Prediction of Prognosis and Surgical Indications for Pulmonary Metastasectomy From Colorectal Cancer

GENERAL THORACIC

Prediction of Prognosis and Surgical Indications for Pulmonary Metastasectomy From Colorectal Cancer Toshihiko Iizasa, MD, Makoto Suzuki, MD, Shigetoshi Yoshida, MD, Shinichiro Motohashi, MD, Kazuhiro Yasufuku, MD, Akira Iyoda, MD, Kiyoshi Shibuya, MD, Kenzo Hiroshima, MD, Yukio Nakatani, MD, and Takehiko Fujisawa, MD Departments of Thoracic Surgery and Diagnostic Pathology, Graduate School of Medicine, Chiba University, Chiba, Japan

Background. Treatment of pulmonary metastases from colorectal cancer by excision has increased rapidly, but reports on indications and prognostic factors are inconsistent. We sought to identify poor prognostic factors preoperatively and to retrospectively evaluate preoperative clinical indications for surgery. Methods. A total of 75 patients with colorectal cancer had pulmonary metastases excised from 1986 to 2003. Tumor size, number, laterality, hilar or mediastinal lymphadenopathy, and carcinoembryonic antigen level were possible risk factors for metastatic tumors, with primary site of colorectal tumor, disease-free interval, and hepatectomy for liver metastasis possible risk factors for primary tumors. Prognostic factors in univariate and multivariate analyses also included age and sex. Results. Five-year survival rates were 41.3% after pulmonary excision and 73.1% after primary colorectal resection. Three factors identified as significant by univariate log-rank test for overall survival after pulmonary resection were carcinoembryonic antigen (p < 0.0001), tumor

laterality (p ⴝ 0.0205), and number of pulmonary metastases (p ⴝ 0.0028). Multivariate analysis found that carcinoembryonic antigen, tumor number, tumor size, and patient’s age were also independent prognostic factors. In contrast, carcinoembryonic antigen, number of metastases, and disease-free interval predicted prognosis after primary colorectal resection. Prior hepatectomy for metastases did not influence prognosis after pulmonary metastasectomy. Conclusions. Elevated carcinoembryonic antigen level and multiple metastases are preoperative predictors of poor prognosis after resection of pulmonary metastases from colorectal cancer. Survival rate is sufficient to justify pulmonary metastasectomy if there is no local or distant metastatic lesion other than in the liver; if needed, sequential pulmonary and hepatic metastasectomy can be performed.

C

metastasectomy from colorectal cancer [3]. Since then, the number of pulmonary metastasectomy procedures for appropriate patients has increased rapidly, and indications and prognosis factors for excision have been reviewed. We have recently encountered some retrospective studies of prognostic factors for lung metastasectomy from colorectal cancer using univariate and multivariate evaluation [4 –11]. However, the results are not entirely consistent with each other, suggesting the need to evaluate more cases from many facilities to establish surgical indications for these patients. We selected 10 factors that appear to significantly influence prognosis for metastasectomy of lung tumors from colorectal cancer. These include age, sex, lung tumor size, number, laterality, hilar or mediastinal lymphadenopathy, and preoperative carcinoembryonic antigen (CEA) level for metastatic tumors; and primary site of colorectal cancer, disease-free interval, and hepatectomy for liver metastasis for primary tumors. The present retrospective study is an attempt to clarify prognostic factors that can be defined preoperatively and relate them to long-term survival in patients who have

olorectal cancer and primary lung cancer are among the most common malignancies in industrialized countries [1, 2]. The typical pattern of lung metastasis from colorectal cancer is single or multiple nodules rather than miliary tumors or lymphangitis carcinomatosa. No effective chemotherapy regimen has been found for metastatic disease. Hence, a surgical procedure to eliminate pulmonary metastases is generally accepted as the only potentially curative treatment. In favor of surgery is the recent trend toward earlier detection of pulmonary metastases as small peripheral densities with the increasingly common use of screening with spiral or high-resolution computed tomography (CT). The number of possibly resectable cases has increased in the last 20 years as physicians who monitor patients after primary surgery have paid much more attention to the possibility of identifying operable cases of metastatic disease. In 1996, we released a report with 29 cases of pulmonary Accepted for publication Feb 9, 2006. Address correspondence to Dr Fujisawa, Department of Thoracic Surgery, Graduate School of Medicine, Chiba University, 1-8-1 Inohana, Chuo-ku, Chiba 260-8670, Japan; e-mail: [email protected].

© 2006 by The Society of Thoracic Surgeons Published by Elsevier Inc

(Ann Thorac Surg 2006;82:254 – 60) © 2006 by The Society of Thoracic Surgeons

0003-4975/06/$32.00 doi:10.1016/j.athoracsur.2006.02.027

IIZASA ET AL PULMONARY METASTASECTOMY AND COLORECTAL CANCER

undergone pulmonary metastasectomy from colorectal cancer. We evaluate appropriateness of surgical indications and disadvantages for patients with metastatic lung tumors with homogeneous data from a single general thoracic surgery unit.

Material and Methods Patients A total of 75 patients with primary colorectal cancer, including 38 colon cancers and 37 rectal cancers, had pulmonary metastases excised from 1986 to 2003 at the Department of Thoracic Surgery, Chiba University Hospital (Chiba, Japan); patients were followed up through the end of 2004. Our criteria for resection of pulmonary metastases from colorectal carcinoma included unilateral or bilateral excisable lung lesions per preoperative chest radiography, no local recurrence of primary lesions, and no extrapulmonary lesions with the exception of associated prior or simultaneous resectable hepatic metastases. This study includes 29 cases reported previously [3]. Of the 75 patients with colorectal cancer, 46 were women and 29 were men, with an age range of 40 to 82 years (mean age with standard deviation, 60.4 ⫾ 8.8 years). The advanced age group was defined as those patients 60 years of age or older. Preoperative serum concentration of CEA was measured by microparticle enzyme immunoassay (normal upper limit, 5 ng/mL). We decided on surgical procedure according to pulmonary tumor location, size (mean with standard deviation, 2.9 ⫾ 1.6 cm), and number (single, 53 cases; double, 11 cases; triple, 2 cases; more than 3, 9 cases), performing 40 lobectomies, 29 wedge resections, 5 segmentectomies, and 1 pneumonectomy without sternotomy. Because metastatic nodules prefer to grow in the posterior segments, median sternotomy is not appropriate to check the posterior segment, pleural dissemination, and effusion. Lymph node sampling of hilar or mediastinal nodes was performed in 44 patients. The intervals between the liver resection and the lung resection were within 1 year for 9 cases and more than 1 year for 11 cases. Follow-up was every 2 to 3 months for the first year after surgery, then every 3 to 6 months for 2 to 5 additional years, and once a year thereafter. Additional follow-up and survival data were determined by telephone or postal contact with the patient. When distant or local disease recurrence developed, any treatment was permitted. The study protocol was reviewed and approved by the ethics committee for human studies at the Graduate School of Medicine in Chiba University on September 30, 2005. Written informed consent was not obtained or required from all subjects because they are not identified in this retrospective study.

Survival Rate and Statistical Analysis The end point was overall survival calculated as the interval between date of primary or pulmonary surgery and death. Cause of death was confirmed by telephone contact with the doctor who followed the patient. Overall

255

survival was calculated by the Kaplan–Meier method [12]. The difference between survival curves was analyzed using the log-rank test for univariate analyses. Clinical prognostic factors examined in this study included age, sex, preoperative CEA level, tumor laterality, tumor number, tumor size, preoperative lymphadenopathy, primary site of colorectal cancer, disease-free interval, and hepatectomy for liver metastasis. A hilar or mediastinal lymph node size of 1.0 cm or more on the short axis measured with CT was defined as positive for lymphadenopathy preoperatively. Multivariate analysis was calculated in a forward stepwise procedure according to the Cox proportional hazards model [13] using the SPSS statistical package (SPSS version 12.0 for Windows, SPSS Inc, Chicago, IL). A 0.10 level of probability was the significance level used for adding or deleting a covariable from the model. Analysis was regarded as statistically significant at a probability value less than 0.05.

Results Survival and Univariate Analysis Overall 5-year and 10-year survival rates were 41.3% and 35.2% from date of pulmonary excision, respectively. Overall 5- and 10-year survival rates were 73.1% and 40.3% from date of primary colorectal resection, respectively (Fig 1). For overall survival after pulmonary resection, three prognostic factors were determined as significant by univariate p value: CEA, p ⬍ 0.0001 (Fig 2); tumor laterality, p ⫽ 0.0205; and number of pulmonary metastases, p ⫽ 0.0028; log-rank test (Table 1). For overall survival after primary colorectal resection, CEA (p ⫽ 0.0004; Fig 2), tumor laterality (p ⫽ 0.0172), number of pulmonary metastases (p ⫽ 0.0019), hilar or mediastinal lymphadenopathy (p ⫽ 0.0125), primary site (p ⫽ 0.0206), and disease-free interval (p ⬍ 0.0001) each displayed a significant difference (Table 1). For cases of pulmonary metastasectomy with sequential resection after partial hepatectomy, the 5-year survival rate was 31.6% after pulmonary resection and 66.6% after colorectal resection. These rates corresponded to rates for cases with simple pulmonary resection of 44.0% and 75.4%; there was no significant difference (p ⫽ 0.3880 and p ⫽ 0. 3369, respectively; Table 1, Fig 3). The difference in 5-year survival rates between colon and rectum as location for primary lesion was not significant for postpulmonary resection (p ⫽ 0.0557), but it was significant for postcolorectal resection (p ⫽ 0.0206; Table 1, Fig. 4). We could not find any difference of prognosis between the intervals between the liver resection and the lung resection (data not shown).

Multivariate Analyses Multivariate analyses of survival rates after pulmonary resection and colorectal resection were performed according to the Cox proportional hazards model using nine clinical prognostic factors (age, sex, tumor size, tumor number, hilar or mediastinal lymphadenopathy, CEA level, primary site of colorectal cancer, disease-free

GENERAL THORACIC

Ann Thorac Surg 2006;82:254 – 60

256

IIZASA ET AL PULMONARY METASTASECTOMY AND COLORECTAL CANCER

Ann Thorac Surg 2006;82:254 – 60

GENERAL THORACIC

Fig 1. Overall survival curves for patients (n ⫽ 75) with pulmonary metastasis from colorectal cancer. (A) Overall 5- and 10-year survival rates were 41.3% and 35.2%, respectively, after resection of pulmonary metastases. (B) Overall 5- and 10-year survival rates were 73.1% and 40.3%, respectively, after initial colorectal resection.

interval, hepatectomy for liver metastasis). Tumor laterality was excluded as a factor from multivariate analyses because of its considerable statistical correlation with tumor number (p ⬍ 0.0001; ␹2 test). First, each factor was considered independently with a univariate Cox model. Age, CEA, tumor number, tumor size, lymphadenopathy, and primary site were selected for multivariate analysis after pulmonary resection (Table 2). Multivariate analysis after pulmonary resection showed that the independent prognostic factors were age (continuous variable), CEA (elevated versus normal), tumor number (single versus plural), and tumor size (continuous variable; Table 2). Age, CEA, tumor number, lymphadenopathy, primary site, and disease-free interval were selected for analysis with multivariate study after primary tumor resection (Table 3). Multivariate analysis after primary tumor resection demonstrated the prognostic significance of CEA (elevated versus normal), tumor number (single versus plural), tumor size (continuous variable), and diseasefree interval (continuous variable; Table 3).

Fig 2. Survival curves according to serum carcinoembryonic antigen (CEA) concentration. (A) The 5- and 10-year survival rates after pulmonary resection were 11.2% and no survivors, respectively, for patients with elevated CEA (thin line); rates were 65.0% and 59.6%, respectively, for patients without elevation (thick line). (B) The 5- and 10-year survival rates after primary resection were 61.1% and 14.9%, respectively, for patients with elevated CEA (thin line); rates were 81.5% and 65.0%, respectively, for patients without elevation (thick line).

Comment The first and second target organs for metastatic colorectal cancer are the liver or lung. No report has disclosed any data from a well-planned randomized control study comparing surgery and other therapeutic procedures for pulmonary metastasis from colorectal cancer. The reason we selected surgery for these patients is that no effective chemotherapy has been established. Recently, we have encountered many cases of pulmonary metastases from colorectal cancer because of early screening with highly efficient CT such as helical CT, high-resolution CT, and multiplanar reconstruction of CT. Therefore, it is meaningful at this time to reconsider surgical indications for these metastatic lung tumors on the basis of prognostic factors for this patient population. In this study, we retrospectively investigated survival and clinical characteristics in patients who underwent resection of pulmonary metastases at our institution and then evaluated the statistical significance of various prog-

Ann Thorac Surg 2006;82:254 – 60

IIZASA ET AL PULMONARY METASTASECTOMY AND COLORECTAL CANCER

257

Table 1. Univariate Analysis of Patient Characteristics and Overall Survival After Pulmonary Metastasectomy and Primary Colorectal Resection

Characteristic Age ⱖ60 y ⬍60 y Sex Male Female CEA Elevated Within normal Tumor laterality Hemilateral Bilateral Tumor number Single Plural Tumor size ⱕ3 cm ⬎3 cm Lymphadenopathya Positive Negative Primary site Colon Rectum Disease-free interval ⱕ median ⬎ median Hepatectomy for liver metastasis Positive Negative a

Postcolorectal Resection Univariate p Value

5-Year Survival

Univariate p Value

No. of Cases

5-Year Survival

42 33

51.2 27.3

0.0718

78.7 65.9

0.1053

46 29

43.1 37.9

0.5749

76.0 67.4

0.5020

30 45

11.2 65.0

⬍0.0001

61.1 81.5

0.0004

65 10

45.8 0

0.0205

77.0 50.0

0.0172

53 22

50.6 15.7

0.0028

82.5 51.9

0.0019

43 32

45.8 36.1

0.4146

74.0 73.5

0.7169

8 68

15.6 45.1

0.0713

58.3 74.8

0.0125

38 37

52.2 29.5

0.0557

84.4 61.2

0.0206

38 37

35.2 47.7

0.1158

20 55

31.6 44.0

0.3880

66.6 75.4

0.3369

A hilar or mediastinal lymph node size of 1 cm or more in a short axis is defined as positive for metastasis.

nostic factors. We demonstrated some noteworthy results. First, overall 5-year survival rates were 41.3% from pulmonary excision and 73.1% from primary colorectal resection. Second, CEA level, tumor size, tumor number, and patient age were found to be significant independent prognostic factors with multivariate analysis after pulmonary excision. Carcinoembryonic antigen level, tumor number, and disease-free interval were significant for survival after primary colorectal resection. In 1996, we reported on 29 patients with colorectal cancer who underwent surgical excision of pulmonary lung tumors [3]; after that period, reported cases accumulated rapidly. More than 100 cases are currently found in the literature. The current study is noteworthy because it evaluates the surgical process with a consistent protocol at one facility. Most of these studies are retrospective analyses involving many institutions, and there are differences in surgical indications and procedure for each facility [9, 14, 15]. The reported 5-year survival rates of

this surgery were 24% to 63%, and most were around 40% [4 –11, 15–22]. Our criteria for resection of pulmonary metastases from colorectal carcinoma included unilateral or bilateral excisable lung lesions per preoperative chest radiography, no local recurrence of primary lesions, and no extrapulmonary lesions with the exception of associated prior or simultaneous resectable hepatic metastases. These criteria are quite acceptable for surgery of metastatic pulmonary nodules from colorectal cancer based on total prognosis and subjected prognosis, the anatomy, and the literature. Human CEA, originally described as a tumor-specific antigen, is a member of the CEA-related cell adhesion molecule family, which is expressed on the apical surface of colon epithelial cells. Carcinoembryonic antigen expression is associated with progression of colorectal cancers and is maintained in more than 95% of colorectal carcinomas and their metastases, making this glycoprotein the most widely used human tumor marker [23]. In

GENERAL THORACIC

Postpulmonary Resection

258

IIZASA ET AL PULMONARY METASTASECTOMY AND COLORECTAL CANCER

Ann Thorac Surg 2006;82:254 – 60

GENERAL THORACIC

Fig 3. Survival curves according to site of primary tumor. (A) The 5and 10-year survival rates after pulmonary resection were 52.2% and 47.4%, respectively, for patients with primary tumor in the colon (thin line); rates were 29.5% and no survivors, respectively, for patients with primary tumor in the rectum (thick line). (B) The 5- and 10-year survival rates after primary resection were 84.4% and 54.5%, respectively, for patients with primary colon tumors (thin line); rates were 61.2% and 23.6%, respectively, for patients with primary tumors in the rectum (thick line).

this series, data demonstrated that prethoracotomy serum CEA was one of the most significant prognostic factors in both univariate and multivariate analyses for survival after resection of metastatic lung tumors; this is consistent with findings by others [5, 6, 10, 11, 15–18], including our previous report [3]. Elevated serum CEA is the only prognostic factor that has been confirmed to be related to biologic features of primary colorectal cancer. We assume that CEA could be one of the most important factors related to metastasis of colorectal cancer because elevated expression of CEA is associated with poorer prognosis not only with primary colorectal cancer but also with metastatic lung tumors. The mechanism underlying metastasis of colorectal cancer remains obscure in patients with elevated levels of CEA; clinicians should monitor patients for an increase in CEA level after excision of metastatic lung tumors. Tumor size [3, 4, 19, 24], tumor number [5, 7, 8, 10, 14, 16, 17, 19 –22, 24], and lymph node metastasis represent the extent of metastatic lung disease as prognostic factors similar to those included in the TNM classification of Fig 4. Survival curves according to prior hepatectomy for hepatic metastasis. (A) The 5- and 10-year survival rates after pulmonary resection were 31.6% and 31.6%, respectively, for patients who underwent resection of hepatic metastases (thick line); rates were 44.0% and 36.7%, respectively, for patients who did not undergo such surgery (thin line). (B) The 5- and 10-year survival rates after pulmonary resection were 66.6% and 31.7%, respectively, for patients who underwent resection of hepatic metastases (thick line); rates were 75.4% and 43.3%, respectively, for patients who did not undergo such surgery (thin line).

primary lung cancer. It is controversial whether all of those factors are independent prognostic factors as they are for primary lung cancer. Our experience indicates that patients with lymph node metastasis in the hilum or mediastinum have poor prognoses. Some authors have also reported lymph node metastasis to be an independent predictor of poor prognosis after surgical excision of pulmonary metastases from colorectal cancer [5, 7, 9, 18, 21]. We determined that a factor of preoperative lymphadenopathy is not necessarily an independent prognostic factor in this series, but 9 patients with pathologic lymph node metastasis of 44 patients who underwent lymph node dissection showed poorer prognosis than the other patients (data not shown). However, no surgeon believes that systematic lymph node dissection as part of surgery for metastatic lung tumors improves prognosis; hence, we perform a wedge resection for a small peripheral metastasis without inclusion of lymph node dissection. Lymph node metastases to hilum or mediastinum should be considered distant metastases rather than independent nodal factors as in primary lung cancer.

Ann Thorac Surg 2006;82:254 – 60

259

IIZASA ET AL PULMONARY METASTASECTOMY AND COLORECTAL CANCER

Variable Univariate Age Sex (female vs male) CEA Tumor number(single vs plural) Tumor size Lymphadenopathy (positive vs negative) Primary site(rectum vs colon) Disease-free interval Hepatectomy for liver metastasis (no vs yes) Multivariate Age CEA (elevated vs normal) Tumor number (single vs plural) Tumor size CEA ⫽ carcinoembryonic antigen;

Relative Risk

95% CI

p Value

0.951 1.219 3.809 0.354 1.255 0.452 0.513 0.993 0.714

0.911–0.992 0.610–2.436 1.865–7.782 0.174–0.721 1.029–1.531 0.168–1.096 0.256–1.028 0.980–1.006 0.330–1.541

0.0198 0.5755 0.0002 0.0042 0.0249 0.0788 0.0599 0.3033 0.3902

0.952 3.425 0.375 1.249

0.908–0.998 1.612–7.277 0.172–0.817 1.005–1.551

0.0396 0.0014 0.0136 0.0447

CI ⫽ confidence interval.

Venous drainage of the colon and rectum is through the portal vein for the entire colon and the proximal third of the rectum, with drainage through systemic veins for the remaining two thirds of the rectum. The liver is regarded as the first filter of colorectal cancer at time of metastasis; therefore, we justified sequential metastasectomy of lung tumors from colorectal cancer even if liver metastasis had also occurred. We clarified with 20 patients in this study that prior hepatectomy for colorectal metastases does not influence prognosis after resection of lung metastases, a finding that is substantially consistent with other related reports [14, 15, 17, 20, 22, 25]. Moreover, based on venous drainage patterns, we anticipated some difference in prognosis between cases originating in the colon and those originating in the rectum. We could not show any difference in prognosis on the

basis of primary tumor site. This result might be limited by the small number of cases. Cases of pulmonary metastasectomy with eradication of the primary tumor and absence or effective treatment of metastases in other organs are registered in the International Registry of Lung Metastases [26]. However, no surgeon will operate with multiple metastases in both liver and lung simultaneously if the primary lesion is in the breast. Only colorectal cancer could become an acceptable candidate for both liver and lung metastasectomy because of venous drainage of the colon and rectum. In conclusion, there are no reports on total survival time from first surgery for colorectal cancer. We consider the 5-year survival rates found at our institution (41.3% for survival after pulmonary excision and 73.1% for

Table 3. Univariate and Multivariate Analysis of Prognostic Factors and Survival Rates After Colorectal Resection (Cox Proportional Hazards Model) Variable Univariate Age Sex (female vs male) CEA (elevated vs normal) Tumor number (single vs plural) Tumor size Lymphadenopathy (positive vs negative) Primary site (colon vs rectum) Disease-free interval Hepatectomy for liver metastasis (no vs yes) Multivariate CEA (elevated vs normal) Tumor number (single vs plural) Disease-free interval CEA ⫽ carcinoembryonic antigen;

CI ⫽ confidence interval.

Relative Risk

95% CI

p Value

0.950 1.268 3.354 0.341 1.158 0.328 0.451 0.971 0.685

0.912–0.991 0.632–2.545 1.656–6.794 0.168–0.694 0.950–1.410 0.131–0.824 0.226–0.900 0.955–0.987 0.315–1.489

0.0165 0.5031 0.0008 0.0030 0.1464 0.0176 0.0240 0.0005 0.3399

4.346 0.361 0.964

2.074–9.105 0.170–0.766 0.946–0.982

⬍0.0001 0.0080 ⬍0.0001

GENERAL THORACIC

Table 2. Univariate and Multivariate Analysis of Prognostic Factors and Survival Rates After Pulmonary Resection (Cox Proportional Hazards Model)

260

IIZASA ET AL PULMONARY METASTASECTOMY AND COLORECTAL CANCER

GENERAL THORACIC

survival after primary colorectal resection) to be within the permissible range. Prior hepatectomy for metastases from colorectal cancer is not a contraindication for lung metastasectomy. Elevated CEA level and tumor number are the most significant prognostic factors for overall survival after resection of lung metastases from colorectal cancer. Surgery for pulmonary metastasis is recommended if there is no local recurrence and no distant metastasis except for liver. Sequential metastasectomy for liver and lung should be permitted.

References 1. Goldberg RM, Fleming TR, Tangen CM, et al. Surgery for recurrent colon cancer: strategies for identifying resectable recurrence and success rates after resection. Eastern Cooperative Oncology Group, the North Central Cancer Treatment Group, and the Southwest Oncology Group. Ann Intern Med 1998;129:27–35. 2. Midgley R, Kerr D. Colorectal cancer. Lancet 1999;353: 391–9. 3. Fujisawa T, Yamaguchi Y, Saitoh Y, et al. Factors influencing survival following pulmonary resection for metastatic colorectal carcinoma. Tohoku J Exp Med 1996;180:153– 60. 4. Vogelsang H, Haas S, Hierholzer C, Berger U, Siewert JR, Prauer H. Factors influencing survival after resection of pulmonary metastases from colorectal cancer. Br J Surg 2004;91:1066 –71. 5. Pfannschmidt J, Muley T, Hoffmann H, Dienemann H. Prognostic factors and survival after complete resection of pulmonary metastases from colorectal carcinoma: experiences in 167 patients. J Thorac Cardiovasc Surg 2003;126: 732–9. 6. Higashiyama M, Kodama K, Higaki N, et al. Surgery for pulmonary metastases from colorectal cancer: the importance of prethoracotomy serum carcinoembryonic antigen as an indicator of prognosis. Jpn J Thorac Cardiovasc Surg 2003;51:289 –96. 7. Ishikawa K, Hashiguchi Y, Mochizuki H, Ozeki Y, Ueno H. Extranodal cancer deposit at the primary tumor site and the number of pulmonary lesions are useful prognostic factors after surgery for colorectal lung metastases. Dis Colon Rectum 2003;46:629 –36. 8. Ike H, Shimada H, Ohki S, Togo S, Yamaguchi S, Ichikawa Y. Results of aggressive resection of lung metastases from colorectal carcinoma detected by intensive follow-up. Dis Colon Rectum 2002;45:468 –75. 9. Saito Y, Omiya H, Kohno K, et al. Pulmonary metastasectomy for 165 patients with colorectal carcinoma: a prognostic assessment. J Thorac Cardiovasc Surg 2002;124: 1007–13.

Ann Thorac Surg 2006;82:254 – 60

10. Rena O, Casadio C, Viano F, et al. Pulmonary resection for metastases from colorectal cancer: factors influencing prognosis. Twenty-year experience. Eur J Cardiothorac Surg 2002;21:906 –12. 11. Sakamoto T, Tsubota N, Iwanaga K, Yuki T, Matsuoka H, Yoshimura M. Pulmonary resection for metastases from colorectal cancer. Chest 2001;119:1069 –72. 12. Kaplan EL, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958;53:457– 81. 13. Cox DR. Regression models and life tables. J Royal Stat Soc Series B 1972;34:187–220. 14. Kobayashi K, Kawamura M, Ishihara T. Surgical treatment for both pulmonary and hepatic metastases from colorectal cancer. J Thorac Cardiovasc Surg 1999;118:1090 – 6. 15. Regnard JF, Grunenwald D, Spaggiari L, et al. Surgical treatment of hepatic and pulmonary metastases from colorectal cancers. Ann Thorac Surg 1998;66:214 –9. 16. McAfee MK, Allen MS, Trastek VF, Ilstrup DM, Deschamps C, Pairolero PC. Colorectal lung metastases: results of surgical excision. Ann Thorac Surg 1992;53:780 –5. 17. 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. 18. Inoue M, Kotake Y, Nakagawa K, Fujiwara K, Fukuhara K, Yasumitsu T. Surgery for pulmonary metastases from colorectal carcinoma. Ann Thorac Surg 2000;70:380 –3. 19. Goya T, Miyazawa N, Kondo H, Tsuchiya R, Naruke T, Suemasu K. Surgical resection of pulmonary metastases from colorectal cancer. 10-year follow-up. Cancer 1989;64: 1418 –21. 20. van Halteren HK, van Geel AN, Hart AA, Zoetmulder FA. Pulmonary resection for metastases of colorectal origin. Chest 1995;107:1526 –31. 21. Okumura S, Kondo H, Tsuboi M, et al. Pulmonary resection for metastatic colorectal cancer: experiences with 159 patients. J Thorac Cardiovasc Surg 1996;112:867– 84. 22. Yano T, Hara N, Ichinose Y, Yokoyama H, Miura T, Ohta M. Results of pulmonary resection of metastatic colorectal cancer and its application. J Thorac Cardiovasc Surg 1993;106: 875–9. 23. Hammarström S, The carcinoembryonic antigen (CEA) family: structures, suggested functions and expression in normal and malignant tissues. Semin Cancer Biol 1999;9:67– 81. 24. Shirouzu K, Isomoto H, Hayashi A, Nagamatsu Y, Kakegawa T. Surgical treatment for patients with pulmonary metastases after resection of primary colorectal carcinoma. Cancer 1995;76:393– 8. 25. Lehnert T, Knaebel HP, Duck M, Bulzebruck H, Herfarth C. Sequential hepatic and pulmonary resections for metastatic colorectal cancer. Br J Surg 1999;86:241–3. 26. The International Registry of Lung Metastases. Long-term results of lung metastasectomy: prognostic analyses based on 5206 cases. J Thorac Cardiovasc Surg 1997;113:37– 49.