Fruquintinib Plus Sintilimab in Patients with Treatment-Naive and Previously Treated Advanced Renal Cell Carcinoma: Results from a Phase Ib/II Clinical Trial.

Background: Antiangiogenic inhibitors plus immune checkpoint inhibitors have synergistic antitumor activity and have improved treatment outcomes in patients with renal cell carcinoma (RCC).

Objective: We report the RCC cohort from a phase Ib/II study in Chinese patients evaluating the efficacy and safety of fruquintinib plus sintilimab in treating advanced clear cell RCC (ccRCC).

Patients And Methods: Eligible patients had pathologically confirmed advanced ccRCC.

A gender breakdown of unexpected benefits generated by work from home in STEM fields - A qualitative analysis of the WiMPBME Task Group survey.

Background: Working from home during the Covid-19 pandemic was perceived differently by men and women working in STEM fields. The aim of this paper is to highlight the unexpected benefits generated by working from home during the pandemic.

Methods: Qualitative methodology was used to analyze data, collected via survey.

Analysis of performance degradation of an active pretensioning seatbelt in a real-world vehicle braking test.

Objective: Previous research has established the effectiveness of active pretensioning seatbelts (APS), also termed motorized pretensioning seatbelts, in mitigating forward leaning and out-of-position displacement during pre-crash scenarios. In the Chinese market, APS trigger times are typically set later than those reported in the literature. This study investigates the real-world performance of APS systems with delayed trigger times under emergency braking conditions.

Spatially Resolved Tumor Ecosystems and Cell States in Gastric Adenocarcinoma Progression and Evolution.

Gastric cancer (GC) is a major cause of global cancer mortality with high levels of heterogeneity. To explore geospatial interactions in tumor ecosystems, we integrated 2,138 spatial transcriptomic regions-of-interest (ROIs) with 152,423 single-cell expression profiles across 226 GC samples from 121 patients. We observed pervasive expression-based intratumor heterogeneity, recapitulating tumor progression through spatially localized and functionally ordered subgroups associated with specific immune microenvironments, checkpoint profiles, and genetic drivers such as SOX9.

Avoiding catastrophic overfitting in fast adversarial training with adaptive similarity step size.

Adversarial training has become a primary method for enhancing the robustness of deep learning models. In recent years, fast adversarial training methods have gained widespread attention due to their lower computational cost. However, since fast adversarial training uses single-step adversarial attacks instead of multi-step attacks, the generated adversarial examples lack diversity, making models prone to catastrophic overfitting and loss of robustness.