This is actually an easier task than I think Paul has. You all have demonstrated enormous tenacity, because I think to sit through nine hours and still be awake is a challenge of itself.

I would remind you, as Allan Lerner, who was my first mentor, suggested to me 34 years ago, if every patient that we see with melanoma just lived long enough, they would all relapse and die of this disease.

That sobering thought actually hadn't been drilled in by anything so much as the recent report from Rona McKai(?), but echoed by a patient we just saw a month ago, who had an allograft of kidney from an unfortunate auto accident, whose donation happened a decade after a thin, primary melanoma, unknown to have metastatic disease, and then evolved widespread donor metastatic melanoma. So, this is a very clever bugger of a tumor.

I think the fact that we have any progress may, in some measure, be worth noting. So, I think, as Peg has talked about in the morning, the case fatality of this disease has fallen.

That is a small source of comfort. Death and relapse are more accurately predicted now and I think this is also, to us, a boon.

For metastatic disease, I think there is no question in anybody's mind here who has ever used hydozile-2, it works. It doesn't work very often, but it works.

Although Lex won't agree, I think we have settled on high dose interferon alpha as an adjuvant therapy for high risk melanoma.

TOP

So, I think the progress that I see, as I look at this disease, is that many clinically useless regimens have been discarded.

The unfortunate truth is that we haven't understood why we failed. I think this is really a critical caveat for us in all new trials, that we study every single one to learn at least, if we fail, why we failed.

Prognostic factors have been refined, but the dismal admission is, we still have no serum markers, we still have no way simply, as in many other tumors, to follow the course of the burden of disease, and this is a critical deficit, as we will come back to discuss.

We have, as we have heard this morning in elegant talks in the start of the day, new genes that are associated with disease and progression, but the unfortunate truth is that we have not a single one of them that has made it as a foundation for a therapeutic intervention.

TOP

So, for IL-2 and for interferon alpha, I think the unfortunate truth that prohibits real advance of this field is that we don't know the mechanisms of either of these putative immunological therapies.

They are both approved at only the highest doses, something we could come back to muse about, but this sort of chemotherapy-like development that got these into approval at the highest dosages has clearly been a major impediment to the adoption of either of these regimens across the world.

TOP

We have a need of markers of disease progression, of the disease itself, because these would allow us to cull away from the patients that we treat 100 of, the 40 or 50 who will never, and would never relapse, even if we didn't treat them.

We need to find mechanisms of action to follow because, in fact, if we could do this, we could further cull the treated population to remove those who are not destined to respond to these therapy with a therapeutic index that would accordingly rise.

TOP

We have a promise from I think many new disease markers that we have heard about this morning.

We have heard multiple CD8 and CD4 epitopes that are lining up in such numerous fashion that the real problem we face is how to address these, how to incorporate these into clinical trials.

We have heard from many other groups, of course, that there are data to suggest that immune responses are correlated with disease outcome, and these I will come back to discuss.

TOP

The leads for the future for us, at least, inside ECOG, come from a trial that has accrued 112 patients as of last week, is formally closed as of next week, called E6096. This is a two-factorial trial which tests the three available peptides of Mart, melan-A, GP100 and tyrosinase, given together in all the recipients of this therapy in this trial, 28 to receive the peptides alone, the peptides plus GM-CSF, the peptides plus interferon alpha, or the peptides plus both of those cytokines.

The good news is that we have immune responses from a third of the patients that we have assayed out past the third month of therapy, and that we have correlation of immune responses with disease outcome.

TOP

So, this is the design of the trial, two by two factorial, as you see here, and I have belabored that probably long enough.

TOP

As we submitted the abstract to ASCO, these were the data. We now have 51 patients analyzed past three months of therapy. The trends hold up. We are still analyzing this data and still assaying the serum and the lymphocytes, and all of this will be mature, I hope, by the next month's presentation at ASCO.

TOP

The leads for the future are that we have many new CD4 epitopes. Jeff has already mentioned the CD4 isotope, Mart melan-A 5173 that Azan Zarour defined in 1998.

We can select these that bind to many, many broader MHC class II specificities, therefore, will not have the limitation that we presently have of selecting patients according to MHC-2 phenotype.

We have the plan to augment from CD8 to CD4 epitopes, our vaccine trials. Inside the University of Pittsburgh cancer institute, we have more than half completed a trial testing the helper and the killer epitopes of melan-A.

Craig Slingluff has written and, within the next month, will activate a trial that brings together 12 CD8 epitopes, half a dozen CD4 epitopes, of the available lineage and cancer testes antigens.

Finally, we have the ability to induce Th1 immunity, to polarize dendritic cells to get potentially better responses

TOP

and, as I mentioned already -- I don't need to twice mention -- we have phase IIs that have grown out of these phase I trials.

TOP

The most interesting approaches of the last month have been working with Bill Bigby and Billy Dade in Pittsburgh, to think about how the incredibly powerful new technology of SELDI-TOF-MS that was discussed earlier today, may be brought to bear on the problems that we face in treating patients with melanoma.

With the samples that Steve Rosenberg has provided to Mike, this analysis is already well underway. With ECOG's willingness, we will expand this to actually ask questions about patients who entered the last intergroup trial, 880 of whom we have serial serums from, over a period of a year or two of their entry into this trial, now more than a third of whom have relapsed, and a substantial fraction who have died, where we can ask, as Finestein says, retroactively, whom it was, was going to relapse at the outset on the basis of the serum marker profile and then, with the intervention of either GMK as a vaccine, perhaps a placebo control, or from the interferon alpha in the active arm, what was the difference that was induced by the interferon in the proteomic profile that was associated with response to this therapy.

TOP

So, beyond this, we have a vast breadth of things that are exciting to me to think about. One is that Don Morton said, there are three therapies for melanoma -- as you have all heard this at SSO -- surgery, surgery and surgery.

So, thinking twice or thrice about this, we wondered why this might be the case. In fact, I had better not go forward now, but I will show you in a minute, there may be data to suggest that the Th2 bias of patients who have active metastatic disease is what the problem is.

In fact, when you can reduce a tumor, whether you reduce it maybe by chemotherapy, certainly, in this case, by surgery, you can actually restore the capacity to induce a Th1 phenotype.

Certainly, there are many cytokines that have the possibility, as Mike Atkins has talked about, to repolarize the T cell response. With the CpGs, we hope to induce the plasmacytoid CD for better immunization with ESO1 vaccine trials shortly to commence in Pittsburgh, and with DC1 induced either by cytokines or NK cells with Powell Kalinski(?) with Mike Lotze, we will test the role that these play in melanoma patients over the very near term.

TOP

So, this is the data from Tatsumi Storkas and from our group, suggesting that patients who have active disease in the black spots never develop a Th1 response, and that it is only those who have no apparent disease -- the open symbols -- that we see a Th1 response to a variety of antigens.

TOP

So, what have we done in the clinical trials that we have undertaken? I think the cooperative groups in the United States have decided upon an evidence-based phase III program that, for the present, this will be based on the evidence we have with high dose interferon alpha, trying to dissect the activity of this regimen using only the induction phase in E1697, which is now intergroup and international with the participation of Peter Hersey's group and many others from Australia.

As Larry Flaherty has talked about, and Mike as well, a SWOG 0008 based upon the data with chemobiotherapy still makes sense, because there may be differences in the NED situation from the active disease situation, in the responsiveness to chemobiotherapy that deserved to be tested at this time.

In our own institution, we decided that the mechanism question is the biggest question, and that a neoadjuvant approach of high dose interferon in patients before resection of palpable nodal disease would allow us to get to this.

We have now treated 11 patients with high dose interferon before resection of their disease, with a surgical biopsy before this, and they have already begun SAGE and expression array analyses to ask what, in fact, is induced in the tumor tissue, in the nodal tissue, that may be associated with the interferon effect.

Finally, as I talked about already, the 1694 serum bank will be a huge asset, if it allows us to get to the question of the proteomic profile of disease, a quantitation of disease and, finally, the impact of interferon in this subset.

TOP

So, the response to vaccines and other biologicals, I think, are clearly best tested in the NED situation.

We have taken this from the 1694 trial evidence. We brought this now into the SWOG 0008 intergroup trial. 1696, which I showed you already, is the peptide plus or minus GM-CSF plus or minus interferon intervention in metastatic disease, is the template from which we have derived the intergroup 4697 trial that Dr. Atkins already talked about, testing the same three peptides, plus or minus GM-CSF in resected, NED patients, and where 600 of these patients, now approaching our 300 mark, will be required to ask this question powerfully enough to get to a clinical conclusion.

TOP

So, the origins and the outcomes of progress to date, I think we have had to be willing to get the rubber to the pavement.

The traction and the data that we need will only come from phase III trials. The big question is what ought to be the next phase III trial.

Clearly, the consumption of resources, time, effort, has daunted us over the last several years. We have been in a stall and we haven't really developed a replacement in the largest single subset, the patient group with a single node involved. So, I think this will clearly be part of the topic for discussion tomorrow.

Translation from the events to the adjuvant sphere is the second major conclusion. Clearly, here the risk is that we expose a lot of patients who can never benefit.

So, we have got to be very careful that we neither are accelerating disease or having unseen toxicities from the interventions that we are going to test in the adjuvant setting.

Translation studies in the human, I think, is the essence of the question. I think with the exception of this elegant data that Jim Mulé just presented, we are trying to avoid mouse models, to avoid cell line data, and I think the real model has two feet.

The critical thing here is for relevant intermediate end points, immunological, genomic, proteomic and cellomic, and there I will stop.

TOP