Abstract
Aims
To demonstrate that radiotherapy (RT) is a valid alternative to surgery in men ≤70 years old with localized prostate cancer.
Methods
From 1988 to 2009, 214 patients with T1-2 N0 M0 prostate cancer were treated with RT. The effects of patient- and treatment-related risk factors on toxicity were investigated.
Results
Median follow-up was 105 months (range 14.2-180). The 5-, 10-, and 15-year biochemical relapse-free survival for all 214 patients was 80%, 61.9%, and 57.5%, respectively. In bivariate analysis, age (≤65 vs 65-70 years) was not a significant factor for biochemical relapse, while radiation dose was (p = 0.05) in multivariate analysis. Cancer-specific survival rates at 5, 10, and 15 years were 98.4%, 93.2%, and 69.7%, respectively. Median overall survival (OS) was 167 months (95% confidence interval 147.3-186.7). The OS rates at 5, 10, and 15 years were 91.8%, 75.8%, and 42.5%, respectively. Acute genitourinary (GU) and gastrointestinal (GI) toxicities occurred in 105 (49%) and 98 patients (45.8%), respectively, with only 2 cases of grade III GI toxicity. Late GU and GI toxicities occurred in 17 (7.9%) and 20 (9.3%) patients, respectively, with 1 grade III GI toxicity and 2 grade III GU toxicities. Risk factors for late toxicity were age and RT dose and technique, which were unrelated to acute toxicity.
Conclusions
Age ≤70 years does not consistently confer a negative prognosis for localized prostate cancer. Radiotherapy appears to be a viable alternative to surgery, offering excellent long-term cancer control.
Introduction
Treatment of early-stage prostate cancer has long been controversial. Optimal therapy remains undefined due to lack of well-designed randomized trials comparing treatment modalities. Outcomes after radical prostatectomy (RP) and radiotherapy (RT), which are both offered for localized prostate cancer, are similar, owing to modern RT techniques that deliver higher doses to target volumes, sparing organs at risk (1, 2). Young, healthy men are, however, being preferentially managed with RP, as most physicians consider it standard treatment, which could reflect an underlying assumption by patients and physicians that RP offers better long-term disease control (3).
Who are considered young patients with prostate cancer? Some authors selected patients under 50 years of age, while others reported patients were young if they were aged under 65 (4-9). In some studies, young age was identified in itself as a negative prognostic factor as younger men were at relatively increased risk for long-term prostate cancer recurrence because of their longer life expectancy (9). Sometimes disease in younger patients was aggressive, with features indicating poor prognosis, but the majority of studies that described this aggressive type of disease were conducted before the prostate-specific antigen (PSA) era, when prostate cancer was diagnosed at more advanced stages.
The present retrospective study was conducted with the aim to assess outcomes after RT in men under 70 years of age with localized prostate cancer as, in clinical practice, these men are preferentially treated with RP (6, 7) rather than RT and conflicting data were provided by a few series analyzing outcomes after RT alone (10-12).
Methods
Patients
We retrospectively analyzed 214 patients aged ≤70 years (median 68, range 51-70) with clinical stage T1-2 N0 M0 prostate cancer who were treated with radical RT between January 1988 and December 2009. All patients had refused RP or were ineligible because of counterindications. All had given their informed consent to their data being used in this retrospective study. Prostate adenocarcinoma was diagnosed by prostate needle biopsy in 190 patients (88.8%) and by transurethral resection of the prostate in 24 (11.2%). Gleason scores were available for 169 (78.9%) patients. The pretreatment work-up in all included history and clinical examination, serum PSA levels, and a transrectal prostatic ultrasound scan; prostate magnetic resonance imaging (MRI) scan was performed in later patients in the series, according to accrual date. Bone scintigraphy and abdominal and pelvic computed tomography (CT) scans were performed in patients with adverse prognostic factors. In addition to RT, diverse schedules of hormone therapy (HT) consisting of a luteinizing hormone-releasing hormone analogue, an antiandrogen, or both were administered to 146 patients for a median of 5.5 months (range 1-36 months).
Radiotherapy
Doses and technical modalities changed over the decades, with the dose to the prostate gradually increasing from 60 Gy to 76 Gy: total dose was 60-68 Gy in 1.8-2 Gy daily fractions in 48 patients (22.4%) and 70-76 Gy in 166 (77.6%). Seminal vesicles received 60 Gy prophylaxis when Roach formula (13) identified a >15% risk of involvement. In 62/214 patients (29%) with adverse prognostic factors, pelvic lymph node prophylaxis was administered as 45-50.4 Gy in 1.8 Gy daily fractions, followed by a boost to the prostate.
A total of 86 patients (40.2%) received conventional RT (2D RT) to the prostatic volume with 4 customized shielded fields (box technique) or with rotational technique (14). A total of 128 patients (59.8%) received 3- or 5-field 3D conformal RT (CRT) to the prostate gland ± the seminal vesicles (clinical target volume [CTV]) according to Roach formula. The planning target volume consisted of CTV plus 1 cm in all directions except for the posterior, where it was 5 mm, to reduce the risk of rectal toxicity. Immobilization devices were used.
Follow-up
All patients were monitored weekly during RT for acute side effects, and checked up every 3 months for the first year, every 6 months until the fifth year, and yearly thereafter. Patients who did not attend check-ups regularly were phoned and interviewed about symptoms and disease status. Late side effects, disease recurrence, and serum PSA were monitored. Three consecutive PSA increases after a nadir had been achieved with RT were defined as biochemical relapse (15). Biochemical failure was recorded as the time midway between the nadir and the first rising PSA. Acute and late toxicities were graded according to Radiation Therapy Oncology Group/European Organization for Research and Treatment of Cancer (RTOG/EORTC) score (16).
Statistical analysis
Patients were stratified into 2 groups according to age: ≤65 years and >65 years (4, 5).
The χ2 test examined bivariate associations between patient characteristics and toxicity data. Survival was measured from the end of RT and patients were censored at their last follow-up or when the event occurred.
Univariate analysis estimated time-related outcome measures. Survival curves were calculated using the Kaplan-Meier product-limit method followed by log-rank test to evaluate differences in expected event probability between groups. The Cox proportional hazard regression model was used for multivariate analysis. Risk factors for biochemical relapse, distant metastases, and cancer-specific survival (CSS) included age (≤65 vs 65-70 years), initial PSA (≤10 ng/mL vs >10 ng/mL), radiation dose (≤70 Gy vs >70 Gy), Gleason score (<7 vs ≥7), and PSA nadir (<1 ng/mL vs ≥1 ng/mL).
The impacts of age, radiation dose, RT technique (2D vs 3D), and HT on toxicity were also analyzed.
Statistical significance was set at p≤0.05. All p values were 2-sided. Statistical analyses were performed using SPSS (Statistical Package for the Social Sciences) release 20.0, August 2011 (SPSS Inc., Chicago, IL, USA).
Results
Patient demographics and clinical data are shown in Table I. The median follow-up was 105 months (range 14.2-180). Within 12 months of ending RT, PSA was <1 ng/mL in 173 patients (80.8%). Biochemical relapse occurred in 65 patients (30.4%) at a median of 46 months (range 4-257). Local relapse developed in 8 patients (3.7%), regional relapse in 2 (0.9%) and distant metastases in 28 (13%) at median of 71.5 months (range 13-166).
Patient characteristics
EBRT = external beam radiation therapy; HT = hormone therapy; iPSA = initial prostate-specific antigen; PSA = prostate-specific antigen; RT = radiotherapy.
Values are n (%).
The 5-, 10-, and 15-year biochemical relapse-free survival (bRFS) was 80% (95% confidence interval [CI] 74.4%-85.6%), 61.9% (95% CI 53.8%-70.1%), and 57.5% (95% CI 47.9%-67.1%), respectively (Fig. 1). Univariate analysis (Tab. II) showed that significant risk factors for biochemical relapse were initial PSA >10 ng/mL (p = 0.006) and PSA nadir (p = 0.023).

Biochemical relapse-free survival, metastases-free survival, and cancer-specific survival in 214 patients ≤70 years of age who were treated with radiotherapy for localized prostate cancer.
Univariate and multivariate analysis for biochemical relapse-free survival, cancer-specific survival, and metastases-free survival
CI = confidence interval; HR = hazard ratio; iPSA = initial prostate-specific antigen; PSA nadir = prostate-specific antigen within 12 months of follow-up; RT = radiotherapy.
When nadir PSA was <1 ng/mL, bRFS was 84.4% at 5 years (95% CI 78.8%-90.1%) and 66.7% at 10 and 15 years (95% CI 57.6%-75.7%). When nadir PSA was ≥1 ng/mL, bRFS was 60.9% at 5 years (95% CI 45.2%-76.6%), 43% at 10 years (95% CI 25.7%-60.3%), and 28.7% at 15 years (95% CI 8.8%-48.6%).
In Cox proportional hazards regression model, initial PSA and radiation dose emerged as independent factors (p<0.001 and p = 0.05, respectively) for biochemical relapse (Tab. II).
The metastases-free survival (MFS) rate was 94.4% at 5 years (95% CI 91.2%-97.6%), 83.9% at 10 years (95% CI 77.6%-90.2%), and 76.9% at 15 years (95% CI 68%-85.8%). Univariate analysis showed low RT doses (p = 0.037) and PSA nadir >1 ng/mL (p = 0.019) worsened prognosis. Cox regression analysis with distant metastases as the dependent variable showed that initial PSA >10 ng/mL was the only significant factor (p = 0.041) for poor prognosis (Tab. II).
At the last follow-up, 93 patients (43.5%) were alive and disease-free and 46 (21.5%) were alive with disease progression: 37 (17.3%) with biochemical relapse, 3 (1.4%) with local relapse, 1 (0.47%) with loco-regional relapse, and 5 (2.3%) with metastases. During follow-up, 72 patients (33.6%) died. Causes of death were disease progression in 25 (11.7%), another cancer in 8 (3.7%), and other causes (mainly cardiovascular diseases) in 39 (18.2%).
The mean CSS of all patients was 165 months (95% CI 159.7-170.3 months); the median was not reached. The CSS rate was 98.4% at 5 years (95% CI 96.7%-100%), 93.2% at 10 years (95% CI 88.9%-97.4%), and 69.7% at 15 years (95% CI 58.6%-80.8%). Univariate analysis showed that only PSA nadir was a significant factor for CSS (p<0.001). When nadir PSA was <1 ng/mL, CSS was 99.3% at 5 years (95% CI 98%-100%) and 97.2% at 10 years (95% CI 94%-100%). When nadir PSA ≥1 ng/mL, CSS was 91.6% at 5 years (95% CI 82.6%-100%) and 78.7% at 10 years (95% CI 64.6%-92.9%). No other covariate reached significance (Tab. II).
Mean and median overall survival (OS) were 143 (95% CI 136.1-150.1) and 167 (95% CI 147.3-186.7) months, respectively. The OS rate was 91.8% at 5 years (95% CI 88.1%-95.5%), 75.8% at 10 years (95% CI 69.1%-82.5%), and 42.5% at 15 years (95% CI 32.3%-52.7%).
Analysis of only the subgroup of 57 patients aged ≤65 confirmed the results of the whole series as far as regards bRFS: the 5-, 10-, and 15-year bRFS was 80.1% (95% CI 69.5%-90.6%), 68.3% (95% CI 55.2%-81.5%), and 63.8%, respectively (95% CI 48.8%-78.8%) (Fig. 2). The MFS rate was 90.6% at 5 years (95% CI 82.7%-98.5%), 74.3% at 10 years (95% CI 61.6%-87.1%), and 67.2% at 15 years (95% CI 52.2%-82.2%). The CSS rate was 98.0% at 5 years (95% CI 94.2%-100%), 91.5% at 10 years (95% CI 83.5%-99.5%), and 59.1% at 15 years (95% CI 41.8%-76.4%). Univariate analysis identified the PSA nadir as significant for metastases and CSS while in multivariate analysis the initial PSA level emerged as a significant risk factor for relapse (Tab. III).

Biochemical relapse-free survival, metastases-free survival, and cancer-specific survival in 57 patients ≤65 years of age who received radiotherapy for localized prostate cancer.
Univariate and multivariate analysis for biochemical relapse-free survival, cancer-specific survival, and metastases-free survival in subjects aged ≤65 years
CI = confidence interval; HR = hazard ratio; iPSA = initial prostate-specific antigen; PSA nadir = prostate-specific antigen within 12 months of follow-up; RT = radiotherapy.
Grade I-II acute genitourinary (GU) toxicity occurred in 105 patients (49.1%). There were no cases of acute grade III or IV GU toxicity. Late GU toxicity developed in 17 patients (7.9%), 2 of whom (0.9%) had grade III (Tabs. IV and V).
Acute RTOG radiation morbidity scoring criteria stratified by age
GI = gastrointestinal; GU = genitourinary; RTOG = radiation therapy oncology group.
RTOG/EORTC late radiation morbidity scoring criteria stratified by age
EORTC = european organization for research and treatment of cancer; GI = gastrointestinal; GU = genitourinary; RTOG = radiation therapy oncology group.
Acute gastrointestinal (GI) toxicity developed in 98 patients (45.8%). Grade I-II occurred in 96 patients (44.9%) and Grade III in 2 (0.9%). Late GI toxicity occurred in 20 patients (9.3%), reaching grade III in one (Tabs. IV and V).
Acute GU and GI toxicity was not linked to the development of late toxicities. Neither acute GU nor acute GI correlated with age, RT dose or technique, or HT. Late GU and GI toxicity correlated closely with 2D RT technique (p<0.0001), RT dose ≥70 Gy (p = 0.000), and age >65 (p = 0.019).
Discussion
In evaluating RT as first-line therapy in patients with intracapsular prostate cancer, the present study is, as far as we know, the first to report on early-stage prostate cancer in patients aged ≤70 years. This age group belongs to the younger part of the prostate cancer population (17).
Radical prostatectomy is preferential treatment because it is believed to offer a better chance of cure and leaves RT as a rescue therapy when biochemical or locoregional relapse occurs. In the present series of patients, who were not candidates for RP or who had refused surgery, RT alone successfully treated localized prostate cancer, irrespective of the patient's age.
Results after RT in young prostatic cancer patients were analyzed in few studies with divergent results, usually with smaller cohorts and shorter follow-ups than ours. Our 5-year bRFS rate was in line with other reports (4, 8, 10, 12), although comparisons are difficult because patient selection criteria and follow-ups were different and results were confused by additional HT in some cases. Some studies reported that age impacted on outcome (11) and others, like ours, did not observe any correlation (4, 10, 12). In fact, age did not impact significantly on any of our outcome measures (bRFS, MFS, CSS) and the probabilities of biochemical and distant MFS are in line with those reported for patients with the same characteristics who underwent RP (18).
Although the main limitation of the present study is its retrospective character, it provides evidence to support the use of RT alone in early-stage prostate cancer in patients under 70 years of age. Since the accrual period ran from 1988 to 2009, one may object that only about 10 patients were recruited per year, but our low accrual rate confirms that RP is preferred by the majority of patients and urologists. In any case, the present observations demonstrate that RT is a valid alternative when RP is counterindicated or refused.
In fact, we are able to report here for the first time that the 15-year probability of biochemical relapse after RT alone or with HT is 42.5% and that RT provided a mean of over 13 years cancer-free survival. Since many of our patients were treated before 2006, when Phoenix criteria were introduced (19), we used American Society for Therapeutic Radiology and Oncology (ASTRO) (15) criteria to define biochemical relapse. One might hypothesize that our outcome rates would be better with Phoenix criteria, but their application in this cohort is not feasible. The ASTRO criteria could mask some RT successes as some patients might have received salvage therapy after a normal increase in PSA due to androgen blockade withdrawal.
Despite biochemical relapse in approximately 50% of our patients, 46/65 (67%) are alive after rescue HT. Today rescue HT is not the only salvage therapy after RT and even RT itself may be offered. In patients with an isolated local recurrence, brachytherapy or external beam RT with the new techniques, such as tomotherapy or stereotactic body RT (SBRT), are reported to give good results (20, 21). Modern imaging methods like MRI and/or choline positron emission tomography-CT allow detection of local, regional, and distant relapses and are helpful in the definition of target volumes when RT is performed as rescue therapy (22). For oligometastatic regional or distant disease, high-dose RT may be proposed with or without HT. All these approaches may improve not only survival but also quality of life in prostate cancer patients.
In the present study, risk factors for biochemical relapse and metastases were analyzed in univariate and multivariate analysis, which were not performed for local and regional relapse, due to the low number of events. Outcomes were not stratified by risk categories because the cohort included relatively few patients at intermediate to high risk since this was a series of early-stage cancers, which are generally low risk.
As in other reports, a basal PSA value >10 ng/mL (23) and a nadir >1 ng/mL (24, 25) were risk factors for biochemical relapse in univariate analysis, while low RT doses together with PSA nadir >1 ng/mL emerged as risk factors in multivariate analysis. The RT doses gradually increased in the 11-year accrual period in the present study and, as higher doses are known to be associated with better outcomes in prostate cancer patients (12, 26), it was hardly surprising to see that a low RT dose was also linked to risk of metastases in our patients. Present results, referring to patients treated between 1988 and 2009, would have been improved by administering high-dose RT (as is our current policy), but this RT level only became safe with modern progress in RT techniques (intensity-modulated RT, volumetric RT, tomotherapy, image-guided RT). In the present series, 65 (30%) cases received doses over 70 Gy (72, 74, and 76 Gy in 3, 36, and 26 cases, respectively). Since all are alive and disease-free except for 4 who died of other causes and 1 of relapse, our survival rate confirms that high-dose RT improved outcome (4, 10, 12, 27). Indeed, Zelefsky et al (12) emphasized that high-dose RT was necessary for optimal outcomes and that patient age did not have a significant impact on long-term biochemical outcome after 3D CRT. Interestingly, Kupelian et al (28) and others (29, 30) observed that with RT dose ≥72 Gy no differences emerged in biochemical relapse-free survival rates after external beam RT or RP for localized prostate cancer.
In our series, the overall toxicity rate was low as the majority of patients received 3D CRT (128/214). Although acute toxicity was reported to be a significant independent predictor of late toxicity (31), with the risk of acute reactions depending on age, radiation dose, and technical modalities, in our cohort, no correlation was observed. Klayton et al (10) found no significant difference between age groups (395 patients ≤60 years vs 1773 patients >60 years) for early ≥ grades 2 and 3 GI or GU toxicity. Late GI toxicity (≥ grade 2) was slightly but significantly higher in older men, 10.1% vs 5.7% (p = 0.009). This result does not differ from our study with 5.1% late GI toxicity, which as expected was closely linked to high-dose RT (RT >70 Gy, p = 0.000), 2D techniques (p<0.0001), and age over 65 years (p = 0.012). Despite these links, the effect of age on RT-induced toxicity is not clear and might reflect individual biological differences (32).
In conclusion, although prospective studies using high-dose RT, modern apparatus, and careful patient selection are needed to confirm the present findings, radical RT offers men with early-stage prostate cancer excellent long-term disease control, showing that age does not always confer a negative prognosis.
Footnotes
Acknowledgment
The authors thank Dr. G.A. Boyd for English language editing.
Financial support: None.
Conflict of interest: None.
