Epidermal growth factor receptor (EGFR) is a member of the ErbB family, a family of tyrosine kinase receptors with growth-promoting effects 1. It is expressed in several carcinomas 23 and high levels of expression are a common feature of the malignant phenotype in many solid human tumours 4. It is activated by up to four different ligands, most commonly epidermal growth factor (EGF) and transforming growth factor-α(TGF-α), and forms homodimeric complexes or heterodimeric complexes with other members of the ErbB family of receptors. Ligand binding and dimerization causes autophosphorylation of the intracytoplasmic domains and activation of the intracellular tyrosine kinase. Activated EGFR (phosphorylated EGFR [pEGFR]) stimulates a number of different signal transduction pathways, such as the Ras/mitogen-activated protein kinase pathway, the phosphoinositide-3 kinase (PI3K)/Akt pathway and the phospholipase-Cγ/protein kinase C pathway. The activation of these pathways initialize with the recruitment of different adaptor proteins, which bind to different phosphotyrosine residues of the cytopasmic tail of EGFR, and continues with a highly complex network of enzymes, proteins and small-molecule secondary messengers 5. The signal transduction pathways activated by pEGFR play important roles in various cellular processes, such as cell proliferation, differentiation, adhesion, migration and apoptosis 46.

EGFR is reported to be expressed in 14% to 91% of patients with breast cancer 789101112, and in several studies it has also been associated with poor prognosis 7, although its prognostic value remains unclear. Recently, EGFR has once again come to the fore, because of the development of several novel drugs that target EGFR. Because EGFR has not proven to be a useful prognostic/predictive marker of clinical response to EGFR-targeted therapies 913, other prognostic/predictive markers have been proposed, including the activated form of EGFR (pEGFR) 14. Previously, EGFR phosphorylation was found to be associated with poor prognosis in non-small-cell lung cancer patients and was suggested to be an important predictor of clinical outcome 1516.

The purpose of the present study was to gain further insight into the prognostic significance of EGFR and pEGFR expression in invasive breast cancer. This was assessed in relation to the classical clinicopathological parameters, clinical outcome and the expression of biological markers of invasion and angiogenesis (phosphorylated Akt [pAkt], urokinase plasminogen activator receptor [uPAR], matrix metalloproteinase [MMP]-14 and vascular endothelial growth factor receptor [VEGFR]-1/Flt-1).

EGFR and pEGFR proteins were immunodetected in the membrane of the malignant cells in 18 (11.3%) and 55 cases (35.7%), respectively (Figure 1). Of 154 patients, eight (5.2%) were both EGFR and pEGFR positive, 10 (6.5%) were EGFR positive only, 47 (30.5%) were pEGFR positive only, and 89 (57.8%) were completely negative.

Based on expressions of EGFR, ER and c-erbB-2, only two carcinomas were classified as breast carcinomas of basal-like phenotype (EGFR-positive, ER-negative and c-erbB-2 negative). None of the studied proteins was associated with menopausal status, tumour size, lymph node status, tumour stage, PR hormonal status and c-erbB-2 protein.

EGFR immunoexpression was related to nuclear grade (P = 0.001) and inversely related to ER receptor status (P = 0.005). Immunoexpression of pEGFR was associated with pAkt, uPAR, MMP-14 and VEGFR-1/Flt-1 (P = 0.008, P = 0.049, P = 0.025 and P = 0.016, respectively). Coexpression of EGFR/pEGFR was statistically associated with pAkt, MMP-14 and VEGFR-1/Flt-1 (P = 0.011, P = 0.039 and P = 0.019, respectively).

With regard to their prognostic significance, EGFR and pEGFR were found to have an unfavourable associated with overall survival (P = 0.046 and P = 0.032, respectively), but not with disease-free survival (univariate analysis). Interestingly, patients with synchronous expression of EGFR and pEGFR had worse prognosis (P = 0.019) than did EGFR-positive or pEGFR-positive patients and totally negative patients (Figure 2). Moreover, in multivariate analysis adjusted for numerous factors (age, menopausal status, histological type, histological grade, tumour size, lymph node status, ER, PR, c-erbB-2, EGFR, pEGFR and EGFR/pEGFR phenotype), tumour size (P = 0.010; B coefficient = 6.550; standard error = 0.728, 95% confidence interval = 1.573 to 27.281) and EGFR/pEGFR coexpression (P = 0.013; B coefficient = 1.520; standard error = 0.168, 95% confidence interval = 1.093 to 2.113) were of independent prognostic significance for overall survival.

This study examined the expression of EGFR and its activated form, pEGFR, in invasive breast carcinoma. We demonstrated that EGFR and pEGFR are both expressed in the membrane of 11.3% and 35.7% of tumour cells. The greater rate of positivity for pEGFR than for EGFR could be attributed to the EGFR downregulation process. Once EGFR is activated it undergoes internalization, resulting in a marked decrease in nonactivated membrane-bound EGFR 24. In previous studies, EGFR immunopositivity exhibits great variety 789101112. The wide range of EGFR expression (14% to 91%) may be accounted for by use of different methods and different criteria for assessment, as well by the presence of basal-like carcinomas that consistently over-express EGFR.

None of the studied proteins was associated with menopausal status, tumor size, lymph node status, tumour stage and PR hormonal status, which is in accordance with the majority of reports on EGFR expression 925. The absence of any association between EGFR and c-erbB-2 expression may be due to variations in c-erbB-4 expression, which antagonizes the influence of c-erbB-2 in tumors 7.

With regard to EGFR, we showed that its expression was positively associated with nuclear grade (P = 0.001) and inversely associated with ER hormonal status (P = 0.005), which is in accordance with a large number of studies 2526. This consistent finding in much of the literature had led to the plausible hypothesis that EGFR signalling may be associated with endocrine resistance or insensitivity 272829. Alternatively, this finding may be accounted for by the strong association between EGFR expression and loss of differentiation in breast cancer.

EGFR activation has an antiapoptotic effect through PI3K pathway 1, a fact that was corroborated by the pEGFR relationship to pAkt (P = 0.014). EGFR can lead to activation of PI3K both directly and indirectly through Ras; it induces downstream activation of phosphoinositide-dependent protein kinase-1 and -2 that phosphorylate thr308 and ser473 on Akt, respectively. Akt's antiapoptotic role is well known, via phosphorylating and sequestering downstream targets including the FOXO family of forkhead transcription factors, the proapoptotic Bad and the protease caspase-9, and by activating the pro-survival transcriptional regulator protein nuclear factor-κB.

Expression of pEGFR was associated with VEGFR-1/Flt-1 and MMP-14. VEGFR-1/Flt-1 is a specific endothelial cell receptor to which the angiogenic factors VEGF-A and VEGF-B bind; it promotes differentiation and vascular maintenance 30. Indeed, in tumour progression EGFR upregulates VEGFR; thus, it is implicated in angiogenesis. In addition, in human cancer cells the EGFR autocrine pathway controls the production of several proangiogenic growth factors, including VEGF 4. It is also known that several EGFR inhibitors, such as monoclonal antibodies, result in a concurrent downregulation of tumour-induced, VEGF-mediated angiogenesis 4. Therefore, the above-mentioned relationship implies an angiogenetic role of activated EGFR. The angiogenetic ability of pEGFR has further confirmed by the pEGFR association with pAkt, which is known to modulate angiogenesis via activation of endothelial nitric oxide synthase 31. MMP-14, as well as most of the MMPs, may promote angiogenesis by at least two different mechanisms: by degrading barriers and allowing endothelial cell invasion; or by liberating factors that promote or maintain the angiogenic phenotype 32. In addition, it promotes invasion and metastasis by degrading ectracellular matrix 33. It appears that the above-mentioned relationship between EGFR and MMP-14 reflects the well established interaction between EGFR pathway and the MMPs 1. EGFR activation is able to upregulate MMPs, whereas MMPs participate in several pathways of EGFR activation, such as the ectodomain shedding of EGFR transmembrane precursor 4, G protein-coupled receptor-mediated transactivation, and uPAR-mediated transactivation of EGFR 1.

The combination of EGFR/pEGFR was associated with pAkt, MMP-14 and VEGFR-1, which may be accounted for by the parallel relationship between pEGFR and those biological parameters.

Another observation in the present study, that of the parallel relationship between pEGFR and uPAR, supports the existence of a uPAR-mediated EGFR transactivation pathway 1 and enforces the invasive effect of pEGFR. The uPAR glycoprotein, receptor of plasminogen activation system, plays a central role in extracellular matrix degradation; thus, it participates in tumour invasion and metastasis 34. Previously, in breast cancer, it has been related to tumour aggressiveness and patients poor survival 21.

With regard to the association of EGFR and pEGFR with prognosis and survival, we found that EGFR and its active form are significantly associated with overall survival (P = 0.035 and P = 0.016, respectively) but not with disease-free survival in univariate analysis. Findings from previous studies regarding prognostic significance of EGFR are controversial. Tsuitsu and coworkers 11 reported the prognostic value of EGFR for overall and disease-free survival, whereas other investigators failed to confirm its prognostic significance in either the entirety of studied cases 789 or in lymph node positive 8 or -negative cases 3536. On the other hand, there is no report about pEGFR expression and prognosis in breast cancer, as far as we know, whereas in our study pEGFR appeared to have an unfavourable impact on overall survival. We also observed that EGFR-positive/pEGFR-positive tumours had worse prognosis in contrast to EGFR-negative and/or pEGFR-negative ones. Moreover, EGFR/pEGFR was found to be an independent prognostic indicator for overall survival. EGFR and pEGFR coexpression appears to be more representative of EGFR dynamics in tumour than solely EGFR or pEGFR expression. Probably, in EGFR-positive/pEGFR-positive tumours, the total number of receptors is greater; thus, there are more receptors available for subsequent phosphorylation. Another possible explanation might be the existence of an inhibitory mechanism, which impedes autophosphorylation of several receptors and permits only a few receptors to become phosphorylated and initiate EGFR-dependent signalling cascades.

Nieto and coworkers 37 also studied EGFR and pEGFR expression in invasive breast cancer, but they failed to attribute any prognostic value to pEGFR expression. To the best of our knowledge, there are no published data confirming the prognostic value of pEGFR or EGFR/pEGFR coexpression in breast cancer. Interestingly, Arteaga and Baselga 14 have noticed that EGFR-positive colon carcinomas with simultaneous expression of pEGFR, coexpressed the EGFR ligand transforming growth factor-α and markers of tumour proliferation, which is in contrast to EGFR-positive/pEGFR-negative carcinomas. Thus, EGFR content in EGFR-positive/pEGFR-negative tumours does not reflect the level of receptor activation.

In the present study, simultaneous expression of both forms of EGFR emerged as a more promising prognostic marker in invasive breast carcinomas. However, more research is needed to clarify the importance of EGFR/pEGFR immunoexpression to prognosis, as well as the applicability of pEGFR expression and amplification in targeted therapies.

EGFR = epidermal growth factor receptor; ER = oestrogen receptor; MMP = matrix metalloproteinase; pAkt = phosphorylated Akt; pEGFR = phosphorylated epidermal growth factor receptor; PI3K = phosphoinositide-3 kinase; PR = progesterone receptor; uPAR = urokinase plasminogen activator receptor; VEGFR = vascular endothelial growth factor receptor.

The authors declare that they have no competing interests.

CM was responsible for the statistical analysis and the manuscript preparation. LN provided overall guidance, direction and critical review of the study design. KK was responsible for performing the immunohistochemical protocols, manuscript preparation, data collection and editing. CZ and IT were responsible for the evaluation of the immunohistochemical staining. PB was responsible for data collection. IG was responsible for supervising the immunohistochemical protocols and editing.