Spatial transcriptomic profiling of bone marrow in myeloproliferative neoplasms

Thesis

This thesis presents a multi-modal, spatially informed profile of bone marrow in myeloproliferative neoplasms (MPNs), integrating spatial transcriptomics and machine learning based image analysis.

In Chapter 3, I develop a pipeline for processing spatial transcriptomic data from bone marrow biopsies. This includes segmentation of non-uniform cells and construction of graphs that preserve neighbourhood relationships across heterogeneous morphology. Cell annotation is improved through multi-layer segmentation and integration with scRNA-seq references, including one enriched for stromal cells. Macrostructures such as bone, adipocytes, vasculature and fibrosis maps are incorporated to provide spatial context.

Chapter 4 presents a spatial transcriptomic analysis of bone marrow from MPN patients and healthy donors. I identify ten cellular neighbourhoods (CNs), several aligning with known regions. CN2, associated with higher fibrosis, may represent early fibrosis or a cellular correlate. Spatial expansion is a MF feature and, in ET and MF, expanded CNs become more distant from CN5 (a lymphoid aggregate), indicating impaired immunosurveillance. At higher fibrosis grades, I observe novel associations between Osteogenic Mesenchymal Stem Cells (Osteo-MSCs) and megakaryocytes. Gene expression suggests these Osteo-MSCs may correlate with a known profibrotic subset. CN1, CN3, and CN4 resist expansion and maintain spatial organisation, while CN3, CN4, and CN9, enriched for HSPCs, are flanked by distinct lineages, supporting spatially encoded niches. Finally, megakaryocyte morphology reflects transcriptional state. This work supports integrating spatial transcriptomics into haematopathology workflows and clinical decision-making.

Authors: Thomas, E

• DOI: 10.5287/ora-ovdjxvejp
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: xenium; myeloproliferative neoplasms; MPN; fibrosis; haematology; blood cancer; spatial transcriptomics; bone marrow; megakaryocytes; pathology; spatial biology
• Subjects: Hematology; Pathology; Fibrosis; Myeloproliferative disorders; Spatial analysis (Statistics); Bioinformatics