| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
J Thorac Cardiovasc Surg 2008;135:1234-1236
© 2008 The American Association for Thoracic Surgery
Invited Commentary |
The authors are clearly presenting some very impressive results here in terms of a 70% five-year survival in the 25% of patients who were complete pathologic responders. That is really sort of best in class. Also, they really demonstrate the fact that en bloc resection results in a very low local recurrence rate.
I have 2 or 3 questions, however, and I think Dr DeMeester is going to address those.
First, you have described candidates for neoadjuvant therapy as having locally advanced disease proved by both endoscopic ultrasonographic analysis and PET scans, yet there are no data in the presentation or in the manuscript on the TNM staging of these patients. Can you provide some information as to their pretreatment TNM status?
Dr DeMeester. The problem with pretreatment staging is that the clinical staging for esophageal cancer is terrible. Many of these patients in the earlier time frame did not undergo PET scans or endoscopic ultrasonography. We do that routinely now, and as you know, our preference is primary surgical resection in anyone who we think has limited local regional disease, which for us, is patients who had any depth of tumor with less than 5 obvious nodes determined by means of endoscopic ultrasonographic analysis or PET scan. For anyone with 5 or more nodes, we know that the systemic failure rate with surgical intervention alone in our experience exceeds 80%, and therefore those patients will need chemotherapy either before or after surgical intervention. We talk to the patient about that, and many of these patients are selected then to have preoperative chemoradiotherapy, followed by surgical intervention. All of our patients who have neoadjuvant therapy have extensive local-regional disease. In fact, if you look at the final pathologic staging, some 60% of the en bloc group had stage II or III disease after neoadjuvant therapy, and therefore the majority of these patients started with significant disease.
Dr Whyte. My next question sort of relates to how patients were assigned to receive an en bloc resection or a transhiatal resection. In the presentation we heard that there is use of the number of lymph nodes to sort of ascribe one operation versus another. However, in the manuscript you talk about it both being related to the age and the presence or absence of comorbidities. I am a little confused then as to whether you used the presence or absence of lymph nodes or the number of lymph nodes to help determine whether patients receive an en bloc resection or a transhiatal resection.
Dr DeMeester. I am sorry if that was confusing in the presentation. We are merely demonstrating that, as should be no surprise, en bloc resection removes far more lymph nodes than transhiatal resection, but the number of lymph nodes has nothing to do with the selection of the operation. The operation is selected on our assessment of the patient's physiologic ability to withstand an en bloc procedure, which is a bigger deal, and therefore patients who are older than 70 or 75 years and have any renal, cardiac, or pulmonary dysfunction are selected for a transhiatal resection. That then brings in the concern that patients in the transhiatal group have higher comorbidities and are dying off as a consequence of these comorbidities and not their cancer. We also prepared cancer-only survival statistics and have demonstrated that the statistics and the significance between the curves are the same. In fact, all deaths, except for one in the transhiatal group, were from cancer. Therefore the comorbidities are what we use to select the procedures, but that had no effect on the outcome from the procedures.
Dr Whyte. So you are not using the presence or absence of nodal disease or the number of nodes to assign someone to en bloc versus transhiatal resection.
Dr DeMeester. No. We use the number of nodes to determine whether we should select patients for neoadjuvant therapy, but that has nothing to do with whether we select them for transhiatal or en bloc resection. Our preference is always en bloc resection, but we will downstep to a transhiatal resection in patients who we fear will not handle the en bloc resection based on comorbidities.
Dr Whyte. My last question then is this: Why don't you do an en bloc resection on everyone? Your mortalities are pretty similar. In terms of the ASA class you presented, they are roughly the same in the transhiatal versus the en bloc resection. As for the presence or absence of comorbidities, the numbers are fairly similar. You had 11 of 40 patients who had significant comorbidities in your en bloc resection group. I am thinking that persons who undergo a transhiatal resection have a laparotomy, a gastric mobilization, a pretty extensive mediastinal dissection, and then a neck incision, and the patients undergoing en bloc resection get a bit more extensive abdominal dissection and then perhaps a rather limited thoracotomy but, in addition, a much more precise mediastinal dissection. I am wondering why you do not do an en bloc esophagectomy in everyone, particularly because you are clearly convinced it is the right operation. Your data indicate that they have better long-term survival, and the mortalities are somewhat similar. Why not do an en bloc resection on everyone? Then we can really see—sort of take out this transhiatal versus en bloc issue—and focus on whether the potential increased morbidity of the bigger operation is more than made up for by the better long-term survival in these patients.
I really enjoyed the paper, and it was well presented. Thanks very much.
Dr DeMeester. We are pretty much coming to that same conclusion. Essentially everyone who comes for primary surgical resection has an en bloc resection at this point. I have done 1 transhiatal resection maybe in the last 3 years, and that was in a guy who had a previous pneumonectomy. Therefore essentially everyone gets an en bloc resection for primary therapy. Part of this experience now is teaching us that even after neoadjuvant therapy, we need to push the en bloc resection because part of our attitude initially, as I think was present in many surgeons, was that if you had neoadjuvant therapy, that was supposed to take care of the problem, and you could probably just get by with a transhiatal resection. Therefore 5 or 10 years ago, when we were first doing this, often we would do a transhiatal resection in a patient who could have had an en bloc resection, but we thought they have had neoadjuvant therapy, the chemoradiation increases the potential morbidity particularly for a thoracotomy, more respiratory complications, maybe it is not worth doing that. These data now will tell us that anybody who comes with neoadjuvant therapy at USC will get an en bloc resection if at all possible.
Dr Robert Cerfolio (Birmingham, Ala). Steve, excellent work. I congratulate you and your efforts to continue to collect prospective data on esophageal cancer. It is a lot of work, and we appreciate everything you have contributed to the field and to our patients.
This thing about complete responders—the only way to know who is a complete responder is to take the cancer out and look. We have generated some data on 170 patients now—we prefer Ivor–Lewis esophagogastrectomy, as you do—and have shown that the change in the maximal standardized uptake value helps predict who will be a complete responder and who will not. When the primary tumor decreases by greater than 75%, that patient is very likely—has a greater than 90% chance of being a complete responder. I am wondering whether you have any data that would corroborate that, and then, if you do, if you would comment on the fact that I get a lot of medical oncologists who say, "If he is a complete responder, he does not need surgery." The ones who refuse or decline it because they are a "complete responder" all seem to return within 9 to 12 months with bad local and often distant metastatic disease. Can you tell me how you respond to your oncologists and push the patient to get the operation and if you have maximal standardized uptake value data?
Dr DeMeester. We are just starting to look at some of the maximal standardized uptake value data. We do not certainly have that going back in this study population, particularly because a number of these patients had their operations before PET scans were widely used. I will say, though, that I am not convinced that PET scanning is ever going to get down to the level of the cell. There is just no chance that you are going to predict 100% of the time who has no residual cancer. I do not think there is any test short of the pathologist with his microscope sorting that out. The reality is, if you leave cancer behind, the patients are going to die from their cancer. It is going to come back. Therefore I do not see a way with any of our current studies to rule out or to find those patients who do not need surgical intervention because of complete pathologic response. We know that the mucosal disease clears up first. If you cannot clear mucosal disease with neoadjuvant therapy, then you have real problems. In our data 60% of patients had residual disease in the esophagus, and therefore the majority of patients do, and it should be resected with an en bloc procedure.
Dr Cerfolio. Therefore you tell the oncologists that you do not know whether the patients are complete responders no matter what their repeat endoscopic ultrasonographic tests and repeat PET scans show, and thus you, like us, recommend surgical resection.
Dr DeMeester. Repeat endoscopic ultrasonography has been shown to be of no value after neoadjuvant therapy, and therefore I do not even offer it. I tell the oncologists that 60% of these patients have residual disease in the esophagus and 40% will have nodal disease, and if you read a recent article by Prenzel using immunohistochemistry, 75% of patients after neoadjuvant therapy have either histologic or immunohistochemical evidence of nodal disease. The vast majority of these patients have disease and should have an esophagectomy. Now the issue you brought forward is that many oncologists have taken an approach that surgical intervention should not be offered after neoadjuvant therapy because if you get a complete pathologic response, your survival is pretty good, and if you did not and you get a transhiatal resection, your survival is horrible, and therefore why should they undergo an operation at all after neoadjuvant therapy? That is the emphasis of this paper. Your survival is not horrible if you do a good operation. Fifty percent of patients were alive at 5 years after neoadjuvant therapy with an en bloc resection, and even with residual disease, it was an impressive survival. These patients should have an esophagectomy, and it should be en bloc.
Dr Cerfolio. Our data are similar, and we agree.
Dr David Follette (Sacramento, Calif). Steven, I enjoyed your paper, and it is one of the first to corroborate some data from East Germany regarding complete responders and survival in squamous cell carcinoma of the esophagus, which we do not see that much of anymore. We really still do not know the benefit of neoadjuvant therapy because the Walsh study is one of the only ones to show the advantage and, as was mentioned in the talk, is severely flawed. In my mind, what is coming to the fore is this issue of a complete response or a partial response. We have seen good long-term survival in persons who are nonresponders to neoadjuvant therapy with good survival rates with just definitive chemotherapy and radiation therapy. My question is, you briefly mentioned to Rob that you do not do repeat ultrasonographic analysis on these persons, but do you use CT scanning restaging after neoadjuvant, and if they are nonresponders or if the tumor is fed and growing with the neoadjuvant, how do you handle that group of patients?
Dr DeMeester. We definitely restage them, looking for systemic disease. If a patient has systemic disease, we do not offer him or her esophagectomy. If the patient does not have systemic disease, the extent of the local regional disease does not dissuade me from moving forward. It is important to separate out squamous tumors from adenocarcinomas. All these were adenocarcinomas. You cannot really do an en bloc resection for squamous tumors. You have millimeters between the trachea and major vessels, and therefore it is a completely different disease, and for those patients, I think definitive chemoradiotherapy often is an advantage. It is totally different for a distal esophageal adenocarcinoma. Most Americans have centimeters of fat between any significant structure and the esophagus, and you can get a very excellent resection and take out the lymph nodes, and there is no comparison between the results of good operations and definitive chemoradiotherapy in patients with distal esophageal adenocarcinoma.
Dr Alex Little (Dayton, Ohio). I have one quick follow-up question to something that was put before. Is the difference in the operation because of the extent of the abdominal dissection or because there was not a thoracic nodal dissection?
Dr DeMeester. It is brought out in the manuscript, but our abdominal dissection is identical whether you have a transhiatal or an en bloc resection, and therefore it is the mediastinal dissection. As you saw, about 45% of patients had mediastinal and abdominal nodes in the en bloc group, and therefore there is a high prevalence of mediastinal nodes that I just do not think you can clear well, leading to the higher recurrence rate and the failure of survival in the transhiatal group.
Dr Little. If I might, I have one quick comment to ask you to react to, and it has already been alluded to. I think fully convincing evidence will not be put before us until all comers are randomized: to compare a relatively fit with a relatively unfit group sort of stacks the deck ahead of time.
Dr DeMeester. That would be true if they were dying from their comorbidities, but as I said, there was 1 death from comorbidities, a non–cancer-related death, in the transhiatal group. All the rest of the deaths were from cancer, and if we analyzed it based on cancer-related deaths only, the statistical significance is actually even greater. In my mind the issue is closed. It is decided. The argument is over. En bloc esophagectomy is superior to a transhiatal resection.
Related Article
J. Thorac. Cardiovasc. Surg. 2008 135: 1228-1236.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
