Disease Subtypes Research Articles

SALL4 as a Useful Marker for the Distinction of Various Gestational Trophoblastic Disease Subtypes: Choriocarcinoma From Other Trophoblastic Lesions and Early Complete Hydatidiform Mole From Partial Mole and NonMolar Villi.

The distinction between choriocarcinoma and residual trophoblastic cell proliferation from a complete hydatidiform mole/invasive mole (CHM/IM) without villi is challenging on curettage materials. We investigated whether SALL4 immunostaining could help differentiate various gestational trophoblastic diseases. Placental site nodules (PSN; n=10), atypical PSN (APSN; n=8), placental site trophoblastic tumors (PSTT; n=9), epithelioid trophoblastic tumors (ETT; n=5), gestational choriocarcinomas (n=31), partial hydatidiform moles (PHM; n=13), CHM/IM (n=47), and nonmolar products of conception (POC) (n=26) were included. SALL4 immunostaining was quantified (0 [1% to 10%], [11% to 100%]) and characterized (scattered single-cell or clustered nuclear positivity) in 2 locations: cytotrophoblast/intermediate trophoblast and villous stromal fibroblasts. A diffuse (11% to 100%) and clustered pattern of SALL4 immunostaining in cytotrophoblast/intermediate trophoblast was statistically associated with choriocarcinomas (74.2%, 23/31) as compared with PSN (0/10; P<0.0001), APSN (0/8; P=0.0002), PSTT (0/9; P<0.0001), ETT (0/5; P=0.0034), PHM (0/13; P<0.0001), CHM/IM (0/47; P<0.0001), and nonmolar POC (0/26; P<0.0001). Most nonchoriocarcinoma samples showed no SALL4 expression; when present, it was of low level (1% to 10%) and with a scattered single-cell staining in 3/9 PSTT (33%), 1/13 PHM (7.7%), 19/47 CHM/IM (40%), and 1/26 nonmolar POC (1.7%). These results were confirmed using a validation cohort. In addition, 66% (31/47) of CHM/IM villous stromal fibroblasts showed SALL4 expression (11% to 100%) (all before 14 gestational weeks), whereas this level of expression was never observed in PHM (0/13), nor in nonmolar POC (0/26; P<0.0001). Finally, a clustered and >10% SALL4 immunostaining in cytotrophoblast/intermediate trophoblast favors choriocarcinoma diagnosis. SALL4 expression in >10% villous stromal fibroblasts before 14 gestational weeks favors CHM/IM rather than PHM and nonmolar POC.

Trigeminal Neuralgia: Disease Characterization and Prediction of Response to Surgical Intervention.

Trigeminal neuralgia (TN) is a highly heterogeneous condition with a wide choice of successful treatment options. However, differences between subtypes are poorly understood and it remains unknown which patients will respond to different treatments. This review aims to summarize the current state of the TN field and explore the problem of predicting surgical outcomes. Attempts have been made to standardize classification, outcome reporting, and treatment guidelines. Identification of risk factors for poor surgical treatment outcomes has led to the development of prognostic scoring systems to predict outcomes and guide management. Advances in imaging techniques including diffusion tensor imaging show promise in better understanding the underlying pathophysiology and predicting surgical outcomes. Progress is hampered by lack of understanding of the pathophysiology in TN and differences between disease subtypes. Further work needs to be done to create accurate prediction tools for widespread use, including use of advanced imaging and computing tools, multicenter collaboration, and prospective validation.

A generative deep neural network for pan-digestive tract cancer survival analysis

BackgroundThe accurate identification of molecular subtypes in digestive tract cancer (DTC) is crucial for making informed treatment decisions and selecting potential biomarkers. With the rapid advancement of artificial intelligence, various machine learning algorithms have been successfully applied in this field. However, the complexity and high dimensionality of the data features may lead to overlapping and ambiguous subtypes during clustering.ResultsIn this study, we propose GDEC, a multi-task generative deep neural network designed for precise digestive tract cancer subtyping. The network optimization process involves employing an integrated loss function consisting of two modules: the generative-adversarial module facilitates spatial data distribution understanding for extracting high-quality information, while the clustering module aids in identifying disease subtypes. The experiments conducted on digestive tract cancer datasets demonstrate that GDEC exhibits exceptional performance compared to other advanced methodologies and can separate different cancer molecular subtypes that possess both statistical and biological significance. Subsequently, 21 hub genes related to pan-DTC heterogeneity and prognosis were identified based on the subtypes clustered by GDEC. The following drug analysis suggested Dasatinib and YM155 as potential therapeutic agents for improving the prognosis of patients in pan-DTC immunotherapy, thereby contributing to the enhancement of cancer patient survival.ConclusionsThe experiment indicate that GDEC outperforms better than other deep-learning-based methods, and the interpretable algorithm can select biologically significant genes and potential drugs for DTC treatment.

Intelligent strategy for severity scoring of skin diseases based on clinical decision-making thinking with lesion-aware transformer

Skin diseases are numerous in types and high in incidence, posing a serious threat to human health. Accurately assessing the severity of skin diseases helps dermatologists in making personalized treatment decisions. However, focusing solely on the skin lesion itself and ignoring the true state of the surrounding skin can lead to distorted results. Assessing the severity of the condition should be a holistic process. Specifically, dermatologists need to compare the abnormal skin with surrounding skin to conduct the diagnosis. To imitate such diagnosis practice of dermatologists, we propose LSATrans, a Transformer based framework customized for severity scoring of skin diseases. Different from the Standard Self-Attention module, we propose the Lesion-aware Self-Attention (LSA) module. LSA can capture the visual features of both lesion and normal surrounding skin areas and include their relationship in modeling. In addition to LSA, the proposed LSATrans also introduces a contrastive learning strategy for further optimization. We first evaluated the performance of LSATrans in scar, atopic dermatitis, and psoriasis scoring tasks, and it achieved mean absolute errors of 0.5895, 0.5614, and 0.5416 respectively in these three tasks. Furthermore, we conducted additional validation of LSATrans’s performance in two distinct skin disease diagnosis tasks, where it demonstrated remarkable outcomes with AUCs of 0.9774 and 0.9801, respectively, in the classification of common skin diseases and subtypes of skin diseases. These results are better than existing methods, indicating that LSATrans is expected to become a universal, accurate and objective intelligent tool for scoring the severity of skin diseases.

DDAVP response and its determinants in bleeding disorders: a systematic review and meta-analysis.

Desmopressin (DDAVP) can be used to prevent or stop bleeding. However, large inter-individual variability is observed in DDAVP response and determinants are largely unknown. In this systematic review and meta-analysis we aim to identify the response to DDAVP, and the factors that determine DDAVP response in patients. We included studies with patients with any bleeding disorder receiving DDAVP. First and second screening round and risk of bias assessment were performed by independent reviewers. The main outcome was proportion of patients with complete (factor level > 50 U/dL), or partial (30-50 U/dL) response to DDAVP. Determinants of response including disease type, age, sex, Von Willebrand factor (VWF) and factor VIII (FVIII) mutations, and baseline factor levels were investigated. In total, 591 articles were found and 103 were included. Of these, 71 articles (1772 patients) were suitable for the study's definition of response. Meta-analysis showed a pooled response proportion of 0·71 [0·64;0·78] and a significant difference in response between disease subtypes. For hemophilia A, baseline FVIII:C was a borderline significant determinant of response. In von Willebrand disease (VWD) type 1 patients, VWF:Ag, VWF:Act and FVIII:C were significant determinants. A large variation in response was observed for specific mutations in VWF and F8. Response to DDAVP varied between disease subtypes, and was largely determined by the baseline levels of FVIII:C for hemophilia A and VWF:Ag for VWD. Our findings highlight the significant differences in response and emphasize the need for a standardized response definition and further research into response mechanisms.

Connecting electronic health records to a biomedical knowledge graph to link clinical phenotypes and molecular endotypes in atopic dermatitis

Precision medicine is defined by the U.S. Food & Drug Administration as “an innovative approach to tailoring disease prevention and treatment that considers differences in people’s genes, environments, and lifestyles”. To succeed in providing personalized medicine to patients, it will be necessary to integrate medical, biological and molecular data in order to identify all complex disease subtypes and understand their pathobiological mechanism. Since biomedical knowledge graphs (BKGs) are limited to the integration of prior knowledge data and do not integrate real-world data (RWD) that would allow for the incorporation of patient level information, we propose a first step towards using RWD, BKGs and graph machine learning (ML) to enable a fully integrated precision medicine strategy. In this study, we established a link between RWD and a BKG. Our methodology introduced a novel patient representation using graph ML applied to the BKG. This approach facilitated the interpretation and extension of ML findings, particularly in disease subtype identification with molecular data contained in the BKG. We applied our innovative methodology to deepen our understanding of atopic dermatitis, a condition with a complex underlying pathophysiological mechanism. Through our analysis, we identified seven subgroups of patients each characterized by clinical and genomic characteristics.

Symmetry in Genetic Distance Metrics: Quantifying Variability in Neurological Disorders for Personalized Treatment of Alzheimer’s and Dementia

This paper aims to adapt and apply genetic distance metrics in biomedical signal processing to improve the classification and monitoring of neurological disorders, specifically Alzheimer’s disease and frontotemporal dementia. The primary objectives are: (1) to quantify the variability in EEG signal patterns among the distinct subtypes of neurodegenerative disorders and healthy individuals, and (2) to explore the potential of a novel genetic similarity metric in establishing correlations between brain signal dynamics and clinical progression. Using a dataset of resting-state EEG recordings (eyes closed) from 88 subjects (36 with Alzheimer’s disease, 23 with frontotemporal dementia, and 29 healthy individuals), a comparative analysis of brain activity patterns was conducted. Symmetry plays a critical role in the proposed genetic similarity metric, as it captures the balanced relationships between intra- and inter-group EEG signal patterns. Our findings demonstrate that this approach significantly improves disease subtype identification and highlights the potential of the genetic similarity metric to optimize the predictive models. Furthermore, this methodology supports the development of personalized therapeutic interventions tailored to individual patient profiles, making a novel contribution to the field of neurological signal analysis and advancing the application of EEG in personalized medicine.

Amygdala-centered fusional connections characterized nonmotor symptoms in Parkinson's disease.

The importance of nonmotor symptoms in understanding the pathogenesis of the heterogeneity of Parkinson's disease has been highlighted. However, the validation of specific brain network biomarkers in nonmotor symptom subtypes is currently lacking. By performing a new approach to compute functional connectivity with structural prior using magnetic resonance imaging, the present study computed both functional connectivity and fusional connectivity features in the nonmotor symptom subtypes of Parkinson's disease, one characterized by cognitive impairment with late onset and the other depression with early onset. The functional connectivity and fusional connectivity features centered at the left amygdala were both detected. The fusional features significantly enhanced the classification performance. The amygdala-postcentral and amygdala-orbital frontal features were critical for cognitive impairment with late onset detection, while the amygdala-temporooccipital features were crucial for depression with early onset detection. Additionally, the fusional connectivity features between the amygdala and the junction sulcus of parietooccipital and temporooccipital regions contributed significantly to differentiating cognitive impairment with late onset and depression with early onset. The within-subtype correlation analysis revealed that age at onset and cognitive scores were associated with features of amygdala-somatosensory/visual-motor processing areas in cognitive impairment with late onset, while related to features of amygdala-emotional processing areas in depression with early onset. Our findings highlighted distinct amygdala-centered fusional connectivity features related to diverse nonmotor symptoms in Parkinson's disease, offering new insights for pathogenesis-targeted treatments for specific Parkinson's disease subtypes.

P1352 Segmental Mucosal Microbiome Variations Across Different Inflammation Extents in Ulcerative Colitis

Abstract Background Ulcerative colitis (UC) is a chronic inflammatory condition characterized by a relapsing-remitting course and represents a major subtype of inflammatory bowel disease (IBD). The gut microbiome plays a pivotal role in the pathogenesis and progression of UC. However, the potential differences in the mucosal microbiome composition across colonic segments, depending on the extent of inflammation in UC, remain poorly explored. Methods we investigated the differences in the mucosal microbiome according to colonic segments in patients with proctitis (R), Lt side colitis (L), pancolitis (P) and remission (RE). A total of 36 UC patients who underwent endoscopic evaluation between March 2023 and July 2024 were finally enrolled. The patients were allocated by disease extent to R (n=18,50.0%), L (n=5, 13.9%), P (n=3, 8.3%) and RE (n=10, 27.8%) groups. The mucosal microbiomes of ascending colon (AC), descending colon (DC), and rectum (RT) were evaluated. Total genomic DNA was extracted from the samples, and the character of the microbiome was determined by sequencing the V3-V4 region of the 16s rRNA gene (Miseq). Results The mucosal microbiome of the RE group exhibited similar alpha diversity across all colonic segments. In contrast, the R and L groups demonstrated increased species richness and evenness in the rectum and descending colon compared to the RE group. However, P group showed consistently low species richness and diversity across all segments compared to the RE group, and with no significant segmental differences within the group. Taxonomic composition analysis at the genus level revealed that segments with active inflammation were enriched with Clostridium innocuum, Campylobacteria, Fusobacteriota, and Alistipes, while Serratia, Rhizorhapis, and Sediminibacterium were depleted compared to the RE group. In beta diversity, the microbiome composition was not different depending on the segment in the RE and P group. In the R and L group, the microbiome composition of the rectum and descending colon were similar but separate from the ascending colon. Conclusion In summary, the mucosal microbiome in inflamed segments differs from that in segments with remission. In cases of pancolitis, the mucosal microbiome exhibited no segmental differences but demonstrated a distinct overall pattern compared to other groups. Evaluating the segmental mucosal microbiome may offer valuable clinical insights into the progression of UC and aid in predicting the extent of inflammation.

Survival Outcomes Associated With Radiological Progressive Disease Subtypes in Patients With Atezolizumab and Bevacizumab-Treated HCC.

To assess the relationship between survival outcomes and subtypes of radiological progressive disease (PD) in patients with hepatocellular carcinoma (HCC) treated with atezolizumab and bevacizumab (Atezo/Bev). A total of 462 patients with Atezo/Bev-treated HCC diagnosed with radiological PD during follow-up were enrolled. PD was classified into three categories: progression or emergence of intrahepatic lesions (PD-IH), macroscopic vascular invasion (PD-MVI), and extrahepatic spread lesions (PD-EHS). We defined PD-multiple as the presence of two or more PD categories. Subsequent analysis was categorized into the "PD-IH or PD-EHS" and "PD-MVI or PD-multiple" groups. The median progression-free survival (PFS) durations for patients with PD-IH, PD-MVI, PD-EHS, and PD-multiple were 5.3, 3.2, 3.9, and 3.5 months (p = 0.003). Patients with "PD-IH or PD-EHS" and "PD-MVI or PD-multiple" had median PFS of 5.2 and 3.5 months (p < 0.001). Median overall survival (OS) for PD-IH, PD-MVI, PD-EHS, and PD-multiple was 22.3, 15.1, 19.4, and 14.2 months (p = 0.002). The OS for patients with "PD-IH or PD-EHS" and "PD-MVI or PD-multiple" was 21.4 and 14.5 months (p < 0.001). Multivariate analysis demonstrated that ECOG-PS ≥ 1 (hazard ratio (HR), 1.508), α-fetoprotein levels ≥ 100 ng/mL (HR, 1.293), albumin-bilirubin grade ≥ 2 (HR, 1.573), liver cirrhosis (HR, 1.361), and PD subtypes PD-MVI or PD-multiple (HR, 1.735) were independently associated with OS. Patients with HCC undergoing Atezo/Bev treatment, diagnosed with PD-multiple (not solely based on IH or EHS) or PD-MVI, experienced poor prognosis, specifically in terms of OS.

P0851 Managing tuberculosis infection risk in inflammatory bowel disease patients before biological therapy initiation – Single-center experience

Abstract Background Ulcerative colitis (UC) and Crohn’s disease (CD) represent distinct subtypes of chronic inflammatory bowel disease (IBD) with increasing global incidence. New studies suggest that adopting a top-down strategy, characterized by the early initiation of biological therapy, yields superior disease control and enhances clinical outcomes. The utilization of biological therapy increases the susceptibility to opportunistic infections, such as tuberculosis (TB). We are witnessing an unexpected rise in TB infection incidence worldwide, posing additional challenges in managing patients with IBD. This study aimed to determine the prevalence of TB infection, including active TB infection (ATBI) and latent TB infection (LTBI), among IBD patients prior to initiating biological therapy. Additionally, the study aimed to investigate whether the presence of TB infection influenced the selection of biological drug. Methods In this retrospective study, we examined a cohort of 558 patients diagnosed with IBD, comprising 303 males and 255 females, with a mean age of 42±16 years. These patients were initiated on biological therapy between 2017 and 2024 at the Clinic for Gastroenterohepatology, University Clinical Center of Serbia. Screening for TB infection was conducted using interferon gamma release assey (IGRA) tests and chest X-rays for all participants. Results Out of the total 558 patients in the study, 306 of them were diagnosed with CD, while 252 had UC. A mere 19 patients (3,41%) tested positive for IGRA, with 2 patients (0,36%) having indeterminate results. Among the IGRA-positive patients, 18 of them were identified as having LTBI, while only 1 presented with ATBI. In the IGRA-positive group, 12 patients were diagnosed with UC, while 7 had CD. Among those with LTBI, the majority, comprising 15 patients, were started on vedolizumab therapy, with 13 of them receiving antituberculosis prophylaxis before initiation of the biological drug. One patient had ATBI and underwent full antituberculosis treatment before starting infliximab therapy. In cases where patients had indeterminate IGRA test results, additional diagnostic assessments were carried out, and TB infection was ruled out before commencing infliximab therapy. Notably, none of the patients in whom biological therapy was commenced developed ATBI during the follow-up period. Conclusion Screening for TB infection is imperative before initiating biological therapy in IBD patients, as the reactivation of LTBI caused by immunosuppression can lead to fatal complications. In our study, the majority of patients with LTBI were treated with vedolizumab, which has demonstrated exceptional safety profile for patients with LTBI, even in the absence of previous antituberculosis prophylaxis.

Hepatocellular CMPK2 promotes the development of metabolic dysfunction-associated steatohepatitis.

Metabolic dysfunction-associated steatohepatitis (MASH), a progressive subtype of metabolic dysfunction-associated steatotic liver disease (MASLD), has limited pharmacological treatment options. Therefore, we aimed to identify novel therapeutic targets. The Gene Expression Omnibus (GEO) database and human liver tissues obtained from patients with MASH were used to identify differentially expressed genes in MASH. The functional role of cytidine/uridine monophosphate kinase 2 (CMPK2) was assessed in mice with hepatocyte-specific overexpression, conditional knockout mice, and several murine MASH models. CMPK2 inhibitors were discovered through surface plasmon resonance imaging coupled with indirect enzyme activity detection. CMPK2, a critical enzyme involved in mitochondrial DNA synthesis, exhibited significant upregulation in the livers of obese individuals with MASH and mice with diet-induced MASH. Hepatocyte-specific Cmpk2 deletion substantially mitigated liver injury, inflammation, and fibrosis in mice. Inhibition of CMPK2, either through genetic manipulation or pharmacological intervention with nordihydroguaiaretic acid (NDGA), suppressed NOD-like receptor family pyrin domain containing 3 (Nlrp3) inflammasome activation and subsequent hepatic pyroptosis. Furthermore, NDGA alleviated diet-induced metabolic disorders, inflammation, and fibrosis in vivo. These findings establish CMPK2 as a critical mediator in the progression from metabolic dysfunction-associated steatotic liver (MASL) to MASH and highlight its potential as a therapeutic target for metabolic diseases. CMPK2 exhibits upregulated in the MASH stage but not in the early stages of MASLD. Our study demonstrated that diet-induced MASH phenotypes, including liver injury, inflammation, and fibrosis were alleviated in hepatocyte-specific Cmpk2-knockout mice. These findings suggest that CMPK2 serve as a critical link in the progression of steatotic liver to steatohepatitis, offering novel mechanistic insights MASH development. Furthermore, this discovery identified CMPK2 as a promising target for the development of therapeutic drugs for MASH.

P0168 Differential expression of enteric glial cell markers in inflamed and non-inflamed tissue of patients with Ulcerative Colitis during clinically active disease and remission

Abstract Background Ulcerative colitis (UC), a major subtype of inflammatory bowel disease, is characterized by chronic inflammation of the colon and rectum. Enteric glial cells (EGC) play a crucial role in gut barrier maintenance and may contribute to UC pathophysiology. This study aims to investigate EGC and their associated proteins in patients during clinically active disease and in clinical remission following anti-inflammatory treatment. Methods Colonic or rectal biopsies were obtained at baseline from both inflamed and non-inflamed segments in 14 patients with clinically active UC initiating anti-inflammatory therapy and from the same segments when the patients were in clinical and endoscopic remission during follow-up. As control groups, colonic biopsies from 16 patients with irritable bowel syndrome (IBS)-Mixed and 16 healthy controls were included. Immunofluorescence staining assessed two EGC markers: the glial fibrillary acidic protein (GFAP+) and S100 calcium-binding protein B (S100β+). Relative estimates of inflammatory proteins in biopsies were analysed using Olink technology, In vitro, the EGC cell line CRL-2690 was exposed to interleukins (IL)-4, IL-6, and IL-18 at varying concentrations and durations, with GFAP expression analysed by western blot. Results In patients with UC, immunofluorescence analysis revealed significantly higher GFAP+ and S100β+ EGC counts in inflamed biopsies during active disease compared to macroscopically non-inflamed biopsies obtained during clinical and endoscopic remission. Compared to baseline biopsies from macroscopically non-inflamed mucosa, GFAP expression significantly decreased during follow-up (Figure 1), while S100β levels remained unaltered. Regardless of mucosal inflammatory status, patients with UC exhibited higher EGC counts than IBS-mixed patients and healthy controls. Our findings also showed a significant upregulation of EGC-associated pro-inflammatory proteins, such as IL-6, IL-8 and TNF in inflamed biopsies from patients with clinically active UC, compared to biopsies obtained when the patients were in remission as well as to healthy controls. Notably, these protein levels decreased during inactive UC, approaching levels observed in the healthy controls. In vitro, IL-6 upregulated GFAP expression dose- and time-dependently, while IL-4 and IL-18 induced expression in a less predictable pattern. Conclusion Elevated EGC counts, and pro-inflammatory proteins characterise inflamed mucosa in UC during clinical active disease highlighting their potential roles in UC related inflammation. These findings may suggest that EGC could be explored as a potential therapeutic target and may contribute to the discovery of novel biomarkers for disease monitoring.

P0026 Multi-omics phenotyping characterizes perturbed molecular divergence underlying different clinical scenarios of under-treatment inflammatory bowel disease

Abstract Background Clinically heterogeneous spectrum and molecular phenotypes of inflammatory bowel disease (IBD) remain to be comprehensively elucidated. Methods This study set out to explore the serum molecular profiles (I) underlying the disease subtypes, in association with (II) elevated fecal calprotectin and (III) disease activity states, (IV) upon treatment escalation, and (V) in patients who needed treatment escalation. The serum proteome, metabolome, and lipidome of 75 IBD patients were profiled. Single- and multi-omic data analysis was performed to determine differential analytes and integrative biosignatures. Results In (I), chronic inflammation along with tight junction disturbances promoting impaired barrier function might underly the differences between IBD subtypes. Compared to Crohn’s disease, ulcerative colitis showed a lower abundance of hypoxanthine correlating negatively with cell adhesion-related proteins of ITGB2, CD200, PRELP, and positively with lysophosphatidylcholines [LPC(16:0-SN1), LPC(O-18:1)]; and up-regulation of bile acids and FGF19. For (II), while proteins related to inflammatory-promoting signaling pathways, sphingomyelins (SMs), and LPC were increased; bile acids, amino acids, and triacylglycerols (TGs) were decreased. Not only in (II), TGs were a major class change in (III), whose abundance was consistently higher in the active state. This, along with decreased SMs levels, correlated strongly with the up-regulation of inflammatory-inducing proteins (CXCL9, CXCL1, ITGB2, IL17A). In (IV), treatment escalation promoted augmentations of signaling pathways-related proteins and triggered significant metabolome responses, especially decreased amino acids. Noticeably, the levels of TGs were consistently increased in the post-treatment escalation group, thereby suggesting treatment response with relation to (III). For (V), down-regulation of signaling pathways-related proteins and higher levels of SMs and PCs was observed. A consistent decrease of TGs in patients who needed treatment escalation along with lower level of TGs in elevated calprotectin group could imply for the inflammatory level. Conclusion This study provided the multi-ome profiles of IBD patients in five clinical scenarios and laid a foundation for opening new doors for disease management and treatment. References M. Agrawal, K.H. Allin, F. Petralia, J.-F. Colombel, T. Jess, Multiomics to elucidate inflammatory bowel disease risk factors and pathways, Nature Reviews Gastroenterology &amp; Hepatology 19(6) (2022) 399-409.

OP08 Histology-derived cell-cell interaction networks, linked to transcriptome profiling, define path to resolution in UC

Abstract Background Despite therapeutic advancements, the cellular and molecular heterogeneity of UC hinders effective treatment. We applied machine learning-based approaches to whole scanned images (WSI) of H&amp;E stained slides prepared from formalin-fixed, paraffin embedded intestinal mucosal biopsies from the etrolizumab clinical trials, aiming to identify cellular interactions and molecular signatures that define disease subtypes and treatment responses. We hypothesised that the cellular-basis of non-response could be derived by linking cell-cell relationships with paired transcriptomic profiles. Methods We analysed 2297 WSI paired with molecular data from 1272 UC patients from the etrolizumab PhIII clinical trial including various cohorts, treatment arms, and timepoints. Sections were segmented using a convolutional neural network, extracting 8 cell types and their coordinates. For each WSI, ~500 interaction features from an overlaid network were engineered using our developed software, CellMaskProfiler. The topology of the cell-cell interaction networks allowed clustering of WSI with similar profiles. Cell count features and clinical data enabled the annotation of clusters. Paired molecular data was used to perform differential expression. Results Using the cell-cell topology, we identified 7 distinct UC subtypes characterised by specific cellular interactions, molecular profiles and clinical features resolving into, intact, and neutrophil, eosinophil, plasma, or lymphocyte dominated profiles. WSI with a high proportion of epithelial cells and more connections between epithelial cells types were strongly associated with lower Robarts Histology Index, indicating milder disease or intact epithelium, while strong interaction with neutrophils showed the opposite effect. Looking at different treatment arms allowed to relate baseline biopsy WSI characteristics with probability of treatment response highlighting the probability of response based on initial WSI subtype. Using the cellular topology allowed to characterise the transition between diseased and healthier clusters. Adjusting for cellularity effects and comparing baseline data for response at induction revealed genes characterising response. Looking at transitions between clusters revealed similar genes. Conclusion This study introduces a novel feature engineering approach to unravel the cellular and molecular complexity of WSI in UC. The findings underscore the significance of cellular interactions in UC pathology and provide new insights into the molecular mechanisms underlying different disease states and demonstrates that the structural topology (form) derived from WSI can be effectively linked to the transcriptomic profiles (function) to define probability of remission.

68Ga]Ga‑FAPI PET/CT in the evaluation of Langerhans cell histiocytosis: comparison with [18F]FDG PET/CT.

We aimed to explore the value of [68Ga]Ga‑FAPI PET/CT for the evaluation of Langerhans cell histiocytosis (LCH) in comparison with [18F]FDG PET/CT. Thirty-two patients pathologically diagnosed with LCH were enrolled in this study. [68Ga]Ga‑FAPI and [18F]FDG PET/CT were performed within 1 week to identify disease extent and status. The detectability and intensity of the involved organs were compared between these two tracers. Thirty patients had active disease while two had non-active disease. In patients with active disease, the most commonly involved organ was bone (27/30), and [68Ga]Ga-FAPI PET/CT detected more osseous lesions (106/106) than [18F]FDG PET/CT (52/106). [68Ga]Ga-FAPI also identified liver, skin, and salivary gland involvement, which were often missed by [18F]FDG. Although both tracers detected thymus and pituitary gland involvement, [68Ga]Ga-FAPI demonstrated higher image contrast and more diagnostic confidence. Lymph node involvement, however, was not visualized by [68Ga]Ga-FAPI. Due to the superior sensitivity of [68Ga]Ga-FAPI, approximately 30% (10/30) of patients experienced reclassification in disease status or subtype. Furthermore, [68Ga]Ga‑FAPI appeared to be advantageous in response assessment. [68Ga]Ga-FAPI PET/CT outperforms [18F]FDG PET/CT in detecting osseous and extra-nodal lesions in LCH, providing a valuable tool for precise disease evaluation and treatment planning.

Emerging Connections Between Inflammatory Bowel Disease Subtypes and Sequential Changes in Total Serum Bilirubin.

Emerging Connections Between Inflammatory Bowel Disease Subtypes and Sequential Changes in Total Serum Bilirubin.

Diagnostic Cut-Off Values Based on Lipid Layer Pattern for Dry Eye Disease Subtypes Assessment.

Background: The aim of the present study was to establish a cut-off value of the Lipid Layer Pattern (LLP) between participants with different subtypes of Dry Eye Disease (DED) including Deficient Dry Eye (ADDE), Evaporative Dry Eye (EDE), and Mixed Dry Eye (MDE). Methods: 240 participants diagnosed with DED according to the Tear Film and Ocular Surface Society in the Dry Eye Workshop II guidelines were included in the study. Tear Meniscus Height (TMH) using the Tearscope illumination and Meibomian Gland Loss Area (MGLA) using the Keratograph 5M were assessed to categorize the participants into an ADDE group, EDE group, or MDE group. Then, the LLP was assessed using the Tearscope following the Guillon (LLP-G) and Colour (LLP-C) schemes. Results: Receiver Operating Characteristics (ROC) showed that both LLP-G and LLP-C have no diagnostic potential in distinguishing between ADDE and EDE participants (both p ≥ 0.724). However, to differentiate the ADDE participants from the MDE, ROC procedures showed a good diagnostic potential with cut-off values of Closed Meshwork-Wave (AUC ± SD = 0.609 ± 0.049, p = 0.038, sensitivity: 23.9%; specificity: 76.1%) and Grey-White (AUC ± SD = 0.611 ± 0.050, p = 0.034, sensitivity: 40.7%; specificity: 73.9%) for LLP-G and LLP-C, respectively. Also, a significant potential to distinguish between the EDE from MDE participants was found, with cut-off values of Closed Meshwork (AUC ± SD = 0.604 ± 0.049, p = 0.043, sensitivity: 40.8%; specificity: 76.1%) and Grey-White (AUC ± SD = 0.604 ± 0.051, p = 0.038, sensitivity: 44.7%; specificity: 73.9%) for LLP-G and LLP-C, respectively. Conclusions: Using the Tearscope, both LLP-G and LLP-C has diagnostic potential to distinguish MDE participants from the other subtypes of DED.

Tumour DNA methylation markers associated with breast cancer survival: a replication study

BackgroundTumour DNA methylation has been investigated as a potential marker for breast cancer survival, but findings often lack replication across studies.MethodsThis study sought to replicate previously reported associations for individual CpG sites and multi-CpG signatures using an Australian sample of 425 women with breast cancer from the Melbourne Collaborative Cohort Study (MCCS). Candidate methylation sites (N = 22) and signatures (N = 3) potentially associated with breast cancer survival were identified from five prior studies that used The Cancer Genome Atlas (TCGA) methylation dataset, which shares key characteristics with the MCCS: comparable sample size, tissue type (formalin-fixed paraffin-embedded; FFPE), technology (Illumina HumanMethylation450 array), and participant characteristics (age, ancestry, and disease subtype and severity). Cox proportional hazard regression analyses were conducted to assess associations between these markers and both breast cancer-specific survival and overall survival, adjusting for relevant participant characteristics.ResultsOur findings revealed partial replication for both individual CpG sites (9 out of 22) and multi-CpG signatures (2 out of 3). These associations were maintained after adjustment for participant characteristics and were stronger for breast cancer-specific mortality than for overall mortality. In fully-adjusted models, strong associations were observed for a CpG in PRAC2 (per standard deviation [SD], HR = 1.67, 95%CI: 1.24–2.25) and a signature based on 28 CpGs developed using elastic net (per SD, HR = 1.48, 95%CI: 1.09–2.00).ConclusionsWhile further studies are needed to confirm and expand on these findings, our study suggests that DNA methylation markers hold promise for improving breast cancer prognostication.

Very-Low-Absorbable Geraniol for the Treatment of Irritable Bowel Syndrome: A "Real-World" Open-Label Study on 1585 Patients.

The objective of this study was to evaluate the efficacy of a very-low-absorbable geraniol formulation, administered as a food supplement, in patients with irritable bowel syndrome (IBS) in a real-world setting in Italy. This open-label study was conducted in Italy on patients diagnosed with IBS and treated for 4 weeks with 240 mg/day of Palmarosa essential oil, absorbed on 960 mg of ginger root powder to obtain a very-low-absorbable geraniol formulation. Baseline characteristics, including demographic and symptoms were recorded using the IBS Severity Scoring System (IBS-SSS). After 28 ± 7 days, the patients were asked to complete the IBS-SSS questionnaire again. The primary objective was to confirm the effects of a very-low-absorbable geraniol formulation on self-reported symptoms of IBS and the quality of life of affected individuals. The secondary objective was to confirm the effect of the treatment on the different IBS subtypes. A total of 1585 patients were included in the study, with a mean age of 44.8 years and 56.4% women. Following the 4-week supplementation period, significant decreases were observed in the patients' IBS-SSS (-67.9%) and all the primary IBS symptoms, such as abdominal distention (-82.3%), unsatisfaction with bowel habits (-46.2%), and interference with quality of life (QoL) (-64.9%) (all p < 0.01). The patients' stool type improved significantly. Treatment was effective in all IBS subtypes. Treatment with very-low-adsorbable geraniol food supplement was associated with improvements in symptoms and bowel habits in all IBS subtypes in a real-world setting in Italy. These findings support the use of geraniol as an effective option for patients with IBS regardless of the disease subtype.