Corrigendum: Tpr deficiency disrupts erythroid maturation with impaired chromatin condensation in zebrafish embryogenesis.

[This corrects the article DOI: 10.3389/fcell.2021.709923.].

Ciliogenesis-associated Kinase 1 Promotes Breast Cancer Cell Proliferation and Chemoresistance via Phosphorylating ERK1.

Ciliogenesis-associated kinase 1 (CILK1) plays a key role in the ciliogenesis and ciliopathies. It remains totally unclear whether CILK1 is involved in tumor progression and therapy resistance. Here, we report that the aberrant high-expression of CILK1 in breast cancer is required for tumor cell proliferation and chemoresistance. Two compounds, CILK1-C30 and CILK1-C28, were identified with selective inhibitory effects towards the Tyr-159/Thr-157 dual-phosphorylation of CILK1, pharmacological inhibition of CILK1 significantly suppressed tumor cell proliferation and overcame chemoresistance in multiple experimental models. Large-scale screen of CILK1 substrates confirmed that the kinase directly phosphorylates ERK1, which is responsible for CILK1-mediated oncogenic function. CILK1 is also indicated to be responsible for the chemoresistance of small-cell lung cancer cells. Our data highlight the importance of CILK1 in cancer, implicating that targeting CILK1/ERK1 might offer therapeutic benefit to cancer patients.

Enhanced Lightweight YOLOX for Small Object Wildfire Detection in UAV Imagery.

Target detection technology based on unmanned aerial vehicle (UAV)-derived aerial imagery has been widely applied in the field of forest fire patrol and rescue. However, due to the specificity of UAV platforms, there are still significant issues to be resolved such as severe omission, low detection accuracy, and poor early warning effectiveness. In light of these issues, this paper proposes an improved YOLOX network for the rapid detection of forest fires in images captured by UAVs. Firstly, to enhance the network's feature-extraction capability in complex fire environments, a multi-level-feature-extraction structure, CSP-ML, is designed to improve the algorithm's detection accuracy for small-target fire areas. Additionally, a CBAM attention mechanism is embedded in the neck network to reduce interference caused by background noise and irrelevant information. Secondly, an adaptive-feature-extraction module is introduced in the YOLOX network's feature fusion part to prevent the loss of important feature information during the fusion process, thus enhancing the network's feature-learning capability. Lastly, the CIoU loss function is used to replace the original loss function, to address issues such as excessive optimization of negative samples and poor gradient-descent direction, thereby strengthening the network's effective recognition of positive samples. Experimental results show that the improved YOLOX network has better detection performance, with mAP@50 and mAP@50_95 increasing by 6.4% and 2.17%, respectively, compared to the traditional YOLOX network. In multi-target flame and small-target flame scenarios, the improved YOLO model achieved a mAP of 96.3%, outperforming deep learning algorithms such as FasterRCNN, SSD, and YOLOv5 by 33.5%, 7.7%, and 7%, respectively. It has a lower omission rate and higher detection accuracy, and it is capable of handling small-target detection tasks in complex fire environments. This can provide support for UAV patrol and rescue applications from a high-altitude perspective.

Mutant kri1l causes abnormal retinal development via cell cycle arrest and apoptosis induction.

Damage to the ribosome or an imbalance in protein biosynthesis can lead to some human diseases, such as diabetic retinopathy (DR) and other eye diseases. Here, we reported that the kri1l gene was responsible for retinal development. The kri1l gene encodes an essential component of the rRNA small subunit processome. The retinal structure was disrupted in kri1l mutants, which resulted in small eyes. The boundaries of each layer of cells in the retina were blurred, and each layer of cells was narrowed and decreased. The photoreceptor cells and Müller glia cells almost disappeared in kri1l mutants. The lack of photoreceptor cells caused a fear of light response. The development of the retina started without abnormalities, and the abnormalities began two days after fertilization. In the kri1l mutant, retinal cell differentiation was defective, resulting in the disappearance of cone cells and Müller cells. The proliferation of retinal cells was increased, while apoptosis was also enhanced in kri1l mutants. γ-H2AX upregulation indicated the accumulation of DNA damage, which resulted in cell cycle arrest and apoptosis. The kri1l mutation reduced the expression of some opsin genes and key retinal genes, which are also essential for retinal development.

SLKIR: A framework for extracting key information from air traffic control instructions Using small sample learning.

In air traffic control (ATC), Key Information Recognition (KIR) of ATC instructions plays a pivotal role in automation. The field's specialized nature has led to a scarcity of related research and a gap with the industry's cutting-edge developments. Addressing this, an innovative end-to-end deep learning framework, Small Sample Learning for Key Information Recognition (SLKIR), is introduced for enhancing KIR in ATC instructions. SLKIR incorporates a novel Multi-Head Local Lexical Association Attention (MHLA) mechanism, specifically designed to enhance accuracy in identifying boundary words of key information by capturing their latent representations. Furthermore, the framework includes a task focused on prompt, aiming to bolster the semantic comprehension of ATC instructions within the core network. To overcome the challenges posed by category imbalance in boundary word and prompt discrimination tasks, tailored loss function optimization strategies are implemented, effectively expediting the learning process and boosting recognition accuracy. The framework's efficacy and adaptability are demonstrated through experiments on two distinct ATC instruction datasets. Notably, SLKIR outperforms the leading baseline model, W2NER, achieving a 3.65% increase in F1 score on the commercial flight dataset and a 12.8% increase on the training flight dataset. This study is the first of its kind to apply small-sample learning in KIR for ATC and the source code of SLKIR will be available at: https://github.com/PANPANKK/ATC_KIR .

Corrigendum: Mutation of Gemin5 causes defective hematopoietic stem/progenitor cells proliferation in zebrafish embryonic hematopoiesis.

[This corrects the article DOI: 10.3389/fcell.2021.670654.].

Assessment and analysis of accents in air traffic control speech: a fusion of deep learning and information theory.

Introduction: Enhancing the generalization and reliability of speech recognition models in the field of air traffic control (ATC) is a challenging task. This is due to the limited storage, difficulty in acquisition, and high labeling costs of ATC speech data, which may result in data sample bias and class imbalance, leading to uncertainty and inaccuracy in speech recognition results. This study investigates a method for assessing the quality of ATC speech based on accents. Different combinations of data quality categories are selected according to the requirements of different model application scenarios to address the aforementioned issues effectively. Methods: The impact of accents on the performance of speech recognition models is analyzed, and a fusion feature phoneme recognition model based on prior text information is constructed to identify phonemes of speech uttered by speakers. This model includes an audio encoding module, a prior text encoding module, a feature fusion module, and fully connected layers. The model takes speech and its corresponding prior text as input and outputs a predicted phoneme sequence of the speech. The model recognizes accented speech as phonemes that do not match the transcribed phoneme sequence of the actual speech text and quantitatively evaluates the accents in ATC communication by calculating the differences between the recognized phoneme sequence and the transcribed phoneme sequence of the actual speech text. Additionally, different levels of accents are input into different types of speech recognition models to analyze and compare the recognition accuracy of the models. Result: Experimental results show that, under the same experimental conditions, the highest impact of different levels of accents on speech recognition accuracy in ATC communication is 26.37%. Discussion: This further demonstrates that accents affect the accuracy of speech recognition models in ATC communication and can be considered as one of the metrics for evaluating the quality of ATC speech.

Collaborative optimization model and algorithm for airport capacity and traffic flow allocation.

How to efficiently utilize the existing airport capacity without physical expansion and considerable economic inputs to meet air traffic needs is one of the important tasks of air traffic management. To improve the efficiency of capacity utilization, it is necessary to find the actual airport capacity properly. In this work, taking Shuangliu International Airport as an example, a methodology for capacity estimation is proposed that combines the empirical method with an analytical approach that uses historical performance data from the airport to construct a capacity envelope to approximate the airport's actual capacity to the greatest extent, establishes a collaborative optimization model that reflects the inherent relations between airport capacity and arrival and departure traffic demand, adopts an improved optimization algorithm to solve the model, and generates an optimal flight allocation scheme. Priority ratio is introduced to dynamically adjust management preferences for arrival and departure traffic demand to further reveal the synergy mechanism between departure and arrival traffic flow demand and the airport capacity. The result shows that the Flight On-time Performance rate is lifted by 6% in the case study which proves the feasibility of the proposed method, demonstrating its value for maximizing airport capacity and traffic flow demand without requiring expansions on airport scales.

Research on automatic pilot repetition generation method based on deep reinforcement learning.

Using computers to replace pilot seats in air traffic control (ATC) simulators is an effective way to improve controller training efficiency and reduce training costs. To achieve this, we propose a deep reinforcement learning model, RoBERTa-RL (RoBERTa with Reinforcement Learning), for generating pilot repetitions. RoBERTa-RL is based on the pre-trained language model RoBERTa and is optimized through transfer learning and reinforcement learning. Transfer learning is used to address the issue of scarce data in the ATC domain, while reinforcement learning algorithms are employed to optimize the RoBERTa model and overcome the limitations in model generalization caused by transfer learning. We selected a real-world area control dataset as the target task training and testing dataset, and a tower control dataset generated based on civil aviation radio land-air communication rules as the test dataset for evaluating model generalization. In terms of the ROUGE evaluation metrics, RoBERTa-RL achieved significant results on the area control dataset with ROUGE-1, ROUGE-2, and ROUGE-L scores of 0.9962, 0.992, and 0.996, respectively. On the tower control dataset, the scores were 0.982, 0.954, and 0.982, respectively. To overcome the limitations of ROUGE in this field, we conducted a detailed evaluation of the proposed model architecture using keyword-based evaluation criteria for the generated repetition instructions. This evaluation criterion calculates various keyword-based metrics based on the segmented results of the repetition instruction text. In the keyword-based evaluation criteria, the constructed model achieved an overall accuracy of 98.8% on the area control dataset and 81.8% on the tower control dataset. In terms of generalization, RoBERTa-RL improved accuracy by 56% compared to the model before improvement and achieved a 47.5% improvement compared to various comparative models. These results indicate that employing reinforcement learning strategies to enhance deep learning algorithms can effectively mitigate the issue of poor generalization in text generation tasks, and this approach holds promise for future application in other related domains.

Proteolytic activation of angiomotin by DDI2 promotes angiogenesis.

The scaffolding protein angiomotin (AMOT) is indispensable for vertebrate embryonic angiogenesis. Here, we report that AMOT undergoes cleavage in the presence of lysophosphatidic acid (LPA), a lipid growth factor also involved in angiogenesis. AMOT cleavage is mediated by aspartic protease DNA damage-inducible 1 homolog 2 (DDI2), and the process is tightly regulated by a signaling axis including neurofibromin 2 (NF2), tankyrase 1/2 (TNKS1/2), and RING finger protein 146 (RNF146), which induce AMOT membrane localization, poly ADP ribosylation, and ubiquitination, respectively. In both zebrafish and mice, the genetic inactivation of AMOT cleavage regulators leads to defective angiogenesis, and the phenotype is rescued by the overexpression of AMOT-CT, a C-terminal AMOT cleavage product. In either physiological or pathological angiogenesis, AMOT-CT is required for vascular expansion, whereas uncleavable AMOT represses this process. Thus, our work uncovers a signaling pathway that regulates angiogenesis by modulating a cleavage-dependent activation of AMOT.

Cthrc1 deficiency aggravates wound healing and promotes cardiac rupture after myocardial infarction via non-canonical WNT5A signaling pathway.

Cardiac fibroblasts are crucial for scar formation and cardiac repair after myocardial infarction (MI). Collagen triple helix repeat containing 1 (CTHRC1), an extracellular matrix protein, is involved in the pathogenesis of vascular remodeling, bone formation, and tumor progression. However, the role and underlying mechanism of CTHRC1 in post-MI wound repair are not fully clear. Bioinformatics analysis demonstrated CTHRC1 up-regulation in cardiac fibroblasts after ischemic cardiac injury. Serum levels of CTHRC1 were increased in MI mice and CTHRC1 expression was up-regulated in cardiac fibroblasts after MI. In vitro results showed that the induction of CTHRC1 expression in cardiac fibroblasts was mediated by canonical TGFß1-Smad2/3 signaling axis. Moreover, CTHRC1 improved wound healing and boosted cardiac fibroblast activation in vitro. Cthrc1 deficiency aggravated cardiac function and reduced collagen deposition as well as increased mortality attributable to cardiac rupture after MI. Consistent with above phenotypes, reduced the levels of myocardial CD31, α-smooth muscle actin, collagen I, and collagen III was observed, whereas myocardial expression of matrix metalloproteinase 2 and matrix metalloproteinase 9 were increased in Cthrc1 knockout mice post-MI. Above effects could be partly reversed by rCTHRC1 protein or rWNT5A protein. Our study indicates that cardiac fibroblast-derived, canonical TGFß1-Smad2/3-dependent CTHRC1 could improve wound repair and prevent cardiac rupture after MI via selectively activating non-canonical WNT5A-PCP signaling pathway.

Targeting WWP1 ameliorates cardiac ischemic injury by suppressing KLF15-ubiquitination mediated myocardial inflammation.

Rationale: Previous studies have suggested that myocardial inflammation plays a critical role after ischemic myocardial infarction (MI); however, the underlying mechanisms still need to be fully elucidated. WW domain-containing ubiquitin E3 ligase 1 (WWP1) is considered as an important therapeutic target for cardiovascular diseases due to its crucial function in non-ischemic cardiomyopathy, though it remains unknown whether targeting WWP1 can alleviate myocardial inflammation and ischemic injury post-MI. Methods: Recombinant adeno-associated virus 9 (rAAV9)-cTnT-mediated WWP1 or Kruppel-like factor 15 (KLF15) gene transfer and a natural WWP1 inhibitor Indole-3-carbinol (I3C) were used to determine the WWP1 function in cardiomyocytes. Cardiac function, tissue injury, myocardial inflammation, and signaling changes in the left ventricular tissues were analyzed after MI. The mechanisms underlying WWP1 regulation of cardiomyocyte phenotypes in vitro were determined using the adenovirus system. Results: We found that WWP1 expression was up-regulated in cardiomyocytes located in the infarct border at the early phase of MI and in hypoxia-treated neonatal rat cardiac myocytes (NRCMs). Cardiomyocyte-specific WWP1 overexpression augmented cardiomyocyte apoptosis, increased infarct size and deteriorated cardiac function. In contrast, inhibition of WWP1 in cardiomyocytes mitigated MI-induced cardiac ischemic injury. Mechanistically, WWP1 triggered excessive cardiomyocyte inflammation after MI by targeting KLF15 to catalyze K48-linked polyubiquitination and degradation. Ultimately, WWP1-mediated degradation of KLF15 contributed to the up-regulation of p65 acetylation, and activated the inflammatory signaling of MAPK in ischemic myocardium and hypoxia-treated cardiomyocytes. Thus, targeting of WWP1 by I3C protected against cardiac dysfunction and remodeling after MI. Conclusions: Our study provides new insights into the previously unrecognized role of WWP1 in cardiomyocyte inflammation and progression of ischemic injury induced by MI. Our findings afford new therapeutic options for patients with ischemic cardiomyopathy.

Association between antibiotic exposure and the risk of infertility in women of childbearing age: A case-control study.

BACKGROUND: Based on self-report questionnaires, two previous epidemiological studies investigated the association between the exposure of women to antibiotics and their fertility. However, biomonitoring studies on low-dose antibiotic exposure, mainly from food and water, and its relation to the risk of infertility are missing. METHODS: Based on a case-control study design, 302 women with infertility (144 primary infertility, 158 secondary infertility) and 302 women with normal fertility, all aged 20-49 years, were recruited from Anhui Province, China, in 2020 and 2021. A total of 41 common antibiotics and two antibiotic metabolites in urine samples were determined by liquid chromatography-triple quadrupole tandem mass spectrometry (LC-QqQ-MS/MS). RESULTS: Twenty-eight antibiotics with detection rates from 10% to 100% in both cases (median concentration: ∼2.294 ng/mL) and controls (∼1.596 ng/mL) were included in the analysis. Logistic regression analysis revealed that after controlling for confounding factors, high concentrations of eight individual antibiotics (sulfamethoxazole, sulfaclozine, sulfamonomethoxine, penicillin G, chlorotetracycline, ofloxacin, norfloxacin, and cyadox) and four antibiotic classes (sulfonamides, tetracyclines, quinoxalines, and veterinary antibiotics) were related to a high risk of female infertility, with odds ratios (ORs) ranging from 1.30 to 2.86, except for chlorotetracycline (OR = 6.34), while another nine individual antibiotics (sulfamethazine, azithromycin, cefaclor, amoxicillin, oxytetracycline, pefloxacin, sarafloxacin, enrofloxacin, and florfenicol) and classes of chloramphenicol analogs and human antibiotics were related to a reduced risk of infertility, with ORs ranging from 0.70 to 0.20. Based on restricted cubic spline models after controlling for confounding factors, we observed that the relationship between all of the above protective antibiotics and infertility was nonlinear: A certain concentration could reduce the risk of female infertility while exceeding a safe dose could increase the risk of infertility. CONCLUSION: These results provide preliminary evidence that the effects of antibiotics on female fertility vary based on the active ingredient and usage and imply the importance of exposure dose. Future studies are needed to verify these results by controlling for multiple confounding factors.

Multi-scale analysis of the impacts of meteorology and emissions on PM2.5 and O3 trends at various regions in China from 2013 to 2020 3. Mechanism assessment of O3 trends by a model.

A multiscale analysis of meteorological trends was carried out to investigate the impacts of the large-scale circulation types as well as the local-scale key weather elements on the complex air pollutants, i.e., PM2.5 and O3 in China. Following accompanying papers on synoptic circulation impact and key weather elements and emission contributions (Gong et al., 2022a; Gong et al., 2022b), an emission-driven Observation-based Box Model (e-OBM) was developed to study the impact mechanisms on O3 trend and quantitatively assess the effects of variation in the emissions control over 2013-2020 for Beijing, Chengdu, Guangzhou and Shanghai. Compared with the original OBM, the e-OBM not only improves the performance to simulate the hourly O3 peak concentration in daytime, but also reasonably reproduces the maximum daily 8-hour average (MDA8) O3 concentrations in the four cities. Based upon the sensitivity experiments, it is found that the meteorology is the dominant driver for the MDA8 O3 trend, contributing from about 32 % to 139 % to the variations. From the mechanistic point of view, the variations of meteorology lead to the enhancement of atmospheric oxidation capacity and the acceleration of O3 production. Further evaluation to the emission changes in four cities shows that the O3-precursors relationships of the four cities have been changed from the VOC-limited regime in 2013 to the transition regime or near-transition regime in 2020. Though the NOx/VOCs ratios have been obviously decreased, the emission reductions up to 2020 were still not enough to mitigate O3 pollution in these cities. It is emphasized in this study that the strengthened control measures with maintaining a certain ratio of NOx and VOCs should be implemented to further curb the increasing trend of O3 in urban areas.

Overexpression of Pleckstrin Homology Domain-Containing Family A Member 4 Is Correlated with Poor Prognostic Outcomes and Immune Infiltration in Lower-Grade Glioma.

Introduction: The global incidence of brain tumors, the most common of which is lower grade glioma (LGG), remains high. Pleckstrin homology domain-containing family A member 4 (PLEKHA4) has been reported to be related to tumor invasion and growth. However, its role and correlation with immunity in LGG remain elusive. Methods: We evaluated the expression pattern, prognostic value, biological functions, and immune effects of PLEKHA4 in LGG. We also analyzed the association between PLEKHA4 levels in different tumors, patient prognosis, and its role in tumor immunity. Depending on the type of research data, we used statistical methods such as Student's t-tests, Mann-Whitney U tests one-way ANOVA tests Kruskal-Wallis tests Pearson's or Spearman's correlation analysis Chi-square and Fisher's exact tests in this paper. Results and Conclusions. The results revealed that PLEKHA4 levels were markedly elevated in most tumors (such as LGG). High PLEKHA4 levels are associated with poor overall survival (OS), progression-free interval (PFI) rates, and disease-specific survival (DSS) in LGG patients. Cox regression analysis and nomograms showed that PLEKHA4 levels are independent prognostic factors for LGG patients. According to functional enrichment analysis, PLEKHA4 levels in LGG are associated with immune infiltration and immunotherapy. In conclusion, PLEKHA4 is a potential prognostic marker and immunotherapy target for LGG.

Activation of Wnt/ß-catenin signaling by Zeb1 in endothelial progenitors induces vascular quiescence entry.

The establishment of a functional vasculature requires endothelial cells to enter quiescence during the completion of development, otherwise pathological overgrowth occurs. How such a transition is regulated remains unclear. Here, we uncover a role of Zeb1 in defining vascular quiescence entry. During quiescence acquisition, Zeb1 increases along with the progressive decline of endothelial progenitors' activities, with Zeb1 loss resulting in endothelial overgrowth and vascular deformities. RNA sequencing (RNA-seq) and assay for transposase-accessible chromatin sequencing (ATAC-seq) analyses reveal that Zeb1 represses Wif1, thereby activating Wnt/ß-catenin signaling. Knockdown of Wif1 rescues the overgrowth induced by Zeb1 deletion. Importantly, local administration of surrogate Wnt molecules in the retina ameliorates the overgrowth defects of Zeb1 mutants. These findings show a mechanism by which Zeb1 induces quiescence of endothelial progenitors during the establishing of vascular homeostasis, providing molecular insight into the inherited neovascular pathologies associated with human ZEB1 mutations, suggesting pharmacological activation of Wnt/ß-catenin signaling as a potential therapeutical approach.

Heterogeneities of zebrafish vasculature development studied by a high throughput light-sheet flow imaging system.

Zebrafish is one of the ideal model animals to study the structural and functional heterogeneities in development. However, the lack of high throughput 3D imaging techniques has limited studies to only a few samples, despite zebrafish spawning tens of embryos at once. Here, we report a light-sheet flow imaging system (LS-FIS) based on light-sheet illumination and a continuous flow imager. LS-FIS enables whole-larva 3D imaging of tens of samples within half an hour. The high throughput 3D imaging capability of LS-FIS was demonstrated with the developmental study of the zebrafish vasculature from 3 to 9 days post-fertilization. Statistical analysis shows significant variances in trunk vessel development but less in hyaloid vessel development.

Phosphorylation of CENP-R by Aurora B regulates kinetochore-microtubule attachment for accurate chromosome segregation.

Error-free mitosis depends on accurate chromosome attachment to spindle microtubules via a fine structure called the centromere that is epigenetically specified by the enrichment of CENP-A nucleosomes. Centromere maintenance during mitosis requires CENP-A-mediated deposition of constitutive centromere-associated network that establishes the inner kinetochore and connects centromeric chromatin to spindle microtubules during mitosis. Although previously proposed to be an adaptor of retinoic acid receptor, here, we show that CENP-R synergizes with CENP-OPQU to regulate kinetochore-microtubule attachment stability and ensure accurate chromosome segregation in mitosis. We found that a phospho-mimicking mutation of CENP-R weakened its localization to the kinetochore, suggesting that phosphorylation may regulate its localization. Perturbation of CENP-R phosphorylation is shown to prevent proper kinetochore-microtubule attachment at metaphase. Mechanistically, CENP-R phosphorylation disrupts its binding with CENP-U. Thus, we speculate that Aurora B-mediated CENP-R phosphorylation promotes the correction of improper kinetochore-microtubule attachment in mitosis. As CENP-R is absent from yeast, we reasoned that metazoan evolved an elaborate chromosome stability control machinery to ensure faithful chromosome segregation in mitosis.

The molecular subtypes of triple negative breast cancer were defined and a ligand-receptor pair score model was constructed by comprehensive analysis of ligand-receptor pairs.

Background: Intercellular communication mediated by ligand-receptor interactions in tumor microenvironment (TME) has a profound impact on tumor progression. This study aimed to explore the molecular subtypes mediated by ligand-receptor (LR) pairs in triple negative breast cancer (TNBC), identify the most important LR pairs to construct a prognostic risk model, and study their effect on TNBC immunotherapy. Methods: LR pairs subclasses of TNBC were categorized by consensus clustering based on LR Pairs in METABRIC dataset. Least absolute shrinkage and selection operator (LASSO) Cox regression and stepwise Akaike information criterion (stepAIC) were conducted to build a LR pairs score model. The relationship between LR pairs score and immune cell infiltration, stromal score and immune score associated with TME was analyzed, and the prediction of drug therapy and immunotherapy efficacy by LR pairs score was evaluated. Results: According to the expression pattern of 145 TNBC prognostic LR pairs, the samples were divided into three subclasses with different survival outcomes, copy number variation (CNV), TME immune cell infiltration, stromal score and immune score. The LR pairs score model constructed in the METABRIC dataset was composed of four LR pairs, and its predictive significance for TNBC prognosis was verified in GSE58812 and GSE21653 cohorts. In addition, LR pairs score was negatively correlated with several immune pathways regulating immunity and immune score, and related to the sensitivity of anti-neoplastic drugs and the effect of anti-PD-L1 therapy. Conclusion: Our study confirmed the impact of LR pairs on the molecular heterogeneity of TNBC, characterized three LR pairs subtypes with different survival outcomes and TME patterns, and proposed a LR pairs score system with predictive significance for TNBC prognosis and anti-PD-L1 therapeutic effect, which provides a potential evaluation scheme for TNBC management.

A Model Incorporating Axillary Tail Position on Mammography for Preoperative Prediction of Non-sentinel Lymph Node Metastasis in Patients with Initial cN+ Breast Cancer after Neoadjuvant Chemotherapy.

RATIONALE AND OBJECTIVES: This study aimed to develop a model incorporating axillary tail position on mammography (AT) for the prediction of non-sentinel Lymph Node (NSLN) metastasis in patients with initial clinical node positivity (cN+). METHODS AND MATERIALS: The study reviewed a total of 257 patients with cN+ breast cancer who underwent both sentinel lymph node biopsy (SLNB) and axillary lymph node dissection (ALND) following neoadjuvant chemotherapy (NAC). A logistic regression model was developed based on these factors and the results of post-NAC AT and axillary ultrasound (AUS). RESULTS: Four clinical factors with p<0.1 in the univariate analysis, including ycT0(odds ratio [OR]: 4.84, 95% confidence interval [CI]: 2.13-11.91, p<0.001), clinical stage before NAC (OR: 2.68, 95%CI: 1.15-6.58, p=0.025), estrogen receptor (ER) expression (OR: 3.29, 95%CI: 1.39-8.39, p=0.009), and HER2 status (OR: 0.21, 95%CI: 0.08-0.50, p=0.001), were independent predictors of NSLN metastases. The clinical model based on the above four factors resulted in the area under the curve (AUC) of 0.82(95%CI: 0.76-0.88) in the training set and 0.83(95% CI: 0.74-0.92) in the validation set. The results of post-NAC AUS and AT were added to the clinical model to construct a clinical imaging model for the prediction of NSLN metastasis with AUC of 0.87(95%CI: 0.81-0.93) in the training set and 0.89(95%CI: 0.82-0.96) in the validation set. CONCLUSIONS: The study incorporated the results of post-NAC AT and AUS with other clinal factors to develop a model to predict NSLN metastasis in patients with initial cN+ before surgery. This model performed excellently, allowing physicians to select patients for whom unnecessary ALND could be avoided after NAC.