Role of targeted therapy in the treatment of advanced prostate cancer


John M. Fitzpatrick, Department of Surgery, Mater Misericordiae Hospital and University College, Dublin, Ireland.


Over the past decade, the treatment of advanced prostate cancer has developed significantly, and perhaps the most dramatic shift came in 2004 with the demonstration that docetaxel-based chemotherapy significantly improved overall survival in patients with castration-resistant prostate cancer. This led to a significant expansion of the role of chemotherapy in the management of prostate cancer. In addition, there is now considerable progress being made in the development of more effective antiandrogens, cytochrome P17 inhibitors, novel chemotherapy regimens, targeted therapies, and immunotherapies that can complement existing therapies and may soon become integrated into the treatment paradigm. Progress in our understanding of molecular signalling pathways that play an important role in prostate cancer has stimulated the investigation of targeted therapies, including antiangiogenic agents, bone-targeted agents, and specific inhibitors of key signalling molecules and chaperone proteins. For the most part, targeted agents are being combined with chemotherapy, similar to the approach taken in other solid tumours. Various therapeutic vaccine strategies also appear to have potential in the treatment of advanced prostate cancer. However, the development of new approaches to the treatment of prostate cancer presents many challenges that will demand collaboration and consensus building with respect to biomarkers for patient selection, clinical endpoints, and trial designs.


androgen-deprivation therapy


PSA-doubling time


(metastatic) castration-resistant prostate cancer


overall survival


Southwestern Oncology Group


androgen receptor


vascular endothelial growth factor (receptor)


endothelin A receptor


phosphatidylinositol (-3-OH kinase)


IGF 1 (receptor)


mammalian target of rapamycin


phosphatase and tensin homologueue


mitogen-activated protein kinase


Response Evaluation Criteria in Solid Tumors


Satraplatin and Prednisone Against Refractory Cancer


progression-free survival


hazard ratio


tyrosine kinase inhibitor


Cancer and Leukaemia Group B


time to progression


nuclear factor-κB ligand


histone deacetylase


cytotoxic T-lymphocyte-associated antigen 4


prostatic acid phosphatase


granulocyte macrophage-colony stimulating factor


Prostate Cancer Clinical Trials Working Group


 Prostate cancer is the leading cause of cancer and second leading cause of cancer-related deaths among men in Europe and the USA [1]. Since the introduction of PSA level testing as a diagnostic marker ≈20 years ago, the annual incidence of prostate cancer has increased dramatically, but, fortunately, widespread PSA screening has resulted in earlier diagnosis, and prostate cancer-specific mortality has actually decreased in the USA since it peaked in the early 1990s [1]. Widespread screening of PSA levels has also resulted in a dramatic stage migration. Most patients are now diagnosed with low-risk, clinically localized disease that can be treated effectively with surgery and radiation [2,3]. Nevertheless, about 10–20% of patients will be diagnosed with locally advanced or metastatic disease while others will progress despite surgery, radiation, and androgen-deprivation therapy (ADT) [3,4], and advanced prostate cancer remains a significant treatment challenge.

Advanced prostate cancer is a remarkably heterogeneous disease, and treatment decisions are complex. Over the course of their disease, which may be 10–15 years in the PSA era [5], many patients will receive multiple therapies, including various forms of radiation therapy, hormonal therapy, and chemotherapy. Moreover, when making treatment decisions, a large number of variables must be considered, including PSA levels and PSA-doubling time (PSADT), Gleason score, disease stage and symptoms, prognosis, patient age and expectations, and the potential adverse effects of treatment on quality of life. Added to this complex picture is the challenge of precisely imaging and localizing disease burden, which can often be difficult with available imaging methods. Consequently, optimal care requires a multidisciplinary approach with input from the urologist, medical oncologist, radiation oncologist, radiologist, and pathologist who must collaborate in the diagnostic assessment and in making treatment decisions throughout the course of the disease.

Over the past decade, the treatment of advanced prostate cancer has developed significantly. ADT, either by surgical castration or chemical castration with a LHRH agonist, remains the cornerstone of treatment for advanced metastatic disease. However, most patients will progress within several years [6], and effective treatment options for castration-resistant prostate cancer (CRPC) remain limited. In the late 1990s, chemotherapy with mitoxantrone-based regimens was shown to provide some palliative benefits in CRPC [7,8], but no therapy had been shown to improve overall survival (OS). The most dramatic shift in the treatment paradigm came in 2004 with the demonstration in the TAX 327 and Southwestern Oncology Group (SWOG) 9916 trials that docetaxel-based chemotherapy provides a significant survival benefit in patients with CRPC [4,9]. This has led to a significant expansion of the role of chemotherapy in the management of prostate cancer, which is now being investigated in the neoadjuvant and adjuvant settings as well as in the management of locally advanced hormone-sensitive disease. Another important advance has been the development of prognostic nomograms beginning with the Kattan nomogram published in 1998, and more recently a nomogram was developed from the TAX 327 trial [10,11]. This has led to a gradual refinement in our ability to predict the prognosis of individual patients, select patients for clinical trials, and determine the best treatment approach for individual patients. However, many questions remain regarding the optimal sequence, timing, and duration of available therapies, and improved therapies are clearly needed. Fortunately, there has been considerable progress in the development of novel targeted therapies, but it remains to be seen how these will complement existing therapies and become integrated into the treatment paradigm.


Molecular signalling pathways that play an important role in prostate cancer include those associated with the androgen receptor (AR), vascular endothelial growth factor receptor (VEGFR), endothelin A receptor (ETAR), phosphatidylinositol-3-OH kinase (PI3K), epidermal growth factor receptor, and the IGF 1 receptor (IGF-1R). In particular, the PI3K/Akt/mammalian target of rapamycin (mTOR) pathway has a prominent role in the development and progression of prostate cancer (Fig. 1). This pathway is frequently activated in prostate cancer cells via reduced expression of the tumour suppressor protein, phosphatase and tensin homologueue (PTEN) [12]. Moreover, several receptors can activate this pathway, including VEGFR and IGF-1R, and the AR has shown significant molecular crosstalk with the PI3K/Akt/mTOR pathway.

Figure 1.

Selected signalling pathways with clinical relevance in prostate cancer. EGFR, epidermal growth factor receptor; T, testosterone; PLC-β, phospholipase C β; DAG, 1,2-diacylglycerol; PKC, protein kinase C; MEK, MAPK/extracellular signal-regulated kinase kinase.

A large body of preclinical and clinical evidence has implicated the PI3K/Akt/mTOR pathway in the regulation of prostate cancer growth and progression [12]. Inactivating mutations and/or deletion (i.e. loss of heterozygosity) of PTEN occur in most primary prostate tumours and many commonly used cell lines (e.g. PC-3 and LNCaP). Loss of PTEN function results in constitutive activation of the PI3K/Akt/mTOR pathway, as evidenced by an increase in phospho-Akt, and this correlates with higher pathological tumour stage and Gleason score, increased incidence of lymph node metastases, increased risk of biochemical recurrence, and androgen-independent growth. For CRPC, there is evidence that the PI3K/Akt/mTOR pathway may up-regulate expression and activation of the AR, particularly at low androgen concentrations, thereby driving tumour growth despite castrate levels of testosterone [13–15]. In turn, the AR can activate mTOR independent of PI3K/Akt activation [16].

Increased expression of VEGF and stimulation of the angiogenic pathway is also critically important to the growth and progression of prostate cancer. A wide variety of signalling molecules and growth factors can stimulate VEGF secretion, including hypoxia inducible factor-1, Src (a proto-oncogene tyrosine-protein kinase), Akt/protein kinase B, and IGF-1 [17]. Binding of VEGF to VEGFR-2 activates a cascade of downstream signalling pathways, including both the PI3K/Akt/mTOR pathway and the mitogen-activated protein kinase (MAPK) pathway in vascular endothelial cells [18]. Activation of the PI3K/Akt/mTOR pathway prolongs survival of endothelial cells, and the MAPK pathway stimulates proliferation, thus promoting neovascularization. In addition, tumour cells may express VEGF receptors, which allows VEGF to directly stimulate tumour growth and enhance survival [17]. Prostate cancer cells also secrete VEGF-C, which binds to VEGFR-3 and stimulates lymphangiogenesis. In a prostate cancer xenograft model, inhibition of VEGF-C or VEGFR-3 significantly reduced lymphangiogenesis and regional lymph node metastasis [19], and increased VEGF-C expression in primary prostate tumours was associated with increased lymph node metastasis and increased survival under oxidative stress [20,21]. Notably, protection from apoptosis associated with VEGF-C was mediated by mTOR [21]. Therefore, both VEGFR-2 and VEGFR-3 are important for the growth and progression of prostate cancer.

The IGF-1R is among the many receptor tyrosine kinases that regulate cellular proliferation, differentiation, and survival, and accumulating evidence over the past decade has indicated a critical role for IGF-1R in prostate cancer development and progression. Like many other growth factor receptors, IGF-1R signals through the PI3K/Akt and Ras-MAPK pathways. In prostate cancer, as in many other solid tumours, circulating IGF-1 levels and IGF-1R signalling are frequently up-regulated, and this is associated with more advanced and aggressive disease [22,23]. In preclinical models, IGF-1 significantly increased the invasiveness of DU145 cells, and this effect could be inhibited either by blocking IGF-1Rs or with specific inhibitors of MAPK or PI3K [24].

A common element to many of these pathways is the intracellular protein tyrosine kinase, Src, which is activated by any number of cell surface growth factor receptors (e.g. epidermal growth factor receptor, VEGFR, and IGF-1R), plays a role in AR signalling, and interacts with a variety of downstream signalling pathways (e.g. PI3K/Akt, Ras/Raf, and MAPK) [25,26]. Src family kinases are highly expressed in prostate cancer where they regulate tumour growth and metastasis, androgen-induced proliferation, and the transition to androgen-independent growth [25]. Src family kinase activity is also up-regulated in CRPC [27]. This is consistent with studies showing that inhibition of Src activity suppresses androgen-independent growth and metastasis [28]. In addition, Src is involved in regulating IGF-1R expression. In vitro studies have shown that androgens stimulate up-regulation of IGF-1R, and this is dependent on Src activity [29]. Src signalling also regulates bone metabolism, primarily osteoclast activity, and appears to be involved in tumour metastasis to bone. Therefore, Src is an important target in prostate cancer that has implications for the development of CRPC and bone metastases.

Finally, the ETAR also plays a key role in prostate cancer by regulating tumour growth, angiogenesis, and metastasis to bone [30]. Prostate cancer cells typically produce endothelin 1, which binds to ETAR expressed on both prostate cancer cells and osteoblasts, thereby stimulating cellular proliferation and migration, increasing resistance to apoptosis, and promoting secretion of proangiogenic factors (e.g. VEGF) from tumour cells. In particular, increased proliferation and survival of osteoblasts results in the formation of painful osteoblastic lesions characterized by new bone formation. It is also noteworthy that endothelin 1 may contribute to bone pain by directly stimulating nociceptors through the ETAR.

A wide range of molecularly targeted therapies are currently in clinical development to target these specific pathways. However, because of the diversity of advanced prostate cancers and its capacity to adapt to changing conditions, the best approach may be to combine targeted agents in an effort to inhibit multiple pathways.


Novel hormonal therapies

Disease progression despite effective ADT is a significant event in the natural history of prostate cancer because CRPC is clinically aggressive and is associated with a poor prognosis (median survival of 15–20 months) [4,9,10]. However, tumour progression despite castrate levels of testosterone does not indicate that tumour growth is completely androgen independent. Indeed, CRPC is often associated with increased expression of ARs, suggesting that the tumour may still be dependent on androgen signalling to sustain growth, and intratumoural androgen biosynthesis may fuel that growth. This hypothesis is further supported by evidence that addition of estramustine (a mustard-oestradiol conjugate with hormonal effects) to chemotherapy delays time to PSA progression and improves OS in patients with CRPC [31]. Currently available antiandrogens that are often used with LHRH agonists, including flutamide, bicalutimide, and nilutimide, exhibit only moderate binding affinity for the AR [32]. This has led to the development of more potent AR inhibitors. Two novel AR inhibitors, MDV3100 (Medivation) and RD162 have shown antitumour activity in murine models of CRPC, and both are orally bioavailable [33]. MDV3100 showed promising activity in a phase I/II study in 140 patients with CRPC. Among 114 patients treated with doses ranging from 30 to 360 mg/day and followed for >12 weeks, 57% of chemotherapy naïve patients (65 patients) and 45% of patients previously treated with chemotherapy (49) had a >50% PSA level decline [34]. A dose of 240 mg/day was recommended for further study. A phase III, randomized, placebo-controlled trial in patients with CRPC previously treated with docetaxel-based chemotherapy (AFFIRM; NCT00974311) is currently ongoing [35].

An alternative strategy to block androgen-dependent growth of CRPC is to inhibit testosterone biosynthesis. Abiraterone acetate (CB7630; Cougar Biotechnology, Los Angeles, CA, USA) is a potent, orally bioavailable, small molecule inhibitor of cytochrome P17, which catalyses two key reactions (17-α hydroxylase and 17,20 lyase) involved in androgen biosynthesis [36], thereby dramatically reducing both adrenal and intratumoural androgen production. A phase I/II study of abiraterone in 54 patients with CRPC showed that among 42 patients treated at the phase II dose (1000 mg/day), 67% had a ≥50% decline in serum PSA level, and 19% had a ≥90% decline from baseline [37]. In addition, nine of 24 patients (37.5%) with measurable disease had a partial response by Response Evaluation Criteria in Solid Tumors (RECIST). These results clearly support the hypothesis that CRPC remains sensitive to endocrine therapy in most patients, and show that abiraterone has promising activity for the treatment of CRPC. Based on these results, several phase II and phase III trials have been initiated. A phase III trial in patients previously treated with docetaxel (NCT00638690) has recently completed enrolment of 1158 patients; OS is the primary endpoint. A second phase III trial has been initiated in patients with asymptomatic or mildly symptomatic metastatic disease who have not received prior chemotherapy or ketoconazole (NCT00887198) [38,39].

A key question is whether these novel hormonal therapies will redefine CRPC and shift the treatment paradigm. Conceivably, the introduction of these novel hormonal therapies either at development of CRPC or earlier in the treatment paradigm, perhaps as an adjunct to LHRH agonists, could delay the need for chemotherapy and potentially alter the natural history of prostate cancer.

Chemotherapy agents

The current standard of care for treating CRPC is systemic chemotherapy with a docetaxel-based regimen. This is based on two landmark studies demonstrating that docetaxel plus prednisone (TAX 327) or docetaxel plus estramustine (SWOG 9916) significantly improved OS compared with mitoxantrone plus prednisone [4,9,40], which had been the standard for more than a decade based on evidence that it provides palliative benefits (i.e. reduced pain and improved quality of life) in patients with CRPC [41–43]. The demonstration that chemotherapy can, indeed, improve survival in this patient population and to an extent similar to that seen in other chemotherapy-sensitive tumour types has refuted the notion that CRPC is chemotherapy refractory and reinvigorated efforts to develop more effective chemotherapy regimens and determine the best agents to use as second-line therapy. Several agents have been investigated in the second-line setting, including satraplatin, mitoxantrone, vinorelbine, ixabepilone, and patupilone (Table 1) [44–50].

Table 1.  Selected clinical studies of chemotherapy for CRPC
DrugClassNStudy design/patientsResultsReference
  1. NIH, National Institutes of Health; PR, partial response; pt, patient.

Satraplatin + prednisoneOral platinum analogue950Randomized, phase III (SPARC)PFS benefit (HR 0.69; P < 0.001)Sternberg et al.[50]
Control: prednisone + placeboPain progression (HR 0.67; P < 0.001)
All pts failed prior chemotherapy; 51% pts failed prior docetaxelPSA response rate: 25% vs 12% (P < 0.001)
Objective response rate: 7% vs. 1% (P < 0.002)
Cabazitaxel (XRP6258)Taxane25Phase I
Pts with solid tumours
PR (RECIST) in 2 pts with prostate cancerMita et al.[47]
Cabazitaxel + prednisone 720Randomized, phase III (TROPIC) vs mitoxantrone + prednisone
All pts failed prior docetaxel
Primary endpoint: OS
NCT00417079 [44]
IxabepiloneEpothilone42Phase II
chemotherapy naïve mCRPC
PSA response rate: 33%
Median PFS: 6 months
Median OS: 18 months
Hussain et al.[46]
Ixabepilone ± estramustine 92Phase IIPSA response rate: 48% monotherapyGalsky et al.[45]
 Chemotherapy naïve CRPC69% with estramustine
 PR rate: 32% monotherapy
 48% with estramustine
Ixabepilone + prednisone 41Randomized, phase IIPSA response rate: 17% vs 20%Rosenberg et al.[49]
vs mitoxantrone + prednisone
Docetaxel-refractory pts
Ixabepilone + mitoxantrone + prednisone 36Phase I, dose-escalation
mCRPC pts previously treated with docetaxel
PSA response rate: 31% overall 43% in pts receiving ≥30 mg/m2 ixabepiloneRosenberg et al.[48]

Satraplatin (GPC Biotech, Martinsried/Munich, Germany) is an oral platinum analogue that has shown activity against CRPC cell lines and taxane-resistant cell lines in vitro and may be a good candidate for use in CRPC after failure of docetaxel [51]. Results from a prematurely terminated, randomized, phase III trial combining satraplatin with prednisone as first-line therapy for CRPC suggested that it has promising antitumour activity compared with prednisone alone [52], and this led to a randomized, placebo-controlled, phase III trial of this regimen as second-line therapy. The Satraplatin and Prednisone Against Refractory Cancer (SPARC) trial enrolled 950 patients who had failed prior chemotherapy (51% had received prior docetaxel), and patients were randomized to satraplatin (80 mg/m2× 5 days q5 weeks) plus prednisone or prednisone plus placebo [53]. The primary endpoint was progression-free survival (PFS), defined by radiological progression, symptomatic progression (i.e. increased pain or need for radiation to bone), skeletal events, or death. This study showed a highly significant improvement in PFS (hazard ratio [HR] 0.69; P < 0.001) and pain progression (HR 0.67; P < 0.001). In addition, patients in the satraplatin arm had a 25% PSA response rate, a 7% objective tumour response rate, and a 24% pain response rate, which were all significantly better than with prednisone alone (12%, 1%, and 14%, respectively) [53]. Unfortunately, there was no OS benefit, and satraplatin was not approved.

Cabazitaxel (XRP6258; Sanofi-Aventis) is a novel taxane that is being investigated in CRPC. One of the key features of cabazitaxel is a low affinity for the multidrug resistance transporter, P-glycoprotein, and studies have shown that cabazitaxel is active in patients with advanced prostate cancer and in taxane-resistant metastatic breast cancer [47,54]. Therefore, an open-label randomized phase III trial (TROPIC) comparing cabazitaxel/prednisone with mitoxantrone/prednisone is being conducted in 720 patients with CRPC who were previously treated with docetaxel [44]; the primary endpoint in this trial is OS.

The epothilones are also being actively investigated for the treatment of CRPC. In general, epothilones appear to be less susceptible to resistance mechanisms such as P-glycoprotein or tubulin mutations that render cells resistant to taxanes [55]. Ixabepilone (BMS-247550; Bristol Myer Squibb) is a synthetic epothilone B analogue that has shown activity against taxane-refractory prostate cancer cell lines in vitro[56]. In two phase II studies in chemotherapy naïve patients with CRPC, ixabepilone (35 or 40 mg/m2 q3 weeks) produced a ≥50% PSA level decline in 33% and 48% of patients, which improved to 69% when ixabepilone was combined with estramustine [45,46,57]. These promising results prompted further study in docetaxel-refractory patients. In this setting, a randomized phase II trial (41 patients) showed that ixabepilone/prednisone has activity similar to that of mitoxantrone/prednisone (17% and 20% of patients, respectively, had a ≥50% decline in PSA levels) [49]. Based on these results, a subsequent phase I study investigated the triple combination of ixabepilone plus mitoxantrone and prednisone in 36 patients previously treated with docetaxel-based chemotherapy. The recommended phase II dose of ixabepilone was 35 mg/m2 combined with 12 mg/m2 mitoxantrone every 3 weeks plus 5 mg prednisone twice daily. Pegfilgrastim was administered because of the high incidence of grade 3/4 neutropenia. Overall, this combination yielded a ≥50% PSA level decline in 31% of patients, and among patients who received ≥30 mg/m2 ixabepilone, 43% had a ≥50% PSA level decline [48].

Targeted agents as adjuncts to chemotherapy

Given the recent success of chemotherapy for the treatment of CRPC, there is a strong rationale to incorporate targeted agents into established regimens in an effort to further improve clinical outcomes. For the most part, targeted agents are being combined concurrently with chemotherapy, similar to the approach taken in other solid tumours, but the value of this approach has yet to be shown in CRPC, and several important questions remain unanswered.

Angiogenesis inhibitors as a class appear to have potential in the treatment of prostate cancer, as they do in many other solid tumours, because angiogenesis and lymphangiogenesis play important roles in the process of tumour progression and metastasis. Microvessel density and plasma VEGF levels are both independent prognostic factors in prostate cancer [58–60], and, in addition to stimulating neovascularization of the tumour, there is evidence to suggest that tumour-derived VEGF may directly stimulate tumour growth via activation of VEGFRs expressed on tumour cells [61]. Several classes of antiangiogenic agents are being investigated in prostate cancer (Table 2) [62–77]. These include: anti-VEGF antibodies such as bevacizumab (Genentech and Roche); VEGF Trap, aflibercept (Sanofi-Aventis and Regeneron); several VEGFR tyrosine kinase inhibitors (TKIs), including the multitargeted TKIs sorafenib (BAY 43–9006; Bayer and Onyx) and sunitinib (SU11248; Pfizer); and thalidomide or its derivative lenalidomide (Celgene).

Table 2.  Selected clinical studies of antiangiogenic agents in prostate cancer
DrugClassNStudy design/patientsResultsReference
  1. NIH, National Institutes of Health; PR, partial response; BRM, biological response modifier; pt, patient.

Bevacizumab + docetaxel + estramustineVEGF mAb79Phase II (CALGB 90006)PSA response rate: 65%Picus et al.[74]
Chemotherapy naïve CRPCPR rate: 53% in 17 pts
Median TTP: 10 monthsArmstrong et al.[65]
Median OS: 21 months
Bevacizumab + docetaxel/prednisone 1020Randomized, phase III (CALGB 90401)
Chemotherapy naïve CRPC
Primary endpoint: OS
NCT00110214 [63]
Bevacizumab + docetaxel/prednisone + thalidomide 54Phase IIPSA response rate: 87%Ning et al.[73]
Chemotherapy naïve mCRPCObjective response rate: 28%
Median PFS: 18 months
Bevacizumab + docetaxel 20Phase IIPSA response rate: 55%Di Lorenzo et al.[69]
Previously treated with docetaxel
Thalidomide + docetaxelBRM75Randomized, phase IIPSA response rate: 53% vs 37%Dahut et al.[67]
vs. docetaxel (n= 25)Median PFS: 5.9 vs. 3.7 months
Chemotherapy naïve mCRPCOS rate at 18 months: 68% vs. 43%
Thalidomide + docetaxel + estramustine 20Phase IIPSA response rate: 90%Figg et al.[71]
Chemotherapy naïve CRPCPR rate: 20% in 10 pts
Median TTP: 7.2 months
Thalidomide + paclitaxel + estramustine 38Phase I/IIPSA response rate: 76%Mathew et al.[72]
Previously treated with chemotherapyObjective response rate: 28% in 18 pts
SunitinibVEGFR multikinase inhibitor34Phase II in CRPCPSA response rate: 6% in both groupsDror Michaelson et al.[70]
Chemotherapy naïve (n= 17)
Previously treated (n= 17)
Sunitinib 36Phase IIPSA response rate: 12%Sonpavde et al.[76]
Previously treated with docetaxelMedian PFS: 4.5 months (19.4 weeks)
SorafenibVEGFR multikinase inhibitor22Phase IINo PSA responseDahut et al.[68]
mCRPC (59% pts had received prior chemotherapy)2 pts had regression in bone lesions
Sorafenib 28Phase IIPSA response rate: 4%Chi et al.[66]
Chemotherapy naïve CRPC
Sorafenib 55Phase IIPSA response rate: 4%Steinbild et al.[77]
Chemotherapy naïve CRPC
Aflibercept + docetaxel/ prednisoneVEGF1200Randomized, phase III (VENICE)Primary endpoint: OSNIH
Trapvs placebo + docetaxel/prednisone NCT00519285 [62]
Chemotherapy naïve CRPC 

To date, bevacizumab has been most extensively studied in CRPC, and although it has shown only modest single-agent activity, it appears to be active when used in combination with chemotherapy. Several studies have combined bevacizumab with standard docetaxel-based regimens as first-line therapy for CRPC. In a phase II trial conducted by the Cancer and Leukaemia Group B (CALGB 90006), treatment with bevacizumab (15 mg/kg q3 weeks) plus docetaxel/estramustine resulted in a >50% PSA response in 79% of patients, partial response in 14 of 33 (42%) patients with measurable disease, median time to progression (TTP) of 10 months, and median OS of 21 months [64,65,74]. Based on these favourable results, a randomized, placebo-controlled, phase III trial (CALGB 90401; NCT00110214) was initiated to test whether the addition of bevacizumab (15 mg/kg q3 weeks) to standard docetaxel/prednisone can improve OS [65]. Accrual of more than 1000 patients is complete, and the results are eagerly awaited [63]. Another phase II trial that combined bevacizumab with docetaxel/prednisone and thalidomide (200 mg/day) also showed a high PSA response rate. Among 54 patients with metastatic CRPC (mCRPC), a median Gleason score of 8, and rising PSA levels (median 99 ng/mL; median PSADT 1.6 months), 47 patients (87%) had a ≥50% PSA level decline, and median PFS was 18 months [73]. Moreover, among 29 patients with measurable disease, 15 (28%) had an objective response (one complete and 14 partial). Finally, an Italian study has investigated bevacizumab in the second-line setting. In this small phase II study, 20 patients who had failed first-line docetaxel were treated with bevacizumab (10 mg/kg) plus docetaxel (60 mg/m2) every 3 weeks, and 11 patients (55%) achieved a >50% PSA response, including four patients who had not responded to prior docetaxel [69]. Taken together, these data suggest that bevacizumab may enhance the activity of both first- and second-line chemotherapy for CRPC, but the results of randomized trials are needed to confirm the clinical benefit of bevacizumab in this setting.

Several phase II studies investigating thalidomide in combination with taxane-based chemotherapy have reported high PSA response rates (ranging from 53% to 90%) [67,71,72]. Among them was a randomized phase II study that showed a survival benefit when thalidomide was combined with weekly docetaxel. Unfortunately, this study was conducted before the q3-week docetaxel regimen was shown to be superior to the weekly regimen. Investigators from the National Cancer Institute randomized 75 chemotherapy naïve patients with mCRPC to weekly docetaxel (30 mg/m2) alone or combined with thalidomide (200 mg/day) [67]. At a median follow-up of 26 months, they reported an improvement in PFS (median 5.9 months vs 3.7 months with docetaxel alone), and the 18-month survival rate was higher in the group that received thalidomide (68% vs 43%) [67]. Subsequently, at a median follow-up of 47 months, they reported a significant improvement in median OS (26 months vs 15 months; P= 0.041) [64,78]. Other studies have combined thalidomide with a combination of paclitaxel (100 mg/m2/week × 2 out of 3 weeks) and estramustine (140 mg/m2 q8 h × 5 days for 2 out of 3 weeks), and although this regimen resulted in favourable PFS (median 7.2 months) in one study [71], the toxicity of estramustine has dampened enthusiasm for this drug. In addition, thalidomide is associated with an increased risk of thromboembolic events, prompting use of prophylactic anticoagulants.

Available data on the activity of other antiangiogenic agents in CRPC (e.g. aflibercept, sorafenib, sunitinib, and lenalidomide) are more limited and preliminary. Most studies investigating these agents as monotherapy have not reported significant activity based on PSA response rates. However, phase II data suggest that they may have some utility when combined with chemotherapy. It is also important to note that in the case of sorafenib, PSA response may not be a useful measure of antitumour activity based on evidence that sorafenib may increase PSA secretion from prostate tumours [68]. This highlights one of the potential challenges when assessing the clinical activity of cytostatic agents in the setting of prostate cancer, which relies heavily on PSA as a biomarker of response. Sorafenib is a multitargeted TKI that inhibits b-Raf, platelet-derived growth factor receptor, and VEGFR-2, and it has shown both antiproliferative and antiangiogenic activity in various tumour types [79]. Several phase II studies of sorafenib monotherapy (400 mg twice daily) in CRPC have reported only modest antitumour activity based on PSA response but improvements in bone metastases have been observed [66,68,77]. Ongoing studies are combining sorafenib with standard chemotherapy regimens.

Sunitinib is another mulitargeted TKI specific for VEGFR and the platelet-derived growth factor receptor that has shown clinical benefit in the treatment of metastatic RCC and gastrointestinal stromal tumours. Preclinical studies have shown that sunitinib (either alone or combined with low-dose docetaxel) inhibits the growth of human CRPC (DU-145) in xenograft models [80]. Two recent phase II studies have investigated the activity of sunitinib monotherapy in CRPC in both the first- and second-line settings [70,76]. Although PSA response rates were low in both studies, there was evidence of radiological responses in the absence of PSA decline, suggesting that PSA response may not be the best endpoint for assessing the activity of sunitinib. Ongoing studies include a phase I/II safety and pharmacokinetic study (NCT00137436) and a pilot phase II study (NCT00879619) of sunitinib combined with standard docetaxel/prednisone as first-line therapy for CRPC. Sunitinib is also being investigated as monotherapy in the second-line setting for CRPC and in patients undergoing radical prostatectomy in an effort to inhibit distant metastasis.

Finally, aflibercept (VEGF Trap) is a recombinant fusion protein consisting of the human VEGF extracellular domains fused to the Fc domain of an IgG antibody, and it has higher affinity for VEGF than bevacizumab in vitro[81]. Aflibercept has shown promising activity in combination with various chemotherapy regimens for the treatment of advanced solid tumours (primarily colorectal, breast, and ovarian cancer) and is currently being investigated in combination with standard docetaxel/prednisone for the first-line treatment of mCRPC. This randomized, placebo-controlled, phase III trial (VENICE; NCT00519285) will enroll 1200 patients, and OS is the primary endpoint [62].

Bone-targeted agents

The bone is clearly an important target in advanced metastatic prostate cancer given that most patients will develop bone metastases at some point during the course of their disease, and most disease-related symptoms are directly related to bone metastases. For that reason, several bone-targeted agents have been developed, most notably the bisphosphonates such as zoledronic acid (Novartis), which are highly effective at stabilizing the bone and preventing skeletal complications in patients with bone metastases. Moreover, there is growing evidence that combining bone-targeting radiopharmaceuticals with chemotherapy may result in better outcomes in patients with bone metastases [82,83].

More recently, a nuclear factor-κB ligand (RANKL) inhibitor, denosumab (Amgen), has been developed for the treatment of bone metastases. RANKL is involved in the regulation of bone metabolism and is overexpressed in osteoblasts associated with prostate cancer bone metastases in mice [84]. Denosumab (120 mg s.c. q4 weeks) has been shown to significantly reduce and delay skeletal-related events (defined as pathological fractures, radiation or surgery to bone, or spinal cord compression) similar to zoledronic acid in patients with bone metastases from breast cancer and other solid tumours [85,86]. A phase III, randomized, non-inferiority trial in patients with bone metastases from CRPC is currently comparing the benefit of denosumab and zoledronic acid based on time to first on-study skeletal-related event (NCT00321620) [87]. A randomized phase II study also reported that denosumab (180 mg q4 weeks) was significantly more effective than continued therapy with an IV bisphosphonate at reducing urinary N-telopeptide levels, a biomarker of bone resorption, in patients with bone metastases (Table 3) [87–97]. In addition to these studies, a large randomized, placebo-controlled trial was conducted to investigate the benefit of denosumab in patients receiving ADT, which can cause significant bone loss. Treatment with denosumab (60 mg q6 months) significantly increased bone mineral density of the lumbar spine at 24 months (+5.6% vs −1%; P < 0.001) and significantly reduced the incidence of new vertebral fractures (1.5% vs 3.9%; P= 0.006) compared with placebo [96,97]. Another large placebo-controlled study designed to determine whether denosumab can prolong bone metastasis-free survival in men with CRPC is currently ongoing (NCT00286091) [88].

Table 3.  Selected clinical studies of bone-targeted agents in development for prostate cancer
DrugClassNStudy design/patientsResultsReference
  1. NIH, National Institutes of Health; uNTX, urinary N-telopeptide; BP, bisphosphonate; BMD, bone mineral density; NS, statistically nonsignificant.

DenosumabRANKL mAb  111Randomized, phase II
vs IV BPs
Patients with bone metastases
SREs 8% vs 17% pts
At 13 week, uNTX reduced to
<50 nmol/L. All pts: 71%
denosumab vs 29% IV BPs
PC: 69% denosumab vs 19% IV BPs
Fizazi et al.[92,93]
Denosumab 1468Randomized, double-blind, phase III vs placebo
Patients receiving ADT
At 24 months, lumbar spine BMD: + 5.6% with denosumab vs
−1.0% with placebo. New vertebral fractures: 1.5% vs 3.9%
Smith et al.[96,97]
Denosumab 1904Randomized, double-blind, phase III vs zoledronic acid
CRPC with bone metastases
Primary endpoint: time to first on-study SRE.
NCT00321620 [87]
Denosumab 1435Randomized, double-blind, phase III vs placebo nonmetastatic CRPCPrimary endpoint: bone metastasis-free survival.
NCT00286091 [88]
AtrasentanETAR antagonist288Randomized, phase II
vs. placebo
Asymptomatic mCRPC
196 vs 129 days (evaluable; P= 0.02)
155 vs. 71 days (PSA; P= 0.002)
Median TTP: 183 days 10 mg
artrasentan vs. 137 days placebo (ITT analysis);
Carducci et al.[91]
Atrasentan 941Randomized, phase III
vs placebo CRPC
PSADT increased (P= 0.03)
Median OS: HR 0.92 (P= NS)
Median TTP: 764 vs 671 days, (P= NS)
PSA progression: 254 vs. 253 days
Nelson et al.[95]
Atrasentan 809Randomized, phase III
vs placebo mCRPC
Median TTP: HR 1.26 (P= 0.007) in prespecified subset of compliant ptsCarducci et al.[90]
Docetaxel/prednisone ± atrasentan 930Randomized, phase III
vs placebo (SWOG S0421)
CRPC with bone metastases
Primary endpoints: OS and PFS
NCT00134056 [89]
ZD4054312Randomized, phase II
vs. placebo
CRPC with bone metastases and no pain or mild pain
Median TTP: NS
Median OS: 24.5 months
10 mg ZD 4054 vs 17.3 months placebo (P= 0.008)
James et al.[94]
Docetaxel ± ZD4054 1044Randomized, phase III
vs placebo (ENTHUSE M1C)
CRPC with bone metastases
Primary endpoint: OS
NCT00617669 [100]
Best supportive care ± ZD4054 580Randomized, phase III
vs placebo (ENTHUSE M1)
CRPC with bone metastases and no pain or mildly symptomatic
Primary endpoint: OS
NCT00554229 [101]
Existing therapy ± ZD4054 1500Randomized, phase III
vs placebo (ENTHUSE M0)
nonmetastatic CRPC
Primary endpoints: PFS and OS
NCT00626548 [102]

Based on the importance of ETAR in the biology of prostate cancer and development of osteoblastic lesions, two oral selective ETAR antagonists (atrasentan and ZD4054) have been developed and have shown clinical benefits in patients with advanced prostate cancer (Table 3). Two large randomized, placebo-controlled, phase II and phase III trials of atrasentan (Abbott Laboratories) in CRPC have reported significant delays in PSA progression or PSADT and disease progression in bone based on biochemical markers (e.g. bone alkaline phosphatase); however, both trials failed to show significant improvements in TTP, based on intent-to-treat analysis of radiological or clinical progression, or improvements in OS [91,95,98]. A second randomized, placebo-controlled, phase III study in patients with mCRPC was closed because of unexpected early progression in most patients, but showed a significant improvement in TTP in a prespecified subset analysis of protocol-compliant patients [90,98]. Moreover, an exploratory subset analysis of patients with bone metastases at baseline (59%) also suggested a significant delay in TTP, thus lending support to the hypothesis that these agents may provide the greatest benefit in patients with bone disease [98]. Studies investigating atrasentan (10 mg/day) combined with standard docetaxel/prednisone are currently ongoing, including a large, randomized, phase III trial being conducted by the SWOG S0421 and NCT00134056 that will assess PFS and OS [89].

In addition, results of a randomized phase II trial suggested that ZD4054 (Astrazeneca) at a dose of 10 mg/day may improve OS despite lack of improvement in TTP (primary endpoint) in patients with bone metastases from CRPC [94]. Based on these encouraging results, three large phase III trials have been initiated to determine whether addition of ZD4054 to standard therapy can delay bone metastasis or improve OS in patients with mCRPC and nonmetastatic CRPC [99–102]. ENTHUSE 33 (NCT00617669) is similar to the SWOG S0421 trial described above and will investigate whether addition of ZD4054 to docetaxel can improve survival in patients with bone metastases.

Other targeted approaches

Several other targeted approaches to the treatment of CRPC are also being explored based on evidence implicating various signalling pathways in tumour growth and progression or novel approaches to overcome resistance mechanisms (Table 4) [103–109]. For example, the cytoprotective chaperone protein, clusterin, has been shown to inhibit apoptosis and protect tumour cells from the cytotoxic effects of radiation and chemotherapy, and clusterin is overexpressed in various human malignancies, including breast, lung, colon, and prostate cancer [110,111]. This led to the development of a modified clusterin antisense oligonucleotide, OGX-011 (OncoGeneX), which has been shown to enhance apoptosis in response to docetaxel in a docetaxel-refractory PC-3dR cell line [112]. Based on a phase I study that established the feasibility of combining OGX-011 with docetaxel and showed activity in patients with advanced prostate cancer [110], a randomized phase II study was conducted in 82 patients with CRPC. The combination of OGX-011 (640 mg/week) with standard docetaxel (q3 week)/prednisone did not significantly improve PSA response rate or PFS but significantly improved median OS by 10 months (adjusted HR = 0.54; P= 0.04) [106]. Although this intriguing result will require confirmation in a larger randomized phase III trial, it suggests that this approach may have merit.

Table 4.  Selected clinical studies of other targeted agents in development for prostate cancer
DrugClassNStudy design/patientsResultsReference
  1. PR, partial response; NTX, N-telopeptide.

OGX-011 + docetaxel/ prednisoneClusterin antisense oligonucleotide82Randomized, phase IIPSA response rate: 58% vs 54%Chi et al.[106]
vs docetaxel/prednisoneMedian PFS: 7.3 vs 6.1 months
Chemotherapy naïve CRPC (adjusted HR 0.54; P= 0.04)Median OS: 27.5 vs 16.9 months
EverolimusmTOR inhibitor15Phase I, dose-findingReduced phospho-S6 stainingLerut et al.[108]
Neoadjuvant before prostatectomy in pts with localized disease
IMC-A12IGF-1R mAb30Phase IIPrimary endpoint: TTPNIH
Chemotherapy naïve mCRPCOngoingNCT00520481 [103]
CP-751871 + docetaxel/ prednisone  200Phase IIPrimary endpoint: PSA response (PSAWG criteria)NIH
vs docetaxel/prednisoneNCT00313781 [104]
Chemotherapy naïve CRPC and previously treated with docetaxelOngoing 
DasatinibSrc inhibitor46Phase IIImprovement in PSADT: 80%Yu et al.[109]
Chemotherapy naïve mCRPC>35% decrease from baseline 
NTX: 57% 
Dasatinib + docetaxel 32Phase I/IIPSA response rate: 41%Araujo et al.[105]
   Chemotherapy naïve CRPCObjective response rate: 57% in 21 pts 
   Improvement in bone scans: 32% 
Dasatinib + docetaxel/ prednisone 1380Randomized, phase IIIPrimary endpoint: OSNIH
vs docetaxel/prednisoneOngoingNCT00744497 [127]
AZD0530 28Phase IIPSA response rate: 0%Lara et al.[107]
CRPC (9 pts[32%] had received previous docetaxel)Median PFS: 2 months 

Other novel agents being investigated in prostate cancer include: mTOR inhibitors, including rapamycin, temsirolimus (Wyeth), and everolimus (Novartis); several IGF-1R inhibitors, including monoclonal antibodies and small molecule TKIs; the Src TKI, dasatinib; and several histone deacetylase (HDAC) inhibitors, including vorinostat (Merck), trichostatin A, and panobinostat (Novartis). A key target in prostate cancer is the PI3K/Akt/mTOR pathway. As described above, this pathway is frequently activated in prostate cancer cells via PTEN loss, which is associated with increased tumour stage and grade, increased risk of recurrence, and development of androgen independent growth [12,113]. This pathway is also activated by VEGFR and IGF-1R. Because of its central role in this pathway, mTOR is an attractive therapeutic target, and several mTOR inhibitors are being evaluated [12]. Preclinical studies with everolimus and temsirolimus have reported growth inhibition of prostate cancer cells lines, particularly PC-3 and LNCaP cells, which are PTEN deficient [114,115]. Everolimus was also shown to sensitize PC-3 and DU145 cells to the cytotoxic effects of radiation in vitro, and the radiosensitizing effect was greatest for PC-3 cells [114]. Moreover, in a xenograft model of bone metastases, everolimus inhibited tumour growth in bone, and the combination of everolimus with docetaxel and zoledronic acid showed additive antitumour activity [116]. Given the evidence of crosstalk between the AR and the PI3K/Akt/mTOR pathway, these agents could have great potential in CRPC. However, to date, limited clinical data are available. A phase II study in patients with localized prostate cancer who were treated with everolimus before prostatectomy showed biological activity [108], but the clinical benefits of mTOR inhibitors in prostate cancer have not yet been determined.

Several HDAC inhibitors have also shown promising activity in preclinical studies. For example, in vitro studies have shown that vorinostat has radiosensitizing effects on DU145 cells, decreased AR and PSA expression in LNCaP cells, and had synergistic antiproliferative effects when combined with antiandrogens [117]. Preclinical evidence also suggests that panobinostat may inhibit angiogenesis by reducing expression of hypoxia inducible factor-1α in PC-3 cells, and trichostatin A may modulate activity of the PI3K/Akt pathway by down-regulating Akt phosphorylation. Notably, the capacity of HDAC inhibitors to inhibit AR expression and activity could be an important biological effect in prostate cancer and supports investigation of these agents in CRPC [118].

Given its role in tumorigenesis and tumour progression, the IGF-1 pathway is another important target in prostate cancer, and several different approaches are being pursued such as growth hormone receptor antagonists (e.g. pegvisomant), IGF-1R monoclonal antibodies (e.g. CP-751871 and IMC-A12), small molecule TKIs (e.g. GSK1904529A), and antisense oligonucleotides (e.g. ATL1101) [119]. One approach to inhibit this pathway is to reduce serum levels of IGF-1. The pegylated human growth hormone analogue pegvisomant, which is used to treat acromegaly, blocks growth hormone receptor signalling, thereby normalizing serum IGF-1 levels. Several anti–IGF-1R antibodies are currently in development. Preclinical studies have shown that these antibodies decrease AR signalling and sensitize prostate cancer xenografts to ADT [120]. A phase I study of IMC-A12 (ImClone) in patients with various solid tumours provided early evidence of clinical activity, and one patient with prostate cancer had a decline in PSA levels and disease stabilization [121]. A phase II study of IMC-A12 monotherapy in chemotherapy naïve patients with mCRPC (NCT00520481) is currently ongoing [103]. Another IGF-1R antibody, CP-751871 (Pfizer), is currently being evaluated in a large phase II study (NCT00313781) combined with standard docetaxel/prednisone in patients with either previously untreated or docetaxel-refractory CRPC [104,122]. Other molecules are in earlier stages of development. In preclinical studies, the IGF-1R TKI GSK1904529A (GlaxoSmithKline) was shown to inhibit proliferation of cell lines from various solid tumours and decrease growth of human tumour xenografts in mice [123]. Similarly, studies with the IGF-1R antisense oligonucleotide ATL1101 (Antisense Therapeutics) in PC3 and LNCaP xenograft models have shown decreases in serum PSA levels, tumour growth inhibition, and delay of tumour progression in castrated mice [124].

Inhibition of Src also appears to hold promise for the treatment of prostate cancer given the key role of Src in many different signalling pathways and in bone metastasis. Among several small molecule Src inhibitors currently in clinical development, only dasatinib (Bristol-Myers Squibb) and AZD0530 have been evaluated in prostate cancer [25,125]. Dasatinib has activity against Bcr/Abl and was originally developed for the treatment of chronic myelogenous leukaemia, but it is also a potent inhibitor of Src. Both dasatinib and AZD0530 have been shown to inhibit growth of prostate cancer cell lines in vitro and in animal models and to reduce bone resorption markers in patients with CRPC [125]. In an animal model of C4–2B bone metastases, treatment with dasatinib significantly decreased serum PSA levels and tumour growth in bone [126]. These encouraging data led to clinical studies in patients with CRPC. A phase II study of dasatinib monotherapy reported clinical activity based on PSADT, objective response, and bone resorption markers [109]. Subsequently, a phase I/II study combining dasatinib (50–120 mg/day) with docetaxel (60–75 mg/m2 q3 weeks) in patients with mCRPC showed a PSA response in 13 of 32 patients (41%) and an objective tumour response in 12 of 21 patients (57%) with measurable disease [105]. In addition, one-third of patients had an improvement on bone scans, and two-thirds had stable disease in bone for ≥6 weeks. A randomized phase III trial of docetaxel/prednisone with or without dasatinib (NCT00744497) and a phase II study of dasatinib monotherapy in patients who failed prior chemotherapy (NCT00570700) are currently ongoing [127,128]. In contrast, a phase II study of AZD0530 monotherapy (175 mg/day) in 28 patients with CRPC, including nine patients (32%) previously treated with docetaxel, failed to show any PSA responses [107], despite evidence of effective Src inhibition in tumour tissue in the previous phase I study [129].


Because prostate cancer is a relatively slow-growing tumour, immunotherapy is a logical strategy and has the potential to be less toxic than chemotherapy. Two primary approaches to immunotherapy for prostate cancer have been pursued, namely therapeutic vaccines and treatment with immunomodulatory antibodies directed against cytotoxic T-lymphocyte-associated antigen 4 (CTLA4).

Three distinct vaccine strategies have been investigated (Table 5) [130–135]. The first is the dendritic cell-based vaccine, sipuleucel-T (Dendreon), that uses the patient’s own dendritic cells, which are isolated via leukapheresis and loaded ex vivo with a tumour-associated antigen (prostatic acid phosphatase [PAP]) fused with granulocyte macrophage-colony stimulating factor (GM-CSF). The GM-CSF portion of the antigen serves to enhance cellular uptake of PAP and activates the dendritic cells so they become mature antigen-presenting cells capable of inducing a robust immune response to PAP when reinfused. This vaccine has shown the greatest promise to date, and is poised to become the first USA Food and Drug Administration-approved vaccine for advanced cancer. Three randomized, placebo-controlled, phase III trials of sipuleucel-T have been conducted in patients with asymptomatic mCRPC. The first study (D9901) in 127 patients showed a trend toward improved clinical or radiological TTP and a significant 4.5-month improvement in median OS (25.9 vs 21.4 months with placebo; P= 0.01) [136]. Subsequently, an identical phase III trial (D9902A) was completed, and an integrated analysis of D9901 and D9902A (225 patients) showed that sipuleucel-T resulted in a significant survival benefit compared with placebo (HR 1.50; P= 0.011), and the vaccine was well tolerated [133]. Most of the side-effects were transient flu-like symptoms of mild-to-moderate severity. The third trial (IMPACT), involving 512 patients, is assessing OS as the primary endpoint and is still ongoing, but interim survival data presented at the 2009 meeting of the AUA showed a significant 22% improvement in OS [130,137]. Although these studies have not shown a survival benefit compared with standard chemotherapy, sipuleucel-T appears to provide significant clinical benefit in asymptomatic patients with CRPC, who may not be treated with chemotherapy in the absence of bone pain.

Table 5.  Selected clinical studies of immunotherapeutic strategies being investigated in prostate cancer
DrugClassNStudy design/patientsResultsReference
  1. PR, partial response; DC, dendritic cell.

Sipuleucel-TDC-based225Randomized, placebo-controlled, phase III (D9901 + 9902 A)OS (HR = 1.5; P= 0.011)Higano et al.[133]
Asymptomatic, chemotherapy naïve CRPC  
Sipuleucel-T 512Randomized, placebo-controlled, phase III (IMPACT)Median OS: 25.8 vs 21.7 months (HR = 0.78; P= 0.032)NIH
NCT00065442 [135]
Asymptomatic, chemotherapy naïve CRPC Drake et al.[130]
 Webcast et al.[137]
GVAXCellular626Randomized, phase III (VITAL-1) vs docetaxel/prednisonePrimary endpoint: OSNIH
vaccineStudy closed based on futility analysis for efficacyNCT00089856 [140]
Asymptomatic, chemotherapy naïve mCRPCDrake et al.[130]
GVAX + docetaxel 408Randomized, phase III (VITAL-2) vs docetaxel/prednisonePrimary endpoint: OSNIH
Study closed due to imbalance in deathsNCT00133224 [139]
Symptomatic, chemotherapy naïve mCRPC Drake et al.[130]
ProstVacVF ± GM-CSF or fowlpox–GM-CSF
ProstVacVF + GM-CSF
Recombinant 25Phase IIPSA response rate: 4%Gulley et al.[132]
vacciniaAsymptomatic, chemotherapy naïve mCRPC>50% decline 
virus vaccine16% >30% decline 
 1 pt had PR by RECIST 
122Randomized, phase IINo difference in median PFSGodfrey et al.[138]
  vs placebo (P= 0.56) 
Chemotherapy naïve mCRPCMedian OS: 24.5 vs 16 months 
  (HR = 0.6; P= 0.01) 
IpilimumabCTLA4 12Phase IPSA response rate: 17%Small et al.[134]
Ipilimumab + GM-CSF  24Phase IPSA response rate: 50% amongFong et al.[131]
6 patients treated at highest dose level. 
1 PR in liver metastases 

The second approach is an irradiated cellular vaccine called GVAX (Cell Genesys) that consists of exogenous prostate cancer cells engineered to secrete GM-CSF, and a third strategy involves injection of a recombinant vaccinia virus construct that encodes PSA and other costimulatory molecules. The latter vaccine, called ProstVac-VF (Bavarian Nordic), has also shown activity in a small phase II study [132] and a survival benefit compared with placebo in a randomized phase II trial in 122 men with asymptomatic mCRPC [138]. Based on those results, a phase III trial is planned. By contrast, GVAX has been compared with standard docetaxel/prednisone in asymptomatic patients (VITAL-1) or combined with docetaxel and compared with docetaxel/prednisone in symptomatic patients (VITAL-2) [130,139,140]. Both of these randomized phase III trials were designed with OS as the primary endpoint; unfortunately, both were terminated due to either safety concerns or a futility analysis indicating lack of efficacy [139,140].

Studies with the anti-CTLA4 monoclonal antibody, ipilimumab (Medarex/Bristol-Myers Squibb) in patients with CRPC have also reported promising results [56]. CTLA4 is a costimulatory molecule expressed on T cells that down-regulates T-cell activation. Therefore, blockade of CTLA4 enhances T-cell activation and may overcome tumour-induced immune tolerance to tumour-associated antigens. Two phase I studies have reported clinical activity in CRPC, and further studies are ongoing. In a phase I study of ipilimumab alone (single infusion of 3 mg/kg), two of 12 patients had a PSA response [134]. In a second phase I study combining ipilimumab with GM-CSF, three of six patients treated with 3 mg/kg ipilimumab had a >50% PSA level decline, and one patient had a partial response in liver metastases [56,131]. This combination was shown to enhance expansion of circulating, activated, CD8+ T cells, including T cells specific for tumour-associated antigens. However, treatment with ipilimumab is associated with significant autoimmune side-effects (e.g. hypophysitis and colitis) that may require treatment with steroids. Studies combining ipilimumab with vaccines (e.g. GVAX) are underway.


Development of new agents for the treatment of advanced prostate cancer continues to present many challenges, not least of which is determining appropriate clinical endpoints when measurable disease is often lacking, defining disease progression is controversial, and the disease often has a prolonged natural history [141]. OS remains the benchmark in phase III trials, but is not practical as an early measure of clinical benefit. Assessment of PSA response is widely used to rapidly assess the antitumour activity of novel agents and regimens in phase I and phase II studies, but PSA response may not be the best measure of antitumour activity for all classes of targeted agents and is not a good indicator of long-term clinical benefit or a validated surrogate for OS in advanced prostate cancer [142,143]. Objective response is also useful as an early indicator of antitumour activity; however, less than half of patients with CRPC have measurable target lesions of >2 cm in diameter, and RECIST does not reflect the common clinical manifestations of advanced prostate cancer [144]. In phase II studies, PFS may be a good endpoint to select promising new therapies for phase III trials, but there continues to be a lack of consensus on how best to define PFS in CRPC. Various parameters have been used to define disease progression, including standard radiographical measures of tumour growth (for patients with measurable metastatic disease), PSA level, and pain, but there is no standard definition and reported studies vary widely. Individually, neither radiographical progression nor PSA progression correlate well with OS [145]. However, there is evidence to suggest that if all three of these criteria are applied in a composite definition of disease progression, PFS can be a good surrogate for OS [146]. This highlights the need for consensus criteria for response and disease progression and development of validated composite clinical endpoints in the setting of CRPC. Perhaps it is time for new surrogate endpoints in the era of targeted therapy.

To this end, The Prostate Cancer Clinical Trials Working Group (PCWG2) recently published updated consensus recommendations regarding trial design and clinical endpoints for investigating new therapies for CRPC [147]. These guidelines suggest changing the focus of phase II studies from measuring response to measuring TTP. Moreover, they establish consensus criteria for defining progression of disease (for both trial eligibility and for assessing treatment benefit) based on changes in PSA level, soft-tissue lesions, bone lesions, and symptoms. The PCWG2 definition of PSA progression (a ≥25% increase from nadir and an absolute increase of at least 2 or 5 ng/mL) was independently analysed using a time-varying analysis to determine whether it alone can predict OS in patients with hormone-sensitive or CRPC, and this analysis showed a strong and statistically significant association with OS (P < 0.001) [148]. Based on this analysis, PSA progression may be a suitable endpoint for a phase II study, but a composite definition of disease progression is recommended by the PCWG2 for phase III trials, and patients should remain on treatment until radiographic or symptomatic progression [147].

Selection of patients for enrolment in clinical trials of new agents is another important feature of the study design that can affect the outcome and its interpretation. This is particularly relevant in prostate cancer because of the heterogeneous nature of the disease. Significant progress has been made in the development of accurate prognostic nomograms, and the PSA Working Group has recently published guidelines on PSADT and its utility as a predictive marker of clinical outcome [149]. However, beyond that there is a need to better select patients for treatment with targeted agents based on the molecular phenotype of their tumour. This approach is rapidly becoming a reality in breast, colorectal, and lung cancer but has yet to be applied in prostate cancer. The goal is to identify subgroups of patients who may benefit from a specific therapeutic approach.


The multidisciplinary management of advanced prostate cancer will no doubt continue to develop as chemotherapy and targeted agents are increasingly integrated into the treatment paradigm at all stages of disease to complement surgery, radiation, and hormonal therapies. Although the development of effective chemotherapy regimens for CRPC has led to significant improvements in OS, prognosis remains poor, and better treatment options are needed. Research in this area is focusing on more potent antiandrogens, new hormonal therapies that can completely abrogate testosterone biosynthesis, exploration of existing and novel chemotherapy agents as second-line therapy, and integration of antiangiogenic agents, bone-targeted agents, and agents that target specific signalling pathways within the tumour. As preclinical and clinical research illuminates the many signalling molecules and pathways that contribute to prostate cancer growth and metastasis, androgen independence, and formation of bone metastases, new strategies are being developed to target those pathways and the subtle ways in which they interact. Evaluating these new strategies and determining which ones merit further study in phase III trials will benefit from use of composite endpoints to assess TTP. Future studies will also need to evaluate the best way to combine these agents, thereby manipulating multiple pathways simultaneously, and identify biomarkers for patient selection so treatment can be tailored to the molecular phenotype of the tumour.


Financial support for medical editorial assistance was provided by Pfizer, Inc. We thank Jeffrey S. Riegel, PhD, of Accuverus, Inc. for medical editorial assistance with manuscript preparation.


Karim Fizazi is on the Advisory Board of Astra Zeneca, Sanofi Aventis, Antigen, Novartis and Roche. Cora N. Sternbery receives honoraria and research support from Pfizer, G.A. and Cougar Biotech. John M. Fitzpatrick is on the Advisory Board of Sanofi Aventis. Majid Tabesh is an Employee of Pfizer.