Tamoxifen induces hypercoagulation and alterations in ERα and ERß dependent on breast cancer sub-phenotype ex vivo.

Tamoxifen shows efficacy in reducing breast cancer-related mortality but clinically, is associated with increased risk for thromboembolic events. We aimed to determine whether breast tumour sub-phenotype could predict propensity for thrombosis. We present two ex vivo Models of Tamoxifen-therapy, Model 1 in which treatment recapitulates accumulation within breast tissue, by treating MCF7 and T47D cells directly prior to exposure to blood constituents; and Model 2 in which we recreate circulating Tamoxifen by treating blood constituents prior to exposure to cancer cells. Blood constituents included whole blood, platelet-rich plasma and platelet-poor plasma. Hypercoagulation was assessed as a function of thrombin activity, expression of CD62P and CD63 activation markers defined as an index of platelet activation, and platelet morphology; while oestrogen receptor expression was assessed using immunocytochemistry with quantitative analysis. We determined, in concert with clinical studies and contrary to selected laboratory investigations, that Tamoxifen induces hypercoagulation, dependent on sub-phenotypes, with the T47D cell line capacity most enhanced. We determined a weak positive correlation between oestrogen receptor expression, and CD62P and CD63; indicating an association between tumour invasion profiles and hypercoagulation, however, other yet unknown factors may play a predictive role in defining hypercoagulation.

Ex vivo interaction between blood components and hormone-dependent breast cancer cells induces alterations associated with epithelial-mesenchymal transition and thrombosis.

The prevalence of breast cancer is steadily increasing with metastasis and thromboembolic complications identified as the most common causes of death. The acquisition of an aggressive phenotype by hormone-dependent breast cancers is mediated by Transforming Growth Factor Beta 1 (TGF-ß1) expression and is associated with epithelial-mesenchymal transition (EMT) and, potentially, increased propensity for thrombosis. We investigated this phenomenon by assessing the effect of platelet-rich plasma (PRP) and whole blood (WB) on parameters of EMT and hypercoagulation in vitro. MCF-7 breast cancer cells were cultured under standard conditions, followed by co-culture with PRP or WB. Cells were processed for real-time PCR (TGF-ß1 and vimentin), electron microscopy or immunocytochemistry (TGF-ß1). Micrographs were qualitatively assessed, and real-time PCR data analyzed with PAST Statistical Software. The addition of blood components heightened TGF-ß1 immunolocalization and significantly increased corresponding gene expression. Both PRP and WB significantly increased vimentin expression and induced a shape change from a typical epithelial phenotype to a spindle-shape morphology, indicative of EMT. Fibrin fiber, network and plaque formation indicated a hypercoagulatory environment. The results thus show that in preparation for hematogenous metastasis, hormone-dependent breast cancer cells assume an aggressive phenotype associated with EMT, simultaneously increasing the propensity for the formation of thrombo-emboli.

Breast cancer cell-induced platelet activation is compounded by tamoxifen and anastrozole in vitro.

Platelet-tumour cell interaction is implicated in the initiation of breast cancer-associated thrombosis, with hormone-therapy (Tamoxifen/Anastrozole), increasing this risk. However, recent in vitro research indicates that Tamoxifen inhibits platelet activation, while the effects of Anastrozole on platelet activation are not well characterised. This study investigated platelet activation caused by Tamoxifen or Anastrozole-treated breast cancer cells in vitro. MCF7 and T47D cells were pre-treated with Tamoxifen or Anastrozole to mimic the effects of the drugs in vivo, and co-cultured with whole blood. Platelet activation was determined using flow cytometry. Platelet (CD41a+CD62P+) was determined using an interval gating strategy. Platelet morphology was visualised using scanning electron microscopy. Our results support clinical findings, showing that hormone-therapy is associated with platelet activation. Tamoxifen-treated MCF7 cells increased P-selectin expression, with ultrastructural analysis showing fully spread platelets. Conversely, Tamoxifen-treated T47D cells decreased P-selectin expression with platelets showing signs of early aggregation. Anastrozole pre-treatment decreased P-selectin expression, with treated MCF7 cells inducing platelet membrane folds and lamellipodia extension, and treated T47D cells inducing platelet aggregation and fibrin network formation indicating hypercoagulation. The findings support clinical studies. Hormone-therapy augments tumour cell-induced platelet activation, which may be linked to cell phenotype. This may have clinical implications for treatment strategies.

p53 and Ha-ras mutations in chemically induced hamster buccal pouch carcinomas.

Well-differentiated squamous cell carcinomas were induced in hamster buccal pouch epithelium by twice weekly topical applications of N-methyl-N-benzylnitrosamine (MBN) or 7,12-dimethylbenz[a]anthracene (DMBA) over a period of 15 weeks. Each of the 22 tumors induced (14 MBN and eight DMBA) were evaluated by single-strand conformation polymorphism and DNA sequencing to identify mutations in conserved exons (E5-E8) of the p53 tumor suppressor gene and codons 12/13 and 61 of Ha-ras. In addition, Northern blot analysis of 10 MBN tumors and five DMBA tumors was performed to determine whether the mdm-2 gene was overexpressed, p53 mutations were detected in five of 14 (35%) MBN-induced carcinomas and in two of eight (25%) DMBA-induced carcinomas. Ha-ras mutations were detected in three of 14 (21%) MBN-induced carcinomas and in three of eight (37%) DMBA-induced carcinomas. One MBN-induced carcinoma exhibited a mutation in both the p53 and Ha-ras genes. The majority (five of seven of p53/Ha-ras mutations induced by MBN were G->A transitions and two of these occurred at hamster p53 codon 248, which corresponds to human p53 codon 245, a known mutational tumor 'hot spot'. A->T transversion at Ha-ras codon 61 accounted for three of five (60%) DMBA-induced mutations. There was no evidence of mdm-2 overexpression in any of the tumors evaluated. Overall, the results provide additional support for the validity of the hamster buccal pouch model of oral carcinogenesis, as applied to sequential cellular and molecular analysis and cancer chemoprevention studies.