Abstract 3642: Phospho-CHK1 as a prognostic biomarker in ovarian cancer and a potential target in platinum-resistant disease

Purpose: Ovarian cancer frequently relapses due to inherent or acquired resistance to platinum-based therapy. To increase therapeutic response, the checkpoint kinases 1 and 2 (CHK1, CHK2) are being targeted in early clinical trials in combination with DNA damaging agents due to their roles in sensing DNA damage and inducing cell cycle arrest. Due to the limited studies on the use of biomarkers to overcome platinum resistance, we set out to characterize potential targets in a cohort of ovarian tumors and paired platinum sensitivity ovarian cancer cell lines. Methods: Total and phospho-protein expression were measured in a panel of 12 ovarian cancer cell lines using antibody arrays in order to characterize signaling pathways. Differentially expressed proteins were measured in a cohort of 128 pre-treatment malignant ovarian tumour lysates using reverse phase protein arrays. Sulforhodamine B cytotoxicity assays were used to assess the DNA damage response relevance of potential targets in paired platinum-sensitive and -resistant ovarian cancer cell lines. Results: Cell line expression clustering identified two groups that differed in patient history: predominantly platinum-based chemotherapy or chemotherapy-naïve background. The overexpressed proteins in the former cluster comprised mainly DNA damage response proteins including p-CHK1 (Ser317) and p-CHK2 (Ser516) implicating these proteins as potential mediators of platinum resistance. Though p-CHK2 had no prognostic value, high p-CHK1 levels were associated with poor overall survival in univariate analysis (medians 21 vs 38 months; corrected P=0.03). In multivariate analysis using a Cox proportional hazard model with other significant factors from univariate analysis (platinum sensitivity, stage, grade, residual disease, and histology), high p-CHK1 tumors had a relative risk of 3.0 (95% CI 1.1 - 8.0, P=0.03), 5.6 for platinum sensitivity (95% CI 2.8-11.1, P<0.001), 1.9 for tumor stage (95% CI 1.2-3.1, P=0.005), and 1.5 for histology (95% CI 1.1-1.9, P=0.004). CHK1 is phosphorylated at Ser317 during DNA damage and accordingly the DNA damage marker p-H2AX (Ser139) was highly expressed in the high p-CHK1 tumor group (P=0.008). The CHK1/2 inhibitor AZD7762 addition had variable effects on the cisplatin dose response when comparing the platinum-sensitive to the paired platinum-resistant cell lines. 50 nM AZD7762 (5.7% growth inhibition) and cisplatin treatment of the platinum resistant PEO4 cells induced a cisplatin concentration response (combination IC50=1.7μM vs cisplatin IC50=9.8μM) similar to that of its paired platinum sensitive PEO1 cell line (cisplatin IC50=2.7μM). Conclusion: This is the first study to identify p-CHK1 as an independent prognostic ovarian cancer biomarker and supports CHK1 as a therapeutic target in platinum-resistant ovarian cancer.

Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3642. doi:1538-7445.AM2012-3642

The role of tandem duplicator phenotype in tumour evolution in high-grade serous ovarian cancer

High-grade serous ovarian carcinoma (HGSOC) is characterized by genomic instability, ubiquitous TP53 loss, and frequent development of platinum resistance. Loss of homologous recombination (HR) is a mutator phenotype present in 50% of HGSOCs and confers hypersensitivity to platinum treatment. We asked which other mutator phenotypes are present in HGSOC and how they drive the emergence of platinum resistance. We performed whole-genome paired-end sequencing on a model of two HGSOC cases, each consisting of a pair of cell lines established before and after clinical resistance emerged, to describe their structural variants (SVs) and to infer their ancestral genomes as the SVs present within each pair. The first case (PEO1/PEO4), with HR deficiency, acquired translocations and small deletions through its early evolution, but a revertant BRCA2 mutation restoring HR function in the resistant lineage re-stabilized its genome and reduced platinum sensitivity. The second case (PEO14/PEO23) had 216 tandem duplications and did not show evidence of HR or mismatch repair deficiency. By comparing the cell lines to the tissues from which they originated, we showed that the tandem duplicator mutator phenotype arose early in progression in vivo and persisted throughout evolution in vivo and in vitro, which may have enabled continual evolution. From the analysis of SNP array data from 454 HGSOC cases in The Cancer Genome Atlas series, we estimate that 12.8% of cases show patterns of aberrations similar to the tandem duplicator, and this phenotype is mutually exclusive with BRCA1/2 carrier mutations.

Model-based global sensitivity analysis as applied to identification of anti-cancer drug targets and biomarkers of drug resistance in the ErbB2/3 network

High levels of variability in cancer-related cellular signalling networks and a lack of parameter identifiability in large-scale network models hamper translation of the results of modelling studies into the process of anti-cancer drug development. Recently global sensitivity analysis (GSA) has been recognised as a useful technique, capable of addressing the uncertainty of the model parameters and generating valid predictions on parametric sensitivities. Here we propose a novel implementation of model-based GSA specially designed to explore how multi-parametric network perturbations affect signal propagation through cancer-related networks. We use area-under-the-curve for time course of changes in phosphorylation of proteins as a characteristic for sensitivity analysis and rank network parameters with regard to their impact on the level of key cancer-related outputs, separating strong inhibitory from stimulatory effects. This allows interpretation of the results in terms which can incorporate the effects of potential anti-cancer drugs on targets and the associated biological markers of cancer. To illustrate the method we applied it to an ErbB signalling network model and explored the sensitivity profile of its key model readout, phosphorylated Akt, in the absence and presence of the ErbB2 inhibitor pertuzumab. The method successfully identified the parameters associated with elevation or suppression of Akt phosphorylation in the ErbB2/3 network. From analysis and comparison of the sensitivity profiles of pAkt in the absence and presence of targeted drugs we derived predictions of drug targets, cancer-related biomarkers and generated hypotheses for combinatorial therapy. Several key predictions have been confirmed in experiments using human ovarian carcinoma cell lines. We also compared GSA-derived predictions with the results of local sensitivity analysis and discuss the applicability of both methods. We propose that the developed GSA procedure can serve as a refining tool in combinatorial anti-cancer drug discovery.

GnRH receptor activation competes at a low level with growth signaling in stably transfected human breast cell lines

Background

Gonadotrophin releasing hormone (GnRH) analogs lower estrogen levels in pre-menopausal breast cancer patients. GnRH receptor (GnRH-R) activation also directly inhibits the growth of certain cells. The applicability of GnRH anti-proliferation to breast cancer was therefore analyzed.

Methods

GnRH-R expression in 298 primary breast cancer samples was measured by quantitative immunofluorescence. Levels of functional GnRH-R in breast-derived cell lines were assessed using ¹²⁵I-ligand binding and stimulation of ³H-inositol phosphate production. Elevated levels of GnRH-R were stably expressed in cells by transfection. Effects of receptor activation on in vitro cell growth were investigated in comparison with IGF-I and EGF receptor inhibition, and correlated with intracellular signaling using western blotting.

Results

GnRH-R immunoscoring was highest in hormone receptor (triple) negative and grade 3 breast tumors. However prior to transfection, functional endogenous GnRH-R were undetectable in four commonly studied breast cancer cell lines (MCF-7, ZR-75-1, T47D and MDA-MB-231). After transfection with GnRH-R, high levels of cell surface GnRH-R were detected in SVCT and MDA-MB-231 clones while low-moderate levels of GnRH-R occurred in MCF-7 clones and ZR-75-1 clones. MCF-7 sub-clones with high levels of GnRH-R were isolated following hygromycin phosphotransferase transfection. High level cell surface GnRH-R enabled induction of high levels of ³H-inositol phosphate and modest growth-inhibition in SVCT cells. In contrast, growth of MCF-7, ZR-75-1 or MDA-MB-231 clones was unaffected by GnRH-R activation. Cell growth was inhibited by IGF-I or EGF receptor inhibitors. IGF-I receptor inhibitor lowered levels of p-ERK1/2 in MCF-7 clones. Washout of IGF-I receptor inhibitor resulted in transient hyper-elevation of p-ERK1/2, but co-addition of GnRH-R agonist did not alter the dynamics of ERK1/2 re-phosphorylation.

Conclusions

Breast cancers exhibit a range of GnRH-R immunostaining, with higher levels of expression found in triple-negative and grade 3 cancers. However, functional cell surface receptors are rare in cultured cells. Intense GnRH-R signaling in transfected breast cancer cells did not markedly inhibit growth, in contrast to transfected HEK 293 cells indicating the importance of intracellular context. GnRH-R signaling could not counteract IGF-I receptor-tyrosine kinase addiction in MCF-7 cells. These results suggest that combinatorial strategies with growth factor inhibitors will be needed to enhance GnRH anti-proliferative effects in breast cancer

Features of the reversible sensitivity-resistance transition in PI3K/PTEN/AKT signalling network after HER2 inhibition

Systems biology approaches that combine experimental data and theoretical modelling to understand cellular signalling network dynamics offer a useful platform to investigate the mechanisms of resistance to drug interventions and to identify combination drug treatments. Extending our work on modelling the PI3K/PTEN/AKT signalling network (SN), we analyse the sensitivity of the SN output signal, phospho-AKT, to inhibition of HER2 receptor. We model typical aberrations in this SN identified in cancer development and drug resistance: loss of PTEN activity, PI3K and AKT mutations, HER2 overexpression, and overproduction of GSK3β and CK2 kinases controlling PTEN phosphorylation. We show that HER2 inhibition by the monoclonal antibody pertuzumab increases SN sensitivity, both to external signals and to changes in kinetic parameters of the proteins and their expression levels induced by mutations in the SN. This increase in sensitivity arises from the transition of SN functioning from saturation to non-saturation mode in response to HER2 inhibition. PTEN loss or PIK3CA mutation causes resistance to anti-HER2 inhibitor and leads to the restoration of saturation mode in SN functioning with a consequent decrease in SN sensitivity. We suggest that a drug-induced increase in SN sensitivity to internal perturbations, and specifically mutations, causes SN fragility. In particular, the SN is vulnerable to mutations that compensate for drug action and this may result in a sensitivity-to-resistance transition. The combination of HER2 and PI3K inhibition does not sensitise the SN to internal perturbations (mutations) in the PI3K/PTEN/AKT pathway: this combination treatment provides both synergetic inhibition and may prevent the SN from acquired mutations causing drug resistance. Through combination inhibition treatments, we studied the impact of upstream and downstream interventions to suppress resistance to the HER2 inhibitor in the SN with PTEN loss. Comparison of experimental results of PI3K inhibition in the PTEN upstream pathway with PDK1 inhibition in the PTEN downstream pathway shows that upstream inhibition abrogates resistance to pertuzumab more effectively than downstream inhibition. This difference in inhibition effect arises from the compensatory mechanism of an activation loop induced in the downstream pathway by PTEN loss. We highlight that drug target identification for combination anti-cancer therapy needs to account for the mutation effects on the upstream and downstream pathways.

Sprouty 2 is an independent prognostic factor in breast cancer and may be useful in stratifying patients for trastuzumab therapy

BACKGROUND

Resistance to trastuzumab is a clinical problem, partly due to overriding activation of MAPK/PI3K signalling. Sprouty-family proteins are negative regulators of MAPK/PI3K signalling, but their role in HER2-therapy resistance is unknown.

PATIENTS AND METHODS

Associations between Sprouty gene expression and clinicopathological features were investigated in a breast cancer microarray meta-analysis. Changes in expression of Spry2 and feedback inhibition on trastuzumab resistance were studied in SKBr3 and BT474 breast carcinoma cell lines using cell viability assays. Spry2 protein expression was measured by quantitative immunofluorescence in a cohort of 122 patients treated with trastuzumab.

RESULTS

Low gene expression of Spry2 was associated with increased pathological grade, high HER2 expression, and was a significant independent prognostic factor. Overexpression of Spry2 in SKBr3s resulted in enhanced inhibition of cell viability after trastuzumab treatment, and the PI3K-inhibitor LY294002 had a similar effect. Low Spry2 expression was associated with increased risk of death (HR = 2.28, 95% CI 1.22-4.26; p = 0.008) in trastuzumab-treated patients, including in multivariate analysis. Stratification of trastuzumab-treated patients using PTEN and Spry2 was superior to either marker in isolation.

CONCLUSION

In breast cancers with deficient feedback inhibition, combinatorial therapy with negative regulators of growth factor signalling may be an effective therapeutic strategy.

Trastuzumab and pertuzumab produce changes in morphology and estrogen receptor signaling in ovarian cancer xenografts revealing new treatment strategies

PURPOSE

The aim of this study was to investigate the antitumor effects of HER2-directed combination therapy in ovarian cancer xenograft models to evaluate their potential. The combinations of trastuzumab and pertuzumab, and trastuzumab and aromatase inhibitor therapy were investigated.

EXPERIMENTAL DESIGN

The effects of trastuzumab, pertuzumab, and letrozole on growth response, apoptosis, morphology, and gene and protein expression were evaluated in the SKOV3 ovarian cancer cell line xenograft and a panel of five human ovarian xenografts derived directly from clinical specimens.

RESULTS

The combination of HER2-directed antibodies showed enhanced antitumor activity compared with single antibody therapy in the SKOV3 xenograft model. Apoptosis, morphology, and estrogen-regulated gene expression were modulated by these antibodies in both spatial and temporal manners. A panel of ovarian cancer xenografts showed differential growth responses to the combination of trastuzumab and pertuzumab. High HER2 expression and increasing HER3 protein expression on treatment were associated with growth response. In trastuzumab-treated SKOV3 tumors, there was a change in tumor morphology, with a reduction in frequency of estrogen receptor alpha (ERα)-negative clear cell areas. Trastuzumab, but not pertuzumab, increased expression of ERα in SKOV3 xenografts when analyzed by quantitative immunofluorescence. ERα and downstream signaling targets were modulated by trastuzumab alone and in combination. Trastuzumab enhanced the responsiveness of SKOV3 xenografts to letrozole when given in combination.

CONCLUSIONS

These data suggest that trastuzumab in combination with pertuzumab could be an effective approach in high HER2-expressing ovarian cancers and could also enhance sensitivity to endocrine therapy in ERα-positive ovarian cancer.

Abstract P2-09-12: Phosphoproteomic Pathway Profiling of Breast Cancers for Biomarker and Target Discovery for Personalized Therapy

Introduction: The control and activation of signaling pathways has been shown to influence response to targeted therapy (PTEN and PI3K for trastuzumab, KRAS and MAPK for cetuximab). Single tissue biomarkers (eg HER2-receptor status) are often poor surrogates for the activation state of downstream pathways such as MAPK and PI3K. We have used phosphoproteomic profiling in order to reclassify breast tumors for therapeutic response prediction.

Methods: Eleven cell lines were profiled using phosphoproteome arrays (250 targets) and reclassified using unsupervised clustering. Sensitivity to therapy predicted by pathway activation status was correlated with sensitivity to therapies in vitro measured by the AlamarBlue cell viability assay. Target specificity was assayed using in-cell western blotting. Phosphoprotein clusters were validated in 107 breast tumors by reverse phase protein arrays (RPPA), and the expression of targets validated in independent cohorts of cancers by quantitative immunofluorescence (AQUA) on tissue microarrays.

Results: Phosphoproteomic profiling revealed 3 main clusters of breast tumors with overlap with existing molecular phenotypes (luminal, basal, and mixed) and different profiles of pathway activation (Cluster A: anti-apoptotic, Cluster B: growth factor-dependent, Cluster C: cell cycle-dependent). Good to excellent correlations between pathway activation and in vitro sensitivity to targeted therapies were shown for several pathways and individual molecules within pathways, including components of NFkB, STAT, and cell cycle. These biomarkers were shown to be present in clinical cohorts by RPPA and AQUA, and related to molecular phenotype. Conclusions: Phosphoproteomic profiling can be used to reclassify tumors in order to make a priori predictions about therapeutic targeting. These clusters exist in real clinical specimens and are therefore also promising biomarkers for targeted therapy.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-09-12.

Abstract P5-05-06: The Role of Sprouty 2 in HER2 Signaling in Breast Cancer; Decreased Expression Is Associated with Poor Outcome Including in Trastuzumab-Treated Patients

Purpose: Resistance to trastuzumab is a frequent clinical problem, in part due to the overriding activation state of the MAPK/PI3K signaling pathways. Sprouty-family proteins are negative regulators of MAPK/PI3K signaling, but their role in HER2 signaling and resistance to therapy is unknown.

Experimental design: The association of Sprouty 2 gene expression with grade, HER2-status, and survival was investigated in a meta-analysis of 1107 breast tumors from six published microarray studies. Sprouty regulation in response to HER2/HER3 signaling was studied using qRT-PCR. Changes in expression of Spry2 and feedback inhibition on trastuzumab-resistance were studied using full-length/dominant-negative transfection or chemical inhibition in SKBr3 and BT474 cell lines in cell viability assays. Finally, expression of Spry2 was measured in a cohort of 122 patients treated with trastuzumab by quantitative fluorescence microscopy (AQUA).

Results: Low Spry2 gene expression was associated with high HER2 protein expression. Spry2 was regulated through HER2/HER3 signaling as a delayed early gene, an effect reversed by treatment with the dimerization inhibitor pertuzumab. Overexpression of Spry2 in the SKBr3 trastuzumab-resistant cell line resulted in synergistic inhibition of cell viability with trastuzumab. Restitution of feedback inhibition with the PI3K-inhibitor LY294002 had a similar effect. Low Spry2 expression was associated with increased risk of death (HR = 2.28, 95% CI 1.22 - 4.26; p=0.008) in trastuzumab-treated patients, including in multivariate analysis (Cox regression analysis, p=0.002).

Conclusions: These results suggest that combinations of negative regulators of growth factor signaling may be an effective therapeutic strategy in breast cancers with deficient feedback inhibition.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P5-05-06.

Abstract P6-07-02: A Systems Analysis of Oscillations in the MAPK Pathway Reveals New Mechanisms of Feed-Forward and Feedback Signaling Control

Oscillations occur in diverse areas of cellular biology, from mechanotransduction to circadian control. Their physiological, pathological, and therapeutic significance in breast cancer remains undetermined, and their existence has yet to be described in the key oncogenic signaling pathways. Mathematical modeling lends itself to detailed study of these complex dynamic phenomena. We observed damped oscillations in pERK with approximately 20 minute periodicity in MCF7 breast carcinoma cell lines after stimulation with heregulin or epidermal growth factor (EGF). In order to establish mechanisms by which these oscillations might occur, we used an ordinary differential equation-based mathematical model of MAPK/PI3K signaling to simulate oscillatory behaviour in MAPK signaling. Both modification of feed-forward (HER2 overexpression) and feedback (abrogation of early gene transcription, EGFR phosphorylation by pERK, crosstalk between pAKT and RAF, and binding of pERK with SOS) loops resulted in the emergence of sustained oscillations in phospho-ERK. When these simulations were tested experimentally, MCF7 cells stably transfected with HER2 (MCF7/HER2-18) showed sustained oscillations in pERK when stimulated with heregulin, which were abrogated by inhibition of HER2 heterodimerisation with pertuzumab. Likewise, suppression of transcriptional feedback with cycloheximide resulted in sustained oscillations in pERK of increased amplitude. We have therefore described oscillations in MAPK signalling for the first time in breast cancer cell lines. Since the duration and amplitude of these oscillations is altered both by addition of the HER2 oncogene, and by inhibition of feedback control, the functional consequences of changes in oscillations may have fundamental consequences both in oncogenesis and therapeutic efficacy.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P6-07-02.

A gene on the HER2 amplicon, C35, is an oncogene in breast cancer whose actions are prevented by inhibition of Syk

Background:

C35 is a 12 kDa membrane-anchored protein endogenously over-expressed in many invasive breast cancers. C35 (C17orf37) is located on the HER2 amplicon, between HER2 and GRB7. The function of over-expressed C35 in invasive breast cancer is unknown.

Methods:

Tissue microarrays containing 122 primary human breast cancer specimens were used to examine the association of C35 with HER2 expression. Cell lines over-expressing C35 were generated and tested for evidence of cell transformation in vitro.

Results:

In primary breast cancers high levels of C35 mRNA expression were associated with HER2 gene amplification. High levels of C35 protein expression were associated with hallmarks of transformation, such as, colony growth in soft agar, invasion into collagen matrix and formation of large acinar structures in three-dimensional (3D) cell cultures. The transformed phenotype was also associated with characteristics of epithelial to mesenchymal transition, such as adoption of spindle cell morphology and down-regulation of epithelial markers, such as E-cadherin and keratin-8. Furthermore, C35-induced transformation in 3D cell cultures was dependent on Syk kinase, a downstream mediator of signalling from the immunoreceptor tyrosine-based activation motif, which is present in C35.

Conclusion:

C35 functions as an oncogene in breast cancer cell lines. Drug targeting of C35 or Syk kinase might be helpful in treating a subset of patients with HER2-amplified breast cancers.

Dynamic changes in gene expression in vivo predict prognosis of tamoxifen-treated patients with breast cancer

INTRODUCTION

Tamoxifen is the most widely prescribed anti-estrogen treatment for patients with estrogen receptor (ER)-positive breast cancer. However, there is still a need for biomarkers that reliably predict endocrine sensitivity in breast cancers and these may well be expressed in a dynamic manner.

METHODS

In this study we assessed gene expression changes at multiple time points (days 1, 2, 4, 7, 14) after tamoxifen treatment in the ER-positive ZR-75-1 xenograft model that displays significant changes in apoptosis, proliferation and angiogenesis within 2 days of therapy.

RESULTS

Hierarchical clustering identified six time-related gene expression patterns, which separated into three groups: two with early/transient responses, two with continuous/late responses and two with variable response patterns. The early/transient response represented reductions in many genes that are involved in cell cycle and proliferation (e.g. BUB1B, CCNA2, CDKN3, MKI67, UBE2C), whereas the continuous/late changed genes represented the more classical estrogen response genes (e.g. TFF1, TFF3, IGFBP5). Genes and the proteins they encode were confirmed to have similar temporal patterns of expression in vitro and in vivo and correlated with reduction in tumour volume in primary breast cancer. The profiles of genes that were most differentially expressed on days 2, 4 and 7 following treatment were able to predict prognosis, whereas those most changed on days 1 and 14 were not, in four tamoxifen treated datasets representing a total of 404 patients.

CONCLUSIONS

Both early/transient/proliferation response genes and continuous/late/estrogen-response genes are able to predict prognosis of primary breast tumours in a dynamic manner. Temporal expression of therapy-response genes is clearly an important factor in characterising the response to endocrine therapy in breast tumours which has significant implications for the timing of biopsies in neoadjuvant biomarker studies.

Cancer systems biology

Cancer is a complex and heterogeneous disease, not only at a genetic and biochemical level, but also at a tissue, organism, and population level. Multiple data streams, from reductionist biochemistry in vitro to high-throughput "-omics" from clinical material, have been generated with the hope that they encode useful information about phenotype and, ultimately, tumour behaviour in response to drugs. While these data stand alone in terms of the biology they represent, there is the enticing prospect that if incorporated into systems biology models, they can help understand complex systems behaviour and provide a predictive framework as an additional tool in understanding how tumours change and respond to treatment over time. Since these biological data are heterogeneous and frequently qualitative rather than quantitative, at the present time a single systems biology approach is unlikely to be effective; instead, different computational and mathematical approaches should be tailored to different types of data, and to each other, in order to test and re-test hypotheses. In time, these models might converge and result in usable tractable models which accurately represent human cancer. Likewise, biologists and clinicians need to understand what the requirements of systems biology are so that compatible data are produced for computational modelling. In this review, we describe some theoretical approaches (data-driven and process-driven) and experimental methodologies which are being used in cancer research and the clinical context where they might be applied.

Modulation of HER3 is a marker of dynamic cell signaling in ovarian cancer: implications for pertuzumab sensitivity

This study was designed to evaluate the expression of HER receptors as a marker of sensitivity to the humanized anti-HER2 monoclonal antibody pertuzumab in ovarian cancer cells. In a recent clinical trial, low levels of HER3 mRNA have been shown to associate with pertuzumab response when combined with gemcitabine. We sought to define how pertuzumab modulated HER expression levels in ovarian cancer using cell line models to better understand differential and dynamic receptor expression in therapeutic response. Changes in HER3 mRNA expression were also assessed in pertuzumab-treated xenografts. HER3 mRNA and, to a lesser extent, HER2, were down-regulated after stimulation both with heregulin-beta1 and epidermal growth factor in a range of ovarian cancer cell lines either growth sensitive or growth resistant to pertuzumab. Pertuzumab reversed this down-regulation and the magnitude of the reversal correlated with pertuzumab sensitivity. The change in HER3 mRNA expression correlated inversely to how much the extracellular signal-regulated kinase and phosphoinositide 3-kinase pathways were dynamically activated with stimulation. Finally, up-regulation of HER3 mRNA was found in cancer xenografts treated with pertuzumab. We conclude that HER3 mRNA is down-regulated by both heregulin-beta1 and epidermal growth factor activation. This suggests that in some tumors, low HER3 mRNA expression is driven by, or dependent on, growth factor. HER3 mRNA expression is effectively reversed in pertuzumab-sensitive tumors. These data are consistent with low HER3 mRNA identifying a pertuzumab-sensitive phenotype.

Functional restoration of BRCA2 protein by secondary BRCA2 mutations in BRCA2-mutated ovarian carcinoma

Acquired platinum resistance is a serious problem in the treatment of ovarian carcinomas. However, the mechanism of the drug resistance has not been elucidated. Here, we show functional significance of restoration of BRCA2 protein by secondary BRCA2 mutations in acquired drug resistance of BRCA2-mutated ovarian carcinoma. Three ovarian cancer cell lines (PEO1, PEO4, and PEO6) were derived from a BRCA2 mutation [5193C>G (Y1655X)] carrier with ovarian carcinoma with acquired cisplatin resistance and a secondary BRCA2 mutation [5193C>T (Y1655Y)] that canceled the inherited mutation. PEO1 was BRCA2 deficient and sensitive to cisplatin and a poly(ADP-ribose) polymerase inhibitor, AG14361, whereas PEO4 was resistant. PEO4 and PEO6, derived from ascites at the time of relapse with cisplatin resistance, had the secondary mutation and were BRCA2 proficient. In vitro cisplatin/AG14361 selection of PEO1 led to restoration of BRCA2 due to another secondary BRCA2 mutation. BRCA2 depletion sensitized BRCA2-restored PEO1 clones and PEO4 to cisplatin/AG14361. Thus, restoration of BRCA2 due to secondary BRCA2 mutation is involved in acquired drug resistance of BRCA2-mutated ovarian carcinoma.

How can systems pathology help us personalize cancer therapy?

Cancer is a complex and heterogeneous disease which changes over time, and in the face of therapeutic intervention. Single tissue biomarkers, while partially successful in helping us understand which patients will respond to therapy, cannot hope to capture this amazing complexity. Systems pathology, which combines measurements made on tissues with new mathematical modeling approaches, permits the testing of new agents and biomarkers in silico through computational analysis. These approaches help us to refine pathological measurements and improve decision making about therapies for clinical trial planning and ultimately personalized therapy.

Gonadotropin-releasing hormone analog structural determinants of selectivity for inhibition of cell growth: support for the concept of ligand-induced selective signaling

GnRH and its receptor are expressed in human reproductive tract cancers, and direct antiproliferative effects of GnRH analogs have been demonstrated in cancer cell lines. The intracellular signaling responsible for this effect differs from that mediating pituitary gonadotropin secretion. The GnRH structure-activity relationship is different for the two effects. Here we report a structure-activity relationship study of GnRH agonist antiproliferative action in model cell systems of rat and human GnRH receptors stably expressed in HEK293 cells. GnRH II was more potent than GnRH I in inhibiting cell growth in the cell lines. In contrast, GnRH I was more potent than GnRH II in stimulating inositol phosphate production, the signaling pathway in gonadotropes. The different residues in GnRH II (His(5), Trp(7), Tyr(8)) were introduced singly or in pairs into GnRH I. Tyr(5) replacement by His(5) produced the highest increase in the antiproliferative potency of GnRH I. Tyr(8) substitution of Arg(8) produced the most selective analog, with very poor inositol phosphate generation but high antiproliferative potency. In nude mice bearing tumors of the HEK293 cell line, GnRH II and an antagonist administration was ineffective in inhibiting tumor growth, but D-amino acid stabilized analogs (D-Lys(6) and D-Arg(6)) ablated tumor growth. Docking of GnRH I and GnRH II to the human GnRH receptor molecular model revealed that Arg(8) of GnRH I makes contact with Asp(302), whereas Tyr(8) of GnRH II appears to make different contacts, suggesting these residues stabilize different receptor conformations mediating differential intracellular signaling and effects on gonadotropin and cell growth. These findings provide the basis for the development of selective GnRH analog cancer therapeutics that directly target tumor cells or inhibit pituitary gonadotropins or do both.

Antiestrogen therapy is active in selected ovarian cancer cases: the use of letrozole in estrogen receptor-positive patients

PURPOSE

To evaluate the efficacy of the aromatase inhibitor letrozole in preselected estrogen receptor (ER)-positive relapsed epithelial ovarian cancer patients and to identify markers that predict endocrine-sensitive disease.

EXPERIMENTAL DESIGN

This was a phase II study of letrozole 2.5 mg daily until clinical or marker evidence of disease progression in previously treated ER-positive ovarian cancer patients with a rising CA125 that had progressed according to Rustin's criteria. The primary end point was response according to CA125 and response evaluation criteria in solid tumors (RECIST) criteria. Marker expression was measured by semiquantitative immunohistochemistry in sections from the primary tumor.

RESULTS

Of 42 patients evaluable for CA125 response, 7 (17%) had a response (decrease of >50%), and 11 (26%) patients had not progressed (doubling of CA125) following 6 months on treatment. The median time taken to achieve the CA125 nadir was 13 weeks (range 10-36). Of 33 patients evaluable for radiological response, 3 (9%) had a partial remission, and 14 (42%) had stable disease at 12 weeks. Eleven patients (26%) had a PFS of >6 months. Subgroup analysis according to ER revealed CA125 response rates of 0% (immunoscore, 150-199), 12% (200-249), and 33% (250-300); P = 0.028, chi(2) for trend. Expression levels of HER2, insulin-like growth factor binding protein 5, trefoil factor 1, and vimentin were associated with CA125 changes on treatment.

CONCLUSIONS

This is the first study of a hormonal agent in a preselected group of ER-positive ovarian cancer patients. A signature of predictive markers, including low HER2 expression, predicts response.

Progressive loss of estrogen receptor alpha cofactor recruitment in endocrine resistance

Differential expression of estrogen receptor-alpha (ERalpha) cofactors has been implicated in endocrine resistance in breast cancer. Using a three-stage MCF-7 cell-based model that emulates the clinical manifestation of acquired endocrine resistant breast cancer we now show, using a combination of chromatin immunoprecipitation and RNA interference, that there is a progressive loss of ERalpha cofactor recruitment to the estrogen-dependent pS2 gene and reduced requirement for cofactor expression. Maximal estrogen induced pS2 induction requires ERalpha and cofactor recruitment in MCF-7 cells, but in the progression to endocrine resistance these requirements are altered and expression has become less dependent on cofactors. Additionally, in estrogen-resistant MCF-7 cells there is a global loss of requirement of individual cofactors for proliferative cell growth indicating that other genes have lost the need for transcriptional cofactors. This loss of the requirement for cofactors may represent an important mechanism for gene misregulation in cancer.