Triple negative breast cancer: outcomes, subtypes and implications for treatment
Walsh, Elaine M.
Walsh, Elaine M.
Loading...
Identifiers
http://hdl.handle.net/10379/15495
https://doi.org/10.13025/16810
https://doi.org/10.13025/16810
Repository DOI
Publication Date
2019-11-30
Type
Thesis
Downloads
Citation
Abstract
Triple negative breast cancer (TNBC) is a heterogeneous subtype representing 15-20% of all breast cancers and is associated with a poor prognosis, despite the advances made in other breast cancer subtypes. TNBC is defined by the absence of the predictive targets oestrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2). Chemotherapy is the mainstay of treatment, as the use of targeted therapy has not improved outcome for this subtype of breast cancer. TNBCs are more common in patients with germline breast cancer 1 gene (BRCA1) mutations and therefore the role of DNA damaging agents has been explored as a therapeutic option for TNBCs. Anthracycline-taxane-based combination chemotherapy is currently considered the optimal regimen for patients with TNBC, but platinum-based neoadjuvant chemotherapy (NACT) has been shown to be associated with improved rates of pathological complete response (pCR) and improved survival in TNBC. The aims of this body of work are to examine the benefit of carboplatin-based NACT, with respect to pCR rates and outcomes for patients diagnosed with TNBC at University Hospital Galway (UHG) over a 16-year period; to determine if the presence of modifiable and non-modifiable TNBC risk factors can be predictive of response to NACT or associated with patient outcomes; to identify predictive and prognostic biomarkers in TNBC, with a focus on predicting TNBC subtypes most likely to respond to carboplatin-based chemotherapy; and to investigate if TNBC subtypes demonstrate different innate drug sensitivity patterns with respect to chemotherapy agents. A database of 355 patients with TNBC was constructed. Extensive details pertaining to clinical and pathological features, as well as outcome data were recorded. Forty-seven percent of all patients received adjuvant systemic chemotherapy and 27% received NACT. Among patients treated with NACT, a pCR breast/axilla was observed in 38% of patients. A pCR breast/axilla was more common in patients who received carboplatin-based NACT (55%) compared to those treated with anthracycline-taxane combination chemotherapy (28%) (p=0.011) By multivariable analysis, using non-pCR as the baseline parameter, carboplatin use was associated with an 83% increased incidence of pCR breast/axilla (p=0.002). Grade 3 histology was also independently associated with pCR and was associated with a 95% increased incidence of pCR breast/axilla (p<0.001). After a follow-up of 30 months (range 5-126 months), pCR breast/axilla was found to be an independent predictor of survival, with respect to disease free survival (DFS), metastasis free survival (MFS) and breast cancer specific survival (BCSS). Modifiable risk factors for TNBC were studied: 60% of patients were classified as overweight and 28% were obese. Body mass index (BMI) was significantly associated with survival. Increasing BMI was associated with improved DFS (p=0.05), MFS (p=0.036) and BCSS (p=0.012) by univariate analysis, an effect which was most pronounced in post-menopausal patients. By multivariable analysis, increasing BMI was associated with improved BCSS: post-menopausal patients with increased BMI were 61% less likely to die of breast cancer compared to patients with normal BMI (p=0.045). Stromal tumour infiltrating lymphocytes (sTILs) and androgen receptor (AR) were evaluated as potential predictive and prognostic biomarkers. Among patients treated with NACT, sTILs, tumour grade and the use of carboplatin were found to be predictive for pCR by multivariable analysis using non-pCR as the baseline parameter. The likelihood of a pCR increased as sTILs increased: sTILs ≥50% increased the likelihood of a pCR breast/axilla by 86% (p=0.030). This effect was most pronounced among patients who did not receive a carboplatin-based regimen where sTILs >10% increased the likelihood of a pCR breast/axilla by 85% (p=0.009); and sTILs ≥50% increased the likelihood of a pCR breast/axilla by 94% (p=0.014). The drug sensitivity patterns of four TNBC cell lines were studied in vitro with chemotherapy drugs used in everyday clinical practice in both two dimensional (2-D) and three dimensional (3-D) models, in order to study the differences between culture models and to investigate drug sensitivity patterns among TNBC subtypes. This work demonstrated different chemotherapy responses among TNBC subtypes. Both the basal-like 1 (BL1) and the luminal androgen receptor (LAR) subtypes were sensitive to carboplatin, docetaxel, paclitaxel and doxorubicin, at small doses and as single agents. The basal-like 2 (BL2) subtype was sensitive only to a three-drug regimen of carboplatin, paclitaxel and cetuximab. The mesenchymal stem-like (MSL) subtype was only sensitive to high dose combination chemotherapy with carboplatin and paclitaxel. The results of the work presented in this thesis add to the body of evidence showing improved rates of pCR with the incorporation of carboplatin in the neoadjuvant regimen for patients with TNBC. This work also supports the role of sTILs as a predictive biomarker in TNBC, and suggests that carboplatin-based chemotherapy could be reserved for low sTIL tumours. Prognostic candidate risk factors have been identified that require further evaluation: this study demonstrates that BMI has important prognostic value in post-menopausal patients. In addition, this thesis has shown that responses to chemotherapy vary by both TNBC subtype and 2-D versus 3-D modelling in vitro, and has identified potential alternative approaches for chemoresistant TNBCs, with novel drug combinations and high dose therapies.
Publisher
NUI Galway
Publisher DOI
Rights
Attribution-NonCommercial-NoDerivs 3.0 Ireland