|
|
||||||||
Eur J Cardiothorac Surg 2004;26:1205-1210
© 2004 Elsevier Science NL
a Departments of Cardiothoracic Surgery, Hematology/Oncology and Respiratory Medicine, Radiation Oncology, and Biomathematics, University Hospital, Münster, Germany
b Department of Thoracic Surgery, Lung Clinic, Hemer, Germany
Received 11 December 2003; received in revised form 25 July 2004; accepted 3 August 2004.
* Corresponding author. Tel.: +49 251 834 7401; fax: +49 251 834 8316. (E-mail: michael.semik{at}ukmuenster.de).
| Abstract |
|---|
|
|
|---|
| 1. Introduction |
|---|
|
|
|---|
We report on an interim analysis in this closed trial to assess the impact of preoperative chemoradiation additional to chemotherapy on resectability, on necessity and feasibility of complex resections, and on surgical morbidity and mortality in the neoadjuvant treatment setting of stage III NSCLC.
| 2. Patients and methods |
|---|
|
|
|---|
All patients were staged according to the International Staging System [7]. Since the trial started before the re-classification of the Mountain criteria in 1997, centrally located T3 N0/1-tumors were included as stage IIIA neoplasms [8]. These were defined by diffuse infiltration less than 2cm from the main carina or radiographically proven infiltration of the parietal pericard. N2-disease at mediastinoscopy led to inclusion as stage IIIA too in patients without T4-tumors. Peripherally located T3 N0/1-tumors with chest wall involvement were excluded from the study. The inclusion criteria for stage IIIB were N3-disease proven by mediastinoscopy, diffuse mediastinal involvement found at mediastinoscopy or in computed tomography, or infiltration of the main carina, the spine, the pulmonary artery, the aorta, or the superior vena cava in computed tomography or MRI. Patients with supraclavicular N3-disease, infiltration of the myocardium, ipsilateral satellite pulmonary nodules in the same lobe, or malignant pleural effusion were not eligible. After staging, the patients were stratified according to center and stage (IIIA vs IIIB) and randomized to receive either treatment A or B (Fig. 1).
|
2.2. Statistical methods
To evaluate statistical significance of proportion we performed two-sided Fisher's Exact Test. Two-sided P-values less than 0.05 were considered significant. In addition, where appropiate, we calculated two-sided exact 95% confidence intervals (CI) for binominal proportions.
2.3. Study design and treatment
All patients started with 3 cycles of chemotherapy consisting of cisplatin 55mg/m2 (day 1+4) and etoposide 100mg/m2 (days 14) as shown in Fig. 1. The interval between the cycles was 3 weeks.
In arm A, patients were then scheduled to combined chemo- and radiotherapy. Five days a week, patients received fractions of 1.5Gy twice daily (intervall at least 6h) until the target volume dose (TVD) of 45Gy was reached (days 6482). Concurrently carboplatin (100mg/m2) and vindesine (3mg absolute) were administered on days 64, 71, and 78. Four to eight weeks after completion of combined chemoradiotherapy, patients free of distant metastases were scheduled for surgery.
In arm B, patients without distant metastases were referred for surgery without further neoadjuvant therapy and received after complete resection radiotherapy with a TVD of 54Gy (1.8Gy daily).
Patients with unresectable tumors, refusal of surgery or incomplete resection went on in arm A to further twice-daily radiation up to a TVD of 24Gy. In arm B these patients received a TVD of 68.4Gy (1.8Gy daily).
2.4. Eligibility for surgery
Patients were considered eligible for surgery if staging demonstrated an improved or stable disease, and medical as well as functional operability was given. These judgements were made in each participating site by a local interdisclipinary conference involving the thoracic surgeon, the pneumologist, the medical oncologist and the radiooncologist. Thoracotomy was performed with the aim of achieving complete resection of the tumor (R0, resection margin microscopically free of tumor cells) and complete mediastinal lymph node sampling or dissection [10].
| 3. Results |
|---|
|
|
|---|
|
Arm A (n=130) complete response, 12 (9.2%); partial response, 92 (70.8%); no change, 24 (18.5%). Arm B (n=143) complete response, 5 (3.5%); partial response, 94 (65.7%); no change, 40 (28%). Moreover, despite staged as progressive disease, in arm A 2 patients (1.5%) and in arm B 4 patients (2.8%) underwent thoracotomy. The difference of response rates between both treatment arms was not statistically significant. Response rates were in arm A 80% and arm B 69.2%.
3.2. Feasibility of surgery
Surgery followed the end of chemoradiotherapy in arm A or the end of sole chemotherapy in arm B after an interval of 45 (median; interquartile interval, 17.5) and 36 (median; interquartile interval, 14) days, respectively. In the majority of cases (134/273, 49%) lobectomy was sufficient for tumor removal. There were no differences between arm A (63/130, 48%) and arm B (71/143, 50%). The necessity of pneumonectomy was also similar among the groups (arm A/arm B: 43/130, 33% vs 45/143, 31%) as was the amount of bilobectomies (arm A/arm B: 7/130, 5.4% vs 6/143, 4.2%) and non-resectional thoracotomies (arm A/arm B 12/130, 9.2% vs arm B 12/143, 8.4%).
Complete resection of the primary tumor with systematic lymphadenectomy or lymph node sampling was possible in 78% of patients (212/273). Resectability according to the treatment groups and further details of surgery are shown in Table 2. As outlined, 121 patients had complex resections. These have been defined by the requirement of extended surgical procedures such as bronchus sleeve resection (n=39), tracheobifurcal resection (n=5), intrapericardial resection (n=78), resection of the chest wall (n=3) or the diaphragm (n=3), partial resection/replacement of the vena cava (n=3), aorta (n=2), pulmonary artery (n=22) or atrium of the heart (n=8), or sole extended pneumectomy (n=13). Fifty-one of 121 patients underwent two or more of these extended procedures.
|
Main cause of surgical mortality was pneumonia with consecutive respiratory insufficiency (n=7) and bronchial stump insufficiency (n=5). Despite employing different surgical strategies including bone spongiosis (n=2), transpericardial occlusion (n=1), myoplasty (n=1), fibrin glue occlusion (n=2), chest tube and antibiotics (n=1), 5 out of 7 patients (pneumonectomy, n=3; bilobectomy, n=1; lobectomy, n=1) with bronchial stump insufficiency (arm A, 4/arm B, 1) died due to this problem (14, 154, 170, 279 and 403 days after surgery). Considering late surgical mortality (>30 days postoperatively) further two patients succumbed to pneumonia, one at day 34 (arm A) and one at day 33 (arm B) after perioperative cerebellar infarction.
Sixty-two percent of patients (n=170) (arm A 80, 62%; arm B 90, 63%) recovered without complications. Those with complications are outlined in Table 3. Moreover, in 13 patients prolonged chest tube therapy (n=8; arm A, 5, arm B, 3), transient neurologic disorder (n=3; arm A, 1; arm B, 2), and chylothorax (n=2; arm A, 1; arm B, 1) were reported. Further complications, occuring once each, comprised in arm A pneumothorax, and putride bronchitis, and in arm B perioperative occurrence of stroke, bronchus stenosis, lung embolism, seroma of chest wall, cardiac failure with lung edema, and cardiac arrest with resuscitation.
|
| 4. Discussion |
|---|
|
|
|---|
In the present interim analysis, bimodality induction seemed to be superior to sole neoadjuvant chemotherapy, particularly as there is a trend toward an improved resectability rate in patients pretreated with chemotherapy and chemoradiation in contrast to mere chemotherapy. A substantial proportion of patients were encountered with complex resections (121/273, 44%; IIIA 38/105, 36%; IIIB 83/168, 49%). Two recent trials, employing preoperative chemoradiation, reported similar rates of complex resections of 3040% in stage IIIA and almost 60% in stage IIIB [2,12,13]. Apart from an advanced tumor stage, fibrosis and scaring related to the neoadjuvant treatment can render lung resection more complex. However, the need of complex resections as well as the necessity of pneumonectomies and lobectomies was more or less similar between the two treatment arms.
In addition to complete resection three phase II-trials, employing chemoradiation prior to surgery, emphasized a favorable histomorphologic response in the resection specimens, either of the mediastinal lymph nodes [2,3] or the mediastinal lymph nodes and additionally the primary tumor [5], as the most important predictor for long-term outcome. This gives rise to speculation on the contribution of surgery to the long-term outcome in the trimodality treatment setting. Thus, a large phase III-trial in stage IIIA (N2) disease (Intergroup trial 0139) is adressing the additional impact of surgery in this treatment approach. The significance of this issue is emphasized by mortality rates in conjunction with trimodality therapy that range about 10% and are clustering with surgery [2].
Particularly radiotherapy employed sequentially or simultaneously to neoadjuvant chemotherapy has been assumed to significantly increase surgical risk as it may promote fibrosis, bleeding and bronchial stump insufficiency [14]. In fact, bronchial stump insufficiency (3.8 vs 2.1%) as well as significant bleeding requiring re-thoracotomy (1.5 vs 0.7%) dominated in arm A of the present trial. As patients with these complications did not cluster substantially beyond or beneath the median time interval of 45 days between the end of chemoradiation and surgery, there is no evidence that performing surgery too late or too early in the time course of treatment has precipitated these adverse events. A monoinstitutional trial enrolling IIIB patients, who were preoperatively treated with radiation twice-daily in addition to chemotherapy, reported on a bronchial stump insufficiency rate of only 4.8% [4]. Another trial exclusively enrolling patients with stage IIIA-disease in a multi-institutional approach with conventional radiotherapy in addition to preoperative paclitaxel-containing chemotherapy reported a rate of 12.5% [15]. When applying a preoperative target volume dose of 60Gy concomittant to chemotherapy, bronchial insufficiency developed even in 23% of cases [14]. These differences between the trials may depend on dosing and application of radiotherapy, but also on different surgical techniques. We experienced no continued occurrence of this complication, when in the multi-center setting of our trial participating institutions consistently sealed the bronchial stump with horizontal mattress sutures and protected the stump with a flap of intercostal muscle or other viable tissue. The importance of an appropriate surgical technique in this regard has also been emphasized by others [13,16,17].
Beyond bronchial stump insufficiency pneumonia was the major cause for surgical mortality with three patients in arm A and four patients in arm B. All but two succumbed early after surgery (<30 days). Moreover, one patient died due to right heart failure 3 days after right-sided pneumectomy, and another expired due to a dehiscent suture of the pulmonary vessels. Albeit surgery in this case violated the protocol he was evaluated as surgical mortality.
In conclusion, overall surgical risk was similar with or without preoperative radiation in this phase III-trial. Preoperative bimodality induction seems to be superior compared to sole preoperative chemotherapy in order to achieve complete resection (n.s.). Bronchial stump insufficiency dominated in the patient cohort with neoadjuvant chemotherapy followed by chemoradiation, and had an impact on surgial mortality. Adopting appropriate surgical techniques, bronchial stump insufficiency could be almost completely eliminated.
| Appendix A. Conference discussion |
|---|
|
|
|---|
Dr Semik: I mentioned that we had 385 patients evaluated for surgery, and 112 patients who were not operated were locally or distantly progressed on the one side, and otherwise functionally inoperable, so these patients were excluded from surgery.
The toxicity is mainly due to esophagitis (grades 3 and 4) in arm A. This reached nearly 20%. So this is an relevant side effect of that bimodality induction. Concerning the nearly equal follow-up in both treatment arms (in the intermediate evaluation), nowadays we proceeded more following arm B without chemoradiation because of that side effect.
Dr T. Dosios (Athens, Greece): I realize that you have more bronchopleural fistula, or, what you said, bronchial stump insufficiencies, in the arm with radiotherapy, but you didn't find any difference in early mortality and long-term survival. So my question is, do we have to put our patients in radiotherapy preoperatively, and do you think that this higher rate of bronchopleural fistula that is about 5% or something like that is due to radiotherapy?
Dr Semik: Concerning the last question, I think the radiochemotherapy may be responsible, but after our learning curve in the study group, the complication of that kind was normalized. So if you are experienced in thoracic surgery and operating on those pretreated patients, I think you can drop the complication rate to a normal level. This would be no reason to stop that kind of therapy.
Dr Dosios: If there is no advantage, why continue giving radiotherapy?
Dr Semik: Up to now it seems that there is no relevant benefit for the bimodality induction, but the final analysis is still ongoing. Later I will report you about the final results and if any subgroups with preoperative radiochemotherapy improved in longterm survival.
Dr K. Athanassiadi (Athens, Greece): I would also like to ask you a similar question about the bronchial fistula you had. What was the total percentage of fistulae you had after pneumonectomy? How do you protect the bronchus after either chemotherapy or radiotherapy or both and I would like to know if this is your strategy in all cases?
Dr Semik: In the early phase of our study we didn't protect the bronchus anastomosis routinely, but after that higher incidence in the beginning of our study, we did it routinely. We had 3 bronchial stump insufficiencies after pneumonectomy, one after bilobectomy, and one after lobectomy.
Dr Athanassiadi: And how did you protect your stump? What did you use?
Dr Semik: Different techniques were used for bronchial stump or anastomoses protection, responsible was the local thoracic surgeon of the participating hospitals. Normally, pedicled flaps of pleural tissue, intercostal muscle or mediastinal thymic tissue were used.
Dr O. Kshivets (Siauliai, Lithuania): My question is do you plan to changeto the cisplatin regimen for example on gemzar? It's less toxic and more effective. And I have some comments. I advise you to give these patients radiotherapy less than 50Gy. You treated with 54Gy according to the last randomized treatment. Gy. According to the last You say the radiologistmeta-analysis of randomized combined treatment trials of operated lung cancer patients in the USA the radiologists recommended less than 50Gy, because more than 50Gy, you will have a great majority of complications in connection with radiotherapy.
Dr Semik: In our arm A, the patients only received 45Gy preoperatively.
| Acknowledgments |
|---|
The study is supported by the Deutsche Krebshilfe, Germany.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Beshay, G. Carboni, B. Hoksch, M. A. Reymond, and R. A. Schmid The role of muscle flap in preventing bronchus stump insufficiency after pneumonectomy for malignant pleural mesothelioma in high-risk patients Interactive CardioVascular and Thoracic Surgery, August 1, 2008; 7(4): 621 - 625. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. Venuta, M. Anile, D. Diso, M. Ibrahim, T. De Giacomo, M. Rolla, V. Liparulo, and G. F. Coloni Operative complications and early mortality after induction therapy for lung cancer Eur. J. Cardiothorac. Surg., April 1, 2007; 31(4): 714 - 717. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Fujita, N. Katakami, Y. Takahashi, K. Hirokawa, A. Ikeda, C. Tabata, T. Mio, and M. Mishima Postoperative complications after induction chemoradiotherapy in patients with non-small-cell lung cancer. Eur. J. Cardiothorac. Surg., June 1, 2006; 29(6): 896 - 901. [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 |