Intended for healthcare professionals

Clinical Review State of the Art Review

Advances in the management of castration resistant prostate cancer

BMJ 2016; 355 doi: (Published 17 October 2016) Cite this as: BMJ 2016;355:i4405
  1. Chad R Ritch, assistant professor1,
  2. Michael S Cookson, professor and chairman2
  1. 1Department of Urology, University of Miami, Miller School of Medicine, Miami, FL, USA
  2. 2Department of Urology, University of Oklahoma College of Medicine and Stephenson Cancer Center, Oklahoma City, OK, USA
  1. Correspondence to: M S Cookson michael-cookson{at}


Docetaxel based chemotherapy showed survival benefit and emerged as the mainstay of treatment for castration resistant prostate cancer (CRPC) in 2004. However, therapeutic options have expanded rapidly since 2011. The spectrum of new agents is broad and includes drugs that target the androgen axis (enzalutamide, abiraterone), immunotherapy (sipuleucel-T), bone seeking radionuclides (radium-223), and second line chemotherapy (cabazitaxel). In addition, new agents have been developed to reduce skeletal related events (denosumab). Given that docetaxel was the standard first line treatment for metastatic CRPC, the newer oral agents that affect the androgen axis were initially approved in the post-docetaxel setting. However, subsequent randomized trials have led to their approval in the pre-chemotherapy setting as well. Patients with CRPC are clinically heterogeneous, ranging from patients who are asymptomatic and do not have metastases to those with substantial symptoms and both bony and visceral metastases. CRPC is a clinically challenging disease entity, therefore, with a wide array of treatment options and multiple possible sequencing combinations depending on the individual patient. This review will summarize the findings of the randomized trials that led to the approval of the therapies for CRPC. It will also discuss recent guidelines and provide suggestions for sequencing of drugs based on the best available evidence.


The management of castration resistant prostate cancer (CRPC) has evolved over the past decade. Before 2010 only studies of docetaxel chemotherapy showed improved survival, resulting in its widespread approval by the US Food and Drug Administration (FDA) and adoption as first line therapy worldwide.1 2 More recently, however, several large randomized trials have led to the approval of multiple new agents for CRPC (table 1).

Table 1

Summary of approved therapeutic agents used in castration resistant prostate cancer

View this table:

Figure 1 shows a schematic outline of various common clinical scenarios for CRPC along with the appropriate therapeutic agent for each scenario. Given the rapid growth of treatment options, the varying mechanisms of action (novel androgen receptor antagonists, novel chemotherapeutics, androgen synthesis inhibitors, immune therapy, bone seeking radionuclide) and different modes of administration (injection, oral, infusion), many practitioners are clinically unfamiliar with these therapies. Furthermore, as CRPC is a clinically heterogeneous group, the indications for and appropriate sequencing of these drugs are challenging.3 Patterns of resistance are emerging, and studies are ongoing to determine what mechanisms exist to promote resistance and whether a method exists to identify and select candidates more appropriately for a particular therapeutic agent.4 Equally important are ongoing studies designed to define the appropriate sequencing and potential combination of agents in CRPC.3


Fig 1 Schematic outline for management of castration resistant prostate cancer (CRPC). *Based on data from STRIVE study39

Recently, the concept of chemotherapy in the hormone sensitive state has emerged as a viable treatment approach owing to the survival benefit noted in a randomized trial of 790 men. The trial compared docetaxel plus androgen deprivation therapy (ADT) versus ADT alone and showed a significantly longer median overall survival in the docetaxel plus ADT group (57.6 v 44.0 months; hazard ratio 0.61, 95% confidence interval 0.47 to 0.81; P<0.001).5 Significant benefit was noted in the subgroup with high volume disease, defined as the presence of visceral metastases or four or more bone lesions with at least one beyond the vertebral bodies and pelvis (49.2 v 32.2 months; P=0.0012). These findings were confirmed by data from the STAMPEDE trial, which showed improved survival in men who received docetaxel at the time of starting long term ADT; in addition, the benefit was seen in all patients rather than being limited to those with high volume disease.6

Therefore, current CRPC treatment algorithms based on stratification before or after chemotherapy may change over time owing to earlier exposure to chemotherapy in the castration sensitive state. New agents targeting the androgen axis class, as well as other novel targets such as those involving DNA repair and immune checkpoint pathways, are being explored as potential therapeutic options for patients with CRPC.7 8 9 This review will discuss the contemporary management of CRPC based on current guidelines from the United States and Europe and review the evidence supporting the guidelines’ recommendations.10 11 12 13 It will also summarize evidence from recent studies that has affected the treatment of CRPC but has yet to be incorporated into the most recent versions of the guidelines, along with emerging data investigating new therapies for CRPC.

Incidence and prevalence

Prostate cancer is the most common solid organ malignancy in men and remains the second leading cause of cancer related deaths in men.14 In the United States, an estimated 26 120 deaths were attributed to prostate cancer in 2016, and many of these occurred in patients with metastatic CRPC.14 Data suggest that 10-20% of patients with metastatic prostate cancer develop CRPC within five years of follow-up and that median survival from the occurrence of castration resistance is approximately 14 (range 9-30) months.15 16

In addition, patients with non-metastatic CRPC are at high risk for progression of disease. Approximately 15-33% can develop metastasis within two years, thereby adding to the future burden of mortality in this population.17 18

In the United Kingdom, the incidence of CRPC from 1999 to 2009 was estimated to be 3.8 per 100 person years among all patients with prostate cancer.17 The mean age was 76.8 years, and the mean survival after classification as castration resistant was 13.5 months.17 However, data from patients enrolled in clinical trials who subsequently develop CRPC suggests that median survival may actually be longer at about two to four years.19 20

The prevalence CRPC has been estimated to be 17.8% among patients with prostate cancer.15 A simulated model used data from several large clinical trials to estimate the epidemiology of various prostate cancer disease states and projected the incidence and prevalence of metastatic CRPC in the United States in 2020 to be 43 211 and 76 431 cases respectively.21 Such simulated models may be useful in clinical research and healthcare policy by forecasting the projected effect of a new drug on the incidence and prevalence of CRPC.

Definition and disease state

The treatment for metastatic hormone sensitive prostate cancer is based on suppression of circulating testosterone—that is, androgen deprivation therapy. The goal of ADT is to decrease circulating testosterone to “castrate levels,” which corresponds to a serum measurement of less than 50 ng/dL. In response to castrate levels of testosterone, a decrease in cancer cell proliferation occurs with subsequent induction of apoptosis. Despite the anti-proliferative response to ADT, cancer cells eventually develop resistance and most cases will manifest signs and/or symptoms of progression.22 23 At the point at which measurable progression of disease is present in the setting of castrate levels of testosterone, in the form of either sequential rises in prostate specific antigen (PSA) or imaging findings (computed tomography, magnetic resonance imaging, or radionuclide bone scintigraphy), the term “castration resistant” prostate cancer is applied. The term “castration resistant” is preferred over the previously used nomenclature such as “androgen independent” and “hormone refractory,” because despite absence of circulating testosterone the cancer is still functionally associated with androgens and the androgen receptor.23 24

Studies have found mutations in the androgen receptor gene that allow the receptor to remain functionally active in the setting of castrate testosterone levels.25 26 Furthermore, mutations of the androgen receptor ligand binding domain can lead to structural changes that allow activation of the receptor by ligands other than testosterone and dihydrotestosterone, including hydrocortisone and progesterone.25 26 27 28 Therefore, when serum testosterone falls to castrate levels, prostate cancer does not actually function independently of androgen and can still respond to androgen manipulation. With respect to biochemical progression after castration, the definition of PSA failure is based on the Prostate Cancer Clinical Trials Working Group 3 as a 25% or greater increase and an absolute increase of 2 ng/mL or more from the nadir PSA, confirmed thee or more weeks later.29

Sources and selection criteria

We searched PubMed from January 1990 to May 2016 using combinations of the following key words: “prostate cancer”, “androgen independent”, “hormone refractory”, “castrate resistant”, “metastases”, “anti-androgen”, “androgen deprivation therapy”, “bicalutamide”, “docetaxel”, “cabazitaxel”, “abiraterone”, “enzalutamide”, “radium-223”, and “sipuleucel-T”. We prioritized peer reviewed publications from prospective randomized controlled trials (RCTs) for inclusion. Specifically, for each approved therapeutic agent, we prioritized the relevant RCTs, as well as phase I and phase II studies, describing the research methods and results that led to that agent’s approval for use in CRPC. In addition, we included peer reviewed articles from retrospective reviews of large cohort studies, sub-analyses of data from RCTs, and high quality systematic reviews and meta-analyses. We included only full text, English language, peer reviewed publications. We also included data and references from peer reviewed organizational guidelines (American Urological Association (AUA), European Association of Urology (EUA), and National Comprehensive Cancer Network (NCCN)).10 11 12 13 29 We used to find active clinical trials relevant to the section on “emerging therapies.”

Treatment of non-metastatic castration resistant disease

As previously mentioned, about a third of patients with non-metastatic disease are at risk of progression within two years.18 On the basis of currently available guidelines, no approved agents exist for first line treatment of asymptomatic, non-metastatic CRPC. However, despite biochemical disease progression, continued ADT is recommended because the androgen receptor remains functionally active in this disease state. The lack of strong recommendation for a single agent by any guideline is due to the paucity of level I data showing a significant survival advantage with any particular therapy in non-metastatic CRPC. Furthermore, the potential side effects and cost associated with therapy, relative to any benefits, mean that their use is not warranted.10 11 13 Active and intense investigation of androgen receptor targeted therapy, such as ARN-509, is under way to determine its efficacy in this subset of patients with non-metastatic CRPC.30

For patients who do not elect to proceed with observation, the first generation anti-androgens (flutamide, bicalutamide, and nilutamide) and first generation androgen synthesis inhibitors (ketoconazole with steroid) may be used.10 11 Patients with a rising PSA on ADT may have up to a 50% decline in PSA with the addition of high dose (150 mg/day) bicalutamide as second line therapy.31 Although it is not the approved dosing regimen, increasing bicalutamide to 150 mg in patients on combined androgen blockade has been reported by one study to decrease PSA by at least 50%.32

Other manipulations beyond addition of anti-androgens have shown limited efficacy in non-metastatic CRPC. A recent phase II trial combining bicalutamide with dutasteride failed to show benefit over bicalutamide alone in patients with non-metastatic CRPC.33 In patients treated with combined androgen blockade, studies suggest that discontinuation of the anti-androgen may produce a modest reduction in PSA.34 35 36 Oral ketoconazole is used to inhibit androgen synthesis and may lead to a greater than 50% decline in PSA with a mean duration of 14.5 months.37 Other studies report a 56% response rate with the combination of ketoconazole, hydrocortisone, and dutasteride and a median duration of 20 months.38 The factor limiting the use of ketoconazole is the toxicity—mainly adrenal insufficiency, nausea, hepatotoxicity, and fatigue—which can occur with grade 3 level severity. Owing to the toxicity profile and marginal efficacy, the use of ketoconazole is rare in most countries where the newer agents are readily available.

Approved agents for metastatic CRPC are now being evaluated in the non-metastatic disease state of early CRPC. Recently, results of the STRIVE study were published; 396 men with non-metastatic (n=139) or metastatic (n=257) CRPC were randomly assigned to enzalutamide (160 mg/day) or bicalutamide (50 mg/day). ADT was continued in both arms.39 The primary endpoint was progression-free survival (PFS). Enzalutamide reduced the risk of progression or death by 76% compared with bicalutamide (hazard ratio 0.24, 0.18 to 0.32; P<0.001). The median PFS was 19.4 months with enzalutamide compared with 5.7 months with bicalutamide. Enzalutamide resulted in significant improvements in all key secondary endpoints including time to PSA progression (hazard ratio 0.19, 0.14 to 0.26; P<0.001), proportion of patients with a greater than 50% PSA response (81% v 31%; P<0.001), and radiographic PFS in metastatic patients (hazard ratio 0.32, 0.21 to 0.50; P<0.001).39 The benefits of enzalutamide over bicalutamide were shown in both the non-metastatic and metastatic groups. The researchers concluded that enzalutamide significantly reduced the risk of prostate cancer progression or death compared with bicalutamide in patients with non-metastatic or metastatic CRPC. Therefore, emerging data suggest a possible benefit of this agent in the non-metastatic disease state.

Treatment of metastatic castration resistant disease

Androgen synthesis inhibitors

Abiraterone acetate plus prednisone

Abiraterone inhibits the androgen synthesis pathway through blockade of the enzymes 17-α-hydroxylase and C17,20-lyase. The randomized double blind COU-AA-302 trial compared 1000 mg abiraterone acetate plus 5 mg prednisone twice daily against placebo plus prednisone in men with asymptomatic or minimally symptomatic metastatic CRPC without previous chemotherapy.40 The co-primary endpoints were radiographic PFS and overall survival. The final analysis showed a significantly longer median overall survival in the abiraterone acetate group than in the placebo group (34.7 (95% confidence interval 32.7 to 36.8) versus 30.3 (28.7 to 33.3) months; hazard ratio 0.81, 0.70 to 0.93; P=0.0033).41 Overall, the drug was well tolerated and no significant differences were seen in the occurrence of any adverse events between treatment and placebo in the interim analysis. On final analysis, the most common grade 3-4 adverse events were cardiac disorders (8% abiraterone acetate versus 4% placebo), increased alanine aminotransferase (6% v <1%), and hypertension (5% v 3%). Low dose oral prednisone is needed with abiraterone, to limit the side effects associated with excess mineralocorticoid.42 On the basis of the results of this trial, the AUA guidelines recommend abiraterone acetate for patients with asymptomatic or minimally symptomatic metastatic CRPC, as well as patients with symptoms with no previous docetaxel therapy and good performance status (as defined by the Eastern Cooperative Oncology Group).10 11 43

The COU-AA-301 trial compared abiraterone acetate plus prednisone versus placebo plus prednisone in patients with previous docetaxel therapy.44 An overall survival advantage was seen for patients taking abiraterone acetate in the interim analysis; on the final analysis, before unblinding and crossover at a median follow-up of 20.2 months, median overall survival with abiraterone acetate remained significantly better than with placebo (15.8 v 11.2 months; hazard ratio 0.74, 0.64 to 0.86; P<0.001).45 In a sub-analysis of patients with symptoms, which looked at pain control in the COU-AA-301 trial (post-docetaxel chemotherapy), those in the abiraterone acetate plus prednisone arm experienced more palliation (45% v 28.8%; P<0.001) and faster median time to palliation of pain (5.6 v 13.7 months; P=0.002) than did those in the placebo arm.46 The AUA guidelines therefore recommend abiraterone acetate for patients with symptomatic metastatic CRPC with previous docetaxel therapy and good performance status.10 11 Abiraterone acetate may also be offered to patients with symptomatic metastatic CRPC with previous docetaxel therapy and poor performance status, with the caveat that most of the patients in these trials had good performance status. Therefore, caution should be used with abiraterone acetate in patients with liver disease and congestive heart failure, as there were reports of fluid retention, cardiac events, and hepatotoxicity in those trials.47 Owing to the relatively low toxicity profile of abiraterone acetate compared with chemotherapy, it may be fairly well tolerated despite poor performance status. However, in a recent retrospective analysis of patients in a Canadian registry treated with abiraterone acetate, poor performance status was an independent predictor of worse overall survival on multivariate analysis, suggesting that abiraterone acetate should be used before a significant decline in performance status.48

First generation agents (ketoconazole)

Historically, oral ketoconazole may be used to suppress gonadal, as well as adrenal, androgen synthesis via inhibition of the CYP-11A and CYP-17A enzymes in the steroidogenesis pathway. A dose of 400 mg of oral ketoconazole and hydrocortisone every eight hours may be used to achieve a decline in PSA.37 49 Therefore, ketoconazole is listed in the AUA guidelines among the agents that can be used in metastatic CRPC but is rarely used in practice owing to toxicity and is not considered first line therapy.

Androgen receptor blockers


Enzalutamide is an androgen receptor inhibitor and is approved by the FDA for use in patients with metastatic CRPC who have received previous chemotherapy on the basis of the results of the AFFIRM trial, as well as in those without previous chemotherapy on the basis of the results of the PREVAIL trial.50 51 The AFFIRM trial investigated the use of enzalutamide versus placebo in 1199 men with metastatic CRPC and previous chemotherapy.50 Median overall survival was significantly longer with enzalutamide (160 mg daily) than with placebo (18.4 v 13.6 months; hazard ratio 0.63, 0.53 to 0.75; P<0.001). The PREVAIL trial was designed to compare enzalutamide with placebo in 1717 patients with no previous chemotherapy and co-primary endpoints of overall survival and radiographic PFS.51 It found a 29% reduction in risk of death (hazard ratio 0.71, 0.60 to 0.84; P<0.001) and an 81% reduction in the risk of radiographic PFS (0.19, 0.15 to 0.23; P<0.001) for enzalutamide compared with placebo. In addition, follow-up analysis of the AFFIRM trial showed significant benefit for enzalutamide with respect to pain progression, time to first skeletal related event, and health related quality of life.52 53

The more common side effects with enzalutamide versus placebo in the AFFIRM trial included fatigue, hot flashes, diarrhea, musculoskeletal pain, and headache. A similar side effect profile was noted in the PREVAIL trial, as well as increased rates of hypertension and falls in the enzalutamide arm. Notably, in the AFFIRM trial, five (0.6%) patients receiving enzalutamide had seizures compared with no patients in the placebo group. However, no difference in the incidence of seizures between enzalutamide and placebo was seen in the PREVAIL trial. Caution should be used when considering enzalutamide in patients with a history of seizures and stroke. Enzalutamide is recommended in patients with minimally symptomatic and symptomatic metastatic CRPC with or without previous chemotherapy and with good performance status.10 11 It may be used in patients with poor performance status; however, most patients in the AFFIRM and PREVAIL trials had good performance status.

Recent data from the TERRAIN trial showed that enzalutamide was superior to bicalutamide in men with asymptomatic/minimally symptomatic metastatic CRPC.54 Patients (n=375) were randomized to either enzalutamide or bicalutamide following progression on ADT. Patients receiving enzalutamide had improved median PFS compared with patients in the bicalutamide group (15.7 v 5.8 months; hazard ratio 0.44, 0.34 to 0.57; P<0.001).54 Further supporting these findings, the STRIVE trial (as discussed previously) showed similar results, with a decreased risk of prostate cancer progression or death for enzalutamide compared with bicalutamide in men with metastatic and non-metastatic CRPC.39

First generation agents (bicalutamide, nilutamide, flutamide)

Bicalutamide 50 mg/day has been shown to produce a PSA response in 20-40% of men with CRPC.55 Variable response rates, ranging from 20% to 57.1%, have been observed with second line anti-androgen in other studies, and the duration of response is approximately 11-27 months.56 Efficacy has also been shown with nilutamide 200-300 mg/day as a second line agent, with 64% of CRPC patients having an initial reduction in PSA.57

Discontinuation of the anti-androgen, referred to as the anti-androgen withdrawal phenomenon, may produce a decrease in PSA. Anti-androgen withdrawal was first described with flutamide and later in reports with bicalutamide and nilutamide.34 35 36 The Southwest Oncology Group (SWOG) conducted a study of anti-androgen withdrawal in 201 patients with CRPC (SWOG 9426) and reported a 21% PSA response rate.58 Most studies of anti-androgen withdrawal report a response in 15-30% of patients, with a duration of about three to six months.59



Sipuleucel-T is an immunotherapy that infuses the patient’s cultured, isolated antigen presenting cells, which have been activated against prostatic acid fused to granulocyte macrophage colony stimulating factor, thereby activating a host antigen specific T cell response against prostate cancer. The RCT of sipuleucel-T versus placebo (IMPACT trial) showed a 22% reduction in the risk of death with sipuleucel-T (hazard ratio 0.78, 0.61 to 0.98; P=0.03) in men with asymptomatic metastatic CRPC (18% previous exposure to chemotherapy) and a median survival improvement of 4.1 months compared with placebo (25.8 v 21.7 months).60 The AUA guideline for CRPC therefore recommends sipuleucel-T as standard therapy in minimally symptomatic/asymptomatic metastatic CRPC and does not recommend use in symptomatic disease or in patients with poor performance status.10 11 No significant difference was seen in time to disease progression, and few patients in the treatment (2.6%) and placebo (1.3%) arms achieved a PSA decline of at least 50%. Follow-up analysis of the IMPACT trial showed that a lower baseline PSA was associated with longer survival.61 Sipuleucel-T was well tolerated, and the most common adverse events with treatment included fevers, chills, fatigue, nausea, and headache. Sipuleucel-T is approved in the United States but is not currently available in Europe.12



Docetaxel leads to apoptosis through inhibition of microtubule assembly and mitotic arrest and was approved by the FDA as first line therapy for metastatic CRPC in 2004, on the basis of the results of the TAX-327 and SWOG 9916 trials.1 2 As a result, the AUA CRPC guideline categorizes patients on the basis of previous docetaxel or no previous docetaxel use. In TAX-327, patients receiving docetaxel every three weeks had a significantly decreased risk of death (hazard ratio 0.76, 0.62 to 0.94; P=0.009) compared with those randomized to receive mitoxantrone.1 Median survival was 18.9 months in the docetaxel every three weeks arm compared with 16.4 months in the mitoxantrone arm. The SWOG 9916 trial, which compared docetaxel plus estramustine with mitoxantrone plus prednisone, showed a 20% reduction in the risk of death and a longer median survival of 17.5 months with docetaxel plus estramustine versus 15.6 months (P=0.02) for mitoxantrone plus prednisone.2 Most patients had bony metastatic disease and good performance status, and 45-64% had pain symptoms. Adverse events occurred in both studies, with significantly more grade 3 and 4 toxicities in the docetaxel arms.

Docetaxel is therefore recommended in patients with asymptomatic/minimally symptomatic metastatic CRPC, as well as those with symptomatic metastatic CRPC and good performance status.10 11 Given that 35% of patients experienced reduction in pain in TAX-327 (compared with 22% in the mitoantrone arm; P=0.01), docetaxel may also be considered, albeit with caution, in patients with poor performance status specifically due to symptoms directly related to tumor burden such as rapid weight loss, weakness, and pain caused by direct tumor involvement and metastasis.

Studies have also shown that intermittent therapy, or a docetaxel re-challenge, may produce responses in patients with metastatic CRPC exposed to previous docetaxel therapy. A small phase II study of 45 patients reported a 25% partial PSA response after docetaxel re-treatment, which was well tolerated.62 The rationale behind this approach is that it allows patients who initially have a good response but in whom treatment is limited by reversible side effects to recover for a “holiday” period and then obtain additional treatment with an agent that has been shown to be efficacious. As more patients may now be treated at the time of presentation with hormone sensitive metastatic prostate cancer, the concept of a re-challenging at the time of progression in patients with metastatic CRPC may become more common.


Cabazitaxel, another taxane based chemotherapeutic agent, was one of the first treatments to be approved by the FDA after docetaxel. The TROPIC trial compared cabazitaxel with mitoxantrone plus prednisone in men with metastatic CRPC who had received previous docetaxel chemotherapy.63 An increase in overall survival of 2.4 months was seen in the cabazitaxel arm, corresponding to a 30% reduction in relative risk of death (hazard ratio 0.70, 0.59 to 0.83; P<0.001). Side effects were similar to those with docetaxel and included significant grade 3 or greater neutropenia in 82% and febrile neutropenia in 8% of patients. Cabazitaxel is recommended for use in patients with metastatic CRPC with good performance status and previous exposure to docetaxel. The AUA guideline recommends co-administration with agents providing neutrophil growth factor support.10 11 Additional toxicities of note include diarrhea (47%), fatigue 37%), nausea 37%), and vomiting (23%).

Radionuclide therapy


Radium-223 is an α emitting, bone seeking calcium mimetic that selectively targets and binds to increased areas of bone turnover in bone metastases. The drug is administered by intravenous injection at four week intervals for a total of six injections. The ALSYMPCA trial was a randomized, double blind, phase III study that compared six injections of radium-223 against placebo in men with CRPC and bony metastases who had received, were not eligible to receive, or had declined docetaxel chemotherapy.64 Median overall survival was longer with radium-223 than placebo (14.9 v 11.3 months; hazard ratio 0.70, 0.58 to 0.83; P<0.001). Subsequent subgroup analysis showed a survival benefit with radium-223, irrespective of previous docetaxel use.65 In addition, a significant improvement in median time to first symptomatic skeletal event was seen for radium-223 versus placebo (15.6 v 9.8 months; hazard ratio 0.66, 0.52 to 0.83; P<0.001). Radium-223 was well tolerated and associated with fewer adverse events than placebo. Although the difference was not significant, a slightly higher rate of diarrhea (25% v 15%) was seen with radium than with placebo. Other known side effects include nausea, vomiting, peripheral edema, and hematologic abnormalities (anemia, leukopenia, thrombocytopenia, neutropenia). A meaningful improvement in quality of life was also noted for radium compared with placebo.64 The AUA guideline recommends radium-223 for use in patients with CRPC with bony metastases, no known visceral metastases, and good performance status, irrespective of previous docetaxel use; in addition, it may also be offered to patients in this category with poor performance status.11

Older generation radionuclides (strontium, samarium)

Bone seeking radionuclides such as samarium-153 have been used for palliation of bone pain with modest results.66 A randomized, double blind, placebo controlled trial found that patients who received samarium-153 had significant improvement in pain scores compared with those given placebo.67 Bone seeking radionuclides may be used as an alternative to standard therapy in patients with metastatic CRPC and good, or poor, performance status with no previous docetaxel therapy, on the basis of the AUA guidelines.10 11 In addition, it may be used in a palliative setting for patients with metastatic CRPC with poor performance status and previous docetaxel use when symptoms result from bone pain.10 11

Sequencing of therapy

To date, no consensus exists regarding the optimal sequencing of therapies in CRPC. As mentioned previously, substantial changes are likely in the coming years owing to the major effect created by the CHAARTED and STAMPEDE trials showing survival benefit from earlier exposure to chemotherapy and likely altering the current CRPC treatment algorithms which are based on previous docetaxel use.5 6 10 11 12 68

The lack of head to head comparisons in CRPC means that the optimal agent for a given scenario is unclear. When selecting therapy for CRPC patients, therefore, several factors should be considered to customize the sequence of agents. Firstly, the presence or absence of metastatic disease is important and should be sub-stratified into visceral (for example, lung or liver) and non-visceral bone metastases. Radium-223 is well tolerated and has shown good efficacy in patients with bone metastases and no visceral metastases and can therefore be considered for early use in patients with a high burden of bone metastases with accompanying painful symptoms. As mentioned previously, owing to a long history of use before the emergence of other therapies, previous treatment with docetaxel chemotherapy should be considered; however, this is likely to change in the coming years as a result of the impending impact of chemotherapy and hormonal therapy in castration sensitive disease.

Secondly, the degree and severity of symptoms should be considered. The severity of symptoms a patient is experiencing may be an important driver of decision making in men with CRPC. An example of the importance of symptoms in sequencing is in the decision to use the approved immune based therapy. Sipuleucel-T was approved in patients with asymptomatic or minimally symptomatic metastatic CRPC.60 Importantly, patients were not evaluated in the setting of symptomatic disease, so current use of sipuleucel-T is in patients with minimal symptoms.

Thirdly, patients’ performance status is another important factor when deciding on appropriate treatment and which agent to use. For symptomatic patients with good performance status, most therapies are well tolerated and starting first line agents with established efficacy and a lower toxicity profile (for example, enzalutamide and abiraterone acetate) is generally preferable to maintain the performance status, rather than agents with the potential for side effects that may adversely affect performance status (for example, neutropenia from docetaxel). With regard to oral therapies (enzalutamide and abiraterone acetate), drug specific side effects should be considered when deciding which agent to start with; for example, in patients with signs of liver disease, congestive heart failure, or contraindications to steroid use, enzalutamide may be preferred. By contrast, patients with a history of seizures may be better served with abiraterone acetate.

Finally, recent evidence suggests that drug resistance may be present in patients who do not respond to abiraterone acetate and/or enzalutamide owing to the presence of androgen receptor variant AR-V7. A prospective study of 62 patients with advanced prostate cancer found that 39% of patients treated with enzalutamide and 19% of those treated with abiraterone acetate had detectable AR-V7 in circulating tumor cells.4 Patients who were positive for AR-V7 had lower PSA response rates and shorter time to progression. Therefore, in patients who rapidly progress on either oral therapy, the possibility of cross resistance due to AR-V7 should be considered, although this variant is not routinely tested for in clinical practice. However, studies have shown that patients who progress on enzalutamide may still have some response to abiraterone acetate, although the response is modest (11% with ≥0% PSA decline) and average duration of response is short (15.4 weeks).69 Studies in patients treated with abiraterone acetate have also shown that androgen receptor gene aberrations, such as copy number gain (duplication of a specific nucleotide sequence), may lead to resistance to treatment and worse overall survival, as well as decreased PFS.70 Patients with AR-V7 splice variants (functionally altered version of AR due to alternative splicing of RNA) may respond better to docetaxel chemotherapy, and future studies will need to be completed to determine whether treatment decisions based on molecular profiling will improve outcomes.

Treatment to minimize skeletal related events

Given that patients with CRPC have previous exposure to ADT and are typically older than those with hormone-naive disease, they are at risk for bone complications. Up to 90% of men with metastases have radiographically detectable bone lesions and, therefore, are at high risk for skeletal morbidity.1 2 71 Calcium and vitamin D may play a role in bone health, and these supplements are recommended as preventive therapy against skeletal events.10 Denosumab, a human monoclonal antibody directed against RANK ligand to inhibit osteoclast mediated bone destruction, has been shown to increase bone mineral density and decrease the incidence of vertebral fractures in men with non-metastatic prostate cancer, compared with placebo.72 Before denosumab, zoledronic acid (a bisphosphonate) was shown to significantly decrease the incidence of skeletal related events compared with placebo.73 However, a subsequent trial comparing denosumab with zoledronic acid showed a slightly longer time to skeletal related events with denosumab (20.7 v 17.1 months).74 Denosumab was associated with more significant hypocalcemia and, therefore, should be supplemented with calcium when administered. Both denosumab and zoledronic acid are associated with osteonecrosis of the jaw, so a dental examination is recommended before therapy is started. In addition, zoledronic acid is associated with decline in renal function and requires dose modification and close monitoring in patients with renal insufficiency.75 The AUA CRPC guideline recommends either denosumab or zoledronic acid for prevention of skeletal related events in men with metastases, although denosumab is the preferred first line agent owing to superiority in head to head comparison.10 11

Emerging therapies and future directions

The recent surge in novel treatments for CRPC has continued to drive additional therapeutic advances. Active research is taking place in the areas of targeted therapy, DNA repair, immunotherapy, and novel androgen inhibitors. Table 2 is a summary of actively recruiting, current phase III clinical trials in CRPC (

Table 2

Summary of active, recruiting phase III trials for castration resistant prostate cancer (as of May 2016,

View this table:

Despite enthusiasm, several eagerly anticipated studies have failed to produce promising results. Ipilimumab, a fully human monoclonal antibody that binds to the inhibitory cytotoxic T lymphocyte antigen 4 (CTLA-4), in combination with radiotherapy, failed to show a benefit in overall survival compared with placebo.8 Cabozantinib, a tyrosine kinase inhibitor, showed promising results with improved bone scan responses in a phase II non-randomized study in men with CRPC pretreated with chemotherapy, but it ultimately failed to show survival efficacy in phase III studies.76 77 Orteronel (TAK-700), a 17,20-lyase inhibitor, also failed to show improvement in overall survival in phase III studies.78 DCVAC, a novel dendritic cell based immunotherapy, is being studied in combination with docetaxel and has shown some activity in phase I/II studies.79 Data from the recent phase II trial, TOPARP, investigating the poly(adenosine diphosphate (ADP)–ribose) polymerase (PARP) inhibitor olaparib in metastatic CRPC showed good responses in patients with defects in DNA repair genes who had not responded to previous therapies.7 These data are exciting, as they are revolutionizing the landscape for a personalized approach to treating CRPC.7

In the non-metastatic CRPC field, much excitement exists about the ongoing studies of ARN-509 (Janssen, Raritan, NJ, USA) and ODM-201 (Bayer, Pittsburgh, PA), both investigational androgen receptor blockers.30 80 Both trials started recently and are actively recruiting participants. Enzalutamide is also being studied in patients with non-metastatic CRPC in the ongoing PROSPER (NCT02003924) and EMBARK (NCT02319837) trials (table 2).


Several guidelines exist to help clinicians in the management of CRPC.10 11 12 13 All are peer reviewed and based on high quality level I evidence from RCTs and data from well designed cohort studies and retrospective reviews. The AUA ( guideline on CRPC specifically discusses the disease state of CRPC. This is in contrast to the NCCN ( and EAU ( guidelines, which tackle the broader topic of advanced prostate cancer, including CRPC.

In contrast to the other guidelines, those of the AUA are formatted using the “index patient” classification, with the goal of personalizing treatment based on the patient’s symptoms, disease burden, and previous docetaxel use and performance status.10 11 The benefit of this approach is that the guidelines may be readily translatable to patients seen in clinical practice. Limitations include the presence of outliers who do not perfectly match any particular “index patient” and the artificial stratification based on docetaxel use, which will likely change in the coming years owing to the earlier use of chemohormonal therapy in hormone-naive prostate cancer.

The EAU guidelines make specific recommendations for medical therapy in CRPC but are not based on a particular patient scenario.12 Sipuleucel-T is not available in Europe; therefore, although it is mentioned in the publication, the EAU guidelines notably differ from those of the AUA and NCCN in that sipuleucel-T is not listed as recommended therapy in CRPC. All three guidelines recommend observation or clinical trial in the non-metastatic CRPC setting for most patients. However, in addition to anti-androgen manipulation and ketoconazole, the NCCN guideline also lists diethylstilbestrol, corticosteroids, and other estrogens as options for non-metastatic CRPC and in metastatic CRPC.13 Although these older agents have limited use in the United States and Europe, they may be more readily available and economically feasible in developing countries. The results of STRIVE (enzalutamide) and ongoing clinical trials will likely move newer oral agents that affect the androgen axis into the earlier non-metastatic CRPC disease state. The NCCN guidelines also stratify metastatic CRPC by location of metastases (visceral versus non-visceral) and specifically recommend docetaxel as first line treatment in patients with visceral metastases. The three guidelines are generally well aligned and are excellent tools for clinical guidance in CRPC. They may be used individually according to the preference of the clinician and relevance to their patient population and geographic locale.


Important advances have been made in the treatment of CRPC since the approval of docetaxel in 2004. Over the past decade, level I evidence has shown survival benefit in CRPC with an additional five agents (sipuleucel-T, abiraterone acetate, enzalutamide, cabazitaxel, and radium-223). All of these agents have shown a median survival benefit of two to five months. The best sequence for these multiple agents in CRPC remains to be elucidated; however, a stepwise, systematic approach to therapy based on type of metastases (bone versus visceral), symptoms, performance status, previous use of docetaxel, and side effect profile can help to customize and optimize treatment. Recent data supporting the earlier use of chemotherapy in hormone-naive prostate cancer, which has shown a significant survival advantage over ADT alone, will alter the future treatment paradigms in CRPC. The rapid development of new therapies and ongoing drug discovery offer potential for improved quality of life and survival in patients with CRPC. Finally, the ability to tailor therapies on the basis of clinical characteristics coupled with molecular profiles holds promise for further enhancing the survival of men with CRPC.

Questions for future research

  • What is the optimal sequence of agents in CRPC?

  • What other mechanisms exist for resistance to androgen inhibitors and chemotherapy in CRPC?

  • What role will novel biomarkers, such as AR-V7, play in personalizing treatment for CRPC?

  • Can certain novel anti-androgens be used earlier in the castration sensitive phase of prostate cancer, similar to chemotherapy?

  • Are there novel imaging tools that can be used to monitor disease response in CRPC?

Glossary of abbreviations

ADT—androgen deprivation therapy

AUA—American Urological Association

CRPC—castration resistant prostate cancer

EUA—European Association of Urology

FDA—Food and Drug Administration

NCCN—National Comprehensive Cancer Network

PFS—progression-free survival

PSA—prostate specific antigen

RCT—randomized controlled trial

SWOG—Southwest Oncology Group


  • Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: none.

  • Contributors: MSC conceptualized the review. MSC and CRR did the literature search, reviewed the referenced articles, and developed the outline. CRR wrote the first draft of the article, and MSC revised the first draft. MRC is the guarantor.

  • Provenance and peer review: Commissioned; externally peer reviewed.

  • Patient involvement:No patients were asked for input in the creation of this article.


View Abstract