Pharma Strategy Blog

Commentary on Pharma & Biotech Oncology / Hematology New Product Development

This weekend heralds the annual American Society of Clinical Oncology (ASCO) Genitourinary (GU) meeting in San Francisco, although ASCO held their press briefing today to provide an update on some of the key topics.

For those of you interested in Alpharadin (radium-225) in castrate-resistant prostate cancer (CRPC), check out the update of Dr Oliver Sartor’s presentation, which is covered on Biotech Strategy Blog.

The key topic that most interested me though, was Dr Howard Scher’s update on Medivation’s Androgen Receptor antagonist, MDV3100, in CRPC.  Previously, Medivation announced that the data showed an improvement in median overall survival (OS) of 4.8 months and this is still solid (Note: J&J’s abiraterone was approved by the FDA based on an OS of 3.9 months in the same population and must be taken with prednisone).

Three new things were important in this presentation though:

  1. There has been some previous concern about the risk of seizures, after they were reported in an earlier trial, but that was at much higher doses.  In this study, the now standard (and much lower) 160 mg dose of MDV3100, demonstrated low levels of seizures (0.6%), which is very reassuring and not something I think many will worry too much about.
  2. MDV3100 has a nice effect not only on OS, but also median time to confirmed PSA Progression, i.e. 8.3 months vs. 3.0 months for placebo (HR 0.248, P<0.0001). Yes, I had to do a double take at that HR – it’s quite phenomenal!
  3. Aside from PSA drops, patients often like to know if their tumour is shrinking or not as evidence of activity and progress. Dr Scher showed the soft tissue response by CT/MRI imaging. There was a 28.9% response rate with MDV3100 compared with 3.8% for placebo (P<0.0001).

From this data we can definitely say that patients lived longer, saw a positive impact on their PSA levels, and felt better compared to placebo. In terms of serious adverse events, there were fewer in the MDV3100 arm (28.4%) versus the placebo arm (33.6%). There were also slightly more discontinuations in the placebo (7.0%) than MDV3100 (3.8%) cohort.

Overall, I wasn’t at all surprised when the host, Dr Nicholas Vogelzang (Medical Director of the Developmental Therapeutics Committee of US Oncology) exuberantly said he had only one comment to Dr Scher’s presentation of the MDV3100 data,

“Wow, that’s very impressive! It’s unprecedented.”

For once, I thought that ‘impressive’ was actually an understatement to apply to a cancer drug.

I also talked to Dr David Hung, CEO of Medivation afterwards. Many readers will remember my interview with Dr Charles Sawyers, the co-inventor of MDV3100, last year about the science behind the development. It was nice to see Medivation’s side of the R&D story, which has gone pretty rapidly so far.

PSB: Are you going to be filing soon based on this data?

David Hung: We are having a pre-NDA meeting with the FDA. Once we have that meeting we will be able to give much more concrete guidance on when we will be filing.

PSB: Some of the pre-chemo trials have started, would they be due to report some data soon?

David Hung: We haven’t given any timelines on any of our other trials.

PSB: When I interviewed with Charles Sawyers previously, he said that many pharma companies were not interested in what is now MDV3100. What did you see in it when many others said “no”?

David Hung: Charles didn’t approach me. I found him! I had read, with great interest, his work on the AR. I was very familiar with his Nature Medicine publication in 2004 showing that overexpression of the AR is a significant molecular change in patients with castration resistant disease. While I think a lot of people thought that targeting the AR would create just another AR antagonist, like casodex, the data suggested to me there was more here.

Because, in Charles’ lab by being able to over-express the AR, we were able to much more carefully assay and screen compounds for their ability to block androgen receptor signaling very thoroughly. And we found in the process that a number of compounds in the series that we were testing had ability to not only block just AR binding by testosterone, which is something that Casodex does, but unlike Casodex these compounds were able to inhibit nuclear translocation as well as DNA binding and activation by the AR.

I am an oncologist by training and was pretty familiar with this area, so when I saw the compounds and saw the data in more detail, I didn’t agree that it would just be another casodex like molecule. I thought the mechanisms suggested that this drug could be special, so when I went ahead and licensed the drug back in 2005. We then took the program forward rapidly through development. We had to do all the standard pharmacokinetics, metabolism, tox, formulation work, then take it into a clinical trial as quickly as we could, led by Howard Scher. So, we were able to develop the molecule very quickly.

One of the differences with MDV3100 over weaker AR antagonists such as bicalutamide, is it’s ability to target splice variants. This was a surprising but important finding. I asked Dr Hung about them:

PSB: Does that potentially mean that the patients in the current trial data presented by Dr Scher, may actually do better over time or is the 4.8 months OS probably going to be the final number?

David Hung: Well, I won’t know the answer to that until I unblind the PREVAIL trial. What is very interesting from our phase 1 / 2 data is that the time to PSA progression in post-chemo patients in that data set is about 203 days. Yet, the time to PSA progression in the pre-chemo patients was 4x longer than that, 812 days, suggesting that the drug may have even more robust activity upstream than it does downstream. Downstream it already has robust activity. We will be greatly looking forward to seeing the PREVAIL data, because that is the pre-chemo population. If we can recapitulate our phase 1 / 2 results, that would be great news for patients.

PSB: At AUA last year, I heard from Charles Sawyers that if you inhibit the androgen receptor, you often activate the PI3-Kinase pathway. His colleague Neil Rosen had also noticed that if you inhibit PI3K, you activate androgen receptor in prostate models. So Charles was saying in their joint paper that the logical thing to do would be to combine an androgen receptor inhibitor and a PI3K-inhibitor to potentially reduce the resistance and hopefully improve outcomes. Is that the kind of combination you might consider in the future?

David Hung: We actually are. You point out exactly the kind of things that we think about. We look to see how our drug works and we look to see what mechanisms might possibly complement our drug. That is the way we think about potential combination studies that we might do.

All this is very exciting news for both Medivation (and commercial partner Astellas), as well as patients with advanced prostate cancer. I hope that the discussions with the FDA go well and we will see a filing, perhaps even with Accelerated or Priority Review in the near future. Based on the data so far, the data clearly shows that MD3100 can make a difference to the lives of men with advanced prostate cancer.

Next week, I’ll be at the American Association for Cancer Research (AACR) Special Conference on Prostate Cancer, jointly chaired by Charles Sawyers (MSKCC) and Arul Chinnaiyan (Michigan) to learn more about the biology of prostate cancer. It promises to be both a timely and exciting meeting!


ResearchBlogging.orgChen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R, Rosenfeld MG, & Sawyers CL (2004). Molecular determinants of resistance to antiandrogen therapy. Nature medicine, 10 (1), 33-9 PMID: 14702632

Carver, B., Chapinski, C., Wongvipat, J., Hieronymus, H., Chen, Y., Chandarlapaty, S., Arora, V., Le, C., Koutcher, J., Scher, H., Scardino, P., Rosen, N., & Sawyers, C. (2011). Reciprocal Feedback Regulation of PI3K and Androgen Receptor Signaling in PTEN-Deficient Prostate Cancer Cancer Cell, 19 (5), 575-586 DOI: 10.1016/j.ccr.2011.04.008

6 Responses to “Update on Medivation’s MDV3100 in advanced prostate cancer”

  1. Biomaven

    Interesting to compare with Zytiga:


    Zytiga survival was 14.8 months (drug+prednisone)  vs
    10.9 months (placebo+prednisone) as against 18.4 vs. 13.6 in this trial. HR’s
    were similar between these trials. So both arms in the Zytiga trial did
    considerably worse – either an effect of the prednisone (seems somewhat
    unlikely) or the Zytiga trial population had a worse prognosis.

    Mysteriously though, time to PSA progression was better in
    the Zytiga trial (10.2 months drug vs 6.6 months placebo)  than in this trial
    (8.3 versus 3.0 months). Seems unlikely that there would be a different
    definition of PSA  progression, but otherwise the results are very
    mysterious. Any clues as to what is going on here?Peter (“Biomaven”)


    • maverickny

      Hi Peter, it’s always tricky to compare different drugs from different studies as there are so many confounding factors. The absolutes are not always useful in this context, so we tend to compare the relative differences over placebo. It’s not ideal, but it’s a start.

      Using this approach, we see that the median OS numbers for MDV3100 and Zytiga over placebo in the post chemo setting are 4.8 and 3.9 months, respectively. Around 4 months is therefore a consistent number. I would expect this difference to increase a lot more in the pre chemo setting, certainly more than we saw for Provenge. As you can see, the PSA progression is also around 4 or 5 months, which makes sense in the context of the OS.

      Hope this helps, Sally.

  2. Biomaven

    I take your point about comparing across trials, but still if you look at time between medians from PSA progression to death, it’s only 0.7 months for the Zytiga trial and 10.1 months for the MDV3100 trial. That’s a pretty dramatic difference.


    • maverickny

      That’s correct, Peter.  My guess is that the difference might be possibly be explained by the splice variants, which as Dr Hung noted in the audio clip are one of the factors responsible for resistance in CRPC.  MDV3100 targets splice variants, abiraterone does not. It’s one of the seemingly small differences between the drugs that may actually have a meaningful clinical impact for patients.

  3. Mguiltinan

    Sally thanks for your MDVN reports which I’ve enjoyed and especially your interviews. Any idea as to when we could get an interim peek at the PFS? The Prevail trial is way over powered at 1700 patients imo. I believe MDV3100 will be even more effective as a pre-chemo drug so a peek could be revealing. I thought the PFS data was impressive in the Affirm trial. So I’m hoping we don’t have to wait for Prevail’s OS data. Do you think the FDA would consider approving MDV3100 in a pre-chemo indication with only strong interim PFS data and well before the OS arrives?

    • maverickny

      The pre-chemo data for both MDV3100 and abiraterone is taking a long while to mature, which suggests to me that the patients are doing remarkably well.  There is no doubt in my mind the data for both drugs will be better prior to chemotherapy, but how much is anyone’s guess.  The timing and availability of the interim PREVAIL data is clearly going to be dependent upon events, so we won’t know anything until that point.

      The magnitude of the difference versus placebo will probably determine whether PFS will be enough or not, but who knows how the FDA will react to the DSMC’s advice? These things are notoriously difficult to predict. My guess is that a PFS of 6 months or less will mean waiting for OS, more than that and the DSMC will be under pressure to recommend unblinding the results. Of course, the danger is that once the study is unblinded and patients cross over, we won’t actually be able to derive true OS.

      The good news is that the longer the abiraterone and MDV3100 trials are delayed in terms of interim results, the more likely patients will be doing very well on therapy.

Comments are closed.

error: Content is protected !!