SLIDES & TRANSCRIPTS
Monday, May 12, 2003

Working Group C: Stem Cell Transplantation

Frederick R. Applebaum, M.D.
Stella M. Davies, Ph.D.

DR. DAVIES: I will start by saying that I challenged Fred to cross dress with me, and I wanted him to talk about pediatrics and I would talk about adults, but he wasn't going to go for it.

In our group, we talked separately about some of the adult issues and some of the pediatric issues. So, I am going to cover the pediatric issues.

We had some fairly simple questions which cover the water front. Who should we transplant? What should we transplant them with, and how should we transplant them?

TOP

So, starting first with who do we transplant, in pediatrics, I think the main controversy is who should be transplanted in CR2, in contrast, perhaps, to the adult groups where there is a lot of discussion about the place of transplantation in CR1.

So, children in CR2, many of them will be retrievable with chemotherapy, and the likelihood of that retrieval seems to be related to the duration of CR1.

We have traditionally been very aggressive with transplantation in those who relapse early, less than 36 months of CR1, and really not entirely resolved which of the late relapses should be transplanted, and are there biology studies, prognostic markers that we could use, perhaps, expression studies, to identify the children not destined not to be cured with chemotherapy who should be transplanted.

As we considered this, and looking at previous BMT studies, it has been notable that randomization to a transplant strategy versus chemotherapy has been challenging.

This has been an apples and oranges kind of study where doctors and families have a lot of difficulty accepting equipoise in this circumstance.

We noted within the ECOG study that only about 50 percent of the cases eligible for randomization actually were randomized.

I am not sure that a truly generalizable randomized study is ever going to be feasible. That is something we should aspire to.

In single center studies of pediatric transplantation, there have been a number of studies that have shown that unrelated donor stem cell sources can be equivalent to a sibling donor stem cell source.

When we looked at this in registry settings and cooperative group settings, I think the results have been somewhat disappointing and inferior outcomes have been seen.

I think it would be interesting to ask why that is happening and whether we can raise the level of outcome with aggressive, uniform supportive care, uniform recommendations, perhaps selection of centers to perform transplantation, to see if we can get equivalent outcomes.

Currently, we have different recommendations for utilization of unrelated donor stem cell sources and sibling stem cell sources, which perhaps biologically don't make sense.

Our transplant protocols, our treatment protocols should perhaps be viewed not as the ultimate answer to a cure for the children with the worst disease, but as a platform for incorporation of new agents and immune modulation, both before and after the transplantation. I think that is another valuable goal for the cooperative group, to bring in both those pieces.

TOP

So, what do we transplant patients with? Obviously, in pediatrics, there has been a great deal of interest and work in the last few years about the role of umbilical cord blood.

Most of the early investigations have been early phase studies looking at feasibility, looking at unit selection, and with very little emphasis on single disease and particular activities.

Mitch Cairo showed us some nice data regarding potential ex vivo manipulation of cord blood, and there is a lot of work to be done in this area, looking at ex vivo manipulation for expansion, genetic modification, identification of cell subsets, that might expand the utilization of umbilical cord blood.

I mentioned earlier we need to assess the place of unrelated donor grafts and see if they are interchangeable with sibling donor grafts, and improve our outcomes within large-scale studies.

TOP

[No text is associated with this slide.]

TOP

Our last question was how do we transplant them, what post-transplant immune modulations might we use to improve outcome and reduce relapse.

We started by agreeing within the room that graft versus leukemia is probably a real effect in ALL, and immune modulation would be a potentially valuable approach to improving outcome and reducing relapse.

TOP

Kirk Schultz showed us this nice overhead, that I borrowed from him, regarding possible immune interventions, of which there are many, which could improve the outcome of transplantation in the sense of intentionally reducing relapse, and also improving immune reconstitutions. Death from late infection remains a problem, and Kirk started with the most likely, moving to the most interesting immunization in normal children.

TOP

So, where does this take us to in terms of needs? A lot of work, a lot of studies. We need studies of the in vivo and ex vivo manipulation of umbilical cord blood.

We need better ways to identify the children in CR2, who will not benefit from chemotherapy, and that can include minimal residual disease studies, as earlier alluded to, understanding of blast biology, and introduction of new agents, with better induction regimens, with the aim of bringing children to transplant with a deeper remission and with less toxicity.

Currently, the very toxic induction regimens will mean a significant number of children won't get to transplant and those that do have significant residual organ toxicity.

We need to consider improvements in uniformity of our outcomes in cooperative group studies. Do we need to raise the level of performance of some centers, perhaps select limited centers for some studies.

Finally, there is a multitude of potential studies of post-BMT immune modulation that can reduce relapse and improve our immune reconstitution.

TOP

DR. APPELBAUM: Turning now to adults, at the outset, it should be understood that ALL, which stands for either acute lymphoblastic leukemia or acute lymphocemic leukemia or acute lymphoid leukemia -- I am not sure yet. In AML, we decided it was acute myeloid leukemia, and I don't know why we haven't made the same change in ALL, so that we could all have the same titles to our chapters.

I propose, whoever it is, that if we make it acute myeloid leukemia and acute lymphoid leukemia, and it would be a lot easier. Anyway, that is beside the point.

In adult ALL, it is not a very common disease, and the number of individuals who get transplanted are even less common.

So, this is not an area where we are likely to have large randomized studies addressing the major issues, such as which is the best preparative regimen in ALL, and what is the best form of GVHD prophylaxis in ALL, or what is the best way to prevent infections post-transplant in ALL.

Those kinds of questions are more likely to be asked and answered in broad studies of leukemia than they are in adult ALL, just because of numbers, which make it daunting.

Nonetheless, what we heard this morning from Jacob and from Marty, from the MRC and the ECOG study, is that the data appear convincing in a very large study, which is unlikely ever to be replicated again.

I mean, they put on over 1,000 patients, now, 1,500 patients, that allogeneic transplantation for adult ALLs, if you have a matched sibling, appears to be superior to conventional chemotherapy or autologous transplantation, lumping the two together, if you don't have a matched sibling.

So, from that, with all the problems of the data and all the problems of interpretation, it is the best we have.

So, the current state of the art would be, if you have a matched sibling and you have ALL and you are in first remission, that you ought to be transplanted.

If you have ALL and you don't have a matched sibling and you have really high risk disease, such as PH positive ALL, you probably should be transplanted in first remission using an unrelated donor, if such a donor can be found. Those are the conclusions that I think we were given from Jake and Marty this morning.

The corollary to that is that there are a number of individuals who have matched siblings and who are transplanted who probably didn't need it, probably would have done well with chemotherapy since study after study after study of adult ALLs over the last two decades have shown 37, 38, 39, 37 percent of patients are cured with conventional chemotherapy and probably never needed the transplant.

Likewise, since unrelated donor transplantation can cure even very high risk patients, there are probably many patients who do not have PH positive ALL, but have high risk disease nonetheless, who would have benefited from an unrelated donor transplant, could we have identified them.

So, given this data, then we would say that one very high priority for improving the outcome of patients with ALL is to identify those patients who do not need an allogeneic transplant because they are likely to be cured with conventional chemotherapy or, contrary, are likely to benefit even from an unrelated donor transplant because they are not likely to be cured with chemotherapy.

We need to develop methods to do that beyond our relatively crude measures today, and those would involve, for example, the genotyping, using the kinds of stuff that Cheryl showed us earlier this morning with the ray analyses.

It may be that it will be done with phenotyping, looking at other issues beyond genotyping, or it may be done by measurements of minimal residual disease.

So, we would put this as a very high priority if we were going to improve the overall outcome of allogeneic transplantation. That is a first major point.

TOP

As far as autologous transplantation is concerned, we heard, I think, from Charlie that there currently is a study going on, which is a large randomized study, the ECOG MRC study, and until that study is done, it would not seem that there is an indication to start another large randomized study of autologous transplantation, given the relative equivalence of past randomized studies.

However, there have been some interesting pilots that might be considered. One pilot which we mentioned would be to look at the issue of autologous transplantation followed by non-ablative allogeneic transplantation, in order to get the benefits of the allogeneic effect without the toxicities that we see with unrelated donor allogeneic transplants.

In the consortium studies that Steve Forman mentioned briefly, we have carried out unrelated donor, non-ablative transplants in patients over the age of 50, with a 100 day mortality of about nine percent, and an overall non-relapse mortality of 18 percent in patients with a variety of high risk diseases.

One could then imagine carrying out an auto allo transplant for patients with very high risk CR1 or CR2, to find out if it could improve outcomes.

As we have also mentioned, there are, in autologous transplantations, a number of issues about what is the best preparative regimens, are there ways to purge bone marrow, are there ways to apply immunotherapy post-transplant, as Steve suggested.

It is highly unlikely, or virtually impossible, that any of these can easily be addressed in prospective randomized trials in the near future, unless we had compelling data to get all cooperative groups in the United States and in Europe together, to do such a trial.

In order to get that kind of compelling data, one would need at least phase II studies with very powerful intermediate end points.

I think this is a recurrent theme that we face, and that is that as investigators who want to move things from the laboratory to the clinic and from phase II to phase III, we are going to increasingly be relying on intermediate end points, because it is the only way we can get the necessary data to move forward.

At the same time, every time we do that, we are going to be butting heads with CTEP and with the FDA, who is not very happy about using intermediate end points, even though they may seem logical, because the tie between the intermediate end point and true survival or disease free survival in every one of these circumstances.

So, this is going to be a repetitive process of education between investigators and regulatory officials, as we try to do this.

TOP

Finally, as far as other areas of research are concerned, we talked about those areas which, if we had deep pockets and wanted to really invest in the preclinical laboratory studies that would allow us to move the agenda forward in the treatment of ALL, that we would probably fund these areas.

We would fund research into the genetics of drug metabolism, of toxicity, and of GVHD. We have been impressed in our own center by the improvement we have made in the outcome of transplantations for CML in chronic phase, for MDS, for myelofibrosis, just by looking at the metabolism of busulfan.

Using busulfan where we target now specifically to 900 nanograms per ml over four days, and use peripheral blood stem cells, we no longer see any age effect in transplantation for CML between the ages of 15 and ages of 65.

So, the cure rates in patients with matched siblings, age 65 is the same as for someone of the age 15, likewise in myelodysplasia.

Myelofibrosis, now we have disease free survivals in a group of patients that are in age 55 to 60 of 70 percent, going out to six years. I think the drug metabolism made a big difference there.

There still are patients who suffer unexpected severe toxicities, which we don't entirely understand and some of that may be the metabolism of other drugs, such as cyclophosphamide, which we are looking at, or some of it may be the genotype response in toxicities.

Finally, we know that there are, in this area, differences in immunomodulatory genes that regulate, for example, the promoters for TNF, the promoters for IL-1, which may predict who will and who will not develop more graft versus host disease.

You have already heard in the past talk about trying to understand better the immunology of GVL, finding T cells from the donor that react relatively or absolutely specifically with leukemia associated, or specific genes.

Those may be fusion products, but more likely, will be up-regulated self antigens such as WT1 or AF1Q, which are highly up-regulated in virtually all ALLs, and not increased in normal hematopoietic stem cells, and may be a very viable target for a T cell response.

Finally, we have heard a lot, in virtually every talk throughout the day, about the need to assess the risk of relapse.

Some of it you heard again, it may be the phenotype of the leukemia, but dynamically, probably the best way to do that is going to be looking at response to induction chemotherapy, the amounts of measurable disease afterwards.

That may be done by better defining flow techniques or by the development of a leukemia chip, which could have every VDJ rearrangement, for example, on it and would make it much easier to find the specific rearrangement for that individual.

Then, very quickly have a PCR assay for them later on and guide therapy, allowing those who are at high risk to get transplanted early.

Probably the application in this area is the least in terms of stretch. I mean, it is clearly doable and would, I think, very quickly translated into improved outcomes from transplantation, saving toxicities from those who don't need it, and allowing those who do need it to get the transplant earlier. That was the gist of our discussion.

TOP