Wishful thinking about abiraterone acetate in CRPC

An article from Thursday’s “The Pink Sheet”, posted on the OncologySTAT web site today, is already asking questions about the potential future use of abiraterone acetate in patients with non-metastatic, castration-resistant prostate cancer (CRPC).

To access this article (“Zytiga cleared fast in refractory CRPC, can it also move to earlier use?“), you do need to register with the OncologySTAT web site, but there is no cost for doing so.

The use of abiraterone in men with non-metastatic CRPC has not yet been investigated in any significant clinical trial.

The Southwest Oncology Group has planned a small Phase II clinical trial of abiraterone in men with a PSA > 4 ng/ml after prior treatment with androgen deprivation therapy, but this trial is not yet enrolling patients.

It is possible that Johnson & Johnson may delay initiation of a Phase III trial in non-metastatic CRPC until they have data from the ongoing, fully enrolled trial in men with metastatic, chemotherapy-naïve CRPC, but we don’t think this is likely because that ongoing trial is unlikely to report results until some time in 2012. We would expect to see initiation of a randomized, double-blind Phase III trial of abiraterone in men with non-metastatic CRPC some time in the next 6 months.

As we have previously noted, the increasing number of agents available for the treatment of advanced forms of prostate cancer is starting to complicate thinking about the best management strategies for patients with progressive disease. The article referenced above suggests, for example, that abiraterone will almost certainly be seen as the “go to” therapy for men following first-line, docetaxel-based chemotherapy in preference to cabazitaxel. Why? Because abiraterone seems to have a superior safety profile compared to cabazitaxel and a longer survival benefit in men who have stopped responding to (or who are just refractory to) treatment with docetaxel + prednisone. Will the median survival benefit of abiraterone followed by cabazitaxel be additive (4.6 months for abiraterone + 2.4 months for cabazitel for a total of 7.0 months) after docetaxel? We may not know that for quite a while.

From the viewpoint of men with or at risk for advanced forms of prostate cancer, such complications are good. Just a decade ago there were no drugs at all that offered a survival benefit once standard hormone therapy was unable to control disease progression.

The article from “The Pink Sheet” also confirms the widespread rumor that the cost for abiraterone acetate is $5,000 per month or $60,000 per year of treatment.

12 Responses

  1. To complicate the matters, I think there is one more path. What course of treatment a patient who responded to docetaxel but progressed later should follow? I think at one point the docetaxel treatment should be repeated but when I am not sure.

  2. Summer:

    It is certainly reasonable for a patient who responds well to docetaxel to consider undergoing a second course of docetaxel-based chemotherapy. As far as I am aware, there are no good data (as yet) to guide us as to whether that should be done before or after a course of abiraterone or any other form of treatment for CRPC.

  3. In the abiraterone trial, about 30% of the patients had two prior docetaxel regimens (see Section 16 of the abiraterone prescribing information). And I’m sure some patients had docetaxel re-treatment after the study. Since this was a large trial, even these subsets should be able to give us some clue about which sequence works best. I hope the peer-reviewed publication will analyze these subsets and talk about post study therapies as well.

  4. Summer:

    I’d be surprised if the primary report of the trial data analyzed the subset data from this study. However, a secondary report from the trial may do this.

  5. Median survival figures from clinical trials are an average based on patients who respond to a therapy but also including results for those patients who do not respond. That’s a good figure of merit for evaluating drugs in clinical trials. For abiraterone, the median survival benefit was 4.6 months. However, that figure does not tell us a whole lot about how patients will fare in actual clinical practice where oncologists will typically be able to tell fairly quickly whether a patient is responding to a drug or not, and, if not, will be able to try another drug or approach. Moreover, that median survival benefit figure for the trial does not tell us the length of response for those patients who do respond. In fact, such successful responses are often frozen at a pre-determined end point, as they were for Provenge at 36 months.

    Do we know the percentage of patients who had strong responses to abiraterone in the trial? Do we have figures describing the length of response among those who did respond?

  6. Jim:

    What happens in clinical trials is actually not that dissimilar from what happens in clinical practice today for advanced disease. Patients are followed until progression and then another form of therapy is introduced.

    However, it would be highly unusual for data to be reported in the published results of the Phase III abiraterone trial giving time to abiraterone failure in the men abiraterone + prednisone arm of the trial with a less than median overall survival. It would be equally unusual for the study results to report time to treatment failure in the men on the abiraterone + prednisone arm of the study with a higher than median survival. Such data can be calculated, but the truth is that such data already reflect where the median survival time occurs, so it is a pre-prejudiced data point. What will be much more important is the range of the survival times around the median, which may be as low as a month or two for the least responsive patients and run into as much as 2 years for the most responsive ones. I would be very surprised to see a large number of men in this trial (who had, after all, already failed at least two types of hormone therapy and one cycle of docetaxel) demonstrate a long-term survival of > 2 years, although there clearly are likely to be a few. When 775/1,195 men in the trial had died, the median overall survival of the men in the abiraterone + prednisone arm of the trial was 15.8 months (as compared to 11.2 months on placebo + prednisone).

  7. Thanks, and I appreciate the points in your 1 PM response.

    I am wondering though how energetically the trial doctors would be reaching beyond the standard of care approaches for patients who were progressing. In talking to a couple of leading clinical trials doctors, I’ve had the impression that they were a bit reluctant to take steps that might tend to interfere with their science. Any thoughts on that?

  8. Jim: I am not sure that I understand your question.

    How one treats a patient in a clinical trial is highly defined by the trial protocol. If one deviates from the protocol, the patient data gets excluded from the analysis. It’s hard enough to enroll patients for clinical trials. Of course “clinical trial doctors” are “reluctant” to interfere with the trial protocols, which are complex and carefully structured for good reason. No good “clinical trial doctor” is going to deviate from the trial protocol until a patient has failed in a specific trial. After a patient fails in a specific trial, they can go on to do whatever else seems appropriate. And remember that individual doctors usually do not know (any more than the patients do) whether a specific patient is on active drug or placebo in a double-blind randomized trial.

  9. I appreciate that good point about the blinding in your 6:20 am post, but I’m still concerned about the vigor of treatment after patients have progressed. Ideally from a science viewpoint, nothing would be done after progression so that the natural course of the disease after treatment or placebo would be the clearest, with minimal interference with the survival endpoint. That of course would be unethical and clearly not in the patient’s interest. From a patient’s viewpoint, anything practical with a reasonable potential for benefit should be tried. As research continues to identify promising options and to gain genetic and other indicators about when each option is likely to work, this disturbance of trial outcomes would seem to be a growing issue. As a hypothetical, suppose that a combination, which I’m making up from thin air — no basis, of GM-CSF (Leukine), pomegranate extract, selenium, abiraterone and Zometa were to emerge as a high-response survival extender for late-stage patients with a certain genetic/epigenetic biology that was not known at trial launch. Unless that advance would have been endorsed by those institutional leaders who consider themselves the elite and enlightened, I’m wondering about the chances that patients in the trial would be treated with such a combination after they had progressed. As you note, trials and treatment are becoming more complicated.

  10. Jim:

    This is hardly a new question. In the treatment of advanced breast cancer this is now a constant problem because dozens of possible drugs and drug combinations can and have been tried as treatments for advanced and refractory disease. However, very few of them have ever shown a really clinically significant outcome.

    Let us assume that we call your speculative combination therapy (described above) as “Cocktail A” and that in June 2012 data is reported at ASCO that Cocktail A is associated with a 12-month survival benefit compared to cabazitaxel in the treatment of men with metastatic, castration-resistant and docetaxel-resistant prostate cancer. OK? That is clearly is significant survival benefit.

    So in the meantime, patients in a Phase III clinical trial of New Drug X + prednisone vs. cabazitaxel + prednisone in metastatic, castration-resistant and docetaxel-resistant prostate cancer are starting to fail therapy. So as of June 2012, your concern is that the men who are failing on either New Drug X + prednisone or cabazitaxel + prednisone are immediately given Cocktail A as opposed to any other putative combination (such as, for example, mitoxantrone + prednisone).

    Since nearly every leading expert in treatment of mCRPC would have been at ASCO when those data were announced, it seems highly likely that Cocktail A would very rapidly become the treatment of choice for metastatic, castration-resistant, docetaxel-resistant prostate cancer. This is a really very small treating community. The word goes around a community like that about highly effective new treatments in a matter of days. (Look how fast we all knew about the results of the abiraterone trial presented in Europe last October … a matter of a few days at worst.) It would be considered unethical to withhold such treatment from a patient who had had progressive disease on New Drug X + prednisone (or on the other arm of the trial).

    Might it impact the overall survival of the patients treated with New Drug X + prednisone or cabazitaxel + prednisone in the Phase III clinical trial? Yes, it might, but theoretically it should affect both groups of patients by the same amount, and so if there were to be a survival benefit from the use of New Drug X + prednisone compared to placebo + prednisone, this should still be evident.

  11. Interesting discussion. This hypothetical situation is not too far away from what’s currently happening in mCRPC trials..

    I think currently prostate cancer drug developers do everything in their power to mitigate such interactions, which might be borderline unethical. For example, Medivation, the developer of MDV3100, issued a Form 8-K that includes the following statement after J&J announced good results:

    “Medivation has specific measures in place to mitigate any potential impact of the abiraterone acetate data on the ongoing AFFIRM trial. However, for competitive reasons, Medivation will not be discussing publicly what those specific measures are.”

    I’m personally very curious what these measures are and withholding such information from the patients in the trial should be at least frowned upon if not unethical.

    Another practical issue is that not all current prostate cancer trials are double blind. For example, this OGX-011 Phase III trial is open label.

    So after abiraterone becomes widely available, would currently enrolled control arm patients be more inclined to declare progression to get out the trial sooner? What would be the effect of that on the trial? Who knows? Do OncoGeneX and TEVA do anything to mitigate this? I’m sure. Is that unethical? We don’t even know what they do. So how can we discuss it?


  12. I am very interested in the clinical characteristics of the patients that failed to respond to abiraterone.

    Also, I wonder if a significant portion of them had AR-Vs. I think that these questions would help to carve out niche populations who could benefit from cabazitaxel or abiraterone.

    Any thoughts about sequencing?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: