Purpose: Immune checkpoint blockade agents were shown to provide a survival advantage in urothelial carcinoma, while some patients got minimal benefit or side effects. Therefore, we aimed to investigate the prognostic value of m6A methylation regulators, and developed a nomogram for predicting the response to atezolizumab in urothelial carcinoma patients.Methods: A total of 298 advanced urothelial carcinoma patients with response data in the IMvigor210 cohort were included. Differential expressions of 23 m6A methylation regulators in different treatment outcomes were conducted. Subsequently, a gene signature was developed in the training set using the least absolute shrinkage and selection operator (LASSO) regression. Based on the multivariable logistic regression, a nomogram was constructed by incorporating the gene signature and independent clinicopathological predictors. The performance of the nomogram was assessed by its discrimination, calibration, and clinical utility with internal validation.Results: Six m6A methylation regulators, including IGF2BP1, IGF2BP3, YTHDF2, HNRNPA2B1, FMR1, and FTO, were significantly differentially expressed between the responders and non-responders. These six regulators were also significantly correlated with the treatment outcomes. Based on the LASSO regression analysis, the gene signature consisting of two selected m6A methylation regulators (FMR1 and HNRNPA2B1) was constructed and showed favorable discrimination. The nomogram integrating the gene signature, TMB, and PD-L1 expression on immune cells, showed favorable calibration and discrimination in the training set (AUC 0.768), which was confirmed in the validation set (AUC 0.755). Decision curve analysis confirmed the potential clinical usefulness of the nomogram.

Background: The existing studies on primary bladder lymphoma (PBL) are retrospective analyses based on individual cases or small series studies, and the research on PBL is not unified and in-depth enough at present because of the scarcity of PBL and the lack of relevant literature. This study is designed to develop and validate nomograms for overall survival (OS) and cancer-specific survival (CSS) prediction in patients with PBL. Methods: According to the Surveillance, Epidemiology, and End Results (SEER) database, 405 patients diagnosed with PBL from 1975 to 2016 were collected and randomly assigned to training (n = 283) and validation (n = 122) cohort. After the multivariable Cox regression, the OS and CSS nomograms were developed. The discrimination, calibration and clinical usefulness of the nomograms were assessed and validated, respectively, by the training and validation cohort. Furthermore, all of the patients were reclassified into high- and low-risk groups and their survival were compared through Kaplan-Meier method and log-rank test. Results: Age, subtype, Ann Arbor stage, radiation and chemotherapy were identified as independent prognostic factors for OS and age, sex, and subtype for CSS, then corresponding nomograms predicting the 3- and 5-year survival were constructed. The presented nomograms demonstrated good discrimination and calibration, which the C-index in the training and validation cohort were 0.744 (95% confidence interval [CI], 0.705–0.783) and 0.675 (95% CI, 0.603–0.747) for OS nomogram and 0.692 (95% CI, 0.632–0.752) and 0.646 (95% CI, 0.549–0.743) for CSS nomogram, respectively. Furthermore, the nomograms can be used to effectively distinguish Patients with PBL at high risk of death. The clinical usefulness of the nomograms was visually displayed by decision curve analysis. Conclusion: We updated the baseline characteristics of patients with PBL and constructed OS and CSS nomograms to predict their 3- and 5-year survival. Using these nomograms, it would be convenient to individually predict the prognosis of patients with PBL and provide guidance for clinical treatment.

Objectives We aimed to use machine learning (ML) algorithms to risk stratify the prognosis of critical pulmonary embolism (PE). Material and methods In total, 1229 patients were obtained from MIMIC-IV database. Main outcomes were set as all-cause mortality within 30 days. Logistic regression (LR) and simplified eXtreme gradient boosting (XGBoost) were applied for model constructions. We chose the final models based on their matching degree with data. To simplify the model and increase its usefulness, finally simplified models were built based on the most important 8 variables. Discrimination and calibration were exploited to evaluate the prediction ability. We stratified the risk groups based on risk estimate deciles. Results The simplified XGB model performed better in model discrimination, which AUC were 0.82 (95% CI: 0.78–0.87) in the validation cohort, compared with the AUC of simplified LR model (0.75 [95% CI: 0.69—0.80]). And XGB performed better than sPESI in the validation cohort. A new risk-classification based on XGB could accurately predict low-risk of mortality, and had high consistency with acknowledged risk scores. Conclusions ML models can accurately predict the 30-day mortality of critical PE patients, which could further be used to reduce the burden of ICU stay, decrease the mortality and improve the quality of life for critical PE patients.