Angiogenesis as a Target for Cancer Therapy by Beverly Teicher, PhD

SLIDES & TRANSCRIPTS
Tuesday, September 14, 2000

Angiogenesis as a Target for Cancer Therapy
Beverly A. Teicher, PhD

Slide 1:

DR. SAXMAN: Thank you.

The last speaker for this morning is Dr. Beverly Teicher. Dr. Teicher is well known in the field of angiogenesis and is currently doing oncology research at Eli Lilly and Company in Indianapolis.

DR. TEICHER: Thank you, Scott. It is very nice to be here. I am not often asked to give introductory talks. I usually stick pretty close to the data. So, in preparing this talk I have borrowed illustrations from many of the very good reviews that colleagues have prepared in this area of angiogenesis lately. Additionally I have looked for some illustrations in the area of small cell lung cancer.

Slide 2:

The process that we are talking about is angiogenesis. As we all know, in order for tumors to grow beyond the size of 2 cubic millimeters in volume, they must develop a vasculature so that, as the tumor cell population grows, the malignant cells secrete angiogenic factors which stimulate nearby vasculature endothelial cells to proliferate and to form tubes and capillaries to promote the continued growth and expansion of the tumor.

This new blood vessel formation also allows tumor cells to extravasate into the bloodstream, travel through the bloodstream , establish colonies again at distal sites, and grow metastatic disease and establish a vasculature once again.

Slide 3:

This illustration is from a perspective written by Jones and Harris which illustrates the excellent science that has gone into elucidating the process of angiogenesis in recent years. One of the important recent findings is the recognition that there are angiogenic factors occurring naturally and angiogenic inhibitory factors which occur naturally.

Among the most important and widely studied angiogenic factors in the tumor are vascular endothelial growth factor which has now been recognized to be a family of proteins, VEGF-C being the most prominent studied to date in tumors. Basic fibroblast growth factor, the angiogenic enzyme, thymidine phosphorylase, also known as platelet-derived endothelial growth factor angiopoietin-1 which, although it does not induce proliferation of endothelial cells in cell culture, is recognized to induce the stabilization of new blood vessels as they grow in vivo. Along with the elucidation of these various angiogenic factors has come the elucidation of the receptors for these factors on the endothelial cells, and these receptors have certainly become prominent targets for anti-angiogenic agent development.

The VEGF receptor FlT-1 which is VEGF receptor one is currently being targeted in the clinic. The VEGF receptor 2, which is also called KDR or Flk-1 for the murine form, is one of the most prominent targets for new agents in the clinic.

Tie-2 is the receptor for angiopoietin-1. It is also recognized to be the receptor for angiopoietin-2, which is one of the naturally occurring angiogenic inhibitory factors. Among the other very prominent factors that are targeted in angiogenesis are the matrix metalloproteases through matrix metalloprotease inhibitors. Also, the adhesion molecules involved in the angiogenic properties of basic fibroblast growth factor, the integrins, Alpha-B, Beta-3 and Alpha-V, Beta-5.

Among the naturally occurring angiogenic inhibitors, thrombospondin is probably the most powerful and potent agent recognized at this point. As I mentioned, angiopoietin-2 and of course, the proteins endostatin, angiostatin, and an ever-growing list as we read in the biotech literature.

Slide 4:

In terms of targets, boiling this down to something a little simpler: in the endothelial cells the extracellular matrix proteases are certainly valid targets. The adhesion molecules, Alpha V, Beta-3 and Alpha V, Beta-5, and the growth factors and the growth factor receptors. There are also agents which directly inhibit the proliferation of endothelial cells.

Slide 5:

Why is vascular targeting an interesting and different target in the treatment of malignant disease? It is believed that agents which target the angiogenic process may have a good hope of working because we are looking at a normal cell and a normal process within the body. Therefore, it may be a more stable process in terms of a therapeutic attack.

It is recognized that the endothelium is a single compartment, that the cells involved are not malignant and therefore, perhaps are genetically more stable. It is a good target because the cells are accessible. Therefore protein molecules which may have trouble penetrating through cell layers in tumors may find a good therapeutic target here.

There are many angiogenic markers which are recognized. There have been no reports as yet, as far as I know, of the development of resistance of endothelial cells to any of the agents that are being studied so that the vascular compartment is an interesting and perhaps a more accessible target for malignant disease than the tumor compartment or the malignant cell compartment, which we have been looking at over these years, although there are many redundant processes involved in angiogenesis. So, we are going to have troubles there, I think.

Slide 6:

In looking at the kinetics of response of vasculature to anti-angiogenic therapies, I think that is a very important thing to keep in mind in terms of therapeutic approaches.

When we treat malignant cells, we treat acutely. These cells are proliferating rapidly or relatively rapidly, and we hope acutely for a response in the tumor.

This doesn't often happen in terms of angiogenesis, but if we are looking at a tumor inhibitory situation where the micrometastatic disease has not yet begun the angiogenic process, we may be able to treat micrometastatic disease and hope for quick responses. In most situations, it is likely that antiangiogenic therapy will have to be given chronically or semichronically and that the regression of that vasculature is going to be a relatively slow process. So longer treatment times and more frequent therapy may be necessary.

Slide 7:

Among the agents which are currently in clinical trial, that have been described under the very broad umbrella of antiangiogenic agents, I have several slide examples, and these are from the NCI database.

Marimastat, these are matrix metalloprotease inhibitors. Marimastat is well into Phase III randomized trials. It is a matrix metalloprotease inhibitor. The Bayer compound, Bay 125966, also, is well into Phase III trials, the Agouron compound, AG-3340. I am sure that many people in this room have worked with these agents or are currently working with these agents: COL 3, neovastat and the Bristol-Myers compound.

Slide 8:

Among agents which inhibit the proliferation of endothelial cells directly are TNP-470 which is in Phase II clinical trials, both alone and in combination with cytotoxic therapies, thalidomide, squalamine, combretastatin-4A prodrug, and endostatin. Phase I clinical trials of endostatin should start, I believe, next one to two weeks.

Slide 9:

Among agents which block the signal transduction of the angiogenic signal are the VEGF antibodies in Phase II clinical trials. The two Sugen compounds, SU5416 and SU6668, these are compounds which block the VEGF receptor signaling of the KDR receptor by binding to the ATP site in this molecule.

Sugen compound 5416, as you know, is given intravenously, and the Sugen compound 6668 is given orally. Also, the Novartis compounds and interferon alpha continue to be explored as a potential antiangiogenic agent.

Slide 10:

Agents which are involved in the inhibition of the integrin signaling pathway include vitexin, the human form of LM609 which is an antibody to the integrin receptor which is in Phase II clinical trials, as well as a compound from Merck which I am not familiar with, EDM 121974, which is a small molecule integrin blocking agent.

From the NCI, CAI continues to be explored as an inhibitor of calcium influx. The cytokine interleukin 12 may act directly as an antiangiogenic agent, but it is its up regulation of interferon gamma that is believed to be intimately involved in the mechanism of action of that agent, and IM 862.

<-a>Slide 11:

Well, is small cell lung cancer a particularly good target for antiangiogenic agents? I think most investigators would say that it is, because it responds well to chemotherapy initially, and many investigations have taken advantage of that response to then come on with maintenance therapy using the anti-angiogenic factors.

Slide 12:

I have searched a bit of literature looking for angiogenic studies involving small cell lung cancer, and this report from Salvin in the Mattson group was one that was particularly interesting where they looked at serum levels of vascular endothelial growth factor in small cell lung cancer patients and separated the groups as having serum levels of VEGF less than the mean or greater than the mean and showing that those patients that had lower levels of VEGF had a better survival. They also then divided the expression of VEGF by stage of limited disease and extensive disease and again found that the limited disease patients did better. When they divided their patients by stage and serum levels of vascular endothelial growth factor, they found that the limited stage disease patients with low levels of VEGF in the serum did best.

Slide 13:

There have also been several reports looking at microvessel counting in patients with lung cancer. In this paper from Poland there was a group of small cell lung cancer patients that were examined. They found that although there were clearly good levels of micro-vessel counts in all of the different types of lung cancer that the samples with small cell lung cancer had the highest level of microvascular counts.

Slide 14:

In looking into the literature for cases where antiangiogenic therapy may have been tried and not called that at the time in small cell lung cancer, again the studies by Mattson from Finland showed quite a lot of work with interferon alpha. In this particular study, interferon alpha was used as a maintenance therapy after chemotherapy in small cell lung cancer patients or CAP was used as a maintenance therapy or there was no maintenance therapy. As you can see, although there is not a big difference in the survival of patients here, the patients who receive maintenance therapy of interferon alpha did do better than the other patients.

In subsequent studies, the same group has looked at interferon alpha in conjunction with chemotherapy and found that because of toxicity of those combination therapies that that strategy did not work as well.

Slide 15:

There are several large clinical trials, which many of you are probably participating in, of Marimastat in small cell lung cancer patients in this randomized Phase III trial which is well under way.

Slide 16:

With the Bayer compound, the biphenyl matrix metalloprotease inhibitor, a similar randomized Phase III trial in small cell lung cancer is also well under way.

Slide 17:

As it turns out, lung cancer has had a very prominent role preclinically in the study of angiogenesis,and this is a non-small cell lung cancer model, the Lewis lung carcinoma which as you know has figured very prominently in the study of angiogenesis agents at the preclinical level. I would like to just go briefly through some of this data.

In this study from my laboratory, we looked at AGM 1470, which is an older name for TNP 470 as an endothelial cell proliferation inhibitor, and the tetracycline matrix metalloprotease inhibitor minocycline as a combination antiangiogenic therapy.

We examined several different schedules of administration of this antiangiogenic combination in conjunction with cytotoxic therapy in this tumor. This is the classical Lewis lung carcinoma, which metastasizes avidly to the lungs from the subcutaneous implant.

The most effective chemotherapeutic agent against the Lewis lung carcinoma is cyclophosphamide, which produces about 20 days of tumor growth delay in this tumor.

We then looked at the antiangiogenic combination, starting therapy very early in the life of the tumor on day 4, when the tumor is just a seed and beginning to explode in its angiogenic activity. We are starting the angiogenic agent combination 3 days later on day 7, when the tumor is actually a fairly well-established nodule. The cytotoxic chemotherapy was administered on days 7-11.

So we learned a lesson that cancer researchers learn again and again, and that is the tumor burden is very important. If we started the antiangiogenic therapy early on day 4 and treated at days 4-11 or 4-18, we obtained the greatest enhancement in tumor growth delay, but even if we had to limit the antiangiogenic therapy to the same 5-day period that we gave the cytotoxic therapy, we still had tumor growth delay of 29 days, which was better than cytotoxic chemotherapy alone.

It was only when we administered the antiangiogenic therapy for the full 2-week period of days 4-18, which is really the full exponential growth phase of this tumor, that we obtained the greatest tumor response and with this therapeutic regimen. Forty to fifty percent of the animals were cured of the Lewis lung carcinoma.

Slide 18:

From these same animals we counted lung metastases on day 20. The untreated control animals bearing this tumor begin dying from lung metastases on days 21 through 25.

So day 20 is very late in the disease of this particular tumor. The control animals had a mean number of about 20 lung metastases on the external surface per lung. The antiangiogenic therapies alone here, the TMP 470 plus minocycline, did not impact on the number of lung metastases that we saw in these animals and actually only had a modest impact, the numbers in parentheses being the percent of mets that are large enough to be vascularized. It only had a modest effect in decreasing the percentage of vascularized lung mets in these animals.

The cytotoxic agent cyclophosphamide decreased the number of lung metastases to 12 from 20, but when we gave the combination of the antiangiogenic therapy along with cyclophosphamide, we saw a marked reduction in number of lung metastases, so that in the group that received 2 weeks of therapy, many of the animals had no lung metastases. So, there was only a mean number of two mets, and again, about half of those animals were cured.

Slide 19:

Based on these studies and others, we concluded that antiangiogenic agents can potentiate cytotoxic therapies including both chemotherapy and radiation therapy.

I can only speak to the antiangiogenic agents that I have tested, and there is an ever-growing number of those that I have not tested. So I cannot make this as a general statement across the board, only for those that I have studied.

Slide 20:

At the time these data were generated, it seemed very unusual to many people that if you decrease the vasculature, if you inhibit the growth of vasculature in the tumor, that you could potentiate therapies which must be delivered through the vasculature and through cell layers to the malignant cells in the tumor.

Slide 21:

So we decided to address that question, and to address that question we used the assay illustrated here. We grew tumors in animals, Lewis lung carcinoma among others, and treated those animals with various anti-angiogenic therapies for 5 days to a week and then treated them with a single dose of cytotoxic therapies. The next day, we injected a trace amount of a fluorescent dye into the animals intravenously, Hoescht 33342, and 20 minutes later excised the tumors, prepared a single cell suspension and were able to obtain a fluorescent gradient of single cells from the tumor.

This was an assay originally developed by Ralph Duran and Dye Chaplan at the British Columbia Cancer Institute and was set up in my laboratory by Dr. Sylvia Holden.

From these cells we did a sterile sort, sorting the 10 percent brightest cells, which we believed to be enriched in populations of tumor cells that were near tumor vasculature, and the 20 percent dimmer cells, which we believed to be enriched in tumor cell populations that were distal from the vasculature at the time that we did the experiment.

Slide 22:

We plated these cells in culture and did a colony formation assay to determine how many of the cells were killed in each of those subpopulations.

We were doing this experiment for many treatment modalities, many chemotherapeutic agents, radiation and hyperthermia, and each time we did it, one control was to examine the fluorescence distribution in the control tumors, versus those from animals that had received various therapies, to see whether or not the therapy we administered altered the distribution of the fluorescent dye. We never saw a change in the distribution of the fluorescent dye in the tumors until we had animals that we treated with TMP 470 and minocycline, an antiangiogenic combination.

In those tumors, we saw a much greater uptake of the dye, much higher fluorescence and intensity throughout the tumor cell population, so that the 10 percent brightest cells in the antiangiogenic treated tumors were much brighter than the bright cells from the control tumors. The 20 percent dim cells in the antiangiogenic treated animals were much brighter than the dim cells from the control tumors. This shift applied even to the mean of the population, so that altogether there was about a fivefold greater uptake of fluorescent dyes into the tumors of the antiangiogenic treated animals.

Slide 23:

At that time, we were able to measure platinum by atomic absorption. We decided to look at platinum uptake from a single injection of cisplatin in animals treated with antiangiogenic therapy, compared to animals who did not receive antiangiogenic therapy.

The open bars here are the relative level of platinum in animals receiving only cis-platinum. In the striped bars are the level of platinum in the tumors of animals treated with TMP 470 and minocycline for 5 days prior to platinum administration. We found an increase of seven-fold in the level of platinum in the tumors of animals that received the antiangiogenic combination.

We also saw increases in platinum levels in liver, kidney, brain, heart, gut, skin, muscle and lungs of these animals.

Slide 24:

We were fortunate at that time to have some C14 labeled cyclosphosphamide in the lab and did the same sort of experiment with administration of C14 cyclophosphamide. In that case, we saw about a three-to-four-fold increase in C14 levels in the tumors of animals treated with antiangiogenic therapy and then radiolabeled cyclophosphamide. We saw also increases in C14 levels in liver, kidney, brain, heart, gut, skin, muscle and lungs of those animals.

Slide 25:

We have now carried this experiment out with—in addition to cyclophosphamide and cisplatin—with carbo-platinum and paclitaxel. We also saw increases in oxygenation of tumors in animals treated with antiangiogenic agents. Of course, we have the initial observation of the increased level of the fluorescent dye in those tumors.

Slide 26:

What is happening to the vasculature in these tumors? Here is vascular staining of the Lewis lung carcinoma stained with CD31 or with factor 8, the control tumors and tumors from animals that had been treated with TMP470 and minocycline. As you can see, there is a decreased amount of staining in the red color here of vasculature in the antiangiogenic agent treated tumors.

Slide 27:

I was very fortunate to have a visiting scientist in my lab from Japan, Yoshi Kagegi, who counted intratumoral vessels in Lewis lung carcinoma samples after treatment with a variety of potential anti-angiogenic agents. That was TNP470, TNP470 and minocycline, Suramin, Suramin and TNP470, genestein and TNP470 and genestine. We stained samples with either CD31 or factor 8. Yoshi found that with each of these antiangiogenic combinations there was a decrease in the number of countable, stainable intratumoral vessels to about one-half to one-third the number in the control tumors, both with CD31 staining and with factor 8 staining.

That is the last slide.

Slide 28:

So, in conclusion, small cell lung cancer certainly is a very good disease target for antiangiogenic therapies, and I am sure there will be lots of discussion of this in the breakout sessions.

Thank you.

Slide 29:

Question and Answer

DR. SAXMAN: There is time for a question of two.

DR. GANDARA: Why do you think there was increased blood flow? Actually the question is increased blood concentration with the agents, the cyclophosphamide, and the blood flow?

DR. TEICHER: No, but Rockesh Jayne has done quite a lot of nice work along those lines, and Rockesh was very concerned about that data when he first showed it. When he did studies in the window chamber, where he has a model where he grows tumors in the window chamber and can follow the growth of the vessels and also look at fluorescence distribution of fluorescently labeled cells or proteins or small molecules in that system, what he found was that the tumor blood vessels, which are pretty lousy to begin with, become even worse in structure, and in some cases there are no endothelial cells along the channels. He called them channels where blood is flowing, and in fact these blood vessels are very, very leaky. So there is a very rapid diffusion, a more rapid diffusion of small molecules and proteins into tumors that were treated with these agents and again, I cannot say that this is true for all antiangiogenic agents. There is such a wide variety of targets, I can only say for the ones that I have worked with.

DR. GANDARA: There was, also, increased concentration of chemotherapeutic agents in the tissue.

DR. TEICHER: That is correct, and I feel it is important to show that data for two reasons. One is that the caveat is that the mouse that I am working with is a rapidly growing host. I am working with 6-to-8-week old animals. So these animals can actually grow over the course of the experiment. So they have growing vasculature throughout their bodies, but also it is true that there is a potential of these types of things increasing the whole body toxicity, normal tissue toxicity of cytotoxic therapies, and I found that to be true when we attempted to use this strategy in the high dose setting. There was increased toxicity.