̽��ϵ��

Topic

Reproducible Transcriptomic States of Plasma-Cell-Rich Lymphomyeloid Aggregates in High Grade Serous Ovarian Carcinoma

Interdisciplinary Studies

Date & location

• Wednesday, April 8, 2026
• 9:00 A.M.
• Clearihue Building, Room B007

Examining Committee

Supervisory Committee

• Dr. Brad Nelson, Department of Biochemistry and Microbiology, ̽��ϵ�� (Supervisor)
• Dr. Farouk Nathoo, Department of Mathematics and Statistics, UVic (Co-Supervisor)
• Dr. Lisa Reynolds, Department of Biochemistry and Microbiology, UVic (Member)
• Dr. Sanjeena Dang, School of Mathematics and Statistics, Carleton University (Outside Member)
• Dr. Ali Bashashati, Department of Pathology and Laboratory Medicine, University of British Columbia (Outside Member)

External Examiner

• Dr. Katey Enfield, Department of Pathology and Laboratory Medicine, University of British Columbia

Chair of Oral Examination

• Dr. Jürgen Ehlting, Department of Biology, UVic

Abstract

Background: Plasma cells are emerging as prognostic markers in high-grade serous ovarian cancer (HGSC), yet their spatial organization and niche-associated transcriptional programs remain poorly characterized. We aimed to define plasma cell rich LMAs (sense immune regions enriched for plasma cells) in HGSC and define their transcriptional signatures. Given the dependence of plasma cells on niche derived signals, we hypothesized that plasma cell rich LMAs would segregate into reproducible aggregate level signatures dominated by distinct contextual programs.

Methods: We profiled 14 primary, untreated HGSC tumours using a modified 10x Genomics Visium workflow to obtain spatially resolved gene expression, B-cell receptor (BCR), and T-cell receptor (TCR) data. Immune lineages were inferred at the spot level using gene signatures, and LMAs were identified using a density-based KDE/DBSCAN framework. Transcriptional programs associated with plasma cell rich LMAs were assessed by within-section differential expression and pathway enrichment. We applied a similar workflow to an external cohort of 8 HGSC tumours to assess generalizability.

Results: LMAs were detected in 4 of 14 tumours (18 LMAs total), and plasma cell rich LMAs were observed in 2 tumours (5 plasma cell rich LMAs) within our internal cohort. Plasma cell rich LMAs were observed in 5 out of 8 tumours within the external cohort (12 plasma cell rich LMAs). Across internal and external cohorts, plasma-cell–rich LMAs segregated into four recurrent transcriptional signatures: (A) reduced biosynthesis/energy metabolism, (ii) stromal and extracellular matrix remodeling, (iii) antigen presentation and (iv) high biosynthesis/energy metabolism. Mixed signatures were also observed in a subset of LMAs.

Significance: These recurrent LMA programs define distinct spatial immune niches in HGSC and generate testable hypotheses about how microenvironmental context may constrain or support antibody-associated immune activity. In larger clinically annotated cohorts, LMA program frequency and localization can be evaluated for association with clinical outcomes.