PI3K pathway activation in breast cancer is associated with the basal-like phenotype and cancer-specific mortality

Abstract

Breast cancer is a common malignancy with current biological therapies tailored to steroid hormone (ER, PR) and HER2 receptor status. Understanding the biological basis of resistance to current targeted therapies and the identification of new potential therapeutic targets is an ongoing challenge. The PI3K pathway is altered in a high proportion of breast cancers and may contribute to therapeutic resistance. We undertook an integrative study of mutational, copy number and expression analyses of key regulators of the PI3K pathway in a cohort of 292 invasive breast cancer patients with known treatment outcomes. The alterations identified in this cohort included PIK3CA mutations (12/168, i.e. 7%), PIK3CA copy number gain (28/209, i.e. 14%), PTEN loss (73/258, i.e. 28%) and AKT activation (62/258, i.e. 24%). Overall at least 1 parameter was altered in 72% (139/193) of primary breast cancers. PI3K pathway activation was significantly associated with ER negative (p = 0.0008) and PR negative (p = 0.006) status, high tumor grade (p = 0.032) and a “basal-like” phenotype (p = 0.01), where 92% (25/27) of tumors had an altered pathway. In univariate analysis, PI3K pathway aberrations were associated with death from breast cancer; however, this relationship was not maintained in multivariate analysis. No association was identified between an activated pathway and outcome in tamoxifen- or chemotherapy-treated patients. We concluded that >70% of breast cancers have an alteration in at least 1 component of the PI3K pathway and this might be exploited to therapeutic advantage especially in “basal-like” cancers.

The phosphatidylinositol 3 kinase (PI3K) pathway regulates many cellular functions, mainly associated with cell proliferation, survival and migration.1 Activation of PI3K can occur in response to a variety of extracellular signals through growth factor receptor- or integrin-mediated pathways. Upon receptor activation, the p85 regulatory subunit of Class 1a PI3K is recruited to phosphotyrosine residues within the receptor. Once recruited to the membrane, the p110α catalytic subunit, encoded by PIK3CA, phosphorylates phosphatidylinositol biphosphate to generate phosphatidylinositol triphosphate (PIP3), which recruits phosphoinositide-dependent kinase 1 (PDK1) and V-Akt murine thymoma viral oncogene homolog 1 (AKT) to the cell membrane. PDK1 and PDK2 (likely the TORC2 complex) then phosphorylate AKT leading to the regulation of an array of downstream pathways including forkhead rhabdomyosarcoma, GSK3β and BAD with consequent effects on cell proliferation, cell metabolism and apoptosis, respectively. PIP3 is a substrate for phosphatase and tensin homologue deleted on chromosome 10 (PTEN), a well-described tumor suppressor protein, which dephosphorylates PIP3, inactivating the PI3K pathway.

Alterations in the PI3K pathway have been documented in many human cancers.2 In breast cancer, PIK3CA activating mutations, in exons 9 and 20, have been identified in ∼24% (range, 12–40%) of patient specimens.3–16 These mutations in the p110α catalytic subunit can activate the pathway and contribute to mammary tumor progression.15PIK3CA amplification has only been described in 2 breast cancer studies with a frequency of 1% and 9%, respectively.12, 16 Molecular alterations and loss of PTEN have also been described in ∼24% (range, 4–48%) of breast cancers5, 6, 14, 17–23 and increased activation of AKT has been identified in ∼41% (range, 9–76%) of breast cancer patients.5, 17, 19, 21, 22, 24–28 Despite the number of studies exploring the PI3K pathway in breast cancer, the association of specific molecular alterations with clinicopathological markers and outcome is often contradictory and unclear with a consensus yet to be reached. This may be due in part to the widely different frequencies of specific aberrations reported in published studies.

PIK3CA mutations have been associated with positive hormone receptor status in 2 studies,3, 14 but not in others,7, 11, 15 and with both favorable9 and poor patient outcome.10 PTEN loss has been associated with high grade and negative hormone receptor status in some studies17, 20 but not in others,22, 23 whereas AKT activation has been associated with high tumor grade,19 PR negativity and HER2 positive status28 but other studies have identified no such associations.12, 17 Furthermore, there has been no published integrative study determining the combined effects of aberrations in these 3 pivotal members of the pathway.

Breast cancer remains the most common invasive cancer of women in industrialized nations and is second only to lung cancer as the leading cause of cancer-related death in women. Breast cancer is heterogenous but can be classified into at least 4 subtypes by gene expression profiling: Luminal A, Luminal B, HER2+ and “basal-like” tumors,29, 30 with each subgroup reflecting a different biology, behavior and outcome. A surrogate classification of these subgroups has been attempted using immunohistochemistry, whereby patients were subclassified according to hormone receptor (ER, PR), HER2, cytokeratin 5/6 (CK5/6) and epidermal growth factor receptor (EGFR) status.31 Current breast cancer treatment is tailored in great measure by this subtype classification. Hormone receptor positive cancers are treated predominantly with endocrine therapy (ER+/PR+/HER2−) or Trastuzumab (HER2+), but despite the advances in treatment and outcome, understanding the development of therapeutic resistance and developing strategies to combat it is a continuing challenge. “Basal-like” cancers (triple negative for ER, PR and HER2 and positive for CK5/6 or EGFR) have no identified target for biological therapy and have recently been a focus of studies searching for potential new therapeutic targets. Thus, the aim of our study was to analyze the overall status of the PI3K pathway, by defining mutational status in exons 9 and 20, copy number and expression of key molecules within the pathway, in a cohort of breast cancer patients of known phenotype, treatment response and outcome status.

Material and Methods

Patients and tumors

Cases were drawn from the St. Vincent's Campus Outcome Cohort, which comprised 292 invasive ductal carcinomas of the breast from patients treated by a single surgeon (P.C.) between February 1992 and August 2002 at St. Vincent's Hospital, Sydney, Australia. Ethics approval was granted for the use of pathology specimens and cognate clinicopathological data (Human Research Ethics Committee of St. Vincent's Hospital, Sydney). A more detailed description of the clinicopathological characteristics of the cohort is published elsewhere.32 Forty percent of tumors were >20 mm, 45% were Grade >2, 43% were lymph node positive, 68% were ER positive, 57% were PR positive and 18% were HER2 FISH positive (>2.2 ratio of HER2:chr.17 centrosome). Median age was 54 years and patients were treated with endocrine therapy (49%), chemotherapy (38%) or both (24%). Cases were prospectively followed-up for a median of 64 months and the outcome events measured were as follows: recurrence (local or distant) (25%), metastasis (23%) and all deaths were recorded but only breast cancer-related deaths (18%) were considered for survival analyses.

DNA extraction

One core from each tumor block, from an area previously selected by a specialist breast pathologist (S.A.O'T.) as the most representative of the tumor, was used to extract DNA. The Agencourt formapure 96-well format kit (Beckman Coulter, Beverly, MA), a paramagnetic bead-based technology, was used to extract DNA from formalin-fixed, paraffin-embedded tissue from 288 individual patients. A significant number of cases (n = 120 of 288, i.e. 42%) were unsuitable for the mutational analysis, and the main reasons were the poor quality of DNA caused by years of storage of these tissues, the fixation protocol employed or the small amount of tissue sometimes available for DNA extraction.

PIK3CA mutational analysis

Primers were designed to avoid amplification of a known PIK3CA pseudogene for exons 9 and 20 of the PIK3CA gene: 9F CTGTAAATCATCTGTGAATC, 9R ATTTTAGCACTTA CCTGTGAC; 20F CAATCTTTTGATGACATTGC, 20R TG GAATCCAGAGTGAGCTTT. These 2 exons contain the 3 hotspots for PIK3CA mutations in cancer. Codons 542 and 545 in exon 9 and codon 1047 in exon 20. PCR reactions were undertaken using 10–50 ng of DNA, 0.4 μmol/l of each primer, 200 μmol/l deoxynucleotide triphosphates, 3 mmol/l MgCl₂, 1× PCR II buffer and 1.5 U of Amplitaq Gold DNA Polymerase (Applied Biosystems, Foster City, CA). The PCR conditions used were as follows: 94°C (10 min) for 1 cycle, 94°C (30 sec), 50°C (30 sec), 72°C (30 sec) for 40 cycles and a final extension step of 72°C (10 min). All PCR products were sequenced with the ABI 3100 Genetic analyzer (Applied Biosystems, Foster City, CA). All mutations were confirmed by analyzing the product of a second, independent PCR.

PIK3CA gene copy number

A subset of tumors (n = 209) was analyzed for PIK3CA gene copy number by real time quantitative PCR (qPCR) on the Roche Lightcycler 480 platform (Roche Diagnostics; F. Hoffmann-La Roche Ltd., Mannheim, Germany). RPLPO, located at 12q24, was used as an endogenous control, as previously described.33RPLPO yielded similar results as the use of a gene copy number control from an unaltered region on the same chromosome as PIK3CA. Primers for both genes were designed by the Roche primer software using the universal probe library system. qPCR reactions were completed using 50 ng of DNA, 0.2 μmol/l of each primer, 0.2 μl of probe and 5 μl of master mix. qPCR conditions used were as follows: 95°C (7 min) for 1 cycle, 95°C (5 sec), 50–60°C (0.4°C per second) (30 sec), 72°C (15 sec) for 50 cycles and a final extension step of 40°C (30 sec). The efficiency of both assays was analyzed from the slope of a standard curve; RPLPO and PIK3CA qPCR assays had an efficiency of 95% and 97%, respectively. Fold changes were calculated with the Delta Ct method. Fold changes above the mean ± 1 SD of all the samples analyzed, i.e. >1.381, were considered copy number gain. This is in agreement with a previously published copy number threshold based on a pool of normal breast tissue.34, 35 The coefficient of variation for triplicate estimates was <0.04.

Tissue specimens

A total of 18 tissue microarrays (TMA) containing 2 cores of each tumor sample were constructed from the formalin-fixed, paraffin-embedded tumor material from each patient in the cohort as previously reported.32 TMAs were produced using the MTA-1 Manual Tissue Arrayer (Beecher Instruments, Woodland, CA, USA).

Immunohistochemistry

Three-micrometer sections of each TMA were cut, deparaffinized and used for immunostaining. The following antibodies were used: pAKT (Ser473) rabbit monoclonal antibody clone 736E11 (Biocare Medical, Concord, CA, USA) and PTEN mouse monoclonal antibody clone 6H2.1 (Dako, Denmark). For pAKT, antigen retrieval was performed using DAKO solution pH 6.0 (s1699; DAKO, Carpentaria, CA) at 125°C for 4 min in a Pascal pressure chamber (DAKO), the antibody was incubated for 90 min at a 1:50 dilution. For PTEN, antigen retrieval was performed using DAKO solution pH 9.0 (s2367; DAKO) at 125°C for 30 sec in a Pascal pressure chamber, and antibody incubation was 90 min at a 1:200 dilution. A DAKO autostainer was used for immunostaining (DAKO). As secondary antibody a FLEX envision system (DAKO) was used to improve staining. Reactions were developed using diaminobenzidine and sections were counterstained, dehydrated and mounted. For pAKT, sections incubated with nonimmune serum (IgG) and LNCAP prostate cancer cells treated with LY29402 were used as negative controls36 and untreated LNCAP cells as positive controls. For PTEN, sections of human breast and prostate tumors were used as positive controls and the same sections incubated with IgG were used as negative controls. Immunohistochemistry protocols, scoring and analysis for p27^Kip1, p21^Cip1/Waf1, cyclin E1, cyclin D1 and Ki67 are described elsewhere37 (McNeil C.M., unpublished data).

Immunohistochemical scoring

Scoring was completed by 4 independent investigators (E.L.-K., S.A.O'T., E.K.A.M., M.R.Q.) blinded to the clinical and pathological information, 3 of whom were specialist breast pathologists; in cases of discrepancy a consensus was reached by conferencing. All tissue areas were examined for both intensity (1+ to 3+) and the proportion of positive cells. A histoscore was calculated by multiplying the percentage of positively stained cells and each category of staining intensity (Fig. 1). PTEN showed cytoplasmic localization and loss of expression was defined as the lack of any tumor cell staining. The data were dichotomized as present or absent staining. pAKT demonstrated predominantly cytoplasmic staining, although in a few cases nuclear staining was apparent. For immunostaining of pAKT the mean histoscore value was used to determine the cutoff value of H > 22.

Figure 1. Immunohistochemical staining of paraffin-embedded breast cancer tissue. Photomicrographs for PTEN expression: negative (IgG) and positive controls (a) and tumor scoring ranging from negative to intensities 1+, 2+ and 3+ (c). Photomicrographs for pAKT expression: negative (IgG) and positive controls (b) and tumor scoring ranging from negative to intensities 1+, 2+ and 3+ (d).

Statistical analyses

To assess the independence of 2 categorical variables, the chi-square test was applied when the variables were dichotomous, the Mann–Whitney test was applied when one variable was dichotomous and the other continuous and the Spearman Rank correlation was applied when both variables were continuous. Kaplan–Meier survival curves and Cox proportional hazard ratio (HR) were estimated to obtain risks of recurrence, metastasis and death, after adjusting for other confounder variables. The results were considered significant at the two-sided p of 0.05 level. Statview version 5.0 was used for the analysis (Abacus Systems, Berkeley, CA).

Results

PIK3CA genetic alterations and PTEN loss

To assess the frequency of genetic lesions in the PIK3CA gene or loss of PTEN protein, both central players in the PI3K pathway, we analyzed PIK3CA mutations by sequencing, PIK3CA copy number by qPCR and PTEN expression levels by immunohistochemistry.

Overall, 12 of 168 (7%) breast cancers assessed harbored PIK3CA mutations. All mutations were at previously reported hotspot codons in exons 9 and 20, i.e. E542K (n = 1), E545K (n = 2), Q546R (n = 1) and H1047R (n = 8).

Copy number gain was identified in 28 of 209 (14%) cancers analyzed. Mutations and copy number gains were almost exclusive events with only 1 cancer encompassing both mutation and copy number gain of the PIK3CA gene (p = 0.748). Neither mutations nor copy number gain were associated with any clinicopathological parameter or outcome.

PTEN, a negative regulator of the PI3K pathway, was lost in 73 of 258 (28%) cancers and was associated with high tumor grade (p = 0.009), tumor size >20 mm (p = 0.009), ER negativity (p < 0.001) and PR negativity (p = 0.0002) (Table 1). PTEN loss was not associated with recurrence, metastasis or breast cancer-specific death.

Table 1. Association between the PI3K pathway and clinicopathological parameters

AKT activation

As a measure of PI3K pathway activation, we also examined pAKT expression levels by immunohistochemistry. pAKT was positive in 62 of 258 (24%) specimens and was positively associated with high tumor grade (p = 0.011), ER negativity (p = 0.040), PR negativity (p = 0.014) and HER2 positivity (p = 0.016). pAKT was also associated with Ki67, a marker of cell proliferation (p = 0.005) (Table 1) and breast cancer-specific death (p = 0.03).

PI3K pathway activation and clinicopathological parameters

To assess the overall status of the PI3K pathway, we considered PIK3CA mutational and copy number data, PTEN expression and AKT activation together. PTEN loss and PIK3CA mutations were mutually exclusive events (p = 0.0368); all PIK3CA mutations were identified in PTEN-expressing tumors. No association was identified between PTEN loss and PIK3CA amplification (p = 0.982). A negative correlation was identified between expression of PTEN and activation of AKT by the Spearman Rank correlation (p = 0.0001). Low PTEN expression was associated with high AKT activation. When all 4 parameters were combined (PIK3CA mutations, PIK3CA copy number gain, PTEN loss and AKT activation) 139 of 193 (72%) breast cancer patients had an alteration in 1 or more components of the PI3K pathway. These 193 patients had similar clinicopathological characteristics as the overall 292 tumors and therefore were representative of the whole cohort (data not shown). Figure 2 describes the overlap between the alterations in the 4 parameters analyzed. Over 75% (105/139) of cancers had only 1 altered parameter, 23% (32/139) had 2 alterations and only 1% (2/139) had 3 altered parameters. This measure of pathway activation was associated with high tumor grade (p = 0.032), ER negative (p = 0.0008) and PR negative (p = 0.006) status (Table 1).

Figure 2. Venn diagram representing the overlap between AKT activation measured by pAKT positive staining, PTEN loss measured by PTEN negative staining, PIK3CA mutation analyzed by sequencing and PIK3CA copy number gain analyzed by qPCR. The figure shows that although there is a slight overlap between the alterations (24%), many cancers have only 1 altered parameter.

Of the 193 patients, with at least 1 altered parameter or wild type for all four, 124 had data for all 4 parameters analyzed. The clinicopathological characteristics of this subgroup of 124 patients were not statistically significantly different to the 193 group of patients (data not shown). All statistical analysis were then undertaken for this group to determine relationships with clinicopathological parameters, subtypes and outcome. The results and conclusions remained unchanged when the analysis was confined to this subgroup of 124 patients (data not shown).

PI3K pathway activation and breast cancer subtypes

When cases were distributed into 5 phenotypes according to ER/PR/HER2, CK5/6 and EGFR status,31, 38i.e. Luminal A (ER+ and/or PR+, HER2−), Luminal B (ER+ and/or PR+, HER2+), HER2 (ER− and PR−, HER2+), “basal-like” (ER−, PR−, HER2−, CK5/6+ and/or EGFR+) and unclassified (negative for all markers), there was no significant relationship with PIK3CA mutation or amplification. PTEN loss was associated with the “basal-like” phenotype (p < 0.001) and high PTEN expression with Luminal A cancers. High pAKT levels were associated with the HER2 phenotype (p = 0.04) and low pAKT levels with the Luminal A phenotype (p = 0.01). The activated pathway as measured by PIK3CA mutations, PIK3CA copy number gain, AKT activation or PTEN loss was associated with “basal-like” tumors (p = 0.01) and the inactive pathway with Luminal A cancers (p = 0.005) (Table 2).

Table 2. Association between the PI3K pathway and breast cancer subtypes

PI3K pathway activation and cell cycle markers

We next analyzed the relationship between components of the PI3K pathway and key cell cycle regulators that had previously been assessed in these cancers as downstream markers of pathway activation. PTEN loss was associated with increased cyclin E1 expression (p = 0.019) and low expression of cyclin D1 (p = 0.019), p21^Cip1/Waf1 (p = 0.019) and p27^Kip1 (p = 0.024). AKT activation was associated with increased cyclin E1 (p = 0.008) and p21^Cip1/Waf1 expression (p = 0.0005). An activated PI3K pathway was associated with high expression of cyclin E1 (p = 0.009), but not with expression of cyclin D1 (p = 0.158) or the Cdk inhibitors p27^Kip1 (p = 0.112) and p21^Cip1/Waf1 (p = 0.724) (Fig. 3).

Figure 3. Correlation between expression of cell cycle markers cyclin D1, cyclin E1, p21^Cip1/Waf1 and p27^Kip1 and PI3K pathway status. Box plots from a Mann–Whitney analysis of the association between an active or inactive PI3K pathway and expression, measured by immunohistochemistry, are shown.

PI3K pathway activation, disease outcome and treatment

In a univariate analysis an aberrant PI3K pathway was associated with breast cancer-specific death (HR = 4.17, 95% CI = 1.28–13.59, p = 0.017) (Fig. 4a). However, when the cohort was stratified on ER status the relationship approached significance for ER+ tumors (n = 117) (HR = 6.61, 95% CI = 0.86–50.38, p = 0.068) but not ER− tumors (n = 71) (HR = 1.40, 95% CI = 0.32–6.04, p = 0.649). In a multivariate analysis incorporating all clinicopathological and molecular parameters reported in Table 3, PI3K pathway activation was not an independent predictor of breast cancer-related death in a model that included lymph node status, tumor grade, PR and HER2 status (HR = 2.61, 95% CI = 0.80–8.57, p = 0.113). However, when tested in a multivariate model against pAKT the activated pathway was a better prognostic marker than pAKT alone (Table 3, model 2).

Figure 4. Kaplan–Meier curves illustrating the relationship between an activated PI3K pathway and disease progression and treatment. (a) Recurrence, metastasis and cancer-specific death stratified by an active (A: alteration of PIK3CA, PTEN or AKT) or inactive (I: wild type for parameters studied) pathway. Number of patients at risk, hazard ratios, 95% CI and p value of the association by log rank testing are shown. (b) Breast cancer-specific death in patients treated with adjuvant endocrine (tamoxifen) or chemotherapy and untreated patients, stratified by an active or inactive pathway.

Table 3. Multivariate analysis for clinicopathological parameters, hormone receptor status and PI3K pathway

Next, patients were distributed into 3 treatment categories, i.e. endocrine therapy, chemotherapy and no adjuvant treatment, and the associations between an active PI3K pathway and breast cancer-specific death were assessed. In patients treated with endocrine therapy (n = 93), an activated pathway was not associated with breast cancer-specific death (HR = 3.79, 95% CI = 0.50–28.79, p= 0.197). A similar result was apparent in the chemotherapy-treated patients (n = 75) (HR = 1.97, 95% CI = 0.58–6.76, p = 0.279). However, in the group of patients that received no adjuvant treatment (n = 67), an activated pathway was associated with a markedly inferior outcome to patients with no pathway aberrations where no patient relapsed. There is no p value as there are no events in the inactive pathway group (Fig. 4b).

The same association between an active PI3K pathway and breast cancer-specific death was then calculated as crude 5-year survival rate. For endocrine treated patients the crude 5-year survival rate was 69% (11/16), for chemotherapy treated patients it was 89% (16/18) and for patients with no adjuvant treatment it was 54% (6/11), of all breast cancer-specific death events recorded in each therapy group of patients.

Discussion

Previous articles indicate that the PI3K pathway is altered in a high proportion of breast cancer patients.3, 17, 39 Here, we report the first combined analysis of several pivotal molecules within the PI3K pathway in a single cohort of breast cancer patients in an attempt to derive an overall measure of pathway activation and relate this to clinicopathological features and outcome of the disease. We detected PIK3CA mutations or copy number gain, PTEN loss or AKT activation in >70% of breast cancer patients, an association between an active pathway and worse outcome and a high prevalence of alterations in the “basal-like” phenotype, which identifies the PI3K pathway as a potential therapeutic target in this subset of breast cancers.

We detected a relatively low frequency of PIK3CA mutations, i.e. 7% compared to earlier estimates of 12–40%.3–16 The use of sequencing is the standard method to detect mutations and although pyrosequencing has a higher analytical sensitivity which could aid in the mutation detection in DNA from tumor tissue contaminated with normal cells, all previous breast cancer studies have used sequencing as their detection method. Some additional mutations could be present in other exons not analyzed but exons 9 and 20 contain the highest frequency of PIK3CA mutations in cancer.14 Each mutation was confirmed in an independent PCR reaction. Our Australian cohort is representative of previously published population-based breast cancer cohorts, with respect to clinicopathological parameters such as tumor grade, tumor size, ER, PR and HER2 status and outcome (McNeil C.M., unpublished data). Therefore, neither the methodology nor the cohort characteristics explain this lower proportion of PIK3CA mutations. However, the frequency of copy number gain was 14%, higher than that detected in previous studies (1–9%).12, 16PIK3CA mutation and amplification were almost mutually exclusive events and since they represent independent mechanisms for elevating PI3K kinase activity, a higher frequency of copy number gain could perhaps explain the lower prevalence of PIK3CA mutations in this cohort compared to those in other reports. Also, the poor quality of the DNA derived from paraffin embedded samples and the failure to enrich for tumor cells before sequencing could be a reason for the low frequency of mutations reported here. We identified AKT activation in 24% of patients, slightly lower levels that some publications although the range of published data is highly variable5, 17, 19, 21, 22, 24–28 and a frequency of PTEN loss of 28%, similar to the average of previous reports on breast cancer patients.5, 6, 14, 17–23

In our study, PTEN loss was exclusive of PIK3CA mutations, as described previously,14 suggesting that these alterations represent redundant mechanisms for activating the PI3K pathway. Recent data have shown that the enhanced proliferation and survival resulting from PTEN loss are unaffected by PIK3CA knockdown and highlight an important role for the p110β isoform.40, 41 This may reflect different basal activities for p110α and β. However, it should be noted that in a recent paper, PTEN loss and PIK3CA mutation were often concordant in breast cancer patients,3 indicating additional work is required to clarify the functional contribution of these pathway aberrations as recently highlighted in the review of Yuan and Cantley.42 Furthermore, these workers3 identified that PTEN loss, but not PIK3CA mutation, was significantly associated with AKT activation. We report similar findings, where PIK3CA mutation and/or copy number gain was not associated with AKT S473 phosphorylation, in contrast to PTEN loss. These findings raise the possibility that alternative PI3K effectors to AKT, such as Rac, play an important role in PIK3CA mutated cancers.

To examine the overall activation status of the pathway, PIK3CA mutational status, PIK3CA copy number gain, PTEN loss and activation of AKT as measured by pAKT levels were combined and by this definition >70% of cancers had an activated pathway. Only 25% of cancers had overlapping alterations, emphasizing the need of an integrative analysis of the PI3K pathway. This activation was associated with a worse patient outcome although it was not an independent predictor of relapse or death from breast cancer implying that the reported associations with tumor grade (data not shown) and hormone receptor status were likely confounding. Pathway activation was identified in half of the hormone receptor-positive patients and was associated with an adverse outcome, although this did not reach statistical significance. In hormone receptor-negative disease pathway activation was identified in a higher percentage of cases (85%) but was not associated with outcome.

Confirmation of the activated pathway was measured by studying selected downstream cell cycle targets. The strongest association was between AKT activation, PTEN loss and pathway activation with increased cyclin E expression. This could be a rate limiting molecule in cell cycle progression due in part to a decrease of GSK3β phosphorylation, which also leads to stabilized cyclin E protein levels.43

The PI3K pathway was not predictive of response to endocrine therapy in our cohort of breast cancer patients; however, an association was identified when only PIK3CA status and AKT activation were combined (data not shown). This latter relationship is in agreement with previous data from in vitro studies which show that activation of PI3K/AKT pathway in breast cancer cells activates ERα and protects breast cancer cells from tamoxifen-induced apoptosis.44 Other studies have shown an association between individual members of the PI3K pathway and tamoxifen resistance.6, 26, 45 Despite 1 report showing PTEN loss was associated with relapse in patients treated with tamoxifen,6 we found no association, possibly because in our cohort PTEN is commonly lost in hormone receptor negative tumors and was thus less common in endocrine-treated patients. Analysis of pAKT and PIK3CA may be beneficial in identifying tamoxifen-resistant patients and this requires further testing in the context of a randomized treatment trial where treatment and outcome data are more robust. However, a poor outcome was also observed in patients not treated with adjuvant therapy, suggesting that aberrations in the PI3K pathway are associated with poor outcome independent of treatment, i.e. they are prognostic rather than predictive markers.

Another important discovery from the study was the association between an activated PI3K pathway and the “basal-like” phenotype of breast cancers. In HER2 cancers an activated pathway described by PIK3CA mutations or PTEN loss has been previously associated with a poor prognosis after trastuzumab therapy,46, 47 confirming the importance of the PI3K pathway in this subgroup of cancers and pointing to the need for further characterization of PI3K pathway inhibitors as complementary treatment strategies for trastuzumab-resistant disease. In “basal-like” cancers the PI3K pathway is emerging as an attractive candidate for targeted therapy. Previous studies have shown that PTEN inactivation leads to the development of mammary tumors that display pathological and molecular features associated with the “basal-like” subtype of breast cancer48, 49 and 2 further studies have shown an association of cyclin E overexpression and p27^Kip1 loss with a basal phenotype.50, 51 Furthermore, a recent publication has reported that PTEN alterations and an active PI3K pathway are associated with “basal-like” breast cancers.52 These data, together with the results presented here, point to the importance of further analyzing PI3K pathway aberrations in “basal-like” cancers and the effect of specific PI3K inhibitors on this cancer subtype.

In conclusion, over 70% of breast cancers in our study had an altered PI3K pathway which was associated with breast-cancer specific death, and we identified this pathway as a potential therapeutic target in “basal-like” breast cancers. Further studies of the integrated PI3K pathway need to be undertaken in larger cohorts of breast cancer patients, preferably in the context of randomized treatment trials, to confirm these findings.

Acknowledgements

The authors thank Ms. Alice Boulghourjian for assistance with processing the tissue, Ms. Anne Holliday for data management and Assoc. Prof. Adrienne Morey for help with the HER2 FISH analysis.