Association of Severe Vitamin D Deficiency with Hospitalization in the Previous Year in Hospitalized Exacerbated COPD Patients.

Purpose: Vitamin D deficiency (VDD, 25-hydroxyvitamin D < 20 ng/mL) has been reported associated with exacerbation of chronic obstructive pulmonary disease (COPD) but sometimes controversial. Research on severe vitamin D deficiency (SVDD, 25-hydroxyvitamin D < 10 ng/mL) in exacerbation of COPD is limited. Patients and Methods: We performed a retrospective observational study in 134 hospitalized exacerbated COPD patients. 25-hydroxyvitamin D was modeled as a continuous or dichotomized (cutoff value: 10 or 20 ng/mL) variable to evaluate the association of SVDD with hospitalization in the previous year. Receiver operator characteristic (ROC) analysis was performed to find the optimal cut-off value of 25-hydroxyvitamin D. Results: In total 23% of the patients had SVDD. SVDD was more prevalent in women, and SVDD group tended to have lower blood eosinophils counts. 25-hydroxyvitamin D level was significantly lower in patients who were hospitalized in the previous year (13.6 vs 16.7 ng/mL, P = 0.044), and the prevalence of SVDD was higher (38.0% vs 14.3%, P = 0.002). SVDD was independently associated with hospitalization in the previous year [odds ratio (OR) 4.34, 95% CI 1.61-11.72, P = 0.004] in hospitalized exacerbated COPD patients, whereas continuous 25-hydroxyvitamin D and VDD were not (P = 0.1, P = 0.9, separately). The ROC curve yielded an area under the curve of 0.60 (95% CI 0.50-0.71) with an optimal 25-hydroxyvitamin D cutoff of 10.4 ng/mL. Conclusion: SVDD probably showed a more stable association with hospitalization in the previous year in hospitalized exacerbated COPD patients. Reasons for lower eosinophil counts in SVDD group needed further exploration.

ABCG2 polymorphism and rivaroxaban pharmacokinetics in healthy individuals after a single dose.

Rivaroxaban is a direct factor Xa inhibitor. Its interindividual variability is large and may be connected to the occurrence of adverse drug reactions or drug inefficacy. Pharmacogenetics studies concentrating on the reasons underlying rivaroxaban's inadequate response could help explain the differences in treatment results and medication safety profiles. Against this background, this study evaluated whether polymorphisms in the gene encoding the ABCG2 transporter modify the pharmacokinetic characteristics of rivaroxaban. A total of 117 healthy volunteers participated in two bioequivalence experiments with a single oral dose of 20 mg rivaroxaban, with one group fasting and the other being fed. Ultra-high-performance liquid chromatography coupled with mass spectrometry was employed to determine the plasma concentrations of rivaroxaban, and the WinNonlin program was used to calculate the pharmacokinetics parameters. In the fasting group, the rivaroxaban pharmacokinetic parameters of Vd (508.27 vs 334.45 vs 275.59 L) and t1/2 (41.04 vs 16.43 vs 15.47 h) were significantly higher in ABCG2 421 A/A genotype carriers than in ABCG2 421 C/C and 421 C/A genotype carriers (P<0.05). The mean values of Cmax (145.81 vs 176.27 vs 190.19 ng/mL), AUC0-t (1193.81 vs 1374.69 vs 1570.77 ng/mL·h), and Cl (11.82 vs 14.50 vs 13.01 mL/h) for these groups were lower, but this difference was not statistically significant (P>0.05). These findings suggested that the ABCG2 421 A/A genotype may impact rivaroxaban parameters after a single dose in healthy subjects. This finding must be validated before it is applied in clinical practice.

Sample size calculation for comparing two ROC curves.

Biomarkers are key components of personalized medicine. In this paper, we consider biomarkers taking continuous values that are associated with disease status, called case and control. The performance of such a biomarker is evaluated by the area under the curve (AUC) of its receiver operating characteristic curve. Oftentimes, two biomarkers are collected from each subject to test if one has a larger AUC than the other. We propose a simple non-parametric statistical test for comparing the performance of two biomarkers. We also present a simple sample size calculation method for this test statistic. Our sample size formula requires specification of AUC values (or the standardized effect size of each biomarker between cases and controls together with the correlation coefficient between two biomarkers), prevalence of cases in the study population, type I error rate, and power. Through simulations, we show that the testing on two biomarkers controls type I error rate accurately and the proposed sample size closely maintains specified statistical power.

Nonparametric receiver operating characteristic curve analysis with an imperfect gold standard.

This article addresses the challenge of estimating receiver operating characteristic (ROC) curves and the areas under these curves (AUC) in the context of an imperfect gold standard, a common issue in diagnostic accuracy studies. We delve into the nonparametric identification and estimation of ROC curves and AUCs when the reference standard for disease status is prone to error. Our approach hinges on the known or estimable accuracy of this imperfect reference standard and the conditional independent assumption, under which we demonstrate the identifiability of ROC curves and propose a nonparametric estimation method. In cases where the accuracy of the imperfect reference standard remains unknown, we establish that while ROC curves are unidentifiable, the sign of the difference between two AUCs is identifiable. This insight leads us to develop a hypothesis-testing method for assessing the relative superiority of AUCs. Compared to the existing methods, the proposed methods are nonparametric so that they do not rely on the parametric model assumptions. In addition, they are applicable to both the ROC/AUC analysis of continuous biomarkers and the AUC analysis of ordinal biomarkers. Our theoretical results and simulation studies validate the proposed methods, which we further illustrate through application in two real-world diagnostic studies.

Development and external validation of machine learning-based models to predict patients with cellulitis developing sepsis during hospitalisation.

OBJECTIVE: Cellulitis is the most common cause of skin-related hospitalisations, and the mortality of patients with sepsis remains high. Some stratification models have been developed, but their performance in external validation has been unsatisfactory. This study was designed to develop and compare different models for predicting patients with cellulitis developing sepsis during hospitalisation. DESIGN: This is a retrospective cohort study. SETTING: This study included both the development and the external-validation phases from two independent large cohorts internationally. PARTICIPANTS AND METHODS: A total of 6695 patients with cellulitis in the Medical Information Mart for Intensive care (MIMIC)-IV database were used to develop models with different machine-learning algorithms. The best models were selected and then externally validated in 2506 patients with cellulitis from the YiduCloud database of our university. The performances and robustness of selected models were further compared in the external-validation group by area under the curve (AUC), diagnostic accuracy, sensitivity, specificity and diagnostic OR. PRIMARY OUTCOME MEASURES: The primary outcome of interest in this study was the development based on the Sepsis-3.0 criteria during hospitalisation. RESULTS: Patient characteristics were significantly different between the two groups. In internal validation, XGBoost was the best model, with an AUC of 0.780, and AdaBoost was the worst model, with an AUC of 0.585. In external validation, the AUC of the artificial neural network (ANN) model was the highest, 0.830, while the AUC of the logistic regression (LR) model was the lowest, 0.792. The AUC values changed less in the boosting and ANN models than in the LR model when variables were deleted. CONCLUSIONS: Boosting and neural network models performed slightly better than the LR model and were more robust in complex clinical situations. The results could provide a tool for clinicians to detect hospitalised patients with cellulitis developing sepsis early.

Phase 1 pharmacokinetic and safety study of soticlestat in participants with mild or moderate hepatic impairment or normal hepatic function.

This phase 1, open-label, three-arm study (NCT05098054) compared the pharmacokinetics and safety of soticlestat (TAK-935) in participants with hepatic impairment. Participants aged ≥18 to <75 years had moderate (Child-Pugh B) or mild (Child-Pugh A) hepatic impairment or normal hepatic function (matched to hepatic-impaired participants by sex, age, and body mass index). Soticlestat was administered as a single oral 300 mg dose. Pharmacokinetic parameters of soticlestat and its metabolites TAK-935-G (M3) and M-I were assessed and compared by group. The incidence of treatment-emergent adverse events (TEAEs) and other safety parameters were also monitored. The pharmacokinetic analyses comprised 35 participants. Participants with moderate hepatic impairment had lower proportions of bound and higher proportions of unbound soticlestat than participants with mild hepatic impairment and normal hepatic function. Total plasma soticlestat pharmacokinetic parameters (maximum observed concentration [Cmax], area under the concentration-time curve from time 0 to time of last quantifiable concentration [AUClast], and AUC from time 0 to infinity [AUC∞]) were approximately 115%, 216%, and 199% higher with moderate and approximately 45%, 35%, and 30% higher with mild hepatic impairment, respectively, than healthy matched participants. Moderate hepatic impairment decreased the liver's ability to metabolize soticlestat to M-I; glucuronidation to M3 was also affected. Mild hepatic impairment resulted in a lower total plasma M-I exposure, but glucuronidation was unaffected. TEAEs were similar across study arms, mild, and no new safety findings were observed. A soticlestat dose reduction is required for individuals with moderate but not mild hepatic impairment.

Bioequivalence study of low dose drospirenone/ethinyl estradiol 3 mg/0.03 mg film tablets under fasting conditions in Turkish healthy female subjects.

This bioequivalence research aims to evaluate the relative bioavailability and pharmacokinetic characteristics of ethinyl estradiol and drospirenone in the test preparation in comparison to the reference preparation during fasting conditions. A liquid chromatography method with tandem mass spectrometry was used to determine the concentrations of drospirenone and ethinyl estradiol in plasma. The pharmacokinetic parameters that were analyzed were the maximum plasma concentration (Cmax), time to achieve Cmax (tmax), elimination half life, and area under the concentration time curve of plasma (AUC0-t, AUC0-∞ for ethinyl estradiol, and AUC0-72h for drospirenone). Both the AUC and Cmax parameters were determined to be between 80.00% and 125.00% (90% confidence intervals), which is the acceptable range. Based on the study findings, it was concluded that the test formulation, which includes 3 mg of drospirenone and 0.03 mg of ethinyl estradiol, demonstrated bioequivalence when compared to the reference formulation.

A two-tier feature selection method for predicting mortality risk in ICU patients with acute kidney injury.

Acute kidney injury (AKI) is one of the most important lethal factors for patients admitted to intensive care units (ICUs), and timely high-risk prognostic assessment and intervention are essential to improving patient prognosis. In this study, a stacking model using the MIMIC-III dataset with a two-tier feature selection approach was developed to predict the risk of in-hospital mortality in ICU patients admitted for AKI. External validation was performed using separate MIMIC-IV and eICU-CRD. The area under the curve (AUC) was calculated using the stacking model, and features were selected using the Boruta and XGBoost feature selection methods. This study compares the performance of a stacking model using two-tier feature selection with a model using single-tier feature selection (XGBoost: 85; Boruta: 83; two-tier: 0.91). The predictive effectiveness of the stacking model was further validated by using different datasets (Validation 1: 0.83; Validation 2: 0.85) and comparing it with a simpler model and traditional clinical scores (SOFA: 0.65; APACH IV: 0.61). In addition, this study combined interpretable techniques and causal inference to analyze the causal relationship between features and predicted outcomes.

[Predictive effect of the dual-parametric MRI modified maximum diameter of the lesions with PI-RADS 4 and 5 on the clinically significant prostate cancer].

OBJECTIVE: To assess the rationality of the maximum lesion diameter of 15 mm in prostate imaging reporting and data system (PI-RADS) as a criterion for upgrading a lesion from category 4 to 5 and improve it to enhance the prediction of clinically significant prostate cancer (csPCa). METHODS: In this study, the patients who underwent prostate magnetic resonance imaging (MRI) and prostate biopsy at Peking University First Hospital from 2019 to 2022 as a development cohort, and the patients in 2023 as a validation cohort were reviewed. The localization and maximum diameter of the lesion were fully evaluated. The area under the curve (AUC) and the cut-off value of the maximum diameter of the lesion to predict the detection of csPCa were calculated from the receiver operating characteristics (ROC) curve. Confounding factors were reduced by propensity score matching (PSM). Diagnostic efficacy was compared in the validation cohort. RESULTS: Of the 589 patients in the development cohort, 358 (60.8%) lesions were located in the peripheral zone and 231 (39.2%) were located in the transition zone, and 496 (84.2%) patients detected csPCa. The median diameter of the lesions in the peripheral zone was smaller than that in the transition zone (14 mm vs. 19 mm, P < 0.001). In the ROC analysis of the maximal diameter on the csPCa prediction, there was no statistically significant difference between the peri-pheral zone (AUC=0.709) and the transition zone (AUC=0.673, P=0.585), and the cut-off values were calculated to be 11.5 mm for the peripheral zone and 16.5 mm for the migrating zone. By calcula-ting the Youden index for the cut-off values in the validation cohort, we found that the categorisation by lesion location led to better predictive results. Finally, the net reclassification index (NRI) was 0.170. CONCLUSION: 15 mm as a criterion for upgrading the PI-RADS score from 4 to 5 is reasonable but too general. The cut-off value for peripheral zone lesions is smaller than that in transitional zone. In the future consideration could be given to setting separate cut-off values for lesions in different locations.

Deep Learning-Based Vascular Aging Prediction From Retinal Fundus Images.

Purpose: The purpose of this study was to establish and validate a deep learning model to screen vascular aging using retinal fundus images. Although vascular aging is considered a novel cardiovascular risk factor, the assessment methods are currently limited and often only available in developed regions. Methods: We used 8865 retinal fundus images and clinical parameters of 4376 patients from two independent datasets for training a deep learning algorithm. The gold standard for vascular aging was defined as a pulse wave velocity ≥1400 cm/s. The probability of the presence of vascular aging was defined as deep learning retinal vascular aging score, the Reti-aging score. We compared the performance of the deep learning model and clinical parameters by calculating the area under the receiver operating characteristics curve (AUC). We recruited clinical specialists, including ophthalmologists and geriatricians, to assess vascular aging in patients using retinal fundus images, aiming to compare the diagnostic performance between deep learning models and clinical specialists. Finally, the potential of Reti-aging score for identifying new-onset hypertension (NH) and new-onset carotid artery plaque (NCP) in the subsequent three years was examined. Results: The Reti-aging score model achieved an AUC of 0.826 (95% confidence interval [CI] = 0.793-0.855) and 0.779 (95% CI = 0.765-0.794) in the internal and external dataset. It showed better performance in predicting vascular aging compared with the prediction with clinical parameters. The average accuracy of ophthalmologists (66.3%) was lower than that of the Reti-aging score model, whereas geriatricians were unable to make predictions based on retinal fundus images. The Reti-aging score was associated with the risk of NH and NCP (P < 0.05). Conclusions: The Reti-aging score model might serve as a novel method to predict vascular aging through analysis of retinal fundus images. Reti-aging score provides a novel indicator to predict new-onset cardiovascular diseases. Translational Relevance: Given the robust performance of our model, it provides a new and reliable method for screening vascular aging, especially in undeveloped areas.

Pharmacokinetics of Trazodone in Hispaniolan Parrots (Amazona ventralis).

The objective of this study was to establish the pharmacokinetics of a single oral dose of trazodone in the Hispaniolan Amazon parrot (Amazona ventralis). Trazodone is a selective serotonin antagonist and reuptake inhibitor used commonly in both human and veterinary medicine as an antidepressant behavioral modification medicine. A single oral dose of compounded trazodone hydrochloride solution (20 mg/mL) at 50 mg/kg was administered to a total of 7 healthy adult Hispaniolan Amazon parrots. The 7 healthy adult parrots ranged in age from 10 to 15 years and weighed 228 to 323g. Blood was collected at baseline (2 weeks before study) and at 1, 2, 4, 6, 10, and 14 hours post-drug administration. Plasma concentrations of both trazodone and its active metabolite m-chlorophenylpiperazine (mCPP) were measured via liquid chromatography tandem mass spectrometry. Noncompartmental pharmacokinetic analysis was completed. The half-life (t1/2) ± SD of trazodone for the Hispaniolan parrots was 1.89 ± 0.49 hours, and the t1/2 ± SD of mCPP metabolite was 1.9 ± 0.55 hours. Maximum serum drug concentrations, or Cmax (ng/mL), were 738.3 ± 285.3 for trazodone. Times to achieve Cmax (hours) for trazadone and the mCPP metabolite were 1 hour and 2 hours postdosing, respectively. While this study did not establish the behavioral effects of trazodone, no adverse side effects were observed throughout the 48-hour period following drug administration and blood collection. Our results indicate that the oral administration of a 50-mg/kg single dose of trazodone to Hispaniolan parrots may be considered a safe dose. Plasma concentrations are comparable to previously published values in humans, dogs, horses, and pigeons (Columba livia domestica) for up to 14 hours following dosing. This study indicates that further studies are needed to establish the pharmacodynamics and the efficacy of trazodone in the medical management of behavioral problems in psittacine species.

Detecting depression severity using weighted random forest and oxidative stress biomarkers.

This study employs machine learning to detect the severity of major depressive disorder (MDD) through binary and multiclass classifications. We compared models that used only biomarkers of oxidative stress with those that incorporate sociodemographic and health-related factors. Data collected from 830 participants, based on the Patient Health Questionnaire (PHQ-9) score, inform our analysis. In binary classification, the Random Forest (RF) classifier achieved the highest Area Under the Curve (AUC) of 0.84 when all features were included. In multiclass classification, the AUC improved from 0.84 with only oxidative stress biomarkers to 0.88 when all characteristics were included. To address data imbalance, weighted classifiers, and Synthetic Minority Over-sampling Technique (SMOTE) approaches were applied. Weighted random forest (WRF) improved multiclass classification, achieving an AUC of 0.91. Statistical tests, including the Friedman test and the Conover post-hoc test, confirmed significant differences between model performances, with WRF using all features outperforming others. Feature importance analysis shows that oxidative stress biomarkers, particularly GSH, are top ranked among all features. Clinicians can leverage the results of this study to improve their decision-making processes by incorporating oxidative stress biomarkers in addition to the standard criteria for depression diagnosis.

The effect of itraconazole, a strong CYP3A4 inhibitor, on the pharmacokinetics of the first-in-class ACKR3/CXCR7 antagonist, ACT-1004-1239.

Cytochrome P450 (CYP) 3A4 is an enzyme involved in the metabolism of many drugs that are currently on the market and is therefore a key player in drug-drug interactions (DDIs). ACT-1004-1239 is a potent and selective, first-in-class ACKR3/CXRC7 antagonist being developed as a treatment for demyelinating diseases including multiple sclerosis. Based on the human absorption, distribution, metabolism, and excretion (ADME) study results, ACT-1004-1239 is predominantly metabolized by CYP3A4. This study investigated the effect of the strong CYP3A4 inhibitor, itraconazole, on the pharmacokinetics of single-dose ACT-1004-1239 in healthy male subjects. In the open-label, fixed-sequence DDI study, a total of 16 subjects were treated. Each subject received a single dose of 10 mg ACT-1004-1239 (Treatment A) in the first period followed by concomitant administration of multiple doses of 200 mg itraconazole and a single dose of 10 mg ACT-1004-1239 in the second period. We report a median of difference in tmax (90% confidence interval, CI) of 0.5 h (0.0, 1.0) comparing both treatments. The geometric mean ratio (GMR) (90% CI) of Cmax and AUC0-∞ was 2.16 (1.89, 2.47) and 2.77 (2.55, 3.00), respectively. The GMR (90% CI) of t1/2 was 1.46 (1.26, 1.70). Both treatments were well-tolerated with an identical incidence in subjects reporting treatment-emergent adverse events (TEAE). The most frequently reported TEAEs were headache and nausea. In conclusion, ACT-1004-1239 is classified as a moderately sensitive CYP3A4 substrate (i.e., increase of AUC ≥2- to <5-fold), and this should be considered in further clinical studies if CYP3A4 inhibitors are concomitantly administered.

Investigating the Bioavailability and Insulin-like Growth Factor-I Release of Two Different Strengths of Somapacitan: A Randomised, Double-Blind Crossover Trial.

STUDY DESIGN AND OBJECTIVE: Randomised, double-blind, crossover trial to confirm bioequivalence of somapacitan, a long-acting growth hormone (GH), in 5 mg/1.5 mL and 10 mg/1.5 mL strengths in equimolar doses. METHODS: Healthy participants were randomised (1:1:1) to subcutaneous somapacitan treatment in one dosing period with 5 mg/1.5 mL and two periods with 10 mg/1.5 mL. Eligibility criteria included age 18-45 years and body mass index 18.5-24.9 kg/m2. Exclusion criteria included history of GH deficiency, previous GH treatment, weight > 100.0 kg and participation in any clinical trial of an investigational medicinal product within 45 days or five times the half-life of the previous investigational product before screening. Area under the curve from time 0 until last quantifiable observation (AUC0-t), maximum serum concentration (Cmax), time to Cmax and terminal half-life of somapacitan and safety were assessed. RESULTS: In total, 33 participants were randomised. For AUC0-t, estimated treatment ratio (ETR) (5 mg/1.5 mL versus 10 mg/1.5 mL) was 0.95 (90% confidence interval [CI] 0.89-1.01). Point estimate and 90% CIs were within the acceptance range (0.80-1.25). For Cmax, ETR was 0.77 (90% CI 0.68-0.89). Point estimate and 90% CIs were outside the acceptance range (0.80-1.25). Mean insulin-like growth factor-I (IGF-I) and IGF-I standard deviation score concentration-time curves for each strength were almost identical. No new safety issues were identified. CONCLUSIONS: Bioequivalence criterion for somapacitan 5 mg/1.5 mL and 10 mg/1.5 mL was met for AUC0-t but not for Cmax. The two strengths had equivalent IGF-I responses. TRIAL REGISTRATION: ClinicalTrials.gov, NCT03905850 (3 April 2019).

Somapacitan is a long-acting growth hormone used to treat people with growth hormone deficiency. Somapacitan is injected under the skin with an injection pen. The dose is based on a person's body weight and how they respond to treatment. We compared two strengths of injection pen, containing either 5 or 10 mg of somapacitan per 1.5 mL. For both strengths, participants were given the same dose. We wanted to understand whether the body absorbs these different strengths into the bloodstream in the same way. We also measured levels of insulin-like growth factor-I (IGF-I), a hormone formed when growth hormone is present in the blood, to see the effect of different strengths of somapacitan on the body. In our study, 33 healthy adults received one round of injection using the somapacitan 5 mg pen and two rounds using the somapacitan 10 mg pen, all at least 3 weeks apart. We found no differences in the amount of somapacitan being absorbed into the bloodstream, nor how fast it was absorbed. The peak amount of somapacitan in the bloodstream was higher in people using the 10 mg pen. There were no differences in IGF-I levels following use of either injection pen. Overall, our results show both strengths of somapacitan lead to similar responses in the body. Having different strength options could allow doctors to adjust the dose of somapacitan more easily, depending on a patient's response to treatment.

Reduced hepatic impairment study to evaluate pharmacokinetics and safety of zavegepant and to inform dosing recommendation for hepatic impairment.

Zavegepant, a high-affinity, selective, small-molecule calcitonin gene-related peptide (CGRP) receptor antagonist, is approved in the United States for acute treatment of migraine in adults. The effects of moderate hepatic impairment (8 participants with Child-Pugh score 7-9 points) on the pharmacokinetics of a single 10-mg intranasal dose of zavegepant versus eight matched participants with normal hepatic function were evaluated in a phase I study. Pharmacokinetic sampling determined total and unbound plasma zavegepant concentrations. Moderate hepatic impairment increased the exposure of total zavegepant (~2-fold increase in AUC0-inf and 16% increase in Cmax) versus normal hepatic function, which is not considered clinically meaningful. The geometric least squares mean ratios (moderate impairment/normal) of plasma zavegepant AUC0-inf and Cmax were 193% (90% confidence interval [CI]: 112, 333; p = 0.051) and 116% (90% CI: 69, 195; p = 0.630), respectively. The geometric mean fraction unbound of zavegepant was similar for participants with moderate hepatic impairment (0.13; coefficient of variation [CV] 13.71%) versus those with normal hepatic function (0.11; CV 21.43%). Similar exposure findings were observed with unbound zavegepant versus normal hepatic function (~2.3-fold increase in AUC0-inf and 39% increase in Cmax). One treatment-emergent adverse event (mild, treatment-related headache) was reported in a participant with normal hepatic function. No dosage adjustment of intranasal zavegepant is required in adults with mild or moderate hepatic impairment.

Diagnostic Accuracy of Serum Procalcitonin to Diagnose Sepsis in Advanced Solid Tumor Patients with Fever.

BACKGROUND: Diagnosis of infection in advanced solid tumor patients can be challenging since signs and symptoms might be overlapping due to paraneoplastic condition. Delay diagnosis of existing infection can lead to more severe conditions and increased mortality. Procalcitonin (PCT) has been used to support the diagnosis of bacterial infection and sepsis. Unfortunately, PCT also increases in malignancy even without an infection. We investigated the diagnostic accuracy of PCT in advanced solid tumor patients with fever to diagnose sepsis. METHODS: A cross-sectional study was conducted in solid advanced tumor patients with fever patients who were admitted to Cipto Mangunkusumo Hospitals, Indonesia between June 2016 and April 2018. Sepsis was defined using 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference criteria. The diagnostic accuracy of PCT was determined using the receiver operating characteristic (ROC) curve. RESULTS: A total of 194 subjects were enrolled in this study. 60.3% were female with a mean age of 49.47±12.87 years old. 143 patients (73.7%) with advanced solid tumors. Among this latter group, 39 patients (27%) were sepsis. The ROC curve showed that the levels of PCT for sepsis in advanced solid tumor patients with fever were in the area under the curve (AUC) 0.853 (95%CI 0.785 - 0.921). The Cut-off of PCT in advanced solid tumor patients with fever to classify as sepsis was 2.87 ng/mL, with a sensitivity of 79.5%, and a specificity of 79.8%. CONCLUSION: PCT has good diagnosis accuracy in advanced solid tumor patients with fever to classify as sepsis, however a higher cut-off compared to non-cancerous patients should be used.

HRGCNLDA: Forecasting of lncRNA-disease association based on hierarchical refinement graph convolutional neural network.

Long non-coding RNA (lncRNA) is considered to be a crucial regulator involved in various human biological processes, including the regulation of tumor immune checkpoint proteins. It has great potential as both a cancer biomolecular biomarker and therapeutic target. Nevertheless, conventional biological experimental techniques are both resource-intensive and laborious, making it essential to develop an accurate and efficient computational method to facilitate the discovery of potential links between lncRNAs and diseases. In this study, we proposed HRGCNLDA, a computational approach utilizing hierarchical refinement of graph convolutional neural networks for forecasting lncRNA-disease potential associations. This approach effectively addresses the over-smoothing problem that arises from stacking multiple layers of graph convolutional neural networks. Specifically, HRGCNLDA enhances the layer representation during message propagation and node updates, thereby amplifying the contribution of hidden layers that resemble the ego layer while reducing discrepancies. The results of the experiments showed that HRGCNLDA achieved the highest AUC-ROC (area under the receiver operating characteristic curve, AUC for short) and AUC-PR (area under the precision versus recall curve, AUPR for short) values compared to other methods. Finally, to further demonstrate the reliability and efficacy of our approach, we performed case studies on the case of three prevalent human diseases, namely, breast cancer, lung cancer and gastric cancer.

Performance evaluation of ML models for preoperative prediction of HER2-low BC based on CE-CBBCT radiomic features: A prospective study.

To explore the value of machine learning (ML) models based on contrast-enhanced cone-beam breast computed tomography (CE-CBBCT) radiomics features for the preoperative prediction of human epidermal growth factor receptor 2 (HER2)-low expression breast cancer (BC). Fifty-six patients with HER2-negative invasive BC who underwent preoperative CE-CBBCT were prospectively analyzed. Patients were randomly divided into training and validation cohorts at approximately 7:3. A total of 1046 quantitative radiomic features were extracted from CE-CBBCT images and normalized using z-scores. The Pearson correlation coefficient and recursive feature elimination were used to identify the optimal features. Six ML models were constructed based on the selected features: linear discriminant analysis (LDA), random forest (RF), support vector machine (SVM), logistic regression (LR), AdaBoost (AB), and decision tree (DT). To evaluate the performance of these models, receiver operating characteristic curves and area under the curve (AUC) were used. Seven features were selected as the optimal features for constructing the ML models. In the training cohort, the AUC values for SVM, LDA, RF, LR, AB, and DT were 0.984, 0.981, 1.000, 0.970, 1.000, and 1.000, respectively. In the validation cohort, the AUC values for the SVM, LDA, RF, LR, AB, and DT were 0.859, 0.880, 0.781, 0.880, 0.750, and 0.713, respectively. Among all ML models, the LDA and LR models demonstrated the best performance. The DeLong test showed that there were no significant differences among the receiver operating characteristic curves in all ML models in the training cohort (P > .05); however, in the validation cohort, the DeLong test showed that the differences between the AUCs of LDA and RF, AB, and DT were statistically significant (P = .037, .003, .046). The AUCs of LR and RF, AB, and DT were statistically significant (P = .023, .005, .030). Nevertheless, no statistically significant differences were observed when compared to the other ML models. ML models based on CE-CBBCT radiomics features achieved excellent performance in the preoperative prediction of HER2-low BC and could potentially serve as an effective tool to assist in precise and personalized targeted therapy.

Deep learning-based pathological prediction of lymph node metastasis for patient with renal cell carcinoma from primary whole slide images.

BACKGROUND: Metastasis renal cell carcinoma (RCC) patients have extremely high mortality rate. A predictive model for RCC micrometastasis based on pathomics could be beneficial for clinicians to make treatment decisions. METHODS: A total of 895 formalin-fixed and paraffin-embedded whole slide images (WSIs) derived from three cohorts, including Shanghai General Hospital (SGH), Clinical Proteomic Tumor Analysis Consortium (CPTAC) and Cancer Genome Atlas (TCGA) cohorts, and another 588 frozen section WSIs from TCGA dataset were involved in the study. The deep learning-based strategy for predicting lymphatic metastasis was developed based on WSIs through clustering-constrained-attention multiple-instance learning method and verified among the three cohorts. The performance of the model was further verified in frozen-pathological sections. In addition, the model was also tested the prognosis prediction of patients with RCC in multi-source patient cohorts. RESULTS: The AUC of the lymphatic metastasis prediction performance was 0.836, 0.865 and 0.812 in TCGA, SGH and CPTAC cohorts, respectively. The performance on frozen section WSIs was with the AUC of 0.801. Patients with high deep learning-based prediction of lymph node metastasis values showed worse prognosis. CONCLUSIONS: In this study, we developed and verified a deep learning-based strategy for predicting lymphatic metastasis from primary RCC WSIs, which could be applied in frozen-pathological sections and act as a prognostic factor for RCC to distinguished patients with worse survival outcomes.

Explainable Deep Learning Model for Predicting Serious Adverse Events in Hospitalized Geriatric Patients Within 72 Hours.

Background: The global aging population presents a significant challenge, with older adults experiencing declining physical and cognitive abilities and increased vulnerability to chronic diseases and adverse health outcomes. This study aims to develop an interpretable deep learning (DL) model to predict adverse events in geriatric patients within 72 hours of hospitalization. Methods: The study used retrospective data (2017-2020) from a major medical center in Taiwan. It included non-trauma geriatric patients who visited the emergency department and were admitted to the general ward. Data preprocessing involved collecting prognostic factors like vital signs, lab results, medical history, and clinical management. A deep feedforward neural network was developed, and performance was evaluated using accuracy, sensitivity, specificity, positive predictive value (PPV), and area under the receiver operating characteristic curve (AUC). Model interpretation utilized the Shapley Additive Explanation (SHAP) technique. Results: The analysis included 127,268 patients, with 2.6% experiencing imminent intensive care unit transfer, respiratory failure, or death during hospitalization. The DL model achieved AUCs of 0.86 and 0.84 in the validation and test sets, respectively, outperforming the Sequential Organ Failure Assessment (SOFA) score. Sensitivity and specificity values ranged from 0.79 to 0.81. The SHAP technique provided insights into feature importance and interactions. Conclusion: The developed DL model demonstrated high accuracy in predicting serious adverse events in geriatric patients within 72 hours of hospitalization. It outperformed the SOFA score and provided valuable insights into the model's decision-making process.