[No text is associated with this slide].

TOP

As Dr. Buzaid mentioned, the AJCC divided stage IV melanoma into three subgroups,

TOP

M1A, M1B, M1C, which do slightly differ but, except for the M1A group, there are very few long-term survivors, and we are not sure that all of these patients in this M1A group who just had soft tissue metastases, actually had stage IV disease.

 

TOP

I think it is fair to say, in looking at the treatment of stage IV melanoma, that there is no therapy that is shown to have a survival advantage. Therefore, the results may be as influenced as much by patient selection as by treatment approach.

TOP

With that as a background, let's go into immunotherapy. This is a cartoon from Dr. Lotze, which shows the adaptive anti-adaptor tumor immune response which starts with an innate immune system killing some tumor cells, activate NK cells, killing tumor cells, tumor antigen being released, picked up by dendritic cells, activating T cells.

As you can see, in various spots along the way, cytokines play a role. In order to get an anti-tumor response that eradicates the tumor, you need to have a sustained TH1 response.

However, the immune system doesn't really work that way and actually goes to great lengths to shut off itself in order to avoid excessive toxicity. So, this is a problem that has to be overcome.

TOP

So, these are the players, the cytokines that Vern asked me to talk about, IL-2, interferon, GM-CSF, IL-12, IL-18.

They have varied mechanisms of action, as highlighted from the prior slide. Some of them expand T cells, some have anti-proliferative, some anti-angiogenic effects, GM-CSF activates DCs.

Some have already been FDA approved, such as IL-2 and interferon, but none have any significant anti-tumor activity greater than 15 percent, and the ones that we know the most about their anti-tumor activity are also the ones that are the most toxic.

TOP

So, this is the data for high dose IL-2 that led to its approval, a response rate of 15 percent. Durable responses, as shown here in these response duration curves, where the median duration of response was about nine months and, for the complete responders, it has not yet been reached.
There are no patients who responded who progressed after the 30 month time point, indicating that most of these patients out here are probably cured.

TOP

In looking at these responders, the response did not appear to be related to tumor burden or sites of disease, different from what we would have expected with a stage IV population treated with an immunotherapy.

The median tumor burden was actually quite sizeable with 65 percent of these responders having visceral involvement, including 35 percent beyond the lung.

TOP

This is the survival curve, median survival 12 months, and 11 percent of these patients were alive at five years.

TOP

In looking at high dose IL-2, I think that, although it appears to be useful, it is toxic in patients, expensive and impractical. Therefore, its use is limited to selected patients treated at experienced centers.

The caveat is that it is uncertain whether this data will hold up in an era of heavy prior interferon use, and the data from the NCI surgery branch actually suggested that response rates were about half in patients who had received prior interferon.

TOP

So, what about other agents and combinations of cytokines?

TOP

Well, IL-2 and interferon, a combination that was studied extensively in the late 1980s, early 1990s, doesn't do any better and, at low doses, is pretty inactive.

 

TOP

What about GM-CSF. Well, there is data from Lynn Spitler that suggests if you compare CM-CSF to historical controls, you see an improved disease free survival,

 

TOP

and improved survival, assuming you pick the correct historical controls.

As we have heard from the staging studies, if you just stage people different, any therapy, including just time or placebo, can be better than historical controls.

TOP

That has prompted ECOG to carry out a confirmatory trial of this data, a trial that is led by David Lawson looking at HLA-A2 negative patients who have stage IV disease resected to NED, GM-CSF or a placebo and in patients who are HLA-A2 positive, they have a peptide vaccine added to the mix, but the major comparison is GM-CSF versus placebo.

TOP

In data from Steve O'Day looking at maintenance after biochemotherapy combining GM-CSF with IL-2, it appears that there was an impressive overall survival in this group of patients of 18-1/2 months and time to progression of 8.3 months, indicating that there may be some value in adding GM-CSF to IL-2 in patients who have obtained a response to biochemotherapy, something that Larry Flaherty may mention in the following talk.

TOP

What about IL-12? There were a lot of studies with IL-12 carried out in melanoma and kidney cancer, and there were very few responses. As shown here, a rare response.

TOP

Part of the reason is shown here, that within a few weeks of receiving IL-12, you can no longer make gamma interferon in response to IL-12.

TOP

What Jared Gallo(?), my former colleague, hypothesized was that that was because of these monocytes and macrophages no longer making IL-15 or IL-18 in response to IL-12, and if you gave IL-2, you could substitute for that lost IL-15 and induce these various reactions.

TOP

In a trial that we carried out, we gave IL-12 by itself for a while and then, at week three, added IL-2 to IL-12, to see if we could restore the ability to make gamma interferon.

TOP

As shown here on this slide, the first dose, a lot of gamma interferon, by the time you got to the third week, as shown here in the open circles, you are no longer making gamma interferon.

When you add IL-2, the gamma interferon comes back up and is maintained even at week six, when IL-2 is added to IL-12.

TOP

What you see with this combination are some patients, including this patient with metastatic melanoma, seeing impressive tumor shrinkage.

TOP

This patient who had cutaneous metastases showing a regression of the melanoma cells and replacement by a lymphocytic infiltrate.

TOP

In a study that is being carried out by Bill Carson in the CALGB, they are looking at IL-12 combined with alpha interferon.

The hypothesis is that IL-12 induced gamma interferon up-regulates levels of JAK-STAT signalling intermediates in PBMCs and tumor cells. Therefore, this may result in enhanced sensitization to low dose interferon, something that is being studied in CALGB, and we will know whether that is, in fact, the case soon.

TOP

What about IL-18? Well, we have initiated a trial with five days of IL-18 in patients with various types of tumor.

This is the first melanoma patient who was treated on IL-18, and this patient has shown a regression of her lung metastases, as well as a regression of her cutaneous and subcutaneous metastases. So, we are batting 1000 percent in melanoma with IL-18.

One solution would be to stop. Another solution would be to try other patients. So, we are looking for other patients once we get the IL-18 back to our clinic, so that we can see whether or not this is, in fact, something that we can see more frequently.

TOP

If enhancing gamma interferon release is what is important, than it might be worth considering combinations of IL-18 and IL-12, which have synergistic effects on gamma interferon release, and hopefully won't prove too toxic.

TOP

So, I am going to spend the remainder of my talk talking about this particular question. Why can't we get over the 15 percent response hump in patients with melanoma using cytokine therapy.

TOP

There are a number of reasons listed here, including inadequate dose.

TOP

As we can see here, low doses of IL-2 or IL-2 and interferon are essentially inactive.

TOP

In data that came out of the NCI surgery branch where, if you look here in the top arrow, is the low affinity receptor for IL-2 and the bottom arrow is the high affinity receptor for IL-2, and you look at various doses of IL-2 and the blood levels that you see, you would probably want to be in between these two arrows where you would activate the high affinity receptor and activate T cells, but not activate the low affinity receptor on NK cells and produce toxicity.

You fall pretty much below this arrow pretty quickly on everything except the high dose IL-2 and that is just in the peripheral blood.

If you really want to activate T cells at the tumor site, you may need higher levels than this in order to get the IL-2 there.

So, it is possible we are not getting enough IL-2 in with therapies that are tolerated.

TOP

That leads to the potential approach of using ways of dissociating toxicity from anti-tumor effects and either blocking here or blocking here or blocking here to block the toxicity without interfering with the anti-tumor effects.

TOP

We have looked at a number of ways of doing this, steroids, pentoxifyllines, cytokine antagonists, selective signal transduction inhibitors, all of which don't do a good job of dissociating toxicity from anti-tumor effects, but there are still some interesting molecules that might be able to be explored including N(methyl)arginine, leukotriene B4 receptor antagonist, and selective superoxide dismutate mimetics, that might potentially allow us to give higher doses of IL-2 without getting into too much toxicity.

TOP

Tachyphylaxis to biologic effects, I have mentioned that we see this in IL-12. We don't know to what extent this is something that is also seen with other cytokines, and it may interfere with our ability to get a sustained immune response.

TOP

Inadequate patient selection, this is something that was brought up by some of the talks this morning.

 

TOP

There are some features that have been shown to predict for response to cytokine based therapy. Patient characteristics include vitiligo, history of autoimmunity, perhaps disease stage, with patients with earlier stages of metastatic disease doing better.

There might also be some tumor characteristics, either gene expression profiling of the tumor cells, or gene expression profiling of the immune infiltrates that may be important, as well as potential features of immune function, either general, such as the TCR zeta chain level, or specific such as T cell reactivity against a tumor. These need to be investigated and this is, in fact, one of the things we are investigating in our spore project.

TOP

I want to call to attention data from Franco Marincola and Ina Wang from the NCI surgery branch, which looked at gene expression profiling and IL-2 response using fine needle aspirates from patients undergoing IL-2 therapy.

They were able to identify 30 genes that were predictive of response. Half related to T cell regulation, particularly interferon response factors, suggesting that immune responsiveness might be predetermined by a tumor microenvironment conducive to immune recognition. This is the type of stuff that we need to look at further when we do immunotherapy studies.

What about inadequate tumor immunogenicity.

TOP

[No text is associated with this slide].

TOP

We know that, if IL-2 works by cytotoxic T cells, and these T cells have to be able to recognize the tumor, and perhaps one way of improving that recognition is to improve the immunogenicity of the tumor or to create a better immune response before you give the cytokines to deal with this particular interaction.

TOP

There are a number of approaches that have looked at this, combining IL-2 with vaccines. The NCI surgery branch saw a 42 percent response rate when they combined high dose IL-2 with GP100 209 2M vaccine.

This has led to a confirmatory trial carried out by a consortium of NCI investigators looking at high dose IL-2 plus or minus vaccines, as well as a trial within the cytokine working group, looking at vaccine plus various schedules of high dose IL-2, where the vaccine is given every three weeks,

TOP

and the IL-2 is either given weeks one and three, seven and nine, or after each vaccination, a trial led by Jeff Sosman.

TOP

In the first part of this trial, where we looked at just cohort three to mirror the NCI surgery branch data, presented by Jared Golab at ASCO, there was a lot response rate, however, indicating that this was unlikely to be the answer.

TOP

Perhaps we need more potent immunogenic approaches, approaches that combine multiple antigens, not just one specific peptide antigen, with cells that are better able to present those cells to the immune system, such as this fusion vaccine that David Avigon and Frank Haluska are going to be investing, the comparison of peptide vaccines as a trial at the DJHCC. If we can generate an immune response, then adding cytokines may be particularly useful.

TOP

Immune suppression, particularly monocyte macrophage-induced immune suppression may be something to look at.

TOP

There is data to suggest that monocytes and macrophages make reactive oxygen metabolites, and that these may inhibit NK and T cell function, and things like histamine may restore that.

TOP

That has been shown to some people's way of thinking in studies combining histamine with low dose IL-2 in organs such as the liver, where there may be a lot of monocytes and macrophages. It appears that that histamine plus IL-2 may work better than low dose IL-2 alone, and this needs to be confirmed with a trial that is undergoing.

TOP

What about tumor induced immune suppression, a very important area?

TOP

There are a number of ways in which tumors inactivate the immune system, such as loss of T cell signalling molecules with the TCR beta chain, the absence of co-simulatory molecules on tumors, HLAG expression, production by tumor cells of inhibitory cytokines, or down modulation of HLA Class I molecules, therefore, leading to the loss of antigen presentation.

TOP

This just shows you that if you present tumor antigens in the wrong context without co-stimulatory molecules, you may result in energy rather than activation.

TOP

More specifically, if tumors have hLAG on their surface, this can activate T and then K cells via ILT2 and induce activated immune cell death in T cells via a CD8 mechanism.

TOP

Just also to show you that patients with advanced disease may lack memory cells, as in this patient compared to a normal, and particularly lack cells that have KIR on their surface, meaning that they may be very naive and may not be able to mount any immune response.

TOP

[No text is associated with this slide].

TOP

What about tumor resistance? Well, this could include loss of interferon signalling molecules, so they can't up-regulate their antigens in response to interferons, absence of any death receptors on the tumor cells -- it was mentioned before, absence of the apoptotic machinery, such as Apaf-1 or procaspase-8, or over expression of ant-apoptotic molecules such as FLIIP and these others mentioned here or, something that hasn't been shown yet in melanoma but worth looking at, expression of granzyme inhibitors.

TOP

In conclusion, I think I would have to say, with regard to cytokine therapy for stage four melanoma, there is no proven effective, generally available, cytokine therapy.

Therefore, investigational approaches can be used first line. I would recommend, if you are using cytokines, that you should look at patients who are stage M1A or Stage IV NED patients, where their immune system may still be relatively intact.
The approaches to consider are combinations with other cytokines, potentially immune enhances, vaccines, toxicity reducing agents, DNA demethylating agents or histone deacetylase inhibitors.

The key to all of this is accurate immune monitoring and patient selection, as I am sure Jeff Weber and others may mention.

TOP

The final slide, I just want to thank all these people who provided slides that I may have used for this talk. Thank you very much.

TOP