- Email: [email protected]
33: Targeting cancer-related inflammation
S9
EACR-23 Oral Presentations, Sunday 6 July 2014 / European Journal of Cancer 50, Suppl. 5 (2014) S3–S11 the last decade, there have been numerous reports about their function and polarization. It has been suggested that TAM follow a transcriptional program termed M2 polarization that was initially attributed to IL4 stimulation of macrophages. However, previous reports were often not comprehensive and sometimes they were even contradictory. We have started to use genomewide transcriptome analysis of TAM derived from different murine syngeneic and spontaneous tumor models. Already our initial analyses clearly revealed that the previous interpretation of TAM activation being related to the socalled M2 polarization program is not apparent on the genome-wide level. In contrast, we identify completely new biological processes that highlight TAM reprogramming. Moreover, there seems to be a rather tumor type specific alteration of TAM transcription and function. Since targeting the tumor microenvironment therapeutically is a very attractive alternative or complementary approach in cancer therapy, a better understanding of TAM transcription and function on a global level will be a prerequisite for such endeavors. No conflict of interest.
tumor angiogenesis and tumor macrophage infiltrate as well as reducing paraneoplastic thrombocytosis. We are now investigating rational combinations of anti-IL-6 antibodies with other treatments. The chemokine receptor CCR4 is abnormally expressed on malignant cells as well as on leukocytes in human renal cell cancer, RCC. Patient plasma levels of the CCR4 ligands and their ratio reflect this abnormality and have prognostic significance. Both a small molecule CCR4 inhibitor and an anti-CCR4 antibody had anti-tumor activity in an RCC model with a novel mechanism of action on tumor-associated macrophages. Conflict of interest: Corporate-sponsored research: Affitech AS.
32 Proffered Paper: Vessel co-option in colorectal cancer liver metastases mediates resistance to conventional anti-angiogenic therapy
34 Progress toward a biomarker discovery-to-development pipeline in clinical proteomics
1
1
2
1
3
3
V. Thompson , S. Frentzas , P. Vermeulen , S. Foo , Z. Eltahir , G. Brown , D. Cunningham3 , A.R. Reynolds1 . 1 The Institute of Cancer Research, Tumour Biology Team, London, United Kingdom, 2 GZA Hospitals St. Augustinus, Translational Cancer Research Unit, Antwerp, Belgium, 3 The Royal Marsden Hospital, Oncology, London, United Kingdom Introduction: The growth of metastatic tumors is considered to require sprouting angiogenesis and this hypothesis has driven the development and clinical application of vascular endothelial growth factor (VEGF) targeted agents, such as bevacizumab. Responses to such agents in the clinic, including in colorectal cancer (CRC) patients, are variable. The biology behind the variable responses are not yet understood and we have no way of selecting patients who may benefit. It is now evident that tumors can also utilize a number of ‘VEGF-independent’ angiogenesis mechanisms, the existence of which may limit the efficacy of conventional therapy. One such mechanism is ‘vessel co-option’, whereby tumors hijack existing local blood vessels as they invade into host tissue. Published studies reported this mechanism to occur in ~28−30% of CRC liver metastases. The purpose of this study is: (1) examine whether vessel co-option is associated with a lack of response to Bevacizumab in advanced colorectal cancer patients, (2) identify whether inhibition of pathways associated with vessel co-option sensitizes to antiangiogenic therapy in vivo. Materials and Methods: Here we present data from a retrospective study of patients with CRC liver metatases that were treated with bevacizumab whereby responses to therapy were measured by radiological morphology (contrast enhanced CT) or pathology (cell viability). The presence of vascular co-option was quantified by histopathological analysis of resected specimens. Further to this, we modeled this growth pattern in vivo by intrahepatic injection of HT29luc colorectal cancer cells into CB17-SCID mice. Candidate proteins believed to be involved were stable knocked down using shRNA lentiviral technology. Results: We demonstrate, in this patient cohort, the presence of vessel cooption in liver metastases was strongly associated with lack of response to bevacizumab. We also show that a similar result is observed in a mouse model of colorectal cancer liver metastasis. Interestingly, when we reduce actin nucleation activity, via stable shRNA targeted knockdown of ArpC3 (p21ARC), in HT29 colorectal cancer cells, vessel co-option is compromised and sensitivity to VEGF-targeted therapy is improved in vivo. Conclusion: This data demonstrate a potential role for vessel co-option as a negative predictive biomarker for anti-angiogenic therapy and may also have consequences for the development of novel therapies for targeting tumor angiogenesis. No conflict of interest. 33 Targeting cancer-related inflammation F. Balkwill1 . 1 Barts and The London School of Medicine and Dentistry, London, United Kingdom Cancers are driven by complex networks of cytokines and chemokines that enable communication between the malignant cells and the many other cell types of the tumor microenvironment. Pre-clinical experiments have shown that targeting these networks with therapeutic antibodies or small molecule inhibitors can decrease tumor growth and spread. The challenge now is to translate these promising results to clinical trial. This talk will focus on the cytokine IL-6 and the chemokine receptor CCR4. IL-6 is a key regulator of an inflammatory cytokine network of highgrade serous ovarian cancer, HGSC. Clinical, pre-clinical and in silico experiments showed that antibodies to IL-6 can have multiple actions within the tumor microenvironment including reductions in cytokine production,
Sunday 6 July 2014
14:00−15:45
Symposium
Cancer Biomarkers
S.A. Carr1 . 1 Broad Institute of MIT and Harvard, Proteomics, Cambridge, USA Better biomarkers are urgently needed to improve diagnosis, guide molecularly targeted therapy, and monitor activity and therapeutic response across a wide spectrum of disease. Proteomics methods based on mass spectrometry hold special promise for the discovery of novel biomarkers that might form the foundation for new clinical blood tests, but to date little has been delivered. Proteomics-based biomarker discovery has been severely hampered by inadequate number and quality of samples, poor study design, and technological approaches lacking sufficient sensitivity, quantitative precision, and capacity to analyze statistically relevant numbers of samples. Another key problem has been the lack of robust quantitative methods to credential candidate protein biomarkers in larger numbers of patient samples prior to clinical evaluation. This problem extends into biology where the lack of highly specific affinity reagents for novel candidate proteins and modified peptides with sufficient sensitivity, specificity, reproducibility and throughput has significantly hampered our ability to understand dynamic, protein-based biological processes. In our biomarker studies we are addressing both of these serious barriers. In the discovery phase, we are employing multiplexed, quantitative MS technologies that enable analysis of larger numbers of patient samples with improved precision leading to better-qualified candidates. For verification of candidate biomarkers, we are developing targeted mass spectrometry-based technologies to screen and quantify low abundance proteins and modified peptides in a variety of biological contexts including human tissue and plasma. These assays are highly multiplexed, sensitive and specific and do not suffer from interferences that plague immunoassays. Targeted MSbased approaches are helping to usher in a new era of quantitative biology where proteins of interest and their modifications can be robustly measured in any biological context. We are applying these quantitative approaches in the context of a generalizable proteomics-based discovery-through-verification pipeline to identify early biomarkers of cardiovascular injury, breast and ovarian cancer. The same technologies are also being brought to bear to understand response and resistance to therapy. These studies are beginning to demonstrate that modern proteomic technologies when coherently integrated can yield novel, credentialed protein and peptide biomarker candidates of sufficient merit to warrant real clinical evaluation and to shed light on biological function. No conflict of interest. 35 Personalised proteotypes and their association with disease R. Aebersold1 . 1 ETH Zurich, Institute of Molecular Systems Biology, Zurich, ¨ Switzerland Powerful genomic technologies are now capable of determine genetic variability at a genomic level and at unprecedented speed, accuracy and (low) cost, and Genome Wide Association Studies (GWAS) have indicated the (frequently weak) linkage of specific loci to disease phenotypes. The question how genetic variability is translated into phenotypes is fundamental to biology and of critical practical importance for medicine. In this presentation we will discuss emerging computational and quantitative proteomic technologies to relate genotypic variation to the proteotype, the quantitative state of the proteome of a specific (disease) tissue. We will show that precise and highly reproducible proteotype measurements can be now generated at high throughput by the targeted proteomic methods selected reaction monitoring (SRM) and, at higher throughput, by SWATH-MS. We will discuss the principles of these mass spectrometric methods, discuss the computational challenged they pose for data analysis and demonstrate with selected applications their ability of proteotype data to
EACR-23 Oral Presentations, Sunday 6 July 2014 / European Journal of Cancer 50, Suppl. 5 (2014) S3–S11 the last decade, there have been numerous reports about their function and polarization. It has been suggested that TAM follow a transcriptional program termed M2 polarization that was initially attributed to IL4 stimulation of macrophages. However, previous reports were often not comprehensive and sometimes they were even contradictory. We have started to use genomewide transcriptome analysis of TAM derived from different murine syngeneic and spontaneous tumor models. Already our initial analyses clearly revealed that the previous interpretation of TAM activation being related to the socalled M2 polarization program is not apparent on the genome-wide level. In contrast, we identify completely new biological processes that highlight TAM reprogramming. Moreover, there seems to be a rather tumor type specific alteration of TAM transcription and function. Since targeting the tumor microenvironment therapeutically is a very attractive alternative or complementary approach in cancer therapy, a better understanding of TAM transcription and function on a global level will be a prerequisite for such endeavors. No conflict of interest.
tumor angiogenesis and tumor macrophage infiltrate as well as reducing paraneoplastic thrombocytosis. We are now investigating rational combinations of anti-IL-6 antibodies with other treatments. The chemokine receptor CCR4 is abnormally expressed on malignant cells as well as on leukocytes in human renal cell cancer, RCC. Patient plasma levels of the CCR4 ligands and their ratio reflect this abnormality and have prognostic significance. Both a small molecule CCR4 inhibitor and an anti-CCR4 antibody had anti-tumor activity in an RCC model with a novel mechanism of action on tumor-associated macrophages. Conflict of interest: Corporate-sponsored research: Affitech AS.
32 Proffered Paper: Vessel co-option in colorectal cancer liver metastases mediates resistance to conventional anti-angiogenic therapy
34 Progress toward a biomarker discovery-to-development pipeline in clinical proteomics
1
1
2
1
3
3
V. Thompson , S. Frentzas , P. Vermeulen , S. Foo , Z. Eltahir , G. Brown , D. Cunningham3 , A.R. Reynolds1 . 1 The Institute of Cancer Research, Tumour Biology Team, London, United Kingdom, 2 GZA Hospitals St. Augustinus, Translational Cancer Research Unit, Antwerp, Belgium, 3 The Royal Marsden Hospital, Oncology, London, United Kingdom Introduction: The growth of metastatic tumors is considered to require sprouting angiogenesis and this hypothesis has driven the development and clinical application of vascular endothelial growth factor (VEGF) targeted agents, such as bevacizumab. Responses to such agents in the clinic, including in colorectal cancer (CRC) patients, are variable. The biology behind the variable responses are not yet understood and we have no way of selecting patients who may benefit. It is now evident that tumors can also utilize a number of ‘VEGF-independent’ angiogenesis mechanisms, the existence of which may limit the efficacy of conventional therapy. One such mechanism is ‘vessel co-option’, whereby tumors hijack existing local blood vessels as they invade into host tissue. Published studies reported this mechanism to occur in ~28−30% of CRC liver metastases. The purpose of this study is: (1) examine whether vessel co-option is associated with a lack of response to Bevacizumab in advanced colorectal cancer patients, (2) identify whether inhibition of pathways associated with vessel co-option sensitizes to antiangiogenic therapy in vivo. Materials and Methods: Here we present data from a retrospective study of patients with CRC liver metatases that were treated with bevacizumab whereby responses to therapy were measured by radiological morphology (contrast enhanced CT) or pathology (cell viability). The presence of vascular co-option was quantified by histopathological analysis of resected specimens. Further to this, we modeled this growth pattern in vivo by intrahepatic injection of HT29luc colorectal cancer cells into CB17-SCID mice. Candidate proteins believed to be involved were stable knocked down using shRNA lentiviral technology. Results: We demonstrate, in this patient cohort, the presence of vessel cooption in liver metastases was strongly associated with lack of response to bevacizumab. We also show that a similar result is observed in a mouse model of colorectal cancer liver metastasis. Interestingly, when we reduce actin nucleation activity, via stable shRNA targeted knockdown of ArpC3 (p21ARC), in HT29 colorectal cancer cells, vessel co-option is compromised and sensitivity to VEGF-targeted therapy is improved in vivo. Conclusion: This data demonstrate a potential role for vessel co-option as a negative predictive biomarker for anti-angiogenic therapy and may also have consequences for the development of novel therapies for targeting tumor angiogenesis. No conflict of interest. 33 Targeting cancer-related inflammation F. Balkwill1 . 1 Barts and The London School of Medicine and Dentistry, London, United Kingdom Cancers are driven by complex networks of cytokines and chemokines that enable communication between the malignant cells and the many other cell types of the tumor microenvironment. Pre-clinical experiments have shown that targeting these networks with therapeutic antibodies or small molecule inhibitors can decrease tumor growth and spread. The challenge now is to translate these promising results to clinical trial. This talk will focus on the cytokine IL-6 and the chemokine receptor CCR4. IL-6 is a key regulator of an inflammatory cytokine network of highgrade serous ovarian cancer, HGSC. Clinical, pre-clinical and in silico experiments showed that antibodies to IL-6 can have multiple actions within the tumor microenvironment including reductions in cytokine production,
Sunday 6 July 2014
14:00−15:45
Symposium
Cancer Biomarkers
S.A. Carr1 . 1 Broad Institute of MIT and Harvard, Proteomics, Cambridge, USA Better biomarkers are urgently needed to improve diagnosis, guide molecularly targeted therapy, and monitor activity and therapeutic response across a wide spectrum of disease. Proteomics methods based on mass spectrometry hold special promise for the discovery of novel biomarkers that might form the foundation for new clinical blood tests, but to date little has been delivered. Proteomics-based biomarker discovery has been severely hampered by inadequate number and quality of samples, poor study design, and technological approaches lacking sufficient sensitivity, quantitative precision, and capacity to analyze statistically relevant numbers of samples. Another key problem has been the lack of robust quantitative methods to credential candidate protein biomarkers in larger numbers of patient samples prior to clinical evaluation. This problem extends into biology where the lack of highly specific affinity reagents for novel candidate proteins and modified peptides with sufficient sensitivity, specificity, reproducibility and throughput has significantly hampered our ability to understand dynamic, protein-based biological processes. In our biomarker studies we are addressing both of these serious barriers. In the discovery phase, we are employing multiplexed, quantitative MS technologies that enable analysis of larger numbers of patient samples with improved precision leading to better-qualified candidates. For verification of candidate biomarkers, we are developing targeted mass spectrometry-based technologies to screen and quantify low abundance proteins and modified peptides in a variety of biological contexts including human tissue and plasma. These assays are highly multiplexed, sensitive and specific and do not suffer from interferences that plague immunoassays. Targeted MSbased approaches are helping to usher in a new era of quantitative biology where proteins of interest and their modifications can be robustly measured in any biological context. We are applying these quantitative approaches in the context of a generalizable proteomics-based discovery-through-verification pipeline to identify early biomarkers of cardiovascular injury, breast and ovarian cancer. The same technologies are also being brought to bear to understand response and resistance to therapy. These studies are beginning to demonstrate that modern proteomic technologies when coherently integrated can yield novel, credentialed protein and peptide biomarker candidates of sufficient merit to warrant real clinical evaluation and to shed light on biological function. No conflict of interest. 35 Personalised proteotypes and their association with disease R. Aebersold1 . 1 ETH Zurich, Institute of Molecular Systems Biology, Zurich, ¨ Switzerland Powerful genomic technologies are now capable of determine genetic variability at a genomic level and at unprecedented speed, accuracy and (low) cost, and Genome Wide Association Studies (GWAS) have indicated the (frequently weak) linkage of specific loci to disease phenotypes. The question how genetic variability is translated into phenotypes is fundamental to biology and of critical practical importance for medicine. In this presentation we will discuss emerging computational and quantitative proteomic technologies to relate genotypic variation to the proteotype, the quantitative state of the proteome of a specific (disease) tissue. We will show that precise and highly reproducible proteotype measurements can be now generated at high throughput by the targeted proteomic methods selected reaction monitoring (SRM) and, at higher throughput, by SWATH-MS. We will discuss the principles of these mass spectrometric methods, discuss the computational challenged they pose for data analysis and demonstrate with selected applications their ability of proteotype data to