Interpreting the data from the S9346/INT-0162 trial in hormone-sensitive, metastatic disease

We have already reported data, presented by Dr. Maha Hussain yesterday, that appear to show that — at least in men with clearly metastatic prostate cancer — continuous and complete androgen deprivation (CAD) extends survival significantly better than intermittent androgen deprivation (IAD).

The data presented by Dr. Hussain at ASCO were discussed, immediately following her presentation, by Dr. William Oh. There is no formal abstract of Dr. Oh’s presentation that you can be referred to. Following Dr. Oh’s presentation there was a special session organized by ASCO specifically for the “open” discussion of the results of this trial.

Before we get into a detailed discussion of these data, we do want to point out two very simple facts:

  • This trial has now, for the first time ever, established that the median survival of men with metastatic prostate cancer who have a good initial response to hormonal therapy is of the order of 5.8 to 6.5 years. This fact has got lost in the controversy over the other trial results, but it is a key piece of new information that is very important indeed. On the other hand,
  • We are regrettably none the wiser about the appropriate timing of initiation of hormonal therapy than we have  been for most of the past 20 years. Should it be started immediately after the patient’s PSA starts to rise in indication of recurrent disease? Should we wait until there are either clear signs and symptoms of metastasis? Should we set “standard” PSA levels at which to start hormonal therapy (e.g., at 20 or 50 ng/ml) unless signs and symptoms of disease demonstrate the need to start hormonal therapy at lower PSA levels in individual patients? We really don’t have a clue.

Now, moving forward, there appear to be several critical issues about this trial and its results that were extremely worrisome to many in the audience at ASCO.

First, the results seem to be counter-intuitive, particularly the subgroup analysis which showed that men with more extensive metastatic disease did better on IAD than men with minimal metastatic disease. These observations led many in the audience to question whether there was something about the way that the trial was conducted that could have led to a result that is actually inaccurate.

Dr. Hussain, who is widely recognized to be a highly reputable and skilled clinical investigator, fully acknowledged that that this trial was initiated 17 years ago, when we certainly knew less about the progression and treatment of metastatic prostate cancer than we do today. However, she was also very clear that the trial was designed and executed to the highest possible standards given the knowledge available at the time of trial initiation, that every attempt was made to assure appropriate data collection to a very high level of quality, and that “the data are the data.” They had surprised her just as much as they had surprised most of the rest of the audience.

Second, Dr. Hussain, Dr. Oh, and others all agreed that we need to be careful not to over-interpret the difference in results between the men with extensive and minimal metastatic disease. This trial was not originally designed to identify such a difference as a primary endpoint. The difference between these two groups of patients is based on a subgroup analysis. Also, there are many who feel that the way the difference between extensive and minimal metastatic disease was defined back in the early to  mid 1990s would not be the way we did that today. However, the difference did appear to be very clear, and Dr. Hussain simply “stuck to her guns” and repeated that “the data are the data” even if they make many of us feel distinctly uncomfortable.

The third, and perhaps the most critical of all the issues was the practical, day-to-day, clinical implications of the results of this trial. What are ASCO and other guideline-developing organizations such as the National Comprehensive Cancer Network (NCCN) going to make of these data and are they going to tell practicing physicians that CAD is now established as the standard of care for the treatment of men with hormone-sensitive, metastatic prostate cancer?

In thinking through this third issue, The “New” Prostate Cancer InfoLink considers that there are a number of very important factors that need to be taken carefully into account:

  • CAD is well known to come with risk for serious side effects that affect quality of life. Newly diagnosed patients need to be made very clear about this in discussing hormonal therapy with their physicians.
  • This trial — whatever its flaws may or may not be — is the only large, randomized clinical trial that has ever been conducted in men with metastatic, hormone-sensitive prostate cancer and sufficiently powered to prove any difference in survival benefit between IAD and CAD. And Dr. Hussain is correct in her statement that we need to understand that “the data are the data.” They show that IAD is inferior to CAD in extending the survival of this well-defined set of men.
  • The only other trial ever carried out and sufficiently powered to demonstrate a difference in survival between men treated with CAD or IAD was the NCIC CTG PR.7 trial. The results of this trial, first presented over a year ago at the Genitourinary Cancers symposium, showed that IAD was non-inferior to CAD — but the men enrolled into this trial were not metastatic at the time of entry into the trial and, furthermore, that trial randomized patients at the time of trial entry. It did not randomize them after they had clearly shown a good and stable response to their first cycle of androgen deprivation.
  • Quality of life and extent of life are key issues that are subject to something known as “physician bias.” In other words, there is a tendency in the academic medical community to see survival as a more important outcome than quality of life. However, from a practical, day-to-day patient management perspective, the wishes of the patient need to be paramount. Some men may wish to just stay alive for as long as possible; others may say that they will happily sacrifice a few months of life to have a better quality of their remaining life.

The “New” Prostate Cancer InfoLink believes that the trial conducted by Dr. Hussain and her colleagues is a very important trial, and congratulates the investigators (and all the participants) on their 17 years of hard work. What it has shown us is that (yet again) what we hope and expect to be true may not turn out to be the case. However, what it also does not do is establish the idea that all men with metastatic, hormone-sensitive prostate cancer should be placed on CAD as opposed to IAD. It tells us simply that a patient and his doctor, in considering their options for the treatment of metastatic, hormone-sensitive prostate cancer today, need to discuss carefully the risks and benefits of CAD and of IAD. Yes, we now have evidence that IAD is less effective at extending survival than CAD in these men, but the decision whether to sacrifice quality of life for a few months of additional survival is one that only a fully informed patient and his family are capable of making — with appropriate guidance from a fully and well-informed clinician.

23 Responses


    I have read the other article about this trial. As a survivor of 12 1/2 years of a challenging case of non-metastatic prostate cancer treated only with IADT3 plus supporting drugs and lifestyle tactics, I have closely followed the thinking and both formally and informally reported clinical experience of doctors using ADT extensively in medical oncology practices dedicated to prostate cancer that have a high proportion of patients with advanced and challenging cases.

    These doctors have strongly favored intermittent therapy for many years for those of us without detected metastases (as determined by conventional tests rather than the much more accurate tests now emerging), but my impression is that they have also consistently stated that men with metastatic cases generally need to be on continuous ADT. This trial is consistent with that view.

    I’m curious whether DHT was monitored during this trial, but the answer will probably require an inquiry to Dr. Hussain. I’m curious whether any of the patients were taking a 5-alpha reductase inhibitor drug as well, either finasteride of Avodart.

    Thanks for posting these articles. I hope to have some other thoughts and questions shortly.


    The information published does not indicate whether the researchers called for a drug to help prevent decrease of bone density, perhaps also countering bone mets to a degree, for the trial participants. The doctors I have followed closely, who work with many such patients, would almost routinely employ a powerful bone density drug, such as Zometa, for metastatic patients. These days denosumab might be used instead. In the early days of the trial, other bisphosphonate drugs would have been available.

    By the way, a few years ago one of the doctors stated he had never encountered a case of osteonecrosis of the jaw in his patients on Zometa, and he attributed that success to dosing no more frequently than every three months. I do not know whether he has updated that experience.

    Another expert medical oncologist is now emphasizing trans-dermal estrogen to protect bone density in such patients and others with less advanced cases.

  3. Dear Jim:

    (1) Neither serum testosterone (T) levels nor DHT levels were assayed during the conduct of this trial. It is my understanding that the investigators wanted to assy the serum T levels but that the National Cancer Institute was unwilling to cover this additional cost.

    (2) It is possible that a few patients in this trial also took dutasteride at some point during this study, but if they did it would have been an “off protocol” use and if this had been discovered the patients would have been inelegible for inclusion in ther final trial data.

    (3) This ntrial was initially designed in 1993. My memory is that back then there was no data to support the idea that drug therapy should be use to protect bone denisty in men or women recieving hormonal therapies for cancer. Indeed, I am not sure when it was that Aredia was first approved for that use, but it might not have been until 1995 or later.

  4. There is also the results from the Spanish randomized study that supposedly showed that IAD was not inferior to CAD in men with advanced and metastatic disease, BUT the cancer mortality was inferior.

  5. Did the participants in this trial receive combination or triple blockade therapy, or was it single therapy alone?

    I also noticed that the intermittent track was given hormone therapy for 7 months, before taking a break. I’d be curious (of course, purely speculating), whether the more common 12-15 month “on” cycles today would make a positive difference in overall survival for those on IADT?

  6. Thank you for yet another thoughtful, insightful interpretation of a very complex subject, Mike.

    I particularly liked your comment:

    … from a practical, day-to-day patient management perspective, the wishes of the patient need to be paramount. Some men may wish to just stay alive for as long as possible; others may say that they will happily sacrifice a few months of life to have a better quality of their remaining life.

    Would that all medical people understood this and discussed options more carefully.

    Although I understand why scientific studies use the median in interpretation and discussion, I always try to have a look at the fate of those to the right of the median. So, we see from this study that of those men who were randomized to one of the arms 23 and 29 per cent respectively survived to live at least 10 years. To me that’s an important point because of the prevailing view that ADT fails after 3 years and death follows shortly after that.

    It is also important, I think, to see that about 40 percent of the deaths did not occur from prostate cancer, but from other causes, showing that even in cases such as these, there is still a substantial risk of dying from something other than cancer.

  7. Your comment, Sitemaster, that you have learned that testosterone was not tested to determine the effectiveness of the LHRH agonist (presuming that an LHRH agonist was prescribed in the trial) makes one wonder whether the LHRH agonist was providing the expectation of castrate level testosterone. It makes little sense to prescribe an LHRH agonist (or more recently the GnRH antagonist degarelix/Firmagon) and not determine whether that medication has brought testosterone down to clinically castrate level.

    Also, using the consideration that the trial began in 1993 and medications not changed in years hence, this would mean that there was no consideration for those men in the intermittent portion being prescribed a 5AR inhibitor during their IAD to inhibit returning testosterone from converting to dihydrotestosterone/DHT — a procedure that can prolong the period of IAD.

    I continue to find the trial having not yet proven to me that CAD can absolutely prolong survival longer than IAD with the protocols of ADT/IAD now being employed by those medical oncologists who specialize specifically in research and treatment of recurring and advanced prostate cancer. Like Jim Waldenfels, I am in my 16th year since beginning an ADT/IAD protocol, and though metastasis must be present somewhere in my system with recent elevating PSA despite Firmagon, nilutamide, and Avodart, I remain asymptomatic with imaging unable to identify metastasis location as well as circulating tumor cells not present in my blood stream.

  8. Gerry:

    I was under the impression that that study was not actually sufficiently powered to draw really meaningful conclusions.

  9. Tom:

    All patients in this trial received bicalutamide + goserelin acetate when they were “on” therapy (in the continuous therapy and the intermittent therapy arms).

    The time of therapy issue was a matter of intense discussion after the results of this trial were presented. As noted proviously, we might well structure such a trial differently if we were to start it today, but it is most unlikely that anyone is going to recommit to funding a trial like this that –today — might well have to last 25 years.

  10. Sitemaster:

    Very interesting. Is the intense after-presentation discussion available online, by chance?

  11. I honestly don’t know if it is available on line, but I don’t think so.

    I do know that it was being audio-recorded by ASCO, so it is possible that it is or will be made available in that format.



    In this trial the participant pool was narrowed to those patients able to achieve a decrease in PSA to ≤ 4.0, a value which I’ve seen used in other trials as an indicator of a good response to ADT. All those achieving that response were then randomized to either CAD or IAD. This threshold of ≤ 4.0, while certainly reasonable for this trial at the time, is in sharp contrast to the threshold used by doctors I think of as expert in ADT who have large practices dedicated to prostate cancer patients. Some of these doctors have had their results published in a major journal (PMID: 10706649); only combined blockade (LHRH agonist and an antiandrogen) was featured in that study, the same as in the Hussain trial.

    In my own case, following the approach of the doctors I referred to in my responses of 6/4 10:20 am and 10:32 am, as well as in the paragraph above, I was dearly hoping to achieve a PSA below 0.05 and keep it there for a year. (It took a while — 31 months, but I did.) That was back in 2001-2002, and the doctors’ clinical experience with many patients on triple blockade had led them to believe that achieving that kind of success — both threshold and duration — was important if a patient were to have substantial success with intermittent therapy. While they have since basically eliminated the duration aspect, as I understand it, thereby allowing less continuous time on ADT, there is now some thought that achieving a PSA of ≤ 0.01 is a superior indication of success. (After two cycles in which I achieved a nadir of ≤ 0.01, my nadir for this third cycle was 0.02 after 19 months of ADT3. However, with the aid of a late boost from low-dose thalidomide and Celebrex, plus lifestyle tactics, I’m now in my 26th month of vacation — not bad!)

    These same doctors are generally not satisfied with the patient’s ADT response even if the PSA drops to 0.5 and levels out. Their approach seems to be to switch drugs, dosages, or other tactics to see if ADT3 can further reduce the PSA. If it cannot, rather than using intermittent therapy, they appear to me to prefer trying other approaches, perhaps including second-line ADT (e.g., ketoconazole, transdermal estrogen), or other drugs such as Leukine.

    I’m wondering what proportion of current metastatic patients would be able to achieve a decline in PSA to ≤ 0.05. I suppose the proportion would be small, but I’ll bet they would do well on IAD if they succeeded in reducing their PSAs to such levels.

  13. Dear Jim:

    Respectfully, the study you refer to above is a retrospective analysis of data from a purely observational cohort of carefully selected and highly motivated patients based on a premise developed several years after the design of the study reported by Hussain et al.

    We simply do not know what the “best practice” might be today for the practice of intermittent therapy — or in whom it is most appropriate to attempt such a strategy. As I have clearly noted earlier, there are patients with metastatic, hormone-sensitive disease in whom it may be entirely appropriate to initiate intermittent hormone therapy once their PSA has been driven to a low (or to an undetectable) level by a “loading cycle” of hormone therapy, but the truth is that we do not know how to select the most appropriate patients; nor do we know how best to manage these patients over time in terms of time “on” and “off” hormone therapy. Even the most experienced practitioners are simply exercising their best clinical judgement, which is very different from the data generated by HUssain et al. in their 17-year-long, randomized, multi-center clinical trial. The data published by Strum et al. may well be hypothesis generating, but they have never been formally tested in any type of randomized trial compared to a “standard” drug treatment protocol. As a consequence, we really have no idea whether this type of treatment is “better” than any other type of long-term, hormonal therapy (intermittent or continuous).

  14. I had a nice chat with one of the researchers in this report tonight and he is pretty convinced that we have level 1 data to go by here. It’s his thought that if a man has metastatic disease in the bones then it’s time to say goodbye to testosterone. And as much as a man doesn’t want to hear it, castration is the best way about it.

    I have always questioned IHT and ADT3 for myself and thus far I can’t find good level 1 data on either. But this one is very tough to dispute and it appears that IHT is inferior (without the double negatives). I am on an HT vacation but I am being told that since I didn’t have mets detected in the bones that’s OK in my case. (I was stage pT3b using AJCC 2002.)

    After the many years that IHT and ADT3 have been available, those touting them have failed to prove their efficacy in level 1 studies. This study has flaws but it’s still a pretty telling story.

    I guess it will still come down to a patient’s preference. I am enjoying my HT vacation ….

  15. Is there any information about the primary treatment of the study cohort?

  16. Ralph:

    Ree primary treatment data … I suspect that the only qualification would have been the onset of metastasis and minimal prior neoadjuvant hormone therapy. Other than that, the patients probably included the entire gamut of patient types. The breakdown will presumably be given in the final publication, but Dr. Hussain did not provide these these details as part of her study presentation. She only had about 20 minutes to present the key data from a complex study.

  17. Surprised to read the view on castration. When I spoke to my oncologist and suggested this might be a good option for me, since I really don’t like taking pills/having quarterly shots, he was unequivocal in his response.

    He said that the failure rate with orchidectomy was no different to ADT, but ADT gave more options. (Including income options, I thought to myself).

  18. Terry:

    I am not sure, but you may be over-interpreting if you assume that the term “castration” was intended to mean orchiectomy/ochidectomy. The physician may simply have meant “castration” to imply either surgical or continuous medical castration. Perhaps Tony can clarify this for us.

  19. Ralphy and Sitemaster:

    I’m replying about Ralphy’s June 7, 2012 at 11:45 am question about primary treatments for men in the study group. Earlier I had some questions about the protocol and found the usual details on the web site. Of course the site provides nothing about results, but it does clarify the ground rules, including prior treatment issues.

  20. Thanks for the link James. After reading some prior publications on SWOG 9346, what is evident is that the myth about 18 to 24 months ADT survival is just that: a myth. Either continuous or intermittent ADT resulted in a longer survival benefit. Also evident is the obvious associated response of PSA to survival. I would still like to see an analysis of primary treatments subgroup to see if debulking by reducing the tumor load at the source has any significance in these results.

    I want to add that I still consider intermittent ADT an important protocol with significant benefits in both QOL and survival over continuous ADT in spite of the reported results.

  21. Ralph:

    The “myth” of 18-24 months survival on ADT (actually 18 to 36 months when I go back to look at David Crawford’s and Mario Eisenberger’s slides from the late 1980s and early 1990s) is still not a “myth” for the men who do not respond well initially to hormone therapy, and it is worth bearing in mind that men being diagnosed that long ago were almost certainly at higher risk for extensive disease than the vast majority of men being diagnosed with metastatic disease today.

    It seems to me now that several factors may be critical to the use of IHT and the selection of appropriate patients for that type of therapy, as follows:

    (1) Can one initially drive the patient’s PSA down to < 0.1 ng/ml and, simultaneously, drive his serum T level down to 10 ng/ml (or whatever is the predetermined re-start criterion for going back on therapy)?

    (2) How long should the initial “on therapy” period of hormone therapy be before the patient first comes “off therapy”?

    (3) How long the patient has a PSA level of < 10 ng/ml (or whatever other measure is prerdefined) and he needs to go back on therapy? (Would this be better of assayed using serum T levels as opposed to PSA levels?)

    (4) What is the patient's PSA doubling time when his PSA starts to rise again during the off=therapy period.

    (5) Should one exclusively use maximal androgen deprivation (i.e., an LHRH agonist and an antiandrogen) to carry out hormone therapy or are the data sufficiently compelling to use either single androgen deprivation with an LHRH agonist or 3ADT with the addition of dutasteride? None of these questions have been resolved as yet, so we don't even have an establshed baseline method (a "best practice") for the treatment of men by using hormone therapy (with or without and antiandrogen or dutasteride).

  22. Mike:

    The current concept of over-treatment is based on the fact that the great majority of men are diagnosed with earlier stages of the disease and many should be able to forgo or postpone treatment. In the ’80s many men were diagnosed with advanced disease. That included cases of metastatic disease which would respond poorly to ADT. My point is that the 18- to 24-month response to ADT is a myth for most of those diagnosed these days and such misinformation is not consistent with current diagnosis. In the previous post I was just wondering if primary treatment showed a long-term benefit.

  23. What were the criteria for the resumption of treatment in the intermittent group? The report just states that “Patients with rising PSA of progressive disease begin CAD therapy…”. Is this information available anywhere?

Leave a Reply

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

You are commenting using your 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: