Value of Systematic Mediastinal Lymph Node Dissection During Pulmonary Metastasectomy

Value of Systematic Mediastinal Lymph Node Dissection During Pulmonary Metastasectomy Florian Loehe, MD, Sonja Kobinger, MD, Rudolf A. Hatz, MD, Thomas Helmberger, MD, Udo Loehrs, MD, and Heinrich Fuerst, MD Departments of Surgery and Clinical Radiology, and Institute of Pathology, Klinikum Grosshadern, University of Munich, Munich, Germany Background. Systematic mediastinal lymph node dissection is the accepted standard when curative resection of bronchial carcinoma is performed. However, mediastinal lymph node dissection is not routinely performed with pulmonary metastasectomy, in which only enlarged or suspicious lymph nodes are removed. The incidence of malignant infiltration of mediastinal lymph nodes in patients with pulmonary metastases is not known. Methods. Sixty-three patients who underwent 71 resections through a thoracotomy for pulmonary metastases of different primary tumors were studied prospectively. Selected patients showed no evidence of tumor progression or extrathoracic metastases and pulmonary metastasectomy was planned with curative intent. All patients underwent preoperative helical computed tomography (CT) scanning. Only patients with no evidence of suspicious mediastinal lymph nodes on the CT scan (less than 1 cm in the short axis) were included in this study. A mediastinal lymph node dissection was performed routinely with metastasectomy. Results. In 9 patients (14.3%) at least one mediastinal lymph node revealed malignant cells in accordance with the resected metastases. When compared with the preoperative CT scan, additional pulmonary metastases were detected in 16.9% of performed operations. There was a trend toward an improved survival rate in patients without involvement of the mediastinal lymph nodes. The number of pulmonary metastases had no influence on survival. Conclusions. On a patient-by-patient basis, the frequency of misdiagnosed mediastinal lymph node metastases is about the same as compared with non small cell bronchial carcinomas. Systematic mediastinal lymph node dissection reveals a significant number of patients, who otherwise are assumed free of residual tumor. The knowledge of metastases to mediastinal lymph nodes after complete resection of pulmonary metastases could influence the decision for adjuvant therapy in selected cases. (Ann Thorac Surg 2001;72:225 9) 2001 by The Society of Thoracic Surgeons Pulmonary metastases are often best treated by resection if the primary tumor is under control and metastatic disease at other sites is absent [1]. The most consistent prognostic factor predicting survival is complete surgical removal of all metastases [2, 3]. It has been reported that thoracotomy and palpating the deflated lung during resection of pulmonary lesions presumed to be metastases often reveals additional pulmonary lesions, which were not detected on the preoperative computed tomography (CT) chest scan. When compared with preoperative CT chest scans, additional malignant pulmonary lesions were found in up to 30% of patients by thoracotomy [4, 5]. Thoracotomy and systematic nodal dissection performed in curative resection of bronchial carcinoma revealed metastases to mediastinal lymph nodes in 10% to 36%, which appeared to be of normal size on the preoperative CT scan [6, 7]. Because dissection of mediastinal lymph nodes is usually not performed routinely during Accepted for publication March 1, 2001. Address reprint requests to Dr Loehe, Department of Surgery, University of Munich, Marchioninistr 15, D-81377 Munich, Germany; e-mail: floehe@hotmail.com. metastasectomy, no studies have reported on additional lymph node metastases when the CT chest scan shows only normal-sized mediastinal lymph nodes (less than 1 cm in short axis). Therefore, the purpose of the present study was to investigate the incidence of malignant mediastinal lymph nodes when compared with preoperative CT chest scans in patients with pulmonary metastases. Material and Methods During a 33-month period (January 1996 to September 1998) a total of 71 pulmonary resections and ipsilateral mediastinal lymph node dissections through a thoracotomy were performed on 63 consecutive patients with pulmonary metastases. In 41 patients (65.1%) lung metastases were the first site of metastatic disease; 22 patients (34.9%) had a history of surgically treated metastases at other locations. In this study only patients with no evidence of extrathoracic metastases and residual or local recurrent disease at the site of primary tumor were included. The 24 women and 39 men had a mean age of 59.3 12.5 years (range 22.2 to 82.7 years). In 8 patients (12.7%) 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00 Published by Elsevier Science Inc PII S0003-4975(01)02615-7

226 LOEHE ET AL Ann Thorac Surg PULMONARY METASTASECTOMY AND MEDIASTINAL LYMPH NODE DISSECTION 2001;72:225 9 Table 1. Primary Malignancy, Number of Resections, and Additional Pulmonary Lesions Compared With Preoperative CT Scan Primary Malignancy Patients Resections Additional Lesions Colon/rectum 18 (28.6%) 20 (28.2%) 4 (22.2%) Renal carcinoma 16 (25.4%) 18 (25.4%) 6 (33.3%) Nasopharyngeal 8 (12.7%) 8 (11.3%) 3 (37.5%) Sarcomas 8 (12.7%) 9 (12.7%) 3 (33.3%) Gynecological 6 (9.5%) 7 (9.9%) 3 (42.8%) Miscellaneous a 7 (11.1%) 9 (12.7%) 0 Total 63 71 19 a Thyroid carcinoma (n 2), testicular tumor (n 2), pancreatic carcinoma (n 2), and bladder carcinoma (n 1). CT computed tomography. pulmonary metastases were bilateral. Pulmonary metastases were discovered on chest radiographs or CT scans performed during routine follow-up. The mean interval between surgical therapy of the primary tumor and appearance of pulmonary metastases was 33.8 39.8 months (range 1 to 204 months). The distribution of primary tumors is shown in Table 1. Chest radiographs were obtained in all patients preoperatively, and intravenous contrast-enhanced helical CT chest scans (slice thickness 8 mm) were performed on state of the art spiral scanners no more than 3 weeks before the planned operation. CT chest scans were analyzed preoperatively by our radiologist (T.H.). The number of pulmonary lesions identified preoperatively was documented and was compared with the number of palpable lesions found during the operation. All palpable pulmonary lesions were radically resected and examined histologically. Furthermore, histologic findings of the dissected mediastinal lymph nodes were compared with the preoperative CT chest scan. Procedure Wedge or segmental resections of the pulmonary nodules were the operative procedures of choice. A lobectomy was performed when either a solitary metastasis was located deeply and centrally in the lobe or when multiple nodules were confined to one lobe. Pneumonectomy was performed if no other choice was available to achieve radical metastasectomy. The surgical approach was a standard anterolateral thoracotomy. Only monolateral procedures were performed. Patients with bilateral metastases (8 of 63 cases) underwent delayed sequential contralateral thoracotomy within 8 weeks. After mobilization of the lung by dissecting the pulmonary ligament and adhesions, the lung was palpated in atelectasis and every suspicious nodule was resected. Wedge resection was the preferred procedure, but when necessary extended resections were performed to achieve an R0-resection (no microscopic or macroscopic malignant disease left after metastasectomy). Mediastinal lymph nodes were dissected systematically at the following locations according to the American Thoracic Society [8]: L1 through L4 (superior and inferior pre- and paratracheal), L5 (subaortic), L6 (paraaortic), L8 (paraesophageal), L9 (pulmonary ligament), and L10 (hilar). Statistical Analysis Differences between the groups were compared with the 2 test. A Kaplan Meier analysis of cumulative survival was calculated and difference in survival was compared using the log rank test. Differences were considered significant when p was less than 0.05. Results The mean disease-free interval (DFI) until the occurrence of pulmonary metastases was 32.2 33.4 months (range 0 to 146 months). We performed 71 pulmonary resections (right side 41; left side 30) in 63 patients. Table 2 lists types of operative procedures for radical resection of all palpable metastases. No intraoperative complication developed. Three patients (4.8%) had postoperative minor complications (1 had superficial wound infection and 2 had pneumonia) that were treated successfully. Major complications developed in 3 patients (4.8%): a Mallory Weiss gastric bleeding was treated successfully by endoscopic intervention, a postoperative pleural empyema had to be drained operatively, and 1 patient with a postoperative myocardial infarction died. Therefore, the total mortality rate was 1.6%. The average hospital stay was 10.8 5.9 days (range 1 to 43 days). A single pulmonary lesion was detected on the preoperative CT chest scan in 39 cases (54.9%). Two pulmonary lesions in the same lung were identified in 13 cases (18.3%), and three or more suspicious lesions in the same lung were identified in 19 cases (26.8%). All CT scans demonstrated complete resectability without evidence of enlarged mediastinal lymph nodes (less than 1 cm in short axis). In all 71 pulmonary resections, histologic examination proved lung metastases in accordance with the primary tumor. There was a discrepancy between the number of identified pulmonary lesions on the preoperative CT scan and the findings on digital palpation of the deflated lung during the operation in 19 resection procedures (26.8%) involving 19 different patients (30.2%). In one case, the number of manually palpated lesions was less than those identified on the preoperative CT scan. De- Table 2. Type of Pulmonary Resections Performed for Radical Metastasectomy Resection n % Wedge 37 52.1 Segmental 9 12.7 Lobectomy 20 28.2 Bilobectomy 1 1.4 Pneumonectomy 4 5.6 Total 71 100

Ann Thorac Surg LOEHE ET AL 2001;72:225 9 PULMONARY METASTASECTOMY AND MEDIASTINAL LYMPH NODE DISSECTION 227 Table 3. Primary Malignancy, Localization and Number of Pulmonary Metastases, Type of Resection, and Level of Metastatic Mediastinal Lymph Nodes Primary Malignancy Localization of Metastases Number of Metastases Resection Level of LN Renal cell carcinoma Left upper lobe 1 Lobectomy L5/L11 Renal cell carcinoma Right lung 5 5 wedge L1 L4 Renal cell carcinoma Right lung 6 6 wedge L10 Squamous cell carcinoma of the pharynx Right middle lobe 1 1 wedge L3/L10 Squamous cell carcinoma of the pharynx Right upper lobe 1 1 wedge L1 L4 Adenocarcinoma of the rectum Left upper lobe 1 Lobectomy L9/L10 Endometrial carcinoma Left lung 4 4 wedge L10 (adenocarcinoma) Ewing sarcoma Left upper lobe 1 Lobectomy L4 Scirrhous adenocarcinoma of the breast Right upper lobe 1 1 wedge L10 pending on the primary tumor, the incidence of additional pulmonary lesions, which were found during the operation, differed between 0% and 42.8% (Table 1). The histologic examination revealed a total number of 22 additional metastases after 71 pulmonary resections (16.9%) involving 12 different patients (19%). None of the performed preoperative CT chest scans suggested malignant alteration by enlargement of the mediastinal lymph nodes. Macroscopically the collected mediastinal lymph nodes were not suspicious, but histologic examination gave evidence of malignant cells in at least one mediastinal lymph node in accordance with the resected lung metastases in 9 cases (12.7%) involving 9 patients (14.3%) (Table 3). The incidence of positive mediastinal lymph nodes depending on the primary tumor ranged between 5.6% (colon/rectum) and 33.3% (gynecological malignancies), but the difference was not significant. There was only a tendency (p 0.49) toward an increasing incidence of nodal involvement between patients with a solitary pulmonary metastasis (9.7%) and patients with two or more pulmonary metastases (17.6%). The mean postoperative follow-up (n 62 patients) was 22.6 11.7 months (range 3 to 48 months), during which 27 of 62 patients died form tumor progression. The overall survival rate was 56.5%. For analyzing the cumulative survival of our patients, a Kaplan Meier curve was calculated (Fig 1). There was no significant difference (p 0.37) in the cumulative survival depending on the mediastinal lymph node status, but there was a trend toward an improved survival in patients without histologic evidence of tumor infiltration of the mediastinal lymph nodes (mean survival 32.9 11.8 months) as compared with patients with positive lymph nodes (mean survival 22.7 15.4 months). The number of pulmonary metastases had no influence on survival. During the observed follow-up period 9 of 20 patients (45%) with a solitary pulmonary metastasis and 18 of 43 patients (41.9%) with two or more pulmonary metastases died. The mean DFI in the group of survivors was 37.6 37.2 months (range 0 to 146 months) and did not differ significantly from the mean DFI of the patients who died (22 21.9 months; range 0 to 88 months). Comment Almost 30% of patients who have curative resection of a malignancy will develop pulmonary metastases [9]. Therefore, all newly detected pulmonary lesions after a history of malignant disease are strongly suspicious for metastases. Only histologic examination of the resected pulmonary nodule provides the correct diagnosis and may influence further therapeutic approaches. Complete surgical resection of pulmonary metastases is now an accepted treatment for improving the overall and disease-free survival and is performed routinely in selected patients. The overall survival rate after complete pulmonary metastasectomy is 36% at 5 years and 26% at 10 years [3]. Survival in patients operated on for solitary metastasis compared with patients with two or more metastases is considered to be significantly higher [5, 9, 10]. In contrast, the number of pulmonary metastases and DFI had no influence on survival in our study. Nevertheless, there is general agreement that a complete resection Fig 1. Kaplan Meier cumulative survival for patients after pulmonary metastasectomy with positive mediastinal lymph nodes ( ) and without nodal involvement ( ), log rank p 0.37.

228 LOEHE ET AL Ann Thorac Surg PULMONARY METASTASECTOMY AND MEDIASTINAL LYMPH NODE DISSECTION 2001;72:225 9 of all pulmonary metastatic lesions is crucial for longterm survival. Incomplete resection decreases survival significantly [2, 3, 5, 9]. In our study, the overall incidence of intraoperatively detected additional pulmonary lesions as compared with the preoperative CT scan was 26.8%, revealing additional pulmonary metastases in 16.9% of performed operations. In 1997, Girard and colleagues [5] observed in a study of 346 patients with pulmonary metastases an overall incidence of additional metastases in 30%. Again our data suggest the recommendation that complete surgical exploration by thoracotomy and palpating the lung should remain the procedure of choice in patients undergoing therapeutic pulmonary metastasectomy [11]. In our study the frequency of wedge and segmental resections (70%) was the same as reported from an analysis of 5,206 cases of pulmonary metastasectomy [3]. Other investigations also showed that a pneumonectomy was necessary in up to 8% of patients for achieving radical metastasectomy [2, 12]. To evaluate the incidence of metastatic involvement of macroscopically unsuspicious mediastinal lymph nodes, we performed a mediastinal lymph node dissection also in patients undergoing wedge resections for pulmonary metastasectomy. There was no difference when comparing the perioperative morbidity and mortality in our study with the results reported by other investigations about pulmonary metastasectomies without mediastinal lymph node dissection [2, 3]. Systematic nodal lymph node dissection must be performed routinely in patients undergoing pulmonary resection of lung cancer for accurate intrathoracic staging, because the sensitivity of CT scans to correctly predict lymph node involvement is only 73% [13, 14]. Several studies have already shown that metastatic involvement of lymph nodes in patients with bronchial carcinoma is frequently misdiagnosed by preoperative CT scan, revealing a false-negative prediction between 14% and 18% [15 17]. All these studies were based on the assumption that 1 cm in the short axis in the transversal plane represents the upper limit of a normal mediastinal lymph node [18, 19]. However, routine mediastinal lymph node dissection accompanying pulmonary metastasectomy is not an accepted standard curative concept in patients with pulmonary metastases. An analysis of the International Registry of Lung Metastases based on 5,206 cases revealed in 5% metastases to hilar or mediastinal lymph nodes, but only suspicious lymph nodes were removed and lymph node dissection was performed in less than 9% of patients [3]. Another study suggested that local recurrence after complete pulmonary metastasectomy seemed to be decreased when a mediastinal lymph node dissection was performed simultaneously [20]. Factors predicting the incidence of mediastinal lymph node involvement in patients with pulmonary metastases are not known. In addition, our investigation, because of the small sample size, does not allow a valid statistical analysis to identify risk factors or predictors for the possibility of nodal involvement or clinical outcome. In our study, we performed routine mediastinal lymph node dissection and found metastases to mediastinal lymph nodes in 12.7% of pulmonary resections; 70% of the pulmonary metastasectomies were performed as wedge and segmental resections in our study. Collecting L11 through L14 lymph nodes could have raised the operative complication rate in these cases. Furthermore, routine dissection of L7 lymph nodes involves an extension of the lymph node dissection to the contralateral mediastinum and could have increased the operative morbidity in our preliminary study. Therefore, routine evaluation L7 and L11 through L14 was not included in our preliminary investigation. The incidence of malignant infiltration of unsuspicious mediastinal lymph nodes in patients undergoing pulmonary metastasectomy is not known. The assessment of L7 and L11 through L14, which is required for a complete systematic node dissection, would have raised the frequency of nodal involvement documented on a patientby-patient basis. The findings of our preliminary study suggest that long-term survival of patients with metastases to mediastinal lymph nodes could be impaired when compared with patients without metastatic lymph node infiltration, whereas DFI and number of pulmonary metastases had no influence on survival. In accordance with other studies, we found additional pulmonary metastases when compared with the number predicted on the preoperative CT scan in up to 17%. Therefore, thoracotomy and digital palpation of the lung should remain the standard operative procedure to achieve curative metastasectomy. Furthermore, based on our results, we suggest that a routine mediastinal lymph node dissection for histologic examination performed with complete metastasectomy provides useful information to assess the state of tumor progression, permitting more accurate oncologic staging. References 1. Rusch VW. Pulmonary metastasectomy. Current indications. Chest 1995;107(Suppl):322S 31S. 2. Girard P, Ducreux M, Baldeyrou P, et al. Surgery for lung metastases from colorectal cancer: analysis of prognostic factors. J Clin Oncol 1996;14:2047 53. 3. Long-term results of lung metastasectomy: prognostic analyses based on 5206 cases. The International Registry of Lung Metastases. J Thorac Cardiovasc Surg 1997;113:37 49. 4. McCormack PM, Ginsberg KB, Bains MS, et al. Accuracy of lung imaging in metastases with implications for the role of thoracoscopy. Ann Thorac Surg 1993;56:863 6. 5. Girard P, Spaggiari L, Baldeyrou P, et al. Should the number of pulmonary metastases influence the surgical decision? Eur J Cardiothorac Surg 1997;12:385 92. 6. Gross BH, Glazer GM, Orringer MB, et al. Bronchogenic carcinoma metastatic to normal-sized lymph nodes and significance. Radiology 1988;166:71 4. 7. Gdeedo A, Van Schil P, Corthouts B, et al. Comparison of imaging [(i)tnm] and pathological TNM [ptnm] in staging of bronchogenic carcinoma. Eur J Cardiothorac Surg 1997;12: 224 7. 8. Martini N, Flehinger BJ, Zaman MB, Beattie EJ. Results of resection in non-oat cell carcinoma of the lung with mediastinal lymph node metastases. Ann Surg 1983;198:386 97.

Ann Thorac Surg LOEHE ET AL 2001;72:225 9 PULMONARY METASTASECTOMY AND MEDIASTINAL LYMPH NODE DISSECTION 229 9. Roberts DG, Lepore V, Cardillo G, et al. Long-term follow-up of operative treatment for pulmonary metastases. Eur J Cardiothorac Surg 1989;3:292 6. 10. MacAfee MK, Allen MS, Trastek VF, et al. Colorectal lung metastases: results of surgical excision. Ann Thorac Surg 1992;53:780 6. 11. Diederich S, Semik M, Lentschig MG, et al. Helical CT of pulmonary nodules in patients with extrathoracic malignancy. CT surgical correlation. AJR Am J Roentgenol 1999; 172:353 60. 12. Mountain CF, McMurtrey MJ, Hermes KE. Surgery for pulmonary metastasis: a 20-year experience. Ann Thorac Surg 1984;38:323 30. 13. Graham AN, Chan KJ, Pastorino U, Goldstraw P. Systematic nodal dissection in the intrathoracic staging of patients with non-small cell lung cancer. J Thorac Cardiovasc Surg 1999; 117:246 51. 14. Weng E, Tran L, Rege S, et al. Accuracy and clinical impact of mediastinal lymph node staging with FDG-PET imaging in potentially resectable lung cancer. Am J Clin Oncol 2000; 23:47 52. 15. Izbicki JR, Thetter O, Karg O, et al. Accuracy of computed tomographic scan and surgical assessment for staging of bronchial carcinoma. J Thorac Cardiovasc Surg 1992;104: 413 20. 16. Arita T, Kuramitsu T, Kawamura M, et al. Bronchogenic carcinoma: incidence of metastases to normal sized lymph nodes. Thorax 1995;50:1267 9. 17. Arita T, Matsumoto T, Kuramitsu T, et al. Is it possible to differentiate malignant mediastinal nodes from benign nodes by size? Reevaluation by CT, transesophageal echocardiography, and nodal specimen. Chest 1996;110:1004 8. 18. Glazer GM, Gross BH, Quint LE, et al. Normal mediastinal lymph nodes: number and size according to American Thoracic Society mapping. AJR Am J Roentgenol 1985;144: 261 5. 19. Kiyono K, Sone S, Sakai F, et al. The number and size of normal mediastinal lymph nodes. AJR Am J Roentgenol 1988;150:771 6. 20. Kamiyoshihara M, Hirai T, Kawashima O, et al. The surgical treatment of metastatic tumors in the lung: is lobectomy with mediastinal lymph node dissection suitable treatment? Oncol Rep 1998;5:453 7. INVITED COMMENTARY Systematic nodal dissection (SND) has become standard practice when undertaking pulmonary resection for lung cancer with curative intent. It is accepted by the International Association for the Study of Lung Cancer (IASLC) as an important step in the intrathoracic staging of lung cancer. In practice it has two component steps; the first to excise all mediastinal fat and the nodes contained therein, submitted for histological analysis labeled in accordance with an internationally accepted nodal chart, and the second step to extend this evaluation to the hilar, interlobar, lobar and segmental nodes in a centrifugal manner until the extent of resection has been determined. SND will demonstrate unexpected N 2 disease in up to 25% of cases and ensures that complete resection has been performed with the minimum resection of lung parenchyma. As the resection of pulmonary metastases has become established for the curative treatment of selected patients with varied primary tumors in most parts of the body surgeons have experienced the occasional unpleasant surprise when unexpected involved nodes have been found at thoracotomy. Undoubtedly earlier estimates of the incidence of nodal deposits have been an underestimate as nodal dissection has not been systematic. We all should therefore be grateful to Dr Loehe and colleagues for doing the study many felt was necessary but no one else was prepared to undertake. Their study, published in this issue of The Annals of Thoracic Surgery provides the answer to the question as to whether SND should be routine at pulmonary metastasectomy. The answer is a definite yes, with unexpected nodal disease being identified in N 1 and N 2 stations in 1 in 7 patients (14.3%). No doubt the debate will now follow the same lines as that over SND in lung cancer. Does SND contribute to our attempts at cure, or is it an investigation that merely gives insight into the prospects for cure? Should the findings influence decisions regarding adjuvant therapy, and will this improve the cure rate? Is there increased morbidity with this extension to the surgical procedure? Can nodal dissection be applied selectively according to cell-type, number, size and site of metastases? Is this another argument against the resection of pulmonary metastases by video-assisted techniques? Loehe and colleagues have established that SND is necessary and it is for the thoracic surgical community to respond with further evidence to answer these questions. Peter Goldstraw, FRCS Department of Thoracic Surgery Royal Brompton and Harefield NHS Trust Sydney St London SW3 6NP, England e-mail: p.goldstraw@rbh.nthames.nhs.uk. 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00 Published by Elsevier Science Inc PII S0003-4975(01)02682-0