Development of Prediction Model Including MicroRNA Expression for Sentinel Lymph Node Metastasis in ER-Positive and HER2-Negative Breast Cancer.

BACKGROUND: The aim of our study is to find microRNAs (miRNAs) associated with sentinel lymph node metastasis (SLNM) and to develop a prediction model for SLNM in ER-positive and HER2-negative (ER+/HER2-) breast cancer. PATIENTS AND METHODS: In the present study, only ER+/HER2- primary breast cancer was considered. The discovery set for SLNM-associated miRNAs included 10 tumors with and 10 tumors without SLNM. The training and validation sets both included 100 tumors. miRNA expression in tumors was examined comprehensively by miRNA microarray in the discovery set and by droplet digital PCR in the training and validation sets. RESULTS: In the discovery set, miR-98, miR-22, and miR-223 were found to be significantly (P < 0.001, fold-change > 2.5) associated with SLNM. In the training set, we constructed the prediction model for SLNM using miR-98, tumor size, and lymphovascular invasion (LVI) with high accuracy (AUC, 0.877). The accuracy of this prediction model was confirmed in the validation set (AUC, 0.883), and it outperformed the conventional Memorial Sloan Kettering Cancer Center nomogram. In situ hybridization revealed the localization of miR-98 expression in tumor cells. CONCLUSIONS: We developed a prediction model consisting of miR-98, tumor size, and LVI for SLNM with high accuracy in ER+/HER2- breast cancer. This model might help decide the indication for SLN biopsy in this subtype.

Vasculogenic mimicry is associated with trastuzumab resistance of HER2-positive breast cancer.

BACKGROUND: Trastuzumab is a drug that targets the receptor tyrosine kinase HER2 and is essential for the treatment of HER2-positive breast cancer. Resistance to the drug leads to severe consequences, including disease recurrence, tumor enlargement, and metastasis. We hypothesized that trastuzumab treatment might be associated with phenotypic switching in HER2-positive breast cancer cells (BCCs), enabling them to escape and survive the effect of trastuzumab. METHODS: We conducted comprehensive immunophenotyping to detect phenotypic changes in HER2-positive BCCs treated with trastuzumab, based on criteria determined a priori. Based on immunophenotyping results, we characterized the vascular phenotypes of HER2-positive BCCs by western blotting, real-time RT-PCR, and tube formation assay. The vascular phenotype of tumor cells from clinical samples was evaluated by staining with periodic acid-Schiff and an anti-CD31 antibody. We explored small molecule inhibitors that suppress tube formation and determined the inhibitory mechanism. RESULTS: Out of 242 cell surface antigens, 9 antigens were significantly upregulated and 3 were significantly downregulated by trastuzumab treatment. All upregulated antigens were related to endothelial and stem cell phenotypes, suggesting that trastuzumab treatment might be correlated to switching to a vascular phenotype, namely, vasculogenic mimicry (VM). Several VM markers were upregulated in trastuzumab-treated cells, but these cells did not form tubes on Matrigel, a functional hallmark of VM. Upon analysis of three trastuzumab-resistant HER2-positive cell lines, we found that all three cell lines showed tube formation on Matrigel in the presence of angiogenic growth factors including EGF, FGF2, IGF1, or VEGF. Clinically, VM channels significantly increased in surviving cancer cell clusters of surgically removed tumors pretreated with trastuzumab and chemotherapy compared to both surgically removed tumors without prior systemic treatment and tumors biopsied before presurgical treatment with trastuzumab. Finally, we found that salinomycin completely suppressed VM in all three trastuzumab-resistant cell lines through disruption of actin cytoskeletal integrity. CONCLUSIONS: VM promotes metastasis and worsens patient outcomes. The present study indicates that HER2-positive BCCs can exhibit VM in an angiogenic microenvironment after eventually acquiring trastuzumab resistance. The clinical finding supports this in vitro observation. Thus, targeting VM might provide a therapeutic benefit to patients with HER2-positive breast cancer.

Prognostic value of tumor cell DNA content determined by flow cytometry using formalin-fixed paraffin-embedded breast cancer tissues.

PURPOSE: The use of formalin-fixed paraffin-embedded (FFPE) tumor tissues in flow cytometry (FCM)-based determination of tumor cell DNA content is more complicated than the use of fresh-frozen tissues. This study aimed to accurately measure tumor cell DNA content from FFPE tissues by separating tumor cells from stromal cells through FCM and investigating its prognostic impact. METHODS: We separately measured the DNA contents of tumor cells and stromal cells by gating with pan-cytokeratin and vimentin (FCMC/V). We evaluated tumor cell DNA contents [DNA index (DI)] of 290 FFPE tumor tissues and classified them into low and high DI groups, using a cutoff DI value determined through an unbiased computational method. RESULTS: The distribution of DI was bimodal, and a cutoff value was determined at a DI of 1.26. The high-DI tumors were associated with aggressive phenotypes and had significantly worse distant recurrence-free intervals (DRFI) than low-DI tumors. Multivariate analysis revealed that lymph node metastasis, Ki67, and DI were independent factors affecting DRFI. Accordingly, patients with low-DI/low-Ki67 tumors had excellent outcomes compared with other tumor types. Multiploid tumors were associated with increased lymphocytic infiltration and aggressive phenotypes. CONCLUSIONS: The DI of FFPE tumors could be precisely determined through FCMC/V. A combination of DI and Ki67 analyses may be able to predict the prognoses of breast cancer patients with greater accuracy.

Clinicopathological analysis of homologous recombination-deficient breast cancers with special reference to response to neoadjuvant paclitaxel followed by FEC.

PURPOSE: This study aimed to elucidate the clinicopathological characteristics of breast tumors with homologous recombination deficiency (HRD) and the sensitivity to neoadjuvant paclitaxel followed by fluorouracil, epirubicin, and cyclophosphamide (P-FEC). METHODS: Tumor biopsy samples obtained before P-FEC from 141 patients with stages II-III breast cancer including the luminal (n = 76), luminal-HER2 (n = 13), HER2 (n = 17), and triple-negative (TNBC, n = 35) subtypes were subjected to assay for HRD score using the OncoScan CNV FFPE Assay Kit. HRD score was a simple sum of NtAI, LOH, and LST (cutoff, 42). TNBCs were also subjected to the gene expression assay using the Affymetrix microarray (U133 plus 2.0) and to the BRCA1 promoter methylation assay using the methylation-specific real-time PCR. RESULTS: Of the 141 breast tumors, 45 samples (32%) had high HRD scores and were associated with high histological grade (P = 0.001), negative progesterone receptor (P = 0.018), high Ki67 index (P = 0.032), and BRCA1 promoter methylation (P = 3.6e-07). The proportion of tumors with high HRD scores was significantly higher in the TNBC subtype than the others (P = 0.006). In the TNBC subtype, but not the others, high HRD scores were significantly (P = 0.001) associated with a low pathological complete response rate to P-FEC. Among the molecular TNBC subtypes, a majority of tumors belonging to the basal-like 1, immunomodulatory, mesenchymal, mesenchymal stem-like, but not luminal androgen receptor (LAR), subtypes had high HRD scores. CONCLUSIONS: Approximately one-third of sporadic breast tumors show a high HRD score, indicating the presence of homologous recombination dysfunction, and they are characterized by biologically aggressive phenotypes, most commonly in the TNBC subtype, and less sensitive to P-FEC.

PAM50 for prediction of response to neoadjuvant chemotherapy for ER-positive breast cancer.

PURPOSE: There is an urgent need for the development of a predictor of response to chemotherapy for ER-positive breast cancer which is less chemosensitive than for ER-negative breast cancer in order to avoid unnecessary chemotherapy. In the present study, intrinsic subtyping by PAM50 was evaluated for its ability to predict a response to chemotherapy. PATIENTS AND METHODS: For this study, 124 patients with ER-positive breast cancer treated with neoadjuvant sequential paclitaxel and FEC (NAC) were evaluated. Tumor biopsy specimens obtained before NAC were subjected to intrinsic subtyping (IS) by gene expression (GE) using PAM50 (PAM50-IS) or immunohistochemistry (IHC-IS). RESULTS: Of the PAM50-ISs (Luminal A, Luminal B, HER2-enriched, and Basal-like), GE-Luminal A showed the lowest pCR rate (1.9%), and multivariate analysis revealed that GE-Luminal A was a significant (P = 0.031) predictor of non-pCR independently of other clinicopathological parameters, including Ki67, and tumor-infiltrating lymphocytes. Of the IHC-ISs, on the other hand, IHC-Luminal A was not significantly associated with pCR. We also found that breast tumors with low ER levels (1-9%), like ER-negative tumors, were mostly GE-HER2-enriched and GE-Basal-like, and more sensitive to NAC than those with high ER levels (≥ 10%). CONCLUSIONS: GE-Luminal A intrinsically subtyped by PAM50 was the least sensitive to NAC and very unlikely to attain pCR. IHC-Luminal A identified by IHC, on the other hand, was not significantly predictive of pCR. In addition, PAM50 revealed that tumors with low ER (1-9%) were more like ER-negative tumors than ER-positive tumors, and most such cases should therefore would better be treated with chemotherapy.

Mutational Analysis of AKT1 and PIK3CA in Intraductal Papillomas of the Breast with Special Reference to Cellular Components.

The pathologic feature of intraductal papillomas is defined as a papillary structure composed of a fibrovascular stromal core lined by luminal epithelial cells and myoepithelial cells. We used droplet digital PCR for the mutational analysis of AKT1 (E17K) and PIK3CA (H1047R, E542K, and E545K) in 60 papillomas. AKT1 and PIK3CA mutations were detected in 12 (20%) and 17 (28%) of the papillomas, respectively. In five tumors harboring mutations, mutational analysis of AKT1 or PIK3CA was performed separately using luminal epithelial cells and myoepithelial cells sorted using anti-cytokeratin 19 antibody and anti-α smooth muscle actin antibody. The two types of cells from a given papilloma had the identical mutation. Three patients with the PIK3CA mutation-positive papilloma developed breast cancers at the resection site of the papilloma, but none of these subsequent breast cancers had the PIK3CA mutation. These results indicate that a papilloma stems from a bipotent progenitor cell that contains the AKT1 or PIK3CA mutation and proliferates and differentiates to form the papilloma. Papilloma can be a risk factor for developing breast cancer but is unlikely to be its obligate precursor.

Construction of a novel multi-gene assay (42-gene classifier) for prediction of late recurrence in ER-positive breast cancer patients.

PURPOSE: Prediction models for late (> 5 years) recurrence in ER-positive breast cancer need to be developed for the accurate selection of patients for extended hormonal therapy. We attempted to develop such a prediction model focusing on the differences in gene expression between breast cancers with early and late recurrence. METHODS: For the training set, 779 ER-positive breast cancers treated with tamoxifen alone for 5 years were selected from the databases (GSE6532, GSE12093, GSE17705, and GSE26971). For the validation set, 221 ER-positive breast cancers treated with adjuvant hormonal therapy for 5 years with or without chemotherapy at our hospital were included. Gene expression was assayed by DNA microarray analysis (Affymetrix U133 plus 2.0). RESULTS: With the 42 genes differentially expressed in early and late recurrence breast cancers in the training set, a prediction model (42GC) for late recurrence was constructed. The patients classified by 42GC into the late recurrence-like group showed a significantly (P = 0.006) higher late recurrence rate as expected but a significantly (P = 1.62 × E-13) lower rate for early recurrence than non-late recurrence-like group. These observations were confirmed for the validation set, i.e., P = 0.020 for late recurrence and P = 5.70 × E-5 for early recurrence. CONCLUSION: We developed a unique prediction model (42GC) for late recurrence by focusing on the biological differences between breast cancers with early and late recurrence. Interestingly, patients in the late recurrence-like group by 42GC were at low risk for early recurrence.

Highly sensitive detection of ESR1 mutations in cell-free DNA from patients with metastatic breast cancer using molecular barcode sequencing.

PURPOSE: We aimed to develop a highly sensitive method to detect ESR1 mutations in cell-free DNA (cfDNA) using next-generation sequencing with molecular barcode (MB-NGS) targeting the hotspot segment (c.1600-1713). METHODS: The sensitivity of MB-NGS was tested using serially diluted ESR1 mutant DNA and then cfDNA samples from 34 patients with metastatic breast cancer were analyzed with MB-NGS. The results of MB-NGS were validated in comparison with conventional NGS and droplet digital PCR (ddPCR). RESULTS: MB-NGS showed a higher sensitivity (0.1%) than NGS without barcode (1%) by reducing background errors. Of the cfDNA samples from 34 patients with metastatic breast cancer, NGS without barcode revealed seven mutations in six patients (17.6%) and MB-NGS revealed six additional mutations including three mutations not reported in the COSMIC database of breast cancer, resulting in total 13 ESR1 mutations in ten patients (29.4%). Regarding the three hotspot mutations, all the patients with mutations detected by MB-NGS had identical mutations detected by droplet digital PCR (ddPCR), and mutant allele frequency correlated very well between both (r = 0.850, p < 0.01). Moreover, all the patients without these mutations by MB-NGS were found to have no mutations by ddPCR. CONCLUSION: In conclusion, MB-NGS could successfully detect ESR1 mutations in cfDNA with a higher sensitivity of 0.1% than conventional NGS and was considered as clinically useful as ddPCR.

Endocrine sensitivity of estrogen receptor-positive breast cancer is negatively correlated with aspartate-ß-hydroxylase expression.

Although prognostic markers for early estrogen receptor (ER)-positive breast cancer have been extensively developed, predictive markers for adjuvant endocrine therapy are still lacking. Focusing on the mechanisms underlying endocrine resistance, we investigated whether the endocrine sensitivity of ER-positive breast cancer cells was correlated with the expression of aspartate-ß-hydroxylase (ASPH), which is involved in the development of hepatocellular carcinoma. ASPH expression in ER-positive and tamoxifen-resistant breast cancer cells was upregulated by the MAPK and phosphoinositide-3 kinase (PI3K) pathways, which both play pivotal roles in endocrine resistance. In the clinical setting, ASPH expression was negatively correlated with recurrence-free survival of luminal B breast cancer patients that received adjuvant endocrine therapy, but not in patients that did not receive adjuvant endocrine therapy. Luminal B breast cancer is one of the intrinsic molecular subtypes identified by the Prediction Analysis of Microarray 50 (PAM50) multiple gene classifier, and because of its poor response to endocrine therapy, chemotherapy in addition to endocrine therapy is generally required after surgical resection. Our results suggest that the endocrine sensitivity of luminal B breast cancer can be assessed by examining ASPH expression, which promotes the consideration of a prospective study on the association between ASPH expression at the mRNA and protein levels in luminal B breast cancer and subsequent response to endocrine therapy.

Detection of ESR1 mutations in plasma and tumors from metastatic breast cancer patients using next-generation sequencing.

PURPOSE: Liquid biopsy using digital PCR (dPCR) has been widely used for the screening of ESR1 mutations, since they are frequently identified in the hotspot. However, dPCR is limited to the known mutations. Therefore, we aimed to analyze the utility of next-generation sequencing (NGS) to discover novel ESR1 mutations. METHODS: Whole exon sequencing of the ESR1 gene using NGS was performed in 16 primary and 47 recurrent tumor samples and 38 plasma samples from hormone receptor-positive metastatic breast cancer patients. Functional analyses were then performed for the novel mutations we detected. RESULTS: We identified no mutations in primary tumors and six mutations in five recurrent tumors, including three types of known mutations (Y537C, Y537N, and D538G) and two novel mutations (E279V and G557R). We also identified seven mutations in five plasma samples, including three types of known mutations (S463P, Y537S, and D538G) and one mutation not reported in COSMIC database (L536H). All nine patients with ESR1 mutations were treated with aromatase inhibitors (AIs) prior to sampling, and the mutations were frequently detected in patients who received AI treatments in the metastatic setting. Among the three novel mutations (E279V, L536H, and G557R), L536H, but not E279V and G557R, showed ligand-independent activity. All three mutant proteins showed nuclear localization and had no relation with non-genomic ER pathways. CONCLUSIONS: Although the molecular mechanisms of the E279V and G557R mutations remain unclear, our data suggest the utility of NGS as a liquid biopsy for metastatic breast cancer patients and the potential to identify novel ESR1 mutations.

A small-molecule inhibitor of SMAD3 attenuates resistance to anti-HER2 drugs in HER2-positive breast cancer cells.

PURPOSE: Resistance against anti-HER2 drugs in HER2-positive breast cancer is a major obstacle to the improving prognosis. Transforming growth factor ß (TGFß) is a cytokine involved in the acquisition of more malignant phenotypes through epithelial-mesenchymal transition (EMT) and cancer stem cell (CSC) properties. The aim of this study was to investigate the effects of TGFß and its downstream SMAD pathway on resistance to anti-HER2 drugs. METHODS: HER2-positive breast cancer cell lines were stimulated with TGFß for 14 days. Then, the sensitivity to trastuzumab and lapatinib and the expression levels of various EMT and CSC markers were examined. The correlation of nuclear SMAD3 expression in untreated breast tumor tissues with trastuzumab efficacy in neoadjuvant settings was examined. The effect of a small-molecule inhibitor of SMAD3 (SIS3) on resistance to anti-HER2 drugs was explored. RESULTS: We found that continuous activation of the TGFß-SMAD3 pathway induced resistance to anti-HER2 drugs and CSC traits in HER2-positive breast cancer cells. The induction of drug resistance by TGFß required strong activation of SMAD3. In fact, activated SMAD3 regulated multiple genes that harbor SMAD-binding elements and are involved in trastuzumab resistance. Nuclear SMAD3 expression in tumor tissue was inversely correlated with sensitivity to neoadjuvant treatment with trastuzumab. SIS3 not only prevented the acquisition of resistance to anti-HER2 drugs but also restored trastuzumab sensitivity in trastuzumab-resistant cells. CONCLUSIONS: This study indicates that the TGFß-SMAD3 pathway plays an important role in the induction and maintenance of resistance to anti-HER2 drugs. Thus, SMAD3 is a potential therapeutic target that can inhibit resistance and restore sensitivity to anti-HER2 drugs.

Protective effect of naturally occurring anti-HER2 autoantibodies on breast cancer.

Anti-HER2-autoantibodies (HER2-AAbs) are found in breast cancer patients as well as healthy individuals. However, the clinical relevance of the antibodies is unknown. We established an enzyme-linked immunosorbent assay with high sensitivity and quantified serum HER2-AAbs in 100 healthy women, 100 untreated patients with ductal carcinoma in situ (DCIS), and 500 untreated patients with invasive breast carcinoma (IBC). The associations between the levels of HER2-AAbs and breast cancer risk, and recurrence-free survival, were examined. High levels of HER2-AAbs were significantly associated with a reduced risk of DCIS (odds ratio [OR] 0.19, P = 4.6 × 10(-7)) or IBC (OR 0.31, P = 3.7 × 10(-7)). Subgroup analysis of IBC revealed a stronger association of HER2-AAbs with a reduced risk of the hormone receptor (HR)(-)/HER2(+) subtype (OR 0.12) than the other subtypes (HR(+)/HER2(-) [OR = 0.32], HR(+)/HER2(+) [OR 0.38], and HR(-)/HER2(-) [OR 0.29]). When we set the cutoff of HER2-AAbs at 20 ng/mL, recurrence-free survival of HER2-AAb-positive patients (N = 74) was significantly better than that of HER2-AAb-negative patients (N = 426) (P = 0.015). Univariate and multivariate analyses demonstrated that HER2-AAbs, as well as histological grade, were independently and significantly (P = 0.0065 and 0.049, respectively) associated with recurrence-free survival. Our exploratory study suggests a protective effect of naturally occurring HER2-AAbs on the development of primary and recurrent breast cancer. Further studies on HER2-AAbs are warranted.

One-Step Nucleic Acid Amplification Assay for Detection of Axillary Lymph Node Metastases in Breast Cancer Patients Treated with Neoadjuvant Chemotherapy.

BACKGROUND: The accuracy of one-step nucleic acid amplification (OSNA) for the detection of lymph node (LN) metastasis in breast cancer patients has been well established. This study aimed to evaluate its accuracy for patients treated with neoadjuvant chemotherapy (NAC). METHODS: For this study, 300 LNs, 115 sentinel lymph nodes (SLNs), and 185 non-SLNs from 88 breast cancer patients treated with NAC were examined by means of histology (hematoxylin and eosin staining and pancytokeratin immunostaining) and OSNA. RESULTS: The diagnostic accuracy, sensitivity, and specificity of OSNA were respectively 92.3, 88.5, and 93.3 % for all LNs, and the corresponding values were 87.8, 75.0, and 91.2 % for SLNs and 95.1, 97.3, and 94.6 % for non-SLNs. The diagnostic accuracy of OSNA was significantly lower for SLNs than for non-SLNs (P = 0.021), which was attributable to the low sensitivity for detection of micrometastases (micromets) due to lower CK19 mRNA expression detected by in situ hybridization (ISH) in SLN micromets than in non-SLN micromets. For primary breast tumors, CK19 mRNA expression showed a significant reduction after NAC (P = 0.040). CONCLUSIONS: The diagnostic accuracy of OSNA for NAC-treated patients is similar to that for NAC-nontreated patients, but its accuracy is significantly lower for SLNs than for non-SLNs. The findings obtained with CK19 mRNA ISH suggest that most SLN micromets cannot be detected by OSNA due to the reduced expression of CK19 mRNA induced by NAC.

Mutational analysis of MED12 in fibroadenomas and phyllodes tumors of the breast by means of targeted next-generation sequencing.

We aimed to analyze MED12 mutation in fibroadenomas (FAs) and phyllodes tumors (PTs) of the breast, which are closely related and consist of epithelial and stromal components. Targeted deep-sequencing using next-generation sequencing was performed in FAs (n = 58) and PTs (n = 27). The frequency of MED12 mutant tumors was significantly higher (P = 0.016) in PTs (74.1 %) than in FAs. (46.6 %). As for FAs, this frequency was significantly higher (P = 0.001) for intracanalicular type (69.0 %) than for other histological subtypes such as pericanalicular, organoid, and mastopathic types (24.1 %). Laser microdissection study revealed that stromal cells, but not epithelial cells, harbored MED12 mutations in both FAs and PTs. MED12 mutation is implicated in the pathogenesis of both FAs and PTs. The similarly high frequency of MED12 mutation in intracanalicular type FAs suggests that they are most closely related to PTs. It is thus speculated that FAs with MED12 mutation are more likely to progress to PTs.

PIK3CA mutations in serum DNA are predictive of recurrence in primary breast cancer patients.

We attempted to develop a highly sensitive and specific method for the detection of circulating tumor DNA (ctDNA) using a digital PCR (dPCR) assay for PIK3CA mutations (i.e., H1047R, E545K, and E542K) in primary breast cancer patients. The sensitivity of the dPCR assay for the mutant alleles was examined using cell lines with PIK3CA mutations and proved to be 0.01 %. Serum samples were collected pre-operatively from 313 stage I-III breast cancer patients, of whom 110 were found to have PIK3CA mutant tumors. The serum samples from these patients with PIK3CA mutant tumors were subjected to the dPCR assay, and 25 (22.7 %) were found to be positive. No PIK3CA mutant ctDNA was detected in the serum samples of 50 healthy women and 30 breast cancer patients with PIK3CA non-mutant tumors. The patients with PIK3CA mutant ctDNA were dichotomized into mutant ctDNA-high (ctDNA(high)) and ctDNA-low (ctDNA(low)) groups based on the median. The ctDNA(high) patients exhibited significantly shorter recurrence-free survival (RFS; P = 0.0002) and overall survival rates (OS; P = 0.0048) compared to those exhibited by the combined ctDNA(low) patient and ctDNA-free patient group. Multivariate analysis revealed that ctDNA(high) status significantly predicted poor RFS and OS and did so independently of conventional histological parameters. These results suggest that dPCR is a highly sensitive and specific method for the detection of PIK3CA mutant ctDNA and that ctDNA(high) but not ctDNA(low) status is a significant and independent prognostic factor for primary breast cancer patients.

BRCA1 promoter methylation of normal breast epithelial cells as a possible precursor for BRCA1-methylated breast cancer.

The breast cancer susceptibility gene 1 (BRCA1) and glutathione S-transferase P1 (GSTP1) promoters are reportedly often methylated in breast cancer tissues. Their methylation status in surrounding normal breast tissues has not been examined thoroughly although this may well be important for a better understanding of breast carcinogenesis. In this study, BRCA1 and GSTP1 promoter methylation was examined by methylation-specific PCR assay. Patients with BRCA1-methylated (n = 15) or BRCA1-unmethylated (n = 15) tumors and those with GSTP1-methylated (n = 9) or GSTP1-unmethylated (n = 11) tumors were included in the present study. Methylation status of manually micro-dissected normal epithelial cells from the formalin-fixed paraffin-embedded sections of normal breast tissues adjacent to and distant from the tumors was examined at multiple sites (n = 1-5). Of the 15 patients with BRCA1-methylated tumors, 9 harbored BRCA1 promoter methylation in at least one site of the normal breast tissues. However, no BRCA1 promoter methylation was observed at any site of the normal tissues of the 15 patients with BRCA1-unmethylated tumors. No GSTP1 promoter methylation was observed in the normal tissues regardless of the methylation status of the tumors. The presence of BRCA1 promoter methylation in the normal tissues was confirmed in the epithelial cells enriched with the magnetic-activated cell sorting method. Our findings suggest that a small proportion of normal breast epithelial cells with BRCA1 promoter methylation can be precursor cells from which BRCA1-methylated breast tumors may originate. This does not apply to GSTP1 promoter methylation.

Comparison of efficacy of 95-gene and 21-gene classifier (Oncotype DX) for prediction of recurrence in ER-positive and node-negative breast cancer patients.

We recently developed a 95-gene classifier (95(GC)) for the prognostic prediction for ER-positive and node-negative breast cancer patients treated with only adjuvant hormonal therapy. The aim of this study was to validate the efficacy of 95(GC) and compare it with that of 21(GC) (Oncotype DX) as well as to evaluate the combination of 95(GC) and 21(GC). DNA microarray data (gene expression) of ER-positive and node-negative breast cancer patients (n = 459) treated with adjuvant hormone therapy alone as well as those of ER-positive breast cancer patients treated with neoadjuvant chemotherapy (n = 359) were classified with 95(GC) and 21(GC) (Recurrence Online at http://www.recurrenceonline.com/ ). 95(GC) classified the 459 patients into low-risk (n = 285; 10 year relapse-free survival: 88.8 %) and high-risk groups (n = 174; 70.6 %) (P = 5.5e-10), and 21(GC) into low-risk group (n = 286; 89.3 %), intermediate-risk (n = 81; 75.7 %), and high-risk (n = 92; 64.7 %) groups (P = 2.9e-10). The combination of 95(GC) and 21(GC) classified them into low-risk (n = 324; 88.9 %) and high-risk (n = 135; 65.0 %) groups (P = 5.9e-14), and also showed that pathological complete response rates were significantly (P = 2.5e-6) higher for the high-risk (17.9 %) than the low-risk group (3.6 %). In addition, we demonstrated that 95(GC) was calculated on a single-sample basis if the reference robust multi-array average workflow was used for normalization. The prognostic prediction capability of 95(GC) appears to be comparable to that of 21(GC). Moreover, their combination seems to result in the identification of more low-risk patients who do not need chemotherapy than either classification alone. The patients in the high-risk group were found to be more chemo-sensitive so that they can benefit more from adjuvant chemotherapy.

Epigenetic regulation of cancer stem cell genes in triple-negative breast cancer.

Expression of specific breast cancer stem cells (BCSCs) is seen in aggressive tumors, but their regulation is unclear. Epigenetic changes influence gene expression and are implicated in breast cancer progression. We hypothesized that promoter methylation regulates specific BCSC-related genes [CD44, CD133, CD24, MSH1 (alias, Musashi-1), and ALDH1] and that this epigenetic profile can identify aggressive subtypes, such as triple-negative breast cancer (TNBC). Methylation analysis was performed using MassARRAY EpiTYPER sequencing; CpG-rich sites were identified in the promoter regions of BCSC genes, except ALDH1. These sites were screened by treatment with 5-aza-2'-deoxycytidine in four TN and five non-TNBC cell lines. The specific regulatory CpG site demonstrating the most significant inverse correlation between CpG site methylation and mRNA expression was identified for CD44, CD133, and Musashi-1, but not for CD24. Methylation of CD44, CD133, and Musashi-1 was evaluated in 91 American Joint Committee on Cancer stage I to III primary breast cancer tumors, and these sites were significantly hypomethylated in TNBC versus non-TNBC. The IHC staining of primary tumors with the highest and lowest methylation levels revealed the strongest staining in hypomethylated specimens, suggesting that hypomethylation leads to gene activation. We demonstrate that methylation is a significant mechanism regulating CD44, CD133, and Musashi-1, and that gene hypomethylation correlates with TNBC. Assessment of epigenetic changes in BCSC genes may provide a more accurate classification of TNBC and could be developed as potential therapeutic targets.

Clinicopathological analysis of breast ductal carcinoma in situ with ALDH1-positive cancer stem cells.

OBJECTIVE: The aim of this study was to elucidate the clinicopathological characteristics of breast ductal carcinomas in situ (DCIS) with aldehyde dehydrogenase 1 (ALDH1)-positive cancer stem cells (CSCs). METHODS: DCIS (n = 194) were subjected to immunohistochemical staining and results were examined for associations with various clinicopathological parameters. The ALDEFLUOR assay for breast cancer cell lines was also performed to determine the proportion of CSCs according to intrinsic subtype. RESULTS: DCIS with ALDH1-positive CSCs were significantly (p < 0.05) more likely to be estrogen receptor-α (ER)-negative, progesterone receptor (PgR)-negative, Ki67-positive and HER2-positive tumors. Luminal A subtype (8.6%) showed a significantly (p < 0.001) lower ALDH1 positivity than the other subtypes [luminal B (50.0%), luminal HER2 (36.8%), HER2 (35.3%) and triple-negative subtype (26.7%)]. Double immunostaining revealed that ALDH1-positive CSCs did not overlap with ER-, PgR- or Ki67-positive tumor cells but did overlap with HER2-positive tumor cells. The percentage of ALDEFLUOR-positive CSCs was lower in the luminal A cell lines (0.02% for T-47D and 0.4% for MCF7) than in the luminal HER2 (9.1% for BT-474 and 9.5% for MDA-MB-361), HER2 (13.9% for AU565 and 33.2% for SK-BR-3) and triple-negative cell lines (28.4% for MDA-MB-231 and 30.7% for MDA-MB-468). CONCLUSIONS: Our results suggest that most CSCs in DCIS are in the G0 phase (Ki67 negative) and do not express ER or PgR, but they do express HER2 in HER2-positive tumors.

Detection of Ultra-Rare ESR1 Mutations in Primary Breast Cancer Using LNA-Clamp ddPCR.

ESR1 mutations in breast cancer are one of the mechanisms of resistance to aromatase inhibitors. These mutations are common in metastatic breast cancer; however, these are rare in primary breast cancer. However, these data have been analyzed mainly in formalin-fixed, paraffin-embedded tissue; thus, rare mutations that may be present in primary breast cancer may be overlooked. In this study, we developed a highly sensitive mutation detection method called locked nucleic acid (LNA)-clamp droplet digital PCR (ddPCR) and validated it. The mutation detection sensitivity was substantiated to 0.003%. Then, we used this method to analyze ESR1 mutations in fresh-frozen (FF) tissues of primary breast cancer. cDNA extracted from the FF tissues of 212 patients with primary breast cancers were measured. Twenty-eight ESR1 mutations were found in twenty-seven (12.7%) patients. Sixteen (7.5%) patients had Y537S mutations and twelve (5.7%) had D538G mutations. Two mutations with a variant allele frequency (VAF) of ≥0.1% and twenty-six mutations with a VAF of <0.1% were found. By using this LNA-clamp ddPCR, this study demonstrated the presence of minor clones with a VAF of <0.1% in primary breast cancer.