| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2006;81:2008-2013
© 2006 The Society of Thoracic Surgeons
a Department of Thoracic Surgery, Scientific Institute H San Raffaele, Milan, Italy
b Department of Pathology, Scientific Institute H San Raffaele, Milan, Italy
Accepted for publication January 3, 2006.
* Address correspondence to Dr Melloni, Unità Operativa di Chirurgia Toracica, Ospedale San Raffaele, Via Olgettina 60, Milan 20132, Italy (Email: giulio.melloni{at}hsr.it).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
METHODS: We identified 81 patients who underwent surgical resection between 1991 and 2004. The microsatellite instability was determined by immunohistochemical evaluation of MSH2 and MLH1 in 117 lesions (41 primary tumors and 76 lung metastases).
RESULTS: Overall 3-, 5-, and 10-year survival rates were 50%, 42%, and 30%, respectively. Univariate analysis showed that stage of the primary tumor (p = 0.037), radicalness of the resection (p = 0.019), and stratification into groups according to the International Registry of Lung Metastases classification (p = 0.039) were prognostic factors. Multivariate analysis showed that stage of the primary tumor (p = 0.030) and the radicalness of the resection (p = 0.014) were independent prognostic factors. All tumors displayed preserved expression of MSH2 and MLH1 and were considered microsatellite stable lesions.
CONCLUSIONS: Pulmonary resection of metastases from colorectal carcinoma results in long-term survival in selected patients. Complete resection, stage of the primary tumor and stratification into groups according to the International Registry of Lung Metastases classification were prognostic factors. All the metastases and the corresponding primary tumors were microsatellite stable lesions. This finding seems to demonstrate that pulmonary metastases are infrequent in colorectal carcinomas with microsatellite instability.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
In recent years, several series have sought to define prognostic factors in patients with metastases from colorectal cancer. The identification of prognostic factors is particularly important in patients with advanced colorectal cancer, a disease whose clinical behavior and biology are poorly understood. Although completeness of resection is consistently reported to affect survival [3], the prognostic significance of other variables such as number of pulmonary metastases, prethoracotomy carcinoembryonic antigen (CEA) serum level, and presence of lymph node metastasis remains controversial [1–6, 8]. In addition, the value of repeated pulmonary metastasectomies and staged resections of colorectal cancer metastases to both liver and lung has not been completely evaluated [9]. Therefore, these traditional prognostic parameters cannot provide absolute criteria to predict survival after pulmonary resection for metastases from colorectal cancer. For this reason, research efforts have been directed toward studying the biological characteristics of colorectal cancer lung metastases [2].
Microsatellite instability (MSI) is a genetic alteration defined by the change of length of repetitive sequences of the microsatellite DNA from the tumor, compared with that of reference tissue from the same subject. Microsatellite instability is caused by defects in mismatch repair genes. The most common affected genes are hMLH1 and MSH2. A percentage ranging from 15% to 20% of colorectal cancers is a MSI tumor whereas the remaining colorectal cancers lacking MSI are called microsatellite stable (MSS) tumors [11, 12]. Studies have shown that patients with primary colorectal cancers displaying MSI have a survival advantage over those displaying MSS, and some authors have therefore posited the presence of biological differences between these two types of tumors [11–13]. However, at present, MSI status has been extensively studied only in colorectal cancers and in other solid tumors, but never systematically in pulmonary metastases. The determination of MSI status of pulmonary metastases from colorectal cancer might be important for a better biological characterization of this advanced neoplastic disease.
In this study, we analyze our experience with pulmonary resection for metastases from colorectal cancer. The aims were to search for factors influencing prognosis and to determine the MSI status both in the primary tumors and in the corresponding pulmonary metastases.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
36 months and single metastasis); group 2, resectable, one risk factor (DFI < 36 months or multiple metastases); group 3, resectable, two risk factors (DFI < 36 months and multiple metastases); group 4, unresectable. Operative mortality was defined as any death during hospitalization or within 30 days from surgery. Late mortality was defined as any subsequent death. Follow-up data were obtained from the patients' hospital visit or telephone interview.
Immunohistochemistry
Immunohistochemical studies were performed on 4-µm thick formalin-fixed paraffin-embedded tissue sections using an indirect immunoperoxidase method for MLH1 and MSH2 (non-biotin detection system; Biogenex, San Ramon, California). In all cases, sections comprised an area of normal tissue adjacent to the tumor. Representative sections from each case were incubated with monoclonal antibodies against MLH1 (G168-728; Pharmingen, San Diego California) and MSH2 (FE11; Calbiochem, Cambridge, Massachusetts).
Two pathologists (C.D. and G.A.) evaluated the lesions independently. Lymphocytes and normal bronchial and colonic epithelium adjacent to tumor served as built-in positive controls. Neoplastic lesions were read as positive when nuclear staining in tumor tissue was present (MSS tumors) or as negative when the nuclear stain was absent (MSI tumors). Because we identified only MSS tumors (see Results), two MSI colorectal cancers (1 MLH1 negative and 1 MSH2 negative) previously characterized by one of the authors [16] were run concurrently to test the reliability of the immunohistochemical techniques.
Statistical Analysis
Survival plots were reconstructed according to the Kaplan-Meier method. Simple Cox regression analysis was used to estimate the odds ratios for the variables evaluated at baseline. Survival rates of patients grouped according to selected variables were compared by log-rank statistics. On the basis of the univariate analysis, multivariable Cox regression analyses were performed to adjust for potential confounders using a forward stepwise method to introduce significant covariates into the model. Reported p values are two-sided. Results of analyses were considered significant at a level of p less than 0.05. All confidence intervals were calculated at the 95% level.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
A total of 88 operations were performed in the 81 patients. Eighty-two operations involved a one-side thoracotomy for single or multiple unilateral metastases. Six operations involved a sternotomy or sequential thoracotomies for bilateral synchronous metastases. Of the 88 operations, 66 were carried out through muscle-sparing thoracotomy, 20 through standard posterolateral thoracotomy, and 2 through median sternotomy. The type of resection included 61 wedge resections, 26 lobectomies, and 1 pneumonectomy. Six patients of 81 (7%) underwent redo operations for recurrent pulmonary metastases. Median DFI was 30 months (range, 1 to 118). Systematic mediastinal lymph node dissection was performed concurrently with all 27 major resections (lobectomies and pneumonectomies). Lymph node sampling was added to the last 12 wedge resections but had not been included in the previous 49 resections.
Mediastinal lymph node metastases were found in 4 patients (10%), whereas the remaining 35 patients showed no evidence of lymph node metastases. Resection of the pulmonary metastases was complete in 74 patients (91%). The remaining 7 patients (9%) underwent incomplete resection because of diffuse metastatic infiltration of the entire lung undetected by the preoperative CT scan (3 patients) and presence of mediastinal lymph node metastases (4 patients). On the basis of pathologic assessment, 44 patients had a single metastasis and 37 had two or more. Median size of the resected pulmonary metastases was 2 cm (range, 1 to 13 cm). The IRLM stage of the patients was 1 in 21 patients (26%), 2 in 38 (47%), 3 in 15 (18%), and 4 in 7 (9%). The MSI status was determined by immunohistochemistry both in the primary colorectal cancer and in the corresponding metastases in 41 patients and in an additional 35 pulmonary metastases for which the primary tumor material was not available at our hospital. All of these 117 tumors (76 metastases and 41 primary tumors) were considered MSS lesions (normal expression of both MSH2 and MLH1), and no tumor was considered MSI (total lack of expression of MSH2 or MLH1, or both).
Sixty-one patients (75%) underwent adjuvant 5-fluorouracil–based chemotherapy. There was no operative mortality. Complications occurred in 7 patients (9%) and included atrial fibrillation in 3, persistent air leak in 3, and digestive hemorrhage in 1. Median hospital stay was 7 days (range, 3 to 25). The follow-up was complete for all 81 patients. The median follow-up was 20 months (range, 4 to 154). Overall 3-, 5-, and 10-year survival rates were 50%, 42%, and 30%, respectively (Fig 1), and the median survival was 37 months. Thirty-six patients (43%) are currently alive with no evidence of recurrence, 11 are alive with evidence of disease, 4 died of unrelated causes, and 30 relapsed and died.
|
|
|
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
In this series, the overall 5-year survival rate was 42%, which can be positively compared with data in the literature, where it ranges from 27% to 61% [1–10]. Our experience, therefore, confirms the effectiveness of surgically treating pulmonary metastases from colorectal cancer in that it provides a significant advantage in terms of survival. The 5-year survival of patients with untreated advanced colorectal cancer is in fact lower than 5%. Chemotherapy only rarely achieves long-term survival in advanced colorectal cancer patients [1, 2]. The efficacy of new types of treatment, such as percutaneous radiofrequency ablation, has still to be fully evaluated. Therefore, surgical resection is at present to be considered the treatment of choice for local control of pulmonary metastases from colorectal cancer.
Surgery of metastases is based on the principle of performing a complete resection while preserving as much healthy pulmonary tissue as possible. The majority of authors believe an open procedure allowing for the manual palpation of the lung to be necessary [5, 6]. Nevertheless, the availability of increasingly advanced CT leads some surgeons to use video-assisted thoracoscopic surgery in the treatment of lung metastases [8]. At present, however, there are no studies on the treatment of lung metastases comparing open surgery with video-assisted thoracoscopic surgery. Our present strategy is to prefer the use of open procedures such as muscle-sparing thoracotomies while limiting the use of video-assisted thoracoscopic surgery to pulmonary biopsies. The need for intraoperative manual palpation is stressed by recent studies showing that it is still indispensable, despite the use of helical CT, when the purpose is to perform a complete resection [17].
The role of systematic lymph node dissection associated with lung resection in the treatment of pulmonary metastases from colorectal cancer has not been established as yet. Strategies used in clinical practice are diversified: no lymph node dissection, systematic lymph node dissection, lymph node sampling, and biopsy only of volume increased lymph nodes. In the literature, the reported incidence of mediastinal lymph node metastases is between 14% [18] and 19% [5]. In our experience, we detected mediastinal lymph node metastases in 10% of patients (4 of 41). Therefore, it is likely that a significant percentage of metastasectomies performed without lymph node dissection or sampling and considered complete resections are actually incomplete resections. The presence of involved lymph nodes strongly influences the ensuing therapeutical strategy, suggesting the need for adjuvant chemotherapy. On the basis of the above considerations, we currently perform lymph node dissection in all major resections and lymph node sampling in all sublobar resections.
In our study, statistical analysis identified complete resection, colon tumor stage, and IRLM stage as prognostic factors.
Several series [1–10], as well as our experience, indicate the need to perform complete resections in the treatment of pulmonary metastases from colorectal cancer. In this study, 9% of resections were considered incomplete because of the presence of mediastinal lymph node metastases or a diffuse metastatic infiltration of the entire lung. Therefore, the inability to detect both mediastinal metastases and pulmonary neoplastic infiltrations represented limits in the preoperative work-up.
Few studies in the literature assess primary tumor features as a prognostic factor. In the series of Vogelsang and coworkers [3], primary tumor stage and lymph node involvement were survival predictors after lung metastasectomy. In our experience, patients with a T1/T2 primary tumor showed a significantly longer survival as compared with patients with a T3/T4 tumor. Other studies [10, 19] have, however, failed to demonstrate the impact of primary tumor features on survival.
As reported by other authors [5, 20], our experience also indicates that DFI and the number of metastases did not influence prognosis on an individual basis. On the other hand, the association of the above two factors with lesion resectability, based on the system of prognostic grouping proposed by the IRLM, was able to identify significant differences in the survival of patients who underwent resection owing to lung metastases from colorectal cancer (5-year survival was 45%, group 1; 47%, group 2; 32%, group 3; and 0%, group 4). Therefore, according to our experience, adequate candidates for surgery would be not only those showing resectable disease, DFI 36 months or longer, and single metastasis but also those showing resectable disease and DFI less than 36 months and/or multiple metastases.
The second aim of our study was to investigate the presence of MSI both in the primary tumors and in the lung metastases for a better biological characterization of advanced colorectal cancer. A better knowledge of the genetic alterations characterizing pulmonary metastases from colorectal cancer is important, not only to clarify the events leading to the acquisition of the metastatic phenotype but also to predict patients' differential responses to treatment and clinical outcomes. We analyzed 117 lesions by means of immunohistochemical evaluation. In 41 patients, both the primary colorectal cancer and the corresponding metastasis were studied, whereas in the remaining 35 patients, who underwent colon resection in other hospitals, only pulmonary metastases were assessed. This analysis was carried out since MSI has been extensively studied only in primary colorectal cancers, but never systematically in pulmonary metastases.
Microsatellite instability is a well-known genetic alteration, first identified in colorectal cancer [12] and subsequently recognized in nearly all solid tumors. In colorectal cancer, MSI has been reported with a frequency ranging from 15% to 20% [11, 12]. Several studies have shown that primary colorectal cancers with MSI have distinct clinicopathologic features distinguishing them from those with MSS. Microsatellite instability colorectal cancers are more frequently located in the proximal colon and are associated with a high prevalence of mucinous and medullary histologies and high numbers of tumor-infiltrating lymphocytes [12, 21]. The gold standard for MSI determination is the polymerase chain reaction test [22, 23]. However, several studies have demonstrated the high correlation between MSI phenotype determined by polymerase chain reaction and complete absence of MSH2 or MLH1 expression using immunohistochemical analysis [22, 23]. Other studies have attempted to assess the impact of MSI on clinical outcome without, however, producing conclusive results. Nevertheless the most recent and largest analyses suggest that primary colorectal cancers with MSI have a significant, stage-independent, multivariate survival advantage over those with MSS [11–13, 21]. Moreover, in vitro and clinical studies indicate that the MSI status of a patient's colorectal cancer may indicate differences in 5-fluorouracil–based chemosensitivity [12, 21]. These findings might be a result of intrinsic biological differences between MSI tumors and MSS tumors.
In our experience, both primary tumor and the metastases had the same MSS phenotype. Although the number of cases in our series is limited, some considerations may be made: (1) the detection of a MSS phenotype in all lung metastases suggests that metastases are infrequent in the colorectal cancers with MSI; (2) since the MSS status of a primary colorectal cancer remains unchanged in its corresponding metastasis, it is likely that MSI does not play a direct role in metastatic process; (3) it is likely that the detection of a MSS phenotype in all lung metastases represents an indirect confirmation of the greater biological aggressiveness of the primary colorectal cancers displaying MSS compared with those displaying MSI.
According to the published data, primary colorectal cancers with MSS have a more aggressive behavior because they are less immunogenic than MSI tumors [26]. Colorectal cancers with MSI present a higher number of tumor-infiltrating lymphocytes if compared with MSS tumors. Tumor-infiltrating lymphocytes in primary colorectal cancers with MSI are also activated and cytotoxic [16, 24]. There is also evidence in the literature that colorectal cancers with MSI produce a host immune response to malignant cells. It is attractive to think that such a reaction is responsible for the better prognosis in patients with this type of tumors [[23, 25].
In conclusion, surgical treatment of pulmonary metastases from colorectal cancer offers a significant advantage in terms of survival in a selected group of patients. Complete resection, some features of the primary tumor, and stratification into groups according to IRLM classification were prognostic factors. All metastases and the corresponding primary tumors studied by immunohistochemical analysis showed a MSS phenotype. This finding seems to demonstrate that pulmonary metastases are infrequent in colorectal cancers with MSI phenotype and shows that MSS status is not modified in the metastatic process. Further studies are necessary, however, to define the clinical significance of these initial results. Evidence strongly suggests that a better understanding of the biology of such lesions will shed new light on the treatment of pulmonary metastases from colorectal cancer.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Pfannschmidt, H. Dienemann, and H. Hoffmann Surgical Resection of Pulmonary Metastases From Colorectal Cancer: A Systematic Review of Published Series Ann. Thorac. Surg., July 1, 2007; 84(1): 324 - 338. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
