Primary treatment of the prostate improves local palliation in men who ultimately develop castrate-resistant prostate cancer
Correspondence: Manish I. Patel, Urological Cancer Outcomes Centre, Sydney Medical School, University of Sydney, Australia.
- To determine whether local treatment of primary prostate cancer gives palliative benefit to men who later develop castrate-resistant prostate cancer (CRPC).
- Local treatments of primary prostate cancer are defined as radical retropubic prostatectomy (RRP) or external beam radiation therapy (EBRT).
Patients and Methods
- Patient records were reviewed in five different hospitals in Sydney, Australia, and 263 men with CRPC were identified.
- Eligible patients comprised men who had progressive disease during androgen deprivation therapy with castrate levels of testosterone.
- Clinical and pathological data were reviewed and evaluated using the chi-squared test and relative risk analysis to determine the relationship between previous local prostate treatment and complications secondary to local disease.
- The end-point was complications and morbidity attributed to cancer progression locally (i.e. from the prostate).
- Primary treatment of the prostate by either RRP or EBRT significantly reduces the incidence of local complications compared to no primary treatment (32.6% vs 54.6%; P = 0.001).
- RRP showed a significantly lower level of local complications compared to EBRT (20.0% vs 46.7%; P = 0.007).
- The most common local complications were bladder outlet obstruction (35.0%) and ureteric obstruction (15.2%).
- The present retrospective analysis supports the hypothesis that primary local prostatic treatment gives palliative benefit to men who later develop CRPC.
- RRP was associated with the lowest local complication rate experienced at the stage of metastatic disease.
Prostate cancer is the most common male urological malignancy in developed countries and has the highest age-specific mortality rate. With our ageing population, prostate cancer will continue to be a major health care concern because the incidence of prostate cancer increases with age .
For men who are diagnosed with advanced prostate cancer, quality of life is substantially affected not only by the effects of metastatic disease, but also by problems caused by local progression of the cancer. Moreover, local recurrence or progression is considered to be associated with increased morbidity and mortality [2, 3]. Therefore, focusing on the local control of cancer even in the advanced setting may potentially enhance quality of life, slow down the systemic progression of the disease and potentially provide a survival advantage [4, 5].
Several studies have investigated the impact of previous local prostatic treatments for primary prostate cancer in the setting of advanced disease. Among them, Thompson et al.  suggested a potential survival advantage in men with previous local treatment before the development of metastatic disease, although this was disputed by Halabi et al. . Few studies have investigated the relationship between previous local prostatic treatment and complications as a result of local disease progression when men later progress to a castrate-resistant state. We studied whether definitive local treatment of primary prostate cancer provides a palliative benefit to men who later develop castrate-resistant prostate cancer (CRPC).
Patients and Methods
In total, 263 men with metastatic CRPC treated between March 2006 and March 2011 were identified from an urological oncology database of patients from five different hospitals in Sydney, Australia (Westmead Public and Private, Hornsby, St George Private and Sydney Adventist Hospitals). Eligible patients were men who had progressive disease (defined by a rising PSA level or the presence of new metastases) despite treatment with androgen deprivation therapy (ADT) and with castrate levels of testosterone. Men were divided into three groups. Group 1 received previous local treatment of primary prostate cancer by radical retropubic prostatectomy (RRP; n = 45) with or without postoperative radiotherapy; group 2 received definitive external beam radiotherapy (EBRT; n = 45) (not adjuvant, salvage or palliative); and group 3 received no initial local prostate therapy (Nil; n = 173) (these patients were treated with watchful waiting or ADT). All eligible patients were treated with ADT either before or on the diagnosis of metastatic prostate cancer. RRP was performed by consultant urologists from community and academic hospitals in New South Wales, whereas EBRT was conducted by radiation oncologists in major teaching hospitals or high-volume private hospitals in Sydney. The decision to administer a particular type of local treatment depended on the subjective clinical judgments of the treating urologists, oncologists and radiation oncologists at the time of the initial diagnosis.
The end-point of the present study was the development of local complications and morbidity attributed to local cancer progression (i.e. from the prostate) after systemic treatment with androgen deprivation. Complications of local disease were defined as ureteric obstruction, bladder outlet obstruction (BOO), haematuria, pelvic pain or prostatitis. Although ureteric obstruction was confirmed by radiological imaging, the other complications were recorded on the basis of the patient's symptoms and were evaluated by the corresponding oncologists. The diagnosis of prostatitis was based on the prolonged symptoms (i.e. urinary frequency and lower abdominal/back pain). Ureteric obstruction was subdivided into hydronephrosis (either unilateral or bilateral) or the requirement of stent insertion. BOO was also subdivided into acute urinary retention or the need for TURP.
Clinical and pathological data were reviewed and evaluated using the chi-squared test, and the relative risk between previous local prostate treatment and complications secondary to local disease recurrence was examined. Statistical analyses were performed using SPSS, version 17 (SPSS Inc., Chicago, IL, USA). Statistical analyses and procedures modelled the Gleason score and the time in years between the initial diagnosis and the diagnosis of CRPC as categorical variables. Age at diagnosis of prostate cancer and age at diagnosis of CRPC were modelled as continuous variables. Any association between local prostatic treatment and the rate of local complications was assessed by the chi-squared test and the 95% CI was determined.
Of the 263 men evaluated in the present study, 83 patients had not died. The mean (range) age at diagnosis of prostate cancer in the study population was 67.6 (48.2–91.1) years and the mean (range) time between the initial diagnosis of prostate cancer and a diagnosis of CRPC was 3.7 (0.1–21.7) years. Patient and pathological characteristics are listed in Table 1.
Table 1. Patient characteristics
|Median age at diagnosis (years)||63.2||67.9||70.2|| |
|Median age at CRPC (years)||70.5||75.9||74.6|| |
|Clinical stage at diagnosis, n|
|Gleason score, n|
|Time to CRPC (years) , n|
|Maximal metastatic involvement at CRPC, n|
Group 1 comprised 45 men. They all presented with localized disease. Of the 45 men who had RRP as their primary treatment for prostate cancer, nine (20%) received adjuvant EBRT after RRP and five (11%) received delayed salvage EBRT.
Group 2 comprised 45 men. In this group, 38 (84%) presented with localized disease; four (9%) presented with locally advanced disease; and three (7%) presented with metastatic disease but still received definitive radiation to the prostate gland.
Group 3 comprised 173 men. Of these, 113 (65%) presented with metastatic disease and started ADT immediately and 44 (25%) initially had localized disease and were managed by watchful waiting before developing advanced disease and the subsequent commencement of ADT. Of the 173 men who did not have definitive treatment of their prostate gland (group 3), 34 (20%) required delayed palliative EBRT because of local prostatic symptoms.
Among patients with a documented Gleason score (n = 179), 65% had a high Gleason score (defined as a Gleason score ≥ 8) and most (76%) of those patients received ADT alone.
When the men (n = 263) developed CRPC, 74 (28%) received no chemotherapy for various reasons, including refusal of chemotherapy, poor performance status, palliation or entry into clinical trials.
Patients who received local therapy to the prostate (groups 1 and 2) showed a longer mean time to develop CRPC (8 years) compared to patients in group 3 (4 years). This finding reflects the differences among the groups with respect to clinical stage at presentation and Gleason scores. When men reached CRPC, most (99%) of the study population showed radiological evidence of bony metastases, whereas some patients also had nodal and/or visceral metastases.
Table 2 shows that primary treatment of the prostate by either RRP or EBRT (groups 1 and 2) significantly reduced the incidence of subsequent local complications compared to that in patients who had no primary treatment (group 3) (32.6% vs 54.6%; P = 0.001). The estimated relative risk of local complication among men who had no local prostatic treatment compared to that in patients receiving local treatment is 1.68 (95% CI, 1.21–2.33). In addition, RRP (group 1) showed a significantly lower probability of local complications compared to EBRT (group 2) (20.0% vs 46.7%; P = 0.007). EBRT (group 2) also showed a trend of less local complications compared to no local treatment (group 3), although the result was not statistically significant (P = 0.4). The most common local complications were BOO and ureteric obstruction.
Table 2. Type of local treatment and complications secondary to local disease
|Group 1: RRP (n = 45)†||20.0%||6 (13.3)||2 (4.4)||1 (2.2)||0 (0)||0 (0)||1 (2.2)||0 (0)|
|Group 2: EBRT (n = 45)*||46.7%||8 (17.8)||16 (35.6)||3 (6.7)||2 (4.4)||1 (2.2)||.0%||2 (4.4)|
|Group 3: Nil (n = 173)†||54.3%||26 (15.0)||74 (42.8)||8 (4.6)||1 (0.6)||1 (0.6)||8 (4.6)||8 (4.7)|
|Total||46.8%||40 (15.2)||92 (35.0)||12 (4.6)||3 (1.2)||2 (0.8)||9 (3.4)||10 (3.8)|
Group 3, with no previous local prostatic treatment, had the highest incidence of local complication of which BOOwas also most common (42.8%; P < 0.001). Although ≈33% of the men with EBRT developed BOO, there were two men (4.4%) who developed BOO in the RRP group. As shown in Table 3, the groups with EBRT (15.6%) or no local treatment (12.1%) have a higher probability of developing acute urinary retention (P = 0.18) and require at least one TURP (P = 0.007). Although no patient required TURP in the RRP group, two patients developed BOO and acute urinary retention as a postoperative complication and required bladder neck contracture incision.
Table 3. Presence of local treatment and complications secondary to local disease
|Group 1 (RRP)||2 (4.4)||0 (0)|
|Group 2 (EBRT)||7 (15.6)||7 (15.6)|
|Group 3 (Nil)||21 (12.1)||39 (22.5)|
|Total||30 (11.4)||46 (17.5)|
The incidence of ureteric obstruction was not substantially different between the groups (P = 0.9). Ureteric obstruction rates were 13.3% in group 1 (RRP) compared to 17.8% and 15.0% for patients in groups 2 and 3, respectively (Table 2). Not all patients with evidence of ureteric obstruction required intervention. There were six patients requiring percutaneous nephrostomy tubes, with one in group 1 (RRP), one in group 2 (ERBT) and four in group 3 (Nil). The rates of stent insertion and bilateral hydronephrosis were also lower in the RRP group, although they were not statistically significant (P = 0.2 and 0.8, respectively) (Table 4).
Table 4. Type of local treatment and complications as a result of ureteric obstruction
|Group 1 (RRP)||6 (13.3)||2 (4.4)|
|Group 2 (EBRT)||8 (17.8)||7 (15.6)|
|Group 3 (Nil)||24 (13.9)||12 (6.9)|
|Total||38 (14.4)||21 (8.0)|
Overall, the incidence of heavy haematuria was higher in the men with no local treatment (group 3) (4.6%). The incidence of prostatitis was negligible. Pelvic pain rates were 6.7% in group 2 (EBRT) compared to 2.2% and 4.6% for patients in groups 1 and 3, respectively (P = 0.8) (Table 2).
The findings of the present study show that patients who receive local treatment of primary prostate cancer by either RRP or EBRT have significantly less local complications when they later develop CRPC compared to those patients who do not receive treatment of the prostate at diagnosis. RRP showed the most effective palliation effect, with lower incidences of BOO compared to the EBRT and no local treatment groups (Table 2). RRP did not show any benefit over the other groups in terms of ureteric obstruction, pelvic pain, haematuria or urinary incontinence. The lack of benefit of RRP (group 1) in preventing ureteric obstruction is possibly because most obstructions were a result of lymphadenopathy for which data are unavailable.
There are a number of retrospective studies that have examined the utility of RPR with respect to reducing local complication in patients found to have node positive disease at the time of surgery. A secondary analysis of a randomized double-blind prospective phase III study was provided by Thompson et al. . The initial study compared orchiectomy plus the antiandrogen flutamide with orchiectomy plus placebo in men with metastatic carcinoma of the prostate. Of the 1286 men in the original trial, there were 919 men with no previous local prostatic treatment and 367 men who had received local therapy to the prostate (148 men with RRP and 219 men with EBRT). The impact of previous local therapy to the prostate on cancer death and disease progression was investigated. Thompson et al.  reported a signicant reduction in death, as well as disease progression, in those men with previous RRP compared to no previous local therapy to prostate. However, an increased risk of death was reported in those men who underwent EBRT . Cheng et al.  conducted a retrospective study on the effect of RRP and surgical castration on complications in patients with node-positive metastatic prostate cancer compared to those with EBRT and surgical castration. Cheng et al.  reported a lower incidence of ureteric obstruction and BOO in patients who were treated with RRP and surgical castration, although the results were not statistically significant. Schmeller et al.  showed that the addition of RRP to ADT in patients with node-positive metastatic disease offered a lower incidence of BOO requiring TURP, although there was no reduction in local recurrence rate. Gjertson et al.  reported that none of 24 patients found to be node positive after RRP and bilateral lateral pelvic lymph node dissection subsequently developed local complications (defined as gross haematuria, urinary retention or hydronephrosis) during a median follow-up period of 74 months. Grimm et al.  reported a benefit with regard to both local progression and palliation (i.e. TURP) in node-positive metastatic prostate cancer treated with RRP compared to ADT alone. A retrospective study conducted in 1990 at the Johns Hopkins Hospital  examined 120 consecutive patients with node-postive metastatic prostate cancer who were treated with ADT, EBRT or RRP. The incidence of symptomatic local progression, in terms of BOO, ureteric obstruction and haematuria, was much lower in the patients treated with RRP. The rate of local progression requiring surgical intervention was 3% vs 43% and 46% in the groups with EBRT and ADT, respectively. Among those patients treated with RRP, there was no TURP performed, although one patient required bladder neck incision as a result of postoperative complication. The present study comprised patients who were treated with RRP for presumed localized prostate cancer. The early eradication of the diseased organ may contribute to the difference in the later incidence of local complication, especially BOO.
Although, in the present study, EBRT showed a lower level of local complications compared to no local treatment, it was also associated with a higher incidence of ureteric obstruction, pelvic pain and prostatitis compared to RRP. Widmark et al. . conducted an open randomized trial comparing ADT with and without EBRT. It was found that the group with ADT and EBRT had a slightly higher incidence of urethral stricture and urinary urgency at 5 years after treatment compared to the ADT-only group. However, a more recent prospective randomized trial of locally advanced prostate cancer conducted by Warde et al.  reported a lower rate of local progression in patients receiving ADT + EBRT (14/603 patients) compared to ADT alone (97/602 patients), with low rates of toxicity in both arms.
The present retrospective study was limited by imperfect comparisons between the treatment groups. There may be discrepancies among the groups for variables that are not available for analysis, such as different tumour stages at diagnosis and patient characteristics. Other limitations include its design as a non-randomized retrospective comparison and the subjective reasons for choosing one primary treatment over another. In addition, a validated questionnaire for quality of life and morbidity was not provided.
The stage and co-morbidity information at the time of initial treatment was unavailable and was not analyzed. It is anticipated that those receiveing RRP or primary EBRT were at a localized stage compared to those who had watchful waiting, of whom some would have had metastatic disease.
These conclusions should be considered with caution and as a guide for future trials in this field.
Radical prostatectomy is normally reserved for men with localized prostate cancer and a high chance of cure. Increasingly, RRP and multimodality therapy utilizing RRP is being used for high-risk patients who have a lower chance of cure. The findings of the present study support the concept that, even if a cure is not achieved in these patients, there is still significant palliative value in undergoing surgery. This has a higher probability of becoming an even more significant issue in the context of the survival benefits seen with the current treatment options for patients with metastatic CRPC . However, a prospective trial would be required to validate this suggestion.
In conclusion, the present retrospective analysis supports the hypothesis that primary local prostatic treatment provides a palliative benefit to men who later develop CRPC. RRP was associated with the lowest local morbidity rates experienced at CRPC stage and may achieve good local palliation in men who ultimately are not cured by this treatment modality.
Conflict of Interest
androgen deprivation therapy
bladder outlet obstruction
castrate-resistant prostate cancer
external beam radiation therapy
radical retropubic prostatectomy