Childbirth after perioperative systemic therapy in patients with breast cancer: a retrospective single institutional study in Japan.

BACKGROUND: The number of breast cancer patients of childbearing age has been increasing. Therefore, we investigated the characteristics and the childbearing status of the patients who received systemic therapy for breast cancer during their childbearing age to better understand the clinical impact of childbirth. METHODS: Female patients with breast cancer younger than 40 years old who underwent surgery and received perioperative systemic therapy from 2007 to 2014 were included in this study. We compared the characteristics of patients with and without childbirth after treatment. RESULT: Of 590 patients, 26 delivered a child, and 355 did not bear a child during the median observation period of 8.1 years, whilst 209 had unknown childbirth data. The childbirth group had a lower mean age at surgery (32.2 vs. 35.1, P < 0.001). The proportion of patients who desired childbirth and used assisted reproductive technology was significantly higher in the childbirth group (65.4 vs. 23.9% and 45.2 vs. 5.1%, respectively, P < 0.001). The patients in the childbirth group had significantly less advanced disease (P = 0.002). In the childbirth group, the age at childbirth was significantly older in patients who received combined endocrine therapy and chemotherapy (40.8 years) than in patients who received either alone (endocrine therapy: 36.9 years, chemotherapy: 36.7 years, P = 0.04). However, survival was not different between those with and without childbirth. CONCLUSION: It is critical to recognize the desire for childbirth in patients with breast cancer who are receiving systemic therapy and to provide them with necessary fertility information before treatment to support their decision-making.

Prognostic significance of histological therapeutic effect in preoperative chemotherapy for breast cancer.

To establish a prognostic prediction system, we examined the relationships between prognosis and histological therapeutic effect or ypTNM classification in 258 breast cancer patients who received neoadjuvant chemotherapy. The case distribution according to therapeutic effect was nine patients (3.5%) with Grade 0, 169 (65.5%) with Grade 1, 58 (22.5%) with Grade 2, and 22 (8.5%) with Grade 3. The 5-year overall survival (OS) rate by therapeutic effect was 56% in Grade 0, 81% in Grade 1, 87% in Grade 2, and 96% in Grade 3. The higher the therapeutic effect the better the prognosis, with a significant difference among the groups (P = 0.008). The case distribution according to ypTNM classification was 20 patients (7.8%) with Stage 0, 83 (32.2%) with Stage I, 77 (29.8%) with Stage II, and 78 (30.2%) with Stage III. The 5-year OS rate by ypTNM classification was 95% in Stage 0, 94% in Stage I, 89% in Stage II, and 59% in Stage III. While prognosis was mostly comparable in Stages 0 and I, in the other stages it became significantly worse as residual cancer increased (P < 0.001). The prognosis of breast cancer patients with neoadjuvant chemotherapy can be predicted by histological therapeutic effect and staging classification of residual cancer.

Four types of ipsilateral breast tumor recurrence (IBTR) after breast-conserving surgery: classification of IBTR based on precise pathological examination.

We classified ipsilateral breast tumor recurrences (IBTRs) based on strict pathological rules. Ninety-six women who were surgically treated for IBTR were included. IBTRs were classified according to their origins and were distinguished based on strict pathological rules: relationship between the IBTR and the primary lumpectomy scar, surgical margin status of the primary cancer, and the presence of in situ lesions of IBTR. The prognosis of these subgroups were compared to that of new primary tumors (NP) in the narrow sense (NPn) that occurred far from the scar. Distant-disease free survival of IBTR that occurred close to the scar with in situ lesions and a negative surgical margin of the primary cancer (NP occurred close to the scar, NPcs) was similar to that of NPn. In contrast, IBTR that occurred close to the scar without in situ lesions (true recurrence (TR) that arose from residual invasive carcinoma foci, TRinv) had significantly poorer prognosis than NPn. IBTR that occurred close to the scar with in situ lesions and a positive surgical margin of the primary cancer (TR arising from a residual in situ lesion, TRis) had more late recurrences than NPcs. Precise pathological examinations indicated four distinct IBTR subtypes with different characteristics.

Characteristics, treatment trends, and long-term outcomes of Japanese patients with pregnancy-associated breast cancer (PABC).

PURPOSE: To clarify the characteristics, treatment trends, and long-term outcomes of patients with pregnancy-associated breast cancer (PABC). METHODS: PABC includes breast cancer diagnosed during pregnancy (PBC) and breast cancer diagnosed within 1 year after childbirth or during lactation (LBC). We compared clinical characteristics of 126 patients with LBC and 49 patients with PBC who underwent surgery at our hospital from 1946 to 2018. Survival was compared between patients with LBC and those with PBC in terms of breast cancer-specific disease-free survival (BC-DFS) and overall survival (OS). RESULTS: Patients with LBC were more likely to have family history, lymph node metastasis, lymphatic invasion, and to receive chemotherapy than patients with PBC. Patients with LBC showed poorer BS-DFS and OS than patients with PBC. Among patients with LBC, those treated after 2005 were older at surgery, had a smaller tumor size, received more systemic therapy, and had a more favorable prognosis than patients treated before 2004. Family history, breast cancer within 1 year after childbirth, and surgery before 2004 as well as cStage, lymph node metastasis, and lymphatic invasion were significantly associated with poor prognosis in patients with LBC. In the multivariate analysis for BC-DFS and OS among patients with PABC, LBC vs PBC did not remain as an independent prognostic factor while cStage remained. CONCLUSION: Patients with LBC had a poorer prognosis than those with PBC, most likely due to disease progression rather than biological characteristics. Early detection and optimization of systemic treatments are critical for improving the outcomes of patients with LBC.

Assessment of axillary node status by ultrasound after neoadjuvant chemotherapy in patients with clinically node-positive breast cancer according to breast cancer subtype.

The use of sentinel node biopsy (SNB) following neoadjuvant chemotherapy (NAC) for patients with cN1 breast cancer is controversial. Improvements of negative predictive value (NPV) by axillary ultrasound (AUS), which corresponds to the accurate prediction rate of node-negative status after NAC, would lead to decreased FNR of SNB following NAC. In this study, we retrospectively investigated the accurate prediction rate of NPV by AUS after NAC in patients with cytologically node-positive breast cancer treated between January 2012 and December 2016. Of 279 eligible patients, the NPV was 49.2% in all patients, but varied significantly by tumor subtype (p < 0.001) and tumor response determined by magnetic resonance imaging (MRI) (p = 0.0003). Of the 23 patients with clinically node negative (ycN0) by AUS and clinical complete response in primary lesion by MRI, the NPV was 100% in patients with HR±/HER2+ or HR-/HER2- breast cancer. In conclusion, regarding FNR reduction post-NAC, it will be of clinical value to take tumor subtype and primary tumor response using MRI into account to identify patients for SNB after NAC.

Metastatic ovarian cancer spreading into mammary ducts mimicking an in situ component of primary breast cancer: a case report.

BACKGROUND: Accurate diagnosis of metastatic tumors in the breast is crucial because the therapeutic approach is essentially different from primary tumors. A key morphological feature of metastatic tumors is their lack of an in situ carcinoma component. Here, we present a unique case of metastatic ovarian carcinoma spreading into mammary ducts and mimicked an in situ component of primary carcinoma. To our knowledge, this is the second case (and the first adult case) confirming the in situ-mimicking growth pattern of a metastatic tumor using immunohistochemistry. CASE PRESENTATION: A 69-year-old Japanese woman was found to have a breast mass with microcalcifications. She had a known history of ovarian mixed serous and endocervical-type mucinous (seromucinous) carcinoma. Needle biopsy specimen of the breast tumor revealed adenocarcinoma displaying an in situ-looking tubular architecture in addition to invasive micropapillary and papillary architectures with psammoma bodies. From these morphological features, metastatic serous carcinoma and invasive micropapillary carcinoma of breast origin were both suspected. In immunohistochemistry, the cancer cells were immunoreactive for WT1, PAX8, and CA125, and negative for GATA3, mammaglobin, and gross cystic disease fluid protein-15. Therefore, the breast tumor was diagnosed to be metastatic ovarian serous carcinoma. The in situ-looking architecture showed the same immunophenotype, but was surrounded by myoepithelium confirmed by immunohistochemistry (e.g. p63, cytokeratin 14, CD10). Thus, the histogenesis of the in situ-like tubular foci was could be explained by the spread of metastatic ovarian cancer cells into existing mammary ducts. CONCLUSION: Metastatic tumors may spread into mammary duct units and mimic an in situ carcinoma component of primary breast cancer. This in situ-mimicking growth pattern can be a potential pitfall in establishing a correct diagnosis of metastasis to the breast. A panel of breast-related and extramammary organ/tumor-specific immunohistochemical markers may be helpful in distinguishing metastatic tumors from primary tumors.

Ductal carcinoma in situ and sentinel lymph node metastasis in breast cancer.

BACKGROUND: The impact of sentinel lymph node biopsy on breast cancer mimicking ductal carcinoma in situ (DCIS) is a matter of debate. METHODS: We studied the rate of occurrence of sentinel lymph node metastasis in 255 breast cancer patients with pure DCIS showing no invasive components on routine pathological examination. We compared this to the rate of occurrence in 177 patients with predominant intraductal-component (IDC) breast cancers containing invasive foci equal to or less than 0.5 cm in size. RESULTS: Most of the clinical and pathological baseline characteristics were the same between the two groups. However, peritumoral lymphatic permeation occurred less often in the pure DCIS group than in the IDC-predominant invasive-lesion group (1.2% vs. 6.8%, p = 0.002). One patient (0.39%) with pure DCIS had two sentinel lymph nodes positive for metastasis. This rate was significantly lower than that in patients with IDC-predominant invasive lesions (6.2%; p < 0.001). CONCLUSIONS: Because the rate of sentinel lymph node metastasis in pure DCIS is very low, sentinel lymph node biopsy can safely be omitted.

Association of Parity and Infant Feeding Method with Breast Density on Mammography.

BACKGROUND: Mammography (MMG) is widely used for the screening and diagnosis of breast cancer. High breast density on MMG prevents breast cancer detection. In the present study, we analyzed the effects of parity and infant feeding method on mammographic breast density. MATERIALS AND METHODS: Ninety women diagnosed as having breast cancer were analyzed. Using the results of medical questionnaires, subjects were grouped according to their parity and infant feeding method, into either nulliparity or parity, and formula-feeding or breastfeeding. Each group consisted of 30 consecutive women from 45 to 49 years. Nulliparous group was unaffected by either parity or breastfeeding, formula-feeding group was affected by parity but not breastfeeding, and breastfeeding group was affected by both parity and breastfeeding. Mammographic breast density was evaluated on the contralateral and cancer-free breast, and was classified into ≥ 50% dense as high breast density and < 50% dense according to the Boyd quantitative scale. The association of parity and infant feeding method with breast density on MMG was analyzed using the Fisher's exact test. RESULTS: Regarding parity, high breast densities were observed in 40% and 17% of nulliparous and parous breasts, respectively. Nulliparous breasts had significantly higher density than parous breasts (p  =  0.010). Regarding infant feeding method, 17% of breasts in both the formula-feeding and breastfeeding groups had high densities. Infant feeding method was not associated with mammographic breast density (p  =  1.0). CONCLUSION: Nulliparity is associated with higher breast density, and parity, regardless of infant feeding method, induces atrophic changes of breast tissue.

Efficacy of radiation boost after breast-conserving surgery for breast cancer with focally positive, tumor-exposed margins.

Many patients with positive margins following breast-conserving surgery (BCS) undergo re-excisions that aim to remove residual disease from the breast, which brings a tremendous emotional burden in addition to financial consequences. We sought to determine whether re-excisions could be safely avoided without compromising local control and survival by using whole-breast radiation therapy (WBRT) with a tumor bed boost in patients with early-stage breast cancer with focally positive, tumor-exposed margins after BCS. All patients with ductal carcinoma in situ (DCIS) and/or invasive breast cancer (IBC) who had pathologically tumor-exposed margins following BCS, without re-excision and treated with WBRT with tumor bed boost between March 2005 and December 2011, were included. The radiotherapy consisted of WBRT at a dose of 50 Gy in 25 fractions, followed by a tumor bed boost with an additional dose of 16 Gy in eight fractions. A total of 125 patients fulfilled the eligibility criteria; of the 125 patients, 1 had bilateral breast cancer, resulting in 126 cases. Invasive disease was found in 102 (81%) cases and purely ductal carcinoma in situ (DCIS) disease in 24 (19%) cases. The 10-year ipsilateral breast tumor recurrence (IBTR) -free survival, progression-free survival (PFS), and 10-year overall survival (OS) rates were 95%, 92.5% and 96%, respectively. Patients with early-stage breast cancer who receive BCS and have focally positive, tumor-exposed margins can avoid re-excision by undergoing WBRT followed by a sufficient dose of tumor bed boost, without negatively impacting local control and survival.

Correction to: Histological classification of breast tumors in the General Rules for Clinical and Pathological Recording of Breast Cancer (18th edition).

In the original publication of the article, the author group and acknowledgements were published incorrectly.

Carcinoma in a solitary intraductal papilloma of the breast.

Herein is reported a rare case of carcinoma arising from papilloma of the breast. A 63-year-old postmenopausal woman noticed a nodule approximately 1 cm in diameter in her left breast. Ultrasonography indicated a mass with a solid pattern within an intracystic tumor measuring 1.5 x 1.5 x 1.4 cm in diameter located near the left nipple. On total image analysis malignancy could not be denied, therefore lumpectomy with resection of the surrounding tissue was performed. Histologically the tumor consisted of cancerous and papilloma components. The cancer cells had high-grade nuclear atypia, were irregular, and contained abundant eosinophilic cytoplasm with a thin vascular stalk. In contrast, the tumor cells had no atypia, and had a thick stroma in the papilloma components. Both lesions could be distinguished clearly from each other. In addition, a transition from papillary to cancerous elements in some areas was seen. An additional partial mastectomy was performed after the lumpectomy but no carcinoma foci were noted in the excised tissue. Possible occurrence of cancerous change in solitary intraductal papilloma of the breast was suspected.

Skin invasion and prognosis in node negative breast cancer: a retrospective study.

BACKGROUND: The impact of skin invasion in node negative breast cancer is uncertain. METHODS: We determined the prognosis in 97 node negative breast cancer patients (case group) who had tumors with skin invasion. Then we compared these patients with 4500 node negative invasive breast cancer patients treated surgically in the same period. RESULTS: Patients with skin invasion tended to be older, had more invasive lobular carcinoma and larger tumor size, and were less likely to have breast conserving surgery than those in the control group. The 5-year disease-free survival rate in the case group was 94.0%. There was no significant difference in the 10-year disease-specific overall survival rates in terms of skin invasion in node negative patients (90.7% in the case group, 92.9% in the control group; p = 0.2032). CONCLUSION: Results suggest that skin invasion has no impact on survival in node negative invasive breast cancer patients. The adjuvant regimens which the individual institute applies for node negative breast cancer should be used regardless of skin invasion.

Sequential Sentinel Node Scintigraphy With Planar and SPECT/CT Images Revealed Contralateral Drainage From Ipsilateral Breast Tumor Relapse in a Patient With Bilateral Breast Cancer.

Radio-guided sentinel node (SN) biopsy is routinely performed in patients with early breast cancer. However, repeated SN scintigraphy in ipsilateral breast tumor relapse (IBTR) often shows extra-axillary drainage, including contralateral axilla. A patient diagnosed with bilateral breast cancer, of which one was IBTR, was studied by sequential radio-guided SN mapping, radiocolloid injection to an IBTR breast and scanning, and then radiocolloid injection to the other breast and scanning. This revealed the lymphatic flow from the IBTR breast to the contralateral axilla. Sequential method may help to depict contralateral axillary lymphatic flow from an IBTR breast in patients with bilateral breast cancer.

Diagnostic ultrasonography and mammography for invasive and noninvasive breast cancer in women aged 30 to 39 years.

PURPOSE: To confirm which modality, ultrasonography (US) or mammography (MMG), is useful to detect breast cancer in women aged 30 to 39 years, and to compare the sensitivity and findings of these two modalities for invasive carcinoma and ductal carcinoma in situ (DCIS) in the diagnostic setting. METHODS: We retrospectively evaluated the sensitivity and findings of these two modalities in 165 patients aged 30 to 39 years, who underwent surgery at the Cancer Institute Hospital between 2001 and 2003. US or MMG were performed after obtaining information on the other modalities previously used and physical examination. The abnormal findings of US were defined as mass lesions and focal hypoechoic areas due to breast cancer. The abnormal findings of MMG were defined as category 3 to 5 (Japanese Mammography Guidelines) masses, calcifications, and other findings due to cancer. RESULTS: Of 165 patients, 147 patients (89%) mammographically showed dense breasts. Histologically, 146 (88%) were invasive carcinomas and 19 (12%) were DCIS. In all carcinomas, the sensitivity of US (95%) was higher than that of MMG (85%). The sensitivity of US for invasive carcinoma (99%) was higher than that of MMG (85%). On the other hand, the sensitivity of MMG for DCIS (89%) was much higher than that of US (68%). CONCLUSIONS: US is more sensitive to detect breast cancers than MMG in this age range, especially for invasive carcinoma. On the other hand, MMG is useful for detecting DCIS, especially when it manifests with microcalcifications.

Diagnostic mammography and ultrasonography for palpable and nonpalpable breast cancer in women aged 30 to 39 years.

PURPOSE: To investigate the relationship between the tumor size of breast cancer by palpation and the sensitivity of mammography (MMG) and ultrasonography (US), and which modality can detect nonpalpable breast cancer in women aged 30 to 39 years. METHODS: We retrospectively evaluated the tumor size by palpation, breast density, and the sensitivity of MMG and US in 165 patients aged 30 to 39 years. Palpation, US, and MMG were performed with prior knowledge of the results of other modalities. The tumor size on palpation were classified into Tnp; nonpalpable, T1p; 2 cm or less, T2p; more than 2 cm, but not more than 5 cm, and T3p; more than 5 cm. RESULTS: Of 165 patients, 147 patients (89%) showed mammographically dense breasts. Of 165 cancers, 14 (8%) were Tnp, 40 (24%) were T1p, 82 (50%) were T2p, and 29 (18%) were T3p. The sensitivity of MMG was 57% (8 of 14) for Tnp, 78% (31 of 40) for T1p, 90% (74 of 82) for T2p, and 97% (28 of 29) for T3p. The sensitivity of US was 43% (6 of 14) for Tnp and 100% for palpable cancers. Of 14 nonpalpable cancers, 4 (29%), 4 (29%), and 2 (14%) could be detected by only MMG, bloody nipple discharge, and US. CONCLUSIONS: The sensitivity of MMG depends on the tumor size on palpation in this age range. MMG fails to detect relatively large palpable cancers. On the other hand, US can detect all palpable cancers. However, the sensitivity of US declines for nonpalpable cancers. For the detection of nonpalpable cancers, MMG, US, and nipple discharge are complementary modalities.

Cystoid Macular Edema during Treatment with Paclitaxel and Bevacizumab in a Patient with Metastatic Breast Cancer: A Case Report and Literature Review.

We present a case of a metastatic breast cancer patient with cystoid macular edema (CME) occurring during treatment with paclitaxel and bevacizumab. She had a history of neoadjuvant chemotherapy and partial mastectomy plus axillary lymph node dissection for stage IIB left-breast cancer. Twenty-four months later, she was diagnosed with multiple bone metastases and underwent chemotherapy with paclitaxel and bevacizumab. Thirty-three months after the initiation of the chemotherapy, she noticed bilateral blurred vision. The retinal thickening with macular edema was observed by optical coherence tomography, resulting in a diagnosis of CME. With cessation of paclitaxel and administrating ocular instillation of a nonsteroidal anti-inflammatory drug, her macular edema gradually reduced and disappeared in a month. While CME caused by chemotherapy is very rare, taxane may cause ocular adverse events such as CME. It is important to urge patients to consult an ophthalmologist promptly when they have visual complaints during taxane chemotherapy.

Surgical excision without whole breast irradiation for complete resection of ductal carcinoma in situ identified using strict, unified criteria.

BACKGROUND: The definition of complete resection of ductal carcinoma in situ (DCIS) is difficult to standardize because of the high variety of surgical breast conserving procedures, specimen handling, and pathological examinations. Using strictly controlled criteria in a single institute, the present study aimed to determine the ipsilateral breast cancer rate when radiotherapy is omitted following complete resection of DCIS. METHODS: We retrospectively examined 363 consecutive DCIS patients who underwent breast-conserving surgery, and of these, 125 (34.4%) had complete resection according to the criteria. We finally included 103 patients who omitted radiotherapy. Ipsilateral and contralateral breast cancer events were assessed. RESULTS: The median follow-up period was 118 months. The incidences of ipsilateral and contralateral breast cancer and ipsilateral invasive breast cancer at 10 years were 10.8%, 9.1%, and 3.6%, respectively. No patient died of breast cancer. CONCLUSION: If complete resection of DCIS can be ensured, the annual incidence of ipsilateral breast cancer, even without irradiation, can be limited to approximately 1%, which equals the incidence of contralateral breast cancer.

Tolerability and safety of classic cyclophosphamide, methotrexate, and fluorouracil treatment in Japanese patients with early breast cancer.

BACKGROUND: Few reports have addressed the feasibility and safety of classic Cyclophosphamide, Methotrexate, and Fluorouracil (CMF) therapy in Japanese female breast cancer patients. METHODS: Twenty-four Japanese patients who received classic CMF, identical to the originally described treatment regimen were studied in terms of treatment dose, treatment delay, and toxicity. RESULTS: Classic CMF was not discontinued in any of the cases. The median delay in treatment was 14 days, and the mean administered dose of cyclophosphamide was 98.2% of the planned dose. None of the patients suffered severe side-effects such as febrile neutropenia; however, in 22 patients in whom the effect of CMF on hair loss could be assessed, 7 (31.8%) had to wear hats or wigs. CONCLUSIONS: Classic CMF is a feasible and safe regimen in Japanese breast cancer patients. In Japan, this regimen is still available for some specific groups of early breast cancer patients.

Frozen section is superior to imprint cytology for the intra-operative assessment of sentinel lymph node metastasis in stage I breast cancer patients.

BACKGROUND: A standard intra-operative procedure for assessing sentinel lymph node metastasis in breast cancer patients has not yet been established. PATIENTS AND METHODS: One hundred and thirty-eight patients with stage I breast cancer who underwent sentinel node biopsy using both imprint cytology and frozen section were analyzed. RESULTS: Seventeen of the 138 patients had sentinel node involvement. Results of imprint cytology included nine false negative cases (sensitivity, 47.1%). In contrast, only two cases of false negatives were found on frozen section (sensitivity, 88.2%). There were two false positive cases identified by imprint cytology (specificity, 98.3%). On the other hand, frozen section had 100% specificity. CONCLUSION: These findings suggest that frozen section is superior to imprint cytology for the intra-operative determination of sentinel lymph node metastasis in stage I breast cancer patients.

Optimal surveillance for postoperative metastasis in breast cancer patients.

PURPOSE: To establish an optimal surveillance schedule after surgery for breast cancer, patients included in an institutional database were retrospectively investigated with respect to the first metastatic site and timing of recurrence. PATIENTS AND METHODS: We investigated 11,676 pT1-4pN0-2M0 breast cancer patients treated from 1985 to 2009 and followed up until June 2014. Our surveillance protocol included physician visits and examinations with bone scans, liver echography, chest roentgenography and laboratory tests. We evaluated the liver, bones, lungs and pleura as surveillance covering sites (SCS) in addition to parameters such as time points exceeding 80 % with respect to the accumulated percentage of patients of recurrence and the number of surveillance per one recurrence (NSR), calculated by dividing the number of patients at risk of recurrence at the start of a particular time frame by the number of patients of recurrence at SCS within that period. RESULTS: There were a total of 1,962 recurrent patients, including 601 patients with locoregional recurrence, nine patients with recurrence in the opposite breast, 1,349 patients with recurrence at distant sites and three unknown patients. The number of patients with the bones, lungs, liver and pleura as the first site of recurrence was 447, 324, 144 and 69, respectively, and 72.9 % of the distant metastatic lesions belonged to SCS. The five-year overall survival rate after recurrence among the patients with single recurrent site was longer than that observed among the patients with multiple sites of recurrence (43.3 vs 25.3 %; p < 0.0001). In addition, more than 80 % of the patients of liver metastasis were detected within 5 years after surgery, while 80 % of the patients of pleura metastasis were detected within 10 years. The NSR was below 200 for the 10-year period, as was the NSR of the patients with lymph node metastasis and a positive hormone receptor status. In contrast, the NSR of the patients with a negative hormone receptor status was above 200 after 5 years. CONCLUSIONS: In this study, the prognosis of the patients with a single site of recurrence was superior to that of the patients with multiple sites. Curable patients with distant metastases included those with single metastatic sites. The optimal surveillance schedule should be established taking into consideration that the incidence of metastasis differs among metastatic sites during follow-up.