What Are the Most Promising Novel Targets for Therapy? by Kenneth Foon, MD

SLIDES & TRANSCRIPTS
Tuesday, February 15, 2000

What Are the Most Promising Novel Targets for Therapy?
Kenneth Foon, MD

Slide 1:

DR. ERLICHMAN: Thank you, Alex.

First of all I would like to thank Dr. Petrelli for his excellent audiovisual skills. When I asked the speakers to come up and talk on their topics for 15 minutes I knew I was giving them a challenge to talk about such a large topic in 15 minutes, but I didn't think I was also going to give them the audiovisual challenge.

At any rate, it is my pleasure now to introduce Dr. Ken Foon who is known to many of you. He is the Director of the Barrett Cancer Center at the University of Cincinnati. He has long experience in the field of immunology and immunotherapy as it pertains to cancer, and he is going to address some of the issues and some of the areas that are pertinent to colon cancer.

DR. FOON: Thank you, Chuck, and thank you for inviting me.

Chuck asked me to review the topic of immunotherapy in colon cancer, and he said that I had 15 minutes, and Bob Mayer as I walked in saw my slides, and he almost passed out, and he said, "You have 7 minutes." So I pared this down to about 10 or 11 slides. And I will be talking very fast. I don't know if it is 7 or 15, but I won't talk quite as fast, but the immunotherapy of colon cancer could be divided into serotherapy, and that really is limited to monoclonal antibody infusional therapy, and although there are some radiolabeled antibody trials in Phase I type trials, I think the only real important study is the 17-1A study, and I will briefly talk about that.

Cytokine therapy really has not taken off in colon cancer, and I don't think there is much of a topic here. Finally, vaccines I think are probably the most exciting area of immunotherapy in colon cancer.

Slide 2:

The problem with trying to develop vaccine therapies is that tumor antigens are self-antigens for the most part, and not recognized as foreign, and so the vaccines have to break this immune tolerance against tumor antigens,

Slide 3:

and they have to activate the various arms of the immune system

Slide 4:

One approach, of course, and the most popular and the oldest approach is using whole tumor cells. That is either using autologous cells or allogeneic cells or antigen supplemented tumor cells by using viral oncolysates, and these kinds of trials have been done for years.

Slide 5:

I think the most important trial comes out of the trial published last year in the Lancet where 254 patients were randomized to injections postsurgically. These were all B&C patients, Stage II and III patients of autologous tumor cells with BCG.

They saw no benefit for Stage III patients, but the Stage II patients did appear to have a significantly longer recurrence-free period. This trial has not been reproduced, and this trial has the problem that these are autologous cells and quality control is very difficult to reproduce with autologous cells.

Slide 6:

Defined antigens or vaccines that you shed antigens from cultured cells as has commonly been done for malignant melanoma, and I am not aware of trials for colon cancer using gangliosides more typical for malignant melanoma that have been proposed for colon cancer as well. Proteins and peptides and anti-idiotype antibodies I will more or less the talk with because that is my area of interest.

Slide 7:

The 17-1A, while of course it is not a vaccine, is a passive infusion of a mouse monoclonal antibody, 189 patients with Dukes' C colon cancer postsurgically were randomized. This is a German study. They were randomized to no therapy versus 17-1A infusions that included one large 500-milligram infusion and then I believe it was monthly infusions followed by 400-milligram monthly infusions, and the median follow-up is now 7 years. The first paper was published in the Lancet, suggesting a prolonged survival and disease-free interval, and the follow-up in the Journal of Clinical Oncology 7 years later has supported that. So it would appear at least based on this one trial that there was a significantly improved survival in those patients that were treated with 17-1A Dukes' C colon cancer. Remember that these patients were not treated with chemotherapy.

As mentioned by Dr. O'Connell, this study is under investigation in the US in group trials today.

Slide 8:

Other approaches are genetically engineered tumor cells transduced with cytokines, viruses expressing cytokines injected in vivo or viral recombinants carrying tumor antigen genes, and this has been a popular approach primarily from Jeff Schlom's lab.

Slide 9:

This is a study where the CEA gene is expressed in the recombinant vaccinia virus. Patients are immunized with this recombinant vaccinia virus or CEA and they were able to show when they took peripheral blood lymphocytes from these patients that these cells could be stimulated in vitro with the CAP-1 peptide and then they can demonstrate that cells could be generated, CTL that would lyse colon cancer cells that were HLA-2 and expressed CEA, and that was highly specific for that, but also remembering that these are cells that were taken from patients after they were vaccinated, stimulated multiple times in vitro before they actually were able to demonstrate cytolytic T cells. So what that actually means in vivo as an antitumor is another issue.

Slide 10:

Another study from the University of Alabama using the same vaccine, the recombinant vaccinia CEA vaccine showed that CEA induced auto-antibodies, that is antibodies against CEA in 7 of 32 patients, but these were very low titer, low acidity, predominantly IgG1 antibodies.

Slide 11:

In a more recent trial using the canarypox vector, the same concept now using canarypox to express the CEA, they elicited cytotoxicity. The same kind of concept, taking cells from patients, stimulating them in vitro with IL-2 and CAP-1 and demonstrating that the cells would specifically lyse A2 restricted CEA positive tumor cells.

Slide 12:

Another approach would be the dendritic cells. Probably the best work in this area is Ken Lyerly's work from Duke, where dendritic cells are being pulsed with either the CAP-1 peptide or messenger RNA for CEA and they have demonstrated that these dendritic cells can then stimulate PBL in vitro to stimulate cytolytic T cells specifically lysing A-positive tumor cells.

Slide 13:

This has now gone into clinical trials. In the end, what are we looking for? Probably you cannot see this very well from the back. The vaccine, whether it is cells, lysates or proteins is taken up by the antigen presenting cell and is expressed. It processes the protein or expresses the peptide on the cell surface. With the Class II antigens they activate CD4 T cells which secrete the TH2 cells which secrete IL-4 IL-5 and IL-10 which stimulate B cells producing antibodies that have an antitumor effect.

The CD4 T cells and the TH1 cells will secrete IL-2 which helps stimulate the CD8 T cell. Then hopefully with Class I restriction, the T cell receptor will then bind to the tumor cell and kill the tumor cell.

So the avenues would be the CD4 activation, B cell activation, hopefully cytolytic T cell activation.

Slide 14:

Our approach has been using the mirror imaging of the idiotypic network.

Slide 15:

The approach is that we take an antibody that binds to a specific antigen. In this case we are using the 801981 antibody that binds to CEA and our interest here is not the AB1 antibody which could be used theoretically for a serotherapy. We are only interested in the AD1 for its unique binding ability to CEA. We then take the AB1 antibody and inject it into Balb/C syngeneic mice who will generate an anti-idiotypic antibody that binds to the binding region of the AB1. We call this an anti-idiotype or AB2 antibody, and this antibody that binds to the binding region of the AB1 antibody is in fact the internal image of CEA.

So this becomes a surrogate antigen for CEA, a substitute for CEA. CEA itself does not stimulate an immune response in patients. Self-antigen, the immune system doesn't recognize it as foreign.

Our hope was that if we could present the CEA antigen as an anti-idiotype antibody, in a completely different molecular configuration, we could trick the immune system essentially into generating an active immunity against CEA. We showed in mice, rabbits and monkeys that this AB2 antibody that we call 3H1 or CeaVac does generate an anti-CEA, both a CEA-specific humoral immune response and a CEA-specific T cell response in mice, rabbits and monkeys.

We took this into clinical trials with patients with advanced disease and showed in the vast majority of these patients we could break immune tolerance against CEA and generate an active immunity against CEA, both the T cell response and a humoral immune response.

Slide 16:

We chose CEA because of its very high expression in colon cancer. In fact, it is virtually 100 percent expressed in colon cancers and metastatic colon cancers. It is also an intracellular adhesion molecule and may have a key role in invasion and metastasis. So it seemed like an ideal target antigen.

Slide 17:

After we had generated immunity in patients with advanced disease we had always felt that the vaccine approach would be best going into an adjuvant setting.

We wanted to ask the question, could our vaccine generate an active immunity in patients in the adjuvant setting who are being treated with 5-FU leukovorin or in this case we started the study 5 years ago for 5-FU and levamisole.

Slide 18:

So we took patients with resected Stage II, III and IV colon cancer. At least half of these patients were simultaneously treated with 5-FU regimen, and virtually every one of these regimens was a standard 5-FU leukovorin type regimen, and the vaccine was given. We have determined that a 2-milligram dose of the vaccine in our prior study seemed to be a fairly optimal dose, and we gave it. The vaccine itself was either in alum where we were also testing QS21 as a different adjuvant. That is another story. It turns out that both worked very well.

Slide 19:

Looking at our patients in summary, I just updated this a month ago, and the data was published a few months ago in the Journal of Clinical Oncology and just to focus on the C2s and the completely resected Ds, we had eight patients in each of those groups. Six of the eight patients with C2, that is greater than five lymph nodes are still free of disease, at least by CAT scans and CEA, at 21 to 47 months two of them did relapse, and 19 in 24 months.

Of the resected Stage IV D patients five of the eight are still disease free from 17 to 40 months. Three of them have progressed in 9 to 24 months. We have one patient even with the incompletely resected disease -- all of these patients have positive surgical margins -- who still has no progression of the disease at 21 months, although we know there is disease based on the surgical report. Eight of those patients have relapsed.

These are compelling data, and in fact they are very interesting data, but of course, they are nothing more than anecdotes until a Phase III trial is carried out.

Slide 20:

The toxicity has been absolutely minimal. We see a local reaction, possibly some minimal systemic effects.

Slide 21:

All 32 patients that we reported on have generated a very high titer polyclonal antibody, anti-CEA response.

I am not going to show any of this data because there isn't time, but it has been as has been published in JACL.

The AB1 response we quantitate at 58 to 350 micrograms per ml. So it is a huge polyclonal antibody response. It is a mixture of all of these subclasses, IgG-1, 2, 3, and 4, a predominance of IgG-1 and IgG-4. We also see some IgM. We have a TH1 CD4 proliferative response in every single one of these patients, and all these patients' serums generated an antibody-dependent cellular cytotoxicity as a mechanism to kill the tumor cells.

So with we felt compelling clinical data -- I shouldn't say "we" I guess B it is a large we -- but the American College of Surgeons with what they felt was compelling clinical data and certainly very strong immunologic data, I know of no other study where we can project that 100 percent of patients and half of these patients were on 5-FU regimens, will generate a very strong immune response.

Slide 22:

The following trial has been proposed, and that is the American College of Surgeons Oncology Group take Dukes' C patients and randomize them to 5-FU leukovorin versus 5-FU leukovorin plus our vaccine.

Thank you.

(Applause.)