You are on page 1of 7

Original Study

Incidence and Patterns of Distant Metastases for
Patients With Early-Stage Breast Cancer After
Breast Conservation Treatment
Abigail T. Berman,1 Arpi D. Thukral,1 Wei-Ting Hwang,2 Lawrence J. Solin,1
Neha Vapiwala1

This study analyzed the risk and pattern of distant metastases after breast conservation therapy (BCT) for
early-stage breast cancer using a competing risk model. Non– breast cancer deaths were the most common
event and bone was the most common metastatic site. There was no clinically relevant increased risk of
metastases at a specific site based on any patient characteristic examined. Site-specific imaging should be
driven by concerning patient-specific signs or symptoms.
Background: Breast conservation treatment (BCT), consisting of breast conservation surgery followed by definitive
radiation therapy (RT), has been shown to be effective for early-stage breast cancer. Patterns of metastatic failure by
specific anatomic site are not well described in the literature. Methods: A total of 1754 patients with stage I or II
invasive carcinoma of the breast treated with BCT between 1977 and 2003 were identified. Patients were scored
based on first site of metastasis: bone, brain, lung, liver, or other. Non– breast cancer deaths, contralateral breast
cancer, and second malignancies were treated as competing risks events. Cumulative incidence functions for each
competing event were calculated using competing risk methodology. Univariate analysis was performed to determine
the hazard ratio (HR) associated with patient and tumor characteristics. Results: The most common event was
non– breast cancer death (16.5% at 15 years; 95% confidence interval [CI], 13.9%-19.4%). The most common
exclusive first site of metastasis was bone (5.9% at 15 years). The 4 most common anatomic sites of distant
metastases as the first exclusive event were bone (41.1%), lung (22.4%), liver (7.3%), and brain (7.3%). Conclusion:
The present study has demonstrated the site-specific risks of metastases. These data support current clinical practice
of screening for site-specific metastatic disease after BCT based on concerning patient-specific signs or symptoms.

Clinical Breast Cancer, Vol. 13, No. 2, 88-94 © 2013 Elsevier Inc. All rights reserved.
Keywords: Breast cancer, Breast conservation treatment, Metastases

Introduction advantage of improved cosmesis. Despite multimodality treatment,
Breast conservation treatment (BCT), consisting of breast conser- approximately 20% to 30% of women with early-stage breast cancer
vation surgery followed by definitive radiation therapy (RT), has still experience distant metastases.7-10 Breast cancer remains the lead-
been shown to be effective for early-stage breast cancer. Six large ing cause of cancer-related deaths worldwide among women, largely
randomized trials have shown equivalent survival outcomes of BCT secondary to the impact of distant metastases.11
compared with mastectomy.1-6 In addition, BCT has the potential Patterns of metastatic failure by specific anatomic site are not well
described in the literature. In addition, methods to predict which pa-
tients will experience metastases and at which sites remain limited. Pre-
Department of Radiation Oncology, University of Pennsylvania, Philadelphia, PA
Department of Biostatistics and Epidemiology, University of Pennsylvania,
dicting the incidence of metastases at specific sites could allow physicians
Philadelphia, PA to more accurately identify patients for specific screening and prevention
Submitted: Apr 5, 2012; Revised: Oct 22, 2012; Accepted: Nov 8, 2012; Epub: Dec methods and to better tailor individualized clinical management.
5, 2012 The current analysis was performed to examine the incidence and
Address for correspondence: Neha Vapiwala, MD, University of Pennsylvania School patterns of the development of metastases at specific sites—including
of Medicine, TRC 2 West, 3400 Civic Center Boulevard, Philadelphia, PA 19104 brain, lung, bone, and liver—in women with early-stage breast can-
Fax: 214-662-5445; e-mail contact:
cer who were treated with BCT. This study aims to identify patient

1526-8209/$ - see frontmatter © 2013 Elsevier Inc. All rights reserved.
88 Clinical Breast Cancer April 2013

orthovoltage in 8 women (0. the presence or ab. The boost was delivered using electrons of varying energy in 1587 women (90. and a new second malignancy invasive carcinoma of the breast treated with breast conservation surgery occurring before any distant metastases were treated as competing followed by definitive RT at the Hospital of University of Pennsylvania risks events. node-negative disease (66. liver.4% of patients. measure covariate effects but allows other competing events to influ- The surgical treatment in all patients included surgical excision ence failure.12. as indicated for a larger separation between the medial and lateral aspects and both chemotherapy and hormonal therapy were administered in of the chest wall.4%). or third presentation of metastatic spread (Table 3). and photons in 19 women (1%). Overall survival was defined from the start date site-specific distant metastases was performed (Table 4). there were multiple sites of disease at diagnosis of metastasis. or other.3%). Figure 2 shows that the site was recorded. All women had unilateral invasive carcinoma diagnosed after initial breast cancer diagnosis and treatment. or treat- or II (T1-2N0-1) disease. and P values ⬍ . Patient. Second malignancies were non– breast cancers that were between 1977 and 2003.67 Gy) within 4. It can be seen that the risk of death from sence of tumor at reexcision if performed. and treatment characteristics were summarized by Univariate analysis of the patient and tumor characteristics with descriptive statistics. excision. among including a failure probability of 5. 27 years). exclud- of the breast at presentation. Cumulative incidence functions. for these patients were reviewed and the site of metastasis. was further divided into 5 categories.40 Gy). including final margin status of the primary tumor breast cancer death. develop- The study cohort was derived from the 1902 consecutive women with ment of contralateral breast cancer. Patients were scored based on first site of metasta.6-7. Contralateral breast malignancies are more common defined as the presence of metastatic disease diagnosed by radiologic than second malignancies.84 Gy. incidences of each site of metastasis were calculated.1% of patients. and the volume of the boost non– breast cancer deaths begins to exceed breast cancer deaths about field. liver (7. for each com- diagnosed before or during radiation were excluded from analysis (n ⫽ peting event were calculated using competing risk methodology. ie. second. Patients who died of breast cancer but (brain. The The 5-. of the RT to the date of death from any cause or the date censored at ated with the development of metastases. event is shown in Table 2 and displayed in Figure 1.4 years (range. 10-.7. Patients also underwent axillary sites. bone. but higher energies were used 20. Pathologic confirmation was The 4 most common anatomic sites of distant metastases as the not required for the diagnosis of metastatic disease. clinical and pathologic stage I (HR)14 were used to evaluate whether any patient.7% of patients. The most 60-72. Most patients had median dose of the whole breast tangential fields was 46 Gy (mean. 13. occurring 20 years after RT as well. classified by AJCC Staging Manual. whether or not reexcision was performed. Because of limited failures in some distant metastatic (lumpectomy) of the primary tumor. In 21. Median age was bed dose (sum of tangential fields and breast boost) of ⱖ60 Gy. Adjuvant chemotherapy was given in energy used was generally 6-MV photons.9%. All women presented with American Joint Committee on Cancer Fine and Gray’s competing risk regressions for the hazard ratio (AJCC) Staging Manual. 16. The clinical decision regarding the total dose was based common event for patients with early-stage breast cancer was non– on a variety of factors. 5th edition. 4. and whether it was the first. brain.5 to 5 weeks. Therefore the final other was considered as multiple sites of metastases. The most common described.9% (95% CI. hormonal therapy was given in 19. and 15-year cumulative incidence of each competing median total dose to the tumor bed was 63 Gy (mean. The median follow-up time. range. majority of first metastases occur by 15 years but there are first events sis: bone. last follow-up. Patient or Clinical Breast Cancer April 2013 89 . iridium implants in 140 women (8%).1%). One patient in whom a peritoneal metastasis developed as the first In metastatic site–specific analysis.5%). Metastatic disease was years (Table 2). The radiation tumor resection (59. lung. 64.13 134). nonbreast second malignancy (excluding nonmelanoma skin cancers) the failure probabilities before experiencing any events. lung (22. were reclassified as pN2 or pN3 disease (stage tional hazards regression model that provides estimation of HRs to III) based on the AJCC Staging Manual. liver) or multiple sites of first metastases.4%) at 15 1754 patients was 7. ment characteristics were associated with the risk of distant metasta- ically positive lymph nodes.5%). exclusive first site of metastasis itation of the statistical model. at a failure probability of 16.5% (95% CI. Patients with a contralateral breast or a ing nonmelanoma skin cancers. imaging studies and clinical evaluation.9 exclusive first site of distant metastasis was bone at every time point. as N1 disease. This approach is similar to the usual Cox propor- edition. The 54 years (mean. range.13 Patients who had 4 or more patholog. Deaths from causes other than breast cancer. only univariate analyses were performed. The cumulative study cohort consists of 1754 patients.1%) with negative margins at the primary 46.5%). All statistical tests lymph node staging either through sentinel lymph node mapping were 2 sided. tumor. encompassing levels I and II.3%). date of and brain (7. tumor. lung.12. the event of distant metastases and only site of metastatic disease was not included because of the lim. 5th ses at specific sites. Radiation methods at our institution have been previously 17 years after the start of radiotherapy (Figure 1).4%) by 15 years.05 were considered statistically and biopsy or through an axillary lymph node dissection.10 Gy. defined from the start of RT.9%-19. The study population was limited to women who received standard Results tangent radiation fields with a primary tumor bed boost to a total tumor Table 1 lists the patient and tumor characteristics.and tumor characteristics as well as other prognostic factors associ. generally significant. Clinical records first exclusive event were bone (41. 21– 89 years). range 44-52. More without recorded evidence of distant metastases or a contralateral breast than 1 site of metastases occurring within 1-month intervals of each malignancy were excluded from analysis (n ⫽ 13). 54 years. sixth and seventh editions. Patients were scored as distant metastases developing at the time of the first evidence of distant metastatic disease after the Methods start of RT.

5 Hormonal therapy alone 345 19.6 Chemotherapy and hormonal therapy 282 16.4 T2.8 Year of RT Treatment Chemotherapy alone 357 20.3 Cancer Positive 136 7. RT ⫽ radiation therapy ⱕ 2. Currently.1 Isolated lung 2. but the risk is not large enough to justify personalized surveillance after BCT.9 4-9 positive lymph nodes 71 17.2 Isolated brain 0. T1. A number of Positive 1006 57.0 753 42.2 Unknown 178 10. node positive 163 9. the American Society of Clinical Oncology and the National Comprehensive Cancer Network guidelines for patients with early-stage breast cancer include semiannual or annual history.5 Node negative 1160 66. 90 Clinical Breast Cancer April 2013 .3 Negative 429 24.0 356 20.8 Cancer Death.4 patient and tumor characteristics as well as treatment modalities may Negative 387 22.0-5.5 5.9 3.7 1991-2003 834 47.6 tumor characteristics were identified that were associated with an Pathologic Staging increased risk of distant metastases at specific sites. there is ex- Positive 798 45. indicating that physicians should ⬎ 63 854 48.5 Discussion T1.8 Variable 5 Years 10 Years 15 Years Pathologic N Stage Non–Breast Cancer Death 3.6 0.1 Final Margins of Primary Tumor Resection metastases Negative 1022 59.3 tastases and confirm that BCT has favorable results in women with early-stage invasive breast carcinoma through 15 years after treat- Total Radiation Dose to Tumor Bed (Gy) ment.4 Not obtained/unknown 1175 67 60-69 369 21 Adjuvant Therapy 70-89 30 1.2 ⱖ 10 positive lymph nodes 28 6. a significant number of distant metastases do 60-63 900 51.0 4. node negative 254 14.5 Contralateral Breast 3.5 1.1 Tumor Size (cm) Abbreviations: HER2 ⫽ human epidermal growth factor receptor 2.9 These results document the incidence and patterns of distant me- T2.9 2.2 2.3 1-3 positive lymph nodes 317 76.7 Isolated Liver 0. Nonetheless.8 10.7 7.4 6.2 Close 188 11 Unknown 371 21.7 1.3 occur at least 15 years after BCT.7 be mindful of any patient symptoms that could indicate new meta- Estrogen Receptor Status static disease even many years after definitive treatment.9 disease.3 Table 2 Estimated Cumulative Incidence of Non–Breast Unknown 645 36. node positive 296 16.1 As survival rates improve with earlier detection and increased use Progesterone Receptor Status of more efficacious chemotherapy and targeted agents.6 5. Distant Metastases in Breast Cancer After BCT Table 1 Patient and Tumor Characteristics Table 1 (continued) Variable n % Variable n % Age at Diagnosis (y) HER2 Status 20-39 226 12.2 T2 417 23.8 0. Distant Metastasis.1 alter the risk of site-specific distant metastases.8 2.5 pected to be a commensurate increase in the prevalence of early breast Negative 485 27.5 50-59 515 29. node negative 1041 59.8 3.2 Multiple sites of first 1.9 Second Malignancy 2.1 Distant Metastasis Node positive 416 26. and Second Malignancy Using Competing Risk Analysis T1 1337 76.9 Positive 150 8.6 40-49 444 25.7 cancer survivors who are at risk for the development of metastatic Not obtained/unknown 471 26.7 None 768 43. Not obtained/unknown 361 20.2 Isolated bone 3. Contralateral Pathologic T Stage Breast Cancer.4 1977-1990 920 52.2 16.4 8.

5 site overall and as any first site or the exclusive first site of metastasis.1 With the advancement of systemic treatment and improved over- Death (Non–Breast Cancer) 185 10. results of BCT. and can occur in isolation but at a lower risk. There- physical examination. however. Our institution previously reported on the 15-year the use of site-specific metastatic screening. in descend- Isolated Lung 43 2. Third.6 all patient outcomes. Graesslin et al found that age. at risk 1754 1013 509 200 44 3 Years since RT start Abbreviation: RT ⫽ Radiation Therapy. regional. Total 585 distant metastases occur in all 4 major metastatic sites 15 years or more after RT.8 We found that bone metastases were the most common metastatic Isolated Bone 79 4.9 The distant metastasis rate Metastases Sites from the National Surgical Adjuvant Breast and Bowel Project B-06 is also comparable at a distant metastasis rate as the first event at 20 years of n % 25%. astatic site based on a patient or tumor characteristic. and mammogram but no routine laboratory or fore treating physicians should consider a lower clinical threshold for imaging of potential metastatic sites. the The overall rates of distant metastases seen in this study are consistent HRs generally contain large CIs.8 Contralateral breast cancer first Second malignancy first Cumulative Incidence 0.0 0 5 10 15 20 25 No.8 are associated with high rates of lung metastases (18. the current rate of distant metastases are likely less DM than that published in this study.2 0. there were with a large enough increased risk for any given patient or tumor char. it is common for patients to present with multiple sites of metastasis initially. Second. or second malignancy).17 SMN 68 3. Abigail T. Isolated Brain 14 0.9 There are several important conclusions from these data.16 The results of this study rein. DM ⫽ distant metastasis.15. bone is the most common site of metastasis Abbreviations: CLB ⫽ contralateral breast. Fifth. Table 3 Distribution of the First Event by Specific Distant eral breast cancer.5%).0 Non–breast cancer death Distant metastasis first 0. distant recurrent. Therefore these data do not support with previous studies.5 ing order of incidence. development of a given metastatic site.8 necke et al who found that bone is the most likely metastatic site for Multiple sites of first metastases 42 2. and often occurs in isolation. liver.4 all cancer subtypes. This is consistent with data reported by Ken- Isolated Liver 14 0. brain is the least common site nancy. with the exception of basal-type tumors. Lung. First. and Second Malignancy as First Events After BCT 1. multiple statistically significant increases in the risk of a specific met- acteristic to justify site-specific metastatic screening. SMN ⫽ second malig. in which 13% of patients experienced distant metastases Efforts have been made to establish predictive guidelines for the as the first event (defined as local. which CLB 120 6. Contralateral Breast Malignancy. Berman et al Figure 1 Cumulative Incidence of Non–Breast Cancer Death.6 0. Forth. contralat. additional imaging at the time of diagnosing metastasis if there is a force this current practice by showing that there is no site of metastasis clinical suggestion of another site of metastasis.4 0. Clinical Breast Cancer April 2013 91 . Distant Metastases. and brain were the next most common sites.

24 The significant predictors of the development of brain metastasis.00 0 5 10 15 20 25 No. Bone. histologic grade. and tumor characteristics and treatment modalities associated with tients received adjuvant hormonal therapy. lung. We have added to the expanding knowledge of patient despite the majority of tumors being ER⫹/PR⫹. the magnitude 92 Clinical Breast Cancer April 2013 . at risk 1754 1013 509 200 44 3 Years since RT start Abbreviation: RT ⫽ Radiation Therapy. most of the the development of metastases. treatment. Conclusion tion for site-specific metastasis given the presence or absence of ste. and magnetic resonance imaging). Along With Multiple Sites of First Metastases 0.17. bias. and trastuzumab for HER2⫹ breast cancer was intro- dermal growth factor receptor 2 (HER2) status. and liver) in a large cohort of women with early- One limitation of this study is that the adjuvant therapy used stage breast cancer treated with BCT at a single institution and fol- during the era of this cohort is not necessarily representative of mod.23 These conflicting data support this study in the conclusion that it is not appropriate to screen a popula. however.19 others show an unchanged inci. In addition. The presence of steroid hormone receptors in breast cancer pro. incurable out- ern-era adjuvant chemotherapy. progesterone (PR). (brain.20 Brain only Bone only Lung only Liver only 0. grade. affects all patient subgroups. HER2 testing.10 0. hormonal therapy. delay between diagnosis and first metastasis.22. we identified each of metastatic sites. bone. could exist a misclassification bias in that metastases too small for tasis. status of ER. al- static sites. lowed prospectively. static site. number of meta. duced after the majority of patients in our cohort were treated. bone scan. that were suggestive of tumor could in fact be benign. negative human epi. and short disease-free survival are all independent and though trastuzumab has been shown to reduce recurrence risk. This study reports the 15-year incidence of specific metastatic sites roid hormone receptors. however. as biopsy dence in the setting of hormonal therapy. Distant Metastases in Breast Cancer After BCT Figure 2 Cumulative Incidence of Exclusive First Site-Specific Metastasis Including Brain. and Liver as Competing Events. Distant metastasis is a frequent. negative estrogen receptor (ER) status. and HER2. such as anti-HER2 therapy with trastuzumab. The trends in increased use of chemotherapy and the use of hormonal authors then developed a nomogram to predict the probability of therapy and targeted agents could affect both the cumulative proba- brain metastasis in nonbrain metastatic disease: Prognostic points are bility of distant metastases and the relative incidence at each meta- assigned for age. confirmation was not required for classification in our study nor is it strate an increase in the incidence.05 0. Several studies have shown ER/PR presence to decrease the detection by imaging were not detected and radiographic findings incidence of distant metastasis. Lung. There duces conflicting results in predicting the incidence of distant metas. computed tomography. positive (ER⫹) tumors.15 Multiple sites Cumulative Incidence 0.21 ER-negative (ER–) tumors have routinely used in clinical practice. For example.20 and yet others demon. and biological come after BCT and can occur at least up to 15 years after primary therapy. and number With regard to identification of metastases. The risk of multiple sites of metasta- tumors reported in the present study predate the era of routine ses was increased by specific characteristics. only 36% of pa. This potential for misclassification been shown to be associated with early metastasis relative to ER.18 metastatic site by conventional imaging (radiography.

43 .98 .37 1.12 0.19 0.03 0.02 4.03 Unknown vs.02 .44 0 2.96 .51 .62 6. negative NA NA 1.49 .01 0.27 . lung (22.74 1.25 1. tory. N0 1.56 .05 2.6 1.26 .27 .74 .44 .71 .06 1.77 .9% metastases. negative NA NA 0.39 .79 .83 0.17 Final Margins of Primary Tumor Resection Positive vs. has been shown to be effective for early-stage breast ● These data support current clinical practice of screening for site- cancer.01 3. HR ⫽ hazard ratio.93 .06 1. negative 2.08 .07 . 60-63 Gy 1.49 .06 0. negative 1.77 Total Radiation Dose to Tumor Bed ⬎ 63 vs. physical examination.66 3. The authors have stated that they have no conflicts of interest.03 . liver driven by concerning patient-specific site or symptoms.1 .89 .12 0.32 1.02 2.98 .97 ⬍ .66 Unknown vs.51 HER2 Status Positive vs.03 0.56 .55 Unknown vs.41 . specific metastatic disease after BCT based on concerning patient- ● Patterns of metastatic failure by specific anatomic site are not well specific signs or symptoms.29 ⬍ .19 .51 .23 . tumor.79 .51 .32 .14 0.19 .97 .42 .78 .3%). positive 2. Berman et al Table 4 Univariate Analysis of Exclusive Site of First Metastasis and Multiple Sites by Risk Factors Variable Brain Bone Lung Liver Multiple HR P Value HR P Value HR P Value HR P Value HR P Value Age at Diagnosis 0.6 Close vs.33 .79 ⬍ .11 .95 .77 . cific sites.36 2.04 0. positive 1.01 1.57 .25 .91 0 0.68 1.65 0. itive RT.06 0.51 1.73 0.07 0.47 therapy Abbreviations: HER2 ⫽ human epidermal growth factor receptor.37 0.019 1.21 Year of RT Treatment 1991-2003 vs. described in the literature. negative 1. NA ⫽ not available.5 ⬍ .65 .32 .88 3. (7. consisting of breast conservation surgery followed by defin.23 0.29 1.4%). Abigail T.98 .8 .72 . and treat- Clinical Practice Points ment characteristics increased the risk of distant metastases to spe- ● BCT.75 Estrogen Receptor Status Negative vs.13 2.75 .14 2.17 ⬍ .47 0. at 15 years).33 .43 Progesterone Receptor Status Negative vs.67 2.01 1.38 . and mammography should remain the ● The 4 most common anatomic sites of distant metastases as the standard for surveillance after BCT. Clinical Breast Cancer April 2013 93 . ● Using competing risks methodology. and brain (7.92 Adjuvant Therapy Chemotherapy alone 3. positive 3.37 .39 Pathologic T Stage T2 vs.46 .3%).1%).3 1.74 . we identified that the most common event after BCT was non– breast cancer death (16.73 .19 Unknown vs.31 0 2.63 .19 .02 1. T1 2.07 Chemotherapy and hormonal NA NA 1. positive 0. of this risk is too small for the routine surveillance of site-specific ● The most common exclusive first site of metastasis was bone (5.31 .92 1.12 2.67 .5% at Disclosure 15 years).28 0.79 .3 Pathologic N Stage N1 vs.91 .27 1. multiple distinct patient.37 1.4 2.45 Hormonal therapy alone 0.35 .99 .61 1.05 ⬍ .92 .012 0.02 0. RT ⫽ radiation therapy.15 .94 0.011 1.14 ⬍ .71 . 1999-1990 0. Site-specific imaging should be first exclusive event were bone (41.35 .4 . The recommended practice guidelines with routine his.31 1.38 NA NA NA NA 0.57 . ● On univariate analysis.79 0.1 .29 3.7 .01 3.46 0.

Biostatistics comparing a conservative treatment to mastectomy in early breast cancer. the value of breast-conserving therapy in stage I and II breast cancer. Schultz DJ. Trastuzumab plus adjuvant chemotherapy CA Cancer J Clin 2011. Byrd DR. 2012. Jemal A. Compton CC. McArthur H. Risk of early recurrence among Pennsylvania experience. Accessed: November 21. conservation with mastectomy in the treatment of stage I and II breast cancer. 28: 6. Fine JP. Fifteen-year results of breast-conserving surgery receptors to prognosis in early breast cancer. Veronesi U. Yerushalmi R. 19. Danish asp#site. Forman D. Jacobson JA. Oncol 1989. et al. 20:4141-9. Andersen JA. Rose C. Anderson S. treatment of invasive breast cancer. eds. et al. brain metastasis in patients with metastatic breast cancer. postmenopausal women with estrogen receptor-positive early breast cancer treated 10. Radiother 2001. Gage I. or both for 21:269-77. Recht A. Available at: http://www. Arriagada R. Cancer Res 1980. J Clin Oncol 2010. J Clin Oncol 2002. Nomogram to predict subsequent trial. et al. 24. lumpectomy. Ten-year results of a comparison of 2032-7. prevention of ipsilateral breast tumor recurrence after lumpectomy in women with 8. 347:1233-41. for operable HER2-positive breast cancer. Hawkins RA. et al. Staging system for breast cancer: revisions for comparing total mastectomy. 94 Clinical Breast Cancer April 2013 . Solin LJ. Effects of chemo. et al. Fowble BL. National Cancer Comprehensive Cancer Network. Br J 7. N Engl J Med 2002. Fisher B. et al. Kennecke H. radiation therapy. et after primary therapy. et al. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG). Eur J Cancer 1990. et al. Distant Metastases in Breast Cancer After BCT References 12. Ten year results of conservative surgery and Cancer 2002. Smith TJ. 17. surgery and radiation therapy. irradiation for stage I and II breast cancer. N Engl J Med 2005. et al. Breast conservation is the treatment of choice surveillance guidelines can be done in a routine office visit. The American Society of Clinical Oncology recommended breast cancer 3. Bartelink H. Bryant J. in small breast cancer: long-term results of a randomized trial. Danish randomized trial comparing Breast. with adjuvant tamoxifen. Bray F. and lumpectomy plus irradiation for the the 6th edition of the AJCC Cancer Staging Manual. Woods R. Kennecke H. 7th ed.nccn. Santiago RJ. 58:233-40. 33:245-51. 15. Metastatic behavior of breast cancer 5. Twenty-year follow-up of a randomized trial 13. Coutant C. Edge. 61:69-90. 2009. 11:19-25. 83:803-19. Singletary SE. New York. et al. 16. Outcome after extended follow-up in N Engl J Med 1995. Danforth DN. Saphner T. 40:3357-60. Pichon MF. J Natl Cancer Inst Monogr 1992. Regression modeling of competing crude failure probabilities. Int J Radiat Oncol Biol Phys 2004. Ashley P. Bryant J. Sarrazin D. 365:1687-717. Gelman R. Olivotto IA. Lê MG. AJCC Cancer Staging Manual. Tesdale AL. Center MM. et al. 23:6807. a prospective study of operable breast cancer: key factors and a prognostic index. Harris E. 21. Cowan KH. et al. et al. 332:907-11. Int J Radiat Oncol Biol Phys 1991. Salvadori B. Relationship of presence of progesterone 9. Ward E. Blichert-Toft M. Fisher B. Breast Cancer Cooperative Group. 11:15-8. Graesslin O. Abdulkarim BS. Lancet 2005. 1. Prescott RJ. 26:668-70. 112:1437-44. Ten-year results of a randomized trial 14. Global cancer statistics. breast conservation therapy with mastectomy: six years of life-table analysis. Ferlay J. Bryant J. Clinical Practice Guidelines— 4. 14:177-84. et al. et al. Randomized clinical trial to assess subtypes. 11. J Natl Cancer Inst Monogr 1992. 87:8-14. and definitive irradiation for stage I and II breast carcinoma: the University of 22. Fentiman IS. 14:2738-46. van Dongen JA. Perez EA. Long-term outcome following breast-conserving invasive breast cancers of one centimeter or less. Dignam JJ. Surg Clin North Am 2003. 20. J Clin Oncol 2005. Romond EH. Brunet M. 28:3271-7. J Clin Oncol 2010. therapy and hormonal therapy for early breast cancer on recurrence and 15-year 23. NY: Springer. J Clin Oncol 1996. Banfi A. et al. Annual hazard rates of recurrence for breast cancer survival: an overview of the randomised trials. Gray R. SB. EORTC 10801 18. 353:1673-84. 2:85-97. et al. Pallud C. Allred C. Cancer 2008. Wu L. 2. Tamoxifen. Int J Radiat Oncol Biol Phys 1995. Tormey DC.