Characterising transcriptomic and immune heterogeneity in aggressive invasive lobular breast cancers

Abstract

Invasive Lobular Carcinoma (ILC) is the second most common histological subtype of breast cancer (BC) accounting for 10-15% of all BCs. It is a unique disease entity with distinct histological appearances, molecular alterations and clinicopathologic features. It also has a unique tumour immune microenvironment. A subgroup of ILC patients have clinically aggressive disease with metastases occurring early (< 3 years) after primary diagnosis. These patients have limited treatment options. There is thus a need to better understand the molecular basis and transcriptional drivers of aggressive ILC, as well of the immune landscape to help identify potential new drug targets to improve patient outcomes. The project firstly investigated molecular drivers of clinically aggressive ILC at the genomic and transcriptomic level. This identified higher rates of TP53, FAT1 and HER2 alterations and an association between FGFR1 alterations and relapse in the aggressive pleomorphic ILC histological subtype. Using RNA sequencing of pleomorphic ILC (n = 47), survival analysis and the use of a random forest model enabled the generation of prognostic risk scores which further validated as predictors of overall survival in independent validation cohorts. Characterisation of the immune landscape at the histological level through the quantification of stromal TILs in 163 ILCs showed that the majority of ILCs are characterised by a low level of stromal TILs but pleomorphic cases contain higher levels, although stromal TILs were not associated with clinical outcome in both pleomorphic and non-pleomorphic ILC. Further immune cell characterisation using NanoString Digital Spatial Profiling (DSP) technology in a cohort of 20 pleomorphic ILCs identified an association between high levels of CD68+ cells (macrophages) and early relapse. Validation using dual IHC identified the M2-like/M1-like macrophage ratio is associated with poor clinical outcome. Co-culture experiments further showed enhanced growth of ILC cell-line tumour cells when grown with M2-like over M1-like macrophages. NanoString Whole Transcriptome (WTA) analysis of 10 pleomorphic ILCs revealed differences between ‘immune-hot’ and ‘immune-cold’ tumour cells and cancer-associated fibroblasts (CAFs) and identified higher expression of HOXB13 in ‘immune-cold’ tumour cells which was associated with poor outcome in independent ILC cohorts. A further experiment studying matched primary and brain metastases from an in vivo ILC model identified enrichment of pre-existing minor subclones that were enriched in brain metastatic lesions defined by high levels of the EMT associated gene GCNT2.Overall, the project revealed new insights into the transcriptomic and immune heterogeneity in clinically aggressive ILCs and identified new patient stratification biomarkers for further validation.