In Response to the Reuters Article Featuring Marie Huber's Analysis of the Pivotal Provenge Study

To the editor:

As someone who covers the biotech universe*, I frankly am appalled that Ms. Huber’s writings have attracted the attention they have in the popular and medical media. The Reuters’ article you posted to your site contains significant misleading statements and does a disservice to readers, medical practitioners, and prostate cancer patients alike.

Insight: New Doubts About Prostate-Cancer Vaccine Provenge
(Reuters) Mar 30, 2012 – Prostate cancer vaccine Provenge has long incited passions unlike any other cancer therapy.
read article »

Ms. Huber’s specious theory has been and continues to be ridiculed by mainstream immunologists and other scientific thinkers, including none other than Dr. James Gulley, the Director of Clinical Trials at the US National Cancer Institute. As can be seen, for example, in the descriptive material related to Dendreon’s Sipuleucel-T (Provenge) on Wikipedia:

‘This theory was brought up at the CMS Advisory Committee Meeting in November 2010. At that meeting, Dr. James Gulley, Director of Clinical Trials at the National Cancer Institute, dismissed it as follows[15]: “The number of white blood cells that were, the proportion of white blood cells that are removed in terms of the total body white blood cell count is around two percent, so it is not a clinically meaningful amount.” In addition, a recent retrospective analysis of patients on the control arm of the IMPACT trial[16] showed that those treated with salvaged Provenge many months after disease progression had median survival time 23.8 months which was far better than the 11.6 months seen with the pure placebo patients. The median difference between patients on the control arm treated and untreated with salvaged Provenge was a direct contradiction to the Immunodepletion theory.

[15] “An analysis to quantify the overall survival benefit of Sipuleucel-t accounting for the cross-over in the control arm of the IMPACT study“.

‘Lastly, the above same analysis of the control arm of the IMPACT trial showed that the recorded median difference of 4.1 months might have been an underestimation due to the longer life of patients treated with salvaged patients. The presentation[16] stated: “Using the RPSFT model, and assuming that APC8015F (salvaged Provenge) was equally effective as PROVENGE, the median overall survival benefit of PROVENGE in the Phase 3 IMPACT trial was estimated to be 7.8 months, had there been no cross-over to APC8015F (HR=0.60, 95% CI: 0.41, 0.95).”‘

[16] “NCT00779402: Provenge for the Treatment of Hormone Sensitive Prostate Cancer“. US National Institutes of Health.

It must be noted, by the way—and unfortunately, this is either overlooked or ignored by the media—that the Journal of the National Cancer Institute (JNCI) has no relationship whatsoever—NONE—with the United States National Cancer Institute. The JNCI is published by Oxford University Press in England.

As to JNJ’s Zytiga, that drug’s Phase 3 trial was recently stopped because it reached stat sig one of two primary end-points: progression-free survival (PFS). The trial did not reach stat sig on the other primary end point: overall survival (OS). Now, with patients crossing over from the placebo cohort and taking Zytiga, chances are getting slimmer by the day that the trial will reach stat sig on OS. Whether or not Zytiga is approved for pre-chemo PCa patients, then, will depend on the strength of the PFS data and how close the trial comes to achieving stat sig on OS.

Finally, writing in the March 2012 issue of AUA News (p. 42), Dr. Carl A. Olsson had this to say about the paper published by Huber et al. in the JNCI:

“The authors missed the report by Hall et al that “there was no evidence that leukapheresis led to immunodepletion,” citing literature proving that each apheresis removed less than 1% of the total body pool of 1012 lymphocytes and reporting a normal measured cell count after the third apheresis in all men.[3]

“Finally, we are all used to the values of interdisciplinary conferences, which usually combine the experiential qualities of urology, clinical oncology, immunology and other disciplines. However, “healthcare analyst” and “investment management” are talents that appear unseemly in major publications.”

[3] Hall SJ, Klotz L, Pantuck AJ et al: Integrated safety data from 4 randomized, double-blind, controlled trials of autologous cellular immunotherapy with sipuleucel-T in patients with prostate cancer. J Urol 2011; 186: 877.

by Theodore J. Cohen, PhD

*Theodore Jerome Cohen, PhD, writes for Seeking Alpha and is a Dendreon Shareholder. He is the author of Death by Wall Street: Rampage of the Bulls. [ed]

article register

Recent Posts

Recent Comments



  1. @Danhen #37
    “An alleged antitumor product does not show any antitumor activity but prolong “survival'”

    It’s clear your knowledge of the field is lacking. It is well-known that the RECIST criteria commonly used and used in the Provenge trials to detect tumor progression are inadequate to test the efficacy of an immunotherapy.

    Here’s the NEJM paper on Ipilimumab (Yervoy) for treating melanoma:

    Ipilimumab had similar median progression time for both trial arms just like what we saw in the IMPACT trial. However, the survival endpoint for the Ipilimumab trial was highly statistically significant. The FDA approved Ipilimumab based on the survival data.

    Now read this paper and related references from there for a broader discussion on assessing progression for an immunotherapy:

    If that’s too taxing, this commentary published on the JNCI is simpler:

  2. would be sorry if one does not see the clear points. No intention or interest to argue at the preset levels that showed incomplete clinical understandings of the study disease and issues concerning the trial. An alleged antitumor product does not show any antitumor activity but prolong “survival’, a mirage or something between Ponzi’s scheme and Lottery. No financial conflict at all. The simple motivation is to help patients make sound decisions in the presence of a highly questionable and probably detrimental product. thanks for your attention. So long………

  3. Danhen,

    I am still waiting for you to respond with a level of detail comparable to the specific points of Dr. Cohen and Ocyan. They have both presented detailed arguments, but your responses have been mostly general. Please help me, and all readers of this topic,to understand the substantive reasons for your opinion.

    Also, Dr. Cohen has stated that he owns stock in Dendreon. However, you have not addressed the point that you could have investment interest in one or more competing companies. If so, that interest could motivate you to want to argue against Dendreon simply for the purpose of hindering Dendreon. Therefore, if you are going to make such a big case about Dr. Cohen’s Dendreon stock ownership then you are required to reveal any interest in any company that competes with Dendreon.

  4. @Danhen

    Not sure what’s up with you to continue making unwarranted investment assumptions on people who you know nothing about.

    My interest here is to make sure that no patients will be denied treatment because of people who do not understand the nature of data making unwarranted assertions about it.

    Now, about the data. The lesson that I’ve tried to teach and you still have not gotten is this. The short MST for pure placebos, the long MST for patients treated with Frovenge, the highly statistical significant result for patients older 65, and the not significant for patients less than 65 are all aspects of the same underlying issue, subgroup biases. In the case of the Huber’s paper, it is not difficult to find counter-examples and counter-arguments to her assertions. Such assertions cannot be trusted.

    The results that we can trust should be based on analyses of entire data and for prospectively defined hypotheses, i.e., the data that got Provenge approved by the FDA.

    For Ms Huber and people like you to keep making up strawmen assertions, then demand proof to counter them is at best an annoying attempt to waste other people’s time and at worst deceitful!

    This is my last post on this thread.

    Happy Easter. And, so long.

  5. DANHEN,

    I revealed, in a footnote at the bottom of my Letter to the Editor, that I was a Dendreon shareholder. What difference would it make if I owned 100, 10,000, or 100,000 shares? I’ve been in the stock since 2006. Even if you knew the number of shares I owned, how would you know what I paid for the stock? $3.86? $12.56? $35.27? All of the above, at one time or another? How do you know what profits I’ve taken in the last six years? What losses? Did I sell or purchase options? Calls? Puts? Etc., etc., etc.

    The fact is: it’s none of your business. I’ve revealed the one fact that is importantant for people to know in the footnote.

    Even more to the point, as a matter of practicality, just the thought anything I write or do regarding Dendreon can move the price of this stock is fantasy. We’re talking a public traded company, sir, with 147.20 million shares outstanding, of which two-thirds are held by institutions and mutual funds. It has a market capitalization of $1.54 billion and traded 4,751,600 shares last Thursday (the last day the stock traded because of the Good Friday closing). Get a grip on reality!

    As for the sequencing of Provenge and Zytiga, I only can refer those who seek guidance in this area to the Journal of Clinical Oncology and to NCCN, which gave Zytiga a category 2B rating in the pre-chemo setting.

    I will let Ocyan speak for himself.

    Be well.

  6. How fair is it when you accused of others having some detailed conflicts of interest and you avoided disclosing yours? Trying to reverse your investment loss with continuously advocating a sham product at the expense of patient’s interest, a malicious goal! May you have a good luck!

    For those who want to use the product, you certainly have your choice but the risk of not prolonging but shortening your life is highly present with the current uderstandings of the data. The system failed with numerous factors including greedy investors (some even act as patient’s advocates, yakh…) and wishful thinking academic politics.

    How patients who received “pure so-called placebo’ survived (with a median of 12 months) is the key for understanding the mechanism by which Provenge seemed to work: generating a difference in the survival compared to the control. In the modern medical oncology hisotry, no products that work without targeting tumors have been found to affect patients’ clinical outcomes. In contrast, a lots of products that have shown antitumor activity failed to translate into clinical benefits.

    Ocyan, the company used Halabi’s predicted survival in the IMPACT baseline evaluation and claimed that the predicted median survival of patients in the control arm was comparable with the median survival observed in the trial at the same arm. Why don’t you address whether there was a similar comparability at other key time points in the control arm such as at years 1.0, 1.5 and 2.0? In addition, you said “As to any imbalances in prognostic factors that may influence such numbers, nobody here really knows. We don’t have the raw data to compute with”. You were ignoring issues again as you do not need raw data here and most investigators in the prostate field would agree that the atrasentan’s study popuation had more unfavorable prognostic factors than the study population of IMPACT and that few patients in the atrasentan trial (stopped 14 months prior docetaxel approval in 2004) received docetaxel treatment as compared to 50-60% patients did in the IMPACT.

    Use of Provenge before Zytiga in the prechemo setting is purely an assumption, proposed by those who has been misled or ill-willed either academically or finacially. For a sham prodcut, does sequencing matter anymore? Hope Zytiga data tell more in coming months. So long…

  7. @StCold04:

    Mr. Seals was fortunate indeed that he was able to obtain the data needed to file a formal Appeal with Washington State authorities in the matter of Kaiser Permanente initially rejecting his application for Provenge based on their invoking the Huber JNCI paper. Fortunately, State authorities recognized how patently ridiculous Kaiser’s argument was, and took swift action to resolve the matter in Mr. Seals’ favor. He is now undergoing his first Provenge infusion.

    In the asymptomatic/minimally symptomatic mCRPC setting for which Provenge is indicated, the National Comprehensive Cancer Network (NCCN) has given Provenge a category 1 rating, the highest evidence level recommendation, based upon three Phase 3 clinical trials. By contrast, and in Provenge’s “space” – the asymptomatic/minimally symptomatic mCRPC setting – Zytiga (abiraterone) has a category 2B rating, the second lowest rating, based upon one small single arm Phase 2 trial. This has not stopped some physicians from prescribing Zytiga off-label to PCa sufferers in an estimated 22% of cases where patients probably should be treated with Provenge first. Current thinking,however, is that prescribing Zytiga first will delay medication with Provenge. The reason for this stems from the belief that if Zytiga is given first, a patient’s system must first be cleansed of prednisone (which is co-administered with Zytiga) before Provenge can be given. Detox can take up to 60 days.

    Here’s the problem with this. Giving Zytiga first (off-label) and then, having to wait for the patient to detox, could result in the disease progressing to the point where the patient no longer is eligible for Provenge. This issue was addressed in the Journal of Clinical Oncology, which made the following recommendation:

    “The practical dilemma of the appropriate sequence of use of the two new noncytotoxic agents (sipuleucel-T and abiraterone) is being addressed by trials that are under development. For now, given the broader window of applicability of abiraterone and the longer time required to develop an immune response with sipuleucel-T, if both agents are to be used, it seems reasonable to administer sipuleucel-T first with Abiraterone after additional disease progression. Biomarkers to help define the optimal use of immunotherapy are needed.”

    Dendreon and JNJ currently are running sequencing trials to determine whether or not the two treatments can be given concurrently or a detox period must be imposted if Zytiga is given first.

    Of course, when it comes to Dendreon’s Provenge, JNJ’s Zytiga, Medivation’s MDV3100 (now also know by its generic name, ‘enzalutamide’), and other treatments coming down the road, it’s not a case of ‘either/or’ but rather, how do doctors prescribe them them in combination? This is why studies related to sequencing are so important.

    Finally, at last month’s NCCN meeting, the organization gave Provenge a 2a rating (at the bottom of the list) in the category of ‘symptomatic prostate cancer.’ This is an off-label category for Provenge, but if insurers follow NCCN guidelines, they may reimburse for the treatment.

  8. @bobrmd

    The analysis in the ASCO abstract was based on a regression analysis, first assuming some functional model, i.e., the rank-preserving structural failure time (RPSFT) model as mentioned in the abstract, then assuming that Frovenge had the same effect as Provenge and computing a correction of effect.

    So that’s not just a straight comparison between the subgroup of Frovenge against the Provenge arm which would be susceptible to the selection bias that you mentioned. RPSFT preserves the original randomization to avoid biases and do the correction based on some assumption about the goodness of the cross-over protocol. In this case, Dendreon assumed that the cross-over protocol had a good an effect as the treatment protocol. So, that’s a fairly conservative assumption. That is, as far as the RSPFT procedure is concerned, the 7.8 months computed wouldbe close to a lower bound of the difference in MSTs between the trial arms.

    Nonetheless, one could still argue about whether RPSFT is an appropriate model for a post-hoc analysis. But that’s an entirely different can of statistical worms.

  9. @Danhen

    Is it right for anyone to call you dishonest for not revealing all your investments so that they can tell which one may be competing with Dendreon? Just to be clear, that’s a rhetorical question. I don’t really care to know as it’s none of my business. Please stop the nonsense of accusing people of some sort of ulterior motives and just focus on the facts.

    As to the Atrasentan data, I went by what the FDA revealed in their Clinical Review of the Atrasentan NDA. The 95% confidence interval for the MST of the control arm was [17.5,21] months. So the slight improvement of the MST in the 2007 paper to 20.5 months was within statistical error.

    The point in the above is that these median numbers are not exactly cast in concrete. So when you use them, you have to be careful to see the limit of validity in your usage or computations. Making an absolute comparison between two medians for better or worse without considering their confidence intervals ranges from being naive to being nonsensical. Btw, this was one of the things wrong with Ms. Huber’s paper.

    Now back to the Provenge and Atrasentan data. The MST of the control arm was 21.7 months with a 95% confidence interval [17.7,23.8]. This CI overlaps with that of the control arm of the Atrasentan trial with a slight shift to the right. So, generally speaking, the placebo arm of IMPACT performed in line with or just a tad better than what’s to be expected of such patients.

    As to any imbalances in prognostic factors that may influence such numbers, nobody here really knows. We don’t have the raw data to compute with. And, btw, this is another thing wrong with Ms. Huber thinking that there was something wrong with the younger and older than 65 subgroups in IMPACT. What imbalances lurked in those subgroups and how did she account for them?

  10. Danhen:

    I already have stated that I am a shareholder. Beyond that, I have no obligation to reveal personal information to you, financial or otherwise.

    Debate the issues as they stand, on the facts.

  11. It is absolutely fascinating to follow this discussion and I thank you all for posting. I would like to support, very strongly Todd’s quest for the best treatment he and other patients at this stage of the disease deserve! I am a front line FD in Canada whose involvement it the treatment of prostate cancer patients consists of indentifying those who might suffer from one, referring them to urologists , following them with oncologists while they receive hormonal treatment and taking them over after chemo fails to comfort them in their final days… I am making patients very aware of the Provenge while they are still responding to hormonal suppression. There are some of them who are willing to pay for Provenge treatment in US. Provenge is not available in Canada but Zytiga already is! It the cost ~$4300/month. It hurts me however, when patients who refract, return from oncologists telling me that they were recommended Zytiga as the better option than Provenge”! I guess one cannot underestimate the power of BP!

  12. Be honest to the public, disclose how many shares you have in DNDN then continue your discussion. I had and have 0 share, but i have seen prca for years. The IMPACT had about 80 percent of patients with asymptomatic diseae and about 20 percent with minimally asymptomatic disease, all had NO visceral involvement. The atrsentan trial mentioned above had about 50 percent of patients asymptomatic and INCLUDED patients with symptomatic and or visceral disease. The median survival was 20.3-20.5 months based on the publication in Cancer in 2007. Cleay, the IMPACT had a patient population better in outcomes than that in the atrasentan trial at the beginning. More importantly, the atrasentan trial was stopped 14 months before taxotere was marketed in 2004. Again, be honest.

  13. The placebo arm results contained patients who received frovenge. Of course the frovenge patients showed a survival almost if not as good as the provenge treated patients. Thats because the provenge patients included patients as sick as the placebo group who did not receive frovenge. There is a subgroup of patients in any trial who do poorly. The frovenge patients had those weeded out whereas the provenge patients did not. So its not a surprise that frovenge patients did well because they were the sunsgroup that did not rapidly progress.
    Another interesting trial would be to compare frovenge with provenge. I suspect frovenge would do worse because the rapidly progressing group would be in both arms and not just in the provenge grup as in the impact trial.
    The only way to test frovenge versus provenge would be to do a randomized trial. That, like provenge versus placebo in this population will never happen. So there will always be unanswered questions as is the case in most therapies. However, to assume provenge has a median survival benefit of 7.2 months over a true placebo assumed provenge equals frovenge efficacy which raises another set of questions apart from the immunpodepletion theory. For the above reasons, I don’t consider the asco abstract showing a possible 7.2 month survival advantage for provenge over a true placebo very convincing. The frovenge group were a faverable subset.

  14. @Danhen
    “How often do you see patients with asymptomatic early MCRPC decease within one year after metastases occur?”

    First, you asked the wrong question. The IMPACT trial enrolled patients with minimal pain. This included “asymptomatic” patients but there were patients with worse conditions, just that they did not require a systematic treatment for pain.

    Now, let’s assume that your question was about the patients enrolled in the IMPACT trial. Consider a trial with similar patients, the phase-3 trial for Atrasentan on minimal pain metastatic prostate cancer. In that case, the overall survival rate at 1 year was about 70% for both trial arms (Atrasentan is not known to effect survival) or about 30% passed away by end of the first year after enrollment. The median survival time for those patients was about 18.6 months in line with what we know of these patients.

    Now, look at the data in the IMPACT trial and what do you see? The placebo curve for IMPACT estimated about the same, a 72.4% survival rate at 1 year. The MST for the IMPACT placebos was 21.7 months, slightly better than the Atrasentan trial, perhaps thanks to the patients who crossed over to Frovenge.

    So, at least by the data from the Atrasentan phase-3 trial, the placebos in IMPACT behaved in line with or better than the “pure placebos”! Where is the evidence that Immunodepletion harmed them?

    And, how did the Provenge patients perform? 81.1% survived at one year and MST 25.8 months. So, Provenge did act early to keep a higher number of treated patients living past one year comparing to if they were not treated. The Provenge MST was 4.1 months better than the IMPACT placebos and 7.2 months better than pure placebos as estimated in the Atrasentan trial.

    Btw, a post-hoc analysis correcting for the cross-over effect that Dendreon recently presented at ASCO showed that the median difference between Provenge and pure placebos could be as high as 7.8 months. That is in line with the above. You see, objective data don’t lie as they have a way of independently confirming one another.

  15. Frovenge contained more than damaged monocytes. It contained frozen celss incubated with the casette. The cassette contains gmcsf a potent stimulant.

    A vaccine made from immune system cells taken from a patient with prostate cancer and frozen for future use. The cells are treated in the laboratory with a growth factor attached to a protein called prostatic-acid phosphatase (PAP), which is found on prostate cancer cells. When APC8015F is injected into the patient, it may cause T cells (a type of white blood cell) to kill tumor cells that have PAP on them.

  16. Dr. Moolten…one last note. Dendreon is going to run a trial with a true placebo. I covered this in an earlier article on Seeking Alpha but neglected to mention it above. Here is what I said earlier:

    “Corporate presentations at healthcare conferences conducted by Wall Street financial firms don’t often reveal much that the Street doesn’t already know. So, I didn’t expect Greg Schiffman, Dendreon’s (DNDN) chief financial officer, to reveal much in the way of new information when he made a presentation at the Lazard Healthcare Conference on November 15, 2011.

    “Importantly, however, near the end of his presentation, Schiffman mentioned that in 2013, the company will initiate a trial with Provenge in earlier stage prostate cancer, a trial for men who are hormonally sensitive (not castrate resistant, but where the disease has metastasized). In particular, he stressed that the placebo arm in this trial would be a true placebo … that is, one that did not include frozen Provenge, or Frovenge, as it is known. As such, he said, the company anticipated a much higher overall survival, or OS, advantage than what had been observed in the pivotal Provenge trials.”

    I think we all want to see the results from this trial, all be their release several years into the future.

  17. Dr. Moolten,

    You have raised a very important point in your comment regarding the Phase 3 trial design for Provenge. To wit:

    “While Huber et al do not prove a suppressive effect harmed the placebo arm patients in the IMPACT study, they raise troubling doubts. At the very least their paper clearly shows that the Dendreon data are tainted by a poorly constructed placebo arm. Better might have been to not remove cells in placebo patients, but to run a sham apheresis and infuse a bag of saline (no cells). However, this approach would have prevented possible crossover of placebo patients to an active arm later on which utilized the frozen cells.”

    It is an unfortunate fact, perhaps, that the protocol for the Phase 3 trail, which was developed in close coordination with the FDA (the agency, or course, had final approval), specifically included a provision to allow patients to cross over and take a frozen form of Provenge (Frovenge). This was done to encourage patients to enroll in the trial. It gave them additional hope that they would have a chance to receive a form of the experimental treatment and derive whatever benefit from it that was possible. I believe, based on the latest data I have, that some 60% of patients in the trial crossed over and received Frovenge. As was seen in two post-approval analyses that I discussed recently on Seeking Alpha (, the benefit from Frovenge was substantial. In fact, the use of Frovenge was a case where the trial design almost failed the treatment. Put another way, Frovenge was so good that patients crossing over achieved a median life-extension benefit that was almost as good as that provided by Provenge itself.

    One of the major problems the FDA and the pharma industry must come to grips with today is how to handle the crossover problem in trials such as the one conducted for Provenge. We are seeing a similar situation unfold in the case of JNJ’s Zytiga for PCa in the pre-chemo space. As I’m sure you are aware, several weeks ago, the safety board recommended that the Phase 3 trial be stopped and patients in the placebo cohort be allowed to cross over and take Zytiga (which is co-administered with prednisone). The problem that JNJ now faces is this. While their Phase 3 trial was stat sig on ONE of two primary endpoints–progression free survival (PFS)–the study failed to achieve stat sig on the second primary endpoint: overall survival (OS). Now, with patients crossing over from the placebo arm and taking Zytiga, the chances of JNJ being able to demonstrate stat sig for OS are being reduced by the day, jeopardizing approval of the drug in the pre-chemo indication. (Zytiga already is approved for post-chemo applications.)

  18. You obviously have no clinical experience in treating prostate cancer. There are 3o,ooo deaths per year frm prostate cancer. To say that a high death rate in the first year is unacceptable is ridiculous. In the dndn trials,the patients had advanced hormone refractory disease with established mets. The median survivsl of these patients is about 2 years. Many deaths occur early and others later. The survival curves are typical for trials for prostate cancer and in keeping with what we see in other trials for this disease.

    Furthermore, frovenge contained monocytes, most of which were non viable because they were frozen. To claim these monocytes were vital to the immune system is naive. The body has a tremendous capacity to regenerate monocytes. The reason the people who didnt receive frovenge died earlier is because they were too sick to receive it as a result of their disease not because of immunodepletion. Any oncologic trial albeit taxotere, zytiga etc has a group of patients who rapidly progress and never go on to other therapies.

    The effects of leukaphereses on the immune system are minimal if at all. Thats why we don’t and never have used it as a means to control with any degree of success autoimmune diseases or organ transplant. People who suffer major trauma or any surgery with major blood loss also loose most or all of their wbc with no measureable consequences. Furthermore we dont see any excess of immune or major infectious complications in the placebo groups in the dndn trials.

    Its intersting to note that they types of cancers that develop from immune supression for transplant patients include diseases like lymphomas not prostate cancer. We dont see an excess of deaths from prostate cancer in that population even though many people are walking around with it and dont even know it. Autopsy series show undiagnosed prostate cancer can be as high as 30% in 50 year olds and 80% in those over 80 y/o. So why dont we see excess prostate cancer deaths in people on chronic immunosupression like rheumatoid arthritis, crohs disease, ulcerative colitis, organ transplants etc. What makes leukaphereses so dangerous? The whole basis for the leukapheresis hypotheses comes from a subset analysis that contained an imbalance of placebo patients versus control. I believe the ridiculousnes of this theory will be put to the test once again when provenge is reviewed by the european union.

    This argument could go on forever because the only way to convince the critics for certain would be to conduct a trial of provenge versus a true placebo, which will never be done. Even in that case, there will be subsets that show no benefit as there are with all trials when one looks at subsets.
    One more point/ If you read Huber’s on line blog you will see that the table she refers to with the 65y/o subset also contains an afro american subset. That subset had a 46 mont survival benefit from provenge which was tremendous. What is the explanation for that? The simple answer is the same as that for the 65 y/o group, a subset analysis that is not balanced.

    I am not even sure the huber paper was peer reviewed since it is a commentary not scientific paper.

  19. @Danhen

    You seem to take it as a particle of faith that the leukapheresis procedures deplete certain important populations of specialized leukocytes. Please explain what you mean by “something related to the natural course of the disease” when it is mathematically clear that, on a probability basis, each population of specialized mononuclear leukocytes loses about 2% at each leukapheresis and the elementary fact that the human body, even for old men, would replace that 2% within a few days at most. Unless you can prove the basis of your question, any hypothesis you make based on it is not worth discussing. This is the way of science, not blind faith.

    I also don’t understand the constant effort by people like you to question other people’s motives instead of just carrying on an objective discussion based on data. On this forum and others, I’ve tried, from time to time, to explain the proper use of statistics and its meanings in interpreting the summary data often presented for clinical trials. I do not pretend to be an expert in immunology, I simply use known and published facts in my reasoning.

    However, proper use of statistics is something that seemed lacking from Ms. Huber herself along with people who believe her thesis. As an example, her calculations based on medians with “wide confidence intervals” to come to a statement that states “specifically” that 20 billions cells were taken from the leukapheresis is simply wrong! Medians and averages are notoriously fickle and can change fast when a few data points are removed or added, that’s why there is a need to also state what the confidence interval is for such numbers. With such basic lacking, what confidence can we place in Ms. Huber calculations and, by derivation, her hypotheses? Nearly zero.

    Instead of asking people to rejoin medical schools, perhaps it is better that you ask people who poke into clinical data to learn or relearn some basic facts about statistics and how it applies to a clinical trial first before making grandiose statements about them.

  20. To the above two: address the concern 1, do not ignore something related to the natural course of the disease. The high death rates in the first year and the separation of the survival curves during the same period in IMPACT are scientifically and clinically unacceptable. How often do you see patients with asymptomatic early MCRPC decease within one year after metastases occur? Do not pretend to be medical professionals to misrepresent the field. Those you referred as leading experts may not be truly evidence-driven with regard to the phony therapy. They are far below Dr. Rosenberg’s leading role and his academic reputation in immunotherapy. Read some his work about how real immunotherapy targets and kills tumors with evidence of killing (CR
    or PR). Ongoing trials and emerging evidence will help many understand better about the nature of the disease setting. when you talked conflicts of interest others had, do not forget your financial one. Yes, I am naive from patient’s perspective. btw, medical schools are still open for new applicants.

  21. @Danhen
    “about 2/3 of the patients in the control arm crossed over to receive “Frovenge” and those patients had a median survival similar to that of patients received Provenge, suggesting that giving back the mononuclear cells was very important to patients.”

    That may sound right but it is not supported by data. The FDA Clinical Review of the Provenge BLA presented an analysis of the cross-over patients. The median time from randomization to take Frovenge was 5.7 months (range 2.2 months to 31.3 months). Even for older sick patients, the length of time involved before a Frovenge treatment would say that whatever was lost in the leukapheresis procedure should have been replaced by the body. Otherwise, patients should have contracted whatever sickness due to any population of cells sufficiently depleted and died of some non-prostate-cancer causes. On this, first the math of random selection says that such a depleted population of mononuclear cells is unlikely to exist due to the small amounts of cells taken in each leukapheresis procedure relative to the total number of cells (less than 2%). Second, the FDA performed a sensitivity analysis of the data restricted to patients known to die from prostate cancer and the result was statistically significant.

    @david moolten
    “the lack of observable tumor response (however one might want to rationalize it) and the very modest clinical response (if real)”

    First, ask any statistician and she will say that the phrase “lack of observable tumor response” should be qualified with “as being measured in the IMPACT trial”. Failure to be detected in an experiment does not equate to effect not existing.

    Read the FDA Clinical Review carefully and you may see that they provided enough sensitivity analyses on the survival benefit to be convincing. Then, the value of progression as a predictor of survival becomes questionable in this setting. That is, the qualification of “being measured in the IMPACT trial” as applied to disease progression means that tumor response as defined by RECIST might be inappropriate to use for measuring the goodness of an immunotherapy. The disease might stabilize at some point later than when a RECIST criterion was seen. This has been discussed amply elsewhere with a FDA committee recommending to take more scans and allow longer time in testing tumor response.

    About “modest clinical benefit”, nobody seems to ask questions about Taxotere 2 months median difference, yet people seem to somehow get all riled up about Provenge 4 months median difference. Also how many cancer treatments are available today and how many show a better than 4 months median survival benefit?

  22. I’m curious, Dr. Molton. You make the following statement: “If it turns out that Provenge is a bust and that the process by which it came to licensure and clinical use was fatally flawed by bias, then other more successful cell therapies may be delayed or prevented from becoming available to the many patients who desperately need them.” How does this meld with the latest statement from the corporation, to wit:

    “The FDA approval was based on a significant improvement in overall survival shown in three well-designed, randomized, double-blind, controlled clinical trials – including the pivotal IMPACT trial – with remarkably consistent results shown across the trials as well as in numerous patient subgroups.

    “The Provenge clinical data have undergone a rigorous scientific review process that included [an] FDA review process, a Center for Medicare and Medicaid Services national coverage determination, a Technology Assessment and the peer review of multiple publications. As evidenced by the FDA approval, CMS national coverage decision, and NCCN treatment guidelines, the PROVENGE data clearly demonstrated a significant patient survival benefit and support its use as a clinically meaningful treatment option for certain types of advanced prostate cancer.”

    So, with all due respect, just what is the seminal event that you are expecting to occur that will lead to its failure?

    Any by the way, given that you raised the subject, if you would like to read about bias within the FDA regarding the approval process of drugs in general and Provenge in particular, I highly recommend the recent book by Mark Mitchell, ‘The Dendreon Effect: How Felons, Con-Men and Wall Street Insiders Manipulate High-Tech Stocks. The book may be purchased from Amazon:

    Read how two doctors with significant conflicts of interest sat on the Provenge Advisory Committee at the behest of ODAC…how one of those doctors was, at the time of the AC, the co-lead on the Phase 3 trial for a competing drug for PCa…how that same doctor was an officer and member, BOD, of an investment firm that owned 8% of the company that owned the competing drug…how the founder of the investment firm was on the BOD of the biotech company that owned the PCa drug under development (are you getting the picture?)…how those two doctors mentioned above(and a third) wrote letters to FDA Commissioner von Eschenbach arguing for the non-approval of Provenge in the days following the AC (an AC at which, by the way, the committee members overwhelmingly RECOMMENDED approval)…how the three letters to Dr. von Eschenbach were leaked to and published by The Cancer Letter either just before or immediately after the letters were delivered to the FDA in an unprecented attempt to sway the agency…how three weeks after the FDA turned down Dendreon’s BLA in early May 2007, Shearing Plough signed a $440M contract to co-develop the very same drug for which the doctor on the AC was serving as co-lead…and on, and on, and on. Please read this book and then, let’s have a discussion about some real ‘selfish interests’ in this matter.

    And Danhen…please…DNDN has 147.20M shares outstanding, of which two-thirds are held by institutions and mutual funds. Are you really so naive as to think I can affect the share price by means of what I post on a site such as this?

  23. The above comment from David is scientifically reasonable and objective. We shall focus on whether data consistency and reliabilty are present in the trial leading to that approval. Two key problems that bother lots of Oncologists are: 1) patients who only received the true “placebo” had a median survival of 12 months, which is too short to be acceptable for patients with asymptomatic early MCRPC without visceral disease. How come? The JNCI paper shed light on that. 2) about 2/3 of the patients in the control arm crossed over to receive “Frovenge” and those patients had a median survival similar to that of patients received Provenge, suggesting that giving back the mononuclear cells was very important to patients. Do we have another example that 2/3 of patients’ crossing over in oncology trials still hold up a survival difference? The Provenge trial may be the first one, but the difference most likely was not related to the antigen employed, rather to the withholding of the large amount of isolated mononuclear cells that are important for immunity. That withholding was consistently used in all Provenge trials, explaining what the investigators and company claimed for having “consistent” results. Yes, surpressing patients in the control arms consistently will give consistent results.

    What matters is patients’ interest. Those with financial or academic conflict of interest put theirs above patients’ interest. This is the core of the scientific war in immunotherapy.

  24. The idea that removal of white blood cells via apheresis might lead to detrimental effects in blood donors, including immunosuppression, goes back at more than twenty years. Conversely, early efforts to use apheresis in the treatment of autoimmune disease included deliberate removal of lymphocytes in an effort to salubriously exploit suppressive effects. Blood collections (e.g. platelet collection via apheresis) in healthy donors have been limited by both industry practice and federal regulations partly to avoid such possible effects.

    The possible disadvantageous impact of leukapheresis (e.g. during the collection of blood progenitor cells for transplant) on immune function has been reported in various patient subsets, including children.

    There is admittedly uncertainty about the exact nature of such effects, i.e. who is affected and how much, is what is safe for allogeneic harvest in healthy donors safe for autologous collection in patients and vice versa? Still,rather than being a new and crazy theory, immunomodulatory effects secondary to white cell depletion are well known, if not well understood. What is surprising is that the placebo concept went through multiple IRB reviews beforehand and the potential confounding effect was not more carefully considered.

    Apheresis obeys first order kinetics and ordinarily removes ~65% of circulating cells with one volume processed, 85% with two volumes. There is a rebound movement of cells from the extravascular space into the intravascular space which can blunt the effect on cell count.

    So a very substantial effect on the levels of circulating lymphocytes and other immune effector cells does occur with a single leukapheresis procedure. Dendreon patients receive three such procedures during treatment, one every two weeks.

    While Huber et al do not prove a suppressive effect harmed the placebo arm patients in the IMPACT study, they raise troubling doubts. At the very least their paper clearly shows that the Dendreon data are tainted by a poorly constructed placebo arm. Better might have been to not remove cells in placebo patients, but to run a sham apheresis and infuse a bag of saline (no cells). However, this approach would have prevented possible crossover of placebo patients to an active arm later on which utilized the frozen cells.

    What concerns me is that so many people are so passionately opposed to Huber’s report because of their own selfish interests.

    While it may turn out that she is wrong, her hypothesis is eminently reasonable based on what we know about about apheresis and impugns Dendreon’s study design. Given the very high cost of this treatment, the lack of observable tumor response (however one might want to rationalize it) and the very modest clinical response (if real), continued rigorous examination is in order.

    As a firm believer in the future of cell therapy, I believe such rigor is essential, especially given Provenge being a first of its kind in many ways. If it turns out that Provenge is a bust and that the process by which it came to licensure and clinical use was fatally flawed by bias, then other more successful cell therapies may be delayed or prevented from becoming available to the many patients who desperately need them.

  25. Cohen’s responses had lots of untruthful and distorted information, which would help him avoid further financial loss in DNdn. He missed some basic issues about the disease and science. More knowledge would help his understandings. As a responsible oncologist, I feel sad that the society has so many who are driven by making profit from harming other or patients.

  26. Huber’s article was already debunked by leading experts calling it a work of fiction and junk science. I could care less about Huber’s character but when her anti Provenge crusade almost kept me from obtaining the treatment it becomes personal. With all of the drama in the Provenge saga and all of the fluxuations on the street with dndn, to have a stock analyst write such an article makes a guy say hmmm. If it smells like poo. It is usually poo. Just an opinion folks. Todd

  27. Dear Critical, not cynical…

    Other than what I posted in my original letter to the editor, and what Ocyan posted, what more do you want to see? How many different ways would you like us to refute what is the equivalent, basically, to the Flat Earth Theory.

    And for someone to step forward like Mr. Seals, who was almost harmed by Ms. Huber’s actions…well, I think that should give you pause as to what is happening here.

    Please…go back and read, if necessary, Dr. Gulley’s comments at the CMS hearings. He is one of the most respected immunologists in the US, if not the world.

    What I would like to see is for Ms. Huber’s co-authors on her JNCI paper to step forward and be heard. Where are Laura Haynes, Chris Parker, and Peter Iversen in all this? Their names are on the JNCI paper as well. Do they still hold the same professional opinions they had when they signed the release of the document to JNCI (which, again, has no relationship to the US National Cancer Institute)?

    J Natl Cancer Inst. 2012 February 22; 104(4): 273–279.
    Published online 2012 January 9. doi: 10.1093/jnci/djr514 PMCID: PMC3283534

    Copyright © The Author(s) 2012. Published by Oxford University Press.
    Interdisciplinary Critique of Sipuleucel-T as Immunotherapy in Castration-Resistant Prostate Cancer
    Marie L. Huber, Laura Haynes, Chris Parker, and Peter Iversen
    Affiliations of authors: Trudeau Institute, Saranac Lake, NY (LH); Academic Urology Unit, The Royal Marsden NHS Foundation Trust, UK (CP); Department of Urology, University of Copenhagen, Rigshospitalet, Copenhagen, Denmark (PI)

    T. Cohen

  28. I have no personal interest in whether Provenge succeeds or fails, but what strikes me about the 10 comments to this post is that only one (ocyan) attempts to refute Huber’s analysis with data. The other 9 resort to character assassination and conspiracy theories. Come on, folks, this is medical science, not Fox News.

  29. Kaiser innitially denied me because of hubers article. Thanks to some helo from those in the know I was able to show prove that the article was crap. Theor next reason for denial was due to lung metastasis. I showed then that C.N.S. guidelines and Dendreon guidelines showed that Provenge was indicated for me. They then denied Provenge citing N.C.C.N. guidelines. My last apoeal denial was overturned by the state.

  30. Todd, first of all let me congratulate you on your perseverance on getting Kaiser to change their stance on provenge and i am glad that now you are going to get provenge. This is a perfect example of what my previous comment meant. That this fallacy is been propagated by the media and is affecting many desperate men like yourself who must fight for their lives not only against this dreadful disease, but also to get the insurance companies approval and the medical community to accept this paradigm shifting technology amidst these lies and innuendos been bandied about in the press….again kudos to you Todd, and my God bless you.

  31. I view Huber’s commentary from a differrent perspective as I am 48 years old and have battled stage 4 prostate cancer for the last 6 years. I recently became hormone refractory and I am appalled that the misinformation of aa stock analyst has been given so much traction. My Kaiser Permanente Oncologist cited Huber’s junk science for reason to deny me the usr of Provenge. If I was the type of person to do what I am told like a good little submissive patient I may not have recieved the drug. Thanks to a strong will, a loud voice, and the help of some good people, Kaisers decision not to provide Provenge was overturned Friday last. Huber’s article is a caustic work of untruth that could ultimately cause the death of men like myself. Todd

  32. I am a 48 year old man who has been battling stage 4 prostate cancer for the last 6 years. I recently became hormone refractory. Speaking as one who has ran the guantlet trying to obtain Provenge, I am appalled that her commentary has been given so much attention. My Kaiser Permanente Oncologist actually ciited this fraud in his decision not to provide me the medication. After six weeks of research on my part and the help of some good people ( wink, countless hours on the phone, and three appeals, friday last I was notified that the deecion to deny Provenge was overturned by the State review board. Huber’s article is a caistic untrutj that will cost men their lives. Todd

  33. This article provides a lot of information that shows there is, sadly, a very real possibility that Ms. Huber’s article was written only with the intention to stir up controversy.

    Provenge works, and probably better than any other prostate cancer treatment for overall survival given the likely 7.8 month median survival benefit referred to in Dr. Cohen’s article.

    This is not the first time Ms. Huber has done this. However, her claims have been strongly refuted by actual experts at the FDA, CMS, and by other medical specialists, all with far better education and experience than she. This raises the question as to why she persists?

    Unfortunately, a very real possibility is that she has an agenda to frustrate the advance of Provenge usage. If she succeeds in persuading even one patient to not get treatment who is otherwise eligible, then she will have possibly shortened that patient’s life. This should not stand.

  34. As a hiring manager that has read thousands of resumes, the Huber resume paints neither the picture of a trained “scientist” or seasoned “analyst” on wallstreet. Individuals who lead, manage, execute, and are responsible for “all” things she claimes do not gain such authority or experience by job-hopping every 6-9 months. These are the classic signs of an incompetent employee and typically someone that is asked to leave rather than be fired. How exactly did she become a “Managing Partner” at Forstmann during her 1-year tenure there when she had absolutely zero previous experience as a wallstreet analyst? Why did she leave so quickly? Did she possibly have a sexual relationship with Tony Forstmann that went sour? Too many questions about her lack of credentials. She has never held a real job as a scientific researcher or been involved in any clincial research. Research and criticism should be conducted in the lab by experts, not by untrained business majors with failed careers on wallstreet.

  35. The path to new therapies historically has shown how treacherous it can be, especially where there are entrenched big pharma players unwilling to yield to paradigm shifting technologies, and Provenge being one that can bring the best in people but unfortunately also the worst. To wit Ms Huber’s crusade against it, highlighted prominently in the press recently with half truths and innuendos about it’s efficacy. How sad it is that the press is propagating this fallacy and in turn affect the treatment option of desperate men seeking to extend their lives, how sad indeed.

  36. A scientific theory can be examined by what it tries to explain and what predictions it makes. Here are some observations on Ms. Huber’s Immunodepletion theory:

    1. The theory was developed to explain a seemingly anomalous subgroup analysis in the FDA Clinical Review of the Provenge data regarding the subgroups =65. In fact, from a statistical point of view, it is normal that when you consider many subgroups, some are bound to go in the opposite direction of the main result (ie, Provenge works). This is because many subgroups may be imbalanced in favor of the wrong arm. A good indicator of this in the <65 subgroup of the IMPACT trial was that there were 77 and 49 on the Provenge arm and the placebo arm respectively. The trial did a 2 to 1 randomization so you would expect the number of Provenge patients to roughly double the number of placebo patients. Yet, this was not the case: 77 is not anywhere close to doubling 49. What other imbalances may be hidden in this subgroup? That is, the result may look intriguing but it should not be taken seriously until someone delve into the raw data to see if there is any other supporting evidence. The FDA, in fact, did that and dismissed the analysis as statistical noise. Ms. Huber conveniently ignored all that and developed an entire new immunological theory based on some hokie calculations to attack the efficacy of Provenge.

    2. Ms. Huber used various median numbers concerning the leukapheresis procedure to assert that more than 90% of circulating white blood cells were taken from patients and that was detrimental to patients. First, from a mathematical perspective, calculations based on medians with wide confidence intervals are subjected to wide errors. So this more than 90% number was likely wrong to start with. Then, it should be noted that at any given time, only less than 2% of immune cells circulate in the blood. So even 90% of that only adds up to less than 2% overall. The human body regenerates those cells in less than a day.

    3. If the Immunodepletion theory was right, immediately after the leukapheresis procedure, the placebo patients should show a higher rate of adverse events than the Provenge patients. The safety database for all randomized trials of Provenge (D9901, D9902a, P-11, IMPACT) did not show any evidence of that. In fact, Provenge patients showed more incidences of fevers and chills typical of an immunotherapy while placebo patients were far less affected by their procedure.

    4. If the Immunodepletion theory was right, the survival curves for the Provenge and placebo groups should diverge quickly. But they did not, they ran parallel for several months before separating, consistent with the theory that it takes months for the immune system to ramp up with a Provenge treatment.

    5. Many placebo patients crossed over after they experienced disease progression (months after randomization) to take a weak form of Provenge made from their salvaged frozen blood (the blood that the Immunodepeltion theory said to be behind immune cell depletion because of it was withheld from patients). If the theory were right that Provenge was no better than placebo and might even be harmful, then those cross-over patients should not do any better than those who did not cross over because the time lapse means that any depleted cells would have been amply replaced by then. Yet, recent analysis showed that the median survival time for the cross-over patients was 23.8 months compared to just 11.6 months for those who did not cross over. Why this difference if Provenge did not work?

    So, the basis of Ms. Huber's questioning of Provenge efficacy based on an anomalous statistical data point could easily be explained. Then, simple predictions of the theory that she developed were contradicted by clinical data. Lastly, her theory did not make sense given some basic facts we know of the immune system.

    The conclusion is that the Immunodepletion hypothesis as Ms Huber defined is simply an imaginary fancy. It's unfortunate that the JNCI chose to publish such an ill-considered paper. I hope they will realize their mistake and recall it.

  37. I would point out that Ms. Huber’s education level seems to be at odds with her assertions of being a “trained scientist”. Her resume (link: ) seems a bit lite in the area of scientific research. Be aware her first school listed (with her grades) is a girl’s school in the UK. It appears to go to a high school equivalent level. The last real job for the firm P. Schoenfeld Asset Management, would have been at a time when Dendreon share price did very well. Maybe, one should search the SEC Edgar system looking for some relationship between Schoenfeld Asset Mngnt and Dendreon to find some history that might shed some light. Finally, I would point out that the FDA did approve Dendreon and was totally aware of the overall process for Provenge.

Post a Comment

You must be logged in to post a comment