FDA publishes analysis of decision not to approve 5-ARIs for prostate cancer prevention

The “New” Prostate Cancer InfoLink is delighted to see Dr. Pazdur and colleagues from the U.S. Food & Drug Administration (FDA) present their reasoning for non-approval of the 5α-reductase inhibitors (5-ARIs) dutasteride and finasteride for prevention of prostate cancer in a new article in the New England Journal of Medicine.

The full text of this article by Theoret et al. has just been published on line.

The authors present previously unpublished data suggesting that all of the preventive impact of the 5-ARIs is confined to men with Gleason grades of 6 or lower at the time of diagnosis who were already at minimal risk for clinically significant prostate cancer while simultaneously increasing risk for the occurrence prostate cancer of Gleason grade ≥ 8. The authors also report that their analysis of available data does not support either of the two primary theories about why the occurrence of the small increase in risk for higher grade cancers may be an artifact of detection bias.

Whatever one’s personal views about this decision, and of the thinking behind this decision, the public disclosure of the rationale for the decision in such a highly visible journal meets a critical need for transparency of the FDA’s decision processes which we should all applaud and encourage.

11 Responses

  1. I watched the whole live webcast of the hearing last December 1. While I thought the advisory committee’s negative vote was defensible, I was frustrated with what came across as a heavy-handed dismissal of analyses showing that finasteride was very unlikely to promote high grade cancer, and I was amazed at Dr. Patrick Walsh’s faulty understanding of the comparative importance of DHT and testosterone in fueling prostate cancer. I will be reading the full report with interest to see if the heavy handed dismissal, with scant justification, and Dr. Walsh’s erroneous view come through in the NEJM report.

    Thanks so much for posting this, and thanks to for the NEJM for publishing it online for all.

  2. Jim:

    Dr. Walsh’s opinions are not referred to at all in this commentary about why the FDA made the decision it did. You need to remember that while strong opinions (from very different points of view) may well be presented at the public meetings of FDA advisory committees, the FDA tries very hard to make its decisions based only on an objective analysis of the available data.

  3. So what is a man taking 5-ARIs for BPH to do? I have read in different reports that about 3 million men have taken these drugs for BPH or hair loss. And reports that there is a 1% to 3% chance of the drugs causing/leading to high grade cancer imply that there should be 30,000 to 90,000 cases of high grade cancer among these men. Seems like this many cases should have been noticed by the medical community if all the reporting is correct, and the drugs banned for use for BPH. (Obviously, a frustrated layman’s laments.)

  4. Dear “Frustrated Layman” John:

    It is exactly for the reason that you describe that many in the prostate cancer community have a problem with the FDA’s analysis of the situation. If 5-ARIs really do present a risk for high-risk prostate cancer, then at some point after the introduction of finasteride for the treatment of BPH (in 1992), there should have been a massive “bump” in the numbers of men who, after starting treatment with finasteride for a diagnosis of BPH, were diagnosed with high-grade prostate cancer. No such bump exists (at least that I am aware of). Indeed, I have never even seen data to suggest such a possibility, and yet it should be easily identifiable in an analysis of Medicare data.

    My entirely personal opinion is that men being treated with 5-ARIs for BPH are at no increase in risk for prostate cancer on the basis of the available data … and we have long argued that the apparent increase in risk for high grade prostate cancer in the PCPT and REDUCE trials is an artefact of the trial design … but we cannot substantiate these beliefs and we sympathize with your question.

  5. In order for the FDA to come to the conclusion that they did, they not only had to “suspend disbelief” in regard to research design flaws that make the results appear to indicate a (mere) correlation between the drugs and high-grade prostate cancer, but they had to wholely ignore and disregard the damaging and life-altering reality of over-treatment and undue and unnecessary “morbidities” (urinary and fecal incontinence and impotence) of all current standard treatments of prostate cancer — whether low or high risk. The term “morbidity” was used a single time in the defensive justification (with no discussion) published in the NEJM.

    The idea that it might be ethically irresponsible to knowingly cause hundreds of thousands of men to be diagnosed with clinically insignificant cancers that they didn’t have to develop and suffer all the consequences of treatment that may be necessary for high-grade prostate cancer patients to endure, but that provide no value to the men with indolent cancers, was apparently never considered and, subsequently, was never addressed.

    This conclusion just slays me:

    “Therefore, the trade-off inherent in using a 5α-reductase inhibitor for prostate-cancer prevention is the acceptance of one (likely a very, very high overestimate) additional high-grade cancer in order to avert three to four (likely a very, very low underestimate) potentially clinically relevant lower-grade cancers.”

    Even if you choose to buy into the FDAs rationalization for denying 3 in 4 (75%) or 4 in 5 (80%) prostate cancer patients a chance at avoiding over-diagnosis and over-treatment, you must admit that simply ignoring all the consequences of the current standards of treatment is not — and can never be — good science or the ethical practice of medicine.

    I’m ashamed of the members of that FDA panel.

  6. Hi Sitemaster and all,

    I have now studied the NEJM report. I too would like to believe the FDA committees are perfectly objective and wise, but I doubt that. I’m not faulting the process or the committees. Within the limits of practicality and realistic time constraints, I believe they do the best they can. However, I’ve attended a number of these hearings, and human factors are clearly in play. The votes often come at the end of long days of intensive presentations, and everyone in the room is clearly eager to come to a conclusion and get on with their lives. I believe mental fatigue is at times a factor; I know it would be for me. I too believe that they do try very hard to make the decisions objectively, but I’m convinced that sometimes they do fall short.

    I was glad to see that Dr. Walsh’s erroneous view — that testosterone is a more important androgen than DHT for prostate cancer — did at least not make it into the FDA report. However, I’m still convinced that his view helped sway some of the advisory committee members.

    Now for the report itself. My bottom line is that I am very disappointed with the FDA, especially with the new labeling requirement. One sentence in the paper is highly obnoxious from the viewpoint of the actual evidence: “These results suggest that one additional man would receive a diagnosis of high-grade prostate cancer (modified Gleason score, 8 to 10) for every 150 to 200 men treated long-term with a 5 alpha-reductase inhibitor.” That assumes the FDA has a persuasive stand that the 5-ARI drugs do promote high-grade disease, and they do not, as I see it — not even close! At least they indicate that the risk, if they are correct, is quite low. However, I am convinced that not only have they not proved or supported their position adequately, but that the reality in time will prove that the risk of high-grade disease for men taking 5-ARI drugs is actually lower, which is suggested by some of the evidence. Their key statement actually could be read in a positive way — that the drugs make it more likely to detect the disease, but that is clearly neither the intent nor the understanding in the media.

    Here’s a key point: re-emphasizing the point about the number of men treated to produce one case of high-grade disease. Even if the FDA is correct in its belief that the 5-ARI drugs may promote high-grade disease, the increased risk is very small, as indicated by the research: 0.5% for dutasteride, and 0.7% for finasteride.

    While the trial researchers advanced evidence that the drugs actually cut the risk of high-grade disease by 27% in one analysis (using “weighted imputation” applied to patients in the trial who had had prostatectomies — the lower half of the second table in the paper, fourth column, RR 0.73), the FDA took a conservative but defensible second look by defining high-risk not as GS 7-10, as the prevention advocates had done, but as GS 8-10. I don’t have a problem with that approach as a reasonable way of probing the data. However, the FDA leaned heavily on its result, which showed an increased risk of higher grade disease in the 5-ARI groups than for the placebo groups. The problem is this: while the confidence intervals for the reduced risk figures were fairly tight, meaning that the true result (from a heavenly viewpoint, which would include the ideal all inclusive population and not just a sample) was quite likely close to the average reported, the confidence intervals for the FDA GS 8-10 analysis were wide, with the ranges between 0.71 to 2.76. The 0.71 value would be right in line with those believing the 5-ARIs actually lowered the risk of high-grade disease, but, unfortunately, the weighted averages were higher, leading the FDA to conclude that there could well be heightened risk. The root problem was that there were too few patients in the GS 8-10 groups, and that problem meant that a chance result was a lot more likely to occur. The FDA could and should have noted this but failed to do so. This is not advanced statistical science; the flaw is rather easy to spot for those who have some experience with statistics. As I studied by scrawled notes of the December 1 webcast, I hoped that the FDA did have a decent argument behind its heavy-handed rejection of the arguments of prevention proponents. How disappointing that the FDA did not. Shame on the FDA! I believe that ultimately these drugs will prove to slightly reduce high-grade disease as well as making it easier to spot. In the meantime, a great prevention opportunity for thousands of men at higher risk for prostate cancer has been lost.

    Fortunately, as the FDA notes, the drugs can still be prescribed off-label for prevention. However, with malpractice attorneys lapping at their heels, I believe that it will take some courage for doctors to prescribe the drugs for prevention, even if they understand the issues. Sad!

    Here is another key point: the FDA did not even mention the value of preventing lower-risk prostate cancer. Despite the widely acknowledged problem with over-treatment of low-risk disease, the FDA showed little interest at the hearing and none in the paper about the value of the 5-ARI drugs in substantially reducing the numbers of men who would need treatment for low-risk disease. This value was emphasized by medical oncologists at the hearing who were proponents of 5-ARI prevention. Again, shame on the FDA!

    I’m struggling to figure out why the FDA has such a readiness to accept arguments and evidence against 5-ARI prevention and such stubbornness in accepting evidence and arguments in favor.

  7. I need to add a PS — about the prevention advocates’ line of argument that 5-ARI drugs make high-grade disease easier to detect by shrinking the volume of the prostate, thereby enabling each biopsy core to represent a larger proportion of the total prostate.

    The FDA actually mentions this powerful line of thought, backed by analysis, in their paper, noting that a PSA greater than 4.0 was a more sensitive indicator of prostate cancer in the finasteride group than in the placebo group. However, they disregard this evidence and favor their own flawed argument of higher risk in the GS 8-10 group, which I addressed previously. Note that they do not come to grips with the reduced volume argument in their paper, instead just implying that their GS 8-10 analysis trumps it. Neither the FDA nor Dr. Walsh and other opponents of 5-ARI prevention came to grips with the volume argument at the hearing either. In view of the fact that the increase in high-grade detection is so small — less than 1%, it is just common sense that the reduced volume due to 5-ARIs likely accounts for it, or at least for a piece of it, which would make any true increase, if there is any, tiny indeed.

  8. Having experienced only 2 years in an off-period during intermittent androgen deprivation (IAD) off an LHRH agonist and antiandrogen and not “maintaining” with a 5-ARI to inhibit returning testosterone from coming in contact with 5-AR isoenzymes and converting to the more powerful stimulant to prostate cancer cell growth, dihydrotestosterone (DHT), then later experiencing 2 months short of 6 years of IAD while maintaining with dutasteride/Avodart, I will continue to be a proponent of the inclusion of a 5-ARI in androgen deprivation therapy and during IAD. There are just too many papers that support this use. And I continue to disagree with the reasoning of the advisory panel that 5-ARIs could be causal to high-grade prostate cancer. Even with this “panel report” there is no absolute indication that this was found to be so.

  9. Jim Waldenfels wrote, “I’m struggling to figure out why the FDA has such a readiness to accept arguments and evidence against 5-ARI prevention and such stubbornness in accepting evidence and arguments in favor.”

    That’s an easy question to answer … The FDA is under enormous political pressure to not approve drugs that have significant risk for adverse effects in otherwise healthy people.

    I should also point out that the article in the NEJM is not a reflection of any particular point of view presented at the advisory committee hearing or of the decisions taken at the advisory committee meeting on December 1, 2010. This article reflects only the FDA’s rationale for the decision it took not to approve any 5-ARI for the prevention of prostate cancer. The FDA takes what happens at advisory committee meetings into account in reaching its decisions, but it is in no way bound by the decisions of these committees.

  10. Thanks Sitemaster for your reply of June 17 at 6:11 AM. I too understand that the NEJM is providing a welcome service in permitting the FDA viewpoint to be publicized.

    However, I am troubled by what appears to be a discrepancy in the paper. If it is there, it should have been spotted by the FDA before submission of the paper and by the NEJM editors after submission.

    Of the cancers in the trials above Gleason score 6 — the Gleason score 7 to 10 range — most (75%) were Gleason 8 to 10 cancers, stated the FDA. Therefore, the FDA decided to look just at the Gleason 8 to 10 cancers, comparing the 5-ARI drug arms to the placebo arms. However, the second table in the paper shows that the preponderance of cancers were Gleason score 7 cancers (70%), not 75% the other way around! To be specific, the second table in the paper shows that there were 243 cases of cancer ranging from Gleason score 7 to 10 in the finasteride trial; of those, there were 73 cancers ranging from Gleason 8 to Gleason 10; that leaves 243 – 73 = 170 Gleason 7 cancers, which is 170/243 = 70%. I would really appreciate someone else looking at the paper and confirming that amazing discrepancy, as I’m concerned I may be so angry that I’m somehow not reading the information correctly.

  11. Dear Jim:

    What the FDA said was that “75% of the increase in tumors with modified Gleason scores of 7 to 10 observed in the finasteride group involved tumors with a score between 8 and 10.” [Bold italic type added for emphasis.]

    The FDA never suggested that 75% of all tumors of Gleason score 7 to 10 were Gleason 8 or higher.

    If you want to see the details, you need to go back and look at the FDA briefing document for the December 1 hearing.

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 )

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: