Conventional therapy induces tumor immunoediting and modulates the immune contexture in colorectal cancer
Creators
-
Fotakis, Georgios
(Data manager)1
-
Rieder, Dietmar
(Data manager)1
-
Loncova, Zuzana1
- Carollo, Sandro1
- Klieser, Eckhard2
- Daniel, Neureiter2
- Huemer, Florian3
- Hoegler, Sandra4
- Tomberger, Martina5
- Krogsdam, Anne1
- Kenner, Lukas4, 5
- Ziegler, Paul K.6
- Greil, Richard3, 7
- Weiss, Lukas3, 7
- Trajanoski, Zlatko (Contact person)1
- 1. Institute of Bioinformatics, Biocenter, Medical University of Innsbruck, Innsbruck, Austria
- 2. Institute of Pathology, Paracelsus Medical University/Salzburger Landeskliniken (SALK), Salzburg, Austria
- 3. IIIrd Medical Department, Paracelsus Medical University Salzburg; Salzburg Cancer Research Institute-Center for Clinical Cancer and Immunology Trials, Salzburg, Austria
- 4. Department of Pathology, Medical University of Vienna, Vienna, Austria
- 5. CBmed-Center for Biomarker Research in Medicine GmbH, Graz, Austria
- 6. Senckenberg. Institute of Pathology, Goethe University Frankfurt, Frankfurt, Germany
- 7. Cancer Cluster Salzburg, Salzburg, Austria
- 1. Institute of Bioinformatics, Biocenter, Medical University of Innsbruck, Innsbruck, Austria
Description
Cancer immunotherapies for patients with colorectal cancer (CRC) continue to lag behind other solid cancer types with the exception of 4% of patients with microsatellite-instable tumors. Thus, there is an urgent need to broaden the clinical benefit of checkpoint blockers to CRC by combining conventional therapies to sensitize tumors to immunotherapy. However, the impact of conventional drugs on immunoediting and hence, imposing positive selection towards less immunogenic variants, and on the tumor immune contexture in CRC remains elusive.
In this study, we performed comprehensive multimodal profiling using longitudinal samples from metastatic CRC patients undergoing neoadjuvant therapy with mFOLFOX6 and Bevacizumab. Exome-sequencing, RNA-sequencing and multiplexed immunofluorescence imaging was carried out on tumor samples obtained before and after therapy and the data was analyzed using established methods. The results of the analysis were extrapolated to publicly available datasets (TCGA and CPTAC). In order to identify a surrogate marker, an explainable artificial intelligence method was developed using a transformer-based analytical pipeline for the identification of features in H&E images associated with specific biological processes, followed by manual evaluation of highly informative tiles by a pathologist.
We expect that the results of this project will provide a deeper understanding of the tumor-immune interactions and will allow the development of more robust combinatorial therapeutic strategies for MSS CRC.
Notes
This archive contains all the processed data generated from this study. The raw data files will not be made publicly accessible to protect patient privacy.