Explore the vast range of titles available.

Accurate prediction of human pharmacokinetics (PK) for lead compounds is one of the critical determinants of successful drug development. Traditional methods for PK parameter prediction, such as in vitro to in vivo extrapolation and physiologically based pharmacokinetic modeling, often require extensive experimental data and time‐consuming calibration of parameters. Machine learning (ML) has been widely applied to predict ADME and physicochemical properties (ADMEP descriptors), but studies focusing on concentration‐time (C‐t) profile prediction remain limited. In this study, we developed a Long Short‐Term Memory (LSTM) based ML framework to predict C‐t profiles following intravenous (IV) bolus drug administration in humans. The model used ADMEP descriptors generated by ADMETlab 3.0 and dose information as input. A total of 40 drugs were used for training and 18 for testing, with concentration data simulated from published PK models. Our approach achieved R2 of 0.75 across all C‐t profiles, and 77.8% of Cmax, 55.6% of clearance, and 61.1% of volume of distribution predictions within a 2‐fold error range, demonstrating predictive performance comparable to previously published ML methods. Furthermore, model performance was found to be associated with the input dose level and ADMEP descriptors, suggesting the accuracy and confidence of the prediction may be expected in advance via these descriptors. This LSTM‐based framework using a small number of compounds enables efficient prediction of human PK profiles with IV dosing, offering a practical alternative to traditional PK prediction models. It holds promise for improving early‐phase prioritizing lead compounds and reducing reliance on animals in drug development.

Real-time detection of tomato pests and diseases is essential for precision agriculture, as it requires high accuracy, speed, and energy efficiency of edge-computing agricultural robots. This study proposes AgriLiteNet (Lightweight Networks for Agriculture), a lightweight neural network integrating MobileNetV3 for local feature extraction and a streamlined Swin Transformer for global modeling. AgriLiteNet is further enhanced by a lightweight channel–spatial mixed attention module and a feature pyramid network, enabling the detection of nine tomato pests and diseases, including small targets like spider mites, dense targets like bacterial spot, and large targets like late blight. It achieves a mean average precision at an intersection-over-union threshold of 0.5 of 0.98735, which is comparable to Suppression Mask R-CNN (0.98955) and Cas-VSwin Transformer (0.98874), and exceeds the performance of YOLOv5n (0.98249) and GMC-MobileV3 (0.98143). With 2.0 million parameters and 0.608 GFLOPs, AgriLiteNet delivers an inference speed of 35 frames per second and power consumption of 15 watts on NVIDIA Jetson Orin NX, surpassing Suppression Mask R-CNN (8 FPS, 22 W) and Cas-VSwin Transformer (12 FPS, 20 W). The model’s efficiency and compact design make it highly suitable for deployment in agricultural robots, supporting sustainable farming through precise pest and disease management.

Background Chronic hepatitis B (CHB) is a significant global health challenge, with tenofovir disoproxil fumarate (TDF) widely used as an effective treatment option. Despite TDF’s efficacy in suppressing hepatitis B virus (HBV) DNA, it rarely achieves functional cure, requiring hepatitis B surface antigen (HBsAg) clearance or seroconversion, which are an optimal goal of CHB treatment. This study aimed to evaluate the long-term effects of TDF monotherapy on HBsAg clearance rates through a systematic review and meta-analysis, combined with a longitudinal modeling analysis to investigate HBsAg dynamics. Methods Eligible studies published between January 1st, 2008, and September 28th, 2023, in PubMed, EMBASE, and Web of Science were included in the systematic review and meta-analysis. The longitudinal model was developed based on data from 123 subjects in a Phase III trial cohort. Results Twenty-three studies were selected for meta-analysis. The summarized HBsAg clearance rate was near zero and unlikely to increase with extended treatment. The longitudinal model of HBsAg dynamic in CHB patients receiving TDF monotherapy showed a good fitting performance and extrapolation predictive ability. Model-based simulation confirmed that HBsAg clearance remained unlikely with prolonged therapy, with median HBsAg levels reducing by 21% after 168 weeks. Conclusions The consistency between meta-analysis and model simulation outcomes indicated that TDF monotherapy can achieve a limited reduction in HBsAg levels but did not result in functional cure, which reinforced the limited role of TDF monotherapy in comprehensive CHB management. Graphical abstract

With the recent advances in machine learning (ML) and artificial intelligence (AI), data‐driven modeling approaches for pharmacokinetics (PK) and pharmacodynamics (PD) have gained popularity due to their versatility in diverse settings and reduced reliance on prior assumptions. However, most of the ML methods ignore the hidden dynamics behind the data, lacking interpretability. This study investigated the applicability of neural controlled differential equation (NCDE), a novel ML method that is suitable for data‐driven modeling of PK and PD profiles, especially in the setting of multiple dosing. We demonstrated that NCDE was capable of combining differential‐equation‐based dynamics with data‐driven characteristics, flexibly incorporating various types of inputs, and embedding discontinuous dynamics. Moreover, a direct correspondence was identified between the learned dynamics of NCDE and the dynamics behind the data, which highlights the intrinsic interpretability of NCDE. Additionally, the influence of important hyperparameters was systematically investigated, and it was found that L1 regularization and the AdaMax optimizer were useful for stabilizing the training process and leading to a generalizable NCDE model. Together, these findings demonstrate the accuracy, generalizability, and interpretability of NCDE, indicating that NCDE is a reliable method for further application. In the future, NCDE may further facilitate PK and PD prediction in general. Neural controlled differential equations (NCDE), driven by control variables, are capable to learn the discontinuous dynamics in the PK and PD datasets.

PurposeThe aims of this longitudinal model-based meta-analysis (MBMA) were to indirectly compare the time courses of survival probabilities and to identify corresponding potential significant covariates across approved drugs in patients with castration-resistant prostate cancer (CRPC).MethodsA systematic literature review for monotherapy studies in patients with CRPC was conducted up to August 8, 2018. The time courses of progression-free survival (PFS) and overall survival (OS) were fitted with parametric survival models. Covariate analyses were performed to determine the impact of treatment drugs, dosing regimens, and patient characteristics on the survival probabilities. Simulations were carried out to quantify the magnitude of covariate effects.ResultsA total of 146 studies including clinical trials and real-world data on longitudinal survival probabilities in 20,712 patients with CRPC were included in our meta-database. The time courses of PFS and OS probabilities were best described by the log-logistic model. There was no significant difference in median OS and PFS between docetaxel, cabazitaxel, abiraterone acetate, and enzalutamide. There was no significant dose-response relationship in PFS or OS for docetaxel at 50 to 120 mg/m2 every 3 weeks (Q3W) and cabazitaxel at 20 to 25 mg/m2 Q3W. Model-based simulations indicated that PFS probability was associated with chemotherapy, Gleason score, and baseline prostate-specific antigen (BLPSA), while OS probability was associated with chemotherapy, Gleason score, visceral metastasis, Eastern Cooperative Oncology Group performance status, and BLPSA.ConclusionOur modeling and simulation framework can be applied to support indirect comparison, dose selection, and go/no-go decision-making for new agents targeting CRPC.

PurposeOverall survival (OS) has traditionally been the primary endpoint to evaluate drug efficiency in oncology but is often limited by the long observation period and high cost. The aims of this study were to perform a comprehensive meta-analysis at a clinical trial level to investigate the potential surrogate endpoints of OS in patients with castration-resistant prostate cancer (CRPC), and to predict OS based on the relationships associated with the potential surrogate endpoints.MethodsA systematic literature search was conducted in the PubMed database up to August 2018. Correlations between OS and potential surrogate endpoints were determined by linear regression analysis weighted by the square roots of sample size. Simulations were conducted to assess the effect of covariates on the relationships between OS and surrogate endpoints.ResultsA total of 233 studies including clinical trials and real-world data were included in our dataset. The correlations between median OS and potential surrogate endpoints for androgen-targeting therapy (R2 = 0.58–0.92) were generally stronger than those for taxane chemotherapy (R2 = 0.37–0.71). Median radiographic progression-free survival (rPFS) showed the strongest correlations with median OS (R2 = 0.94) in patients treated with novel androgen-targeting therapy.ConclusionThe meta-analysis demonstrated that rPFS might serve as a potential surrogate endpoint of OS and offer opportunity to facilitate the interim analyses and decision-making during the early stage of clinical trials for androgen-targeting agents.

Hemoporfin, a porphyrin derivative photosensitizer, has been approved for the treatment of port‐wine stain (PWS) in adults. However, its optimal dose for the pediatric population remains unclear. This study aimed to explore appropriate dosing for pediatric patients with PWS through population pharmacokinetics (PopPK) and exposure‐response (ER) analysis. Data from a prospective pilot study of hemoporfin photodynamic therapy in pediatric PWS patients, as well as a phase I study in healthy adult volunteers, were utilized for the analysis. The pharmacokinetics of hemoporfin in the pediatric population can be described by a three‐compartment model with linear elimination following allometric scaling rules. Simulations indicated that simply scaling down the approved adult dose of 5 mg/kg based on weight for the pediatric population, which is a common practice among clinicians, may lead to reduced drug exposure in pediatric patients. Mean Cmax and AUC0–30min in pediatric patients were 18.7% and 30.5% lower than those in adults, respectively. A positive relationship was identified between AUC0–30min and the probability of investigators or patients giving high ratings for efficacy, suggesting that improved efficacy may be achieved with higher hemoporfin exposure. A series of dosing regimens were explored to match exposure in the pediatric population to that of the adult population. These findings may accelerate the development of pediatric indications for hemoporfin and help address the unmet medical needs of pediatric patients with PWS.

Pegbing (peginterferon alpha‐2b) is a polyethylene glycol‐modified interferon α‐2b injection that has demonstrated favorable efficacy and safety profiles in the treatment of chronic hepatitis B (CHB). This study aimed to develop a population pharmacokinetic (PopPK) model of Pegbing in both healthy subjects and CHB patients and to investigate the influence of covariates on its pharmacokinetic behavior. Pharmacokinetic data were obtained from a Phase I trial in healthy volunteers and a Phase II trial in CHB patients. A one‐compartment model with a target‐mediated drug disposition (TMDD) component incorporating IFN receptor downregulation was established to describe the pooled data from 28 healthy subjects and 39 CHB patients. Physiologically reasonable parameters were estimated, providing a good description and prediction of the model. Furthermore, the final PopPK model was externally validated using an independent dataset of 115 CHB patients. In the covariate analysis, health status (healthy v.s. CHB) was a significant covariate, affecting the Pegbing absorption rate, creatinine clearance was associated with clearance, and body weight affected the volume of distribution. Compared with healthy subjects, CHB patients exhibited a consistent area under the curve (AUC) but a higher Cmax. A PopPK model of Pegbing in both healthy volunteers and CHB patients was successfully established. Based on the model simulation, covariate‐based dose adjustment is unnecessary for CHB patients with normal renal function.

Immune checkpoint inhibitors (ICIs) have become a vital part of the therapeutic landscape for non‐small cell lung cancer (NSCLC) in recent years benefiting from their remarkable efficacy. However, ICIs are associated with potentially life‐threatening immune‐related adverse events (irAEs). This study aims to quantify dose dependence and additional influencing factors of both any grade and grade greater than or equal to 3 irAEs in patients with NSCLC treated by ICIs. The trial‐level irAE data was collected and pooled from 129 cohorts in 81 clinical studies. A logit‐transformed meta‐regression model was applied to derive the quantitative relationship of irAE rate and ICI exposure. Programmed cell death‐1 (PD‐1) or programmed cell death ligand‐1 (PD‐L1) inhibitors showed no dose dependence in patients with NSCLC, whereas cytotoxic T lymphocyte–associated antigen 4 (CTLA‐4) inhibitors exhibited a statistically significant dose dependence when used alone or combined with PD‐1 or PD‐L1 inhibitors. Besides, therapy line and combination of ICIs with chemotherapy or target therapy were significant covariates. Hopefully, the results of this study can improve clinicians’ awareness of irAEs and be helpful for clinical decisions during ICI treatment for NSCLC.

Retroperitoneal sarcoma (RPS) is a rare malignancy which can be difficult to manage due to the variety of clinical behaviors. In this study, we aimed to develop a parametric modeling framework to quantify the relationship between postoperative dynamics of several biomarkers and overall/progression‐free survival of RPS. One hundred seventy‐four patients with RPS who received surgical resection with curative intent at the Peking University Cancer Hospital Sarcoma Center were retrospectively included. Potential prognostic factors were preliminarily identified. Longitudinal analyses of body mass index (BMI), serum total protein (TP), and white blood cells (WBCs) were performed using nonlinear mixed effects models. The impacts of time‐varying and time‐invariant predictors on survival were investigated by parametric time‐to‐event (TTE) models. The majority of patients experienced decline in BMI, recovery of TP, as well as transient elevation in WBC counts after surgery, which significantly correlated with survival. An indirect‐response model incorporating surgery effect described the fluctuation in percentage BMI. The recovery of TP was captured by a modified Gompertz model, and a semimechanistic model was selected for WBCs. TTE models estimated that the daily cumulative average of predicted BMI and WBC, the seventh‐day TP, as well as certain baseline variables, were significant predictors of survival. Model‐based simulations were performed to examine the clinical significance of prognostic factors. The current work quantified the individual trajectories of prognostic biomarkers in response to surgery and predicted clinical outcomes, which would constitute an additional strategy for disease monitoring and intervention in postoperative RPS.