A Multimodal Ensemble Deep Learning Model for Functional Outcome Prognosis of Stroke Patients.

BACKGROUND AND PURPOSE: The accurate prediction of functional outcomes in patients with acute ischemic stroke (AIS) is crucial for informed clinical decision-making and optimal resource utilization. As such, this study aimed to construct an ensemble deep learning model that integrates multimodal imaging and clinical data to predict the 90-day functional outcomes after AIS. METHODS: We used data from the Korean Stroke Neuroimaging Initiative database, a prospective multicenter stroke registry to construct an ensemble model integrated individual 3D convolutional neural networks for diffusion-weighted imaging and fluid-attenuated inversion recovery (FLAIR), along with a deep neural network for clinical data, to predict 90-day functional independence after AIS using a modified Rankin Scale (mRS) of 3-6. To evaluate the performance of the ensemble model, we compared the area under the curve (AUC) of the proposed method with that of individual models trained on each modality to identify patients with AIS with an mRS score of 3-6. RESULTS: Of the 2,606 patients with AIS, 993 (38.1%) achieved an mRS score of 3-6 at 90 days post-stroke. Our model achieved AUC values of 0.830 (standard cross-validation [CV]) and 0.779 (time-based CV), which significantly outperformed the other models relying on single modalities: b-value of 1,000 s/mm2 (P<0.001), apparent diffusion coefficient map (P<0.001), FLAIR (P<0.001), and clinical data (P=0.004). CONCLUSION: The integration of multimodal imaging and clinical data resulted in superior prediction of the 90-day functional outcomes in AIS patients compared to the use of a single data modality.

Inhibition of the Alternative Complement Pathway May Cause Secretion of Factor B, Enabling an Early Detection of Pancreatic Cancer.

This study aims to elucidate the cellular mechanisms behind the secretion of complement factor B (CFB), known for its dual roles as an early biomarker for pancreatic ductal adenocarcinoma (PDAC) and as the initial substrate for the alternative complement pathway (ACP). Using parallel reaction monitoring analysis, we confirmed a consistent ∼2-fold increase in CFB expression in PDAC patients compared with that in both healthy donors (HD) and chronic pancreatitis (CP) patients. Elevated ACP activity was observed in CP and other benign conditions compared with that in HD and PDAC patients, suggesting a functional link between ACP and PDAC. Protein-protein interaction analyses involving key complement proteins and their regulatory factors were conducted using blood samples from PDAC patients and cultured cell lines. Our findings revealed a complex control system governing the ACP and its regulatory factors, including Kirsten rat sarcoma viral oncogene homolog (KRAS) mutation, adrenomedullin (AM), and complement factor H (CFH). Particularly, AM emerged as a crucial player in CFB secretion, activating CFH and promoting its predominant binding to C3b over CFB. Mechanistically, our data suggest that the KRAS mutation stimulates AM expression, enhancing CFH activity in the fluid phase through binding. This heightened AM-CFH interaction conferred greater affinity for C3b over CFB, potentially suppressing the ACP cascade. This sequence of events likely culminated in the preferential release of ductal CFB into plasma during the early stages of PDAC. (Data set ID PXD047043.).

Evaluation of Alternative Transport Media for RT-qPCR-Based SARS-CoV-2 Testing.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which causes coronavirus disease 2019 (COVID-19), is still rapidly spreading as of March 2022. An accurate and rapid molecular diagnosis is essential to determine the exact number of confirmed cases. Currently, the viral transport medium (VTM) required for testing is in short supply due to a sharp increase in the laboratory tests performed, and alternative VTMs are needed to alleviate the shortage. Guanidine thiocyanate-based media reportedly inactivate SARS-CoV-2 and are compatible with quantitative reverse transcription polymerase chain reaction (RT-qPCR) assays, but the compatibility and the viral detection capacity have not been fully validated. To evaluate the guanidine thiocyanate-based Gene Transport Medium (GeneTM) as an alternative VTM, we prepared 39 SARS-CoV-2-positive and 7 SARS-CoV-2-negative samples in GeneTM, eNAT™, and phosphate-buffered saline (PBS). The cycle threshold (Ct) values of three SARS-CoV-2 targets (the S, RdRP, and N genes) were analyzed using RT-qPCR testing. The comparison of Ct values from the positive samples showed a high correlation (R 2= 0.95-0.96) between GeneTM and eNAT™, indicating a comparable viral detection capacity. The delta Ct values of the SARS-CoV-2 genes in each transport medium were maintained for 14 days at cold (4°C) or room (25°C) temperatures, suggesting viral samples were stably preserved in the transport media for 14 days. Together, GeneTM is a potential alternative VTM with comparable RT-qPCR performance and stability to those of standard media.

Evaluation of Three Automated Extraction Systems for the Detection of SARS-CoV-2 from Clinical Respiratory Specimens.

Severe acute respiratory syndrome coronavirus (SARS-CoV-2) is highly contagious and causes coronavirus disease 2019 (COVID-19). Reverse transcription quantitative polymerase chain reaction (RT-qPCR) is the most accurate and reliable molecular assay to detect active SARS-CoV-2 infection. However, a rapid increase in test subjects has created a global bottleneck in testing capacity. Given that efficient nucleic acid extraction greatly affects reliable and accurate testing results, we compared three extraction platforms: MagNA Pure 96 DNA and Viral NA Small Volume kit on MagNA Pure 96 (Roche, Basel, Switzerland), careGENETM Viral/Pathogen HiFi Nucleic Acid Isolation kit (WELLS BIO Inc., Seoul, Korea) on KingFisher Flex (Thermo Fisher Scientific, Rocklin, CA, USA), and SGRespiTM Pure kit (Seegene Inc., Seoul, Korea) on Maelstrom 9600 (Taiwan Advanced Nanotech Inc., Taoyuan, Taiwan). RNA was extracted from 245 residual respiratory specimens from the different types of samples (i.e., NPS, sputum, and saliva) using three different kits. The 95% limits of detection of median tissue culture infectious dose per milliliter (TCID50/mL) for the MagNA Pure 96, KingFisher Flex, and Maelstrom 9600 were 0.37-3.15 × 101, 0.41-3.62 × 101, and 0.33-1.98 × 101, respectively. The KingFisher Flex platform exhibited 99.2% sensitivity and 100% specificity, whereas Maelstrom 9600 exhibited 98.3-100% sensitivity and 100% specificity. Bland-Altman analysis revealed a 95.2% concordance between MagNA Pure 96 and KingFisher Flex and 95.4% concordance between MagNA Pure 96 and Maelstrom 9600, indicating that all three platforms provided statistically reliable results. This suggests that two modifying platforms, KingFisher Flex and Maelstrom 9600, are accurate and scalable extraction platforms for large-scale SARS-CoV-2 clinical detection and could help the management of COVID-19 patients.

Development of an efficient Sanger sequencing-based assay for detecting SARS-CoV-2 spike mutations.

Novel strains of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) harboring nucleotide changes (mutations) in the spike gene have emerged and are spreading rapidly. These mutations are associated with SARS-CoV-2 transmissibility, virulence, or resistance to some neutralizing antibodies. Thus, the accurate detection of spike mutants is crucial for controlling SARS-CoV-2 transmission and identifying neutralizing antibody-resistance caused by amino acid changes in the receptor-binding domain. Here, we developed five SARS-CoV-2 spike gene primer pairs (5-SSG primer assay; 69S, 144S, 417S, 484S, and 570S) and verified their ability to detect nine key spike mutations (ΔH69/V70, T95I, G142D, ΔY144, K417T/N, L452R, E484K/Q, N501Y, and H655Y) using a Sanger sequencing-based assay. The 5-SSG primer assay showed 100% specificity and a conservative limit of detection with a median tissue culture infective dose (TCID50) values of 1.4 × 102 TCID50/mL. The accuracy of the 5-SSG primer assay was confirmed by next generation sequencing. The results of these two approaches showed 100% consistency. Taken together, the ability of the 5-SSG primer assay to accurately detect key SARS-CoV-2 spike mutants is reliable. Thus, it is a useful tool for detecting SARS-CoV-2 spike gene mutants in a clinical setting, thereby helping to improve the management of patients with COVID-19.

Cancer cell-induced neutrophil extracellular traps promote both hypercoagulability and cancer progression.

INTRODUCTION: Neutrophils can generate extracellular net-like structures by releasing their DNA-histone complexes and antimicrobial peptides, which is called neutrophil extracellular traps (NETs). Various stimuli can induce NET formation. In particular, neutrophils and NET formation are abundant in tumor tissue. This study investigated how cancer cells induce NET formation and whether this NET formation promotes plasma thrombin generation and cancer progression. METHODS: Induction of NET formation by a pancreatic cancer cell line (AsPC-1) was assessed by measuring the histone-DNA complex level. The endogenous thrombin potential (ETP) was measured by thrombin generation assay. In vitro migration, invasion, and tubule formation assays were performed. The circulating levels of NET markers and hypercoagulability markers were assessed in 62 patients with pancreatobiliary malignancy and 30 healthy controls. RESULTS: AsPC-1 significantly induced NET formation in a dose-dependent manner. Conditioned medium (CM) from AsPC-1 also induced NETs. Interestingly, NET-formation was abolished by heat-inactivated CM, but not by lipid-extracted CM, suggesting an important role of protein components. A reactive oxygen species inhibitor did not inhibit cancer cell-induced NET formation, but prostaglandin E1 (PGE1, cyclic adenosine monophosphate inducer) and antithrombin did. NETs significantly increased ETP of normal plasma. Of note, NETs promoted cancer cell migration and invasion as well as angiogenesis, which were inhibited by histone-binding agents (heparin, polysialic acid), a DNA-degrading enzyme, and Toll-like receptor neutralizing antibodies. In patients with pancreatobiliary malignancy, elevated NET markers correlated well with hypercoagulability makers. CONCLUSION: Our findings indicate that cancer cell-induced NET formation enhances both hypercoagulability and cancer progression and suggest that inhibitors of NET formation such as PGE1 and antithrombin can be potential therapeutics to reduce both hypercoagulability and cancer progression.

Thrombotic Risk of Non-Criteria Anti-Phospholipid Antibodies Measured by Line Immunoassay: Superiority of Anti-Phosphatidylserine and Anti-Phosphatidic Acid Antibodies.

BACKGROUND: Increasing interest has focused on the development of new assays more specific for APS and application of multiple combinations of anti-phospholipid antibodies (aPLs). This study explored the thrombotic risk of non-criteria aPLs measured by a new line immunoassay (LIA) and the benefit of additional non-criteria aPLs results to the APS diagnosis. METHODS: LIA were performed to detect 9 aPLs in 180 patients requested for lupus anticoagulant (LA) measurement. Antibodies against anti-cardiolipin (CL), ß2 glycoprotein I (GPI), and ß2GPI domain I were measured by ELISA. RESULTS: The agreement percentages for IgG/IgM anti-CL and anti-ß2GPI between ELISA and LIA (anti-CL 68.2% and 82.6%; anti-ß2GPI 71.7% and 93.2%, respectively). Among 9 aPLs measured by LIA, single presence IgG of anti-phosphatidylserine (odds ratio (OR) 16.477) and anti-phosphatidic acid (OR 9.625) predicted higher thrombotic risk than anti-ß2GPI (OR 5.538). Other aPLs measured by LIA (anti-prothrombin, anti-annexin V, anti-phosphatidylinositol, phosphatidylethanolamine, and anti-phosphatidylglycerol) did not show any significant thrombotic risk. Addition of the 2 non-criteria aPLs (anti-phosphatidylserine and anti-phosphatidic acid) to the established APS criteria increased the diagnostic specificity and accuracy for thrombosis. The positive rates of anti-ß2GPI and anti-phosphatidylserine measured by LIA were quite high in patients with positive anti-ß2GPI domain I. CONCLUSIONS: The anti-phosphatidylserine and anti-phosphatidic acid among non-criteria aPLs have a high likeli-hood as new markers for thrombotic prediction.

Contact System Activation and Neutrophil Extracellular Trap Markers: Risk Factors for Portal Vein Thrombosis in Patients With Hepatocellular Carcinoma.

Although portal vein thrombosis (PVT) commonly occurs in patients with hepatocellular carcinoma (HCC), the hypercoagulability mechanism in patients with HCC is not entirely clear. Recently, tumor-induced formation of neutrophil extracellular traps (NET) has been shown to trigger contact system activation, and contact system activation has been shown to be a new mechanism of thrombosis. Therefore, we investigated whether contact system activation and NET formation occurred in relation to PVT in HCC patients. The circulating levels of NET formation markers (DNA-histone complex, double-stranded DNA, neutrophil elastase) and contact system activation markers (factor XIIa and high-molecular-weight kininogen) were measured in 177 patients who had been diagnosed with HCC and 48 healthy controls. Presence of PVT was confirmed in 77 HCC patients. The levels of NET formation and contact system activation markers were significantly higher in patients than in healthy controls and they increased significantly with the increase in the model for end-stage liver disease (MELD) scores. Of note, these markers were significantly higher in HCC patients with PVT than in those without PVT. These NET formation markers and the contact system activation markers were significant thrombotic risk factors in HCC patients. The well-known liver injury markers (alanine transaminase, prothrombin time) significantly contributed to factor XIIa level. Contact system activation and NET formation are well correlated with liver disease severity and the markers of these can be used as thrombotic risk factors in HCC patients. In addition, therapeutics inhibiting the contact system can be potentially used to manage PVT in HCC patients.

Elevated extracellular trap formation and contact system activation in acute leukemia.

Leukemic cells release their nuclear contents into the extracellular space upon activation. The released nuclear contents, called extracellular traps, can activate the contact system of coagulation. This study accessed the extent of contact system activation, the levels of extracellular traps, and coagulation activation in hematologic malignancies including acute leukemia. In 154 patients with hematologic malignancies (acute leukemia, n = 29; myelodysplastic syndrome, n = 20; myeloproliferative neoplasms, n = 69; plasma cell myeloma, n = 36) and 48 normal controls, the levels of coagulation factors (fibrinogen and factor VII, VIII, IX, and XII), D-dimer, thrombin generation, extracellular trap markers (histone-DNA complex, cell-free dsDNA, leukocyte elastase), and contact system markers (activated factor XII [XIIa], high-molecular-weight kininogen, prekallikrein, bradykinin) were measured. Patients with acute leukemia showed the highest levels of peak thrombin, extracellular trap markers, and factor XIIa. Factor XIIa level was significantly associated with the presence of acute leukemia. The histone-DNA complex and cell-free dsDNA were revealed as significant associated factors with the factor XIIa level. Three markers of extracellular traps and two markers of thrombin generation significantly contributed to the hemostatic abnormalities in hematologic malignancies. Contact system was activated in acute leukemia and its activation was significantly associated with the extent of extracellular trap formation. This finding suggests that extracellular traps might be a major source of contact system activation and therapeutic strategies targeting extracellular trap formation or contact system activation may be beneficial in acute leukemia.

Effects of interactions between common genetic variants and smoking on colorectal cancer.

BACKGROUND: Although genome-wide association studies (GWAS) have identified variants in approximately 40 susceptibility loci for colorectal cancer (CRC), there are few studies on the interactions between identified single-nucleotide polymorphisms (SNPs) and lifestyle risk factors. We evaluated whether smoking could modify associations between these genetic variants and CRC risk. METHODS: A total of 703 CRC patients and 1406 healthy controls were included in this case-control study from the National Cancer Center in Korea. Thirty CRC susceptibility SNPs identified in previous GWAS were genotyped. A logistic regression model was used to examine associations between the SNPs and smoking behaviors by sex. The interaction was estimated by including an additional interaction term in the model. RESULTS: In men, an increased CRC risk was observed for longer durations (OR>28 vs. ≤28years = 1.49 (95% CI = 1.11-1.98)), greater quantities (OR≥20 vs. <20cigarettes/day = 2.12 (1.61-2.79)), and longer pack-years of smoking (OR≥21 vs. <21pack-years = 1.78 (1.35-2.35)). In women, longer pack-years of smoking significantly increased CRC risk (OR≥5 vs. <5pack-years = 6.11 (1.10-34.00)). Moreover, there were significant interactions between smoking status and the polymorphisms rs1957636 at 14q22.3 (P interaction = 5.5 × 10-4) and rs4813802 at 20p12.3 (P interaction = 0.04) in men. Interactions between smoking status and the rs6687758 at 1q41 (P interaction = 0.03), duration and the rs174537 at 11q12.2 (P interaction = 0.05), and pack-years and the rs4813802 (P interaction = 0.04) were also found in women. CONCLUSIONS: Associations between susceptibility SNPs and CRC risk may be modified by smoking behaviors, supporting the existence of gene-smoking interactions.