Prospective Breast Cancer Biobanking

A. Early detection of systemic relapses in breast cancer.

WP 1. Analysis of primary tumor.

Analysis of primary tumor and metastases (Pathology group). In this work package both DNA and RNA from the primary tumors of 125 high-risk (Luminal-B) patients will be analyzed by next-generation sequencing (NGS). Furthermore, the investigators will also analyze DNA and RNA from the available metastatic lesions (currently material available for 3/8) that have appeared, and will appear during the study period.

MiRNA and mRNA will be profiled using PureLink™ miRNA Isolation Kit (Invitrogen), Dynabeads mRNA direct micro kit (Ambion),) and Ion total RNA-seq kit V2 (Life Technologies) on our Ion Proton NGS instrument. Bioinformatic analyses of paired mRNA and microRNA expression profiles will be used to reveal differentially expressed microRNAs between patients with and without recurrences under treatment. These data will be linked to our previous studies and used to identify target candidates for microRNA analysis in exosomes and TEPs.

In addition, NGS will also be performed on DNA from the primary tumors and metastases to verify DNA mutations especially in ER and other known oncogenes, including those examined in WP3. The DNA sequencing will be performed using the Oncomine Comprehensive Assay. This is an amplicon based approach analyzing hotspot SNVs, indels, and CNVs. These analyses will provide us with a biological understanding and background knowledge for each of the individual tumor, and these datasets can also be compared to the recently published data.

WP 2. Circulating tumor celles (CTC).

Circulating tumor cells (Oncology group) Circulating tumor cells are being enriched from peripheral blood samples by density centrifugation and subsequent immunomagnetic depletion of leucocytes. The oncology group has recently developed and published a new depletion method, termed MINDEC (Multi- marker Immuno- magnetic Negative Depletion Enrichment of CTCs), which has superior recovery and enrichment rates. RNA and DNA are isolated simultaneously from the enriched fraction, to allow CTC detection by both RNA and DNA-based approaches. Specific mRNA markers, with high levels in tumor cells and low levels in normal leukocytes, are being pre-amplified and quantified by real-time PCR as surrogate markers for CTCs. Both epithelial-specific markers and markers related to epithelial mesenchymal transition are part of the marker panel. Until now, the investigatorshave analyzed 170 consecutive blood samples from the larger PBCB cohort, enriched by the MINDEC procedure and found evidence for CTCs in about 20% of the samples (from both low- and high-risk patients), a finding that encourages us to continue this project. The investigators have also analyzed blood samples from 30 healthy female volunteers for comparison. All collected blood samples from 125 high-risk patients will be analyzed (around 575 samples). The presence or level of CTCs in the analyzed samples will later be compared to known prognostic factors, treatment effect and disease outcome.

WP 3. circulating tumor DNA (ctDNA).

Tumor-specific mutations can be utilized as markers for ctDNA because they are not present in normal cells and normal plasma DNA. The oncology group has recently demonstrated the clinical relevance of ctDNA measurements in pancreatic cancer. The investigators will now measure ctDNA levels in plasma samples from high-risk breast cancer patients by targeted next-generation sequencing. The recently released "Oncomine breast cfDNA assay" (Thermo Fisher) will be used to detect mutations in a panel of ten genes that are frequently mutated in breast cancer. The assay is based on molecular barcoding of templates and allows reproducible detection of mutations down to 0,1% control samples. Taking the low concentration of cell free DNA (=cfDNA) in plasma into account (typically 10 ng per ml blood), a sensitivity of 0,1% is considered sufficient. The sequencing will be performed on our Ion Proton NGS instrument (Life Technologies). The pre-operative blood sample and yearly follow-up samples from the 125 high-risk patients will be analyzed in this WP. Blood samples from 30 healthy female volunteers have already been collected and will be analyzed for comparison.

The presence and level of tumor-specific mutations will be analyzed in relation to treatment effect and disease outcome. The mutational profile in the primary tumor biopsy and the plasma samples will be compared in order to reveal potential heterogeneity. In addition, longitudinal changes in the mutation profile of ctDNA will be compared with disease development to potentially shed some light on the biological mechanisms causing treatment resistance or late disease relapses.

WP 4. microRNA (miRNA)

Circulating microRNA from exosomes and TEPs (Pathology group) Total RNA will be isolated from exosomes and TEPs from the blood samples taken before and during treatment. The investigators have recently established methods for isolation of total RNA from exosomes (using exoRNeasy serum Plasma kit (Qiagen) and miRCURY RNA isolation kit (Exiqon)) and TEPs. Using the mentioned protocol, total RNA from exosomes has already been isolated from all the 125 high risk patients first visit/before treatment, additionally from the last blood samples collected will from these patients, total RNA will be isolated from both exosomes and TEPs. Furthermore, microRNA profile will be performed on the isolated RNA using the pipeline and platform already established for our Ion Proton instrument. These profiles will be compared with the bioinformatic analysis of mRNA-miRNA profiles from the primary tumor (WP1) and the samples taken before treatment. MicroRNA that are not present in the tissue sample and/or in the blood sample before treatment, but do appear in blood right before a recurrence appears can then be retrospectively traced in previous blood samples in order to see how sensitive these microRNAs are in predicting treatment resistance. MicroRNA profile from TEPs will be compared to the microRNA profile from exosomes and to the mRNA-miRNA profiles from the primary tumor (WP1, to see if TEPs reflects the tumor and if they have the potential to predict relapse.

WP 5. Metabolomics

Metabolomics is the study of small molecules comprising substrates, intermediates and end products of cellular metabolism, such as amino acids, sugars and small organic acids. The metabolic state of cancer cells is substantially altered compared to normal cells, a fact that can be utilized for diagnostic purposes. Specific metabolic signatures from tumor tissue provide additional information for determination of breast cancer subtypes and prediction of outcome1. For example, increased tumor lactate and glycine levels are related to poor prognosis in patients with estrogen receptor (ER) positive cancer. Metabolomic analyses of primary tumors have also demonstrated predictive value in relation to neoadjuvant treatment of patients with locally advanced disease.

Circulating metabolites have also been shown to provide prognostic information in operable breast cancer and further stratify risk within existing genetically determined risk categories. Importantly, metabolomic alterations may arise directly or indirectly from micrometastatic disease, rather than primary tumor. Recently, the investigators have shown that systemic lactate and pyruvate levels predict inferior outcome in patients with operable ER-positive breast cancers. Tumor-adjacent tissue and immunological responses may also contribute to an altered metabolomic profile. There is also evidence that metabolic profiling can be used for patient monitoring in some cancers, although such evidence is still lacking in breast cancer. Therefore, the investigators intend to investigate whether postoperative monitoring by means of metabolic profiling in blood is useful for early detection of breast cancer recurrence.

WP 6. Integrative molecular monitoring for recurrence detection

New technology has revolutionized the level of biological information that can be obtained from clinical samples, represented by the new "omics" terms genomics, transcriptomics, metabolomics, etc. The availability of such big datasets, even in the public domain, has encouraged the development of integrative methods that can extract vital information from multiple combined data sources. Surprising new connections between omics-datasets have been revealed by such approaches, exemplified by a link between cell-free DNA fragmentation and gene expression46. Accordingly, our knowledge about breast cancer has also been extended by integrative approaches, resulting in a more comprehensive understanding of the disease. The prognostic subclassification of breast cancers has for instance been refined and novel tumor-specific antigens identified. Integrated molecular data have even been shown to have a higher prognostic power than separate molecular levels in breast cancer. Thus, a combined analysis of the genetic (ctDNA), transcriptomic (miRNA) and metabolomic data levels in peripheral blood samples is therefore planned in the current project to maximize their joint biomarker potential in operable breast cancer.

Sample size calculations.

The investigators performed sample size calculations using SPSS Sample Power software. These calculations demonstrated that 125 high-risk patients should be sufficient to give a log rank test power of 80%, when testing the prognostic value of ctDNA/CTC detection before surgery. The sample size calculations were based on the qualified assumption that the average 5-year survival rate in the high-risk group is 90%, whereas it is 75% and 95% in the ctDNA/CTC positive and negative subgroups, respectively. Of the estimated 125 patients, 30 were assumed positive for ctDNA/CTC and 95 negative.

--------------------------------------------------------------------------------------------- ------------

B. Improve drug monitoring of tamoxifen in breast cancer

About 75% of all breast cancer belong to the luminal subtypes, which express the hormonal receptors Estrogen Receptor and Progesterone Receptor. These patients are treated with the anti estrogenic drugs; tamoxifen and/or aromatase inhibitors. The two most active tamoxifen metabolites are Z-4OHtam and Z-4OHNDtam (Z-endoxifen) have 30-100 times higher affinity for ER than tamoxifen. These metabolites ultimately constitute the blocking effect at the ER-level aiming to eradicate micro metastatic disease and are responsible for the improved survival following the establishment of this adjuvant systemic treatment. Direct measurement of these metabolites bypasses all disturbances from the diversity of CYP2D6 activity i.e. alternative metabolic pathways, adherence to the drug and inhibiting drug interactions. Our novel LC / MS-MS methodology takes into account all of the above-mentioned variables and provides a functional read-out report of the serum level of the active tamoxifen metabolites in the individual patient. This method can also distinguish between the inactive and active isomers of Endoxifen and 4-OHtam, which are the most active ER-blocking metabolites of tamoxifen. In collaboration with the Oslo Breast Cancer Research Group, the investigators have recently shown in a retrospective observational study that patients with low serum concentrations of Z-4OHtam < 3.26 nM or Z-Endoxifen < 9.00 nM (about 12% of all patients) have significant worse breast cancer specific and total survival than patients with serum concentrations above these thresholds(adjusted HR = 4.3; CI95 = 1.9-13.6) (red curves in Figure 4). Patients with a very high level of these metabolites (approximately 12% of patients) had no breast-specific endpoints Now, the investigators need to validate this discovery in independent patient materials. If validated, this will be of direct clinical benefit for 25% of ER + breast cancer patients planning adjuvant tamoxifen treatment. Such therapeutic drug monitoring (TDM) could identify risk patients with inadequate levels of active tamoxifen metabolites. This could lead to a dose increase or switch to an alternative endocrine treatment form in these patients. Patients with very high metabolite levels can continue on tamoxifen and do not need to switch to an aromatase inhibitor.

Thus, TDM may turn out to be a paradigmatic shift in endocrine tam treatment of ER-positive breast cancer. Importantly, the distance from "bench to bed" in this study is very short due to our recent findings, a feasible method and over 30 years of experience with tamoxifen in the clinical setting.

--------------------------------------------------------------------------------------------- ------------

C. Tumour microenvironment in breast cancer - adipose stromal immune infiltration and interaction with tumour at the growth zone

Background: Breast cancer is still the most common cancer type among women in the Western World, Norway included. Yearly, around 3500 new Norwegian breast cancer patients are diagnosed. During the last 50 years, the incidence of breast cancer in Norway has more than doubled. Interestingly, during the same time frame over-weight and obesity have increased alarmingly. Molecular subtyping of breast cancer tumours has improved the understanding of its intrinsic biology. This has opened up for a more personalized treatment approach and improved the outcome for breast cancer patients over the last 20 years .

In addition to the molecular subtyping, the diagnostics of breast cancer today consists of staging (through the TNM-classification system) and microscopic examination by a pathologist. There is increasing evidence for the prognostic value of evaluating tumour microenvironment (e.g. tumour infiltrating lymphocytes (TILs)) as a part of diagnostics in breast cancer, but as of yet this is not implemented in clinical practice. High numbers of TILs are strongly associated with good prognosis in patients with triple-negative breast cancer, but have not shown any such correlation in luminal breast cancer. Macrophages in breast adipose tissue differentiate into two distinct phenotypes, often classified as M1 and M2 with an increase of the latter in breast cancer. M1 attack and phagocytize cancer cells, while M2 have anti-inflammatory properties and have been shown to have tumour-promoting functions.

Infiltrating tumour associated macrophages (TAMs) in breast cancer are associated with higher proliferation, advanced histological grade and, in the case of triple-negative breast cancer, higher risk of distal metastasis. Dysfunctional adipose tissue with hypertrophy, hyperplasia, and hypoxia is a common feature in our obesogenic environment. In breast adipose tissue the infiltration of macrophages and the aggregation of these into crown like structures (CLS) is correlated with elevated levels of prostaglandin E2 (PGE2), which have been found to silence tumour suppressors and DNA repair genes .

The present project will increase our knowledge of the interaction between the primary breast tumour and the immune cell infiltration in the juxta-tumoral adipose tissue. Our findings will be correlated to molecular breast cancer subtypes, proximity to tumour, markers of proliferation, and ultimately to clinical outcome We reason that this knowledge will provide novel predictive targets for immune-modulating adjuvant cancer therapy. This PhD project has three interrelated work packages (WP) with the ultimate goal of identifying markers of inflammation in juxta-tumoral adipose tissue that are relevant in clinical assessment of adjuvant therapy of breast cancer. In WP 1 and 2, the phenotypical and genomic landscape of juxta-tumoral adipose tissue will be characterized, and in WP 3 we will investigate whether this knowledge can predict clinical outcome.

The project is based on a strong regional collaboration between the Haukeland University Hospital (HUH) and Stavanger University Hospital (SUH). For the current project, we have in a systematic manner collected adipose and tumour tissue from 30 consecutive breast cancer patients undergoing mastectomies at SUH. Surgical specimens of the breast are without delay forwarded to the pathologist who, under morphological control, sample relevant tissue from the invasive front and the core (0,5x1x1 cm) of the primary tumour. Controls from normal breast tissue (2000 mg) from the distant contralateral quadrant will also be collected and verified morphologically by the pathologist. The tissue is collected in duplicates pairs. One of the samples is immediately snap-frozen in liquid nitrogen for preservation, while the other half of each pair is preserved in formalin and fixated in paraffin (FFPE). In addition, three pairs of samples of adipose tissue (each of 1x1x1 cm) with increasing distance from the tumour border will make up a gradient of adipose tissue from the invasive front of the tumour area. The gradients collected are also immediately snap-frozen in liquid nitrogen and one half of each pair stored as FFPE. We have performed a pilot study confirming a sufficient RNA yield from the frozen tissue samples for RNA sequencing.

The tissue is collected as part of the general research biobank, Prospective Breast Cancer Biobank (PBCB) (REK# 2010/1957), an ongoing biobank-project.

The retrospective Stavanger cohort consist of all breast cancer patients diagnosed with first onset invasive breast cancer at the department of Pathology, Stavanger University Hospital, between 1993 - 2004. As SUH is the only hospital in the region this is a true population based biobank, already included are clinicopathological data like: grading, stage, TNM, treatment and proliferation markers (Ki-67, PPH3/MAI); last clinical follow-up took place in 2016.

We have four main hypotheses related to inflammation and immune cell infiltration in breast cancer:

1. The juxta-tumoral adipose tissue close to the tumour border shows an increased level of inflammation and immune cell infiltration.

2. The inflammation and immune cell infiltration attenuate with increased distance from the tumour border.

3. There are differences in the inflammation and immune cell infiltration between different subtypes of breast cancer (Luminal A, Luminal B and Basal like).

4. Increased inflammation and immune cell infiltration are correlated to adverse clinical outcome (i.e. relapse of disease).

4.1 STUDY DESIGN, METHODS AND ANALYSES

We have four main hypotheses related to inflammation and immune cell infiltration in breast cancer:

1. The juxta-tumoral adipose tissue close to the tumour border shows an increased level of inflammation and immune cell infiltration.

2. The inflammation and immune cell infiltration attenuate with increased distance from the tumour border.

3. There are differences in the inflammation and immune cell infiltration between different subtypes of breast cancer (Luminal A, Luminal B and Basal like).

4. Increased inflammation and immune cell infiltration are correlated to adverse clinical outcome (i.e. relapse of disease).

WP 1: MORPHOLOGIC AND GENOMIC MAPPING OF IMMUNE CELL INFILTRATION IN ADIPOSE TISSUE SURROUNDING BREAST CANCER

Background: Inflammation is an important facilitator for tumour development (16). Adipose tissue near breast cancer have been shown to have increased immune cell infiltration, but it is not yet known how this relates to molecular subtypes and/or proximity to tumour. We will use IHC and RNA-sequencing to evaluate the level of inflammation, infiltration of immune cells and immune gene expression signatures in the adipose tissue surrounding breast cancer.

Design: Prospective explorative study Patients, material and power calculation: Adipose tissue is collected from the prospectively collected mastectomy specimens. Tissue samples are collected from three different molecular subgroups (Luminal A, Luminal B and Basal like breast cancer), 10 in each group. Core biopsies are collected in a gradient with increasing distance from tumour border (fig. 1). The Norwegian Genomics Consortium (NGC) has confirmed that the sample sizes that are based on the selection in specific breast cancer subtypes are sufficient for statistical power.

Methods: Slides of formalin fixated tissue from tumour, invasive front and adipose tissue (fig. 1) will be stained using IHC and antibodies against immune cells (total immune cells, CD45; T cells, CD3, CD8, FoxP3; myeloid cells, CD68, HLA-DR). The slides will be evaluated by microscopy and the results will be correlated to molecular subgroups of breast cancer, localization in relation to tumour and proliferation markers (i.e. Ki67, PPH3 and MAI). In parallel, we will perform RNA-sequencing on frozen tissue collected from the same tissue samples as described above. Several analysis approaches will be used on the RNAseq-data. Except for analysing the data exploratory, specific analysis algorithms including deconvoluting immune cells (ABIS, CIBERSORTx) and more specifically looking at tumour immune signatures (TIP analysis), among others.

Expected outcome: We expect to find increased inflammation and immune cell infiltration in the juxta-tumoral adipose tissue and altered expression of inflammatory gene signatures that we can further investigate in WP 2.

WP 2: ASSESSMENT OF INFLAMMATORY REACTION IN JUXTA-TUMOURAL ADIPOSE TISSUE IN BREAST CANCER USING IMMUNOHISTOCHEMISTRY AND IMAGING MASS CYTOMETRY.

Background: Breast cancer cells interact with the surrounding tissue and causes inflammation and immune cell infiltration in the tumour and the adipose tissue. Imaging mass cytometry has shown that immune cell infiltration in tumours correlates with prognosis, but as of yet there are no studies examining whether the adipose tissue surrounding the tumour also show structural patterns of immune cell infiltration. Therefore, we will examine the adipose tissue surrounding breast cancer tumours using Hyperion™ Imaging Mass Cytometry.

Design: Prospective explorative study Patients, material and power calculation: Tissue samples are collected from 30 prospectively obtained mastectomy specimens with three different molecular subgroups (Luminal A, Luminal B and Basal like breast cancer), 10 in each group. Adipose tissue is collected in a gradient away from tumour and from the contralateral quadrant as a control sample (fig. 2). A sample size of 30 patients is deemed to be sufficient for an explorative study of this sort.

Methods: A Hyperion™ IMC panel of up to 50 unique antibodies will be developed based on an already existing and validated panel available at the UoB core facility which we have previously successfully used in other projects. The available panel will be further extended based on findings from WP 1 and first validated on "non-essential" breast tissue samples before used on the above-described study samples. Using Hyperion™ IMC we intend to analyse the levels of inflammatory reaction in the juxta-tumoural adipose tissue. IMC data will be analysed with histoCAT+, MAV and FlowJo (typically used for flow cytometry data analysis but also highly suitable for high-parameter imaging anslysis). Furthermore, we will be able to use gene expression data of known and novel factors, identified in WP 1, and visualize whether the expression is generalized or compartmentalized in the adipose tissue. Normal adipose breast tissue from the contralateral quadrant will be used as control. In order to examine the influence of BC on the inflammation of juxta-tumoural adipose tissue of the various molecular subtypes in more detail, we also intend to stain tissue from three tissue segments extending from the tumour in order to determine whether the levels depend on the distance from tumour or the molecular subtype (Fig. 1). The inflammation in the juxta-tumoural adipose tissue will also be assessed morphologically by using IHC. This will allow us to select regions of interest to investigate using IMC.

Expected outcome: We expect to find increased inflammation and immune cell infiltration in the adipose tissue near the tumour and to identify novel biomarkers in the adipose tissue allowing for investigation of the clinical significance in WP 3.

WP 3: INFLAMMATORY REACTION IN JUXTA-TUMOURAL ADIPOSE TISSUE IN BREAST CANCER AND CLINICAL OUTCOME.

Background:

Immunohistochemistry is a cornerstone in breast cancer diagnostics and serves as the main modality for identifying molecular subgroups of breast cancer and proliferation markers such as Ki-67. This is essential in determining what treatment regimen is best suited for the particular patient. Heterogeneity in the tumour tissue (both tumour cells and infiltrating immune cells) have been linked to difference in clinical outcome , but it is not yet known if the same applies for the adipose tissue surrounding the tumour. Thus, we will examine juxta-tumoral adipose tissue for inflammation and immune cell infiltration and correlate this with clinical outcome in breast cancer patients.

Design: Clinical observational study

Patients and Methods: A number of 3500 breast cancer patients are accrued in a biobank with a median follow-up time of 15 years. A large selection of tissue samples from these patients will be assessed with respect to immune cells in the adipose tissue and biomarkers discovered in WP 1 and 2 using IHC. The samples will be stratified into three groups according to molecular subgroups (i.e. Luminal A, Luminal B and Basal like breast cancer). The tissue slides will be analysed and the results correlated with clinical outcome from the patient journals (access to journals is already approved by REK). Using Cox proportional hazard ratio and log-rank test, we will evaluate the effect increased inflammation in adipose tissue has on patient survivability.

Power calculation: We estimated the test power of the difference in recurrence free survival between patients positive and negative for juxta tumoral inflammation as assessed by IHC and IMC. The effect measure is the hazard rate (HR) as calculated by the Cox regression method. For example: 1000 patients with a recurrence rate of 15% will have created at least 150 endpoints. If we have an inflammation rate of 20% we will have an 80% power to detect a survival difference with a hazard ratio (HR) of 1.80 or greater.

Expected outcome: This last work package is designed to give proof of principle that it is feasible to incorporate histological examination of juxta-tumoural adipose tissue as a part of breast cancer diagnostics. In the longer perspective we aim to identify new specific biomarkers of immune infiltration in adipose tissue that correlate to adverse outcome in breast cancer patients. Such biomarkers may serve as predictive tools for identifying patients eligible for adjuvant anti-inflammatory or immune-modulating treatment.

--------------------------------------------------------------------------------------------- ----

D.Monitor side effects, QoL, depression, fatigue and work life participation Background

Treatment of breast cancer employs all surgery, chemotherapy, radiation therapy and various targeted therapy options like antiestrogen and anti HER-2 therapy.These treatment options create bothersome side effects both in a short term and long term perspective. Importantly, more complex relationships between oncological therapy modalities and subjective health complaints in breast cancer patients are recently discovered. Fatigue, anxiety and depression scores are probably functional read outs of the total treatment burden. The anti cancer treatment has also a substantial impact on Quality of Life and social aspects for the cancer survivors which extends to work life participation. Sick leave (SL) five years after the breast cancer diagnosis is more dependent on social factors than on illness (2). Hence, breast cancer hits the women twice; first the disease burden then the sosio-economic worries. Breast cancer survivors have almost 3 x higher risk of receiving disability pension (DP) compared to cancer -free patients in a retrospective register based study (3). However, this study was based on mostly abandoned oncological treatment regimens.

Aims

The aim of this study is to identify risk factors for becoming a long term SL, WAA or DP receiver by an in-depth qualitative analysis and mixed-method approach by triangulating PROM data, clinical/biological data and the NAV-data. Use of NAV-data as endpoints will provide valuable information to clinicians and general practitioners who may optimize the follow-up also in regards to work-life participation. Hence, the investigators have a robust study design to identy fisk factors for high SL, WAA and DP rates. Ultimately, the investigators aim to lower the financial expenses breast cancer generates to society.

Study design / material & methods.

The study design is an observational study where the investigators will in the present study map the rates of and denty fisk factors for high SL, WAA and DP among breast cancer survivors with the up-to date treatment schedules. The investigators will follow a 4-step translational approach:

1. The investigators will first conduct an exploratory qualitative study using document analysis of health and work life policy in the Norwegian context, combined with empirical data from semi- structured interviews and focus group with 20 high risk patients) to be analyzed thematically. The theoretical framework for this qualitative project is derived from symbolic interactionism in which uncovering and understanding meaning in specific contexts is a goal. Since meaningful social and professional interactions are important for psychosocial rehabilitation of breast cancer individuals symbolic interactionism is an appropriate methodological framework. This qualitative approach will create the bases of the approach in the PROM based and NAV-database oriented studies below.

2. NAV- data

The investigators currently collaborate with vice director Anneline Christine Teigen and senior consultant Günter Olsborg in NAV-Rogaland and the Mikrodata section at Statistics Norway (SSB). The aim for this collaboration is gain more knowledge about the use of social security services among breast cancer patients together with the NAV- Rogaland experts and together with SSB use the FD-trygd database to obtain reliable prospective sociodemographic data on SLs (both short & long term), Work Assessment Allowance (WAA) (Arbeidsavklaringspenger) and DPs (both time limited

& life long) for each patients in the present PerMoBreCan study.

3. Patient reported outcome measure (PROM) data is obtained at baseline and thereafter yearly from all PerMoBreCan patients. The PROM data consists of 1.HRQoL instruments (EORTC QLQ-C30,EORTC QLQ-BR23 and FACT B), 2.Hospital Anxiety and Depression Scale, HAD, 3. Fatigue instruments Fatique Impact Scale, FIS; Fatugue Severity Scale (FSS) and VAS-fatigue, 4.Side effects questionnaires, 5. Joint- pain questionnaire, 6. The Mishel Uncertainty in Illness Scale, MUIS; and 7. Food habits questionnaire and the 8.ROMA III questionnaire of bowel complaints (IBD and IBS).

4. Biomarker identification.

All patients attending this study will be screened on circulating biomarkers releated to fatigue. These biomarkers are on the protein level, genetic expression level and also on the epigenetic level. Our long-established pleasant cooperation with prof. Roald Omdal at Stavanger University Hospital, who is a clinical immuologist and transaltional fatigue researcher, is committed to conduct these analyses in his research lab.