• Citer

Deciphering Invasive Lobular Carcinoma Heterogeneity through Multimodal Spatial Analysis

par Serra, Matteo
Président du jury Deplus, Rachel
Promoteur Sotiriou, Christos
Publication Non publié, 2025-09-05

Thèse de doctorat

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

Invasive lobular carcinoma (ILC) is the second most common histological subtype of breast cancer, yet it remains underexplored compared to invasive breast cancer of no special type (NST). While NST and ILC share common treatment strategies, ILC is characterized by distinct molecular and morphological features, including the inactivation of CDH1, a diffuse growth pattern, and a unique metastatic behavior. Despite these differences, current prognostic models rely primarily on proliferation markers, which may not fully capture the biological complexity of ILC. This thesis applies spatial transcriptomics (ST) and integrative multi-omics approaches to investigate ILC heterogeneity, identify new molecular subtypes, and develop a refined prognostic model that incorporates tumor-microenvironment interactions.To characterize the heterogeneity of ILC, ST was applied to a cohort of 43 primary hormone receptor-positive, HER2-negative ILC samples, integrating spatially resolved gene expression data with histopathological annotations. This approach revealed significant intra- and inter-tumor heterogeneity, particularly in the composition of the tumor microenvironment (TME). By clustering ST spots based on gene expression profiles and subsequently integrating this information with morphology, four distinct ILC subtypes (ILC4TME) were identified, each defined by unique microenvironmental and biological features. Some subtypes were associated with a more immune-infiltrated and metabolically active TME, while others were enriched in normal stromal elements or displayed high tumor proliferation rates. The robustness of these subtypes was validated in independent bulk RNA-sequencing datasets (METABRIC and SCAN-B), where they showed differences in prognosis, underscoring their clinical relevance.In a complementary study, this thesis also explored the role of adipose tissue in ILC progression, focusing on its metabolic and immune interactions with tumor cells. Adipose-rich areas in proximity to tumor cells were found to be metabolically active and associated with distinct gene expression signatures. These findings led to the development of ILC-MAP, a prognostic model that integrates metabolic, immune, and proliferation-related features, along with clinicopathological variables. This model demonstrated improved risk stratification compared to traditional prognostic tools when applied to external datasets, highlighting the importance of considering TME- driven processes in predicting ILC outcomes.The findings presented in this thesis challenge the notion that proliferation is the sole driver of ILC aggressiveness. Instead, they highlight the influence of microenvironmental factors, such as immune infiltration, metabolic activity, and tumor-stroma interactions, in shaping disease progression. The identification of distinct ILC subtypes provides a biologically informed classification system that could complement existing clinical and molecular prognostic models. Furthermore, the ability to infer these subtypes from bulk RNA sequencing suggests their potential for clinical implementation, offering a more accessible alternative to spatially resolved techniques.By combining spatial transcriptomics, gene expression analysis, and histopathological annotation, this thesis provides new insights into the biological diversity of ILC, emphasizing the relevance of TME heterogeneity in both classification and prognosis. The findings contribute to a growing body of evidence supporting the need for tumor- microenvironment-informed approaches to better stratify patients and refine therapeutic strategies for ILC.