Mary-Ellen Taplin: Thank you, and I'd like to thank the organizers for inviting me. My first time at this meeting, and it's really an effective way to present and digest and debate information, so it's a pleasure.

My disclosures, so this was my assignment, which I thought was rather ambitious for my 10 to 12 minutes, so bear with me, and I hope to avoid the cow bell because my last slide is the best one. Okay, so here's my caveat. There really is no prospective clinical trial data to guide optimal selection of treatment for first-line mCRPC that's developed after treatment with ADT with either docetaxel or AR-axis inhibition, so I think Dr. de Bono's talk on the first day really went into this in-depth, but when you think about treatment for a disease state, you really need to think about what's driving the disease, and, in mCRPC, we know there's adaption and a selection of resistant clones. Some of those clones are AR-dependent, AR-independent, AR null. What we do know for sure is it's very heterogeneous, and we need a way to understand that biology in individual patients. That really doesn't exist today.

We do have some tools that we use. These include biopsying and genomic sequencing, looking for AR-V7, AR amplification, looking for DDR alterations, looking for MSH high, but, frequently, when we put on our clinical hat, we don't have those tools really at our fingertips when we're sitting with that patient in the clinic, and that's what limits applying these molecular changes to our treatment.

So I'm going to go back to our old standard and letting clinical features guide our therapy selection, and some of the things we've talked about may define more aggressive disease, and that includes the development of metastatic disease in the de novo setting rather than the patients who were treated locally and then, subsequently, developed metastatic disease. We know the location and the amount of disease in the mCRPC setting predicts for more aggressive disease such as tumor burden and visceral disease, particularly liver, and then we know the time to progression to mCRPC if it develops shortly on docetaxel or abi, enza or apa, that predicts for more aggressive disease as the short PSA doubling times. So, talking to you this morning, we're going to think about the disease in terms of aggressive variant or aggressive, clinically aggressive versus not. 

So, if you have a patient sitting in front of you that you think you know has done poorly with their therapy up to this point, some of the things that I think you should consider is an early biopsy looking for neuroendocrine features and then consider these treatment options, and they're not prioritized because of the absence of data as I mentioned, but I would consider chemotherapy early with either cabazitaxel if they've already had docetaxel or consider a regimen of docetaxel and carboplatin, or if you know there's a DDR alteration either in the germline or somatic, you can consider carboplatin or a PARP inhibitor.

We don't know that AR-axis therapy won't work post-chemotherapy in the poor-risk setting, but I have some concern that any treatment responses will be short and limited so the patient should be monitored carefully. You also have radium and lutecium to consider, immunotherapy if it's MSH high, and then always consider clinical trials. So I'm not sure how helpful this is, it's like an NCC menu list of all different things, but, unfortunately, given the state of the field at this time, this is where we are in and this is some treatments that we can debate during the discussion coming up.

This is a little chart I made for you to think about these patients if they've had doce first or if they've had enza, abi or apa first, and I think, in the patients who've had the hormone therapy first, you can certainly consider chemotherapy, and if they've had a favorable response, you could also consider the other hormone therapy, sipuleucel-T and so forth.

So, now, I'm going to switch to a couple of the other topics that I was assigned. One is flare, and here are some incidences that are reported in the literature. It's fairly common after chemotherapy and up to a third of patients, less common in hormone therapy, but certainly observed, and the take-home message is to be aware that flare is a possibility, to inform patients ahead of time of the possibility of PSA flare, to reduce their anxiety and, particularly, avoid changing therapy early on in the course of treatment within the first three months unless you're sure it's clear progression and not flare.

I was also assigned to talk about steroids with abiraterone, but this is going to be covered by Dr. Attard later in the day, so I'm going to skip that and show you a little data that we recently published on using abiraterone without steroids. So this is a paper in Cancer, Dr. Rana McKay is the first author, and we, together with our colleagues at Memorial Sloan Kettering, did a small trial, a 60-patient trial, looking at the use of full-dose abiraterone without any steroids, and we measured toxicity, hypertension, hypokalemia and edema, and you can see the grade 3-4 in the column, and I will note that, compared to the incidences in LATITUDE and STAMPEDE in which prednisone five milligrams qd was used, the incidences were not actually that much different.

There was only three patients in whom the steroids had to be added for toxicities, and I will say that we had very specific algorithms to follow for hypertension and hypokalemia and how to treat them. They're published in the manuscript, so any patient, any of you who are going to use abiraterone without steroids, I would refer you to this publication, and here are some swimmer plots of where the toxicities happen. The top one is hypertension. The hypertension can happen at any time during the course of the abiraterone, whereas the hypokalemia tends to occur early within the first three months, although there was one patient who had a later incidence.

We've heard a lot today of using low-dose abiraterone, but no one's actually showed the data, so this is the data from the paper from University of Chicago in which 34 patients in each arm, a standard dose or a low-dose, and the low-dose was actually given with a low fat diet, and the primary endpoint was at three months, so I will say the caveat is we don't have long-term efficacy with this approach, but at least in the early phase of treatment, the first three months or so, the response rates looked very similar, and that was whether the patients had had prior docetaxel or not.

So in conclusion, there's no validated biomarkers to select therapy in first-line mCRPC after either docetaxel or AR-axis therapy for hormone-sensitive prostate cancer. Clinical parameters of aggressive disease, which include short response to initial therapy, high tumor burden, rapid pace of progression, liver metastasis, poor genomics, may make one consider chemotherapy over other treatments. A favorable response to initial therapy, there are many options that we have, including additional AR-axis therapy, sipuleucel-T and chemotherapy. I would say genomically characterize the tumor early when possible in mCRPC. Abiraterone without steroids is not recommended routinely but can be considered in the appropriate patients with close monitoring, and low-dose abiraterone with food can be considered in special circumstances as well, again, with close monitoring for ongoing efficacy.

No cowbell, and thank you very much.