Application of HepaRG cells for genotoxicity assessment: a review.

There has been growing interest in the use of human-derived metabolically competent cells for genotoxicity testing. The HepaRG cell line is considered one of the most promising cell models because it is TP53-proficient and retains many characteristics of primary human hepatocytes. In recent years, HepaRG cells, cultured in both a traditional two-dimensional (2D) format and as more advanced in-vivo-like 3D spheroids, have been employed in assays that measure different types of genetic toxicity endpoints, including DNA damage, mutations, and chromosomal damage. This review summarizes published studies that have used HepaRG cells for genotoxicity assessment, including cell model evaluation studies and risk assessment for various compounds. Both 2D and 3D HepaRG models can be adapted to several high-throughput genotoxicity assays, generating a large number of data points that facilitate quantitative benchmark concentration modeling. With further validation, HepaRG cells could serve as a unique, human-based new alternative methodology for in vitro genotoxicity testing.

Evaluating the mutagenicity of N-nitrosodimethylamine in 2D and 3D HepaRG cell cultures using error-corrected next generation sequencing.

Human liver-derived metabolically competent HepaRG cells have been successfully employed in both two-dimensional (2D) and 3D spheroid formats for performing the comet assay and micronucleus (MN) assay. In the present study, we have investigated expanding the genotoxicity endpoints evaluated in HepaRG cells by detecting mutagenesis using two error-corrected next generation sequencing (ecNGS) technologies, Duplex Sequencing (DS) and High-Fidelity (HiFi) Sequencing. Both HepaRG 2D cells and 3D spheroids were exposed for 72 h to N-nitrosodimethylamine (NDMA), followed by an additional incubation for the fixation of induced mutations. NDMA-induced DNA damage, chromosomal damage, and mutagenesis were determined using the comet assay, MN assay, and ecNGS, respectively. The 72-h treatment with NDMA resulted in concentration-dependent increases in cytotoxicity, DNA damage, MN formation, and mutation frequency in both 2D and 3D cultures, with greater responses observed in the 3D spheroids compared to 2D cells. The mutational spectrum analysis showed that NDMA induced predominantly A:T → G:C transitions, along with a lower frequency of G:C → A:T transitions, and exhibited a different trinucleotide signature relative to the negative control. These results demonstrate that the HepaRG 2D cells and 3D spheroid models can be used for mutagenesis assessment using both DS and HiFi Sequencing, with the caveat that severe cytotoxic concentrations should be avoided when conducting DS. With further validation, the HepaRG 2D/3D system may become a powerful human-based metabolically competent platform for genotoxicity testing.

CircTLK1: A novel regulator involved in VSMC phenotypic switching and the developmental process of atherosclerosis.

Phenotypic switching of vascular smooth muscle cells (VSMCs) is essential for atherosclerosis development. Circular RNA (circRNA) is a specific non-coding RNA that is produced as a closed-loop structure in mammals, and its specific expression pattern is closely related to its cell type and tissue. To clarify the roles of circTLK1 in VSMC phenotypic switching, we performed qRT-PCR, immunoblotting, and immunostaining. qRT-PCR revealed that circTLK1 was upregulated in both mouse models of atherosclerosis in vivo and PDGF (platelet-derived growth factor)-BB-induced VSMCs in vitro. Furthermore, the overexpression of circTLK1 promoted PDGF-BB-induced VSMC phenotypic switching. Conversely, experiments performed in vivo demonstrate that the knockdown of SMC-specific circTLK1 led to a reduction in the development of atherosclerosis. The relationship between circTLK1 and miR-513a-3p and Krüppel-like factor 4 (KLF4) was detected by RNA immunoprecipitation (RIP), luciferase reporter assay, RNA pull-down, and RNA fluorescence in situ hybridization (RNA FISH). Mechanistically, circTLK1 acted as a sponge for miR-513a-3p, leading to the upregulation of KLF4, a key transcription factor for phenotypic switching. Targeting the circTLK1/miR-513a-3p/KLF4 axis may provide a potential therapeutic strategy for atherosclerosis.

Intraoperative frozen section pathology of vaginal margin in radical hysterectomy on the prognosis and quality of life for patients with IB2-IIA2 cervical cancer: study protocol for a multicenter randomized controlled trial.

BACKGROUND: Several risk factors have been identified that compromise the treatment outcome in patients with early-to-mid-stage cervical cancer (CC) who are primarily treated with radical surgery. However, there is no report on the impact of intraoperative frozen pathology examination of vaginal margins on the prognosis of patients with CC. This study aimed to conduct a randomized controlled trial (RCT) to determine whether selective vaginal resection can reduce the incidence of operative complications and the risk of postoperative radiotherapy. The impact of the length of the vagina removed in radical hysterectomy (RH) on prognosis and quality of life (QoL) for IB2-IIA2 CC patients will be investigated. METHODS: A multicenter, non-inferiority, RCT at 7 institutions in China is designed to investigate the effect of intraoperative frozen pathology exam of vaginal margin in RH on the survival outcomes for patients with IB2-IIA2 CC. Eligible patients aged 18-70 years will be randomly assigned online by one-to-one random allocation to receive intraoperative frozen pathology exam of vaginal margin or not. If frozen pathology indicates positive margin, continue resection of 1 centimeter of vaginal tissue until negative margin is achieved. The primary end point is 2-year disease-free survival (DFS). Adverse events (AEs) caused by further vagina resection, 5-year DFS, 2-year overall survival (OS), 5-year OS and AEs caused by radiotherapy and QoL are secondary end points. A total of 310 patients will be enrolled from 7 tertiary hospitals in China within 3-year period and followed up for 5 years. TRIAL REGISTRATION: Chinese Clinical Trial Registry Identifier: ChiCTR2000035668.

Subtle adjustments for constructing multi-nuclear luminescent lanthanide organic polyhedra with triazole-based chelates.

Controlled self-assembly of predetermined multi-nuclear lanthanide organic polyhedra (LOPs) still presents a challenge, primarily due to the unpredictable coordination numbers and labile coordination geometries of lanthanide ions. In this study, through introducing triazole-based chelates to increase the chelating angle of C2-symmetric linear ligands and stabilize the coordination geometry of Eu(III) centers, M4L6-type (M = EuIII, L = ligand) tetrahedra were efficiently synthesized, especially a biphenyl-bridged ligand which is well known to form M2L3-type helicates. A series of LOPs were formed and characterized by high-resolution electrospray ionization time-of-flight mass spectroscopy (ESI-TOF-MS) and X-ray crystallography. Moreover, the europium complexes exhibit bright emission (luminescence quantum yield up to 42.4%) and circularly polarized luminescence properties (|glum| up to 4.5 × 10-2). This study provides a feasible strategy for constructing multi-nuclear luminescent LOPs towards potential applications.

Integrated lithium niobate microwave photonic processing engine.

Integrated microwave photonics (MWP) is an intriguing technology for the generation, transmission and manipulation of microwave signals in chip-scale optical systems1,2. In particular, ultrafast processing of analogue signals in the optical domain with high fidelity and low latency could enable a variety of applications such as MWP filters3-5, microwave signal processing6-9 and image recognition10,11. An ideal integrated MWP processing platform should have both an efficient and high-speed electro-optic modulation block to faithfully perform microwave-optic conversion at low power and also a low-loss functional photonic network to implement various signal-processing tasks. Moreover, large-scale, low-cost manufacturability is required to monolithically integrate the two building blocks on the same chip. Here we demonstrate such an integrated MWP processing engine based on a 4 inch wafer-scale thin-film lithium niobate platform. It can perform multipurpose tasks with processing bandwidths of up to 67 GHz at complementary metal-oxide-semiconductor (CMOS)-compatible voltages. We achieve ultrafast analogue computation, namely temporal integration and differentiation, at sampling rates of up to 256 giga samples per second, and deploy these functions to showcase three proof-of-concept applications: solving ordinary differential equations, generating ultra-wideband signals and detecting edges in images. We further leverage the image edge detector to realize a photonic-assisted image segmentation model that can effectively outline the boundaries of melanoma lesion in medical diagnostic images. Our ultrafast lithium niobate MWP engine could provide compact, low-latency and cost-effective solutions for future wireless communications, high-resolution radar and photonic artificial intelligence.

Disentangle Then Calibrate with Gradient Guidance: A Unified Framework for Common and Rare Disease Diagnosis.

The computer-aided diagnosis (CAD) for rare diseases using medical imaging poses a significant challenge due to the requirement of large volumes of labeled training data, which is particularly difficult to collect for rare diseases. Although Few-shot learning (FSL) methods have been developed for this task, these methods focus solely on rare disease diagnosis, failing to preserve the performance in common disease diagnosis. To address this issue, we propose the Disentangle then Calibrate with Gradient Guidance (DCGG) framework under the setting of generalized few-shot learning, i.e., using one model to diagnose both common and rare diseases. The DCGG framework consists of a network backbone, a gradient-guided network disentanglement (GND) module, and a gradient-induced feature calibration (GFC) module. The GND module disentangles the network into a disease-shared component and a disease-specific component based on gradient guidance, and devises independent optimization strategies for both components, respectively, when learning from rare diseases. The GFC module transfers only the disease-shared channels of common-disease features to rare diseases, and incorporates the optimal transport theory to identify the best transport scheme based on the semantic relationship among different diseases. Based on the best transport scheme, the GFC module calibrates the distribution of rare-disease features at the disease-shared channels, deriving more informative rare-disease features for better diagnosis. The proposed DCGG framework has been evaluated on three public medical image classification datasets. Our results suggest that the DCGG framework achieves state-of-the-art performance in diagnosing both common and rare diseases.

Novel serum biomarker of Golgi protein 73 for the diagnosis of clinically significant portal hypertension in patients with compensated cirrhosis.

Hepatic venous pressure gradient (HVPG) is the gold standard for evaluating clinically significant portal hypertension (CSPH). However, reliable noninvasive methods are limited. Our study aims to investigate the diagnostic value of serum Golgi protein 73 (GP73) for CSPH in patients with compensated cirrhosis. The study enrolled 262 consecutive patients with compensated cirrhosis from three centers in China from February 2021 to September 2023, who underwent both serum GP73 tests and HVPG measurements. CSPH was defined as HVPG ≥ 10 mmHg. Diagnostic accuracy was evaluated using the areas under the receiver operating characteristic curve (AUC). The prevalence of CSPH was 56.9% (n = 149). There were significant differences between the CSPH and non-CSPH groups in the median serum GP73 level (126.8 vs. 73.1 ng/mL, p < 0.001). GP73 level showed a significant positive linear correlation with HVPG (r = 0.459, p < 0.001). The AUC for the diagnosis of CSPH using serum GP73 alone was 0.75 (95% confidence interval [CI] 0.68-0.81). Multivariate logistic regression analysis determined that the levels of GP73, platelets and international normalized ratio were independently associated with CSPH. The combination of these three markers was termed "IP73" score with an AUC value of 0.85 (95% CI 0.80-0.89) for CSPH. Using 0 as a cut-off value, the specificity and sensitivity of IP73 score were 77.9% and 81.9%, respectively. The IP73 score offers a novel, simple and noninvasive method of assessing CSPH in patients with compensated cirrhosis. A cut-off value of the IP73 score at 0 can distinguish patients with or without CSPH.

Key Challenges for In Vitro Testing of Tobacco Products for Regulatory Applications: Recommendations for the In Vitro Mouse Lymphoma Assay.

The Institute for In Vitro Sciences (IIVS) is sponsoring a series of workshops to develop recommendations for optimal scientific and technical approaches for conducting in vitro assays to assess potential toxicity within and across traditional tobacco and various tobacco and nicotine next-generation products (NGPs), including Heated Tobacco Products (HTPs) and Electronic Nicotine Delivery Systems (ENDS). This report was developed by a working group composed of attendees of the seventh IIVS workshop, 'Approaches and recommendations for conducting the mouse lymphoma gene mutation assay (MLA) and introduction to in vitro disease models', which was held virtually on 21-23 June 2022. This publication provides a background overview of the MLA, and includes the description of assay conduct and data interpretation, key challenges and recommended best practices for evaluating tobacco and nicotine products, with a focus on the evaluation of NGPs, and a summary of how the assay has been used to evaluate and compare tobacco and nicotine products.

MCM2 promotes the stemness of endometrial cancer cells via the Akt/ß-catenin pathway.

Minichromosome maintenance complex component 2 (MCM2) is a member of the MCM family and is involved in various cancers. However, the role of MCM2 in endometrial cancer (EC) remains unclear. In this study, we aim to determine the biological function of MCM2 in EC cells and identify the potential underlying mechanisms. MCM2 expression and prognostic significance were analyzed in TCGA-UCEC datasets. Combining bioinformatics analyses and experiments, stemness-related molecules and phenotypes were examined to evaluate the impact of MCM2 on stemness in EC cells. The major findings of these analyses are as follows: 1) MCM2 is expressed at higher levels in EC tissues than in normal endometrial tissues. High expression of MCM2 is related to the characteristics of poorly differentiated EC. High MCM2 expression is correlated with poor overall survival in EC patients; 2) MCM2 knockdown was found to decrease sphere formation ability, downregulate the expression of stemness-related molecules, and reduce the proportion of CD133+ cells, while MCM2 overexpression elicited the opposite effect in EC cells; 3) MCM2-mediated stemness features are dependent on the activation of Akt/ß-catenin signaling pathways; and 4) MCM2 knockdown increases cisplatin sensitivity in EC cells. MCM2 regulates stemness by regulating the Akt/ß-catenin signaling pathway in EC cells.

Molecular mechanisms of AMPK/YAP/NLRP3 signaling pathway affecting the occurrence and development of ankylosing spondylitis.

BACKGROUND: Investigate the AMPK (protein kinase AMP-activated catalytic subunit alpha 1)/YAP (Yes1 associated transcriptional regulator)/NLRP3 (NLR family pyrin domain containing 3) signaling pathway's role in ankylosing spondylitis (AS) development using public database analysis, in vitro and in vivo experiments. METHODS: Retrieve AS dataset, analyze differential gene expression in R, conduct functional enrichment analysis, collect 30 AS patient and 30 normal control samples, and construct a mouse model. ELISA, IP, and knockdown experiments were performed to detect expression changes. RESULTS: NLRP3 was identified as a significant AS-related gene. Caspase-1, IL-1ß, IL-17A, IL-18, IL-23, YAP, and NLRP3 were upregulated in AS patients. Overexpressing AMPK inhibited YAP's blockade on NLRP3 ubiquitination, reducing ossification in fibroblasts. Inhibiting AMPK exacerbated AS symptoms in AS mice. CONCLUSION: AMPK may suppress YAP expression, leading to NLRP3 inflammasome inhibition and AS alleviation.

Excited-Multimer Mediated Supramolecular Upconversion on Multicomponent Lanthanide-Organic Assemblies.

Upconversion (UC) is a fascinating anti-Stokes-like optical process with promising applications in diverse fields. However, known UC mechanisms are mainly based on direct energy transfer between metal ions, which constrains the designability and tunability of the structures and properties. Here, we synthesize two types of Ln8L12-type (Ln for lanthanide ion; L for organic ligand L1 or L2R/S) lanthanide-organic complexes with assembly induced excited-multimer states. The Yb8(L2R/S)12 assembly exhibits upconverted multimer green fluorescence under 980 nm excitation through a cooperative sensitization process. Furthermore, upconverted red emission from Eu3+ on the heterometallic (Yb/Eu)8L12 assemblies is also realized via excited-multimer mediated energy relay. Our findings demonstrate a new strategy for designing UC materials, which is crucial for exploiting photofunctions of multicomponent lanthanide-organic complexes.

The involvement of hepatic cytochrome P450s in the cytotoxicity of lapatinib.

Lapatinib, an oral tyrosine kinase inhibitor used as a first-line treatment for HER2-positive breast cancer, has been reported to be associated with hepatotoxicity; however, the underlying mechanisms remain unclear. In this study, we report that lapatinib causes cytotoxicity in multiple types of hepatic cells, including primary human hepatocytes, HepaRG cells, and HepG2 cells. A 24-h treatment with lapatinib induced cell cycle disturbances, apoptosis, and DNA damage, and decreased the protein levels of topoisomerase in HepG2 cells. We investigated the role of cytochrome P450 (CYP)-mediated metabolism in lapatinib-induced cytotoxicity using our previously established HepG2 cell lines, which express each of 14 CYPs (1A1, 1A2, 1B1, 2A6, 2B6, 2C8, 2C9, 2C18, 2C19, 2D6, 2E1, 3A4, 3A5, and 3A7). We demonstrate that lapatinib is metabolized by CYP1A1, 3A4, 3A5, and 3A7. Among these, lapatinib-induced cytotoxicity and DNA damage were attenuated in cells overexpressing CYP3A5 or 3A7. Additionally, we measured the production of three primary metabolites of lapatinib (O-dealkylated lapatinib, N-dealkylated lapatinib, and N-hydroxy lapatinib) in CYP1A1-, 3A4-, 3A5-, and 3A7-overexpressing HepG2 cells. We compared the cytotoxicity of lapatinib and its 3 metabolites in primary human hepatocytes, HepaRG cells, and HepG2 cells and demonstrated that N-dealkylated lapatinib is more toxic than the parent drug and the other metabolites. Taken together, our results indicate that lapatinib-induced cytotoxicity involves multiple mechanisms, such as apoptosis and DNA damage; that N-dealkylated lapatinib is a toxic metabolite contributing to the toxic effect of lapatinib; and that CYP3A5- and 3A7-mediated metabolism plays a role in attenuating the cytotoxicity of lapatinib.

Umbilical Cord Blood-Derived M1 Macrophage Exosomes Loaded with Cisplatin Target Ovarian Cancer In Vivo and Reverse Cisplatin Resistance.

We investigated the therapeutic efficacy of umbilical cord blood (UCB)-derived M1 macrophage exosomes loaded with cisplatin (CIS) in ovarian cancer and platinum resistance. M1 macrophages were purified by using CD14 magnetic beads and characterized by flow cytometry. Our analyses included morphology, particle size, particle concentration, potential, drug loading capacity, counts of entry into cells, antitumor effect in vivo, and the ability to reverse drug resistance. A2780, SKOV3, and A2780/DDP, SKOV3/DDP ovarian cancer cells (CIS-sensitive and CIS-resistant cell lines, respectively) were treated with CIS or CIS-loaded M1 macrophage exosomes (M1exoCISs). The encapsulation efficiency of CIS loading into M1 macrophage exosomes was approximately 30%. In vitro, M1exoCIS treatment reduced the CIS IC50 values of both A2780, SKOV3, and A2780/DDP, SKOV3/DDP cells. We evaluated the effect of M1exoCIS on tumor growth using a mouse ovarian cancer subcutaneous transplantation tumor model inoculated with A2780/DDP cells. M1exoCIS was observed in the liver, spleen, and tumor sites 24 h posttreatment; the fluorescence intensity of M1exoCIS is higher than that of CIS. After 7 days, M1exoCIS significantly inhibited the growth of subcutaneously transplanted tumors compared with CIS alone and had a longer survival time. Moreover, the toxicity test shows that M1exoCIS has less hepatorenal toxicity than CIS. To investigate the mechanism of M1exoCIS targeting, homing, and reversing drug resistance, we performed RT-PCR, Western blotting, and Proteome Profiler Human Receptor Array analyses. We found that A2780 and A2780/DDP cells expressed the integrin ß1/CD29 receptor, while M1 exosomes expressed integrin ß1/CD29. In addition, M1exos carries long noncoding RNA H19, implicated in PTEN protein upregulation and miR-130a and Pgp gene downregulation, leading to the reversal of CIS drug resistance. Therefore, UCB-derived M1exoCIS target tumor sites of ovarian cancer in vivo and can be used to increase the CIS sensitivity and cytotoxicity.

Handling Open-Set Noise and Novel Target Recognition in Domain Adaptive Semantic Segmentation.

This paper studies a practical domain adaptive (DA) semantic segmentation problem where only pseudo-labeled target data is accessible through a black-box model. Due to the domain gap and label shift between two domains, pseudo-labeled target data contains mixed closed-set and open-set label noises. In this paper, we propose a simplex noise transition matrix (SimT) to model the mixed noise distributions in DA semantic segmentation, and leverage SimT to handle open-set label noise and enable novel target recognition. When handling open-set noises, we formulate the problem as estimation of SimT. By exploiting computational geometry analysis and properties of segmentation, we design four complementary regularizers, i.e., volume regularization, anchor guidance, convex guarantee, and semantic constraint, to approximate the true SimT. Specifically, volume regularization minimizes the volume of simplex formed by rows of the non-square SimT, ensuring outputs of model to fit into the ground truth label distribution. To compensate for the lack of open-set knowledge, anchor guidance, convex guarantee, and semantic constraint are devised to enable the modeling of open-set noise distribution. The estimated SimT is utilized to correct noise issues in pseudo labels and promote the generalization ability of segmentation model on target domain data. In the task of novel target recognition, we first propose closed-to-open label correction (C2OLC) to explicitly derive the supervision signal for open-set classes by exploiting the estimated SimT, and then advance a semantic relation (SR) loss that harnesses the inter-class relation to facilitate the open-set class sample recognition in target domain. Extensive experimental results demonstrate that the proposed SimT can be flexibly plugged into existing DA methods to boost both closed-set and open-set class performance.

Construction of an Immunophenoscore-Related Signature for Evaluating Prognosis and Immunotherapy Sensitivity in Ovarian Cancer.

Ovarian cancer (OC) is the deadliest gynecological malignancy in the world, and immunotherapy is emerging as a promising treatment. Immunophenoscore (IPS) is a robust biomarker distinguishing sensitive responders from immunotherapy. In this study, we aimed to construct a prognostic model for predicting overall survival (OS) and identifying patients who would benefit from immunotherapy. First, we combined The Cancer Genome Atlas (TCGA) and The Cancer Immune Atlas (TCIA) data sets and incorporated 229 OC samples into a training cohort. The validation cohort included 240 OC samples from the Gene Expression Omnibus (GEO) cohort. The training cohort was divided into high- and low-IPS subgroups to obtain differentially expressed genes (DEGs). DEGs with OS were identified by Univariate Cox regression analysis. The least absolute shrinkage and selection operator (LASSO) Cox regression was used to construct the prognostic model. Then, immune and mutation analyses were performed to explore the relationship between the model and the tumor microenvironment (TME) and tumor mutation burden (TMB). Eighty-three DEGs were obtained between the high-and low-IPS subgroups, where 17 DEGs were associated with OS. The five essential genes were selected to establish the prognostic model, which showed high accuracy for predicting OS and could be an independent survival indicator. OC samples that were divided by risk scores showed distinguished immune status, TME, TMB, immunotherapy response, and chemotherapy sensitivity. Similar results were validated in the GEO cohort. We developed an immunophenoscore-related signature associated with the TME to predict OS and response to immunotherapy in OC.

Complex Blood Supply Patterns in Cesarean Scar Pregnancy: Insights from Digital Subtraction Angiography Imaging.

BACKGROUND Understanding the blood supply pattern of cesarean scar pregnancy (CSP) can effectively help to determine the best choice of treatment. The aim of this study was to investigate the blood supply pattern and outcomes of patients with CSP through digital subtraction angiography (DSA) imaging. MATERIAL AND METHODS This was a retrospective cohort study. Patients were divided into 2 groups according to the type of CSP. The DSA images of these patients were reviewed, including the type of blood supply, dominant vessel, and collateral blood supply to the gestational sac. The clinical outcomes were analyzed between the 2 groups. RESULTS Thirty-seven patients with type I and 29 patients with type II CSP were enrolled in this study. Type II CSP showed a higher proportion of rich blood supply than type I (44.83% vs 29.72%, P>0.05). Compared with type II CSP, type I CSP tended to have bilateral dominant blood supply predominance (67.57% vs 41.38%, P<0.05). The incidence of collateral blood supply was 5.41% in the type I CSP group and 31.03% in the type II CSP group (P<0.05). In the type II CSP group, multiple collateral blood vessels were found in 4 patients. The superior vesicle artery was the most common source of collateral blood supply in both groups. Two patients with type II CSP suffered massive bleeding during surgery after uterine artery embolization (UAE). None of the patients received a hysterectomy. CONCLUSIONS UAE is safe and effective for both types of CSP. The blood supply pattern is more complex and abnormal in type II CSP. More attention should be paid to the collateral blood supply to achieve complete embolization during the UAE procedure in the case of type II CSP.

Dynamic feature splicing for few-shot rare disease diagnosis.

Annotated images for rare disease diagnosis are extremely hard to collect. Therefore, identifying rare diseases under a few-shot learning (FSL) setting is significant. Existing FSL methods transfer useful and global knowledge from base classes with abundant training samples to enrich features of novel classes with few training samples, but still face difficulties when being applied to medical images due to the complex lesion characteristics and large intra-class variance. In this paper, we propose a dynamic feature splicing (DNFS) framework for few-shot rare disease diagnosis. Under DNFS, both low-level features (i.e., the output of three convolutional blocks) and high-level features (i.e., the output of the last fully connected layer) of novel classes are dynamically enriched. We construct the position coherent DNFS (P-DNFS) module to perform low-level feature splicing, where a lesion-oriented Transformer is designed to detect lesion regions. Thus, novel-class channels are replaced by similar base-class channels within the detected lesion regions to achieve disease-related feature enrichment. We also devise a semantic coherent DNFS (S-DNFS) module to perform high-level feature splicing. It explores cross-image channel relations and selects base-class channels with semantic consistency for explicit knowledge transfer. Both low-level and high-level feature splicings are performed dynamically and iteratively. Consequently, abundant spliced features are generated for disease diagnosis, leading to more accurate decision boundary and improved diagnosis performance. Extensive experiments have been conducted on three medical image classification datasets. Our results suggest that the proposed DNFS achieves superior performance against state-of-the-art approaches.