Previously presented in abstract form at the Genitourinary Cancers Symposium, 6 March 2010.
Patient-reported quality of life during radiation treatment for localized prostate cancer: results from a prospective phase II trial
Version of Record online: 13 APR 2012
© 2012 BJU INTERNATIONAL
Volume 110, Issue 11, pages 1690–1695, December 2012
How to Cite
Chen, R. C., Zhang, Y., Chen, M.-H., McMahon, E., Loffredo, M., McPherson, C. P., Nguyen, A. U., Nguyen, P. L. and D'Amico, A. V. (2012), Patient-reported quality of life during radiation treatment for localized prostate cancer: results from a prospective phase II trial. BJU International, 110: 1690–1695. doi: 10.1111/j.1464-410X.2012.11117.x
- Issue online: 6 DEC 2012
- Version of Record online: 13 APR 2012
- Accepted for publication 18 January 2012
- prostate cancer;
- radiation therapy;
- quality of life;
- clinical trial
Study Type – Therapy (cohort)
Level of Evidence 2a
What's known on the subject? and What does the study add?
Patient-reported quality of life (QoL) in prostate cancer is recognized as an important outcome, and has been shown in multiple studies to capture the incidence and timing of patient symptoms more accurately than physician-graded toxicity reports. Although the long-term QoL after completing radiation therapy (RT) has been previously studied, patient experience during RT is not well described in the literature.
The present study collected patient-reported QoL during RT in a prospective phase II clinical trial. The study describes in detail the time course and severity of gastrointestinal and genitourinary symptoms during radiation, providing clinically useful information for patients and physicians considering RT during the treatment decision-making process.
- • To evaluate data collected from a phase II trial to describe the time course and trajectory of patient-reported acute urinary and bowel symptom development during radiation therapy (RT) for prostate cancer.
PATIENTS AND METHODS
- • In all, 100 patients with intermediate- or high-risk prostate cancer received 72 Gy of RT to the prostate and seminal vesicles, with 6 months of concurrent androgen deprivation therapy; a rectal balloon was used for prostate immobilization.
- • Patients completed the validated Prostate Cancer Symptom Indices questionnaire every 1–2 weeks, reporting urinary and bowel symptoms on a four- or five-point Likert scale.
- • A score of ≥3 in a symptom is associated with clinically meaningful distress. Cumulative incidence of each symptom is reported. Bonferroni corrections of P values were used to adjust for multiple comparisons.
- • Urinary symptoms were frequent at baseline and worsened during treatment. By the end of RT, 28–50% of patients developed clinically meaningful obstructive and irritative urinary symptoms.
- • Acute bowel symptoms were less frequent. Each bowel symptom increased in frequency by 9–26% from baseline to end of RT.
- • Urinary incontinence was rare.
- • Overall, symptom burden at the end of treatment was modest.
- • Urinary symptoms were common during RT, and bowel symptoms were less frequent.
- • These results inform patients and physicians during the decision-making process about potential patient quality of life experiences during RT, and also provide a benchmark for comparative effectiveness studies against newer treatments and technologies.
Prostate Cancer Symptom Indices
quality of life
Health-related quality of life (QoL) is increasingly recognized as an important outcome after all forms of cancer treatment [1,2]. In prostate cancer, numerous studies over the past 10 years have measured patient-reported urinary and bowel symptoms after radiation therapy (RT) [3–5]. While physician-assessed toxicity reporting has been useful, a number of studies have shown that physicians might under-report the frequency and/or severity of treatment-related symptoms [6,7]. Reports by the patients themselves can best capture the true incidence and timing of symptom development and convey its effects on QoL, and, to this end, the National Cancer Institute has identified patient-reported outcomes as a priority area for future research .
Previous QoL studies on prostate cancer have focused on long-term symptoms after RT, but acute symptom development during treatment has been largely ignored in research. During RT, acute urinary symptoms such as frequency, dysuria and nocturia are sometimes experienced, as are bowel side-effects such as diarrhoea and urgency. The frequency and severity of urinary and bowel symptoms during treatment, and the timing of symptom development, are not well described. This information can help physicians counsel their patients before treatment, and therefore inform the decision-making process.
In a prospective phase II trial, 100 patients with intermediate- or high-risk prostate cancer were treated with three-dimensional conformal RT to 72 Gy, with 6 months of androgen deprivation therapy; during the boost portion of the RT, an intrarectal balloon was used daily for prostate immobilization [9–11]. Patients completed a validated prostate cancer-specific QoL instrument every 1–2 weeks during treatment. The goal of the present study is to report the trajectory of patient-reported urinary and bowel symptoms during RT. The incorporation of QoL life measurement during RT to capture acute patient experiences in a prospective clinical trial setting is novel and a currently understudied area of research.
PATIENTS AND METHODS
In all, 100 patients with biopsy-proven prostate cancer were enrolled in an institute review board-approved phase II Dana Farber-Harvard Cancer Center study between September 2001 and November 2004. All patients had at least one feature indicative of intermediate- or high-risk disease: PSA level >10 ng/mL, Gleason score ≥7, or clinical or radiographic T3 disease. Exclusion criteria were history of abdominoperineal resection, prostatic-rectal fistula, inflammatory bowel disease or an anal canal stricture.
Patients received three-dimensional conformal RT using a four-field technique with 15 MV photons, to a total dose of 72 Gy. The details of RT have been described previously . Radiation targeted the prostate and seminal vesicles, but not the pelvic lymph nodes. During the boost portion of RT (15 treatments), a rectal balloon was used daily for prostate immobilization. Androgen deprivation therapy consisted of leuprolide and an antiandrogen (flutamide or bicalutamide), starting 2 months before RT and was given for a total duration of 6 months.
The validated Prostate Cancer Symptom Indices (PCSI) questionnaire was given to patients at baseline (before initiating RT) and every 1–2 weeks during RT [13,14]. The ‘urinary obstruction and irritation’ scale of the PCSI includes the following five questions:
- 1In the past week, how easy has your urine flow been? (Urine flow)
- 2In the past week, how often did you urinate at night? (Nocturia)
- 3In the past week, how many times did you urinate? (Frequency)
- 4In the past week, how often have you felt pain or burning during urination? (Dysuria)
- 5In the past week, how often did you have the feeling that it is urgent to pass your urine? (Urgency)
In addition, a question assessing urinary incontinence was included.
The ‘bowel problems’ scale of the PCSI includes the following seven items:
- 1In the past week, have you had diarrhoea or loose watery stools? (Diarrhoea)
- 2In the past week, have you had a sense of urgency that you move your bowels? (Urgency)
- 3In the past week, have you had any tenderness or pain when you move your bowels? (Pain)
- 4In the past week, have you had bleeding with your bowel movements? (Bleeding)
- 5In the past week, have you had abdominal cramping or pain? (Cramping)
- 6In the past week, have you passed mucus from your rectum? (Mucus)
- 7In the past week, have you had the feeling that you have an urge to move your bowels, but have nothing to pass? (Tenesmus)
Patients self-reported the frequency of each symptom on a four- or five-point Likert scale. A score ≥3 in any symptom was previously found to be associated with patient distress and was therefore deemed a ‘clinically significant’ problem a priori.
Descriptive statistics were used to report the frequency and severity of individual urinary and bowel symptoms during RT. Cumulative incidence of clinically meaningful symptoms (score ≥3) was reported. Generalized estimating equations were used to assess the statistical significance of changes in symptom frequency over time . All P values were two-sided. Bonferroni correction of P values was used to adjust for multiple comparisons .
Linear regression multivariable analysis was used to assess the potential association between age, prostate volume and volume of rectum receiving ≥70 Gy (rectal V70) with the overall urinary obstruction and irritation score and the bowel problems score, as calculated per Clark et al. 
Patient and clinical characteristics are shown in Table 1. The median age was 67 years and 76% of the prostate cancers were clinically staged T1 or T2a. The volume of rectum receiving high dose radiation (≥70 Gy) was limited to less than 10 cc according to the protocol specifications.
|Percentage of 100 patients (unless noted otherwise)|
|Median (range) age, years||67 (46–80)|
|Biopsy Gleason score|
|6 or less||29|
|Initial PSA level, ng/mL|
|≥10 to <20||18|
|Rectal V70, mL|
Figure 1A shows, for the entire patient cohort, the frequency and severity of urinary symptoms at baseline and at week 7–8, which represents the end of the RT course. Most patients reported some degree of nocturia (72%), urinary frequency (75%) and urgency (62%), even at baseline. By week 7–8, an additional 20–25% of patients reported having slow urinary flow (20%), nocturia (24%), frequency (22%) and urgency (21%), while an additional 49% of patients developed some degree of dysuria. Figure 1B and C show the outcomes for subgroups of patients with no baseline symptom (Fig. 1B) and those with baseline symptoms (Fig. 1C). Development and/or progression of symptoms was seen in both subgroups. Patients in general had mild to moderate symptoms (scores 2–3), and less than 15% of patients reported symptoms of higher severity with the exception of nocturia (35% reported nocturia more than three times at week 7–8).
The trajectory of urinary symptom development during RT is shown in Fig. 2. Based on prior work by Talcott et al.  using the PCSI, we defined a score ≥3 to be a clinically meaningful endpoint, representing a degree of symptoms that caused patient distress. From baseline to week 7–8, there was a steady development in each of the urinary symptoms with time. The biggest increases were seen in clinically significant nocturia, urinary frequency and urgency. Urinary incontinence was rare, only reported by one patient (data not shown). All five symptoms showed statistically significant increases from baseline to week 7–8.
Figure 3A shows the frequency and severity of bowel symptoms at baseline and week 7–8. By contrast to urinary symptoms, most patients did not have baseline bowel symptoms. Less than 10% of patients reported ‘frequent’ or ‘very frequent’ symptoms at week 7–8, except for rectal urgency (12%). Outcomes for the subgroup of patients with no baseline bowel symptoms are presented in Fig. 3B; too few patients had baseline bowel symptoms for meaningful results to be shown for that group. The time course of bowel symptom development is shown in Fig. 4. All symptoms except cramping showed statistically significant increases in frequency from baseline to week 7–8.
To assess the overall symptom burden, we examined the mean score across the five items in the urinary obstructive and irritative scale (Fig. 5A) and the seven items in the bowel problems scale (Fig. 5B) by week. While 85% of patients had a mean score ≥2 across all urinary symptoms by week 7–8, only 34% had a mean score ≥3 and 1% had a mean ≥4. For the bowel symptoms, 33% of patients had a mean score ≥2 across all symptoms by week 7–8; one patient had a mean score ≥3 and no patient had a mean ≥4.
Multivariable analysis demonstrated no significant association of age, prostate volume or rectal V70 with the bowel problems score. Age and prostate volume were not significantly associated with the urinary obstructive and irritative score.
Using a validated QoL instrument, the present study described patient-reported frequency, severity and time course of urinary and bowel symptom development during an 8-week course of RT for localized prostate cancer. The goal of the present study was to provide patients and physicians with information regarding acute treatment-related symptoms, to facilitate treatment decision-making and informed consent for RT. To our knowledge, this information is currently not well described in the published literature.
Radiation therapy is an effective, first-line treatment modality for localized prostate cancer. However, patients can experience acute urinary and bowel toxicity due to incidental radiation dose to the bladder, urethra and rectum. To help patients anticipate the side-effects they might experience during treatment, it is essential to quantify the frequency and severity of these symptoms through research. Several studies have reported acute treatment-related toxicity based on physician report [18–22]. For example, Matzinger et al.  reported the acute toxicity of patients receiving three-dimensional conformal or intensity-modulated RT on a clinical trial by the European Organization for Research and Treatment of Cancer (EORTC). Overall, 35% of patients developed grade 2 urinary toxicity, and another 6% Grade 3; 18% developed Grade 2 bowel toxicity and 1% Grade 3. While these reports are beneficial, information obtained from the patients themselves could capture more fully the actual extent and significance of treatment-related symptoms. Because physician assessments can under-report patient symptoms [6,7], there is increasing recognition of the value of patient-reported outcomes to provide complementary information. The National Cancer Institute and the cooperative groups have called for incorporation of patient-reported outcomes in all clinical trials .
In a phase II trial which treated intermediate- or high-risk patients with prostate cancer with RT and 6 months of concurrent androgen deprivation therapy, we measured patient-reported urinary and bowel symptoms every 1–2 weeks during RT using the validated Prostate Cancer Symptom Indices questionnaire [14,15,23]. Patients reported a steady increase of urinary and, to a lesser extent, bowel symptoms throughout 8 weeks of treatment. Symptom development and progression affected both patients with and without baseline symptomatology. Comparing baseline to end of treatment (week 7–8) for clinically significant (score ≥3) symptoms, ≈ 28–50% of patients developed each of the urinary symptoms, and 8–26% developed bowel symptoms. As treating physicians prescribed alpha-blockers (such as tamsulosin), non-steroidal anti-inflammatory, and/or anti-diarrhoeal medication when deemed medically necessary, these rates of symptoms occurred inclusive of medical intervention.
We found two other studies in the literature which used validated QoL instruments to examine acute patient-reported symptoms during RT for prostate cancer [24,25]. In general, the results of the present study are consistent with these two prior studies. Pinkawa et al.  administered the Expanded Prostate Cancer Index Composite (EPIC) to patients before starting radiation and on the last treatment day. The mean urinary bother score worsened by 22 points (on a 100-point scale) and the bowel bother score worsened by 23 points. Geinitz et al.  used the EORTC Prostate Cancer Module (PR-25) pre-treatment, and at 40 and 60 Gy. Mean urinary symptom score worsened by 14 points from baseline to 60 Gy, and bowel symptoms score by three points. However, some relevant symptoms are not included in the PR-25, including frequency, urgency, pain and mucus. Neither of these studies comprehensively described the time course of symptom development during radiation. Further, the clinical meaning of these score changes is difficult to interpret, and neither study described the development and severity of individual urinary and bowel symptoms.
An important limitation of the present study is the use of three-dimensional conformal RT to a total dose of 72 Gy. Today, most patients receive intensity-modulated RT (IMRT) and to doses of 75.6 Gy or higher – which could result in acute QoL experiences that differ from what has been reported by the present study. However, to our knowledge, the present study provides the first clinical trial report in the literature regarding acute QoL changes during RT for prostate cancer, and therefore offers patients and physicians much needed information that is not currently available. Our data add to and complement the currently available information (physician-reported acute toxicity from clinical trials) to help in the counselling of patients before RT. Whether IMRT to higher doses results in significantly different outcomes requires future research.
While we described the acute symptom development during RT, the present study was not designed to examine symptom resolution over time or the long-term QoL effects of RT for prostate cancer. These have been reported by other studies [23,25]. In our limited exploratory multivariable analysis, we did not find a significant association between rectal V70 and the overall bowel problems score. Further, the overall incidence of acute bowel symptoms was low. This is most likely due to the narrow range of V70 values which was required by study protocol (all radiation plans must have V10 <10 mL). It is possible that patients treated with higher doses to the rectum could experience increased morbidity; therefore, the potential association between dose to organs and patient-reported QoL deserves further study.
In summary, we comprehensively described acute patient-reported urinary and bowel symptoms during RT for prostate cancer. Most patients developed acute urinary symptoms during treatment, but the overall symptom burden was relatively modest. Bowel symptoms developed less frequently. The results of the present study provide data from the patient's perspective, which can be used by patients and physicians during the treatment decision-making process. In addition, as patient-reported outcome measurement becomes increasingly incorporated in clinical trials, the results of the present study may be helpful for future comparative effectiveness studies to examine whether newer treatments or technologies might improve upon these outcomes.
CONFLICT OF INTEREST
Paul Nguyen received a research grant from Varian Medical Systems.
- 17Simultaneous inferences and other topics in regression analyses-1, in Applied Linear Regression Models, Homewood, IL: Irwin Professional Publishing, 1983: 150–3, , .