A randomized trial of external beam radiotherapy versus cryoablation in patients with localized prostate cancer

Quality of life outcomes




A recent randomized trial to compare external beam radiation therapy (EBRT) to cryoablation for localized disease showed cryoablation to be noninferior to external beam EBRT in disease progression and overall and disease-specific survival. We report on the quality of life (QOL) outcomes for this trial.


From December 1997 through February 2003, 244 men with newly diagnosed localized prostate cancer were randomly assigned to cryoablation or EBRT (median dose 68 Gy). All patients received neoadjuvant antiandrogen therapy. Patients completed the EORTC QLQ C30 and the Prostate Cancer Index (PCI) before treatment and at 1.5, 3, 6, 12, 18, 24, and 36 months post-treatment.


Regardless of treatment arm, participants reported high levels of QOL with few exceptions. cryoablation was associated with more acute urinary dysfunction (mean PCI urinary function cryoablation = 69.4; mean EBRT = 90.7; P < .001), which resolved over time. No late arising QOL issues were observed. Both EBRT and cryoablation participants reported decreases in sexual function at 3 months with the cryoablation patients reporting poorer functioning (mean cryoablation = 7.2: mean EBRT = 32.9; P < .001). Mean sexual function score was 15 points lower at 3 years for the cryoablation group and 13% more of the cryoablation men said that sexuality was a moderate or big problem.


In this randomized trial, no long-term QOL advantage for either treatment was apparent with the exception of poorer sexual function reported by those treated with cryoablation. Men who wish to increase their odds of retaining sexual function might be counseled to choose EBRT over cryoablation. Cancer 2009. © 2009 American Cancer Society.

The optimal strategy for managing localized prostate cancer has not been established and currently includes several potential options: active surveillance, radical prostatectomy, external beam radiation (EBRT), brachytherapy, and cryoablation.1, 2 Other modalities such as photodynamic therapy and high-intensity focused ultrasound are still being investigated.3-;5 No single modality has proven superiority for survival, and, consequently, differences in quality of life (QOL) become important considerations in determining clinical benefit.6, 7 Unfortunately, few published randomized clinical trials compare QOL outcomes8, 9 and current knowledge on the impact of the modalities upon QOL is largely based on nonrandomized designs.10 This seminal report used a randomized study design to prospectively evaluate QOL outcomes for 2 primary treatments of prostate cancer: EBRT and cryoablation.

Since the mid 1990s, there has been resurgence in the use of cryoablation as a definitive treatment for prostate cancer. Publications of prostate cryoablation have shown promising results with 5-year biochemical control rates ranging from 52%-92%.11, 12 We previously reported a phase 2 cryoablation series with 3-year QOL outcomes that were comparable to other treatment options.13

Herein, we present the QOL results of our single institution, randomized trial comparing EBRT and cryoablation for clinically localized prostate cancer. The cancer control outcomes, which showed cryoablation to be noninferior to EBRT in disease progression and overall disease-specific survival, are reported separately.14 QOL outcomes were included as secondary measures and were collected prospectively throughout the duration of the trial. We had hypothesized that while there may be differences in specific domains of QOL immediately after treatment, over the long-term, there would be no differences between the 2 groups of participants.


Patient Eligibility

Eligibility criteria included: histologically proven adenocarcinoma of the prostate; biopsy staged T1-3; no evidence of nodal or distant metastases; pretreatment prostate-specific antigen (PSA) ≤20; and gland volume ≤60 mL. We use the term “biopsy staged” rather than clinical stage because it more accurately reflects how patients were staged. Biopsies comprise bilateral sampling of base, mid and apical zones of the prostate and both neurovascular bundles and seminal vesicles. A single reference pathologist reviewed all biopsies. As there was no unanimity among the trial physicians regarding clinical staging, the biopsy information was incorporated into the staging using the American Joint Committee on Cancer 1992 nomenclature. Therefore, if a patient did not have palpable abnormalities but had positive biopsies in the gland bilaterally, he was considered stage bT2C. If the biopsies showed microscopic extraprostatic extension or seminal vesicle involvement, the staging was denoted as bT3c. Patients were excluded if they had: 1) clinically bulky T3 tumor; 2) undergone prior pelvic radiation; 3) received previous androgen deprivation therapy (ADT) at any time; or 4) if transurethral resection of prostate had been performed in less than 3 months.

Local urologists and radiation oncologists identified potential candidates. Recruitment was facilitated by the use of educational sessions attended by patients and family, with presentations by urologists, radiation oncologists, medical oncologists, and psychologists. Before study entry, candidates underwent a physical examination and a transrectal ultrasound scan (TRUS)-guided 10-core biopsy. Patients with a Gleason score ≥8 underwent lymph node dissection to rule out nodal involvement.

Treatment Plan

The original protocol, and the subsequent hormone and radiation dosing changes, as described below, were approved after institutional review by the Conjoint Health Research Ethics Board at the University of Calgary. All patients provided written informed consent.

ADT using luteinizing hormone-releasing hormone (LH-RH) agonist was provided to all patients before definitive local treatment. Initially, a single 3-month depot of LH-RH agonist was given, followed by local treatment between 90 days and 120 days after the first injection. In 2001, the local standard of care changed to include 6 months of ADT pre-EBRT using 2 consecutive 3-month depots of LH-RH. The protocol was amended to reflect this change with local treatment commencing between 180 days and 210 days after the first injection.

All patients randomized to the EBRT study arm were treated with a standard 4 field box technique (2 Gy/day, 5 days/week) using high-energy megavoltage x-rays of ≥10 MV. The prescribed radiation dose was 68 Gy. The dose was increased to 70 Gy in early 2000 and finally to 73.5 Gy in late 2002 in response to changing standards of practice.15 The prostate gland and the seminal vesicle base were included in the clinical target volume (CTV). If there was a positive biopsy of the seminal vesicles (ie, stage bT3c), then they were included entirely in the CTV. Elective pelvic nodal radiation was not given. CT simulation was done on all patients.

The cryoablation procedure used TRUS guidance with argon/helium third-generation equipment was used in all cases.16 Thermosensor monitoring, urethral warming, and saline injections were routinely applied to separate anterior rectal wall from the posterior prostate, and 2 freeze-thaw cycles were used in all cases. Unlike EBRT, the cryoablation procedure may be repeated. Therefore, the protocol allowed cryoablation patients with early PSA failure to undergo an adjuvant repeat cryoablation procedure within the first 6 months without considering the patient a failure. Repeat cryoablation for rising PSA was performed only if a repeat prostate biopsy was positive.

Post-treatment follow-up included PSA measurements, digital rectal examination (DRE), and QOL assessment. Routine radiological assessment was not done, and follow-up bone scan was conducted only for clinical indication. The protocol mandated that all patients undergo 10 core biopsy 36 months post-treatment.

Quality of Life Assessment

All patients were administered the EORTC QLQ C3017 and the UCLA Prostate Cancer Index18 before randomization and at 1.5, 3, 6, 12, 24, and 36 months.

The EORTC QLQ C3017 questionnaire was developed by the European Organization for the Research and Treatment of Cancer, Quality of Life Study Group and includes several domains that, together, are important in assessing QOL. It represents a “core” questionnaire intended for use with patients having any type of cancer and, therefore, addresses questions of a general nature. Domains include: physical activity, emotional state, social interaction, global health status/QOL, and several symptom scales, including nausea, sleep disturbance, loss of appetite, constipation, diarrhea, and financial impact. The reliability coefficients ranged from 0.54 to 0.86 before treatment to 0.522 to 0.89 during treatment among 305 patients from 13 countries. The EORTC QLQ C30 exceeded the acceptable reliability coefficient of greater than is 0.7 in 4 of 9 scales before treatment and in 8 of 9 scales after treatment.17 The EORTC QLQ C30 has been the most widely used measure of QOL in prostate cancer trials.19

The Prostate Cancer Index18 is an instrument that assesses function and bother in 3 organ systems: sexual, urinary, and bowel. The instrument was originally developed for a large-scale QOL study of men treated for localized prostate cancer18 as a result of 3 focus groups with prostate cancer patients and spouses. Litwin et al18 found it to be sensitive in distinguishing between radical prostatectomy, and radiation treatments, and conservative management for localized prostate cancer.

Statistical Considerations

The assessment of compliance was based upon the number of completed forms divided by the number of expected forms, accounting for deaths at each time point. If a patient completed more than 1 valid questionnaire for a given time window, the questionnaire that was completed first during this period was retained for analysis.

Descriptive statistics and graphical displays created in Stata 9 (StataCorp LP, College Park, Tex.) were used to illustrate changes in patient's QOL over time. The data was analyzed on an available case basis and missing data was not imputed. A t test was performed at 3 months and 36 months for all domain and subscale scores of the EORTC-QLQ-C30, as well as the urinary, bowel, and sexual function subscales of the Prostate Cancer Index to look for significant short-term and long-term differences in QOL on those domains between cryoablation and EBRT patients.

Patients were assessed for potency, and the analysis of sexual function was restricted to those men who reported being potent at baseline. With approximately 100 patients left at 36 months who had been potent at baseline, there was 70% power to detect a 10-point different in sexual function score with alpha = .05 and standard deviation (SD) = 20 (or 96% power to detect a 15-point difference or 100% power to detect 20-point difference or greater).

With respect to urinary and bowel function, there was 77% power to detect a 5-point difference, 90% power to detect a 6-point different, and 97% power to detect a 7-point difference using n1 = 117, n2 = 114, alpha = .05, SD = 14.

Fisher exact test was used at each time point to examine statistically significant differences between treatment groups for the proportion of patients who reported 1) adverse symptoms, 2) problems with urinary, bowel, or sexual function, and 3) use of assisted and/or unassisted intercourse. Given the multiple testing required, caution needs to be taken when interpreting the results and p values are used for guidance only.



Between December 1997 and February 2003, 764 patients were assessed for eligibility and 244 were randomized. Thirteen patients did not receive their assigned treatment (Fig. 1). Of the 114 EBRT patients, 58 received doses of less than 70 Gy, whereas 56 received 70 Gy or higher. Neoadjuvant hormone therapy duration was 3 months in 160 patients and 6 months in 71 patients. One cryoablation patient underwent a second procedure within the 6 months after the initial treatment. Table 1 shows the characteristics of the study population by treatment assignment. Six patients have been lost to follow-up.

Figure 1.

Flow of patients assessed for eligibility including reasons for nonparticipation. RT indicates radiation therapy; RP, radical prostatectomy; WW, watchful waiting.

Table 1. Patient Demographics, Biopsy, and Clinical Variables
 Cryoablation, n=122EBRT, n=122
  • EBRT indicates external beam radiation therapy; PSA, prostate-specific antigen; IQR, interquartile range.

  • *

    Low risk, PSA<10 ng/mL, Stage<T2b, Gleason<7; Moderate risk, only 1 of the following: PSA≥10, Stage≥T2b, Gleason≥7; High risk, 2 or 3 of the following: PSA≥10, Stage≥T2b, Gleason≥7.

Age, y  
Biopsy stage, No. (%)  
 T2a22 (18.0)20 (16.4)
 T2b28 (23.0)23 (18.9)
 T2c49 (40.2)57 (46.7)
 T3a17 (13.9)18 (14.8)
 T3b,c6 (4.0)4 (3.3)
Gleason score, No. (%)  
 4-55 (4.1)2 (1.6)
 637 (30.3)42 (34.4)
 769 (56.6)65 (53.3)
 8-1011 (9.0)13 (10.7)
Baseline PSA, ng/mL  
 Median [IQR]8.1 [5.7-10.9]9.0 [6.6-12.5]
Risk category*, No. (%)  
 Low10 (8.2)10 (8.2)
 Intermediate36 (29.5)28 (23.0)
 High76 (62.3)84 (68.8)

Compliance With QOL Questionnaire Completion

The percentage of patients completing the questionnaires at each time point is shown in Table 2. Participants were highly compliant in completing QOL questionnaires with completion rates ranging from 100% to a low of 48% at 1.5 months (Table 2). The low completion at 1.5 months and 3 months was not a result of poor patient compliance but was a result of the variability in the follow-up schedules of the radiation oncologists.

Table 2. Quality of Life Questionnaire Compliance Over Time for the EORTC-QLQ-C30 and PCI, Including the Number of Expected and Completed Forms, the Percentage of Patients Completing Forms, and the Time Windows for Inclusion of Forms
Follow-up, mo01.53612182436
  1. EBRT indicates external beam radiation therapy; PCI, Prostate Cancer Index.

Time window, moPrior to treatment(0-2)(2-4.5)(4.5-9)(9-15)(15-21)(21-30)(30-42)

Quality of Life


The EORTC-QLQ-C30 domain and subscale scores are shown in Figure 2. There were slight changes in scores for both treatment groups at 6 weeks. Single-item symptom questions are shown in Table 3. There did not appear to be an increase in symptoms after either treatment with the exception of insomnia, and a greater proportion of patients in the EBRT group experienced this symptom at 3 months and 24 months compared with the cryoablation group (Table 3). There were no visual differences in mean scores (Fig. 2) or statistically significant differences in proportions (Table 3) at 3 years across any of the domains or subscales scores.

Figure 2.

Mean scores for the 5 functional scales (physical, role, emotional, cognitive, social),* global health status scale, and 3 symptom scales (fatigue, nausea/vomiting, pain)** of the EORTC-QLQ-C30, by treatment group. * A higher score indicates better function. ** A higher score indicates a greater severity of symptoms.

Table 3. Percentage of Patients Over Time Reporting Adverse Symptoms (Quite a bit or a lot of) as Listed on the EORTC-QLQ-C30
Follow-up, mo01.53612182436
  • EBRT indicates external beam radiation therapy.

  • *

    Differences between cryoablation and EBRT that are statistically significant at P < .05

Shortness of breath        
Appetite loss        

Prostate Cancer Index

The average urinary, bowel, and sexual function scores of the Prostate Cancer Index for the 2 treatment groups are shown in Figure 3. Visually, there are differences immediately after treatment between the 2 groups, with a larger decline in the cryoablation patients across all 3 domains. At 3 months post-treatment, cryoablation patients experienced lower urinary function scores (mean cryoablation = 69.4; mean EBRT = 90.7; P < .001), similar bowel function scores (mean cryoablation = 83.7; mean EBRT = 81.4; P = .366), and the cryoablation patients experienced lower sexual function scores (mean cryoablation = 7.2; mean EBRT = 32.9; P < .001).

Figure 3.

Mean scores for the urinary, bowel, and sexual function scales of the UCLA Prostate Cancer Index, by treatment group. Note: Mean sexual function score is restricted to those men who reported being potent at baseline.

At 36 months post-treatment, the EBRT patients experienced slightly lower urinary function scores (mean cryoablation = 93.0; mean EBRT = 88.6, P = .043), there was no difference in bowel function scores (mean cryoablation = 88.1; mean EBRT = 84.1, P = .092), and the cryoablation patients experienced lower sexual function scores (mean cryoablation = 16.0; mean EBRT = 36.7; P < .001). Patients in both groups recovered back to their baseline levels of urinary and bowel function by 36 months post-treatment. However, both groups of patients did not recover in terms of their sexual function.

The proportion of patients reporting a significant amount of urinary bother increased in the cryoablation group from 1% at baseline to 29.3% at 6 weeks after surgery, with no long-term effects at 36 months (Table 4). There was a short-term difference in proportions between the EBRT group and cryoablation group; however, there were no long-term differences in urinary bother (Table 4). There was a slight increase in the patients who reported bowel bother in the EBRT group, from 2.6% at baseline to 10.3% at 36 months after treatment reporting this problem (Table 4). Moderate to severe problems with sexual function were reported by about 12.5% of the men in both groups at baseline, with the percentage increasing to over 53% in the cryoablation group and over 34% in the radiation therapy group reporting this problem at 36 months of follow-up (Table 4).

Table 4. Percentage of Patients Reporting a Moderate or Big Problem With Urinary, Bowel, or Sexual Function
Follow-up, mo01.53612182436
  • EBRT indicates external beam radiation therapy.

  • *

    Differences between cryoablation and EBRT that are statistically significant at P < .01.

  • Differences between cryoablation and EBRT that are statistically significant at P < .05.

Urinary bother        
Bowel bother        
Sexual bother for patients potent at baseline        

At the time of recruitment into the study, 113 men reported being potent (56 cryoablation and 57 radiation therapy patients). Of these 113 men, 62% in the cryoablation study arm and 55% in the radiation therapy study arm were having unassisted intercourse at baseline (Table 5). By 36 months after baseline, 22% of cryoablation patients and 36% of radiation therapy patients were having unassisted or assisted intercourse (Table 5).

Table 5. Percentage of Patients Over Time Reporting Assisted and Unassisted Intercourse (Among Those Patients Who Were Potent at Baseline)
CharacteristicsNo. (%)No. (%)No. (%)No. (%)No. (%)No. (%)No. (%)No. (%)
  • EBRT indicates external beam radiation therapy.

  • *

    Differences between cryoablation and EBRT that are statistically significant at P < .01.

  • Differences between cryoablation and EBRT that are statistically significant at P < .05.

Follow-up, mo01.53612182436
Unassisted intercourse        
 Cryoablation34 (61.8)0 (0)0 (0)2 (3.9)2 (3.9)3 (5.9)2 (4.4)3 (6.7)
 EBRT30 (54.6)13 (44.8)*12 (27.3)*23 (41.1)*21 (39.6)*20 (41.7)*19 (38.0)*15 (30.0)*
Assisted intercourse        
 Cryoablation0 (0)0 (0)1 (2.0)3 (5.9)7 (13.7)6 (11.8)9 (19.2)8 (17.8)
 EBRT0 (0)0 (0)2 (4.6)2 (3.6)4 (7.4)5 (10.4)5 (9.8)5 (10.2)
Unassisted or assisted intercourse        
 Cryoablation34 (61.8)0 (0)1 (2.0)5 (9.8)9 (17.7)9 (17.7)10 (21.3)10 (22.2)
 EBRT30 (54.6)13 (44.8)*14 (31.8)*24 (42.9)*24 (44.4)*23 (47.9)*21 (41.2)18 (36.0)


There does not appear to be a long-term QOL advantage for either EBRT or cryoablation with the exception of sexual function. Men treated with EBRT reported better sexual function at 6 weeks post-treatment and continued to reported better function than the men treated with cryoablation at each assessment out to 36 months. Overall, although men did experience some declines in their QOL immediately after treatment, they had regained their baseline levels of QOL by 1 year, and, subsequently, they maintained this level. Neither treatment was associated with a late onset of difficulties. At the 6-week assessment, we observed increases in pain and declines in social function, urinary, bowel and sexual function with cryoablation participants showing slightly poorer function compared with EBRT. Fatigue increased at 6 weeks in both study arms with EBRT participants reporting slightly more.

Although EBRT patients reported slightly lower urinary function scores than the cryoablation patients (cryoablation = 93.0; mean EBRT = 88.6, P = .043), the difference was not thought to be clinically significant for 2 reasons. First, a clinically meaningful difference has not been established for the Prostate Cancer Index, and the difference observed in this study was smaller than the often-used ½ standard deviation threshold.20 Second, there were a similar proportion of EBRT patients reporting urinary bother at 36 months compared with baseline, indicating that, most likely, there were no long-term affects on urinary function in this group.

Sexual function was the one domain in which dramatic losses were reported at 6 weeks, while showing some recovery over time, and continued to be lower at 3 years for both study arms. It should be noted that we report data only for those men who were potent at baseline. The cryoablation men reported greater losses in function at 6 weeks and, at each assessment time, scored lower than the EBRT participants. Mean sexual function score was 15 pointes lower at 3 years for the cryoablation group, and 13% more of the cryoablation men said that sexuality was a moderate or big problem. At 3 years, only 22% of the cryoablation men reported having assisted or unassisted intercourse, whereas 36% of the EBRT men reported having unassisted or assisted intercourse. We did not see the declines over time in sexual function in the EBRT study arm as reported in other studies,21 but we did see improvements over time in sexual function in the cryoablation study arm as previously reported.13

It should be noted that all patients in this trial received ADT. It has previously been shown that patients receiving even limited duration ADT report worse long-term outcomes across multiple QOL domains compared with those who receive primary treatment alone.10, 22 Therefore, it might be the case that the QOL of the patients in this trial may have been exacerbated by ADT.

The current study is the first prospective report of QOL outcomes in a randomized clinical trial of 2 treatments for localized prostate cancer. There are only 2 other randomized trials comparing primary treatments for prostate cancer reporting QOL outcomes.8, 9 However, both these trials were cross sectional in design.

It is difficult to compare results across studies for several reasons. The other 2 trials did not report baseline levels of QOL and the exact time because treatment or randomization is not known. Akakura et al8 reported QOL outcomes of 22 patients who underwent radical prostatectomy and 24 patients who underwent EBRT. Similar to the current trial, all patients in that trial received ADT. However, patients included in that trial had extracapsular as well as bilateral disease. QOL was measured only at 1 time at least 3 years post-treatment. Unfortunately, premorbid QOL and equivalency of the 2 groups was not known. Response rate was less than 50% in both study arms. They concluded that there was less QOL disturbance for the EBRT patients.

Steineck et al9 evaluated the QOL of patients treated by radical prostatectomy or watchful waiting for men with localized prostate cancer. Men were asked to complete a mailed questionnaire at least 12 months after surgery and 14 months after randomization for the watchful waiting study arm. Questionnaires were returned by 166 men assigned to the surgery study arm for a response rate of 88% and by 189 of men assigned to the watchful waiting study arm for a response rate of 86%. It was concluded that although there were differences in the risks for sexual and urinary dysfunction, there were not differences in well being or QOL.

Some limitations of this study should be noted. Although the completion rate of the QOL questionnaires was exceptionally high for a study of this kind, the completion rate was only 48% at the 6 week assessment and 77% at the 3 month assessment for the EBRT group. There is always a concern that those missing patients may vary in some important way such as being too ill to complete the questionnaires. We have no reason to believe that there was any systematic bias as the lower rate was due to variability in the scheduling of follow-up visits of the treating radiation oncologists. Of course this only affects the short-term QOL results, not the overall or long-term findings.

It should be further recognized that this is a single institution study and the results may not generalize to other centers. cryoablation remains operator-dependent and complication rates have varied according to site, technique, and indications, although treatment is becoming more standardized.

In conclusion, both cryoablation and EBRT appear to offer high long-term QOL. Men who wish to increase their odds of retaining sexual function might be counseled to choose EBRT over cryoablation.

Conflict of Interest Disclosures

This research was support by a grant from the National Cancer Institute of Canada and the Calgary Health Region. This study was supported by the National Cancer Institute of Canada, the Alberta Cancer Board and the Calgary Health Region.

Dr. Bejoy Thomas is a postdoctoral fellow (2007-2009) funded by the Alberta Heritage Foundation for Medical Research, Alberta, Canada.

Bryan J. Donnelly is on the medical advisory board of Endocare.

John C. Rewcastle is a former employee of Endocare and is currently an employee of EDAP TMS, owner of Firma Medical.