Explore the vast range of titles available.

Over the past two decades, elucidation of the genetic defects that underlie cancer has resulted in a plethora of novel targeted cancer drugs. Although these agents can initially be highly effective, resistance to single-agent therapies remains a major challenge. Combining drugs can help avoid resistance, but the number of possible drug combinations vastly exceeds what can be tested clinically, both financially and in terms of patient availability. Rational drug combinations based on a deep understanding of the underlying molecular mechanisms associated with therapy resistance are potentially powerful in the treatment of cancer. Here, we discuss the mechanisms of resistance to targeted therapies and how effective drug combinations can be identified to combat resistance. The challenges in clinically developing these combinations and future perspectives are considered.Single-agent therapies targeting specific dysregulated pathways in cancer can be highly effective, but drug resistance frequently develops. Here, Bernards and colleagues discuss the mechanisms underlying resistance to targeted therapies, and assess how these can be suppressed by using tailored combination therapies.

Industrial supply chains continue to face inefficiencies, fragmented decision‐making, and disruptions that weaken resilience and competitiveness. Traditional mechanisms often struggle to balance cost minimization, timely delivery, and adaptability under uncertainty, creating a need for advanced solutions. However, existing AI approaches address these challenges in isolation—predictive, prescriptive, and collaborative mechanisms are rarely unified into a single adaptive framework, leaving a critical gap in achieving end‐to‐end supply chain intelligence. This paper proposes an integrated AI‐enabled framework that combines predictive modeling, prescriptive modeling, and collaborative decision‐making to enhance efficiency in industrial supply chains. The predictive layer employs Long Short‐Term Memory (LSTM) networks for demand forecasting and equipment failure prediction, capturing temporal dependencies and irregular demand shocks with high accuracy. The proposed LSTM model demonstrated superior predictive accuracy, achieving an MAE of 0.01755, an MAPE of 0.0283, RMSE of 0.0235, and an MSE of 0.00006, values significantly lower than those reported by existing methods, confirming its capability to capture complex temporal patterns with high fidelity. The prescriptive layer applies Reinforcement Learning (RL) to dynamically optimize routing, scheduling, and inventory strategies, ensuring cost reduction and lead time improvement under varying operational scenarios. Collaborative decision‐making is achieved through decision fusion and rule‐based heuristics, which integrate outputs from predictive and prescriptive models while embedding domain expertise to ensure realistic and implementable strategies. Efficiency improvement paths are evaluated using key performance indicators, including cost reduction, lead time optimization, inventory turnover, service level enhancement, and sustainability metrics, thereby demonstrating measurable operational gains. The novelty of this research lies in its holistic integration of predictive, prescriptive, and collaborative layers into a unified AI‐enabled framework, moving beyond isolated optimization approaches to establish adaptive ecosystems capable of continuous learning and improvement. Industrial supply chains continue to face inefficiencies, fragmented decision‐making, and disruptions that weaken resilience and competitiveness. Traditional mechanisms often struggle to balance cost minimization, timely delivery, and adaptability under uncertainty, creating a need for advanced solutions. This paper proposes an integrated AI‐enabled framework that combines predictive modeling, prescriptive modeling, and collaborative decision‐making to enhance efficiency in industrial supply chains. The predictive layer employs Long Short‐Term Memory (LSTM) networks for demand forecasting and equipment failure prediction, capturing temporal dependencies and irregular demand shocks with high accuracy. The proposed LSTM model demonstrated superior predictive accuracy, achieving MAE of 0.01755, MAPE of 0.0283, RMSE of 0.0235, and MSE of 0.00006, values significantly lower than those reported by existing methods, confirming its capability to capture complex temporal patterns with high fidelity. The prescriptive layer applies Reinforcement Learning (RL) to dynamically optimize routing, scheduling, and inventory strategies, ensuring cost reduction and lead time improvement under varying operational scenarios. Collaborative decision‐making is achieved through decision fusion and rule‐based heuristics, which integrate outputs from predictive and prescriptive models while embedding domain expertise to ensure realistic and implementable strategies. Efficiency improvement paths are evaluated using key performance indicators, including cost reduction, lead time optimization, inventory turnover, service level enhancement, and sustainability metrics, thereby demonstrating measurable operational gains. The novelty of this research lies in its holistic integration of predictive, prescriptive, and collaborative layers into a unified AI‐driven framework, moving beyond isolated optimization approaches to establish adaptive ecosystems capable of continuous learning and improvement.

Hepatocellular carcinoma (HCC)—the most common form of liver cancer—is an aggressive malignancy with few effective treatment options 1 . Lenvatinib is a small-molecule inhibitor of multiple receptor tyrosine kinases that is used for the treatment of patients with advanced HCC, but this drug has only limited clinical benefit 2 . Here, using a kinome-centred CRISPR–Cas9 genetic screen, we show that inhibition of epidermal growth factor receptor (EGFR) is synthetic lethal with lenvatinib in liver cancer. The combination of the EGFR inhibitor gefitinib and lenvatinib displays potent anti-proliferative effects in vitro in liver cancer cell lines that express EGFR and in vivo in xenografted liver cancer cell lines, immunocompetent mouse models and patient-derived HCC tumours in mice. Mechanistically, inhibition of fibroblast growth factor receptor (FGFR)  by lenvatinib treatment leads to feedback activation of the EGFR–PAK2–ERK5 signalling axis, which is blocked by EGFR inhibition. Treatment of 12 patients with advanced HCC who were unresponsive to lenvatinib treatment with the combination of lenvatinib plus gefitinib (trial identifier NCT04642547) resulted in meaningful clinical responses. The combination therapy identified here may represent a promising strategy for the approximately 50% of patients with advanced HCC who have high levels of EGFR. EGFR inhibition and lenvatinib treatment of liver cancer cells in vitro and in in vivo mouse models has potent anti-proliferative effects, and lenvatinib plus gefitinib treatment of 12 patients with advanced liver cancer resulted in meaningful clinical responses.

Liver cancer remains difficult to treat, owing to a paucity of drugs that target critical dependencies 1 , 2 ; broad-spectrum kinase inhibitors such as sorafenib provide only a modest benefit to patients with hepatocellular carcinoma 3 . The induction of senescence may represent a strategy for the treatment of cancer, especially when combined with a second drug that selectively eliminates senescent cancer cells (senolysis) 4 , 5 . Here, using a kinome-focused genetic screen, we show that pharmacological inhibition of the DNA-replication kinase CDC7 induces senescence selectively in liver cancer cells with mutations in TP53 . A follow-up chemical screen identified the antidepressant sertraline as an agent that kills hepatocellular carcinoma cells that have been rendered senescent by inhibition of CDC7. Sertraline suppressed mTOR signalling, and selective drugs that target this pathway were highly effective in causing the apoptotic cell death of hepatocellular carcinoma cells treated with a CDC7 inhibitor. The feedback reactivation of mTOR signalling after its inhibition 6 is blocked in cells that have been treated with a CDC7 inhibitor, which leads to the sustained inhibition of mTOR and cell death. Using multiple in vivo mouse models of liver cancer, we show that treatment with combined inhibition of of CDC7 and mTOR results in a marked reduction of tumour growth. Our data indicate that exploiting an induced vulnerability could be an effective treatment for liver cancer. CDC7 inhibition selectively induces senescence in hepatocellular carcinoma cells with TP53 mutations, which enables the selective apoptotic cell death of these senescent cells using inhibitors of mTOR signalling.

Existing research is limited, and evidence suggests that urinary incontinence may be induced by certain medications, highlighting the urgent need for systematic studies. This study uses the reporting odds ratio (ROR) to evaluate the reports of drug-induced urinary incontinence in the FAERS database from the first quarter of 2004 to the fourth quarter of 2024.Univariate analysis, LASSO regression, and multivariate regression analysis were conducted to further explore the risk factors for drug-induced urinary incontinence. Bonferroni correction was applied to the results of the multiple comparisons. Additionally, the Weibull distribution test was used to assess the temporal characteristics of drug-induced urinary incontinence. Multivariate regression analysis ultimately identified 19 medications as independent risk factors for drug-induced urinary incontinence, including neuropsychiatric drugs (13/19), gastrointestinal and metabolic drugs (2/19), musculoskeletal system drugs (2/19), cardiovascular drugs (1/19), and urogenital and sex hormone drugs (1/19). Over half (57.12%) of the cases of drug-induced urinary incontinence occurred within 30 days after the initiation of medication. This study provides important insights for clinicians in preventing drug-induced urinary incontinence. However, future mechanistic studies and randomized controlled trials are needed to further elucidate and validate these findings.

The blood-brain barrier (BBB), as a crucial gate of brain-blood molecular exchange, is involved in the pathogenesis of multiple neurological diseases. Oxidative stress is caused by an imbalance between the production of reactive oxygen species (ROS) and the scavenger system. Since oxidative stress plays a significant role in the production and maintenance of the BBB, the cerebrovascular system is especially vulnerable to it. The pathways that initiate BBB dysfunction include, but are not limited to, mitochondrial dysfunction, excitotoxicity, iron metabolism, cytokines, pyroptosis, and necroptosis, all converging on the generation of ROS. Interestingly, ROS also provide common triggers that directly regulate BBB damage, parameters including tight junction (TJ) modifications, transporters, matrix metalloproteinase (MMP) activation, inflammatory responses, and autophagy. We will discuss the role of oxidative stress-mediated BBB disruption in neurological diseases, such as hemorrhagic stroke, ischemic stroke (IS), Alzheimer’s disease (AD), Parkinson’s disease (PD), traumatic brain injury (TBI), amyotrophic lateral sclerosis (ALS), and cerebral small vessel disease (CSVD). This review will also discuss the latest clinical evidence of potential biomarkers and antioxidant drugs towards oxidative stress in neurological diseases. A deeper understanding of how oxidative stress damages BBB may open up more therapeutic options for the treatment of neurological diseases.

Accumulated data implicate that long noncoding RNA (lncRNA) plays a pivotal role in rheumatoid arthritis (RA), potentially serving as a competitive endogenous RNA (ceRNA) for microRNAs (miRNAs). The lncRNA myocardial infarction-associated transcript (MIAT) has been demonstrated to regulate inflammation. However, the role of MIAT in the inflammation of RA remains inadequately explored. This study aims to elucidate MIAT’s role in the inflammation of lipopolysaccharide (LPS)-induced macrophages and to uncover the underlying molecular mechanisms. We observed heightened MIAT expression in LPS-induced J774A.1 cells and collagen-induced arthritis mouse models, in contrast to the expression pattern of miR-30a-5p. Silencing MIAT resulted in increased expression of the inflammatory cytokines IL-1β and TNF-α. Simultaneously, MIAT interference significantly impeded macrophage autophagy, evidenced by decreased expression of autophagy-related markers LC3-II and Beclin-1, alongside increased levels of p62 in LPS-induced J774A.1 cells. Notably, MIAT functioned as a ceRNA, sponging miR-30a-5p and exerting a negative regulatory influence on its expression. SOCS1 emerged as a target of miR-30a-5p, modulated by MIAT. Mechanistically, inhibiting miR-30a-5p reversed the impact of MIAT deficiency in promoting LPS-induced inflammation, while SOCS1 knockdown countered the cytokine inhibitory effect induced by silencing miR-30a-5p. In summary, this study indicates that lncRNA MIAT suppresses inflammation in LPS-induced J774A.1 macrophages by stimulating autophagy through the miR-30a-5p/SOCS1 axis. This suggests that MIAT holds promise as a potential therapeutic target for RA inflammation.

Background Chronic obstructive pulmonary disease (COPD) is a major global public health concern, and early screening and identification of high-risk populations are critical for reducing the disease burden. Although several studies have explored the application of machine learning methods in COPD risk prediction, existing models often have limited feature dimensions and insufficient interpretability. Identifying key risk factors and constructing reliable predictive models remain challenges in clinical practice. Objective This study aims to integrate multidimensional features based on data from the National Health and Nutrition Examination Survey (NHANES) and to compare the performance of different machine learning models in COPD risk prediction. The goal is to identify the optimal model and enhance its clinical applicability through interpretability analysis. Methods This study utilized data from the NHANES collected between 2009 and 2018. After systematic feature selection and preprocessing, three models were developed: multivariate binary logistic regression, XGBoost, and Multilayer Perceptron (MLP). Model training and evaluation were performed using stratified five-fold cross-validation. Model performance was comprehensively assessed based on accuracy, precision, recall, F1 score, and the area under the receiver operating characteristic curve (AUC). To enhance model transparency, the SHapley Additive Explanations (SHAP) method was employed to interpret key features and their influence trends within the MLP model. Results The MLP model demonstrated the best performance across all evaluation metrics, achieving an average accuracy of 0.937, precision of 0.6624, recall of 0.6535, and F1 score of 0.657 in stratified five-fold cross-validation. The performance gap between the training and testing sets was minimal, indicating no obvious overfitting. SHAP analysis identified smoking years, asthma, age, dietary health status, total protein, red cell distribution width (RDW), BMI, marital status, secondhand smoke exposure, and total bilirubin as important predictive features. Furthermore, dependence plots revealed critical risk inflection points for key continuous variables. Conclusion Based on large-scale and multidimensional feature data, this study constructed a COPD risk prediction model with favorable performance and enhanced interpretability. The findings suggest that the MLP model has the potential to effectively identify individuals at high risk for COPD and may offer value in clinical applications. Future studies are warranted to integrate longitudinal follow-up data and multimodal information to further improve predictive accuracy and clinical interpretability, thereby providing a more robust foundation for early screening and personalized interventions in COPD.

The immune system has a central role in eliminating detrimental factors, by frequently launching inflammatory responses towards pathogen infection and inner danger signal outbreak. Acute and chronic inflammatory responses are critical determinants for consequences of kidney diseases, in which inflammasomes were inevitably involved. Inflammasomes are closely linked to many kidney diseases such as acute kidney injury and chronic kidney diseases. Inflammasomes are macromolecules consisting of multiple proteins, and their formation initiates the cleavage of procaspase-1, resulting in the activation of gasdermin D as well as the maturation and release of interleukin-1β and IL-18, leading to pyroptosis. Here, we discuss the mechanism in which inflammasomes occur, as well as their roles in inflammatory kidney diseases, in order to shed light for discovering new therapeutical targets for the prevention and treatment of inflammatory kidney diseases and consequent end-stage renal disease.

The nod-like receptor family pyrin domain containing 3 (NLRP3) is currently the most widely studied inflammasome and has become a hot topic of recent research. As a macromolecular complex, the NLRP3 inflammasome is activated to produce downstream factors, including caspase-1, IL-1β, and IL-18, which then promote local inflammatory responses and induce pyroptosis, leading to unfavorable effects. A growing number of studies have examined the relationship between the NLRP3 inflammasome and cardiovascular diseases (CVDs). However, some studies have shown that the NLRP3 inflammasome is not involved in the occurrence of certain diseases. Therefore, identifying the mechanism of action of the NLRP3 inflammasome and its potential involvement in the pathological process of disease progression is of utmost importance. This review discusses the mechanisms of NLRP3 inflammasome activation and the relationship between the inflammasome and CVDs, including coronary atherosclerosis, myocardial ischemia/reperfusion, cardiomyopathies, and arrhythmia, as well as CVD-related treatments.