Final Report Summary - EPI-FEM-CARE (EPIGENETICS FOR FEMALE PERSONALIZED CANCER CARE)
Executive Summary:
Every year in the EU ~45,300 and 330,000 women are diagnosed with ovarian and breast cancer respectively & 28,800 and 90,000 of these two groups of women will die as a consequence of these diseases. Currently there are no tools available that allow for:
a) Effective screening of ovarian and/or breast cancer of sufficient sensitivity and specificity to avoid potential over-diagnosis,
or;
b) Stratification of patients into optimal personalised therapy regimes in ovarian/breast cancer.
EpiFemCare ADVANCES
Progress in personalised cancer medicine will only be possible with the development of bioassays involving the analysis of easy accessible biomaterials that contain stable target molecules reflective of disease. We have established and clinically validated a series of blood tests based upon DNA methylation technology that will facilitate both early detection and prediction of therapeutic outcome in breast and ovarian cancer.
CONSORTIUM
Our pan-European academic-industrial consortium has thrived through a collaboration of teams with diverse clinical, scientific & industrial expertise. Project partners have worked with the latest state-of-the-art technologies and, integrally, the best available cohort and clinical trial sample sets available. GATC-Biotech and Genedata are Europe’s leading providers of DNA sequencing and bioinformatics for biomarker development and together with the leading clinical academic groups in gynaecological oncology – University College London, Charles University Prague and Ludwig-Maximilians University Munich – have developed blood based biomarkers for breast and ovarian cancers.
The clinical partners have provided access to unique cohort through the prospective collection of more than 1880 serum samples from women with breast or ovarian cancers, benign conditions or from healthy volunteers. In addition, serum samples collected from >200,000 women well in advance of disease (UK Collaborative Trial of Ovarian Cancer Screening) or before and during treatment (SUCCESS Trial) were provided by the clinical partners for analysis.
IMPACT
The EpiFemCare project has developed new and improved tools and technologies that will contribute to enabling the uptake of personalised medicine into clinical practice. These tools and technologies focus on improving results in samples containing very low concentrations of DNA from tumours and are critical to improving the performance of liquid biopsy products in the future. They include new workflows for quality-controlled and standardized processing of Reduced Representation Bisulfite Sequencing (RRBS) that have been implemented in Genedata’s Profiler software and highly standardised, optimised workflows for cell-free DNA extraction and targeted sequencing within the GATCLIQUID for oncology products at GATC Biotech.
Project Context and Objectives:
2.1 Project Context
In the European Union and the United States, one in every three individuals faces a cancer diagnosis in their lifetime. In women, approximately half of all cancers originate in the female reproductive organs, i.e. breast, ovary, endometrium and cervix. Breast cancer (BC) is the most common cancer in women, affecting as many as 1 in 8 women. The high mortality rate is related to its tendency to spread: one-third of axillary, node-negative BC patients develop local or distant metastases, even in the absence of tumor spread at the time of primary diagnosis.
Ovarian cancer (OC) has low prevalence but is the most common cause of death from gynecological malignancies. In the early stages, women are mostly asymptomatic or present with non-specific symptoms, making diagnosis difficult. However, while patients with stage I disease have a 5-year survival of >90%, diagnosis is most often made at stage IIIC when 5-year survival is less than 40%. The majority of patients with advanced stage disease undergo recurrence despite aggressive surgery and chemotherapy with platinum and taxanes.
Cervical cancer incidence and associated deaths have been reduced by up to 80% through introduction and widespread uptake of triennial cervical screening with the Papanicolaou test (or Pap smear). In contrast, breast and ovarian cancers continue to pose considerable challenges in terms of prediction and early detection. Unlike cervical cancer, the abnormal cells are not directly accessible. Furthermore, in the case of ovarian cancer, the cell of origin is not well defined.
Ovarian cancer
No effective screening method for OC is available at present. Beyond screening, once a pelvic mass is identified in a patient, a non-invasive test that could help to triage patient care would be highly desirable. Biochemical markers, transvaginal ultrasound, scoring systems and models are widely used to discriminate between benign and malignant disease. The most commonly used marker, CA125, is also expressed in numerous benign conditions, and it is positive in only about 50% of early-stage ovarian cancers. The Risk of Malignancy Algorithm (ROMA), based on assessment of CA125 together with human epididymis secretory protein 4 (HE4), was initially thought to be superior to CA125 alone. However, other groups have suggested that further validation is required. Risk of Malignancy Index (RMI), a score based on ultrasound variables as well as on menopausal status and CA125, is widely used at present, mainly in the United Kingdom. Although the RMI index allows the referring gynaecologist to send a patient to experts based on an objective assessment, its sensitivity is as low as 78%. Transvaginal ultrasound by an expert operator using a formal scoring model system is a highly sensitive and ideal second-stage diagnostic method; however, it is highly dependent on individual expertise. As a result, discrimination between a benign and malignant ovarian tumor remains a significant challenge in clinical practice.
Availability of a biomarker or a panel of biomarkers that can detect OC in its earliest stages with both high sensitivity and specificity would improve outcomes. In addition, they may be used as prognostic or predictive indicators, or as novel targets for cancer treatment.
Breast cancer
The lack of diagnostic markers detectable in early BC is a critical issue in patient management. In clinical practice, since early BC does not cause symptoms, mammography or ultrasound imaging are used for screening. Mammography screening, which detects characteristic masses and/or micro-calcifications, is routinely performed in patients older than 45-50. Intensified breast screening can also be used for younger women when other factors, such as genetic profile or family history, indicate higher risk of disease.
Despite the fact that mammography screening for BC does save some lives, there is a continuing debate about its usefulness. Recent data demonstrate an improved treatment outcome for only 3 13% of women in whom BC was detected by screening mammography. Consequently, it would seem that 87-97% of women with a screening-detected BC do not receive a clear treatment benefit, indicating that these cancers are still not being detected early enough. In short, concerns about the potential risks and harms of the current breast screening program continue, as should the development of novel methods for early detection.
2.2 EpiFemCare’s objectives
The EpiFemCare project aimed to devise a set of bioassays – or liquid biopsies – to meet the unmet needs described above. The project’s objectives were to establish and clinically validate a series of blood tests based upon DNA methylation technology that would facilitate both early detection and prediction of therapeutic outcome in breast and ovarian cancer, as follows:
1. Screening and early detection of breast and ovarian cancer;
2. Diagnosis of ovarian cancers through discrimination of benign and cancerous growths within the pelvis; and
3. Personalised care through monitoring the efficacy of treatment and determining the right drug regimen for each woman.
DNA methylation is a chemical modification of DNA known to respond to the environment and consequentially capable of modifying gene expression. Selection of DNA methylation as the basis for the EpiFemCare blood tests offers key advantages over other potential biomarkers, including genetic mutation, proteins, lipids and RNA. DNA methylation is: (1) biologically and technically stable, (2) has the capacity to be amplified, and (3) demonstrates cancer specificity when compared to normal or benign tissues. Technology platforms for the analysis of DNA methylation have evolved rapidly over the past few years to the point where epigenome-wide analysis can now be performed with relative ease. The EpiFemCare project used technology capable of the sensitive detection of cell free DNA methylation in serum against an excess background of normal contaminating DNA – a feature that is crucial for the development and clinical success of blood based DNA methylation tests.
2.3 EpiFemCare workplan
Work Package 1: Orchestration and establishment of a breast cancer collection optimised for epigenetic analyses
Work Package 2: Orchestration and establishment of an ovarian cancer collection optimised for epigenetic analyses
Work Package 3: Discovery of cancer specific CpGs (Phase 1)
Work Package 4: Data management and analysis of the data generated
Work Package 5: Development and application of DNAme serum assays
Work Package 6: Knowledge Dissemination, IP Use & exploitation
Work Package 7: Project Management and Administration
2.4 EpiFemCare partners
EpiFemCare brings together 6 leading research organisations and SMEs from across Europe. The project was led by Professor Martin Widschwendter from University College London, who is an internationally renowned researcher in epigenetics, Head of the UCL Department of Women’s Cancer at the Institute for Women’s Health and Consultant Gynaecological Surgeon.
Partners in the project included: University College London, GATC Biotech AG, Genedata AG, Euram Limited, Ludwig-Maximilians-University (Munich) and Charles University (Prague).
Project Results:
3.1 Project Organisation
The project was organised into three phases with specific endpoints to establish DNAme based markers for diagnosis, early detection and monitoring of treatment for ovarian cancer & breast cancer:
PHASE 1 - Epigenome-wide discovery of cancer specific differentially methylated regions (DMRs) in ovarian and breast cancer.
Project partners collected cancer tissue and serum from patients with all histotypes, grades and stages of breast & ovarian cancer as well as healthy women and women with benign conditions. These samples were analysed alongside white blood cell DNA and tissue from other cancers to ensure that DMRs selected were specific to breast and ovarian cancers. Two complimentary technologies were applied for biomarker discovery.
PHASE 2 - Application of the best performing DMR regions in development of a serum based assay.
Clinical tests for the best performing cancer specific DMRs in serum identified in Phase 1 were developed using two alternative (and more clinically practicable, i.e. high-throughput) technologies: Digital MethyLight and targeted bisulfite sequencing. Targeted bisulfite sequencing was subsequently selected and used to analyse a number of small sample sets and hence to facilitate selection of the most promising DMRs (more details below) which were then employed in Phase 3.
PHASE 3 - Testing the serum test performance in serial samples from clinical trials. We applied the DMRs from Phase 2 to analyse serum samples from clinical trials including the United Kingdom Clinical Trial of Ovarian Cancer Screening (UKCTOCS) trial. Samples analysed included those predating breast or ovarian cancer diagnosis and were used to determine the potential of the EpiFemCare tests to as a screening method both breast and ovarian cancer. The breast cancer specific test was applied to samples from the SUCCESS trial, before and after adjuvant treatment, to assess the secondary potential of the EpiFemCare test to predict therapeutic response. The ovarian cancer specific test was applied to clinical samples prospectively collected during the project from women who undergoing (neo)-adjuvant chemotherapy for ovarian cancer.
3.2 Prospective sample collection
Project partners in London (University College London), Prague (Charles University) and Munich (Ludwig Maximilians Universität) collected serum and matched tissue (where possible) from women with breast or ovarian cancers or benign conditions as well as those with other cancers (such as endometrial cancer) and healthy volunteers.
We collected 409 serum samples from women diagnosed with BC and non-invasive breast malignant diseases; in 105 cases matching tissue was also collected. The collection of samples with a benign histology includes 62 serum samples and serum was also collected from 172 healthy women.
For ovarian cancer, a total of 193 serum samples from women diagnosed with ovarian cancer (malignant or borderline) were collected; in 78 cases matching tissue was also collected. The collection of samples with a benign histology includes 561 serum samples from women with those conditions that provide a positive test-result using current strategies of ovarian cancer screening (i.e. conditions such as serous and mucinous cystadenoma, endometriosis, fibroid, tubovarial abscess, functional cysts, dermoids, ovarian fibromas, non-gynaecological benign conditions and endometrial polyps). Tissue was available in 72 cases. Serum was also collected from 199 healthy women.
Serum sample collection was also widened to include women with metastases to the ovary (32 women), other malignant ovarian conditions such as non-epithelial malignant disease (26 women), non-gynaecological malignancies including vulval and cervical cancers (26 women) and endometrial cancers (205 women) in order to test the specificity of potential markers to discriminate ovarian cancer from other malignant conditions of the female genital tract.
Protocols for sample collection were established at the start of the project and were designed to mimic conditions under which samples in the existing sample sets (UKCTOCS and SUCCESS) were collected.
3.3 Phase 1 - results
Two complementary technical approaches were applied for biomarker discovery in Phase 1: an array based approach for tissue analysis and a sequencing approach. The array based approach used the Illumina Infinium Human Methylation450 BeadChip Array technology to interrogate the methylation status of ~485,000 sites in cancer and control tissue genomes from The Cancer Genome Atlas (TCGA). Genedata Expressionist® for Genomic Profiling was used to develop a pipeline for selecting the most promising cancer-specific DMRs that fulfil the strict specificity criteria for a serum based test.
In parallel, Reduced Representation Bisulfite Sequencing (RRBS) was used to identify and confirm cancer-specific methylated DNA in matching tissue samples. RRBS enhances sequencing coverage of CpG dinucleotides and, by removing CpG-poor, constitutively-methylated, intergenic regions, provides data at single base-pair resolution for CpG islands, promoters and enhancer elements. The protocol includes digestion of DNA with restriction endonuclease MspI leading to high sequencing coverage for the areas subsequently represented – the reduced representation genome. This is a particular advantage for comparative DNAme profiling as most fragments from this reduced representation genome will be sequenced in all RRBS analyses of a given species.
Genedata then developed new multi-step algorithms for scanning the RRBS data for single read methylation patterns that are unique to tumor samples, providing superior specificity and sensitivity for both the discovery and clinical utility of DNA methylation markers. Two approaches were used to rank the patterns for candidate biomarkers for cancer diagnostics:
- A pooled approach of the read per tissue scored patterns based on over-representation within the tumor;
- A non-pooled approach scored patterns based on prevalence within different tumor samples and absence from controls, respectively.
3.4 Phase 2 & 3 – results
The most promising DMRs identified in Phase 1 were selected for further analysis in an assay that would be appropriate for clinical application (Phases 2 and 3). The assay selected was a targeted ultra-deep bisulfite sequencing method performed at GATC Biotech using a two-step PCR approach similar to the recently published BisPCR2 method.
Following development of the clinical assay, selected DMRs were tested in a series of sample sets to narrow down and prioritise selection of the best performing biomarker candidates. For ovarian cancer, two test sets were used (Test Set 1 comprised 59 serum samples, Test Set 2 92 serum samples). DMRs were subsequently explored for their clinical utility in three different settings to fulfil the project’s objectives.
A set of 250 prospectively collected serum samples was used to validate the test in the diagnosis of ovarian cancer through discrimination of benign and cancerous growths within the pelvis. In addition:
- UKCTOCS samples were used to investigate the screening and early detection aspect. A total of 172 samples from the UKCTOCS trial were analysed. This sample set included 43 samples from the control arm collected at the time of recruitment from women who developed an invasive epithelial ovarian cancer within two years after recruitment and 132 samples from the multimodal screening arm collected at the time of recruitment from women who did not develop any cancer within the first five years after recruitment. Samples in the each arm were matched to cases with regards to age at recruitment, center and month of recruitment; and
- Patients undergoing (neo)-adjuvant chemotherapy (NACT) to reduce the tumour load prior to debulking surgery for advanced ovarian cancer were also sampled to investigate the test’s potential for personalising treatment. In these patients, blood was drawn and serum separated before each of three neoadjuvant chemotherapy courses. Full sample sets and linked clinical data were obtained from 25 women in total.
In the case of breast cancer, two sets of serum samples were used to develop the assay: Test Set 1 comprised 30 serum samples and Test Set 2 78 samples. The developed assay was then validated as follows:
- Samples from a total of 419 women in the SUCCESS study (two samples from each volunteer – one prior and one after completion of adjuvant systemic treatment) were analysed; and
- A total of 936 samples were then analysed from within the UKCTOCS sample set. These included all 234 UKCTOCS serum samples from women with BC who provided a serum sample up to three years in advance of breast cancer diagnosis and who subsequently died from breast cancer, 234 women who developed breast cancer within three years of diagnosis and survived BC and 468 women who did not develop any breast cancer over the entire 10-15 year follow up period.
Potential Impact:
4.1 Potential impact – contribution to impacts foreseen in the Work Programme
The EpiFemCare project has contributed directly to the impacts foreseen in the Work Programme 2012: Development of technologies with a view to patient group stratification for personalised medicine applications (FP7-HEALTH-2012-INNOVATION-1).
The EpiFemCare project has developed new and improved tools and technologies that will contribute to enabling the uptake of personalised medicine into clinical practice. These tools and technologies focus on improving results in samples containing very low concentrations of DNA from tumours and are critical to improving the performance of liquid biopsy products in the future.
A specific feature of the EpiFemCare project was the focus on SME-targeted research and innovation. The research intensive SMEs GATC Biotech AG and Genedata AG played leading roles in the project and the new tools they developed during the project will benefit them and help to improve the competitiveness of Europe in this field in the future.
SME Focus – Genedata
For the EpiFemCare project, Genedata scientists managed, processed and analysed terabytes of epigenetic and clinical data from next-generation sequencing, microarray, and qPCR experiments to help identify, confirm and clinically validate biomarkers that should lead to significant improvements in early cancer detection and subsequent patient care.
Working with the Genedata Profiler platform to analyse tumour samples as well as liquid biopsies, the Genedata team developed and implemented new workflows for quality-controlled and standardized processing of Reduced Representation Bisulfite Sequencing (RRBS), which were critical for the success of the project. Genedata scientists detected DMRs of CpG sites and annotated the identified DMRs with nearby genes and CpG islands. Detecting highly diluted tumour specific DNA fragments in blood samples is very challenging. The Genedata team solved this problem by developing new algorithms for scanning the RRBS data for single read methylation patterns that are unique to tumour samples, providing superior specificity and sensitivity for both the discovery and clinical utility of DNA methylation markers. The scientists then ranked the resulting patterns to isolate cancer-specific candidate biomarkers for clinical diagnostics. All the experimental data, including NGS reads data and clinical annotations, was managed by Genedata.
SME Focus – GATC Biotech
The sequencing of several thousands of cell-free DNA samples within the EpiFemCare project led to the optimisation of highly standardised workflows for oncology products at GATC and contributed to launch of the GATCLIQUID service line. GATC Biotech has developed specialised protocols for cell-free DNA extraction that are capable of handling the very low concentrations and highly fragmented nature of the cell-free DNA. Further GATC Biotech developed protocols for RRBS and targeted bisulfite sequencing for the discovery and validation of epigenetic markers in the EpiFemCare samples.
4.2 Potential impact – for the scientific and clinical community
The EpiFemCare project has built on existing knowledge to demonstrate the huge potential of circulating tumour DNA and the analysis of epigenetic modification (methylation) in serum for identification of cancer biomarkers. The DNA methylome was selected as the target for personalised medicine due to the fact that it offers several advantages over gene-expression microarrays and proteomic approaches because (i) the DNA molecule is very stable; (ii) it can be compared with absolute reference points which greatly simplifies the design of internal references; (iii) abnormal methylation patterns in cancer cells differ qualitatively from normal cells, not just quantitatively allowing for the development of assays with high specificity and sensitivity in the presence of large amount of normal background DNA; (iv) methylation assays for individual markers tend to be universal whereas genetic mutation assays usually have to be tailored to the individual tumour; and finally (v) DNA methylation patterns are fairly stable over time and do not fluctuate in response to short-term stimuli, as gene-expression profiles do.
We have demonstrated that reduced representation bisulfite sequencing offers a cost effective alternative to whole-genome bisulfite sequencing for the identification of differentially methylated regions in DNA samples and that, in comparison to microarray-based methods, next-generation sequencing enables single nucleotide resolution, analysis of more CpG sites, and analysis of samples containing as little as 30-50ng DNA.
In the field of bioinformatics, project partners have developed the novel bioinformatic tools required to identify the most relevant and most promising cancer-specific methylated DNA regions that show a consistent methylation pattern across all linked CpGs of a particular region in the cancer sample and not in any normal sample (for example in white blood cells).
4.3 Socio-economic impact and the wider societal implications of the project so far
Cervical cancer serves as a shining example of the potential of an early detection strategy: the lifetime risk of dying from cervical cancer has fallen by more than 80% for women who attend cervical cancer screening programmes. Mammography screening for breast cancer is able to save some lives, but treatment outcome is improved for only 3-13% of women attending screening; the large majority do not receive a clear treatment benefit. No effective population-based ovarian cancer screening is available. In both cases, therefore, the availability of a simple, non-invasive test that would identify women with early stage cancer would be beneficial.
A cancer screening programme is only considered to be effective if >75% of the population participates. Current screening programs face problems with uptake mainly due to the inconvenience, pain and embarrassment a screening test imposes on women and the fact that multiple different strategies are needed for different cancers. The use of a blood test to provide a liquid biopsy result for women entering the screening programme, as designed in the EpiFemCare project, would be less invasive and imposing than current methodologies.
Beside individualised early detection, monitoring adjuvant treatment is likewise important. Breast cancer is a heterogeneous disease with diverse morphologies, molecular characteristics, clinical behaviour, and response to therapeutics. As we delve more into the complex nature of this disease, it becomes imperative to determine appropriate prognostic and predictive markers that can be used by physicians and patients for informed decision making.
The EpiFemCare project has investigated whether cancer-specifically methylated DNA in serum is able to act as a surrogate for active cancer and will facilitate future clinical studies immensely. Treatment of breast cancer in the EU costs about €6.73 billion every year and the majority of this goes into systemic treatment. A test which (i) is able to reduce the need (due to early detection) and (ii) allows the monitoring of adjuvant treatment reduces the costs of treatment as well as the costs associated with treatment-related side effects.
4.4 Dissemination activities
The EpiFemCare team has undertaken a number of activities to facilitate the dissemination of the project results. In order to allow publicise the project and its objectives and to promote the free flow of project information, we established at the start of the project a website dedicated to the EpiFemCare project (www.epifemcare.eu). The site had two separate sections: a consortium members’ only private section and a public section. Throughout the project we have used the private section as a repository for project specific documents and information. The public section has hosted information about the project, including its objectives and an outline of the workplan, as well as details about the project partners and emerging project publications and news. The website also holds videos produced by one of the project partners, Euram Ltd, that describe the scientific background to the project in lay terms, its aims and the overall strategy for achieving those aims. Towards the end of the project, the website was redeveloped as a vehicle for presenting the project’s results.
The scientific researchers have made significant efforts throughout the project to engage with the scientific/technical community and have prepared and submitted 9 publications to peer reviewed journals at the time of writing this report. More publications are expected in the coming months. Professor Widschwendter has worked closely with the patient advocacy group of the European Society of Gynaecological Oncology, ENGAGe, to develop awareness of the project and the EpiFemCare team has also developed a leaflet explaining the science behind the project for a lay audience (available via our website). Altogether, a total of 91 activities have been undertaken by the project to develop awareness about the project and present results.
List of Websites:
Address of the project public website
www.epifemcare.eu
Contact details
Prof Martin Widschwendter, MD
Professor in Women's Cancer &
Consultant Gynaecological Oncologist
Department of Women's Cancer,
University College London,
Medical School Building, Room 340,
74 Huntley Street,
London WC1E 6AU,
United Kingdom
Every year in the EU ~45,300 and 330,000 women are diagnosed with ovarian and breast cancer respectively & 28,800 and 90,000 of these two groups of women will die as a consequence of these diseases. Currently there are no tools available that allow for:
a) Effective screening of ovarian and/or breast cancer of sufficient sensitivity and specificity to avoid potential over-diagnosis,
or;
b) Stratification of patients into optimal personalised therapy regimes in ovarian/breast cancer.
EpiFemCare ADVANCES
Progress in personalised cancer medicine will only be possible with the development of bioassays involving the analysis of easy accessible biomaterials that contain stable target molecules reflective of disease. We have established and clinically validated a series of blood tests based upon DNA methylation technology that will facilitate both early detection and prediction of therapeutic outcome in breast and ovarian cancer.
CONSORTIUM
Our pan-European academic-industrial consortium has thrived through a collaboration of teams with diverse clinical, scientific & industrial expertise. Project partners have worked with the latest state-of-the-art technologies and, integrally, the best available cohort and clinical trial sample sets available. GATC-Biotech and Genedata are Europe’s leading providers of DNA sequencing and bioinformatics for biomarker development and together with the leading clinical academic groups in gynaecological oncology – University College London, Charles University Prague and Ludwig-Maximilians University Munich – have developed blood based biomarkers for breast and ovarian cancers.
The clinical partners have provided access to unique cohort through the prospective collection of more than 1880 serum samples from women with breast or ovarian cancers, benign conditions or from healthy volunteers. In addition, serum samples collected from >200,000 women well in advance of disease (UK Collaborative Trial of Ovarian Cancer Screening) or before and during treatment (SUCCESS Trial) were provided by the clinical partners for analysis.
IMPACT
The EpiFemCare project has developed new and improved tools and technologies that will contribute to enabling the uptake of personalised medicine into clinical practice. These tools and technologies focus on improving results in samples containing very low concentrations of DNA from tumours and are critical to improving the performance of liquid biopsy products in the future. They include new workflows for quality-controlled and standardized processing of Reduced Representation Bisulfite Sequencing (RRBS) that have been implemented in Genedata’s Profiler software and highly standardised, optimised workflows for cell-free DNA extraction and targeted sequencing within the GATCLIQUID for oncology products at GATC Biotech.
Project Context and Objectives:
2.1 Project Context
In the European Union and the United States, one in every three individuals faces a cancer diagnosis in their lifetime. In women, approximately half of all cancers originate in the female reproductive organs, i.e. breast, ovary, endometrium and cervix. Breast cancer (BC) is the most common cancer in women, affecting as many as 1 in 8 women. The high mortality rate is related to its tendency to spread: one-third of axillary, node-negative BC patients develop local or distant metastases, even in the absence of tumor spread at the time of primary diagnosis.
Ovarian cancer (OC) has low prevalence but is the most common cause of death from gynecological malignancies. In the early stages, women are mostly asymptomatic or present with non-specific symptoms, making diagnosis difficult. However, while patients with stage I disease have a 5-year survival of >90%, diagnosis is most often made at stage IIIC when 5-year survival is less than 40%. The majority of patients with advanced stage disease undergo recurrence despite aggressive surgery and chemotherapy with platinum and taxanes.
Cervical cancer incidence and associated deaths have been reduced by up to 80% through introduction and widespread uptake of triennial cervical screening with the Papanicolaou test (or Pap smear). In contrast, breast and ovarian cancers continue to pose considerable challenges in terms of prediction and early detection. Unlike cervical cancer, the abnormal cells are not directly accessible. Furthermore, in the case of ovarian cancer, the cell of origin is not well defined.
Ovarian cancer
No effective screening method for OC is available at present. Beyond screening, once a pelvic mass is identified in a patient, a non-invasive test that could help to triage patient care would be highly desirable. Biochemical markers, transvaginal ultrasound, scoring systems and models are widely used to discriminate between benign and malignant disease. The most commonly used marker, CA125, is also expressed in numerous benign conditions, and it is positive in only about 50% of early-stage ovarian cancers. The Risk of Malignancy Algorithm (ROMA), based on assessment of CA125 together with human epididymis secretory protein 4 (HE4), was initially thought to be superior to CA125 alone. However, other groups have suggested that further validation is required. Risk of Malignancy Index (RMI), a score based on ultrasound variables as well as on menopausal status and CA125, is widely used at present, mainly in the United Kingdom. Although the RMI index allows the referring gynaecologist to send a patient to experts based on an objective assessment, its sensitivity is as low as 78%. Transvaginal ultrasound by an expert operator using a formal scoring model system is a highly sensitive and ideal second-stage diagnostic method; however, it is highly dependent on individual expertise. As a result, discrimination between a benign and malignant ovarian tumor remains a significant challenge in clinical practice.
Availability of a biomarker or a panel of biomarkers that can detect OC in its earliest stages with both high sensitivity and specificity would improve outcomes. In addition, they may be used as prognostic or predictive indicators, or as novel targets for cancer treatment.
Breast cancer
The lack of diagnostic markers detectable in early BC is a critical issue in patient management. In clinical practice, since early BC does not cause symptoms, mammography or ultrasound imaging are used for screening. Mammography screening, which detects characteristic masses and/or micro-calcifications, is routinely performed in patients older than 45-50. Intensified breast screening can also be used for younger women when other factors, such as genetic profile or family history, indicate higher risk of disease.
Despite the fact that mammography screening for BC does save some lives, there is a continuing debate about its usefulness. Recent data demonstrate an improved treatment outcome for only 3 13% of women in whom BC was detected by screening mammography. Consequently, it would seem that 87-97% of women with a screening-detected BC do not receive a clear treatment benefit, indicating that these cancers are still not being detected early enough. In short, concerns about the potential risks and harms of the current breast screening program continue, as should the development of novel methods for early detection.
2.2 EpiFemCare’s objectives
The EpiFemCare project aimed to devise a set of bioassays – or liquid biopsies – to meet the unmet needs described above. The project’s objectives were to establish and clinically validate a series of blood tests based upon DNA methylation technology that would facilitate both early detection and prediction of therapeutic outcome in breast and ovarian cancer, as follows:
1. Screening and early detection of breast and ovarian cancer;
2. Diagnosis of ovarian cancers through discrimination of benign and cancerous growths within the pelvis; and
3. Personalised care through monitoring the efficacy of treatment and determining the right drug regimen for each woman.
DNA methylation is a chemical modification of DNA known to respond to the environment and consequentially capable of modifying gene expression. Selection of DNA methylation as the basis for the EpiFemCare blood tests offers key advantages over other potential biomarkers, including genetic mutation, proteins, lipids and RNA. DNA methylation is: (1) biologically and technically stable, (2) has the capacity to be amplified, and (3) demonstrates cancer specificity when compared to normal or benign tissues. Technology platforms for the analysis of DNA methylation have evolved rapidly over the past few years to the point where epigenome-wide analysis can now be performed with relative ease. The EpiFemCare project used technology capable of the sensitive detection of cell free DNA methylation in serum against an excess background of normal contaminating DNA – a feature that is crucial for the development and clinical success of blood based DNA methylation tests.
2.3 EpiFemCare workplan
Work Package 1: Orchestration and establishment of a breast cancer collection optimised for epigenetic analyses
Work Package 2: Orchestration and establishment of an ovarian cancer collection optimised for epigenetic analyses
Work Package 3: Discovery of cancer specific CpGs (Phase 1)
Work Package 4: Data management and analysis of the data generated
Work Package 5: Development and application of DNAme serum assays
Work Package 6: Knowledge Dissemination, IP Use & exploitation
Work Package 7: Project Management and Administration
2.4 EpiFemCare partners
EpiFemCare brings together 6 leading research organisations and SMEs from across Europe. The project was led by Professor Martin Widschwendter from University College London, who is an internationally renowned researcher in epigenetics, Head of the UCL Department of Women’s Cancer at the Institute for Women’s Health and Consultant Gynaecological Surgeon.
Partners in the project included: University College London, GATC Biotech AG, Genedata AG, Euram Limited, Ludwig-Maximilians-University (Munich) and Charles University (Prague).
Project Results:
3.1 Project Organisation
The project was organised into three phases with specific endpoints to establish DNAme based markers for diagnosis, early detection and monitoring of treatment for ovarian cancer & breast cancer:
PHASE 1 - Epigenome-wide discovery of cancer specific differentially methylated regions (DMRs) in ovarian and breast cancer.
Project partners collected cancer tissue and serum from patients with all histotypes, grades and stages of breast & ovarian cancer as well as healthy women and women with benign conditions. These samples were analysed alongside white blood cell DNA and tissue from other cancers to ensure that DMRs selected were specific to breast and ovarian cancers. Two complimentary technologies were applied for biomarker discovery.
PHASE 2 - Application of the best performing DMR regions in development of a serum based assay.
Clinical tests for the best performing cancer specific DMRs in serum identified in Phase 1 were developed using two alternative (and more clinically practicable, i.e. high-throughput) technologies: Digital MethyLight and targeted bisulfite sequencing. Targeted bisulfite sequencing was subsequently selected and used to analyse a number of small sample sets and hence to facilitate selection of the most promising DMRs (more details below) which were then employed in Phase 3.
PHASE 3 - Testing the serum test performance in serial samples from clinical trials. We applied the DMRs from Phase 2 to analyse serum samples from clinical trials including the United Kingdom Clinical Trial of Ovarian Cancer Screening (UKCTOCS) trial. Samples analysed included those predating breast or ovarian cancer diagnosis and were used to determine the potential of the EpiFemCare tests to as a screening method both breast and ovarian cancer. The breast cancer specific test was applied to samples from the SUCCESS trial, before and after adjuvant treatment, to assess the secondary potential of the EpiFemCare test to predict therapeutic response. The ovarian cancer specific test was applied to clinical samples prospectively collected during the project from women who undergoing (neo)-adjuvant chemotherapy for ovarian cancer.
3.2 Prospective sample collection
Project partners in London (University College London), Prague (Charles University) and Munich (Ludwig Maximilians Universität) collected serum and matched tissue (where possible) from women with breast or ovarian cancers or benign conditions as well as those with other cancers (such as endometrial cancer) and healthy volunteers.
We collected 409 serum samples from women diagnosed with BC and non-invasive breast malignant diseases; in 105 cases matching tissue was also collected. The collection of samples with a benign histology includes 62 serum samples and serum was also collected from 172 healthy women.
For ovarian cancer, a total of 193 serum samples from women diagnosed with ovarian cancer (malignant or borderline) were collected; in 78 cases matching tissue was also collected. The collection of samples with a benign histology includes 561 serum samples from women with those conditions that provide a positive test-result using current strategies of ovarian cancer screening (i.e. conditions such as serous and mucinous cystadenoma, endometriosis, fibroid, tubovarial abscess, functional cysts, dermoids, ovarian fibromas, non-gynaecological benign conditions and endometrial polyps). Tissue was available in 72 cases. Serum was also collected from 199 healthy women.
Serum sample collection was also widened to include women with metastases to the ovary (32 women), other malignant ovarian conditions such as non-epithelial malignant disease (26 women), non-gynaecological malignancies including vulval and cervical cancers (26 women) and endometrial cancers (205 women) in order to test the specificity of potential markers to discriminate ovarian cancer from other malignant conditions of the female genital tract.
Protocols for sample collection were established at the start of the project and were designed to mimic conditions under which samples in the existing sample sets (UKCTOCS and SUCCESS) were collected.
3.3 Phase 1 - results
Two complementary technical approaches were applied for biomarker discovery in Phase 1: an array based approach for tissue analysis and a sequencing approach. The array based approach used the Illumina Infinium Human Methylation450 BeadChip Array technology to interrogate the methylation status of ~485,000 sites in cancer and control tissue genomes from The Cancer Genome Atlas (TCGA). Genedata Expressionist® for Genomic Profiling was used to develop a pipeline for selecting the most promising cancer-specific DMRs that fulfil the strict specificity criteria for a serum based test.
In parallel, Reduced Representation Bisulfite Sequencing (RRBS) was used to identify and confirm cancer-specific methylated DNA in matching tissue samples. RRBS enhances sequencing coverage of CpG dinucleotides and, by removing CpG-poor, constitutively-methylated, intergenic regions, provides data at single base-pair resolution for CpG islands, promoters and enhancer elements. The protocol includes digestion of DNA with restriction endonuclease MspI leading to high sequencing coverage for the areas subsequently represented – the reduced representation genome. This is a particular advantage for comparative DNAme profiling as most fragments from this reduced representation genome will be sequenced in all RRBS analyses of a given species.
Genedata then developed new multi-step algorithms for scanning the RRBS data for single read methylation patterns that are unique to tumor samples, providing superior specificity and sensitivity for both the discovery and clinical utility of DNA methylation markers. Two approaches were used to rank the patterns for candidate biomarkers for cancer diagnostics:
- A pooled approach of the read per tissue scored patterns based on over-representation within the tumor;
- A non-pooled approach scored patterns based on prevalence within different tumor samples and absence from controls, respectively.
3.4 Phase 2 & 3 – results
The most promising DMRs identified in Phase 1 were selected for further analysis in an assay that would be appropriate for clinical application (Phases 2 and 3). The assay selected was a targeted ultra-deep bisulfite sequencing method performed at GATC Biotech using a two-step PCR approach similar to the recently published BisPCR2 method.
Following development of the clinical assay, selected DMRs were tested in a series of sample sets to narrow down and prioritise selection of the best performing biomarker candidates. For ovarian cancer, two test sets were used (Test Set 1 comprised 59 serum samples, Test Set 2 92 serum samples). DMRs were subsequently explored for their clinical utility in three different settings to fulfil the project’s objectives.
A set of 250 prospectively collected serum samples was used to validate the test in the diagnosis of ovarian cancer through discrimination of benign and cancerous growths within the pelvis. In addition:
- UKCTOCS samples were used to investigate the screening and early detection aspect. A total of 172 samples from the UKCTOCS trial were analysed. This sample set included 43 samples from the control arm collected at the time of recruitment from women who developed an invasive epithelial ovarian cancer within two years after recruitment and 132 samples from the multimodal screening arm collected at the time of recruitment from women who did not develop any cancer within the first five years after recruitment. Samples in the each arm were matched to cases with regards to age at recruitment, center and month of recruitment; and
- Patients undergoing (neo)-adjuvant chemotherapy (NACT) to reduce the tumour load prior to debulking surgery for advanced ovarian cancer were also sampled to investigate the test’s potential for personalising treatment. In these patients, blood was drawn and serum separated before each of three neoadjuvant chemotherapy courses. Full sample sets and linked clinical data were obtained from 25 women in total.
In the case of breast cancer, two sets of serum samples were used to develop the assay: Test Set 1 comprised 30 serum samples and Test Set 2 78 samples. The developed assay was then validated as follows:
- Samples from a total of 419 women in the SUCCESS study (two samples from each volunteer – one prior and one after completion of adjuvant systemic treatment) were analysed; and
- A total of 936 samples were then analysed from within the UKCTOCS sample set. These included all 234 UKCTOCS serum samples from women with BC who provided a serum sample up to three years in advance of breast cancer diagnosis and who subsequently died from breast cancer, 234 women who developed breast cancer within three years of diagnosis and survived BC and 468 women who did not develop any breast cancer over the entire 10-15 year follow up period.
Potential Impact:
4.1 Potential impact – contribution to impacts foreseen in the Work Programme
The EpiFemCare project has contributed directly to the impacts foreseen in the Work Programme 2012: Development of technologies with a view to patient group stratification for personalised medicine applications (FP7-HEALTH-2012-INNOVATION-1).
The EpiFemCare project has developed new and improved tools and technologies that will contribute to enabling the uptake of personalised medicine into clinical practice. These tools and technologies focus on improving results in samples containing very low concentrations of DNA from tumours and are critical to improving the performance of liquid biopsy products in the future.
A specific feature of the EpiFemCare project was the focus on SME-targeted research and innovation. The research intensive SMEs GATC Biotech AG and Genedata AG played leading roles in the project and the new tools they developed during the project will benefit them and help to improve the competitiveness of Europe in this field in the future.
SME Focus – Genedata
For the EpiFemCare project, Genedata scientists managed, processed and analysed terabytes of epigenetic and clinical data from next-generation sequencing, microarray, and qPCR experiments to help identify, confirm and clinically validate biomarkers that should lead to significant improvements in early cancer detection and subsequent patient care.
Working with the Genedata Profiler platform to analyse tumour samples as well as liquid biopsies, the Genedata team developed and implemented new workflows for quality-controlled and standardized processing of Reduced Representation Bisulfite Sequencing (RRBS), which were critical for the success of the project. Genedata scientists detected DMRs of CpG sites and annotated the identified DMRs with nearby genes and CpG islands. Detecting highly diluted tumour specific DNA fragments in blood samples is very challenging. The Genedata team solved this problem by developing new algorithms for scanning the RRBS data for single read methylation patterns that are unique to tumour samples, providing superior specificity and sensitivity for both the discovery and clinical utility of DNA methylation markers. The scientists then ranked the resulting patterns to isolate cancer-specific candidate biomarkers for clinical diagnostics. All the experimental data, including NGS reads data and clinical annotations, was managed by Genedata.
SME Focus – GATC Biotech
The sequencing of several thousands of cell-free DNA samples within the EpiFemCare project led to the optimisation of highly standardised workflows for oncology products at GATC and contributed to launch of the GATCLIQUID service line. GATC Biotech has developed specialised protocols for cell-free DNA extraction that are capable of handling the very low concentrations and highly fragmented nature of the cell-free DNA. Further GATC Biotech developed protocols for RRBS and targeted bisulfite sequencing for the discovery and validation of epigenetic markers in the EpiFemCare samples.
4.2 Potential impact – for the scientific and clinical community
The EpiFemCare project has built on existing knowledge to demonstrate the huge potential of circulating tumour DNA and the analysis of epigenetic modification (methylation) in serum for identification of cancer biomarkers. The DNA methylome was selected as the target for personalised medicine due to the fact that it offers several advantages over gene-expression microarrays and proteomic approaches because (i) the DNA molecule is very stable; (ii) it can be compared with absolute reference points which greatly simplifies the design of internal references; (iii) abnormal methylation patterns in cancer cells differ qualitatively from normal cells, not just quantitatively allowing for the development of assays with high specificity and sensitivity in the presence of large amount of normal background DNA; (iv) methylation assays for individual markers tend to be universal whereas genetic mutation assays usually have to be tailored to the individual tumour; and finally (v) DNA methylation patterns are fairly stable over time and do not fluctuate in response to short-term stimuli, as gene-expression profiles do.
We have demonstrated that reduced representation bisulfite sequencing offers a cost effective alternative to whole-genome bisulfite sequencing for the identification of differentially methylated regions in DNA samples and that, in comparison to microarray-based methods, next-generation sequencing enables single nucleotide resolution, analysis of more CpG sites, and analysis of samples containing as little as 30-50ng DNA.
In the field of bioinformatics, project partners have developed the novel bioinformatic tools required to identify the most relevant and most promising cancer-specific methylated DNA regions that show a consistent methylation pattern across all linked CpGs of a particular region in the cancer sample and not in any normal sample (for example in white blood cells).
4.3 Socio-economic impact and the wider societal implications of the project so far
Cervical cancer serves as a shining example of the potential of an early detection strategy: the lifetime risk of dying from cervical cancer has fallen by more than 80% for women who attend cervical cancer screening programmes. Mammography screening for breast cancer is able to save some lives, but treatment outcome is improved for only 3-13% of women attending screening; the large majority do not receive a clear treatment benefit. No effective population-based ovarian cancer screening is available. In both cases, therefore, the availability of a simple, non-invasive test that would identify women with early stage cancer would be beneficial.
A cancer screening programme is only considered to be effective if >75% of the population participates. Current screening programs face problems with uptake mainly due to the inconvenience, pain and embarrassment a screening test imposes on women and the fact that multiple different strategies are needed for different cancers. The use of a blood test to provide a liquid biopsy result for women entering the screening programme, as designed in the EpiFemCare project, would be less invasive and imposing than current methodologies.
Beside individualised early detection, monitoring adjuvant treatment is likewise important. Breast cancer is a heterogeneous disease with diverse morphologies, molecular characteristics, clinical behaviour, and response to therapeutics. As we delve more into the complex nature of this disease, it becomes imperative to determine appropriate prognostic and predictive markers that can be used by physicians and patients for informed decision making.
The EpiFemCare project has investigated whether cancer-specifically methylated DNA in serum is able to act as a surrogate for active cancer and will facilitate future clinical studies immensely. Treatment of breast cancer in the EU costs about €6.73 billion every year and the majority of this goes into systemic treatment. A test which (i) is able to reduce the need (due to early detection) and (ii) allows the monitoring of adjuvant treatment reduces the costs of treatment as well as the costs associated with treatment-related side effects.
4.4 Dissemination activities
The EpiFemCare team has undertaken a number of activities to facilitate the dissemination of the project results. In order to allow publicise the project and its objectives and to promote the free flow of project information, we established at the start of the project a website dedicated to the EpiFemCare project (www.epifemcare.eu). The site had two separate sections: a consortium members’ only private section and a public section. Throughout the project we have used the private section as a repository for project specific documents and information. The public section has hosted information about the project, including its objectives and an outline of the workplan, as well as details about the project partners and emerging project publications and news. The website also holds videos produced by one of the project partners, Euram Ltd, that describe the scientific background to the project in lay terms, its aims and the overall strategy for achieving those aims. Towards the end of the project, the website was redeveloped as a vehicle for presenting the project’s results.
The scientific researchers have made significant efforts throughout the project to engage with the scientific/technical community and have prepared and submitted 9 publications to peer reviewed journals at the time of writing this report. More publications are expected in the coming months. Professor Widschwendter has worked closely with the patient advocacy group of the European Society of Gynaecological Oncology, ENGAGe, to develop awareness of the project and the EpiFemCare team has also developed a leaflet explaining the science behind the project for a lay audience (available via our website). Altogether, a total of 91 activities have been undertaken by the project to develop awareness about the project and present results.
List of Websites:
Address of the project public website
www.epifemcare.eu
Contact details
Prof Martin Widschwendter, MD
Professor in Women's Cancer &
Consultant Gynaecological Oncologist
Department of Women's Cancer,
University College London,
Medical School Building, Room 340,
74 Huntley Street,
London WC1E 6AU,
United Kingdom