| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Division of Thoracic Surgery, Rush University Medical Center, Chicago, Illinois
Accepted for publication May 28, 2009.
* Address correspondence to Dr Faber, University Thoracic Surgeons, 1725 W. Harrison, Suite 774, Chicago, IL 60612 (Email: pfaber{at}rush.edu).
Presented at the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Methods: Clinical records of patients with non–small cell lung cancer undergoing sleeve lobectomy between 1983 and 2008 were reviewed for age, sex, type of sleeve resection, clinicopathologic TNM stage, complications, and 90-day mortality. Chemotherapy and radiation therapy regimens were recorded for the patients undergoing neoadjuvant treatment. Kaplan-Meier survival curves were compared.
Results: There were 64 patients identified as having undergone sleeve resection for non–small cell lung cancer. Of the 64 total patients, 43 did not receive concurrent neoadjuvant chemoradiotherapy [NCR] versus 21 patients who did [CRS]. All of the CRS patients underwent platinum-based chemotherapy and radiation (range, 2,000 to 6,100 cGy). Thirteen patients (62%) were downstaged, with 4 complete responders. The 90-day mortality was 2.7% (2 patients) in the NCR group and 0% in the CRS group. The incidence of major complications in the NCR group was 46.5% (20 of 43) with 4.7% (2 of 43) anastomosis-related complications (stenosis, 1; bronchovascular fistula, 1). The incidence of major complications in the CRS group was 42.9% (9 of 21) with no anastomosis-related problems. Five-year survival in the NCR group was 48% compared with 41% in the CRS group (p = 0.63). There were 9% (4 of 43) of patients with local recurrence in the NCR group versus 10% (2 of 21) of patients in the CRS group (p = 0.65).
Conclusions: Anastomosis-related complications were not increased among the patients receiving neoadjuvant therapy compared with those who did not. In addition, local recurrence was also similar between the two groups. Furthermore, the survival of the two groups was not statistically different. Sleeve lobectomy after chemoradiotherapy for advanced non–small cell lung cancer can be performed with acceptable morbidity and mortality.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Since the adoption of neoadjuvant therapy for patients with clinically advanced disease, the safety of sleeve lobectomy remains a concern. It is thought that neoadjuvant radiation, either alone or with chemotherapy with its attendant detrimental effects on the bronchial blood supply, may lead to a higher risk of anastomotic complications after sleeve resections. Therefore, the purpose of this study was to evaluate the anastomotic outcomes associated with sleeve lobectomy in patients who have undergone neoadjuvant chemoradiotherapy.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
Age, sex, forced expiratory volume in 1 second, lung diffusion capacity for carbon monoxide, pretreatment and postoperative stages, type of neoadjuvant therapy (for the CRS group), type of sleeve resection, 90-day mortality, major complications, and survival were extracted. Major complications were divided into the broad categories of cardiac, pulmonary, and anastomotic. Cardiac complications included arrhythmias and myocardial ischemia or infarction. Pulmonary complications included respiratory failure or adult respiratory distress syndrome requiring intubation, pneumonia, or atelectasis requiring bronchoscopy. Prolonged air leak was defined as an alveolar air leak that persisted for longer than 7 days. Bronchoscopic confirmation that the anastomosis was not the source of air leak was made in each of the patients with prolonged air leaks. Anastomotic complications included stenoses or dehiscences leading to a bronchopleural or bronchovascular fistula.
Clinical Considerations
All patients received a computed tomography scan of the chest and upper abdomen and preoperative pulmonary function tests. Positron emission tomography scan replaced bone scans in the evaluation of metastatic disease for the patients later in the series. Selective mediastinoscopy was performed if there was a suspicion of N2 involvement based on the computed tomography (node > 1 cm in short axis) or positron emission tomography scan (standardized uptake value > 2.5) criteria.
Neoadjuvant therapy consisted of platinum-based chemotherapy (cisplatin/5-fluorouracil or etoposide, or carboplatin/Taxol) with concurrent radiation. Mean dose of radiation received was 4,000 cGy (range, 2,000 to 6,100 cGy). The dose increased during the course of the study to our current standard of 4,500 cGy. All patients receiving preoperative chemoradiotherapy were restaged with at least a chest computed tomography scan to rule out interval progression of disease before operative intervention. The clinical preoperative staging was used to assess whether a patient had been downstaged if mediastinoscopy was not performed before neoadjuvant therapy.
The technique of sleeve resection has been described in detail elsewhere [11]. All resections were performed through a posterolateral thoracotomy. End-to-end anastomoses were constructed with interrupted 4-0 braided absorbable sutures. Size discrepancies were corrected by appropriate suture spacing; no telescoping was attempted. The use of tissue flaps to cover the anastomosis was at the discretion of the individual surgeon. A lymphadenectomy was performed in all cases. On the right, this included levels 2, 4, 7, 8, 9, 10, and 11. On the left, levels 5, 6, 7, 8, 9, 10, and 11 were dissected. All patients underwent a bronchoscopy at the completion of the case to evaluate the anastomosis and to clear the airway of blood and secretions.
Tumor recurrence was also recorded and compared. Local recurrence was defined as that which occurred in the ipsilateral chest or mediastinum. Particularly, local recurrence at the anastomosis was assessed. Distant recurrence was defined as that which occurred in the contralateral chest or elsewhere.
Statistical Methods
The two groups were compared with respect to the demographics and clinical outcomes using Student's t tests or Fisher's exact tests as appropriate. Kaplan-Meier survival curves were generated and log-rank analyses were performed to compare the survivals. A probability value of less than 0.05 was considered statistically significant. Statistical analysis was performed using SAS version 9.1 software (SAS Institute Inc, Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
|
|
All 21 patients in the CRS group received their planned radiation dose. However, because the study spans a period of 25 years, the dose of radiation therapy was not standardized for all patients. One patient early in the experience received a dose of 2,000 cGy. As newer studies advocating higher doses were published and our center became more comfortable with increasing doses of preoperative radiation, the patients began receiving doses of 4,500 cGy, our current standard. One patient received 6,100 cGy as part of definitive chemoradiation treatment before being reconsidered for resection. The mean dose received by the CRS group was 4,000 cGy.
In the NCR group there were 25 right upper lobectomies, 2 right middle lobectomies, and 2 right middle and upper bilobectomies, as well as 5 left upper lobectomies and 9 left lower lobectomies performed. Pulmonary arterioplasty was performed in conjunction with 2 left upper lobectomies. In the CRS group there were 19 right upper lobectomies, 1 left upper lobectomy, and 1 left lower lobectomy performed. One of the right upper lobectomies required an arterioplasty.
There were 4.7% (2 of 43) perioperative deaths in the NCR group (pneumonia, 1; bronchopulmonary artery fistula, 1) and 0% in the CRS group (p = 0.45). The overall morbidity was 46.5% (20 of 43) in the NCR group and 42.9% (9 of 21) in the CRS group (p = 0.50). Table 3 shows the incidence of postoperative complications. In the NCR group there was one each of the following anastomotic complications: a stenosis and a bronchopulmonary artery fistula. The patient with stenosis underwent serial dilatations until his death owing to recurrent metastatic cancer. The patient with the bronchopulmonary artery fistula experienced adult respiratory distress syndrome after a completion pneumonectomy and subsequently died. This completion pneumonectomy was the only one performed in either group during the study period. In the CRS group, there were no anastomotic complications.
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
After these reports, sleeve resections became the standard operation for all patients with suitable anatomy, regardless of their ability to tolerate a pneumonectomy. At this time, it also became common for patients with N2 disease to undergo neoadjuvant therapy. With the effect of the chemotherapy and radiation therapy on the healing of any suture line, the safety of bronchoplastic resections after neoadjuvant therapy has become a concern. Rendina and colleagues [12] reported their experience with bronchovascular procedures after neoadjuvant chemotherapy. Of the 68 patients who underwent platinum-based chemotherapy, 27 subsequently had a sleeve lobectomy. There were no anastomotic stenoses or disruptions. However, in North America, most programs have adopted concurrent chemoradiotherapy as neoadjuvant therapy for advanced NSCLC. The addition of radiation therapy is postulated to potentiate the healing complications at the anastomosis that chemotherapy alone may also present. In the experience of Rea and associates [13] with sleeve lobectomy, 21.6% (43 of 199) of patients underwent neoadjuvant therapy: 7% (14 of 199) chemotherapy, 5% (10 of 199) chemoradiotherapy, and 9.5% (19 of 199) radiation therapy. Univariate analysis only identified neoadjuvant radiation therapy as being predictive for early bronchial complications.
In our experience of 21 patients who underwent resection after concurrent radiation and platinum-based chemotherapy, a sleeve lobectomy was possible with reasonable mortality and morbidity. There were no perioperative deaths among the patients who received neoadjuvant chemoradiotherapy. This compared favorably with the 4% mortality in the NCR group. This incidence of mortality also was comparable to the results demonstrated by Deslauriers and coworkers (1.3%) [4], Ma and associates (3.5%) [5], and Ferguson and Lehman (4.1%) [9]. When looking at trials of patients who did receive neoadjuvant therapy, similar outcomes were seen. Burfeind and colleagues [14], compared 19 patients who underwent neoadjuvant therapy before a bronchoplastic resection with 54 patients who did not. They had no deaths in the neoadjuvant group. In the review by Cerfolio and associates [15] of pulmonary artery sleeve resections, one half of the 42 patients studied received neoadjuvant therapy. There was only 1 perioperative mortality.
When morbidity was assessed, the complication rate of 42.9% in the neoadjuvant therapy group was similar to the overall morbidity of 46.5% in the resection only group. These figures were also within range of the 31% morbidity reported in the meta-analysis by Ma and associates [5]. Burfiend and colleagues [14] noted a 42% incidence of morbidity after sleeve resections after neoadjuvant therapy. In particular, no anastomotic complications were seen in the neoadjuvant group being presented here as compared with a 4.7% incidence of complications in the resection only group. This is consistent with the findings of Deslauriers and coworkers (2.6%) [4] and Ludwig and associates (6.9%) [6]. Burfiend and colleagues [14] had 1 patient who experienced a bronchopleural fistula requiring a completion pneumonectomy, and Cerfolio and associates [15] reported no anastomotic complications. In contrast to the neoadjuvant therapy group in this investigation, in both studies by Burfeind and colleagues [14] and Cerfolio and coworkers [15], not all of the patients underwent neoadjuvant chemoradiotherapy; some underwent chemotherapy only [14], and some received no neoadjuvant therapy at all [15].
Although there were only 2 perioperative mortalities in the entire series reported in this article, one was owing to a bronchopulmonary artery fistula. Other series have also documented this complication as a not infrequent cause of perioperative mortality [16]. The other complication that occurred in this series was bronchial stenosis. In certain series, this appears to be the most common anastomotic complication [6, 8]. Bronchopleural fistulae were not observed in either group in this study. This result was very surprising, particularly among the patients receiving neoadjuvant therapy, as this would have been expected to predispose these patients to this complication. Patients undergoing sleeve lobectomy without neoadjuvant therapy have reported incidences of bronchopleural fistulae between 0.5% and 6% [7, 12, 13, 17, 18]. No clear reason for this fortuitous outcome could be identified. As mentioned earlier, routine flap coverage was not used for all the anastomoses, including within the neoadjuvant group. The sleeve lobectomies were generally performed in the standard fashion that is often presented in other manuscripts or textbooks, and this technique has become standardized in the academic institution of the authors [11].
In addition to the acceptable short-term results, this study also demonstrated satisfactory long-term outcomes. Patients in the NCR group had an overall 5-year survival of 49%. This was similar to the rates published from large sleeve lobectomy series without a significant subset of patients receiving neoadjuvant therapy, which have demonstrated survivals ranging from 39% to 54% [4–7, 17]. As expected, based on the advanced stage of disease, those who received neoadjuvant therapy had a lower overall 5-year survival of 41%. This figure is acceptable given the fact that the majority of the patients receiving neoadjuvant therapy were those who initially had locoregionally advanced disease. In fact, it compares well with other retrospective sleeve lobectomy series that have included a significant number of patients receiving neoadjuvant therapy. Those series demonstrated an overall long-term survival ranging from 39% to 60% [12, 13, 15]. This report's figure of 41% survival also comes close to the meta-analyses of sleeve lobectomies in all-comers, which showed a 5-year overall survival of approximately 50% [5, 9].
Yamamoto and colleagues [18] demonstrated in their analysis that, although overall survival was not different on the basis of having received neoadjuvant therapy, the disease-free survival was significantly different. Presumably, there was greater recurrence among the neoadjuvant therapy group indicating a higher stage of disease. Similarly, in this series, there was a much higher percentage of recurrences in patients who underwent neoadjuvant therapy. Among the CRS patients, 62% had a recurrence as compared with 21% in the NCR group. Although the incidence of recurrence among the NCR group was consistent with other reports of recurrence (up to two thirds of patients, with local recurrence in up to a third of patients), the incidence of recurrence in the neoadjuvant group was exceedingly high when compared with other reports of distant metastases after sleeve lobectomy [4, 8, 13, 17]. This was attributed mostly to the fact that the patients in each group were not matched. A vast majority of patients (76%) in the neoadjuvant therapy cohort had stage III disease, whereas patients in the resection only group were fairly equally distributed among stages I, II, and III. This group of mostly stage III patients would certainly be expected to fare less well than the patients with an earlier stage. This is supported by the findings of other investigations suggesting that preoperative chemotherapy is a risk factor for decreased long-term survival [8]. Also, when we looked at the pattern of recurrence, the majority of them were distant in nature (11 distant recurrences of 13 total). This finding was consistent with other reports of recurrences [13, 18].
Only 2 patients had a local recurrence in the neoadjuvant therapy group, compared with 4 patients in the resection only group, which was not a significant difference. Rea and coworkers [13] had a similar experience in that of all the patients that died in their series, only 11% of them died of locoregional recurrences. The majority of patients died of distant metastases, and even the number of non-cancer deaths was more then two times that of deaths because of local disease. Therefore, sleeve lobectomy even in advanced stage NSCLC appears effective in locoregional control as part of trimodality therapy.
This study does have limitations. It is a retrospective study with a modest number of patients, limiting statistical power. A type II error therefore cannot be ruled out. The number of patients in the study is small despite a long study period, but that is because only patients with NSCLC are included. The vast majority of sleeve lobectomies performed in this period were for endobronchial carcinoid tumors. However, even with a relatively small number of patients, it compares favorably with other recent series [14, 15]. The fact that the neoadjuvant group of patients had a higher percentage of advanced stage cancer obviously affects the long-term survival analyses. Owing to the long study period, the dose of neoadjuvant radiation received by the patients was also not standardized. Thus, even though all patients completed their scheduled neoadjuvant therapy, it may be hard to generalize this study to patients currently receiving up to 6,000 cGy of neoadjuvant radiotherapy.
Five of the patients receiving neoadjuvant chemoradiotherapy had persistent N2 disease after induction therapy. There is controversy in the literature as to whether those patients should be offered resection or definitive chemoradiation alone [4]. At the time of performance of those resections, this debate had not yet surfaced. However, the fact that 4 of the 5 patients with positive posttherapy N2 involvement died within 2 years of their resection does lend credence to the fact that caution should be exercised in offering a sleeve resection to this subset of patients. A question can also be raised as to whether invasive preoperative staging (eg, with cervical mediastinoscopy) should have been done in every patient to accurately categorize those who were downstaged after neoadjuvant chemoradiotherapy. Regardless of pathologic staging preoperatively, all the patients in the CRS group completed the planned neoadjuvant therapy, and thus the primary end points of perioperative mortality, morbidity, and anastomotic complications should be unaffected.
Neoadjuvant therapy for stage IIIA NSCLC has become common. A prospective, randomized trial of sleeve lobectomy with and without neoadjuvant chemoradiotherapy would thus be unlikely to happen. A randomized trial of sleeve lobectomy versus pneumonectomy after neoadjuvant trial may be unethical if patients will be subjected to a more morbid pneumonectomy rather than a parenchyma-sparing resection. Therefore, despite the limitations of a retrospective series, studies such as this one may remain our best effort to add to the current literature and allow for the formation of recommendations regarding sleeve resections after neoadjuvant therapy.
Based on the review of the authors' experience, sleeve lobectomy in the setting of preoperative chemoradiotherapy is not associated with a higher incidence of anastomosis-related complications. Furthermore, it can be performed with acceptable morbidity and mortality and can provide good long-term survival. Finally, satisfactory locoregional control of disease can be achieved with sleeve resection after neoadjuvant chemoradiotherapy.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This is an important paper. Hopefully, if published, this will actually be the fifth paper this year that will talk about the feasibility of safely performing either pneumonectomies or sleeve lobectomies in the face of chemotherapy and radiation. So I think it's a very important contribution to the literature. Our group has done this in the past, Ben Daly has reported a smaller series, and Rendina's group in Rome has also published a series that does include patients who had chemotherapy and radiation.
I have two specific questions. One is a technical one, which I'm sure you know I'm going to ask. Did you use muscle flaps routinely, even when you just did a bronchial sleeve, or did you only use it when you did a bronchial and vascular sleeve, because, again, your results are excellent with so few anastomotic issues.
DR MILMAN: Actually, use of flaps was not standardized. One surgeon routinely performed pleural flaps where he interposed the flap between the bronchial anastomosis and the PA (pulmonary artery) on all sleeve resections. The other surgeons did not routinely use any flaps.
DR KRASNA: The other question relates to two of the other techniques that you mentioned but you didn't discuss them in your subanalysis. You were comparing those patients who had chemorads to those who did not, but within the group who had chemotherapy and radiation—and, again, your outcomes are so good, it's hard to dissect out—did you perceive along the way in the change over that 20-year period that doing mediastinoscopy first had any effect on the dissection or injury to the airway? Again, you didn't have any anastomotic leaks, so you can't say that you compromised the bronchial vasculature. Likewise, did you perceive a difference with a higher radiation dose that I know you now use at Rush over that time period in terms of postoperative function?
I enjoyed your paper.
DR MILMAN: One, as I said, mediastinoscopy was not used in all patients. Two, this paper included only sleeve lobectomies. We did not include sleeve pneumonectomies or any other resections that were closer to the carina. So we did not appreciate any difference in terms of doing mediastinoscopy first and how it affected our resection. As far as I know, there has been no difference in terms of being able to do the resection with higher doses of radiation.
DR ROBERT J. CERFOLIO (Birmingham, AL): What dose of radiation are you using now?
DR MILMAN: Right now it's still 4,000, but we're trying to rev up to 5,000 to 6,000.
DR CERFOLIO: Where did you get 4,000? I mean 4,500 is the lowest dose that most use and most of us are using 6,000 or now even 7,200. Where did you get 4,000?
DR MILMAN: We do have a range between 2,000 and 6,100, but the majority of the patients received 4,000.
DR L. PENFIELD FABER (Chicago, IL): Our initial phase II trial of concurrent chemoradiation for neoadjuvant therapy was begun in 1986. We started with 4,500 cGy to decrease side effects of radiation. This was based on the earlier study of radiation alone as neoadjuvant therapy, which resulted in a high number of postoperative complications. That study was carried out in the late 1960s. Doctor Cerfolio, I would suggest you review that manuscript.
DR CERFOLIO: I have a reference in my paper.
DR FABER: Then why ask the question?
DR CERFOLIO: No, I still ask the question.
DR FABER: We utilized 4,500 cGy as the radiation was concurrent with chemotherapy and we wanted to minimize complications. Radiation and chemotherapy are synergistic and this dose of radiation works.
DR CERFOLIO: Well, there's your answer—but many have used higher doses, so if patients are not resected they have maximized their medical therapy without significant time gaps in the radiation.
DR KESHAUDAS PAHUJA (Stoughton, MA): I have a couple of technical questions. One, were you always able to bring the bronchus intermedius to the main bronchus? Did you do something to try to bring them together, any additional maneuvers, you know, bring the chin down or putting a chin stitch or possibly dividing the pericardium?
DR MILMAN: No, just with circumferential mobilization of the mediastinal pleura and the division of the inferior pulmonary ligament, we were able to have the bronchus intermedius come up to meet the main stem bronchus and we did not have any problems with it.
DR PAHUJA: I assume you had no tension on any of the anastomoses?
DR MILMAN: Nothing that led to dehiscence, no.
DR CERFOLIO: Talk to me about the use of Vicryl. Are you still using Vicryl or are you using PDS?
DR MILMAN: We are still using Vicryl.
DR CERFOLIO: And you're happy and your knots are on the outside?
DR MILMAN: We try to have them all on the outside. The posterior wall sometimes ends up with the knots being on the inside.
DR FABER: If it ain't broke, don't fix it.
DR CERFOLIO: Well, yes. If you have no anastomotic complications without buttressing and with using Vicryl, I think that's a credit to the incredible technical skill of the surgeons. I'm not sure if all of us mere humans can duplicate that using no muscles and knots on the inside of the lumen. I think the message for maybe the mere mortals, like me, and the majority of us in the audience is, I still like to use the muscle because I think it prevents problems, but this is evidence that if you do it really well, maybe you don't need to and maybe Vicryl is just as good as PDS.
DR FABER: Better.
DR CERFOLIO: Better ... well then, once again, there's your answer.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Gomez-Caro, M. Boada, N. Reguart, N. Vinolas, F. Casas, and L. Molins Sleeve lobectomy after induction chemoradiotherapy Eur J Cardiothorac Surg, January 4, 2012; (2012) ezr184v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
