These are, indeed, exciting times for patients with metastatic prostate cancer and the clinicians who care for them. After years of relatively little clinical progress, we have suddenly found ourselves in the enviable situation of multiple treatment approaches demonstrating overall survival benefit by different modalities with nonoverlapping toxicities and mechanisms of action. However, these new opportunities raise some interesting concerns regarding the integration of all these options. In particular, is there one approach, sequence, or combination of treatments that improves outcomes better than another? How do we judge if a treatment is working in an individual patient? Is the underlying biology of castrate resistant prostate cancer (CRPC) similar in all patients or should therapy be tailored to their clinical or genetic phenotype? And finally, are there still unmet needs in CRPC?
Current guidelines (from the National Comprehensive Cancer Network [NCCN]) for treatment of men with metastatic CRPC are increasingly nuanced and are now based on a few key cancer features that to this point in time have been relatively de-emphasized in our clinical trials: the presence of cancer pain and visceral, specifically hepatic, metastases. Neither pain nor visceral disease has figured prominently in the development of these most recent advances, none of which has a specific indication for their management. We believe that in 2012, these guidelines provide the highest evidence-based clinical pathway for therapeutic decision-making in patients with metastatic CRPC. Yet even here, there are still many gaps in our knowledge regarding these recommendations. For example, in patients with hepatic metastases or cancer pain, docetaxel(Drug information on docetaxel) chemotherapy is the recommended first line of treatment, yet its effectiveness in comparison to that of abiraterone acetate (Zytiga) in this population is not known. Conversely, patients with asymptomatic or minimally symptomatic CRPC stand the greatest likelihood to receive all treatment strategies in some type of sequence, including sipuleucel-T (Provenge), docetaxel/prednisone, abiraterone acetate/prednisone, and cabazitaxel (Jevtana)/prednisone, but are all four strategies in sequence better than any two or three? Clearly, more comparative effectiveness research is needed in the field to support our practices.
Oncology is rapidly transitioning into the era of personalized, biomarker-driven medicine, based on individual tumor profiles and predictive sensitivity to targeted agents, exemplified by the predictive value of KRAS mutations for epidermal growth factor inhibitors in colorectal cancer, HER2 amplification for anti-HER2 therapeutics in breast cancer, BCR-ABL inhibitors for Philadelphia chromosome–positive chronic myelogenous leukemia, BRAF mutations for selective Raf kinase inhibitors in metastatic melanoma, and ALK fusions for lung adenocarcinomas. While prostate cancer has lagged behind in predictive medicine, development of biomarkers in the setting of these new agents is imperative to identify the unmet opportunities in subsets of patients. We believe that given the increasing complexity and cost of newly approved therapies, men with CRPC will likely have their future therapy individually tailored based on both predictive biomarkers and prognostic risk groups. It is already well appreciated that many factors are important in determining survival in these patients, ranging from the presence of painful metastases and visceral metastatic sites to elevated PSA, lactate dehydrogenase, and alkaline phosphatase levels.[1,2] These factors can be compiled into nomograms or risk groups that can risk-stratify patients and assist in the selection (or development) of therapies based on toxicity profiles, the likelihood of immediate clinical impact, and timing of subsequent treatment. For example, sipuleucel-T is FDA-approved specifically for metastatic-CRPC patients without significant pain (ie, relatively good-risk patients). Developing therapies specifically in a poor-risk setting (ie, cancer pain and/or hepatic metastases) would potentially address an unmet need. As healthcare practices necessitate greater consciousness of costs by providers, we will need data to support more tailored approaches to the various clinical subsets of metastatic CRPC. One form of data will come from prospective randomized controlled trials, but these are inevitably slow, costly, and specific. Another source of clinical data to inform practice patterns can come from prospective registry cohorts. While not as definitive as a prespecified clinical trial, registries can inform us, in times of changing practice patterns, of trends worthy of further interrogation. In addition to being hypothesis-generating, registries could identify clinical phenotypes of patients that emerge from the current and evolving treatment landscape. Clearly, today’s changing menu of management options would seem to justify an investment into prospective registries.
With the emergence of multiple disparate treatment options it will be equally important to judge not just which patients are likely to benefit from a therapy, but which ones actually have an improved outcome. Surrogate markers of response have been difficult to develop in many fields of cancer management, but perhaps none more so than prostate cancer. Our traditional measures of objective response rate, progression-free survival, and even pain and PSA response have not passed the bar of surrogacy for overall survival.[5,6] Yet for the individual patient, how else can we decide to continue with one therapy or move onto another? Even on a population basis, to rely solely on overall survival as an endpoint may become increasingly onerous, given the expected improved survival with today’s sequential therapies. In the future, we need to identify biomarkers that can help this individual tailoring of systemic therapy and address key biologic questions in the pathophysiology of CRPC progression. For example, can adrenal androgen precursor levels predict for sensitivity to abiraterone acetate and can androgen receptor (AR) splice variants predict sensitivity to the novel antiandrogen MDV3100? Can bone turnover biomarkers select for patients most likely to develop skeletal events and thus benefit from bone microenvironment agents such as zoledronic acid(Drug information on zoledronic acid) (Zometa), denosumab (Xgeva), or radium-223 (Alpharadin)? Are men with PTEN-null or AR-independent CRPC sensitive to PI3 kinase inhibitors? Is there a role for continuation of novel hormonal therapies beyond progression? Does targeting of the epithelial compartment in castrate-resistant disease, such as with hormonal agents, induce an epithelial-mesenchymal transition (EMT) and render highly plastic tumor cells more sensitive to antimesenchymal/stem cell therapies? Is there cross-resistance between novel hormonal therapies and taxanes, given their potentially overlapping mechanism of action on AR? These and many more questions are among the important challenges that the field is facing in order to understand the placement of agents in the therapeutic arsenal, and to maximize patient care. As the number of active agents available to treat metastatic CRPC expands, it is logical to consider the potentially greater benefit of these agents in men with more minimal residual disease after local therapy or following biochemical-only recurrence. None of the therapies discussed in this review have cured men with metastatic disease, yet the clinical benefits seen in this setting support an earlier use. As Drs. Crawford and Flaig point out in their article in this issue of ONCOLOGY, future development in the minimal disease setting of combination approaches with novel androgen blockade, chemotherapy, or immunotherapy are justified.
While we have been met with success over the past 2 years, we need to rechallenge our past assumptions and failures. We must support research into more intelligent screening practices, more selective local therapy application, and more biomarker-driven therapeutics (including surgery and radiation in this approach), and should promote the application of science to our clinical trial designs and patient selection, to optimize care. What is to come in the decade ahead? We think that the future is bright, but we will need to move beyond broad treatment approaches for all and towards more enrichment strategies that result in greater clinical benefit. For now, our menu of options has expanded dramatically. We must choose wisely how we order our treatments in patients. Our guidelines are just a start, but a good one that should not be ignored.
Financial Disclosure: Dr. Armstrong has served as a speaker for Dendreon and sanofi-aventis. He has served as a consultant to Dendreon, Amgen, and Bristol-Myers Squibb (BMS), has served on the advisory board of Amgen, and has received research funding from Dendreon, sanofi-aventis, Medivation, Imclone, BMS, and Active-Biotech. Dr. Moul serves on the speakers bureaus of sanofi-aventis, Dendreon, Janssen, and Ferring.