Prospective Evaluation of AI R&D Tool for Patient Stratification - Mechanism of Action Evaluation in Triple Negative Breast Cancer (PEAR-MET)

This is a UK-based observational study that aims to discover novel predictive biomarkers with the potential to guide treatment decision making and prolong PFS and OS in patients with advanced TNBC. Patients will undergo a mandatory, study-specific core needle biopsy or fine needle aspiration of the breast tumour or metastasis before commencing their next line of therapy. The research sample will be run on Pear Bio's test whilst the patient receives therapy as per their physician's choice. This study will not use Pear Bio's tool to inform the choice of treatment, with the treating oncologist being blinded to the test results. Treatment response data will be collected at multiple timepoints to conduct analyses on the study's secondary and tertiary objectives.

Fresh tissue resections that arrive at Pear Bio's lab will undergo processing, cell culture and various drug dosing and omics assays (depending on extracted cell numbers). Tumour samples will be processed using a cell isolation kit to retrieve a viable single-cell suspension. A minimum of 100,000 cells (10,000 viable cells per chip) will be used for staining with live and dead cell-tracking dyes. In parallel, blood vials will be processed for PBMCs and further effector cell extraction (flow cytometry, Dynabeads, etc). The remaining cells will be used for sequencing (DNA/RNA), fixed for immunofluorescence characterisation of biomarkers/receptor status or used for further omics assays (if cell numbers allow). This may include tumour mutational burden and microsatellite instability testing.

The stained cells will be cultured in a biomimetic hydrogel within Pear Bio's organ-on-a-chip to provide a physiological 3D environment for drug dosing experiments. Using a microfluidic device, samples in each chip will be exposed to approved therapies (either as monotherapy or combination therapies, as outlined below) over multiple days.

In parallel, PBMCs will be extracted from whole blood, characterised and sorted via flow cytometry and fluorescence-activated cell sorting (FACS) or magnetic beads selection. Cells of interest (e.g. CD8+ T cells) will be used for culture in Pear Bio's chips jointly with cells isolated from the matched tumour sample. To test immunotherapies, tumour cells will be co-cultured with immune cells in a modified organ-on-a-chip architecture. Chips receiving immunotherapies may be tested for tumour mutational burden and/or microsatellite instability.

Confocal microscopy will be conducted daily to collect 3D image data of the cells and track their position and behaviour over time. At the end of the assay, the 3D cell cultures will be fixed for further 3D immunofluorescence analyses or used for embedding, sectioning and assessment of spatial transcriptomics. For targeted therapies, RNAseq, IF and other omics data will be integrated to confirm drug MoA and identify other potential therapeutic targets. Concurrently, 3D image data is processed through a computer vision pipeline to measure functional metrics of the ex vivo 3D cell cultures, including cell viability, cell culture width and cell migration, both at a bulk tumour level and at a single-cell resolution. For immunotherapies, additional metrics such as immune cell infiltration and immune cell killing will be recorded. A patient report is then generated to outline an individual patient sample's response to each therapy tested.