IDIGH – CRP Joint Research Day

Title of Talk: Development of Multimodal AI tools to Personalize the Care of Transplant Recipients

Dr. Mamatha Bhat is a Hepatologist and Clinician-Scientist at the University Health Network’s Ajmera Transplant Centre, Scientist at TGHRI and Associate Professor of Medicine at the University of Toronto. Dr. Bhat completed her medical school and residency training at McGill University. She then completed a Transplant Hepatology fellowship at the Mayo Clinic in Rochester, Minnesota, followed by a CIHR Fellowship for Health Professionals, through which she completed a PhD in Medical Biophysics. The goal of Dr. Bhat’s research program is to improve long-term outcomes of liver transplantation by developing tools of Artificial Intelligence integrating clinical and ‘omics data, and has been funded by CIHR, Terry Fox research institute, Canadian Liver Foundation, American society of Transplant among others. She has published over 170 papers in journals such as Lancet Digital Health, Journal of Hepatology, JAMA Surgery, Gut and Hepatology. Dr. Bhat has been the recipient of recognitions such as the 2024 William Goldie Prize from the Department of Medicine, 2024 Gary Levy Research excellence award from the Ajmera transplant centre, 2022 CASL Research Excellence award and the 2021 American Society of Transplantation Basic Science Career Development Award. Dr. Bhat is also Director of the Clinician-scientist training program in the Dept of Medicine at U of T, Partnership & Engagement Lead for the Temerty Centre for AI Research and Education in Medicine (T-CAIREM), and past Chair of the International Liver Transplant Society Basic and Translational Science Research committee.

IDIGH Program Co-Leader
Scientist, RI-MUHC, Infectious Diseases and Immunity in Global Health Program, Centre for Outcomes Research and Evaluation
Associate Professor, Department of Family Medicine, Faculty of Medicine and Health Sciences, McGill University
Département de médecine générale, Division of Secondary Care, MUHC

---

Title of Talk: HPV-related cancers in people living with HIV

Dr. Alexandra de Pokomandy is a family physician specialized in HIV care and epidemiologist. Her clinical practice is at the Chronic Viral Illness Service (CVIS) of the McGill University Health Centre (MUHC) – Glen site. Her main research interests are in prevention of comorbidities in people living with HIV, with a particular focus on cancers related to human papillomavirus (HPV) and women’s health. In addition to clinical supervision offered to medical students and residents at the CVIS, she supervises graduate students and teaches epidemiology in the graduate program of the Family Medicine Department at McGill University.

Scientist, RI-MUHC, Infectious Diseases and Immunity in Global Health Program
Bruce Kaufman McGill IBD Research Chair
Professor, member of the division of Gastroenterology and Hepatology (MUHC).

---

Title of Talk: Uncontrolled inflammation in the GI tract: from pathophysiology to therapeutic solutions

His clinical activity is centered on the management of Inflammatory Bowel Disease (IBD). He worked at Hôpital Saint-Louis (APHP and Université Paris Cité), as head of the department of Gastroenterology, and director of lab team Intestinal Immunity in Inflammation and Cancer at INSERM U1160 at the Saint-Louis Research Institute. He founded the REMIND group in 2006, a national network aimed to develop translational research on IBD. He was the first scientific officer of ECCO, and is an IOIBD member.

His research focuses on inflammatory bowel diseases, a group of immune-mediated inflammatory disorders characterized by uncontrolled inflammation of the intestinal mucosa. He focuses on the role of innate and adaptive immunity and is interested in a variety of models to study the causes and mechanisms of these diseases, such as the postoperative situation where some patients relapse and others do not, or the response to advanced therapies. The main goals are to better characterize disease subtypes, to identify biomarkers that predict success or failure of advanced therapies, and to identify optimal combinations and novel therapeutic targets.

Title of Talk: Mucosal ecology in oncology : the  bladder cancer case

Dr. Éva Michaud specializes in mucosal immunology and its implications in cancer therapy. She completed her Ph.D. in Immunology from Université Lyon-Saint-Étienne, France, where she investigated the role of secretory IgA in inflammatory bowel diseases (IBD). Her research delved into understanding the mechanisms connecting dysbiosis and pathogenic immune responses at mucosal surfaces, aiming to identify novel therapeutic targets.

Continuing her academic journey, Dr. Michaud pursued post-doctoral research at McGill University, focusing on immune and microbial determinants of response to radiation therapy in muscle-invasive bladder cancer. Her work explored the interplay between gut microbiome composition and immune responses in bladder cancer, shedding light on potential avenues for improving treatment outcomes. In addition to her research endeavors, Dr. Michaud is passionate about education and mentorship. She has contributed significantly to curriculum design and teaching in immunology and vaccinology at several French academic institutions, empowering students with knowledge and practical skills in immunology and genetics.

Dr. Michaud’s contributions to the field have been recognized through numerous awards and accolades, including Best Research Presentation at previous Canadian Urological Association Meetings and the Aaron & Susan Lieberman & Family Award in Urology Research..

Title of Talk: Epigenetic drivers of malignant bone sarcomas

Title of Talk: IL-18 modulates Treg cell functional adaptation, alleviating immune suppression in melanoma tumour microenvironments

Abstract: Background/Rationale: Despite representing only 4% of all skin cancers, melanoma remains the leading cause of skin cancer-related mortality due to its propensity for metastasis and resistance to immune checkpoint inhibitors (ICI). Improving ICI efficacy requires a deeper understanding of the mechanisms governing lymphocyte infiltration and activity within the tumour microenvironment (TME). Alarmins, released upon tissue injury by hematopoietic and non-hematopoietic cells, promote local T-cell activation. Specifically, interleukin (IL)-18 helps to orchestrate an IFN-γ-driven type 1 immune response, modulating Treg suppression and promoting tumour cell clearance. We aim to investigate whether IL-18 disrupts the functional stability of tumour-infiltrating Treg cells, thereby diminishing effector suppression and augmenting α-PD-1 efficacy.

Methodology: The YUMMER 1.7 melanoma cell line was implanted into wild-type (control) mice and in mice with a T-cell-specific deletion of the IL-18R (CD4ΔIL-18R). Following tumour implantation, mice were treated with α-PD-1 or PBS. To further dissect the dynamics of IL-18-driven Treg responses, we established a tamoxifen-inducible model with a Treg-specific deletion of the IL-18R (Foxp3ΔIL-18R). At clinical endpoint, tumours were removed, and the dynamics of T-cell responses, relative to Foxp3⁺ Treg cells, in tumour and lymphoid compartments were analyzed by high-dimensional flow cytometry.

Results/Significance: We report that CD4ΔIL-18R deficient mice exhibit a significantly poorer response to α-PD-1 treatment relative to control mice. Tumour-infiltrating Treg cells exhibit IL-18-driven functional reprogramming, differentiating into IFN-γ⁺ Th1-like cells with reduced suppression of CD4⁺ T-cells in vitro. In Foxp3ΔIL-18R mice, IL-18R deletion leads to a bona fide conversion of Foxp3⁺ Treg cells into Foxp3⁻ exTreg cells.

Title of Talk: Understanding Early Host-Pathogen Interactions In Pulmonary Mycobacterial Infections

Abstract: Introduction: Patient-derived organoids (PDOs) have emerged as valuable preclinical models for studying tubo-ovarian carcinomas (OC), offering a platform to study molecular mechanisms and drug testing. However, published rate success for PDOs are low and mostly limited to fresh tissue. Here we present our experience using a range of tissue sources and techniques to optimize the yield of experimental models.

Hypothesis: We hypothesize that representative PDOs can be developed using fresh tissue (ftPDOs) and cryopreserved (ctPDOs) OC tissues and that resuscitated (rPDOs) can be successfully regrown for OC research.

Methodology: PDO development protocols were implemented using ftPDO, fresh ascites (aPDO), ctPDO, and rPDO. The morphology and molecular profile of rPDOs, ctPDOs, and fresh PDOs were compared to those of their respective parent tumor.

Results: A total of 35 PDOs (22 ftPDOs, 13 aPDOs) were developed from OC patients with various histological subtypes and cancer stages. Among the 22 ftPDOs, one reached passage 8 and another reached passage 20, both of endometrioid histotype. One aPDO reached passage five. Three rPDO cases have shown promising regrowth. Three ctPDOs are under development. Ongoing investigations aim to determine factors influencing PDO growth success. Pending immunohistochemical results are expected to show conserved morphology and protein expressions among the samples.

Contribution: The optimized cryopreservation and resuscitation strategies contribute to long-term accessibility to PDOs and tumor tissues for OC research, contributing to therapeutic discoveries.

Title of Talk: Ovarian Organoid Development Using Fresh and Cryopreserved Tissue, and Resuscitation of Patient-Derived Organoids

Abstract:Introduction: Patient-derived organoids (PDOs) have emerged as valuable preclinical models for studying tubo-ovarian carcinomas (OC), offering a platform to study molecular mechanisms and drug testing. However, published rate success for PDOs are low and mostly limited to fresh tissue. Here we present our experience using a range of tissue sources and techniques to optimize the yield of experimental models.

Hypothesis: We hypothesize that representative PDOs can be developed using fresh tissue (ftPDOs) and cryopreserved (ctPDOs) OC tissues and that resuscitated (rPDOs) can be successfully regrown for OC research.

Methodology: PDO development protocols were implemented using ftPDO, fresh ascites (aPDO), ctPDO, and rPDO. The morphology and molecular profile of rPDOs, ctPDOs, and fresh PDOs were compared to those of their respective parent tumor.

Results: A total of 35 PDOs (22 ftPDOs, 13 aPDOs) were developed from OC patients with various histological subtypes and cancer stages. Among the 22 ftPDOs, one reached passage 8 and another reached passage 20, both of endometrioid histotype. One aPDO reached passage five. Three rPDO cases have shown promising regrowth. Three ctPDOs are under development. Ongoing investigations aim to determine factors influencing PDO growth success. Pending immunohistochemical results are expected to show conserved morphology and protein expressions among the samples.

Contribution: The optimized cryopreservation and resuscitation strategies contribute to long-term accessibility to PDOs and tumor tissues for OC research, contributing to therapeutic discoveries.

Title of Talk: Novel Overexpression Platform to find Chemoresistance Genes for Combination Therapy

Abstract: Overcoming chemoresistance remains a critical challenge in cancer treatment, preventing progress towards precision oncology. While loss-of-function studies have been informative, the contribution of gene overexpression to chemoresistance is less understudied. This research addresses this gap by employing a novel functional genomics platform to systematically identify key chemoresistance genes, with the potential to translate this basic research to better precision oncology in clinical settings.

We developed the Bxb1-landing pad human ORFeome-integrated system for a proteome-wide Gene Overexpression (BOGO) platform, which enables single-copy ORF integration in a reusable and unbiased fashion across ~18K genes in cancer cell models. The BOGO platform was applied to HeLa cells treated with 16 chemotherapeutic agents. Drug-specific gene clusters were validated using single-cell RNA sequencing, with construction of a common resistance network. Functional studies were conducted using lysosomal activity and cancer cell viability assays under combination treatments to assess the clinical relevance of identified resistance genes.

Our screening identified enriched functional modules in autophagy, apoptosis, and Wnt signaling pathways, with key associated genes including TRADD, BCL2, and DAPK3. An autophagy-associated synergistic drug combination was selected for further evaluation. The results demonstrated that this combination more effectively reversed lysosome activity compared to individual chemotherapeutic treatments. The drug pair displayed significantly increased efficacy in colon and pancreatic cancer cells, with no impact on non-cancerous cells. This study presents an innovative platform to uncover chemoresistance drivers and identify synergistic drug combinations, offering a new framework for future cancer therapies.

Title of Talk: Proteomic Insights on LRV2 and Exosome-Mediated Rewiring of Macrophages in Leishmania Infection

Abstract: Leishmania RNA virus 2 (LRV2) has been implicated in exacerbating disease severity in cutaneous leishmaniasis, yet its precise role in host immune modulation remains unclear. LRV2 is transmitted via exosomes, which influence host-pathogen interactions by delivering virulence factors. Previous studies have established that Leishmania exosomes are critical mediators of immune evasion and disease progression, enriching virulence factors such as GP63 to modulate macrophage signaling, inhibit microbicidal functions, and influence inflammatory cell recruitment. However, the impact of exosome carrying LRV2 on host immune responses and its contribution to infection dynamics remains unexplored. We isolated exosomes from Leishmania major ^LRV2+(Uzbekistan field strain) and its LRV2- counterpart. Mouse bone marrow-derived macrophages were exposed to either Leishmania, their respective exosomes, or both, followed by proteomic analysis at an early infection stage. Our proteomic analysis reveals that while exosomes alone initially enhance macrophage activation, their proteomic profile is enriched in pathways related to phagocytosis and endocytic trafficking, potentially facilitating Leishmania entry. In contrast, co-infections with L. major and exosomes elicited the strongest immunosuppressive proteomic signatures, with LRV2+ co-infections further amplifying this effect through downregulation of key pro-inflammatory signaling pathways involved in signal transduction, innate immunity, and metabolism. The shift from an activated to a suppressed proteomic state may contribute to an environment conducive to both Leishmania entry and persistence. Given the global burden of leishmaniasis and its impact on vulnerable populations, understanding how LRV2 and exosomes contribute to disease severity is crucial for improved diagnostic, therapeutic, and vaccine strategies against this neglected tropical disease.

Title of Talk: Early-life RSV infection leads to durable and sex-specific alterations to the lung immune environment

Abstract: Background: Respiratory syncytial virus (RSV) is a leading cause of lower tract respiratory infection (LTRI) in infants under the age of one and is strongly linked to the development of allergic airways disease (AAD). While recently approved immunoprophylaxis reduces severe RSV-attributable LTRI, evidence of their ability to prevent RSV infection or reduce the long-term outcomes linked to RSV infection of any severity remains limited.

Previously, we demonstrated that neonatal RSV infection increases the acute release of IL-33 in the lung, a key cytokine that activates ILC2s – early drivers of T2 inflammation. Our data show that RSV infection of neonates also durably alters the lung, driving an increase in ILC2s selectively in adult female mice. Our data further show that early-life RSV infection leads to a selective increase in ILC2s in adult female mice. We aim to determine the immune mechanisms underlying this sex-specific effect, focusing on IL-33 as a potential therapeutic target to prevent AAD post-early life RSV.

Results: IL-33 levels peak at 12 hours post-RSV infection and are significantly higher in females. By postnatal day 50, IL-33 increases again, but only in RSV-infected females, accompanied by a selective expansion of ILC2s and elevated type 2 cytokine production.

Conclusions: These findings demonstrate that early-life RSV infection induces both acute and delayed IL-33 elevation, particularly in females. By 40 days post-infection, this corresponds to a selective expansion of KLRG1⁻ ILC2s, suggesting sex-specific immune responses that may drive enhanced inflammation and AHR. These results support the development of targeted therapies for RSV-associated AAD.

Title of Talk: Mapping the immune landscape of gastroesophageal adenocarcinoma patients with a high neutrophil-to-lymphocyte ratio

Abstract: INTRODUCTION: Previously, tumour staging was almost the sole predictor of outcomes. However, emerging data have emphasized the role of inflammation in cancer progression and outcomes. A high neutrophil-to-lymphocyte ratio (NLR>4) has repeatedly been associated with worse oncological outcomes. However, there is a knowledge gap about its underlying biology, restricting its use as a therapeutic avenue. This study thus aims to characterize the contrasts in immune landscapes of gastroesophageal adenocarcinoma (GEA) patients with high and low NLR.

METHODS: (1) Baseline circulating NLR were calculated through clinical tests. Patient tumour samples were stained for neutrophils and lymphocytes to assess intratumoural NLR. (2) Peripheral blood mononuclear cells (PBMCs) and neutrophils from GEA patients were co-cultured for 5 days, after which PBMC proliferation was assessed via cell counting. PBMCs were then co-cultured for 24hrs with A549-GFP cells and cytotoxicity was examined through EVOS imaging. (3) A 96-analyte ELISA was performed on patient plasma samples to establish a NLR-based cytokine profile. (4) Single cell RNA sequencing was performed on GEA tumour tissues.

RESULTS: (1) Both circulating and intratumoural NLR are indicative of poor clinical outcomes. (2) Circulating immune cells from high NLR patients are functionally different. (3) High circulating NLR underlines a pro-inflammatory immune landscape. (4) The tumour immune microenvironment is shaped by the patient’s NLR.

SIGNIFICANCE: Our findings provide insights on the immunological profile of high vs low NLR GEA patients, thus allowing a better understanding on why high NLR is associated with worse outcomes and how it can be exploited therapeutically.

Title of Talk: Fragmentomic features of plasma cell-free DNA in osteosarcoma patients

Abstract: Background/Rational: Plasma cell-freeDNA (cfDNA) molecules may have distinct fragmentomic characteristics, including 5′ end 4-mer motifs, fragment size, and nucleosomal patterns, that may have disease-specific prognostic and predictive value. It has been previously demonstrated that in osteosarcoma (OS) patients enrichment of short cfDNA fragments is associated with poor survival (PMID: 36735493). Here, we aim to explore whether OS-derived cfDNA has other disease-specific fragmentomic characteristics.

Methods: We analyzed publicly available whole-genome sequencing (WGS) data from cfDNA of three OS patients (SRA PRJNA917431) alongside our previously published cfDNA WGS data from 7 leiomyosarcoma, 12 leiomyoma, 3 dedifferentiated liposarcoma, and 3 well-differentiated liposarcoma patients (PMIDs: 29463554, 32232185, 34986184). Fragmentomic analysis was conducted by extracting the first four nucleotides from the 5′ fragment ends of plasma DNA molecules, calculating the frequency of each of the 256 possible motifs and normalizing these frequencies to the total number of reads in each sample. We applied multiclass differential analysis of the frequency of these motifs in 5 different types of tumors using nonparametric method with resampling and the permutation plug-in method to estimate the false discovery rate (SAMseq, samr R package).

Results/Significance: We identified 64 4-mer motifs enriched in OS cfDNA compared to cfDNA derived from other tumor types. Motifs linked to DNASE1L3 nuclease activity were decreased, while TAAA, AAAA, and TTTT were elevated in OS compared to other tumor types. These findings highlight distinct cfDNA fragmentomic patterns in OS, which provides a rationale for further investigation to explore their potential clinical utility in larger patient cohorts.

Title of Talk: Transcriptional Signatures of Immunosenescence and Inflammation in Adult Localized Scleroderma.

Abstract: Background/Rationale: Localized scleroderma (LS) is a fibrosing autoimmune skin disease with poorly defined pathogenic mechanisms. Environmental triggers are implicated, but how these exposures lead to sustained immune dysregulation remains unclear. In systemic autoimmune diseases, immunosenescence (age-associated immune remodeling) has emerged as a mechanistic link between environmental stressors and chronic inflammation. We hypothesized that LS exhibits transcriptional and histologic features of immunosenescence, particularly in clinically active or inflammatory disease subsets.

Methodology: RNA sequencing was performed on skin biopsies from 60 adults with LS and 22 age- and sex-matched healthy controls. Differential gene expression, unsupervised clustering, Gene Ontology and pathway enrichment (KEGG, MSigDB Hallmarks), gene set enrichment analysis (GSEA), and GTM-based cell-type deconvolution were conducted. Immunohistochemistry for p16 (CDKN2A) was used to assess senescent cell accumulation.

Results/Significance: LS skin demonstrated 704 differentially expressed genes, including enrichment of senescence-associated transcripts (CDKN2A, CXCL10, LAG3) and T cell activation markers (STAT1, IFNG). Clustering resolved three molecular subsets: inflammatory, fibrotic, and mixed. The inflammatory subset displayed robust activation of interferon signaling, senescence pathways, and Th1/Th17 signatures. Fibrotic LS exhibited a transcriptome partially overlapping with controls. p16 protein levels were significantly increased in LS compared to controls, supporting senescent cell accumulation. These findings provide evidence that immunosenescence, coupled with T cell–mediated inflammation, underlies transcriptional heterogeneity in LS. Senescence signatures were most prominent in active inflammatory lesions. Targeting senescence-associated pathways may offer a novel therapeutic strategy in LS, particularly for patients with refractory inflammation.

Title of Talk: From Conversation to Classification: An AI Emotional Tone Module for Psychosocial Risk Detection in Breast Cancer

Abstract:

Background

Emotional distress is common among oncology patients yet remains underreported due to clinical constraints, stigma, and insufficient detection methods. MARVINA is an AI-powered chatbot in development to assist breast cancer patients with symptom management. To enhance MARVINA’s future capabilities, we developed an emotional tone classification module to detect distress and facilitate future psychosocial interventions.

Methods

We adapted an annotated corpus of 3,750 messages from cancer support forums, classifying emotional tone into four categories: very negative (1,000), negative (1,000), neutral (1,000), and positive (750). Using a 70/15/15 split for training, validation, and testing, we fine-tuned a pre-trained DistillBERT model. Class weighting addressed label imbalance, and vocabulary expansion mimicked informal patient interactions, including emojis and URLs. Evaluation metrics included precision, recall, accuracy, and confidence scores.

Results/Significance

The classifier achieved an overall accuracy of 85%, with particularly strong performance in identifying very negative (89% precision/86% recall) and positive content (82% precision/85% recall). For example, the input, “I’m exhausted. The chemo is unbearable, and every day feels worse than the last. I’ve lost my hair, my strength, and honestly, my hope,” was classified as very negative with 83% confidence. The input, “I’m feeling much better today after the treatment 😊. Hope everything will go well in the future!” was classified as positive with 90% confidence.

This module represents a foundational step toward integrating psychosocial risk detection into MARVINA. Future directions include deploying tailored interventions and alerts to support emotional well-being in oncology care.

Title of Talk: Towards the development of an mCODE radiomics-dosiomics extension and knowledge base

Abstract: Background/Rationale: The scattered nature of health data, along with the lack of standardization interoperability, limits the potential of Artificial Intelligence (AI) incorporating medical images and radiomics to automate outcomes assessment in radiotherapy (RT) treatments. Establishing a standardized lexicon and data structure could enhance multicenter clinical studies. Our goal is to structure patient data relevant to RT research and create a knowledge base (KB), a machine-readable repository, with an ontology as a domain, including radiomics and medical images. Methodology: We aim to identify the essential data elements needed to encode radiomics and dosiomics information and develop the ontology. We are building our study on Minimal Common Oncology Data Elements (mCODE), an international initiative to improve interoperability by establishing a core set of structured data elements. We propose an extension to link patients’ medical image data, radiomics, and dosiomics with their health records. A review of the existing literature on the standardization of radiomics and dosiomics methods was conducted to include the minimum parameters that would impact their acquisition. We included data elements recommended by the Image Biomarker Standardisation Initiative (IBSI) guidelines. Results/Significance: Currently, our proposed extension consists of 14 data element sets with 99 attributes. The radiomics extension includes 11 sets and 54 attributes, while the dosiomics portion has 3 sets and 15 attributes. We also included an Imaging Study Profile. An mCODE-compliant ontology for multicenter radiomics and dosiomics studies has not yet been implemented. This study represents an initial step to standardize medical information for radiation oncology research.

Title of Talk: Whole genome sequencing revealed lethal prostate cancer signatures in primary tumors and serial liquid biopsies

Abstract: Background/Rational: Molecular characterization of primary prostate cancer (PCa) is essential for better understanding driver genes and the heterogeneity therein. However, additional modalities such as longitudinal testing in liquid biopsies are necessary to discern mechanisms of progression to lethal stages. The aim is to identify driver genes of lethal PCa in primary tumours and trace relevant molecular changes longitudinally in blood.

Methodology: Banked fresh frozen prostate tissues from radical prostatectomy cases (lethal PCa=80, disease-free=35) were processed to identify tumour foci and macro-dissect cores of high cellularity (>75%) for DNA extraction. Serial plasma collections from three lethal cases were used to isolate cell-free DNA (cfDNA). Sequencing of tumours, matched blood DNAs, and cfDNAs were performed to identify copy number variations (CNVs) and mutations.

Results/Significance: Already established and novel CNVs were found in lethal PCa. Clinical analysis revealed genomic deletions associated with rapid recurrence, metastases, castration-resistance/CRPC, and short overall survival. Comparing both lethal and disease-free patients showed shared alterations associated with PCa development and disease recurrence, and also pinpoint genomic signatures specifically associated with the PCa severity. Longitudinal cfDNA analysis revealed rising circulating tumor DNA (ctDNA) fraction at progression. Alterations associated with emerging PCa cell subtypes were detected in ctDNAs when metastatic patients reached the late stage. The identification of novel genomic changes in tumor DNA and ctDNAs may improve our understanding of lethal PCa. The ctDNA findings support the importance of liquid biopsy to monitor progression and can lead to the development of novel tests and more effective treatments for PCa.

Title of Talk: Increased Prevalence of Renal Cysts in Hereditary Leiomyomatosis and Renal Cell Cancer Syndrome: A Retrospective Cohort Study

Abstract: Hereditary Leiomyomatosis and Renal Cell Cancer (HLRCC) is a genetic syndrome caused by pathogenic variants in the fumarate hydratase (FH) gene, and characterized by cutaneous leiomyomas (CL), uterine leiomyomas (UL), and renal cell carcinoma (RCC). While renal cysts are not formally part of the HLRCC phenotype, evidence suggests a possible association. We aimed to assess the prevalence and features of renal cysts in an HLRCC cohort.

We reviewed magnetic resonance imaging (MRI) data from HLRCC patients followed at our OncoGenodermatology clinic (December 2022-January 2025). Demographics, renal cyst and HLRCC features were recorded. Cyst prevalence was compared to general population data using prevalence ratios (PRs). Logistic regression analyzed predictors of cysts, using adjusted odds ratios (aORs) and 95% confidence intervals (CIs).

Seventy-two patients were included (68.06% female, 98.61% White, median age: 44.5 years [interquartile range: 36, 55]). Renal cysts were present in 59.72%, significantly higher than the 27% prevalence reported in an MRI-based general population study (p <0.0001, PR 2.18; 95% CI 1.78-2.67). Cysts were more frequent in males (65.22%) and older patients (30.00% in patients <30 years vs 100.00% in ≥70). 67.44% had multiple. Among 19 evaluated, 36.84% had Bosniak II cysts. Increasing age was an independent predictor (aOR per year: 1.07; 95% CI 1.03–1.12). No associations were observed with FH variant type, CL, UL, or RCC. Our findings suggest renal cysts may be under-recognized HLRCC features, highlighting the utility of MRI surveillance for early intervention. Longitudinal studies are needed to evaluate renal cysts as potential RCC precursors.

Title of Talk: Predicting Pain in CBCT Images

Abstract: Background/Rationale: Bone metastases (BMs) are uniquely aggressive and painful cancers. Using medical imaging alone, it is difficult to distinguish painful BMs from painless ones. Past studies have been conducted to make this distinction by correlating CT radiomics features with pain data extracted from doctors’ notes. Our purpose is thus to investigate the correlation between cone-beam CT (CBCT) delta-radiomics features and pain levels in thoracic bone metastases patients undergoing radiotherapy in order to predict pain progression.

Methodology: Of 343 treatment plans obtained for metastatic thoracic spine lesions, 6 contained a sufficient number of CBCTs and associated pain data for our study. Lesion centre points were obtained from expert-labelled planning CT images and mapped to registered CBCTs. Doctor-reported pain data were extracted from documents in patients’ electronic health records. PyRadiomics was used to extract over 90 radiomics features from the CBCT images.

Results / Significance: Significant trends relating delta radiomics to pain were not observed. This result is due to the fact that, while pain data is often recorded during a patient’s initial consult, it is scarcely recorded throughout treatment. To overcome this barrier, we are training a pain-prediction model on first-day treatment CBCT scans (which offer much more pain data). Once validated, we will test this model on the subset of intra-treatment CBCT images with corresponding reported pain. The additional data afforded to this alternate model (trained only on first-day treatment CBCTs) will help address the limitations of our delta-radiomics model.

Title of Talk: Examining references to patient-reported outcome and experience measures in HIV clinical guidelines: A scoping review

Abstract: The author has chosen not to share the abstract with the public.

Title of Talk: Probing the Molecular Functions of CD109 in HNSCC Using UltraID Technology

Abstract:INTRODUCTION: Head and Neck Squamous Cell Carcinoma (HNSCC) is the sixth most common cancer. Risk factors include tobacco, alcohol, and HPV-16 infection. Most patients present with advanced disease with five-year survival rates below 50%. Despite available treatments, outcomes remain poor, underscoring the need for innovative therapies. Recent insights have identified potential biomarkers like CD109 which is highly expressed in SCC and possibly involved in tumor progression via its TGF-β signaling, cell proliferation and epithelial-mesenchymal transition. My study focuses on CD109’s role in tumorigenesis by comparing its protein interactions in normal versus cancerous cells. Utilizing UltraID, a smaller, faster, and more specific variant of the biotin ligase BirA, this project leverages proximity labeling technology to biotinylate protein lysine amines within a 10 nm radius.

METHODS: A plasmid containing CD109 fused to UltraID was cloned and transfected into HACAT, SCC9 and UM-SCC-38 cells. Expression was induced with doxycycline, and biotin was added to label CD109-interacting proteins. A streptavidin bead pulldown will isolate biotinylated proteins and will then be sent for mass spectrometry analysis.

RESULTS: The CD109-BirA UltraID plasmid was successfully transfected into all three cell lines. Expression was confirmed post-doxycycline induction, and optimal doxycycline concentration and incubation time were established to maximize labeling while maintaining viability. Next step is to perform streptavidin pulldown and mass spectrometry to characterize CD109’s interactome.

CONCLUSIONS: This study aims to define CD109’s role in HNSCC through UltraID-based proximity labeling. Upcoming mass spectrometry results will provide insights into CD109-mediated signaling, potentially identifying novel therapeutic targets for HNSCC.

Title of Talk: Conjugation of polyethylene-glycol to whole inactivated influenza virions enhances hemagglutinin specific humoral responses

Abstract:The author has chosen not to share the abstract with the public.

Title of Talk: Pharmacological targeting of the hexosamine biosynthesis pathway in leiomyosarcoma and desmoid type fibromatosis

Abstract:INTRODUCTION: Desmoid-type fibromatosis (DTF) is a locally aggressive, non-metastatic tumor managed with surveillance, hormonal therapy, or targeted agents. In contrast, leiomyosarcoma (LMS) is a highly malignant sarcoma with poor prognosis and limited treatment options. Recent studies implicate metabolic reprogramming, particularly the hexosamine biosynthesis pathway (HBP), in tumor progression. HBP produces UDP-GlcNAc, a substrate for glycosylation that regulates oncogenic signaling. While HBP activation is described in several cancers, its role in LMS remains unclear. We identified GFPT2, the HBP rate-limiting enzyme, as upregulated in LMS and correlated with poor outcomes. Transcriptional analyses suggest HBP contributes to LMS progression via aberrant O-GlcNAcylation mediated by O-GlcNAc transferase (OGT).

METHODS: We used LMS cell lines (SK-LMS-1, SK-UT-1B, LMS03) and one DTF line (DES23) to characterize OGT/O-GlcNAc expression by Western blot. Cells were treated with OGT inhibitors OSMI-1 (competitive) and ST045849 (allosteric) for 24h. Cell viability was assessed using CellTiter-Glo.

RESULTS: All cell lines expressed OGT and showed dose-dependent sensitivity to OGT inhibition. IC50 values for OSMI-1 ranged from 31.18 µM (SK-UT-1B) to ~400 µM (DES23). ST045849 was more potent, with LMS03 being the most responsive (49.03 µM).

CONCLUSION: OGT inhibition reduced viability in LMS and DTF cells, supporting HBP as a therapeutic target. Ongoing work includes testing next-generation inhibitors and combination strategies, and identifying O-GlcNAcylated proteins to refine therapeutic potential in sarcomas.