Cancer and Metastasis Reviews

, Volume 29, Issue 4, pp 751–759

The phosphatidyl inositol 3-kinase/AKT signaling pathway in breast cancer


    • Instituto Nacional Enfermedades Neoplásicas
  • Hernán Cortes-Funes
    • Hospital Universitario 12 de Octubre
  • Henry L. Gomez
    • Instituto Nacional Enfermedades Neoplásicas
  • Eva M. Ciruelos
    • Hospital Universitario 12 de Octubre

DOI: 10.1007/s10555-010-9261-0

Cite this article as:
Castaneda, C.A., Cortes-Funes, H., Gomez, H.L. et al. Cancer Metastasis Rev (2010) 29: 751. doi:10.1007/s10555-010-9261-0


The phosphatidyl inositol 3-kinase (PI3K)/Akt pathway mediates the effects of a variety of extracellular signals in a number of cellular processes including cell growth, proliferation, and survival. The alteration of integrants of this pathway through mutation of its coding genes increases the activation status of the signaling and can thus lead to cellular transformation. The frequent dysregulation of the PI3K/Akt pathway in breast cancer (BC) and the mediation of this pathway in different processes characteristically implicated in tumorigenesis have attracted the interest of this pathway in BC; however, a more comprehensive understanding of the signaling intricacies is necessary to develop clinical applications of the modulation of this pathway in this pathology. We review a series of experiments examining the contribution of alteration of integrants of this signaling network to human BC and we make an update of the information about the effect of the modulation of this pathway in this cancer.


PIK3CAPTENp110αAktmTORbreast cancer

1 Introduction

Breast cancer is the second most prevalent cause of cancer death in the western countries and displays a complex etiology. Alterations in pathways regulating cell signaling, growth, proliferation and apoptosis are central to this neoplasia. Thus, activation of the PI3K/Akt pathway is thought to play a key role in the regulation of malignant breast cancer (BC) cells growth [1].

PI3Ks are lipid kinases divided into three classes according to sequence homology, substrate preference and tissue distribution. Class I PI3Ks are further divided into class IA and IB, being the first one the most frequently implicated in cancer (Fig. 1). Into class IA, the p110α and its regulatory subunit, p85, are the members most related to the regulation of cell division and tumorigenesis, and will be the main topics of this review. The other three class I isoforms are p110 β, δ, and γ; at the contrary to α isoform, they have no known oncogenic mutations, but there is some evidence of the potential oncogenic role of their wild-type overexpression [2].
Fig. 1

PI3K signaling. For explanation, see text

p110α is composed of five domains: a kinase with a C-terminal domain, an adaptor binding domain that joins the kinase, a Ras-binding domain (RBD), a C2 domain that joins the p85 subunit, and a helical domain. It is encoded by the PIK3CA gene that consists of 20 exons located on chromosome 3q26.3 [2]. p85 is also composed of five domains that include the N-terminal Src homology-2 (SH2) and the inter-SH2 domains. In its basal state, this subunit is bound to and inhibits p110α, and the minimal fragment capable of performing this regulation is the binding of the mentioned domains [2].

The signaling cascade is initiated when the activation of a tyrosine-kinase receptor, such as epidermal growth factor receptor-1 and insulin-like growth factor receptor, occurs [3]. Some studies indicate that Flk-1/KDR and a subpopulation of estrogen receptor (ER)-α can also trigger this pathway [4]. This phosphorilated tyrosine attracts p85 subunit and relieves the inhibition it exerts over p110α with its subsequent recruitment to the plasma membrane and activation [2].

In a different and p85 independent way, Ras can interact with RBD domain of p110α with its tethering to cell membrane and activation. It has been recently reported that a novel protein, PI3K-interacting protein-1, interacts with p110α and acts as another regulator of its function as well [2].

Activation of p110α leads to the phosphorylation of phosphatidylinositol—4, 5-biphosphate (PIP2) with the production of phosphatidylinositol 3,4,5-triphosphate (PIP3). PIP3 acts as a second messenger through the recruitment of adaptor and effectors’ proteins containing a Pleckstrin homology (PH) domain, like Akt and Pdk1, to cellular membranes [2].The phosphatase and tensin homolog deleted on chromosome 10 (PTEN), with the participation of p110β, dephosphorylates PIP3 to PIP2. PTEN also acts directly in a nuclear setting and modify processes related to cell cycle, apoptosis and chromosomal integrity.

Activation of p110α leads to the phosphorylation of phosphatidylinositol—PIP2 with the production of phosphatidylinositol PIP3. PIP3 acts as a second messenger through the recruitment of adaptor and effectors’ proteins containing a PH domain, like Akt and Pdk1, to cellular membranes [2]. The phosphatase and tensin homolog deleted on PTEN, with the participation of p110β, dephosphorylate PIP3 to PIP2. PTEN, which is a dual lipid and protein phosphatase, also acts directly in a nuclear setting and modify processes related to cell cycle, apoptosis and chromosomal integrity.

Akt is a serine–threonine kinase that has three family members, Akt1, 2, and 3, coded by AKT1, 2 and 3, respectively. It has different localizations inside the cell and different activities in the regulation of cellular processes [1, 5]. When Akt gets the cell membrane, it is phosphorylated at Thr308 and Ser473 by Pdk1 and the mammalian target of rapamycin (mTOR) complex (mTORC)-2, respectively [2]. Activated Akt moves to cytoplasm and nucleus, where it phosphorylates and activates target proteins involved in different cellular functions.

Akt activates mTORC-1, a complex also modulated by ERK, which produces the stimulation of protein synthesis and cell growth by regulating the ribosomal p70S6 kinase (S6K) and the eukaryotic translation initiation factor 4E (eIF4E) [2, 6]. Akt also stimulates cell cycle progression and proliferation by modulating cell cycle inhibitors, such as p21, p27kip1 and GSK3, and cell cycle stimulators, such as c-myc and cyclin D1 [7]. Akt also regulates programmed cell death through inhibition of both pro-apoptotic genes (FasL and Bim) and proteins (BAD and BAX), stimulation of anti- apoptotic proteins (NF-αK) and degradation of the tumor suppressor protein p53 [8]. Through upregulation of the transcription repressor Snail, Akt is implicated in the induction of epithelial-mesenchymal transition and achievement of invasive properties [9, 10]. Akt/mTORC1 induces the cell expression of HIF-1 [11] and VEGF [12] and is essential for angiogenesis [13]; and also appears to induce the cell expression of CXCR4 protein, a chemokine receptor, that has been shown to play key roles in stroma support for tissue specific metastasis [14].

It has also been described some feedback inhibition mechanisms inside this complex pathway, one of the most studied and with evidence of clinical relevance is the ability of S6K to phosphorylate and inhibit insulin receptor substrate 1 (IRS-1), an upstream member of the PI3K/Akt pathway [2, 3, 15].

Taken together, this information show the implication of the members of this pathway in processes intimately related with cancer development and metastasis [2].

2 Mutational alterations in PI3K/Akt pathway and breast cancer

The particular importance of upstream PI3K/Akt pathway members like ERα and HER2, the evidence of preclinical mammary tumorigenic activity of PI3K/Akt members, and the high rates of deregulation of the PI3K/Akt pathway in BC [1618] have attracted the interest of studying this pathway in mammary tumors. Activation of PI3K/Akt pathway in BC has shown to be as frequent as 70% [17]; and some studies suggest its association with aggressive features like high histological grade, basal like [17, 19, 20] and HER2 phenotypes [2023], and with a poor clinical outcome [17, 20, 2225].

The best known genetic alterations associated to an aberrant activation of this pathway in BC are the presence of activating point mutations at PIK3CA, loss of PTEN activity, and mutation of AKT1. The coexistence of the three mentioned alterations in BC is an unresolved topic [1, 17, 2632]. They have also been observed in in situ lesions [3336], and a recent study has shown the acquisition of PI3KCA mutations during tumor progression [37]. Additionally, tumors with mutant PIK3CA often possess mutations or alterations in other components of the PI3K/Akt pathway, such as Ras and HER2/HER3, creating an amplified signal that could efficiently enhance oncogenic transformation [38].

Mutations in PIK3CA, the gene coding for p110α, are somatic and the majority map in to three “hotspots” (80% of all mutations). They produce substitution of a single amino acid in their products, so the first two hotspots mutations involve exon 9 and usually produce a change of E542 and E545 in the helical domain to lysine. The third hotspot mutation is located in exon 20 and frequently produces the substitution of His1047 in the kinase domain of p110α by arginine. The first two mutations relief the inhibitory interaction between p85 and the helical domain of p110α, and the third mentioned mutation allows easier access of p110α to the membrane, all of them producing an increased activity [2].

Preclinical studies have found that PIK3CA activity induces mammary tumorigenesis and also angiogenesis [39]. PIK3CA mutations are found in 28% to 40% BC tumors, being His1047Arg the most common mutation [31, 4042]. However, their relation to specific phenotype or prognosis in BC is unclear. Controversially, some studies have found a negative prognosis value for the mutated status of PIK3CA [24, 41, 43], while others have shown no prognostic value [15, 30, 40, 41, 4446], or even have associated the mutations to a more favorable clinical outcome [28, 31, 42, 47].

The first study evaluated 547 breast tumors and 41 BC cell lines and identified PIK3CA mutation rates of 21% and 39%, respectively. It found an association of the mutated status to hormone receptor (HR)-positive and HER2 overexpression/amplification. However, no associations were found between PIK3CA mutational status and survival [30]. The latest study evaluated 590 breast tumors and identified 32.5% mutations. It found an association of mutations with HR-positive, HER2-negative and lower tumor grade. Survival analysis found a significantly better clinical outcome, including overall survival [28]. The association to HR positivity observed in both studies was strong and has also been described by other researchers [27, 30, 31, 42, 47].

Inactivation of PTEN leads to loss of its lipid phosphatase activity and accumulation of PIP3, and has shown to promote mammary tumorigenesis in preclinical studies [48]. It can be produced by somatic mutations leading to either protein truncation, deletions, transcription repression or epigenetic silencing [2]. PTEN mutations have been observed in less than 5% of primary BC, although loss of expression rate is as high as 48% and the most frequent causes are promoter hypermethylation and deletions [27, 30, 31, 34, 4956].

The role of PTEN loss in defining prognosis in BC is also controversial: some studies have shown no prognostic value [16, 50, 57, 58] while others have found PTEN loss with a poor prognosis [51, 5356, 59]. PTEN loss and its association with HR or HER2 status is also uncertain [27, 30, 31, 50, 51, 55].

Lastly, it has been described a mutation in the PH domain of AKT1 (E17K) that produces single amino acid substitution that alters the enzyme lipid binding and presumably leads to constitutive membrane localization, which confers mammary transforming activity in preclinical studies [2, 26, 60]. This AKT1 mutation, although infrequent (1.4–8%), appears to be associated with positive HR status [28, 30].

Additionally, some PI3K/Akt pathway gene signatures have been developed. So, a PTEN gene signature including 246 genes was evaluated in 351 early breast tumors and tested in two independent breast tumor data sets of 295 and 99 samples. It did not show any relation with either ER or HER2 status but a correlation with poor outcome [58]. An analysis of more than 1700 BC samples identified and evaluated a gene expression signature induced by PIK3CA mutations. This signature was correlated ultimately with down-regulation of the PI3K/AKT/mTOR pathway (probably through a negative feedback loop), and associated to a better clinical outcome in the ER(+)/HER2(−) tumors [61].

3 PI3K/Akt pathway implication in breast cancer treatment

In the last decade some studies have evaluated the predictive role of PI3K/Akt pathway over the response to diverse anticancer agents. So, some in vitro studies suggested that an activation of PI3K/Akt pathway could play a role in resistance to chemotherapy [48, 62], but it has not been supported by the inability of PIK3CA status to predict pathologic responses to induction chemotherapy with anthracycline and taxanes [63]. Contradictory, a recent study found Akt phosphorylation status could have a predictive role for adjuvant paclitaxel [64].

A more encouraging area is this pathway’s predictive role in hormonal and HER2-directed therapy. The activation of the PI3K/Akt pathway has been associated with resistance to endocrine therapy in preclinical studies [65]; in this sense, Akt activation and loss of PTEN activity in patients with ER-positive BC have been related to resistance to endocrine therapy [23, 25, 55, 6670].

Preclinical studies have also found that deregulation of PI3K/Akt through PTEN loss or PIK3CA mutations leads to resistance to the anti-HER2 agents trastuzumab and lapatinib; and the use of PI3K inhibitors restore the sensitivity to them in these cells (although the preclinical data is controversial for the TK inhibitor) [7177]. Among three small cohorts of HER2-positive metastatic BC patients treated with trastuzumab-containing regimens, the presence of PIK3CA mutations or low PTEN expression were associated to lesser sensitivity to these therapies [29, 74, 78, 79].

A number of drugs targeting the PI3K/Akt signaling pathway have been developed and found to block tumor growth not only by targeting tumor cells but also via their effects on tumor vasculature [80]. It appears that their effect as monotherapy is low and some data of combination with other drugs are encouraging [81]. In addition to combining them with chemotherapy, hormonal, and antiHER-2 treatments, some preclinical studies have found synergic effect when added to IGF1R or MEK inhibitors, due to their feedback upregulation of the PI3K/Akt pathway [8286]. On the other hand, some studies have also started to test predictive factors to PI3K inhibitors in order to find more sensitive BC subpopulations, but the results are insufficient to make conclusions [58, 84, 8795].

This drugs have been grouped in four sets: dual PI3K-mTOR inhibitors (enzymes with similar structure), PI3K inhibitors (that don’t inhibit mTOR), mTOR catalytic site inhibitors, and Akt inhibitors. All of them have demonstrated activity in BC cells and some reversed the resistance to antiHER2 agents in preclinical studies [9698]. Akt inhibitors are in current clinical trials but there is not information of their activity in BC at this moment.

3.1 Dual PI3K-mTOR inhibitors

These compounds theoretically are able to turn off the PI3K pathway activity avoiding any feedback stimulus [81]. A recent phase I study has evaluated the oral PI3K inhibitor BEZ235 in 59 patients with advanced solid tumors, including 13 breast (22%) tumors, and has found a good safety profile, two partial responses (PRs) and 14 four-month stabilizations (SD). Into the BC patients, one PR and 4 SD ≥ 4 months were obtained [99].

3.2 PI3K inhibitors

GDC-0941 is a pan-class I PI3K inhibitor that has been evaluated in a phase I trial with 13 patients with solid tumors. It showed a good toxicity profile and some evidence of antitumor activity mesured by PET-CT [100]. BKM120, another pan-class I PI3K inhibitor, has been evaluated in a phase I study with 30 patients with advanced solid tumors, including eight with BC. It has shown a favorable toxicity profile and one PR and five minor responses, three of them in patients with BC. In addition, 12 patients (40%) across all doses experienced disease stabilization for ≥4 months [101].

3.3 mTOR inhibitors

This family of drugs, with their predecessor rapamycin, inhibits mTORC1, but not mTORC2 and could lead the theoretically undesired IGF-IR/IRS-1-dependent activation of Akt and MAPK [102, 103]. They were the first PI3K/Akt pathway inhibitor group entering to BC clinical trials, and data from clinical trials with everolimus, temsirolimus, and ridaforolimus are known.

A phase II study evaluating everolimus in minimally pretreated patients with metastatic BC found an acceptable tolerability, and obtained a response rate of 12%, with one complete response and three PRs, in 30 evaluable patients [104]. Another randomized phase II study evaluated the combination of daily oral everolimus plus letrozole versus letrozole alone for 4 months as induction treatment on 270 postmenopausal women with ER-positive early BC. Clinical response rate with everolimus and letrozole was significantly higher than letrozole alone when evaluated by physical exam (68% vs 59%) and by ultrasound (58% vs 47%). However, everolimus was associated with an increase in grade 3/4 adverse events (22.6% vs 3.8%) [105].

In a phase II study, 109 heavily pretreated patients with advanced BC-received temsirolimus, that was well tolerated and lead to an objective response rate of 9.2% (ten PRs) [106]. Despite that, the interim analysis of a phase III placebo-controlled trial of letrozole with or without temsirolimus reported no improvement in PFS [107]. A phase I study of the combination of temsirolimus and capecitabine in 12 patients with heavily pretreated metastatic BC found a good toxicity profile with a clinical benefit rate of 44% [108].

Two phase I studies evaluated the combination of everolimus, trastuzumab and either paclitaxel or vinorelbine in a total of 79 patients with metastatic BC, almost all trastuzumab resistant (99%). Results showed an acceptable safety profile and encouraging efficacy with 4% complete responses, 21.5% partial responses, and 54% stabilizations [109111]. Recently a phase I/II trial of the combination of trastuzumab and everolimus in 47 trastuzumab-resistant HER2-positive BC patients, found rates 34% clinical benefit, including nine patients that previously received lapatinib treatment [112]. Another phase II study evaluated the combination of everolimus, weekly paclitaxel and trastuzumab in 55 patients with trastuzumab and taxane resistant HER2-positive metastatic BC. This study confirmed a good toxicity profile and found a 20% PR and 56% SD rates in the first 25 evaluable patients [113].

Results of a single-arm phase II trial evaluating the combination of ridaforolimus with trastuzumab in 34 patients with trastuzumab-resistant HER2-positive metastatic BC has been reported (Table 1). Results showed good tolerance and encouraging clinical activity of five (15%) PRs and 13 (38%) SD [114]. Other phase I study evaluated the combination of ridaforolimus and the IGF-1R antibody dalotozumab in 62 patients with advanced solid tumors, and found a good tolerance and promising anticancer activity. Among 23 BC patients included, activity was found in five patients (two PRs, 1 SD for 9 months, and two partial metabolic responses on PDG-PET scan).
Table 1

Results of clinical trials with PI3K/Akt pathway targeting drugs

Therapeutic agent

Clinical scenario

Comments of efficacy

Dual PI3K-mTOR inhibitors


Phase I study in advanced solid tumors (n = 59, including 13 BC)

Among BC, 1 PR and 4 SD of at least 4 months [99]

PI3K inhibitors


Phase I study in advanced solid tumor (n = 30, including 8 BC)

Among BC: 1 PR and 5 minor responses [101]

mTOR inhibitors (only phase II-III)


Phase II study in minimally pretreated metastatic BC patients (n = 30)

Response rate of 12% (1 CR and 3 PRs) [104]

Phase II study of neoadjuvant letrozole with or without everolimus in postmenopausal women with ER positive BC (n: 270)

The clinical response rate significantly higher in the everolimus arm when evaluated by physical exam (68% vs 59%) and by ultrasound (58% vs 47%) [105]

Phase I/II trial of the combination of trastuzumab and everolimus in trastuzumab-resistant HER2-2 BC patients (n = 47)

PR of 15% and clinical benefit of 34% [112]

Phase II study of the combination of everolimus, weekly paclitaxel and trastuzumab in trastuzumab and taxane- resistant HER2+ metastatic BC (n = 55)

Clinical benefit in the first 25 evaluable patients: PR of 20% and SD of 56% [113]


Phase II study in heavily pretreated advanced BC (n = 109)

Objective response rate of 9.2% (10 PRs) [106]

Phase III study of letrozole with or without temsirolimus in postmenopausal advanced BC (n = 992)

No improvement in PFS [107]


Phase II study of the combination of ridaforolimus with trastuzumab in trastuzumab-resistant HER2+ metastatic BC (n = 34)

Clinical benefit: PR of 5 (15%) and SD of 13 (38%) [114]

4 Conclusions

PI3K/Akt signaling pathway is key in the development of BC; however, its prognostic and predictive value is still not clear. Inhibitors of this pathway are in current development and some of them have entered onto clinical trials in BC.


TRANSBIG traineeship programme to Carlos A. Castaneda.

Copyright information

© Springer Science+Business Media, LLC 2010