J Clin Oncol 24: by American Society of Clinical Oncology INTRODUCTION

Transcription

1 VOLUME 24 NUMBER 18 JUNE JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T Complete Excision of Primary Breast Tumor Improves Survival of Patients With Metastatic Breast Cancer at Diagnosis Elisabetta Rapiti, Helena M. Verkooijen, Georges Vlastos, Gerald Fioretta, Isabelle Neyroud-Caspar, André Pascal Sappino, Pierre O. Chappuis, and Christine Bouchardy From the Geneva Cancer Registry, Institute for Social and Preventive Medicine, University of Geneva; and the Department of Obstetrics and Gynecology, Unit of Senology and Gynecologic Oncology, Service of Oncology, and Service of Medical Genetics, Geneva University Hospitals, Geneva, Switzerland. Submitted September 13, 2005; accepted February 9, Supported by PROSPER Grant No from the Swiss National Science Foundation (H.M.V). Authors disclosures of potential conflicts of interest and author contributions are found at the end of this article. Address reprint requests to Christine Bouchardy, MD, Geneva Cancer Registry, 55, Bd de la Cluse, CH-1205 Geneva, Switzerland; Christine.BouchardyMagnin@imsp.unige.ch by American Society of Clinical Oncology X/06/ /$20.00 DOI: /JCO A B S T R A C T Purpose Surgery of the primary tumor usually is not advised for patients with metastatic breast cancer at diagnosis because the disease is considered incurable. In this population-based study, we evaluate the impact of local surgery on survival of patients with metastatic breast cancer at diagnosis. Methods We included all 300 metastatic breast cancer patients recorded at the Geneva Cancer Registry between 1977 and We compared mortality risks from breast cancer between patients who had surgery of the primary breast tumor to those who had not and adjusted these risks for other prognostic factors. Results Women who had complete excision of the primary breast tumor with negative surgical margins had a 40% reduced risk of death as a result of breast cancer (multiadjusted hazard ratio [HR], 0.6;, 0.4 to 1.0) compared with women who did not have surgery (P.049). This mortality reduction was not significantly different among patients with different sites of metastasis, but in the stratified analysis the effect was particularly evident for women with bone metastasis only (HR, 0.2;, 0.1 to 0.4; P.001). Survival of women who had surgery with positive surgical margins was not different from that of women who did not have surgery. Conclusion Complete surgical excision of the primary tumor improves survival of patients with metastatic breast cancer at diagnosis, particularly among women with only bone metastases. J Clin Oncol 24: by American Society of Clinical Oncology INTRODUCTION Metastatic breast cancer is considered an incurable disease. 1 The vast majority of patients with metastatic breast cancer do not survive beyond 5 years after diagnosis. 2,3 Treatment for women diagnosed with metastatic breast cancer, therefore, is palliative. Systemic therapy is the treatment of choice, including chemotherapy, hormonal treatment, and biologic therapy. 4,5 The removal of the metastatic lesions is suggested in selected patients. 6 The surgery of breast tumors with distant metastasesisindicatedonlytopreventlocalcomplications. 7-9 It is generally accepted that local therapy provides no survival advantage once metastases have occurred and that, in fact, tumor excision may further stimulate the growth of the metastases. 10,11 These arguments, however, have never been sustained with randomized clinical trials. The one observational study that has been conducted found that surgery of the primary tumor can actually improve survival of metastatic breast cancer. 12 Furthermore, as imaging modalities improve and smaller foci of metastatic disease can be detected, management of the primary tumor in patients presenting with stage IV disease assumes greater clinical relevance. In this population-based study we determined the impact of local surgery on survival of women with metastatic breast cancer at diagnosis. METHODS We used information of the Geneva Cancer Registry, which records all incident cases of cancer diagnosed in the canton of Geneva (approximately 420,000 inhabitants) since Information about the patients comes from several sources, including all university and private laboratories of pathology, hospitals, and practitioners, and is considered accurate, as attested by its low percentage ( 2%) of patient cases recorded from death certificates only

2 Rapiti et al Information recorded for each patient includes sociodemographic data, diagnostic circumstances, diagnostic modalities, pathologic features of the tumor, treatment during the first 6 months, survival, and cause of death. In addition to passive follow-up (routine examination of death certificates and hospital records), the registry regularly assesses survival through an active follow-up performed routinely each year using the files of the Cantonal Population Office, which is in charge of the registration of the resident population. For all deceased patients, the registry s medical staff systematically consults medical files, writes to practitioners to assess cause of death, and codes the cause according to the WHO classification. 14 This study included all women residents in the Swiss canton of Geneva with metastatic breast cancer at diagnosis (stage IV of the TNM classification system 15 : any T, any N, and presence of distant metastasis) for the period 1977 to Of 4,845 women with invasive breast cancer recorded during the study period, 317 (7%) presented with distant metastasis at diagnosis. Women who were registered on the date of their death (n 17) were excluded from the study. The final study included 300 patients. Sociodemographic variables considered were age ( 50, 50 to 59, 60 to 69, 70 to 79, or 80 years), and period of diagnosis (1977 to 1981, 1982 to 1986, 1987 to 1991, or 1992 to 1996). Socioeconomic class was based on the woman s last occupation or, for the unemployed, that of the spouse. Four levels were considered as follows: low (manual employees, and skilled and unskilled workers), middle (nonmanual employees and administrative staff), high (professionals, executives, and administrators), and unknown. 16 The method of discovery of the cancer was grouped into three categories (symptoms, screening, other). The health care sector where treatment took place was categorized as private (private clinics), public (public university hospitals), or unknown. Staging was based on clinical TNM classification. The primary tumor was classified as T0-1 (nonpalpable, 2 cm), T2 ( 2to 5 cm), T3 ( 5cm), T4 (invasion to chest wall/skin and inflammatory carcinoma), and Tx (unknown). Lymph node invasion was classified as N0 (no invasion), N1 (movable axillary), N2 (fixed axillary), N3 (infraclavicular, internal mammary, or supraclavicular) and Nx (unknown). The histologic grade was grouped into four categories: well differentiated, moderately differentiated, poorly differentiated/undifferentiated, and unknown. Estrogen receptors status was categorized as positive (when 10% of tumor cells expressed estrogen receptors), negative, or unknown. Surgical treatment (mastectomy or breast conservation surgery) was divided in four groups: surgery with negative margins, surgery with positive margins, surgery with unknown margins, and no surgery. Other treatments considered were radiotherapy of the breast or chest wall (yes/no), chemotherapy (yes/no), and hormonal treatment (yes/no). Additional data on metastases at the time of diagnosis were extracted retrospectively from the clinical files by trained registrars. The diagnosis of metastasis was classified as confirmed (if based on a histologic or cytologic examination) or suspected (if based on imaging only). We created four categories of metastatic disease according to metastatic site involvement: bone metastases (yes/no), visceral metastases (lungs, pleura, mediastinum, peritoneum, liver, and so on; yes/no), soft tissues metastases (lymph nodes, skin, subcutaneous tissues; yes/no), and CNS metastases (brain, cranial nerves, and so on; yes/no). Women with more than one metastatic site involved could belong to more than one category. The number of metastatic sites was classified as 1, 2, 3 or more. We calculated 5-year breast cancer specific survival by the actuarial method (intervals in months and SE according to Greenwood) 17 and compared the overall survival curves with the log-rank test. We used Cox proportional hazards analysis to calculate both the unadjusted and adjusted mortality risks (hazard ratio [HR]) and the s. Only deaths as a result of breast cancer were considered in all analyses. Patients who died as a result of causes other than breast cancer were censored at the time of death. To evaluate the independent contribution of each variable on mortality from breast cancer, covariates significantly associated (P.05) with prognosis from breast cancer in the univariate analysis were included in a multivariate analysis. Interaction tests were done to evaluate if the effect of surgery varied according to metastatic site. 18 Survival curves were plotted as estimated from the Cox multivariate model. All the analyses were done with SPSS software (version 11; SPSS Inc, Chicago, IL). RESULTS The actuarial breast cancer specific survival of women diagnosed with metastatic disease was 58% at 1 year (, 53% to 64%), 31% at 3 years (, 26% to 37%), and 16% at 5 years (, 11% to 20%). Table 1 shows the characteristics of the patients, tumors, metastases, and treatments. The mean age at diagnosis was 67.4 years (standard deviation, 14.3 years). Thirty-eight women (13%) were younger than 50 years at diagnosis; 137 women (46%) were older than 70 years. Symptoms of the primary tumor or at the metastatic site(s) revealed the diagnosis in the majority of the women (76%). Only 4% were diagnosed through screening or surveillance (mostly clinical examination). Breast cancer diagnosis was only clinical (no microscopic verification) in 13% (n 40) of the women. Another 19% (n 57) were diagnosed by cytology only. A high proportion of breast cancers was reported with unknown grade (70%; n 211) or with unknown estrogen receptor status (63%; n 190). The distant metastases were cytologically confirmed for 43% of the patients (n 129). The sites of metastasis were bones in 64% of the patients (n 193) and viscera in 52% (n 155); 78 women presented with both bone and visceral involvement. Sixty-one patients (20%) had chemotherapy, 131 (44%) had hormonal treatment, 44 (15%) had a combination of both. Sixteen patients (5%) had surgery without other therapies. Forty-seven women (16%) did not receive any treatment. Overall, 173 patients (58%) did not receive any kind of surgery of the primary tumor, whereas 127 (42%) had either mastectomy (n 87) or tumorectomy (n 40). Among women who underwent surgery, 48% (n 61) had negative surgical margins, 26% (n 33) had positive margins, and 26% (n 33) had unknown margins. Axillary lymph node dissection was performed in 24% (n 73). The mean number of removed lymph nodes was 13 (range, one to 34). Most women who had tumorectomy had no adjuvant radiotherapy (73%; n 29). Women who had surgery of the breast tumor were younger than women who did not undergo surgery (mean age, 61.8 years for women who underwent surgery v 71.6 for women who did not undergo surgery; P.0001). They were more often treated in the private sector (35% v 22%, P.013), had a lower T stage (proportion of T1, 15% v 8%; T2, 24% v 17%; T3, 15% v 10%; and T4, 29% v 51%; P.003), and lower N stages (proportion of N0, 14% v 15%; N1, 50% v 31%; N2, 20% v 16%; and N3, 3% v 14%; P.0005). As expected, tumor characteristics were better assessed among women who underwent surgery, with a lower proportion of unknown estrogen receptors status (41% v 80%; P.0001) and unknown grade (46% v 88%; P.0001). Patients who underwent surgery had only one metastatic localization more often (61% v 41%; P.001) and visceral metastasis less often (43% v 58%; P.007). Use of local radiotherapy was higher among women who underwent surgery (21% v 5%; P.0001), and was associated mainly with breastconserving surgery. Use of chemotherapy was lower among women who underwent surgery (53% v 74%; P.0002), whereas hormonal therapy use was similar (43% v 40%; P.637). Five-year breast cancer specific survival was 27% (, 16% to 39%) for women who had surgery with negative margins, 16% (, 3% to 28%) for women who had surgery with positive margins, 12% (, 1% to 23%) for women who had surgery with 2744 JOURNAL OF CLINICAL ONCOLOGY

3 Surgery in Metastatic Breast Cancer Table 1. Main Patient, Tumor, and Treatment Characteristics of 300 Women With Metastatic Breast Cancer at Diagnosis and Their Effect on Breast Cancer Mortality in Univariate Analysis (Geneva, 1977 to 1996) Characteristic No. of Patients % Unadjusted Hazard Ratio Adjusted Hazard Ratio Age groups, years to to to to to to to to 8.2 Socioeconomic class Low Middle to to 1.9 High to to 1.4 Unknown to to 1.3 Period of diagnosis to to to to to to 1.3 Method of discovery Symptoms Screening to to 1.5 Other to to 0.9 Sector of care Private Public to to 1.1 Unknown Clinical T T T to to 2.6 T to to 2.6 T to to 2.6 Tx to to 3.3 Grade Well differentiated Moderately differentiated to to 6.5 Poorly/undifferentiated to to 5.2 Unknown to to 5.0 Regional lymph node involvement N N to to 2.4 N to to 2.5 N to to 4.3 Nx to to 2.5 Estrogen receptors Negative Positive to to 0.9 Unknown to to 1.3 Cytologic/histologic confirmation of metastasis No Yes to to 1.7 Number of sites of metastasis to to to to 1.9 Bone metastasis No Yes to to 1.5 Soft metastasis No Yes to to 1.3 Visceral metastasis No Yes to to 2.1 (continued on following page)

4 Rapiti et al Table 1. Main Patient, Tumor, and Treatment Characteristics of 300 Women With Metastatic Breast Cancer at Diagnosis and Their Effect on Breast Cancer Mortality in Univariate Analysis (Geneva, 1977 to 1996) (continued) Characteristic No. of Patients % Unadjusted Hazard Ratio Adjusted Hazard Ratio CNS metastasis No Yes to to 3.2 Radiotherapy Yes No to to 2.5 Chemotherapy Yes No to to 1.7 Hormonal treatment Yes No to to 3.7 Local surgery None Yes, margins negative to to 1.0 Yes, margins positive to to 2.1 Yes margins unknown to to 1.7 Hazard ratios adjusted for age, period of diagnosis, method of discovery, lymph nodes involvement, estrogen receptor status, cytologic/histologic confirmation, No. of sites of metastases, visceral metastases, metastases of CNS, radiotherapy, hormonal treatment, and surgery. Reference. P.01. P.05. P.001 unknown margin status, and 12% (, 7% to 17%) for women who did not have surgery (log-rank test 19.46; 3 df; P.0002). Table 1 lists the effect of patient and tumor characteristics on mortality from breast cancer in univariate analysis. Age, period of diagnosis, method of discovery, grade, regional lymph node involvement, estrogen receptor status, cytologic/histologic confirmation of the metastases, number of metastatic sites involved, presence of visceral and CNS metastases, radiotherapy, and hormonal therapy were linked significantly to prognosis. Socioeconomic class, sector of care, presence of bone or soft metastases, and chemotherapy were not associated with survival. Surgery of primary tumor was also linked significantly to prognosis (P.0003): women who had surgery with negative margins had a lower risk of dying as a result of breast cancer compared with women who did not have surgical treatment (HR, 0.5;, 0.3 to 0.7), whereas no difference was shown for women with positive surgical margins (HR, 0.8;, 0.5 to 1.1), or unknown margin status (HR, 0.8;, 0.6 to 1.3). The results of the multiadjusted model are listed in Table 1. Age, method of discovery, regional lymph nodes involvement, visceral or nervous system metastases, and hormonal treatment remained significantly linked to breast cancer prognosis. Local surgery also remained independently linked to prognosis (P.048). Patients with surgery of the primary tumor with negative margins had a reduced risk of death from breast cancer compared with patients who did not undergo surgery (HR, 0.6;, 0.4 to 1.0; P.0485). For the other women who underwent surgery, the mortality risk was not significantly different from those of women who did not undergo surgery (HR, 1.3; 95% CI, 0.8 to 2.1 for surgery with positive margins; HR, 1.1;, 0.7 to 1.7 for surgery with unknown margin status). Figure 1 shows the survival curves of patients by surgery derived from the Cox model accounting for other prognostic factors. The effect of surgery on breast cancer mortality was not different for patients with only bone metastasis and patients with metastasis at other sites ( 2, 1.253; 3 df; P.74), After stratifying by site of metastasis, however, we observed that the positive effect of surgery with negative margins was evident particularly for women who had only bone metastases at diagnosis (adjusted HR, 0.2;, 0.1 to 0.4; P.001), whereas for women diagnosed with metastases at other sites the impact on survival was not statistically significant (adjusted HR, 0.7; :, 0.4 to 1.2; P 0 310; Table 2). To rule out the presence of putative biases related to selection of patients for the surgical group, we performed additional subanalyses. First, we repeated the analysis excluding women who survived 30 days or less (n 28) because they did not survive long enough to undergo surgery. For the whole group of patients, a similar reduction in mortality risk from breast cancer was observed, but this was no longer significant (HR, 0.7;, 0.5 to 1.1; P.094). In the subgroup of women with bone metastases only, surgery remained a significant determinant of survival (P.001), and again the effect of surgery with negative margins was associated with a strong reduced risk of breast cancer death (HR, 0.2;, 0.1 to 0.5). Next, we evaluated the effect of surgery in the two subgroups of women with and without axillary dissection. For women with negative margins, the risk of death from breast cancer was reduced in both groups, but the effect seems more evident for the group who also received axillary dissection, although this difference was not statistically significant (HR, 0.2;, 0.02 to 1.9 v HR, 0.7;, 0.4 to 1.2 for women with and without axillary dissection, respectively). Then, we evaluated the effect of surgery in the 192 patients with tumors more than 5 cm and positive lymph nodes (N 1); that is, those patients most likely to undergo complete preoperative work-up for metastatic spread. The results were similar to those of the whole study population (HR, 0.6;, 0.3 to 1.1 for women with surgery with negative margins v HR, 1.0;, 0.5 to 1.9 for women with 2746 JOURNAL OF CLINICAL ONCOLOGY

5 Surgery in Metastatic Breast Cancer positive margins v HR, 1.2;, 0.6 to 2.2 for women with unknown margins). Finally, we excluded from the analysis 12 women whose metastases were diagnosed between 1 and 2 months after the surgery. Again, the direction of our findings did not change, although statistical significance of the results was affected (HR, 0.7;, 0.4 to 1.0 for surgery with negative margins v HR, 1.4;, 0.8 to 2.3 for surgery with positive margins v HR, 1.1;, 0.6 to 1.7 for surgery with unknown margins as compared with no surgery; P.065). DISCUSSION Fig 1. Five-year adjusted specific survival according to local surgery in 300 women with metastatic breast cancer at diagnosis; Geneva 1977 to Only deaths from breast cancer are considered. Survival curves are derived from a Cox model that includes age, period of diagnosis, method of discovery, lymph node involvement, estrogen receptor status, cytological confirmation, number of localizations of metastases, visceral metastasis, metastasis of central nervous system, surgery, radiotherapy, and hormonal therapy. This population-based study shows that total surgical removal of breast tumor improves prognosis of women with metastatic breast cancer. Women who underwent surgery had a 50% reduction in breast cancer mortality compared with women who did not have surgery. This improvement in survival is observed only if the entire primary tumor was removed with free surgical margins. The effect of surgery was not significantly different among patients with different sites of metastasis. However, results of a stratified analysis suggested a greater effect among women with only bone metastases at diagnosis. Currently, breast surgery in metastatic breast cancer patients is limited to salvage mastectomy performed in patients with local problems due to large or ulcerating tumors. Removal of the primary tumor is commonly deemed inappropriate once metastases have occurred because survival is determined by the evolution of the distant metastases and not by the local disease. 7-9 Moreover, biologic data, partly sustained by experimental studies and partly by physicians personal experiences, suggest that the growth of distant metastases may be stimulated after the primary tumor has been removed. 10,11,19 In this study, the surgical removal of primary breast cancer in metastatic patients does not impair survival but actually improves prognosis if done with negative surgical margins. Our results confirm those previously observed for other cancer sites, such as renal cell cancer, 20 colorectal cancer, 21 gastric cancer, 22 and melanoma, 23 for which a reduction of the tumor burden has been reported to be effective in terms of increased patient survival. With regard to breast cancer, only one previous study investigated the impact of surgery of metastatic breast cancer on survival. 12 That large retrospective study is based on the National Cancer Data Base of the American College of Surgeons Table 2. Multiadjusted Effect of Local Surgery on Breast Cancer Mortality Among 300 Women With Metastatic Breast Cancer at Diagnosis According to Site of Metastasis (Geneva, 1977 to 1996) Characteristic No. of Patients % Crude Hazard Ratio Multiadjusted Hazard Ratio P (Wald test) Bone metastasis only 94 Local surgery.001 None Yes, margins negative to to 0.4 Yes, margins positive to to 2.5 Yes margins unknown to to 3.0 Other sites 206 Local surgery.310 None Yes, margins negative to to 1.2 Yes, margins positive to to 2.8 Yes, margins unknown to to 2.0 Hazard ratios adjusted for age, period of diagnosis, method of discovery, lymph nodes involvement, estrogen receptor status, cytologic/histologic confirmation, No. of sites of metastases, surgery, radiotherapy, hormonal treatment. Reference. Visceral, soft tissues or central nervous system

6 Rapiti et al Commission on Cancer which included 16,000 women with stage IV breast cancer at diagnosis. The authors found that patients with surgical resection of the primary tumor had lower adjusted breast cancer mortality risk compared with patients with no surgery. In addition, patients who underwent surgery with free margins had better prognosis than those with positive margins (HR, 0.612; 95% CI, to for negative margins v HR, 0.751;, to for positive margins, as compared with no operation). The results of this study also provide some evidence of an additional beneficial effect of axillary dissection when performed together with the tumor removal. It has been suggested that the total tumor burden plays a central role in survival and that the primary tumor can be considered as another metastatic site. 12 Therefore, the removal of the primary tumor could be viewed as part of a multimodal strategy to prevent additional metastasization of cancer cells. Consistent with this hypothesis, recent studies found a strong correlation between the level of circulating tumor cells and prognosis of metastatic breast cancer. 24,25 Researchers have shown previously that patients with metastatic spread limited to the skeleton have a relatively prolonged survival and a favorable response to systemic therapy 26 compared with patients with involvement of other metastatic sites, 27 but the reasons are unclear Interestingly, the results of our study show that the patients with bone metastases benefit the most from surgical removal of the primary breast tumor. Although this is not a randomized study, population-based observational studies are a valuable tool when evaluating the benefit of surgery in metastatic breast cancer patients. In such studies, it is particularly important to look into the possibility of selection bias due to unrecorded factors. In such a case, the differences found in the mortality risk between women who underwent surgery versus women who did not undergo surgery could reflect lower use of surgery in breast cancer patients with putative poorer disease prognosis. However, the direction and magnitude of the protective effect of surgery was confirmed after adjusting for the main cancer prognostic factors and in subgroup analyses that explored the possibility of selection bias by excluding patients with short survival, delayed metastatic disease, and more advanced tumors. In conclusion, this study provides additional evidence that for women with metastatic breast cancer at diagnosis, primary tumor removal with negative margins significantly improves survival, especially in patients with only bone metastases. Well-designed prospective studies are needed to re-evaluate the treatment paradigm no surgery of the primary tumor in breast cancer with metastases at diagnosis and to determine the impact of breast surgery on outcome of these patients. REFERENCES 1. Wood WC, Muss HB, Solin LJ, et al: Cancer of the breast, in DeVita VT Jr, Hellman S, Rosenberg SA (eds): Cancer Principles & Practice of Oncology. Philadelphia, PA, Lippincott Williams & Wilkins, 2005, pp Falkson G, Holcroft C, Gelman RS, et al: Tenyear follow-up study of premenopausal women with metastatic breast cancer: An Eastern Cooperative Oncology Group study. J Clin Oncol 13: , Donegan WL: Stage IV carcinoma, in Donegan WL, Spratt JS (eds): Cancer of the Breast. Philadelphia, PA, Saunders, Elsevier Science, 2002, pp Bernard-Marty C, Cardoso F, Piccart MJ: Facts and controversies in systemic treatment of metastatic breast cancer. Oncologist 9: , Hortobagyi GN: Treatment of breast cancer. N Engl J Med 339: , Singletary SE, Walsh G, Vauthey JN, et al: A role for curative surgery in the treatment of selected patients with metastatic breast cancer. Oncologist 8: , National Health and Medical Research Council: Clinical practice guidelines for the management of advanced breast cancer. Prepared by the isource National Breast Cancer Centre Advanced Breast Cancer Working Group. Endorsed January National Cancer Institute: Breast Cancer (PDQ): Treatment. Stage IIIB, inoperable IIIC, IV, recurrent, and metastatic breast cancer. HealthProfessional/page8 9. Canadian Cancer Society: Canadian Cancer Encyclopedia. Breast cancer. Treatment by stage. Stage IV breast cancer treatment. Stage IV (any T, any N, M1) Coffey JC, Wang JH, Smith MJ, et al: Excisional surgery for cancer cure: Therapy at a cost. Lancet Oncol 4: , Baum M, Demicheli R, Hrushesky W, et al: Does surgery unfavourably perturb the natural history of early breast cancer by accelerating the appearance of distant metastases? Eur J Cancer 41: , Khan SA, Stewart AK, Morrow M: Does aggressive local therapy improve survival in metastatic breast cancer? Surgery 132: , Bouchardy C: Switzerland, Geneva. In Parkin DM, Whelan SL, Ferlay J, et al (eds): Cancer Incidence in Five Continents, Vol VII. Lyon, France, International Agency for Research on Cancer, 1997, pp WHO: ICD-O International Classification of Diseases for Oncology (ed 1). Geneva, Switzerland, World Health Organization, UICC: TNM Classification of Malignant Tumors. New York, NY, International Union Against Cancer, Bouchardy C, Schüler G, Minder C, et al: Cancer risk by occupation and socioeconomic group among men: A study by the Association of Swiss Cancer Registries. Scand J Work Environ Health 28:1-88, 2002 (suppl 1) 17. Greenwood M: The Natural Duration of Cancer. London, United Kingdom, Her Majesty s Stationary Office, Hill C, Com-Nougué C, Kramar A, et al: Analyse Statistique des Données de Survie. Paris, France, Médecine-sciences Flammarion, O Reilly MS, Holmgren L, Shing Y,: Angiostatin: A novel angiogenesis inhibitor that mediates the suppression of metastases by a Lewis lung carcinoma. Cell 79: , Flanigan RC, Salmon SE, Blumenstein BA, et al: Nephrectomy followed by interferon alfa-2b compared with interferon alfa-2b alone for metastatic renal-cell cancer. N Engl J Med 345: , Rosen SA, Buell JF, Yoshida A, et al: Initial presentation with stage IV colorectal cancer: How aggressive should we be? Arch Surg 135: , Hallissey MT, Allum WH, Roginski C, et al: Palliative surgery for gastric cancer. Cancer 62: , Essner R, Lee JH, Wanek LA, et al: Contemporary surgical treatment of advanced-stage melanoma. Arch Surg 139: , Cristofanilli M, Budd GT, Ellis MJ, et al: Circulating tumor cells, disease progression, and survival in metastatic breast cancer. N Engl J Med 351: , Cristofanilli M, Hayes DF, Budd GT, et al: Circulating tumor cells: A novel prognostic factor for newly diagnosed metastatic breast cancer. J Clin Oncol 23: , Sherry MM, Greco FA, Johnson DH, et al: Metastatic breast cancer confined to the skeletal system: An indolent disease. Am J Med 81: , Briasoulis E, Karavasilis V, Kostadima L, et al: Metastatic breast carcinoma confined to bone: Portrait of a clinical entity. Cancer 101: , Kang Y, Siegel PM, Shu W, et al: A multigenic program mediating breast cancer metastasis to bone. Cancer Cell 3: , Fidler IJ: The pathogenesis of cancer metastasis: The seed and soil hypothesis revisited. Nat Rev Cancer 3: , Goodison S, Kawai K, Hihara J, et al: Prolonged dormancy and site-specific growth potential of cancer cells spontaneously disseminated from nonmetastatic breast tumors as revealed by labeling with green fluorescent protein. Clin Cancer Res 9: , JOURNAL OF CLINICAL ONCOLOGY

7 Surgery in Metastatic Breast Cancer Acknowledgment We thank Hyma Schubert for her help in collecting the data and Stina Blagojevic for her technical and editorial assistance. Authors Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest. Author Contributions Conception and design: Elisabetta Rapiti, Christine Bouchardy Provision of study materials or patients: Georges Vlastos, André Pascal Sappino, Pierre O. Chappuis Collection and assembly of data: Gerald Fioretta, Isabelle Neyroud-Caspar Data analysis and interpretation: Elisabetta Rapiti, Helena M. Verkooijen, Georges Vlastos, Gerald Fioretta, André Pascal Sappino, Pierre O. Chappuis Manuscript writing: Elisabetta Rapiti, Helena M. Verkooijen, Christine Bouchardy Final approval of manuscript: Elisabetta Rapiti, Helena M. Verkooijen, Georges Vlastos, Gerald Fioretta, Isabelle Neyroud-Caspar, André Pascal Sappino, Pierre O. Chappuis, Christine Bouchardy