Objective:To develop a machine learning prediction model based on peripheral blood indicators for assessing 30-day mortality risk in elderly patients diagnosed with hemophagocytic lymphohistiocytosis (HLH).Methods:A retrospective cohort study was conducted, enrolling elderly patients (age≥65 years) diagnosed HLH at the First Affiliated Hospital of Nanjing Medical University between January 1, 2015, and November 30, 2023. Demographic characteristics, clinical manifestations, and laboratory parameters at admission were collected. The study included 204 elderly HLH patients with a median age of 70 (68-75) years, comprising 134 males (65.69%) and 70 females (34.31%). Using computer-generated random numbers, the data was randomly divided into the training and validation cohorts at a 7∶3 ratio. Based on 30-day survival outcomes, patients in the training cohort were categorized into the death and survivor groups. Predictive variables were screened through univariate analysis and the Boruta algorithm, with prediction models constructed using 11 machine learning algorithms. Model performance was evaluated using the following metrics: area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, positive predictive value, negative predictive value, F1-score, calibration curve, and decision curve analysis. SHAP analysis was employed for model interpretation.Results:Comparison between the death and survivor groups in the training cohort identified 25 significant indicators ( P<0.05) through univariate analysis. Boruta algorithm-based screening further identified nine predictive variables: urea, ferritin, creatinine (CREA), D-dimer (D-D), platelet (PLT), activated partial thromboplastin time (APTT), aspartate aminotransferase (AST), creatine kinase (CK), and alanine aminotransferase (ALT). Among the 11 algorithms, the top five models by AUC in the training cohort were: XGBoost(AUC=1.000), AdaBoost(AUC=1.000), GBDT(AUC=1.000), DT(AUC=0.967), and RF(AUC=0.945). In the validation cohort, the top five performers by AUC were: RF(AUC=0.812), LR(AUC=0.792), LightGBM(AUC=0.769), AdaBoost(AUC=0.746), and GBDT(AUC=0.742). Thus, the RF model demonstrated optimal performance. SHAP analysis indicated urea as the most significant contributor to prediction outcomes. Conclusion:A machine learning model based on routine laboratory indicators can accurately predict the 30-day mortality risk in elderly HLH patients.

Objective:To develop a machine learning prediction model based on peripheral blood indicators for assessing 30-day mortality risk in elderly patients diagnosed with hemophagocytic lymphohistiocytosis (HLH).Methods:A retrospective cohort study was conducted, enrolling elderly patients (age≥65 years) diagnosed HLH at the First Affiliated Hospital of Nanjing Medical University between January 1, 2015, and November 30, 2023. Demographic characteristics, clinical manifestations, and laboratory parameters at admission were collected. The study included 204 elderly HLH patients with a median age of 70 (68-75) years, comprising 134 males (65.69%) and 70 females (34.31%). Using computer-generated random numbers, the data was randomly divided into the training and validation cohorts at a 7∶3 ratio. Based on 30-day survival outcomes, patients in the training cohort were categorized into the death and survivor groups. Predictive variables were screened through univariate analysis and the Boruta algorithm, with prediction models constructed using 11 machine learning algorithms. Model performance was evaluated using the following metrics: area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, positive predictive value, negative predictive value, F1-score, calibration curve, and decision curve analysis. SHAP analysis was employed for model interpretation.Results:Comparison between the death and survivor groups in the training cohort identified 25 significant indicators ( P<0.05) through univariate analysis. Boruta algorithm-based screening further identified nine predictive variables: urea, ferritin, creatinine (CREA), D-dimer (D-D), platelet (PLT), activated partial thromboplastin time (APTT), aspartate aminotransferase (AST), creatine kinase (CK), and alanine aminotransferase (ALT). Among the 11 algorithms, the top five models by AUC in the training cohort were: XGBoost(AUC=1.000), AdaBoost(AUC=1.000), GBDT(AUC=1.000), DT(AUC=0.967), and RF(AUC=0.945). In the validation cohort, the top five performers by AUC were: RF(AUC=0.812), LR(AUC=0.792), LightGBM(AUC=0.769), AdaBoost(AUC=0.746), and GBDT(AUC=0.742). Thus, the RF model demonstrated optimal performance. SHAP analysis indicated urea as the most significant contributor to prediction outcomes. Conclusion:A machine learning model based on routine laboratory indicators can accurately predict the 30-day mortality risk in elderly HLH patients.

Objective:To assess the prognostic value of methylation of interferon regulatory factor 6 ( IRF6) gene promoter in patients diagnosed with Kidney renal clear cell carcinoma (KIRC). Methods:The primary lesions of fifty KIRC patients who were diagnosed at the First Affiliated Hospital of Nanjing Medical University from January 2016 to January 2020 were collected. The expression of IRF6 protein was determined with an immunohistochemical method. The correlation between the level of IRF6 expression and survival and/or metastasis status was analyzed. The mRNA and protein levels of the IRF6 in KIRC and normal renal tissues were compared by using bioinformatic tools. The difference in the methylation rate of the IRF6 gene promoter between tumor and adjacent tissues was analyzed by searching the online databases. Statistical analysis was carried out for the methylation status of the IRF6 gene promoter region to select those negatively correlated with the overall survival (OS) among the patients. In vitro experiments were conducted with cell lines to verify the correlation between the status of promoter methylation and transcription level of the IRF6 gene. Results:The mRNA and protein levels of the IRF6 gene in KIRC tissues were significantly lower than those of the normal controls, and this was more prominent in patients who had died or developed metastasis. The extent of IRF6 gene promoter methylation in the KIRC tissues was much higher compared with that of the adjacent normal renal tissues. There was a significant negative correlation between the methylation of the IRF6 gene promoter and mRNA level of the IRF6 ( R= -0.52). The higher methylation degree in the IRF6 gene promoter regions cg12034118 and cg16030177, the shorter the OS and worse prognosis in the patients. Only twenty CpG sites in cg12034118 were confirmed to be highly methylated in KIRC cell lines. The transcription level of the IRF6 gene was upregulated in a time- and dose-dependent manner after the treatment with demethylation reagent 5-azadeoxycytidine. Conclusion:The methylation of IRF6 gene promoter in the renal tissues of KIRC patients is closely correlated with the OS. Cg12034118 may provide a promising biomarker for laboratory detection, and its high methylation rate has certain reference value for the prognosis.

[This corrects the article DOI: 10.1016/j.apsb.2021.07.006.].

Given the opposing effects of Akt and AMP-activated protein kinase (AMPK) on metabolic homeostasis, this study examined the effects of deletion of Akt2 and AMPKα2 on fat diet-induced hepatic steatosis. Akt2–Ampkα2 double knockout (DKO) mice were placed on high fat diet for 5 months. Glucose metabolism, energy homeostasis, cardiac function, lipid accumulation, and hepatic steatosis were examined. DKO mice were lean without anthropometric defects. High fat intake led to adiposity and decreased respiratory exchange ratio (RER) in wild-type (WT) mice, which were ablated in DKO but not Akt2−/− and Ampkα2−/− mice. High fat intake increased blood and hepatic triglycerides and cholesterol, promoted hepatic steatosis and injury in WT mice. These effects were eliminated in DKO but not Akt2−/− and Ampkα2−/− mice. Fat diet promoted fat accumulation, and enlarged adipocyte size, the effect was negated in DKO mice. Fat intake elevated fatty acid synthase (FAS), carbohydrate-responsive element-binding protein (CHREBP), sterol regulatory element-binding protein 1 (SREBP1), peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α), peroxisome proliferator-activated receptor-α (PPARα), PPARγ, stearoyl-CoA desaturase 1 (SCD-1), phosphoenolpyruvate carboxykinase (PEPCK), glucose 6-phosphatase (G6Pase), and diglyceride O-acyltransferase 1 (DGAT1), the effect was absent in DKO but not Akt2−/− and Ampkα2−/− mice. Fat diet dampened mitophagy, promoted inflammation and phosphorylation of forkhead box protein O1 (FoxO1) and AMPKα1 (Ser485), the effects were eradicated by DKO. Deletion of Parkin effectively nullified DKO-induced metabolic benefits against high fat intake. Liver samples from obese humans displayed lowered microtubule-associated proteins 1A/1B light chain 3B (LC3B), Pink1, Parkin, as well as enhanced phosphorylation of Akt, AMPK (Ser485), and FoxO1, which were consolidated by RNA sequencing (RNAseq) and mass spectrometry analyses from rodent and human livers. These data suggest that concurrent deletion of Akt2 and AMPKα2 offers resilience to fat diet-induced obesity and hepatic steatosis, possibly through preservation of Parkin-mediated mitophagy and lipid metabolism.

Objective:To Clone the silencer sequences of human stimulator of interferon genes (STING) and evaluate its activity in HEK293T and HeLa, and to preliminarily investigate the transcriptional regulatory mechanisms, screening and verifying the possible binding elements for the silencer sequences of STING.Methods:The human STING-5-1a(-124~+ 267, transcription start site, TSS: 0) and STING-5-2a(-124~+ 168) regions were amplified by PCR, subcloned into pGL3-Basic plasmid, and the luciferase activity was detected in HEK293T and HeLa; the human STING silencer region STING-5-1b(+ 169~+ 267) was subcloned into the pGL3-Control plasmid (pGL3-C-5-1b-positive/negative) and STING-5-1b is divided into two complementary elements, subcloned into pGL3-Control vector, named pGL3-C-STING 5-1b-α(+ 169~+ 209) and pGL3-C-STING-5-1b-β(+ 210~+ 267), detecting the relative luciferase activity of the above plasmid in HEK293T and HeLa. Using bioinformatics methods to predict the transcription factor binding site of human STING silencer, make site-directed mutagenesis at the predicted binding site, and detect luciferase activity. Knockdown of THAP1 and TEAD1 by a siRNA strategy and detect the expression level of STING by western blot analysis. Chromosome immunoprecipitation assay further verified the binding of transcription factor to hSTING silencer.Results:It was verified that plasmids mentioned above were constructed correctly by nucleotide sequencing. The relative luciferase increased after truncating the STING-5-1b(+ 169~+ 267) fragment. The relative luciferase activity of pGL3-C-5-1b-positive recombinant plasmid decreased ( P<0.05), compared with pGL3-Control plasmid. Among them, STING-5-1b-β(+ 210~+ 267) fragment play a major inhibitory role. Using bioinformatics software to predict that the transcription factors PR domain zinc finger protein 4 (PRDM4), Thanatos-associated protein 1 (THAP1), TEA Domain Transcription Factor 1 (TEAD1), Nuclear receptor subfamily 4 group A member 1 (NR4A1), Krueppel-like factor 4 (KLF4) and Forkhead box protein O3(FOXO3) may bind to the human STING silencer region (+ 210~+ 267). After transfecting the mutant recombinant plasmid of the transcription factors into HEK293T and HeLa, the relative luciferase activity of THAP1-Mut and TEAD1-Mut were significantly increased, suggesting that STING silencers may contain binding sites of THAP1 and TEAD1. Knockdown of THAP1 and TEAD1 by a siRNA strategy significantly enhanced the transcription activity. Chromosome immunoprecipitation assay showed that the transcription factors TEAD1 and THAP1 combined with hSTING silencer region in the cells. Conclusions:The hSTING silencer luciferase reporter plasmid was successfully constructed. By the activity comparison, it is speculated that the core silencer region of human STING is located in the + 210~+ 267 element, which may contain several potential transcription factor binding sequences.The potential binding sites for transcription factors that may be contained in the DNA, and use Western blot and chromosome immunoprecipitation assays to further confirm the combination of transcription factors TEAD1 and THAP1 with hSTING silencer, laying the foundation for subsequent research.

Objective To assess the effects of axial spinous process-muscle-vascellum complex transplantation for posterior atlantoarial fusion.Methods Data of 27 cases with altantoarial disease who were treated by posterior atlantoarial fusion using axial spinous process-muscle-vascellum complex transplantation from June 2015 to June 2016 were retrospectively analyzed.There were 19 males and 8 females aged from 9 to 68 years old (mean,41.0±15.4 years old).Two cases were diagnosed with atlanto-axial instability.Fourteen cases were diagnosed with atlas fracture and eleven cases were diagnosed atlanto-axial fracture.All the 27 patients suffered from neck pain or limitations of cervical motion.All patients were assessed clinically by atlantoaxial reduction and bone graft fusion.The pre-operative and post-operative atlanto-dens interval (ADI),visual analogue scale (VAS),Japanese Orthopaedic Association scores (JOA),improvement rate of JOA score and axial symptoms were measured and statistically analyzed.Complications were recorded.Clinical outcome of latest follow-up was evaluated by X-ray and CT scan.Results The time of operation was 2.0-2.5 h and blood loss was 150-300 ml.All the patients were followed-up for 9 to 18 months (mean,11.5±2.1 months).The VAS of neck pain improved from 3.6±2.7 (range,2.0-5.0) pre-operatively to 1.4±0.2 (range,0-2.0) 12 months postoperatively (P=0.000).The JOA score improved from 11.7± 1.9 (range,10.0-15.0) pre-operatively to 15.3±0.6 (range,14.0-17.0)12 months post-operatively (P=0.000).The improvement rate of JOA score at the latest follow-up was 54.1％± 12.4％,including 23 cases (85.19％) excellent,and 4 cases (14.81％) good.The results of axial symptoms were no-symptom in 22 cases (81.48％) and mild symptoms in 5 cases (18.52％).Postoperative cervical spine X-ray and CT showed that the sagittal cervical spine alignment was restored.There was statistically significant difference between ADI of 4.3±1.1 mm (range,3.9-4.5 mm) pre-operatively to 2.5± 0.4 mm (range,2.1-2.6 mm) 12 months post-operatively,which was improved significantly (P=0.000).There were no complications found during the follow-up.Conclusion The application of axial spinous process-muscle-vascellum complex transplantation for posterior atlantoaxial fixation can preserve the dynamic function of muscles and reduce the postoperative pain,as well as avoid donor site morbidity.

Objective To explore the biomechanical properties of posterior occipital condyle screws compared with common occipitocervical fusion internal fixation and it's impacts upon stress of hypoglossal canals.Methods Finite element models based on the occipitocervical CT data of one 28-year-old male healthy volunteer were built,including normal model,instability model,internal fixation model by occipital condyle screws,internal fixation model by occipital plate screws,and internal fixation model by transarticular screws.Fifty N gravity and 1.5 N · m torque were exerted upon the surface of occipital bone so that the models could perform lateral bending,flexion,extension,and rotational motions.The motion range and stress distribution of internal fixation were compared under varying conditions among different occipitocervical fusion models.In addition,the impact of occipital condyle screw upon hypoglossal canals was examined.Results Compared with instability model,the motion range in the internal fixation model by occipital condyle screws declined by 96.8％,95.6％,95.0％ and 98.5％ respectively in lateral bending,flexion,extension and rotation.In the internal fixation by occipital plate screws,the motion range decreased by 96.3％,95.7％,98.4％ and 99.6％ respectively.In the internal fixation by transarticular screws,the motion range exhibited a decline of 95.7％,94.0％,94.3％ and 98.9％,respectively.The stress peaks in the occipital condyle screw were 192.4 MPa,201.6 MPa and 187.6 MPa under lateral bending,flexion,and rotation conditions,respectively.The stress peaks in the occipital plate screw were 279.6 MPa,213.7 MPa,and 154.1 MPa,respectively.The stress peaks in the transarticular screw were 232.4 MPa,220.9 MPa,and 224.5 MPa,respectively.The stress impact peak of occipital condyle screw on the hypoglossal canals wall was 12.96 MPa,and the content deformationunder the hypoglossal canal was 0.64％.Conclusions The occipital condyle screw internal fixation has similar stability with common occipitocervical fusion fixations.The occipital condyle screw has more uniform stress distribution and less effect on the hypoglossal canals,and hence is safe and reliable as anchor point on the cranial side in occipitocervical fusion.

BACKGROUND: Circulating levels of cell-free DNA increase in many pathologic conditions. However, notable discrepancies in the quantitative analysis of cell-free DNA from a large number of laboratories have become a considerable pitfall, hampering its clinical application. METHODS: We designed a novel recombinant DNA fragment that could be applied as an internal standard in a newly developed and validated duplex real-time PCR assay for the quantitative analysis of total cell-free plasma DNA, which was tested in 5,442 healthy adults and 200 trauma patients. RESULTS: Compared with two traditional methods, this novel assay showed a lower detection limit of 0.1 ng/mL, lower intra- and inter-assay CVs, and higher accuracy in the recovery test. The median plasma DNA concentration of healthy males (20.3 ng/mL, n=3,092) was significantly higher than that of healthy females (16.1 ng/mL, n=2,350) (Mann-Whitney two-sample rank sum test, P<0.0001). The reference intervals of plasma DNA concentration were 0-45.8 ng/mL and 0-52.5 ng/mL for healthy females and males, respectively. The plasma DNA concentrations of the majority of trauma patients (96%) were higher than the upper normal cutoff values and were closely related to the corresponding injury severity scores (R²=0.916, P<0.0001). CONCLUSIONS: This duplex real-time PCR assay with a new internal standard could eliminate variation and allow for more sensitive, repeatable, accurate, and stable quantitative measurements of plasma DNA, showing promising application in clinical diagnosis.
Adult ; DNA/*blood/standards ; Female ; Healthy Volunteers ; Humans ; Male ; Middle Aged ; Real-Time Polymerase Chain Reaction/*methods/standards ; Reference Values ; Wounds and Injuries/blood

Objective To investigate the different diagnostic value of serum Chitinase 3-like 1 protein(CHI3L1)and Alpha-fe-toprotein(AFP)in diagnosing hepatocellular carcinoma (HCC).Methods One hundreds HCC patients confirmed by histopa-thology were recruited between December,2015 to April,2016 from the First Affiliated Hospital of Nanjing Medical Univer-sity.Simultaneously,100 patients with chronic hepatitis B and 59 healthy individuals,matched by sex and age with HCC pa-tients,were recruited as control groups.Serum CHI3L1and AFP were measured in different groups and the difference were analyzed by STATA 1 2.0 Statistical software.The ability of these two items in differentiating different group was analyzing by ROC curve using MedCal Ver 15.2.2 software.Results Serum CHI3L1 were significantly differences in the three groups using Kruskal-Wallis analysis (χ2=93.19,P=0.000),the differences were further compared in different two groups using Mann-Whitney analysis.The results showed that serum CHI3L1 in HCC group were significantly higher than in chronic hepatitis B and healthy control group (P=0.000,z=8.766,7.400).Serum AFP were significantly differences in the three groups using Kruskal-Wallis analysis (χ2=147.54,P=0.000),and the differences were further compared in different two groups using Mann-Whitney analysis.The results showed that serum AFP in HCC group were significantly higher than in chronic hepatitis B and healthy control group (P=0.000,z=10.938,9.033).The ROC curve analysis of serum CHI3L1 and AFP for differentiating HCC group from CHB group showed that CHI3L1 yield AUC of 0.859 (95% CI:0.803~0.904) with 85% sensitivity,79% specificity and 76.8 pg/ml cut-off value,AFP yield AUC of 0.948 (95% CI:0.904~0.974)with 85% sensitivity,98% specificity and 7.6 ng/ml cut-off value,in distinguishing HCC with CHB group,the power of AFP was superior to that of CHI3L1 (P=0.006).The ROC curve analysis of serum CHI3L1 and AFP for differentiating HCC group from healthy individuals group showed that CHI3L1 yield AUC of 0.852 (95% CI:0.787~0.903)with 85% sensi-tivity,76% specificity and 76.8 pg/ml cut-off value,AFP yield AUC of 0.929 (95% CI:0.878~0.964)with 84% sensitivi-ty,100% specificity and 7.8 ng/ml cut-off value,in distinguishing HCC with healthy individuals group,the power of AFP was also superior to that of CHI3L1 (P=0.045).Conclusion Serum CHI3L1 similar to AFP has much power ability to di-agnosis HCC,but AFP was superior to CHI3L1.