Abstract:Metastases are the primary cause of cancer-related death, and improving the means of predicting and targeting their development is one of the major goals in cancer research. While surgical resection and neo-adjuvant therapy can cure well-confined primary tumors, our ability to effectively treat cancer is largely dependent on our capacity to interdict the process of metastasis. The recent accumulation of 'omics data from metastatic tumors provides an unprecedented opportunity to develop machine learning models to predict the molecular changes during metastasis and explore the patterns of metastasis formation. With this in mind, we developed MetMapper, a deep learning model trained on primary and metastatic tumors from > 13 000 patients integrating data from 11 published data resources. MetMapper can predict the transcriptomic changes of a primary tumor when it metastasizes to different distant organs. The results were extensively validated using transcriptomics data from matched primary and metastatic tumor biopsies extracted from the same patients. Furthermore, MetMapper's predictions revealed that the non-random patterns of cancer metastases can be partly explained by the degree of transcriptome reprogramming needed during metastasis: primary tumors tend to metastasize to organs that require minimal changes to their transcriptomes. Using MetMapper, we derived a metastatic potential score for patient tumors and demonstrate that this score can be used to stratify patients into high and low survival groups across different indications. The predicted metastatic potential of patient tumors significantly correlates with experimentally characterized metastatic potential of cancer cell lines. Additionally, by performing in-silico perturbations of genes and oncogenic pathways that can alter the metastatic potential of patient tumors, we identified genomic features that are highly associated with metastases to specific organs, some of which were reported by existing pan-cancer clinical sequencing studies. Our results demonstrate the utility of MetMapper as a novel AI-powered methodology for investigating mechanisms and patterns of metastatic dissemination, as well as forecasting metastatic outcomes of patient tumors. Citation Format: Gang Li, Evan Béal, Dean Sumner, Giorgio G. Galli, Viviana Cremasco, Joshua M. Korn, Frank Dondelinger, David Ruddy, Audrey Kauffmann, Slavica Dimitrieva. Predicting metastatic transcriptomes of patient tumors with deep learning [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular s); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl) nr 896.

A metastasis map of human cancer cell lines

A multi-modal single-cell and spatial expression map of metastatic breast cancer biopsies across clinicopathological features

Identifying Metabolomic Features That Predict Metastasis of Melanoma from a Primary Site

Single-cell lineages reveal the rates, routes, and drivers of metastasis in cancer xenografts

Abstract IA23: Single cell-resolution lineage tracing of tumors in a mouse xenograft model reveals rapid and multidirectional metastasis

Abstract 896: Predicting metastatic transcriptomes of patient tumors with deep learning

Mapping the breast tumor microenvironment: proximity analysis reveals spatial relationships between macrophage subtypes and metastasis-initiating cancer cells

Identifying a gene signature of metastatic potential by linking pre-metastatic state to ultimate metastatic fate

Discovery of Metastatic Regulators using a Rapid and Quantitative Intravital Chick Chorioallantoic Membrane Model

Transcriptomic analysis identifies organ-specific metastasis genes and pathways across different primary sites

Metabolic Mechanisms Regulating Distinct Steps of the Melanoma Metastatic Cascade

Mapping the niche of breast cancer metastases in lung and liver

Integrative Analysis of Membrane Proteome and MicroRNA Reveals Novel Lung Cancer Metastasis Biomarkers

Spatiotemporal Immune Landscape of Colorectal Cancer Liver Metastasis at Single-Cell Level

Integrated cross-species transcriptional network analysis of metastatic susceptibility

MetFinder: A Tool for Automated Quantitation of Metastatic Burden in Histological Sections From Preclinical Models

Integrating Single-cell and Bulk RNA-seq to Construct a Metastasis-related Model for Evaluating Immunotherapy and Chemotherapy in Uveal Melanoma

For robust big data analyses: a collection of 150 important pro-metastatic genes

Epco-42. Single-Cell Multiomic Analysis Of Brain Metastases Across Multiple Primary Tumor Types