13. Sequencing Therapies in Metastatic Castration—Resistant Prostate Cancer

The approval of docetaxel in 2004 ended a nearly 60 year drought in which nothing was shown to prolong life for patients with metastatic castration-resistant prostate cancer (mCRPC), and ushered in a new era of treatment for mCRPC [1, 2]. Since then, six therapeutic agents (including docetaxel) have been approved for the treatment of metastatic castration-resistant prostate cancer (mCRPC)—all on the basis of Phase III data indicating a survival advantage with these drugs. These agents include chemotherapeutics (e.g. docetaxel and cabazitaxel), androgen receptor (AR) directed agents (e.g. abiraterone and enzalutamide), immunotherapeutics (i.e. sipuleucel-t) and radiopharmaceuticals (i.e. radium-223) (Fig. 13.1) [1‐9].

Since docetaxel’s approval in 2004, many subsequent approvals in the mCRPC therapeutic space have been predicated on prior docetaxel exposure (Fig. 13.2). For instance, the Phase III cabazitaxel study mandated that patients be post-docetaxel, and as such it remains only approved in docetaxel-treated patients [5]. Similarly, the approvals for abiraterone and enzalutamide were initially granted post-docetaxel, with approval for docetaxel-naïve patients only occurring after Phase III trials in that patient population were completed [3, 4, 8, 9]. While this regulatory framework was likely born out of the desire to select a patient population in which a survival benefit could be quickly demonstrated, it is not reflective of the current treatment landscape in which many options exist for men with mCRPC. Furthermore, the Phase III studies testing these newer agents did not postulate biologic rationales for the specific therapeutic sequence being tested (i.e. pre- or post-docetaxel) or incorporate biomarkers to address the issue of optimal drug sequence, and there remains no historical or biologic basis for why AR targeting agents, such as abiraterone or enzalutamide, or cabazitaxel would only work if given after docetaxel. An acknowledgment of this point is reflected in the Prostate Cancer Working Group 3 (PCWG3) guidelines regarding the conduct of clinical trials in men with mCRPC. The PCWG3 guidelines provide a revised therapeutic framework, emphasizing the sequential use of approved agents, rather than selecting drugs based on a whether or not an individual has received prior docetaxel [10].

While some of these drugs have seemingly distinct mechanisms of action (e.g. radium-223 and sipuleucel-t), there is considerable mechanistic overlap between others (e.g. abiraterone and enzalutamide). As such, it is not surprising that evidence of cross-resistance between many of these therapies is being increasingly recognized. At the heart of this cross-resistance is the fact that many of our approved therapies rely on inhibiting the same target: the lineage-survival oncogene AR [12, 13]. In addition, increasing evidence indicates that prostate cancer evolves over the course of treatment, with more resistant subclones emerging, resulting in an inherently more difficult to treat disease [14‐17].

Questions surrounding the optimal mCRPC treatment paradigm and how to sequence the six available mCRPC drugs remain. In this chapter, we will outline our current understanding of how to best utilize the drugs approved for men with mCRPC, and highlight some of the controversies surrounding when to use each of these agents.

Taxanes (i.e. docetaxel and cabazitaxel) remain the only class of chemotherapeutics that result in improved overall survival compared to active controls for men with mCRPC [1, 2]. In recent years, several effective oral AR-directed therapies have been approved in the pre-docetaxel space, and as such practice patterns are beginning to shift towards delayed use of docetaxel in favor of these less toxic, and more easily administered agents [3, 8, 18].

Docetaxel still remains an important therapeutic option, however, and under certain circumstances may be preferred over the use of drugs like abiraterone or enzalutamide. For instance, docetaxel may be more appropriate if a patient requires rapid palliation or control of visceral metastases [19]. Docetaxel has been reported to achieve a pain response after 27 days and result in improved quality of life after 43 days [20]. To put this in context, the time to pain palliation with abiraterone has been reported to be 5.6 months [21]. While cross-trial comparisons are bias-prone, this difference is quite dramatic. It should also be noted that in both the abiraterone and enzalutamide pre-docetaxel Phase III studies, only patients who were asymptomatic or minimally symptomatic were permitted on study [3, 8]. Further complicating the issue is the recent data indicating that docetaxel results in substantial survival gains when used for hormone-sensitive metastatic prostate cancer, raising the question of whether the early use of docetaxel for men with mCRPC should be liberalized [22, 23].

For now the question regarding whether to sequence docetaxel before or after next-generation AR-directed therapies (i.e. abiraterone or enzalutamide) is not fully answered. Additional clinical trials are needed before any definitive conclusions regarding how to best utilize docetaxel in the context of effective and readily available oral agents remain.

For the majority of men with mCRPC, AR-signaling still constitutes an important driver of disease progression. Both tissue androgens and AR increase in tumor tissue as prostate cancer transitions to the castration-resistant state, and AR-regulated genes such as PSA are often expressed at high levels in men with mCRPC [24]. These observations led to renewed interest in targeting AR-signaling in mCRPC patients and subsequently prompted the development of drugs like abiraterone (an inhibitor of extragonadal androgen biosynthesis) and enzalutamide (a pure AR-antagonist) [3, 4, 8, 9, 25].

Both abiraterone and enzalutamide were initially developed in the post-docetaxel space, with their respective pivotal Phase III trials demonstrating a survival advantage compared to controls [4, 9]. Given that these studies were limited to a population that was exposed to prior docetaxel, initial drug approval was limited to patients that had already received docetaxel. Subsequent approvals in the pre-docetaxel space was only granted following publication of additional Phase III data showing a survival benefit when these drugs were used pre-docetaxel [3, 8, 26]. Consensus guidelines, such as the National Comprehensive Cancer Network (NCCN) Prostate Cancer Guidelines and the American Urological Association (AUA), recommend reserving docetaxel for patients with symptomatic, rapidly progressive or visceral disease [19, 27]. As such, practice patterns are likely shifting toward the earlier (i.e. pre-docetaxel) use of both abiraterone and enzalutamide [18].

In regard to the question of whether to use abiraterone or enzalutamide first, there are no clinical trials directly comparing these two agents to one another. Both drugs function similarly by inhibiting the ligand (e.g. testosterone, dihydrotestosterone [DHT]) AR interaction, and it is probably safe to assume that both agents have comparable efficacy when used as a first line treatment for mCRPC. The decision to choose one agent over the other may ultimately hinge on considerations independent of any anti-tumor effect. Abiraterone requires concurrent treatment with prednisone in order to blunt the mineralocorticoid side effects associated with its use [4, 8]. Therefore, in patients for whom steroids are contraindicated or undesirable (e.g. those with diabetes), enzalutamide may be preferred. It is also notable that in the pre-docetaxel enzalutamide Phase III trial patients with visceral disease were permitted; whereas, in the pre-docetaxel abiraterone Phase III trial these men were excluded [3, 8]. There is, however, no clear evidence that abiraterone is ineffective in treating visceral disease.

Radium-223 is a novel alpha emitting radiopharmaceutical that possesses intrinsic bone homing properties similar to that of other alkaline earth elements, such as calcium, and is approved for the treatment of symptomatic, bone-metastatic CRPC. Radium-223 is not expected to affect soft tissue metastases, and on this basis, patients with visceral metastases or nodal metastases >3 cm in the short axis were excluded from the Phase III trial that led to radium-223’s approval [7]. As such, it is only approved for mCRPC patients with metastatic disease predominantly affecting bone.

In contrast to abiraterone and enzalutamide, radium-223 was not tested in separate Phase III trials specifically designed for men that were pre- and post-docetaxel [7]. At baseline, fifty-seven percent of men enrolled to the radium-223 arm and placebo arm of the Phase III study had received prior docetaxel and the remainder were docetaxel naïve. As with abiraterone or enzalutamide, the indication for radium-223 is therefore agnostic of prior docetaxel treatment status.

In addition to location of metastatic disease, another consideration before initiating radium-223 is the fact that it did not clearly result in improved pain endpoints in its Phase III trial as assessed by validated pain assessment instruments. Given that a Phase II trial did document a palliative benefit, however, an ongoing observational study was designed to assess radium-223 effect on pain (clinicaltrials.gov: NCT02398526) [28]. Until the results of this study are reported, it should not be assumed that radium-223 would lead to a meaningful improvement in pain.

Cabazitaxel is a newer taxane that was shown to result in prolonged survival compared to mitoxantrone when used after docetaxel [5]. As such, it is only approved for use following disease progression on a docetaxel-based regimen. While overall cabazitaxel resulted in a decreased risk of death compared to mitoxantrone, there was a higher risk of death within 30 days of receiving the last dose, likely reflecting the toxicity of this agent. This is in contrast to radium-223, which demonstrated comparable frequency of grade 3 and 4 adverse events compared to placebo. Whether cabazitaxel would prove to be as toxic in a less heavily pre-treated group of patients remains to be seen, and prospective studies assessing its effectiveness pre-docetaxel are currently underway (FIRSTANA trial [clinicaltrials.gov: NCT01308567] and TAXYNERGY trial [clinicaltrials.gov: NCT01718353]). Depending on the results of these trials, cabazitaxel use could potentially be approved for docetaxel-naïve patients, but for now its use in this patient population remains off-label.

Sipuleucel-T is an ex vivo autologous immunotherapy product and remains the only cancer vaccine shown to improve overall survival in Phase III testing [6, 29]. It is designed to produce an immune response toward the prostate antigen PAP. The Phase III study only included mCRPC patients with asymptomatic disease and an anticipated life expectancy of ≥6 months. The median survival of the placebo and sipuleucel-t arms was long at 21.7 months and 25.8 months, respectively, likely reflecting the relatively good prognosis of the patients enrolled to the Phase III study [6]. Similar to the radium-223 Phase III trial, enrollment to the sipuleucel-t Phase III study was not limited based on prior docetaxel exposure, and ultimately 15.5 and 12.3 % of patients randomized to sipuleucel-t and placebo arms, respectively, were post-docetaxel. Interestingly no difference in disease progression was observed between study groups, and while this might call into question the overall survival benefit observed on this study, it should be noted that two additional randomized trials reported a survival advantage with sipuleucel-t [30, 31].

Based on the aforementioned, sipuleucel-t is not approved for patients with symptomatic disease, and given that only men with a life expectancy ≥6 months were included in the pivotal Phase III trial, it is not appropriate for patients with rapidly progressive disease or those expected to live <6 months. While it is technically approved pre- or post-docetaxel, in practice most men that have progressed on docetaxel are likely in need of a therapy that will result in objective tumor control.

Surgical or medical castration (i.e. androgen deprivation therapy; ADT) as a treatment for prostate cancer was the first example of an effective targeted cancer therapy, and to this day inhibiting AR-signaling, primarily through disrupting ligand-AR interactions, remains the mainstay of treating advanced disease [25, 32]. Until recently, progression beyond frontline ADT was felt to represent an “androgen independent” or “hormone refractory” state; however, with the recognition that the AR transcriptional program is still operative in men that progress beyond ADT, the nomenclature has shifted such that these men are now referred to as having castration-resistant prostate cancer. As discussed above, many of the newer agents approved for the treatment of mCRPC function to inhibit AR-signaling, and not surprisingly, their long-term use can lead to the emergence of a drug resistant phenotype—manifested as diminished clinical activity when these drugs are used sequentially.

As the repertoire of approved agents for the treatment of mCRPC increases, so do the number of choices we face regarding which drug to select for any given patient. At this point, all approved therapies (except cabazitaxel) are indicated pre- and post-docetaxel. Recently presented consensus guidelines on the conduct of mCRPC clinical trials encourages investigators to avoid defining trial cohorts on the basis of chemotherapy treatment status—reflecting the new mCRPC therapeutic landscape [10]. To date, it remains unclear if an optimal sequence of mCRPC drugs exists; however, specific presenting clinical features may help steer the choice of when to use each agent.

Docetaxel is probably most appropriate if a rapid palliative response is needed (e.g. patients with rapidly progressive disease or visceral metastases) [2, 20]. On the other hand, abiraterone and enzalutamide both have excellent activity in patients with mCRPC, and in contrast to docetaxel, are generally better tolerated—often making them more attractive choices for first-line mCRPC therapy, particularly for patients with significant comorbidities or impaired performance status [3, 4, 8, 9]. Evidence of cross-resistance between drugs that inhibit AR-signaling is becoming more apparent, and emerging clinical data indicates that adaptive changes in AR expression and/or emergence of constitutively active AR-Vs may drive resistance to chronic exposure to AR-directed drugs. Preliminary data indicates that docetaxel may be a better choice in the face of a resistant phenotype; however, whether there is an optimal sequence with which to use abiraterone, enzalutamide and docetaxel remains to be seen [69].

Sipuleucel-t, radium-223 and cabazitaxel have somewhat more restricted indications. Sipuleucel-t is only approved for asymptomatic patients, and as such is probably only useful in a small subset of mCRPC patients that present with slowly progressing disease [6]. It is important to note that sipuleucel-t will not prevent disease progression, so the appropriateness of delaying the initiation of drugs that can control disease (e.g. abiraterone, enzalutamide, docetaxel) must be considered. Radium-223 only targets bone metastases, and is not approved for patients with visceral disease or metastatic adenopathy >3 cm in size—limiting its usefulness in heavily pre-treated patients that are more likely to have soft tissue metastases [7]. Cabazitaxel is only approved following progression on docetaxel. In addition, it is relatively toxic and affords only a modest improvement in overall survival [5].

With so many agents to choose from, work towards developing predictive biomarkers should be prioritized. Ongoing efforts to characterize the molecular landscape of mCRPC through metastatic biopsy programs will likely play an important role in defining which group of patients stand to benefit from specific therapeutic agents. For instance, in a recent report by Robinson et al. [70] patients with mCRPC underwent a targeted biopsy followed by an integrative genomic assessment. In addition to documenting several previously described genomic aberrations (e.g. AR, ETS genes, TP53 and PTEN), this study also found a higher than expected frequency of biallelic loss of DNA damage repair pathway genes (e.g. BRCA1, BRCA2 and ATM). This finding provides a rational basis for ‘precision oncology’ trials testing agents that either induce DNA damage or impair the DNA damage repair machinery (e.g. platinum chemotherapy, PARP inhibitor) in those with evidence of biallelic loss of DNA damage repair pathway genes.

Ultimately, prospective trials are needed to delineate the role each of the aforementioned drugs will play in treating mCRPC. Specific questions that need to be addressed include: (I) Are combinations of drugs better than their sequential use? (II) Is one sequence of drug use better than another? (III) Should every mCRPC patient receive every approved drug? (IV) Can biomarkers (e.g. AR expression level, AR-V status) lead to improved patient-drug selection? There has been a lot of progress made in the treatment of mCRPC over the past decade. Many questions remain, but with continued work we will learn how to most effectively use the plethora of agents now approved for mCRPC.