Cancer is amongst the leading causes of disease-related morbidity and mortality. A major challenge in cancer treatment is the development of biology-informed, personalised treatment strategies. Recent advances in artificial intelligence (AI) and next-generation sequencing (NGS) technologies have shed further insights into disease biology and treatment pathways, thus identifying new, precision medicine-based therapeutic opportunities.

The biological mechanisms leading to cancer development and progression arise from complex and plastic networks of dysregulated cellular programs involving many signalling pathways and effector molecules. Cancer cells alter their surrounding environment via cell-cell interactions with non-tumor cells or by secreting cytokines, chemokines and other factors. This reprogramming of the tumour microenvironment (TME) is critical for cancer progression, invasion, and metastasis. Moreover, there are increasing studies that show that both innate and adaptive immune cell types contribute to tumorigenesis and treatment resistance when present within the TME. Understanding the crosstalk between cancer cells and the surrounding TME will inform on mechanisms of sensitivity and resistance to treatment, including immunotherapy (IO) and targeted therapies.

Spatially resolved-Omics is an emerging field that characterises cell types by gene/protein expressions within their spatial context in the tissue organisation. Recent high profile spatial transcriptomics studies have uncovered specific cell identities that define the surrounding TME.

The MOSAIC study, a collaborative initiative across industry and top oncology hospitals, proposes to go way beyond current cancer molecular profiling projects by combining the generation and analysis of multiple data modalities (3 essential mandatory modalities: Clinical Data, Hematoxylin and Eosin (H&E) microscopic image, Spatial transcriptomics; up to 3 high priority data modalities depending on technical feasibility and sample size: bulk Ribonucleic Acid Sequencing (RNAseq), bulk Whole Exome Sequencing (WES), Single-cell transcriptomics; and potentially other optional data modalities and follow-up experiments such as single-cell omics, immunohistochemistry and spatial proteomics or other molecular profiling of proteins and molecules) on a minimum of 2,000 tumour samples across a different cancer indications. This will generate broad molecular and cellular profiling data of the tumour and its microenvironment from cancer patients, integrated with clinical data, at an unprecedented scale and resolution.

This study will enroll patients diagnosed with one of the eligible cancer indications and for which a formalin fixed paraffin embedded (FFPE) tumor sample from already performed biopsy and/or surgical resection is available within their local pathology archive or their affiliate centers archives.

The MOSAIC study expects to have a strong impact for patients in terms of new targeted therapeutic drug discovery, identification of patient subgroups requiring either specific treatment or broader clinical care and identification of novel treatment response and resistance mechanisms.

Inclusion Criteria:

1. That the patient was over the age of 18 when the sample was taken/consented to.

2. Availability of patient informed consent or non-opposition form to perform exploratory research matching at least one of the MOSAIC objectives (unless authorisation is granted by the local Institutional Review Board / Independent Ethics Committee (IRB/IEC) for the use of samples in the study according to local regulations and law).

3. Have a confirmed diagnosis based on international criteria for the relevant tumor type.

4. Confirmed formalin fixed and paraffin embedded (FFPE) tissue availability to generate at least the 3 core data modalities, and preferentially all MOSAIC data modalities.

5. Confirmed availability of associated clinical data.

6. Qualification of the paraffin tissue block meeting all of the following:

   • Being of the expected tumor type
   • For solid tumors (all cancer indications except diffuse large B cell lymphoma (DLBCL)): Tumor cell content ranging from 40% to 80% on an hematoxylin and eosin (H&E) section within a specified area as dictated by the lab protocol specific to the technique utilized
   • For DLBCL, a minimum of 80% of high grade component on an H&E section within a specified area as dictated by the lab protocol specific to the technique utilized
   • Wherever possible, the remaining tissue thickness must be over 125 micrometers (indicative range)
   • Tumor sample must be <10 years old

In addition, each sub-cohort within each cancer indication will have specific inclusion criteria (e.g., disease stage; sampling site; treatment received…).

Exclusion Criteria:

1. Samples without a preserved tissue architecture, such as cytologies and cytoblocks.

In addition, each sub-cohort within each cancer indication may have specific exclusion criteria (e.g. histological subtype; history of immunosuppression; etc…).

Patients across different cancer indications, which include, but are not restricted to histologically confirmed solid tumors (all types and subtypes) and diffuse large B cell lymphoma (DLBCL) and for which there is a formalin fixed and paraffin embedded (FFPE) tumor sample from already performed biopsy and/or surgical resection available.

Pseudonymized, preprocessed, and minimized data belonging to a subset of deceased patients is going to be made available to researchers on the European Genome-Phenome Archive (EGA). This project is called MOSAIC-Window.

In addition, following the ten years of the MOSAIC research study, the collected and generated data within MOSAIC will be anonymized in compliance with applicable laws and regulation, and a public version of the MOSAIC data will be released for research purposes. A dedicated committee of the study will decide on the conditions and modalities of the anonymization.

MOSAIC-Window start date: Q3 2024

Prior to uploading the data to the EGA, Owkin ensures that all personal data embedded in the datasets are pseudonymized and encrypted both in transit and at rest. Owkin is responsible for obtaining and demonstrating the existence of data subjects' informed consent or any other suitable legal basis for processing personal data, obtaining appropriate ethical or other approvals for collecting the data, and complying with applicable Data Protection Laws.

Researchers from institutions outside the MOSAIC Consortium will be able to ask for access to this data, specifying the research questions they want to answer. A Data Access Committee (DAC) will evaluate the researcher's application and decide whether to give access to the MOSAIC-Window dataset or not. If the application is successful, the researcher will agree with the terms and conditions defined by the DAC (e.g. creative common license, no re-identification of patients, etc.) and then will be able to download the data for analysis.