Melanoma 2003

SLIDES & TRANSCRIPTS
Monday, May 5, 2003

The European Perspective for Treatment of Metastatic Melanoma: Implications for Future Clinical Trials

Ulrich Keilholz, M.D.

Slide 1:

Thank you. I am very grateful to the organizers for giving me the next two hours to discuss the approach to stage IV melanoma in 20 different countries.

Maybe I should start in the first 15 minutes with the EORTC perspective and then see what Frank and Vern will do to me.
I also got this eight-page behavior list. I still remember one line of it. I should talk about results and implications and not how we got there.

Slide 2:

So, this is the result. I will not tell you how we got there, but this is the results of a randomized EORTC study on 350 patients, with a negative outcome.

We randomized patients to receive TGIC cisplatin with or without IL-2, and with mature follow-up, there is no statistically significant benefit.

Let's go back to what we thought we should need to see to have meaningful results.

Slide 3:

The intended difference that was felt to be clinically relevant five years ago was an improvement from 10 to 20 percent in overall survival at two years.

So, we thought we would be smart and not look for a difference in median survival, because we knew that 50 percent of patients would progress right through treatment, but look at the two-year increment.

Now the observation is an improvement from 12.8 to 17.6 percent in two years, which was not significant, or 8.5 to 12.7 percent at three years, which is not significant and which we thought five years ago would not be clinically relevant, and I think that still holds today. So, an increased number of patients would not change the picture.

Slide 4:

Now, to go to that extremely throughput laptop-based method of looking at prognostic factors, once you have clinical factors -- the lab factors are different -- and that is AJCC N stage.

With the logrank test, it was not significant, what the end stage of the patient is, in terms of overall survival.

So, that may be the reason that we only have very few patients with N1A because usually in N1A with just two metastatic sites, a low volume disease, we in Europe would rather do experimental treatment than chemotherapy.

Slide 5:

We went back one step further and just looked at two factors, serum LDH and how the patient felt, that is, Karnofsky index.

If we look at initial performance in LDH and have a combination factor, the blue being normal LDH and 100 percent Kanofsky index, the green elevated LDH and Karnofsky index of 90 percent or less, and the red one just one of those bad factors.

It gives a clear distinction of the lines and actually it is rather decisive. In the good prognostic factor group, regardless of metastatic side or number of metastases, there is a 15 percent level of long-term survival.

Slide 6:

Now, we can look back to a more mature trial, that was the first trial. This is a small patient population, trying to ask the question in another way, to add cisplatin to interferon alpha and IL-2.

There was a doubling in response rate and time to progression, but no effect on overall survival, and that still holds five years later, still the same curves.

There are a few patients out here and those patients, two of them, had a partial response to treatment and that was surgically completed. The other patients had a complete response to treatment.

So, those with clearance of tumor upon this treatment do have a likelihood to survive five years later,

Slide 7:

but still, this is a low percentage of patients. AJCC M stage, again, was of borderline significance,

Slide 8:

and our clinical factor performance of LDH again showed the nice divergence of the curve and was able to define the good prognostic group of patients that will be out there five years later.

Slide 9:

So, the conclusion of these prognostic factors is AJCC M substaging may not be optimal for stratification in clinical trials with systemic treatments because a lot of the N1A patients will not receive systemic treatments, but the combination of LDH and performance status is a relevant factor.

Slide 10:

We still have one randomized phase II study out there, and we did not dare, in the old days, to open a protocol just to compare DTIC to a four drug treatment, but we, rather, looked whether there was cross resistance.

So, we randomized 90 patients to receive DTIC, platin, interferon alpha, IL-2, or to receive two cycles of DTIC and then go on to the four drug treatment to look for cross resistance.

That trial has completed accrual and treatments are completed, but the analysis is ongoing and I can't tell you today whether there is cross resistance or not.

Slide 11:

So, what have we done now? We have compared IL-2 without interferon alpha, plus or minus chemotherapy.

We have compared chemotherapy plus or minus IL-2 or IL-2 and interferon alpha,

Slide 12:

and both of those randomized trials are negative for overall survival.

Slide 13:

We have not done, with a Heidel's(?) IL-2 schedule this comparison directing chemotherapy against biotherapy. In a way, this is addressed in the international study by Maxim comparing DTIC with interferon alpha-2-histamine, but this is a low dose regimen. So, we will see what that is, but Heidel's IL-2 regimen has not been tested.

Slide 14:

What did we learn so far from these chemoimmunotherapy trials? There is a response rate with interferon alpha, IL-2 that is approximately 18 percent in phase III trials.

The response rate, but not survival, is increased by chemotherapy. However, the converse is not true. With the addition of IL-2 to chemotherapy, at least the same decrescendo regimen that has been used here, does not confer benefit to a response rate, and not to overall survival.

The subgroup of patients with normal LDH and 100 percent Karnofsky index, there is a long-term survival up to 20 percent, regardless of metastatic side. That is interesting for future developments.

Slide 15:

This sounds like a study that should have been taking place 50 years ago, that is an international multi-center, three arm randomized phase II study to compare IL-2 regimens -- civ, decrescendo, subcutaneous -- with a primary end point of IL-2 pharmacokinetics, and secondary end point of serum IL-2 receptor induction, and melanoma-specific CTL induction.

This trial was not performed 15 years ago. This trial was just completed, and it shows how illogical some of the development of IL2 has been, that things like this were not clarified in the beginning. Maybe the field would have developed differently.

Slide 16:

This trial is also completed in accrual. We are studying the melanoma specific CTL induction and, so far, there is very little CTL induction, only one patient out of 21, and that is in contrast to the combination of interferon alpha and IL-2, where we saw a higher rate of induction of melanoma specific CTLs.

Slide 17:

So, what will be the fate of IL-2 in Europe? It will be different from the United States. Academic interest persists in this molecule, but there is no license for melanoma in Europe.

The patent in Europe is about to expire this year, which is not true for the United States. So, there is no further company development plan. That is what we are left with at this point.

Slide 18:

The major academic interest is to use it as an adjunct to vaccines, and also the IL-2 subcutaneous plus or minus histamine in liver metastases patients has major accrual in Europe. I think 2,000 of the patients come from Europe and one third from the United States.

Slide 19:

So, a change in topics. This goes now to the chemotherapy approach. There is a randomized phase III study of fotemustine versus dacarbazine as first line treatment of disseminated melanoma, to look whether dacarbazine is really the only drug. That was reported at ASCO last year and also as ASMO last year.

Slide 20:

It is a little too small. It is 112 versus 117 patients. The response rate of fotemustin was 15.2 percent, DTIC 6.8 percent, which was borderline significant.

The response duration was not really different.

Slide 21:

The overall survival was 7.3 months with fotemustin, 5.6 month with DTIC, again, borderline significance.
Time to brain metastases with fotemustin, which crosses the blood brain barrier, was much longer than DTIC, but again, borderline significance.
So, maybe with a doubling of patients it could have been significant, but at least it shows that there is another drug that achieves results as impressive as DTIC.

Slide 22:

So, what is the current status of fotemustin in Europe? It is not available here. It is licensed in France and Switzerland and, in those two countries, it is, in a number of institutions, it is the first treatment option for metastatic melanoma patient.

Slide 23:

You know, they cause a lot of hematotoxicity, so they may be the last treatment option, because after that, there is almost no meaningful way to give another cytotoxic drug. That is the problem with this compound.

Slide 24:

Still, we have an EORTC trial in preparation with fotemustin in ocular melanoma metastatic to the liver, of fotemustin intravenously into the liver artery to see whether local application is more active. That is based on phase II results from Lausanne.

Slide 25:

How about vaccine studies in Europe? MAGE-3 protein is tested with different adjuvants.

There will be a randomized phase II study in the EORTC where the end point is stabilization rate in immune responses. So, we are not looking for a response rate. We changed the end point of the study.

It is a randomized phase II looking at immune response and also clinical stabilization rate. We know that, with a vaccine in patients with overt stage IV, we will not see much in terms of response and action.

HSP90, phase III in preparation. That again will be in phase III extracapsular disease, stage IV NED.

There are various peptides, academic peptide trials with various adjuvants. The problem there is that, because of the patenting situation, there is only very little company interest, and also there are various definitions of GNP production, whether the peptide is produced under a GMP condition, or every single amino acid is produced under GMP conditions, and that makes multicenter trials in multiple countries difficult because every authority in every country has a different viewpoint here.

Slide 26:

For the vaccine trials, we do follow the SBT recommendations on vaccine monitoring that were made one-and-a-half years ago in this city, to do two of those three assays, ELISPOT, CFC and tetramer assay.

Slide 27:

That can lead to results like this one, that was a cohort study looking at the role of adjuvants in peptides. That was the example of tyrosinase peptides with adjuvant in non-metastatic melanoma patients.

Nine patients with no adjuvant, three patients had a response after six vaccinations. GM-CSF alone did not change that picture.

If you add KLH, if you get in the problem of immunodominance, you get beautiful responses to KLH but none any more to the peptide.

The combination of the T helper function of KLH and GM-CSF gives early T cell responses in two thirds of the patients. So, that is maybe more active, and that is in line to follow with other adjuvants, other combinations, to see the frequency of induction of T cells, but also the kinetics of induction, and further, of course, what Jeff Weber was mentioning, the induction of T cell memory.

Just to complement Jeff's thought, we rarely find memory T cells in peripheral blood, but we frequently find them in the bone marrow. Maybe the blood is not the compartment to look for T cell memory responses.

Slide 28:

The clinical vaccine philosophy in Europe is to find suitable vaccines from stage IV trials to be used in earlier disease stages -- that is, stage III, stage IV NED patients.

An emerging issue may be vaccines for maintenance of responses induced by other treatment modalities.

To go on this stage IV NED issue, this is just three patients.

Slide 29:

This is three years. These are all the recurrences in these patients, most of them cutaneous, subcutaneous, but this is a lymph node metastases, this is a brain metastases.

The vaccine was started here, tyrosinase peptide and GM-CSF at that time. This is three years later. There is a single recurrence in the regional lymph node, same patient as this one, and this is a single small bowel metastases.

At that time, the patient had a T cell response and the T cells were lacking CCRN. CCRN is the direction of the T cells to the small bowel.

The melanoma cells did express CCRN. So, the melanoma cells did have direction to the small bowel, but the T cell says not. So, that is another potential mechanism of immune escape, for the T cells to be in the wrong compartment.

We now have reported 20 patients in this situation and in a number of others, so some interesting clinical response. We have to wait another two years to really sort that out.

Slide 30:

So, what is the current stage IV treatment approach in Europe? Well, the first line is experimental treatment in academic centers and trials.

Outside first line, maybe DTIC or fotemustin, otherwise, the second line is DTIC or fotemustin. Combination treatments are now days only mainly used with palliative intent if there are symptoms.

That is thoughts and perspectives. Do you want me now to discuss the guidelines in 20 different countries? Maybe not.