scores were not statistically different from the baseline scores in any treatment group. This between-group comparison was, of course, prone to

7 620 I. J. Radiation Oncology Biology Physics Volume 51, Number 3, 2001 considerable biases. It is our opinion, however, that it is the best information available at present. All of these men were treated by the same three physicians at a single institution during a short interval. No other prospective HRQOL study from a single institution comparing different treatments has been reported. How is one to interpret these results? Do the observed HRQOL changes mean anything? Sound interpretation of HRQOL research requires distinguishing between statistical significance and clinical significance. A brief comment relating the magnitude of the HRQOL changes observed in this study and the clinical significance is, therefore, in order. What magnitude of change represents an important effect on well-being? More specifically, does a decrease in the FACT-P score by 15 points amount to a clinically meaningful reduction in a prostate cancer patient s quality of life, both clinically and as perceived by the patient? Two approaches to answering this question are discussed. Relying on either method supports our contention that the HRQOL changes seen after IB or RP are clinically meaningful. Sloan and others (30) at the Mayo Clinic have described one such approach. These authors classified the effect sizes as small, moderate, or large. Sloan et al. used the empiric rule that the standard deviation for any variable is roughly one-sixth of the range. For a two-sample t test, the small, moderate, and large effect sizes correspond to 0.2, 0.5, and 0.8 standard deviations. They admit that this process may be overly conservative. They report that in their experience, clinicians are unlikely to consider HRQOL differences as clinically significant unless the magnitude of the change with any particular HRQOL instrument is 10% of the theoretical range. This magnitude of change would correspond to a moderately large effect size by Cohen s method. Sloan et al. view this as the level of change which most physicians would be comfortable categorizing as clinically important. Another approach has come from investigators in Canada (31). Redelmeier et al. coined the phrase minimally important difference (MID) to denote the smallest difference in health status that signifies an important difference rather than a minor difference in patient symptoms. The method used to arrive at the MID depends on patients relating themselves to other patients with a similar diagnosis. The patient populations studied to date include patients with chronic obstructive pulmonary disease, rheumatoid arthritis, and, most recently, patients with head-and-neck cancer (31 33). In the most recent report, these investigators reviewed previous MID studies and found that across instruments and diseases, the MID values remain constant at 5 10% of the theoretical range (32). These two different methods arrive at similar conclusions, suggesting that a HRQOL instrument should change by at least 5%, and probably by 10%, of the theoretical range to be meaningful to both clinicians and patients. We believe that the magnitude of HRQOL changes seen in the IB group (17.9 units; 11.5% of the theoretical range) and the RP group (20.6 units; 13% of theoretical range) represent clinically meaningful decreases. The HRQOL decline reported by men in the EBRT group was of a smaller magnitude (7.6 units; 5% of theoretical range) and may not represent a major change in health status. More work examining the MID in patients with prostate cancer is necessary. Despite a prospective design using a reliable and valid HRQOL instrument, the present study has several deficiencies. First and foremost, this study did not randomize the participants to a particular treatment; the treatment choice was made by the individual patients with input from the physicians caring for them. As with almost any other report, the RP group was younger, on average, than the patients receiving IB or EBRT. There were also differences in the PSA levels and Gleason scores among the groups. Adjusting for potential confounding variables did not alter the relationship between the early HRQOL decreases and the treatment received (Table 5). The extent to which other variables, not measured in this study, may further confound the association between treatment and HRQOL is unknown. A second weakness of this report was the small sample size. Although the number of participants in each group did allow for the ability to detect changes of points in the total HRQOL score within groups, the power to detect smaller changes (2 5 points) was limited. As outlined above, HRQOL changes of this magnitude are likely to be considered minor by patients and physicians alike. Finally, some investigators may criticize the lack of a generic HRQOL measure in this report. This criticism has some validity, but must be balanced with the desire to decrease participant burden. The FACT-G component of the FACT-P has been validated as a HRQOL measure in a wide cross-section of cancer patients. Given that most generic HRQOL instruments could not distinguish prostate cancer patients from controls, it was decided to exclude a generic tool. The results of this analysis have important implications. Men who are considering RP or IB as treatment for localized prostate should be counseled that their HRQOL will likely decrease soon after treatment. Men considering EBRT can be counseled that any decreases in their HRQOL after treatment are likely to be quite modest. As many individual men are presented with different treatment options, they are most interested in how the different treatments compare with one another. Although it is tempting to use the data from this report to directly compare the HRQOL changes observed among the treatments, such a comparison would be prone to considerable biases. At the least, this work suggests that the HRQOL changes are likely to be treatment-specific, further emphasizing the importance of a randomized trial comparing the different treatment options in this population of men.

8 Quality of life after treatment for prostate cancer W. R. LEE et al. 621 REFERENCES 1. Stewart AL, Hays RD, Ware JEJ. The MOS short-form general health survey, Reliability and validity in a patient population. Med Care 1988;26: Kaasa S. Measurement of quality of life in clinical trials. Oncology 1992;49: Greenlee RT, Murray T, Bolden S, et al. Cancer statistics, CA Cancer J Clin 2000;50: Millikan R, Logothetis C. Update of the NCCN Guidelines for treatment of prostate cancer. Oncology 1997;11: Arterbery VE, Frazier A, Dalmia P, et al. Quality of life after permanent prostate implant. Semin Surg Oncol 1997;13: Borghede G, Sullivan M. Measurement of quality of life in localized prostatic cancer patients treated with radiotherapy: Development of a prostate cancer-specific module supplementing the EORTC QLQ-C30. Qual Life Res 1996;5: Borghede G, Karlsson J, Sullivan M. Quality of life in patients with prostatic cancer: Results from a Swedish population study. J Urol 1997;158: Brandeis JM, Litwin MS, Burnison CM, et al. Quality of life outcomes after brachytherapy for early stage prostate cancer. J Urol 2000;163: Clark JA, Rieker P, Propert KJ, et al. Changes in quality of life following treatment for early prostate cancer. Urology 1999; 53: Esper P, Mo F, Chodak G, et al. Measuring quality of life in men with prostate cancer using the functional assessment of cancer therapy-prostate instrument. Urology 1997;50: Joly F, Brune D, Couette JE, et al. Health-related quality of life, and sequelae in patients treated with brachytherapy, and external beam irradiation for localized prostate cancer. Ann Oncol 1998;9: Krongrad A, Granville LJ, Burke MA, et al. Predictors of general quality of life in patients with benign prostate hyperplasia or prostate cancer. J Urol 1997;157: Krongrad A, Litwin MS, Lai H, et al. Dimensions of quality of life in prostate cancer. J Urol 1998;160: Lee WR, McQuellon RP, Case LD, et al. Early quality of life assessment in men treated with permanent source interstitial brachytherapy for clinically localized prostate cancer. J Urol 1999;162: Lee WR, McQuellon RP, Harris-Henderson K, et al. A preliminary analysis of health-related quality of life in the first year after permanent source interstitial brachytherapy (PIB) for clinically localized prostate cancer. Int J Radiat Oncol Biol Phys 2000;46: Litwin MS, Hays RD, Fink A, et al. Quality-of-life outcomes in men treated for localized prostate cancer. JAMA 1995;273: Litwin MS, Hays RD, Fink A, et al. The UCLA Prostate Cancer Index: Development, reliability, and validity of a health-related quality of life measure. Med Care 1998;36: Litwin MS, McGuigan KA, Shpall AI, et al. Recovery of health related quality of life in the year after radical prostatectomy: Early experience. J Urol 1999;161: Lubeck DP, Litwin MS, Henning JM, et al. Measurement of health-related quality of life in men with prostate cancer: the CaPSURE database. Qual Life Res 1997;6: Lubeck DP, Litwin MS, Henning JM, et al. Changes in health-related quality of life in the first year after treatment for prostate cancer: results from CaPSURE. Urology 1999;53: Shrader-Bogen CL, Kjellberg JL, McPherson CP, et al. Quality of life and treatment outcomes: Prostate carcinoma patients perspectives after prostatectomy or radiation therapy. Cancer 1997;79: Talcott JA, Rieker P, Clark JA, et al. Patient-reported symptoms after primary therapy for early prostate cancer: Results of a prospective cohort study. J Clin Oncol 1998; 16: Litwin MS. Health related quality of life in older men without prostate cancer. J Urol 1999;161: Barry MJ, Fowler FJJ, O Leary MP, et al., for the Measurement Committee of the American Urological Association. The American Urological Association Symptom Index for benign prostatic hyperplasia. J Urol 1992;148: Cella DF, Tulsky DS, Gray G, et al. The Functional Assessment of Cancer Therapy Scale: Development and validation of the general measure. J Clin Oncol 1993;11: Beahrs OH, Henson DE, Hutter RVP, for the American Joint Committee on Cancer. Manual for staging of cancer. 4 th ed. Philadelphia: JB Lippincott; Blasko JC, Ragde H, Luse RW, et al. Should brachytherapy be considered a therapeutic option in localized prostate cancer? Urol Clin North Am 1996;23: Grimm PD, Blasko JC, Ragde H, et al. Does brachytherapy have a role in the treatment of prostate cancer. Hematol Oncol Clin North Am 1996;10: Bice WS, Prestidge BR, Prete JJ, et al. Clinical impact of implementing the recommendations of AAPM Task Group 43 on permanent prostate brachytherapy using I-125. Int J Radiat Oncol Biol Phys 1998;40: Sloan JA, O Fallon JR, Suman VJ, et al. Incorporating quality of life measurement in oncology clinical trials. Proc Biometrics Sect Am Stat Assoc 1997: Redelmeier DA, Guyatt GH, Goldstein RS. Assessing the minimal important difference in symptoms: A comparison of two techniques. J Clin Epidemiol 1996;49: Ringash J, O Sullivan B, Bezjak A, et al. Minimal important difference for quality of life measures in about five to ten percent of the instrument range. Am Soc Clin Oncol Proc 2000;19:434a. 33. Wells GA, Tugwell P, Kraag GR, et al. Minimum important difference between patients with rheumatoid arthritis: The patient s perspective. J Rheumatol 1993;20:

9 622 I. J. Radiation Oncology Biology Physics Volume 51, Number 3, 2001 APPENDIX FACT G (Version 3) Below is a list of statements that other people with your illness have said are important. By circling one number per line, please indicate how true each statement has been for you during the past 7 days. PHYSICAL WELL-BEING 1. I have a lack of energy I have nausea Because of my physical condition, I have trouble meeting the needs of my family I have pain I am bothered by side effects of treatment I feel ill I am forced to spend time in bed Looking at the above 7 questions, how would you say your PHYSICAL WELL-BEING affects your quality of life?... so SOCIAL/FAMILY WELL-BEING 9. I feel distant from my friends I get emotional support from my family I get support from my friends and neighbors My family has accepted my illness Family communication about my illness is poor I feel close to my partner (or the person who is my main support) Have you been sexually active during the past year? No Yes If yes: I am satisfied with my sex life Looking at the above 7 questions, how would you say your SOCIAL/FAMILY WELL-BEING affects your quality of life? so FACT G Questionnaire Please indicate how true each statement has been for you during the past 7 days. RELATIONSHIP WITH DOCTOR 17. I have confidence in my doctor(s) My doctor is available to answer my questions Looking at the above 2 questions, how would you say your RELATIONSHIP WITH THE DOCTOR affects your quality of life? so

10 Quality of life after treatment for prostate cancer W. R. LEE et al. 623 EMOTIONAL WELL-BEING 20. I feel sad I am proud of how I m coping with my illness I am losing hope in the fight against my illness I feel nervous I worry about dying I worry that my condition will get worse Looking at the above 6 questions, how would you say your EMOTIONAL WELL-BEING affects your quality of life? so FUNCTIONAL WELL-BEING 27. I am able to work (include work in home) My work (include work in home) is fulfilling I am able to enjoy life I have accepted my illness I am sleeping well I am enjoying the things I usually do for fun I am content with the quality of my life right now Looking at the above 7 questions, how would you say your FUNCTIONAL WELL-BEING affects your quality of life? so FACT P Questionnaire (Version 3) Please indicate how true each statement has been for you during the past 7 days. ADDITIONAL CONCERNS 35. I am losing weight I have a good appetite I have aches and pains that bother me I have certain areas of my body where I experience significant pain My pains keep me from doing things I want to do I am satisfied with my present comfort level I am able to feel like a man I have trouble moving my bowels I have difficulty urinating I urinate more frequently than usual My problems with urinating limit my activities I am able to have and keep an erection Looking at the above 12 questions, how would you say these ADDITIONAL CONCERNS affect your quality of life? so