Final Report Summary - RATHER (Rational Therapy for Breast Cancer: Individualized Treatment for Difficult-to-Treat Breast Cancer Subtypes)
Executive Summary:
RATHER (Rational therapy for breast cancer: individualised treatment for difficult-to-treat breast cancer subtypes) aimed to identify new therapeutic targets for two difficult to treat subtypes of breast cancer. A quarter of all breast cancer patients are diagnosed with these difficult-to-treat subtypes of breast cancer; Invasive Lobular Carcinoma (ILC), an aggressive form of breast cancer that is resistant to standard therapies (10% of cases globally), or Triple Negative (TN) breast cancer which lacks the common targets for treatment; estrogen, progesterone and HER2 receptors (15% of cases globally). Both are linked with poor prognosis and there are currently no effective therapies available.
RATHER was a large multidisciplinary collaborative project funded by the EU under the FP7 programme (€6 million) led by University College Dublin. The consortium comprised 8 partners from the academic and private sectors; University College Dublin (Ireland), OncoMark (Ireland), Agendia (the Netherlands), Netherlands Cancer Institute (the Netherlands), Institut Curie (France), Vall d’Hebron Hospital (Spain), Cambridge University (UK), and Lund University (Sweden).
The RATHER consortium aimed to better understand difficult-to-treat breast cancers by studying kinases, which are activated in many cancer types and which have been shown to play a role in cancer. In cancer, there is a strong correlation between genetic mutations and a patient’s response to drugs. Therefore, the RATHER consortium looked to investigate mutations in kinases to deliver personalised therapeutic interventions to the 25% of patients who currently lack effective targeted therapies.
To survey kinases to identify novel biomarkers for new molecular diagnostic assays or novel drug targets for new therapeutic approaches, the consortium used a multidisciplinary approach applying comprehensive genomic, transcriptomic and functional analysis and data integration, leading to the validation of targets / biomarkers, the selection of lead candidates, and the initiation of a Phase Ib/II clinical trial.
The RATHER project integrated clinical and preclinical research, establishing collaborations between academia and industry delivering a number of important results and outputs throughout the duration of the project. The results from the RATHER project have been published in international scientific journals and presented at international conferences, with a number of manuscripts still in preparation pending further validation of results. Work on the promising targets identified during the project will be brought to completion beyond RATHER, facilitated by the relationships and collaborations established during this project.
Some major outputs from the RATHER project include one ongoing clinical trial (POSEIDON) directly funded by the RATHER project, 3 clinical trials entered into by RATHER consortium partners (GELATO, I-SPY2 and SUBITO), a validated assay for ILC (MammaPrint), a new biomarker signature for ILC (immune/hormone related), a new biomarker signature for TN breast cancer (BRCAness), a new biomarker for TN breast cancer (CDK7), a new treatment strategy for ILC (MAP3K1/MAP2K4 genotype-dependent) and a new treatment strategy for TN breast cancer (CDK7 & BCL-2/XL-targeted).
In summary, the RATHER consortium have published 27 articles (19 peer-reviewed original articles, 3 review articles, 4 abstracts) and 1 book chapter, have delivered 57 presentations (32 international, 25 national) and presented 28 posters. Four patents have been filed.
Project Context and Objectives:
In the RATHER project, we aimed to identify and validate novel kinase targets for therapy for TN and ILC subtypes of breast cancer, as well as to develop molecular diagnostic assays and test a novel kinase inhibitor in a clinical trial.
Overall RATHER Objectives:
• To identify novel kinase targets for therapy
• To validate these targets within in vitro preclinical models
• To develop a prognostic gene expression profile for ILC
• To develop predictive and prognostic clinical assays for breast cancer
• To test a PI3K inhibitor +/- endocrine therapy in a Phase I/II clinical trial
• To generate new knowledge, protect resulting intellectual property and commercialise findings (SMEs)
WP1: Selection of suitable tissues, processing of materials and collection of clinical data
In this WP, we aimed to select 150 TN and 150 ILC tumour samples from the frozen tissue collections of NKI and CAM and isolate DNA, RNA and protein lysates from these samples. Matching formalin-fixed, paraffin-embedded (FFPE tissues were also collected for TMA construction.
• Task 1.1. Selection of tumour samples for molecular analyses
• Task 1.2. DNA, RNA and protein extraction of tumour samples
• Task 1.3. Patient data collection
WP2: Kinome re-sequencing 150 ILC and 150 TN breast cancer
Here, we aimed to sequence all 518 kinases and 68 additional genes that are frequent targets of mutation in breast cancer in all 300 tumour samples, using high-throughput next generation sequencing, together with exon capture technology.
• Task 2.1. Generation of raw kinome sequence data
• Task 2.2. Analysis of raw large-scale genomic sequence data
• Task 2.3. Validation of mutations
WP3: Reverse phase protein lysate arrays (RPPA) on ILC and TN tumour samples
In WP3, we used RPPA technology to assess the activation status of kinases as judged by phospho-epitope quantitation.
• Task 3.1. Validation of additional phosphor-specific antibodies
• Task 3.2. Retrieval of protein lysates and concentration equilibration
• Task 3.3. Protein array generation and antibody revelation
• Task 3.4. Quantification and analysis of the data
WP4: Tissue microarray (TMA) analysis of ILC and TN tumour samples
In WP4, we validated findings from WP3 on TMAs to validate that the differential activations seen in the previous WP take place in the tumour cells rather than in the stroma.
• Task 4.1. TMA construction
• Task 4.2. Protein expression analysis
• Task 4.3. In situ hybridization (ISH) on TMAs
• Task 4.4. Manual and automated image analysis of TMA data
WP5: Genomic analysis of ILC and TN tumours using SNP arrays to evaluate copy number and allele variation
In WP5, we comprehensively analysed copy number gains and losses in these tumours, with a special emphasis on gains and losses in loci encoding kinases.
• Task 5.1. Data pre-processing
• Task 5.2. Copy number estimation and determination of loss of heterozygosity (LOH) status
• Task 5.3. Identification of summarized regions of interest
• Task 5.4. Integrated measurement of copy number and gene expression
WP6: Gene expression analysis using DNA microarrays on ILC and TN tumour samples
In WP6, we performed full genome gene expression analysis of all 300 tumours via gene arrays and RNA sequencing. In addition, a commercial predictive assay for lobular cancers was also developed.
• Task 6.1. Identification of predictive genes and integrative signatures
• Task 6.2. Development of prognostic profile for ILC
• Task 6.3. RNA sequencing analysis
WP7: Validation of kinase targets for therapy in preclinical models
In WP7, we used siRNA, shRNA and small molecule/antibody-based drugs to validate the candidate kinase targets for therapy in appropriate cell line models of TN / ILC breast cancer.
• Task 7.1. Identifying driver mutations in breast cancer cell line models
• Task 7.2. Finding small molecule (antibody-based) drugs that are toxic in cells having oncogenic driver mutations
• Task 7.3. Selecting optimal kinase targets for guidance of kinase inhibitor treatment in the clinical trial
WP8: Development of standardized biomarker assays for breast cancer patient stratification
In WP8, the SMEs developed biomarker assays for breast cancer patients, particularly for ILC patients.
• Task 8.1. Development of specific companion diagnostic assays
• Task 8.2. Development of a prognostic test for ILC
• Task 8.3. Validation of a unique mechanistically-anchored breast cancer biomarker panel
• Task 8.4. Digital image analysis of kinase alterations
WP9: Initiation of phase I/II clinical trial to evaluate tumour genotype-drug response relationships
In WP9, we are performing a phase I/II clinical trial (POSEIDON) to study the efficacy of a selective PI3K inhibitor in ductal and lobular breast cancer patients.
• Task 9.1. Phase Ib dose-escalation study
• Task 9.2. Randomised phase II trial
• Task 9.3. Translational substudies
WP10: Data Integration
In WP10, we performed an integrative bioinformatics analysis on the data generated in WPs 2-6. The outcome of this will be a list of candidate targets for therapy in either ILC or TN breast cancer.
• Task 10.1. Integration of data on pathway control / systems biology
WP11: Project and database management
In WP11, the coordinator and project management team were responsible for managing the project, including all reporting requirements, and the database.
• Task 11.1. Project management
• Task 11.2. Preparation of website and communications portal
• Task 11.3. Database management
• Task 11.4. Ethics management
Project Results:
Sample collection and distribution was a key part of the RATHER project and the main aim of WP1. Tumour samples were selected from NKI and CAM biobanks based on a number of pre-specified criteria. A total of 145 ILC and 176 TN samples were selected and then processed for RNA extraction, DNA extraction and protein extraction, and distributed to partners along with matching clinical, survival and pathology data. The DNA was used for sequencing in WP2 and SNP array analysis in WP5, the RNA was used for gene expression analysis in WP6 and the extracted protein was used for RPPA analysis in WP3. Formalin-fixed paraffin embedded (FFPE) tissue blocks were also collected for all patients to generate TMAs for use in WP4. The clinical and survival data allowed correlation of kinome mutations with clinical outcome.
WP2 aimed to identify signalling pathways that drive the development of breast cancer and those that influence therapy response through the sequencing of the ILC and TN tumour samples obtained in WP1. NKI have developed a capture system that enables high throughput sequencing of all protein kinases, common oncogenes and tumour suppressor genes, a selection of lipid kinases and components of the PI3K pathway. NKI successfully completed sequencing for 143 ILC and 150 TN samples and identified a number of kinase mutations; the most common mutations found in ILC were CDH1 and PIK3CA; in TN the most common found was TP53. Overall, in ILC a total of 987 protein altering somatic variations were identified in 393 genes while TN samples had 1,208 protein-altering somatic variants in 425 genes. The results suggest that understanding the spectrum of mutations in breast cancer will help to stratify patients.
In WP3, reverse phase protein lysate array (RPPA) analysis was used to examine protein expression and phosphorylation of kinases. The RATHER project allowed Institut Curie to significantly increase the amount of validated antibodies that can be applied to RPPA. Over 500 new primary antibodies were tested over the course of the project, with over 150 validated for use in RPPA. These antibodies have not only been used in the RATHER project but continue to be used for many other cancer-related projects. RPPA data was used to identify differences in protein expression / activation between ILC and TN tumours and to identify subgroups within ILC. A number of proteins were identified that are differentially expressed between ILC and TN breast cancers. 4 subgroups of ILCs were identified that show biologically distinct features in terms of cell signalling pathway activation status. Proteins identified may have a prognostic value in IHC. Activation of the PI3K / Akt pathway and its correlation with patient survival was determined by RPPA, which further justified the clinical interest of the POSIEDON clinical trial (WP9). Additionally, a list of (phopho-) proteins associated with patient outcome was established that could potentially serve as prognostic biomarkers of therapeutic targets. 10 (phospho-)proteins of which a higher expression was associated with a better prognosis were identified and 9 (phospho-) proteins associated with a worse outcome were also identified (Michaut et al., 2016). Protein biomarkers identified have to be validated in an independent cohort of ILCs, which is ongoing in a new project in Institut Curie outside of the RATHER project.
The use of immunohistochemistry (IHC) in clinical cohorts is of critical importance in determining the utility of a biomarker in clinical practice. Despite the wide use of IHC as a biomarker validation tool, no universally accepted standardization guidelines have been developed to determine the acceptability of particular antibodies for IHC prior to its use. In WP4, we developed an IHC and Western blot (WB) workflow to ensure antibody specificity. Following successful antibody validation, we progressed to IHC on breast cancer tissue microarrays (TMA) which allowed the evaluation of the expression level, mutation and activation status of kinases. TMAs were constructed containing 142 ILC and 152 TN patient samples, in combination with clinical information, that were available for use within the whole RATHER project. Higher levels of AKT-3 amplification were identified in TN samples in comparison to ER-positive breast cancers. AKT-3 amplification along with protein expression was negatively associated with Recurrence Free Survival (RFS) in TN. It is hypothesised that AKT-3 amplification could represent a potentially relevant oncogenic event in a subset of TN breast cancers that may select cells sensitive to Akt-3 inhibitors. A scientific publication describing the molecular alterations of Akt-3 has been published (O’Hurley et al., 2013) in addition to a review article (O’Hurley et al., 2014) and a book chapter (Linn et al., 2013) discussing biomarkers and the use of IHC for biomarker validation.
As cancer is a genetic disease with alterations in DNA copy number characteristic of solid tumours, WP5 aimed to profile the genomic alterations for both ILC and TN tumour samples. Gains and losses were seen in both ILC and TN samples, with significant differences identified between ILC and TN breast cancers. The ILC data was integrated in WP10 and included in the Michaut et al., 2016 publication that presents an integrated molecular portrait of ILC.
WP6 aimed to quantify kinase expression levels to discover important information about the relative abundance of kinases and therefore, their relative activity. Transcriptomic analysis was used to quantify and identify gene signatures predictive of pathway alterations. The expression levels of 41 kinase or kinase-related genes were shown to be correlated with ILC patient outcome. These results provide a list of interesting targets for downstream validation. Additional work was carried out in WP6 on TN breast cancer. Approximately 10-15% of TN breast cancers are mutant for the tumour suppressor BRCA1. Deficiencies in BRCA1 result in genomic instability. These tumours are classified as either BRCA1-like or non-BRCA1-like, with the BRCA1-like subgroup having a significantly worse prognosis. In the WP6, we identified differences in gene expression, mutation and prognosis between BRCA1-like and non-BRCA1-like tumours, identifying a ‘BRCAness’ (BRCA1-like) signature providing insight into the genomic instability in BRCA1-like samples. Genes most commonly upregulated in BRCA1-like tumours included those associated with DNA replication, recombination and repair suggesting they may be susceptible to therapies targeted to DNA and cell cycle pathways such as PARPi. This work was published in Molecular Oncology (Severson, TM et al., 2015) and a patent has been submitted for the 77 BRCAness signature genes. Agendia is currently focussed on developing a companion diagnostic test based on this BRCAness signature. The BRCAness signature has also been accepted as a qualifying biomarker in the ISPY2 trial (Investigation of Serial Studies to Predict Your Therapeutic Response with Imaging and Molecular Analysis) and is currently being assessed to predict response in patients treated with the PARPi-inhibitor ABT-888. Results are outlined in a recent publication in NPJ Breast Cancer (Wolf, D. et al., 2017). BRCA1-like validation in the I-SPY2 trial has been instrumental for the start of the phase III SUBITO trial (NCT02810743), ‘Substantially improving the cure rate of high risk BRCA1-like breast cancer’. If a positive SUBITO study outcome is seen, high-dose alkylating chemotherapy with autologous stem cell rescue will be reintroduced and reimbursed for stage II, HER2 negative, BRCA1-like breast cancer patients in the Netherlands. WP6 also aimed to develop and validate a new profile, or to optimise existing profiles such as MammaPrint, specifically for the ILC subgroup. MammaPrint is a prognostic breast cancer test developed and commercialised by Agendia, which predicts the risk of breast cancer recurrence based on gene expression analysis. This study looked to identify lobular tumours that require more aggressive intervention i.e. patients who are like to develop early metastasis (within 5 years), as these patients can be most successfully treated with systemic adjuvant chemotherapy. MammaPrint’s prognostic profile was tested for it’s predictive power within a small set of ILC samples. Results indicated that MammaPrint may be sufficient in identifying a subgroup of ILC patients that would benefit from more aggressive treatment. The performance of MammaPrint in a larger set of ILC samples was carried out in WP8.
WP7 aimed to validate kinase targets in preclinical models for therapy for ILC and TN breast cancer. Kinase targets identified in WPs 2-6 were functionally interrogated to analyse whether they drivers of the oncogenic phenotype in TN and ILC. CDK7 was found to be closely linked to poor prognosis in the TN breast cancer context, with evidence demonstrated in the RATHER project of elevated CDK7 expression as a candidate biomarker of poor prognosis and candidate therapeutic target in TN breast cancer. We also offer a preclinical proof of concept for combining CDK7 and BCL-2/BCL-XL inhibitors as a mechanism-based therapeutic strategy to improve TNBC treatment. These results are outlined in a publication in Cancer Research (Li, B. et al., 2017). UCD are now working with Carrick Therapeutics to aid development of their novel CDK7 inhibitor. In WP7, a new potential treatment strategy was also proposed for the treatment of ILC, identifying candidate drug treatment regimens for ILC based on their genotype. We predicted that ILCs that have a mutation in either component of their JNK pathway will be sensitive to MEKi monotherapy, but that tumours that are wild-type for both MAP2K4 and MAP3K1 will require co-treatment with a pan-HER inhibitor such as Dacomitnib. Validation of these results in animal models of ILC is required before translation into clinical therapies. An article detailing this work has been published in Cell Research (Xue, Z., et al. 2018) and NKI are collaborating with a pharmaceutical company to conduct future clinical trials.
The primary aim of WP8 was to develop standardised biomarker assays to be used in Phase I/II clinical trials or in other diagnostic settings to facilitate patient stratification for receipt of relevant kinase inhibitors or alternative therapies. A secondary aim of this WP was to examine the utility of the developed array-based prognostic tool for ILC. Agendia are currently developing a companion diagnostic test for the BRCAness signature for TN breast cancer (output of WP6). They have also studied the prognostic value of the MammaPrint test for breast cancer patients with early-stage ILC with study results validating MammaPrint as an independent factor for this subgroup of patients. The increase in incidence of ILC over the past decades and the modest representation of ILC in the MammaPrint development dataset called for a stratified survival analysis dedicated to this specific subgroup. Under RATHER, Agendia successfully completed the first steps of translating MammaPrint from microarray to a Next generation Sequencing Platform, a technology already validated in the clinical context. Outside of the RATHER project, Agendia has further developed MammaPrint on NGS and received the necessary CE mark followed by a soft launch of the product in February 2018. OncoMark and UCD worked on the validation of a unique mechanistically-anchored breast cancer biomarker panel (OncoMasTR) that showed great prognostic potential for early stage breast cancer patients, predicting whether they could safely forego chemotherapy. OncoMark separately validated the potential of 3 biomarkers within this panel (plus 3 control genes) in a gene-expression based assay. OncoMark received separate funding from an SME Instrument grant and private investors and have developed this signature into a qPCR test which has been successfully clinically validated. This test has recently been CE marked. UCD examined the potential of 4 biomarkers from the panel in an IHC-based prognostic assay, with further funding awarded from Science Foundation Ireland to complete the validation studies initiated during RATHER.
A major success for the RATHER consortium was the launch of POSEIDON, a Phase Ib/II clinical trial (WP9) to evaluate the safety and effectiveness of GDC-0032 when given alongside tamoxifen. This trial is sponsored by NKI in collaboration with Genentech, EurocanPlatform and the RATHER project. It combines traditional endocrine therapy (Tamoxifen) with a novel drug GDC-0032 (Taselisib), an isoform selective inhibitor of a kinase (PI3K). This novel drug inhibits PI3K, a kinase that is often mutated in ILC. This therapeutic approach has the potential to halt the growth of cancer driven by this kinase. Previous trials with non-selective PI3K pathway inhibitors have experienced relatively low single agent anti-tumour activity, as well as toxicities. Therefore, the isoform-selective inhibitor (GDC-0032 / Taselisib) shows promising activity and low toxicity in Phase Ia clinical trials. The trial is currently ongoing and will extend past the RATHER project.
Phase Ib has been completed, with the recruitment of 30 patients. Taselisib in combination with tamoxifen was generally well tolerated with no unexpected toxicities observed. A recommended dose for phase II was determined as 4mg Taselisib for 21 days in combination with 20mg Tamoxifen continuously. Phase II is ongoing. To date, 90 patients have been recruited to Phase II of the trial at a number of hospital sites across the Netherlands, the UK and Spain, with the first site to open imminently in France.
The POSEIDON trial will extend beyond the RATHER project; however, it has already provided some relevant outputs:
• It is Genentech’s first ever investigator-led multinational trial in Europe (and first non-US).
• Results for Phase Ib have been presented at ASCO 2016 (Baird, R.).
• Recruitment of Phase Ib was completed in approximately 13 months, demonstrating that clinical trials with academic leadership can be competitive despite restrictions.
• Biopsies from 10 of 30 patients in Phase Ib have been obtained to perform molecular studies, impressive data when compared to competitive trials promoted by pharma companies.
WP10 represented the key integrating force within the RATHER consortium, specifically focussing on analysing and integrating data from the other WPs, with all data managed in a relational database developed in WP11. The main objective of WP10 was to develop and apply a computational framework to enable the integration of data from the different WPs (particularly WPs 2-6). Publicly available data was also integrated with the data generated from the RATHER project. An integrated molecular portrait of ILC was established, containing a comprehensive view of the genomic, transcriptomic and proteomic landscapes of the RATHER biological samples, with a particular focus on the kinome. These results have been published in Scientific Reports (Michaut et al., 2016) with some of the key findings outlined below:
• Mutations in CDH1 and in the PI3K pathway are the most frequent molecular interactions in ILC
• Two main subtypes of ILC were identified:
o An immune-related subtype with mRNA up-regulation of PD-L1, PD-1 and CTLA-4 and greater sensitivity to DNA-damaging agents in representative cell line models
o A hormone-related subtype, associated with Epithelial to Mesenchymal Transition (EMT) and gain of chromosomes 1q and 8q and loss of chromosome 11q.
• Three groups with different clinical outcomes were identified, including a group with extremely good prognosis, using somatic mutation rate and eIF4B protein level.
• A comprehensive overview of the molecular alterations driving ILC and links with therapy response was presented.
The molecular characterisation of ILC and identification of the two subtypes has led to a phase II proof-of-concept trial assessing the efficacy of carboplatin and atezolizumab in metastatic ILC. The trial is called the GELATO trial (NCT03147040) which is sponsored by NKI in collaboration with Roche Pharma. Patients are currently being recruited to this trial.
Due to the complex nature of the RATHER programme and the associated data generated, project and data management activities have been key. WP11 aimed to ensure the governance and coordination of the whole project, to create a dedicated consortium webpage to facilitate communication within and outside the consortium and to design, create and populate a secure database with data generated during the project. Project management activities included the negotiation of the consortium agreement, organisation of consortium meetings and TCs, writing of scientific and final reports and allocation of funds. These activities proceeded smoothly with no major problems. A dedicated RATHER website was developed which was update regularly to disseminate activities such as project meetings, publications and press releases. A ‘Filestore’ Repository database was developed to host the data generated by the RATHER consortium providing a comprehensive view of the projects clinical samples and cell lines. The consortium plan to upload the data to a public repository such as the European Molecular Biology Laboratory’s European Bioinformatics Institute (EMBL-EBI) European Genome-Phenome Archive (EGA) following the completion of the RATHER project.
Potential Impact:
The RATHER research programme has contributed and is continuing to contribute significantly to our understanding of the mechanisms underlying two difficult-to-treat subtypes of breast cancer (which account for almost a quarter of cases diagnosed, leading to widespread impact at multiple levels.
Over the course of the RATHER project, the consortium published 27 articles; 19 peer-reviewed original articles, 3 review articles, 4 abstracts and 1 book chapter. Articles were published by the RATHER consortium in high profile journals such as Cancer Research (IF 9.329) Cell Research (IF 15.973) Scientific Reports (IF 4.122) Molecular Oncology (IF 5.331) FEBS (IF 4.530) and Clinical Cancer Research (IF 10.199).
Publications in traditional media have included an article in the Irish Independent newspaper by then EU Commissioner for Research, Innovation and Science, Marie Geoghegan Quinn, discussing the aims of the RATHER project in an article entitled “Cancer knows no boundaries – but we can slow death rate”. OncoMark wrote an article for the European Commission entitled “Targeting the drivers of difficult-to-treat breast cancers” for World Cancer Day in 2015. The RATHER consortium have also released and been featured in a number of press releases both in Ireland and Europe-wide.
The consortium also made a large effort to disseminate the RATHER project and it’s findings through the development of a dedicated RATHER website, which was kept up-to-date describing the latest outputs and meetings of the RATHER consortium. Prof. Rene Bernards also presented an overview of the RATHER project at the AACR annual meeting in 2012 entitled “ Coordination of Collaborative Translational Cancer Research in Europe: Towards Evidence-Based Therapies and Care”.
The RATHER consortium were encouraged to present their work at conferences and meetings (national and international) to promote their findings and to disseminate the work of the RATHER consortium. Overall, RATHER consortium partners gave 25 national presentations, 32 international presentations and 28 poster presentations at international & national conferences. The RATHER consortium organised a symposium entitled “Leveraging the power of systems medicine in personalised oncology” which was held in UCD, Ireland in 2015 and attended by the scientific community. The RATHER coordinator, Prof. Gallagher, also chaired and hosted a Euroscience Open Forum in 2012 (ESOF12) entitled “The true cost of personalised cancer medicine”.
The RATHER consortium also understood the importance of outreach activities; to engage the public in scientific research, to raise awareness of breast cancer and the recent advances in breast cancer treatment and to disseminate the findings of the RATHER project. The consortium members regularly participated in outreach activities such as public seminars, lab tours, educational outreach and patient engagement activities.
Following the conclusion of the RATHER project, the consortium will continue to disseminate the findings of the project. The RATHER research findings will be publicly disseminated through the provision of infographics and slides highlighting the outputs and progress of breast cancer research, through announcements by relevant charitable organisations including the Irish Cancer Society, regular updates on social media platforms including Twitter and Facebook via the OncoMark, Conway Institute, UCD School of Medicine, UCD Research and other appropriate research accounts of the RATHER partners. Following publication of the final manuscripts, news highlights will be provided across partner University websites, via radio interview and national newspaper pieces.
Awards have also been given to members of the group including NovaUCD Innovation Award to Prof. William Gallagher (2011) and EU Women Innovator Award (2nd Place) to Prof. Laura van’t Veer (NKI/AG). RATHER hosted a scientific symposium in January 2015, entitled ‘Leveraging the power of systems medicine in personalised oncology’, which was well attended. The agenda featured international leaders in the field of personalised medicine.
In terms of Intellectual Property, 4 patents have been filed by RATHER partners during the RATHER project:
• Combination therapy - combined MAP2K4/MAP3K1 and MEK/ERK inhibition – Inventors: René Bernards (NKI), Zheng Xue (NKI) (2016)
• Methods and means for subtyping invasive lobular breast cancer – Inventors: Tycho Bismeijer (NKI), Magali Michaut (NKI), René Bernards (NKI), Lodewyk Wessels (NKI) (2016)
• A method of predicting risk of recurrence of cancer – Inventors: Adrian Bracken (TCD), Fiona Lanigan (UCD), William Gallagher (UCD) (2016)
• Methods for molecular classification of BRCA-like breast and/or ovarian cancer – Inventors: Justine Peeters (AG), Iris Simon (AG), Tesa Severson (NKI), Sabine Linn (NKI) (2016)
RATHER has generated a large number of potential biomarkers for new molecular diagnostic assays and potential drug targets for new therapeutic approaches; however, for translation and use in the clinic, these need to undergo further validation and their performance needs to be determined in clinical trials. The RATHER partners have leveraged funding from other sources to fund further validation studies and have initiated clinical trials, other than POSEIDON, on a number of the potential targets.
Institut Curie have been awarded €90k from various sources in France to validate the protein biomarkers identified in WP3. If validated, these protein biomarkers will be patented whenever possible and published in peer-reviewed journals. UCD have initiated a collaboration with Carrick Therapeutics worth €150k to aid development of a CDK7 inhibitor for treatment of TN breast cancer. The BRCAness signature for TN breast cancer developed by NKI has been validated in the international I-SPY2 trial which has led to the initiation of a phase II trial, SUBITO, which is costing almost €3 million. This trial will change how TN breast cancer is treated in the Netherlands if successful. NKI have also leveraged over €1.5 million for additional development and validation work. The discovery of two ILC subtypes has led to the initiation of a Phase II proof-of-concept trial, GELATO, looking at combining carboplatin and atezolizumab therapies in ILC. Additionally, the new treatment strategies for ILC and TN breast cancer will have to be validated in animal models. Collaborations and discussions with pharmaceutical companies are ongoing to conduct future clinical trials in this space. UCD have been awarded €2.5 million to further develop the OncoMasTR panel as an IHC-based assay for use in breast and prostate cancer, with the funded programme being entitled OPTi-PREDICT. The RATHER Coordinator, Prof Gallagher, is also the Director of BREAST-PREDICT, a €7.5 million collaborative cancer research centre funded by the Irish Cancer Society which was launched in 2013 to work towards personalised breast cancer medicine. OncoMark received a €2.75 million grant from the SME Instrument to develop the OncoMasTR panel as a prognostic qPCR test for early-stage breast cancer patients. Additionally, UCD received €150k from Breast Cancer now to determine the role of epigenetic reader BRC3 in ILC alongside $448,480 from the Susan G Komen Foundation to further study BET inhibition as a therapeutic strategy for ILC.
The work under RATHER has had a large impact on the SME partners. For Agendia, the project ensured the first steps of translating MammaPrint from microarray to NGS. This translation allowed for assay decentralisation and subsequent increase market penetration to countries where a centralised test cannot be offered due to government regulations. With this decentralising strategy, MammaPrint is keeping up with technical advances and customers wishes. Outside the RATHER project, Agendia has further developed MammaPrint on NGS and received the necessary CE mark followed by a soft launch of the product in February 2018. The RATHER project has also allowed Agendia to validate MammaPrint in ILC, paving the way for this test to be used in clinical practice in this subgroup. OncoMark used the feasibility data on the OncoMasTR panel to win an H2020 SME Instrument grant worth €2.75 million, along with €2.1 million from private investors, which has brought the OncoMasTR test to commercialisation.
Following validation and clinical trials of the RATHER outputs, there will be significant and long-term impact for patients. The outputs of the RATHER project will lead to personalised treatment for difficult-to treat-breast cancers, particularly ILC and TN breast cancer. This will lead to better survival rates and an improved quality-of-life for patients.
The RATHER consortium addressed key, as yet unanswered, issues relating to a lack of therapeutic (and associated molecular diagnostic) options for ILC and TN breast cancer. Using multiple state-of-the-art approaches, in an integrated manner, we shed much needed biological insight into both of these difficult-to-treat subtypes of breast cancer, with the key aim of identifying lead candidate kinase targets for therapeutic intervention, alongside development of tailored molecular diagnostic approaches for prediction of therapeutic response.
By identifying activated/mutated kinases in ILC and TN breast cancer, clinical assays for patient stratification can be developed. This will promote the use of personalised therapies for these subgroups, thus avoiding unnecessary treatment of patients who will not benefit from specific treatments.
List of Websites:
www.ratherproject.com
Contact Coordinator: Prof. William Gallagher
William.gallagher@ucd.ie
+353 (0)1 7166743
RATHER (Rational therapy for breast cancer: individualised treatment for difficult-to-treat breast cancer subtypes) aimed to identify new therapeutic targets for two difficult to treat subtypes of breast cancer. A quarter of all breast cancer patients are diagnosed with these difficult-to-treat subtypes of breast cancer; Invasive Lobular Carcinoma (ILC), an aggressive form of breast cancer that is resistant to standard therapies (10% of cases globally), or Triple Negative (TN) breast cancer which lacks the common targets for treatment; estrogen, progesterone and HER2 receptors (15% of cases globally). Both are linked with poor prognosis and there are currently no effective therapies available.
RATHER was a large multidisciplinary collaborative project funded by the EU under the FP7 programme (€6 million) led by University College Dublin. The consortium comprised 8 partners from the academic and private sectors; University College Dublin (Ireland), OncoMark (Ireland), Agendia (the Netherlands), Netherlands Cancer Institute (the Netherlands), Institut Curie (France), Vall d’Hebron Hospital (Spain), Cambridge University (UK), and Lund University (Sweden).
The RATHER consortium aimed to better understand difficult-to-treat breast cancers by studying kinases, which are activated in many cancer types and which have been shown to play a role in cancer. In cancer, there is a strong correlation between genetic mutations and a patient’s response to drugs. Therefore, the RATHER consortium looked to investigate mutations in kinases to deliver personalised therapeutic interventions to the 25% of patients who currently lack effective targeted therapies.
To survey kinases to identify novel biomarkers for new molecular diagnostic assays or novel drug targets for new therapeutic approaches, the consortium used a multidisciplinary approach applying comprehensive genomic, transcriptomic and functional analysis and data integration, leading to the validation of targets / biomarkers, the selection of lead candidates, and the initiation of a Phase Ib/II clinical trial.
The RATHER project integrated clinical and preclinical research, establishing collaborations between academia and industry delivering a number of important results and outputs throughout the duration of the project. The results from the RATHER project have been published in international scientific journals and presented at international conferences, with a number of manuscripts still in preparation pending further validation of results. Work on the promising targets identified during the project will be brought to completion beyond RATHER, facilitated by the relationships and collaborations established during this project.
Some major outputs from the RATHER project include one ongoing clinical trial (POSEIDON) directly funded by the RATHER project, 3 clinical trials entered into by RATHER consortium partners (GELATO, I-SPY2 and SUBITO), a validated assay for ILC (MammaPrint), a new biomarker signature for ILC (immune/hormone related), a new biomarker signature for TN breast cancer (BRCAness), a new biomarker for TN breast cancer (CDK7), a new treatment strategy for ILC (MAP3K1/MAP2K4 genotype-dependent) and a new treatment strategy for TN breast cancer (CDK7 & BCL-2/XL-targeted).
In summary, the RATHER consortium have published 27 articles (19 peer-reviewed original articles, 3 review articles, 4 abstracts) and 1 book chapter, have delivered 57 presentations (32 international, 25 national) and presented 28 posters. Four patents have been filed.
Project Context and Objectives:
In the RATHER project, we aimed to identify and validate novel kinase targets for therapy for TN and ILC subtypes of breast cancer, as well as to develop molecular diagnostic assays and test a novel kinase inhibitor in a clinical trial.
Overall RATHER Objectives:
• To identify novel kinase targets for therapy
• To validate these targets within in vitro preclinical models
• To develop a prognostic gene expression profile for ILC
• To develop predictive and prognostic clinical assays for breast cancer
• To test a PI3K inhibitor +/- endocrine therapy in a Phase I/II clinical trial
• To generate new knowledge, protect resulting intellectual property and commercialise findings (SMEs)
WP1: Selection of suitable tissues, processing of materials and collection of clinical data
In this WP, we aimed to select 150 TN and 150 ILC tumour samples from the frozen tissue collections of NKI and CAM and isolate DNA, RNA and protein lysates from these samples. Matching formalin-fixed, paraffin-embedded (FFPE tissues were also collected for TMA construction.
• Task 1.1. Selection of tumour samples for molecular analyses
• Task 1.2. DNA, RNA and protein extraction of tumour samples
• Task 1.3. Patient data collection
WP2: Kinome re-sequencing 150 ILC and 150 TN breast cancer
Here, we aimed to sequence all 518 kinases and 68 additional genes that are frequent targets of mutation in breast cancer in all 300 tumour samples, using high-throughput next generation sequencing, together with exon capture technology.
• Task 2.1. Generation of raw kinome sequence data
• Task 2.2. Analysis of raw large-scale genomic sequence data
• Task 2.3. Validation of mutations
WP3: Reverse phase protein lysate arrays (RPPA) on ILC and TN tumour samples
In WP3, we used RPPA technology to assess the activation status of kinases as judged by phospho-epitope quantitation.
• Task 3.1. Validation of additional phosphor-specific antibodies
• Task 3.2. Retrieval of protein lysates and concentration equilibration
• Task 3.3. Protein array generation and antibody revelation
• Task 3.4. Quantification and analysis of the data
WP4: Tissue microarray (TMA) analysis of ILC and TN tumour samples
In WP4, we validated findings from WP3 on TMAs to validate that the differential activations seen in the previous WP take place in the tumour cells rather than in the stroma.
• Task 4.1. TMA construction
• Task 4.2. Protein expression analysis
• Task 4.3. In situ hybridization (ISH) on TMAs
• Task 4.4. Manual and automated image analysis of TMA data
WP5: Genomic analysis of ILC and TN tumours using SNP arrays to evaluate copy number and allele variation
In WP5, we comprehensively analysed copy number gains and losses in these tumours, with a special emphasis on gains and losses in loci encoding kinases.
• Task 5.1. Data pre-processing
• Task 5.2. Copy number estimation and determination of loss of heterozygosity (LOH) status
• Task 5.3. Identification of summarized regions of interest
• Task 5.4. Integrated measurement of copy number and gene expression
WP6: Gene expression analysis using DNA microarrays on ILC and TN tumour samples
In WP6, we performed full genome gene expression analysis of all 300 tumours via gene arrays and RNA sequencing. In addition, a commercial predictive assay for lobular cancers was also developed.
• Task 6.1. Identification of predictive genes and integrative signatures
• Task 6.2. Development of prognostic profile for ILC
• Task 6.3. RNA sequencing analysis
WP7: Validation of kinase targets for therapy in preclinical models
In WP7, we used siRNA, shRNA and small molecule/antibody-based drugs to validate the candidate kinase targets for therapy in appropriate cell line models of TN / ILC breast cancer.
• Task 7.1. Identifying driver mutations in breast cancer cell line models
• Task 7.2. Finding small molecule (antibody-based) drugs that are toxic in cells having oncogenic driver mutations
• Task 7.3. Selecting optimal kinase targets for guidance of kinase inhibitor treatment in the clinical trial
WP8: Development of standardized biomarker assays for breast cancer patient stratification
In WP8, the SMEs developed biomarker assays for breast cancer patients, particularly for ILC patients.
• Task 8.1. Development of specific companion diagnostic assays
• Task 8.2. Development of a prognostic test for ILC
• Task 8.3. Validation of a unique mechanistically-anchored breast cancer biomarker panel
• Task 8.4. Digital image analysis of kinase alterations
WP9: Initiation of phase I/II clinical trial to evaluate tumour genotype-drug response relationships
In WP9, we are performing a phase I/II clinical trial (POSEIDON) to study the efficacy of a selective PI3K inhibitor in ductal and lobular breast cancer patients.
• Task 9.1. Phase Ib dose-escalation study
• Task 9.2. Randomised phase II trial
• Task 9.3. Translational substudies
WP10: Data Integration
In WP10, we performed an integrative bioinformatics analysis on the data generated in WPs 2-6. The outcome of this will be a list of candidate targets for therapy in either ILC or TN breast cancer.
• Task 10.1. Integration of data on pathway control / systems biology
WP11: Project and database management
In WP11, the coordinator and project management team were responsible for managing the project, including all reporting requirements, and the database.
• Task 11.1. Project management
• Task 11.2. Preparation of website and communications portal
• Task 11.3. Database management
• Task 11.4. Ethics management
Project Results:
Sample collection and distribution was a key part of the RATHER project and the main aim of WP1. Tumour samples were selected from NKI and CAM biobanks based on a number of pre-specified criteria. A total of 145 ILC and 176 TN samples were selected and then processed for RNA extraction, DNA extraction and protein extraction, and distributed to partners along with matching clinical, survival and pathology data. The DNA was used for sequencing in WP2 and SNP array analysis in WP5, the RNA was used for gene expression analysis in WP6 and the extracted protein was used for RPPA analysis in WP3. Formalin-fixed paraffin embedded (FFPE) tissue blocks were also collected for all patients to generate TMAs for use in WP4. The clinical and survival data allowed correlation of kinome mutations with clinical outcome.
WP2 aimed to identify signalling pathways that drive the development of breast cancer and those that influence therapy response through the sequencing of the ILC and TN tumour samples obtained in WP1. NKI have developed a capture system that enables high throughput sequencing of all protein kinases, common oncogenes and tumour suppressor genes, a selection of lipid kinases and components of the PI3K pathway. NKI successfully completed sequencing for 143 ILC and 150 TN samples and identified a number of kinase mutations; the most common mutations found in ILC were CDH1 and PIK3CA; in TN the most common found was TP53. Overall, in ILC a total of 987 protein altering somatic variations were identified in 393 genes while TN samples had 1,208 protein-altering somatic variants in 425 genes. The results suggest that understanding the spectrum of mutations in breast cancer will help to stratify patients.
In WP3, reverse phase protein lysate array (RPPA) analysis was used to examine protein expression and phosphorylation of kinases. The RATHER project allowed Institut Curie to significantly increase the amount of validated antibodies that can be applied to RPPA. Over 500 new primary antibodies were tested over the course of the project, with over 150 validated for use in RPPA. These antibodies have not only been used in the RATHER project but continue to be used for many other cancer-related projects. RPPA data was used to identify differences in protein expression / activation between ILC and TN tumours and to identify subgroups within ILC. A number of proteins were identified that are differentially expressed between ILC and TN breast cancers. 4 subgroups of ILCs were identified that show biologically distinct features in terms of cell signalling pathway activation status. Proteins identified may have a prognostic value in IHC. Activation of the PI3K / Akt pathway and its correlation with patient survival was determined by RPPA, which further justified the clinical interest of the POSIEDON clinical trial (WP9). Additionally, a list of (phopho-) proteins associated with patient outcome was established that could potentially serve as prognostic biomarkers of therapeutic targets. 10 (phospho-)proteins of which a higher expression was associated with a better prognosis were identified and 9 (phospho-) proteins associated with a worse outcome were also identified (Michaut et al., 2016). Protein biomarkers identified have to be validated in an independent cohort of ILCs, which is ongoing in a new project in Institut Curie outside of the RATHER project.
The use of immunohistochemistry (IHC) in clinical cohorts is of critical importance in determining the utility of a biomarker in clinical practice. Despite the wide use of IHC as a biomarker validation tool, no universally accepted standardization guidelines have been developed to determine the acceptability of particular antibodies for IHC prior to its use. In WP4, we developed an IHC and Western blot (WB) workflow to ensure antibody specificity. Following successful antibody validation, we progressed to IHC on breast cancer tissue microarrays (TMA) which allowed the evaluation of the expression level, mutation and activation status of kinases. TMAs were constructed containing 142 ILC and 152 TN patient samples, in combination with clinical information, that were available for use within the whole RATHER project. Higher levels of AKT-3 amplification were identified in TN samples in comparison to ER-positive breast cancers. AKT-3 amplification along with protein expression was negatively associated with Recurrence Free Survival (RFS) in TN. It is hypothesised that AKT-3 amplification could represent a potentially relevant oncogenic event in a subset of TN breast cancers that may select cells sensitive to Akt-3 inhibitors. A scientific publication describing the molecular alterations of Akt-3 has been published (O’Hurley et al., 2013) in addition to a review article (O’Hurley et al., 2014) and a book chapter (Linn et al., 2013) discussing biomarkers and the use of IHC for biomarker validation.
As cancer is a genetic disease with alterations in DNA copy number characteristic of solid tumours, WP5 aimed to profile the genomic alterations for both ILC and TN tumour samples. Gains and losses were seen in both ILC and TN samples, with significant differences identified between ILC and TN breast cancers. The ILC data was integrated in WP10 and included in the Michaut et al., 2016 publication that presents an integrated molecular portrait of ILC.
WP6 aimed to quantify kinase expression levels to discover important information about the relative abundance of kinases and therefore, their relative activity. Transcriptomic analysis was used to quantify and identify gene signatures predictive of pathway alterations. The expression levels of 41 kinase or kinase-related genes were shown to be correlated with ILC patient outcome. These results provide a list of interesting targets for downstream validation. Additional work was carried out in WP6 on TN breast cancer. Approximately 10-15% of TN breast cancers are mutant for the tumour suppressor BRCA1. Deficiencies in BRCA1 result in genomic instability. These tumours are classified as either BRCA1-like or non-BRCA1-like, with the BRCA1-like subgroup having a significantly worse prognosis. In the WP6, we identified differences in gene expression, mutation and prognosis between BRCA1-like and non-BRCA1-like tumours, identifying a ‘BRCAness’ (BRCA1-like) signature providing insight into the genomic instability in BRCA1-like samples. Genes most commonly upregulated in BRCA1-like tumours included those associated with DNA replication, recombination and repair suggesting they may be susceptible to therapies targeted to DNA and cell cycle pathways such as PARPi. This work was published in Molecular Oncology (Severson, TM et al., 2015) and a patent has been submitted for the 77 BRCAness signature genes. Agendia is currently focussed on developing a companion diagnostic test based on this BRCAness signature. The BRCAness signature has also been accepted as a qualifying biomarker in the ISPY2 trial (Investigation of Serial Studies to Predict Your Therapeutic Response with Imaging and Molecular Analysis) and is currently being assessed to predict response in patients treated with the PARPi-inhibitor ABT-888. Results are outlined in a recent publication in NPJ Breast Cancer (Wolf, D. et al., 2017). BRCA1-like validation in the I-SPY2 trial has been instrumental for the start of the phase III SUBITO trial (NCT02810743), ‘Substantially improving the cure rate of high risk BRCA1-like breast cancer’. If a positive SUBITO study outcome is seen, high-dose alkylating chemotherapy with autologous stem cell rescue will be reintroduced and reimbursed for stage II, HER2 negative, BRCA1-like breast cancer patients in the Netherlands. WP6 also aimed to develop and validate a new profile, or to optimise existing profiles such as MammaPrint, specifically for the ILC subgroup. MammaPrint is a prognostic breast cancer test developed and commercialised by Agendia, which predicts the risk of breast cancer recurrence based on gene expression analysis. This study looked to identify lobular tumours that require more aggressive intervention i.e. patients who are like to develop early metastasis (within 5 years), as these patients can be most successfully treated with systemic adjuvant chemotherapy. MammaPrint’s prognostic profile was tested for it’s predictive power within a small set of ILC samples. Results indicated that MammaPrint may be sufficient in identifying a subgroup of ILC patients that would benefit from more aggressive treatment. The performance of MammaPrint in a larger set of ILC samples was carried out in WP8.
WP7 aimed to validate kinase targets in preclinical models for therapy for ILC and TN breast cancer. Kinase targets identified in WPs 2-6 were functionally interrogated to analyse whether they drivers of the oncogenic phenotype in TN and ILC. CDK7 was found to be closely linked to poor prognosis in the TN breast cancer context, with evidence demonstrated in the RATHER project of elevated CDK7 expression as a candidate biomarker of poor prognosis and candidate therapeutic target in TN breast cancer. We also offer a preclinical proof of concept for combining CDK7 and BCL-2/BCL-XL inhibitors as a mechanism-based therapeutic strategy to improve TNBC treatment. These results are outlined in a publication in Cancer Research (Li, B. et al., 2017). UCD are now working with Carrick Therapeutics to aid development of their novel CDK7 inhibitor. In WP7, a new potential treatment strategy was also proposed for the treatment of ILC, identifying candidate drug treatment regimens for ILC based on their genotype. We predicted that ILCs that have a mutation in either component of their JNK pathway will be sensitive to MEKi monotherapy, but that tumours that are wild-type for both MAP2K4 and MAP3K1 will require co-treatment with a pan-HER inhibitor such as Dacomitnib. Validation of these results in animal models of ILC is required before translation into clinical therapies. An article detailing this work has been published in Cell Research (Xue, Z., et al. 2018) and NKI are collaborating with a pharmaceutical company to conduct future clinical trials.
The primary aim of WP8 was to develop standardised biomarker assays to be used in Phase I/II clinical trials or in other diagnostic settings to facilitate patient stratification for receipt of relevant kinase inhibitors or alternative therapies. A secondary aim of this WP was to examine the utility of the developed array-based prognostic tool for ILC. Agendia are currently developing a companion diagnostic test for the BRCAness signature for TN breast cancer (output of WP6). They have also studied the prognostic value of the MammaPrint test for breast cancer patients with early-stage ILC with study results validating MammaPrint as an independent factor for this subgroup of patients. The increase in incidence of ILC over the past decades and the modest representation of ILC in the MammaPrint development dataset called for a stratified survival analysis dedicated to this specific subgroup. Under RATHER, Agendia successfully completed the first steps of translating MammaPrint from microarray to a Next generation Sequencing Platform, a technology already validated in the clinical context. Outside of the RATHER project, Agendia has further developed MammaPrint on NGS and received the necessary CE mark followed by a soft launch of the product in February 2018. OncoMark and UCD worked on the validation of a unique mechanistically-anchored breast cancer biomarker panel (OncoMasTR) that showed great prognostic potential for early stage breast cancer patients, predicting whether they could safely forego chemotherapy. OncoMark separately validated the potential of 3 biomarkers within this panel (plus 3 control genes) in a gene-expression based assay. OncoMark received separate funding from an SME Instrument grant and private investors and have developed this signature into a qPCR test which has been successfully clinically validated. This test has recently been CE marked. UCD examined the potential of 4 biomarkers from the panel in an IHC-based prognostic assay, with further funding awarded from Science Foundation Ireland to complete the validation studies initiated during RATHER.
A major success for the RATHER consortium was the launch of POSEIDON, a Phase Ib/II clinical trial (WP9) to evaluate the safety and effectiveness of GDC-0032 when given alongside tamoxifen. This trial is sponsored by NKI in collaboration with Genentech, EurocanPlatform and the RATHER project. It combines traditional endocrine therapy (Tamoxifen) with a novel drug GDC-0032 (Taselisib), an isoform selective inhibitor of a kinase (PI3K). This novel drug inhibits PI3K, a kinase that is often mutated in ILC. This therapeutic approach has the potential to halt the growth of cancer driven by this kinase. Previous trials with non-selective PI3K pathway inhibitors have experienced relatively low single agent anti-tumour activity, as well as toxicities. Therefore, the isoform-selective inhibitor (GDC-0032 / Taselisib) shows promising activity and low toxicity in Phase Ia clinical trials. The trial is currently ongoing and will extend past the RATHER project.
Phase Ib has been completed, with the recruitment of 30 patients. Taselisib in combination with tamoxifen was generally well tolerated with no unexpected toxicities observed. A recommended dose for phase II was determined as 4mg Taselisib for 21 days in combination with 20mg Tamoxifen continuously. Phase II is ongoing. To date, 90 patients have been recruited to Phase II of the trial at a number of hospital sites across the Netherlands, the UK and Spain, with the first site to open imminently in France.
The POSEIDON trial will extend beyond the RATHER project; however, it has already provided some relevant outputs:
• It is Genentech’s first ever investigator-led multinational trial in Europe (and first non-US).
• Results for Phase Ib have been presented at ASCO 2016 (Baird, R.).
• Recruitment of Phase Ib was completed in approximately 13 months, demonstrating that clinical trials with academic leadership can be competitive despite restrictions.
• Biopsies from 10 of 30 patients in Phase Ib have been obtained to perform molecular studies, impressive data when compared to competitive trials promoted by pharma companies.
WP10 represented the key integrating force within the RATHER consortium, specifically focussing on analysing and integrating data from the other WPs, with all data managed in a relational database developed in WP11. The main objective of WP10 was to develop and apply a computational framework to enable the integration of data from the different WPs (particularly WPs 2-6). Publicly available data was also integrated with the data generated from the RATHER project. An integrated molecular portrait of ILC was established, containing a comprehensive view of the genomic, transcriptomic and proteomic landscapes of the RATHER biological samples, with a particular focus on the kinome. These results have been published in Scientific Reports (Michaut et al., 2016) with some of the key findings outlined below:
• Mutations in CDH1 and in the PI3K pathway are the most frequent molecular interactions in ILC
• Two main subtypes of ILC were identified:
o An immune-related subtype with mRNA up-regulation of PD-L1, PD-1 and CTLA-4 and greater sensitivity to DNA-damaging agents in representative cell line models
o A hormone-related subtype, associated with Epithelial to Mesenchymal Transition (EMT) and gain of chromosomes 1q and 8q and loss of chromosome 11q.
• Three groups with different clinical outcomes were identified, including a group with extremely good prognosis, using somatic mutation rate and eIF4B protein level.
• A comprehensive overview of the molecular alterations driving ILC and links with therapy response was presented.
The molecular characterisation of ILC and identification of the two subtypes has led to a phase II proof-of-concept trial assessing the efficacy of carboplatin and atezolizumab in metastatic ILC. The trial is called the GELATO trial (NCT03147040) which is sponsored by NKI in collaboration with Roche Pharma. Patients are currently being recruited to this trial.
Due to the complex nature of the RATHER programme and the associated data generated, project and data management activities have been key. WP11 aimed to ensure the governance and coordination of the whole project, to create a dedicated consortium webpage to facilitate communication within and outside the consortium and to design, create and populate a secure database with data generated during the project. Project management activities included the negotiation of the consortium agreement, organisation of consortium meetings and TCs, writing of scientific and final reports and allocation of funds. These activities proceeded smoothly with no major problems. A dedicated RATHER website was developed which was update regularly to disseminate activities such as project meetings, publications and press releases. A ‘Filestore’ Repository database was developed to host the data generated by the RATHER consortium providing a comprehensive view of the projects clinical samples and cell lines. The consortium plan to upload the data to a public repository such as the European Molecular Biology Laboratory’s European Bioinformatics Institute (EMBL-EBI) European Genome-Phenome Archive (EGA) following the completion of the RATHER project.
Potential Impact:
The RATHER research programme has contributed and is continuing to contribute significantly to our understanding of the mechanisms underlying two difficult-to-treat subtypes of breast cancer (which account for almost a quarter of cases diagnosed, leading to widespread impact at multiple levels.
Over the course of the RATHER project, the consortium published 27 articles; 19 peer-reviewed original articles, 3 review articles, 4 abstracts and 1 book chapter. Articles were published by the RATHER consortium in high profile journals such as Cancer Research (IF 9.329) Cell Research (IF 15.973) Scientific Reports (IF 4.122) Molecular Oncology (IF 5.331) FEBS (IF 4.530) and Clinical Cancer Research (IF 10.199).
Publications in traditional media have included an article in the Irish Independent newspaper by then EU Commissioner for Research, Innovation and Science, Marie Geoghegan Quinn, discussing the aims of the RATHER project in an article entitled “Cancer knows no boundaries – but we can slow death rate”. OncoMark wrote an article for the European Commission entitled “Targeting the drivers of difficult-to-treat breast cancers” for World Cancer Day in 2015. The RATHER consortium have also released and been featured in a number of press releases both in Ireland and Europe-wide.
The consortium also made a large effort to disseminate the RATHER project and it’s findings through the development of a dedicated RATHER website, which was kept up-to-date describing the latest outputs and meetings of the RATHER consortium. Prof. Rene Bernards also presented an overview of the RATHER project at the AACR annual meeting in 2012 entitled “ Coordination of Collaborative Translational Cancer Research in Europe: Towards Evidence-Based Therapies and Care”.
The RATHER consortium were encouraged to present their work at conferences and meetings (national and international) to promote their findings and to disseminate the work of the RATHER consortium. Overall, RATHER consortium partners gave 25 national presentations, 32 international presentations and 28 poster presentations at international & national conferences. The RATHER consortium organised a symposium entitled “Leveraging the power of systems medicine in personalised oncology” which was held in UCD, Ireland in 2015 and attended by the scientific community. The RATHER coordinator, Prof. Gallagher, also chaired and hosted a Euroscience Open Forum in 2012 (ESOF12) entitled “The true cost of personalised cancer medicine”.
The RATHER consortium also understood the importance of outreach activities; to engage the public in scientific research, to raise awareness of breast cancer and the recent advances in breast cancer treatment and to disseminate the findings of the RATHER project. The consortium members regularly participated in outreach activities such as public seminars, lab tours, educational outreach and patient engagement activities.
Following the conclusion of the RATHER project, the consortium will continue to disseminate the findings of the project. The RATHER research findings will be publicly disseminated through the provision of infographics and slides highlighting the outputs and progress of breast cancer research, through announcements by relevant charitable organisations including the Irish Cancer Society, regular updates on social media platforms including Twitter and Facebook via the OncoMark, Conway Institute, UCD School of Medicine, UCD Research and other appropriate research accounts of the RATHER partners. Following publication of the final manuscripts, news highlights will be provided across partner University websites, via radio interview and national newspaper pieces.
Awards have also been given to members of the group including NovaUCD Innovation Award to Prof. William Gallagher (2011) and EU Women Innovator Award (2nd Place) to Prof. Laura van’t Veer (NKI/AG). RATHER hosted a scientific symposium in January 2015, entitled ‘Leveraging the power of systems medicine in personalised oncology’, which was well attended. The agenda featured international leaders in the field of personalised medicine.
In terms of Intellectual Property, 4 patents have been filed by RATHER partners during the RATHER project:
• Combination therapy - combined MAP2K4/MAP3K1 and MEK/ERK inhibition – Inventors: René Bernards (NKI), Zheng Xue (NKI) (2016)
• Methods and means for subtyping invasive lobular breast cancer – Inventors: Tycho Bismeijer (NKI), Magali Michaut (NKI), René Bernards (NKI), Lodewyk Wessels (NKI) (2016)
• A method of predicting risk of recurrence of cancer – Inventors: Adrian Bracken (TCD), Fiona Lanigan (UCD), William Gallagher (UCD) (2016)
• Methods for molecular classification of BRCA-like breast and/or ovarian cancer – Inventors: Justine Peeters (AG), Iris Simon (AG), Tesa Severson (NKI), Sabine Linn (NKI) (2016)
RATHER has generated a large number of potential biomarkers for new molecular diagnostic assays and potential drug targets for new therapeutic approaches; however, for translation and use in the clinic, these need to undergo further validation and their performance needs to be determined in clinical trials. The RATHER partners have leveraged funding from other sources to fund further validation studies and have initiated clinical trials, other than POSEIDON, on a number of the potential targets.
Institut Curie have been awarded €90k from various sources in France to validate the protein biomarkers identified in WP3. If validated, these protein biomarkers will be patented whenever possible and published in peer-reviewed journals. UCD have initiated a collaboration with Carrick Therapeutics worth €150k to aid development of a CDK7 inhibitor for treatment of TN breast cancer. The BRCAness signature for TN breast cancer developed by NKI has been validated in the international I-SPY2 trial which has led to the initiation of a phase II trial, SUBITO, which is costing almost €3 million. This trial will change how TN breast cancer is treated in the Netherlands if successful. NKI have also leveraged over €1.5 million for additional development and validation work. The discovery of two ILC subtypes has led to the initiation of a Phase II proof-of-concept trial, GELATO, looking at combining carboplatin and atezolizumab therapies in ILC. Additionally, the new treatment strategies for ILC and TN breast cancer will have to be validated in animal models. Collaborations and discussions with pharmaceutical companies are ongoing to conduct future clinical trials in this space. UCD have been awarded €2.5 million to further develop the OncoMasTR panel as an IHC-based assay for use in breast and prostate cancer, with the funded programme being entitled OPTi-PREDICT. The RATHER Coordinator, Prof Gallagher, is also the Director of BREAST-PREDICT, a €7.5 million collaborative cancer research centre funded by the Irish Cancer Society which was launched in 2013 to work towards personalised breast cancer medicine. OncoMark received a €2.75 million grant from the SME Instrument to develop the OncoMasTR panel as a prognostic qPCR test for early-stage breast cancer patients. Additionally, UCD received €150k from Breast Cancer now to determine the role of epigenetic reader BRC3 in ILC alongside $448,480 from the Susan G Komen Foundation to further study BET inhibition as a therapeutic strategy for ILC.
The work under RATHER has had a large impact on the SME partners. For Agendia, the project ensured the first steps of translating MammaPrint from microarray to NGS. This translation allowed for assay decentralisation and subsequent increase market penetration to countries where a centralised test cannot be offered due to government regulations. With this decentralising strategy, MammaPrint is keeping up with technical advances and customers wishes. Outside the RATHER project, Agendia has further developed MammaPrint on NGS and received the necessary CE mark followed by a soft launch of the product in February 2018. The RATHER project has also allowed Agendia to validate MammaPrint in ILC, paving the way for this test to be used in clinical practice in this subgroup. OncoMark used the feasibility data on the OncoMasTR panel to win an H2020 SME Instrument grant worth €2.75 million, along with €2.1 million from private investors, which has brought the OncoMasTR test to commercialisation.
Following validation and clinical trials of the RATHER outputs, there will be significant and long-term impact for patients. The outputs of the RATHER project will lead to personalised treatment for difficult-to treat-breast cancers, particularly ILC and TN breast cancer. This will lead to better survival rates and an improved quality-of-life for patients.
The RATHER consortium addressed key, as yet unanswered, issues relating to a lack of therapeutic (and associated molecular diagnostic) options for ILC and TN breast cancer. Using multiple state-of-the-art approaches, in an integrated manner, we shed much needed biological insight into both of these difficult-to-treat subtypes of breast cancer, with the key aim of identifying lead candidate kinase targets for therapeutic intervention, alongside development of tailored molecular diagnostic approaches for prediction of therapeutic response.
By identifying activated/mutated kinases in ILC and TN breast cancer, clinical assays for patient stratification can be developed. This will promote the use of personalised therapies for these subgroups, thus avoiding unnecessary treatment of patients who will not benefit from specific treatments.
List of Websites:
www.ratherproject.com
Contact Coordinator: Prof. William Gallagher
William.gallagher@ucd.ie
+353 (0)1 7166743
