Wenzhou Central Hospital

Background: The FAM (family with sequence similarity) gene family, implicated in various malignancies, remains underexplored in microsatellite-stable (MSS) colorectal cancer (CRC). MSS CRC represents the majority of CRC cases and lacks reliable prognostic biomarkers as well as effective immunotherapeutic strategies. Methods: Transcriptome profiles and matched clinical data from The Cancer Genome Atlas-COADREAD cohort were used as a training set, while GSE29623 and GSE39582 served as independent validation cohorts. A prognostic signature (FAM gene Family-associated expression scoring (FAMscore)) was constructed based on FAM family genes using Multivariate-Cox regression. Associations between FAMscore, survival outcomes, tumor immune microenvironment features, and predicted drug sensitivity were explored through computational analyses. Results: The FAMscore stratified MSS CRC patients into high- and low-risk groups with significantly different (P ≤ 0.05) overall survival (OS) in the training cohort and consistent trends in validation datasets. A high FAMscore was associated with an immunosuppressive tumor immune microenvironment, characterized by increased regulatory T cells and M2 macrophages. Exploratory analyses suggested differential predicted sensitivity to selected targeted agents. Conclusions: FAMscore is associated with prognosis and immune landscape features in MSS CRC. These findings are exploratory and hypothesis-generating, and further experimental and prospective validation is required.

<title>Abstract</title> Background Diagnosing peritoneal carcinomatosis (PC) remains challenging. This study aims to evaluate the value of ascites chromosomal instability (CIN) analysis in the diagnosis of PC. Methods We performed low-coverage whole-genome sequencing on 50 human ascites samples. CIN was profiled using ultrasensitive chromosomal aneuploidy detection (UCAD), and its diagnostic performance was validated against pathological standards. Results CIN analysis detected peritoneal metastases with 90.0% sensitivity and 82.0% overall accuracy. In malignant ascites, 72.7% were CIN-positive, showing recurrent alterations (e.g., 3q gain, 14q loss) that formed co-occurring gain/loss modules. Unsupervised clustering identified three CIN subtypes with distinct genomic instability patterns. Conclusion Ascites CIN analysis shows promise as a highly sensitive, minimally invasive diagnostic tool for PC. The identified CIN patterns and subtypes provide insights into peritoneal metastasis biology and hold promise for clinical stratification.

The purpose of the study is to evaluate the effectiveness of acupoint therapy in the comprehensive postoperative management of patients with liver cancer, aiming to reduce pain, decrease the need for opioid analgesics, and improve quality of life. The study was conducted in China at Wenzhou Central Hospital. A total of 156 patients who had undergone liver resection surgery for liver cancer were included. The patients were randomized into two groups: the experimental group ( n = 79), which received standard postoperative rehabilitation combined with acupoint therapy, and the control group ( n = 77), which received only standard postoperative rehabilitation. Patient assessment was conducted at three time points: before treatment initiation (T0), on the 4th day of therapy (T1), and 2 weeks after completion of the course (T2). Twelve nurses participated in the study and recorded sleep quality, appetite, need for assistance, behavioral distress signs, and frequency of PRN interventions. The results of this study confirm the effectiveness of acupoint therapy as an adjunctive method in the rehabilitation process following liver cancer surgery. The practical value lies in the potential for reducing the pharmacological burden, decreasing the frequency of opioid-related side effects, and accelerating patient recovery.

Detection and treatment of breast cancer is critical for human health. Molecular subtyping is pivotal for guiding personalized treatment and prognosis predictions. Conventional subtyping relies on biopsy-based molecular testing, which are invasive for patients and resource-intensive for clinical practice. Deep-learning approaches have been researched for non-invasive subtype categorization directly from screening mammograms. However, the mammograms are normally taken with craniocaudal (CC) and mediolateral oblique (MLO) views, where the unwanted radiation dose is a concern. The motivation of this study was to investigate the performance of deep learning-based networks on single view mammograms. Single-view mammography images (CC, MLO) were studied for four deep learning networks (ResNet-18, GoogLeNet, MambaOut, and ConvNeXt) on the molecular subtype tasks. Input images received typical resizing, normalization, augmentation, training, testing and five-fold cross-validation. Both CC and MLO views demonstrated decent classification performances across molecular subtypes and across networks. Overall, MLO view exhibited a moderate advantage-higher F1 in most comparisons and a slight AUC improvement in about half-yet effects differed by network. For HER2-enriched, MLO tended to perform exceptionally well (notably with ResNet-18 and MambaOut), whereas Triple-Negative results demonstrated architecture-dependent view effects, including MLO-specific recall decreases in ConvNeXt and GoogLeNet. For Luminal A/B, sensitivity patterns were diverse among architectures, with CC often competitive and sometimes superior. Collectively, these results suggest single-view feasibility: either CC or MLO alone contains discriminative signal for subtype prediction, providing a non-invasive adjunct when a paired view is unavailable or suboptimal for radiation dose concern, and potentially reducing acquisition and reading burden in routine workflows.

BackgroundLaparoscopic Roux-en-Y gastric bypass (LRYGB) is a complex bariatric surgical procedure that demands advanced technical skills. However, traditional training models (virtual simulators, cadavers, and porcine organs) are Limited by issues concerning realism, cost, and accessibility. Although 3D-printed models offer promising potential solutions, their application in LRYGB training remains underexplored.Methods3D-printed silicone models of stomach and small intestine were fabricated using patient-derived anatomical data. The physical properties of these models (tensile strength, conductivity, and rheology) were compared with those of human and porcine tissues. Face and content validity , as well as technical realism, were evaluated through expert questionnaires. Construct validity was assessed using the Modified Bariatric Objective Structured Assessment of Technical Skills (MBOSATS). Trainees participated in repeated simulation sessions to evaluate their learning curves.ResultsThe silicone models demonstrated tensile strengths comparable to those of human tissues. Additionally, they exhibited superior aesthetics compared to porcine models (P < 0.05). Experts highly rated the face and content validity of the models, with median scores ranging from 4 to 5 out of 5. MBOSATS scores effectively differentiated between expertise levels (experts: 188.2 ± 1.48 vs. Intermediates: 145.10 ± 11.45 vs. residents: 64.1 ± 4.19, P < 0.001). In terms of training effectiveness, trainees achieved proficiency after 4–5 sessions, and post-training performance of residents on porcine models matched that of intermediates (P > 0.05). The models cost $2,000, significantly less than cadaver or animal alternatives.ConclusionsThe 3D-printed silicone models exhibit realistic mechanical properties, process high validity, and have proven effective in expediting skill acquisition during LRYGB training. They serve as a cost-effective and ethical alternative to traditional training tools, offering particular advantages in resource-constrained settings.Trial registrationThis study was prospectively registered in the Chinese Clinical Trial Registry (ChiCTR) under registration number ChiCTR2300078726 on 15 December 2023. The trial protocol was approved prior to the enrollment of the first participant, ensuring adherence to ethical standards and transparency in reporting health care intervention outcomes.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12893-025-03141-y.

Psoriasis is a chronic inflammatory skin condition with significant systemic and psychosocial impacts. Emerging evidence suggests a potential link between gut microbiota and psoriasis, prompting growing interest in gut microbiota-targeted therapies (GMTT) as a potential novel treatment strategy. Aims: This study aims to conduct a systematic review and meta-analysis to evaluate the efficacy and safety of GMTT in patients with psoriasis. Methods: PubMed, Embase, Web of Science, Cochrane, CNKI and Wanfang were searched up to May 2024 for randomized controlled trials (RCTs) assessing GMTT for psoriasis. Outcomes included Psoriasis Area and Severity Index (PASI), Dermatology Life Quality Index (DLQI), C-reactive protein (CRP) and adverse events. The risk of bias of the included RCTs was independently assessed using the Cochrane RoB tool. A random-effects model was used to calculate pooled standardized mean differences (SMDs) and 95% confidence intervals (CIs). Sensitivity, subgroup analyses and publication bias analyses were performed to assess the stability of the results and the sources of heterogeneity. Analyses were conducted using Review Manager 5.4 and STATA 15.1. The GRADE approach was used to assess the certainty of evidence. Results: Eight RCTs involving 409 patients were included. GMTT was associated with a modest reduction in PASI scores (SMD = -0.52, 95% CI [-0.98, -0.06], I² = 79%, p = 0.03). Subgroup analyses showed greater efficacy with antibiotic intervention, interventions < 3 months, Asian populations and plaque psoriasis. No significant differences were observed in adverse events (RR = 0.99, 95% CI [0.50, 1.94], I² = 23%, p = 0.97), DLQI or CRP. Funnel plots and Egger’s test indicated no significant publication bias, but sensitivity analysis revealed some instability. Conclusions: GMTT may offer modest benefits for psoriasis symptoms, particularly in specific subgroups. However, larger RCTs are needed to confirm these findings due to current limitations such as small sample sizes and regional biases.