Current Therapy and Future Directions in the Multimodality Treatment of Locally Advanced Non-Small Cell Lung Cancer by David Gandara, MD

SLIDES & TRANSCRIPTS
Wednesday, June 14, 2000

Current Therapy and Future Directions in the Multimodality Treatment of Locally Advanced Non-Small Cell Lung Cancer
David R. Gandara, MD

Slide 1:

DR. SAXMAN: I'd like to start off the morning program by introducing Dr. David Gandara, Professor of Medicine at the University of California, Davis. He is also chair of the Southwest Oncology Group Lung Cancer Committee. David is going to talk about, "Current Therapy and Future Directions in the Multimodality Treatment of Locally Advanced Non-Small Cell Lung Cancer."

DR. GANDARA: Thank you, Scott. It's really a pleasure to be here for this State of the Science meeting, and to share what I think will be a very exciting day and a half. Never have there been so many opportunities for advancing survival in lung cancer.

Dr. Saxman said that non-small cell lung cancer was near and dear to our hearts. He asked me to make this presentation from the standpoint of the eyes of a medical oncologist. The old saying goes that beauty is in the eye of the beholder. If that is true, then non-small cell lung cancer has not been a pretty picture up until now. Hopefully, that is changing.

The data that I will review, and my opinion about the current status of therapy and where we are going in clinical research, reflects studies that most of you in the room have published. So I hope I do them justice and set the stage for this meeting.

Slide 2:

The first thing I would like to say is that the clinical management of advanced non-small cell lung cancer has evolved and is evolving, not only in academic centers but in the community as well, from a single specialty or modality into a combined modality approach. This point cannot be overemphasized. It has resulted not only in academic centers, but also in many community hospitals in the development of multidisciplinary thoracic oncology programs, by teams that prospectively make decisions about patient management.

Slide 3:

The critical need for multidisciplinary evaluation and combined modality therapy is reflected by these five-year survival results adapted from Mountain. Locally advanced disease, broadly interpreted as Stage IIIA or IIIB clinical stage demonstrates very poor survival.

Slide 4:

The other point that needs to be emphasized before going into actual study results is that there are very distinct clinical subgroups even within Stage IIIA, as currently defined. Most of these are candidates for combined modality therapy. Stage T₄, based on malignant pleural effusion really has a natural history and therapeutic options identical to that of Stage IV disease. But all the other categories are candidates for combined modality therapy, their prognosis and their therapy defined by specific findings in terms of tumor and lymph node status.

Slide 5:

Within the last ten years a number of research studies have looked at the role of surgery, radiotherapy, and chemotherapy, and I would like to review some of that data with you now.

Slide 6:

I'll focus on two areas. Although the primary focus of this meeting is combinations of chemotherapy and radiation, I think an important paradigm in terms of telling us where we are and what we have learned about the biology of the disease really comes from the preoperative studies that have been done. This also presents a window of opportunity for studying novel therapeutics in that we have pre-treatment tissue and post-treatment tissue in sufficient quantity.

Slide 7:

In my opinion, this is where we are. This is the current status in terms of preoperative therapy for Stage III disease. Preoperative chemotherapy appears to improve survival compared to surgery alone in some patient subsets. Preoperative chemoradiotherapy results in higher complete radiographic and pathologic complete response rates than chemotherapy alone, but also shows higher toxicity. Whether chemoradiotherapy followed by surgery is superior to the definitive chemoradiation without surgery in Stage IIIA N2 disease is unclear, and it is the subject of the ongoing intergroup trial chaired by Dr. Albain. Lastly, a point that is worth emphasizing, trials of preoperative therapy represent an in vivo model in which we can both clinically and through laboratory correlates assess sensitivity or resistance to new therapeutic regimens.

Slide 8:

These are updated five-year survival data from Jack Roth and Raphael Rosell's small randomized trials of surgery alone versus preoperative chemotherapy plus surgery. Although the patient numbers are very small, at five years there continues to be a difference in survival. This is really not definitive information, but I think it leads us down several paths in terms of follow-up studies.

Slide 9:

This is the intergroup trial, Intergroup 0139, of preoperative chemotherapy and radiation followed by surgical resection, versus definitive chemotherapy and radiation. Both arms of this trial are based on Phase II studies performed by the Southwest Oncology group, and I think it really gives us long-term toxicity and survival data as a very good basis for these two arms.

Slide 10:

In particular, SWOG 8805 of preoperative chemotherapy and radiation prior to surgical resection in pathologically staged patients, both N₂ and Stage IIIB without plural effusion, gives us some important information about how we are attacking both local control and distant failure in this disease. This is one of the largest experiences, especially of chemotherapy and radiation. Importantly, 40% of the patients had pathological Stage IIIB disease. The five-year survival rate in both arms is about 20%. So this represents curative therapy.

And importantly in this study, with two cycles of cisplatin plus etoposide and radiation to 45 Gray, there was a major pathologic impact on the tumor. Complete pathologic response occurred in 21% of patients, an additional 37% of patients having only a few microscopic foci of disease left behind. So a major pathologic response in almost 60% of patients. Considering what we have felt to be the chemosensitivity or radiation sensitivity of this disease, these are impressive results.

Another important outcome from this study, which I think I will close with as a question later, is that the brain was a major site of relapse in this trial. This has been true for all similarly designed studies, as well as studies of definitive chemoradiation. It raises the question of the potential value of prophylactic cranial irradiation.

Slide 11:

The other important finding from this relatively large Phase II trial, is that 26% of patients had only stable disease by CT scan so they did not meet radiographic criteria for response. And yet of these patients who went to surgical resection, the major pathologic response, pathologic CR or near CR, was actually found in almost half the patients. This tells us that CT scanning is not a good way of assessing the efficacy of combined modality therapy.

Overall, the long-term survival in this trial did not correlate with radiographic response. The best predictor of long-term survival was eradication of mediastinum and lymph node disease by this preoperative therapy.

Slide 12:

This slide shows the survival curve of patients who initially had positive nodal disease. If the disease had been eradicated by the preoperative chemotherapy and radiation, the five-year survival rate was 33%. If there was residual disease, it was 11%. To me, this is the best rationale for the current intergroup study, because it raises the question of whether post-chemoradiation surgery is therapeutic or merely prognostic, telling us who has already been cured perhaps by the chemotherapy and radiation.

Slide 13:

This same paradigm has been followed in the treatment of superior sulcus tumor. This is an abstract presented both in the plenary session at the American Thoracic Society as well as ASCO a few weeks ago, in which the identical therapy to SWOG 8805 was employed prior to surgical resection in patients with superior sulcus tumor, who did not have mediastinal involvement.

Again, it was a large trial requiring the entire intergroup effort, 116 patients. There was pathologic CR or near CR in 57.5% of patients, downstaging in 63% of patients, with unprecedented survival for superior sulcus tumors, 50% in both T₃ and T₄ subsets.

Drs. Rusch and Ethan Dmitrovsky, who is in the audience, are performing molecular correlative studies to see if we can determine who benefited, and who perhaps did not, from this approach. The question has been raised whether this represents a new standard of care in the United States. We will never have enough patients to do a randomized trial on superior sulcus tumors. This is the first large prospective, multi-institutional trial.

Slide 14:

Now moving to definitive chemotherapy and radiation, in my opinion, this is where we are today. The goals of local control and eradication of distant micrometastases are often in conflict in terms of the chemotherapy and radiation regimens that have been designed. Sequential platinum-based chemotherapy followed by radiotherapy improves survival compared to radiation therapy alone. I think almost all would agree to that. Recently, it has been reported that the concurrent use of chemotherapy and radiation is superior to sequential therapy. I'll briefly review these two recent trials. However, various combinations of full dose chemotherapy and concurrent chemoradiation may offer the best outcome. Again, I'll talk about some recent studies in this regard.

Slide 15:

These are the approaches: chemotherapy given completely before radiation, chemotherapy given concurrently with radiation, and then two ways of sequencing chemotherapy and concurrent radiation, either induction first, or consolidation after.

Slide 16:

The confirmatory trial for proof of principle that chemotherapy followed by radiation is superior to radiation alone is RTOG study 8808 which had three arms; 1) standard radiotherapy, 2) twice daily hyperfractionated radiotherapy, and 3) cisplatin plus vinblastine, a drug which we would not consider to be active today in non-small cell lung cancer, followed by radiation.

Slide 17:

This study did confirm the superiority of chemotherapy prior to radiation, but the five-year results of this trial, with less than 10% survival in all arms of the study, show clearly that we need to move forward from this approach.

Slide 18:

Recently there have been two randomized trials reported, this one in full publication in the Journal of Clinical Oncology by the West Japan group led by Dr. Furuse, comparing concurrent versus sequential chemoradiotherapy. Although the actual regimens could be criticized for the chemotherapy and the radiation split course that was given in the concurrent arm, the survival is clearly better. The five-year survival in the concurrent arm crosses that threshold of 10%.

Slide 19:

At ASCO a few weeks ago, Dr. Curren presented the preliminary results of RTOG 9410 B that same sequential regimen, the same regimen given concurrently from day 1, and then a concurrent chemotherapy with hyperfractionated radiotherapy.

Slide 20:

This is a preliminary analysis. The best survival was with the concurrent once-daily radiation, but not reaching statistical significance. Also importantly, the previous encouraging data regarding hyperfractionated radiotherapy with this particular regimen did not achieve better survival, but was more toxic. Interestingly though, there was better local control.

Slide 21:

Where do we move forward from these data? In my opinion, these are the four areas to explore in the immediate future. The first is to optimize the currently available therapies, new chemotherapeutic agents, improved radiotherapy, and reduce toxicity.

Second, optimize therapy for individual patients or special patient subsets. We have an opportunity now for what I refer to as molecular profiling, to be able to assess a set of genes, perhaps by microarray, and determine what might be the best therapeutic approach for an individual patient sitting in our office.

We need improved staging. Also, how do we better define these patient subsets, at least in Stage III disease, for different therapeutic strategies? It is imperative that our future trials maximize the patient and scientific resources available to us. Incorporation of correlative studies is imperative. Tomorrow we will hear a presentation on functional imaging, a better way perhaps of assessing the impact of our new therapies. Lastly, and most importantly, there are now a host of novel therapeutic agents directed against newly defined molecular targets. How should they best be studied? I'll spend the last few minutes discussing those issues.

Slide 22:

These are the relatively new chemotherapeutic agents that have been incorporated into chemoradiation protocols, and also preoperative protocols in this disease.

Slide 23:

They are all more active than the drugs we have had before. The critical issues B and Dr. Turrisi is not with us today, so I had to put a little of my radiation hat on to address some of these issues -- but how do you incorporate these drugs in regard to dose? Is your goal radiosensitization with concurrent therapy, or higher cytotoxic doses to eradicate distant micrometastases?

We know from some previous studies, such as the Schaake-Koning trial for EORTC, that low dose concurrent cisplatin in that trial improved survival not because it impacted on distant micro metastases, but only through better local control. What should the appropriate schedule be? And what should be the sequence with radiation?

Slide 24:

I would like to highlight only two recent studies incorporating these new chemotherapeutic agents, because they offer us both sides of the coin in the paradigm of combinations of sequential and concurrent therapy. The first of those is a randomized Phase II trial by Everett Vokes, CALGB 9431, in which a novel design was used of induction chemotherapy with an investigation drug plus cisplatin, followed by concurrent chemoradiation with reduced doses of the investigational agent.

Slide 25:

This was given in a relatively rapid fashion, with the ability to assess for toxicity, although not designed to definitely compare these arms. The three drugs were gemcitabine, paclitaxel, vinorelbine given at full dose and reduced dose.

Slide 26:

Again, remembering that this study was not designed based on sample size to give us definitive comparisons for efficacy, the results were about the same.

Slide 27:

But there were differences in toxicity with each of these new agents incorporated with platinum and radiation.

This is the toxicity of the concurrent therapy -- neutropenia was greater with paclitaxel; thrombocytopenia, gemcitabine; neuropathy with vinorelbine. And esophagitis, usually the dose limiting toxicity of definitive chemotherapy and radiation in 50% of the patients with the incorporation of the potent radiosensitizer, gemcitabine. Is gemcitabine like methotrexate, a very potent radiosensitizer but as potent for normal tissues as it is for the tumor?

Slide 28:

In the second of these studies in the paradigm, SWOG 9504, which was presented at this year's ASCO meeting, the opposite sequence was done. Concurrent chemotherapy and radiation with cisplatin plus etoposide and radiation, in this case followed by the new taxane, docetaxel. Radiation was given by conventional once a day fractionation.

Slide 29:

The rationale for this trial was to maintain an effective core of concurrent chemoradiation where we had five-year survival and toxicity data from our previous SWOG trial, but to substitute docetaxel for what would have been two continued cycles of cisplatin/etoposide following chemoradiation at a time in which we had observed in the previous study a large number of relapses. The principle here is taxane sequencing prior to the emergence of clinical drug resistance based on potential molecular mechanisms of taxane use in this setting, and the clinical activity of this particular drug, docetaxel, in second line therapy.

Slide 30:

The pathologic staging requirements are relatively unique for this study for Stage IIIB disease: N₃ disease, requiring pathologic documentation, T₄ pathologic documentation, or clear cut involvement on CT or MRI of other structures, and pleural effusion cytology negative. So a relatively homogeneous group of patients compared to most other clinically staged trials in IIIB disease.

Slide 31:

The survival in 83 patients in this trial is very encouraging. The median survival, 22 months; the two-year survival, 50%. As you can see, a considerable proportion of patients are on the long-term end of the curve, and there have been few relapses within the last year.

Slide 32:

To compare this to the opposite approach, Dr. Vokes study of induction followed by concurrent therapy, this is the toxicity analysis of the SWOG trial. Neutropenia during the consolidation docetaxel is very common, but of brief duration. The incidence of pneumonitis was 10%. You will recall in Vokes trial with all three agents it was 10-15%. There were three fatal late pneumonitis cases, two probably radiation-related. This should be addressed in our discussions as we go through this session. Esophagitis occurred much less frequently than in the opposite approach, at least based on this comparison.

Slide 33:

Current cooperative group trials in the United States are looking at the issues that I described earlier. In CALGB; how do we optimize schedule, concurrent versus sequential/concurrent. RTOG; reduced toxicity, amifostine. ECOG; optimize radiotherapy, hyperfractionated versus standard.

Slide 34:

These are the trial designs: CALGB, concurrent chemoradiation alone, or induction followed by concurrent therapy.

Slide 35:

RTOG 9801, induction chemotherapy followed by concurrent with hyperfractionated radiotherapy with or without amifostine for cytoprotection, and the

Slide 36:

ECOG trial of induction chemotherapy followed by either standard radiotherapy or hyperfractionated accelerated radiotherapy.

Slide 37:

In the year 2000, where do we stand with definitive chemotherapy and radiation? This is from Kathy Pister's presentation as an ASCO discussant last month. We have moved forward. This is her analysis. I guess this is a bit of a mini metaanalysis B radiation alone, sequential chemotherapy followed by radiation, concurrent, and then the two paradigms I talked about, combining sequential and concurrent. Two-year survivals have moved up. Toxicity has also moved up.

Slide 38:

In the future it is quite likely that these sorts of manipulations of classic chemotherapeutic agents and radiation will take us forward, but at a very slow pace. Instead, it is the incorporation of novel therapeutic agents that I think will really take us to the next level for survival.

Agents with unique mechanisms of action, molecular targets independent of classic chemotherapy and radiation, many of which surprisingly have synergistic interactions with chemotherapy and radiation.

Slide 39:

And there are a host of new classes of these agents entering clinical trials. I won't belabor these.

Slide 40:

I would like to close by talking a little bit about how we might incorporate these agents.

In my opinion, it is unlikely that they will be used as single agents. Instead they will be used in combination with the standard therapies for that clinical setting. One issue is how will they be used, concurrently with or following chemoradiation, or in the terms of surgical resection, preceding or following surgical resection, which would allow a window of opportunity to test biologic correlates and tumor tissue.

Slide 41:

I would like to close just by showing an example of one these, and how it is being employed in clinical trials sponsored by the National Cancer Institute. Tirapazamine, a hypoxic cytotoxin that is selective for reduced and hypoxic cells, has preclinical synergism with both chemotherapy and radiation. There is also a positive randomized clinical trial in advanced stage disease.

Slide 42:

This is the study design for a Phase II trial by the California Cancer Consortium, NCI-sponsored, incorporating tirapazamine into chemoradiotherapy for unresectable Stage III disease. Patients are treated with chemoradiation with tirapazamine, followed by chemotherapy with tirapazamine.

Slide 43:

Important ancillary correlative studies - I emphasize this point as an example of what I think we need to be doing in the future: functional imaging, laboratory correlates, functional imaging with FDG for tumor metabolism and oxygen-15 water for perfusion, tissue correlates for what might be important molecular correlates of activity for both tirapazamine and the chemotherapy employed, serum analyzed serially for what may be a marker of tirapazamine effect, PAF-1 or plasminogen activating factor, and also analysis in serum of free circulating DNA for beta tubulin.

Slide 44:

The rationale for the PET ancillary study is shown here, and the inadequacy of CT scanning for restaging and preoperative therapy compared to pathologic response. Recent data suggest that PET restaging may predict for survival after chemotherapy and radiation, and the potential use of functional imaging to assess chemosensitivity in preoperative patients.

Slide 45:

These are the data from this year's ASCO from McManus, 56 patients with Stage III disease were treated with definitive chemotherapy and radiation. A PET scan was performed at baseline and repeated two months after. Twenty-four of these patients, or 43%, achieved a PET scan CR. This predicted improved survival at two years, 84%, versus anything less than a CR by PET, 31%. CT scan response did not predict survival.

Slide 46:

Laboratory correlative studies. This is the approach used by the California Cancer Consortium. We call it a three-tiered approach in which new agents are studied pre-clinically in vitro or in vivo. In all patients entered in all studies there are certain baseline tissues available, and serial serum samples which would be obtained, and then in selected patients, a serial biopsy.

Slide 47:

In this trial of tirapazamine with chemotherapy and radiation, there are correlative studies ongoing of taxane response and resistance. Again, I would like to close just by highlighting two of those; p27 and beta tubulin.

Slide 48:

This is a proposed p27 pathway for oncogenesis and therapeutic response. Dr. Gumerlock will cover this in one of the breakout sessions later today. But there is a growing consensus that there is a central role for this CDK inhibitor in treatment-related apoptosis by a variety of inducers including taxane therapy.

Slide 49:

These are the data from Paul Gumerlock at UC-Davis, for the first time demonstrating that p27 induction is independent of the bcl-2 phosphorylation previously described as a potential mechanism for taxanes. Here at low dose, no bcl-2 phosphorylation is seen, but strong induction of p27.

Slide 50:

Again, this will be covered in more detail later today.

This is the p27 experiment showing that mitotic phase p27, which should not be present, is present, is strongly induced, and is associated in flow cytometry with apoptosis.

Slide 51:

Beta tubulin mutation -- striking data presented last year, published in the Journal of Clinical Oncology by the Rosell group. Patient tumor specimens analyzed for mutational status of beta tubulin. One-third were mutated. None of these patients, 0 of 16, responded to paclitaxel. The median survival was very short; so one of our most effective chemotherapeutic agents

Slide 52:

and a potential predictor for response.

At this year's ASCO meeting, Dr. Rozell and Dr. Monzo presented data showing that beta tubulin could also be isolated, analyzed and mutational status determined from serum specimens of free circulating DNA.

Slide 53:

Paul Gumerlock at our institution has shown for instance that k-ras can be identified. These are extraction protocols. This is the preferred method, method 3, in normal blood donors. So this is the normal k-ras gene in a blood donor.

Slide 54:

Serum tumor DNA then offers considerable potential for application in clinical trials to identify a molecular profile when tissue is unavailable. Change in gene status; data from the Baylin group on change in methylation status as a tumor marker to identify residual disease after surgical resection. And it does not require tissue.

Slide 55:

So this can be done routinely in our correlative studies.

The hypothesis of these laboratory clinical correlative studies are shown here: that patient response and survival will be influenced by the underlying molecular profile in individual patient tumors, or perhaps subsets. That this information can be exploited to optimize new therapeutic approaches to the cancer patient. And the potential application of new technologies such as microarray.

Slide 56:

Slide 57:

These are two trials, either proposed or ongoing, further evaluating novel therapeutic agents. I may have this design a little wrong. The ECOG people will tell me. But the randomization is to thalidomide or not, following chemoradiation. Thalidomide will also be given with the chemoradiation.

Slide 58:

This is a proposed intergroup trial of the EGFR inhibitor ZD1839 following chemoradiotherapy. SWOG, NCI Canada, the North Central Group, and proposes to test chemoradiation followed by placebo or this novel therapeutic agent.

Slide 59:

In closing, there are many unanswered questions; probably more unanswered than answered at this point. What is the optimal staging and restaging? What is the role of functional imaging? Defining patient subsets for different therapeutic options. Preoperative therapy versus definitive chemoradiotherapy. The optimal scheduling of chemoradiotherapy. The best chemotherapy and radiation.

The issues, I think a very cogent one for these sorts of discussions, is prophylactic brain radiation, the role of predictive molecular markers, and how to best integrate novel therapeutic agents.

Thank you very much.

[Applause.]

Slide 60:

DR. SAXMAN: We have time for questions.

DR. GANDARA: I am reminded by the quote by Enrico Fermi, who said, "Before hearing this lecture I was confused. After hearing this lecture, I remain confused, but on a higher level." The Nobel laureate had some meaning. For me, it's still six o'clock in the morning, California time, but you should be more awake.

DR. SAXMAN: I had a question, David. One of the issues that has been discussed has to do with combinations of chemotherapy as radiosensitizers given concurrently with radiation in locally advanced disease. For reasons that you stated, most of the regimens build on some platinum-based analogue. The problem with that of course is that the platinum’s given concurrently with radiation have some significant toxicity, and it's often difficult to add new chemotherapeutics, at least in reasonable doses.

So one of the questions that has arisen is are we stuck with platinum-based analogues as concurrent therapy, as sort of our foundation for building upon this? Or are there ways, and is it reasonable to test other chemotherapeutics given in higher doses, but without building on the platinum data?

DR. GANDARA: Not with cisplatin, but it creates both a scientific and an ethical dilemma. In limited small cell lung cancer where you have a potentially curable subset, and at least in advanced stage disease, and actually in Stage III disease as well, where metaanalysis suggest that platinum is the critical factor for improved survival, it requires innovative trial designs to decide how we test, in this case, non-platinum regimens, or how we integrate novel therapeutic agents?

The novel therapeutic agents, I think in particular, we should not be stuck on use of MTD. Many of these drugs will likely have threshold effects for biologic activity. I think in our deliberations today we need to go through a number of trial designs about how to incorporate them into a stage of disease where some of the patients are potentially curable. So we're not stuck with platinum, but from a clinical standpoint right now, the data suggests it offers the best chance for long-term survival.

DR. WILLEY: What are the criteria you use to select patients for sequential pathological testing?

DR. GANDARA: The first thing is having tissue access. That is the reason surgery wasn't the highlight of this State of the Science Meeting. I felt we needed to spend some time on the pre-operative approach, because it gives you tissue from the surgical resection post-treatment, and you will have some sort of a biopsy pre-treatment.

Dr. Saxman mentioned that small cell lung cancer was particularly difficult in terms of getting serial tissue biopsies, but non-small cell lung cancer is not far behind. Many of the initial diagnoses now are made with fine needle aspiration, small amounts of material. Also there is inability in many instances to have fresh tissue, unless it is built into prospective design.

So for Stage III non-small cell lung cancer, it would require trying to obtain some fresh tissue in your initial diagnosis, for instance by bronchoscopy, and then having some post-treatment tissue. In a pre-operative trial, you should have that. In a definitive chemoradiation trial, we rarely have it, and we need either a surrogate tissue, or another way of measuring what's going in the tumor. That's why I bring up the issue of circulating DNA, at least if the DNA were a way to do that.

It's a considerable challenge. In our studies we have pushed very hard to do serial biopsies when they are feasible. It again raises ethical issues for patient care. It also raises considerable cost issues. The National Cancer Institute for the first time now, has a translational research fund which will assist in studies such as the one I described, in actually paying the radiologist, the pathologist, everything else that goes along with trying to do these sorts of correlative studies. So it's not an easy answer.

DR. BUNN: There is a bit of a Catch-22 in the meeting design here. The meeting is about locally advanced disease, and then we're talking about incorporating new agents. Of course the question is, are new agents going to be incorporated first in Stage IV disease? Because it's a little less complicated to have just one agent with the experimental approach versus two.

I think as we have discussions during the meeting, it's going to be important. It's a bit daunting for companies developing these products to think about going to the FDA where you have two modalities with the new agent, as opposed to just one modality. Part of our discussion today should be whether these agents going to be tested right away in locally advanced disease? As you mentioned, there are some reasons why that might be particularly useful when you can get some tissue. Or whether these agents are really going to wait until we get done with Stage IV disease, to see which ones sort out before they are incorporated into Stage III disease.

DR. GANDARA: That is an excellent point. In view of the time I had allotted, I actually took out some slides, and some of these issues were on them. With many of the novel therapeutics, it is likely that the greatest impact will not be in advanced widely metastatic disease. It will be in earlier stage disease, perhaps even as some sort of prolonged maintenance therapy, which has not usually been thought of as effective with classic chemotherapy but which may be for some of these agents.

The opportunity for studying these agents, in large Phase III trials at least, such that you would have a definitive answer for a novel therapeutic, are limited. We need to move forward though from just interactions of chemoradiation, and reserve some of our studies and some of our NCI-sponsored trials for incorporation of the novel therapeutics in earlier stage disease -- and it really means surgically resected patients earlier than Stage III. That may be the optimal place to show proof of principle for some of the novel therapeutic agents we will be talking about.

DR. SHEPHERD: I certainly agree with you that we should be moving these into earlier stage disease. I think one of the problems that we have to recognize is that development costs for these agents are huge, and the pharmaceutical industry doesn't want to do a Stage II disease, or even a Stage III trial when they know that they won't get their answer from it until maybe five or ten years down the road, and they are trying to bring these agents to market.

So I think we have to sympathize a bit with the financial aspects of developing these new agents. But there must be other ways then to support doing the trials in earlier stage disease, where I think most of us feel that they will probably have their greatest role to play.

DR. GANDARA: I agree entirely, Frances. And again, that's the reason I spent a considerable amount of time today talking about pre-operative therapy, because if it didn't compromise the surgical curability, a window of opportunity study with a novel therapeutic could be done. Then you would have the tumor and the bucket to analyze for biologic effects on the presumed molecular target.

It would be worthwhile having comments from some of the people like Jack Roth about what is an acceptable delay before surgical therapy in a patient who is not undergoing what we consider relatively effective chemotherapy, but a completely untested novel agent. And would we even want to do that in early stage patients? This is all assuming we have adequate toxicity data already on the agent from advanced stage disease.

DR. ROTH: I think we certainly have delays of months in terms of our induction regimens. Obviously, the patient is being treated during that time. But certainly with the usual radiation/chemotherapy regimens, we have at least a month delay between the end of therapy and the beginning of surgery. That's certainly very acceptable, and perhaps even somewhat longer delays might be acceptable as well. But I think it does bring up an interesting issue in terms of surrogate endpoints, and the need to develop these as possible approval points for some of these agents. I would just ask you, David, and some of the others, what are potential valid surrogate endpoints that we might have for these studies?

You have pointed out for example, pathologic complete response, or almost complete response, as correlating very nicely with survival. Would that be an adequate surrogate endpoint at the end of that induction study for example, with some biologic agent? For example, if you were comparing this in a randomized study and you saw a significant increase in the pathologic CR rate, would that be an acceptable endpoint for the study?

DR. GANDARA: I think if the pathologic complete response rate correlates with improved survival, then that would actually be a fairly clear-cut one. The issue is how do we incorporate functional imaging? Could functional imaging be a surrogate endpoint? Obviously if we can measure something in the serum or in the tumor at a pre- and post-biopsy, there are a myriad of potential surrogate endpoints that could be utilized, depending on the agent in question.

These are issues that we have to think about, both from a scientific standpoint and science, and many of these agents come from industry. What will the FDA think about surrogate endpoints? Often we are trying to get a drug approved, and how do we approach it from that standpoint, too?

Although I'm not a prostate cancer physician, there has been tremendous controversy about even the use of something like PSA for prostate cancer. In lung cancer we are nowhere near as far along as we are with PSA and prostate cancer.

DR. SAXMAN: Thank you very much. Our next speaker this morning I want to particularly thank for coming and being part of the meeting and giving the presentation. As any of you who read the original agenda that was sent out know, Dr. Drew Turrisi was going to give a talk next. Unfortunately, Dr. Turrisi had emergency surgery last week. I was told he had a coronary artery bypass surgery. I'm also told that he's fortunately doing well, but could not be here today.

So I think we are very lucky to have Dr. Salgia here this morning. He graciously agreed to do this at the last minute, but I think it's going to be extremely valuable for our discussions later this afternoon.

So Ravi Salgia is Assistant Professor of medicine at the Dana-Farber Cancer Institute. He is going to speak with us this morning about molecular abnormalities in lung cancer and their potential for novel therapeutics.