Ophthalmic care may not align with patient need: An analysis on state-wide patient needs and provider density between 2008 and 2022

PURPOSE

This study aims to assess the extent to which the demand for ophthalmologic care among patients at the state level is reflected in Google Trends data, serving as an indicator of patient desire in ophthalmology.

METHODS

For each state, patient interest in ophthalmologic care was estimated using the Google Trends resource measuring web search and YouTube search rates for multiple ophthalmologic terms. We compared the change in search for ophthalmologic terms over time and used ordinary least squares regression to evaluate whether search interest for ophthalmologic terms was able to predict the rate of practicing ophthalmologists in each state. We also compare the changing rates of searches across the web and YouTube to evaluate the resources patients are most likely to utilize.

RESULTS

From 2008 to 2022, web search rates for general ophthalmology related terms increased by 43.98%, while search interest for retinal specific terms increased by 19.51%. YouTube specific results for general ophthalmology terms increased by 55.83% while search for retinal terms fell by 58.48%. Ophthalmologic and retinal specific search interest was not significantly associated with either outcome.

CONCLUSIONS

Our findings suggest that patient information needs, demographic elements, and the educational backgrounds of residents and fellows - those important factors - are surprisingly poorly correlated with ophthalmology provider density. Furthermore, we observed no noteworthy correlation between the search interest in ophthalmology and the overall density of ophthalmologists or retinal specialists. This implies that there is a pressing need to explore and implement strategies aimed at better aligning these influencing factors the choices made by ophthalmologists in selecting their practice locations to bridge the gap between healthcare availability and public interest.

Comparison of Novel Proteomic Expression Profiles for Radiation Exposure in Male and Female C57BL6 Mice

There is a need for point-of-care diagnostics for future mass casualty events involving radiation exposure. The development of radiation exposure and dose prediction algorithms for biodosimetry is needed for screening of large populations during these scenarios, and exploration of the potential effects which sex, age, genetic heterogeneity, and physiological comorbidities may have on the utility of biodosimetry diagnostics is needed. In the current study, proteomic profiling was used to examine sex-specific differences in age-matched C57BL6 mice on the blood proteome after radiation exposure, and the usefulness of development and application of biodosimetry algorithms using both male and female samples. Male and female mice between 9-11 weeks of age received a dose of total-body irradiation (TBI) of either 2, 4 or 8 Gy and plasma was collected at days 1, 3 and 7 postirradiation. Plasma was then screened using the SomaScan v4.1 assay for ∼7,000 protein analytes. A subset panel of protein biomarkers demonstrated significant (FDR < 0.05 and |logFC| > 0.2) changes in expression after radiation exposure. All proteins were used for feature selection to build predictive models of radiation exposure using different sample and sex-specific cohorts. Both binary (prediction of any radiation exposure) and multidose (prediction of specific radiation dose) model series were developed using either female and male samples combined or only female or only male samples. The binary series (models 1, 2 and 3) and multidose series (models 4, 5 and 6) included female/male combined, female only and male only respectively. Detectable values were obtained for all ∼7,000 proteins included in the SomaScan assay for all samples. Each model algorithm built using a unique sample cohort was validated with a training set of samples and tested with a separate new sample series. Overall predictive accuracies in the binary model series was ∼100% at the model training level, and when tested with fresh samples, 97.9% for model 1 (female and male) and 100% for model 2 (female only) and model 3 (male only). When sex-specific models 2 and 3 were tested with the opposite sex, the overall predictive accuracy rate dropped to 62.5% for model 2 and remained 100% for model 3. The overall predictive accuracy rate in the multidose model series was 100% for all models at the model training level and, when tested with fresh samples, 83.3%, 75% and 83.3% for Multidose models 4-6, respectively. When sex-specific model 5 (female only) and model 6 (male only) were tested with the opposite sex, the overall predictive accuracy rate dropped to 52.1% and 68.8%, respectively. These models represent novel predictive panels of radiation-responsive proteomic biomarkers and illustrate the utility and necessity of considering sex-specific differences in development of radiation biodosimetry prediction algorithms. As sex-specific differences were observed in this study, and as use of point-of-care radiation diagnostics in future mass casualty settings will necessarily include persons of both sexes, consideration of sex-specific variation is essential to ensure these diagnostic tools have practical utility in the field.

Tetrahydropalmatine promotes random skin flap survival in rats via the PI3K/AKT signaling pathway

ETHNOPHARMACOLOGICAL RELEVANCE

Flap necrosis is the most common complication after flap transplantation, but its prevention remains challenging. Tetrahydropalmatine (THP) is the main bioactive component of the traditional Chinese medicine Corydalis yanhusuo, with effects that include the activation of blood circulation, the promotion of qi, and pain relief. Although THP is widely used to treat various pain conditions, its impact on flap survival is unknown.

AIM OF THE STUDY

To explore the effect and mechanism of THP on skin flap survival.

MATERIALS AND METHODS

In this study, we established a modified McFarlane flap model, and the flap survival rate was calculated after 7 days of THP treatment. Angiogenesis and blood perfusion were evaluated using lead oxide/gelatin angiography and laser Doppler, respectively. Flap tissue obtained from zone II was evaluated histopathologically, by hematoxylin and eosin staining, and in assays for malondialdehyde content and superoxide dismutase activity. Immunofluorescence was performed to detect interleukin (IL)-6, tumor necrosis factor (TNF)-α, hypoxia-inducible factor (HIF)-1α, Bcl-2, Bax, caspase-3, caspase-9, SQSTM1/P62, Beclin-1, and LC3 expression, and Western blot to assess PI3K/AKT signaling pathway activation and Vascular endothelial growth factor (VEGF) expression. The role played by the autophagy pathway in flap necrosis was examined using rapamycin, a specific inhibitor of mTOR.

RESULTS

Experimentally, THP improved the survival rate of skin flaps, promoted angiogenesis, and improved blood perfusion. THP administration reduced the inflammatory response, oxidative stress, and apoptosis in addition to inhibiting autophagy via the PI3K/AKT/mTOR pathway. Rapamycin partially reversed these effects.

CONCLUSION

THP promotes skin flap survival via the PI3K/AKT signaling pathway.

An Updated Simplified Severity Scale for Age-Related Macular Degeneration, Incorporating Reticular Pseudodrusen: Age-Related Eye Disease Study Report No. 42

PURPOSE

To update the Age-Related Eye Disease Study (AREDS) Simplified Severity Scale for risk of late age-related macular degeneration (AMD), including incorporation of reticular pseudodrusen (RPD), and to perform external validation on the AREDS2.

DESIGN

Post hoc analysis of two clinical trial cohorts: AREDS and AREDS2.

PARTICIPANTS

Participants with no late AMD in either eye at baseline in AREDS (n=2719) and AREDS2 (n=1472).

METHODS

Five-year rates of progression to late AMD were calculated according to levels 0-4 on the Simplified Severity Scale, following two updates: (i) non-central GA considered part of the outcome rather than a risk feature, and (ii) scale separation according to RPD status (determined by validated deep learning grading of color fundus photographs).

MAIN OUTCOME MEASURES

Five-year rate of progression to late AMD (defined as neovascular AMD or any GA).

RESULTS

In the AREDS, following the first scale update, the five-year rates of progression to late AMD for levels 0-4 were 0.3%, 4.5%, 12.9%, 32.2%, and 55.6%, respectively. Following both updates, the proportion progressing to late AMD by five years was 8.4% in participants without RPD and 40.6% in those with RPD. As the final Simplified Severity Scale, the five-year progression rates for levels 0-4, respectively, were 0.3%, 4.3%, 11.6%, 26.7%, and 50.0%, for participants without RPD at baseline, and 2.8%, 8.0%, 29.0%, 58.7%, and 72.2%, for participants with RPD at baseline. In external validation on the AREDS2, for levels 2-4, the progression rates were similar, at 15.0%, 27.7%, and 45.7% (RPD absent) and 26.2%, 46.0%, and 73.0% (RPD present), respectively.

CONCLUSIONS

The AREDS AMD Simplified Severity Scale has been modernized with two important updates. The new scale for individuals without RPD has five-year progression rates of ∼0.5%, 4%, 12%, ∼25%, and 50%, such that the rates on the original scale remain accurate. The new scale for individuals with RPD has five-year progression rates of 3%, 8%, ∼30%, ∼60%, and ∼70%, i.e., approximately double for most levels. This scale fits updated definitions of late AMD, has increased prognostic accuracy, appears generalizable to similar populations, but remains simple for broad risk categorization.

Diet changes due to urbanization in South Africa are linked to microbiome and metabolome signatures of Westernization and colorectal cancer

Transition from traditional high-fiber to Western diets in urbanizing communities of Sub-Saharan Africa is associated with increased risk of non-communicable diseases (NCD), exemplified by colorectal cancer (CRC) risk. To investigate how urbanization gives rise to microbial patterns that may be amenable by dietary intervention, we analyzed diet intake, fecal 16 S bacteriome, virome, and metabolome in a cross-sectional study in healthy rural and urban Xhosa people (South Africa). Urban Xhosa individuals had higher intakes of energy (urban: 3,578 ± 455; rural: 2,185 ± 179 kcal/d), fat and animal protein. This was associated with lower fecal bacteriome diversity and a shift from genera favoring degradation of complex carbohydrates (e.g., Prevotella) to taxa previously shown to be associated with bile acid metabolism and CRC. Urban Xhosa individuals had higher fecal levels of deoxycholic acid, shown to be associated with higher CRC risk, but similar short-chain fatty acid concentrations compared with rural individuals. Fecal virome composition was associated with distinct gut bacterial communities across urbanization, characterized by different dominant host bacteria (urban: Bacteriodota; rural: unassigned taxa) and variable correlation with fecal metabolites and dietary nutrients. Food and skin microbiota samples showed compositional differences along the urbanization gradient. Rural-urban dietary transition in South Africa is linked to major changes in the gut microbiome and metabolome. Further studies are needed to prove cause and identify whether restoration of specific components of the traditional diet will arrest the accelerating rise in NCDs in Sub-Saharan Africa.

PubTator 3.0: an AI-powered literature resource for unlocking biomedical knowledge

PubTator 3.0 (https://www.ncbi.nlm.nih.gov/research/pubtator3/) is a biomedical literature resource using state-of-the-art AI techniques to offer semantic and relation searches for key concepts like proteins, genetic variants, diseases and chemicals. It currently provides over one billion entity and relation annotations across approximately 36 million PubMed abstracts and 6 million full-text articles from the PMC open access subset, updated weekly. PubTator 3.0's online interface and API utilize these precomputed entity relations and synonyms to provide advanced search capabilities and enable large-scale analyses, streamlining many complex information needs. We showcase the retrieval quality of PubTator 3.0 using a series of entity pair queries, demonstrating that PubTator 3.0 retrieves a greater number of articles than either PubMed or Google Scholar, with higher precision in the top 20 results. We further show that integrating ChatGPT (GPT-4) with PubTator APIs dramatically improves the factuality and verifiability of its responses. In summary, PubTator 3.0 offers a comprehensive set of features and tools that allow researchers to navigate the ever-expanding wealth of biomedical literature, expediting research and unlocking valuable insights for scientific discovery.

Independent and interactive associations of heart rate and obesity with type 2 diabetes mellites: A population-based study

BACKGROUND

Although obesity and heart rate (HR) were closely related to the prevalence and development of type 2 diabetes mllitus (T2DM), few studies have shown a co-association effect of them on T2DM. We aimed at assessing the interactive effects of HR and obesity with prevalence of T2DM in Chinese population, providing the exact cutpoint of the risk threshold for blood glucose with high HR.

MATERIALS AND METHODS

In the Risk Evaluation of cAncers in Chinese diabeTic Individuals: a lONgitudinal study (REACTION) cohorts (N = 8398), the relationship between HR and T2DM was explored by linear regression, logistic regression, and restricted cubic spline, and odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. Interaction terms between HR and body mass index (BMI) and HR and waist circumference (WC) were introduced into the logistic regression model.

RESULTS

In those with HR > 88.0 beats/min, fasting plasma glucose and oral glucose tolerance tests were significantly correlated with HR, and the prevalence of T2DM was highly correlated with HR (all p < .05). There were interactive associations of HR and obesity in patients with T2DM with HR < 74 beats/min.

CONCLUSION

High HR was in interaction with obesity, associating with prevalence of T2DM. The newly subdivided risk threshold for HR with T2DM might be HR > 88 beats/minute.

IL-36 Regulates Neutrophil Chemotaxis and Bone Loss at the Oral Barrier

Tissue-specific mechanisms regulate neutrophil immunity at the oral barrier, which plays a key role in periodontitis. Although it has been proposed that fibroblasts emit a powerful neutrophil chemotactic signal, how this chemotactic signal is driven has not been clear. The objective of this study was to investigate the site-specific regulatory mechanisms by which fibroblasts drive powerful neutrophil chemotactic signals within the oral barrier, with particular emphasis on the role of the IL-36 family. The present study found that IL-36γ, agonist of IL-36R, could promote neutrophil chemotaxis via fibroblast. Single-cell RNA sequencing data disclosed that IL36G is primarily expressed in human and mouse gingival epithelial cells and mouse neutrophils. Notably, there was a substantial increase in IL-36γ levels during periodontitis. In vitro experiments demonstrated that IL-36γ specifically activates gingival fibroblasts, leading to chemotaxis of neutrophils. In vivo experiments revealed that IL-36Ra inhibited the infiltration of neutrophils and bone resorption, while IL-36γ promoted their progression in the ligature-induced periodontitis mouse model. In summary, these data elucidate the function of the site-enriched IL-36γ in regulating neutrophil immunity and bone resorption at the oral barrier. These findings provide new insights into the tissue-specific pathophysiology of periodontitis and offer a promising avenue for prevention and treatment through targeted intervention of the IL-36 family.

Advancing entity recognition in biomedicine via instruction tuning of large language models

MOTIVATION

Large Language Models (LLMs) have the potential to revolutionize the field of Natural Language Processing, excelling not only in text generation and reasoning tasks but also in their ability for zero/few-shot learning, swiftly adapting to new tasks with minimal fine-tuning. LLMs have also demonstrated great promise in biomedical and healthcare applications. However, when it comes to Named Entity Recognition (NER), particularly within the biomedical domain, LLMs fall short of the effectiveness exhibited by fine-tuned domain-specific models. One key reason is that NER is typically conceptualized as a sequence labeling task, whereas LLMs are optimized for text generation and reasoning tasks.

RESULTS

We developed an instruction-based learning paradigm that transforms biomedical NER from a sequence labeling task into a generation task. This paradigm is end-to-end and streamlines the training and evaluation process by automatically repurposing pre-existing biomedical NER datasets. We further developed BioNER-LLaMA using the proposed paradigm with LLaMA-7B as the foundational LLM. We conducted extensive testing on BioNER-LLaMA across three widely recognized biomedical NER datasets, consisting of entities related to diseases, chemicals, and genes. The results revealed that BioNER-LLaMA consistently achieved higher F1-scores ranging from 5% to 30% compared to the few-shot learning capabilities of GPT-4 on datasets with different biomedical entities. We show that a general-domain LLM can match the performance of rigorously fine-tuned PubMedBERT models and PMC-LLaMA, biomedical-specific language model. Our findings underscore the potential of our proposed paradigm in developing general-domain LLMs that can rival SOTA performances in multi-task, multi-domain scenarios in biomedical and health applications.

AVAILABILITY AND IMPLEMENTATION

Datasets and other resources are available at https://github.com/BIDS-Xu-Lab/BioNER-LLaMA.

[A survey on the current situation of serum vitamin A and vitamin D levels among children aged 2-<7 years of 20 cities in China]

Objective: To investigate serum vitamin A and vitamin D status in children aged 2-<7 years in 20 cities in China. Methods: A cross-sectional study was conducted. A total of 2 924 healthy children aged 2-<7 years were recruited from September 2018 to September 2019 from 20 cities in China, categorized by age groups of 2-<3 years, 3-<5 years, and 5-<7 years. The demographic and economic characteristics and health-related information of the enrolled children were investigated. Body weight and height were measured by professional staff members. The serum vitamin A and vitamin D levels were detected by high-performance liquid chromatography-tandem mass spectrometry. Chi-square test and Logistic regression were applied to analyze the association between vitamin A and vitamin D deficiency and insufficiency as well as their underlying impact factors. Results: The age of the 2 924 enrolled children was 4.33 (3.42, 5.17) years. There were 1 726 males (59.03%) and 1 198 females (40.97%). The prevalences of vitamin A and vitamin D deficiency in enrolled children were 2.19% (64/2 924) and 3.52% (103/2 924), respectively, and the insufficiency rates were 29.27% (856/2 924) and 22.20% (649/2 924), respectively. Children with both vitamin A and vitamin D deficiencies or insufficiencies were found in 10.50% (307/2 924) of cases. Both vitamin A (χ2=7.91 and 8.06, both P=0.005) and vitamin D (χ2=71.35 and 115.10, both P<0.001) insufficiency rates were higher in children aged 3-<5 and 5-<7 years than those in children aged 2-<3 years. Vitamin A and vitamin D supplementation in the last 3 months was a protective factor for vitamin A and D deficiency and insufficiency, respectively (OR=0.68 and 0.22, 95%CI 0.49-0.95 and 0.13-0.40, both P<0.05). The rates of vitamin A and D insufficiency was higher in children with annual household incomes <60 000 RMB than in those with annual household incomes ≥60 000 RMB (χ2=34.11 and 10.43, both P<0.01). Northwest and Southwest had the highest rates of vitamin A and vitamin D insufficiency in children aged 2-<7 yeas, respectively (χ2=93.22 and 202.54, both P<0.001). Conclusions: Among 20 cities in China, children aged 2-<7 years experience high rates of vitamin A and vitamin D insufficiency, which are affected by age, family economic level, vitamin A and vitamin D supplementation, and regional economic level. The current results suggest that high level of attention should be paid to vitamin A and vitamin D nutritional status of preschool children.

Functional Brain Network Alterations in Patients With Systemic Lupus Erythematosus With Different Cognitive Function States: A Graph Theory Analysis Study

OBJECTIVE

The study aimed to investigate the characteristics of brain functional network disruption in patients with systemic lupus erythematosus (SLE) with different cognitive function states by using graph theory analysis and to explore their relationship with clinical data and neuropsychiatric scales.

METHODS

Resting-state functional magnetic resonance imaging data were collected from 38 female SLE patients and 44 healthy controls. Based on Montreal Cognitive Assessment (MoCA) scores, SLE patients were divided into a high MoCA group (MoCA-H; MoCA score, ≥26) and a low MoCA group (MoCA-L; MoCA score, <26). The matrix of resting-state functional brain networks of subjects in the 3 groups was constructed by using the graph theory approach. The topological properties of the functional brain networks, including global and local metrics, in the 3 groups were calculated. The differences in the topological properties of networks between the 3 groups were compared. In addition, Spearman correlation analysis was used to explore the correlation between altered topological properties of brain networks and clinical indicators, as well as neuropsychiatric scales in SLE patients in the MoCA-L group.

RESULTS

At the global level, in the sparsity threshold range of 0.10 to 0.34, the values of small-world properties were greater than 1 in all 3 groups, indicating that functional brain networks of both 3 groups had small-world properties. There were statistically significant differences in the characteristic path length, global, and local efficiency between 3 groups ( F = 3.825, P = 0.0260; F = 3.722, P = 0.0285; and F = 3.457, P = 0.0364, respectively). Systemic lupus erythematosus patients in the MoCA-L group showed increased characteristic path length ( t = 2.816, P = 0.00651), decreased global ( t = -2.729, P = 0.00826), and local efficiency ( t = -2.623, P = 0.0109) compared with healthy controls. No statistically significant differences in local metrics were found between the MoCA-H group and the healthy control, MoCA-L groups. At the local level, there was statistically significant difference in the node efficiency among the 3 groups ( P < 0.05 after Bonferroni correction). Compared with healthy controls, SLE patients in the MoCA-L group showed decreased node efficiency in left anterior cingulate paracingulate gyrus, bilateral putamen, bilateral pallidum, and left Heschl gyrus. No statistically significant differences in the local metrics were found between the MoCA-H, MoCA-L, and healthy control groups. Correlation analysis in SLE patients in the MoCA-L group showed that the characteristic path length was positively correlated with C4 levels ( r = 0.587, P = 0.007), the global and local efficiencies were negatively correlated with C4 levels ( r = -0.599, P = 0.005; r = -0.599, P = 0.005, respectively), and the node efficiency in the bilateral putamen was negatively correlated with C4 levels ( r = -0.611, P = 0.004; r = -0.570, P = 0.009). The node efficiency in the left pallidum was negatively correlated with disease duration ( r = -0.480, P = 0.032). The node efficiency in the left Heschl gyrus was negatively correlated with IgM levels ( r = -0.478, P = 0.033). No correlation was noted between other network metrics, clinical indicators, and neuropsychological scales.

CONCLUSIONS

The topological properties of functional brain networks were disrupted in SLE patients with low MoCA scores, suggesting that altered topological properties of the brain networks were associated with cognitive function in SLE patients. Correlation between altered topological properties of the brain networks and clinical indicators was noted in SLE patients with low MoCA scores, suggesting that altered topological properties of brain networks in SLE patients may have clinical significance as imaging markers for monitoring disease changes in patients with SLE.

L-Borneol promotes skin flap survival by regulating HIF-1α/NF-κB pathway

ETHNOPHARMACOLOGICAL RELEVANCE

The clinical application of skin flaps in surgical reconstruction is frequently impeded by the occurrence of distant necrosis. L-Borneol exhibits myogenic properties in traditional Chinese medicine and is used in clinical settings to promote wound healing and conditions such as stroke. Nevertheless, the precise mechanism by which borneol exerts its protective effects on skin flap survival remains unclear.

AIM OF THE STUDY

To explore the potential of L-borneol to promote skin flap survival and elucidate the underlying mechanisms.

MATERIALS AND METHODS

Thirty-six male Sprague-Dawley rats were randomly divided into three groups: a high-dose (200 mg/kg L-borneol per day), a low-dose (50 mg/kg/day), and control group (same volume of solvent). In each rat, a modified rectangular McFarlane flap model measuring 3 × 9 cm was constructed. Daily intragastric administration of L-borneol or solvent was performed. The flap was divided into three square sections of equal size, namely Zone I (the proximal zone), Zone II (the intermediate zone), and Zone III (the distal zone). The survival rate was quantified, and the histological state of each flap was evaluated on the seventh day following the surgical procedure. The assessment of angiogenesis was conducted using lead oxide/gelatin angiography, whereas the evaluation of blood flow in the free flap was performed using laser Doppler flow imaging. Superoxide dismutase activity was detected using the water-soluble tetrazolium salt-8 method. The quantities of vascular endothelial growth factor, interleukin (IL)-1β, IL-6, and tumour necrosis factor-α were determined using immunohistochemistry. The levels of nuclear transcription factor-κB, hypoxia-inducible factor-1, B-cell lymphoma-2 (BCL-2), and BCL-2-associated X (BAX) were determined by Western blotting technique.

RESULTS

Flap survival rate significantly improved and neutrophil recruitment and release were enhanced after treatment with the compound. Angiogenesis was promoted. L-borneol protected against oxidative stress by increasing superoxide dismutase activity and decreasing malondialdehyde content. It downregulated the hypoxia-inducible factor nuclear transcription factor-κB pathway, leading to the inhibition of several inflammatory factors. Simultaneously, it facilitated the expression of vascular endothelial growth factor and BCL-2.

CONCLUSION

The study shows that L-borneol may promote skin flap survival by inhibiting HIF-1α/NF-κB pathway.

Rivaroxaban down-regulates pyroptosis and the TLR4/NF-κB/NLRP3 signaling pathway to promote flap survival

BACKGROUND

Flap placement remains the primary method for wound repair, but postoperative ischemic flap necrosis is of major concern. This study explored whether rivaroxaban, a factor Xa inhibitor, enhanced flap survival.

METHODS

Thirty-six rats were randomly divided into control, low-dose rivaroxaban (3 mg/kg/day), and high-dose rivaroxaban (7 mg/kg/day) groups. On postoperative day 7, the flap survival rate was analyzed and the average survival area calculated. After the rats were euthanized, immunological and molecular biological techniques were employed to assess vascular regeneration, pyroptosis, and inflammation.

RESULTS

Rivaroxaban upregulated VEGF expression, in turn enhancing angiogenesis, and it downregulated IL-1β, IL-6, and TNF-α expression, thereby mitigating inflammation. The drug also suppressed TLR4, NF-κB p65, NLRP3, caspase-1, and IL-18 syntheses, thus inhibiting pyroptosis.

CONCLUSIONS

Rivaroxaban enhanced random flap survival by down-regulating the TLR4/NF-κB/NLRP3 signaling pathway to suppress pyroptosis, promoting vascular regeneration and inhibiting inflammation.

Higher-order combinatorial chromatin perturbations by engineered CRISPR-Cas12a for functional genomics

Multiplexed genetic perturbations are critical for testing functional interactions among coding or non-coding genetic elements. Compared to double-stranded DNA cutting, repressive chromatin formation using CRISPR interference (CRISPRi) avoids genotoxicity and is more effective for perturbing non-coding regulatory elements in pooled assays. However, current CRISPRi pooled screening approaches are limited to targeting 1-3 genomic sites per cell. To develop a tool for higher-order ( > 3) combinatorial targeting of genomic sites with CRISPRi in functional genomics screens, we engineered an Acidaminococcus Cas12a variant -- referred to as mul tiplexed transcriptional interference AsCas12a (multiAsCas12a). multiAsCas12a incorporates a key mutation, R1226A, motivated by the hypothesis of nicking-induced stabilization of the ribonucleoprotein:DNA complex for improving CRISPRi activity. multiAsCas12a significantly outperforms prior state-of-the-art Cas12a variants in combinatorial CRISPRi targeting using high-order multiplexed arrays of lentivirally transduced CRISPR RNAs (crRNA), including in high-throughput pooled screens using 6-plex crRNA array libraries. Using multiAsCas12a CRISPRi, we discover new enhancer elements and dissect the combinatorial function of cis-regulatory elements. These results instantiate a group testing framework for efficiently surveying potentially numerous combinations of chromatin perturbations for biological discovery and engineering.

GeneGPT: augmenting large language models with domain tools for improved access to biomedical information

MOTIVATION

While large language models (LLMs) have been successfully applied to various tasks, they still face challenges with hallucinations. Augmenting LLMs with domain-specific tools such as database utilities can facilitate easier and more precise access to specialized knowledge. In this article, we present GeneGPT, a novel method for teaching LLMs to use the Web APIs of the National Center for Biotechnology Information (NCBI) for answering genomics questions. Specifically, we prompt Codex to solve the GeneTuring tests with NCBI Web APIs by in-context learning and an augmented decoding algorithm that can detect and execute API calls.

RESULTS

Experimental results show that GeneGPT achieves state-of-the-art performance on eight tasks in the GeneTuring benchmark with an average score of 0.83, largely surpassing retrieval-augmented LLMs such as the new Bing (0.44), biomedical LLMs such as BioMedLM (0.08) and BioGPT (0.04), as well as GPT-3 (0.16) and ChatGPT (0.12). Our further analyses suggest that: First, API demonstrations have good cross-task generalizability and are more useful than documentations for in-context learning; second, GeneGPT can generalize to longer chains of API calls and answer multi-hop questions in GeneHop, a novel dataset introduced in this work; finally, different types of errors are enriched in different tasks, providing valuable insights for future improvements.

AVAILABILITY AND IMPLEMENTATION

The GeneGPT code and data are publicly available at https://github.com/ncbi/GeneGPT.

Tetrandrine promotes the survival of the random skin flap via the PI3K/AKT signaling pathway

Flaps are mainly used for wound repair. However, postoperative ischemic necrosis of the distal flap is a major problem, which needs to be addressed urgently. We evaluated whether tetrandrine, a compound found in traditional Chinese medicine, can prolong the survival rate of random skin flaps. Thirty-six rats were randomly divided into control, low-dose tetrandrine (25 mg/kg/day), and high-dose tetrandrine (60 mg/kg/day) groups. On postoperative Day 7, the flap survival and average survival area were determined. After the rats were sacrificed, the levels of angiogenesis, apoptosis, and inflammation in the flap tissue were detected with immunology and molecular biology analyses. Tetrandrine increased vascular endothelial growth factor and Bcl-2 expression, in turn promoting angiogenesis and anti-apoptotic processes, respectively. Additionally, tetrandrine decreased the expression of Bax, which is associated with the induction of apoptosis, and also decreased inflammation in the flap tissue. Tetrandrine improved the survival rate of random flaps by promoting angiogenesis, inhibiting apoptosis, and reducing inflammation in the flap tissue through the modulation of the PI3K/AKT signaling pathway.

Improving large language models for clinical named entity recognition via prompt engineering

IMPORTANCE

The study highlights the potential of large language models, specifically GPT-3.5 and GPT-4, in processing complex clinical data and extracting meaningful information with minimal training data. By developing and refining prompt-based strategies, we can significantly enhance the models' performance, making them viable tools for clinical NER tasks and possibly reducing the reliance on extensive annotated datasets.

OBJECTIVES

This study quantifies the capabilities of GPT-3.5 and GPT-4 for clinical named entity recognition (NER) tasks and proposes task-specific prompts to improve their performance.

MATERIALS AND METHODS

We evaluated these models on 2 clinical NER tasks: (1) to extract medical problems, treatments, and tests from clinical notes in the MTSamples corpus, following the 2010 i2b2 concept extraction shared task, and (2) to identify nervous system disorder-related adverse events from safety reports in the vaccine adverse event reporting system (VAERS). To improve the GPT models' performance, we developed a clinical task-specific prompt framework that includes (1) baseline prompts with task description and format specification, (2) annotation guideline-based prompts, (3) error analysis-based instructions, and (4) annotated samples for few-shot learning. We assessed each prompt's effectiveness and compared the models to BioClinicalBERT.

RESULTS

Using baseline prompts, GPT-3.5 and GPT-4 achieved relaxed F1 scores of 0.634, 0.804 for MTSamples and 0.301, 0.593 for VAERS. Additional prompt components consistently improved model performance. When all 4 components were used, GPT-3.5 and GPT-4 achieved relaxed F1 socres of 0.794, 0.861 for MTSamples and 0.676, 0.736 for VAERS, demonstrating the effectiveness of our prompt framework. Although these results trail BioClinicalBERT (F1 of 0.901 for the MTSamples dataset and 0.802 for the VAERS), it is very promising considering few training samples are needed.

DISCUSSION

The study's findings suggest a promising direction in leveraging LLMs for clinical NER tasks. However, while the performance of GPT models improved with task-specific prompts, there's a need for further development and refinement. LLMs like GPT-4 show potential in achieving close performance to state-of-the-art models like BioClinicalBERT, but they still require careful prompt engineering and understanding of task-specific knowledge. The study also underscores the importance of evaluation schemas that accurately reflect the capabilities and performance of LLMs in clinical settings.

CONCLUSION

While direct application of GPT models to clinical NER tasks falls short of optimal performance, our task-specific prompt framework, incorporating medical knowledge and training samples, significantly enhances GPT models' feasibility for potential clinical applications.

PubTator 3.0: an AI-powered Literature Resource for Unlocking Biomedical Knowledge

PubTator 3.0 (https://www.ncbi.nlm.nih.gov/research/pubtator3/) is a biomedical literature resource using state-of-the-art AI techniques to offer semantic and relation searches for key concepts like proteins, genetic variants, diseases, and chemicals. It currently provides over one billion entity and relation annotations across approximately 36 million PubMed abstracts and 6 million full-text articles from the PMC open access subset, updated weekly. PubTator 3.0's online interface and API utilize these precomputed entity relations and synonyms to provide advanced search capabilities and enable large-scale analyses, streamlining many complex information needs. We showcase the retrieval quality of PubTator 3.0 using a series of entity pair queries, demonstrating that PubTator 3.0 retrieves a greater number of articles than either PubMed or Google Scholar, with higher precision in the top 20 results. We further show that integrating ChatGPT (GPT-4) with PubTator APIs dramatically improves the factuality and verifiability of its responses. In summary, PubTator 3.0 offers a comprehensive set of features and tools that allow researchers to navigate the ever-expanding wealth of biomedical literature, expediting research and unlocking valuable insights for scientific discovery.

[Clinical characteristics and associated factors of mild cognitive impairment in patients with common cardiovascular diseases]

Objective: To analyze the clinical characteristics of patients with common cardiovascular diseases (CVD, including hypertension, coronary heart disease, atrial fibrillation, and heart failure) combined with mild cognitive impairment (MCI) and explore the potential risk factors of MCI in patients with CVD. Methods: A total of 2 294 patients with common cardiovascular diseases who met the criteria at Cardiology Medical Center in Beijing Anzhen Hospital, Capital Medical University, from June 1, 2021, to January 5, 2022, were retrospectively included. The patients were divided into the normal cognitive function group (1 107 cases) and the MCI group (1 187 cases). Demographic information and CVD status were collected. The information of cognitive function were collected using the Montreal Cognitive Assessment (MoCA) and the Mini-Mental State Examination (MMSE) scales. The difference between normal cognitive function and MCI were compared and analyzed. The logistic regression analysis was used to explored risk factors of MCI in CVD patients. Results: A total of 2 294 patients aged (60.6±10.4) years were included, among whom there were 29.99% (688 cases) females. Compared with patients in the normal cognitive function group, patients in the MCI group were older [ (57.9±11.4) vs (63.1±8.9) years old, P<0.001], with a higher proportion of women [26.47% (293 cases) vs 33.28% (395 cases), P<0.001]; there was a higher proportion of patients suffering from hypertension in the MCI group [59.62% (660 cases) vs 64.62% (767 cases), P=0.014］, and more components of CVD [(1.68±0.62) vs (1.74±0.65) components, P=0.017]. The risk factors of MCI in patients with common CVD were increased age, increased depression score, combined with hypertension, and ≥3 common components of CVD, with OR (95%CI) of 1.043 (1.032-1.054), 1.021 (1.004-1.037), 1.151 (1.142-3.439), and 1.137 (1.023-1.797), respectively (all P values <0.05). Increasing education level was observed to be associated with reduced risk of MCI with OR (95%CI) of 0.319 (0.271-0.378) (P<0.05). Conclusions: The incidence of MCI was high in CVD patients. The risk factors of MCI in CVD patients included hypertension and≥3 common components of CVD.

Ang1/Tie2/VE-Cadherin Signaling Regulates DPSCs in Vascular Maturation

Adding dental pulp stem cells (DPSCs) to vascular endothelial cell-formed vessel-like structures can increase the longevity of these vessel networks. DPSCs display pericyte-like cell functions and closely assemble endothelial cells (ECs). However, the mechanisms of DPSC-derived pericyte-like cells in stabilizing the vessel networks are not fully understood. In this study, we investigated the functions of E-DPSCs, which were DPSCs isolated from the direct coculture of human umbilical vein endothelial cells (HUVECs) and DPSCs, and T-DPSCs, which were DPSCs treated by transforming growth factor beta 1 (TGF-β1), in stabilizing blood vessels in vitro and in vivo. A 3-dimensional coculture spheroid sprouting assay was conducted to compare the functions of E-DPSCs and T-DPSCs in vitro. Dental pulp angiogenesis in the severe combined immunodeficiency (SCID) mouse model was used to explore the roles of E-DPSCs and T-DPSCs in vascularization in vivo. The results demonstrated that both E-DPSCs and T-DPSCs possess smooth muscle cell-like cell properties, exhibiting higher expression of the mural cell-specific markers and the suppression of HUVEC sprouting. E-DPSCs and T-DPSCs inhibited HUVEC sprouting by activating TEK tyrosine kinase (Tie2) signaling, upregulating vascular endothelial (VE)-cadherin, and downregulating vascular endothelial growth factor receptor 2 (VEGFR2). In vivo study revealed more perfused and total blood vessels in the HUVEC + E-DPSC group, HUVEC + T-DPSC group, angiopoietin 1 (Ang1) pretreated group, and vascular endothelial protein tyrosine phosphatase (VE-PTP) inhibitor pretreated group, compared to HUVEC + DPSC group. In conclusion, these data indicated that E-DPSCs and T-DPSCs could stabilize the newly formed blood vessels and accelerate their perfusion. The critical regulating pathways are Ang1/Tie2/VE-cadherin and VEGF/VEGFR2 signaling.

Machine-learning-based performance comparison of two-dimensional (2D) and three-dimensional (3D) CT radiomics features for intracerebral haemorrhage expansion

AIM

To investigate the value of non-contrast CT (NCCT)-based two-dimensional (2D) radiomics features in predicting haematoma expansion (HE) after spontaneous intracerebral haemorrhage (ICH) and compare its predictive ability with the three-dimensional (3D) signature.

MATERIALS AND METHODS

Three hundred and seven ICH patients who received baseline NCCT within 6 h of ictus from two stroke centres were analysed retrospectively. 2D and 3D radiomics features were extracted in the manner of one-to-one correspondence. The 2D and 3D models were generated by four different machine-learning algorithms (regularised L1 logistic regression, decision tree, support vector machine and AdaBoost), and the receiver operating characteristic (ROC) curve was used to compare their predictive performance. A robustness analysis was performed according to baseline haematoma volume.

RESULTS

Each feature type of 2D and 3D modalities used for subsequent analyses had excellent consistency (mean ICC >0.9). Among the different machine-learning algorithms, pairwise comparison showed no significant difference in both the training (mean area under the ROC curve [AUC] 0.858 versus 0.802, all p>0.05) and validation datasets (mean AUC 0.725 versus 0.678, all p>0.05), and the 10-fold cross-validation evaluation yielded similar results. The AUCs of the 2D and 3D models were comparable either in the binary or tertile volume analysis (all p>0.5).

CONCLUSION

NCCT-derived 2D radiomics features exhibited acceptable and similar performance to the 3D features in predicting HE, and this comparability seemed unaffected by initial haematoma volume. The 2D signature may be preferred in future HE-related radiomic works given its compatibility with emergency condition of ICH.

[Performance verification and clinical application evaluation of D-dimer assay]

Objective: To evaluate the clinical application performance of a domestic D-dimer assay reagent (ADX D-dimer). Methods: A total of 546 residual sodium citrate anticoagulated plasma samples (530 of which were used for comparability validation and 16 for sample preparation of other validation components) were selected after the completion of clinical testing at Peking Union Medical College Hospital from Jun 2022 to May 2023. According to the American Clinical Laboratory Standards Institute (CLSI) guidelines, national health industry standards and relevant references, the performance of ADX D-dimer used in Sysmex CS 5100 fully automated coagulation analyzer which included accuracy, precision, linear range, carryover rate, interference resistance capability and reference interval were validated and the agreement compared with two mainstream imported detection reagents (reagent A: Vidas D-dimer reagent; reagent B: Innovance D-dimer detection reagent) was evaluated. The clinical diagnostic efficacy of the ADX D-dimer was evaluated using the ELISA D-dimer (reagent A) test results as criteria. Results: The linear correlation coefficient of the 6-point calibrated absorbance and target value was 0.998, the bias of accuracy met the requirements (-2.8%-8.4%), and the coefficient of variation (CV) of within-run and between-day precision of the two levels were 1.0%-2.7% and 2.7%-4.1%, respectively, which were less than the requirements of the manufacturer's statement and the national health industry standard. The linear range within 0.33-9.69 mg/L FEU was verified and the carryover rate was 0. There was no significant interference with the assay results at bilirubin F≤0.22 g/L, bilirubin C≤0.22 g/L, hemoglobin≤5.5 g/L and celiac≤2 800 FTU. The manufacturer's reference interval≤0.5 mg/L FEU was verified suitable for this laboratory. For 358 samples without suspicious heterophilic antibody whose D-dimer levels range from 0.06 to143.63 mg/L FEU, the correlation between ADX D-dimer and another two assay was good, with r values being 0.968 and 0.975, respectively, the percentage of deviation and relative deviation beyond the 95% confidence interval was 3.4%-4.5% and 5.3%-7.0%. The correlation between ADX D-dimer and ELISA D-dimer was better than that of reagent B in the concentration range of 0.06-1.00 mg/L FEU (r=0.858, 0.134). For 172 samples with heterophilic antibody, the correlation between ADX D-dimer and ELISA D-dimer was still good(r=0.827), with the percentage of deviation and relative deviation being 6.4% (11/172). The diagnostic efficacy was evaluated using 530 samples, and the sensitivity, specificity, positive predictive value, negative predictive value of ADX D-dimer was 97.4%, 77.6%, 91.9%, 91.9%. The area under the curve was 0.976 (95%CI: 0.964-0.987, P<0.001). Conclusion: The ADX D-dimer reagent has superior assay and diagnostic performance, and can meet the needs of clinical laboratories.

A deep network DeepOpacityNet for detection of cataracts from color fundus photographs

BACKGROUND

Cataract diagnosis typically requires in-person evaluation by an ophthalmologist. However, color fundus photography (CFP) is widely performed outside ophthalmology clinics, which could be exploited to increase the accessibility of cataract screening by automated detection.

METHODS

DeepOpacityNet was developed to detect cataracts from CFP and highlight the most relevant CFP features associated with cataracts. We used 17,514 CFPs from 2573 AREDS2 participants curated from the Age-Related Eye Diseases Study 2 (AREDS2) dataset, of which 8681 CFPs were labeled with cataracts. The ground truth labels were transferred from slit-lamp examination of nuclear cataracts and reading center grading of anterior segment photographs for cortical and posterior subcapsular cataracts. DeepOpacityNet was internally validated on an independent test set (20%), compared to three ophthalmologists on a subset of the test set (100 CFPs), externally validated on three datasets obtained from the Singapore Epidemiology of Eye Diseases study (SEED), and visualized to highlight important features.

RESULTS

Internally, DeepOpacityNet achieved a superior accuracy of 0.66 (95% confidence interval (CI): 0.64-0.68) and an area under the curve (AUC) of 0.72 (95% CI: 0.70-0.74), compared to that of other state-of-the-art methods. DeepOpacityNet achieved an accuracy of 0.75, compared to an accuracy of 0.67 for the ophthalmologist with the highest performance. Externally, DeepOpacityNet achieved AUC scores of 0.86, 0.88, and 0.89 on SEED datasets, demonstrating the generalizability of our proposed method. Visualizations show that the visibility of blood vessels could be characteristic of cataract absence while blurred regions could be characteristic of cataract presence.

CONCLUSIONS

DeepOpacityNet could detect cataracts from CFPs in AREDS2 with performance superior to that of ophthalmologists and generate interpretable results. The code and models are available at https://github.com/ncbi/DeepOpacityNet ( https://doi.org/10.5281/zenodo.10127002 ).

[Clinical characteristics of epileptic seizure in neurofibromatosis type 1 in 15 cases]

Objective: To summarize the clinical characteristics of epileptic seizure associated with neurofibromatosis type 1 (NF1). Methods: From January 2017 to July 2023 at Children's Hospital Capital Institute of Pediatrics, medical records of patients with both NF1 and epileptic seizure were reviewed in this case series study. The clinical characteristics, treatment and prognosis were analyzed retrospectively. Results: A total of 15 patients(12 boys and 3 girls) were collected. Café-au-lait macules were observed in all 15 patients. There were 6 patients with neurodevelopmental disorders and the main manifestations were intellectual disability or developmental delay. The age at the first epileptic seizure was 2.5 (1.2, 5.5) years. There were various seizure types, including generalized tonic-clonic seizures in 8 patients, focal motor seizures in 6 patients, epileptic spasm in 4 patients, tonic seizures in 1 patient, absence in 1 patient, generalized myoclonic seizure in 1 patient and focal to bilateral tonic-clonic seizure in 1 patient. Among 14 patients whose brain magnetic resonance imaging results were available, there were abnormal signals in corpus callosum, basal ganglia, thalamus or cerebellum in 6 patients, dilated ventricles of different degrees in 3 patients, blurred gray and white matter boundary in 2 patients, agenesis of corpus callosum in 1 patient and no obvious abnormalities in the other patients. Among 13 epilepsy patients, 8 were seizure-free with 1 or 2 antiseizure medications(ASM), 1 with drug resistant epilepsy was seizure-free after left temporal lobectomy, and the other 4 patients who have received 2 to 9 ASM had persistent seizures. One patient with complex febrile convulsion achieved seizure freedom after oral administration of diazepam on demand. One patient had only 1 unprovoked epileptic seizure and did not have another seizure without taking any ASM. Conclusions: The first epileptic seizure in NF1 patients usually occurs in infancy and early childhood, with the main seizure type of generalized tonic-clonic seizure and focal motor seizure. Some patients have intellectual disability or developmental delay. Most epilepsy patients achieve seizure freedom with ASM.

Neoadjuvant nivolumab plus chemotherapy versus chemotherapy for resectable NSCLC: subpopulation analysis of Chinese patients in CheckMate 816

BACKGROUND

Neoadjuvant nivolumab plus chemotherapy significantly improved event-free survival (EFS) and pathologic complete response (pCR) versus chemotherapy alone in patients with resectable non-small-cell lung cancer (NSCLC) in the global phase III CheckMate 816 study. Here, we report post hoc exploratory efficacy, safety, and surgical outcomes in the Chinese subpopulation of this study.

METHODS

Adults with stage IB-IIIA resectable NSCLC were randomized to receive nivolumab 360 mg plus chemotherapy or chemotherapy alone every 3 weeks for three cycles followed by surgery. Primary endpoints included EFS and pCR (both per blinded independent review). EFS and pCR results were from 14 October 2022, and 16 September 2020, database locks, respectively.

RESULTS

The Chinese subpopulation comprised 97 patients (nivolumab plus chemotherapy, 44; chemotherapy, 53). At 38.2 months of minimum follow-up, median EFS was not reached [95% confidence interval (CI) 23.4 months-not reached] in the nivolumab plus chemotherapy arm and 13.9 months (95% CI 8.3-34.3 months) in the chemotherapy arm (hazard ratio 0.47, 95% CI 0.25-0.88). pCR rates were 25.0% (95% CI 13.2% to 40.3%) and 1.9% (95% CI 0.0% to 10.1%), respectively (odds ratio 11.05; 95% CI 1.41-86.49). Of 97 Chinese patients, 36 (82%) in the nivolumab plus chemotherapy arm and 41 (77%) in the chemotherapy arm underwent definitive surgery. Grade 3-4 treatment-related adverse events occurred in 18/43 patients (42%) treated with nivolumab plus chemotherapy and 22/53 patients (42%) treated with chemotherapy.

CONCLUSIONS

Consistent with findings in the global study population of CheckMate 816, neoadjuvant nivolumab plus chemotherapy improved EFS and pCR versus chemotherapy in the Chinese subpopulation without impacting treatment tolerability or the feasibility of surgery. These findings support the use of nivolumab plus chemotherapy as a standard neoadjuvant treatment option for Chinese patients with resectable NSCLC.

Deep-GA-Net for Accurate and Explainable Detection of Geographic Atrophy on OCT Scans

Objective

To propose Deep-GA-Net, a 3-dimensional (3D) deep learning network with 3D attention layer, for the detection of geographic atrophy (GA) on spectral domain OCT (SD-OCT) scans, explain its decision making, and compare it with existing methods.

Design

Deep learning model development.

Participants

Three hundred eleven participants from the Age-Related Eye Disease Study 2 Ancillary SD-OCT Study.

Methods

A dataset of 1284 SD-OCT scans from 311 participants was used to develop Deep-GA-Net. Cross-validation was used to evaluate Deep-GA-Net, where each testing set contained no participant from the corresponding training set. En face heatmaps and important regions at the B-scan level were used to visualize the outputs of Deep-GA-Net, and 3 ophthalmologists graded the presence or absence of GA in them to assess the explainability (i.e., understandability and interpretability) of its detections.

Main Outcome Measures

Accuracy, area under receiver operating characteristic curve (AUC), area under precision-recall curve (APR).

Results

Compared with other networks, Deep-GA-Net achieved the best metrics, with accuracy of 0.93, AUC of 0.94, and APR of 0.91, and received the best gradings of 0.98 and 0.68 on the en face heatmap and B-scan grading tasks, respectively.

Conclusions

Deep-GA-Net was able to detect GA accurately from SD-OCT scans. The visualizations of Deep-GA-Net were more explainable, as suggested by 3 ophthalmologists. The code and pretrained models are publicly available at https://github.com/ncbi/Deep-GA-Net.

Financial Disclosures

The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Opportunities and challenges for ChatGPT and large language models in biomedicine and health

ChatGPT has drawn considerable attention from both the general public and domain experts with its remarkable text generation capabilities. This has subsequently led to the emergence of diverse applications in the field of biomedicine and health. In this work, we examine the diverse applications of large language models (LLMs), such as ChatGPT, in biomedicine and health. Specifically, we explore the areas of biomedical information retrieval, question answering, medical text summarization, information extraction and medical education and investigate whether LLMs possess the transformative power to revolutionize these tasks or whether the distinct complexities of biomedical domain presents unique challenges. Following an extensive literature survey, we find that significant advances have been made in the field of text generation tasks, surpassing the previous state-of-the-art methods. For other applications, the advances have been modest. Overall, LLMs have not yet revolutionized biomedicine, but recent rapid progress indicates that such methods hold great potential to provide valuable means for accelerating discovery and improving health. We also find that the use of LLMs, like ChatGPT, in the fields of biomedicine and health entails various risks and challenges, including fabricated information in its generated responses, as well as legal and privacy concerns associated with sensitive patient data. We believe this survey can provide a comprehensive and timely overview to biomedical researchers and healthcare practitioners on the opportunities and challenges associated with using ChatGPT and other LLMs for transforming biomedicine and health.

GRSF1 antagonizes age-associated hypercoagulability via modulation of fibrinogen mRNA stability

Age-associated hypercoagulability is accompanied by the increase of plasma levels of some coagulation factors including fibrinogen which may contribute to the increased risk of cardiovascular, cerebrovascular, and thrombotic diseases in elderly people. However, the underlying mechanism of increased plasma fibrinogen concentration during aging is still elusive. GRSF1 belongs to the heterogeneous nuclear ribonucleoproteins F/H (hnRNP F/H) subfamily. Here, we report that GRSF1 attenuates hypercoagulability via negative modulation of fibrinogen expression. We demonstrated that GRSF1 negatively regulated fibrinogen expression at both mRNA and protein levels. GRSF1 directly interacted with the coding region (CDS) of FGA, FGB, and FGG mRNAs, and decreased their stability thus mitigating fibrinogen expression. We further identified that only a few G-tracts within the Fib C domain of FGA, FGB, and FGG CDS and the qRRM2 domain of GRSF1 were required for their interaction. Moreover, we confirmed hypercoagulability and the decrease of GRSF1 expression level during mice aging. Functionally, GRSF1 overexpression in old mice liver decreased fibrinogen plasma level, reduced hypercoagulability, and mitigated blood coagulation activity, whereas GRSF1 knockdown in young mice liver increased fibrinogen plasma level and promoted blood coagulation activity. Collectively, our findings unveil a novel posttranscriptional regulation of fibrinogen by GRSF1 and uncover a critical role of GRSF1 in regulating blood coagulation activity.

MedCPT: Contrastive Pre-trained Transformers with large-scale PubMed search logs for zero-shot biomedical information retrieval

MOTIVATION

Information retrieval (IR) is essential in biomedical knowledge acquisition and clinical decision support. While recent progress has shown that language model encoders perform better semantic retrieval, training such models requires abundant query-article annotations that are difficult to obtain in biomedicine. As a result, most biomedical IR systems only conduct lexical matching. In response, we introduce MedCPT, a first-of-its-kind Contrastively Pre-trained Transformer model for zero-shot semantic IR in biomedicine.

RESULTS

To train MedCPT, we collected an unprecedented scale of 255 million user click logs from PubMed. With such data, we use contrastive learning to train a pair of closely integrated retriever and re-ranker. Experimental results show that MedCPT sets new state-of-the-art performance on six biomedical IR tasks, outperforming various baselines including much larger models, such as GPT-3-sized cpt-text-XL. In addition, MedCPT also generates better biomedical article and sentence representations for semantic evaluations. As such, MedCPT can be readily applied to various real-world biomedical IR tasks.

AVAILABILITY AND IMPLEMENTATION

The MedCPT code and model are available at https://github.com/ncbi/MedCPT.

Microstructure and Corrosion Behavior of Zinc/Hydroxyapatite Multi-Layer Coating Prepared by Combining Cold Spraying and High-Velocity Suspension Flame Spraying

The study aims to enhance the corrosion resistance and bioactivity of Mg alloy substrates through the development of a zinc/hydroxyapatite multi-layer (Zn/HA-ML) coating. The Zn/HA-ML coating was prepared by depositing a cold-sprayed (CS) Zn underlayer and a high-velocity suspension flame sprayed (HVSFS) Zn/HA multi-layer and was compared with the CS Zn coating and the Zn/HA dual-layer (Zn/HA-DL) coating. Phase, microstructure, and bonding strength were examined, respectively, by X-ray diffraction, scanning electron microscopy, and tensile bonding testing. Corrosion behavior and bioactivity were investigated using potentiodynamic polarization, electrochemical impedance spectroscopy, and immersion testing. Results show that the HVSFS Zn/HA composite layers were mainly composed of Zn, HA, and ZnO and were well bonded to the substrate. The HVSFS HA upper layer on the CS Zn underlayer in the Zn/HA-DL coating exhibited microcracks due to their mismatched thermal expansion coefficient (CTE). The Zn/HA-ML coating exhibited good bonding within different layers and showed a higher bonding strength of 27.3 ± 2.3 MPa than the Zn/HA-DL coating of 20.4 ± 2.7 MPa. The CS Zn coating, Zn/HA-DL coating, and Zn/HA-ML coating decreased the corrosion current density of the Mg alloy substrate by around two-fourfold from 3.12 ± 0.75 mA/cm2 to 1.41 ± 0.82mA/cm2, 1.06 ± 0.31 mA/cm2, and 0.88 ± 0.27 mA/cm2, respectively. The Zn/HA-ML coating showed a sixfold decrease in the corrosion current density and more improvements in the corrosion resistance by twofold after an immersion time of 14 days, which was mainly attributed to newly formed apatite and corrosion by-products of Zn particles. The Zn/HA-ML coating effectively combined the advantages of the corrosion resistance of CS Zn underlayer and the bioactivity of HVSFS Zn/HA multi-layers, which proposed a low-temperature strategy for improving corrosion resistance and bioactivity for implant metals.

Opportunities and Challenges for ChatGPT and Large Language Models in Biomedicine and Health

ChatGPT has drawn considerable attention from both the general public and domain experts with its remarkable text generation capabilities. This has subsequently led to the emergence of diverse applications in the field of biomedicine and health. In this work, we examine the diverse applications of large language models (LLMs), such as ChatGPT, in biomedicine and health. Specifically we explore the areas of biomedical information retrieval, question answering, medical text summarization, information extraction, and medical education, and investigate whether LLMs possess the transformative power to revolutionize these tasks or whether the distinct complexities of biomedical domain presents unique challenges. Following an extensive literature survey, we find that significant advances have been made in the field of text generation tasks, surpassing the previous state-of-the-art methods. For other applications, the advances have been modest. Overall, LLMs have not yet revolutionized biomedicine, but recent rapid progress indicates that such methods hold great potential to provide valuable means for accelerating discovery and improving health. We also find that the use of LLMs, like ChatGPT, in the fields of biomedicine and health entails various risks and challenges, including fabricated information in its generated responses, as well as legal and privacy concerns associated with sensitive patient data. We believe this survey can provide a comprehensive and timely overview to biomedical researchers and healthcare practitioners on the opportunities and challenges associated with using ChatGPT and other LLMs for transforming biomedicine and health.

[Esophageal dysmotility detection in patients with gastroesophageal reflux-related cough and its influence on the efficacy of anti-reflux therapy]

Objective: To investigate the characteristics of esophageal dysmotility in patients with an initial diagnosis of acid/non-acid gastroesophageal reflux-related cough (GERC), and its correlation with the therapeutic response to anti-reflux treatments to search for the useful indicators to screen patients with chronic cough suitable for anti-reflux therapy. Methods: A total of 173 patients with suspicious GERC who attended the Chronic Cough Specialist Clinic of Department of Pulmonary and Critical Care Medicine, Tongji Hospital, School of Medicine, Tongji University between June 2020 and December 2022 were retrospectively selected for the study. The age of the patients was (45.1±14.6) years old, including 87 males and 86 females. Their demographic characteristics, clinical manifestations, and the results of high-resolution manometry (HRM) and multichannel intraluminal impedance-pH monitoring (MII-pH) were collected. Information on the etiological identification process and final diagnosis was also recorded. The recruited cases were grouped according to therapeutic outcomes and divided into cases with a favourable response to conventional anti-reflux treatment, cases with a favourable response to intensified anti-reflux treatment, and cases with no response to anti-reflux treatment. Factors influencing the efficacy of anti-reflux treatment were investigated. Differences between groups were compared using the χ2 test, Student-Newman-Keuls test, and Kruskal-Wallis H(K) test, where applicable. Logistic regression analysis using forward stepwise regression based on maximum likelihood estimation was used to screen for influence factors. Results: The 175 patients with suspicious GERC included 45 (26.0%) patients who responded to conventional anti-reflux treatment, 54 (31.2%) who responded to intensified anti-reflux treatment and 74 (42.8%) who did not respond to anti-reflux therapies. Esophageal dysmotility was present in 52.0% of patients (90/173), but was less common in patients who responded to conventional anti-reflux treatment (χ2=8.09, P=0.018). Although the majority of reflux episodes were non-acid (136/173, 78.6%), the proportion of acid reflux (χ2=19.49, P<0.001) and acid exposure time (H=11.04, P=0.004) were significantly higher in patients who responded to conventional anti-reflux treatment. The patients with acid and non-acid GERC had comparable proportion of esophageal dysmotility (64.9% vs. 48.5%, χ2=3.11, P=0.078), with a shorter break [2.4 (0.7, 5.6) cm vs. 6.1 (1.4, 10.0) cm, Z=-2.39, P=0.017], longer upper esophageal sphincter [(4.1±0.9) cm vs. (3.7±1.3) cm, t=-2.09, P=0.038], higher percentage of normal esophageal contractions [60.0% (17.8%, 90.0%) vs. 30.0% (0, 80.0%), Z=-2.14, P=0.032], and lower percentage of large break [10.0% (0, 40.0%) vs. 50.0% (0, 100.0%), Z=-2.92, P=0.004] in the patients with non-acid GERC. The mean resting pressure of the lower esophageal sphincter was significantly lower (H=7.49, P=0.024), while the percentage of ineffective esophageal contractions was markedly higher (H=8.60, P=0.014) in the patients who responded to intensified anti-reflux treatment and in the patients who did not respond to the anti-reflux therapies. Multifactorial logistic regression analysis identified the percentage of ineffective contraction as an independent factor predicting the efficacy of conventional anti-reflux treatment, with a cut-off value of≤45% and a moderate predictive value (AUC=0.67, P=0.004). Conclusions: Esophageal dysmotility is common in GERC patients with different characteristics in acid and non-acid GERC. The percentage of ineffective esophageal contraction may be a useful indicator for selecting an anti-reflux strategy and predicting treatment outcomes.

Measurement of oesophageal hiatus surface area by multiplanar reconstruction of MDCT: relationship with lower oesophageal sphincter pressure and acid reflux

AIM

To evaluate the relationship between oesophageal hiatus surface area (OHSA) and gastro-oesophageal reflux disease (GERD).

MATERIALS AND METHODS

Patients who underwent 24-h pH monitoring, oesophageal high-resolution manometry, and upper abdominal contrast-enhanced multidetector computed tomography (MDCT) during 2014-2021 were enrolled. Patients with a hiatus hernia (HH) on MDCT or who had a history of gastro-oesophageal surgery were excluded. Multiplanar reconstruction (MPR) of the MDCT image was used for the measurement of OHSA. Correlations of OHSA with acid exposure time (AET) and lower oesophageal sphincter (LOS) pressure of all patients were analysed.

RESULTS

Seventy-eight patients were included in the study. OHSA was much less in the AET <4% group than in the AET >6% group (1.61 ± 0.42 versus 2.09 ± 0.55 cm2, p<0.001). Correlation analysis reveals that OHSA correlated positively with AET (correlation coefficient = 0.47, p<0.001). Receiver operating characteristic (ROC) curve analysis reveals that OHSA can significantly distinguish patients in different groups divided by AET (area under the ROC curve [AUC] = 0.76, 95% confidence interval [CI]: 0.63-0.90). OHSA was not related to LOS pressure (correlation coefficient = -0.268, p=0.051). There was no difference in OHSA between the low LOS pressure group and the normal LOS pressure group (1.84 ± 0.61 versus 1.74 ± 0.50 cm2, p=0.52).

CONCLUSIONS

OHSA significantly correlated with AET but has no relationship with LOS pressure. It may be an independent risk factor of GERD.

BioREx: Improving biomedical relation extraction by leveraging heterogeneous datasets

Biomedical relation extraction (RE) is the task of automatically identifying and characterizing relations between biomedical concepts from free text. RE is a central task in biomedical natural language processing (NLP) research and plays a critical role in many downstream applications, such as literature-based discovery and knowledge graph construction. State-of-the-art methods were used primarily to train machine learning models on individual RE datasets, such as protein-protein interaction and chemical-induced disease relation. Manual dataset annotation, however, is highly expensive and time-consuming, as it requires domain knowledge. Existing RE datasets are usually domain-specific or small, which limits the development of generalized and high-performing RE models. In this work, we present a novel framework for systematically addressing the data heterogeneity of individual datasets and combining them into a large dataset. Based on the framework and dataset, we report on BioREx, a data-centric approach for extracting relations. Our evaluation shows that BioREx achieves significantly higher performance than the benchmark system trained on the individual dataset, setting a new SOTA from 74.4% to 79.6% in F-1 measure on the recently released BioRED corpus. We further demonstrate that the combined dataset can improve performance for five different RE tasks. In addition, we show that on average BioREx compares favorably to current best-performing methods such as transfer learning and multi-task learning. Finally, we demonstrate BioREx's robustness and generalizability in two independent RE tasks not previously seen in training data: drug-drug N-ary combination and document-level gene-disease RE. The integrated dataset and optimized method have been packaged as a stand-alone tool available at https://github.com/ncbi/BioREx.

Evaluation of a Deep-Learning Auto-Segmentation Model of Cardiac Substructures

PURPOSE/OBJECTIVE(S)

Increasing evidence has suggested that limiting dose not only to the whole heart but also to cardiac substructures can potentially reduce cardiac toxicities. Manual contouring of cardiac substructure can be challenging and time-consuming. To address this concern, we developed a deep learning (DL) model, trained on convolutional neural network algorithms in large external datasets, for auto-segmentation of cardiac substructures. This study aimed to evaluate the quality of the cardiac substructure contours generated by the DL algorithm.

MATERIALS/METHODS

We identified 28 patients with esophagus or gastroesophageal junction cancer from a single institution who received radiation to the esophagus between January 2017 and December 2022. For each case, the DL-generated cardiac substructures (4 heart chambers - left/right atrium [L/RA] and L/R ventricle [L/RV], 4 coronary arteries - L common [LCA], L anterior descending [LAD], L circumflex [LCx], and R common [RCA], and great vessels - ascending aorta [AA], pulmonary artery [PA], and superior vena cava [SVC]) were modified by two radiation oncologists (RO) using the contouring atlas developed by Duane et al. Spatial overlapping of the contours were then assessed using the Dice similarity coefficient (DSC), 95th percentile Hausdorff distance (HD-95), and normalized surface dice at 2 mm tolerance (NSD-2).

RESULTS

The mean values of DSC, HD-95, and NSD-2 are shown in Table 1. Overall, the mean DSC, HD-95, and NSD-2 for the heart chambers ranged from 0.82 to 0.92, 0.40 cm to 1.52 cm, and 0.68 to 0.85, respectively. Ranges of the mean DSC, HD-95, and NSD-2 for the coronary vessels were 0.41 to 0.74, 0.18 cm to 0.98 cm, and 0.66 to 0.77, respectively. Lastly, comparison of the great vessel contours yielded the following ranges for mean DSC, HD-95, and NSD-2 respectively: 0.72 to 0.92, 0.30 cm to 1.64 cm, and 0.65 to 0.83.

CONCLUSION

Our study demonstrates that auto-segmentation of cardiac substructures by DL-powered models can be comparable to manual contours for certain cardiac substructures, namely the four heart chambers and great vessels. Further improvement of the DL on contouring of coronary vessels would be needed prior to the autosegmentation model being widely adopted.

Development and Validation of a Deep Learning-Based Auto-Delineation of Target Volume and Organs at Risk in Pancreatic Cancer Radiotherapy

PURPOSE/OBJECTIVE(S)

The delineation of the clinical target volume (CTV), gross target volume (GTV) and organs at risk (OARs) is a crucial and laborious in pancreatic cancer radiotherapy. In this work, we propose and evaluate a three-dimensional (3D) novel convolutional neural network (CNN) for automatic and accurate CTV, GTV and OARs in pancreatic cancer.

MATERIALS/METHODS

A total of 120 computed tomography (CT) scans patients with pancreatic cancer were collected. A novel 3D CNN network, called ResUNet3D, was developed to achieve auto-delineation. 96 patients chosen randomly were used for training, 12 patients for validation, and 12 patients for testing. Meanwhile, the Dice similarity coefficient (DSC) and 95th percentile Hausdorff distance (HD95%) were used to assess the performance.

RESULTS

The DSC values for the test data were 80.9±8.6%, 77.5±5.6%, 94.5±1.3%, 66.2±13.4%, 73.6±7.6%, 79.0±8.7%, 94.1±1.9%, 94.6±1.4%, 87.3±5.8% for CTV, GTV, liver, duodenum, spinal cord, bowel, kidney left, kidney right, stomach. The corresponding HD95% values were 10.7±6.9mm, 7.8±5.7mm, 11.6±5.6mm, 18.6±5.6mm, 2.7±0.7mm, 17.7±8.6mm, 3.9±1.4mm, 3.7±1.9mm, 13.4±5.7mm, respectively. The average delineation time for one patient's CT images was within 5 seconds.

CONCLUSION

The experimental results demonstrate that the CTV, GTV and OARs delineated for pancreatic cancer by ResUNet3D achieved a close agreement with the ground truth. ResUNet3D could significantly reduce the radiation oncologists' contouring time.

Pilot Study of a Novel Ring Gantry-Based PET/CT Linear Accelerator in Patients with Prostate Cancer Receiving [18F]-DCFPyL for PSMA PET Imaging

PURPOSE/OBJECTIVE(S)

The RefleXion X1® system is a hybrid PET imaging-radiotherapy system that uses real-time positron emissions from a PET tracer to deliver biologically guided radiotherapy (BgRT). This study (NCT05470699) evaluated the hypothesis that the X1 PET imaging subsystem would be able to detect [18F]-DCFPyL PSMA PET signal sufficient to generate a deliverable BgRT plan in patients with prostate cancer.

MATERIALS/METHODS

Patients with prostate cancer scheduled for a diagnostic [18F]-DCFPyL PSMA PET scan as part of standard of care were eligible. Upon completion of the diagnostic PSMA PET scan, images were transferred to the radiotherapy planning system for target identification and contouring. If at least one PET avid tumor lesion was identified, the patient was then scanned on the X1 unit. BgRT planning was performed on each X1 scanned patient. The target lesion volume, activity concentration (AC) and normalized target signal (NTS) were acquired. Successful and deliverable BgRT plans required that the target AC was ≥ 5 kBq/ml and NTS ≥ 2.7.

RESULTS

Twenty-six patients underwent [18F]-DCFPyL PET scans (13 with rising PSA after surgery or radiotherapy, 6 with known metastases and 7 with newly diagnosed high-risk prostate cancer). Median (range) PSA was 3.40 (0.04-122). In 16 patients a PET avid tumor was identified and contoured for planning (4 lymph nodes, 5 bone, 6 prostate gland, and 1 prostate bed). In 13 patients the target lesion was visualized on the X1 PET scan, while in 3 patients the target lesion was too close to the bladder to be clearly visualized. BgRT planning was feasible and met standard of care published SBRT organ dose constraints in 8 patients (3 prostate gland, 3 bone, 2 lymph nodes). BgRT planning was not feasible in 8 patients due to insufficient AC, low NTS or proximity of the target lesion to the PET avid bladder. The accompanying table compares median (range) target volume, AC and NTS for feasible versus not feasible plans.

CONCLUSION

This is the first study to investigate the feasibility of using [18F]-DCFPyL PET imaging for BgRT plan generation on the X1 system in patients with prostate cancer. Lesions that are relevant to radiotherapy of prostate cancer can be visualized including lymph node and bone metastases. A dedicated BgRT workflow with PSMA PET imaging on the X1 at 60 minutes post injection will result in higher target AC and will optimize BgRT planning. PET avid lesions < 1 cm or close to the bladder may make BgRT planning challenging. [18F]-DCFPyL-guided BgRT is technically feasible using the RefleXion X1. BgRT using targeted PET radiopharmaceuticals to biologically guide external beam radiotherapy represents a promising new dimension in radiation oncology and warrants further investigation.

Radiotherapy for Patients with Locally Advanced Esophageal Squamous Cell Carcinoma Receiving Neoadjuvant Immunotherapy Combined with Chemotherapy

PURPOSE/OBJECTIVE(S)

With the success of immunotherapy in advanced esophageal cancer, neoadjuvant chemo-immunotherapy (CIT) is being increasingly used for local staged esophageal cancer, especially in the context of clinical trials, which brings similar pCR with neoadjuvant chemoradiotherapy and shows promising results. However, there is still a part of potentially operable patients can't undergo surgery after neoadjuvant chemo-immunotherapy. The follow-up treatment and prognosis of this population remain unclear.

MATERIALS/METHODS

Patients pathologically diagnosed with ESCC, clinical stage T1-3N+M0 or T3-4aNanyM0(AJCC 8th), PS 0-1 were retrospectively enrolled from 1/2020 to 6/2021 in Zhejiang Cancer Hospital. All patients firstly received PD-1 inhibitors (Camrelizumab, Sintilimab or Tislelizumab) plus chemotherapy (albumin paclitaxel,260 mg/m²on day 1 plus carboplatin AUC = 5 on day 1) every 3 weeks for 2-4 cycles. For those patients who did not receive surgery, definitive radiotherapy with 50.4Gy/28F or 50Gy/25F was adopted using VMAT, concurrent with chemotherapy or alone. The concurrent chemotherapy regimens included weekly TC (paclitaxel 50 mg/m 2, d1, carboplatin AUC = 2, d1) or S1 (60mg bid d1-14,29-42). The survival outcomes and treatment toxicity were recorded and analyzed.

RESULTS

A total of 56 eligible patients were finally identified from 558 patients who were treated in department of thoracic surgery, 31 patients showed no response to neoadjuvant CIT (6 with PD and 25 with SD), 25 patients achieved PR but did not receive surgery due to poor performance status or refuse to operation. Median age was 66(IQR 56-72) and 55(98.2%) were males. 12(19.6%) were stage II and 44(80.4%) were stage III. Among all the patients,25 (44.6%) received radiotherapy alone, and 31 (55.4%) received chemoradiotherapy after neoadjuvant CIT. The median follow-up was 11.8 months (IQR 8.6-20.1). The median PFS and OS were 16.5 months (95CI 12.9-21.4) and 18.6 months (95CI 11.2-NA), respectively. In the subgroup analysis, the median PFS for patients with PR to CIT was 20.2 moths (95CI:17.23-NA), and 12.9 moths (95CI: 0.68-20.4) for patients with SD or PD, HR was 0.45 (95CI:0.22- 0.93, P = 0.027). No significant difference was observed for patients received radiotherapy alone or chemoradiotherapy with HR = 1.36(95CI:0.69-2.71, P = 0.37). The most common AEs observed during this study were anemia (98.2%), Leukopenia (83.9%), Thrombocytopenia (53.6%). Adverse events of grade≥3 radiation-induced pneumonitis and esophagitis were 12.5% and 32.1%, especially, 6 patients (10.7%) died from esophageal fistula and 2 patients (3.6%) died from grade 5 pneumonitis.

CONCLUSION

For local advanced ESCC patients after neoadjuvant CIT who did not receive surgery, definitive radiotherapy was an optional treatment strategy. However, those patients with no response to CIT also showed poor response to radiotherapy, and particular attention should be paid to treatment related toxicity, especially esophageal fistula.

Dosimetric Plan Evaluation of Biology Guided Radiotherapy Using [18F]-DCFPyL PSMA Radiotracer in Patients with Prostate Cancer

PURPOSE/OBJECTIVE(S)

The X1 system represents a cutting-edge solution in radiotherapy delivery, with its capability to perform Biology Guided Radiotherapy (BgRT). The system utilizes real-time positron emission tomography (PET) signal as biological fiducials to provide tracked dose delivery and is initially available for use with [18F]-Fluorodeoxyglucose (FDG). The aim of this research study is to assess the quality of BgRT treatment plans for prostate cancer using patients' PSMA PET images obtained on the X1 system.

MATERIALS/METHODS

Sixteen patients with at least one PET-avid tumor identified on their whole-body diagnostic PSMA PET scan were selected. These patients were scanned on X1 following their diagnostic scan without additional radiotracer administration. Based on the X1 PET images, a BgRT plan was created for each patient, with the prescription dose determined by the location of treatment sites. The planning objectives of organs-at-risk (OARs) were established in accordance with the 2018 Timmerman guidelines. Target coverage objective was the dose covering 95% (D95%) of the planning target volume (PTV) to be higher than 100%. The following parameters were analyzed: PTV D95%, the minimal dose (Dmin) of gross tumor volume (GTV), plan maximum dose (Dmax), conformity index (CI), gradient index (GI), and maximum point dose (D0.03cc) to the nearest OARs. The X1 BgRT planning system also generated dose volume histogram (DVH) bounds, which model variations in BgRT delivery. The low boundary of GTV Dmin, representing the minimum GTV dose in the worst-case scenario, was recorded.

RESULTS

BgRT plans were created for all patients, except for one where the target signal was indistinguishable from the bladder. The prescription dose was 2700 cGy or 3000 cGy in 3 fractions for lymph node lesions, 2400 cGy to 3000 cGy in 3 fractions for bone metastasis, and 4500 cGy in 5 fractions for lesions in prostate. All plans met the dose constraints for OARs as per the Timmerman guidelines. The Dmax of all plans was 129.9% ± 6.9% (mean ± standard deviation). The PTV D95% and GTV Dmin were 101.7% ± 1.0% and 111.0% ± 7.6%, respectively. The low boundary of GTV Dmin was 95.9% ± 5.8%. The CI and GI were 1.22 ± 0.11 and 9.40 ± 2.12, respectively. The D0.03cc to nearest OARs was 84.6% ± 25.4%. The estimated treatment time was 699 ± 228 seconds.

CONCLUSION

This study is a pioneering effort to evaluate the quality of BgRT plans for prostate cancer patients using the [18F]-DCFPyL PSMA radiotracer. Our results showed that all BgRT plans met the planning objectives defined in the Timmerman protocol. BgRT with [18F]-DCFPyL represents a promising treatment modality for patients with prostate cancer. Further research is needed to validate this approach, including a comprehensive assessment of the dosimetric and tracking accuracy through physical measurements.

Different Radiation Dose of Neoadjuvant Chemoradiation for Resectable Thoracic Esophageal Squamous Carcinoma: A Randomized Phase II Clinical Trial

PURPOSE/OBJECTIVE(S)

Radiation dose used in the neoadjuvant chemoradiotherapy (nCRT) for patients with locally advanced esophageal squamous cell carcinoma (ESCC) varies in clinical practice, no prospective studies have been conducted on this issue. The hypothesis of this study is that higher radiation dose in nCRT was associated with improved progression free survival in resectable ESCC.

MATERIALS/METHODS

A phase II prospective randomized controlled trial was conducted from February 22, 2018 to February 22, 2021 in Zhejiang Cancer Hospital (NCT03381651). Patient with locally resectable advanced thoracic ESCC (cT2-T4aNxM0, T1-3N+M0) were randomized 1:1 to receive nCRT with different radiation doses of 50.4 Gy/28F and 41.4 Gy/23F, weekly TC regimen (paclitaxel 50 mg/m2, d1, carboplatin AUC = 2, d1) was administered as concurrent chemotherapy. The primary endpoint was 2-year progression free survival (PFS), and the secondary endpoints included complete pathological response (pCR) rate, overall survival (OS), local recurrence free survival (LRFS), and distant metastasis free survival (DMFS). The last follow-up date was February 22, 2022 and data analysis was performed from May 1, 2022, to October 31, 2022.

RESULTS

A total of 147 patients were enrolled in the prospective study, including 72 patients in 50.4 Gy/28F group and 75 patients in 41.4 Gy/23F group. Among all the patients, 101 patients finally underwent surgical resection, 23/46(50%) achieved pCR in the high-dose group, while 18/55(32.7%) achieved pCR in the low-dose group (P = 0.10). In intention-to-treat population analysis, 2-year PFS in the high-dose and low-dose groups were 53.5% (95% CI: 42.9-66.7%) and 45.5% (95% CI: 35.4%-58.5%), respectively, and hazard ratio (HR) was 0.72 (95% CI: 0.46-1.12, P = 0.14). Median PFS were 44.4 months (95% CI: 16-NA) versus 20.8 months (95% CI: 15-NA) in high and low dose groups. 2-year OS in the high and low-dose groups were 61.8% (95% CI: 51.2 -74.5%) and 60.5% (95% CI: 50.3 -72.8%), respectively, with a HR = 0.78(95% CI: 0.48-1.27, P = 0.32). Median OS were not arrived in both groups up to the last follow-up. In addition, 2-year LRFS were 87.9% (95% CI: 78.4-98.5%) and 74.8% (95% CI: 63.3-88.3%) in high and low-dose groups, while 2-year DMFS were 70.6% (95% CI: 58.3-85.5%) and 73.6% (95% CI: 61.7-87.8%), respectively. No significant statistical difference was observed between the two groups in terms of side effects during nCRT and postoperative complications.

CONCLUSION

Our results showed that neoadjuvant high-dose radiotherapy failed to improve overall survival in ESCC. However, it was found that pCR and local tumor control rates tended to be increased at 50.4 Gy/28F, compared to 41.4 Gy/23F, although the trend was not statistically significant. There was no significant difference in treatment side effects between the two groups. Longer-term follow-up is required to verify our findings.

The radiomic biomarker in non-small cell lung cancer: 18F-FDG PET/CT characterisation of programmed death-ligand 1 status

AIM

To present an integrated 2-[18F]-fluoro-2-deoxy-d-glucose (18F-FDG) positron-emission tomography (PET)/computed tomography (CT) radiomic characterisation of programmed death-ligand 1 (PD-L1) status in non-small-cell lung cancer (NSCLC).

MATERIALS AND METHODS

In this retrospective study, 18F-FDG PET/CT images and clinical data of 394 eligible patients were divided into training (n=275) and test sets (n=119). Next, the corresponding nodule of interest was segmented manually on the axial CT images by radiologists. After which, the spatial position matching method was used to match the image positions of CT and PET, and radiomic features of the CT and PET images were extracted. Radiomic models were built using five different machine-learning classifiers and the performance of the radiomic models were further evaluated. Finally, a radiomic signature was established to predict the PD-L1 status in patients with NSCLC using the features in the best performing radiomic model.

RESULTS

The radiomic model based on the PET intranodular region determined using the logistic regression classifier preformed best, yielding an area under the receiver operating characteristics curve (AUC) of 0.813 (95% CI: 0.812, 0.821) on the test set. The clinical features did not improve the test set AUC (0.806, 95% CI: 0.801, 0.810). The final radiomic signature for PD-L1 status was consisted of three PET radiomic features.

CONCLUSION

This study showed that an 18F-FDG PET/CT-based radiomic signature could be used as a non-invasive biomarker to discriminate PD-L1-positive from PD-L1-negative in patients with NSCLC.

[The official-private game of ginseng trade between China and Korea in modern times]

As culturally friendly countries, China and the Korean Peninsula have had active medical exchanges throughout history, which had a significant impact on the traditional medicine industry and daily life of both sides. Ginseng is an important commodity in the drug trade between China and Korea, and its profits are lucrative and an important source of fiscal revenue for the Korean government.In modern times, lured by huge profits, some Korean and Chinese businessmen have engaged in ginseng smuggling, even government officials.In order to safeguard the interests of the official camp, the two governments have always severely punished smuggling, but ginseng smuggling has not disappeared, and the game between official and private ginseng has become a unique phenomenon in the medical exchanges between China and Korea.

Andrographolide loaded montmorillonite attenuated enterotoxigenic Escherichia coli induced intestinal barrier injury and inflammation in a mouse model

Montmorillonite (MMT), a natural absorbent agent, has widely been accepted for its antidiarrhea function in human and farm animals; however, its specific physicochemical property limits its biological function in practical use. In the current study, raw MMT was loaded by andrographolide, namely andrographolide loaded montmorillonite (AGP-MMT). The microstructure of AGP-MMT was observed by scanning electron microscope (SEM) and X-ray diffraction (XRD). The effect of AGP-MMT on the growth performance, intestinal barrier and inflammation was investigated in an enterotoxigenic Escherichia coli (ETEC) challenged mice model. The results show that the microstructure of MMT was obviously changed after andrographolide modification: AGP-MMT exhibited a large number of spheroid particles, and floccule aggregates, but lower interplanar spacing compared with MMT. ETEC infection induced body weight losses and intestinal barrier function injury, as indicated by a lower villus height and ratio of villus height/crypt depth, whereas the serum levels of diamine oxidase (DAO), D-xylose and ETEC shedding were higher in the ETEC group compared with the CON group. Mice pretreated with AGP-MMT showed alleviated body weight losses and the intestinal barrier function injury induced by ETEC challenge. The villus height and the ratio of villus height/crypt depth, were higher in mice pretreated with AGP-MMT than those pretreated with equal levels of MMT. Pretreatment with AGP-MMT also alleviated the increased concentration of serum tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β), and the corresponding genes in the jejunum induced by ETEC infection in mice. The protein and mRNA levels of IL-1β were lower in mice pretreated with AGP-MMT than those with equal levels of MMT. The results indicate that AGP-MMT was more effective in alleviating intestinal barrier injury and inflammation in mice with ETEC challenge than MMT.

[Traditional Chinese medicine for treatment of echinococcosis: a review]

Echinococcosis is a zoonotic parasitic disease caused by infection with Echinococcus species. As the drug of first choice for treatment of echinococcosis, albendazole suffers from problems of large doses and remarkable adverse reactions in clinical therapy. Development of novel drugs against echinococcosis is of urgent need. Recently, great advances have been achieved in the research on traditional Chinese medicine for treatment of echinococcosis. This review summarizes the progress of researches on traditional Chinese medicine for treatment of echinococcosis, aiming to provide insights into development of anti-echinococcosis drugs.

The association of FGF-21 with the risk of newly diagnosed type-2 diabetes mellitus: a cross-sectional study in Southern China

BACKGROUND

This study investigated the relationship between fibroblast growth factor 21 (FGF-21) and newly diagnosed type-2 diabetes mellitus (T2DM).

METHODS

In this cross-sectional study, FGF-21 and T2DM risk were analyzed using restricted cubic splines with univariate or multivariate logistic regression analysis. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated via logistic regression analysis. Cluster and subgroup analyses were conducted to evaluate the associations between FGF-21 and diabetes in different subpopulations. Nomograms and ROC curves were used to explore the clinical utility of FGF-21 in the diabetes assessment model.

RESULTS

High levels of FGF-21 were significantly associated with a high risk of T2DM after adjusting for confounding factors in both the total population and subpopulations (P for trend < 0.001). In the total population, the ORs of diabetes with increasing FGF-21 quartiles were 1.00 (reference), 1.24 (95% CI 0.56-2.80; quartile 2), 2.47 (95% CI 1.18-5.33; quartile 3), and 3.24 (95% CI 1.53-7.14; quartile 4) in Model 4 (P < 0.001), and the trend was consistent in different subpopulations. In addition, compared with the model constructed with conventional noninvasive indicators, the AUC of the model constructed by adding FGF-21 was increased from 0.668 (95% CI: 0.602-0.733) to 0.715 (95% CI: 0.654-0.777), indicating that FGF-21 could significantly improve the risk-assessment efficiency of type-2 diabetes.

CONCLUSION

This study demonstrated that a high level of circulating FGF-21 was positively correlated with diabetes, and levels of FGF-21 could be an important biomarker for the assessment of diabetes risk.

The efficacy of hydrogel containing zinc oxide-loaded and minocycline serum albumin nanopartical in the treatment of peri-implantitis

BACKGROUND

We conducted this animal study to assess the efficacy of the novel hydrogel containing zinc oxide-loaded and minocycline serum albumin nanoparticals (Mino-ZnO@Alb NPs) on peri-implantitis in an experimental mouse model.

MATERIAL AND METHODS

Mino-ZnO@Alb NPs was prepared as previously reported. The peri-implantitis model was successfully established in rats, and the rats were divided into three groups randomly: Mino-ZnO@Alb NPs (Mino-ZnO) group, minocycline group, and untreated group. Four weeks later, clinical and radiographic assessments were performed to evaluate soft tissue inflammation and bone resorption level. Histologic analysis was performed to estimate the amount of remaining supporting bone tissue (SBT) around implants. ELISA tests were used to determine the concentration of inflammation factor interleukin-1-beta (IL-1β) and anti-inflammation factor tumor necrosis factor-alpha (TNF-α) around implants.

RESULTS

After one month, the Mino-ZnO group showed better results than the other two groups in regards to the results of bleeding on probing, probing pocket depth, bleeding index and gingival index. X-ray showed that SBT at mesial and distal sites around implants in the other two groups was significantly lower compared with that of Mino-ZnO group. The quantity of osteoclasts in peri-implant tissues of the Mino-ZnO group was less than that in the minocycline and untreated groups. IL-1β in the Mino-ZnO group was lower than that in the other two groups. TNF-α level was the opposite.

CONCLUSIONS

Mino-ZnO@Alb NPs can effectively treat peri-implantitis and promote soft tissue healing, and may act as a promising product.

Venous and arterial thromboembolism after colorectal cancer in the Netherlands: Incidence, predictors, and prognosis

BACKGROUND

Colorectal cancer (CRC) is the third most prevalent cancer type. CRC-patients are at increased risk of venous and arterial thromboembolism (TE), but the magnitude of the risks, their predictors and consequences are not exactly known.

OBJECTIVES

We aimed to determine incidence, predictors and prognosis of TE after incident CRC in a large, unselected population.

METHODS

Using data from Statistics Netherlands and the Netherlands Comprehensive Cancer Organization, all incident CRC-patients were identified between 2013 and 2018 plus a sample of 1:2 age- and sex-matched control subjects. Incidence rates and cumulative incidences for TE were estimated. Predictor variables for TE were explored by univariable Cox regression. The association between TE and all-cause mortality was evaluated by multivariable time-dependent Cox regression.

RESULTS

68,238 incident CRC-patients were matched to 136,476 controls. CRC-patients had a 1-year cumulative venous TE (VTE) incidence of 1.93 % (95%CI 1.83-2.04), versus 0.24 % (95%CI 0.21-0.27) in controls (HR 8.85; 95%CI 7.83-9.99). For arterial TE (ATE), this was 2.74 % (95%CI 2.62-2.87) in CRC versus 1.88 % (95%CI 1.81-1.95) in controls (HR 1.57; 95%CI 1.47-1.66). Cancer stage, surgery, chemotherapy and asthma were predictors for VTE, whereas age, prior ATE and Parkinson's disease were predictors for ATE. CRC patients with TE had an increased risk of all-cause mortality (VTE HR; 3.68 (95%CI 3.30-4.10, ATE HR; 3.05 (95%CI 2.75-3.39)) compared with CRC-patients without TE.

CONCLUSIONS

This Dutch nationwide cohort study adds detailed knowledge on the risk of VTE and ATE, their predictors and prognosis in CRC-patients. These findings may drive TE prophylactic management decisions.

[Establishment and validation of a risk prediction model for disseminated intravascular coagulation patients with electrical burns]

Objective: To establish and validate a risk prediction model of disseminated intravascular coagulation (DIC) by the screening independent risk factors for the occurrence of DIC in patients with electrical burns. Methods: The retrospective case series study was conducted. The clinical data of 218 electrical burn patients admitted to Baogang Hospital of Inner Mongolia from January 2015 to January 2023 who met the inclusion criteria were collected, including 198 males and 20 females, with the age of (38±14) years. The patients were divided into DIC group and non DIC group based on whether they were diagnosed with DIC during the treatment period. The following data of patients of two groups were collected and compared, including age, gender, total burn area, full-thickness burn area, injury voltage, whether osteofascial compartment syndrome occurred within 1 day after injury, duration of stay in burn intensive care unit, total length of hospital stay, whether combined with inhalation injury and multiple injuries, whether shock occurred upon admission, the abbreviated burn severity index score, and the acute physiology and chronic health evaluation Ⅱ score. The laboratory examination data of the patients within 24 hours after admission were also collected, including blood routine indexes: white blood cell count (WBC), hemoglobin level, platelet count (PLT), and neutrophil count; coagulation indexes: activated partial thromboplastin time (APTT), prothrombin time, thrombin time, and levels of D-dimer and fibrinogen (FIB); blood biochemistry indexes: aspartic transaminase, alanine transaminase, direct bilirubin, total bilirubin, total protein, albumin, blood glucose, creatinine, and urea nitrogen; blood gas analysis indexes: blood pH value, arterial partial pressure of oxygen, arterial partial pressure of carbon dioxide, bicarbonate, and base excess; and cardiac zymogram indexes: levels of myoglobin, troponin, lactate dehydrogenase, creatine kinase (CK), and α-hydroxybutyrate dehydrogenase. Data were statistically analyzed with chi-square test, Fisher's exact probability test, independent sample t test, and Mann-Whitney U test. For the variables with statistically significant differences in single factor analysis, the least absolute value selection and shrinkage operator (LASSO) regression was used to reduce the dimension, and the predictive factors for DIC in 218 patients with electrical burns were screened. The above-mentioned predictors were included in multivariate logistic regression analysis to find out the independent risk factors for DIC in 218 patients with electrical burns, and to draw the prediction model nomograms. The performance of the prediction model was evaluated by the receiver operating characteristic (ROC) curve and the area under the ROC curve, and the prediction model was validated by the calibration curve and clinical decision curve analysis (DCA). Results: Compared with those in non DIC group, the total burn area, full-thickness burn area, total length of hospital stay, and the proportions of high voltage caused injury, occurrence of osteofascial compartment syndrome within 1 day after injury, combination of inhalation injury, and occurrence of shock upon admission of patients in DIC group were significantly increased/prolonged (with Z values of -2.53, -4.65, and -2.10, respectively, with χ2 values of 11.46, 16.00, 7.98, and 18.93, respectively, P<0.05). Compared with those in non DIC group, the APTT, level of D-dimer, myoglobin, WBC, PLT, and levels of FIB, total bilirubin, and CK of patients within 24 hours after admission in DIC group were significantly prolonged/increased (with Z values of -2.02, -4.51, and -3.82, respectively, with t values of -3.84, -2.34, -2.77, -2.70, and -2.61, respectively), and the level of total protein and blood pH value were significantly reduced (t=-2.85, Z=-2.03), P<0.05. LASSO regression analysis was carried out for the above 17 indicators with statistically significant differences. The results showed that injury voltage, the occurrence of shock upon admission, the occurrence of osteofascial compartment syndrome within 1 day after injury, and levels of D-dimer and total protein within 24 hours after admission were predictive factors for the occurrence of DIC in 218 patients with electrical burns (with regression coefficients of 0.24, 0.52, 0.35, 0.13, and -0.001, respectively). Multivariate logistic regression analysis showed that injury voltage, the occurrence of shock upon admission, the occurrence of osteofascial compartment syndrome within 1 day after injury, and D-dimer level within 24 hours after admission were independent risk factors for DIC in 218 patients with electrical burns (with odds ratios of 3.33, 4.24, 2.68, and 1.38, respectively, with 95% confidence intervals of 1.43-7.79, 1.78-10.07, 1.17-6.13, and 1.19-1.61, respectively, P<0.05). Based on the aforementioned four independent risk factors, the nomogram of prediction model for evaluating the probability of DIC in patients was drawn. The area under the ROC curve of prediction model was 0.88, and the 95% confidence interval was 0.82-0.95, indicating that the model had good predictive ability; the curve of prediction model tended to be near the ideal curve, indicating that the model had a high calibration degree; the clinical DCA of prediction model showed that the threshold probability of patients ranged from 4% to 97%, indicating that the model had good predictive ability. Conclusions: The injury voltage, the occurrence of shock upon admission, the occurrence of osteofascial compartment syndrome within 1 day after injury, and D-dimer level within 24 hours after admission are independent risk factors for the occurrence of DIC in patients with electrical burns. The prediction model established based on the above indicators can provide early warning for the occurrence of DIC in these patients.

[A case of intellectual developmental disorder with behavioral abnormalities and craniofacial dysmorphism with or without seizures caused by PHF21A gene variation and review of literature]

Objective: To discuss the clinical and genetic features of intellectual developmental disorder with behavioral abnormalities and craniofacial dysmorphism with or without seizures (IDDBCS). Methods: The clinical and genetic records of a patient who was diagnosed with IDDBCS caused by PHF21A gene variation at Children's Hospital Capital Institute of Pediatrics in 2021 were collected retrospectively. Using " PHF21A gene" as the keyword, relevant articles were searched at CNKI, Wanfang Data and PubMed from establishment of databases to February 2023. Clinical and genetic features of IDDBCS were summarized in the combination of this case. Results: An 8 months of age boy showed overgrowth (height, weight and head circumference were all higher than the 97^th percentile of children of the same age and sex) and language and motor developmental delay after birth, and gradually showed autism-like symptoms like stereotyped behavior and poor eye contact. At 8 months of age, he began to show epileptic seizures, which were in the form of a series of spastic seizures with no reaction to adrenocorticotropic hormone but a good response to vigabatrin. Physical examination showed special craniofacial appearances including a prominent high forehead, sparse eyebrows, broad nasal bridge, and downturned mouth with a tent-shaped upper lip. The patient also manifested hypotonia. Whole exome sequencing showed a de novo heterogeneous variant, PHF21A (NM_001101802.1): c.54+1G>A, and IDDBCS was diagnosed. A total of 6 articles (all English articles) were collected, involving this case and other 14 patients of IDDBCS caused by PHF21A gene variation. Clinical manifestations were intellectual disability or developmental delay (15 patients), craniofacial anomalies (15 patients), behavioral abnormalities (12 patients), seizures (9 patients), and overgrowth (8 patients). The main pathogenic variations were frameshift variations (8 patients). Conclusions: IDDBCS should be considered when patients show nervous developmental abnormalities, craniofacial anomalies, seizures and overgrowth. PHF21A gene variation detection helps to make a definite diagnosis.