With increasing incidence of breast cancer, the cancer registry data has been more important to make health policies. The Korean Ministry of Health and Welfare started a nationwide, hospital-based cancer registry (Korea Central Cancer Registry) in 1980.(1) Many other attempts such as population-based regional cancer registry programs and Korea Breast Cancer Registration Program have been made to collect data on the incidences of breast cancer in Korea. These data has helped us understand the incidence and the trend of breast cancer in Korea. But the stage-specific detailed information is still insufficient. Furthermore the proportion of stage IV among overall breast cancer is very small, so the data about stage IV breast cancer in Korea have been insufficient. According to the surveillance, epidemiology and end result (SEER) cancer statistics review from National Cancer Institute, the proportion of stage IV is about 6% in overall breast cancer patient in SEER 9 Registries for 1998-2002.(2) By contrast, the Korean Breast Cancer Society showed the proportion of stage IV breast cancer is 1.9% (122/6,334) in the Nationwide Korean Breast Cancer Data of 2002.(3)

With regard to treatment, stage IV breast cancer shows very different feature from other stage breast cancer in that it almost depends on the medical, systemic therapy. Because distant metastasis of primary breast cancer is frequently found in specific organs, evaluation of distant metastasis at the initial presentation is required for staging and adequate management. To find the predictive characteristics of these stage IV patients, it could help the proper work up of stage IV breast cancer.

Therefore we conducted this study to figure out the predictive factors of stage IV breast cancer.

We reviewed the clinicopathologic findings from prospectively collected database of 1,424 consecutive patients with primary breast cancer who were underwent primary therapy in National Cancer Center, Korea from October 2000 to January 2005. We excluded the patients who received therapy for breast cancer in outside hospital. 38 patients were diagnosed stage IV breast cancer at the initial presentation. We followed up them until September 2006.

We analyzed the age, menopausal status, clinical stage, clinical tumor size, clinical lymph node status, estrogen receptor status (ER), progesteron receptor status (PR), HER2/neu status, p53 expression, histologic grade, histopatholgic type, aspartate transferase (AST) level, alkaline phosphatase (ALP) level, alanine transferase (ALT) level and CA15-3 level at the commencement of therapy. Clinical nodal status was evaluated by chest CT, positron emission tomography (PET) and ultrasonography. In this study, the clinical stage was determined by sixth edition TNM staging by the American Joint Committed on Cancer. HER2/neu expression can be measured and graded with the immunohistochemical (IHC) technique or fluororescence in situ hybridization (FISH).

The diagnosis of metastatic breast cancer was confirmed in a satisfactory fashion. These potential sites (e.g., bone, liver, and lung) should be evaluated. A diagnosis of bone metastasis was defined as positive findings by bone scan. The possibility of liver metastasis was evaluated by CT scan or ultrasonography. Lung metastasis was evaluated by either chest X-ray or chest CT scan. Less common sites of metastasis (e.g., brain metastasis) was evaluated if indicated based on the patients'symptoms. PET scan was also used helpful to find metastasis.

The proportion of stage IV was calculated according to age, menopausal status, tumor size, lymph node status, hormonal receptor status, HER2/neu, histologic grade and laboratory data (e.g. AST, ALT, and CA15-3).

The pearson's chi-square test were used to determine the differences in clinicopathological features between the subgroups of patients. Survival estimates were computed using the Kaplan-Meier method. Multivariate logistic regression was used to evaluate the association between distant metastasis and clinicopathologic factors. The SPSS version 11.0 was used for statistical calculations. All p values were two-tailed. A p value of 0.05 was taken to be significant and the 95% confidence interval was calculated.

The proportion of stage IV was 2.7% (38/1424) in overall patients with primary breast cancer at National Cancer Center in Korea (Table 1). The clinicopathologic characteristics of patients were shown in Table 2. A median age of stage IV breast cancer was 49 yr (range: 28-71 yr). All patients of stage IV breast cancer had invasive ductal carcinoma except two invasive lobular carcinomas. By contrast, there was no stage IV breast cancer in patients with mucinous, medullary, papillary, tubular, and apocrine carcinoma. Palliative mastectomy was performed in seven cases in these patients. The median tumor size was 3 cm (range 1-13 cm) in these 38 patients.

Patients with T0-T1b tumor showed no metastasis at the initial presentation. Only 1.2% of patients with T1c tumors showed distant metastases (6/491). Metastases were found 0.9% of T1 tumors (6/672), 2.4% of T2 tumors (13/535), 8.3% of T3 tumors (4/48), and 27.1% of T4 tumors (13/48). Two patients with metastasis could not be measured by T stage. By contrast, the patients with T0 tumor didn't showed metastasis (p<0.001) (Table 2).

According to the hormonal receptor status, metastases were found in 1.9% (19/979) of ER or PR+ tumors and 4.1% (17/419) of ER-/PR- tumors (p=0.027).

Among node negative breast cancer, only one patient showed distant metastasis at the initial presentation (0.1%) (Table 2). In the patient, the size of tumor was 1.4 cm in ultrasonography. But lung metastasis was shown in chest CT. In node positive patients, metastases were found in 2.7% of T1c tumors (5/183), in 3.9% of T2 tumors (13/333), 10.8% of T3 tumors (4/37), and 27.3% of T4 tumors (12/44) (p<0.001).

The most common metastatic site was bone (47.4%) followed by lung (44.7%) and liver (36.8%). Single organ metastasis was in 19 of 38 (50.0%) and two or more organs were involved in the others (Table 3).

Tumor size (≥2 cm) (p<0.001), positive lymph node (p< 0.001), hormonal receptor positive status (p=0.027), high level of ALP, AST and CA15-3 (p<0.001) were significant factors of distant metastasis on univariate analysis (Table 4). On multivariate analysis, statistically significant predictors of distant metastasis were tumor size (≥2 cm) (p=0.026), positive lymph node (p<0.001), high level of ALP (p=0.013), AST (p=0.003) and CA15-3 (p=0.025) (Table 5).

They received chemotherapy in 73.7% (28/38), radiotherapy in 31.6% (12/38) and palliative surgery in 18.4% (7/38). A median survival was 38.5 months (Fig 1). The median survival was 32.5 months in patients with palliative operation and 24.5 months in patients without palliative operation. It showed that palliative mastectomy didn't afford statistically significant survival advantage (p=0.199).

In patients with single organ metastasis, the median survival was shown differently with respect to the organ affected by the cancer: 38.5 months with bone metastasis, 22.0 months with lung, 29.0 months with liver and 1.0 month with brain (Table 3). Median survival was 38.5 months in patients with bone metastasis only and 22.5 months in patients with other organ metastasis. But it couldn't show significant difference of survival between patients with bone metastasis only and patients with other organ metastasis (p=0.1734) (Fig 2).

The Korean Ministry of Health and Welfare started a nationwide, hospital-based cancer registry (Korea Central Cancer Registry) in 1980.(1) And the Korean Breast Cancer Society started the estimation of baseline data for breast cancer in Korea in 1996 to evaluate the trend and character of breast cancer in Korea.

The Korean Breast Cancer Society showed the proportion of stage IV breast cancer is 1.9% (122/6,334) in the Nationwide Korean Breast Cancer Data of 2002.(3) By contrast, the SEER program reported 6.0% of women in SEER registry for 1998-2002 in the United States showed distant metastasis.(2) In fact, the short-term experience of cancer registry in Korea hasn't showed the reliability of data about breast cancer. With the low proportion of stage IV among overall breast cancer, we could not obtain the detailed information about stage IV breast cancer in Korea.

Therefore we reviewed 1,424 consecutive primary breast cancer patients who undertook primary therapy in National cancer center in Korea and identified 38 patients with distant metastasis at the initial presentation. It was different from other multi-center analysis in that we could reduce the missing data by the review of consecutive patients from single institute. We could also obtain more information of stage IV breast cancer than other hospital in Korea because more patients with advanced stage visited our hospital than others.

The proportion of stage IV among overall breast cancer was 2.7% (38/1,424). It was higher than 1.9% of the Korean Breast Cancer Society. The reasons of high proportion in our data would be unclear. But it seems that we reduced the missing data by a review of consecutive patients and the more patients of advanced cancer visited our hospital. Both data from the Korean Breast Cancer Society and our hospital show the lower proportion of stage IV among overall breast cancer than SEER. We think that the small sized breast in Korean female can let herself detect breast cancer earlier than western female. By contrast, the female in western country would have problem about early detection of breast cancer by self-examination because of huge-sized breast. And the medical insurance system in U.S. would make the patients of low social economic status more difficult for approach to the hospital. Therefore, it seems that the proportion of stage IV breast cancer at the initial presentation is higher than Korean data.

Usually, the work-up including chest X-ray, liver ultrasound, bone scan, and laboratory findings (AST, ALT, ALP and CA15-3) is performed in patients with breast cancer to rule out lung, liver and bone metastases. Multivariate logistic regression analysis revealed that statistically significant predictors of distant metastasis were tumor size (≥2 cm) (p=0.026), positive lymph node (p<0.001), higher level of ALP (p=0.013), AST (p=0.003), CA15-3 (p=0.025).

According to the relation between stage and metastatic location, Myers et al.(4) reported 3.1% of bone metastasis rate by a review of nine studies, 0.5% of lung metastasis rate by a review of four studies and 0.6% of a liver metastasis rate by a review of two studies. They evaluated the detection rates of metastasis according to stage by only tumor size and nodal status. The detection rates of metastasis among patients with stage I breast cancer were 0.5% for bone scan, 0.0% for liver ultrasonography and 0.1% for chest X-ray. Among patients with stage II, the detection rates were 2.4%, 0.4% and 0.2% respectively, and among patients with stage III disease these were 8.3%, 2.0% and 1.7% respectively. By the same way, our data showed the detection rates of metastasis among patients with stage I were 0.0% for bone, 0.0% for liver and 0.2% for lung. Among patients with stage II, the detection rates were 0.6%, 0.3%, 0.8% respectively, and among patients with stage III, these were 7.0%, 5.0%, 4.5% respectively.

It is well known that the tumor marker CA15-3 is a good marker to establish the extent of disease in patients with breast cancer. Nicolini et al.(5) showed that in breast cancer patients the CEA-TPA-CA15-3 tumor marker panel has a high value in selecting those patients with bone metastases, or at high risk of developing clinically evident bone metastases. Our data also showed that CA15-3 was statistically significant predictors of stage IV breast cancer (Table 4, 5).

Our study showed ALP, AST and CA 15-5 were also significant predictive factors of stage IV breast cancer. But this data was different from previous study that showed liver enzymes were not useful in the preoperative diagnosis of hepatic metastases.(6)

The role of local therapy in women with stage IV breast cancer needs to be reevaluated, and local therapy plus systemic therapy should be compared with systemic therapy alone in a randomized trial. In one study of 45 patients with limited stage IV disease, following local surgery and chemotherapy 53% of patients were alive and free from disease after 44 months of follow-up.(7) Our study showed the median survival was 38.5 months in these stage IV patients. In patients with single organ metastasis, the survival times differed with respect to the organ affected by metastasis. The prognosis was good in patients with bone metastasis but the prognosis is poor in patients with brain metastasis.

Our study showed that the factors to predict distant metastasis of breast cancer were large size of tumor, positive lymph node status, elevated alkaline phosphatase, aspartate transferase and CA15-3. Therefore breast cancer patients with those clinical characteristics should be carefully evaluated to detect distant metastasis.