Search

Histological special types of breast cancer have distinctive morphological features and account for up to 25 % of all invasive breast cancers. We sought to determine whether at the genomic level, histological special types of breast cancer are distinct from grade- and estrogen receptor (ER)-matched invasive carcinomas of no special type (IC-NSTs), and to define genes whose expression correlates with gene copy number in histological special types of breast cancer. We characterized 59 breast cancers of ten histological special types using array-based comparative genomic hybridization (aCGH). Hierarchical clustering revealed that the patterns of gene copy number aberrations segregated with ER-status and histological grade, and that samples from each of the breast cancer histological special types preferentially clustered together. We confirmed the patterns of gene copy number aberrations previously reported for lobular, micropapillary, metaplastic, and mucinous carcinomas. On the other hand, metaplastic and medullary carcinomas were found to have genomic profiles similar to those of grade- and ER-matched IC-NSTs. The genomic aberrations observed in invasive carcinomas with osteoclast-like stromal giant cells support its classification as IC-NST variant. Integrative aCGH and gene expression analysis led to the identification of 145 transcripts that were significantly overexpressed when amplified in histological special types of breast cancer. Our results illustrate that together with histological grade and ER-status, histological type is also associated with the patterns and complexity of gene copy number aberrations in breast cancer, with adenoid cystic and mucinous carcinomas being examples of ER-negative and ER-positive breast cancers with distinctive repertoires of gene copy number aberrations.

Oncocytic carcinomas are composed of mitochondrion-rich cells. Though recognised by the WHO classification as a histological special type of breast cancer, their status as a discrete pathological entity remains a matter of contention. Given that oncocytic tumours of other anatomical sites display distinct clinico-pathological and molecular features, we sought to define the molecular genetic features of mitochondrion-rich breast tumours and to compare them with a series of histological grade- and oestrogen receptor status-matched invasive ductal carcinomas of no special type. Seventeen mitochondrion-rich breast carcinomas, including nine bona fide oncocytic carcinomas, were profiled with antibodies against oestrogen, progesterone and androgen receptors, HER2, Ki67, GCDFP-15, chromogranin, epithelial membrane antigen, cytokeratin 7, cytokeratin 14, CD68 and mitochondria antigen. These tumours were microdissected and DNA extracted from samples with >70% of tumour cells. Fourteen cases yielded DNA of sufficient quality/quantity and were subjected to high-resolution microarray comparative genomic hybridisation analysis. The genomic profiles were compared to those of 28 grade- and oestrogen receptor status-matched invasive ductal carcinomas of no special type. Oncocytic and other mitochondrion-rich tumours did not differ significantly between themselves. As a group, mitochondrion-rich carcinomas were immunophenotypically heterogenous. Recurrent copy number changes were similar to those described in unselected breast cancers. However, unsupervised and supervised analysis identified a subset of mitochondrion-rich cancers, which often displayed gains of 11q13.1-q13.2 and 19p13. Changes in the latter two chromosomal regions have been shown to be associated with oncocytic tumours of the kidney and thyroid, respectively, and host several nuclear genes with specific mitochondrial function. Our results indicate that in a way akin to oncocytic tumours of other anatomical sites, at least a subset of mitochondrion-rich breast carcinomas may be underpinned by a distinct pattern of chromosomal changes potentially relevant for mitochondria accumulation and constitute a discrete molecular entity.

Caveolin-1 is the principal constituent protein of caveolae, which are specialised plasma membrane invaginations with diverse biological roles. Caveolin-1 is suggested to have tumour suppressive functions and CAV1 gene mutations have been reported in 20% of breast cancers. The aim of the present study was to evaluate the frequency of CAV1 mutations in a large cohort of optimally accrued breast cancers. Two independent series of breast cancer samples were analysed: 82 fresh-frozen grade 3 and 158 formalin-fixed paraffin-embedded invasive ductal carcinomas of no special type were consecutively accrued and subjected to microdissection of neoplastic epithelial cells prior to DNA extraction. Thirty-nine human breast cancer cell lines were also included in this study. The trans-membrane region of CAV1 and adjacent sequences, where mutations are reported to cluster, were amplified by PCR, followed by direct sequencing and mutational analysis. None of the reported CAV1 gene mutations, including CAV1 (P132L), were identified in either clinical samples (95% CI: 0-1.5%) or human breast cancer cell lines analysed. One novel non-synonymous germline polymorphism was detected within a reported region of high mutational frequency. This study does not corroborate the reported frequent occurrence of CAV1 gene mutations, including CAV1 (P132L), in primary human breast carcinomas. Our findings demonstrate that if CAV1 mutations do exist, their overall mutational frequency is substantially lower than positive reports have suggested. Taken together with other studies, which have also failed to identify CAV1 mutations, our data call into question the existence and biological and clinical relevance of CAV1 gene mutations in human breast cancer.

Amplification of 11q13 is found in approximately 15% of breast cancers. Cyclin D1 (CCND1) has been reported to be the 'driver' of this amplicon, however, multiple genes map to the smallest region of amplification of 11q13. Out of these genes, cortactin (CTTN) has been shown to be consistently overexpressed at the mRNA level in tumours harbouring 11q13 amplification. The aims of this study are to define whether CTTN is consistently co-amplified with the main core of the 11q13 amplicon, whether it is consistently overexpressed when amplified and to determine correlations between CTTN amplification and overexpression with clinicopathological features of breast cancers and survival of breast cancer patients. CTTN and CCND1 chromogenic in situ hybridisation (CISH) probes and a validated monoclonal antibody against CTTN were applied to a tissue microarray of a cohort of breast cancers from patients treated with anthracycline-based chemotherapy. CTTN and CCND1 amplifications were found in 12.3 and 12.4% of cases, respectively. All cases harbouring CTTN amplification also displayed CCND1 amplification. High expression of CTTN was found in 10.8% of cases and was associated with CTTN amplification, expression of 'basal' markers and topoisomerase IIα. Exploratory subgroup analysis of tumours devoid of 11q13 amplification revealed that high expression of CTTN in the absence of CTTN gene amplification was associated with lymph node negative disease, lack of hormone receptors and FOXA1, expression of 'basal' markers, high Ki-67 indices, p53 nuclear expression, and basal-like and triple negative phenotypes. CTTN expression and CTTN gene amplification were not associated with disease-, metastasis-free and overall survival. In conclusion, CTTN is consistently co-amplified with CCND1 and expressed at higher levels in breast cancers harbouring 11q13 amplification, suggesting that CTTN may also constitute one of the drivers of this amplicon. CTTN expression is not associated with the outcome of breast cancer patients treated with anthracycline-based chemotherapy.

Introduction: Despite strong evidence regarding the role of CCND1 amplification and protein overexpression in breast carcinoma, the associations between CCND1 amplification/cyclin D1 overexpression and clinicopathological variables and clinical outcome remain controversial., Aims of the Study: (1) to correlate cyclin D1 expression with gene amplification; (2) to analyse the correlations between CCND1 amplification and overexpression with clinicopathological features and patients' outcome in invasive breast cancer; (3) to define the prevalence and clinical significance of cyclin D1 overexpression and CCND1 amplification in ER positive breast carcinomas (4) to define the prevalence of cyclin D1 overexpression and CCND1 amplification in breast cancers with basal-like immunophenotype., Materials and Methods: CCND1 amplification and protein expression were assessed on a tissue microarray containing 880 unselected invasive breast cancer cases, by means of chromogenic in situ hybridisation using the Spotlight CCND1 amplification probe and immunohistochemistry, using the rabbit monoclonal antibody SP4., Results: A total of 59/613 tumours (9.6%) showed CCND1 amplification and 224/514 (43.6%) showed strong cyclin D1 expression. A strong positive correlation between CCND1 amplification and higher levels of cyclin D1 expression was found (P < 0.001). Basal-like cancers showed infrequent CCND1 amplification and cyclin D1 overexpression (P < 0.001). Both CCND1 amplification and cyclin D1 expression were associated with positive ER status. CCND1 gene amplification was an independent prognostic factor for patients with ER positive breast cancer., Conclusion: Our results demonstrate a strong correlation between CCND1 amplification and its protein expression in breast cancer. However, protein expression is more pervasive than gene amplification and associated with ER expression.

Background: TOP2A gene encodes topoisomerase II alpha, the direct molecular target of anthracyclines. This gene is frequently coamplified with HER2. The aims of this study were to analyse the pattern of TOP2A amplification and protein expression in relation to the molecular subgroups of breast cancers; and to define the impact of TOP2A amplification on the outcome of a series of patients homogeneously treated with adjuvant anthracyclines., Methods: A cohort of 245 patients with early breast cancer homogeneously treated with anthracyclines in the adjuvant setting was selected. A tissue microarray containing these cancers was used to determine HER2 and TOP2A gene copy number by means of chromogenic in situ hybridization. Immunohistochemical staining of topoisomerase II alpha was also performed using a monoclonal antibody (Ki-S1). TOP2A amplification and protein expression were correlated with classical prognostic parameters, expression of immunohistochemical markers and with a gene expression profiling classification using surrogate immunohistochemical markers. Kaplan-Meier method was used to construct survival curves and results were compared with log-rank test., Results: TOP2A amplification was restricted to tumours with HER2 amplification and was significantly associated with ER positivity. In the subgroup of patients with HER2 amplified tumours, TOP2A amplification predicted a better overall survival and disease free survival (P = 0.028 and 0.026, respectively). On multivariate analysis, TOP2A amplification maintained its predictive value for DFS., Conclusion: TOP2A amplification is likely to be a useful marker to predict the subset of patients who will benefit from anthracyclines.