Therapeutic management of bone metastasis in prostate cancer: an update

Transcription

1 Expert Review of Anticancer Therapy ISSN: (Print) (Online) Journal homepage: Therapeutic management of bone metastasis in prostate cancer: an update Fable Zustovich & Davide Pastorelli To cite this article: Fable Zustovich & Davide Pastorelli (2016): Therapeutic management of bone metastasis in prostate cancer: an update, Expert Review of Anticancer Therapy, DOI: / To link to this article: Accepted author version posted online: 25 Sep Submit your article to this journal View related articles View Crossmark data Full Terms & Conditions of access and use can be found at Download by: [Cornell University Library] Date: 27 September 2016, At: 08:59

2 Publisher: Taylor & Francis Journal: Expert Review of Anticancer Therapy DOI: / Review Therapeutic management of bone metastasis in prostate cancer: an update Fable Zustovich 1 and Davide Pastorelli 2 1 Oncologia Medica, ULSS 1 Belluno, Italy 2 Oncologia Medica, ULSS 2 Feltre, Italy Corresponding author Fable Zustovich Oncologia Medica ULSS 1, Belluno, Italy fable.zustovich@ulss.belluno.it Phone: Fax:

3 Abstract Introduction: Bone metastases affect the majority of patients with castration-resistant prostate cancer (CRPC), resulting in significant morbidity and mortality. This review describes the current therapies available for the management of CRPC patients with bone metastases. Areas covered: Studies on the use of currently available therapeutic approaches for palliating pain, delaying skeletal-related events (SREs) and prolonging survival in CRPC patients with bone metastases have been examined. PubMed database was searched in May 2016 starting with the following keywords: ("castration-resistant prostate cancer" OR "CRPC") AND "bone metastases", and approximately 270 results were retrieved. More specific searches were then performed on the epidemiology and molecular pathogenesis (in particular, "vicious cycle" was used as a keyword), the management of pain, SREs and survival. The following keywords were also used individually: abiraterone, cabazitaxel, denosumab, docetaxel, enzalutamide, radium-223, sipuleucel-t, samarium- 153, strontium-89, zoledronate. Randomized-controlled trials, observational studies, reviews, systematic reviews and meta-analyses were selected and articles were excluded if not in English. Expert Commentary: Currently, clear recommendations on the optimal use of the agents available to treat mcrpc are lacking. Therefore, to ensure patients the best treatment, both their clinical characteristics and the features of each product have to be considered. Keywords: Abiraterone; bone metastases; cabazitaxel; castration-resistant prostate cancer; denosumab; docetaxel; enzalutamide; radium-223; zoledronate.

4 1. Introduction Prostate cancer (PCa) is the second most common type of cancer in men and accounts for nearly 20% of all newly diagnosed male tumors. In the US, 180,890 new cases have been estimated to be diagnosed in 2016 and 26,120 men to die of this disease [1]. At diagnosis, approximately 80% of patients present with localized PCa and 4% with distant metastases, the 5-year relative survival rate being 100% and 28% respectively [1]. Due to PCa cell growth dependance on androgens, recurrent or metastatic disease is managed with surgical or pharmaceutical castration. Indeed, androgen deprivation therapy (ADT) is the mainstay therapy in metastatic PCa, with response observed in 80-90% of cases. However, nearly all men eventually progress within one to three years leading to castration-resistant PCa (CRPC) [2]. The majority of patients with metastatic CRPC (mcrpc) develop bone metastases [3] resulting in a significant increase in morbidity and mortality [4,5]. Morbidity and impact on quality of life (QoL) are due to the increased risk of bone fractures, bone pain, nervous tissue compressions and hypercalcemia [5,6]. These complications, collectively referred to as skeletal-related events (SREs), are associated with impaired mobility, general suffering, reduced self-sufficiency, poor QoL, increased mortality, and increased health care costs [7 9]. Greater insight into the pathophysiology of bone metastases has led to the development of new bone-targeted agents aimed at reducing the rate of SREs and prolonging survival. While the number of available treatments that yield significant benefit in these patients has increased in recent years [10], initial survival benefit was observed only with docetaxel-based chemotherapy [11,12]. Therefore, it was approved by FDA in Since 2010, five new agents have emerged following phase III studies that have also gained FDA approval. These therapies include the CYP17 lyase inhibitor abiraterone acetate, the antiandrogen enzalutamide, the microtubule stabilizer cabazitaxel,

5 and the radiopharmaceutical radium 223 [13,14]. Although these drugs act through different mechanisms on different targets, they have been shown to be able to further prolong survival in chemo-naïve patients treated previously with docetaxel [15]. The aim of this review is to describe the current therapies available for patients with CRPC and bone metastases. 2. CRPC and mcrpc: epidemiology, natural history and prognosis Two phase III randomized trials investigated the natural history of nonmetastatic CRPC patients included in the respective placebo arms [16,17]. Results showed that 33% [16] and 46% of [17] men had developed bone metastasis and 21% [16] and 20% [17] had died within 2 years from study entry. Also, a systematic review analyzing data from 71,179 patients observed for up to 12 years reported that 10-20% of PCa patients had experienced progression to CRPC within approximately 5 years and, at the time of CRPC diagnosis, 84% already presented with bone metastases and 33% of those without developed them within 2 years [18]. mcrcp patients had a shorter survival than CRPC cases (9-13 months vs months, respectively) and suffered from a rapid deterioration of QoL, with frequently reported pain and SREs such as bone fractures and spinal cord compression [18]. Pain can occur in each stage of PCa, although its incidence increases up to 90% during the terminal phase of disease [19]. SREs occur in 44% to 80% of PCa men with bone metastases [20 22], but their incidence and severity have been also linked to endocrine therapy [23] and ADT [24]. SREs and bone metastases negatively affect overall survival (OS). Indeed, it has been reported that the 1- and 5-year survival rate in PCa patients without bone metastasis is 87% and 56%, respectively, vs. 47% and 3% in those with bone metastasis and 40% and <1% in those with bone metastasis and SREs [20]. The presence of bone metastases increased the risk of death (hazard ratio [HR]=6.6, 95% confidence

6 interval [CI] [21]), which was further worsened by concomitant SREs (HR=10.2, 95% CI ) [21]). Moreover, the site (axial vs appendicular) [25] and number of sites of bone metastases [26], together with the time from initial diagnosis to first bone metastasis [27], pain intensity, SREs and bone turnover markers [26,28,29] (the latter being predictors also of the risk of SREs development [28,30]) have been shown to be predictors of survival in mcrpc patients. Since men with or without bone metastases have distinct therapeutic options, especially if receiving ADT, the correct, early detection of bone metastasis is important to promptly ensure the most adequate treatment [31]. However, metastatic disease can be misdiagnosed as a result of the insufficient detection limit of standard radiologic imaging [32,33]. Hence, new molecular imaging techniques have been developed, such as magnetic resonance imaging (MRI) of the axial skeleton, PET imaging, and whole-body MRI [32 35]. It is worth noting that in 2015 the Prostate Cancer Clinical Trials Working Group 3 (PCWG3) has advised to increase, in clinical trials conducted in mcrpc patients, the frequency of imaging assessment (by bone scans or CT/MRI) from every 12 weeks to every 8-9 weeks for the first 24 weeks and then every 12 weeks [36]. 3. Pathogenesis of bone metastases Bone metastases display areas of both osteoblastic and osteoclastic (i.e. osteolysis) activity [37], which result in bone frailty, weakened structural integrity and SREs. Hence, understanding the pathogenesis of metastatic disease is of the utmost importance for treatment. In this process, the complex interaction between tumor cells and bone microenvironment plays a central role. Under normal conditions, skeletal integrity is maintained through the dynamic interplay of osteoclasts, deriving from hematopoetic stem cells (HSC) and osteoblasts, from stromal stem cells [38]. After adhering to the bone, osteoclasts form a resorption cavity and release chemotactic cytokines that

7 attract osteoblasts. Once in the resorprion cavity, osteoblasts produce bone matrix and express the receptor activator of nuclear factor-kb ligand (RANKL), which is a key mediator of osteoclasteogenesis. Indeed, upon binding to the RANK transmembrane receptor on osteoclast surface, osteoclasts express mediators of differentiation and survival (e.g. Src), bone adherence and bone catabolism (e.g. cathepsin K). The balance between bone formation and resorption is maintained by the decoy receptor osteoprotegerin (OPG), expressed by osteoblasts, that blocks RANKL, preventing osteoclast activation. In the metastatic cascade (Figure 1), tumor cells home to the bone and adhere to the bone matrix via the interaction between chemokine receptor-4 (CXCR4) on their surface and stromal cell-derived factor 1 (SDF-1) secreted by osteoblasts, competing with HSCs for the place in the bone marrow niche (i.e. onco-niche) [39]. Once established within the bone microenvironment, tumor cells may remain dormant or proliferate. In the latter case, they secrete paracrine mediators that stimulate osteoblasts and stromal cells to proliferate, differentiate, and produce growth factors (e.g. parathyroid hormone-related peptide that induces RANKL expression, interleukin-1 [IL-1] and IL-6). Abnormal levels of RANKL promote excessive activity of osteoclasts that degrade the bone matrix and release, among others, fibroblast growth factor, transforming growth factor-β and tumor necrosis factor-α, which promote the growth and survival of tumor cells, thus completing the so-called vicious circle [40]. In this process, the expression levels of RANKL, RANK and OPG correlate with aggressive, metastatic PCa. Besides, recent in vitro and in vivo studies have highlighted the role of different mirnas [41,42], osteogenic transcription factor Runx2 [43], cyclin A1, P450 aromatase [44] and contactin 1 [45] in CRPC progression and metastasization.

8 4. Pain palliation Standard treatment of symptomatic metastatic bone pain relies on external beam radiation therapy (EBRT) [46], localized at the painful area at risk of fracture, and radiopharmaceuticals combined with analgesics (opioids or nonsteroidal anti-inflammatory drugs [NSAIDs]) [47]. Although EBRT is effective in relieving pain and improving QoL in most patients, pain progression often occurs and retreatment may be required [48,49]. Men with multifocal bone disease not eligible for EBRT are managed with the radiopharmaceuticals strontium-89 chloride (Sr-89) and samarium-153 (Sm-153). Sr-89 is a pure β-emitter calcium-mimetic, with a half-life of 50.5 days and a penetration energy of 2.4 mm, while Sm-153 is a chelating agent that emits also γ-particles, with a half-life of 1.9 days and a penetration energy of 0.6 mm [50]. Both agents are incorporated into areas of high bone turnover (i.e. metastatic sites with high osteoblastic activity) where β-particles induce cancer cell death by DNA single-strand breaks. Both radiopharmaceuticals are indicated for pain palliation in mcrpc patients with bone metastases [50]. Symptom improvement has been reported in 55-80% of patients, with response duration of 2-17 weeks [51,52]. However, despite a reduction in pain and analgesis needed, severe adverse effects (AEs, mostly transient myelosuppression) are common [53]. For this reason, and because they did not provide OS and PFS prolongation, nor SREs or QoL improvement, their use in clinical practice is actually limited [54,55]. Finally, despite, in 2005, a qualitative systematic review reported the increased effectiveness of NSAIDs, alone or combined with opioids, for the treatment of cancer pain [47], it was recently withdrawn by the authors because, although correct at the time of publication, it is out of date and new reports are awaited [56].

9 5. Prevention of skeletal complications 5.1 Zoledronic acid Zoledronic acid is a third-generation bisphosphonate and the first agent approved for the management of bone metastases in patients with CRPC [57]. Bisphosphonates show a chemical structure similar to the normal component of bone matrix pyrophosphate, are absorbed and bind to hydroxyapatite crystals causing inhibition of bone reabsorption by osteoclasts [58]. Zoledronic acid is the only bisphosphonate that has demonstrated significant efficacy and long-term clinical benefit by preventing SREs in PCa patients [59,60]. The administration of zoledronic acid (4 mg every 3 weeks) versus placebo in 643 mcrpc patients resulted in a reduction of the rate of patients having at least 1 SRE (33% vs. 44%; P=0.021). It also displayed consistent efficacy across all secondary endpoints (time to the first SRE, skeletal morbidity rate, proportion of patients with individual SRE events, time to disease progression, objective bone lesion response, bone biochemical markers, and QoL) and improvements of pain and analgesia scores [61]. Pain scores and the use of analgesic drugs favored zoledronic acid, but there were no differences in disease progression or OS [22]. Based on the findings from this study, zoledronic acid was subsequently approved by the US and European regulatory authorities for the prevention of SREs in patients with prostate, breast, and lung cancer. In the past few years, several other trials have examined the use of zoledronic acid in CRPC patients with varying degrees of success. The TRAPEZE trial assessed the efficacy of zoledronic acid or Sr-89 on top of docetaxel in 757 patients [62]. Sr-89 combined with docetaxel improved clinical progressionfree survival (PFS) but did not improve OS, SRE-free interval, or total SREs. Furthermore, zoledronic acid did not prolong PFS or OS but did significantly improve median SRE-free interval and reduced total SREs by about 30%.

10 Other trials have also examined the benefit afforded by zoledronic acid in patients with earlier stages of CRPC. The CALGB90202 phase III trial compared the effect of zoledronic acid vs. placebo in castration-sensitive PCa patients with the aim of detecting a reduction in the risk of first SRE [63]. Zoledronic acid was not associated with increased time to first SRE, the median time to first SRE being 31.9 months in the zoledronic acid group and 29.8 months in the placebo group. OS and rates of AEs were similar between the two groups. In the STAMPEDE trial, a randomized controlled multiarm, multistage trial, the addition of zoledronic acid to ADT in hormone-naïve patients did show no significant benefit in time to first SRE, in the entire population or in the subgroup of patients with bone metastases [64]. The authors concluded that zoledronic acid showed no evidence of survival improvement and should therefore not be part of standard of care for this population [64]. Finally, the ZEUS study investigated the efficacy of zoledronic acid for the prevention of bone metastasis in high-risk non-metastatic PCa patients receiving ADT [65]. Zoledronic acid administered every 3 months was found to be ineffective for the prevention of bone metastases in high-risk localized PCa patients at 4 years (proportion of bone metastasis was 14.7% with zoledronic acid vs. 13.2% in the control group). On the basis of evidence from these trials, CRPC is the only setting where zoledronic acid has been shown to have proven efficacy in the management of bone metastases. 5.2 Denosumab RANKL is the main driver of osteoclast formation, function, and survival [66]. Denosumab is a fully human monoclonal antibody directed against RANKL inhibiting osteoclast-mediated bone destruction, decreasing bone reabsorption and increasing bone mass [67]. This agent is administered quickly via subcutaneous injection compared to zoledronic acid, which is administered intravenously over a period of 30 minutes.

11 The HALT prostate cancer trial was undertaken in patients undergoing ADT and demonstrated the ability of denosumab (60 mg every 6 months) to avoid bone tissue density depletion and to reduce the incidence of new spine fractures compared with placebo. This study did not demonstrate any significant benefit in OS with denosumab [67]. Another second phase III trial enrolled 1,904 patients with mcrpc, compared denosumab (120 mg administered subcutaneously every 4 weeks) with zoledronic acid (4 mg intravenously every 3 weeks) [68]. Denosumab improved the median delay in time to first SRE by 3.6 months (20.7 months vs 17.1 months; HR=0.82; P<0.001 for noninferiority, P=0.008 for superiority). The two groups had similar OS and time to disease progression. OS, disease progression, and rates of AEs and serious AEs were similar in the 2 groups. However, an increased incidence of hypocalcemia (13% in the denosumab group vs 6% in the zoledronic acid group; P<0.0001) was observed. Based on results from this trial, denosumab gained FDA approval for the treatment of mcrpc patients with bone metastases. A posthoc analysis of this phase III trial [68] subsequently showed that denosumab reduced the risk of symptomatic skeletal-related events (SSEs) compared to zoledronic acid [69]. Denosumab therapy significantly reduced the risk of developing the first SSE [HR=0.78; 95% CI, ; P=0.005] and first and subsequent SSEs (rate ratio=0.78; 95% CI, ; P=0.004) compared to zoledronic acid. Another phase III trial examined the effect of denosumab (120 mg subcutaneously, every 4 weeks) or placebo in bone-metastasis-free survival in 1,432 men at high risk of developing bone metastasis (defined as elevated PSA of 8.0 ng/ml or those with short PSA doubling time [PSADT] of 10.0 months or both) [70]. Although no difference in OS or PFS was seen between groups, denosumab significantly increased bone-metastases-free survival by 4.2 months (29.5 vs months, HR=0.85; 95% CI, ; P=0.028). More recently, a post-hoc analysis of 3 phase III randomized trials compared the effect of zoledronic acid to denosumab in preventing SREs in patients with bone metastases [71]. This analysis included a

12 total of 5,543 patients, of which 1901 had PCa. This analysis demonstrated that denosumab administered every 4 weeks was superior to zoledronic acid in preventing SREs in all patients with metastatic bone disease, regardless of their baseline characteristics of ECOG performance status, number of bone metastases, presence or absence of visceral metastases, and urinary N-telopeptide (untx) level [71]. Another post-hoc analysis by the same authors of 3 identical phase III trials demonstrated that untx and serum bone-specific alkaline phosphatase (sbsap) levels ( median levels, compared with <median levels), after 3 months of treatment with denosumab or zoledronic acid were associated with reduced OS (HR for death=1.85, P< and HR=2.44, P<0.0001, respectively), increased risk of disease progression (HR=1.31, P< and HR=1.71, P<0.0001, respectively) and disease progression in bone (HR=1.11, P=0.041 and HR=1.27, P<0.0001, respectively) [72]. These recent findings suggest that untx and sbsap may be considered as noninvasive, early predictors for response and survival in patients with advanced cancer and bone metastases receiving bone antiresorptive agents. 6. Survival benefit 6.1 Docetaxel Docetaxel is a taxane (i.e. mitotic inhibitor) approved by the FDA in 2004 for the treatment of mcrpc in combination with prednisone. Indeed, the TAX 327 randomized phase III trial compared the efficacy and safety of docetaxel (every 3 weeks or weekly) plus prednisone (n=335 and 334, respectively) to mitoxantrone plus prednisone (n=337) [11]. Docetaxel every 3 weeks yielded longer median OS compared to weekly docetaxel and control treatment (18.9, 17.4 and 16.5 months, respectively), and these results were subsequently confirmed in an updated analysis of median OS

13 [73]. The HR for death was 0.76 (95% CI ; P=0.009) upon docetaxel every 3 weeks and 0.91 (95% CI ; P=0.36) upon weekly docetaxel [11]. A significantly better response in terms of pain, serum PSA level and QoL was observed in docetaxel-treated men compared to those given mitoxantrone plus prednisone. However, they also experienced more AEs. In particular, despite the rate of grade 3-4 anemia and thrombocytopenia was low (range: 0-5%) and similar among treatment groups, the frequency of grade 3-4 neutropenia was significantly higher among patients treated with docetaxel every 3 weeks compared to those receiving mitoxantrone (32% vs. 22%, respectively; P<0.05). On the contrary, the latter group experienced significantly more frequently grade 3-4 neutropenia than patients receiving weekly docetaxel (22% vs. 2%, P<0.0015) [11]. In the SWOG 9916 randomized phase III trial, survival prolongation was observed in mcrpc patients treated with docetaxel (every 3 weeks) plus estramustine (n=338) versus mitoxantrone plus prednisone (n=336) (17.5 vs months, P=0.02; death HR=0.80; 95 % CI ) [12]. In the docetaxel arm, time to progression was significantly longer than in the control arm (6.3 vs 3.2 months, P<0.001) but pain relief was similar and bone marrow toxicity more severe [12]. Finally, the impact of early docetaxel treatment in association with hormone-therapy was recently examined in three randomized phase III trials [64,74,75]. The GETUG-AFU 15 study randomized 385 men with metastatic, hormone-sensitive PCa (mhspc) to receive ADT plus docetaxel (n=192) or ADT alone (n=193). Median OS was not significantly improved upon docetaxel (58.9 vs 54.2 months, HR=1.01; 95% CI ), not even with long-term follow-up (60.9 vs 46.5 months; HR=0.9; 95% CI ; P=0.44). However, results from the CHAARTED trial, that randomized 790 mhspc men to receive either ADT plus docetaxel or ADT alone, showed that, in the combination-treated group, the median OS was 13.6 months longer than in the ADT group (57.6 months vs months; HR for death in the combination group =0.61; 95% CI ; P<0.001) [75]. Moreover, the STAMPEDE trial randomized 2,962 men with either high-risk localized (24%), node-positive (15%), or mhspc

14 (61%) to receive ADT alone, ADT plus docetaxel, ADT plus zoledronic acid or ADT plus zoledronic acid and docetaxel. Unlike zoledronic acid, the addition of docetaxel significantly prolonged OS compared to ADT only (81 vs 71 months, HR=0.78; 95% CI ; P=0.006) [64]. Therefore, while the benefit provived by docetaxel in terms of OS is well documented, no clear evidence exists on its effect on pain and the delay or prevention of SREs in bony mcrpc. In this regard, a recent phase III trial demonstrated that docetaxel in combination with zoledronic acid improved median SRE-free interval and reduced total SREs by approximately 30% [62], but further results are warranted. 6.2 Cabazitaxel Cabazitaxel is a semi-synthetic taxane that received FDA approval in 2010 to be used in combination with prednisone for the treatment of mcrpc men previously on docetaxel. Approval was based on results from an open-label phase III study (TROPIC) conducted in mcrpc patients (25% with visceral and 84% with bone metastases) relapsed after first-line docetaxel and randomized to cabazitaxel plus prednisone (n=378) or mitoxantrone plus prednisone (n=377) [76]. Cabazitaxel yielded a significantly prolonged median survival compared to control treatment (15.1 vs 12.7 months, P<0.0001), with a death HR=0.70 (95% CI , P<0.0001) that was similar to that of patients suffering from pain at baseline (45%) (HR=0.76; 95% CI ). Median PFS was also significantly improved in the experimental arm, being 2.8 vs 1.4 months (HR=0.74; 95% CI ; P<0.0001). No improvement in pain relief was observed by either treatments. The most common clinically significant grade 3 AEs were neutropenia and diarrhea; 7 deaths (2% of the total) were recorded for septic neutropenia in the cabazitaxel arm [76]. A subsequent update of this trial demonstrated that, compared to mitoxantrone, this taxane could prolong OS at 2 years (odds ratio=2.11; 95% CI ) but provided similar palliation benefits [77]. In line with these results, a recent review of randomized

15 trials evaluating mcrpc patient-reported outcomes since 2010 found no meaningful impact of cabazitaxel on pain and QoL [78]. Analysis of the QoL and safety data from 112 mcrpc patients treated with cabazitaxel and progressed during or after docetaxel in the UK Early Access Programme showed a trend towards improved QoL; major toxicities were neutropenic sepsis (6.3%) and diarrhea (4.5%) [79]. Despite the fact that cabazitaxel is able to prolong survival in the setting of mcrpc patients with bone metastases, to date no clinical benefit has been reported in term of pain palliation [76,78] and no clear evidence exists on its effect on the delay or prevention of SREs. 6.3 Abiraterone Abiraterone acetate is a potent selective and irreversible inhibitor of the enzyme 17ahydroxylase/C17.20-lyase (CYP17A1), therefore inhibiting androgen synthesis in adrenal glands, testicles, and prostate cancer, the only organs or tissues where this enzyme is expressed [80,81]. Abiraterone is administered orally in the fasted state and requires the concomitant administration of prednisone to avoid toxicity from excess mineralocorticoids associated with CYP17A1 blockade. The COU-AA-301 trial compared abiraterone acetate (1,000 mg per day) plus prednisone (5 mg twice daily) vs. placebo plus prednisone (5 mg twice daily) in 1,195 patients with mcrpc who had relapsed after docetaxel-based chemotherapy [82]. The primary endpoint, OS, was superior for the experimental arm (14.8 vs months) with a 35% risk reduction of death compared with the control arm. Similarly, all secondary endpoints (PSA response, radiologic PFS, and time to progression) were statistically superior in favor of the experimental arm. Post-hoc analyses of COU- AA-301 showed that the incidence of SREs was 29% with abiraterone acetate and 33% with placebo; time to first SRE was significantly longer with abiraterone acetate (median 25.0 vs months; HR=0.615; P=0.0001) [83]. In patients with clinically significant pain at baseline, abiraterone produced

16 significantly superior palliation (45.0% vs. 28.8%; P=0.0005) and faster palliation of pain intensity (median time to palliation 5.6 vs.13.7 months; P=0.0018) [83]. In addition, the evaluation of the subgroups of patients who were asymptomatic at baseline revealed that in the experimental arm the use of analgesics was reduced [83]. A second phase III trial (COU-AA-302) compared abiraterone (1,000 mg/d) plus prednisone (5 mg twice daily) vs. prednisone (5 mg twice daily) plus placebo in 1,088 chemotherapy naïve men with asymptomatic or minimally symptomatic CRPC [84]. Radiographic PFS was significantly improved in the experimental arm (16.5 vs 8.2 months; HR=0.52; 95% CI ; P<0.0001). OS also favoured abiraterone vs. prednisone (median: 35.3 vs months; HR=0.79; P=0.015), but did not reach the prespecified statistical efficacy boundary (αlevel: ) [85]. Finally, a post-hoc analysis of the COU-AA-302 trial examined the safety and efficacy of abiraterone acetate with concomitant bone targeted therapies (BTT) [86]. Superiority of abiraterone acetate was confirmed both with and without BTT. Compared with no BTT use, concomitant BTT significantly improved OS (HR=0.75; P=0.01) and increased the time to ECOG deterioration (HR=0.75; P<0.001) and time to opiate use for cancer-related pain (HR=0.80; P=0.036). 6.4 Enzalutamide Enzalutamide is a potent androgen receptor antagonist and, like abiraterone, it is approved for the treatment of both patients with mcrpc progressing after chemotherapy and chemotherapy-naïve patients. It blocks the nuclear translocation of androgen receptors and inhibits binding to chromosomal DNA, leading to cell apoptosis. In contrast to abiraterone, the oral administration of enzalutamide does not require the concomitant administration of steroids and a fasting state [87]. The AFFIRM phase III trial evaluated enzalutamide (160 mg per day) vs. placebo in 1,199 patients with mcrpc previously treated with a docetaxel regimen [88]. The investigation achieved the primary

17 endpoint set of superiority of OS over placebo (18.4 vs 13.6 months). The superiority of enzalutamide was also observed in the secondary endpoints consisting of time to PSA progression, radiologic PFS, and time to first SRE. PSA progression was delayed by 16.7 months in patients treated with enzalutamide compared with 13.3 months for placebo-treated patients (HR=0.69; P<0.001). Rates of fatigue, diarrhea, and hot flashes were higher in the enzalutamide group. Seizures were reported in five patients (0.6%) receiving enzalutamide. Based on data from the AFFIRM study, this agent was approved by US and European regulatory agencies in the second half of Recently, a post-hoc subanalysis of all randomized patients (n=1199) from the AFFIRM trial was performed. Enzalutamide consistently improved OS, rpfs, and TTPP compared with placebo across PSA subgroups, regardless of baseline PSA level [89]. HRs for improvements in OS ranged from 0.55 to 0.73 across the 4 PSA groups. This post hoc analysis of the AFFIRM trial demonstrates consistent benefits in OS, rpfs, and TTPP with enzalutamide regardless of baseline disease severity, as assessed by PSA. PREVAIL, a phase III trial, was conducted in more than 1,700 patients with asymptomatic or minimally symptomatic progressive metastatic disease who failed ADT but had not yet been treated with chemotherapy. This study has recently been stopped because the interim analysis showed that patients treated with enzalutamide vs. placebo experienced a 29% reduction in risk of death (HR= 0.70; 95% CI ; P=0.0001) and an 81% reduction in risk of radiographic progression (HR=0.186; 95% CI ; P<0.0001) [90]. Although the median time to first SRE was similar in the two arms, the risk of first SRE was significantly decreased with enzalutamide (HR=0.72; 95% CI ; P<0.001) [91]. Median time to progression in Brief Pain Inventory Short Form pain was 5.7 months with enzalutamide vs. 5.6 months with placebo (HR=0.62, 95% CI ; P < ). Furthermore, at week 13, progression of pain was less frequent with enzalutamide compared to placebo (29% vs. 42%, P<0.0001) [91].

18 6.5 Radium-223 Radium-223 (as radium-223 dichloride) is a radiopharmaceutical emitting predominantly α-particles (95.3% of energy, whereas 3.6% is emitted as β-particles and 1.1% as γ-radiation), with an half-life of 11.4 days. It is a calcium-mimetic, bone-seeking drug that selectively targets areas of high bone turnover, where it forms complexes with hydroxyapatite. α-particles have a short range (i.e. 100 μm, <10 cell diameters), that lessens the damage to the surrounding cells, and highly energetic radiation able to induce DNA double-strand breaks in adjacent tumour cells, killing them. As it demonstrated to be effective in both docetaxel-naïve and post-docetaxel patients, radium-223 can be used as monotherapy both in first-line and subsequent lines [92,93]. The pivotal ALSYMPCA trial enrolled 921 patients with mcrpc and metastases in the bone only, who had received, were not eligible for or refused prior docetaxel [92]. Patients were randomized 2:1 to receive 6 injections of radium-223 at 50 kbq/kg, or placebo, with a 4 week-interval between administrations. A composite bone endpoint was introduced, SSEs, defined as a symptomatic fracture, radiation or surgery to the bone, or spinal cord compression, that was advised by PCWG3 to be reported in future trials, as they include only symptomatic events of clear clinical signficance rather than SREs that also include asymptomatic fractures [36]. The study was terminated for efficacy at the prespecified interim analysis (when 314 deaths had occurred; n=809). A significant prolongation of median OS was observed in the radium- 223 arm compared to placebo (14.0 vs 11.2 months, respectively; HR=0.70; 95% CI ; P=0.002). Similar results were achieved in the updated analysis (when 528 deaths had accurred; n=921) and in patients subgrouped based on the use of bisphosphonates, previous docetaxel treatment and performance status (PS) (for ECOG PS 2 statistical significance was not reached). Patients treated with radium-223 experienced a significant clinical benefit also for secondary endpoints, such as the increase in total alkaline phosphatase and PSA level. In particular, a post-hoc analysis unveiled a similar frequency of SSEs in the two arms, being 33% in the radium-223 group and

19 38% in the placebo group, with a delayed time to the first SSE upon radium-223 (15.6 months vs 9.8 months) and a significantly lower risk of SSE (HR=0.66, 95% CI , P= ) independently of baseline characteristics [94]. Also the risk of EBRT and spinal cord compression were reduced in men treated with radium-223, whereas no decrease in the risk of symptomatic pathological bone fractures or the need for tumor-related orthopedic surgical interventions was reported [94]. Compared to placebo, radium-223 improved QoL in a significantly higher percentage of patients and at a significantly higher extent [92,95]. Finally, the frequency of AEs (all, grade 3-4, serious and rate of study drug-discontinuation due to AEs) was lower upon radium-223 than placebo, with a low rate of grade 3 AEs such as myelosuppression and gastrointestinal events [92]. Indeed, it has been shown that the fırst cell layer that can stop radionuclide penetration in the bone marrow, thus preventing hematologic toxicity [96]. Grade 3-4 thrombocytopenia was more common in the subgroup of radium-223-treated men that had received previous docetaxel (9% vs 3%). The risk of grade 2-4 anemia was increased in patients with a greater extent of disease ( 6 metastases or SuperScan) [97]. To date, radium-223 is the only bone-seeking radiopharmaceutical able to prolong OS, prevent SSEs and improve QoL in the setting of mcrpc men with bone metastatic disease. Tables 1 and 2 summarize current ongoing studies exploring the efficacy and safety of radium-223, alone or in combination with other agents, in this complex setting. In this regard, a recent exploratory analysis from the ALSYMPCA trial assessed the efficacy and safety of chemotherapy administered after radium-223 or placebo (n=142 and 64, respectively) [98]. Results showed that administration of chemotherapy after radium-223 was feasible and tolerable for mcrpc patients with bone metastases, even after previous docetaxel. Indeed, docetaxel was given to 70% and 72% of men in the two arms, and mitoxantrone in 16% and 20%, respectively, and had been given before ALSYMCA study beginning to 61% and 58%, respectively. Patients previously treated with radium-223 started chemotherapy 9.1 months after randomization vs 7.5 months in the placebo group, but the duration

20 of treatment and median OS were similar (4.6 and 4.2 months, and 16 and 15.8 months, respectively). Chemotherapy was well tolerated, the hematologic toxicities of grade 3-4 being reported in <10% of patients in both groups. Preliminary results from a single-institution retrospective analysis [99] of toxicity data obtained from patients administered radium-223 alone (n=11) or in combination with next-generation ADT (n=14) indicated a similar profile and modest hematologic toxicities, in the range expected for this group of heavily preatreated patients [92]. Finally, in an early access program setting, patients concomitantly treated with radium-223 and denosumab or abiraterone displayed prolonged survival, along with those presenting with a good ECOG-PS (0-1 vs 2), no pain (vs mild-moderate vs severe) and ALP levels <220 U/L (vs 220 U/L) at baseline [100]. Overall, radium-223 was well tolerated [100,101] and a clinical benefit was reported in terms of percentage of patients experiencing SSEs and ALP and PSA decrease from baseline [100,101]. 6.6 Sipuleucel-T The T-cell activator Sipuleucel-T is is produced by the ex vivo exposure of autologous, patient-specific dendritic cell precursors to a recombinant protein composed of PSA fused to granulocytemacrophage colony-stimulating factor [102]. Sipuleucel-T has been approved in 2010 for the treatment of asymptomatic or minimally symptomatic mcrpc based on the results from a phase III study on mcrpc patients randomized to receive either the vaccine (n=341) or placebo (171 patients) intravenously every 2 weeks (three infusions total) [103]. Sipuleucel-T yielded a survival benefit compared to placebo (25.8 vs 21.7 months, respectively), with a HR for death of 0.78 (95% CI 0.61 to 0.98; P=0.03). However, the time to objective disease progression and bone pain results were comparable between arms. AEs, generally moderate and transitory, were more frequent upon

21 sipuleucel-t and included chills, pyrexia and headache [103]. Similar results were reported in a metaanalysis including the above mentioned study plus two more randomized trials, for a total of 737 participants [104]. Due to the lack of efficacy on disease progression and the cost, the use of sipuleucel-t is limited [105]. 7. Conclusion A better understanding of the mechanisms leading to bone metastases in PCa has allowed for the design and development of novel bone-targeted therapeutic strategies. Since 2004, the approval by FDA of docetaxel, cabazitaxel, abiraterone acetate, enzalutamide, and radium-223 has broaden the spectrum of treatment options for mcrpc patients and changed the goal of therapy from palliation only to also prevention or delay of SREs and prolongation of survival [106]. Although both zoledronic acid and denosumab have shown to be effective in reducing the burden of bone metastatic disease, this benefit has not translated into survival improvement. In contrast, radium-223 represents a novel treatment option for these complex patients. This α-emitter has not only demonstrated efficacy in preventing SSEs and improving QoL, but is also the first bone-seeking agent associated with a significant OS improvement, regardless of previous docetaxel-based therapy. 8. Expert commentary To date, based on the clinical picture of the patient with mcrpc and bone metastases, physicians may choose among a number of therapeutic options to palliate pain, delay skeletal events and prolong survival. However, no clear recommendations are currently available on the best strategy to adopt for delaying the time to SRE and extending survival due to the lack of trials assessing head-to-

22 head comparisons and drug sequencing. Therefore, the optimal approach takes into account the clinical characteristics of patients and the features of each agent available. With regard to prolonging survival, the first-line treatment with abiraterone or enzalutamide proved to be effective in asymptomatic or pauci-symptomatic patients with ECOG-PS 0-1. In contrast, a docetaxel-based chemotherapy regimen appears more appropriate in patients with advanced or rapidly progressing disease, as demonstrated by results from the TAX 327 study, in which about 90% of symptomatic patients presented with bone metastases, 22% with visceral metastases, and 13% with Karnovsky PS <70 [11]. Considering the inclusion criteria of phase III trials, abiraterone could be indicated for patients without visceral disease; if viscerale metastases are documented, docetaxel or enzalutamide could represent a first-line option. Another criterion to differentiate the 2 hormonal agents could be the absence of food interaction for enzalutamide and the lack of concomitant prednisone administration. Moreover, baseline conditions such as seizures and cardiovascular alterations could be considered critical for choosing enzalutamide and abiraterone, since in phase III trials these events resulted in comorbidities. Conversely, radium-223 induced a statistically significant clinical benefit with a seemingly low toxicity profile in mcrpc patients. Due to the demonstrated ability to increase survival, this radioisotope should not be classified like the other bone-seeking therapies used to treat pain or delay SRE in symptomatic patients. In addition, because the pivotal study showed that efficacy was observed in both groups of docetaxel-naïve and post-docetaxel patients, the drug should be evaluated to be used as monotherapy in front-line at first appearance of symptomatic bone metastases, in the absence of visceral disease, or after previous treatments (second-generation hormonal therapy or chemotherapy). Studies conducted in the clinical practice setting confirmed the efficacy and safety observed in the ALSYMPCA trial and documented that an early use is associated with improved efficacy and lower toxicity [93,100]. Also, data from pivotal trial reported the possibility of the concomitant use of radium-223 with radiotherapy and

23 bisphosphonates; international expanded access program data confirmed the efficacy and safety of the association with denosumab [100]. The lack of interaction with other drugs and organs allows radium-223 to be used independent of patient age and presence of comorbidities. Moreover, considering its mechanism of action, which is independent of the androgen receptor, it may be hypothesized to employ this radiopharmaceutical even in patients with primary or acquired resistance to second-generation hormone therapies. Cabazitaxel seems indicated for symptomatic patients or patients with liver metastases or rapidly progressing disease, after docetaxel treatment. Importantly, as cabazitaxel is able to overcome the blood-brain barrier, it may be considered for patients with brain metastases. However, analyses of grade 3-4 toxicities observed in phase III studies indicate that cabazitaxel is not the most appropriate drug for patients with comorbidities associated with blood cell alterations (anemia, neutropenia). Moreover, no data are available on the effect of this taxane in patients with impaired liver function. Denosumab and zoledronic acid have shown to be effective in preventing skeletal complications but not in prolonging survival. However, they might be used in combination with other drugs able to extend OS for higher clinical benefit. Well-known AEs related to zoledronic acid include nephrotoxicity, hypocalcemia and osteonecrosis of the jaw (ONJ) [107]. Because the risk of ONJ increases with trauma and infections, it is recommended an assessment of the dental status of patients before administration of bisphosphonate therapy [107]. A major advantage of denosumab is its subcutaneous method of injection that can be administered quickly compared to zoledronic acid, which has to be given as an IV infusion for 30 minutes and can be challenging to give in an already busy clinical setting. Moreover, denosumab achieves its maximal concentration in blood 5 21 days after the injection contrary to zoledronic acid that shows a rapid decrease in serum. One of the most concerning side effects that include ONJ is similar for both drugs, although hypocalcemia incidence is

24 more frequent, especially during the early months of first administration. Yet, denosumab-related hypocalcemia may be avoided with adequate calcium and vitamin D supplementations [108]. 9. Five-year view To date several unmet needs remain in the management of mcrpc patients with bone metastatic disease. For example, the early and accurate detection of bone metastases, as well as the optimal time of assessment and the criteria to classify both metastases and treatment response need to be further improved. Moreover, the optimal timing and sequencing or combination of agents remains to be defined. Indeed, the rapid and recent developments of novel effective agents has made treatment decision more difficult due to the paucity of studies assessing these objectives and of direct head-tohead comparison trials. According to the National Comprehensive Cancer Network guidelines, docetaxel rechallenge should be avoided and, in patients relapsing after first-line abiraterone and enzalutamide, second-line therapy with the same drugs should be avoided due to possible crossresistance [109]. Radium-223 is indicated for patients with bone metastases either in first- or secondline. Table 1-3 report the list of ongoing observational and interventional (phase I-III) studies including radium-223 and the other agents addressed in the previous pages. Despite the approval of many agents, the OS of mcrpc men with bone metastases has been prolonged of few months only, and therefore new research efforts have to focus on a deeper elucidation of the biology of the disease and also on the mechanisms of resistance. In this regard, several mechanisms of resistance to AR-directed therapies have recently been clarified, including AR and CYP17 lyase upregulation and AR splice variants such as variant 7 (AR-V7). This is a ligandindependent constitutionally-active form of the AR associated with poor outcomes to abiraterone and enzalutamide [110]. The multifunctional androgen synthesis inhibitor galeterone is being investigated in clinical trials for its capability of possibly targeting both full-length and aberrant AR,

25 including AR-V7 [111]. Recently, the phase I and 2 ARMOR studies assessed the efficacy and safety of galeterone in men with treatment-naive nonmetastatic or metastatic CRPC and established a dose for subsequent studies [112]. Currently, the phase 3 study ARMOR3-SV is ongoing, which is an open label randomized controlled trial of galeterone compared to enzalutamide in men with mcrpc espressing the AR-V7 mrna (AR-V7) (NCT ). Another investigational drug currently being tested in mcrpc patients with bone disease is ARN-509 [113]. This is a potent androgen receptor antagonist that targets the AR ligand-binding domain and prevents AR nuclear translocation, DNA binding and transcription of AR gene targets [113]. At present, phase 1 (ClinicalTrials.gov Identifier: NCT , NCT and NCT ) and phase 3 (ClinicalTrials.gov Identifier: NCT and NCT ) trials are ongoing to assess its use, alone or in combination, in mcrpc men chemotherapy-naive or after abiraterone acetate. Indeed, to better tailor treatment, it is crucial to identify the candidates that may benefit the most from a specific sequence of drugs, based on patient comorbidities, age, disease history, previous treatment and prognosis. As resistance is emerging as a further burden in the management of these patients, the identification of prognostic predictors of resistance is of paramount importance, as well as markers of early detection and prognosis. In this regard, it is worth mentioning that sustained AR inhibition has been shown to trigger the activation of alternative, compensatory oncogenic pathways (e.g. of Akt and enhancer of zeste homolog 2), that provide a survival advantage to cancer cells and may induce resistance [114] representing, at the same time, possible novel therapeutic targets.

26 10. Key issues Bone metastases remain a life-threatening complication of mcrpc Since 2004, novel therapeutic strategies have been developed to palliate pain, prevent or delay skeletal complications and prolong survival So far, Radium-223 is the only agent able to provide a significant clinical benefit in terms of skeletal events, survival and QoL in mcrpc patient with bone metastases, regardless of previous therapy and with a good safety profile Several unmet needs remain in the setting of mcrpc patients with bone disease, that diserve further research efforts Due to the lack of clear recommendations, therapy still needs to be tailored on a case-by-case basis. Acknowledgements The authors would like to thank Dr Colin Gerard Egan and Dr Clara Ricci (Primula Multimedia SRL, Pisa, Italy) who provided skillful editorial assistance. Funding Bayor Healthcare Italy provided financial support for editorial services. Declaration of Interest The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

27 Figure legend Figure 1. The "vicious cycle" at the base of the interaction between prostate cancer (PCa) cells, osteoblasts and osteoclasts, in the bone microenvironment, that leads to bone metastases. After escaping from the primary tumor and invading circulation, PCa cells home to the bone driven by chemotactic factors released by osteoblasts (e.g. SDF-1 that binds to CXCR4). Here, through paracrine mediators, PCa cells trigger osteoblast growth which, in turn, stimulate osteoclast activity via the RANKL/RANK/OPG pathway. Following bone matrix degradation by osteoclasts, factors promoting the growth and survival of tumor cells are released, and the circle closes. Abbreviations: SDF-1, stromal-derived factor-1; CXCR4, chemokine receptor 4; RANKL, Receptor activator of nuclear factor kappa-b ligand; OPG, oteoprotegerin; PTHrP, parathyroid hormone-related protein; IL-1, interleukin-1; IL-6, interleukin-6; SREs, skeletal-related events; FGF, fibroblast growth factor, FGF; TGF-β, transforming growth factor-β; TNF-α, tumor necrosis factor-α.