Small Drugs, Huge Impact: The Extraordinary Impact of Antisense Oligonucleotides in Research and Drug Development

Antisense oligonucleotides (ASOs) are an increasingly represented class of drugs. These small sequences of nucleotides are designed to precisely target other oligonucleotides, usually RNA species, and are modified to protect them from degradation by nucleases. Their specificity is due to their sequence, so it is possible to target any RNA sequence that is already known. These molecules are very versatile and adaptable given that their sequence and chemistry can be custom manufactured. Based on the chemistry being used, their activity may significantly change and their effects on cell function and phenotypes can differ dramatically. While some will cause the target RNA to decay, others will only bind to the target and act as a steric blocker. Their incredible versatility is the key to manipulating several aspects of nucleic acid function as well as their process, and alter the transcriptome profile of a specific cell type or tissue. For example, they can be used to modify splicing or mask specific sites on a target. The entire design rather than just the sequence is essential to ensuring the specificity of the ASO to its target. Thus, it is vitally important to ensure that the complete process of drug design and testing is taken into account. ASOs’ adaptability is a considerable advantage, and over the past decades has allowed multiple new drugs to be approved. This, in turn, has had a significant and positive impact on patient lives. Given current challenges presented by the COVID-19 pandemic, it is necessary to find new therapeutic strategies that would complement the vaccination efforts being used across the globe. ASOs may be a very powerful tool that can be used to target the virus RNA and provide a therapeutic paradigm. The proof of the efficacy of ASOs as an anti-viral agent is long-standing, yet no molecule currently has FDA approval. The emergence and widespread use of RNA vaccines during this health crisis might provide an ideal opportunity to develop the first anti-viral ASOs on the market. In this review, we describe the story of ASOs, the different characteristics of their chemistry, and how their characteristics translate into research and as a clinical tool.

A novel deep learning-based 3D cell segmentation framework for future image-based disease detection

Cell segmentation plays a crucial role in understanding, diagnosing, and treating diseases. Despite the recent success of deep learning-based cell segmentation methods, it remains challenging to accurately segment densely packed cells in 3D cell membrane images. Existing approaches also require fine-tuning multiple manually selected hyperparameters on the new datasets. We develop a deep learning-based 3D cell segmentation pipeline, 3DCellSeg, to address these challenges. Compared to the existing methods, our approach carries the following novelties: (1) a robust two-stage pipeline, requiring only one hyperparameter; (2) a light-weight deep convolutional neural network (3DCellSegNet) to efficiently output voxel-wise masks; (3) a custom loss function (3DCellSeg Loss) to tackle the clumped cell problem; and (4) an efficient touching area-based clustering algorithm (TASCAN) to separate 3D cells from the foreground masks. Cell segmentation experiments conducted on four different cell datasets show that 3DCellSeg outperforms the baseline models on the ATAS (plant), HMS (animal), and LRP (plant) datasets with an overall accuracy of 95.6%, 76.4%, and 74.7%, respectively, while achieving an accuracy comparable to the baselines on the Ovules (plant) dataset with an overall accuracy of 82.2%. Ablation studies show that the individual improvements in accuracy is attributable to 3DCellSegNet, 3DCellSeg Loss, and TASCAN, with the 3DCellSeg demonstrating robustness across different datasets and cell shapes. Our results suggest that 3DCellSeg can serve a powerful biomedical and clinical tool, such as histo-pathological image analysis, for cancer diagnosis and grading.

An in vivo pig model for testing novel PET radioligands targeting cerebral protein aggregates

Positron emission tomography (PET) has become an essential clinical tool for diagnosing neurodegenerative diseases with abnormal accumulation of proteins like amyloid-β or tau. Despite many attempts, it has not been possible to develop an appropriate radioligand for imaging aggregated α-synuclein, which is seen in, e.g., Parkinson's Disease. Access to a large animal model with α-synuclein pathology would critically enable a more translationally appropriate evaluation of novel radioligands. We here established a pig model with cerebral injections of α-synuclein preformed fibrils or brain homogenate from postmortem human brain tissue from individuals with Alzheimer's disease (AD) or dementia with Lewy body (DLB) into the pig's brain using minimally invasive surgery and validated against saline injections. In the absence of a suitable α-synuclein radioligand, we validated the model with an unselective amyloid-β tracer [11C]PIB, which has a high affinity for β-sheet structures in aggregates. Gadolinium-enhanced MRI confirmed that the blood-brain barrier function was intact. A few hours post-injection, pigs were PET scanned with [11C]PIB. Quantification was done with Logan invasive graphical analysis and simplified reference tissue model 2 using the occipital cortex as a reference region. After the scan, we retrieved the brains to confirm successful injection using autoradiography and immunohistochemistry. We found four times higher [11C]PIB uptake in AD-homogenate-injected regions and two times higher uptake in α-synuclein-preformed-fibrils-injected regions compared to the saline-injected regions. The [11C]PIB uptake was the same in the occipital cortex, cerebellum, DLB-homogenate, and saline-injected regions. With its large brains and ability to undergo repeated PET scans as well as neurosurgical procedures, the pig provides a robust, cost-effective, and good translational model for assessment of novel radioligands including, but not limited to, proteinopathies.

Medical Photography Underutilized as a Clinical Tool: Key Considerations for Better Adoption (Preprint)

UNSTRUCTURED NA

The Future of Virtual Reality and Deep Learning in Visual Field Testing

The visual field (VF) examination is a useful clinical tool for monitoring a variety of ocular diseases. Despite its wide utility in eye clinics, the test as currently conducted is subject to an array of issues that interfere in obtaining accurate results. Visual field exams of patients suffering from additional ocular conditions are often unreliable due to interference between the comorbid diseases. To improve upon these shortcomings, virtual reality (VR) and deep learning are being explored as potential solutions. Virtual reality has been incorporated into novel visual field exams to provide a portable, 3D exam experience. Deep learning, a specialization of machine learning, has been used in conjunction with VR, such as in the iGlaucoma application, to limit subjective bias occurring from patients' eye movements. This chapter seeks to analyze and critique how VR and deep learning can augment the visual field experience by improving accuracy, reducing subjective bias, and ultimately, providing clinicians with a greater capacity to enhance patient outcomes.

Temperature Profile and Adverse Outcomes After Discharge From the Intensive Care Unit

Background A predictive model that uses the rhythmicity of core body temperature (CBT) could be an easily accessible clinical tool to ultimately improve outcomes among critically ill patients. Objectives To assess the relation between the 24-hour CBT profile (CBT-24) before intensive care unit (ICU) discharge and clinical events in the step-down unit within 7 days of ICU discharge. Methods This retrospective cohort study in a tertiary ICU at a single center included adult patients requiring acute invasive ventilation for more than 48 hours and assessed major clinical adverse events (MCAEs) and rapid response system activations (RRSAs) within 7 days of ICU discharge (MCAE-7 and RRSA-7, respectively). Results The 291 enrolled patients had a median mechanical ventilation duration of 139 hours (IQR, 50-862 hours) and at admission had a median Acute Physiology and Chronic Health Evaluation II score of 22 (IQR, 7-42). At least 1 MCAE or RRSA occurred in 64% and 22% of patients, respectively. Independent predictors of an MCAE-7 were absence of CBT-24 rhythmicity (odds ratio, 1.78 [95% CI, 1.07-2.98]; P = .03), Sequential Organ Failure Assessment score at ICU discharge (1.10 [1.00-1.21]; P = .05), male sex (1.72 [1.04-2.86]; P = .04), age (1.02 [1.00-1.04]; P = .02), and Charlson Comorbidity Index (0.87 [0.76-0.99]; P = .03). Age (1.03 [1.01-1.05]; P = .006), sepsis at ICU admission (2.02 [1.13-3.63]; P = .02), and Charlson Comorbidity Index (1.18 [1.02-1.36]; P = .02) were independent predictors of an RRSA-7. Conclusions Use of CBT-24 rhythmicity can assist in stratifying a patient’s risk of subsequent deterioration during general care within 7 days of ICU discharge.

Pancreatic Cancer Patient-Derived Organoid Platforms: A Clinical Tool to Study Cell- and Non-Cell-Autonomous Mechanisms of Treatment Response

For many years, cell lines and animal models have been essential to improve our understanding of the basis of cell metabolism, signaling, and genetics. They also provided an essential boost to cancer drug discovery. Nevertheless, these model systems failed to reproduce the tumor heterogeneity and the complex biological interactions between cancer cells and human hosts, making a high priority search for alternative methods that are able to export results from model systems to humans, which has become a major bottleneck in the drug development. The emergent human in vitro 3D cell culture technologies have attracted widespread attention because they seem to have the potential to overcome these limitations. Organoids are unique 3D culture models with the ability to self-organize in contained structures. Their versatility has offered an exceptional window of opportunity to approach human cancers. Pancreatic cancers (PCs) patient-derived-organoids (PDOs) preserve histological, genomic, and molecular features of neoplasms they originate from and therefore retain their heterogeneity. Patient-derived organoids can be established with a high success rate from minimal tissue core specimens acquired with endoscopic-ultrasound-guided techniques and assembled into platforms, representing tens to hundreds of cancers each conserving specific features, expanding the types of patient samples that can be propagated and analyzed in the laboratory. Because of their nature, PDO platforms are multipurpose systems that can be easily adapted in co-culture settings to perform a wide spectrum of studies, ranging from drug discovery to immune response evaluation to tumor-stroma interaction. This possibility to increase the complexity of organoids creating a hybrid culture with non-epithelial cells increases the interest in organoid-based platforms giving a pragmatic way to deeply study biological interactions in vitro. In this view, implementing organoid models in co-clinical trials to compare drug responses may represent the next step toward even more personalized medicine. In the present review, we discuss how PDO platforms are shaping modern-day oncology aiding to unravel the most complex aspects of PC.

A Network-Based Analysis of Disease Complication Associations for Obstetric Disorders in the UK Biobank

Background: Recent studies have found that women with obstetric disorders are at increased risk for a variety of long-term complications. However, the underlying pathophysiology of these connections remains undetermined. A network-based view incorporating knowledge of other diseases and genetic associations will aid our understanding of the role of genetics in pregnancy-related disease complications. Methods: We built a disease–disease network (DDN) using UK Biobank (UKBB) summary data from a phenome-wide association study (PheWAS) to elaborate multiple disease associations. We also constructed egocentric DDNs, where each network focuses on a pregnancy-related disorder and its neighboring diseases. We then applied graph-based semi-supervised learning (GSSL) to translate the connections in the egocentric DDNs to pathologic knowledge. Results: A total of 26 egocentric DDNs were constructed for each pregnancy-related phenotype in the UKBB. Applying GSSL to each DDN, we obtained complication risk scores for additional phenotypes given the pregnancy-related disease of interest. Predictions were validated using co-occurrences derived from UKBB electronic health records. Our proposed method achieved an increase in average area under the receiver operating characteristic curve (AUC) by a factor of 1.35 from 55.0% to 74.4% compared to the use of the full DDN. Conclusion: Egocentric DDNs hold promise as a clinical tool for the network-based identification of potential disease complications for a variety of phenotypes.