SLIDES & TRANSCRIPTS
Wednesday, June 20

CHEMOTHERAPY/ALTERNATIVE DELIVERY METHODS SECTION - ANGIOGENESIS INHIBITORS

Fadlo Khuri, MD

Obviously I think Jack Ruckdeschel kicked things off by saying that there is no such thing as a benign lung cancer and these statistics kind of point out that this is a disease that has evolved the wrong way. I am sorry you cannot see the red curve which basically shows that non-small cell lung cancer is now by far the leading cause of cancer-related death among men and by 2010 according to current projections will kill twice as many women as breast cancer.

TOP

And we do have a nice correlation with stages. As you have seen we actually do better in stage I. Kathy Pisters has quoted a 67% 5-year survival for pT1 N0, M0 and pT2 tumors and I think that is probably the best figure we are going to use.

TOP

One thing that was debated earlier, and we added this slide is what is the role of PET? Many of you have been like myself I am sure just stunned by the advances in recent years with PET technology.
Here is a lady with not very impressive mediastinal nodes that actually were PET positive, and she was essentially saved from having surgery. One of the presentations at this year's ASCO was does PET save us from useless thoracotomies in stage I disease?

TOP

Unfortunately, the answer at least from the small trial by Dr. Boyer and others is no. In this randomized trial of about 165 evaluable patients you can see the thoracotomy rate is the same between the two arms as is the relapse and death rate and in reality the investigators were forced to conclude that in patients who appeared to have pathologic stage I or IIA tumors who had a PET scan as part of this randomized trial it did not reduce the incidence of thoracotomy.

TOP

We have come a long way, I am not sure the right way in the antiangiogenic field from a few years ago and I think many of you may recall that in 1997, we were, all supposed to be out of business at least the medical oncologists among us by 1999, when Dr. James Watson, a Nobel Laureate predicted that Dr. Folkman would cure all cancer within 2 years. Fortunately Dr. Folkman himself was the voice of reason saying that Endostatin can completely regress tumors but in mice. No humans have been treated. If you are a mouse and have cancer we can take good care of you. I respectfully disagree because in our experiments we mostly sacrifice the mice. So, I don't know if that qualifies as taking good care. Now, Endostatin has progressed. It was first discovered in 1995. It was licensed to Entremed in 1997. It was introduced in partnership with the NCI in 1998. Its stock went off the charts around here. Phase I clinical studies began at M.D. Anderson November 1999, and my good friend and colleague Roy Herbst presented the data this year at ASCO and saw its stock plummet. So, we have been talking to Roy about considering not taking a career in the financial industry.

TOP

. There are a number of important antiangiogenic agents. The question is where are they in clinical trials and what is the safety data; are any of these ready for prime time? I think probably the furthest along are these two molecules, the anti-VEGF antibody from Genentech. I am going to show you some data that was presented by Russ Devore a few years ago at ASCO and updated by Alan Sandler recently.
SU5416 is an agent that many investigators have been looking at particularly in colorectal cancer and Kaposi's Sarcoma, Gail Eckhardt, Lee Rosen and several others. It looks like it is a selective anti-VEGF agent. SU6668 which is a multi-tyrosine kinase inhibitor probably not only inhibits VEGF but BFGF and PDGF and there are many other molecules including Josh Fidler's all-time favorite interferon alpha which we have heard him claim many times is the only true antiangiogenic. I won't discuss it because of the absence of good data in lung cancer.

TOP

TNP470 is an agent that we have looked at, an agent derived from fumagillin. We have combination studies, and this looks interesting, and there are many other agents including Endostatin. So, this is not meant by any means to be an exhaustive list.

TOP

To try to tie in with what Ethan was talking about, one of the things that COX-2 does is it turns on VEGF and to some lesser degree but also a significant amount BFGF, and we know that certainly in models including Lewis lung cancer models. Also in some transgenics if you treat with COX-2 inhibitors you can basically shut off VEGF expression, and one of the reasons we got into looking at COX-2 expression and inhibitors was frankly to try to tie it in with retinoids because we had shown that RAR beta was overexpressed in a certain number of lung cancers, and those patients did poorly. Those were stage I lung cancers. We were somewhat relieved to find that those same patients had overexpression of COX-2. When we have looked again in unpublished data we have seen that those tend to be the tumors that have the most brisk expression of VEGF. So, there may be a subclass of stage I lung cancer patients that do poorly because they have several important pro-angiogenic factors turned on early.

TOP

This was data where we looked at increasing expression of COX-2 in these different stage I lung tumors. As Ethan showed you can probably break this down even further. The patients with the most abundant expression of COX-2, a small number of patients, three patients who had 50% or more of their cells overexpressing COX-2 also had RAR beta expression in abundance intratumorally. They also had strong VEGF and all three of those patients, all stage I tumors had extrathoracic metastases within 9 months.

TOP

So, for me this helps us to define a subpopulation of patients who probably are presenting with either occult metastatic disease or very strongly pro-angiogenic factors in their tumor that will ultimately determine their outcome. That is why I think with all due respect to the surgeons and the radiation oncologists in the audience, you need us because these patients die of systemic disease.

TOP

Now, in terms of the MMPs this was a very hot area a few years ago and remains a very provocative area. We know that MMP2, MMP9 and MMP14 as well as MMP11 are all expressed to a significant degree and resected in non-small cell lung cancers.

TOP

And these are probably by my review of the literature the three most important MMPs at least in lung cancer, MMP2 which is important for gellatinase A MMP 9 and MMP14 which by a loop is able to activate MMP2 expression, (Cont'd)

TOP

...and there have been a number of trials of MMP2 inhibitors in lung cancer. Unfortunately they have been disappointing. The companies that have developed these agents have tossed back and forth between looking at these agents in early stage disease or, as some other drugs have been developed, in late stage disease in some tumors. They have also debated whether to develop them in trials which strongly overexpress certain MMP inhibitors versus trials in all tumors. I think we know that the results so far have been disappointing.

TOP

This trial by Dr. Smiley, et al, was presented at ASCO this year showing that there was no dose response relationship to the MMP inhibitor Prinomastat versus placebo. There is absolutely no difference in overall survival except maybe a slight trend favoring the placebo and so based on this I don't see an abundant development of MMP inhibitors at this current time even in patients with advanced non-small cell lung cancer. When you see this kind of data in more advanced disease you can understand some of our pharmaceutical partners' reluctance to push this in earlier stage disease.

TOP

What about the agent that probably has had the biggest splash in lung cancer? By my bet that would be the recombinant monoclonal antibody, the VEGF from Genentech. This was humanized to try to avoid immunogenicity, and it seems to be a pan-VEGF antagonist and that is, also, a point of interest.There have been several recent publications that suggest that selective VEGF isomers do have highly selective roles in invasion in lung cancer, perhaps VEGF9 and VEGF8.

TOP

This was the trial that Russ Devore, Alan Sandler and Dave Johnson and others conducted. It was a randomized trial of Taxol and carboplatin versus Taxol and carboplatin with an intermediate dose of recombinant anti-VEGF, 7.5 milligrams per kilogram versus high dose. This was by no means a definitive trial. It was a Phase II trial and the major endpoint was time to progression.They saw some very interesting results which they ultimately categorized as responses.

TOP

For example, this is the lesion that by classic criteria would not have counted as a response because the volume is the same, but as you can see after three cycles this is fully cavitated.

TOP

If you look at time to progression you can see that time to progression was significantly better for the high-dose anti-VEGF than it was for the control. The next curve will demonstrate that.

TOP

If you look at those patients and include the toxicities that are most likely to occur here, including GI bleed, epistaxis, hemoptysis you can see a significant number of hemoptysis taking place in the patients who got the high dose and intermediate dose VEGF as well as significant epistaxis. This seemed to occur disproportionately in those patients who had squamous cell tumors. Now, is there a mechanistic reason underpinning this toxicity or is it just that we tend to see squamous cell cancer as central lesions that generally can cavitate into central vessels? It is not entirely clear to me. At any rate the next trial will basically be designed to look at non-squamous cell tumors and I believe was in the process of being approved at some point by CTEP.

TOP

So, the next class of compounds that I thought was very interesting, and we will have the slides up here eventually to show you the SUGEN compounds. I just want to caution that much of the data for the SUGEN compounds has actually been in colorectal cancer and in melanoma and they have seen with at least SU5416 three complete responses in advanced colorectal cancer, and certainly that is encouraging data. However, in searching the literature for lung cancer responses those are quite rare.

TOP

So, this is SU5416. It is a synthetic small molecule which is an angiogenesis signalling inhibitor. It is fairly specific for inhibition of VEGF. It is administered twice weekly by intravenous infusion. It is pretty well tolerated. The headaches are DLT. Now, most importantly what we are looking for with many of these agents is safety, and we are looking for a safety database that is reassuring. This drug has been given to about 600 patients now, and is in Phase II, III clinical trials and has shown clinical activity in multiple tumor types although as I related lung cancer is not one of those tumor types where we have seen some expression yet.

TOP

Now, this is some limited data culled from the literature from the Mayo Clinic in Roswell Park. The pilot study is in colorectal cancer with various different dosing schedules twice weekly here, small numbers but impressively they saw three complete responses.

TOP

One thing that Paul and others mentioned is that in early stage disease you need a compound, particularly if you have a fully resected tumor and you are in the chemopreventive setting with reasonable tolerability. You don't just need efficacy, but you need a reasonable degree of tolerability. Basically if you look at this agent's tolerability profile you can see that there are in some cases with certain schedules significant numbers of patients with nausea and vomiting. For me this is probably going to be difficult to administer an intravenous compound to early stage disease unless these are medically inoperable patients.

TOP

Now, SU6668, interesting choice of numbers, basically is more of a pan-tyrosine kinase inhibitor, inhibits the TK of VEGF, BFGF and also PDGF. This has been tested and has significant activity in a number of squamous cell and other tumors including tumor regression and at a minimum tumor stasis. And as you can see what is nice about this molecule. You can see a very nice dose response effect at doses that are potentially tolerable, and this agent is in clinical trials, the encouraging thing being that some of these early clinical trials have included patients with non-small cell lung cancer where there has been at least the suggestion of prolongation of time to progression. So, I think this will probably be an agent that is of some interest.

TOP

We started with Endostatin and I am going to show you some of the interesting data with Endostatin while basically pointing out the disclaimer that I am not sure that this agent is quite ready for prime time. Preclinically this looked extremely interesting. What it can do is it can essentially induce a state of tumor dormancy. This is a basically 20 kilodalton C terminal fragment of collagen 18. It is an endothelial cell specific agent. So, you can avoid many of the toxicities that you get when you use classic cytotoxic agents that do not distinguish between normal cells and tumor cells. In some studies by Dr. Folkman and Mike O'Reilly and others it induced a complete regression in animals and prolonged tumor dormancy until the drug was discontinued and then with re-treatment there was no evidence of resistance, so, a very attractive agent. Because you are targeting endothelial cells there is no drug resistance seen and they saw evidence not only of inhibition of angiogenesis but of significant endothelial cell apoptosis.

TOP

This was the M.D. Anderson Endostatin trial. It was a biopsy-driven trial in which patients basically were given the Endostatin daily as 20 minute infusions, and with biopsies at baseline and at 8 weeks. They had multiple imaging studies. I am told the study cost M.D. Anderson about $1.3 million. It is always nice to work with the NCI at our expense.

TOP

So this was a very, very interesting and exciting study, in that Roy Herbst and Jim Abbruzzese and others were able to show that even though none of us would define this as an active agent they were able to at least learn a lot about the sort of non-invasive imaging that I think we are going to need in order to develop these agents further.

TOP

Unfortunately, did the drug work? I will let you make your own conclusions. I am sorry this doesn't show up. Red seems to be an unpopular color here, but as you can see time to progression did not really change when you went from dose level to dose level. These patients essentially generally progressed within 8 weeks. Not all of them were refractory. The requirement was really that they had to have biopsy accessible disease.

TOP

But we did learn a lot in terms of how do you image these patients, and if you look at the imaging by PET scan here you can see for example that this individual whose tumor was growing slowly you could tell that about 8 weeks later or 4 weeks later, or within 4 weeks and 8 weeks you had increased glucose metabolism. This actually presaged overt progression and increased vascular flow to the tumor which indicated that perhaps the Endostatin was not accomplishing what we wanted.

TOP

On the other hand in this woman with melanoma who had a mixed response you see some more interesting things at one of the higher doses. This lady had, at least in this pancreatic lesion, decreasing glucose metabolism quantified over one cycle and two cycles of Endostatin as well as decreasing blood flow. I think this is the kind of vascular and PET imaging that we need to further develop, not only the farnesyltransferase inhibitors and EGFR inhibitors and retinoids but all agents that potentially impact on targets such as tumor vasculature.

TOP

One more point to make. Why is it critical to take a systemic approach to patients with what seemed to be early stage disease? This is a very elegant study by Dr. Cote that he published 3 years ago in JCO looking at bone marrow aspirations. You could see that even in early stage patients if they were CK18 positive they tended to have a far more morbid and mortal course. So, many of these patients whom we and our surgeons think of as pathologic T1 tumors actually have bone marrow metastases indicating again and again what we have heard over the last 2 days that lung cancer is a systemic disease and requires multidisciplinary involvement and probably multi-agent involvement.

TOP

So, in my opinion summarizing the antiangiogenic activity in non-small cell lung cancer this hasn't been a major focus of antiangiogenic drug development. There have been some promising results with the anti-VEGF compound in combination with chemotherapy, but unexpected toxicities. I don't think any of us would have anticipated a high percentage of patients having significant hemorrhage. I think there is potential for non-invasive imaging shown by the Endostatin trial. Where are we going with these agents? I think that if we are going to look in advanced disease we need further combination studies and we probably need combination studies even in early stage disease. My increasing sense of it is that these are not stand-alone agents. We need more safety data about the anti-VEGF compounds, about Endostatin and TNP and other molecules, and we need better surrogate markers of response, not just the non-invasive radiologic but the other compounds.

TOP

And in tipping my hat to my own mentor, Ki Hong, I ultimately think that for the treatment of these patients they are going to have to head down not just the chemotherapy aisle but the chemoprevention aisle of the practicing medical oncologist.
Thank you very much.

TOP

[Discussion not Transcribed]

TOP