In patients with human epidermal growth factor receptor-2 (HER2)-overexpressing breast cancer, treatment with trastuzumab has been shown to markedly improve the outcome. We investigated the role of trastuzumab on brain metastasis (BM) in HER2-positive breast cancer patients. From 1999 to 2006, 251 patients were treated with palliative chemotherapy for HER2-positive metastatic breast cancer at Samsung Medical Center. The medical records of these patients were analysed to study the effects of trastuzumab on BM prevalence and outcomes. Patients were grouped according to trastuzumab therapy: pre-T (no trastuzumab therapy) vs post-T (trastuzumab therapy). The development of BM between the two treatment groups was significantly different (37.8% for post-T vs 25.0% for pre-T, P=0.028). Patients who had received trastuzumab had longer times to BM compared with patients who were not treated with trastuzumab (median 15 months for post-T group vs 10 months for pre-T group, P=0.035). Time to death (TTD) from BM was significantly longer in the post-T group than in the pre-T group (median 14.9 vs 4.0 months, P=0.0005). Extracranial disease control at the time of BM, 12 months or more of progression-free survival of extracranial disease and treatment with lapatinib were independent prognostic factors for TTD from BM.

The achievement of the pathologic complete response (pCR) has been considered a metric for the success of neoadjuvant chemotherapy (NAC) and a powerful surrogate indicator of the risk of recurrence and long-term survival. This study aimed to develop a multimodal deep learning model that combined clinical information and pretreatment MR images for predicting pCR to NAC in patients with breast cancer. The retrospective study cohort consisted of 536 patients with invasive breast cancer who underwent pre-operative NAC. We developed a deep learning model to fuse high-dimensional MR image features and the clinical information for the pretreatment prediction of pCR to NAC in breast cancer. The proposed deep learning model trained on all datasets as clinical information, T1-weighted subtraction images, and T2-weighted images shows better performance with area under the curve (AUC) of 0.888 as compared to the model using only clinical information (AUC = 0.827, P < 0.05). Our results demonstrate that the multimodal fusion approach using deep learning with both clinical information and MR images achieve higher prediction performance compared to the deep learning model without the fusion approach. Deep learning could integrate pretreatment MR images with clinical information to improve pCR prediction performance.

Background: There has been reported that the association between nodal spread and tumor size was disrupted in triple-negative breast cancer (TNBC) and it showed characteristically early relapse. The TNM (tumor–node–metastasis) staging system might not be equally effective as a prognostic indicator for all subtypes. The aim of our study was to evaluate the usefulness of the staging according to subtypes. Patients and methods: We conducted a retrospective analysis of invasive breast cancer patients who received curative surgery at Samsung Medical Center from 2000 to 2004. Relapse-free survivals (RFS) by stage were analyzed. Results: Thousand eight hundred and seventy-nine patients who were available clinicopathologic data were included. These patients were divided into three subtypes: hormone receptor (HR)+, human epidermal growth factor receptor 2+, and triple negative groups. As the stage became more advanced, the slope of each stage of the RFS curves of patients with HR+ and HER2+ steadily increased. In contrast, RFS curves intermingled and showed overlap from stage 1 to 3A in TNBC patients. There was only wide separation of RFS curves between stage 1-3A and 3B-3C in TNBC. Conclusions: The current TNM staging system might not be enough for encompassing the tumor biology and for predicting outcomes to make therapeutic decisions for all BCs, especially for TNBC patients.

Exosomes in blood play an important role in cell-to-cell signaling and are a novel source of biomarkers for the diagnosis and prognosis of diseases. Recently, evidence has accumulated that cytokines are released from encapsulated exosomes and are capable of eliciting biological effects upon contact with sensitive cells. However, there is currently limited information on exosome isolation methods for cytokine research. In this study, we evaluated three exosome isolation methods for their usability, yield, purity, and effectiveness in subsequent cytokine profiling. We found that ultracentrifugation (UC) and Exoquick (EQ), but not exoEasy, yielded appropriate exosome sizes, and EQ had higher exosome extraction efficiency than the other two methods. Although UC generated markedly fewer particles than EQ, it yielded a relatively high purity. Next, we performed a multiplex assay with the ProcartaPlex Immune Monitoring 65-Plex Panel to determine the feasibility of these methods for cytokine profiling. The results indicated significant differences among isolation methods when analyzing exosomal cytokine profiles. We further investigated the changes of exosomal cytokines according to breast cancer progression in triple-negative breast cancer. We found significantly decreased concentrations of MIP-3 alpha, IL-23, M-CSF, Eotaxin-3, BLC, SDF-1 alpha, IL-2R, MDC, FGF-2, IL-22, and IL-31 in exosomes from metastatic breast cancer (MBC) patients.

Lee SJ, et al. Cancer Res Treat. 2011 Jun;43(2):89-95. http://dx.doi.org/10.4143/crt.2011.43.2.89

Leptomeningeal metastasis (LM) usually occurs late during the course of breast cancer. The aim of this study was to characterize the clinical features and outcomes of LM based on breast cancer subtypes in conjunction with brain parenchymal metastases. A retrospective study was performed of breast cancer patients with LM, who received palliative management at Samsung Medical Center between 1995 and 2008. Among the 272 metastatic breast cancer patients with central nervous system (CNS) involvement, 68 patients with LM were identified. The median age was 46 years (range, 24–72 years). The median survival duration from LM to death (LM-OS) was 4.5 months (range, 0.2–26.4 months). Patients surviving for 12 or more months were rarer among triple negative (TN) patients compared to other subtypes (21.7% for HR + ve vs. 27.8% for HER2 + ve vs. 72.7% for TN, P = 0.217). Death caused by CNS involvement appeared to be much more common in TN than in other subtypes (0% for HR + ve vs. 36% for HER2 + ve vs. 64% for TN, P = 0.060). Median survival time from distant metastasis was significantly different among the three groups (28.3 vs. 29.1 vs. 11.8 months, P < 0.0001). However, median survival time from LM did not differ (4.1 vs. 5.9 vs. 3.8 months, P = 0.226). Characteristic manifestations and treatment outcomes of LM may be affected by the unique biology of breast cancer intrinsic subtypes. The different roles of active combined treatment modalities including both systemic chemotherapy and local treatment modalities should be considered to improve outcomes.