Smoothened PaTCHes of Hedgehogs

Posted by Jeff Berk, BOLT International;  I’m spending the week at AACR 2011, in Orlando.  If you’re here, give me a shout!  I really like this meeting.  Not quite as crazy as ASCO, and less pompous than ASH.  There have been a lot of intriguing presentations, most of which I’m sworn to secrecy not to talk about; since I’m down here working for “The Man”.  So you’re probably thinking that this is going to be a pretty worthless blog (what a set up!)


There is a lot of interest here in anything to do with the hedgehog pathway.  PTCH, SMO, Hh, even some more upstream targets.  I had a no holds barred conversation with a guy who knows quite a bit about this subject a couple of months ago, and I want to share some of his thoughts.  I’m censoring it a little to protect the innocent, but hopefully you’ll find some of his insights to be of value anyway:




Thank you very much for fitting me in.


You don’t ask easy questions either.  But, my sweet spot is smoothened hedgehog.



Okay.  So right now let’s start with the overarching question of you and I talked about pancreatic not that long ago and so obviously that is going to be on your list, but what are the tumor types that you think about that have a potential to be impacted by that pathway?


There aren’t that many surprisingly enough.  The only tumor types where it has been proven to be active, any of these, in just one of them to be specific, has been in locally advanced or metastatic basal cell, which is not very many patients.  The other tumor type is the most common children’s brain tumor, medulloblastoma.  So those are the only two that have definitely had any evidence of activity.  So far, and the only big trial that has been done is colon cancer, GDC-0449 and it has been very disappointing.



That was the one just reported at ESMO?


Yep.  It is very disappointing.



Were you there by the way?


No.  I did know the results.  It is Genentech.  I don’t know why they did it.  I think they are feeling their oats.  The biology is kind of iffy.  The biology was very good in the pancreas where it looks like it improved drug transport of gemcitabine.  There is no evidence the drug works by itself at all in any other tumor type except where you have these patched or smoothened mutations and the only two that really are consistent is basal cell, medulloblastoma and a few rare sarcomas, very rare, very rare.  It is just very, very limited.  The only thing you can bet on right now is only those and it is just extremely small number.  It works.  GDC works extremely well.  We can go through the other ones here in a minute.  But right now it is the only thing that you can take to the bank.  Everything else is very speculative that you put it with chemotherapy that it has an anti-stromal so it takes out some of the fibrosis and the support of fibroblast and cancer associated fibroblasts, CAFs.  So you take off the cancer associated fibroblast by inhibiting the hedgehog pathway that you will get kind of a shrinking, shriveling of the stroma.  We call it stromal collapse.  We are hoping for that.  In animal models, the genetic engineered mouse there were fabulous results published in Science by Olive and colleagues.  Other than that, hedgehog is expressed in many tumor types but the only place this is a magic bullet is with patched or smoothened mutations.  I think it is very challenging to know what to do next.  Probably the best thing is being done next are both GDC and Infinity are doing clinical trials in combination with gemcitabine or other drugs in pancreas.  To me that is kind of a last big holdout as whether it makes gemcitabine better in pancreatic cancer.



I have heard different I will call them hand wavings until you give me their blessing, but I have heard different theories on how these inhibitors might be of benefit in terms of targeting stem cell populations.  You haven’t really mentioned that yet today.  Are there population of cells that are involved in regrowth of some of these other tumors where even though overexpression of hedgehog isn’t a main driver or if smoothened is not a main driver of these tumors initially, might that have any impact on resistance mechanisms or regrowth of some of these tumors that are knocked down either by chemotherapy or some other therapy?


You bring up an important point.  …there was a triple threat; one was the mutation, the other one was the stroma and the third was the stem cells.  There is no doubt.  Matra, at Hopkins, and other people have shown that stem cells certainly are support by – they don’t have mutations, but they are dependent on hedgehog signaling pathway –  is probably the smartest way to say it for survival.  So then you say how would I prove that clinically.  So the only way to do it be as you say, knock it back, if you can.  So you look up and say well pancreas cancer we don’t know how to knock it back yet.  You could use it in patients after surgery and hope for the best there.  That will take a bold trial.  There are only 20 percent of people you can do surgery on and then randomized with or without the hedgehog inhibitor.  In breast cancer they are also expressed, but again you are talking about a pretty big adjuvant trial.  You knock it back and there is going to be randomization.  So it is pretty big.  And the answer is it has some potential.  I don’t think it is necessarily too much hand waving.  I would say that is smart and I hadn’t covered that.  It is a potential but I always say, what is the clinical trial design we have to get that?  Just as you said, you knock back and then you randomized.  It could be done.  Somebody is going to have to have some deep pockets to do it.



Come back to and if you would benchmark GDC versus the Infinity drug, IPI-926.  Most of the data I have seen is on GDC 0449 so my assumption is it is just notable further along in clinical trials.  So speaking to what is in the public domain what can you say in terms of the comparison of these two drugs?  Are they more alike or more different?


I can’t say very much (confidentiality issue).  I have both of them.  I think they both work is probably the best I can say.  I think they both work.  GDC is way out in front of IPI-926, but I think that IPI also works.  I have also have the guys at Novartis who think that with LDE 225, the sun rises and sets.  I would say that you are probably missing three others on here.  The only one I am not familiar with is the ENZ-4482.  I actually tried to look that up and I couldn’t find anything on that.



I will tell you what I’ve heard about that in a second,  but what should be on my list that is not?


I can’t say.  That “I can’t” is because we are going to start the phase I trials with them and I am sworn to secrecy.



That is fine.


I would say there is a good solid three other ones; one that will start clinical trials in a week.  But you know it almost doesn’t make too much difference because I think that they are all going to be kind of crowded into just a small use.  Now how do I know these things work?  I know they work because they work on the PTCH mutations, not because they have been active in other tumor types.  It is just the same thing.  So if you don’t get activity there you don’t have a drug, I think, because it says at least you are getting the pathway taken out.  The only ones I am aware of, although a good friend at Novartis says that the 225 is looking good, so only the top three you have there I think is something you actually take to the bank.  I don’t know the Pfizer one and the Exelixis one, again, I couldn’t find it.



I don’t know very much about the Pfizer one.  I do know that BMS opted to not exercise their rights to take forward XL-139, either because they saw a lot of competition ahead of them or because they know that the drug doesn’t work.  I can’t speak to which of those but I can say that BMS that had access to the data package opted to not go for that.  Take that for what it is worth.  The ENZ-4482, my understanding is that it is an miRNA based approach and so there are some questions on how you deliver it and that kind of thing, but that is the only one that mechanistically in any way seemed very different to me than every other molecule I had listed there.


I really appreciate the education on that.  I always learn something speaking with you.  Thank you.  Boy the miRNA is going to have a ways to go on that.  I have to say that if they were just things that you knew for sure, well you know for sure that the top three are in play and you can see that there is a lot of skepticism mounting.  There is no question that the GDC is working and IPI I know for sure.  Both of those are looking good.  GDC has filed, I think, already with the FDA for the basal cell indication.  I don’t know that for sure.



So what would be the reason Novartis would go forward with LDE-225?


First of all, there is no question that you are going to develop mutations.  So it is going to be the AIDS story all over again.  You take these things for fairly long periods of time.  I don’t buy that it is such a small market.  I just buy that the tumor types are kind of on the rare side.



I see.


But so was CML.





Nine billion dollars later.  Of course, you know Gleevec worked in GI stromal tumor and that is uncommon too.  I think the other thing is that a lot of Mohs surgeons, the dermatology community has really gotten into this GDC.  I have been asked by every dermatologist in the world to come talk.  Why?  Because that is Mohs surgery.   I mean one young woman came up to me in Kentucky about a year ago and she said, “Am I going to be put out of business.  I was hoping I could put my kids through college”.  She is a Mohs surgeon, and gets 7,000 bucks a crack.  I said, no don’t worry about it.  But I think the deep ethmoid sinus, down to the dura kind of things it is going to be make these people more operable.  They are going to start using this to put people into some induction and then they are going to do their surgeries, which take literally six or seven days of carving away on people.  I had Mohs surgery but it took me two days to get one off my lip.  It just is really something that you would rather not have.  So you can downstage it.  I think they are doing that.  And then the other thing is all these people are working on topical formulations.  Basal cell is the most common cancer in the world.  And the dermatologist, strangely enough, they actually like creams.  The Mohs surgeons make huge amounts of money.  But if they get something on the bridge of the nose they don’t like to do it.



I want to show my ignorance.  You are calling them a Mohs surgeon?


Mohs was a dermatology surgeon from the University of Wisconsin.  It is not ignorance at all.  What he did was he came up with a way of doing surgery to try to spare normal tissues.  So think of one of these things sitting on the bridge of the nose.  Well you want to try to preserve tissue so what he did is he took a lot of Novocain and then a section lidocaine they section it off, they look at it under the microscope, they let you sit there for an hour.  They process it and they say nope I have got to go to the right deeper so they come back and then they carve you to the right deeper.  And they keep going until they get negative margins.  They keep going in all directions and depths until they get negative margins.  That is called Mohs surgeon.  I got to know Mohs.  Believe it or not, he was 100 when I met him.  He came to a seminar I gave.  He has passed away now, but this is only about five years ago.  Mohs was the smartest guy in the room in terms of questions.  Unbelievable.  I couldn’t believe he was 100 years old.  He was so smart.  My point is he created this whole specialty.  The guys make 20,000 bucks on some ear procedure.  They try to save tissue, but they keep carving away at you and after a day goes by they carve as much as they can and they say, okay, come back tomorrow.  They patch you, put a bandage on you and you come back the next day hurting like a banshee.  They lidocaine it, you feel good again for a few ours, and they keep carving until they get a negative margin.  The problem is that sometimes they just can’t get around these suckers.  The basal cells send deep roots.  That is why when you used to lop them off they never really cured them.  By this methodology they got a much higher cure rate.  They don’t like to give radiation to these because sometimes feel that the things actually grow better.  There are also people with a disease called Gorlins Syndrome.  It is not that uncommon.  I don’t know the incidence but Gorlins Syndrome is where you get a hundred of these a year.  I will tell you that is why people stick after these because there is some chance for chronic use.  I was in a time when they actually priced Novartis’ Gleevec and the guy said oh we will never make any money on it because this disease is so uncommon.  They forgot one thing, it works.  If it works it is going to be a like an antihypertensive and people will take it for long periods of time.



So right now what is the average survival for a patient with these basal cell carcinomas?


Years and years.  For even the locally invasive, we call it extensive local disease, for the metastatic disease it is about ten months.  For the locally invasive it can be for years.  For the Gorlins patients it is from the time they are born for their whole life until one gets out of hand.  They usually go 70 or 80 years.



What I am hearing you say is that if we knock down the hedgehog pathway you are not going to necessarily eliminate these tumors but in terms of their invasiveness, is that what the objective is?


What happens is they do go away but there is a resistant clone and that resistant clone grows out, it is selected for and you are out a year and the thing is growing and the GDC-0449 is not working anymore, so you will switch to another one.



Is that resistance mechanism known?


We know it for medulloblastoma, an acquired SMO mutation.  It is a very good question as usual.  We know it for some of the patients; about a third of the patients, have resistance based on a new clone.  Other mechanism resistance we don’t see a new one that we can detect.  So two-thirds of the time we are not positive that it is due to a resistant clone.  There are a lot of things being looked at like transport or something else.  There is room for other ones because eventually people progress on 0449.



Is any new hedgehog or smoothened inhibitor going to get niched into 449 failures?


I think they all will.  The question is, I don’t know what the strategy of the next people coming along is going to be.  All I can tell you is that many of the patients going on these second and third generations have all progressed on 449.



You brought up CML and Novartis and so immediately I am thinking they came out with imatinib and then BMS comes out with dasatinib and it was first the blast crisis and then moving into second line in chronic and then eventually first line in chronic as well.


You are exactly right.  I am just saying that is why people are pursuing these because the minimalist approach would be I will just take the patients who progressed and I will have nilotinib which is pretty good, or dasatinib which is pretty good, again, because people will take it for longer periods of time.  Usually two courses and out doesn’t apply here.



Very good.  Now in terms of these medulloblastomas, so you made your case to me on why these basal cell carcinomas we are going to turn that into a chronic lifelong therapy commercially, great.


Especially for the Gorlins Syndrome.  That is a big deal.



Now I turn to the medulloblastomas.  Can you make the same case that from a commercial standpoint that use of a smoothened or hedgehog inhibitor is going to have that kind of dramatic impact on chronic usage of a drug?


Yeah, I don’t think there is any question.  First of all, here is the negative.  In medulloblastoma the only patient in the NEJM was a 23-year-old man.  The guy had it for a long time and he was bedridden and he got up in a few days and walked.  His bone metastasis disappeared.  Weren’t allowed to treat a child because in mice this fuses the growth plates in bone.  Of course, one of the fathers called me up because I wasn’t allowed to treat him but he said, “you and I know we would rather have a short son than no son at all”.  The FDA has opened up and allowed GDC-0449 to be utilized by the pediatric oncology branch in children.  I would say that I would be afraid to stop it.  I would say as uncommon as the disease is, it is still the most common cancer in children that it would have a long-term use.  I wouldn’t want to stop it.  It is non-toxic.



That was my next question about what the downsides are to it?


None.  Well, the bone plates is about the only thing.  You brought up the stem cell.  A lot of people felt out there that we were going to have trouble.  We have had people on for more than two and a half years on GDC.  That is no secret.  A lot of people felt if it is an anti-stem cell it would get our bone marrow stem cells.  Not a whisper.  Not even a whisper even in patients who had a lot of prior chemotherapy.  Nothing.



Were you surprised?


I wasn’t surprised myself.  I thought okay we got it on some stem cells but okay, are they really stem cells, and the answer was apparently not because they are not in our bone marrow stem cells.  The second point was that maybe they would hit our stem cells in the brain.  That was a big worry; people would develop Alzheimer’s.  And the same thing, as far as we know.  I just got a picture of one of my patients, you aren’t going to believe, the 90th birthday he just did his first free fall in sky diving.  Of course he has got a guy behind him.  But it is great.  He took out his teeth first.  He looks like an astronaut with his cheeks just punched in and he is so happy.  He has got a left ear because of GDC-0449.  And if any guy is going to lose his marbles it would be him but nothing.  And nobody, people have been on for a long time, no weird behavior, none of that stuff.  So none of those things have come to pass yet.  But it has to be something that has passed a couple of years of treatment to do it because we just aren’t seeing it.  So I am a little skeptical that the expression on some.  Of course, it might take years to come to find out.  I can’t guarantee that.  But its either that or you lose them.  I think that people will continue to take them.  I would never stop them if I had my choice.



Got it.  So final question on this pathway.  If I were Infinity and I needed to jump start, close the gap with Genentech where do you push for early registration or for a hole that FDA will look at and say given the benefit of the doubt we get IPI out there as well?


I personally would go where the gold is.  You are only going to have a couple of choices.  If GDC gets approved, your only two choices are going to be verses GDC, which is almost ridiculous.  I would just take patients who progressed on GDC.  I have no idea.  Genentech let the NCI do the medulloblastoma.  I would go out there and recruit the medulloblastoma and get a second indication.  I might even win with the medulloblastoma indication.  So I do both at the same time.  It is no secret and it was published in JNCI that they have got the IPI-926.  We are going to start a trial with gemcitabine/Abraxane with or without.



In pancreatic ductal adenocarcinoma?


Yeah.  Going for pancreas.  So that is the odd man out.  I know other people are thinking about ovarian to try to get that stem cell.  So it’s in pancreas because of the Science article of Olive and colleagues.



I am trying to remember from a couple of ASCOs ago I thought I saw a GDC-0449 trial in small cell lung cancer.


Oh yeah, they are trying everything.  Nothing has panned out yet.  I think they haven’t been reported the small cell.  I suspect for a reason, but in small cell it takes so long to accrue.  I don’t have any working knowledge of that, but it takes so long to accrue small cell in this country.  It is a disappearing disease.



Categories: basal cell carcinoma, CANCER, colorectal cancer, Gorlin's syndrome, medulloblastoma, pancreatic ductal adenocarcinoma
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>