Activities and Progress 2016

In Kalland´s Group, one key activity has been the generation and characterization of a new experimental model of stepwise prostate tumorigenesis, comprising benign cells (EPT1), pre-malignant mesenchymal type cells (EPT2), tumorigenic (EPT3-N04/EPT3-PT1) and metastatic (EPT3-M1) cells in mice, with different phenotypes and behavior, and applied in studies of transcriptional reprogramming and drug discovery. This is the first experimental prostate cancer model that derived tumorigenic prostate cells by using physiological selection pressure only. Epithelial-to-mesenchymal transition (EMT) was an early feature of the model, and tumor initiating cell (TIC) subpopulations have been characterized among the tumorigenic cells.

Work on the experimental tumorigenesis model has resulted in increased insight into the potential of gene expression reprogramming as a source of cell heterogeneity.

Subpopulations of TICs show activation of the WNT pathway and an autocrine IL6/STAT3 feedback loop associated with tumorigenesis. Recently, this model has been used in a drug discovery and development program, resulting in 5 WNT/β-catenin inhibitor candidates (with patents pending).

Kalland´s team, in collaboration with Gjertsen and others, have initiated an ongoing phase I clinical trial of cryoimmunotherapy for advanced prostate cancer, including an intensive biomarker program with liquid biopsy projects and collaboration with Dr. Pantel, tissue-based biomarker analysis, and mass cytometry profiling.

Within CCBIO, Kalland is collaborating with Gjertsen, Lorens and Akslen.

Gullberg´s Group has studied how different connective tissue cells interact with tumor cells and the extracellular matrix, a process that is similar to wound healing and scarring. In particular, integrins are important regulators of these processes. The group has established a model using A549 lung cancer cells to study tumor-stromal interactions, recently reporting that integrin α11 from fibroblasts is important to stimulate tumor cells to secrete soluble factors influencing immune cell recruitment and tumor growth. Also, integrin α11 was found to be important for stromal stiffness and tumor spread in non-small cell lung cancer.

Gullberg and co-workers have found that integrin α11β1 is the receptor on fibroblasts that mediates contraction of wounds, and that this mechanism is mediated via c-jun N-terminal kinase (JNK). This is an important step forward in the efforts to delineate the molecular mechanism of cell-collagen interactions.

The team is now also working to generate and characterize novel integrin α11 blocking antibodies with potential use as robust biomarkers in tissue analyses, and work is ongoing to establish and characterize an α11 promoter-Cre mouse strain.

Within CCBIO, Gullberg is collaborating with Reed, Akslen, Johannessen, and Gjertsen.

In collaboration with Gullberg and others, Reed´s Group has a focus on the dynamic extracellular matrix and on interstitial fluid pressure (Pif) in tumors and how this can be modified. This is relevant for imaging technologies and for distribution of drugs. Tumors have an elevated Pif that acts as a functional barrier towards transcapillary fluid flux that can block the distribution of cytostatic anti-cancer agents.

The group has reported that integrin α11 has an influence on the interstitial pressure, and subsequently on tumor growth patterns in mice lacking this integrin (breast and prostate cancer models). The results point to important biophysical features of the tumor microenvironment and their importance for cancer progress.

Reed´s group has also been working on the use of improved imaging techniques (DCEMRI) in determining tumor vasculature and transcapillary transport in preclinical models. Lately, the team is studying the interaction between the genetic background of integrins (using various mouse strains) and growth of breast cancer models. Studies on the effect of hyperbaric oxygen on experimental tumor growth are ongoing.

Within CCBIO, Reed is collaborating with Gullberg and Akslen.

on basic and translational aspects of oral cancer with focus on cancer-host interactions, particularly between the surface epithelium and the underlying connective tissue. The team has established novel in vitro assays of human tissue-based 3D cell culture models of normal mucosa and oral cancer tissue, and a new rodent oral cancer model.

In collaboration with Gullberg´s group, integrin α11 has been identified as a key regulator in stromaendothelial cross talk. Johannessen´s team now aims to develop a diagnostic and prognostic biomarker profile that can stratify patients with oral premalignant and malignant lesions for a more individualized therapy, and they have reported a “malignancy index” signature which is now being validated.

The group identified two distinct fibroblast (CAF) subgroups in oral squamous cell carcinoma based on transcriptome analysis of primary cells from human cancers: a CAF subgroup with a gene expression profile closer to normal fibroblasts, having a more motile phenotype and deeper carcinoma cell invasion; and a CAF subgroup with a more divergent gene expression profile, having a more stationary phenotype, with less tumor formation and invasion.This study points to functional heterogeneity within tumor associated fibroblasts.

Within CCBIO, Johannessen collaborates with Gullberg, as well as Junior Investigator Costea.

The Lorens Group works on cellular plasticity, such as stem cell differentiation and transdifferentiation, a critical prerequisite for adult tissue homeostasis and injury repair. Using comparative functional approaches, the team is investigating the regulation of tumor cell plasticity and maintenance of normal adult stem and progenitor cells.

Recent results highlight the Axl receptor tyrosine kinase as a key regulator of both normal adult epithelial stem/progenitor cells and a determinant of carcinoma cell plasticity. The studies on Axl signaling have provided new insights into the regulation of tumor phenotypic heterogeneity and form the basis for the recent clinical translation of novel Axl inhibitors (e.g. BGB324). Importantly, it was recently also reported that Axl-activity could be blocked by low-dose warfarin.

The group continues to study how microenvironmental factors and immune cell challenge illicit tumor cell phenotypic plasticity that engenders acquired resistance to both chemo- and immunotherapeutic agents. On this background, a national investigator-sponsored trial on metastatic melanoma with BGB324 anti-Axl therapy and anti-PD-1 (PI Straume) has been launched, with an intensive biomarkerprogram included.

Within CCBIO, Lorens is collaborating with Gjertsen, Straume and Akslen.

Akslen´s Group has focused on the use of biomarkers for improved molecular classification and grading of malignant tumors, as a better guide for targeted treatment. Studies of human tumor samples (primary and metastatic lesions) are combined with experimental cell and animal models to improve translation. The team is concentrating on studies of the tumor microenvironment, especially tumor-vascular interactions and angiogenesis markers, and the use of precise indicators for tumor proliferation with clinical applications.

The team has reported novel tissue-based angiogenesis biomarkers. As examples, microvessel proliferation was studied in several human tumor types and provides better prognostic information than vascular density. This marker proved valid also in xenograft models of breast cancer. Further, a 32-gene RNA-based expression signature for microvessel proliferation gives prognostic information in endometrial and breast cancer, and was linked to 6p21 amplification.

An 18-gene s ignature was identified for vascular invasion by tumor cells, pointing towards novel mechanisms involved in early metastatic spread, and this signature was strongly prognostic in breast cancer. Studies on the progenitor cell marker Nestin indicated an ability to identify BRCA-1 related breast cancer and the aggressive basal-like phenotype.

During the last year, the team has reported on tumor proliferation markers in breast cancers and how these change from primary tumors to metastases. Data on proliferation is currently used in definition of Luminal B breast cancer and treatment decisions. Whereas 15% of the cases changed from low (primary tumors) to high proliferation (metastases), treatment consequences are currently not clear in guidelines but are being discussed.

Further, a paper on improved definitions of extra-nodal growth in lymph node metastases of breast cancer has led to new national guidelines.

In a collaborative study with Dr. Watnick (Boston), the importance of prosaposin (PSAP) and thrombospondin-1 (TSP-1) for ovarian cancer progression was reported (Wang, 2016), expanding on previous findings from this collaboration.

Within CCBIO, Akslen is collaborating with Kalland, Gullberg, Reed, Lorens, Straume and Gjertsen, as well as Junior Investigators Wik and Krakstad.

Gjertsen´s Group, supported by the Early Phase Clinical Trial Unit at Haukeland University Hospital, has been the initial center for a phase I trial (BGBC003; clinicaltrials.gov) in AML using the novel anti-Axl drug BGB324 (per oral formulation) in collaboration with BerGenBio. The trial is now also recruiting in Houston (Texas) and Germany.

In parallel, more focused small trials in chronic myeloid leukemia has been performed and completed in collaboration with the Nordic CML Study Group, providing a unique material for proof of principle testing of how to monitor signaling in cancer cells as biomarkers for risk and therapy response. Importantly, new instrumentation funded by Bergen Research Foundation in 2015, a mass cytometer, allows multiparametric analysis of single tumor cells.

Through CCBIO and the Helse Bergen Clinical Trials Unit, the team will address clonal evolution in AML through mass cytometric analysis. The team has performed extensive studies of signaling patterns in CML cases. There is a need for more direct biomarker analyses for early kinase inhibitor therapy, based on increasing reports of adverse events. The group has demonstrated that the drug target can be monitored in the actual cancer cells, and suggests that cellular signal systems involved in signaling of BCRABL outline the long time response. This single cell analysis of cellular signaling fit with the blood levels of the drug, and is likely to be a preferred method for future precision medicine with signaling targeted therapy.

In the phase I trial with BGB324, the concepts of single-cell biomarker profiling are tested. Different analysis methods and read-out panels have been developed during 2016. The possibility to employ single cell biomarker technology in drug development is very promising. The strategy is also to move these concepts beyond blood cancers to metastatic solid cancers, based on strong collaborations within CCBIO, and several trials are now prepared.

In collaboration with Kalland, a phase I clinical trial of cryoimmunotherapy has been initiated for patients with metastatic castration resistant prostate cancer. The trial is based upon a dendritic cell based immunotherapy protocol in collaboration with the Haakon Ragde Foundation in Seattle (USA). The associated biobank is used for development of advanced immune-monitoring and circulating tumor cell enumeration as well as organoid cell culture isolation.

Within CCBIO, Gjertsen is collaborating with Lorens, Kalland, Straume, Akslen and Gullberg, as well as Associate PIs Bjørge and McCormack.

Straume´s Group is focusing on the identification of predictive biomarkers for therapy response in academic trials of patients with metastatic melanoma and kidney cancer.

In melanoma, previous results of a clinical trial with the anti-VEGF antibody bevacizumab documented that ~30 % of the patients experienced clinical benefit of the treatment. Based on a screen of multiple candidate markers in tissues of primary tumors and metastases, as well as serum markers, HSP27 tissue expression in metastatic lesions was able to predict therapy response. The team continues to screen for serum-based biomarkers of therapy response, to antiangiogenesis treatment (bevacizumab) and immunotherapy (ipilimumab).

Due to the recently reported role of Axl in immune evasion, a national investigatorsponsored trial on metastatic melanoma with BGB324 anti-Axl therapy and anti-PD-1, directed by Straume, has been launched, with an intensive biomarker-program included. Focus will be on predictive markers of response. In a collaboration with national centers, 150 patients with metastatic melanoma were treated with ipilimumab, a CTLA-4 antibody (phase IV clinical trial). Blood and tissue samples are being studied to identify predictive markers of response. In a trial series of 45 cases with metastatic clear cell renal carcinoma treated by VEGF-inhibition (sunitinib), the team is now working on a set of candidate biomarkers for their predictive value.

In a study that could be of major clinical significance, the Straume Group reported that breast cancer recurrence can be influenced by the timing of surgery. The finding might lead to increased awareness about the role of surgery in high risk patients and increased use of immediate surgery.

Within CCBIO, Straume collaborates with Lorens, Gjertsen and Akslen.

The Bergen Gynecologic Cancer Group (previously led by Salvesen) has made significant efforts in biomarker discovery and validation in gynecologic cancers, at the genetic and protein levels, with special focus on endometrial cancer and hormone receptor regulation and impact. For both estrogen receptor (ER) and progesterone receptor (PR), loss of expression is linked to aggressive disease and poor survival.

ATAD2, a cofactor for ER, was strongly linked to aggressive signatures, while FOXA1, another ER cofactor, showed an unexpected switch in expression from primary tumors to metastatic lesions. Loss of both ER and PR predicted lymph node metastases, and this finding led to determination of ER/PR status for endometrial cancer as a stratifier for lymphadenectomy in a phase 4 implementation trial (MoMaTEC2). Stathmin expression was found to predict clinical response to taxane treatment in endometrial cancer, both in preclinical and clinical settings. This finding has been taken to a phase 2 integrated biomarker trial for paclitaxel treatment in endometrial and ovarian cancer (MoMaTEC2).

The team continues studies on genetic alterations in gynecologic cancer, in collaboration with the Broad Institute (Boston). In particular, data from an extensive molecular profiling of genomic alterations in cervical carcinomas were presented (published in Nature). Similar studies on endometrial cancer, also in collaboration with several other teams, are ongoing. Further, the team has studied different imaging modalities in preclinical and clinical settings in relation to angiogenesis and clinical characteristics. The findings are relevant for preoperative patient stratification.

In summary, several efforts and initiatives within CCBIO, with increased internal and external collaboration, are now up and running. The projects are spanning from matrix biology and plasticity programs, through discovery and validation of biomarkers and signatures, to clinical trials with targeted biomarker panels using liquid biopsy and single cell analysis. In this context, the programs on ethics and economics of biomarker based therapy, are also expanding and are being integrated in the recently established clinical trials.