Normal Tissue Regions Research Articles

Liver tumor segmentation method combining multi-axis attention and conditional generative adversarial networks.

In modern medical imaging-assisted therapies, manual annotation is commonly employed for liver and tumor segmentation in abdominal CT images. However, this approach suffers from low efficiency and poor accuracy. With the development of deep learning, automatic liver tumor segmentation algorithms based on neural networks have emerged, for the improvement of the work efficiency. However, existing liver tumor segmentation algorithms still have several limitations: (1) they often encounter the common issue of class imbalance in liver tumor segmentation tasks, where the tumor region is significantly smaller than the normal tissue region, causing models to predict more negative samples and neglect the tumor region; (2) they fail to adequately consider feature fusion between global contexts, leading to the loss of crucial information; (3) they exhibit weak perception of local details such as fuzzy boundaries, irregular shapes, and small lesions, thereby failing to capture important features. To address these issues, we propose a Multi-Axis Attention Conditional Generative Adversarial Network, referred to as MA-cGAN. Firstly, we propose the Multi-Axis attention mechanism (MA) that projects three-dimensional CT images along different axes to extract two-dimensional features. The features from different axes are then fused by using learnable factors to capture key information from different directions. Secondly, the MA is incorporated into a U-shaped segmentation network as the generator to enhance its ability to extract detailed features. Thirdly, a conditional generative adversarial network is built by combining a discriminator and a generator to enhance the stability and accuracy of the generator's segmentation results. The MA-cGAN was trained and tested on the LiTS public dataset for the liver and tumor segmentation challenge. Experimental results show that MA-cGAN improves the Dice coefficient, Hausdorff distance, average surface distance, and other metrics compared to the state-of-the-art segmentation models. The segmented liver and tumor models have clear edges, fewer false positive regions, and are closer to the true labels, which plays an active role in medical adjuvant therapy. The source code with our proposed model are available at https://github.com/jhliao0525/MA-cGAN.git.

Shear Wave Elastography in Evaluating the Efficacy of Liver Microwave Ablation: An Experimental Study

Background: The effect of microwave ablation on small hepatocellular carcinoma is not significantly different from that of surgical resection. Accurate assessment of the ablation effect is crucial for ensuring the safety and effectiveness of hyperthermia. Objectives: This study aimed to explore the feasibility and accuracy of supersonic shear wave elastography (SWE) in quantitatively evaluating the microwave ablation margin of the liver. Materials and Methods: Three surgeons were each randomly assigned 4 Wuzhishan miniature pigs (WZSPs) to perform 12 ablation procedures at an ablation power of 40 W. Based on the ablation time, the lesions were divided into 15, 30, and 60-second groups. Immediately after ablation, SWE and modulus measurements were performed 5 times for each ablation lesion. The SWE data were expressed as mean ± standard deviation. Within-group and between-group comparisons were made using repeated measures analysis of variance. Results: A total of 144 effective ablations and 131 effective pathological results were obtained. Within the same ablation time, the elastic modulus increased in the surrounding normal tissue, ablation margin, and ablation center regions in a stepwise manner (P < 0.01). However, in the ablation center region, the elastic modulus decreased in a stepwise manner with the shortening of ablation time [60 s (97.16 ± 14.58 kPa) > 30 s (77.84 ± 9.64 kPa) > 15 s (38.92 ± 3.12 kPa)], with statistically significant differences (F = 2,131.832, P < 0.01). The elastic modulus of the ablation margin region at different ablation times remained between 22.68 - 23.56 kPa. Conclusion: The elastic modulus range in the ablation margin region after microwave ablation is relatively fixed. Shear wave elastography aids in the quantitative evaluation of the ablation margin region of the liver and has high practical value in monitoring and evaluating ultrasound ablation.

Methylation-regulated tumor suppressor gene PDE7B promotes HCC invasion and metastasis through the PI3K/AKT signaling pathway.

Hepatocellular carcinoma (HCC) has a high mortality rate, and the mechanisms underlying tumor development and progression remain unclear. However, inactivated tumor suppressor genes might play key roles. DNA methylation is a critical regulatory mechanism for inactivating tumor suppressor genes in HCC. Therefore, this study investigated methylation-related tumor suppressors in HCC to identify potential biomarkers and therapeutic targets. We assessed genome-wide DNA methylation in HCC using whole genome bisulfite sequencing (WGBS) and RNA sequencing, respectively, and identified the differential expression of methylation-related genes, and finally screened phosphodiesterase 7B (PDE7B) for the study. The correlation between PDE7B expression and clinical features was then assessed. We then analyzed the changes of PDE7B expression in HCC cells before and after DNA methyltransferase inhibitor treatment by MassArray nucleic acid mass spectrometry. Furthermore, HCC cell lines overexpressing PDE7B were constructed to investigate its effect on HCC cell function. Finally, GO and KEGG were applied for the enrichment analysis of PDE7B-related pathways, and their effects on the expression of pathway proteins and EMT-related factors in HCC cells were preliminarily explored. HCC exhibited a genome-wide hypomethylation pattern. We screened 713 hypomethylated and 362 hypermethylated mCG regions in HCC and adjacent normal tissues. GO analysis showed that the main molecular functions of hypermethylation and hypomethylation were "DNA-binding transcriptional activator activity" and "structural component of ribosomes", respectively, whereas KEGG analysis showed that they were enriched in "bile secretion" and "Ras-associated protein-1 (Rap1) signaling pathway", respectively. PDE7B expression was significantly down-regulated in HCC tissues, and this low expression was negatively correlated with recurrence and prognosis of HCC. In addition, DNA methylation regulates PDE7B expression in HCC. On the contrary, overexpression of PDE7B inhibited tumor proliferation and metastasis in vitro. In addition, PDE7B-related genes were mainly enriched in the PI3K/ATK signaling pathway, and PDE7B overexpression inhibited the progression of PI3K/ATK signaling pathway-related proteins and EMT. PDE7B expression in HCC may be regulated by promoter methylation. PDE7B can regulate the EMT process in HCC cells through the PI3K/AKT pathway, which in turn affects HCC metastasis and invasion.

Experimental investigation of photothermal conversion and thermal conductivity of broadband absorbing gold nanoblackbodies and graphene oxide nanoparticles for plasmonic photothermal cancer therapy

The choice of nanomaterial is crucial for achieving desired thermal ablation of a tumors, considering nanoparticle dose/concentration and near-infrared (NIR) irradiation parameters during plasmonic photothermal therapy (PPTT). Most studies prioritize use of nanoparticles with high photothermal conversion efficiency (η) for PPTT without emphasizing the thermal conductivity (κ) within a solid tumor. Herein, two broadband absorbing nanoparticles, i.e., gold nanoblackbodies (AuNBs) and graphene oxide (GO) nanoparticles, are compared for η under similar NIR irradiation and nanoparticles dose. Further, the effect of photothermal conversion and κ are studied through tumor-tissue mimicking phantoms mimicking invasive ductal carcinoma embedded with nanoparticles, surrounded by normal tissue region. Under similar conditions, η is higher in AuNBs (73.2%) than GO nanoparticles (63.7%), while κ of GO nanoparticles embedded phantom (0.62 W/mK) surpasses AuNBs embedded phantom (0.59 W/mK) by 5.08%. Photothermal response through tumor-tissue mimicking phantoms shows that GO nanoparticles achieve higher temperatures within the entire tumor region than AuNBs due to their high κ, despite having lower η. This finding suggests that η is not the sole parameter, and it is the coupled effect of η and κ for the choice/suitability of a nanoparticle to attain desired temperatures throughout the tumor for PPTT.

High clonal diversity and spatial genetic admixture in early prostate cancer and surrounding normal tissue

Somatic copy number alterations (SCNAs) are pervasive in advanced human cancers, but their prevalence and spatial distribution in early-stage, localized tumors and their surrounding normal tissues are poorly characterized. Here, we perform multi-region, single-cell DNA sequencing to characterize the SCNA landscape across tumor-rich and normal tissue in two male patients with localized prostate cancer. We identify two distinct karyotypes: ‘pseudo-diploid’ cells harboring few SCNAs and highly aneuploid cells. Pseudo-diploid cells form numerous small-sized subclones ranging from highly spatially localized to broadly spread subclones. In contrast, aneuploid cells do not form subclones and are detected throughout the prostate, including normal tissue regions. Highly localized pseudo-diploid subclones are confined within tumor-rich regions and carry deletions in multiple tumor-suppressor genes. Our study reveals that SCNAs are widespread in normal and tumor regions across the prostate in localized prostate cancer patients and suggests that a subset of pseudo-diploid cells drive tumorigenesis in the aging prostate.

Abstract 7028: Optimization of DNA extraction methods and DNA methylation array quality from FFPE prostate tumor tissues to identify DNA methylation biomarkers

Abstract Prostate cancer is the second leading cause of cancer related deaths among men in the United States. Health disparities in men diagnosed with prostate cancer are observed between patients of African and those of European ancestry. To elucidate molecular mechanisms behind prostate tumorigenesis and racial disparities, researchers have identified genetic alterations such as gene fusions, deletions, and mutations that occur at varying frequencies among different ethnic groups. However, results are inconsistent across studies, suggesting that racial disparities may be multifactorial. Our approach to further understand the mechanisms underlying racial disparities is to compare DNA methylation profiles between cancers from different ethnic groups using formalin-fixed paraffin-embedded (FFPE) prostate cancer tissue samples. For this study, FFPE prostate cancer specimens from 32 Black men and 30 age-matched White men were obtained in the form of tissue sections on glass slides. To profile DNA methylation in these samples, tumor and normal tissue regions (in a subset of cases) were outlined on a hematoxylin & eosin (H&E) reference slide and the respective regions were macro-dissected off the corresponding unstained slides from the same block. The areas of macro-dissected regions were measured using ImageJ, and the number of slides were taken to normalize DNA yield. Next, we compared magnetic bead and column based DNA extraction methods. We evaluated them on their relative DNA yield per tissue size, and identified that the bead-based extraction method had both higher and more consistent DNA yield. In total, we used 95 unique tissue regions to profile global DNA methylation levels using Illumina Epic array. We found that a subset of samples harbored an aberrant DNA methylation pattern. Therefore, we extracted DNA from additional FFPE tissue slides using the bead-based method and repeated the assay with a larger amount of DNA, successfully rescuing the majority of the aberrant samples. Next, we will evaluate associations between DNA methylation levels in different ethnic groups, clinicopathological features, and follow-up data to help identify potential DNA methylation biomarkers and therapeutic targets linked to prostate cancer from different ethnicities. Citation Format: Colton Duran Stensrud, Leonardo Gonzalez-Smith, Claire Stevens, Huan Cao, Jenaye Mack, Olumide Arigbede, Daniel Weisenberger, Sarah G. Buxbaum, Sara M. Falzarano, Suhn Rhie. Optimization of DNA extraction methods and DNA methylation array quality from FFPE prostate tumor tissues to identify DNA methylation biomarkers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7028.

Abstract 5483: Macrospatial ecotype heterogeneity in early-stage breast cancer identified by spatial nucleus barcoding

Abstract Ductal carcinoma in situ (DCIS), the predominant form of early-stage breast cancer and a precursor to invasive breast cancer, remains inadequately understood in terms of spatial microenvironment reprogramming. In this study, we developed a Spatial Nucleus Barcoding (SNuBar) approach for single-nucleus RNA sequencing (SNuBar-RNA) that utilizes a single oligonucleotide adaptor to preserve spatial information. The accuracy and scalability of SNuBar was validated using cell line mixture experiments. We then applied SNuBar-RNA to investigate the macrospatial ecotype heterogeneity in early-stage breast cancer. By analyzing single cells from different macrodissected spatial regions in normal breast tissues, matched normal and DCIS tissues, we identified 13 major cell types and 43 distinct cell states, co-localizing in diverse combinations across four primary topographic areas (adipose-enriched, normal epithelial, activated stroma and tumor). Spatial analysis revealed the reprogramming of vascular and myoepithelial cells in localized zones near premalignant cells, while fibroblasts, T-cells, and myeloid cells exhibited reprogramming across broader fields. Validation using the MERSCOPE assay confirmed stromal heterogeneity between tumor and normal areas. Collectively, our data shows that many diverse cell types are reprogrammed in spatially-defined areas in the early-stage breast cancer microenvironment, relative to the normal breast tissue regions. Citation Format: Kaile Wang, Zhenna Xiao, Yiyun Lin, Runmin Wei, Shanshan Bai, Jie Ye, Rui Ye, Min Hu, Aatish Thennavan, Emi Sei, Alastair Thompson, Savitri Krishnamurthy, Nicholas Navin. Macrospatial ecotype heterogeneity in early-stage breast cancer identified by spatial nucleus barcoding [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 5483.

P160 Integrative transcriptional analysis to identify predictive biomarkers for IBD severity and subtyping

Abstract Background Conventional diagnostic strategies for Inflammatory Bowel Diseases (IBD) heavily rely on invasive biopsy procedures and subjective questionnaires, thereby introducing intrinsic limitations in the context of precision medicine. By understanding the molecular mechanisms behind IBD disease biology, we aim to identify certain signalling pathways relevant for epithelial tissue regeneration under chronic inflammation conditions. Methods To address these limitations, we compared gene expression in a large cohort of IBD patients, utilizing the IBD Plexus database (Crohn’s & Colitis Foundation, US) as well as other publicly available datasets. We further validated our results derived from the transcriptional programs by using patient derived advanced 3D-organoid models (Hospital of Nottingham, UK). These models, derived from ulcerated and normal tissue regions of IBD patients, serve as a sophisticated platform for robust validation, bridging the gap between molecular insights and clinical applications. Results Our primary objective was to discern putative biomarkers that are capable of effectively stratifying IBD subtypes. We further identified the genes that were upregulated exclusively in each subtypes and the gene signature that is elevated in inflamed samples compared to non-inflamed samples. We found that either the chemokine and interleukin (IL-17) signalling pathways which have known roles in recruitment of immune cells thereby perpetuating chronic inflammation and several key chemokines influence the integrity of the epithelial barrier function and regeneration. We are also analysing non-invasive biomarkers by using samples such as blood, plasma and stool and employ machine learning methods to correlate with established clinical indicators such as sCDAI, PRO2, PRO3, and MAYO categories. Conclusion Hence, in identifying specific targets we can introduce novel therapeutic strategies to mitigate inflammation and restore integrity of the gut epithelial barrier. With this research we aim to improve the current IBD diagnostics landscape as well propels it towards precision medicine by offering precise subtyping and predictive markers for disease severity, ultimately contributing to more personalised and effective therapeutics interventions.

Abstract C059: Optimization of DNA extraction methods from FFPE prostate tumor tissues of Black men to identify DNA methylation biomarkers

Abstract Prostate cancer is the second leading cause of cancer related deaths among men in the United States, and health disparities in those diagnosed with prostate cancer are observed between those of African and European ancestry. Our approach to understand the mechanisms underlying these differences is to compare DNA methylation profiles between samples using formalin-fixed parafin-embedded (FFPE) prostate tumor tissues. To profile DNA methylation from FFPE prostate tumor tissues, we tested magnetic bead and column based DNA extraction methods, and evaluated them on their total DNA yield. Additionally, we assessed the relationship between DNA yield and clinical features. Prostate specimens for this study were obtained from the University of Florida in the form of FFPE tissue slides. Both tumor and normal tissue regions were outlined on an H&E reference slide and the respective regions were macro-dissected from unstained slides of the same block. The areas of macro-dissected regions were measured using ImageJ, and the number of slides were taken to further normalize DNA yield. Magnetic bead and column based FFPE DNA extraction kits were used to isolate DNA, and the relative amounts of the isolated DNA per tissue size were evaluated. In addition, various clinical features of the samples were compared to determine if their presence influences DNA yield. Preliminary data suggests that total DNA extracted from both bead and column based methods are relatively comparable, but bead based methods yielded better quality of DNA. We observed minimal correlation between DNA yield and grade group. We detected a potential association between cribriform pattern and DNA yield. We are in the process of profiling DNA methylation using the isolated DNA. Successful completion of this study will identify DNA methylation biomarkers and further develop diagnostic and therapeutic tools for prostate cancer patients from African ancestry. Citation Format: Colton Stensrud, Claire Stevens, Leonardo Gonzalez-Smith, Huan Cao, Sarah Buxbaum, Sara Falzarano, Suhn Rhie. Optimization of DNA extraction methods from FFPE prostate tumor tissues of Black men to identify DNA methylation biomarkers [abstract]. In: Proceedings of the 16th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2023 Sep 29-Oct 2;Orlando, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2023;32(12 Suppl):Abstract nr C059.

A comprehensive evaluation of advanced dose calculation algorithms for brain stereotactic radiosurgery.

Accurate dose calculation is important in both target and low dose normal tissue regions for brain stereotactic radiosurgery (SRS). In this study, we aim to evaluate the dosimetric accuracy of the two advanced dose calculation algorithms for brain SRS. Retrospective clinical case study and phantom study were performed. For the clinical study, 138 SRS patient plans (443 targets) were generated using BrainLab Elements Voxel Monte Carlo (VMC). To evaluate the dose calculation accuracy, the plans were exported into Eclipse and recalculated with Acuros XB (AXB) algorithm with identical beam parameters. The calculated dose at the target center (Dref), dose to 95% target volume (D95), and the average dose to target (Dmean) were compared. Also, the distance from the skull was analyzed. For the phantom study, a cylindrical phantom and a head phantom were used, and the delivered dose was measured by an ion chamber and EBT3 film, respectively, at various locations. The measurement was compared with the calculated doses from VMC and AXB. In clinical cases, VMC dose calculations tended to be higher than AXB. It was found that the difference in Dref showed>5% in some cases for smaller volumes<0.3cm3 . Dmean and D95 differences were also higher for small targets. No obvious trend was found between the dose difference and the distance from the skull. In phantom studies, VMC dose was also higher than AXB for smaller targets, and VMC showed better agreement with the measurements than AXB for both point dose and high dose spread. The two advanced calculation algorithms were extensively compared. For brain SRS, AXB sometimes calculates a noticeable lower target dose for small targets than VMC, and VMC tends to have a slightly closer agreement with measurements than AXB.

Global field of view-based pixel-level recognition method for medical images

Artificial intelligence image processing has been of interest to research investigators in tumor identification and determination. Magnetic resonance imaging for clinical detection is the technique of choice for identifying tumors because of its advantages such as accurate localization with tomography in any orientation. Nevertheless, owing to the complexity of the images and the heterogeneity of the tumors, existing methodologies have insufficient field of view and require expensive computations to capture semantic information in the view, rendering them lacking in universality of application. Consequently, this thesis developed a medical image segmentation algorithm based on global field of view attention network (GVANet). It focuses on replacing the original convolution with a transformer structure and views in a larger field-of-view domain to build a global view at each layer, which captures the refined pixel information and category information in the region of interest with fewer parameters so as to address the defective tumor edge segmentation problem. The dissertation exploits the pixel-level information of the input image, the category information of the tumor region and the normal tissue region to segment the MRI image and assign weights to the pixel representatives. This medical image recognition algorithm enables to undertake the ambiguous tumor edge segmentation task with low computational complexity and to maximize the segmentation accuracy and model property. Nearly four thousand MRI images from the Monash University Research Center for Artificial Intelligence were applied for the experiments. The outcome indicates that the approach obtains outstanding classification capability on the data set. Both the mask (IoU) and DSC quality were improved by 7.6% and 6.3% over the strong baseline.

Characterization of the Molecular Events Underlying the Establishment of Axillary Meristem Region in Pepper.

Plant architecture is a major motif of plant diversity, and shoot branching patterns primarily determine the aerial architecture of plants. In this study, we identified an inbred pepper line with fewer lateral branches, 20C1734, which was free of lateral branches at the middle and upper nodes of the main stem with smooth and flat leaf axils. Successive leaf axil sections confirmed that in normal pepper plants, for either node n, Pn (Primordium n) < 1 cm and Pn+1 < 1 cm were the critical periods between the identification of axillary meristems and the establishment of the region, whereas Pn+3 < 1 cm was fully developed and formed a completely new organ. In 20C1734, the normal axillary meristematic tissue region establishment and meristematic cell identity confirmation could not be performed on the axils without axillary buds. Comparative transcriptome analysis revealed that "auxin-activated signaling pathway", "response to auxin", "response to abscisic acid", "auxin biosynthetic process", and the biosynthesis of the terms/pathways, such as "secondary metabolites", were differentially enriched in different types of leaf axils at critical periods of axillary meristem development. The accuracy of RNA-seq was verified using RT-PCR for some genes in the pathway. Several differentially expressed genes (DEGs) related to endogenous phytohormones were targeted, including several genes of the PINs family. The endogenous hormone assay showed extremely high levels of IAA and ABA in leaf axils without axillary buds. ABA content in particular was unusually high. At the same time, there is no regular change in IAA level in this type of leaf axils (normal leaf axils will be accompanied by AM formation and IAA content will be low). Based on this, we speculated that the contents of endogenous hormones IAA and ABA in 20C1734 plant increased sharply, which led to the abnormal expression of genes in related pathways, which affected the formation of Ams in leaf axils in the middle and late vegetative growth period, and finally, nodes without axillary buds and side branches appeared.

3D CEST MRI with an unevenly segmented RF irradiation scheme: A feasibility study in brain tumor imaging.

To integrate 3D CEST EPI with an unevenly segmented RF irradiation module and preliminarily demonstrate it in the clinical setting. A CEST MRI with unevenly segmented RF saturation was implemented, including a long primary RF saturation to induce the steady-state CEST effect, maintained with repetitive short secondary RF irradiation between readouts. This configuration reduces relaxation-induced blur artifacts during acquisition, allowing fast 3D spatial coverage. Numerical simulations were performed to select parameters such as flip angle (FA), short RF saturation duration (Ts2), and the number of readout segments. The sequence was validated experimentally with data from a phantom, healthy volunteers, and a brain tumor patient. Based on the numerical simulation and l-carnosine gel phantom experiment, FA, Ts2, and the number of segments were set to 20°, 0.3 s, and the range from 4 to 8, respectively. The proposed method minimized signal modulation in the human brain images in the kz direction during the acquisition and provided the blur artifacts-free CEST contrast over the whole volume. Additionally, the CEST contrast in the tumor tissue region is higher than in the contralateral normal tissue region. It is feasible to implement a highly accelerated 3D EPI CEST imaging with unevenly segmented RF irradiation.

Partial erosion on under-methylated regions and chromatin reprogramming contribute to oncogene activation in IDH mutant gliomas

BackgroundIDH1/2 hotspot mutations are well known to drive oncogenic mutations in gliomas and are well-defined in the WHO 2021 classification of central nervous system tumors. Specifically, IDH mutations lead to aberrant hypermethylation of under-methylated regions (UMRs) in normal tissues through the disruption of TET enzymes. However, the chromatin reprogramming and transcriptional changes induced by IDH-related hypermethylation in gliomas remain unclear.ResultsHere, we have developed a precise computational framework based on Hidden Markov Model to identify altered methylation states of UMRs at single-base resolution. By applying this framework to whole-genome bisulfite sequencing data from 75 normal brain tissues and 15 IDH mutant glioma tissues, we identified two distinct types of hypermethylated UMRs in IDH mutant gliomas. We named them partially hypermethylated UMRs (phUMRs) and fully hypermethylated UMRs (fhUMRs), respectively. We found that the phUMRs and fhUMRs exhibit distinct genomic features and chromatin states. Genes related to fhUMRs were more likely to be repressed in IDH mutant gliomas. In contrast, genes related to phUMRs were prone to be up-regulated in IDH mutant gliomas. Such activation of phUMR genes is associated with the accumulation of active H3K4me3 and the loss of H3K27me3, as well as H3K36me3 accumulation in gene bodies to maintain gene expression stability. In summary, partial erosion on UMRs was accompanied by locus-specific changes in key chromatin marks, which may contribute to oncogene activation.ConclusionsOur study provides a computational strategy for precise decoding of methylation encroachment patterns in IDH mutant gliomas, revealing potential mechanistic insights into chromatin reprogramming that contribute to oncogenesis.

Abstract 2873: Evolution of immune cell composition and functionality as pancreatic intraepithelial neoplasia progresses to pancreatic ductal adenocarcinoma

Abstract Pancreatic ductal adenocarcinoma (PDAC) is most often diagnosed at an advanced stage. Newly diagnosed patients therefore have a dismal five-year survival rate of 11%. However, PDAC progresses from pre-invasive pancreatic intraepithelial neoplasia (PanIN) over at least a decade. Throughout this transition, the tissue microenvironment becomes increasingly immunosuppressive. Early PanINs may therefore be more amenable to immune-based interception strategies; however, little is known about the pre-malignant lesion immune microenvironment in PDAC. We hypothesized that the identification of the immune landscape of PanINs will elucidate the immuno-dynamic changes that occur during PanIN-to-PDAC progression and identify novel strategies for intercepting PDAC. Here, we use an inducible mouse model to study the evolution of immunosuppression from PanINs to PDAC. We pair spatial molecular profiling of our mouse model with profiling of human tissue in a cohort of patients with normal tissue, chronic pancreatitis, PanIN, and PDAC. To examine the evolution of the immune microenvironment throughout PanIN-to-PDAC progression, we first optimized a tamoxifen-inducible Pdx1-CreERT2 mouse that controls KRASG12D expression and knocks out p53. The impact of KRASG12D expression on the immune cell landscape in PanIN and PDAC lesions was examined by immunohistochemistry (IHC) and RNA in situ hybridization on mouse pancreas. We used imaging mass cytometry (IMC) of 35 immune markers to better classify and quantify the immune cell subtypes. Our analyses thus far reveal increased Tregs as PanINs progress to PDAC. Furthermore, although CD3+ T cells are recruited to tumors, these immune cells are strictly restricted to the immediate edge of the tumor and predominantly consist of Tregs. For our human analyses, FFPE pancreas sections from treatment naïve patients who had undergone surgical resection without neoadjuvant chemotherapy were evaluated. Each section contained regions of normal tissue, chronic pancreatitis, PanIN and PDAC. These were evaluated using IHC, IMC, and spatial transcriptomics to allow us to spatially evaluate the immune populations associated with lesions and PDAC. The relative density and localization of myeloid and lymphoid cell types in both PanIN and PDAC regions revealed an initial influx of CD8+ and CD4+ T cells to PanINs and a progressively immunosuppressive microenvironment in subsequent stages. While PanINs and PDAC both recruited immune cells, the phenotypes of the immune infiltrates were distinct and revealed unique immune pathways that could contribute to immunosuppression as PanINs develop into PDACs. Our proteomic and transcriptomic data from mouse and human pancreas show that mutant KRAS driven premalignant lesions recruit an evolving immune response that readily becomes immunosuppressive as progression to PDAC occurs. Citation Format: Melissa R. Lyman, Jacob T. Mitchell, Luciane T. Kagohara, Benjamin Barrett, Amanda Huff, Sarah Shin, Gabriella Longway, Anuj Gupta, Lalitya Andaloori, Todd D. Armstrong, Daniel Haldar, Robert Anders, Elizabeth Thompson, Nilo Azad, Won Jin Ho, Elizabeth Jaffee, Elana J. Fertig, Neeha Zaidi. Evolution of immune cell composition and functionality as pancreatic intraepithelial neoplasia progresses to pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2873.

Abstract 76: Reprogramming of the preinvasive breast microenvironment identified by spatial nucleus barcoding

Abstract Ductal-carcinoma-in-situ (DCIS) is the most common early-stage breast cancer and a non-obligate precursor to invasive breast cancer. However our understanding of the spatial microenvironment and the reprogramming of cell types in DCIS remains limited. To address this question, we developed a Spatial Nucleus Barcoding (SNuBar) method that utilizes a single oligonucleotide adaptor to preserve spatial information followed by performing single nucleus RNA sequencing. We validated the accuracy and scalability of SNuBar in cell lines and applied this method to investigate the spatial microenvironment of 4 normal breast tissues, 7 normal adjacent DCIS tissues, and 8 DCIS cancers. The analysis of single cells isolated from 596 macrodissected spatial regions identified 13 major cell types and 43 cell states, that co-localized in different combinations across four main topographic areas. Spatial analysis revealed that vascular and myoepithelial cells were reprogrammed in localized zones near the premalignant cells, while fibroblasts, T-cells and myeloid cells were reprogrammed across broader fields. Our data shows that many diverse cell types are reprogrammed in spatially-defined areas in the DCIS microenvironment, relative to the normal breast tissue regions. Citation Format: Kaile Wang, Zhenna Xiao, Yiyun Lin, Runmin Wei, Jie Ye, Rui Ye, Min Hu, Shanshan Bai, Emi Sei, Aatish Thennavan, Bora Lim, Andrew Futreal, Alastair Thompson, Savitri Krishnamurthy, Nicholas Navin. Reprogramming of the preinvasive breast microenvironment identified by spatial nucleus barcoding [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 76.

Autofluorescence-Raman Spectroscopy for Ex Vivo Mapping Colorectal Liver Metastases and Liver Tissue

IntroductionIdentifying colorectal liver metastases (CRLM) during liver resection could assist in achieving clear surgical margins, which is an important prognostic variable for both disease-free and overall survival. The aim of this study was to investigate the effect of auto-fluorescence (AF) and Raman spectroscopy for ex vivo label-free discrimination of CRLMs from normal liver tissue. Secondary aims include exploring options for multimodal AF-Raman integration with respect to diagnosis accuracy and imaging speed on human liver tissue and CRLM. MethodsLiver samples were obtained from patients undergoing liver surgery for CRLM who provided informed consent (15 patients were recruited). AF and Raman spectroscopy was performed on CRLM and normal liver tissue samples and then compared to histology. ResultsAF emission spectra demonstrated that the 671 nm and 775/785 nm excitation wavelengths provided the highest contrast, as normal liver tissue elicited on average around eight-fold higher AF intensity compared to CRLM. The use of the 785 nm wavelength had the advantage of enabling Raman spectroscopy measurements from CRLM regions, allowing discrimination of CRLM from regions of normal liver tissue eliciting unusual low AF intensity, preventing misclassification. Proof-of-concept experiments using small pieces of CRLM samples covered by large normal liver tissue demonstrated the feasibility of a dual-modality AF-Raman for detection of positive margins within few minutes. ConclusionsAF imaging and Raman spectroscopy can discriminate CRLM from normal liver tissue in an ex vivo setting. These results suggest the potential for developing integrated multimodal AF-Raman imaging techniques for intraoperative assessment of surgical margins.

Abstract No. 51 Differentiation of Irreversible Electroporation Regions in Normal and Tumor Tissues Using MRI Textures

Irreversible electroporation (IRE) is a promising tissue ablation technique for liver cancer patients unsuitable for surgical resection or liver transplantation. The main clinical challenge in IRE ablation is differentiating IRE ablation zones for immediate assessment. Here, we investigated the reproducibility of radiomic analysis of conventional T1w MRI data acquired from liver tumors and normal liver tissues for assessment of IRE outcomes. IRE ablation was performed on twenty-six rabbits' normal liver or tumors. Anatomical and perfusion MRI data were acquired immediately after the completion of the procedure. IRE and RE zones were outlined on perfusion MRI data with histology validation and mask images were translated to T1w images. T1w MRI data were normalized using three levels to examine the effects of different image specificity (8, 16, and 32) while determining the characteristic patterns related to IRE ablation. At each specificity level, quantitative imaging features were computed by analyzing intensity distribution and textural patterns. The relevant variables were identified by following a multi-step feature selection framework that integrates pairwise likelihood, analysis, and LASSO. Multiple classification models were generated using logistic regression, nonlinear support vector machines, and random forest (RF) with a five-fold cross-validation approach. The accuracy, sensitivity, specificity, and receiver operating characteristic (ROC) curves were employed for the evaluation of the developed classification methods. Random forest classification models outperformed logistic regression and support vector machine models. In terms of image specificity levels, the images sampled with sixteen intensity levels facilitated the most accurate and robust performance compared with images with lower and higher details. The random forest classifier equipped with features computed from images with sixteen levels obtained a training accuracy of 90.86% and a validation accuracy of 81.09% for T1w images. The model sensitivity was 92.33% and 81.33%, and specificity was 89.33% and 80.67% for training and validation sets, respectively. Moreover, we recorded the area under the ROC curve as 97.86% for training and 89.47% for the validation cohort. T1w MRI radiomic features reveal IRE ablation-related effects performed on normal liver and tumor tissues, which demonstrates the potential benefit of quantitative analysis of conventional MRI data for immediate assessment of IRE ablation zones in other solid cancers.

A spatial squeeze and multimodal feature fusion attention network for multiple tumor segmentation from PET–CT Volumes

Tumor segmentation is a key step in computer-aided diagnosis. The PET–CT co-segmentation method combines the high sensitivity of PET images and the anatomical information of CT images. For whole-body multiple tumors, such as soft tissue sarcoma, lymphoma, etc., due to the different lesion location and size, it is necessary to segment the tumor area according to the whole body anatomical information. How to effectively leverage whole-body contextual information and the fusion of multimodal information is the key to the problem. To address this issue, we propose a spatial squeeze and multimodal feature fusion attention network for whole-body multiple tumors segmentation based on PET–CT volumes. Our proposed method consists of two parts, a Coronal-Spatial Squeeze Attention Extraction Network (CSAE-Net) and a Precise PET–CT Fusion Attention Segmentation Network (PFAS-Net), respectively. In CSAE-Net, we squeeze a 3D PET–CT volume along the coronal plane into m 2D images, and obtain 3D Coronal Spatial Squeeze Attention Volume based on these 2D images. In PFAS-Net, the input is a 2D axial PET–CT slice, and the previously obtained coronal spatial squeeze attention map is used to guide the segmentation. Moreover, a Multimodal Fusion Attention (MFA) module is proposed to fuse the metabolic information of PET and the anatomical information of CT. We perform experiments on PET–CT datasets of two whole-body multiple tumors, Soft Tissue Sarcoma (STS) and Lymphoma. The results show that our proposed method improved Dice values by 8.03% in STS and 1.74% in Lymphoma. Also the visualization results show that our proposed method is able to suppress high-uptake regions of normal tissues.

Multi-omics profiling and digital image analysis reveal the potential prognostic and immunotherapeutic properties of CD93 in stomach adenocarcinoma.

Recent evidence highlights the fact that immunotherapy has significantly improved patient outcomes. CD93, as a type I transmembrane glycoprotein, was correlated with tumor-associated angiogenesis; however, how CD93 correlates with immunotherapy in stomach adenocarcinoma (STAD) remains unclear. TCGA, GTEx, GEO, TIMER2.0, HPA, TISIDB, TCIA, cBioPortal, LinkedOmics, and ImmuCellAI public databases were used to elucidate CD93 in STAD. Visualization and statistical analysis of data were performed by R (Version 4.1.3), GraphPad (Version 8.0.1), and QuPath (Version 0.3.2). CD93 was highly expressed in STAD compared with adjacent normal tissues. The overexpression of CD93 was significantly correlated with a poor prognosis in STAD. There was a negative correlation between CD93 expression levels with CD93 mutation and methylation in STAD. Our results revealed that CD93 expression was positively associated with most immunosuppressive genes (including PD-1, PD-L1, CTLA-4, and LAG3), immunostimulatory genes, HLA, chemokine, and chemokine receptor proteins in STAD. Furthermore, in STAD, CD93 was noticeably associated with the abundance of multiple immune cell infiltration levels. Functional HALLMARK and KEGG term enhancement analysis of CD93 through Gene Set Enrichment Analysis was correlated with the process of the angiogenesis pathway. Subsequently, digital image analysis results by QuPath revealed that the properties of CD93+ cells were statistically significant in different regions of stomach cancer and normal stomach tissue. Finally, we utilized external databases, including GEO, TISIDB, ImmuCellAI, and TCIA, to validate that CD93 plays a key role in the immunotherapy of STAD. Our study reveals that CD93 is a potential oncogene and is an indicative biomarker of a worse prognosis and exerts its immunomodulatory properties and potential possibilities for immunotherapy in STAD.