Targeted Therapeutics: The GIST Prototype

SLIDES & TRANSCRIPTS
Monday, June 17

Targeted Therapeutics: The GIST Prototype

Allan Van Oosteron, MD, PhD

Slide 1:

DR. VAN OOSTEROM: First, I want to say something, that I thank Murray, Ernest and Scott for inviting me, I am the only non-American speaker here today. I will first tell about how it started. In January 1999, Novartis organized a meeting of some drug developers in Newark. They told us about the first 16 patients they had treated with CML, where they had addressed the bcr-abl and that they saw fabulous responses.

There was also discussed -- which were the other targets for GIST, being kit, and PDGF. During that meeting, some people said, well kit, small cell lung cancer, we can do that in the United States. Where you have growth factor receptors, brain tumors and sarcomas, we can do that in the United States. By the fact, we at EORTC had lot of contacts with many sarcoma investigators. The thought there was that the sarcoma cooperation in the United States was basically dead, and that for studies in sarcoma was better to go to Europe.

It is owing to the courage of people from Novartis who decided that the first target to do was not lung cancer, not breast cancer or colorectal cancer, but small and rare diseases. That we now are now sitting here, with many of my friends which I have known for 25 years and are working in sarcomas, some of their names are here, some of the names are not here because they were not instrumental in the development of imatinib. You have seen them here and that is why I said it is the anti-GIST conspiracy that we have worked on.

Slide 2:

This is the structure of imatinib. Here is the original name. Here is the original name. It is a very difficult structure.

Slide 3:

You will see that it is now, in the United States, called "Gleevec". In Europe, we call it "Glivec". For us, there is an "e". They took a drug which was called glioma-vector. That didn't work out and they had that patent and they therefore decided to give that name to this drug. We call it "STI571". You know why, but it is normally called "imatinib".
It was known to inhibit tyrosine kinases including bcr-abl kinase PDGF and c-kit. It is based on research of an ATP-binding site. It is a highly soluble pill which is orally available, and at this time it is available only as 100 milligram yellow capsules, but I am told that shortly we will have 400 milligram capsules.

Slide 4:

This was the data that was presented. We see that these kinases were inhibited in fairly low concentrations, and these other kinases, it didn't work at all. When I came back to Europe, I asked Jaap and Martine from the EORTC to help to write the phase I protocol, we had agreement that we should also do the phase II and eventually the EORTC should do it if it should go into phase III.

The protocol was written and approved by the EORTC in June 1999, and we started exactly one year and two months later because Novartis had no drug. Because of the fantastic explosion of CML, they had to upscale the drug five times before we got the drug to start this study. Therefore, we had a protocol that was 14 months old. Therefore, we started very late, not too late, of course, but we started late.

Slide 5:

The protocol sets out that this was an introduction in all types of sarcomas, basically over-expressed platelet derived growth factor and for kit, CD117 positivity was a must and CD34 was also a must.

Slide 6:

The way the drug works has been very nicely described recently in the New England Journal of Medicine by Karen Antman, and you see that the stem cell factor comes here and you get dimerization. After dimerization takes place, then there comes phosphorylation, and then this single molecule or molecules are then going further into the cell. Imatinib, due to a structure, blocks this ATP binding so that the signals don't work and they don't function. That depends, of course, on whether the mutation is right, but that we didn't know yet.

Slide 7:

GIST, as you all know -- I think this is for sarcoma people a slide which basically I don't have to address. The only important thing is that most of these GIST are in the gastric. But you can have them in the small intestines, in the colon and even in the esophagus.

The incidence is, to our estimation, four to six per million per year, so for 10 million inhabitants about 50 GIST per year. Since most patients with GIST live a considerable time, just by the fact that we don't know the natural course completely very well, we think that the number of patients with GIST who are living is about five times higher than yearly incidence. You will see the median survival here for locally advanced disease and for metastatic disease. They are basically resistant to radiotherapy and also resistant to chemotherapy.

Slide 8:

We started, basically, so you will understand, in August of 2000, we started with four patients and a 400 milligram dose. That was the dose that was in CML the usual dose, then we increased to 600 milligrams, two times 300 milligrams per day. Since we saw early effects in these patients and we had so many patients who were sent to us in London and Rotterdam, we decided to do four patients extra in this cohort, to be sure that this cohort was safe.

So, we increased, and you will see the last patient in this cohort, since here we saw a considerable amount of toxicity, we started with at least 500 milligram directly, then we did eight patients more to ensure that 400 milligram twice a day was the dose to go for.

Slide 9:

Well, here you will see the DLTs, and I think that is well known to everybody here -- edema, rash, neutropenia and, in the highest dose, nausea, vomiting, and also edema. Here you see the hematologic side effects up to 18 months in our phase I patients. You will see that we have some leukopenia, we have some granulocytopenia, but not much, and we have some very heavy anemia.

Slide 10:

The non-hematological side effects you can see that here, especially in the 1,000 milligram dose cohort, we had lots of non-hematological side effects.

Slide 11:

I think the most important is the responses that we have seen in these patients, despite the fact of course, that in phase I study you basically should not look at response but toxicity, and you will see that we had four responses in the 400 milligram; four in the 600 milligram, nine of the 16 patients treated with 800 milligram, and five of the seven patients who had a GIST in the 1,000 milligram group, you realized that five of the patients in this study didn't have GIST but had other sarcomas that were non-kit positive.

Slide 12:

So, the response rate was very high and here you see the updates as of May 15. Of the 35 cases, we had 22 partial responses, of which there are 17 at this moment still on treatment.

Of the five patients that relapsed despite the fact that they originally had a response, three died and two on the 400 milligram dose had their dose escalated to 800 milligram and then again had a response. Ten patients had stable disease, four of them relapsed, two died, two had their dose escalated from 400 and 600 milligram to 800 and 1,000 milligram, respectively, and have, again, stable disease. Two patients died, implying that, as of four weeks ago, 27 of the 35 patients were still in treatment with 17 PRs and 10 stable disease. Therefore, four had their dose escalated. Nine died of the disease. Of the non-GIST in the phase I, four died and one was alive.

Slide 13:

Here you see the responding patients. Most of you will have seen this. That is how the patient comes in, with his belly full of tumor. This is after eight weeks. This is after six months, and now it is smaller.

Slide 14:

That was a patient of mine. This is a patient of Jaap Verweij. He had here tumor and here large tumors and you will see that after eight weeks those tumors have also essentially come down.

At Harvard and in my institution we have two very dedicated PET people.

Slide 15:

We really say that PET is the end of everything, and I think especially in GIST PET is telling you something informative because here you see the lesion on CT scan, but here you see these same lesions on the PET scan. Here you see the whole patient.

Then, this is after eight days. It remains like that, and it takes about four months until you see really a partial response on the classical CT scans, which I think is very important for the patients and especially I think in neoadjuvant treatment. I think the making of a PET scan can convince the surgeon that he should not operate immediately but first, if the patient is FGF-positive before they start, that they should continue for eight to twelve weeks or maybe 16 weeks -- that is not yet clear -- before they operate and then you can see, at operation, how it goes.

Slide 16:

During the same time that we did this study, the U.S.-Finland group did a phase II trial in which they compared the 400 milligram to 600 milligram, and they also did correlative science studies, imaging, and the biopsies.

Slide 17:

We see in this data that Chuck Blanke presented with me at ASCO last year, they entered 147 patients between July 2000 and April 2001. You will see that most patients were heavily pretreated.

Slide 18:

Here you see the overall response rate, as they reported response rate in their study. It was just the same as in our studies -- about 60 percent partial responses and 20 percent stable diseases.

I think the most important thing from this study, however, is the work that has been performed by Mike Heinrich, and I'm sorry Mike that I have an old slide included not the most recent one; but the most recent one is the one that Charles Blanke reported last year at ASCO.

Slide 19:

I think this is so fundamental that you see there is a difference in response rate for the different mutations, and that is holding up both in the enlarged American study and also the unpublished data from our EORTC studies.

You see that if a patient has an axon 11 mutation, the likelihood of having a response is nearly 80 percent. If the patient isn't axon 11, it is only 40 percent; and if the patients have a wild type, then it is only about 10 percent of those patients that will respond.

This tells you that basically what we should do -- I completely agree with Chris earlier today -- that it might at this moment be very difficult to evaluate patients with immunohistochemistry with the DAKO reagent and then, if it is positive, on a non-antigen retrieved sample that you can go directly and look for the mutation analysis. I think in the future we need to find something to directly detect which mutation it is, because that tells you then the likelihood that the patient can respond.

Slide 20:

Here you will see the upgraded data as Meg VonMehren presented them at ASCO this year, and about 18 percent of their patients have progressed on the 400 to 600 milligram. Some patients they also increased the dosage so, again saw some responses and some of those patients were operated.

Slide 21:

Of course, the question is, why did these patients progress? That is one of the things that we still have to look at.

Again, we need to sample the patients, we need fresh tissue from patients who responded. I can tell you, for example, in my unit, four patients died, and out of all these patients, we have done an autopsy within six hours. We really kept them in the hospital to ensure that we would have tissue of these patients to do these types of studies. I think that, in my opinion, that is the way medical oncologists and surgeons should work at this moment. We need tissue to analyze what is going on. It is important to know whether kit is still present, is it activated despite continuation of imatinib? We have to evaluate the drug levels and other signaling pathways, just as Dave Parkinson today also said.

Then we have to develop rational options for therapies with the above data. I think in the future we will do combination therapies addressing both kit and maybe one of the detected pathways downstream from kit.

Slide 22:

We have here the conclusion that it was very well tolerated and safe for at least 15 months and has the same response and patients still on drug as we have.

Slide 23:

Then the EORTC did a phase II study in GIST and non-GIST, including patients who were c-kit negative, to see if you could do something with the PDGF receptor. As was already alluded to, that didn't work out.

Slide 24:

Here you will see that in the non-GIST patients we have no responses and in the GIST patients,

Slide 25:

as you can see here,

Slide 26:

that most patients respond very well as Ian Judson reported at ASCO this year.

Slide 27:

It was basically Jaap's study but he was not allowed to go this year, because he flies too much over already, according to his boss.

The other thing which I think is important to say is that not all GIST patients respond directly because about half of the response will take up to six months, and we haven't even seen responses up to 12 months in patients, which were stable. Telling you that there might be different mechanisms of action.

Slide 28:

The conclusion of Ian also was that we had 73 percent of the patients who were progressing.

Slide 29:

Then here is the culmination of the whole thing that began now 18 months ago. This trial started in December, both the American groups and the European groups together with Scott Saxman started to talk about a large study which in the United States was an Intergroup study and in Europe was EORTC, Scandinavian, Australian, and Italian group study, in which we compared the 400 milligrams once a day with the 400 milligrams twice a day.

In the United States study which started the 15th of January 2001 and ran through September 1, they entered 746 patients.

Slide 30:

Here is the accrual on the EORTC study and we had 946 patients in total. That tells you a little bit about how Intergroup cooperation can go.

Slide 31:

In conclusion, we have here now in the treatment of this disease, we have in the EORTC 35 patients; you see that we have 77 percent of the patients still on drug after 17 to 21 months. In the U.S./Finland study, we have 73 percent of the patients still on drug. In the EORTC phase II, we have 74 percent.

We have 1,692 patients which are in the two phase III studies, of which we have promised each other, because we weren't allowed enter patients into each other's studies -- I will come back to that -- that we would do meta-analysis with the same entry criteria, et cetera, to exactly know how the score is. To my regret, there are many phase IV studies in which we will learn nothing.

We have to be aware that we have late leukopenia, rash and bleeding and in the case of diseases progression my advice is to dose increase or to do surgery.
The other thing is, what have we learned? What we have learned is that kit should be determined without antigen retrieval and can be done in every center in the world who really wants to do it, because there are allways mast cells present which are the positive controls.

CRs are, in metastatic patients, very rare. We have many PRs and stable diseases, but some of these patients will recur.

We need their tissue to analyze it to see what we can do further. Adjuvant, especially neoadjuvant studies, have started, but for me, the dose and the duration of the treatment remain problematic because I don't know exactly what advice to give at the moment.

Especially for the neoadjuvant treatments, the surgeons sleep quietly and leave the patient to the medical oncologists for the time being, because if they perform a PET scan and the PET scan turns negative, then the treatment works and then they should operate the patient after therapy when the disease is really small.

Mutational analysis is fundamental. I think what Mike Heinrich and Brian Drucker has done at Oregon is remarkable. So, Oregon should be the imantinib place and be renamed I think.

Anyhow, that is a fundamental we should really be looking hard to find a pharmaceutical to find a way to do the mutational analysis study very rapidly and very simply, so that we can promise the patients better things than we can at the moment.

As I said, even in GIST, we should go for combinations to address not only kit but also the pathways further down.

Dave, I think you are absolutely right. Since we have so few complete responses, and that is what our goal should be for metastatic disease, it is not the end. It is only the beginning of the targeted therapy. Everybody thinks it is the end? It is the beginning of the end, I hope.

Finally, in the future, we should increase our cooperation globally between the major groups -- Americans and the EORTC and other groups and industry -- and discuss how we can overcome the problems that block this optimal cooperation at the moment. Thank you very much.

[Applause.]