The immune system´s role in the fight against cancer represents a double edged sword: it protects from tumor development and metastasis (immune surveillance) and also promotes tumor formation by inhibiting anti-tumor immune responses, e.g. via influences on the tumor microenvironment.
Our main research interest is the comprehensive analysis of tumor-host interactions and especially the interplay between the tumor and cells of the immune system. This includes the assessment of infiltrating immune cells and Cytokine/Chemokine levels at the tumor or metastatic site as well as the analysis of local and peripheral humoral and cellular immune responses. Technological advances now allow an unparalleled insight into the makeup of the immunological pathomechanisms in serum samples and archival cancer specimens. The precise quantification and localization of infiltrating immune cell subtypes is now made possible by an automated staining, scanning and analysis high-throughput process. Detailed insights into the cytokine and chemokine milieu of the tumor and surrounding tissue, the precise localization of involved immune cell populations are also possible with this approach. These immunologic phenomena can now be put into a comprehensive context with cellular immune responses in the peripheral blood and antibody responses, measured by multiplex assays, allowing to discern clinically relevant parameters. The overall goal is to determine, which parameters are meaningful to develop novel individualized therapeutic strategies. Of further interest is the prognostic and predictive relevance of these “immune parameters” on patient survival and response rates to ongoing treatments.
Comprehensive analysis of tumor host interactions in the immunological tumor microenvironment and the periphery
Localization and density of immune cells in human metastatic colorectal cancer
The prognostic role of tumor infiltrating lymphocytes has been shown in many tumor entities. A high density of infiltrating immune cells is apparently associated with a more favorable clinical course (Galon et al. 2006, Zhang et al. 2003). The molecular mechanism underlying the clinical impact and the causes of the observed differences in the immunological tumor micro-milieu are not clearly understood. The detailed differential analysis of the immune cells and cytokines involved will provide a direct way to aid clinical decisions and can be used as a therapeutic tool. Therefore, the dedicated analysis of the tumor compartments (stroma, epithelium, invasion front, surrounding tissue) with respect to localization of immune cells, cytokine profiling and expression profiling in the light of the clinical course will yield a better image of the complex relations. Early results on the association between immune cell densities and response to chemotherapy indicate a clear potential for direct clinical applications. Immune cell infiltrates at the invasive margin of CRC liver metastases were analyzed on a high-resolution automated microscopy system with complete tissue sections (Figure 1: Workflow of standardized tissue evaluation.). Densities of immune cells at the invasive margin of liver metastases allowed the prediction of response to chemotherapy and showed a statistically significant association with time to progression (Halama et al., 2011). These findings emphasize that the impact of the local immune response on the clinical course is a general phenomenon, extending from the primary tumor to metastatic lesions. To further broaden the investigated immune cell subsets, an unambiguous marker for NK cells was established as well as marker for other immune cell subsets (mast cells, macrophages etc.). The obtained density data is based on 400 primary colorectal cancer samples and allows the direct evaluation of clinical relevance.
Comprehensive characterization of the tumor immuno-microenvironment in liver metastases and primary colorectal cancer
Detailed insights into the cytokine and chemokine milieu of the tumor and surrounding tissue in combination with the precise localization of involved immune cell populations will yield a detailed understanding of the immunologic pathomechanisms. A combination of high-throughput large scale microscopic quantification with cytokine & chemokine profiling on different morphological compartments of the tumor (colorectal cancer liver metastases and primary tumors) are performed (Figure 2: Workflow for high-throughput large scale microscopic quantification with cytokine & chemokine profiling on different morphological compartments of the tumor.). Differences with respect to clinically relevant immune cell densities can be used for stratification, giving a differential view on the underlying immunological mechanisms. The investigation has produced high-resolution maps of cytrokine & chemokine „niches“ and has identified compartments with unique immunological signatures. Novel laser capture microdissection procedures allow contamination-free single cell separation and therefore allow for the first time a highly specific analysis. Not only analyses of proteins but also on mRNA level are conducted. The above mentioned techniques are all used on clinical samples with knowledge of the underlying clinical course to identify potential predictive patterns or therapeutic interventions. These insights already led to a successful phase I clinical trial (NCT00001) that can be seen as the blueprint for individualized immunological approaches with new therapeutic options.
Clinical relevance of spontaneous humoral immune responses against tumor associated antigens
Autoantibodies, a hallmark of both autoimmunity and cancer, represent an easily accessible surrogate for measuring adaptive immune responses to cancer. We are applying multiplex assays to study the prognostic impact of antibody responses against tumor-associated antigens (TAAs) in different tumor entities, i.e. malignant melanoma.
Malignant melanoma is despite improved detection in early stages still a deadly disease and metastatic relapse is still a common problem with an often fatal outcome. The appearance of autoantibodies in melanoma patients treated with interferon has been associated with improved recurrence-free survival and overall survival. Spontaneous humoral immune responses against tumor-associated antigens (TAA) have been described for various tumor entities. In contrast to responses to TAAs, analyses of autoimmune responses, i.e. against normal tissue structures, were already performed and seem to indicate a prognostic relevance for malignant melanoma (Gogas et al. 2006). This finding however is not without criticism and the role of humoral immune responses remains controversial (Bouwhuis et al. 2009). Antibodies directed against TAAs are however mainly restricted to cancer patients and therefore allow a more specific analysis. A large-scale analysis of spontaneous antibody responses against tumor-associated antigens has not been conducted so far. To study the prognostic significance of spontaneous antibody responses in more detail and to determine the relevance of antigens as targets for immunotherapy, we developed a multiplex assay (Waterboer et al. 2005, Waterboer et al. 2006) for the systematic analysis of serum antibodies against a broad number of TAA (Figure 3: Principle of the multiplex bead based serological assay.). Serum samples of malignant melanoma patients (stage I to IV) and controls were analyzed for spontaneous antibody responses against 29 TAAs. Especially for early stage melanoma patients´ antibody responses against certain antigens were prognostically significant and were associated with shorter time to progession and overall survival (Zörnig and Halama et al., 2014). We now have a panel of more than 50 antigens and are analyzing serum antibody responses in different cancer entities.
KFO227 Subproject 8: Clinical significance of the local and systemic immunologic micro-environment in patients with colorectal cancer liver metastases
It has been recently shown that the clinical outcome of patients with colorectal cancer (CRC) is greatly influenced by immune responses at the tumor site. T cell infiltrates in primary CRC, especially memory T cells, correlate with a more favorable prognosis whereas patients with CRC that have low densities of infiltrating T cells have a poor prognosis, independent of the tumor stage. This applies to the metastatic disease situation. The link between the local immunological microenvironment and the systemic environment has not been systematically analyzed. In contrast to cellular anti-tumor immune responses it is especially unknown to date if tumor associated antigen (TAA)-specific antibody responses mediate an anti-tumor effect in vivo or if they just represent an epiphenomenon of ongoing effective - or ineffective T cell responses. To better understand the immunological effector mechanisms of anti-tumor responses we will analyze the cytokine and chemokine profile in CRC liver metastases as well as immune cell densities and distributions and correlate those findings with cytokine profiles and TAA-specific antibody responses in patients´ sera. Whole-slide imaging of histological sections coupled with automatic image processing, microdissection and multiplex protein quantification technologies have been established, as well as multiplex serology.
Linking these detailed immunological profiles with clinical data allows identifying key biomarkers and factors in the intricate interplay between these compartments. Identification of prognostic and predictive serum markers for treatment decisions will improve patient stratification and subsequent clinical routine.
HIPO H034 Project: Bringing together genomics and immunomics: colorectal cancer liver metastases, their mutational repertoire and immune responses against mutated proteins
Metastatic colorectal cancer is a devastating disease, especially if resection of the metastatic burden is technically not feasible. Understanding the immunological changes from primary tumor to metastatic lesion as well as the targets of the „native” T cells are promising pathways to improve patient care. In this respect, this proposal aims to bring together the data on the whole exome sequencing of twenty colorectal cancer liver metastases, the T cell receptor clonality and the corresponding data on the immune microenvironment, especially on the T cell status. We could identify a broad dynamic range of T cell infiltration at the invasive margin of colorectal cancer liver metastases as well as in other cancer entities. To better understand, what these effector T cells might possibly recognize within the local microenvironment, the joint analysis of the mutations present within the tumor cells and the status of the immunological microenvironment will provide further details about this intricate network. In further steps, the identified frequently mutated proteins (frequent mutations) will be used to generate these altered proteins ex vivo and subsequently test these proteins in T cell assays for recognition by extracted T cells from patient samples.
Our bioinformatic research acitivities focus on the assessment of intratumoral spatial heterogeneity of Ki-67 biomarker expression in neoplasms using complex systems approaches. It has been recognized for some time that heterogeneity of morphology and proliferative rate is common in many neoplasms. A robust method is sought in order to provide the clinician with a computational toolset for assessing the spatial heterogeneity of proliferation on Ki-67 biomarker expression.
Another interest is the visualization of the co-localization of T lymphocytes and macrophages in brightfield immunostains in the framework of Clifford algebras. Clifford algebras appear to be a powerful framework for the geometric transformation of colour information. A robust method is sought in order to visualize the co-localization of lymphocyte marker stains in serial sections after deformable registration.
Characterization of the breast cancer antigen NY-BR-1 and its clinical significance for cancer immunotherapy
Among all cancer cases, breast cancer is one of the most common malignancies among women and the leading cause of death in the developed world (American Cancer Society, 2008). Thus, there is clear need to define new target molecules for the development of novel therapeutic approaches. Antigen-specific immunotherapeutic strategies in cancer, both cell-mediated and antibody based, rely on the identification of target antigens expressed in cancer cells. We previously used the SEREX technique to isolate breast cancer associated antigens and identified a new breast differentiation antigen designated as NY-BR-1. Expression analysis by RT-PCR, quantitative PCR and cDNA microarrays revealed that NY-BR-1 is expressed in normal breast tissue, testis and occasionally in prostate tissue, but not in other normal tissues. Importantly, NY-BR-1 is expressed in >70% of primary breast cancers and expression is maintained in breast cancer metastases (Seil et al. 2007). NY-BR-1´s capability of eliciting spontaneous humoral immune responses was observed in approximately 7% of breast cancer patients, who have detectable levels of anti-NY-BR-1 serum antibodies (Seil et al., 2007). Two NY-BR-1 derived CD8 T cell epitopes have been identified by analyzing CD8 T lymphocytes of HLA-A2 positive breast cancer patients with NY-BR-1 expressing tumors (Jäger et al. 2005). In collaboration with the group of Prof. S. Eichmüller (DKFZ Heidelberg), novel MHC class II epitopes were identified and validated. Due to its tissue-restricted expression pattern and its immunogenicity, NY-BR-1 is a promising target antigen for clinical immunotherapeutic strategies, such as vaccines and T cell based approaches.
Confocal microscopy, FACS analysis and cell surface protein biotinylation assays revealed that NY-BR-1 is expressed in the cytoplasm and at the cell surface of transfected cell lines where it is accessible for monoclonal antibodies in cell culture systems (Seil et al., 2007). These findings qualify NY-BR-1 also as a highly attractive target antigen for antibody-based therapy. The main objectives of this project are to explore the biological function of NY-BR-1 and to evaluate the potential of NY-BR-1 as a target antigen for antibody-based strategies in breast cancer patients. After extensive characterization of NY-BR-1 the translation into immunotherapy with specific CAR constructs (chimeric antigen receptor) is now performed.
Cancer stem cells and chimeric antigen receptor therapy approaches
Our understanding of tumor initiation and maintenance has undergone a dramatic change in the last decade. It has been discovered, that the broad heterogeneity of a malignant tissue is organized in a hierarchical manner on which the so called cancer stem cells (CSCs) reside at the top level. These master regulator cells are known to be notoriously resistant against classical treatments like chemo- or radiotherapy and are thought to be the main driver of metastasis in distant organs. Since they have the unique ability to remain silent and in low number within the circulation, the recurrence of a tumor even years after therapy is also a CSC attribute. We have shown in the last years that immunotherapy using engineered T cells is a powerful tool to eradicate specifically the CSC subfraction which in turn leads to the shrinkage of the remaining bulk tumor tissue and a final cure in mouse model experiments. The engineered T cells in our settings harbor a so called chimeric antigen receptor (CAR) which combines the nearly unlimited binding capacities of a monoclonal antibody with the signaling features of a T cell into one single molecule. This allows us to engraft T cells with a new specificity regardless of the HLA status of the targeted tissue. Our research within the department in the next years will be focused on the improvement of an anti-CSC CAR treatment with respect to specificity and safety. To reach this goal, we aim on three subjects:
- Development of a bi-specific CAR, that activates the T cell only after binding to a pattern of two antigens
- Identification of new targets on human CSCs and their exploitation for immunotherapy
- Intrinsic labelling of CSCs in vivo by non-pathogenic viral vectors