Okay, thank you. Next is the summary of the surgical section and Dr. Rusch is going to do that.

DR. RUSCH: I will try to capture the points that I think were made in the presentations and then give you what is really a single recommendation with several ramifications. First of all I think we heard that lobectomy via thoracotomy remains the standard of care for stage I tumors. We heard that VATS lobectomy potentially yields comparable results to that in carefully selected patients managed by experienced surgeons. We talked about the importance of lymph node sampling versus mediastinal lymph node dissection and talked about the fact that that was undergoing study through a randomized trial. We heard that the technique of sentinel lymph node sampling now appears feasible with a diagnostic accuracy that approaches that of mediastinal lymph node sampling or dissection.

Clearly further investigation is needed to refine and validate this. The potential role of sentinel lymph node biopsy is promising but certainly will require very careful investigation in the management of these subcentimeter lesions in the future.
We have heard from the Japanese surgeons that the routes of lymphatic drainage from the lung I think are well defined thanks to their work in resected specimens and in cadavers. However, this new set of lesions, lesions that are primarily CT detected, presents to the surgeon a new set of questions with respect to surgical management that we really have not had to grapple with in the past.

Both the Japanese experience and several retrospective series including the one presented by Dan Miller from the Mayo Clinic with these very early certainly less than 2 centimeter and predominantly 1 centimeter or less lesions suggests that the size of the primary tumor does not fully define the biological behavior. Clearly a small number of these very small lesions have a highly malignant potential whereas others appear to have an indolent behavior, and rarely metastasize.

The proposed Nuguchi classification may help to distinguish between some of these lesions but needs to be validated in further studies in larger multicenter studies.
I think that in summary that there really is insufficient information about the imaging characteristics, the pathological, morphological features and molecular abnormalities of these very early cancers. So, this is really thematic with what you have just heard from the other two summaries and especially with respect to the subcentimeter invasive cancers or to the lesions that are premalignant and very small. Because of this lack of information we really are not yet poised to design trials of surgical management or for that matter other management, such as radiation oncology intervention or as you heard RFA.

Moreover the number of patients that are currently being diagnosed with these lesions in North America is still undefined and therefore I think that the logical point of departure is to start with a registry. This is very consistent with what you just heard from the pathology group. This is a natural history registry which would link a careful clinical database with, it would serve as a platform from which imaging studies could be carried out and the pathology studies which you have just heard about could, also, be carried out.

Now, this registry obviously needs to be linked to the specimen repository that was just suggested by the pathologists. Most of the specimens acquired from surgical management are obviously going to be paraffin embedded although I think that from a few selected high-volume institutions we could probably obtain snap frozen material as well.

One group that could serve as the point of departure for such a registry would be the American College of Surgeons Oncology Group. I think it is of great importance that we be the repository for perhaps the clinical information and follow-up but also make available to investigators such as all of the people here in this room who have the interest and the expertise in the molecular studies, the imaging studies, the pathology studies, the material that would be gathered from patients who are entered on such a registry.

I would be interested from any of the other surgeons who are still here, particularly Nasser, if you have any comments if I have missed any points here in the summary.
Dr. Altorki: That summarizes it well. No other comments.

DR. RUSCH: Dick?

PARTICIPANT: I agree.

DR. SAXMAN: Valerie, you have experience with these types of registries through the lung cancer study group. Would you like to comment upon what the difficulties are with that type of registry, what obstacles would need to be overcome? I mean in a large sense, not the micro management of it. What are the large issues and problems that you have encountered with these sorts of efforts in the past?

DR. RUSCH: The LCSG T1/N0 registry that I mentioned briefly yesterday was actually a very straightforward undertaking and somewhat different from what we are talking about here. Here I think the structure of what we would do is very important because if we are going to have a registry of such patients, we need to have both the tissue repository that would be available to investigators both inside and outside the group. Also, we need to have the imaging repository that would be available for imaging pathological and clinical correlation and so I think that we would need to take the concept of what was done in the LCSG and expand that and it is more complicated than what we did before.

Actually the T1/N0 registry in the LCSG worked extremely well and so the clinical component of that is the most straightforward. But it is building the other components and making sure that we have a long-term repository that can be sent out to other investigators that is really going to be the challenging part of it.
Parenthetically I spoke to Bob Ginsberg before this meeting, and one of his suggestions was that if we had such a registry within the College group that that should be likened to efforts perhaps through the RTOG to develop a registry of similar format for patients who are undergoing treatment primarily with radiation but we can address that in the next group.

DR. SAXMAN: Possibly a linkage I guess with ACRIN. Ed, would you like to comment?

DR. STAAB: I think ACRIN for sure but again going back to that Lung Imaging Database Consortium Group that was a contract to develop an open source in the database. The express purpose of this was initially to be able to validate all sorts of algorithms and things that we were developing, but we needed a base to do this which would require truth measures that Matt had commented on.
I think that we can certainly use that to build on to link to the other components that Valerie is discussing.

PARTICIPANT: Since I think we are talking about registries for pathology and surgery it seems to me that there is no real merit to duplicating the registries. There are already existing several screening programs that are connected through mechanisms, for example, like the one that the ELCAP has and I think perhaps one single registry for both pathology and surgery that could link the major screening programs in the country would suffice, rather than having multiple duplications here. Perhaps that would be, the SPORE would be one mechanism that would be able to support something like that.

As far as the tissue repository I think we need to give some thought to what Harvey said yesterday. Until such a time that there is one megabank that banks everything, there are screening activities going on already. Tissue banking that has been going on both currently and for the last several years needs to be supported.

PARTICIPANT: I think that what I am hearing is that we need some real concerted coordination to make sure that registries that are being developed are coordinated and that the kind of information and the way the information is stored is similar across registries so that ultimately we can mine the data and develop hypotheses as well as hopefully draw some conclusions. But I think that we need to make sure that there is some coordination and as I say some organization of the data.

DR. SAXMAN: Any other comments?

TOP