Morbidity and Mortality of Major Pulmonary Resections in Patients With Early-Stage Lung Cancer: Initial Results of the Randomized, prospective ACOSOG ZOOSO Trial Mark S. Allen, MD, Gail E. Darling, MD, -ratne T. V. pechet, MD, John D. Mitchell, MD, James E. Herndon II, PhD, Rodney J. Landreneau, MD, Richard l. Inculet, MD, David R. Jones, MD, Bryan F. Meyers, MD, David H. Harpole, MD, Joe B. Putnam, Jr, MD, Valerie W. Rusch, MD, and the ACOSOG Z0030 Study Group* Background.
Little prospective, multiinstitutional data exist regarding the morbidity and mortality after major ulmonary resections for lung cancer or whether a mediastinal lymph node dissection increases morbidity and mortality. Methods. Prospectively collected 30-day postoperative data was analyzed from 11 patients undergoing pulmonary resection who were enrolled from July 1999 to February 2004 in a randomized trial comparing lymph node sampling versus mediastinal lymph node dissection for early stage lung cancer. Results.
Of the 11 patients randomized, 1,023 were included in the analysis. Median age was 68 years (range, 23 to 89 years); 52% were men. Lobectomy was performed in 766 (75%) and pneumonectomy in 42 (4%). Pathologic tage was IA in 424 (42%), 1B in 418 (41%), ‘IA 37 (4%), 11B in 97 (9%), and Ill in 45 (5%). Lymph node sampling was performed in 498 patients and lymph node dissection in 525. Operative mortality was 2. 0% (10 of 498) for lymph node sampling and 0. 76% (4 of 525) for lymph node dissection. Complications occurred in 38% of patients in each group.
Lymph node dissection had a longer median operative time and greater total chest tube drainage (15 minutes, 121 mL, respectively). There was no difference in the median hospitalization, which was 6 days in each group (p , 404. 0(. Conclusions. Complete mediastinal lymphadenectomy dds little morbidity to a pulmonary resection for lung cancer. These data from a current, multiinstitutional cohort of patients who underwent a major pulmonary resection constitute a new baseline with which to compare results in the future. Ann Thorac surg 2006;81 013-20) 2006 by The Society of Thoracic Surgeons Variability exits among surgeons’ opinions as to whether to remove all, some, or none of the mediastinal lymph nodes at the time of pulmonary resection for lung cancer; therefore, practices vary widely throughout the world. Surgeons in some cancer centers perform a complete mediastinal lymphadenectomy, whereas others sample ymph nodes when a visual or palpable abnormality is found.
Unfortunately, despite the fact that surgical staging of mediastinal lymph nodes is thought to be important, most surgeons do not perform a complete lymphadenectomy at the time of lung cancer resection. Reasons that a complete mediastinal lymphadenectomy is not performed include the perception that this may lead to more complications, prolonged hospitalization, or increased mortality, and may not affect long-term survival.
The American College of Surgery Oncology Group Z0030 study is a randomized, multiinstitutional, prospective trial designed to etermine whether long-term survival is affected by mediastinal lymph node dissection (LND) versus lymph node sampling (LNS) at the time of pulmonary resection for lung cancer. A secondary objective is to compare morbidity and mortality between the two arms to determine whether mediastinal LND ad- ung cancer continues to be a major cause of cancer mortality, both in the United States and worldwide.
Because the prognosis of lung cancer is directly proportional to whether lymph node metastasis has occurred, accurate lymph node assessment is crucial. Computed tomography has only limited accuracy, and positron mission tomography, although better, is still not perfect. Visual evaluation of lymph nodes at the time of thoracotomy has also been shown to be inaccurate . Complete excision of lymph nodes with microscopic evaluation is thought to be the most accurate method to determine lymph node metastasis.
Accepted for publication June 24, 2005. *Members of the American College of Surgeons Oncology (ACOSOG) Z0030 study group and their affiliations are listed in the Appendix. Presented at the Forty-first Annual Meeting of The Society of Thoracic Surgeons, Tampa, FL, Jan 24-26, 2005. Winner of the J. Maxwell Chamberlain Memorial Award for General Thoracic Surgery. Address correspondence to Dr Allen, Division of General Thoracic Surgery, Mayo Clinic, 200 First St SW, Rochester, MN 55905; e-mail: allen. [email protected] edu.
Published by Elsevier Inc 0003-4975/06/$32. 00 GENERAL THORACIC J. MAXWELL CHAMBERLAIN MEMORIAL PAPER 1014 CHAMBERLAIN MEMORIAL PAPER ALLEN ET AL MORBIDITY AND MORTALITY OF LYMPHADENECTOMY versely affects the patient’s outcome. This report presents the initial results from this randomized multiinstitutional trial and compares the short-term outcomes between patients who underwent LND and patients who underwent LNS only. Material and Methods From June 1999 to February 2004, 11 patients were randomized into the Z0030 trial.
Eligibility requirements included patients older than 18 years of age, an Eastern Cooperative Oncology Group (ECOG) performance score lower than 3, and a tissue diagnosis of a clinically resectable Tl or T2, NO or nonhilar NI, MO non-smallcell lung cancer (squamous cell carcinoma, large cell carcinoma, or adenocarcinoma, including bronchoalveolar carcinoma) established before randomization. Patients underwent computed tomography of the chest and upper abdomen that included the liver and adrenal lands within 60 days of the pulmonary resection.
Patients who had a prethoracotomy mediastinoscopy were eligible if there were no mediastinal lymph node metastases identified. Patients who did not undergo mediastinoscopy could have no evidence of mediastinal lymph adenopathy by computed tomography criteria (greater than 1. 0 cm in the shortest axis). Eligible patients had to be a candidate for a complete resection of the carcinoma by means of pneumonectomy, lobectomy, bilobectomy, or anatomic segmentectomy, with or without sleeve resection.
Exclusion criteria included patients who had T3 or T4 umors, patients who were treated with pulmonary wedge excision, and patients who received prior chemotherapy or radiation therapy for their cancer. After randomization, retrospective review found 1 55 patients to be ineligible for participation. The reason for ineligibility was minor (eg, timing violation) in 67 of these patients, remaining 88 patients (56 in the LNS group and 33 in the LND group) were excluded.
The reasons for exclusion were incorrect stage in 29 patients, inadequate LNS in 14, benign disease in 6, insufficient documentation in 5, and other reasons in 34 patients. The remaining 1,023 patients orm the basis of this study. All surgeons participating in the trial were general thoracic surgeons and diplomats of the American Board of Thoracic Surgery or the equivalent. There were 102 different surgeons from 63 institutions who enrolled patients into the trial.
The enrolling surgeon was required to read a detailed description of the technique of mediastinal LND and watch a video that demonstrated the technique of a complete mediastinal LND. All operative notes were reviewed by the principal investigators (M. S. A. or G. E. D. ) for completeness of the mediastinal dissection and adherence to the protocol. Lymph nodes were named according to The American Thoracic Society lymph node stations . Lymph node stations sampled for tumors on the right were 2R, 4R, 7, and IOR. For tumors on the left, stations 5, 6, 7, and IOL were sampled.
Any other lymph node that was thought to be suspicious Ann Thorac Surg for metastasis was also biopsied. If all required lymph nodes excised showed no evidence of cancer on frozensection examination, patients were randomized to either LNS only with no further lymph node dissection or to complete mediastinal LND. Complete mediastinal LND for tumors on the right nvolved removing all lymph tissue from an area bounded caudally by the takeoff of the right upper lobe, superiorly by the innominate artery, anteriorly by the superior vena cava, and posteriorly by the trachea.
All tissue was removed from this area, and at the completion of the dissection the trachea, vagus nerve, and the superior vena cava were visible. Lymph nodes in the prevascular area, adjacent to the superior vena cava, were removed as were the lymph nodes in the retrotracheal area. Complete mediastinal LND for tumors on the left involved removing all lymph tissue from the area between the phrenic nerve nteriorly and the vagus nerve posteriorly. Superiorly, all lymph tissue was removed to the top of the aortic arch. The caudal boundary was the Table 1 .
Characteristics of Patients in Each Arm of the Study Variable Number of patients Male Median age (range) (y) ECOG performance score 2 Tumor locationa RUL RML RLL LUL LLL Histology Squamous cell Adenocarcinoma Large cell Bronchoalveolar other NSCLC Type of resectionb Segmentectomy Lobectomy Bilobectomy Pneumonectomy Combination Lymph Node Sampling Arm Dissection Arm 498 257 (52%) 68 (23-89) 525 272 (52%) 67 (37-87) 344 (69%) 139 (28%) 343 (65%) 169 (32%) 3 (2%) 193 (39%) 36 (7%) 88 (17%) 129 (26%) 58 (12%) 193 (37%) 29 (6%) 101 (19%) 144 (27%) 64 (12%) 131 (26%) 196 (39%) 27 (5%) 108 (21%) 141 (27%) 229 (44%) 22 (4%) 32 (6%) 379 (76%) 18 (4%) 45 (9%) 34 (6%) 387 (73%) 24 (5%) 56 (11%) Counts do not sum to 100% because some patients had disease that b involved multiple lobes. in 2 patients. ECOG 0 Eastern Cooperative Oncology Group; LLL 0 left lower lobe; LUL 0 left upper lobe; NSCLC 0 non-small-cell lung cancer; RLL 0 right lower lobe; RML 0 right middle lobe; right upper lobe. 1015 compare groups relative to ordinal variables such as chest tube drainage . Logistic regression was used to assess the relationship between age and key measurements of morbidity. All statistical tests reported in the manuscript are two-sided. The threshold of significance was set at p less than 0. 05.
The protocol was approved by the central Institutional Review Board, and the Institutional Review Board of each institution that enrolled patients. All patients signed informed consent. Results Fig 1. Location of the tumors in all patients. Some patients had cancer that involved multiple lobes. left mainstem bronchus. At the completion of the dissection the aortopulmonary indow was free of lymph tissue and the recurrent nerve was preserved. Regardless of the side of the tumor, all subcarinal, inferior pulmonary ligament, periesophageal, lobar, and intralobar lymph nodes were resected. Complete subcarinal LND included removing all lymph tissue caudal to the carina and both left and right mainstem bronchi.
All lymph nodes adjacent to the inferior pulmonary ligament and the caudal half of the esophagus were also removed. When the dissection was complete, both mainstem bronchi, the posterior pericardium, and the esophagus were free of all lymph tissue. During resection of obar and interlobar lymph nodes were resected. Fisher’s exact test was used to compare treatment regimens relative to rates of operative mortality and morbidity . Operative mortality included all patients who died within the first 30 days after surgery or during the same hospitalization. The Wilcoxon test was used to There were 529 (52%) men and 494 (48%) women randomized to either LNS only (498 patients) or LND (525 patients; Table 1).
The median age was 68 years with a range of 23 to 89 years. The race was white in 955 patients (93%), black in 46 (4%), and other in 22 (2%). An ECOG performance score of zero was recorded in 87 patients (67%), one 308 (30%), and two in 28 (3%). Tumor location is shown in Figure 1; the most common tumor location was the right upper lobe (38%). Mediastinoscopy was performed in 161 (32%) patients who had LNS and 141 (27%) who had LND (not significant). The extent of resection was lobectomy in 766 patients (75%), segmentectomy in 70 (7%), pneumonectomy in 42 (4%), bilobectomy in 42 (4%), and wedge resection in 2 (0. 2%). Combinations of the above resections were performed in 101 patients (10%).
The pulmonary resection was by mens of a thoracotomy in 919 patients (90%), video-assisted thoracoscopic surgery in 66 (6%), nd both in 38 (4%). Complete resection (RO) was performed in 1,002 (98%), macroscopically complete resection (RI) in 19 (2%), and grossly incomplete resection (R2) in 2 (0. 2%). The histologic description was squamous cell carcinoma in 272 patients (27%), adenocarcinoma in 425 (42%), large cell in 49 (5%), bronchoalveolar carcinoma in 68 (6%), and other non- small-cell lung cancers in 209 patients (20%). There were no statistical differences in any of these variables between the LND and the LNS group (Table 1). The pathologic stage is shown in Table 2.
There were 45 patients who had pathologic stage IIIA or 111B; 12 in the LNS arm and 33 in the LND arm. Positive mediastinal lymph nodes were discovered in 20 (3. 8%) patients who had negative sampling and were randomized to LND. Table 2. Pathologic Stage Lymph Node Sampling Stage Lymph Node Dissection Totals No. of Patients % of Total 211 205 13 56 4 8 42. 5 41 . 2 2. 6 11. 3 0. 8 1. 6 213 24 41 9 40. 6 4. 6 7. 8 424 418 37 97 28 17 41 . 5 40. 9 3. 6 9. 5 2. 7 1. 7 1016 Table 3. Reasons for Stage IIIA or 111B LNS N2 disease 2nd cancer same lobeb Chest wall pleura involvedb 02 cm from carinab Phrenic nerve involvedb Trachea involvedb LND Total 20