Purpose Despite breast cancer being the most common cancer in women in the developed world, only a minority of patients (\< 10%) has stage IV disease at diagnosis (1). In addition, 20-30% of patients with early breast cancer will experience distant metastatic relapse (2). Due to advances in available multimodality therapies and a better understanding of tumor biology, survival of stage IV patients is constantly improving (3-5).
Clinically, oligometastatic breast cancer is characterized by solitary/few detectable lesions, usually limited to single organs, in which local therapy with curative intent could impact survival. This population of 'potentially curable' stage IV disease is estimated to be 1-10% of newly diagnosed patients with metastatic breast cancer (5). A multimodal approach is endorsed for these selected patients (5,6). The identification of patients with truly oligometastatic disease is challenging. Most published series refer to an era before modern imaging (i.e. positron emission tomography-computed tomography), and thus many patients were probably understaged, potentially leading to underestimation of the global effect of an aggressive local management.
Patients with oligometastatic disease can be divided into 3 cohorts (7,8): 1) patients who present with oligometastases; 2) patients with residual oligometastases after systemic therapy; and 3) those with relapsed oligometastases after curative locoregional therapy. These different groups have possibly distinct prognoses, and may need differential approaches.
Several series have reported on lung and liver metastases resection in oligometastatic breast cancer and most data are from small series of patients collected over many years. The largest dataset comes from the International Registry of Lung Metastases and presents results of lung metastasectomy in 467 breast cancer patients (9). Complete resection was possible in 84% of patients and led to a median survival of 37 months (5-year OS = 38%, 10-year OS = 22%). Identified prognostic factors for lung resection include disease-free interval and number of metastases (disease-free interval \>36 months and solitary metastases being the most favorable), ER status, size of metastases, completeness of resection, and use of anatomical resection (as opposite to wedge resection) (9-19). Obviously, the reported data refer to a subset of patients who were selected for favorable prognostic factors, and any comparison with surgically untreated patients is threatened by serious biases.
Pulmonary resection in metastatic breast cancer patients, apart from its potential therapeutic value, is also an important diagnostic tool, especially in patients with a suspected first recurrence, allowing for differential diagnosis with second primary lung cancers and benign lesions (19,20). The proportion of lesions proved not to be breast cancer metastases in various series ranges from 7% to 66% (9,14,18,21). As the morbidity and mortality of pulmonary resection has decreased substantially over the last decades, this potentially beneficial procedure can be discussed in a selected group of patients (9,19,21).
Surgery to remove liver metastases is an accepted treatment modality in patients with colorectal cancer (22). Its role in breast cancer is, however, much less recognized. Liver is a common metastatic breast cancer site, but only 5% of patients have isolated liver involvement. In various series of hepatic resection for breast cancer metastases, the reported median survival ranged from 14.5 to 63 months and the 5-year survival from 14% to 61%, in general, comparing well with nonsurgically treated patients. Most of the reported series, however, describe extremely selected patients, constituting 1% or less of metastatic breast cancer patients treated over the respective periods of time (22-26). As an example, in the one of the important series, all patients were asymptomatic and identified through intensive surveillance programs (27). Patients with isolated liver metastasis can potentially be managed with local treatments (surgery, radiofrequency ablation (RFA), transcatheter arterial chemoembolization (TACE), cyberKnife stereotactic radiosurgery). In contrast to liver colorectal metastases, local treatment for breast cancer liver metastasis is not considered a therapeutic option due to common involvement of additional organs; nevertheless, in some selected patients, local approaches have been associated with long-term survival (10,29,30).
Because patients who are potentially eligible for this therapeutic strategy represent only 1%-3% of the total metastatic breast cancer population, a large global collaboration is needed to confirm its impact on long-term survival or cure and to ensure adequate statistical power and strength of the results. Enrolled patients must have comparable clinical and biological characteristics, extent of staging, and frequency of monitoring to avoid selection biases and stage migration, which can substantially affect the outcome.
The number of intervention performed for metastatic breast cancer has dramatically increased over the past 2 decades. Hepatectomy and pulmonary resection for stage IV colorectal cancer is now considered the standard of care for resectable patients with isolated hepatic and/or pulmonary disease and acceptable performance status. However, the indications for resection / intervention of breast cancer origin metastases are not as clearly defined. The aim of this study to focus on emerging data for the intervention (resection and/or radiofrequency ablation (RFA), transcatheter arterial chemoembolization (TACE), cyberKnife stereotactic radiosurgery) of breast cancer metastatic disease to the lung and liver, with a focus on indications for resection / intervention.
Patient acquisition and data collection All patient should be female. All following demographic, clinical and tumour specific data will be collected: patient age, performance status according to the WHO classification at the time of first presentation to the hospital, stage, type and receptor status (ER, PR, HER-2) of the primary tumour, tumor load in the lung and/or liver (uni-or bilateral affection, number of metastases, diameter of the largest tumour size), chronology of the metastases (metachronous or synchronous), presence of further metastatic (synchronous metastases) or recurrent (metachronous) extrahepatic tumour growth and lymph node involvement in the mediastinal region or hepatoduodenal ligament. In addition, treatment-related and follow-up data will be documented: Intervention method (surgical resection, cryosurgery, laser-induced thermotherapy, radio-frequency ablation, ethanol injection, cyber-knife), median follow-up, incidence of tumour recurrence, time to recurrence, overall survival, and the administration of chemotherapy. Survival will be calculated from the date of diagnosis of the lung and/or liver metastases.
Definitions Lung and/or liver metastases will be defined as metachronous if the interval between resection of the primary tumour and first diagnosis of metastases is longer than 3 months. General physical condition of study patients will be measured by the performance status according to the WHO classification.
Prior and during the treatment, all patients will be undergone evaluation with complete medical history, physical examination, baseline blood tests, and imaging evaluations (positron emission tomography (PET-CT) and/or radiography, bone scintigraphy, computed tomography scan, magnetic resonance imaging, ultrasonography). Effects of the treatment will be monitoring by imaging analyses at every 3 months for lesions detected at diagnosis of lung and/or hepatic metastasis or every 6 months without detectable lesions before the treatment.
R0 resection is a complete resection with no microscopic residual tumor, while R1 resection means a complete resection with no grossly visual tumor as defined by surgeon, but microscopic cancer may be left behind. R2 resection indicates a partial resection with grossly visible tumor left behind.
Effects of the treatment were assessed by analyses of maximum response, according to either the criteria of the World Health Organization (WHO) (30) or the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.0 (31) since its announcement in 2004. Responses in bone metastases were evaluated using the methods reported by MD Anderson Cancer Center (32,33), with CR as clear evidence of complete bone recalcification with attainment of near-normal bone architecture or normalization of scan, and partial response as radio- logical evidence of sclerosis in lytic lesions or marked improvement of bone scan. In recent years, MRI was performed when available, and the imaging information was taken into account in evaluation of bone metastases. Responses in pleural fluids were determined as: CR for radiological demonstration of complete disappearance of pleural fluid, and PR for 50% or more decrease. For study inclusion, pathological diagnosis of the primary lesions had to be obtained, whereas tissue or cellular pathology of the metastatic lesions was preferred but not mandatory. In patients who performed surgical resection, a complete histopathological response will be defined as the absence of vital tumour cells. Detection of tumour necrosis or scar formation within the tumour tissue will be regarded as a partial pathohistological response.
Ethics committee approval in Istanbul University Istanbul Medical Faculty can be obtained to participate for this study.
Statistical analysis The characteristics of the patients with assessable data will be compared with X2 test or Wilcoxon rank-sum tests. Overall survival (OS) will be defined as duration from diagnosis of the metastasis to last visit or death. Progression free interval (PFI) will be defined as duration from diagnosis of the metastasis to the point when disease progression will be detected. Complete response (CR) will be achieved by a systemic therapy or no evidence of clinical disease (NED) after a local therapy considered relapse-free, and the duration between induction of relapse free status and the point of relapse detection will be defined as the relapse free interval (RFI); survival time after induction of relapse-free status without relapse will be defined as relapse free survival (RFS). Unrelated death other than from breast cancer will be considered as censored in evaluation of PFI and RFI. Survival curves will be calculated according to the Kaplan-Meier method. The log-rank test will be used to evaluate differences in survival between groups with intervention and no-intervention. A multivariate analysis will be performed by Cox regression and the following variables will be included: age, DFI, performance status (PS), liver metastasis, pulmonary metastasis, presence of other metastasis, hormone receptor status, HER2 status, Ki-67 status, intrinsic subtype, administration of local therapy, and chemotherapy regimens. Statistical significance will be defined as a p value \<0.05. All p values will be calculated for two-sided analyses. For observed survival, all deaths will be considered as events including patients who died from secondary causes without recurrence. Statistical analyses will be performed using SPSS 22.0 (SPSS, Chicago, IL) software.
