1. Background The molecular profiles of cancer specimens are routinely determined as an essential supplement to the histological and immunological typing of tumours. The expression of EGFR (epidermal growth factor receptor) family receptors and ERs (estrogen receptors) and to a lesser extent of PGR (progesterone receptor) has aided the decision on patient management. However, it is uncertain how their expression might relate to the prediction of recurrence, a cryptic indication of silent dissemination. The presence of ER/PR may be associated with a lower risk of early metastasis. But the prediction of benefits of disease-free survival has remained controversial. Some uncertainty can be due to the lack of information concerning the specific ER isoform that has been studied, because ERα and ERβ have opposing effects on cell proliferation. Hence it is imperative that Ki-67 is also measured. Besides, the menopausal status of the patients can affect how reliable the early and late relapse predictions can be. Apart from ER, other factors have been adduced as possible predictive indicators of early or late recurrence. Some studies have advocated the use of cytokeratin 5 and MKI67 (marker of proliferation Ki-67) as valuable adjuncts to the use of the conventional steroid and HER2 to assist in predicting the likelihood of early or late recurrence. ISBN: 9780123878205 1,2 There is increasing realisation of the development of resistance therapies targeting ERBB2, which will influence possible recurrence. 3 Much effort has been expended into the analysis of the subtypes, Luminal A [ER+ PR+ HER2- low KI-67], Luminal B [ER+ PR+ HER2+/- high KI-67], Triple-negative/basal-like [ER- PR- HER2-] and ERBB2-over expressed [ER- PR- HER2+], but there is no consensus or consistency in the findings concerning the association of early or late recurrence with the subtypes. The basal-like breast cancer, which is a triple negative subtype, is characterised by high expression of cytokeratins and MKI67 is linked with higher risk of early recurrence. However, ERBB2 positivity carries the highest risk of early recurrence, whilst luminal A (HER2- low KI-67) and Luminal B (HER2+ and high KI-67), could be liable to late recurrence. 2-4 There is also the contrasting view that Luminal B tumours often show early local recurrence and even metastasis. 5ERBB2 positive breast cancer has poor prognosis since the carcinoma can progress to the metastatic stage even with anti-HER2 therapy. The resistance may be intrinsic or acquired. The intrinsic resistances can be due to genetic alterations such as ERBB2 gene mutation or amplification. 6 The intrinsic nature is also emphasised by the detection of ERBB2 positivity in breast cancer stem cells. 7 But this has to consider that CD44, which is often used as a marker of stem cells, might be contributing to the resistance without any link to the stemness. 8 It has been argued that resistance to treatment might occur in patients with constitutively activated ERBB2 signalling or resistance might be acquired by the secondary activation of ERBB2 signalling by the crosstalk with signalling activated by other biological response modifiers. The signalling by growth factor receptors, including ERBB2, functions by activation of downstream systems such as mTOR. 9,10 Overall, these findings partially repudiate the consistent claims of growth factor family receptors are total and reliable indicators of tumour progression. Several cancer antigens have been putatively associated the breast cancer as indicators of secondary spread and predictors of disease-free as well as overall survival of breast cancer patients. 11 Some cancer antigens are non-specific and may be expressed at significant levels in normal tissues. The cancer antigens CA 15-3 and CA 27.29 received much attention. CA 27.29 is a soluble form of MUC1. The latter is a high molecular weight transmembrane glycoprotein. CA 27.29 was thought to be a sensitive marker and specific for metastatic breast cancer. But this does not appear to consistently hold true. Increased expression of CA 27.29 often seemed to relate to tumour size. Increases in the antigen levels have been found in breast cancer patients who had no evidence of recurrence or metastatic disease. 12,13 Indeed, it is aberrantly expressed in several forms of cancer. This does not necessarily detract from its merit as a breast cancer marker. However, it is often also expressed in benign breast disease. The glycoprotein MUC1, which is down regulated with EMT activation, was suggested as a marker for early detection. However, it has been deemed as unsuitable on account of its low sensitivity and specificity. 14 Cell cycle regulators such as p53 and the E3 ubiquitin ligase MKI67 have been used to monitor the degree of apoptosis. They are associated with a self-limiting phase of growth that is subject to physiological constraints. ISBN: 9780123878205 The loss of cell cycle control is an important element in tumour growth, but it is uncertain if the expression of cell cycle regulators, even prominent ones such as TP53, has enough clinical merit. The TP53 is frequently mutated in breast cancer. It has been suggested that the presence of mutations might indicate by proxy favourable prognosis because that could render the tumours sensitive to chemotherapy with positive response of patients. However, some uncertainties have appeared concerning the clinical relevance of p53. There are reports that Luminal A breast cancer with TP53 mutations have better prognosis. 15 This is a biologically incongruous outcome, because wild type TP53 is said to able to reduce the expression of membrane associated proteins, such as fibronectin. That would be conducive to cell invasion. 16 The influence that TP53 might exert has been reported to depend upon the status of loss of PTEN that regulates AKT signalling. 17 This would implicate the AKT/mTOR in the TP53 predictable outcome. The cyclin dependent kinase p21waf1/cip1, which mediates p53 function, can induce cell cycle arrest or promote tumour growth by suppressing apoptosis. This is another important factor that has to be considered while using TP53 as a clinical marker. From this discussion, it follows that one has to employ markers that completely cover the important aspects of metastatic spread.2. Epithelial mesenchymal transition The major behavioural alterations in cancer cells of accelerated growth, invasion, induction of tumour associated and intratumoral angiogenesis, and metastatic disseminations are closely linked with the process of epithelial mesenchymal transition. A majority of cancers are of epithelial origin and hence the emphasis here is on the crucial activation of EMT. EMT is a developmental programme. But its activation is an essential element in the development, progression, and metastasis of cancer. ISBN 978-0-12-416570-0, ISBN 978-0-12-812896-1 Here I have attempted to focus on the markers that determine the intrinsic potential of tumours for invasion and metastasis. Inherent in EMT is the gain of the ability to invade, preparatory to metastasis. Many molecular targets of therapeutic value have been identified, together with elucidation of the signal transduction networks that they operate. 3. EMT associated factors EMT is characterised by the down regulation of epithelial markers such as CDH1, MUC1, integrins and laminin, and the appearance of mesenchymal markers such as vimentin. Together with these EMT activating transcription factors are also expressed. Several transcription factors have been closely identified with the activation of EMT. Snail/Slug, Twist, and Zeb-1 are most often cited. The expression of these transcription factors is up regulated in many forms of cancer. 18,19 Snail and Slug are zinc-finger transcription factors that control invasion by suppressing the expression of E-cadherin. This occurs in parallel with the up regulation of vimentin. The alterations in the expression of both of these have been closely linked with the activation of EMT [Figure 1]. 19-23 Figure 1 illustrates the regulation of the processes associated with cancer progression and metastasis by mTOR with the mediation of EMT.4. The mTOR signalling system The mechanistic target of rapamycin MTOR is a regulatory system that co-ordinates the function of several genes associated with cancer development, progression and metastasis. The mTOR is a kinase of the PI3K [phosphoinositide-3 kinase] related family of kinases. 24 It exerts its effects via two protein complexes viz. mTORC1 and mTORC2. The mTORC1 complex has three components, viz. RPTOR (regulatory-associated protein of MTOR complex1), G protein β-subunit-like protein (GβL), and PRAS40 (AKT1S1). The mTORC1 effectors RPS6KB1 and (EIF4EBP1 (4EBP1, eukaryotic initiation factor 4E-binding protein 1) regulate ribosomal biosynthesis and cell cycle regulatory proteins required for G1-S transition and cell proliferation. The mTORC2 components are the RICTOR (RPTOR Independent Companion Of MTOR Complex 2) RICTOR , MLST8 (GβL), mSin1 (MAPKAP1), and PROTOR (PRR5), which can activate AKT and might also be involved in cytoskeletal reorganisation. 25,26 This signalling process leads to cell proliferation and survival and possibly also promotes cell motility. A regulatory feedback circuitry operates between mTORC1 and mTORC2 and DEPTOR (DEP-domain-containing and mTOR-interactive protein). Loss of DEPTOR activates the mTOR complexes and S6K (RPS6KB1), SGK1, and AKT1, which lead to cell proliferation and survival. When overexpressed, DEPTOR suppresses the feedback inhibition of AKT1 by mTORC1. The removal of the inhibitory constraint promotes AKT-mediated cell proliferation. Here explored are the appropriateness and practicality of deploying the potential molecular targets and the constituents of the signal transduction systems, which they activate to assess aggressive potential of tumours in the laboratory setting. Figure 2 illustrates the compass of phenotypic control exerted by mTOR and underscores its significance in cancer growth and progression. The biologicaleffects of some growth factors might themselves be influenced by mTOR by modulating their downstream regulators. Note that mTOR can activate the JAK/STAT signalling system, which is involved with a wide variety of biological parameters of cell differentiation, cell proliferation, apoptosis, and cell migration. Among the significant downstream targets of the JAK/STAT pathway are PI3K/AkT and the Raf/ERK/MEK cascade of MAPK signalling. This figure is based on references cited in the text. CSCs: Cancer stem cells, EGFR: Epidermal growth factor receptor, ER: Oestrogen receptor, IGF/IFGR: Insulin-like growth factor/receptor, Pi3k: Phosphoinositide-3 kinase, PRP: Progesterone receptor, RTK: Receptor tyrosine kinas, TGF-β: Transforming growth factor- β, VEGFR: Vascular endothelial growth factor.5. The operation of mTOR signalling in tumour development The evidence that mTOR system is activated by the transcription Snail and Slug factors is rather indirect, if not tenuous, at present. The biological effects of mTOR inhibitors and the phenotypic effects of some growth factors do suggest Snail/Slug activity. 27,28 The operation of the EGFR/mTOR/EMT signalling has been amply demonstrated. For example, inhibition of Twist makes tumours sensitive to the action of EGFR inhibitors. 29 Furthermore, some growth factors can induce invasion by the Raf/Rac/Snail/Slug/EMT signalling route. This route has close links with mTOR. 30 Markers of EMT and mTOR are aberrantly expressed in parallel in many human tumours, but with little or no focus on breast cancer. Among possible complicating factors is the influence of the molecular environment prevalent at the time. For instance, PI3K/AKT/mTOR inhibitors have been studied, but these had to take account of the effect of PI3K/AKT/mTOR inhibitors depended upon the status of PTEN activation. 31,32 When the integrity of PI3K/AKT/mTOR is compromised by mutations of the AKT1, tumours have become resistant to endocrine therapy. 33 The deficiencies of the currently available data and the inadequacies of approach amply justify exploring new vistas, such as EMT and mTOR profiles. The new approaches could offer possible additional aids to supplement the rather limited capacity of the currently employed breast cancer markers to predict the progression and metastasis breast cancer [Figures 1 and 2]. One can reasonably argue that the suppression of MTOR is a potentially viable approach to control tumour growth and metastatic progression.Not only can negative regulators of mTOR such as allosteric inhibitors and RTK (receptor tyrosine kinase) inhibitors, but also dual mTOR/PI3K inhibitors can be deployed for tumour control. The latter would suppress mTOR with concurrent suppression of EMT, CSCs, and the induction of angiogenesis. Metformin has displayed marked anti-tumour effects. It is also a negative regulator of MTOR and indeed actively inhibits intratumoral and tumour associated angiogenesis. So also statins, which have been the subject of much debate, and bisphosphonates, possess significant anti-tumour activity. This effect could occur via inhibition of MTOR, of which the statins are indeed demonstrably capable. DOI: 10.2174/1573394714666180611113834 Angiogenic signals are transduced by several pathways often co-ordinated by MTOR. 6 Hence MTOR inhibition could provide an efficient means for effective cancer control. Nevertheless, the diversity of angiogenic signalling does suggest combination therapy could enhance the total inhibitory effect in comparison with the use of MTOR inhibitors as single agents. The cost effectiveness of targeted molecular therapies and conventional chemotherapeutic agents would need some serious consideration. That exercise does seem eminently worthwhile. There should be no ‘Lakshman Rekhas’ [red lines] where survival and quality of life of patients are concerned. By all accounts, MTOR is likely to appear in the context of cancer biology. Nonetheless, demonstrating MTOR activity and alterations in associated signalling might provide new markers and support the thesis that strategically selected new laboratory tests could be devised for patient management.6. Prospective focus The MTOR inhibitors are currently in clinical use, e.g. rapamycin for the treatment of glioblastoma, temsirolimus for renal cell carcinoma and mantle cell lymphoma, and everolimus for neuroendocrine tumours and breast cancer. Future work may most beneficially focus on the development of novel allosteric inhibitors and dual mTOR/PI3k inhibitors to achieve more efficient molecular targeting, given that many RTK inhibitors are now available. Clinical trials to determine the efficacy of metformin for tumour control and whether this occurs together with mTOR inhibition would also seem worthy of study. There is no paucity of clinical trials with MTOR inhibitors. While planning and designing the trials it is essential to bear in mind that MTOR is a cybernetic regulatory system. Clinical trials to-date with MTOR inhibitors have taken no cognisance of opposing signals that MTOR is known to co-ordinate. For instance, insulin and insulin-like growth factor and their receptors induce cell proliferation. This occurs via MTOR signalling. 6,34 It has been shown that mTORC1 and its downstream effectors RPS6KB1 and 4EBP (EIF4EBP1) can suppress the insulin receptor substrate, which is itself an inhibitor of PIK3CA. Therefore, inhibiting MTOR would in effect promote PI3K/AkT activity and lead to cell proliferation. However, in ‘rapamycin insensitive’ situations mTORC2 would likely promote PI3K and cell proliferation. In this state, inhibitors of MTOR might have beneficial effects. The growth promoting signals of EGF/EGFR can flow down the IRS route. Therefore, combining MTOR inhibitors with inhibitors of RTKs could provide a distinct additive growth suppressor effect. In this context it ought to be stated that mTOR exerts markedly pleiotropic effects on autophagy and apoptosis signalling. Apoptosis may be induced or suppressed apparently in conjunction with the ambient molecular environment. In this signalling scenario it is unsurprising that some clinical trials with conventional inhibitors of MTOR and a few with allosteric inhibitors found no benefits. This could be due to highly restricted primary outcome measures and criteria of clinical evaluation of response adopted, and no less, a consequence of the lack of use of markers of the signalling systems integrated by MTOR. Therefore, the somewhat incomplete and deficient studies should not detract from the proposition that the markers for the MTOR transcription factors and the downstream effectors of MTOR and EMT signalling should be comprehensively evaluated in tumour samples. Several issues need to be resolved so that a reasonable basis is built for introducing new markers and methods in laboratory testing. Objectivity in choosing the appropriate test for practicality such as the economic aspect, the availability of reagents and the establishment of quality control assessment is essential. Furthermore, stringent standardisation is required for the measurement of expression of molecular targets. Heterogeneity of the expression of molecular markers is a recognised source of error in data interpretation. Differences in the methods used can lead to much divergence of estimated expression of the markers. This can happen whilst measuring cellular receptor expression using radioactively labelled ligand binding assays and immunohistochemical techniques to measure receptors levels. Immunohistochemistry would reveal the degree of heterogeneity of expression, but ligand binding assays would not. The extrapolation of the data to predict progression has to take heterogeneity into account since it markedly influences phenotypic behaviour. 35 It is inevitable that there would be crosstalk between the signalling systems that might affect the expression of collateral genes. This could negate the perceived positive merits and could be counterproductive in designing modes of therapy for individual patients. Signalling crosstalk is also a potential source of resistance to therapy. Artificial neural network-based technology would enter the scenario here. The molecular profiles as a panel of markers or as individual markers have to be evaluated for robustness of predictive value. This holds true also for MTOR and EMT constituents employed as markers. The predictive weights of individual markers and the significance of the relative expression of the markers would also have a bearing on the predictive ability. Artificial neural networks algorithms could help in the identification of EMT and mTOR markers beneficial to individual patients. Conflicts of interest: NoneAcknowledgements I thank Professor Garth Nicolson for his critical review on the manuscript, and Mr David Headon foI for his helpful suggestions.