BMS-708163 (Bristol-Myers Squibb) for Alzheimer’s Disease and Mild Cognitive Impairment

Posted by Jeff Berk, BOLT International;


My next meeting is going to be ASCO.  Can I tell you, I’m so beyond thrilled to be getting to go to another meeting in Chicago.  Maybe this time my hotel room will be North of Milwaukee, instead of the 50 miles away from McCormack the last ASCO.  And what community organizer got bribed so that the rest of us have to go there 10 straight years?!  Go Cubs.


Each year, at about this time, BOLT runs a special panel with our Alzheimer’s Disease Thought leaders.  We call it our “Fantasy Formulary” edition.  The essence of the exercise is to force each Panelist to choose a group of pipeline therapies that he would put on his formulary, if his “budget” was limited.  There is a fair amount of projection that takes place, as some drugs may have very limited clinical data available, and some may simply be a product concept.  Since I’m a tease by nature, I’m sharing in this blog the past couple of years’ responses based on desired mechanism of action.  So 2009 and 2010 looked something like this:



  • One of the specific products that was rated very highly was BMS-708163, which is a potent and selective γ-secretase inhibitor.   Since I really like BMS’ development strategy for the drug, I’ve decided to blab on about it tonight.



  • In Ph-I trials it decreased CSF Aβ40 and Aβ42 approximately 30% following daily dose of 100 mg after 28 days and by 60% at daily dose of 150 mg.   BMS-708153 is 190-fold more selective for APP than for Notch:  It has an IC50 for APP cleavage of 0.3 nM and an IC50 of 58 nM for Notch.  The APP potency is much higher than that seen for begacestat (15 nM).



  • BMS-708163 looks to be more “notch sparing” than semagacestat (since deceased).  The drug hits its target in man, lowers A-beta and the gastrointestinal toxicity profile of BMS-708163 is cleaner than semagacestat. There are two top-of-mind reasons why a notch sparing drug will be preferable to one that also inhibits notch.  First notch signaling is required for normal differentiation in the gut.  Notch inhibitors lead to rapid gastrointestinal side effects.  Second, notch is a tumor suppressor in the skin.  There is a fear that inhibiting notch in older patients will led to skin changes, loss of hair color, and potentially tumors – precisely WHY semagacestat is no longer with us).  Earlier use of a secretase inhibitor is predicted to have a better disease modifying outcome, and this safety profile has allowed BMS to take BMS-708163 into a 270-patient placebo controlled prodromal MCI (mild cognitive impairment) trial.

“In the setting of a cancer as long as you don’t kill the person and you kill the cancer you are okay.  You probably only have to treat for weeks to months.  In the setting of Alzheimer’s disease, you have to do this chronically.  The main side effect is an intolerable one.  It is basically GI; notch segment needs to be required for normal differentiation of the gut.  So first of all, you have a pretty immediate and tough to deal with GI problem.  Secondly, you will have long-term effects on the immune system, which at least depending on dose, timing and context may be good or bad, but something that is also very worrisome in an elderly population.  In the skin notch is a tumor suppressor so you could develop skin changes and all sorts of stuff and your hair will turn white potentially.  Notch is such a central signaling molecule even in the living adult in an elderly individual that it is a huge concern”.


“So if we have budget cuts and sorry to tell you we are going to cut your formulary down to one new drug what is your one drug going to be?  I guess I would say a secretase inhibitor.  Which one?  If I had to pick one…  You only get one on your new formulary sorry to say.  If I had to make the choice today I think I would choose the BMS gamma secretase inhibitor”.



“I think the Lilly drug has strong data from say CSF studies that they can lower A-beta, but I think the BMS compound looks quite interesting too.  I think that is fair if you say we don’t really have a way of comparing.  I wouldn’t argue with that.”


“So which drugs do I think would work best there?  I think the drugs that reduce amyloid work best there.   So there are a lot of pharmacotherapies in the pipeline. I think the two big categories are secretase inhibitors and immunotherapeutics.  For secretase inhibitors as far as drugs that are out there already I might say that the BMS gamma secretase inhibitor looks pretty promising.  Now is that based on the fact that they have got the trial population with MCI whereas Lilly doesn’t? It is based on that.  I think that what has been presented on that compound looks pretty good”.


“Now I could make guesses to the leading candidate for each of these.  As of this moment I would probably say the leading candidate for secretase inhibitor would be the BMS gamma secretase inhibitor perhaps.  A lot of people said they expected to have certainly a better notch friendly profile than Lilly’s. Based on what we’ve all seen it looks at this point better”.


“The BMS compound appears to be much more effective in terms of reduction of A-beta generation at the dose that is going to move forward, yeah”.



“I still think I would say in the A-beta therapies I think you would have to raise the notch-sparing GSI’s especially BMS’s pretty high up there.  It hits target in man very clearly.  It reduces A-beta production in the brain and seems to be reasonably well tolerated”.



“What I would say is that this drug it will be able to lower A-beta and the mechanistic based toxicity is likely to be eliminated.  That doesn’t mean it won’t be toxic”.



“I am familiar with this BMS drug and they have presented some nice demonstration of a margin of safety between the concentration that inhibits gamma secretase activity and beta amyloid processing and the activity that inhibits notch signaling.  Now I don’t remember seeing that sort of data for the Lilly drug.  It is probably just that it wasn’t presented in a concise manner.  That is the only other gamma secretase inhibitor that I know is that far along and the way the safety data has been presented is very encouraging but I don’t have any other reason for really concluding that it is going to be any better than the Lilly drug”.


It must be relatively notch sparing, otherwise the people taking the drug wouldn’t have tolerated it for a month, which is what they showed the data for at ICAD.  I think that is a promising drug as far as I can tell”.




  • It makes more sense to treat patients at the MCI stage than once they have suffered further neuronal degeneration.   Panelists really like the BMS strategy of studying BMS-708163 in an MCI trial anchored by an amyloid biomarker (NCT00890890).  Because low Abeta42 has been found to be associated with more rapid conversion from MCI to AD in patients starting with higher cognitive reserve, this trial is screening out patients who have a baseline Abeta42 level <200 ng/ml.

“Notch sparing gamma secretase inhibitors are the next.  My sense is that BMS are running trials in MCI patients.  That is a bold strategy.  I mean what is the basis for thinking that…? I think basically if you really look carefully at the preclinical studies, some of those I have generated, which we have talked about but haven’t published yet, say the likelihood of a gamma secretase inhibitor or any A-beta production inhibitor working once you have got a head full of amyloid are pretty low.  So even when you have MCI all bets are off because you have got a head full of amyloid and you have had it for a long time.  So I think they are taking their best shot.  It is not that I still think therapies targeting production are going to necessarily be the best bet moving forward, at least in the therapeutic setting.  In the prevention I think they might work.  Again, the issue is I don’t know how we do that trial.  But I think of all the preclinical data I have seen, this compound is going to hit its target.  It looks like it will, at least in the short term, be reasonably well tolerated.  That you are not going to get the notch related side effects.  What other side effects you can get I have no idea.  There could be a plethora of them that could be worrisome.  That is sort of where I stand just because I think you have got to give a company credit.  Their preclinical data that was presented at ICAD last year was really quite convincing, and I think best in class at this point for anti-A-beta therapies”.


My sense is that BMS is being very bold in running that drug in an MCI population.  Is there any reason to think that a gamma secretase inhibitor used at the MCI stage?  Is there any data to show use of the gamma secretase inhibitor at the MCI stage modifies the course of disease any differently than using it at the AD stage? No, this was all based on models that people have that is a concern that if you treat patients too late you have brain failure and it is better to treat people early, but that is not based on data.  It is just based on speculation.  There are a few patients in the 1792 study, the original Elan vaccine study – some of these people at autopsy had amyloid that seemed to have been pulled out of their brains – but these people declined and their dementia didn’t get better.  There is few number of subjects though so you have got to be careful not to draw definitive conclusions from that.  But it just raises concerns that if you wait too long and people get too severely affected before you pull amyloid out it is like treating somebody with heart failure with statins.  You can start lowering the blood cholesterol but if they have had three heart attacks and the heart is all scarred up it is too late.  It is that kind of an analogy.  There is no data behind it at all.  BMS is bold but it is not all that bold.  They are looking at MCI.  They are not just taking everybody with MCI.  They are using biomarkers to help identify MCI patients who are most likely to have AD.  The concern is whether they are losing so much statistical power because the earlier you go in the disease the slower the rate of progression.  And the earlier you go in the disease the greater the variance in the rate of progression because you are including people who don’t have AD.  So this is based on the ratio of A-beta 40 to A-beta 42? Yeah they are using CSF A-beta”.


“My next stage of disease is what I would call early AD although many people prefer prodromal AD and that is the population currently being studied in the BMS program.  It is essentially MCI anchored by a biomarker.  I like the idea of MCI anchored by an amyloid biomarker”.

“That is the main difference (between NCT00890890 and a trial in asymptomatic patients) is that now you have got symptoms plus disease, symptoms plus neuropathology and my guess is you will have less effective disease modification because you are later and you may need more symptomatic treatments”.


“I think that secretase inhibitors and immunotherapies are still leading the charge, specifically if I have to choose in the secretase inhibitors I have really have a hard time choosing between the BMS product and the Lilly product.  But on that side I think the secretase inhibitors are showing at least good safety signals.  Whether they change outcome I think BMS was brave to go after the MCI group.  I think that was a worthy endeavor.  They are selecting them on the A-beta 42 marker and if it is low they get randomized, if it is not low enough they do not get randomized.  Was there anything based on data that has been published that supports anything more than it is a big market opportunity? No.  There is no data supporting why to go after those people.  Of course you look at Shah’s paper from late last year and we know very clearly that if you have absolute A-beta under 200 your probability of conversion to Alzheimer’s disease in the next year is north of 90 percent.  So selecting an enriched population like BMS has done is exactly what needs to be done.  So if there is going to be an intervention that works that may be an inflection point in which to do it.  From a scientific standpoint I think that it is a good idea.



  • Well, that’s enough for tonight.  I’ll post the safety comments when I run out of other things to talk about.


Categories: Alzheimer's / MCI, Neurology
Tags: , , , , , , , , , ,
Bookmark the permalink.

Post a comment

Comments on are monitored.

You may use these HTML tags and attributes:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>