WP6 - Molecular Pathology and Clinical Outcome

Work package number 6 Start date or starting event:  Month 1
Work package title Molecular Pathology and Clinical Outcome
Activity Type RTD
Participant number 9 1
8 11 14
Person-months participant
52 31
37 32.4 8


  1. To characterise the (molecular) phenotypes of tumours using a combination of histopathology, immuno-histochemistry, copy number analysis and gene expressionmicroarrays

  2. To evaluate the associations between genetic and environmental factors and (molecular) phenotype of disease

TRANSBIG is a European Network of Excellence funded by the European Union 6th Framework Programme. TRANSBIG governs and initiates the scientific work around the EORTC clinical trial Mindact.

Mindact-Transbig funds have been secured, including the EU for approximately one fifth, as well as various other sponsors, that allow banking of tumor tissue, frozen and in paraffin embedded in 'tissue-microarrays', banking of serum and whole blood, as well as the experimental execution of RNA isolation from frozen tumor tissue and subsequent hybridization and data storage of whole genome expression microarrays. There is no funding for any other research.

COGS will provide the funds to ship blood samples to a central laboratory facility, extract DNA from pheripheral blood banked by Transbig and subsequent costs of experiments for SNP analysis (1,536) SNPs per individual. Transbig will make available to COGS, stored data of whole genome expression arrays of tumor tissue, as well as basic demographic data and results of immunohistochemistry stained tissue-microarrays of patients entered in the Mindact trial following the guidelines set in TRANSBIG for these.

COGS will conduct the data analysis correlating patient's SNP status to tumor characteristics. WP6 aims to unravel the contribution of an individual’s cancer susceptibility confined by genetic make-up (i.e., SNP profile from genome-wide and locus specific association studies, WP2, WP3 and later WP4 respectively, as well as alleles of known high-risk susceptibility genes) together with life- style/environmental factors (WP5) in relation to the development of a specific type of breast, ovarian or prostate cancer (in collaboration with WP2 and WP5) (overall coordination WP6 participant 8). We hypothesize that histological/molecular tumour characteristics are related to the etiologic pathways of cancer. It is known that high-risk alleles such as BRCA gene mutations are associated with specific histological and molecular breast tumour features, and it has also been suggested that, for instance, hormone replacement therapy influences breast tumour characteristics.

Therefore, it would be of great interest to correlate newly identified low-risk SNP variants, individually and combined with high- risk alleles and environmental factors, to tumour histological and molecular tumour profiles, hypothesizing that specific risk factor associations might be stronger for distinct subtypes. Similar relationships are expected also for ovarian and prostate cancer. As such, tumour subtypes may also guide the weighing of the importance of individual genetic variants and epidemiological risk factors. In addition, these specific etiologies may affect the clinical outcome of disease. Unravelling these dependencies will therefore give new insights for the design of cancer preventive and therapeutic interventions (WP7).

COGS WP6 will have the unique possibility to combine world’s largest datasets on germline SNPs (this project), within well-established international consortia with available tumour phenotypic and epidemiological data: standard histopathology and life-style/environmental data in all consortia, but most special unsurpassed large genome-wide expression and copy number data from EU 6th framework project TRANSBIG, and NBAC, OCAC respectively. Collaboration for unselected breast cancer (BCAC, TRANSBIG, NBAC; for WP6 coordinated by participants 7, 8, 11), ovarian cancer (OCAC; for WP6 coordinated by participant 11) and prostate cancer (PRACTICAL; for WP6 coordinated by participant 1) and BRCA1/2 mutation carriers (CIMBA/IBCCS/NBAC; for WP6 coordinated by participant 9).

Description of work.
In WP6, tumour subtypes within breast, ovarian and prostate cancer will be established using 1) standard histopathology including standard immunohistochemical markers and in addition (partly) tissue microarrays (TMAs) harbouring representative core punches from tumours (BCAC, TRANSBIG, OCAC, PRACTICAL); 2) gene expression microarrays (TRANSBIG, NBAC); 3) comparative genomic analysis (CGH) (OCAC, NBAC). Histopathology, immunotypes, gene expression, and genomic profiles will be available to WP6 and have in large part already been obtained, or are being collected in ongoing prospective or retrospective studies.

For evaluation of standard histopathology BCAC up to 15000 cases, TRANSBIG up to 4000, OCAC up to 4000, CIMBA/IBCCS up to 2000 and PRACTICAL up to 10000. Part of these will be available also with additional immunohistochemical data derived from TMAs. TRANSBIG will make available the microarray expression data (44K Agilent arrays) for at least 4000 with maximum of 6000 samples (specified conditions). NBAC will have gene expression microarrays on up to 1000 tumours. Genomic profiling by CGH will be available in OCAC on up to 4000 samples and in NBAC on up to 700 breast cancers.

Task 6.1 Standardization of existing histopathological data:

Standardization of histopathological and immunotyping data is mandatory to allow evaluation across contributing sub-studies. A codebook with standardized description of parameters and a database will be constructed. In addition, a TMA scoring system will implemented.

Task 6.2 Evaluation of cancer subtype (histopathology) within breast, ovarian, and prostate cancer with epidemiological data and genetic loci already identified:
Initially the cancer subtype association analysis will be conducted using defined subtypes (e.g., breast cancer invasive ductal/lobular etc, basal/luminal; ovarian cancer serous/mucinous, prostate cancer gleason score):

  1. For breast cancer, the analysis will include at least ~8 most significant gene loci already identified in the existing GWAs of breast cancer and those identified in the pooled analysis of GWAs conducted by WP3. Genotype data for these genetic variants are already available for ca. 22000 cases and 22000 controls in BCAC. Similar genotype data will be available for CIMBA (12000 BRCA1/2 mutation carriers) and analyzed in tumours from carriers.

  2. For ovarian cancer, several novel genetic loci are expected to emerge from the ongoing GWAs in Cambridge and OCAC by 2009. In the GWA carried out in Cambridge, 20000 SNPs will be genotyped in 5000 cases and 5000 controls.

  3. For prostate cancer, analysis can be carried with about 5 novel gene loci of prostate cancer currently identified in PRACTICAL. Standard statistical methods (regression analysis) will be used to assess both multiplicative as well as additive interaction since the latter is of particular interest for public health. Multiple testing will be accounted for. Besides, analysis of cancer subtype with single genetic loci, analysis of cancer type by multiple genes will also be conducted, restricted to gene-gene interactions identified from novel genetic variants of existing genome wide data and candidate gene analyses. For the analyses, besides using standard statistical methods, new statistical methodology (WP2) will be employed where appropriate. Furthermore, gene-gene-environmental interaction and cancer type etiology will also be conducted.

Task 6.3 Evaluation of cancer subtype (histopathology) with lifestyle / environmental and novel genetic loci identified in this project:
Further cancer type analyses will be carried out to account for the results emerging from the Illumina array of 1500 selected SNPs in 10000 breast cancer cases and 10000 controls, 5000 ovarian cancer cases and 5000 controls, and 10000 prostate cancer cases and 10000 controls (WP3, WP4). Also other confirmed loci from the consortia or their members could be included in the analyses. Two approaches will be pursued: (a) analysis of gxe of environmental/lifestyle risk factors with most significant SNPs for the respective cancer sites, and (b) explorative analysis of gxe of environmental/lifestyle risk factors with the 50 most significant SNPs for the respective cancer sites. The analyses will be carried out as indicated above in task 6.2., including gene-gene-environmental interaction and cancer type etiology.

Task 6.4. Evaluation of cancer molecular subtypes within breast and ovarian cancer (histopathology TMAs): For part of the breast, ovarian and prostate cancers with existing standard histopathological data (approximately one third) TMAs stained with standard and additional immunohistochemical markers are made or under construction in the respective consortia (BCAC, NBAC, CIMBA, OCAC, PRACTICAL). An inventory of available TMAs stained with relevant markers should become available in year 2-3 and when feasible will then be coded and integrated in the COGS analysis following procedures as described under task 6.2.

Task 6.5. Evaluation of cancer molecular subtypes within breast (genomic and expression arrays) and ovarian cancer (genomic and expression arrays) with epidemiological data and genetic loci: This WP will have the unique opportunity to combine the genetic risk associated germline loci with detailed molecular information of the patient’s tumours. TRANSBIG is conducting the clinical trial MINDACT (EORTC 10041/BIG04-3; accrual start Jan 2007-anticipated end 2010 where 6000 breast cancer patients receive risk assessment for recurrence of disease by either clinico-pathological risk assessment, as well as by gene expression profiling for the 70 gene prognosis profile. In addition, for each patient a full genome 44K Agilent array will be made.

The microarray raw data can become available after the end of the recruitment period of MINDACT; the presentation and publication of results related to these data can only be done after the presentation/publication of the MINDACT first efficacy results. In task 6.4 we will integrate these two data sets and analyse them as described in 6.2/6.3. In addition, new data mining algorithms will be employed that will be specifically developed within this project (WP2) to use existing knowledge on molecular subtype (e.g., basal/luminal), to see whether this gives further indications for relevant SNPs from the genome wide association study.

Task 6.4 will also be conducted on an already existing dataset from NBAC, consisting of gene expression profiles (oligo arrays) from ~700 breast cancers, enriched with cases from BRCA1/BRCA2 carriers and non-BRCA1/2 familial cases, as well as of genomic profiles (tiling BAC array CGH) from the same material. This dataset will be extended by molecular profiling of additional tumours to reach a number of ~1000. Furthermore, of the latter, CpG methylation patterns (Illumina) will be available, allowing to provide further evidence of distinct breast cancer subtypes, and making the correlation to constitutional genetic make-up more profound.

Task 6.6. Evaluation of cancer molecular subtypes within breast and ovarian cancer with novel genetic loci and significant epidemiological factors identified in this project: Significant cancer subtype determinants of specific environmental/lifestyle factors and genes will be used to further refine a risk prediction model (WP2), which will be the basis of the implementation of the results (WP7); It is envisaged that the results will lead to prediction model of each of the cancers subtypes in which genes and environmental factors are included.


D 6.1 Standardization of histopathological and immunohisto data; codebook (task 6.1, month 8)

D 6.2 TMA scoring system agreed, report (task 6.1, month 12)

D 6.3 Database for histopathological and immunohisto data (task 6.1, month 12 - Milestone 6.1)

D 6.4 Existing data coded in developed database, report (task 6.1, month 16 - Milestone 6.2)

D 6.5 SNP existing genotype data WP3 added to database (task 6.2, month 16 – Milestone 6.3)

D 6.6 Genotype (G) and tumour type (T)(standard histopathology) analysis, report (task 6.2, month 18)

D 6.7 Genotype/environment interaction (GxE) and (T) analysis, report (task 6.2, month 20)

D 6.8 Draft manuscripts task 6.2 (month 24 – Milestone 6.4)

D 6.9 New G data WP3/WP4 added to database (task 6.3, month 26 – Milestone 6.5)

D 6.10 New G and T(standard histopathology) analysis, report (task 6.3, month 28)

D 6.11 New GxE and T(standard histopathology) analysis, report (task 6.3, month 30)

D 6.12 Draft manuscripts task 6.3 (month 33 – Milestone 6.6)

D 6.13 Coding of available TMA staining in developed database (task 6.4, month 32 - Milestone 6.7)

D 6.14 Initial analysis G and T (TMA), report (task 6.4, month 34)

D 6.15 Initial analysis G and T(expression or CGH-array), report (task 6.5, month 36)

D 6.16 Initial analysis GxE and T (TMA), report (task 6.5, month 38)

D 6.17 Initial analysis GxE and T(expression or CGH-array), report (task 6.5, month 40)

D 6.18 New refined G & E data WP3/4/5 added to database (task 6.6, month 38 Milestone 6.9)

D 6.19 Draft manuscripts task 6.4-5 (month 42 – Milestone 6.10)

D 6.20 Draft manuscripts task 6.6 (month 45 – Milestone 6.10)

Additional information