Objective:This study aims to construct and validate several machine learning (ML) algorithms to predict long-term mortality and identify risk factors in unselected patients post-cardiac surgery.MethodsThe Medical Information Mart for Intensive Care (MIMIC-III) database was used to perform a retrospective administrative database study. Candidate predictors consisted of the demographics, comorbidity, vital signs, laboratory test results, scoring systems, and treatment information on the first day of ICU admission. Four-year mortality was set as the study outcome. We used the ML methods of logistic regression (LR), artificial neural network (NNET), naïve bayes (NB), gradient boosting machine (GBM), adapting boosting (Ada), random forest (RF), bagged trees (BT), and eXtreme Gradient Boosting (XGB). The prognostic capacity and clinical utility of these ML models were compared using the area under the receiver operating characteristic curves (AUC), calibration curves, and decision curve analysis (DCA).ResultsOf 7,368 patients in MIMIC-III included in the final cohort, a total of 1,337 (18.15%) patients died during a 4-year follow-up. Among 65 variables extracted from the database, a total of 25 predictors were selected using recursive feature elimination and included in the subsequent analysis. The Ada model performed best among eight models in both discriminatory ability with the highest AUC of 0.801 and goodness of fit (visualized by calibration curve). Moreover, the DCA shows that the net benefit of the RF, Ada, and BT models surpassed that of other ML models for almost all threshold probability values. Additionally, through the Ada technique, we determined that red blood cell distribution width (RDW), blood urea nitrogen (BUN), SAPS II, anion gap (AG), age, urine output, chloride, creatinine, congestive heart failure, and SOFA were the Top 10 predictors in the feature importance rankings.ConclusionsThe Ada model performs best in predicting 4-year mortality after cardiac surgery among the eight ML models, which might have significant application in the development of early warning systems for patients following operations.

Background Lung squamous cell carcinoma (LSCC) is a common subtype of non-small cell lung cancer. Our study aimed to construct and validate a nomogram for predicting overall survival (OS) for postoperative LSCC patients. Methods A total of 8,078 patients eligible for recruitment between 2010 and 2015 were selected from the Surveillance, Epidemiology and End Results database. The study outcomes were 1-, 2-, and 3-year OS. Analyses performed included univariable and multivariable Cox regression, receiver the operating characteristic (ROC) curves, calibration plots, decision curve analysis (DCA), and Kaplan-Meier survival curves. Results Seven variables were selected to establish the nomogram for LSCC. The area under the curve in ROC curve revealed 0.658, 0.651, and 0.647 in the training cohort and 0.673, 0.667, and 0.658 in the validation cohort at 1-, 2-, and 3-year survival. The calibration curves showed satisfactory consistencies between nomogram-predicted and observed survival probability. DCA showed great clinical net benefit. In risk stratification systems, patients were divided into three risk groups with a significant difference in OS in Kaplan-Meier curves (P༜0.001). Conclusions A prognostic nomogram for OS of postoperative LSCC patients was developed and validated, which may assist clinicians in evaluating prognosis and providing highly individualized therapy.

Sepsis-associated acute kidney injury (SA-AKI) results in significant morbidity and mortality, and ferroptosis may play a role in its pathogenesis. Our aim was to examine the effect of exogenous H2S (GYY4137) on ferroptosis and AKI in in vivo and in vitro models of sepsis and explore the possible mechanism involved. Sepsis was induced by cecal ligation and puncture (CLP) in male C57BL/6 mice, which were randomly divided into the sham, CLP, and CLP + GYY4137 group. The indicators of SA-AKI were most prominent at 24 h after CLP, and analysis of the protein expression of ferroptosis indicators showed that ferroptosis was also exacerbated at 24 h after CLP. Moreover, the level of the endogenous H2S synthase CSE (Cystathionine-γ-lyase) and endogenous H2S significantly decreased after CLP. Treatment with GYY4137 reversed or attenuated all these changes. In the in vitro experiments, LPS was used to simulate SA-AKI in mouse renal glomerular endothelial cells (MRGECs). Measurement of ferroptosis-related markers and products of mitochondrial oxidative stress showed that GYY4137 could attenuate ferroptosis and regulate mitochondrial oxidative stress. These findings imply that GYY4137 alleviates SA-AKI by inhibiting ferroptosis triggered by excessive mitochondrial oxidative stress. Thus, GYY4137 may be an effective drug for the clinical treatment of SA-AKI.

Background Many COVID-19-infected patients have been observed to develop unexplained valvular heart disease (VHD), and the association between COVID-19 and VHD remains inconclusive. Therefore, we conducted a two-sample Mendelian randomization study to infer causality between COVID-19 and VHD from a genetic perspective using COVID-19 genetic tools.Methods This study used genetic variables and summary statistics from COVID-19 and VHD genome-wide association studies (GWAS). Single nucleotide polymorphisms (SNPs) were selected based on the assumption of instrumental variables (IVs). The inverse-variance weighted (IVW) method was used as the main analysis method to summarize the causal effects between exposure and outcome, while the weighted median and weighted mode methods were used as secondary methods. MR-Egger was used to test for horizontal pleiotropy, and the Q-test was used to test for heterogeneity. Sensitivity analysis was conducted using leave-one-out method. Scatterplots, forest plots, and funnel plots were used to visualize the results of MR analysis.Results In this study, seven COVID-19-related SNPs were selected as IVs, and the IVW [odds ratio (OR) = 1.16, 95% confidence interval (CI) = 1.04 − 1.28, P = 0.008], weighted median (OR = 1.21, 95% CI = 1.06 − 1.39, P = 0.006), and weighted mode (OR = 1.27, 95% CI = 1.05 − 1.54, P = 0.047) analysis methods suggested a causal effect of COVID-19 on CHD. MR-Egger indicated no evidence of horizontal pleiotropy (P = 0.589), and the Q-test suggested no heterogeneity (IVW, P = 0.349). Sensitivity analysis indicated robustness of the MR analysis results.Conclusions MR analysis revealed a causal effect of COVID-19 infection on the occurrence of VHD, indicating that patients with COVID-19 had a higher risk of VHD.

Myxoma constitutes the main subtype of all benign cardiac tumors, tending to be more common in women and occurring mostly in the left and right atria. Its classic clinical presentations are intracardiac obstruction, embolization, and systemic or constitutional symptoms, such as fever, in decreasing order. Several imaging techniques such as echocardiography, computed tomography, and angiocardiography contribute to the diagnosis of myxoma, ruling out significant coronary diseases, and assessment of myocardial invasion and tumor involvement of adjacent structures. Surgical resection is the only effective therapeutic option for patients with cardiac myxoma. Here, we report a unique case of a middle-aged man who presented with a giant myxoma and a 3-day history of chest tightness and shortness of breath after physical activity. Subsequently, transthoracic echocardiography revealed a mass of solid echodensity located within the right ventricle, complicated by abnormal hemodynamics. A cardiac computed tomographic angiography showed a large homogeneous density filling defect consuming most parts of the right ventricle and protruding from beat to beat. A surgical resection and histological study later successfully confirmed the diagnosis, and the patient's postoperative recovery course was found to be uneventful.