Quality of life and symptoms in a randomized trial of radiotherapy versus deferred treatment of localized prostate carcinoma†
Article first published online: 13 DEC 2001
Copyright © 2001 American Cancer Society
Volume 92, Issue 12, pages 3111–3119, 15 December 2001
How to Cite
Fransson, P., Damber, J.-E., Tomic, R., Modig, H., Nyberg, G. and Widmark, A. (2001), Quality of life and symptoms in a randomized trial of radiotherapy versus deferred treatment of localized prostate carcinoma. Cancer, 92: 3111–3119. doi: 10.1002/1097-0142(20011215)92:12<3111::AID-CNCR10160>3.0.CO;2-E
Participating centra: Department of Urology, Boden, Piteå, Gällivare, Lycksele, Skellefteå, Sundsvall, Härnösand and Sollefteå, Sweden.
- Issue published online: 13 DEC 2001
- Article first published online: 13 DEC 2001
- Manuscript Accepted: 29 AUG 2001
- Manuscript Received: 22 MAY 2001
- Swedish Cancer Society
- Research Foundation of the Department of Oncology, University of Umeå
- prostatic carcinoma;
- watchful waiting;
- quality of life;
Treatment of localized prostate carcinoma (LPC) using radiotherapy (RT) can induce disturbances in a patient's quality of life (QOL) and urinary and intestinal function. Late symptoms and QOL were evaluated in a randomized trial between RT and deferred treatment (DT).
Quality of life was evaluated with European Organization for Research and Treatment of Cancer's QLQ-C30 (+3) formula. Urinary and intestinal problems were evaluated with a validated symptom specific self-assessment questionnaire, QUFW94. The questionnaires were sent to 108 randomized patients with LPC and to an age-matched control group (n = 68). Mean age was 72 years. Mean total dose was 65 grays (Gy; 62.3–70 Gy). The median follow-up time from randomization was 40.6 months for the RT group and 30.4 months for the DT group.
Social functioning was the only QOL scale in which a significant difference was found between the two patient groups and compared with the control group. Multivariate regression analysis showed that hematuria, incontinence, mucus, and planning of daily activities in response to intestinal problems caused this decrease in QOL in the RT group. A significant increase of intestinal problems was observed in the RT versus DT groups regarding mucus, stool leakage, intestinal blood, and planning of daily activity in response to intestinal problems.
The RT patients showed increased levels of minor intestinal side effects compared with the DT patients and the controls, but the RT patients reported no decreased QOL except for decreased social functioning. This could be because this group developed coping skills or because of a low magnitude of side effects to influence the QOL. Cancer 2001;92:3111–9. © 2001 American Cancer Society.
Patients with localized prostate carcinoma (LPC) are mostly of advanced age, and the tumors are slow-growing; many of these patients will die of other diseases. With deferred treatment, 80% of the patients with small tumors (International Union Against Cancer [UICC], 1978 T0-T2) and well or moderately differentiated tumors will be alive for 15 years after diagnosis.1 No well conducted randomized studies have given any solid proof that the curative treatment methods currently used, such as radical prostatectomy or radiotherapy, can prolong survival as compared with deferred treatment in patients with LPC. This means that any treatment given to this patient group that causes side effects should be monitored carefully to assess side effects so that the given treatment does not induce more problems for the patient than the disease itself.
Radiotherapy with curative intent is commonly used to treat LPC. Late effects are reported from 15–20% of the patients describing mild to moderate urinary and gastrointestinal problems.2, 3 Most studies on treatment-related morbidity after pelvic radiotherapy are based on the physician's estimation of the major side effects.4–7 Fewer studies are published that describe the LPC patient's own opinion of the side effects after pelvic radiation treatment using self-administered questionnaires.8–13
Even if LPC patients are cured, they may still experience side effects, and the level of functioning may be diminished. Patients also may experience psychosocial impairment of life after treatment. Current studies of prostate carcinoma often include an evaluation of the patients' perceptions of their quality of life (QOL),11, 12, 14–17 but so far none to our knowledge has compared symptoms and QOL using a randomized patient population of radiotherapy (RT) and deferred treatment (DT). The current study also compares the two patient groups with an age-matched population of men from the same region (Northern part of Sweden) without LPC.
The objective of this study was twofold. Using the European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 (+3) formula, we studied how the side effects and symptoms in each patient group affected the patients' daily QOL and compared that with a control population. Second, we compared the patients' own evaluation of late (> 1 year) urinary and intestinal symptoms after RT or DT and compared that with an age-matched control population.
The results of this study should show how treatment-induced symptoms influence QOL in RT or DT patients with a slow-growing tumor.
MATERIALS AND METHODS
Patient and Control Populations
Between 1986 and 1996, a total of 166 patients with LPC were included in a randomized trial (Umeå 1) comparing RT with DT. All patients had cytologically or histologically verified PC: T1a-T2, G1-G2, pN0, and M0. (At the end of the study, no lymph node staging was performed if the prostate specific antigen [PSA] value was <10 ng/mL; therefore, some patients were recorded as having a lymph node status of Nx, Table 1). Exclusion criteria were from the protocol previous treatment for PC and other disease having an expected survival shorter than a normal population at the same age, that is an expected survival > 10 years.
|Characteristic||RT (n = 59) (%)||DT (n = 49) (%)|
|Lymph node status|
From the primary group of 166 patients, we excluded 30 patients who had died according to the Swedish population register at the date of the submission of the questionnaire. Of the remaining 136 patients, 11 patients were excluded because of a follow-up less than 6 months or a progressed disease with progressive treatment before the questionnaire.
The answer frequency was 90% in the RT and 85% in the DT group. There were no differences between the groups regarding the tumor characteristics (according to International Union Against Cancer 1992) at the time of diagnosis of the RT (n = 59) and DT (n = 49) patients were T1a, 5%/6%; T1b, 18%/12%; T1c, 2%/0%; T2, 75%/82%, respectively (Table 1, ns). The lymph node status, grade (according to the World Health Organization grading system), and PSA values at the time for diagnosis are presented in Table 1.
The patient groups also were compared with an age-matched control group of PC free men, recruited from the Swedish population register, from the same region as the population of treated patients with PC. The answer frequency in the control group was 49% (n = 68).
The median follow-up time from the randomization date to the time of the questionnaire was 40.6 months for the RT group and 30.4 months for the DT group (P = 0.055).
The mean age at the time of questionnaire was 71.3 years (range, 49.1–83.0 years) in the RT group and 72.8 years (range, 58.9–81.9 years; P = 0.136) in the DT group. The mean age in the control group was 72.2 years (range, 58.9–81.9 years; P = 0.327).
Radiotherapy Treatment Technique
Computed tomography planned RT was given with conventional (n = 31) or conformal (n = 28) four-field box technique, 5 days a week, 2 grays (Gy) per fraction. The average total dose in the patient group primarily treated with RT was 64.8 Gy (range, 62.3–70.0 Gy). Treatment was given with 20.9- or 50-MeV photons. Planning target volume included the prostate (macroscopic target volume) and a 2-cm margin.
Two questionnaires (QUFW94 and QLQ-C30 [+3]) were used. The questionnaires and an information letter were mailed to the patients between 1991 and 1998 and to the controls in May 1995. If no answer was received after 1 month a reminding letter was sent.
The validated self-assessment questionnaire QUFW94,8–10 which is written in Swedish, was designed to evaluate PC patients' side effects after pelvic RT. The questionnaire contains four main categories: general function (physical ability, life situation, marital status, and eating habits), urinary problems, intestinal problems (bowel function), and sexual function (not presented in this article).
The questionnaire contained 38 questions with modified linear-analog scales containing boxes with values between 0 and 10; 0 was labeled on the scale as “no problem/very good function” and 10 was “many problems/very bad function.”
Higher values mean more problems/worse function. Some questions contained only “yes” or “no” answer alternatives. The patients were encouraged to evaluate their symptoms and life situation during the previous week. Some questions requested only a written answer/comment (those questions are not summarized and reported in the current study).
The EORTC QLQ-C30 (+3) Questionnaire
The EORTC has developed the QLQ-C30 (version 1) questionnaire, which has been thoroughly validated and cross-culturally tested in cancer patients.18–24 The questionnaire used in this study, QLQ-C30 (+3), is an interim version, which contains questions from both the original (version 1) and modified questions included in the latest current version 2. In the analysis of the data, only the questions included in the current version 2 are summarized and reported. The questionnaire contains five functioning scales: physical, role functioning, emotional, cognitive, and social functioning. It also includes a global health status/QOL scale. Higher mean scores on these scales reports better functioning and better QOL. Three symptom scales were also included concerning nausea and vomiting, pain, and fatigue. Six single symptom items measured the levels of constipation, diarrhea, loss of appetite, sleep disturbance, dyspnea, and the financial impact. Higher mean scores on the symptom and single items indicate more symptoms/problems.
The statistical analyses were performed using the SPSS version 8.0 software package. Mean values were calculated for all items. To evaluate differences between patient and control groups, we used the nonparametric Mann–Whitney tests. A P value less than 0.05 was regarded as statistically significant.
Multivariate linear regression analysis was used when trying to find the variable(s) best predicting the value of the different dependent variables.
All calculations on the EORTC QLQ-C30 (+3) questionnaire were performed after the scores were linearly transformed to a 0 to 100 scale to facilitate presentation and interpretation of the data according to the EORTC recommendations.
Permission to perform the study was given both from the Ethical Board at Northern University of Sweden and the National Computer Inspection Board in Sweden.
Quality of Life QLQ-C30 (+3)
Social functioning was the only scale in which the RT patients reported (mean, 80.0) a worse outcome than the DT (mean, 92.8) and control groups (mean, 93.0, P = 0.003, Fig. 1). Multivariate linear regression analysis with social functioning as the dependent variable showed that hematuria, urinary incontinence, mucus, and planning of daily activity because of intestinal problems (r2 = 0.578) were the variables that best predicted the worse outcome of the social functioning in the RT group.
No significant differences were observed between the patient groups and the controls on the fatigue, nausea/vomiting, and pain symptom scales (Fig. 2). Diarrhea was the only symptom of the single items that showed a significant difference between patients and controls (Fig. 3), whereas no difference in diarrhea symptoms was observed between the two patient groups (RT, 9.7; DT, 10.1; P = 0.466).
Symptom Evaluation QUFW94
All three groups reported very good physical ability (Table 2). Eighty-five percent in the RT patient group, 92% in the DT patient group, and 87% in control group could “move freely indoors and outdoors without any help” (P = 0.612). A small difference was detected between the RT (mean, 1.8) and DT (mean, 0.7) patients regarding the “limitation in the daily life” caused by their PC (P = 0.001; Table 2). No difference in the occurrence of other diseases regarding the “limitation in the daily life” was reported between the RT (46%) and DT (42%) patient groups and the controls (39%; P = 0.784). The RT patients (mean, 3.5) reported decreased “life situation” compared with the DT patients (mean, 1.9) and controls (mean, 1.8; P < 0.001). The RT patient group used more intestinal medication, mostly for the prevention of diarrhea (P < 0.001; Table 3).
|Variable||RT (n = 59)||DT (n = 49)||Control (n = 68)|
|Mean||95% CI||Mean||95% CI||P valuea||Mean||95% CI||P valueb|
|Limitation in daily activity caused by prostate carcinomac||1.8||1.29–2.39||0.7||0.19–1.26||0.001||—||—||—|
|Existence (daily life)||2.6||2.08–3.18||1.8||1.10–2.40||0.012||1.7||1.19–2.28||0.007|
|Urinary frequency/24 hrs||7.7||6.97–8.37||7.6||6.94–8.25||0.898||6.3||5.78–6.76||0.001|
|Limitation in daily activity caused by urinary problems||1.1||0.63–1.62||0.9||0.27–1.47||0.06||0.3||0–0.63||0.000|
|Urinary problems in general||1.8||1.15–2.42||1.2||0.60–1.86||0.227||1.0||0.49–1.54||0.06|
|Limitation in daily activity caused by intestinal problems||1.3||0.73–1.85||0.3||0.04–0.48||0.001||0.1||0–0.22||0.000|
|Stool frequency/24 hrs||2.1||1.78–2.44||1.3||1.07–1.48||0.000||1.6||1.39–1.76||0.000|
|Intestinal problems in general||2.3||1.58–3.05||1.0||0.40–1.64||0.003||0.5||0.16–0.77||0.000|
|Planning of daily activity caused by intestinal problems||1.9||1.19–2.54||1.0||0.22–1.72||0.004||0.2||0–0.55||0.000|
|Problem||RT (n = 59) (%)||DT (n = 49) (%)||P value RT vs. DT||Control (n = 68) (%)||P valueb (all groups)|
|Limitation in daily activity caused by other diseasec||45.6||41.7||0.686||39.4||0.784|
|Use of pad for urinary incontinence||17.5||2.1||0.011||0.0||< 0.001|
|Use of sanitary shields for stool leakage||8.6||4.1||0.346||0.0||0.055|
|Use of catheter||0.0||2.1||0.272||1.5||0.577|
|Use of intestinal medication||30.4||14.6||0.058||4.5||< 0.001|
The urinary frequency during 24 hours, a possible indication of both bladder irritation and obstruction of urinary flow, did not differ between the two patient groups (P = 0.898), but a significant increase in urinary frequency was detected in both patient groups as compared with the control group (P = 0.001; Table 2). There was no difference between the three groups regarding the night urinary frequency (nocturia, Table 2).
For the question “do you have problems with your urinary tract?” no difference could be detected between the two patient groups (P = 0.227; Table 2). However, there was a slight increase in urinary problems in general in the RT patients as compared with the controls (P = 0.019).
A slight difference could be detected between the RT group (mean, 1.1), DT group (mean, 0.9), and the controls (mean, 0.3) to the question whether their urinary problems in some way influenced their daily activities (P < 0.001; Table 2). Incontinence and hematuria problems/symptoms also showed a significant difference between the two patient groups (Table 2).
Seventeen percent of the RT patients (n = 10) and 2% of the DT patients (n = 1) regularly used pads for urinary leakage (P ≤ 0.001; Table 3). None of the controls regularly used pads.
No difference regarding urinary problems was found between the patients treated with conventional versus conformal technique except in hematuria (conventional mean, 0.43; conformal mean, 0.26; P = 0.041).
The stool frequency per 24-hour period was increased in the RT group compared with both the DT patient group and the controls (P < 0.001; Table 2).
The RT patients reported more intestinal problems on almost all subquestions as well as increased general intestinal problems compared with both the DT patients and the controls (Table 2) except for bowel movements, cramps, nausea, or consistency of the stools (Table 2). The control group reported a very low level of intestinal problems (Table 2). A comparison between the two treatment techniques (conformal vs. conventional) did not show any significant differences on the bowel problem scales.
In this study for the first time to our knowledge, QOL and urinary and intestinal symptoms were compared in patients participating in a randomized trial between RT and DT. The only observed significant difference in QOL was on the social functioning scale in the QLQ-C30 questionnaire. Hematuria, incontinence, mucus, and planning of daily activity because of intestinal problems were the most important factors influencing the social functioning in the multivariate analysis. However, hematuria was a very rare problem even for RT patients whereas only 2 patients (3%) reported a value higher than 1.0 on the scale of 0 to 10 in the QUFW94 formula.
Very few differences were observed in comparisons of QOL between the two treatment groups and compared with the control population. However, significant differences were observed using the more symptom specific QUFW94 questionnaire10 in measures of urinary problems and especially intestinal problems between the two patient groups.
Litwin et al.12 found no difference in QOL among PC treatment groups (surgery, irradiation, and observation) in a nonrandomized trial as compared with a control group. Similar findings with no difference in QOL also were reported by Lilleby et al.14 and by Joly et al.15 comparing patients treated with brachytherapy and external irradiation with controls. Two Scandinavian population studies17, 25 showed the same magnitude of QOL in the age group of men older than 70 years as in the control group in the current study.
The RT patients reported a rather low level of treatment-related urinary late side effects compared with the DT patients. Incontinence and hematuria were the only symptoms that were significantly different between the two patient groups. However, these problems were of mild character, even in the RT group, because the highest reported problem (incontinence) only reached a mean value of 1.5 on a 0 to 10 scale. The relatively high percentage of our RT patients using pads (17%), compared with 2% reported by others,26, 27 may be because those patients adapted to their incontinence problem by using urinary pads; therefore, it would explain the low frequency of reported incontinence problem. A similar level of reported urinary problems also was found in an earlier study comparing PC patients and controls.8 Joly et al.15 also showed that incontinence was one factor that differed between the patients and controls and was reported more often as a problem by the patients than by physicians.
The reported intestinal problems in the RT group were of mild nature, but significant differences in most symptoms between the RT and DT patients were detected. The RT patients are limited in their daily activity caused by intestinal problems to a higher degree than both the DT patients and the controls, mostly regarding the interference with the reported problems of mucus in the RT group.
The two different treatment techniques, conventional four-field box technique versus conformal technique, did not show any difference regarding bowel symptoms in the evaluation, contrary to earlier studies.14, 28 However, the current study consisted of a rather small sample of patients in the two different treatment techniques. These results might suggest/indicate differences in magnitude of change of treatment technique.
In a randomized trial of conformal versus conventional RT in PC patients, Dearnaley et al.,3 found fewer patients with proctitis and rectal bleeding in the conformal group. In that study, modest doses of 64 Gy were used as in our study, and the magnitude and type of symptoms were similar to those in the Dearnaley et al. study.3 Nguyen et al.26 reported decreased changes in bowel function (10% and 34%) among patients receiving conformal treatment despite higher dose (78 Gy) than patients with conventional technique (70 Gy), respectively.
It was a bit surprisingly that the QLQ-C30 formula did not detect any large differences in QOL between the groups despite that the more PC specific symptom questionnaire (QUFW94) detected significant differences between the groups especially regarding intestinal symptoms. This might suggest that these minor symptoms were of such low magnitude that they did not affect the quality of living to the extent that it would be detectable with the EORTC formula. Another explanation also might be that the RT patients had had time to adapt to the RT-induced symptoms and to develop coping behaviors.29–31
In patients with untreated LPC, a high level of fear/worrying for the outcome of the disease was found in an interview study of seven men with untreated LPC (1–3 years after diagnosis). These patients described their situation as “a life in the shadow.” This suggests that living with untreated LPC could be a burden, which also might contribute the marginal difference in QOL between the patient groups (Hedestig O, Sandman P-O, Widmark A, unpublished data).
This study also confirms the important factor that treatment of LPC has consequences that continue to affect the patient for many years after the given treatment. Even if day-to-day life is normalized, symptoms may occur. The knowledge of possible outcomes when selecting treatment for these patients must include both survival and QOL.
This was a study of 108 patients for whom QOL and symptoms were evaluated with a cross-sectional design. Unfortunately no information about QOL and symptoms was available on the patients' general health or physical function before diagnosis and treatment. Conversely, side effects are compared within a randomized population, which will strengthen the results. With this limitation of lack of baseline values, we cannot really be sure how much the RT affected the patients compared with before the start of the treatment.
The relatively low response rate of 49% among the controls also was reported in a QOL study by Litwin et al.12 However, because the questionnaires have been used in other studies including comparisons of patients with LPC and age-matched controls,8–10 we know that the level of reported symptoms in age-matched PC free men are of the same magnitude as in the current control population.
The lack of evaluation of sexual function and the correlation with QOL in the current study is important, because decreased sexual function (erectile dysfunction) after RT against LPC is still common. However, because the survival analyses of the randomized study still are unavailable (closed study), we could not present information on sexual function in a proper way without presenting the outcome and treatment of the progressed disease (i.e., hormones), which would affect the results of the sexual function in this study.
The management of LPC is controversial. Besides DT, the treatment options include prostatectomy and RT (external and/or brachytherapy), but many patients will die from diseases other than LPC. The choice of optimal handling/treatment requires properly counseling the patients about the risks and advantages of each treatment option. The treatment-related morbidity must be minimized, especially when treatment is given to men with low-risk LPC that would not have troubled them for many years even without any treatment.1, 32–36 It may be easier to choose DT for a less healthy patient (e.g., with a history of myocardial infarction) than a healthy man with LPC considering the time to benefit from treatment as well as risk for complications. However, even with a longer life expectancy, an extended life span should be weighed against the risk of complications due to treatment so that the patients and physicians will be better able to make decisions about proceeding with active therapy or deferring treatment for LPC. One question that should be asked is: is it worthwhile to live the remaining years with RT-induced side effects when a patient seldom has any troublesome symptoms from the disease at the time of diagnosis?
The RT patients did not report any decreased QOL, except on social functioning, compared with both the DT patients and the control group using the QLQ-C30 formula. With the more PC treatment specific QUFW94 questionnaire, the RT patients compared with the DT patients and controls reported increased levels of many intestinal symptoms of minor magnitude and frequency. The marginal influence of symptoms on QOL could be because this group developed coping skills or because the magnitude of side effects was too low to influence the QOL. The current study demonstrated that the QLQ-C30 and the QUFW94 formula can be used to measure outcomes in older men with treated or untreated LPC, as well as men without PC.
- 32Natural history of localized prostatic cancer. Lancet 1989; 15: 799–803., , , , , .