**3. Cell of origin**

However, most investigators use the presence (+) or absence (-) of immunohistochemical (IHS) markers or receptor combinations that are expressed by neoplastic cells. In this manner, distinct tumor categories to-date are identified by the expression of estrogen receptors (ER), progesterone receptors (PR), human epidermal growth factor receptors (EGFR) 1 and 2 (Her 1 and 2) and cytokeratin 5, 6 (CK5,6) [15-19, 23-25]. Generally, molecular sub-types correspond to IHC receptor status [26]. These subtypes are of great clinical and research importance as they are utilized to administer and target therapeutic regimes based on predictions of response. Furthermore, these subtypes have been shown to display a wide variety of responses to different treatments [27-30] and are associated with other clinical outcomes, such as patient relapse and overall survival. The most favorable outcomes are noted for the luminal A subtype, which are hormone sensitive [31]. The Her2+ and basal subtypes are noted as more aggressive and have fewer therapeutic options [31-33]. The normal-like and claudin-low are unclassified (negative for all major receptors), and associated with poor prognosis [20-22, 30, 34]. Several small studies support the concept that molecular subtype and tumor receptor status may change during ordinary disease progression and a major study completed by MacFarlane *et al*. in 2008 revealed that 21% of relapsed tumors had changes in either ER/PR or HER2 receptor status [35]. This significant proportion led the author to suggest that biopsies of relapsed/ metastatic breast cancers should be performed routinely. This also should be an important consideration in research. In summary, treatment options considered effective in primary stage cancer may no longer be optimal in later stages of the disease, including relapses and metastasis

Other breast cancer classification approaches are also used to assist in both prognostic and treatment decisions. These include computer models that are based on a combination of several factors and offer individual survival predictions and calculations of treatment benefits [36]. For example, patients undergoing systemic adjuvant therapy can determine optimal treatment through the commercially available computer model *Adjuvant* which has been successfully validated in several cohorts, including the United States and Canada [36, 37]. Other useful classification tools utilized for breast cancer treatment choices include prognostic assessments (such as USC/Van Nuys prognostic index (VNPI)) and general comorbidity assessments [38, 39]. Also consider the case of familial breast cancer (genetic classification) whereby a patient may opt to undergo preventative measures such as mastectomy. Additionally, immunohisto‐ chemistry testing, other than those mentioned earlier, continue to prove favorable as prog‐ nostic markers across various molecular subtypes [40]. For example, in human breast cancer epithelial cell proliferation is considered a significant prognostic marker [41] and could possibly be used as a prediction tool to measure different hormone treatment related risks [42]. Therefore, the immunohistochemical marker, Ki67 (a nuclear protein expressed by cells in all active phases of the cycle except for quiescent or resting cells) is utilized frequently to evaluate proliferation [43]. Such labeling indicates a significant association with higher carcinoma grade, clinical response to endocrine therapy, higher risk of relapse, and worse survival in

as determined by the current classification scheme.

348 Using Old Solutions to New Problems - Natural Drug Discovery in the 21st Century

**2.3. Other classification approaches**

patients with early breast cancer [44].

While scientists have predicted that specific breast cancer types may arise from different types of progenitor cells, it has been difficult to identify the cells of origin, and the topic remains controversial to date. Breast tumors represent a heterogeneous collection of cell populations with different biological properties [45]. Shared molecular features have made it difficult to distinguish the cell populations of breast cancer tumors and only recently have researchers been able to differentiate stem cells from other progenitor cells. Within this context, tumor evolution has been explained by two main theories. The traditional clonal evolution model is based on the premise that all tumor cells have the capacity to undergo self-renewal which is an indication of their potential to undergo tumor progression and drug resistance [46, 47]. The Cancer Stem Cell (CSC) theory emphasizes the ability of only a minor population of tumori‐ genic cells capable of self-renewal and differentiation [20, 48] and this specific sub-set of CSC's gives rise to new tumors which are phenotypically identical to the original tumors [49]. CSC's are believed to be a small population of cells with dysregulated self–renewal properties capable of continuous self-renewal and differentiation and responsible for tumor existence treatment resistance and relapse. Existence of CSC's in various types of tumours, including breast cancer [50],[51] has been identified. ALDH (aldehyde dehydrogenase)-1 is a marker of normal and malignant human mammary stem cells [52], and these cells can also be isolated using the cell surface markers epithelial-specific antigen (ESA), CD44 and the absence of the expression of CD24 [53]. When transformed cells undergo epithelial-mesenchymal transition (EMT), they have been noted to gain properties of stem cells [54]. Although evidence supporting the CSC model was initially obtained from acute myeloid leukemia [55], successive studies maintain that solid tumors, including breast cancer tumors, are also driven and sustained by CSCs [50].

Regardless of origin, an abundance of research has clearly confirmed the existence of cancer stem cells (CSCs) or tumor-initiating cells (TICs) in a variety of human cancers [55-58], including breast cancer [50]. Nonetheless, most of the therapeutic approaches available, inclusive of chemotherapy and radiation, lack the ability to effectively kill these populations [59-62]. This may explain the lack of progress in eliminating cancerous tumors and preventing metastasis and may help to rationalize therapeutic resistance; therefore, the CSC or TIC population has become a target for cancer prevention and therapy [63]. A general consensus exists in the literature that breast cancer re-occurrence is assumed to be caused by a subpopulation of tumor initiating cells possessing stem cell attributes of a tumor as well as resistance to chemotherapy, radiation and other forms of treatment [64-66]. Of great interest is the role of CSC's in tumor relapse and resistance to therapy and recent articles suggesting that such resistance can be overcome.
