Skip to Content

Biogen’s Alzheimer’s Data

If you’re a rabid biotech investor, you already know all about Biogen’s data this morning. If you’re sane, or insane in some other, more interesting direction, then here’s what’s up: last December, the company released some Phase I data suggesting that their amyloid antibody, BIIB037 (aducanumab) was showing better-than-expected responses.
Some of you may be wondering about that. Clinical responses? In a Phase I study? Well, there’s nothing stopping you from collecting the data, and in this case, since Biogen was dose-responsing the subjects, they looked both at the amounts of amyloid (using Lilly’s imaging agent for this purpose) and they also gave the patients standard cognitive tests to see if something was happening. That’s what has everyone so excited, because there does appear to have been a dose-response, by both measurements. I’m glad to hear it – there is a huge need for an Alzheimer’s therapy, of course, and the failure rate has been so brutal in the clinic (arguably 100% – there are approved drugs, but they don’t do much).
Here’s more on this from Matthew Herper at Forbes. The thing is, although this is good news, many investors are taking it as incredible, world-changing news. Biogen stock has been gaining over the last few months on expectations, and today’s data will probably keep the party going. That’s an awful lot to pin on a bunch of Phase Ib data, I’d say. Especially since this is Alzheimer’s. The only phase that matters in Alzheimer’s is Phase III.
Biogen knows this – they’re skipping right ahead to it, and good for them for taking the risk. But a risk it is. Every other Alzheimer’s antibody trial has failed, even though some have tried to pretend otherwise. The data are not in yet on the more preventative trials, where you hit the earliest possible patient population, true. But the results have not been encouraging. Either Biogen is really on to something here (which is what the investors are betting on) or they’re in for a crashing disappointment in a few years. Yep, a few years – that’s how long it’s likely to take to run a large, well-controlled Phase III in this area.
There are other things to think about. One big issue is that the sample sizes are very, very small. Adam Feuerstein (on Twitter) noticed that the cognitive decline in Biogen’s placebo group was larger than one might have expected. And there does seem to be some brain swelling in patients who are APOE4 carriers. All of these things are going to have to be worked out, slowly and expensively, so if you see headlines that Biogen has cured Alzheimer’s, don’t believe them – yet. I hope they have, but no one knows that, or can know that, for some time to come.
As a side issue, the excitement in the company’s stock is part of the general biotech stock run of the last two or three years. It has been a tremendous market, that’s for sure – buying the biotech index in 2010 would have been pretty slick move, although you’d have also lost some hair by now wondering how long the party would continue. My guess is that the people who did buy it back then have already sold and regretted it, and maybe more than once. As someone who works in the industry, I’ve certainly benefited from the runup, so by warning about investor euphoria I’m actually working against my own short-term interests!

50 comments on “Biogen’s Alzheimer’s Data”

  1. MoBio says:

    This is certainly intriguing and one wants to be hopeful — but how many have seen such ‘hints of activity’ in Phase I/II studies only to see that it was noise in Phase III?

  2. johnnyboy says:

    Does anyone know what, if any, is the difference between BIIB’s amyloid antibody and all the other failed amyloid antibodies that came before ?

  3. Edward Taussig says:

    It’s the totality of the data that’s impressive, amyloid clearing, dose-response for both amyloid clearing and cognition, very statistically significant results.
    But is was P1 so too soon to bet the farm, but clearly justified in moving to p3

  4. Anonymous says:

    It’s incredible how humans can become so irrational in their decision making by ignoring probabilities when the potential upside becomes big enough. Like playing the lottery.

  5. Hap says:

    We don’t know the overall probabilities of what could work and what can’t – we only know that what’s been done hasn’t worked at all. It’s possible they found something. (Big lotteries “work” because you can’t ever spend enough to fill your chance of winning (1e-9 chance of winning 1e8 means you’d have to spend 1e9 to be sure of winning, and no one has the lifetime for it – thus you either win before you’ve come closing to spending enough to be sure of winning or (mostly) you lose completely.) In the latter, more probable case, people don’t generally lose enough to care. People are paying for hope, essentially.
    Of course, it’d be better to find out in P2, if you could design a trial in P2 to give you a reasonable assessment of your chances of success. P3 is an expensive place to lose.

  6. Farmhand says:

    @2
    According to the Forbes article, this antibody was derived from people who lived into old age without getting AD. Can’t tell you structurally how it differs, ie what epitope or anything. We should maintain a healthy skepticism since the treatment group size is small and there was significant edema at the high dose, but this sure does look encouraging for a change.

  7. Luigi says:

    According to a source at the meeting – the data were derived from 37 sites in the US. With 165 patients in the baseline group presented by Biogen, that makes 4.46 patients per site – either the 37 is wrong or the site count is cause for great concern.

  8. Luigi says:

    According to a source at the meeting – the data were derived from 37 sites in the US. With 165 patients in the baseline group presented by Biogen, that makes 4.46 patients per site – either the 37 is wrong or the site count is cause for great concern.

  9. Allchemistry says:

    Brain swelling suggests at least that the antibody is targetting to the right tissue.

  10. Anonymous says:

    @6 that seems to me like a clever way of looking for a basal protective antibody.

  11. anonymous says:

    Anyone (Luigi’s source) see what the mini–mental state examination (MMSE) baseline values were? Also, what’s the expected error for the test?
    Same questions for the Clinical Dementia Rating (CDR).
    The results for the latter are rather odd, aren’t they? Placebo group decreased by 2.04 points. But, CDR is only a 0 to 3 scale (0 = no dementia; 3 = severe dementia). So the placebo group (best case scenario) went from 0 to 2.4, i.e. from no dementia to moderate dementia, in one year.
    Really? But the same placebo patients worsened by 3.14 points on the MMSE — which is a 30 point scale. Really?

  12. Luigi says:

    Baseline MMSE 23.2 – 24.8.
    All trial end points reported as the deltas noted in the Herper article.

  13. gcc says:

    I’m very interested in the discovery of this antibody. Here’s a quote from the Forbes article:
    “Al Sandrock, Biogen’s senior vice president and chief medical officer, says that BIIB037 was created by looking for antibodies in patients who survived into old age without getting Alzheimer’s. These patients had anti-amyloid antibodies, and one of them was turned into a drug.”
    Does anybody know if any of this work on anti-amyloid antibodies in elderly people without Alzheimer’s disease has been published?

  14. anonymous says:

    In response to number 2’s comments, I believe it’s targeting the soluble abeta, more like sola and less like bapi. But it’s not quite clear.

  15. Anonymous says:

    “Of course, it’d be better to find out in P2, if you could design a trial in P2 to give you a reasonable assessment of your chances of success. P3 is an expensive place to lose.”
    Exactly. Going straight into Phase 3 is like slamming into a concrete wall and hoping you come out OK on the other side. It’s stupid. The whole point of doing the Phase 2 is to check whether Phase 3 is likely to succeed, but at much lower cost in case of failure, which is always more likely than not.

  16. gcc says:

    I’m very interested in the discovery of this antibody. Here’s a quote from the Forbes article:
    “Al Sandrock, Biogen’s senior vice president and chief medical officer, says that BIIB037 was created by looking for antibodies in patients who survived into old age without getting Alzheimer’s. These patients had anti-amyloid antibodies, and one of them was turned into a drug.”
    Does anybody know if any of this work on anti-amyloid antibodies in elderly people without Alzheimer’s disease has been published?

  17. anonymous says:

    @12 So MMSE went from ~24 to ~21 in the placebo group of 40 patients. That’s from borderline mild cognitive impairment to mild cognitive impairment (variously defined as 18-23 or 19-24). Not much of a worsening; is this even a significant given the very small group of patients? See Halsey et al. (2015) Nature Methods 12:179 on the fickleness of p-values with small sample sizes.
    But what about CDR? How can it change (what was its baseline??) from 2.5 to 0, or 3 to 0.5, in the same patients? That from nearly normal to severely demented.
    Something doesn’t add up…

  18. steve says:

    The antibody is not actually Biogen’s but instead came from a small biotech called Neurimmune http://www.neurimmune.com/products/alzheimers-disease.html It sees a conformational epitope on beta amyloid; it has a high affinity/avidity for insoluble fibrillar Aβ and a 100-fold decreased affinity for Aβ monomers

  19. Chrispy says:

    I just searched for aducanumab on Web of Science and EBSCO and found nothing. Does anyone know if this antibody was called something else early on or was its development entirely stealth?

  20. bio_kruncher says:

    @17 They are reporting CDR-SB. SB=sum of boxes. This score is sum of 6 items each on 0-3 scale so 18 is max worst score.

  21. Anonymous says:

    where’s Lane?

  22. Anonymous says:

    Busy copying and pasting..

  23. anonymous says:

    @20, bio_kruncher: CDR-SB baseline values?

  24. KevinH says:

    @15. Well sure, going full phase 3 is an expensive place to fail. But on the other hand, not going full phase 3 has an opportunity cost. Conducting a phase 2 trial means delaying the phase 3 trial and – more important – the eventual FDA (and other) approvals.
    Given the size of the potential market – and the absolutely incredible, phenomenal, unbelievable truckloads of money a successful drug would bring in – it may well be worthwhile, financially speaking, to risk failing in a premature phase 3 trial in exchange for even a small chance to be a year or two earlier to market.

  25. Equinox_15 says:

    Strategic design of phase III clinical trials based on limited data from phase I(b). If you guess right, achieve market nirvana in less time. Guess wrong and you will still be contributing more financial justification as to why cost of discovery, development and launch of a new drug is >$2 billion (and growing). Either way, its’s a win-win for the industry.

  26. watcher says:

    Is it ethical to jump from Phase 1 to Phase 3? What about extra safety that could come from a larger, longer Phase 2? Such shortcuts very often turn out to be huge mistakes, sometimes when the drug could be useful, yet missed information that could have helped in design and conduct of Phase 3 will contribute to a poorer study design. For me, this is not a good sign looking ahead…and on the stock, I’d short it if I was a risk investor.

  27. anon says:

    IV dosing millions of people?

  28. Anonymous says:

    Taking shortcuts is ALWAYS a mistake in drug development.

  29. Anonymous75 says:

    Is it possible that patients on active were functionally unblinded by the high rate of AEs?

  30. Eric says:

    #3 Edward – I don’t really think they have ‘very statistically significant results’. For both MMSE and CDR-sb p was less than 0.05 at 52 weeks and their was no effect 24 weeks. The amyloid effects are more robust (p less than 0.001) which most people would agree is significant.
    Also slightly worrisome is the fact that there was no dose dependency for MMSE and CDR-sb at the 24 week time point. This could simply mean it was too early to see effects, but it turns out that the cohort sizes changed at each time point. For example, in the placebo group n=37,36,and 23 at baseline, 24 wk, and 52 wk respectively. The numbers go down because subjects discontinued the study due to adverse events/disease progression/lost to follow up.
    Certainly it’s better to see positive effects than not, but I wouldn’t put much stock in the cognition data they reported.

  31. anon says:

    The results announced today were for interim results to 54 weeks in 166 patients (30 weeks for the 6 mg/kg arm).
    The trial long extension intends to dose to 113 weeks. The trial is still recruiting. It began in October 2012, so some of the patients might already have been dosed for over 2 years!
    By the time the phase 3 trial starts, Biogen will likely have a much clearer read out on the effectiveness.
    The risk of another negative AD trial would then be greatly reduced.

  32. annon too says:

    On the last comments about Biotech stocks rising….as an example, put $20,000 within an IRA into a Morningstar 5 star Biotech fund in Aug 2008, reinvested all distributions, added an additional $5000 early in 2014….and today it’s worth in excess of $90,000. Do I think it’s over…well maybe, if the run up depends Phase 1 studies such as this one. But, there are enough wins that appear, such as the new PCSK-9 inhibitors that, if safe, should be big winners.
    In this world, it only takes a few great successes to make up for several losses, and to make good money too.

  33. DrSnowboard says:

    @24 always good to keep the benefit to patients front and centre…

  34. Dado says:

    Given that SAEs and discontinuations were APOE4 and dose dependent (and there were no discontinuations in the placebo group vs 33% or so at 10mpk for example) I wonder whether the positive effect of the treatment on cognitive endpoints is overestimated due to the fact that there would be less APOE4 carriers in the treatment groups, especially at 10mpk, which are well known to have steeper cognitive decline rates compared to non-carriers.

  35. Dado says:

    Given that SAEs and discontinuations were APOE4 and dose dependent (and there were no discontinuations in the placebo group vs 33% or so at 10mpk for example) I wonder whether the positive effect of the treatment on cognitive endpoints is overestimated due to the fact that there would be less APOE4 carriers in the treatment groups, especially at 10mpk, which are well known to have steeper cognitive decline rates compared to non-carriers.

  36. Anonymous says:

    @35, Dado — “overestimated”? Seems likely. Biogen-Idec seems to be drinking the same Kool-Aid that Eli Lilly finds so tasty.

  37. Allchemistry says:

    If anti-amyloid antibodies indeed protect against Alzheimer disease, than vaccination would seem a logical next step.

  38. Anonymous says:

    Since the majority of the AD population is ApoE4 positive (60-75%) the SAEs being so high in this subgroup are problematic. Additionally, the cognition results are not as good IMO as they have been made to look. The MMSE results are not dose dependent in that the 3 mg group had better results than the 10 mg group. Additionally, the 6mg group has not been fully reported yet (why?) and the results matched the placebo group until halfway through the trial based on the slides presented at the meeting. The CDR-SB was only significant at the highest dose level which experienced the highest SAEs.

  39. anon says:

    Do you have the url for the slide presentation?
    I am sure quite a few would be interested.

  40. Wildbiftek says:

    @johnnyboy The document 1742-4933-10-18 dot pdf (Google it) gives a nice survey of the amyloid targeted therapies both past and present. Biogen’s compound squarely falls into that category although it does have some distinctions.

  41. Anonymous says:

    All the rush decisions are made by Biogen CEO who is 68 y old and ready to retire ultra rich. Mr Scangos promised excellent results after phase I on XL999 for prostate cancerwhen he was CEO of Exelixis. XL999 failed after phase III and Exelixis stocks tumbled and the company was forced to lay off 2/3d of his employees. Biogen executives are really successfull trying to gain investor confidence by always giving an upbeat view on company guidance.
    By the time patients find out the results if a drug which will never a miracle cure for AD, Scangos will be 70 years old and one of the top richest in the Boston area!

  42. Anonymous says:

    All the rush decisions are made by Biogen CEO who is 68 y old and ready to retire ultra rich. Mr Scangos promised excellent results after phase I on XL999 for prostate cancer when he was CEO of Exelixis. XL999 failed after phase III and Exelixis stocks tumbled and the company was forced to lay off 2/3d of his employees. Biogen executives are really successfull trying to gain investor confidence by always giving an upbeat view on company guidance.
    By the time patients find out the results if a drug which will never a miracle cure for AD, Scangos will be 70 years old and one of the top richest in the Boston area!

  43. anon says:

    That is startling.
    How could this ever be allowed?
    Why are there not laws requiring options to be tied to actual achievements and not merely press releases?
    Anyone could play the game of how much do you want to bet that there is gold at the bottom of this mine. That is not investing and should not be legal.

  44. Anonymous says:

    I wonder if Biogen conducted the Trials….

  45. Anonymous says:

    I hope this therapy fails miserably in phase III. I’ve had enough with all the MBA, lawyer, and accountant CEOs making all the money. It’s time they learn a lesson and a thing or two about humiliation.

  46. Anonymous says:

    I hope this therapy fails miserably in phase III. I’ve had enough with all the MBA, lawyer, and accountant CEOs making all the money. It’s time they learn a lesson and a thing or two about humiliation.

  47. Anonymous says:

    Well, well, well,
    Looks like a newly released article may throw cold water on Biogen’s data:
    http://www.bloomberg.com/news/articles/2015-03-24/alzheimer-s-debate-revived-even-as-biogen-s-drug-trial-advances

  48. anon says:

    @47
    I never understood how the markets worked.
    This research report on the underlying cause of Alzheimer’s appearing 2 days after the (expected) release of the Biogen results greatly helped clarify the media cycle for me.
    Research has established the central importance of tau in AD over the last 25 years.
    Why now?

  49. warrwim says:

    Apparently up to 1 in 5 individuals are Apoe4 carriers.

  50. anon says:

    Yes, APOE4 is the ancestral genotype and probably conferred (confers) benefits on those with it.
    Half of Alzheimer patients have it.

Comments are closed.