Discussion

Article Outline

• Copyright

Dr Robert Ginsberg (Toronto, Ontario, Canada). I would like to congratulate Dr DeCamp and colleagues on a beautifully presented study. I have a few comments.

The title of the talk appears to assert that this induction chemoradiotherapy improves survival. I’m not sure that this has been proved, because there are no valid data for comparison. Be that as it may, I have a few questions for Dr DeCamp.

First, having had the experience of being in the SWOG study, was there any difference in selection of your patients with positive mediastinal lymph nodes? In other words, did you accept patients with single-station lymph node disease? If so, how many were there? Only 105 patients in a 6-year period were entered into this trial. What happened to the other patients with stage IIIA and IIIB disease at your institution during this time? How often was there multistation bulky nodal disease among your patients?

When you compare your data with those of SWOG to conclude that this accelerated chemoradiation improves survival, I have some questions. In your induction therapy all patients necessarily had to be hospitalized for their infusion, and 40% also had to be hospitalized because of toxicity. This occurred before the operation in the induction phase and after the operation in the adjuvant phase. Please comment on that and its comparison with the SWOG 4-week chemoradiation preoperative protocol. When you add up the figures, your protocol is a 12-week protocol, and many induction chemoradiation protocols nowadays are 4 to 5 weeks of induction chemoradiation followed by a 3- to 4-week rest, followed by surgery. That’s a 10-week protocol.

Your resectability figures were unchanged, your patient compliance figures were unchanged, and your downstaging was unchanged relative to the SWOG trial—a very simple protocol, and the time of treatment, even though it was “accelerated hyperfractionation” of the radiotherapy, was identical to SWOG at 12 weeks.

When you say that you’ve improved survival, what’s the certainty factor? It’s a decade later, and certainly all of us have had patient selection changes in our induction therapy once we have experience with more patients. So how certain can you be that you’ve improved survival? Maybe you’ve just improved selection.

I also would like to know your center’s philosophy on resection of T4 and N3 disease. At operation no attempt was made to dissect the N3 nodes. How do you know that you didn’t have N3 disease left behind, even though the N2 nodes were sterilized?

Also, you had 7 T4 tumors, and none of these patients underwent an extended resection. Four of the T4 tumors were inoperable, and in the other 3 cases, as far as I can tell, there was no extended resection. All these patients underwent simple lobectomy or pneumonectomy. What is the philosophy of your surgical approach after induction chemoradiation? Does it not matter anymore that the original site of disease be removed? Is it okay just to remove the primary site of disease and the regional lymph nodes, leaving N3 nodes that may contain residual tumor?

Finally, now that you’ve prognosticated, how are you going to deal with the 60-year-old patient with squamous cell cancer? And how are you going to deal with the identification of persistent N2 disease, which you presume is also N3 disease, in the patients with initial N3 disease after the chemoradiation, since none of those patients survived?

Dr DeCamp. I appreciate your insightful comments, Dr Ginsberg, and certainly recognize your preeminence in this field.

Regarding your question as to our philosophy about single-station versus multistation, intracapsular versus extracapsular, or bulky disease, we did not discriminate nor did we stratify our outcome analysis according to these factors. It is our philosophy that mediastinal nodal involvement is a marker for systemic disease, and without an effective systemic form of treatment as part of our treatment plan, we do not believe that we need to discriminate between those. All of that being said, I’m a firm believer that there are good IIIBs and bad IIIBs and good IIIAs and bad IIIAs; however, we tend to enroll all patients with reasonable performance status and any N2 or N3 involvement in this accelerated schema.

It is true that this is a toxic regimen. All of the patients are inpatients during the first week of both the induction and the adjuvant portions of treatment for continuous infusion chemotherapy and for adequate hydration to prevent renal toxicity. It’s also true that 40% of patients required unexpected readmission for febrile neutropenia. There were no toxicity-related deaths before induction therapy, and all 105 patients underwent surgery on time. Seven patients could not be operated on, 6 because of progressive disease and 1 because his forced expiratory volume in 1 second was considered to limit resectability. The fact that this is toxic therapy necessitates carefully coordinated care by the surgeon, medical oncologists, and radiation oncologists, and we’re blessed to have this expertise. Though we don’t have a cost-effectiveness analysis, I think it would be an interesting comparison to look at the total cost of this accelerated regimen versus the more standard course.

I do take minor issue with your comment about our schema being similar to SWOG 8805. In their Journal of Clinical Oncology publication, the mean time to complete just the induction portion of therapy in SWOG 8805 was 40 days. Ours was 12 days, and there were no delays in getting our patients to surgery. In the entire cohort of 105 patients, 74% completed all the prescribed treatment in our schema. This was similar to the rate in SWOG 8805; however, only a fraction of their patients had any adjuvant therapy prescribed.

Moving on to T4 and N3, we had 7 patients with T4 disease, 5 on the basis of direct mediastinal invasion and 2 because of superior vena caval involvement. Both cases of superior vena caval involvement were found to be unresectable at the time of exploration, whereas most cases of mediastinal invasion were resectable. We did have one chest wall resection (T3). This is not a series that could be easily compared with others with T4 lesions involving the esophagus, the aorta, or spine. We did not routinely make an attempt to resect N3 disease at the time of our resection. We did make the leap of faith assumption that sterilization of the N2 nodes reflected sterilization of the N3 nodes. However, in our analysis we assumed that if there was residual N2 disease in a patient who originally had N3 disease that there was no downstaging. So the downstaging reflected sterilization of the mediastinal nodes.

Other investigators, including Grunenwald, have advocated an aggressive mode of resection, including sternotomy and bilateral lymphadenectomy. I have a hard time reconciling that as providing any therapeutic benefit. I do see that it would provide prognostic information that would help us better understand what we’re doing with induction chemoradiotherapy, but I think residual mediastinal nodal disease is a marker that there is residual systemic disease.

In terms of the 60-year-old patient with squamous cell carcinoma, obviously clinical judgment goes far beyond age and histologic type. I would continue to advocate careful assessment according to clinical judgment.

In terms of how we might go forward in terms of assessing patients for residual nodal disease, I think the emerging technologies of computed tomography combined with positron-emission tomography and a better understanding of the utility of positron-emission tomography to assess response may help. Another consideration would be the limited use of reoperative mediastinoscopy or videothoracoscopy to evaluate patients for residual disease after induction therapy. Resection may be futile for these patients, and other novel therapies may be more appropriate.

Dr Thomas A. D’Amico (Durham, NC). Dr DeCamp, could you share with us what you believe the role of radiation therapy in induction protocols is? There is no evidence that it adds to long-term survival. How would hyperfractionation then enhance that role?

Second, what percentage of your patients had failure because of brain metastasis, and can your study support the use of prophylactic cranial irradiation in an induction therapy protocol?

Dr DeCamp. I think that if you look at pathologic response at the time of surgery after induction, chemoradiation (SWOG 8805) versus a straight chemotherapy induction regimen, for example, the Cancer and Leukemia Group B (CALGB) 8935 study, you’ll find that there were more overall responses, certainly more complete responses and more downstaging, in the chemoradiation trials. Those patients had more meaningful survival. So I do think the use of radiotherapy with concurrent chemotherapy provides treatment synergy, and it establishes that synergy early in the course of therapy to enhance response and perhaps survival.

I don’t think we have the statistical power to support or refute the use of prophylactic cranial irradiation. We had a less than 10% local failure rate, whereas we had a 50% distant failure rate. Most of those were in the brain. I think we have defined those people who have responded and have potentially benefited from resection. Because we can predict that nearly half are going to go on to have brain metastasis, we would certainly support revisiting the use of prophylactic cranial irradiation in this subset with pathologic response and mediastinal sterilization.

Dr Douglas J. Mathisen (Boston, Mass). Our approach has been similar to yours for IIIA nodal disease and fairly closely mimics your experience. I have one simple question and one that will take a little bit of explanation on your part. We noticed that there were higher incidences of pulmonary embolism and adult respiratory distress syndrome in both the treatment phase and the postoperative phase, and I wonder whether you had any similar experiences.

The more interesting question to me is about persistent nodal disease. We have never taken the approach of doing remediastinoscopy and have accepted that there is some value in removing local disease and doing nodal dissections and have had about a 20% survival rate in that group of patients with persistently positive nodal disease, and invariably somebody makes the comment that the old CALGB study treating nodal disease with chemoradiotherapy achieved roughly a 20% survival rate. My contention has always been that in that group, their 20% survival rate was likely to have been those who had a complete response to chemoradiotherapy. Would you address the issue of those who argue that persistent disease ought to be treated with chemoradiotherapy and use that CALGB study as justification. Obviously we don’t know, but my suspicion has always been that it’s a different group of patients.

Dr DeCamp. First, regarding the adult respiratory distress syndrome issue, 4 or 5 of the 7 deaths were from respiratory failure, adult respiratory distress syndrome, and multiple organ failure. Interestingly, these were fairly evenly divided between lobectomies and pneumonectomies. So it wasn’t just the fact that a patient had undergone a pneumonectomy after chemoradiation. We are acutely aware, even paranoid if you will, about that risk and try to minimize the inspired oxygen fraction during the time of resection. We also use a couple of doses of steroids. I know that’s an anecdotally supported practice, and I don’t have any data to support it. It makes us feel a little better to try to preempt, if you will, that difficult problem.

Regarding the role of surgery even with persistent nodal disease, I agree with you. The old CALGB nonsurgical trial had a substantial local recurrence rate. We have substantially reduced that local recurrence problem. We may be simply shifting the natural history of the disease and demonstrating that with surgery and radiation we provide exquisite local control. The Achilles heel in the treatment schema for these patients is the lack of good, effective systemic therapy. So I do think there is a role for resection. I’m not sure that the role extends to N3 disease. I think that patients who have residual N3 disease are truly nonresponders. I’m not convinced that subjecting them to the morbidity of resection, with or without bilateral lymphadenectomy, is in their best interest.

Dr Douglas E. Wood (Seattle, Wash). I don’t understand the radiobiologic sense of a split course of radiation. I do understand the chemotherapy before and the chemotherapy after, but to me splitting the course of radiation seems to violate at least what I know about the effectiveness of radiation. Can you explain?

Dr DeCamp. You’re absolutely right. We had to take the radiation oncologists out behind the barn and tie them up to get them to agree to manage patients with a split course. It certainly violates the fundamentals of radiobiology. I think it’s really a compromise in terms of multimodality therapy. We wanted to ensure that most patients got to surgery. We think most patients who are cured of lung cancer are going to undergo resection. We are concerned about the added toxicity of, for example, a second cycle of induction therapy before resection. Our schema is clearly a compromise from a radiation oncology standpoint, yet the results are encouraging. Local control is excellent. We are continuing to explore tweaking the schema in an effort to improve survival.