Bethesda Handbook of Clinical Oncology, 2nd Edition



Prostate Cancer

James L. Gulley*

William L. Dahut

*Laboratory of Tumor Immunology and Biology, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland

Medical Oncology Clinical Research Unit, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland


Prostate cancer (CaP) is the most common noncutaneous malignancy occurring in American men and the second most frequent cause of cancer-related mortality in American men. The risk for developing CaP increases progressively with age, and although the median age at diagnosis is 72 years, there is no “peak” age. Black men have a lower median age at diagnosis than do white men. Incidence and mortality rates are higher in blacks than in whites.

Incidence and mortality rates were found to increase after the U.S. Food and Drug Administration (FDA) approved the use of prostate-specific antigen (PSA) screening in 1986. The age-adjusted mortality rates have now dropped below the rate in 1986, possibly because of decline in distant disease incidence (1).

Frequency of clinically aggressive disease varies geographically, although the frequency of occult tumors does not. This suggests a role of environmental factors in the etiology of CaP. Studies of Japanese immigrants to the United States show that the incidence of CaP increases after migration.


  • Age (incidence increases with age)
  • Family history (risk increases twofold with a first-degree relative)
  • Race (in the United States, the incidence is highest among blacks followed by whites, and finally by Asians)
  • Geographic location (lowest in Asia, high in Scandinavia and United States)
  • Dietary fat (putative but not definitive).


Screening for CaP is performed with PSA and/or digital rectal examination (DRE). There is much debate about screening asymptomatic men for CaP. This debate centers on whether biologically and clinically significant cancers are actually being detected at an early enough stage to reduce mortality, or conversely, whether cancers detected by screening would never have caused clinically significant disease if these had not been detected or treated. Autopsy series have shown the rate of occult CaP in men in their 80s to be approximately 75%, and more men die with, rather than from, CaP. A recent study showed a decrease in CaP-specific mortality in men undergoing radical prostatectomy (RP) compared to those who underwent


watchful waiting, but there was no difference in overall survival (OS). Thus, because local treatment is associated with significant morbidity, and no randomized controlled trial has convincingly shown a decrease in mortality from screening or early treatment, screening remains controversial.

Despite the controversy, there is widespread use of PSA for CaP screening in the United States. Advocates of screening recommend that annual screening should commence at the age of 50 for average-risk men and at the age of 40 for African American men and for men with a family history of CaP.


Chemoprevention trials examine the treatment of asymptomatic patients with therapeutic agents that may be capable of preventing cancer. The Prostate Cancer Prevention trial was a study that compared finasteride, a 5-α reductase inhibitor, to a placebo in the prevention of CaP in men older than 50 years who had a normal DRE and a PSA of 3.0 ng per mL or less (2). This trial had more than 9,000 men included in the final analysis. A surprisingly large number of patients in both arms developed CaP. There was a significant decrease in the incidence of CaP in the experimental arm, with 24.8% of placebo-treated men and 18.4% of finasteride-treated men being diagnosed with CaP (p < 0.001). There were several other differences seen between the two arms. Sexual dysfunction was more common in the finasteride-treated men, whereas urinary symptoms were more common in those receiving placebo. The finasteride group showed an increase in Gleason score (GS) of 7 to 10 tumors, with 6.4% of men being affected in comparison to 5.1% of men in the placebo group having this pathology. One possible explanation is that this is the effect of androgen-deprivation therapy (ADT) on the appearance of the pathologic specimen. Longer follow-up is needed to determine the clinical course of these patients.

A large-scale 2 × 2 factorial study compared dietary supplementation with vitamin E and/or β-carotene and/or placebo in male Finnish smokers (3). Although the Finnish trial showed an increase in lung cancer in the β-carotene arm, a secondary analysis showed that there were fewer CaPs and CaP-related deaths in those men who received vitamin E. Another separate, multi-institutional trial tested supplementation with selenium versus placebo in preventing skin cancer (4). In a secondary analysis, men who received selenium experienced only 37% as many CaPs as did men who received placebo, a statistically significant difference. A retrospective study of selenium levels in men with and without CaP also showed a protective effect in men who had the highest amount of selenium.

These interesting retrospective data have led to the development of a prospective placebo-controlled, 2 × 2 factorial study that is examining the effect of selenium and vitamin E in the prevention of CaP—the Selenium and Vitamin E Cancer Prevention Trial (SELECT) (5).


Ninety-five percent of CaPs are adenocarcinomas; other histologies (sarcoma, lymphoma, small cell carcinoma, and transitional carcinoma) are rarely found. Although visceral metastases or osteolytic bone metastases are found in a few patients with metastatic adenocarcinoma of the prostate, a careful examination of pathology should be performed to determine whether a nonadenocarcinoma variant is present in this setting because treatment regimens differ for these variants.

Adenocarcinoma arises in the peripheral zone of the prostate in approximately 70% of cases.

The GS (see Table 14.1) is obtained by examining the histologic architecture of the biopsy. Primary and secondary Gleason grades are given to the biopsy of prostate gland tissue obtained at the time of surgery. The primary grade denotes the dominant histologic pattern, and the secondary grade is the pattern that represents the bulk of the nondominant pattern or a focal high-grade area. Both primary and secondary grades range from 1 (well differentiated) to 5 (poorly differentiated).


The two grades are added together, resulting in the GS (range 2 to 10). Total GS of 2 to 4, 5 to 6, 7, and 8 to 10 represent well-differentiated, moderately differentiated, moderately poorly differentiated, and poorly differentiated tumors, respectively. However, there is a growing belief that the highest score is most predictive of clinical outcome. There is often a change in GS at the time of RP, with 20% of cases being upgraded to a higher score.

TABLE 14.1. Risk of Metastatic Disease

Gleason score

Risk of developing metastatic disease (%)








Signs and Symptoms

Men with local or regional disease can be asymptomatic or can have lower–urinary tract symptoms similar to that of benign prostatic hypertrophy (BPH). Occasionally, men with regional disease have hematuria. The symptoms of the presence of metastatic disease include bone pain or weight loss, and, rarely, spinal cord compression. With the advent of widespread screening with PSA, most men are asymptomatic at the time of diagnosis.

Workup and Staging

CaP is usually detected by an abnormal PSA and/or DRE, followed by a transrectal ultrasound with core biopsy (generally 8 to 12 cores). Historically, a PSA level of >4 ng per mL was used as a threshold for biopsy, but recent data suggest that cancers can be seen with lower PSA levels. A negative biopsy should prompt reassessment in 6 months with PSA or DRE and repeat biopsy as needed.

Age enters into the decision-making process; men with low-volume tumors, with a GS of 5, or older men with major comorbid illnesses and limited life expectancy are less likely to benefit from definitive treatment than are younger and otherwise healthy men. In men with a life expectancy of less than 5 years with asymptomatic CaP, one option is deferral of further workup and treatment until presentation of symptoms; however, many patients prefer a more aggressive course.

If local treatment is planned, a bone scan is indicated for bone pain, T3 or T4, GS >7, or PSA >10 ng per mL. There is no clinical evidence that obtaining a “baseline” bone scan in other patients improves survival.

Computerized tomography (CT) scan or magnetic resonance imaging (MRI) is obtained for T3 and T4 lesions to detect the presence of enlarged lymph nodes in men for whom surgery is considered. If lymph nodes appear enlarged, a fine-needle aspiration (FNA) can be performed, potentially eliminating the need for surgery. CT scans are used for treatment planning for radiation therapy (RT). Baseline laboratory tests include complete blood count (CBC), creatinine level, PSA (if not yet done), and alkaline phosphatase level, and preoperative studies are performed if surgery is being considered (see Tables 14.2 and 14.3).

TABLE 14.2. Staging of Prostate Cancer: Tumor–node–metastasis (TNM) System

TURP, transurethral resection of the prostate; PSA, prostate-specific antigen.

T1   Tumor not palpable or visible by imaging
   T1a   Tumor incidental finding in ≤5% of tissue resected (TURP)
   T1b   Tumor incidental finding in >5% of tissue resected (TURP)
   T1c   Tumor identified by needle biopsy alone (after PSA is found to be elevated)

T2   Tumor confined to the prostate
   T2a   Tumor involves half of a lobe or less
   T2b   Tumor involves more than half of a lobe, but not both lobes
   T2c   Tumor involves both lobes

T3   Tumor extends through the prostatic capsule
   T3a   Extracapsular extension (unilateral or bilateral)
   T3b   Tumor invades the seminal vesicle(s)

T4   Tumor is fixed or invades adjacent structures other than seminal vesicles: bladder neck, external sphincter, rectum, levator muscles, and/or pelvic wall

N0   No regional lymph node metastasis

N1   Metastases in regional lymph node(s)

M0   No distant metastases

M1   Distant metastases (M1a, nonregional LN; M1b, bone(s); M1c, other).

Stage grouping





Any GS





GS 2, 3, or 4




Any GS




Any GS





Any GS





Any GS

Any T



Any GS

Any T

Any N


Any GS

TABLE 14.3. Modified Jewett system

PSA, prostate-specific antigen.

A: Clinically undetectable tumor confined to the prostate gland/incidental at prostate surgery
   A1: Well differentiated, focal involvement
   A2: Moderately or poorly differentiated, involves multiple foci of the gland

B: Tumor confined to the prostate gland
   B0: Nonpalpable, PSA detected
   B1: Single nodule in one lobe of the prostate
   B2: More extensive involvement of one lobe or both lobes

C: Tumor localized to the periprostatic area but extending through the prostatic capsule or involving the seminal vesicles
   C1: Clinical extracapsular extension
   C2: Extracapsular tumor producing bladder outlet or ureteral obstruction

D: Metastatic disease
   D0: Clinically localized disease with persistently elevated acid phosphatase
   D1: Regional lymph nodes only
   D2: Distant lymph nodes, metastases to bone or visceral organs
   D3: D2 patients who relapse after adequate endocrine therapy

Prognostic Factors at the Time of Diagnosis

  • Stage (Table 14.2 and 14.3)
  • Grade (GS)



  • PSA level
  • DNA ploidy in stage C or D1 patients (diploid is better)
  • Age
  • Serum alkaline phosphatase level.



Radical Prostatectomy

RP is performed by the perineal or retropubic approach. Pelvic lymph node dissection may be performed at the time of RP, although it may not be necessary in patients with T1c disease whose PSA is <10 ng per mL and GS is <7; the results will alter therapeutic decision less than 1% of the time in these patients. For patients at high risk of developing positive lymph nodes, pelvic laparoscopic lymph node dissection can be performed initially.

The formula for predicting percentage chance of positive pelvic lymph nodes is

Nerve-sparing Radical Prostatectomy

This is an appropriate procedure for men with small-volume disease in an attempt to conserve potency. However, after nerve-sparing RP, the rates of impotency that have been reported in the community are much higher than the rates in the original reports from certain tertiary


referral centers. Talcott et al. (6) reported data on 94 patients who underwent RP and who were observed before surgery and for up to 12 months after surgery. Nearly 80% of the men who had bilateral nerve-sparing surgery reported erections inadequate for sexual intercourse at 12 months after surgery, compared with 33% of such instances before surgery. Unilateral nerve-sparing surgery provided no benefit in potency. These numbers are significantly lower than those published previously in several retrospective studies. Studies have shown that diagnosis at a younger age and the absence of capsular penetration or seminal vesicle invasion correlate with a greater likelihood of conservation of potency after surgery.

Incontinence rates vary among single-institution experiences, and various definitions of incontinence make interpretation of studies difficult. In one prospective study evaluating urinary control at 3, 12, and 24 months after RP, it was reported that 58%, 35%, and 42% of patients, respectively, wore pads in their underwear, and 24%, 11%, and 15% of patients, respectively, reported “a lot” of urine leakage or escalating symptoms.

The Prostate Cancer Outcomes Study reported on the quality of life of patients following either surgery or RT. As one might expect, for patients with normal baseline function, surgery was associated with (a) an inferior urinary outcome score (approximately a 30-point decrease on a 100-point scale) compared with RT (less than 5-point decrease), (b) a better bowel score (less than 5-point decrease) compared with RT (about a 10-point decrease), and (c) slightly worse sexual function score (about a 40-point decrease) compared with RT (about a 30-point decrease), with all comparisons being statistically significant.

Neoadjuvant Therapy

The role of neoadjuvant hormonal therapy before RP is still under investigation. A recent randomized trial compared neoadjuvant treatment with 3 versus 8 months of leuprolide plus flutamide for clinically confined CaP (7). The proportion of men with negative margins and organ-confined disease was significantly higher in the group receiving 8 months of neoadjuvant hormonal therapy. Although there was a trend in the neoadjuvant group toward a delay in biochemical recurrence, it is too early in this study to determine survival. Thus, in general, neoadjuvant hormonal therapy prior to RP should be used only in the context of a clinical trial.

Extracapsular extension or positive margins in the surgical specimen portends a high incidence of disease recurrence, and such patients can be considered for clinical trials involving


postoperative radiation, chemotherapy, or hormonal treatment. Radiation may delay disease recurrence, but it has yet to show a convincing impact on survival.

After surgery, a detectable PSA level indicates a relapse, either local or systemic, although PSA increase may be slow and not necessarily indicate immediate treatment. In a study of nearly 2,000 men whose PSA was measured every 3 months after surgery, the median time for the development of metastases after detection of an elevated PSA (>0.2 ng per mL) was 8 years (8). PSA doubling time and GS were predictive of time to metastases.

Surgical Complications

Complications include immediate morbidity or mortality (2%) from surgery, impotence (35% to 60%), urinary incontinence (10% to 30%, depending on the definition), urinary stricture, and fecal incontinence (approximately 5%, with the retropubic approach, to 18%, with the perineal approach).

Radiation Therapy

External Beam

The traditional four-field box arrangement is used, with approximately 70 Gy given over 7 to 8 weeks. Prophylactic radiation of pelvic lymph nodes along with neoadjuvant and concurrent ADT has been shown to increase progression-free survival in patients when compared with those who do not receive whole pelvis radiation or short-term adjuvant ADT [Radiation Therapy Oncology Group (RTOG) 9413].

Ten-year Cause-specific Survival Rates with RT:

  • T1, 79%
  • T2, 66%
  • T3, 55%
  • T4, 22%.

Adjuvant Treatment with Androgen-deprivation Therapy

There are compelling data that combining ADT with RT in patients at high risk for recurrent disease improves OS (see Table 14.4). One large trial conducted by the RTOG 8531 randomized 945 patients having either T3 or T1–2, and N1 to RT (prostate and pelvis) with either adjuvant gonadotropin-releasing hormone agonist (GnRH-A) therapy for an indefinite period or GnRH-A therapy at the time of relapse (9,10). The recently presented 10-year OS favored the hormone-treated group (53% versus 38%, p < 0.0043) (11).

TABLE 14.4. Risk Categories for Posttherapy Prostate-specific Antigen Failure (12)





From Beer TM, Hough KM, Garzotto M, et al. Weekly high-dose calcitriol and docetaxel in advanced prostate cancer. Semin Oncol 2001;28(4):49–55, with permission.


T1c, T2a







Gleason score








5-Year risk of biochemical failure




In addition, the European Organization for Research and Treatment of Cancer (EORTC) performed a trial in 412 patients with poorly differentiated (WHO grade 3) or T3–4 cancers (13). Patients were randomized to receive radiation with or without concurrent and adjuvant GnRH-A therapy for 3 years, with the first month of GnRH-A therapy given with


an antiandrogen. The 5-year OS was 78% in the hormonal therapy group versus 62% in the radiation-alone therapy group (p = 0.0002) (14).

A subsequent trial (RTOG 9202) studied the use of neoadjuvant and concurrent hormonal therapy (GnRH-A with antiandrogen) with or without adjuvant GnRH-A therapy for 2 years in 1,554 patients who had a T stage of at least T2b (15). This trial demonstrated an OS advantage at 5 years (81% versus 71%, p = 0.044) in the subset of patients who had GS 8 to 10 tumors.

Another smaller randomized trial that stratified patients by biopsy-proven lymph node metastasis showed similar results, with improvement in median OS in the combined treatment arm seen only in the lymph node–positive tumor patients, further suggesting that patients who are at high risk for systemic disease benefit from ADT (16). Taken together, these trials provide strong evidence that patients at high risk for recurrence who are receiving definitive RT should be treated with long-term hormonal therapy.

Investigations are being initiated to test the addition of adjuvant chemotherapy in this setting.

3-D Conformational or High-dose Radiation Therapy

With careful treatment planning using 3-D conformational techniques, higher doses of radiation (up to 80 Gy over 7 to 8 weeks) can be delivered to the prostate while sparing normal tissue. This approach appears promising in that it may offer higher dosages to the tumor and thus greater efficacy and less toxicity to the surrounding structures.


Interstitial brachytherapy with radioactive palladium or iodine (I 125) seeds has been used for patients with T1 or T2 tumors. Initially, this approach required retropubic implantation (which required laparotomy), but over the last 10 years, CT and/or transrectal ultrasound have been used to guide seed placement, and the procedure is performed on an outpatient basis. Better definition of tumor volume and radiation dosimetry have made this technique more accurate. Initial results have been very promising; however, a randomized trial comparing brachytherapy to surgery had to be closed because of poor accrual.

External Beam and Brachytherapy

The use of combined external beam and brachytherapy has come into increasing use (17,18). A 10-year review of experience with this combination showed a biochemical (i.e., normal PSA) relapse-free survival of 79% in T3 tumors, suggesting a strong role for the combination in these lesions (19). However, there are no randomized trials comparing this combination with external beam or brachytherapy alone, and thus, this treatment approach remains investigational at this point.

Complications of Radiation Therapy

  • Acute (during treatment) complications: cystitis, proctitis, enteritis, and fatigue.
  • Long-term complications: impotence, incontinence (3%), frequent bowel movements (10%, more than with RP), and urethral stricture [RT is delayed by 4 weeks after transurethral resection of the prostate (TURP)].


Cryosurgery involves destruction of CaP cells through probes that subject the prostate tissue to freezing followed by thawing. Some believe that cryosurgery is a good option for men with high-grade tumors (Gleason 8 to 10), high PSA levels (20 to 40 ng per mL), or stage C


tumors, who potentially do not respond well to RT or surgery. Cryosurgery is also an option for local recurrence of CaP. This technique is still in early development, and long-term efficacy is not fully established. Side effects include incontinence, impotence, and injury to the bladder outlet and rectal tissues.


In several European countries, observation of patients with CaP, particularly of low stage or low or moderate grade, is commonplace. Particularly with elderly patients, survival equivalent to aggressive treatment has been reported.

Hormonal Therapy

ADT is used most commonly for metastatic CaP, although it has been used in the treatment of localized disease and in the neoadjuvant, and adjuvant, settings with RT. Androgen blockade is achieved through bilateral surgical castration or depot injections of GnRH-A (e.g., leuprolide, goserelin, and buserelin), which offer equivalent efficacy. Combined androgen blockade can be achieved by adding an oral antiandrogenic agent (e.g., nilutamide, flutamide, and bicalutamide); however, this is controversial, and if a benefit does exist, it is small.

Tumor flare is possible with the use of GnRH-A, which initially causes an increase in luteinizing hormone (LH) and follicle-stimulating hormone (FSH). Because the CaP cells are androgen sensitive, this increase in testosterone-inducing hormones can exacerbate symptoms of CaP and can elevate PSA. Tumor flare can be prevented by the use of oral antiandrogens, which compete with androgens at the androgen receptor site, at low dose for several weeks before beginning therapy. There is a reduced risk of tumor flare with a lower volume of disease. Abarelix is a GnRH antagonist that also may be used in patients who are at high risk for complications associated with tumor flare. Because of the risk of potentially life-threatening allergic reactions, its use is limited to those who are have advanced, symptomatic CaP and who do not have other treatment options.

The use of estrogen [e.g. diethylstilbestrol (DES)] has fallen into disfavor because of its side-effect profile. Therapeutic doses were shown to induce a high frequency of cardiovascular complications and mortality in a population predisposed to cardiovascular disease.

Intermittent androgen ablation (IAA) is a recent approach to treating patients with hormone-sensitive CaP. This involves treatment with hormonal therapy for 2 to 3 months beyond “best response,” followed by a discontinuation of hormonal therapy and restarting hormones at some predetermined point (e.g., based on the PSA value). IAA has the potential advantage of improved quality of life during the time the patient is not receiving hormonal therapy. The effects of these manipulations on a hormonally sensitive tumor, however, are unknown. A large randomized trial by the Southwestern Oncology Group (SWOG) is under way to evaluate the effects of these manipulations.

The rationale for continued androgen blockade, antiandrogen withdrawal, and additional hormonal therapy (e.g., ketoconazole and aminoglutethimide) are discussed in subsequent text about second-line hormonal therapy.

Dosages and Side Effects of Hormonal Therapy

  • Bilateral orchiectomy:Side effects include tumor flare (see subsequent text), impotence, loss of libido, gynecomastia, hot flashes, and osteoporosis.
  • GnRH agonists:
  1. Goserelin acetate (Zoladex), 3.6 mg s.c. every month or 10.8 mg s.c. every 3 months. Side effects are the same as with orchiectomy.



  1. Leuprolide acetate (Lupron), 7.5 mg s.c. every month or 22.5 mg i.m. every 3 months, or 30 mg s.c. every 4 months. Comparable efficacy with long-acting formulations. Side effects are the same as with orchiectomy.
  • GnRH antagonist:
  1. Abarelix (Plenaxis), 100 mg i.m. on days 1, 15, 29, and every 4 weeks thereafter. In addition to orchiectomy-like side effect, patients are at risk for allergic reactions and must be observed for 30 minutes after injection.
  • Oral antiandrogens:
  1. Flutamide (Eulexin), 250 mg PO t.i.d.Side effects include diarrhea, nausea, breast tenderness, hepatotoxicity [liver enzymes levels must be monitored], loss of libido, and impotence.
  2. Bicalutamide (Casodex), 50 mg daily. Side effects include nausea, breast tenderness, hepatotoxicity (transaminases must be monitored), hot flashes, loss of libido, and impotence.
  3. Nilutamide (Nilandron), 150 mg PO daily. Side effects include pulmonary fibrosis (rare), visual field changes (i.e., night blindness or abnormal adaptation to darkness), hepatotoxicity (liver enzymes must be monitored), impotence, loss of libido, hot flashes, nausea, and disulfiram-like reaction.

Hot flashes from hormonal therapy can be treated with clonidine (0.1 mg per day), low-dose estrogens, or possibly antidepressants. Painful gynecomastia can be treated with external beam radiation (electron beam) therapy to the breasts.

Testosterone-lowering therapy causes a decrease in estradiol levels because of decreased peripheral aromatization. Because estradiol is needed to maintain bone density, osteoporosis is increasingly being diagnosed in patients who are receiving ADT (20). This side effect is most likely to occur with long duration of ADT and in men who have had baseline osteoporosis before starting anti-androgen therapy. In general, all men receiving ADT should receive vitamin D (400 to 800 IU) and calcium (1,200 to 1,500 mg) daily. Bone densitometry can be performed to assess the presence and/or severity of osteoporosis. Treatment with bisphosphonates should be considered in patients with low bone mineral density.


Comparison of treatment modalities with respect to primary outcomes [OS and disease-free survival (DFS)] is difficult because of variability in study design, patient selection, and technique of each therapy. No satisfactory randomized trials comparing RT with RP have been conducted.

In general, comparing single-modality studies is difficult because of disparities in treatment groups; men receiving RT tend to be older and have more comorbid illnesses. A major problem is that men receiving RT would have had only clinical staging of their cancer, whereas many patients who undergo surgery would have had pathologic staging that may upstage their tumor, including the discovery of positive pelvic lymph nodes (metastatic disease). One way to compare the two modalities is by PSA and PSA-free survival (also called biochemical relapse-free survival), provided that the patients have reproducible PSA values on repeated measurement before their definitive, primary therapy.

In the “PSA era,” RT and RP appear to have equivalent PSA-free survival in appropriately matched patients at 5 years, but differ in type and frequency of side effects. Brachytherapy appears to be promising, although most studies have been conducted in patients with early stage, low-grade disease only, who are thought to be the most appropriate patients for this treatment modality. One comparison of 3-D conformational RT with 125I implants in comparable patients concluded that both treatment modalities had equivalent efficacy, with higher urinary complications in the brachytherapy group.

Continuing research in quality of life may allow more informed choices among the various treatment modalities.




Treatment decisions should be made with regard to the patient's age and comorbid conditions (i.e., life expectancy) and preferences, as well as tumor-related factors (i.e., stage, grade, and PSA). Enrollment in clinical trials is appropriate at each stage of disease.


Observation is often appropriate for elderly patients with low-grade disease. Definitive treatment should be considered for high PSA levels or GS if life expectancy is long with (a) RT (external beam, 3-D, and brachytherapy) and (b) RP.

Localized Disease (T1b through T2c)

Two factors, in particular, are important in deciding whether to treat localized disease. These are (a) the probability of having organ-confined disease: normograms based on PSA, clinical stage and GS, and (b) the patient's overall life expectancy. Observation can be considered in selected patients.

For a long life expectancy and a reasonable chance of having organ-defined disease, RT (including 3-D/high dose or brachytherapy) or RP remains the standard of care. Surgery is usually reserved for patients younger than 70 years. For a short life expectancy, RT (external beam, 3-D, and brachytherapy), hormonal therapy, or deferring treatment until the presentation of symptoms is recommended.

Patients in whom surgery would pose a great risk (e.g., cardiovascular or pulmonary disease) should have RT (i.e., external beam, 3-D, or brachytherapy). Otherwise, patient preference should play a major role in deciding primary treatment choice, after informed discussions have occurred. Certain patients should be considered for neoadjuvant hormonal therapy in combination with RT (see earlier discussion). The role of neoadjuvant hormonal therapy in surgical patients remains investigational (see earlier).

If pelvic lymph nodes are found to be grossly positive at the time of surgery, prostatectomy should not be performed. A recently updated randomized trial (21) compared men with T1 or T2 lesions and microscopically positive lymph nodes. Men were randomized to immediate hormonal therapy (orchiectomy or goserelin) versus observation (with hormonal therapy started at the time of metastases or symptomatic local recurrences). The difference in OS at 10 years was 72.4% versus 49% (p = 0.025), favoring the immediate ADT group.

If the tumor is found to have extracapsular extension on the pathologic specimen, consideration can be given to enrolling in clinical trials using adjuvant RT, hormonal therapy, or a combination of the two. As yet, these strategies have unproven survival benefits.

There are insufficient data to claim a survival benefit for surgery versus radiation as a primary treatment modality.

Stage III (T3 N0 M0)

RP can be considered for well-differentiated tumors. Other patients should probably receive RT (external beam) with or without neoadjuvant hormonal therapy. Hormonal therapy alone (surgical castration or GnRH-A) is another option. Some patients are treated with brachytherapy followed by external beam RT, but results are preliminary at this point.

T3b, T3c, T4 N0

These patients are unlikely to be cured by surgery. Viable treatment options include hormonal therapy, RT alone, or RT with hormonal therapy. With node-positive disease, patients are considered metastatic at the time of diagnosis, and local therapy should be used primarily


for symptom management. Observation is a viable option for this group of patients, given the side effects of hormonal therapy and RT and their primarily palliative role.

N + or M1

There are no conclusive data that RT or RP improves survival in patients who are node positive. Hormonal therapy remains the standard of care. Surgical castration or medical orchiectomy with GnRH-A offer equivalent efficacy. There is no proven survival benefit of maximal androgen blockade (surgical castration and oral antiandrogen) compared with surgical castration alone. A randomized study of leuprolide with or without flutamide alone showed a slight survival advantage (22); however, a similar study design that used orchiectomy instead of GnRH-A showed equivalent survival (23). In the trial with orchiectomy and flutamide versus flutamide alone, quality of life was shown to be reduced in the group receiving antiandrogen therapy in addition to castration (24). The reason for the discrepancy between the results of two trials is not completely known. A meta-analysis found an OS advantage for the combined blockade at 5 years, but not 2 years; however, the magnitude of the difference is of questionable clinical significance (25).

Another area of controversy is the use of delayed or immediate hormonal therapy in patients who are initially metastatic or in those in whom tumors recur. The Medical Research Council trial (26) randomized patients with locally advanced or asymptomatic metastatic CaP to immediate versus deferred orchiectomy and GnRH-A. A significant difference in survival was noted in the patients, especially in M0 patients. However, this study has often been criticized. The comprehensive, evidence-based review conducted by the Agency for Health Care Policy and Research (AHCPR) concluded that there is no evidence favoring immediate compared with deferred androgen suppression; however, many practitioners tend toward early antiandrogen therapy, and newer evidence favoring early antiandrogen therapy is beginning to mount. Again, quality-of-life issues should play a role in decision analysis.


RT after RP for patients with margin-positive disease has been either used soon after surgery (adjuvant) or delayed until PSA progression occurs (salvage). Not surprisingly, biochemical failure appears to be less likely when adjuvant RT is used. Higher biochemical failure rates were seen in patients with seminal vesicle involvement (adjuvant) and with a GS of 4 or 5, or PSA >2.0 ng per mL (salvage). Long-term follow-up is not available with this approach. A combination of RT and hormonal therapy after local relapse is also being examined. In addition, some patients with a negative metastatic workup and local recurrence after RT can be considered for salvage surgery; the evidence for this is even less defined.


Patients treated with curative intent should have their PSA level ascertained at least every 6 months for 5 years and then annually. Annual DRE is appropriate to detect recurrences.

After treatment with RP, any reproducible, detectable PSA indicates a relapse. PSA failure after RT is defined as three consecutive increases in PSA levels. Treatment for patients who relapse after radiation has not been standardized; participation in clinical trials should be encouraged. Hormonal treatment and salvage surgery (if a metastatic workup is negative and if patient is in good health) are options.

Surveillance and PSA measurement may be optional for patients with a short life expectancy who are treated with observation alone. Treatment should be guided by symptoms. For patients with a longer life expectancy, annual DRE and PSA along with workup and treatment of symptoms are appropriate if watchful waiting is selected as the primary treatment.



For patients with metastatic disease, intensity and type of follow-up are determined by the degree of clinical progression; for patients who respond well to hormonal therapy, follow-up at 3 months (with PSA) is reasonable. Bone scans are indicated depending on clinical symptoms but should not be ordered routinely. Patients with bony metastases are at risk for spinal cord compression, and MRI should be ordered when signs or symptoms are suggestive of this complication.

It is important to note that there is interlaboratory variation in PSA levels.


In this chapter, PSA response rates (PRRs) are shown as percentage of patients with a PSA decline greater than 50%, which is a generally agreed upon criterion (27). Ranges of PRRs are given when available. Confidence intervals are not shown around the PRR but are, in general, wide. Because of differences in patient selection, it is difficult to compare clinical trials by PRRs alone. Quality of life is an extremely important consideration in treating patients with CaP and may be a primary consideration in the choice of chemotherapeutic agents for metastatic androgen-insensitive prostate cancer (AIPC). An important caveat, however, is that some agents (particularly cytostatic agents) may upregulate or downregulate PSA expression independent of their effect on cancer growth.


Metastatic prostate cancer tends to present at the spine or axial skeleton. Visceral metastases are uncommon, and brain metastases are even rarer. CaP cells usually respond to hormonal manipulations that block the production of androgen with durable remissions and significant palliation. Duration of response ranges from 12 to 18 months, with 20% of patients having a complete biochemical response at 5 years. However, ultimately, androgen-independent CaP cells emerge and lead to progression of disease.

The use of maximal androgen blockade is not currently considered as standard of care in the initial treatment of metastatic disease. In patients who are being treated with GnRH-A or who undergo surgical castration, the addition of an antiandrogen agent may produce positive responses in up to 30% to 40% of instances.

Antiandrogen Withdrawal

Once GnRH-A therapy is started for patients with metastatic disease, it should be continued for life. It has been shown that metastatic CaP can reactivate if testosterone levels are allowed to increase or if exogenous testosterone is administered, and anecdotal evidence has shown a worsening of disease in patients who discontinued GnRH-A. Approximately 20% of patients being treated by maximal androgen blockade have PRR upon discontinuation of the oral antiandrogen [range, 15% to 33% relative risk (RR)], although these declines in PSA levels are relatively short lived (median duration, 3 to 5 months). Decreased cancer-related anemia and decreased pain were also reported. Antiandrogen withdrawal response occurs within 4 to 6 weeks, depending on the half-life of the agent.

Second-line Hormonal Therapy

Even after androgen withdrawal has failed, some patients will benefit from switching to antiandrogens or initiating treatment with aminoglutethimide, ketoconazole, or glucocorticoids. A minimum of 1 month is required to assess patients for a response. A few patients are likely to respond, and responses are usually not long in duration, but these agents offer much less toxicity than chemotherapy.



Adrenal Androgen Inhibitors

Adrenal androgen inhibitors work by achieving a “medical adrenalectomy,” which further decreases androgen production. Responses have been seen in patients after antiandrogen withdrawal. It is important to use steroid replacements in patients receiving adrenal androgen inhibitors; this is often started with hydrocortisone 20 mg every morning and 10 mg every evening but is increased to 20 mg PO b.i.d if patients show symptoms of glucocorticoid insufficiency (e.g., fatigue).

Adrenal Androgen Inhibitors: Regimens and Toxicities

Aminoglutethimide + Hydrocortisone (RR = 49%) (28):

  • Aminoglutethimide (Cytadren), 125 mg PO q.i.d., increasing to 250 mg q.i.d. + hydrocortisone, 20 mg PO b.i.d.
  • Side effects: sedation, skin rashes, and fever
  • Rarer side effects: ataxia, hypothyroidism, abnormal liver enzyme levels, and peripheral edema.

Ketoconazole + Hydrocortisone (RR = 35% to 50%) (29):

  • Ketoconazole (Nizoral), 200 mg PO t.i.d., increasing to 400 mg PO t.i.d. + hydrocortisone, 20 mg PO b.i.d.
  • Side effects: impotence, pruritus, nail changes, adrenal insufficiency, nausea, emesis, and hepatotoxicity. Liver transaminases need to be monitored. (Ketoconazole is absorbed at an acidic pH; therefore, the concomitant use of H2blockers, antacids, or proton pump inhibitors should be avoided. Coating agents such as sucralfate may be substituted.)

Corticosteroids (RR = 18% to 22%) (30):

  • Corticosteroids alone have been shown to improve pain in patients with symptomatic bone metastases.
  • Prednisone, 5 mg PO every morning and 2.5 mg PO every evening, increasing to 5 mg PO b.i.d.; side effects are the same as for any medical use of prednisone.


Patients have a median survival of 12 to 18 months after developing androgen-independent prostate cancer. Chemotherapy with docetaxel-containing regimens has only recently been shown to improve survival in patients with AIPC, with median OS of about 18 months (31,32). Participation in clinical trials of novel agents or combinations should be encouraged. Antiangiogenic agents and immunotherapy or tumor vaccines are undergoing studies in AIPC.

Estramustine phosphate (EMP, Emcyt) is a compound that combines estradiol and nitrogen mustard. It works by binding to microtubule-associated proteins. Many phase II trials of single-agent estramustine have been conducted. However, a synergistic effect appears to result when estramustine is combined with other agents that have activity against the microtubule proteins (e.g., vinblastine and the taxanes); therefore, estramustine is usually given in combination.

Docetaxel (Taxotere)/Estramustine (Emcyt) (PRR = 50%) (31):

  • Improved median OS from 15 months (mitoxantrone/prednisone) to 18 months (p= 0.008) is observed.
  • Estramustine, 280 mg PO t.i.d., is administered 1 hour before or 2 hours after meals, on days 1–5, with docetaxel, 60 mg per m2on day 2. Patients were dose escalated to 70 mg per m2 if there was no grade 3 toxicity observed in cycle 1.
  • Dexamethasone, 20 mg PO t.i.d. × 3 doses, is administered, starting the evening of day 1 of each cycle.



  • Cycle is repeated every 21 days.
  • Side effects are grade 3 toxicity: hematologic effect (20%), gastrointestinal effect (20%), cardiovascular effect (15%), infection (14%), flu-like syndrome (10%), and neurologic effect (7%).

Docetaxel (Taxotere) (PRR 45%) (32):

  • Improved median OS from 16.5 months (mitoxantrone/prednisone) to 18.9 months (p= 0.0005) and improved quality of life [Functional Assessment of Cancer Therapy-Prostate (FACT-P) 22% versus 13%; p = 0.009] are observed.
  • Docetaxel 75 mg per m2i.v. is given.
  • Prednisone 5 mg is given twice daily.
  • Cycle is repeated every 21 days.
  • Side effects include grade 3 toxicity, granulocytopenia (32%), infection (5.7%), anemia (4.9%), and fatigue (4.5%), and any grade of toxicity, anemia (67%), neutropenia (41%), fluid retention (24%), sensory neuropathy (30%), nausea (41%), fatigue (53%), myalgia (15%), and alopecia (65%).

Docetaxel (Taxotere) + Calcitriol (PRR = 81%;ORR 53%) (12):

  • Calcitriol 0.5 µg per kg PO in four divided doses was given over 4 hours on day 1.
  • Docetaxel 36 mg per m2i.v. over 15 to 30 minutes was given on day 2.
  • Dexamethasone 8 mg PO was given 12 hours and 1 hour before docetaxel infusion and 12 hours after docetaxel infusion.
  • Treatment was repeated each week for 6 out of every 8 weeks.
  • Therapy was held when platelet count was less than 75,000 or when absolute neutrophil count (ANC) was less than 1,000; docetaxel dose was reduced by 25% if ANC or platelet recovery was greater than 1 week.
  • Patients were maintained on reduced calcium diets (400 to 500 mg daily).
  • Side effects include grade 3 or greater toxicity with leukopenia (41%), neutropenia (24%), anemia (3%), hyperglycemia (24%), and peptic ulcer (11%).

Mitoxantrone (Novantrone) + Prednisone (PRR = 33%) (30,33):

  • This regimen has been shown to improve quality of life, but not DFS or OS, in two randomized controlled trials versus steroids alone.
  • Prednisone, 5 mg PO b.i.d, on day 1, + mitoxantrone, 12 mg per m2i.v., on day 21, delayed if not recovered hematologically.
  • Mitoxantrone was stopped at a cumulative dose of 140 mg per m2. Prochlorperazine was used as an antiemetic.
  • Side effects: cardiac abnormalities in 6% of patients in the mitoxantrone arm only [2% with congestive heart failure (CHF)], neutropenic fever (1.1%), neutropenia (45%), thrombocytopenia (5%), nausea and vomiting (29%), and alopecia (26%); exacerbation of diabetes in one patient.
  • Prednisone, 5 mg PO b.i.d., on day 1, + mitoxantrone, 12 mg per m2i.v., on day 21, delayed if not recovered hematologically.


CNS metastases are relatively rare, but not unheard of, in CaP. Visceral metastases occur in about 20% of patients; most patients have symptoms related to bone metastases.

Bone Metastases

The use of RT to localized painful bone metastases has been shown to provide palliation. Usually, the painful vertebra and the two vertebrae superior and inferior to the lesion are treated with 3,000 cGy radiation in ten fractions. Pain relief occurs in approximately 80% of


patients; side effects generally are limited to fatigue and anemia that is usually reversible. The spinal cord can tolerate radiation up to approximately 5,000 cGy, so re-treating these areas with a radiation of 2,000 cGy can sometimes be attempted, although with caution.

For widespread disease, hemibody irradiation has been used, as has the radioisotope strontium-89 (Metastron), a calcium analog that preferentially localizes in the tumor. Palliation of pain with strontium has been reported in up to 75% of cases and typically occurs after 1 to 3 weeks of treatment and may continue for several months. Toxicities of strontium include the potential for flare (15%) that is often associated with a later response and a reversible thrombocytopenia in 25% of patients that usually resolves by 3 months. Strontium can often be readministered. Samarium-153 lexidronam (Quadramet) is a newer radioisotope with treatment indications similar to those of strontium and a shorter half-life than strontium.

Bisphosphonates, agents that inhibit osteoclastic bone resorption, have been shown to decrease skeletal-related events in patients with advanced androgen-independent disease (34). Careful attention to pain control with narcotics and adjuncts should be maintained in patients with bone metastases.

Spinal Cord Compression

Spinal cord compression, an oncologic emergency, is common in patients with metastatic prostate cancer who have widespread bony metastases. This consists largely of vertebral column metastases impinging on the spinal cord.

More than 90% of patients have pain as an early sign; pain that is worsening is particularly disturbing. There may be pain in the involved spine, muscle weakness, or abnormalities in the neurologic examination. Signs that are often indicative of irreversible damage include weakness and/or sensory loss corresponding to the level of the spinal cord compression. Signs such as genitourinary or gastrointestinal dysfunction (e.g., urinary retention or constipation) or autonomic dysfunction are late signs, and spinal cord compression usually progresses rapidly at this point.

One should have a high index of suspicion for spinal cord compression in patients known to have osseous metastases in CaP, particularly with new signs or any symptoms related to spinal cord compression. Diagnosis requires a thorough history and physical examination, with special attention to the musculoskeletal and neurologic examinations. The standard for diagnosing and localizing epidural cord compression is an MRI, usually with gadolinium. A myelogram is still used in patients with contraindications to MRI (e.g., a pacemaker).

Steroids should be started (dexamethasone, 100 mg i.v., followed by 4 mg i.v. or PO, every 6 hours) as soon as history or neurologic examination suggests spinal cord compression. RT, given as 3,000 cGy in ten fractions to the involved vertebra and to the two superior and two inferior vertebrae, is the usual treatment modality, and early consultation with an radiation oncologist is warranted. Surgical resection of the vertebral body is generally used in patients who have had previous RT of the involved area if these patients require procedures for spinal stability or experience progression despite treatment with steroids and RT or if RT facilities are not locally available. It should also be considered in patients with a rapidly progressive neurologic deficit because the relief from RT is slower (by days) than from a surgical decompressive procedure. A recent randomized trial showed that patients subjected to decompressive surgical resection followed by radiation retained the ability to walk significantly longer than those treated with radiation alone. Neurologic or orthopedic surgeons should be consulted early in the diagnosis of spinal cord compression as well.


  1. Chu KC, Tarone RE, Freeman HP. Trends in prostate cancer mortality among black men and white men in the United States. Cancer2003;97(6):1507–1516.
  2. Thompson IM, Goodman PJ, Tangen CM, et al. The influence of finasteride on the development of prostate cancer. N Engl J Med2003;349(3):215–224.



  1. Albanes D, Heinonen OP, Taylor PR, et al. Alpha-tcopherol and beta-carotene supplements and lung cancer incidence in the alpha-tocopherol, beta-carotene cancer prevention study: effects of base-line characteristics and study compliance. J Natl Cancer Inst1996;88(21):1560–1570.
  2. Clark LC, Combs GF Jr, Turnbull BW, et al. Effects of selenium supplementation for cancer prevention in patients with carcinoma of the skin. A randomized controlled trial. Nutritional prevention of cancer study group. JAMA1996;276(24):1957–1963.
  3. This can be accessed via the internet:
  4. Talcott JA, Rieker P, Clark JA, et al. Patient-reported symptoms after primary therapy for early prostate cancer: results of a prospective cohort study. J Clin Oncol1998;16(1):275–283.
  5. Gleave M, Goldenberg L, Chin JL, et al. Natural history of progression after PSA elevation following radical prostatectomy: update. J Urol2003;169(4):A690.
  6. Pound CR, Partin AW, Eisenberger MA, et al. Natural history of progression after PSA elevation following radical prostatectomy. JAMA1999;281(17):1591–1597
  7. Pilepich MV, Caplan R, Byhardt RW, et al. Phase III trial of androgen suppression using goserelin in unfavorable-prognosis carcinoma of the prostate treated with definitive radiotherapy: report of Radiation Therapy Oncology Group Protocol 852-31. J Clin Oncol1997;15(3):1013–1021.
  8. Lawton CA, Winter K, Murray K, et al. Updated results of the phase III Radiation Therapy Oncology Group (RTOG) trial 85-31 evaluating the potential benefit of androgen suppression following standard radiation therapy for unfavorable prognosis carcinoma of the prostate. Int J Radiat Oncol Biol Phys2001;49(4):937–946.
  9. Pilepich MV, Winter K, Lawton C, et al. Phase III trial of androgen suppression adjuvant to definitive radiotherapy. Long term results of RTOG study 85-31. Proc Am Soc Clin Oncol2003;22:381. Abstract 1530.
  10. Beer TM, Hough KM, Garzotto M, et al. Weekly high-dose calcitriol and docetaxel in advanced prostate cancer. Semin Oncol2001;28(4):49–55.
  11. Bolla M, Gonzalez D, Warde P, et al. Improved survival in patients with locally advanced prostate cancer treated with radiotherapy and goserelin. N Engl J Med1997;337(5):295–300.
  12. Bolla M, Collette L, Blank L, et al. Long-term results with immediate androgen suppression and external irradiation in patients with locally advanced prostate cancer (an EORTC study): a phase III randomised trial. Lancet2002;360(9327):103–108.
  13. Hanks GE, Pajak TF, Porter A, et al. Radiation Therapy Oncology Group. Phase III trial of long-term adjuvant androgen deprivation after neoadjuvant hormonal cytoreduction and radiotherapy in locally advanced carcinoma of the prostate: the radiation therapy oncology group protocol 92-02. J Clin Oncol2003;21(21):3972–3978.
  14. Granfors T, Modig H, Damber JE, et al. Combined orchiectomy and external radiotherapy versus radiotherapy alone for nonmetastatic prostate cancer with or without pelvic lymph node involvement: a prospective randomized study. J Urol1998;159(6):2030–2034.
  15. Dattoli M, Wallner K, True L, et al. Long-term outcomes after treatment with external beam radiation therapy and palladium 103 for patients with higher risk prostate carcinoma—influence of prostatic acid phosphatase. Cancer2003;97(4):979–983.
  16. Critz FA, Williams WH, Levinson AK, et al. Simultaneous irradiation for prostate cancer: intermediate results with modern techniques.J Urol2000;164(3 Pt 1):738–741.
  17. Kovacs G, Galalae R, Loch T, et al. Prostate preservation by combined external beam and HDR brachytherapy in nodal negative prostate cancer. Strahlenther Onkol1999;175:87–88.
  18. Daniell HW. Osteoporosis after orchiectomy for prostate cancer. J Urol1997;157(2):439–444.
  19. Messing E, Manola J, Sarosdy M, et al. Immediate hormonal therapy compared with observation after radical prostatectomy and pelvic lymphadenectomy in men with node positive prostate cancer: results at 10 years of EST 3886. J Urol2003;169(4):1480.
  20. Crawford ED, Eisenberger MA, McLeod DG, et al. A controlled trial of leuprolide with and without flutamide in prostatic carcinoma. N Engl J Med1989;321(7):419–424.
  21. Eisenberger MA, Blumenstein BA, Crawford ED, et al. Bilateral orchiectomy with or without flutamide for metastatic prostate cancer.N Engl J Med1998;339(15):1036–1042.
  22. Moinpour CM, Savage MJ, Troxel A, et al. Quality of life in advanced prostate cancer: results of a randomized therapeutic trial. J Natl Cancer Inst1998;90(20):1537–1544.
  23. Agency for Health Care Policy and Research. Relative effectiveness and cost-effectiveness of methods of androgen suppression in the treatment of advanced prostatic cancer (AHCPR Publication No 99-E012). 1999.



  1. Adib RS, Anderson JB, Ashken MH, et al. Immediate versus deferred treatment for advanced prostatic cancer: initial results of the Medical Research Council trial. Br J Urol1997;79(2):235–246.
  2. Bubley GJ, Carducci M, Dahut W, et al. Eligibility and response guidelines for phase II clinical trials in androgen-independent prostate cancer: recommendations from the Prostate-Specific Antigen Working Group. J Clin Oncol1999;17(11):3461–3467.
  3. Sartor O, Cooper M, Weinberger M. Surprising activity of flutamide withdrawal, when combined with aminoglutethimide, in treatment of hormone-refractory prostate-cancer. J Natl Cancer Inst1994;86(6):463, 222.
  4. Small EJ, Baron A, Bok R. Simultaneous antiandrogen withdrawal and treatment with ketoconazole and hydrocortisone in patients with advanced prostate carcinoma. Cancer1997;80(9):1755–1759.
  5. Tannock IF, Osoba D, Stockler MR, et al. Chemotherapy with mitoxantrone plus prednisone or prednisone alone for symptomatic hormone-resistant prostate cancer: a Canadian randomized trial with palliative end points. J Clin Oncol1996;14(6):1756–1764.
  6. Petrylak D, Tangen C, Hussain M, et al. SWOG 99-16: randomized phase III trial of docetaxel/estramustine versus mitoxantrone/prednisone in men with androgen-independent prostate cancer. Proc Am Soc Clin Oncol2004;23:2.
  7. Eisenberger M, De Wit R, Berry W, et al. A multicenter phase III comparison of docetaxel + prednisone and mitoxantrone + prednisone in patient with hormone refractory prostate cancer. Proc Am Soc Clin Oncol2004;23:2.
  8. Kantoff PW, Halabi S, Conaway M, et al. Hydrocortisone with or without mitoxantrone in men with hormone-refractory prostate cancer: results of the Cancer and Leukemia Group B 9182 study. J Clin Oncol1999;17(8):2506–2513.
  9. Saad F, Gleason DM, Murray R, et al. A randomized, placebo-controlled trial of zoledronic acid in patients with hormone-refractory metastatic prostate carcinoma. J Natl Cancer Inst2002;94(19):1458–1468.