DR. BRENNAN: Questions for Jonathan? My sense is that the second group will come forward with some recommendations which are very complementary to what you have presented, but not duplicatory. Questions for Dr. Fletcher?

It seems to me that one element that is missing from your presentation and Sharon's is the molecular validation of diagnosis, the kind of work that is underway at Stanford, underway at Memorial -- probably other centers, additionally -- to try to provide a molecular correlate to immunohistochemical and microscopic diagnosis. Any comments concerning that?

DR. JONATHAN FLETCHER: I can offer some personal comments, but I sure won't speak for the group on this one.

There are two issues there. We know that, in the ideal world, that for some of these very compelling diagnostic markers, they would be done routinely in every case. It certainly would make good sense, as Sharon mentioned, to have EWS-FLI1 evaluated in every potential Ewing's sarcoma, to have the X;18 translocation or the encoded SYT-SSX fusion evaluated in every potential synovial sarcoma.

One of the problems in making a recommendation that that be done universally, of course, is there will be disagreement, and including amongst our group that that is necessary.

The other one which, from my end -- running a diagnostic cytogenetic service is really the biggie -- is that once you get outside of Marc Ladanyi's group and the groups who are doing this with a passion and a focused effort, those tests in the contemporary world are going to be done badly.

That is my personal feeling. Just because it is a DNA test or a cytogenetic test doesn't mean that it is going to be done with any better quality control. In fact, personally, I think in many cases it will be worse than the standard morphological approaches which we were discussing the fallibility of. So, perhaps that is worth some discussion from the group. I must say, we realized this morning that we sidestepped that a little bit in the breakout session yesterday. So, if there are strong feelings, this would be a good time to bring them up, I think.

DR. BAKER: Jonathan, I thought that was a superb way of looking at this problem; but I want to challenge you about one word you used a couple of times, and that was the idea of selecting out common histological subtypes for evaluation.

I don't know that that is what I would like to see. I clearly would like to see what you propose become real; but, for example -- and only by example -- we would certainly like to see a disease like alveolar soft part sarcoma become targeted, even though it wouldn't meet the criteria of being common.

I think there are other clues in biology that would make some types more relevant than others.

DR. JONATHAN FLETCHER: I agree absolutely. These are all, in some spectrum or some vantage point, uncommon disorders. To say that, well, well-diff liposarcoma makes the grade and alveolar soft part or clear cell sarcoma doesn't is clearly arbitrary.

The logistical issue -- and we discussed that problem in the breakout -- is that to get somebody to do these SAGE analyses, for instance, is a good bit of work. So, from a practical standpoint, you have to start with some subset. If you're going into the group and say, which are we going to do?, it might be that alveolar soft part sarcoma won't but retroperitoneal leiomyosarcoma will; but it is clearly arbitrary.

Certainly, from the standpoint of being unique cell lineages and shedding light on biology and mesenchymal biology in a greater sense, it might be that alveolar soft part sarcoma is the goal. So, we recognize that.

DR. BENJAMIN: I would like to get to the issue of the quality of molecular diagnosis. I think saying that because the majority of people don't know how to do something it will be done badly, therefore, it should not be done, is not the message that we should get out. That goes back to the light microscopy diagnosis of these same tumors. If there are tools that help with diagnosis, we should insist that they be done by qualified individuals.

DR. JONATHAN FLETCHER: One specific recommendation that the group was at least countenancing was that -- amongst a Web resource that would centralize information that is relevant to the sarcomas -- some of it might be more developmental-- that a key aspect would be to define the laboratories that are actively doing RT-PCR, FISH, and sequencing for the various targets of interest or relevant in the sarcomas.

This, of course, is part and parcel of the urgency of having frozen material for every case, such that either you can legislate to make that a routine phenomenon -- that if it is synovial sarcoma, it has to be sent for molecular validation -- or take the softer approach of saying, if you have any concerns about it then it needs to be sent for molecular validation.

One of those two needs to be done and, in my view, that should be one of the recommendations from this group. We need frozen material to make that easy and, ideally, you need a list of -- you can't sidestep this one. If you are going to expect labs that don't focus on tumors in the first place to be doing a good job of this, it is just unrealistic. I think there needs to be a core group of labs that are identified where people are strongly encouraged to send specimens so that they are done properly.

DR. HEALEY: If you don't insist on these molecular tests being done to validate diagnoses, then you are going to fall into the problem that Chris Fletcher discussed yesterday -- of not being reimbursed, of not being countenanced by the agencies and the insurance companies and the like.

Now, I think the goal is as Bob Benjamin just stated eloquently, that you should set the standard high, appropriately, and then force other people to meet that standard one way or another. Setting the bar low and accepting mediocrity is not what this conference is about.

DR. BENJAMIN: Just to get back to the question, it is not really important to me, treating a patient, to know the molecular diagnosis of a patient that Harry Evans says is unquestionably a biphasic synovial sarcoma.

What I need to know is the molecular diagnosis of the patient that he calls unclassified because it doesn't fit the criteria but, in fact, has the same translocation and therefore, would benefit from the same therapy. I think we really need to put this into the routine management, particularly for the cases where the diagnosis is unclear.

DR. JONATHAN FLETCHER: What you are asking for, I am sure, is what everybody can agree to, and the infrastructure needs to be there to support that.

Going on to having these things done routinely is a different story again. One of the logistics -- and I don't know what the real intention is here is, are we equipped to do that, beyond the fact that ideally this is the right pathway.

Just to take FISH as an example, which is one approach, do we really have the resources? One could make a general recommendation that these sorts of adjuncts need to be routinely done, and it is questionable as to whether there are the resources in place now to do them at the adequate level of reproducibility. Therefore, that needs to be set into place.

DR. BORDEN: If you and Sharon are in agreement, I would encourage you to put something like this into the recommendations. This is a research meeting. This is not a clinical practice meeting. As Dr. Healey just said, we should be setting the bar high.

DR. HOGENDOORN: You are listing of resources that are needed. There are a number of things of consideration for me as well, and that is for instance the storage of corresponding normal DNA from these patients, which can act as internal control, but also for stem line abnormalities.

The second thing is the storage of serum for getting into the proteomics area. It could be very valuable to have series of serum at the point of diagnosis available, to see whether we could take up our best diagnostic efforts.
I missed a bit of the proteomics approach in your presentation as well, or do you think it is too far away from us?

DR. JONATHAN FLETCHER: Oh, no. The proteomics approach that I am describing here is a low-throughput one, but it is proteomics nonetheless.

For example, one can use phospho-specific antibodies to evaluate 50 targets at a time. There are plenty of other methods that I guarantee you that, if there is a publication that comes out of this, there is a great deal of enthusiasm all around for bringing in the specific avenues that would be used including the various proteomics options. So, that would all be detailed, but as subsets of these general headings.
DR. BORDEN: This has been an excellent start to the cross-disciplinary discussion that the organizing committee hoped would happen after these presentations. Any other comments, points, corrections? All right, thank you very much.

The next session or the next breakout group is the group which addressed therapeutic interventions, targeted therapies for systemic disease -- Jaap Verweij and Larry Baker.

TOP