AIM: To examine differences in etiology, risk factors, gender, and acute-phase manifestations of retinal artery occlusion(RAO)patients in different age groups, highlighting age-related heterogeneity.

METHODS:This retrospective comparative cohort study enrolled patients with RAO. Patients were divided into <50 y group and ≥50 y group. Etiological classification was based on modified TOAST criteria. Comparisons assessed the prevalence of risk factors, etiological distribution, sex ratio, and, in cases of acute central retinal artery occlusion(CRAO), fundus characteristics and visual outcomes.

RESULTS:This study included a total of 145 RAO patients,which were divided into two groups based on age: <50 y group(15 females and 13 males)with the mean age of 36.64±9.72 y, and ≥50 y group(38 females and 79 males)with the mean age of 63.50±9.54 y. Hypertension(46.4% vs 85.5%)and diabetes(3.6% vs 27.4%)were less prevalent in <50 y group(both P<0.001). The <50 y group was characterized predominantly by stroke of other determined etiology(SOE, 35.7% vs 2.6%, P<0.001), whereas the ≥50 y group was dominated by small artery occlusion(SAO, 35.9%)and large artery atherosclerosis(LAA, 31.6%), the incidence rates were significantly higher in the ≥50 y group than in the <50 y group(both P<0.001). Females were more frequently represented in <50 y group(53.6% vs 32.5%, P=0.039). Multivariable analysis demonstrated that LAA was independently linked to older age and SOE to younger age. Boxcarring occurred exclusively in ≥50 y patients\〖39.5%(17/43)vs 0, P=0.011\〗.

CONCLUSION:The RAO patients<50 y predominantly present with non-atherosclerotic causes and a higher proportion of females, whereas patients≥50 y are mainly affected by atherosclerosis driven etiologies and show signs of reduced vascular compensatory capacity. These findings indicate that age-related differences in RAO should be taken into account when determining etiologic evaluation and management strategies.

ObjectiveTo evaluate the clinical efficacy and safety of Xiaozhen Formula （消疹方） in preventing skin toxicity induced by epidermal growth factor receptor inhibitors （EGFRIs） in patients with EGFR-mutant non-small cell lung cancer （NSCLC）. MethodsA randomized， double-blind， placebo-controlled， multicenter clinical study was conducted. A total of 120 patients with EGFR-mutant NSCLC from seven centers were enrolled and randomly assigned to a treatment group （60 cases） or a control group （60 cases）. On the basis of EGFRI-targeted therapy， patients in the treatment group received Xiaozhen Formula granules， whereas those in the control group received Xiaozhen Formula placebo granules， both at 10 g twice daily for 4 consecutive weeks. The primary outcomes were the grading of skin toxicity evaluated using the National Cancer Institute Common Terminology Criteria for Adverse Events （NCI-CTCAE） version 5.0 and the incidence of rash. Secondary outcomes included the time to onset and time to resolution of the highest-grade rash， the European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire-Core 30 （EORTC QLQ-C30） scores， the Hospital Anxiety and Depression Scale （HADS） scores， and disease control rate， together with safety assessments. ResultsBased on the full analysis set （FAS）， the incidence of skin toxicity in the treatment group was 18.33% （11/60）， significantly lower than 45.00% （27/60） in the control group （P＜0.01）， with the highest severity of skin toxicity in both groups was grade 2. Among patients who developed rash， the resolution rate in the treatment group was 90.91% （10/11）， higher than that 33.33% （9/27） in the control group. The median time to resolution was 15 days （95%CI： 8-17 days） in the treatment group and it was not reached in the control group； the distribution of time to resolution differed significantly between groups （P＜0.001）. After treatment， the treatment group had higher EORTC QLQ-C30 functional scale and global health status/quality-of-life scores， and lower symptom scale， single-item scores， as well as HADS-A， and HADS-D scores than the control group （P＜0.05）. In the FAS analysis， the disease control rate （DCR） was 85.0% （51/60） in the treatment group， lower than 100.0% （60/60） in the control group （P=0.003）， whereas no significant between-group difference was observed in the per-protocol set （PPS） analysis （P=0.464）. The incidence of adverse events did not differ significantly between the two groups （P＞0.05）， and no definite safety signals related to the investigational drug were observed. ConclusionXiaozhen Formula can reduce the incidence of EGFRI-induced skin toxicity， and improve the outcome of rash regression， the patients' quality of life and psychological status in EGFR-mutant NSCLC patients.

Objective To construct a prognostic model of lung adenocarcinoma(LUAD)based on m5C modification-related genes and to explore its clinical value.Methods Based on the LUAD data in TCGA,GSE30219,GSE31210,and GSE50081 cohorts,prognosis-related m5C modification-related genes were screened,and the prognostic model was constructed by using univariate Cox,Lasso,and multivariate Cox regression analyses.Kaplan-Meier curve,ROC curve,and Cox regression were used to observe the robustness and prognostic performance of the model.The correlation between the prognostic model and clinico-pathologic features was further explored.Results A prognostic model consisting of eight m5C modifi-cation-related genes,including CDK1,CDKN1A,NOP2,RRM2,TCL6,TLR8,TRDMT1,and YTHDF2,was constructed.Risk score was an independent risk factor for the prognosis of patients with LUAD,and it is combined with age,T stage,and N stage to constitute a nomogram which can accurately predict the prognosis of patients.The infiltration of macrophages and CD4+/CD8+T cells was significantly reduced in high-risk patients.The risk score in LUAD tissues was significantly higher than that in normal tissues and was positively correlated with T stage and N stage.The risk score of smoking and EGFR wild-type patients was higher than that of non-smoking and EGFR-mutant patients.Conclusion The prognostic model constructed based on m5C modification-related genes has shown good accuracy and stability in predicting the prognosis of patients with LUAD,and it is closely related to clinical features,driver gene mutations,and immune infiltration,which can provide a potential basis for the treatment and prognostic assessment of LUAD.

Objective To observe the value of machine learning(ML)models based on ultrasound radiomics for preoperatively distinguishing atypical parathyroid tumor(APT)/parathyroid carcinoma(PC)and parathyroid adenoma(PA).Methods Totally 330 primary hyperparathyroidism patients who underwent surgical treatments were retrospectively enrolled and categorized into APT/PC group(n=78)and PA group(n=252)according to surgical pathology and clinical follow-up results,also divided into training set(n=231)and test set(n=99)at the ratio of 7∶3.Based on preoperative ultrasound,545 radiomics features were extracted,and recursive feature elimination(RFE),Kruskal-Wallis or analysis of variance methods were used to screen the features,respectively.Support vector machine(SVM),linear discriminant analysis(LDA),least absolute shrinkage and selection operator logistic regression(LRLASSO),also random forest(RF)and decision tree(DT)algorithms were adopted to construct ML models for differentiating APT/PC and PA,respectively.Then the models were trained in training set,their performance were verified in test set,and a 5-fold cross-validation was adopted to screen out the better combinations.Results Compared with Kruskal-Wallis and analysis of variance methods,the distinguishing efficacy of SVM,LDA,LRLASSO,RF and DT models constructed based on features screened out using RFE method in training set(area under the curve[AUC]=0.870,0.878,0.850,0.847,1.000)and test set(AUC=0.856,0.842,0.827,0.847 and 0.704)were all relatively higher.In test set,the AUC of SVM,LDA,LRLASSO and RF models constructed based on the features screened out using RFE method(included 25,23,17 and 23 features)were all higher than that of DT model(8 features)(all P＜0.001).No significant difference of AUC was found between SVM,LRLASSO or RF models and LDA model(all P＞0.05).The AUC of SVM and RF models were higher than that of LRLASSO model(both P＜0.05),while of SVM and RF models were not significantly different(P＞0.05),indicating that SVM,LDA and RF models were better ones.Conclusion SVM,LDA,LRLASSO,RF and DT models based on ultrasound radiomics could effectively distinguish APT/PC and PA preoperatively,among which SVM,LDA and RF models had better diagnostic efficacy.

Objective To construct a prognostic model of lung adenocarcinoma(LUAD)based on m5C modification-related genes and to explore its clinical value.Methods Based on the LUAD data in TCGA,GSE30219,GSE31210,and GSE50081 cohorts,prognosis-related m5C modification-related genes were screened,and the prognostic model was constructed by using univariate Cox,Lasso,and multivariate Cox regression analyses.Kaplan-Meier curve,ROC curve,and Cox regression were used to observe the robustness and prognostic performance of the model.The correlation between the prognostic model and clinico-pathologic features was further explored.Results A prognostic model consisting of eight m5C modifi-cation-related genes,including CDK1,CDKN1A,NOP2,RRM2,TCL6,TLR8,TRDMT1,and YTHDF2,was constructed.Risk score was an independent risk factor for the prognosis of patients with LUAD,and it is combined with age,T stage,and N stage to constitute a nomogram which can accurately predict the prognosis of patients.The infiltration of macrophages and CD4+/CD8+T cells was significantly reduced in high-risk patients.The risk score in LUAD tissues was significantly higher than that in normal tissues and was positively correlated with T stage and N stage.The risk score of smoking and EGFR wild-type patients was higher than that of non-smoking and EGFR-mutant patients.Conclusion The prognostic model constructed based on m5C modification-related genes has shown good accuracy and stability in predicting the prognosis of patients with LUAD,and it is closely related to clinical features,driver gene mutations,and immune infiltration,which can provide a potential basis for the treatment and prognostic assessment of LUAD.

Objective:To explore the lived experiences of family members of terminal ICU patients regarding their humanistic care needs and provide theoretical foundations for developing nursing care plans tailored to their needs.Methods:This study was a descriptive qualitative study. From April to December 2023, 16 family members of terminally ill ICU patients in Shanghai East Hospital, Tongji University were selected for semi-structured interviews using purposive sampling method, and the interview data were qualitatively analyzed using Colaizzi 7-step analysis.Results:The humanistic care needs of family members of terminally ill ICU patients can be categorized into five themes, namely, the need to know the condition at the first time; the need to participate in treatment and decision-making; the need to respect the wishes of terminally ill patients; the need for psychological care; and the need for social support.Conclusions:The humanistic care needs of family members of terminal ICU patients remain largely unmet. Nursing professionals should consider these needs and preferences and provide family members with professional guidance to help them establish positive coping mechanisms.

Objective:To study the impact of fluoride exposure through drinking water on the risk of hypertension among residents in Jishan County, Shanxi Province.Methods:From March to April 2023, a cluster sampling method was used to select permanent residents aged ≥18 years and residing for ≥10 years in 12 villages in drinking water-borne endemic fluorosis areas of Jishan County, Shanxi Province as the survey subjects. A questionnaire survey, physical examination, and morning urinary fluoride level testing were conducted. The least absolute shrinkage and selection operator (Lasso) regression were used to analyze the key influencing factors of hypertension. Restricted cubic spline was used to evaluate the linear relationship between urinary fluoride and hypertension. Logistic regression was used to analyze the impact of urinary fluoride on hypertension.Results:Finally, 2 453 survey subjects were included, aged (62 ± 10) years, including 1 565 patients (63.80%) with hypertension. There were significant differences in the distribution of age, gender, education level, annual household income, body mass index (BMI), and the level and distribution of urinary fluoride between hypertension group and normal blood pressure group ( P < 0.05). The Lasso regression results showed that age, education level, BMI, and urinary fluoride were the key influencing factors of hypertension, with coefficients of 1.04, - 0.12, 0.24 and 0.01, respectively. The results of the restricted cubic spline showed that there was a linear relationship between urinary fluoride and hypertension after adjusting for age, education level, and BMI ( Poverall = 0.018, Pnonlinear = 0.482). The logistic regression results showed that after adjusting for age, education level, and BMI, urinary fluoride > 4.68 mg/L was a risk factor for hypertension ( OR = 1.42, 95% CI: 1.10 - 1.84, P = 0.007). Conclusion:High urinary fluoride is a risk factor for hypertension in drinking water-borne endemic fluorosis areas of Jishan County, Shanxi Province.

Objective To explore the expression and diagnostic value of serum cardiac troponin Ⅰ (cTn Ⅰ ) and neuron-specific enolase protein (NSE) in patients with cryptogenic stroke caused by patent foramen ovale (PFO). Methods A total of 108 patients with PFO admitted to Dazhou Central Hospital from December 2020 to December 2022 were selected and divided into 32 patients with cryptogenic stroke group (stroke group) and 76 patients with non-cryptogenic stroke group (non-stroke group) according to whether the patients caused cryptogenic stroke. According to the National Institutes of Health Stroke Scale (NIHSS),they were divided into mild group (<4points,n=10),moderate group (4～15points,n=15) and severe group (>15points,n=7). levels of fibrinogen (FIB),C-reactive protein (CRP),and alanine aminotransferase (ALT) were measured by an automatic biochemical analyzer in all patients. The serum levels of cTn Ⅰ and NSE were detected by ELISA-diagnostic value of serum cTn Ⅰ and NSE in PFO-induced cryptogenic stroke by ROC curve analysis. Logistic regression was applied to analyze the factors affecting the occurrence of cryptogenic stroke in PFO patients. Results Compared with the non-stroke group,the levels of FIB(3.20±0.36g/L vs 2.95±0.30g/L),CRP(73.58±7.43mg/L vs 62.52±7.11mg/L),ALT (68.58±6.82U/L vs 61.08±6.28U/L),cTn Ⅰ(0.78±0.10ng/L vs 0.58±0.08ng/L)and NSE(37.52±3.82μg/L vs 30.35±3.72μg/L)in the stroke group were increased,and the differences were statistically significant(t=3.722～10.994,all P<0.05). Compared with the mild group,the levels of cTn Ⅰ and NSE in the serum of patients in the moderate and severe groups were increased(t=5.891,7.177;3.458,4.105),the levels of cTn Ⅰ,NSE in the serum of the severe group were significantly higher than those in the moderate group (t=2.474,1.336),the differences were statistically significant (all P<0.05).The AUC (95％ CI) for the combined diagnosis of cryptogenic stroke in PFO patients was[0.981 (0.962～1.000)],which was higher than cTn Ⅰ[0.878(0.864～0.948)],NSE[0.911(0.850～0.972)]diagnosed alone (Z=8.621,6.477,all P<0.001). Logistic regression analysis showed that cTn Ⅰ (OR=2.631,95％CI:1.508～4.591),and NSE(OR=2.841,95％CI:1.514～5.330)were the influencing factors for the occurrence of cryptogenic stroke in PFO patients (P<0.05). Conclusion The serum levels of cTn Ⅰ and NSE in patients with cryptogenic stroke caused by PFO patients are significantly increased,which has certain diagnostic value for patients with cryptogenic stroke.

Non-alcoholic fatty liver disease(NAFLD)is a metabolic disease characterized by abnormal deposition of lipid in hepatocytes.If not intervened in time,NAFLD may develop into liver fibrosis or liver cancer,and ultimately threatening life.NAFLD has complicated etiology and pathogenesis,and there are no effective therapeutic means and specific drugs.Currently,insulin sensitizers,lipid-lowering agents and hep-atoprotective agents are often used for clinical intervention,but these drugs have obvious side effects,and their effectiveness and safety need to be further confirmed.Adenosine monophosphate(AMP)-activated protein kinase(AMPK)plays a central role in maintaining energy homeostasis.Activated AMPK can enhance lipid degradation,alleviate insulin resistance(IR),suppress oxidative stress and inflammatory response,and regulate autophagy,thereby alleviating NAFLD.Natural herbal medicines have received extensive attention recently because of their regulatory effects on AMPK and low side effects.In this article,we reviewed the biologically active natural herbal medicines(such as natural herbal medicine formulas,extracts,polysaccharides,and monomers)that reported in recent years to treat NAFLD via regulating AMPK,which can serve as a foundation for subsequent development of candidate drugs for NAFLD.

Objective To observe the value of machine learning(ML)models based on ultrasound radiomics for preoperatively distinguishing atypical parathyroid tumor(APT)/parathyroid carcinoma(PC)and parathyroid adenoma(PA).Methods Totally 330 primary hyperparathyroidism patients who underwent surgical treatments were retrospectively enrolled and categorized into APT/PC group(n=78)and PA group(n=252)according to surgical pathology and clinical follow-up results,also divided into training set(n=231)and test set(n=99)at the ratio of 7∶3.Based on preoperative ultrasound,545 radiomics features were extracted,and recursive feature elimination(RFE),Kruskal-Wallis or analysis of variance methods were used to screen the features,respectively.Support vector machine(SVM),linear discriminant analysis(LDA),least absolute shrinkage and selection operator logistic regression(LRLASSO),also random forest(RF)and decision tree(DT)algorithms were adopted to construct ML models for differentiating APT/PC and PA,respectively.Then the models were trained in training set,their performance were verified in test set,and a 5-fold cross-validation was adopted to screen out the better combinations.Results Compared with Kruskal-Wallis and analysis of variance methods,the distinguishing efficacy of SVM,LDA,LRLASSO,RF and DT models constructed based on features screened out using RFE method in training set(area under the curve[AUC]=0.870,0.878,0.850,0.847,1.000)and test set(AUC=0.856,0.842,0.827,0.847 and 0.704)were all relatively higher.In test set,the AUC of SVM,LDA,LRLASSO and RF models constructed based on the features screened out using RFE method(included 25,23,17 and 23 features)were all higher than that of DT model(8 features)(all P＜0.001).No significant difference of AUC was found between SVM,LRLASSO or RF models and LDA model(all P＞0.05).The AUC of SVM and RF models were higher than that of LRLASSO model(both P＜0.05),while of SVM and RF models were not significantly different(P＞0.05),indicating that SVM,LDA and RF models were better ones.Conclusion SVM,LDA,LRLASSO,RF and DT models based on ultrasound radiomics could effectively distinguish APT/PC and PA preoperatively,among which SVM,LDA and RF models had better diagnostic efficacy.