- Email: [email protected]
Prognostic significance of lymph node metastasis found during pulmonary metastasectomy for extrapulmonary carcinoma
GENERAL THORACIC
Prognostic Significance of Lymph Node Metastasis Found During Pulmonary Metastasectomy for Extrapulmonary Carcinoma Sina Ercan, MD, Francis C. Nichols III, MD, Victor F. Trastek, MD, Claude Deschamps, MD, Mark S. Allen, MD, Daniel L. Miller, MD, Cathy D. Schleck, BS, and Peter C. Pairolero, MD Divisions of General Thoracic Surgery and Biostatistics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota, and Division of General Thoracic Surgery, Mayo Clinic, Scottsdale, Arizona
Background. The prognostic significance of lymph node metastasis in cancer patients is well documented. Pulmonary metastasectomy in selected patients is associated with improved survival. Little is known about the prognostic significance of lymph node metastases found during pulmonary metastasectomy for extrapulmonary carcinoma metastatic to the lung. Methods. The records of all patients who underwent pulmonary metastasectomy and complete mediastinal lymph node dissection for extrapulmonary carcinomas at our institution from November 1985 through July 1999 were reviewed. Results. Eight hundred eighty-three patients underwent pulmonary metastasectomy. Of these, 70 patients (7.9%) (44 men, 26 women) had concomitant complete lymphadenectomy. Median age was 64 years (range, 33 to 83 years). Median time interval between primary tumor resection and metastasectomy was 34 months (range, 0 to
188 months). Wedge excision was performed in 46 patients, lobectomy in 16, both in 7, and pneumonectomy in 1. Lymph node metastases were found in 20 patients (28.6%) and were classified as intrapulmonary or hilar (N1) in 9, mediastinal (N2) in 8, and both in 3. There were no operative deaths. Median follow-up was 6.6 years (range, 1.1 to 14.6 years). Three-year survival for patients with negative lymph nodes was 69% as compared with only 38% for those with positive lymph nodes (p < 0.001). Conclusions. The presence of lymph node metastases at the time of pulmonary metastasectomy for extrapulmonary carcinoma has an adverse effect on prognosis. Complete mediastinal lymph node dissection should be considered at the time of pulmonary metastasectomy for carcinoma to improve staging and guide treatment.
T
Despite either complete lymphadenectomy or lymph node sampling being the accepted standard in pulmonary resection for primary lung cancer [4], the same is not true for pulmonary metastases. Little information exists regarding the prognostic significance of metastatic disease found in regional lymph nodes at the time of curative pulmonary metastasectomy. To evaluate the influence of lymph node involvement, we retrospectively reviewed our experience with complete mediastinal lymphadenectomy performed at the time of pulmonary metastasectomy.
he inability to prevent and eradicate metastatic disease remains the primary cause of cancer death. The lung is a frequent site of distant metastases, and the presence of pulmonary metastases often signifies uncontrollable tumor spread and eventual death. Nonetheless, some patients do have metastases limited solely to the lungs. Despite years of early skepticism, it is now generally recognized that overall survival can be improved in carefully selected patients undergoing resection of pulmonary metastases [1–3]. Several prognostic indicators have been identified in patients undergoing surgical excision of lung metastases. These include the time interval between resection of the primary tumor and the finding of pulmonary metastases, number of metastases, resectability of metastases, and the cell type of the primary tumor. Accepted for publication June 30, 2003. Presented at the Thirty-eighth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28 –30, 2002. Address reprint requests to Dr Nichols, Division of General Thoracic Surgery, Mayo Clinic and Mayo Foundation, 200 First St, SW, Rochester, MN 55905; e-mail: [email protected].
© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2004;77:1786 –91) © 2004 by The Society of Thoracic Surgeons
Material and Methods Between November 1985 and August 1999, eight hundred eighty-three patients underwent pulmonary resection for metastases from nonsarcomatous extrathoracic primary cancers at the Mayo Clinic in Rochester, Minnesota. Complete mediastinal lymphadenectomy had been performed in 107 of these patients (12.1%). Thoracic lymph node stations were classified according to the International System for Staging Lung Cancer [5]. The record of each patient was reviewed for age, gender, primary 0003-4975/04/$30.00 doi:10.1016/S0003-4975(03)01200-1
tumor date, location, histology, treatment, and current status. The date, location, histology, and treatment of pulmonary metastases, number of pulmonary resections, mediastinal and regional lymph node status, postoperative complications, and long-term survival were also reviewed. Only patients who had complete resection of all known extrapulmonary cancer were included. Pulmonary resection was deemed curative if all known intrathoracic cancer was removed at thoracotomy. Thirty-seven patients (34.6%) were eliminated due to an inadequate primary operation, persistent unresectable extrathoracic disease, unresectable intrathoracic disease, pulmonary pathology different from the primary pathology, insufficient data, or refusal to participate in research. All patients were evaluated preoperatively for local recurrence of their primary cancer and for sites of extrathoracic metastases. Computed tomography (CT) was the procedure most commonly utilized. Head CT and bone scans were utilized if clinically indicated. Barium enema or colonoscopy was performed in colon cancer patients. Positron emission tomography (PET) was not available during the time frame of this study. Operative mortality included patients who died within 30 days after thoracotomy or those who died later but during the same hospitalization. Survival was estimated by the Kaplan-Meier method using the date of the first pulmonary metastasectomy with lymph node dissection as the starting point [6]. The influence of variables on survival was analyzed using the log-rank test for discrete variables [7] and the proportional hazards model of Cox for continuous variables and multivariate models [8]. A p value of less than or equal to 0.05 was considered significant. Approval for this study was granted by Mayo Foundation’s Institutional Review Board.
Clinical Findings There were 70 patients (44 men and 26 women). Median age was 64 years and ranged from 33 to 83 years. The primary cancer was colon in 44 patients, renal cell in 13, malignant melanoma in 6, endometrial and salivary gland carcinoma in 2 each, and bladder, prostate, and testicular carcinoma in 1 each. Fifty-six patients (80.0%) had their primary cancer resection elsewhere. After resection of the primary cancer and before the development of pulmonary metastases, 11 patients were treated with chemotherapy, 5 with radiation therapy, and 5 with a combination of both. Three patients early in the series were evaluated with standard chest radiography and conventional tomography. Sixty-two patients (88.6%) had a chest CT that showed a solitary nodule in 43 patients, two in 11, and three or more in 8. Ten patients had bilateral nodules. Twelve patients had enlarged mediastinal or regional lymph nodes on CT. Except for these patients, no particular concern about the malignant potential of the thoracic lymph nodes was identified preoperatively to account for the complete lymphadenectomy. Twenty-two patients (31.4%) underwent preoperative transthoracic needle aspiration (TTNA). In 21 (95.4%), the TTNA was positive for malignancy and histologically similar to the primary cancer. Eleven patients had
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
1787
bronchoscopy, and 4 were diagnostic for metastatic carcinoma. Seven patients had undergone extrathoracic metastasectomies before pulmonary resection; all were considered to have had complete resection. The carcinoembryonic antigen was measured in 31 of the 44 patients (70.4%) with colon cancer and was elevated (ⱖ5 ng/mL) in 9. The median interval between resection of the primary cancer and the first pulmonary resection was 34 months (range, 0 to 188 months). Mediastinoscopy was performed in 12 patients, including 5 with clinically enlarged mediastinal lymph nodes; none was positive. At the time of initial pulmonary metastasectomy, a lateral thoracotomy was performed in 64 patients, median sternotomy in 1, and simultaneous left thoracoscopic wedge resection and right thoracotomy in 1. Four patients underwent staged bilateral thoracotomies. In no patient was there concern that a primary lung cancer was being resected rather than a pulmonary metastasis. Intraoperatively, no particular concern of the operating surgeon regarding the malignant potential of the thoracic lymph nodes was identified to account for the complete lymphadenectomy. Forty-six patients had wedge excisions; 18 were multiple. A lobectomy was performed in 16 patients, lobectomy combined with wedge excision in 7, and pneumonectomy in 1. Fifteen patients had a second thoracotomy, 4 a third, and 2 a fourth. Altogether, 142 metastases were resected in these 70 patients, and included 98 wedge excisions, two segmentectomies, 27 lobectomies, and one each bilobectomy, sleeve lobectomy, and pneumonectomy. One patient had an associated chest wall resection. At the initial pulmonary metastasectomy, 48 patients had solitary metastases, 12 had two, 5 had three, 4 had four, and 1 had nine. This was greater than the number of nodules detected preoperatively by chest CT. Lymph node metastases were found in 20 patients (28.6%) and were classified as N1 in 9 patients, N2 in 8, and both in 3. The lymph node dissection was done at the time of the second ipsilateral metastasectomy in 2 patients and during a contralateral metastasectomy after two previous ipsilateral metastasectomies in a third. All 3 of these latter patients were N0 at the time of lymphadenectomy.
Results There were no operative deaths. Complications occurred in 7 patients (10%) and included atrial fibrillation in 2 (2.8%), pneumothorax in 2 (2.8%), and recurrent laryngeal nerve paresis, myocardial infarction, and bronchopleural fistula in 1 each. Hospitalization ranged from 4 to 30 days, with a median of 7 days. Follow-up was complete in all patients and ranged from 1 to 14.6 years, with a median of 6.6 years. Twentyseven patients are currently alive, 5 with evidence of recurrent cancer. Forty-three patients are dead, 34 from recurrent cancer. The median disease-free interval from the time of the extrathoracic primary operation to pulmonary metastasectomy for those patients with metastatically involved lymph nodes was 36 months, as compared to 32 months for those patients without lymph node involvement (p ⫽ 0.64). Throughout the study period,
GENERAL THORACIC
Ann Thorac Surg 2004;77:1786 –91
1788
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
GENERAL THORACIC Fig 1. Estimated survival (death due to any cause) of 70 patients undergoing pulmonary metastasectomy and complete mediastinal lymph node dissection. Zero time on the abscissa is date of first pulmonary metastasectomy and lymph node dissection.
metastatic cancer recurred in 43 patients. Recurrence was limited to the lungs in 15 patients, extrathoracic sites only in 9, and both in 19. Tumor recurrence was significantly higher in the patients with metastatically involved lymph nodes (p ⫽ 0.004). The 3-year recurrence free rate was 23% (95% CI, 8% to 66%) in patients with metastatically involved lymph nodes and 49% (95% CI, 36% to 66%) in patients without metastatically positive lymph nodes. Overall, estimated 5- and 10-year survival was 48% (95% CI, 36% to 60%) and 29% (95% CI, 16% to 42%), respectively (Fig 1). Three-year survival for patients without lymph node metastases was 69% (95% CI, 58% to 83%) as compared to only 38% (95% CI, 20% to 68%) for those with metastases (p ⬍ 0.001) (Fig 2). Survival was not improved when only N1 lymph nodes were metastatically involved. Moreover, only 2 patients with metastatically involved lymph nodes survived beyond 4 years. Both of these patients had low-grade renal cell carcinomas.
Fig 2. Estimated survival of 70 patients undergoing pulmonary metastasectomy and complete mediastinal lymphadenectomy without metastatic nodal involvement (negative) and with nodal involvement (positive). Zero time on the abscissa is date of first pulmonary metastasectomy and lymph node dissection. (p ⬍ 0.001).
Ann Thorac Surg 2004;77:1786 –91
The number of pulmonary metastases was not associated with a difference in survival. Three-year survival for patients with a solitary metastasis was 61% (95% CI, 49% to 78%), as compared with 57% (95% CI, 39% to 82%) for those with multiple metastases (p ⫽ 0.36). No difference in lymph node metastases was observed in patients with a solitary metastasis versus multiple pulmonary metastases (p ⫽ 0.11). During the follow-up period, 19 of 48 patients (39.6%) with a solitary metastasis and 15 of 22 (68.2%) with two or more pulmonary metastases died due to recurrent cancer. Age, gender, cell type of the primary tumor, and the extent of resection did not influence survival. Fifteen patients underwent a second thoracotomy for recurrent pulmonary metastases; 9 had a right thoracotomy, 6 a left, and 1 had an associated chest wall resection. At the time of second thoracotomy, 9 patients had a solitary metastasis and 6 had two or more. In 2 patients, both with low-grade metastatic renal cell carcinoma, metastatic lymph node disease was found at the time of the second thoracotomy. Survival in these 15 patients after their second pulmonary metastasectomy was 82% (95% CI, 59% to 100%) at 5 years. Four patients had a third pulmonary resection for recurrent cancer and 2 patients had a fourth. Three of these 4 patients are still alive 2 to 7 years after their last pulmonary metastasectomy. None has evidence of recurrent disease.
Comment Controversy exists as to the frequency with which pulmonary metastases can metastasize to regional lymph nodes. An autopsy series by Abrams and colleagues [9] demonstrated a 33% incidence of mediastinal lymph node metastases in patients with nonpulmonary carcinoma. In 1965, Thomford and colleagues [10] reviewed our experience with pulmonary metastasectomy when they reported 221 pulmonary resections in 205 patients over a 21-year period. Twenty patients (9.8%) had metastatically involved lobar or hilar lymph nodes. Five-year survival was 30.3%, but the prognostic significance of lymph node involvement was not analyzed. A report from the International Registry of Lung Metastases described a 5% incidence of lymph node metastases in 5,206 patients undergoing pulmonary metastasectomy [1]. Of note, however, is that lymph nodes were resected or sampled in only 4.6% of these patients. Putnam [3] suggested that mediastinal lymph node metastases rarely occur from pulmonary metastases; however, his report deals exclusively with metastatic sarcoma. We agree that nodal metastases are much less common with sarcomatous pulmonary metastases, and this is why patients with metastatic sarcoma were not included in this report. In contrast, Martini and McCormack [2] concluded that pulmonary carcinoma metastases are likely to metastasize to regional lymph nodes, especially melanomas. They fail, however, to mention the prognostic significance of this event and did not discuss mediastinal lymph node dissection as part of pulmonary metastasectomy. Our recent experience [11, 12] suggests that metastatic lymph node disease at the time of pulmonary
metastasectomy for carcinoma is present in 8.6% to12.5% of patients. However, until recently, routine lymph node dissection has not been uniformly performed in our practice, as shown in this study by the low rate of lymphadenectomies (12.1%) in our patients undergoing pulmonary metastasectomy. Whereas many of our patients having pulmonary metastasectomy had lymph nodes identified on the final pathology report, it was only these 107 patients who underwent complete mediastinal lymphadenectomy. Except for the 12 patients with mediastinal or regional lymphadenopathy identified on the preoperative chest CT, no particular concern about the mediastinal lymph nodes either preoperatively or intraoperatively was identified to account for the complete lymphadenectomy. Of these 12 patients, 5 underwent mediastinoscopy and all were negative for metastatically involved lymph nodes. Ultimately, only 2 of these 12 patients had lymph node metastasis. To date, few reports have addressed the prognostic significance of metastatically involved lymph nodes with pulmonary metastasectomy. The only factor affecting survival in our series was the presence of lymph node metastasis. Kamiyoshihara and colleagues [13] in a retrospective review of 28 patients undergoing pulmonary metastasectomy found no detectable difference between patients with and without positive regional lymph nodes. Unlike other series, these authors reported no long-term survivors in patients undergoing partial resection (ie, less than lobectomy). More recently, Loehe and colleagues [14] have attempted to clarify the importance of systematic lymph node dissection during pulmonary metastasectomy. In a group of 63 patients, they found a 14.3% incidence of mediastinal lymph node metastases. Similar to their study, we found that the number of pulmonary metastases was not associated with a difference in survival. Also similar was the trend toward an increasing incidence of metastatic lymph node involvement with an increasing number of metastases. In contrast, Loehe and colleagues [14] failed to show a difference in survival in patients with metastatically involved mediastinal lymph nodes (p ⫽ 0.37). Our study demonstrates that patients undergoing pulmonary metastasectomy who are found to have metastatically involved regional lymph nodes have an increased incidence of cancer recurrence and decreased survival. We believe this observation supports systematic lymph node dissection at the time of pulmonary metastasectomy for metastatic carcinoma. Several questions regarding the indications and potential therapeutic role of mediastinal lymphadenectomy or sampling remain [15]. Unanswered is whether regional lymph node dissection at the time of pulmonary metastasectomy contributes to improved survival, or is it simply a prognostic indicator perhaps suggesting the need for adjuvant therapy. Does the number of pulmonary
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
1789
metastases influence the need for lymph node dissection? Accepting the role for lymph node dissection, does this further cloud the role for utilizing video-assisted thoracic surgery in pulmonary metastasectomy? These questions hopefully will be answered in future prospective clinical trials. In conclusion, surgical resection of pulmonary metastases should be offered to patients whose primary tumor is controlled and are good operative risks. All pulmonary and extrathoracic metastases must be amenable to complete surgical resection for a survival benefit to be achieved. Our findings suggest that the presence of metastatic disease in regional lymph nodes has an adverse effect on survival in those patients undergoing curative pulmonary metastasectomy. Therefore, we recommend complete lymph nodes dissection in these patients to better define their prognosis and potentially guide adjuvant therapy.
References 1. Pastorino U, Buyse M, Friedel G, et al. Long-term results of lung metastasectomy: prognostic analysis based on 5206 cases. J Thorac Cardiovasc Surg 1997;113:37–49. 2. Martini N, McCormack PM. Evolution of the surgical management of pulmonary metastases. Chest Surg Clin North Am 1998;8:13–27. 3. Putnam JB. Secondary tumors of the lung. In: Shields TW, LoCicero J, Ponn RB, eds. General thoracic surgery, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2000:1555–76. 4. LoCicero J, Ponn RB, Daly BD. Surgical treatment of nonsmall cell lung cancer. In: Shields TW, LoCicero J, Ponn RB, eds. General thoracic surgery, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2000:1311–41. 5. Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997;111:1710 –7. 6. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–81. 7. Peto R, Peto J. Asymptotically efficient rank andinvariant procedures. J R Statist Soc, Series A 1972;135:185–201. 8. Cox DR. Regression models and life-tables (with discussion). J R Statist Soc, Series B 1972;34:187–220. 9. Abrams HL, Spiro R, Goldstein N. Metastases in carcinoma: analysis of 1000 autopsied cases. Cancer 1950;3:75–85. 10. Thomford NR, Woolner LB, Clagett OT. The surgical treatment of metastatic tumors in the lungs. J Thorac Cardiovasc Surg 1965;49:357–69. 11. McAfee MK, Allen MS, Trastek VF, et al. Colorectal lung metastases: results of surgical excision. Ann Thor Surg 1992;53:780 –6. 12. Cerfolio RJ, Allen MS, Deschamps C, et al. Pulmonary resection of metastatic renal cell carcinoma. Ann Thorac Surg 1994;57:339 –44. 13. 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. 14. Loehe F, Kobinger S, Hatz RA, et al. Value of systemic mediastinal lymph node dissection during pulmonary metastasectomy. Ann Thorac Surg 2001;72:225–9. 15. Goldstraw P. Invited commentary. Ann Thorac Surg 2001;72: 229.
GENERAL THORACIC
Ann Thorac Surg 2004;77:1786 –91
1790
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
Ann Thorac Surg 2004;77:1786 –91
DISCUSSION GENERAL THORACIC
DR HENNING A. GAISSERT (Boston, MA): Only a minority of patients with pulmonary metastatic carcinoma are candidates for surgical resection, and only a minority of those undergoing resection achieve long-term survival. Clearly, we are in need of improving patient selection. Doctor Ercan and associates deserve recognition for reminding us of the metastatic potential of pulmonary metastasis. I thank the authors for providing me with a copy of the manuscript in advance of the meeting. Their investigation originates from an interest for the prognostic significance of lymph node metastasis in patients with resectable pulmonary metastasis. The analysis is performed on a smaller select group of patients among them, and a difference in survival is found. The question arises whether the finding of metastatic nodes in 29% of the select group can be used to make statements about the total group undergoing lung resection. We do not know what criteria led to lymph node dissection. Although long-term survival is comparable with other studies, the node dissection group may have had more advanced disease. The rate of lymph node metastasis, for example, was twice that of the recent prospective study from the University of Munich quoted in your manuscript. Your conclusions should therefore not necessarily be generalized to all patients with resectable pulmonary metastasis. I would like to ask you whether the pattern of recurrence in patients with lymph node metastasis was different from those without. In your manuscript, you did not break down the numbers dependent on its presence or absence. If 29% of all patients with resectable pulmonary metastasis had lymph node disease, one might investigate lymph node spread preoperatively, for example, by mediastinoscopy. If that number drops to 14%, as suggested by the Munich group, one may not stage every patient. Has the knowledge gained from this study changed your practice? And if you now saw a patient with a solitary peripheral pulmonary metastasis and a metastatic hilar lymph node, would you be concerned enough about the lymphatic drainage bed to perform a lobectomy rather than a wedge resection? I enjoyed your presentation and I thank the Society for the opportunity to discuss this paper. DR ERCAN: Thank you, Dr Gaissert, for discussing the paper. We feel that we can generalize our findings to patients with pulmonary metastases from extrapulmonary carcinomas. We realize that only 12% of our nearly 400 metastasectomy patients had complete lymph node dissections. This is likely due to the fact that early in our pulmonary metastasectomy experience lymphadenectomy was not routinely performed. In the literature from autopsy studies, intrathoracic lymph node metastases have been found in 30% of patients with nonpulmonary carcinoma. Our finding of 28.6% is remarkably close to this. The same is not true for patients with sarcoma. For patients with pulmonary metastases where the primary tumor was a carcinoma, we now routinely perform a mediastinal lymphadenectomy. As for your other question, the pattern of recurrence was differentiated as either being intrathoracic or extrathoracic. For the patients who did recur, most of them did not undergo re-resection as they recurred systemically. The recurrences were sometimes only identified radiographically but were obvious. We believe we have shown that the recurrence rate was statistically higher in patients who had lymph node involvement identified during the pulmonary metastasectomy.
As for the patient who had a solitary metastasis with hilar lymph node involvement, the ultimate goal during pulmonary metastasectomy is to remove all obvious disease while at the same time conserve normal lung. Lobectomy would not be our standard approach; rather it would be wedge resection and complete lymphadenectomy. Sometimes, however, the metastasis is located in such a position that resection requires a lobectomy. We do in select patients occasionally perform lobectomy for the resection of pulmonary metastasis. Based on the poor prognosis in patients found to have metastatically involved lymph nodes, our approach to those patients might change in terms of consideration for future resections or the need for adjuvant therapy. DR DOUGLAS E. WOOD (Seattle, WA): I like your paper, Dr Ercan, and it was valuable up until the conclusions. I have to say I disagree completely with where you went from your data to the conclusions. If we look at the percent of patients who had lymph node metastases out of your total cohort, it is actually less than 3%, and I would be hard pressed to recommend that we add a significant surgical procedure for 3% of patients. It is only 20 patients out of 833 patients, not 20 patients out of 70. You selected those patients, I am sure, on the basis of enlarged lymph nodes at the time of surgery or on preoperative staging. The question is, do you think that these are lymph node metastases from the metastases or are these synchronous metastases to the mediastinum along with the pulmonary metastases, implying just a more aggressive biologic behavior? Do you actually think that mediastinal lymphadenectomy confers any therapeutic benefit? DR ERCAN: Thank you, Dr Wood. Unfortunately, of the 883 patients only 107 underwent documented complete lymphadenectomy. Thirty-seven patients were eliminated for a variety of reasons including an inadequate primary operation, persistent or unresectable extrathoracic disease, and unresectable intrathoracic disease. The remaining 70 patients were not selected because of preoperative or intraoperative findings. For patients with worrisome mediastinal nodes preoperatively, mediastinoscopy was performed. None of the operative records showed any concern about the intrathoracic lymph nodes. It is true that one of the surgeons in the study believed more than others in thoracic lymphadenectomy for these types of patients. This fact likely accounts for the overall low number of complete lymph node dissections. According to the outcome, once these patients had lymph node involvement, they clearly showed a worse survival. We cannot be sure that doing the mediastinal dissection added any positive benefit to the outcome. However, the patients were better staged. Given the poor prognosis of patients with metastatically involved lymph nodes, consideration might be given to adjuvant therapy. Ultimately, this might only be answered with larger prospective and randomized studies. DR STEVEN R. DEMEESTER (Los Angeles, CA): I wondered if you are using just histology to look for metastases in the lymph nodes or if you are using immunohistochemistry, and if you are not using it, what higher percentage you might find with immunohistochemistry staining of the lymph nodes, and what that might impart in terms of prognosis for the patients?
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
1791
DR ERCAN: Thank you very much and that is a good point. During the time of this study, immunohistochemical evaluation was not routine. That might be something to look at in the future.
that if you analyzed survival in metastatic renal cell carcinoma or melanoma you would have achieved anywhere near this 3-year survival rate. Were there enough non– colon cancer patients to perform this analysis?
DR RENEE S. HARTZ (New Orleans, LA): I believe my question is consistent with those of Drs DeMeester and Wood in that on one of your slides you showed a 70% survival rate at 3 years after simple metastectomy. This is an astonishing figure, and you did not elaborate on whether or not your univariate and multivariate analyses were performed on the basis of primary tumor cell type. Can this high survival rate be explained by the fact that most of your patients had colonic primaries? It seems unlikely
DR ERCAN: Thank you very much. As you said, the majority of the patients had colonic primaries, and 26 patients had a variety of other tumor types (renal cell cancer in 13 patients and melanoma in 6). Cell type of the primary did not influence survival. Interestingly, there were only 2 patients with metastatically involved lymph nodes who survived beyond 4 years. Both of these patients had low-grade renal cell carcinomas.
GENERAL THORACIC
Ann Thorac Surg 2004;77:1786 –91
Prognostic Significance of Lymph Node Metastasis Found During Pulmonary Metastasectomy for Extrapulmonary Carcinoma Sina Ercan, MD, Francis C. Nichols III, MD, Victor F. Trastek, MD, Claude Deschamps, MD, Mark S. Allen, MD, Daniel L. Miller, MD, Cathy D. Schleck, BS, and Peter C. Pairolero, MD Divisions of General Thoracic Surgery and Biostatistics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota, and Division of General Thoracic Surgery, Mayo Clinic, Scottsdale, Arizona
Background. The prognostic significance of lymph node metastasis in cancer patients is well documented. Pulmonary metastasectomy in selected patients is associated with improved survival. Little is known about the prognostic significance of lymph node metastases found during pulmonary metastasectomy for extrapulmonary carcinoma metastatic to the lung. Methods. The records of all patients who underwent pulmonary metastasectomy and complete mediastinal lymph node dissection for extrapulmonary carcinomas at our institution from November 1985 through July 1999 were reviewed. Results. Eight hundred eighty-three patients underwent pulmonary metastasectomy. Of these, 70 patients (7.9%) (44 men, 26 women) had concomitant complete lymphadenectomy. Median age was 64 years (range, 33 to 83 years). Median time interval between primary tumor resection and metastasectomy was 34 months (range, 0 to
188 months). Wedge excision was performed in 46 patients, lobectomy in 16, both in 7, and pneumonectomy in 1. Lymph node metastases were found in 20 patients (28.6%) and were classified as intrapulmonary or hilar (N1) in 9, mediastinal (N2) in 8, and both in 3. There were no operative deaths. Median follow-up was 6.6 years (range, 1.1 to 14.6 years). Three-year survival for patients with negative lymph nodes was 69% as compared with only 38% for those with positive lymph nodes (p < 0.001). Conclusions. The presence of lymph node metastases at the time of pulmonary metastasectomy for extrapulmonary carcinoma has an adverse effect on prognosis. Complete mediastinal lymph node dissection should be considered at the time of pulmonary metastasectomy for carcinoma to improve staging and guide treatment.
T
Despite either complete lymphadenectomy or lymph node sampling being the accepted standard in pulmonary resection for primary lung cancer [4], the same is not true for pulmonary metastases. Little information exists regarding the prognostic significance of metastatic disease found in regional lymph nodes at the time of curative pulmonary metastasectomy. To evaluate the influence of lymph node involvement, we retrospectively reviewed our experience with complete mediastinal lymphadenectomy performed at the time of pulmonary metastasectomy.
he inability to prevent and eradicate metastatic disease remains the primary cause of cancer death. The lung is a frequent site of distant metastases, and the presence of pulmonary metastases often signifies uncontrollable tumor spread and eventual death. Nonetheless, some patients do have metastases limited solely to the lungs. Despite years of early skepticism, it is now generally recognized that overall survival can be improved in carefully selected patients undergoing resection of pulmonary metastases [1–3]. Several prognostic indicators have been identified in patients undergoing surgical excision of lung metastases. These include the time interval between resection of the primary tumor and the finding of pulmonary metastases, number of metastases, resectability of metastases, and the cell type of the primary tumor. Accepted for publication June 30, 2003. Presented at the Thirty-eighth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28 –30, 2002. Address reprint requests to Dr Nichols, Division of General Thoracic Surgery, Mayo Clinic and Mayo Foundation, 200 First St, SW, Rochester, MN 55905; e-mail: [email protected].
© 2004 by The Society of Thoracic Surgeons Published by Elsevier Inc
(Ann Thorac Surg 2004;77:1786 –91) © 2004 by The Society of Thoracic Surgeons
Material and Methods Between November 1985 and August 1999, eight hundred eighty-three patients underwent pulmonary resection for metastases from nonsarcomatous extrathoracic primary cancers at the Mayo Clinic in Rochester, Minnesota. Complete mediastinal lymphadenectomy had been performed in 107 of these patients (12.1%). Thoracic lymph node stations were classified according to the International System for Staging Lung Cancer [5]. The record of each patient was reviewed for age, gender, primary 0003-4975/04/$30.00 doi:10.1016/S0003-4975(03)01200-1
tumor date, location, histology, treatment, and current status. The date, location, histology, and treatment of pulmonary metastases, number of pulmonary resections, mediastinal and regional lymph node status, postoperative complications, and long-term survival were also reviewed. Only patients who had complete resection of all known extrapulmonary cancer were included. Pulmonary resection was deemed curative if all known intrathoracic cancer was removed at thoracotomy. Thirty-seven patients (34.6%) were eliminated due to an inadequate primary operation, persistent unresectable extrathoracic disease, unresectable intrathoracic disease, pulmonary pathology different from the primary pathology, insufficient data, or refusal to participate in research. All patients were evaluated preoperatively for local recurrence of their primary cancer and for sites of extrathoracic metastases. Computed tomography (CT) was the procedure most commonly utilized. Head CT and bone scans were utilized if clinically indicated. Barium enema or colonoscopy was performed in colon cancer patients. Positron emission tomography (PET) was not available during the time frame of this study. Operative mortality included patients who died within 30 days after thoracotomy or those who died later but during the same hospitalization. Survival was estimated by the Kaplan-Meier method using the date of the first pulmonary metastasectomy with lymph node dissection as the starting point [6]. The influence of variables on survival was analyzed using the log-rank test for discrete variables [7] and the proportional hazards model of Cox for continuous variables and multivariate models [8]. A p value of less than or equal to 0.05 was considered significant. Approval for this study was granted by Mayo Foundation’s Institutional Review Board.
Clinical Findings There were 70 patients (44 men and 26 women). Median age was 64 years and ranged from 33 to 83 years. The primary cancer was colon in 44 patients, renal cell in 13, malignant melanoma in 6, endometrial and salivary gland carcinoma in 2 each, and bladder, prostate, and testicular carcinoma in 1 each. Fifty-six patients (80.0%) had their primary cancer resection elsewhere. After resection of the primary cancer and before the development of pulmonary metastases, 11 patients were treated with chemotherapy, 5 with radiation therapy, and 5 with a combination of both. Three patients early in the series were evaluated with standard chest radiography and conventional tomography. Sixty-two patients (88.6%) had a chest CT that showed a solitary nodule in 43 patients, two in 11, and three or more in 8. Ten patients had bilateral nodules. Twelve patients had enlarged mediastinal or regional lymph nodes on CT. Except for these patients, no particular concern about the malignant potential of the thoracic lymph nodes was identified preoperatively to account for the complete lymphadenectomy. Twenty-two patients (31.4%) underwent preoperative transthoracic needle aspiration (TTNA). In 21 (95.4%), the TTNA was positive for malignancy and histologically similar to the primary cancer. Eleven patients had
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
1787
bronchoscopy, and 4 were diagnostic for metastatic carcinoma. Seven patients had undergone extrathoracic metastasectomies before pulmonary resection; all were considered to have had complete resection. The carcinoembryonic antigen was measured in 31 of the 44 patients (70.4%) with colon cancer and was elevated (ⱖ5 ng/mL) in 9. The median interval between resection of the primary cancer and the first pulmonary resection was 34 months (range, 0 to 188 months). Mediastinoscopy was performed in 12 patients, including 5 with clinically enlarged mediastinal lymph nodes; none was positive. At the time of initial pulmonary metastasectomy, a lateral thoracotomy was performed in 64 patients, median sternotomy in 1, and simultaneous left thoracoscopic wedge resection and right thoracotomy in 1. Four patients underwent staged bilateral thoracotomies. In no patient was there concern that a primary lung cancer was being resected rather than a pulmonary metastasis. Intraoperatively, no particular concern of the operating surgeon regarding the malignant potential of the thoracic lymph nodes was identified to account for the complete lymphadenectomy. Forty-six patients had wedge excisions; 18 were multiple. A lobectomy was performed in 16 patients, lobectomy combined with wedge excision in 7, and pneumonectomy in 1. Fifteen patients had a second thoracotomy, 4 a third, and 2 a fourth. Altogether, 142 metastases were resected in these 70 patients, and included 98 wedge excisions, two segmentectomies, 27 lobectomies, and one each bilobectomy, sleeve lobectomy, and pneumonectomy. One patient had an associated chest wall resection. At the initial pulmonary metastasectomy, 48 patients had solitary metastases, 12 had two, 5 had three, 4 had four, and 1 had nine. This was greater than the number of nodules detected preoperatively by chest CT. Lymph node metastases were found in 20 patients (28.6%) and were classified as N1 in 9 patients, N2 in 8, and both in 3. The lymph node dissection was done at the time of the second ipsilateral metastasectomy in 2 patients and during a contralateral metastasectomy after two previous ipsilateral metastasectomies in a third. All 3 of these latter patients were N0 at the time of lymphadenectomy.
Results There were no operative deaths. Complications occurred in 7 patients (10%) and included atrial fibrillation in 2 (2.8%), pneumothorax in 2 (2.8%), and recurrent laryngeal nerve paresis, myocardial infarction, and bronchopleural fistula in 1 each. Hospitalization ranged from 4 to 30 days, with a median of 7 days. Follow-up was complete in all patients and ranged from 1 to 14.6 years, with a median of 6.6 years. Twentyseven patients are currently alive, 5 with evidence of recurrent cancer. Forty-three patients are dead, 34 from recurrent cancer. The median disease-free interval from the time of the extrathoracic primary operation to pulmonary metastasectomy for those patients with metastatically involved lymph nodes was 36 months, as compared to 32 months for those patients without lymph node involvement (p ⫽ 0.64). Throughout the study period,
GENERAL THORACIC
Ann Thorac Surg 2004;77:1786 –91
1788
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
GENERAL THORACIC Fig 1. Estimated survival (death due to any cause) of 70 patients undergoing pulmonary metastasectomy and complete mediastinal lymph node dissection. Zero time on the abscissa is date of first pulmonary metastasectomy and lymph node dissection.
metastatic cancer recurred in 43 patients. Recurrence was limited to the lungs in 15 patients, extrathoracic sites only in 9, and both in 19. Tumor recurrence was significantly higher in the patients with metastatically involved lymph nodes (p ⫽ 0.004). The 3-year recurrence free rate was 23% (95% CI, 8% to 66%) in patients with metastatically involved lymph nodes and 49% (95% CI, 36% to 66%) in patients without metastatically positive lymph nodes. Overall, estimated 5- and 10-year survival was 48% (95% CI, 36% to 60%) and 29% (95% CI, 16% to 42%), respectively (Fig 1). Three-year survival for patients without lymph node metastases was 69% (95% CI, 58% to 83%) as compared to only 38% (95% CI, 20% to 68%) for those with metastases (p ⬍ 0.001) (Fig 2). Survival was not improved when only N1 lymph nodes were metastatically involved. Moreover, only 2 patients with metastatically involved lymph nodes survived beyond 4 years. Both of these patients had low-grade renal cell carcinomas.
Fig 2. Estimated survival of 70 patients undergoing pulmonary metastasectomy and complete mediastinal lymphadenectomy without metastatic nodal involvement (negative) and with nodal involvement (positive). Zero time on the abscissa is date of first pulmonary metastasectomy and lymph node dissection. (p ⬍ 0.001).
Ann Thorac Surg 2004;77:1786 –91
The number of pulmonary metastases was not associated with a difference in survival. Three-year survival for patients with a solitary metastasis was 61% (95% CI, 49% to 78%), as compared with 57% (95% CI, 39% to 82%) for those with multiple metastases (p ⫽ 0.36). No difference in lymph node metastases was observed in patients with a solitary metastasis versus multiple pulmonary metastases (p ⫽ 0.11). During the follow-up period, 19 of 48 patients (39.6%) with a solitary metastasis and 15 of 22 (68.2%) with two or more pulmonary metastases died due to recurrent cancer. Age, gender, cell type of the primary tumor, and the extent of resection did not influence survival. Fifteen patients underwent a second thoracotomy for recurrent pulmonary metastases; 9 had a right thoracotomy, 6 a left, and 1 had an associated chest wall resection. At the time of second thoracotomy, 9 patients had a solitary metastasis and 6 had two or more. In 2 patients, both with low-grade metastatic renal cell carcinoma, metastatic lymph node disease was found at the time of the second thoracotomy. Survival in these 15 patients after their second pulmonary metastasectomy was 82% (95% CI, 59% to 100%) at 5 years. Four patients had a third pulmonary resection for recurrent cancer and 2 patients had a fourth. Three of these 4 patients are still alive 2 to 7 years after their last pulmonary metastasectomy. None has evidence of recurrent disease.
Comment Controversy exists as to the frequency with which pulmonary metastases can metastasize to regional lymph nodes. An autopsy series by Abrams and colleagues [9] demonstrated a 33% incidence of mediastinal lymph node metastases in patients with nonpulmonary carcinoma. In 1965, Thomford and colleagues [10] reviewed our experience with pulmonary metastasectomy when they reported 221 pulmonary resections in 205 patients over a 21-year period. Twenty patients (9.8%) had metastatically involved lobar or hilar lymph nodes. Five-year survival was 30.3%, but the prognostic significance of lymph node involvement was not analyzed. A report from the International Registry of Lung Metastases described a 5% incidence of lymph node metastases in 5,206 patients undergoing pulmonary metastasectomy [1]. Of note, however, is that lymph nodes were resected or sampled in only 4.6% of these patients. Putnam [3] suggested that mediastinal lymph node metastases rarely occur from pulmonary metastases; however, his report deals exclusively with metastatic sarcoma. We agree that nodal metastases are much less common with sarcomatous pulmonary metastases, and this is why patients with metastatic sarcoma were not included in this report. In contrast, Martini and McCormack [2] concluded that pulmonary carcinoma metastases are likely to metastasize to regional lymph nodes, especially melanomas. They fail, however, to mention the prognostic significance of this event and did not discuss mediastinal lymph node dissection as part of pulmonary metastasectomy. Our recent experience [11, 12] suggests that metastatic lymph node disease at the time of pulmonary
metastasectomy for carcinoma is present in 8.6% to12.5% of patients. However, until recently, routine lymph node dissection has not been uniformly performed in our practice, as shown in this study by the low rate of lymphadenectomies (12.1%) in our patients undergoing pulmonary metastasectomy. Whereas many of our patients having pulmonary metastasectomy had lymph nodes identified on the final pathology report, it was only these 107 patients who underwent complete mediastinal lymphadenectomy. Except for the 12 patients with mediastinal or regional lymphadenopathy identified on the preoperative chest CT, no particular concern about the mediastinal lymph nodes either preoperatively or intraoperatively was identified to account for the complete lymphadenectomy. Of these 12 patients, 5 underwent mediastinoscopy and all were negative for metastatically involved lymph nodes. Ultimately, only 2 of these 12 patients had lymph node metastasis. To date, few reports have addressed the prognostic significance of metastatically involved lymph nodes with pulmonary metastasectomy. The only factor affecting survival in our series was the presence of lymph node metastasis. Kamiyoshihara and colleagues [13] in a retrospective review of 28 patients undergoing pulmonary metastasectomy found no detectable difference between patients with and without positive regional lymph nodes. Unlike other series, these authors reported no long-term survivors in patients undergoing partial resection (ie, less than lobectomy). More recently, Loehe and colleagues [14] have attempted to clarify the importance of systematic lymph node dissection during pulmonary metastasectomy. In a group of 63 patients, they found a 14.3% incidence of mediastinal lymph node metastases. Similar to their study, we found that the number of pulmonary metastases was not associated with a difference in survival. Also similar was the trend toward an increasing incidence of metastatic lymph node involvement with an increasing number of metastases. In contrast, Loehe and colleagues [14] failed to show a difference in survival in patients with metastatically involved mediastinal lymph nodes (p ⫽ 0.37). Our study demonstrates that patients undergoing pulmonary metastasectomy who are found to have metastatically involved regional lymph nodes have an increased incidence of cancer recurrence and decreased survival. We believe this observation supports systematic lymph node dissection at the time of pulmonary metastasectomy for metastatic carcinoma. Several questions regarding the indications and potential therapeutic role of mediastinal lymphadenectomy or sampling remain [15]. Unanswered is whether regional lymph node dissection at the time of pulmonary metastasectomy contributes to improved survival, or is it simply a prognostic indicator perhaps suggesting the need for adjuvant therapy. Does the number of pulmonary
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
1789
metastases influence the need for lymph node dissection? Accepting the role for lymph node dissection, does this further cloud the role for utilizing video-assisted thoracic surgery in pulmonary metastasectomy? These questions hopefully will be answered in future prospective clinical trials. In conclusion, surgical resection of pulmonary metastases should be offered to patients whose primary tumor is controlled and are good operative risks. All pulmonary and extrathoracic metastases must be amenable to complete surgical resection for a survival benefit to be achieved. Our findings suggest that the presence of metastatic disease in regional lymph nodes has an adverse effect on survival in those patients undergoing curative pulmonary metastasectomy. Therefore, we recommend complete lymph nodes dissection in these patients to better define their prognosis and potentially guide adjuvant therapy.
References 1. Pastorino U, Buyse M, Friedel G, et al. Long-term results of lung metastasectomy: prognostic analysis based on 5206 cases. J Thorac Cardiovasc Surg 1997;113:37–49. 2. Martini N, McCormack PM. Evolution of the surgical management of pulmonary metastases. Chest Surg Clin North Am 1998;8:13–27. 3. Putnam JB. Secondary tumors of the lung. In: Shields TW, LoCicero J, Ponn RB, eds. General thoracic surgery, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2000:1555–76. 4. LoCicero J, Ponn RB, Daly BD. Surgical treatment of nonsmall cell lung cancer. In: Shields TW, LoCicero J, Ponn RB, eds. General thoracic surgery, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2000:1311–41. 5. Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997;111:1710 –7. 6. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–81. 7. Peto R, Peto J. Asymptotically efficient rank andinvariant procedures. J R Statist Soc, Series A 1972;135:185–201. 8. Cox DR. Regression models and life-tables (with discussion). J R Statist Soc, Series B 1972;34:187–220. 9. Abrams HL, Spiro R, Goldstein N. Metastases in carcinoma: analysis of 1000 autopsied cases. Cancer 1950;3:75–85. 10. Thomford NR, Woolner LB, Clagett OT. The surgical treatment of metastatic tumors in the lungs. J Thorac Cardiovasc Surg 1965;49:357–69. 11. McAfee MK, Allen MS, Trastek VF, et al. Colorectal lung metastases: results of surgical excision. Ann Thor Surg 1992;53:780 –6. 12. Cerfolio RJ, Allen MS, Deschamps C, et al. Pulmonary resection of metastatic renal cell carcinoma. Ann Thorac Surg 1994;57:339 –44. 13. 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. 14. Loehe F, Kobinger S, Hatz RA, et al. Value of systemic mediastinal lymph node dissection during pulmonary metastasectomy. Ann Thorac Surg 2001;72:225–9. 15. Goldstraw P. Invited commentary. Ann Thorac Surg 2001;72: 229.
GENERAL THORACIC
Ann Thorac Surg 2004;77:1786 –91
1790
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
Ann Thorac Surg 2004;77:1786 –91
DISCUSSION GENERAL THORACIC
DR HENNING A. GAISSERT (Boston, MA): Only a minority of patients with pulmonary metastatic carcinoma are candidates for surgical resection, and only a minority of those undergoing resection achieve long-term survival. Clearly, we are in need of improving patient selection. Doctor Ercan and associates deserve recognition for reminding us of the metastatic potential of pulmonary metastasis. I thank the authors for providing me with a copy of the manuscript in advance of the meeting. Their investigation originates from an interest for the prognostic significance of lymph node metastasis in patients with resectable pulmonary metastasis. The analysis is performed on a smaller select group of patients among them, and a difference in survival is found. The question arises whether the finding of metastatic nodes in 29% of the select group can be used to make statements about the total group undergoing lung resection. We do not know what criteria led to lymph node dissection. Although long-term survival is comparable with other studies, the node dissection group may have had more advanced disease. The rate of lymph node metastasis, for example, was twice that of the recent prospective study from the University of Munich quoted in your manuscript. Your conclusions should therefore not necessarily be generalized to all patients with resectable pulmonary metastasis. I would like to ask you whether the pattern of recurrence in patients with lymph node metastasis was different from those without. In your manuscript, you did not break down the numbers dependent on its presence or absence. If 29% of all patients with resectable pulmonary metastasis had lymph node disease, one might investigate lymph node spread preoperatively, for example, by mediastinoscopy. If that number drops to 14%, as suggested by the Munich group, one may not stage every patient. Has the knowledge gained from this study changed your practice? And if you now saw a patient with a solitary peripheral pulmonary metastasis and a metastatic hilar lymph node, would you be concerned enough about the lymphatic drainage bed to perform a lobectomy rather than a wedge resection? I enjoyed your presentation and I thank the Society for the opportunity to discuss this paper. DR ERCAN: Thank you, Dr Gaissert, for discussing the paper. We feel that we can generalize our findings to patients with pulmonary metastases from extrapulmonary carcinomas. We realize that only 12% of our nearly 400 metastasectomy patients had complete lymph node dissections. This is likely due to the fact that early in our pulmonary metastasectomy experience lymphadenectomy was not routinely performed. In the literature from autopsy studies, intrathoracic lymph node metastases have been found in 30% of patients with nonpulmonary carcinoma. Our finding of 28.6% is remarkably close to this. The same is not true for patients with sarcoma. For patients with pulmonary metastases where the primary tumor was a carcinoma, we now routinely perform a mediastinal lymphadenectomy. As for your other question, the pattern of recurrence was differentiated as either being intrathoracic or extrathoracic. For the patients who did recur, most of them did not undergo re-resection as they recurred systemically. The recurrences were sometimes only identified radiographically but were obvious. We believe we have shown that the recurrence rate was statistically higher in patients who had lymph node involvement identified during the pulmonary metastasectomy.
As for the patient who had a solitary metastasis with hilar lymph node involvement, the ultimate goal during pulmonary metastasectomy is to remove all obvious disease while at the same time conserve normal lung. Lobectomy would not be our standard approach; rather it would be wedge resection and complete lymphadenectomy. Sometimes, however, the metastasis is located in such a position that resection requires a lobectomy. We do in select patients occasionally perform lobectomy for the resection of pulmonary metastasis. Based on the poor prognosis in patients found to have metastatically involved lymph nodes, our approach to those patients might change in terms of consideration for future resections or the need for adjuvant therapy. DR DOUGLAS E. WOOD (Seattle, WA): I like your paper, Dr Ercan, and it was valuable up until the conclusions. I have to say I disagree completely with where you went from your data to the conclusions. If we look at the percent of patients who had lymph node metastases out of your total cohort, it is actually less than 3%, and I would be hard pressed to recommend that we add a significant surgical procedure for 3% of patients. It is only 20 patients out of 833 patients, not 20 patients out of 70. You selected those patients, I am sure, on the basis of enlarged lymph nodes at the time of surgery or on preoperative staging. The question is, do you think that these are lymph node metastases from the metastases or are these synchronous metastases to the mediastinum along with the pulmonary metastases, implying just a more aggressive biologic behavior? Do you actually think that mediastinal lymphadenectomy confers any therapeutic benefit? DR ERCAN: Thank you, Dr Wood. Unfortunately, of the 883 patients only 107 underwent documented complete lymphadenectomy. Thirty-seven patients were eliminated for a variety of reasons including an inadequate primary operation, persistent or unresectable extrathoracic disease, and unresectable intrathoracic disease. The remaining 70 patients were not selected because of preoperative or intraoperative findings. For patients with worrisome mediastinal nodes preoperatively, mediastinoscopy was performed. None of the operative records showed any concern about the intrathoracic lymph nodes. It is true that one of the surgeons in the study believed more than others in thoracic lymphadenectomy for these types of patients. This fact likely accounts for the overall low number of complete lymph node dissections. According to the outcome, once these patients had lymph node involvement, they clearly showed a worse survival. We cannot be sure that doing the mediastinal dissection added any positive benefit to the outcome. However, the patients were better staged. Given the poor prognosis of patients with metastatically involved lymph nodes, consideration might be given to adjuvant therapy. Ultimately, this might only be answered with larger prospective and randomized studies. DR STEVEN R. DEMEESTER (Los Angeles, CA): I wondered if you are using just histology to look for metastases in the lymph nodes or if you are using immunohistochemistry, and if you are not using it, what higher percentage you might find with immunohistochemistry staining of the lymph nodes, and what that might impart in terms of prognosis for the patients?
ERCAN ET AL LYMPH NODE METASTASIS AT METASTASECTOMY
1791
DR ERCAN: Thank you very much and that is a good point. During the time of this study, immunohistochemical evaluation was not routine. That might be something to look at in the future.
that if you analyzed survival in metastatic renal cell carcinoma or melanoma you would have achieved anywhere near this 3-year survival rate. Were there enough non– colon cancer patients to perform this analysis?
DR RENEE S. HARTZ (New Orleans, LA): I believe my question is consistent with those of Drs DeMeester and Wood in that on one of your slides you showed a 70% survival rate at 3 years after simple metastectomy. This is an astonishing figure, and you did not elaborate on whether or not your univariate and multivariate analyses were performed on the basis of primary tumor cell type. Can this high survival rate be explained by the fact that most of your patients had colonic primaries? It seems unlikely
DR ERCAN: Thank you very much. As you said, the majority of the patients had colonic primaries, and 26 patients had a variety of other tumor types (renal cell cancer in 13 patients and melanoma in 6). Cell type of the primary did not influence survival. Interestingly, there were only 2 patients with metastatically involved lymph nodes who survived beyond 4 years. Both of these patients had low-grade renal cell carcinomas.
GENERAL THORACIC
Ann Thorac Surg 2004;77:1786 –91