Ferroptosis, as a form of programmed cell death independent of apoptosis, has been demonstrated that plays a major role in tumorigenesis and cancer treatment. A comprehensive analysis of ferroptosis-related genes (FRGs) may lead to a novel choice for the treatment of Ewing sarcoma (ES). Here, 148 differentially expressed FRGs (DEFRGs) were identified between normal and ES tissue. And the GO and KEGG analyses of DEFRGs indicated that these genes were enriched in cancer and immune-related signaling pathways. Then, the GSE17679 cohort was randomly divided into train and test cohorts. Based on the train cohort, AURKA, RGS4, and RIPK1 were identified as key genes through the univariate Cox regression analysis, the random survival forest algorithm, and the multivariate Cox regression analysis and utilized to establish a prognostic FRG signature. The validation results demonstrated that the gene signature has not only excellent prediction performance and generalization ability but is also good at predicting the response of immunotherapy and chemotherapy. Subsequent analysis indicated that all 3 key genes play key roles in tumor immunity and prognosis of ES. Of these, AURKA was highly associated with EWSR1, which was verified by a single-cell dataset (GSE130019). Therefore, the 3 genes may be potential therapeutic targets for ES. At the end of this study, we also constructed an accurate nomogram that helps clinicians to assess the survival time of ES patients. In conclusion, our study constructed an excellent gene signature, which is helpful in improving the prognosis of ES patients.

Background. Ewing sarcoma (ES) is the second most common pediatric bone tumor with a high rate of metastasis, high recurrence, and low survival rate. Therefore, the identification of new biomarkers which can improve the prognosis of ES patients is urgently needed. Methods. Here, GSE17679 dataset was downloaded from GEO databases. WGCNA method was used to identify one module associating with OVS (overall vital survival) and event. cytoHubba was used to screen out 50 hub genes from the module genes. Then, GSE17679 dataset was randomly divided into train cohort and test cohort. Next, univariate Cox analysis, LASSO regression analysis, and multivariate Cox analysis were conducted on 50 hub genes combined with train cohort data to select pivotal genes. Finally, an optimal 7-gene-based risk assessment model was established, which was verified by test cohort, entire GSE17679, and two independent datasets (GSE63157 and TCGA-SARC). Results. The results of the functional enrichment analysis revealed that the OVS and event-associated module were mainly enriched in the protein transcription, cell proliferation, and cell-cycle control. And the train cohort was divided into high-risk and low-risk subgroups based on the median risk score; the results showed that the survival of the low-risk subgroup was significantly longer than high-risk. ROC analysis revealed that AUC values of 1, 3, and 5-year survival were 0.85, 0.94, and 0.88, and Kaplan-Meier analysis also revealed that P value < 0.0001, indicating that this model was accurate, which was also verified in the test, entire cohort, and two independent datasets (GSE63157 and TCGA-SARC). Then, we performed a comprehensive analysis (differential expression analysis, correlation analysis and survival analysis) of seven pivotal genes, and found that four genes (NCAPG, KIF4A, NUF2 and CDC20) plays a more crucial role in the prognosis of ES. Conclusion. Taken together, this study established an optimal 7-gene-based risk assessment model and identified 4 potential therapeutic targets, to improve the prognosis of ES patients.

Purpose. Surgical site infection is one of the serious complications after lumbar fusion. Early prediction and timely intervention can reduce the harm to patients. The aims of this study were to construct and validate a machine learning model for predicting surgical site infection after posterior lumbar interbody fusion, to screen out the most important risk factors for surgical site infection, and to explore whether synthetic minority oversampling technique could improve the model performance. Method. This study reviewed 584 patients who underwent posterior lumbar interbody fusion for degenerative lumbar disease at our center from January 2019 to August 2021. Clinical information and laboratory test data were collected from the electronic medical records. The original dataset was divided into training set and validation set in a 1 : 1 ratio. Seven machine learning algorithms were used to develop predictive models; the training set of each model was resampled using synthetic minority oversampling technique. Finally, the model performance was assessed in the validation set. Results. Of the 584 patients, 33 (5.65%) occurred surgical site infection. Stepwise logistic regression showed that preoperative albumin level (OR 0.659, 95% CI 0.563-0.756), diabetes (OR 9.129, 95% CI 3.816-23.126), intraoperative dural tear (OR 8.436, 95% CI 2.729-25.334), and rheumatic disease (OR 8.471, 95% CI 1.743-39.567) were significant predictors associated with surgical site infection. The performance of the AdaBoost Classification Trees model was the best among the seven machine learning models, and synthetic minority oversampling technique improved the performance of all models. Conclusion. The prediction model we constructed based on machine learning and synthetic minority oversampling technique can accurately predict surgical site infection, which is conducive to clinical decision-making and optimization of perioperative management.