Extensions to decision curve analysis, a novel method for evaluating diagnostic tests, prediction models and molecular markers

Establishment and validation of a prognostic model based on common laboratory indicators for SARS-CoV-2 infection in Chinese population

BACKGROUND

At the beginning of December 2022, the Chinese government made major adjustments to the epidemic prevention and control measures. The epidemic infection data and laboratory makers for infected patients based on this period may help with the management and prognostication of COVID-19 patients.

METHODS

The COVID-19 patients hospitalized during December 2022 were enrolled. Logistic regression analysis was used to screen significant factors associated with mortality in patients with COVID-19. Candidate variables were screened by LASSO and stepwise logistic regression methods and were used to construct logistic regression as the prognostic model. The performance of the models was evaluated by discrimination, calibration, and net benefit.

RESULTS

888 patients were eligible, consisting of 715 survivors and 173 all-cause deaths. Factors significantly associated with mortality in COVID-19 patients were: lactate dehydrogenase (LDH), albumin (ALB), procalcitonin (PCT), age, smoking history, malignancy history, high density lipoprotein cholesterol (HDL-C), lactate, vaccine status and urea. 335 of the 888 eligible patients were defined as ICU cases. Seven predictors, including neutrophil to lymphocyte ratio, D-dimer, PCT, C-reactive protein, ALB, bicarbonate, and LDH, were finally selected to establish the prognostic model and generate a nomogram. The area under the curve of the receiver operating curve in the training and validation cohorts were respectively 0.842 and 0.853. In terms of calibration, predicted probabilities and observed proportions displayed high agreements. Decision curve analysis showed high clinical net benefit in the risk threshold of 0.10-0.85. A cutoff value of 81.220 was determined to predict the outcome of COVID-19 patients via this nomogram.

CONCLUSIONS

The laboratory model established in this study showed high discrimination, calibration, and net benefit. It may be used for early identification of severe patients with COVID-19.

A computed tomography‑based radio‑clinical model for the prediction of microvascular invasion in gastric cancer

The objective of the present study was to build and validate a radio-clinical model integrating radiological features and clinical characteristics based on information available before surgery for prediction of microvascular invasion (MI) in gastric cancer. The retrospective study included a cohort of 534 patients (n=374 for the training set and n=160 for the test set) who were diagnosed with gastric cancer. All patients underwent contrast-enhanced computed tomography within one month before surgery. The focal area was mapped by ITK-SNAP. Radiomics features were extracted from portal venous phase CT images. Principal component analysis was used to reduce dimensionality, maximum relevance minimum redundancy, and least absolute shrinkage and selection operator to screen features most associated with MI. The radiomics signature was subsequently computed based on the coefficient weight assigned to it. The independent risk factors for MI of gastric cancer were determined using univariate analysis and multivariate logistic regression analysis. Univariate logistic regression analysis was used to assess the association between clinical characteristics and MI status. A radio-clinical model was constructed by employing multi-variable logistic regression analysis, incorporating radiomic features with clinical characteristics. Receiver operating characteristic (ROC) curve analysis, decision curve analysis (DCA) and calibration curves were employed for the analysis and evaluation of the model's performance. The radiomics signature model had moderate recognition ability, with an area under ROC curve (AUC) of 0.77 for the training set and 0.73 for the test set. The radio-clinical model, consisting of rad-score and clinical features, could well discriminate the training set and test set (AUC=0.88 and 0.80, respectively). The calibration curves and DCA further validated the favorable fit and clinical applicability of the radio-clinical model. In conclusion, the radio-clinical model combining the radiomics signature and clinical characteristics may be used to individually predict MI in gastric cancer to aid in the development of a clinical treatment strategy.

Clinical Significance of Combined Detection of CCL22 and IL-1 as Potential New Bronchial Inflammatory Mediators in Children's Asthma

BACKGROUNDS

Severe asthma is a significant health burden because children with severe asthma are vulnerable to medication-related side effects, life-threatening deterioration, and impaired quality of life. However, there is a lack of data to elucidate the role of inflammatory variables in asthma. This study aimed to compare the levels of inflammatory factors in serum and sputum in children with acute and stable asthma to those in healthy children and the ability to predict clinical response to azithromycin therapy.

METHODS

This study recruited 95 individuals aged 1-3 years old and collected data from January 2018 to 2020. We examined serum and sputum inflammatory factors and constructed the least absolute shrinkage and selection operator (LASSO) model. Predictive models were constructed through multifactor logistic regression and presented in the form of column-line plots. The performance of the column-line diagrams was measured by subject work characteristics (ROC) curves, calibration plots, and decision curve analysis (DCA). Then, filter-paper samples were collected from 45 children with acute asthma who were randomly assigned to receive either azithromycin (10 mg/kg, n = 22) or placebo (n = 23). Pretreatment levels of immune mediators were then analyzed and compared with clinical response to azithromycin therapy.

RESULTS

Of the 95 eligible participants, 21 (22.11%) were healthy controls, 29 (30.53%) had stable asthma, and 45 (47.37%) had acute asthma. The levels of interferon-γ (IFN-γ), tumor necrosis factor-a (TNF-α), chemokine CCL22 (CCL22), interleukin 12 (IL-12), chemokine CCL4 (CCL4), chemokine CCL2 (CCL2), and chemokine CCL13 (CCL13）were significantly higher in the acute asthma group than in the stable asthma group. A logistic regression analysis was performed using CCL22 and IL-1 as independent variables. Additionally, IFN-γ, TNF-α, IL-1, IL-13, and CCL22 were identified in the LASSO model. Finally, we found that CCL22 and IL-1 were more responsive in predicting the response to azithromycin treatment.

CONCLUSION

Our results show that CCL22 and IL-1 are both representative markers during asthma symptom exacerbations and an immune mediator that can predict response to azithromycin therapy.

Development of fully automated models for staging liver fibrosis using non-contrast MRI and artificial intelligence: a retrospective multicenter study

Background

Accurate staging of liver fibrosis (LF) is essential for clinical management in chronic liver disease. While non-contrast MRI (NC-MRI) yields valuable information for liver assessment, its effectiveness in predicting LF remains underexplored. This study aimed to develop and validate artificial intelligence (AI)-powered models utilizing NC-MRI for staging LF.

Methods

A total of 1726 patients from Shengjing Hospital of China Medical University, registered between October 2003 and October 2022, were retrospectively collected, and divided into development (n = 1208) and internal test (n = 518) cohorts. An external test cohort consisting of 337 individuals from six centers, registered between June 2015 and November 2022, were also included. All participants underwent NC-MRI (T1-weighted imaging, T1WI; and T2-fat-suppressed imaging, T2FS) and liver biopsies. Two classification models (CMs), named T1 and T2FS, were trained on respective image types using 3D contextual transformer networks and evaluated on both test cohorts. Additionally, three CMs-Clinic, Image, and Fusion-were developed using clinical features, T1 and T2FS scores, and their integration via logistic regression. Classification effectiveness of CMs was assessed using the area under the receiver operating characteristic curve (AUC). A comparison was conducted between the optimal models (OMs) with highest AUC and other methods (transient elastography, five serum biomarkers, and six radiologists).

Findings

Fusion models (i.e., OM) yielded the highest AUC among the CMs, achieving AUCs of 0.810 for significant fibrosis, 0.881 for advanced fibrosis, and 0.918 for cirrhosis in the internal test cohort, and 0.808, 0.868, and 0.925, respectively, in the external test cohort. The OMs demonstrated superior performance in AUC, significantly surpassing transient elastography (only for staging ≥ F2 and ≥ F3 grades), serum biomarkers, and three junior radiologists for staging LF. Radiologists, with the aid of the OMs, can achieve a higher AUC in LF assessment.

Interpretation

AI-powered models utilizing NC-MRI, including T1WI and T2FS, accurately stage LF.

Funding

National Natural Science Foundation of China (No. 82071885); General Program of the Liaoning Provincial Department of Education (LJKMZ20221160); Liaoning Province Science and Technology Joint Plan (2023JH2/101700127); the Leading Young Talent Program of Xingliao Yingcai in Liaoning Province (XLYC2203037).

Preoperative markers for identifying CT ≤2 cm solid nodules of lung adenocarcinoma based on image deep learning

BACKGROUND

The solid pattern is a highly malignant subtype of lung adenocarcinoma. In the current era of transitioning from lobectomy to sublobar resection for the surgical treatment of small lung cancers, preoperative identification of this subtype is highly important for patient surgical approach selection and long-term prognosis.

METHODS

A total of 1489 patients with clinical stage IA1-2 primary lung adenocarcinoma were enrolled. Based on patient clinical characteristics and lung imaging features obtained via deep learning, highly correlated diagnostic factors were identified through LASSO regression and decision tree analysis. Subsequently, a logistic model and nomogram were constructed. A restricted cubic spline (RCS) was used to calculate the optimal inflection point of quantitative data and the differences between the groups.

RESULTS

The three-dimensional proportion of solid component (PSC), sex, and smoking status was identified as being highly correlated diagnostic factors for solid predominant adenocarcinoma. The logistic model had good prediction efficiency, and the area under the ROC curve was 0.85. Decision curve analysis demonstrated that the application of diagnostic factors can improve patient outcomes. RCS analysis indicated that the proportion of solid adenocarcinomas increased by 4.6 times when the PSC was ≥72%. A PSC of 72% is a good cutoff point.

CONCLUSION

The preoperative diagnosis of solid-pattern adenocarcinoma can be confirmed by typical imaging features and clinical characteristics, assisting the thoracic surgeon in developing a more precise surgical plan.

Multitask machine learning-based tumor-associated collagen signatures predict peritoneal recurrence and disease-free survival in gastric cancer

BACKGROUND

Accurate prediction of peritoneal recurrence for gastric cancer (GC) is crucial in clinic. The collagen alterations in tumor microenvironment affect the migration and treatment response of cancer cells. Herein, we proposed multitask machine learning-based tumor-associated collagen signatures (TACS), which are composed of quantitative collagen features derived from multiphoton imaging, to simultaneously predict peritoneal recurrence (TACSPR) and disease-free survival (TACSDFS).

METHODS

Among 713 consecutive patients, with 275 in training cohort, 222 patients in internal validation cohort, and 216 patients in external validation cohort, we developed and validated a multitask machine learning model for simultaneously predicting peritoneal recurrence (TACSPR) and disease-free survival (TACSDFS). The accuracy of the model for prediction of peritoneal recurrence and prognosis as well as its association with adjuvant chemotherapy were evaluated.

RESULTS

The TACSPR and TACSDFS were independently associated with peritoneal recurrence and disease-free survival in three cohorts, respectively (all P < 0.001). The TACSPR demonstrated a favorable performance for peritoneal recurrence in all three cohorts. In addition, the TACSDFS also showed a satisfactory accuracy for disease-free survival among included patients. For stage II and III diseases, adjuvant chemotherapy improved the survival of patients with low TACSPR and low TACSDFS, or high TACSPR and low TACSDFS, or low TACSPR and high TACSDFS, but had no impact on patients with high TACSPR and high TACSDFS.

CONCLUSIONS

The multitask machine learning model allows accurate prediction of peritoneal recurrence and survival for GC and could distinguish patients who might benefit from adjuvant chemotherapy.

Development of a model to predict the risk of multi-drug resistant organism infections in ruptured intracranial aneurysms patients with hospital-acquired pneumonia in the neurological intensive care unit

OBJECTIVE

This study was developed to explore the incidence of multi-drug resistant organism (MDRO) infections among ruptured intracranial aneurysms(RIA) patient with hospital-acquired pneumonia(HAP) in the neurological intensive care unit (NICU), and to establish risk factors related to the development of these infections.

METHODS

We collected clinical and laboratory data from 328 eligible patients from January 2018 to December 2022. Bacterial culture results were used to assess MDRO strain distributions, and risk factors related to MDRO infection incidence were identified through logistic regression analyses. These risk factors were further used to establish a predictive model for the incidence of MDRO infections, after which this model underwent internal validation.

RESULTS

In this study cohort, 26.5 % of RIA patients with HAP developed MDRO infections (87/328). The most common MDRO pathogens in these patients included Multidrug-resistant Klebsiella pneumoniae (34.31 %) and Multidrug-resistant Acinetobacter baumannii (27.45 %). Six MDRO risk factors, namely, diabetes (P = 0.032), tracheotomy (P = 0.004), history of mechanical ventilation (P = 0.033), lower albumin levels (P < 0.001), hydrocephalus (P < 0.001) and Glasgow Coma Scale (GCS) score ≤8 (P = 0.032) were all independently correlated with MDRO infection incidence. The prediction model exhibited satisfactory discrimination (area under the curve [AUC], 0.842) and calibration (slope, 1.000), with a decision curve analysis further supporting the clinical utility of this model.

CONCLUSIONS

In summary, risk factors and bacterial distributions associated with MDRO infections among RIA patients with HAP in the NICU were herein assessed. The developed predictive model can aid clinicians to identify and screen high-risk patients for preventing MDRO infections.

From bytes to bedside: a systematic review on the use and readiness of artificial intelligence in the neonatal and pediatric intensive care unit

PURPOSE

Despite its promise to enhance patient outcomes and support clinical decision making, clinical use of artificial intelligence (AI) models at the bedside remains limited. Translation of advancements in AI research into tangible clinical benefits is necessary to improve neonatal and pediatric care for critically ill patients. This systematic review seeks to assess the maturity of AI models in neonatal and pediatric intensive care unit (NICU and PICU) treatment, and their risk of bias and objectives.

METHODS

We conducted a systematic search in Medline ALL, Embase, Web of Science Core Collection, Cochrane Central Register of Controlled Trials, and Google Scholar. Studies using AI models during NICU or PICU stay were eligible for inclusion. Study design, objective, dataset size, level of validation, risk of bias, and technological readiness of the models were extracted.

RESULTS

Out of the 1257 identified studies 262 were included. The majority of studies was conducted in the NICU (66%) and most had a high risk of bias (77%). An insufficient sample size was the main cause for this high risk of bias. No studies were identified that integrated an AI model in routine clinical practice and the majority of the studies remained in the prototyping and model development phase.

CONCLUSION

The majority of AI models remain within the testing and prototyping phase and have a high risk of bias. Bridging the gap between designing and clinical implementation of AI models is needed to warrant safe and trustworthy AI models. Specific guidelines and approaches can help improve clinical outcome with usage of AI.

Development and validation of a dynamic nomogram for short-term survival in acute heart failure patients with acute kidney injury upon ICU admission

Objective

The objective of this study is to develop and validate an effective prognostic nomogram for predicting the short-term survival rate of patients with acute heart failure (AHF) complicated by acute kidney injury (AKI) who are admitted to the intensive care unit (ICU).

Patients and methods

We conducted an analysis of data from patients of AHF with AKI spanning the period from 2008 to 2019, utilizing the MIMIC-IV database. Patients were randomly divided into training and validation sets. The training set employed the least absolute shrinkage and selection operator regression model to identify predictors of AKI. Subsequently, a dynamic nomogram was constructed using multivariate Cox regression analysis within the training set and was subsequently validated using the validation set. The nomogram's predictive accuracy, calibration, and clinical utility were evaluated through the concordance index (C-index), calibration plots, and decision curve analysis (DCA).

Results

A total of 978 AHF patients with AKI were analyzed. Multivariate analysis identified serum creatinine, race, age, use of human albumin, use of vasoactive drug, and hemoglobin as independent predictors significantly influencing the short-term prognosis of AHF patients with AKI upon ICU admission. The C-index for the training and validation sets were 0.81 (95%CI: 0.74-0.87) and 0.80 (95 % CI: 0.67-0.92), respectively. The calibration plot of the nomogram demonstrated a close alignment between predicted and observed probabilities. Furthermore, the DCA confirmed the clinical utility of the nomogram.

Conclusions

This study presents a dynamic nomogram that incorporates clinical risk factors and can be conveniently utilized to predict short-term prognosis for AHF patients with AKI upon ICU admission.

Subclassification of obesity for precision prediction of cardiometabolic diseases

Obesity and cardiometabolic disease often, but not always, coincide. Distinguishing subpopulations within which cardiometabolic risk diverges from the risk expected for a given body mass index (BMI) may facilitate precision prevention of cardiometabolic diseases. Accordingly, we performed unsupervised clustering in four European population-based cohorts (N ≈ 173,000). We detected five discordant profiles consisting of individuals with cardiometabolic biomarkers higher or lower than expected given their BMI, which generally increases disease risk, in total representing ~20% of the total population. Persons with discordant profiles differed from concordant individuals in prevalence and future risk of major adverse cardiovascular events (MACE) and type 2 diabetes. Subtle BMI-discordances in biomarkers affected disease risk. For instance, a 10% higher probability of having a discordant lipid profile was associated with a 5% higher risk of MACE (hazard ratio in women 1.05, 95% confidence interval 1.03, 1.06, P = 4.19 × 10-10; hazard ratio in men 1.05, 95% confidence interval 1.04, 1.06, P = 9.33 × 10-14). Multivariate prediction models for MACE and type 2 diabetes performed better when incorporating discordant profile information (likelihood ratio test P < 0.001). This enhancement represents an additional net benefit of 4-15 additional correct interventions and 37-135 additional unnecessary interventions correctly avoided for every 10,000 individuals tested.

Enhancing Preoperative Diagnosis of Subscapular Muscle Injuries with Shoulder MRI-based Multimodal Radiomics

RATIONALE AND OBJECTIVES

Rotator cuff injury is a common ailment in the musculoskeletal system, with the subscapularis muscle being the largest and most robust muscle of the rotator cuff. The occurrence of subscapularis muscle tears is more frequent than previously reported. The main objective of this research is to harness the power of artificial intelligence to enhance the precision in diagnosing subscapularis muscle injuries via magnetic resonance imaging of the shoulder joint, prior to surgical intervention. This study seeks to integrate advanced artificial intelligence algorithms to analyze magnetic resonance imaging data, aiming to provide more accurate preoperative assessments, which can potentially lead to better surgical outcomes and patient care and promote technological progress in the field of medical imaging analysis.

METHOD

This is a multicenter study that involves 324 patients from a major medical center serving as both the training and testing groups, with an additional 60 patients from two other medical centers comprising the verifying group. The imaging protocol for all these subjects included a series of shoulder magnetic resonance imaging scans: T1-weighted coronal sequences, T2-weighted coronal, axial, and sagittal images. These comprehensive imaging modalities were utilized to thoroughly examine the shoulder joint's anatomical details and to detect any signs of subscapularis muscle damage. To enhance the diagnostic accuracy before surgical procedures, radiomic analysis was employed. This technique involves the extraction of a multitude of quantitative features from the magnetic resonance imaging, which can provide a more nuanced and data-driven approach to identifying subscapularis muscle injuries. The integration of radiomics in this study aims to offer a more precise preoperative assessment, potentially leading to improved surgical planning and patient outcomes.

RESULT

In the course of this study, a comprehensive extraction of 1197 radiomic features was performed for each imaging modality of every patient. The coronal T1-weighted modality, when assessed within the internal verifying cohort, delivered a diagnostic accuracy of 0.766, coupled with an AUC of 0.803. In the case of the T2-weighted modality, the coronal planes exhibited a diagnostic accuracy of 0.781 and an AUC of 0.844. The axial T2-weighted images recorded an accuracy of 0.719 and an AUC of 0.761, while the sagittal T2-weighted images scored an accuracy of 0.766 and an AUC of 0.821. The amalgamation of these imaging techniques through a multimodal strategy markedly enhanced the accuracy to 0.828, with an AUC of 0.916 for the internal verifying group. The diagnostic performance of the coronal T1-weighted modality in the external verifying cohort yielded an accuracy of 0.833, with an area under the curve (AUC) of 0.819. For the T2-weighted modality, the coronal imaging demonstrated an accuracy of 0.767 and an AUC of 0.794. The axial T2-weighted images had an accuracy of 0.783 and an AUC of 0.797, while the sagittal T2-weighted images achieved an accuracy of 0.833 and an AUC of 0.800. When combining the modalities, the multimodal approach significantly improved the accuracy to 0.867, with an AUC of 0.803 in the external verifying group, indicating a robust diagnostic capability.

CONCLUSION

Our study demonstrates that the application of multimodal radiomic techniques to shoulder magnetic resonance imaging significantly enhances the precision of preoperative diagnosis for subscapularis muscle injuries. This approach leverages the comprehensive data provided by various magnetic resonance imaging modalities to offer a more detailed and accurate assessment, which is crucial for surgical planning and patient care.

The short-term outcomes and risk factors of post-myocardial infarction ventricular septal rupture: a multi-center retrospective Study

OBJECTIVES

Post-myocardial infarction ventricular septal rupture (PIVSR) is one of the most severe types of mechanical complications after acute myocardial infarction (AMI) with high mortality and poor prognosis. The risk factors for short-term mortality of patients with PIVSR may be not widely recognized. We aimed to assess the prevalence and short-term mortality risk predictors of PIVSR.

METHODS

A total of 62 patients with a diagnosis of PIVSR were admitted to three top general public hospitals in Chongqing, China. Clinical characteristics and short-term outcomes of patients with PIVSR were compared. Predictors of PIVSR were assessed using logistic regression analysis.

RESULTS

Mean age was 70.7 ± 10.7 years (38.7% female). The overall in-hospital mortality of PIVSR remained high (71%). Most (47/62) of the patients were in Killip class III or IV at the time of rupture diagnosis. Logistic regression analysis revealed that white blood cell count (WBC, OR 1.619, 95% CI 1.172-2.237, P = 0.005), cardiogenic shock (OR 47.706, 95%CI 2.859-795.945, P = 0.007) and left ventricular ejection fraction (LVEF, OR 0.803, 95%CI 0.689-0.936, P = 0.009) were independent risk factors of in-hospital early mortality. The nomogram developed for predicting the risk of short-term mortality showed a robust discrimination, with an area under the receiver operating characteristic curve (AUC) of 0.956 (95%CI 0.912-1.000).

CONCLUSION

The short-term mortality of PIVSR remained high. WBC, cardiogenic shock, and LVEF were the independent predictive factors of short-term mortality. Our nomogram might be used to predict early mortality of patients with PIVSR.

Machine learning-based prediction model of lower extremity deep vein thrombosis after stroke

This study aimed to apply machine learning (ML) techniques to develop and validate a risk prediction model for post-stroke lower extremity deep vein thrombosis (DVT) based on patients' limb function, activities of daily living (ADL), clinical laboratory indicators, and DVT preventive measures. We retrospectively analyzed 620 stroke patients. Eight ML models-logistic regression (LR), support vector machine (SVM), random forest (RF), decision tree (DT), neural network (NN), extreme gradient boosting (XGBoost), Bayesian (NB), and K-nearest neighbor (KNN)-were used to build the model. These models were extensively evaluated using ROC curves, AUC, PR curves, PRAUC, accuracy, sensitivity, specificity, and clinical decision curves (DCA). Shapley's additive explanation (SHAP) was used to determine feature importance. Finally, based on the optimal ML algorithm, different functional feature set models were compared with the Padua scale to select the best feature set model. Our results indicated that the RF algorithm demonstrated superior performance in various evaluation metrics, including AUC (0.74/0.73), PRAUC (0.58/0.58), accuracy (0.75/0.77), and sensitivity (0.78/0.80) in both the training set and test set. DCA analysis revealed that the RF model had the highest clinical net benefit. SHAP analysis showed that D-dimer had the most significant influence on DVT, followed by age, Brunnstrom stage (lower limb), prothrombin time (PT), and mobility ability. The RF algorithm can predict post-stroke DVT to guide clinical practice.

A novel nomogram for predicting the prognosis of hepatocellular carcinoma patients following immune checkpoint inhibitors treatment beyond progression: a single center study based on Chinese population

Background

Hepatocellular carcinoma (HCC) persists as a dominant cause of cancer-related mortality globally, with a notably rapid escalation in mortality rates. The advent of immunotherapy, particularly immune checkpoint inhibitors (ICIs), has ushered in a new era in the management of liver cancer, albeit with unresolved challenges in the context of treatment beyond progression (TBP) and stratified prognosis in diverse populations. This study aimed to develop and validate a novel nomogram model to identify factors that predict the benefit of continued immunotherapy for hepatocellular carcinoma patients following disease progression in clinical practice.

Methods

This study retrospectively analyzed the efficacy of ICIs in TBP, focusing on the Chinese population with advanced liver cancer. A nomogram was constructed based on four independent risk factors identified through Cox multivariate analysis, aiming to predict patient prognosis post-ICI treatment. The model was validated through receiver operating characteristic (ROC) curve analysis and categorized patients into high-, intermediate-, and low-risk groups, with further validation using calibration plots and decision curve analysis (DCA).

Results

The low-risk group demonstrated significantly enhanced overall survival (OS) compared to the high-risk group, with the nomogram predictions aligning closely with actual outcomes for 6- and 9-month OS. The model exhibited commendable predictive accuracy, achieving a C-index exceeding 0.7 in both training and validation datasets. The DCA underscored the clinical utility of the nomogram-based prognostic model, further substantiated by the area under the ROC curve (AUC).

Conclusions

The developed nomogram presents a potentially valuable tool for predicting the prognosis of HCC patients undergoing ICI therapy beyond progression, particularly within the Chinese demographic. However, the study is constrained by its retrospective, single-center nature and necessitates further validation through large-scale, multicenter clinical studies across varied populations.

Prediction of high-risk prostate cancer based on the habitat features of biparametric magnetic resonance and the omics features of contrast-enhanced ultrasound

Rationale and objectives

To predict high-risk prostate cancer (PCa) by combining the habitat features of biparametric magnetic resonance imaging (bp-MRI) with the omics features of contrast-enhanced ultrasound (CEUS).

Materials and methods

This study retrospectively collected patients with PCa confirmed by histopathology from January 2020 to June 2023. All patients underwent bp-MRI and CEUS of the prostate, followed by a targeted and transrectal systematic prostate biopsy. The cases were divided into the intermediate-low-risk group (Gleason score ≤7, n = 59) and high-risk group (Gleason score ≥8, n = 33). Radiomics prediction models, namely, MRI_habitat, CEUS_intra, and MRI-CEUS models, were developed based on the habitat features of bp-MRI, the omics features of CEUS, and a merge of features of the two, respectively. Predicted probabilities, called radscores, were then obtained. Clinical-radiological indicators were screened to construct clinic models, which generated clinic scores. The omics-clinic model was constructed by combining the radscore of MRI-CEUS and the clinic score. The predictive performance of all the models was evaluated using the receiver operating characteristic curve.

Results

The area under the curve (AUC) values of the MRI-CEUS model were 0.875 and 0.842 in the training set and test set, respectively, which were higher than those of the MR_habitat (training set: 0.846, test set: 0.813), CEUS_intra (training set: 0.801, test set: 0.743), and clinic models (training set: 0.722, test set: 0.611). The omics-clinic model achieved a higher AUC (train set: 0.986, test set: 0.898).

Conclusions

The combination of the habitat features of bp-MRI and the omics features of CEUS can help predict high-risk PCa.

Noninvasive radiomics approach predicts dopamine agonists treatment response in patients with prolactinoma: a multicenter study

RATIONALE AND OBJECTIVES

The first-line treatment for prolactinoma is drug therapy with dopamine agonists (DAs). However, some patients with resistance to DA treatment should prioritize surgical treatment. Therefore, it is crucial to accurately identify the drug treatment response of prolactinoma before treatment. The present study was performed to determine the DA treatment response of prolactinoma using a clinical radiomic model that incorporated radiomic and clinical features before treatment.

MATERIALS AND METHODS

In total, 255 patients diagnosed with prolactinoma were retrospectively divided to training and validation sets. An elastic net algorithm was used to screen the radiomic features, and a fusion radiomic model was established. A clinical radiomic model was then constructed to integrate the fusion radiomic model and the most important clinical features through multivariate logistic regression analysis for individual prediction. The calibration, discrimination, and clinical applicability of the established models were evaluated. 60 patients with prolactinoma from other centers were used to validate the performance of the constructed model.

RESULTS

The fusion radiomic model was constructed from three significant radiomic features, and the area under the curve in the training set and validation set was 0.930 and 0.910, respectively. The clinical radiomic model was constructed using the radiomic model and three clinical features. The model exhibited good recognition and calibration abilities as evidenced by its area under the curve of 0.96, 0.92, and 0.92 in the training, validation, and external multicenter validation set, respectively. Analysis of the decision curve showed that the fusion radiomic model and clinical radiomic model had good clinical application value for DA treatment response prediction in patients with prolactinoma.

CONCLUSION

Our clinical radiomic model demonstrated high sensitivity and excellent performance in predicting DA treatment response in prolactinoma. This model holds promise for the noninvasive development of individualized diagnosis and treatment strategies for patients with prolactinoma.

Disulfidptosis and ferroptosis related genes define the immune microenvironment and NUBPL serves as a potential biomarker for predicting prognosis and immunotherapy response in bladder cancer

Background

Ferroptosis and disulfidptosis are regulatory forms of cell death that play an important role in tumorigenesis and progression. However, few biomarkers about disulfidptosis and ferroptosis related genes (DFRGs) have been developed to predict the prognosis of bladder cancer (BC).

Methods

We conducted a bioinformatics analysis using public BC datasets to examine the prognostic significance of differentially expressed DFRGs. A Lasso regression was employed to create a prognostic prediction model from these DFRGs. Hub DFRGs that play a role in immunotherapy response and immunoregulation were pinpointed. Immunohistochemistry (IHC) experiment was performed to assess NUBPL and c-MYC expression in BC patients who underwent surgery or received immune checkpoint inhibitor (ICI) immunotherapy at Peking University Cancer Hospital.

Results

We constructed a valid model to predict the prognosis of BC based on DFRGs and performed relevant validation, the results demonstrated that the model was an independent prognostic factor for BC. Further analysis indicated that the model score, combined with the expression of various immune factors and tumor mutation burden (TMB), could predict the prognosis for BC. In addition, we also found that NUBPL was strongly associated with prognosis and response to ICI treatment, and NUBPL may influence BC malignant progression through the c-MYC pathway.

Conclusions

Our research findings highlight the satisfactory predictive value of DFRGs in the immune microenvironment and suggest that NUBPL may be a highly promising biomarker for predicting the prognosis and efficacy of ICI treatment in BC patients.

Machine learning for identifying risk of death in patients with severe fever with thrombocytopenia syndrome

Background

Severe fever with thrombocytopenia syndrome (SFTS) has attracted attention due to the rising incidence and high severity and mortality rates. This study aims to construct a machine learning (ML) model to identify SFTS patients at high risk of death early in hospital admission, and to provide early intensive intervention with a view to reducing the risk of death.

Methods

Data of patients hospitalized for SFTS in two hospitals were collected as training and validation sets, respectively, and six ML methods were used to construct the models using the screened variables as features. The performance of the models was comprehensively evaluated and the best model was selected for interpretation and development of an online web calculator for application.

Results

A total of 483 participants were enrolled in the study and 96 (19.88%) patients died due to SFTS. After a comprehensive evaluation, the XGBoost-based model performs best: the AUC scores for the training and validation sets are 0.962 and 0.997.

Conclusion

Using ML can be a good way to identify high risk individuals in SFTS patients. We can use this model to identify patients at high risk of death early in their admission and manage them intensively at an early stage.

Construction of a molecular diagnostic system for neurogenic rosacea by combining transcriptome sequencing and machine learning

Patients with neurogenic rosacea (NR) frequently demonstrate pronounced neurological manifestations, often unresponsive to conventional therapeutic approaches. A molecular-level understanding and diagnosis of this patient cohort could significantly guide clinical interventions. In this study, we amalgamated our sequencing data (n = 46) with a publicly accessible database (n = 38) to perform an unsupervised cluster analysis of the integrated dataset. The eighty-four rosacea patients were partitioned into two distinct clusters. Neurovascular biomarkers were found to be elevated in cluster 1 compared to cluster 2. Pathways in cluster 1 were predominantly involved in neurotransmitter synthesis, transmission, and functionality, whereas cluster 2 pathways were centered on inflammation-related processes. Differential gene expression analysis and WGCNA were employed to delineate the characteristic gene sets of the two clusters. Subsequently, a diagnostic model was constructed from the identified gene sets using linear regression methodologies. The model's C index, comprising genes PNPLA3, CUX2, PLIN2, and HMGCR, achieved a remarkable value of 0.9683, with an area under the curve (AUC) for the training cohort's nomogram of 0.9376. Clinical characteristics from our dataset (n = 46) were assessed by three seasoned dermatologists, forming the NR validation cohort (NR, n = 18; non-neurogenic rosacea, n = 28). Upon application of our model to NR diagnosis, the model's AUC value reached 0.9023. Finally, potential therapeutic candidates for both patient groups were predicted via the Connectivity Map. In summation, this study unveiled two clusters with unique molecular phenotypes within rosacea, leading to the development of a precise diagnostic model instrumental in NR diagnosis.

Nomogram predicting early urinary incontinence after radical prostatectomy

PURPOSE

One of the most frequent side effects of radical prostatectomy (RP) is urinary incontinence. The primary cause of urine incontinence is usually thought to be impaired urethral sphincter function; nevertheless, the pathophysiology and recovery process of urine incontinence remains unclear. This study aimed to identify potential risk variables, build a risk prediction tool that considers preoperative urodynamic findings, and direct doctors to take necessary action to reduce the likelihood of developing early urinary incontinence.

METHODS

We retrospectively screened patients who underwent radical prostatectomy between January 1, 2020 and December 31, 2023 at the First People 's Hospital of Nantong, China. According to nomogram results, patients who developed incontinence within three months were classified as having early incontinence. The training group's general characteristics were first screened using univariate logistic analysis, and the LASSO method was applied for the best prediction. Multivariate logistic regression analysis was carried out to determine independent risk factors for early postoperative urine incontinence in the training group and to create nomograms that predict the likelihood of developing early urinary incontinence. The model was internally validated by computing the performance of the validation cohort. The nomogram discrimination, correction, and clinical usefulness were assessed using the c-index, receiver operating characteristic curve, correction plot, and clinical decision curve.

RESULTS

The study involved 142 patients in all. Multivariate logistic regression analysis following RP found seven independent risk variables for early urinary incontinence. A nomogram was constructed based on these independent risk factors. The training and validation groups' c-indices showed that the model had high accuracy and stability. The calibration curve demonstrates that the corrective effect of the training and verification groups is perfect, and the area under the receiver operating characteristic curve indicates great identification capacity. Using a nomogram, the clinical net benefit was maximised within a probability threshold of 0.01-1, according to decision curve analysis (DCA).

CONCLUSION

The nomogram model created in this study can offer a clear, personalised analysis of the risk of early urine incontinence following RP. It is highly discriminatory and accurate, and it can help create efficient preventative measures and identify high-risk populations.

NEWS2, S/F-ratio and ROX-index at emergency department for the prediction of adverse outcomes in COVID-19 patients: An external validation study

BACKGROUND

In the context of the COVID-19 pandemic, the early and accurate identification of patients at risk of deterioration was crucial in overcrowded and resource-limited emergency departments. This study conducts an external validation for the evaluation of the performance of the National Early Warning Score 2 (NEWS2), the S/F ratio, and the ROX index at ED admission in a large cohort of COVID-19 patients from Colombia, South America, assessing the net clinical benefit with decision curve analysis.

METHODS

A prospective cohort study was conducted on 6907 adult patients with confirmed COVID-19 admitted to a tertiary care ED in Colombia. The study evaluated the diagnostic performance of NEWS2, S/F ratio, and ROX index scores at ED admission using the area under the receiver operating characteristic curve (AUROC) for discrimination, calibration, and decision curve analysis for the prediction of intensive care unit admission, invasive mechanical ventilation, and in-hospital mortality.

RESULTS

We included 6907 patients who presented to the ED with confirmed SARS-CoV-2 infection from March 2020 to November 2021. Mean age was 51 (35-65) years and 50.4% of patients were males. The rate of intensive care unit admission was 28%, and in-hospital death was 9.8%. All three scores have good discriminatory performance for the three outcomes based on the AUROC. S/F ratio showed miscalibration at low predicted probabilities and decision curve analysis indicated that the NEWS2 score provided a greater net benefit compared to other scores across at a 10% threshold to decide ED admission at a high-level of care facility.

CONCLUSIONS

The NEWS2, S/F ratio, and ROX index at ED admission have good discriminatory performances in COVID-19 patients for the prediction of adverse outcomes, but the NEWS2 score has a higher net benefit underscoring its clinical utility in optimizing patient management and resource allocation in emergency settings.

Radiomic prediction for durable response to high-dose methotrexate-based chemotherapy in primary central nervous system lymphoma

BACKGROUND

The rarity of primary central nervous system lymphoma (PCNSL) and treatment heterogeneity contributes to a lack of prognostic models for evaluating posttreatment remission. This study aimed to develop and validate radiomic-based models to predict the durable response (DR) to high-dose methotrexate (HD-MTX)-based chemotherapy in PCNSL patients.

METHODS

A total of 159 patients pathologically diagnosed with PCNSL between 2011 and 2021 across two institutions were enrolled. According to the NCCN guidelines, the DR was defined as the remission lasting ≥1 year after receiving HD-MTX-based chemotherapy. For each patient, a total of 1218 radiomic features were extracted from prebiopsy T1 contrast-enhanced MR images. Multiple machine-learning algorithms were utilized for feature selection and classification to build a radiomic signature. The radiomic-clinical integrated models were developed using the random forest method. Model performance was externally validated to verify its clinical utility.

RESULTS

A total of 105 PCNSL patients were enrolled after excluding 54 cases with ineligibility. The training and validation cohorts comprised 76 and 29 individuals, respectively. Among them, 65 patients achieved DR. The radiomic signature, consisting of 8 selected features, demonstrated strong predictive performance, with area under the curves of 0.994 in training cohort and 0.913 in validation cohort. This signature was independently associated with the DR in both cohorts. Both the radiomic signature and integrated models significantly outperformed the clinical models in two cohorts. Decision curve analysis underscored the clinical utility of the established models.

CONCLUSIONS

This radiomic signature and integrated models have the potential to accurately predict the DR to HD-MTX-based chemotherapy in PCNSL patients, providing valuable therapeutic insights.

Validation of the FIGO2023 staging system for early-stage endometrial cancer

BACKGROUND

In 2023, the International Federation of Gynecology and Obstetrics (FIGO) updated the endometrial cancer staging system (FIGO2023). Our study aimed to validate the prognostic value of FIGO2023 in patients with early-stage EC (Stage I and Stage II).

METHODS

After screening eligible EC patients from the Surveillance, Epidemiology and End Results (SEER) database, Kaplan-Meier cancer-specific survival (CSS) curves were used to evaluate the prognosis of patients with different stages. In addition, AUC, C-index, Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and Decision curve analysis (DCA) were used to comprehensively compare the efficacy of the new and the old staging system in predicting prognosis.

RESULTS

A total of 33,156 patients were enrolled. The introduction of FIGO2023 significantly increased the proportion of stage II patients from 5.53 % to 24.76 %. The FIGO2023 defines different substages for patients, which show significant differences in CSS. Compared with FIGO2009, FIGO2023 performed better in discrimination, goodness of fit and clinical decision making.

CONCLUSION

Compared with FIGO2009, FIGO2023 had a higher accuracy in predicting CSS in patients with early-stage EC in the SEER database.

Factors predicting the return of spontaneous circulation rate of cardiopulmonary resuscitation in China: Development and evaluation of predictive nomogram

Background

This study aimed to construct and internally validate a probability of the return of spontaneous circulation (ROSC) rate nomogram in a Chinese population of patients with cardiac arrest (CA).

Methods

Patients with CA receiving standard cardiopulmonary resuscitation (CPR) were studied retrospectively. The minor absolute shrinkage and selection operator (LASSO) regression analysis and multivariable logistic regression evaluated various demographic and clinicopathological characteristics. A predictive nomogram was constructed and evaluated for accuracy and reliability using C-index, the area under the receiver operating characteristic curve (AUC), calibration plot, and decision curve analysis (DCA).

Results

A cohort of 508 patients who had experienced CA and received standard CPR was randomly divided into training (70 %, n = 356) and validation groups (30 %, n = 152) for the study. LASSO regression analysis and multivariable logistic regression revealed that thirteen variables, such as age, CPR start time, Electric defibrillation, Epinephrine, Sodium bicarbonate (NaHCO3), CPR Compression duration, The postoperative prothrombin (PT) time, Lactate (Lac), Cardiac troponin (cTn), Potassium (K+), D-dimer, Hypertension (HBP), and Diabetes mellitus (DM), were found to be independent predictors of the ROSC rate of CPR. The nomogram model showed exceptional discrimination, with a C-index of 0.933 (95 % confidence interval: 0.882-0.984). Even in the internal validation, a remarkable C-index value of 0.926 (95 % confidence interval: 0.875-0.977) was still obtained. The accuracy and reliability of the model were also verified by the AUC of 0.923 in the training group and 0.926 in the validation group. The calibration curve showed the model agreed with the actual results. DCA suggested that the predictive nomogram had clinical utility.

Conclusions

A predictive nomogram model was successfully established and proved to identify the influencing factors of the ROSC rate in patients with CA. During cardiopulmonary resuscitation, adjusting the emergency treatment based on the influence factors on ROSC rate is suggested to improve the treatment rate of patients with CA.

Prognostic nomogram for patients with advanced unresectable hepatocellular carcinoma treated with TAE combined with HAIC

Background

Hepatocellular carcinoma (HCC) is the most common primary liver cancer and often arises in the context of chronic liver disease, such as hepatitis B or C infection, and cirrhosis. Advanced unresectable HCC (uHCC) presents significant treatment challenges due to its advanced stage and inoperability. One efficient treatment method for advanced uHCC is the use of hepatic arterial infusion chemotherapy (HAIC) combined with transcatheter arterial embolization (TAE).

Patients and Methods

In this study, we conducted a retrospective collection of clinical data, including basic information, radiological data, and blood test parameters, for patients with advanced uHCC who underwent TAE + HAIC treatment from August 2020 to February 2023. A total of 743 cases involving 262 patients were included. Ultimately, the covariates included in the analysis were the Child-Pugh score, extrahepatic metastasis, tumor number, tumor size, and treatment method.

Results

In the study, we performed univariable and multivariable analysis on 23 clinical factors that were screened by LASSO regression, indicating that the five variables aforementionedly were identified as independent factors influencing patient prognosis. Then we developed a nomogram of the sensitive model and calculated concordance indices of prognostic survival models.

Conclusion

Based on the uHCC patient cohort, we have developed a prognostic model for OS in patients who received TAE + HAIC treatment. This model can accurately predict OS and has the potential to assist in personalized clinical decision-making.

Prediction of postoperative complications following transanal total mesorectal excision in middle and low rectal cancer: development and internal validation of a clinical prediction model

PURPOSE

The objective of this study is to develop a nomogram for the personalized prediction of postoperative complication risks in patients with middle and low rectal cancer who are undergoing transanal total mesorectal excision (taTME). This tool aims to assist clinicians in early identification of high-risk patients and in addressing preoperative risk factors to enhance surgical safety.

METHODS

In this case-control study, 207 patients diagnosed with middle and low rectal cancer and undergoing taTME between February 2018 and November 2023 at The First Affiliated Hospital of Xiamen University were included. Independent risk factors for postoperative complications were analyzed using the Least Absolute Shrinkage and Selection Operator (LASSO) regression and multifactorial logistic regression models. A predictive nomogram was constructed using R Studio.

RESULTS

Among the 207 patients, 57 (27.5%) experienced postoperative complications. The LASSO and multifactorial logistic regression analyses identified operation time (OR = 1.010, P = 0.007), smoking history (OR = 9.693, P < 0.001), anastomotic technique (OR = 0.260, P = 0.004), and ASA score (OR = 9.077, P = 0.051) as significant predictors. These factors were integrated into the nomogram. The model's accuracy was validated through receiver operating characteristic curves, calibration curves, consistency indices, and decision curve analysis.

CONCLUSION

The developed nomogram, incorporating operation time, smoking history, anastomotic technique, and ASA score, effectively forecasts postoperative complication risks in taTME procedures. It is a valuable tool for clinicians to identify patients at heightened risk and initiate timely interventions, ultimately improving patient outcomes.

Better performance of the APPLE score for the prediction of very early atrial fibrillation recurrence post-ablation

OBJECTIVE

The benefits of rhythm control in early atrial fibrillation (AF) are increasingly recognized. This study aimed to investigate whether early AF ablation contributes to long-term sinus rhythm maintenance and to identify a suitable predictive score.

METHODS

According to diagnosis-to-ablation time, this study prospectively enrolled 245 patients with very early AF, 262 with early AF, and 588 with late AF for radiofrequency ablation from June 2017 to December 2022. Clinical data, risk scores, and follow-up results were collected and analyzed.

RESULTS

Baseline characteristics were similar among the three cohorts. During a median follow-up period of 26 months, AF recurrence was observed in 61 (24.9%), 66 (25.2%), and 216 (36.7%) patients in the very early, early, and late AF cohorts, respectively. In the multivariable-adjusted model, very early and early AF were associated with a reduced risk of AF recurrence, with hazard ratios of 0.72 (95% confidence interval [CI] 0.52-0.99) and 0.57 (95% CI 0.41-0.78), respectively. The APPLE score demonstrated the highest predictive power for very early AF, with an area under the curve (AUC) of 0.74. However, its predictive power decreased with time from diagnosis, showing low predictive power for late AF (AUC = 0.58). In addition, the time-dependent concordance index showed consistent results. For very early AF, the Akaike information criterion and decision curve analysis showed that APPLE had the highest predictive value.

CONCLUSION

Very early AF ablation was associated with a lower recurrence rate, and the APPLE score provided a higher predictive value for these patients. (URL: https://www.chictr.org.cn/; Unique identifier: ChiCTR-OIN-17013021).

Age and sex specific thresholds for risk stratification of cardiovascular disease and clinical decision making: prospective open cohort study

Objective

To quantify the potential advantages of using 10 year risk prediction models for cardiovascular disease, in combination with risk thresholds specific to both age and sex, to identify individuals at high risk of cardiovascular disease for allocation of statin treatment.

Design

Prospective open cohort study.

Setting

Primary care data from the UK Clinical Practice Research Datalink GOLD, linked with hospital admissions from Hospital Episode Statistics and national mortality records from the Office for National Statistics in England, 1 January 2006 to 31 May 2019.

Participants

1 046 736 individuals (aged 40-85 years) with no cardiovascular disease, diabetes, or a history of statin treatment at baseline using data from electronic health records.

Main outcome measures

10 year risk of cardiovascular disease, calculated with version 2 of the QRISK cardiovascular disease risk algorithm (QRISK2), with two main strategies to identify individuals at high risk: in strategy A, estimated risk was a fixed cut-off value of ≥10% (ie, as per the UK National Institute for Health and Care Excellence guidelines); in strategy B, estimated risk was ≥10% or ≥90th centile of age and sex specific risk distributions.

Results

Compared with strategy A, strategy B stratified 20 241 (149.8%) more women aged ≤53 years and 9832 (150.2%) more men aged ≤47 years as having a high risk of cardiovascular disease; for all other ages the strategies were the same. Assuming that treatment with statins would be initiated in those identified as high risk, differences in the estimated gain in cardiovascular disease-free life years from statin treatment for strategy B versus strategy A were 0.14 and 0.16 years for women and men aged 40 years, respectively; among individuals aged 40-49 years, the numbers needed to treat to prevent one cardiovascular disease event for strategy B versus strategy A were 39 versus 21 in women and 19 versus 15 in men, respectively.

Conclusions

This study quantified the potential gains in cardiovascular disease-free life years when implementing prevention strategies based on age and sex specific risk thresholds instead of a fixed risk threshold for allocation of statin treatment. Such gains should be weighed against the costs of treating more younger people with statins for longer.

A CT-based radiomics model for predicting lymph node metastasis in hepatic alveolar echinococcosis patients to support lymph node dissection

BACKGROUND

Hepatic alveolar echinococcosis (AE) is a severe zoonotic parasitic disease, and accurate preoperative prediction of lymph node (LN) metastasis in AE patients is crucial for disease management, but it remains an unresolved challenge. The aim of this study was to establish a radiomics model for the preoperative prediction of LN metastasis in hepatic AE patients.

METHODS

A total of 100 hepatic AE patients who underwent hepatectomy and hepatoduodenal ligament LN dissection at Qinghai Provincial People's Hospital between January 2016 and August 2023 were included in the study. The patients were randomly divided into a training set and a validation set at an 8:2 ratio. Radiomic features were extracted from three-dimensional images of the hepatoduodenal ligament LNs delineated on arterial phase computed tomography (CT) scans of hepatic AE patients. Least absolute shrinkage and selection operator (LASSO) regression was applied for data dimensionality reduction and feature selection. Multivariate logistic regression analysis was performed to develop a prediction model, and the predictive performance of the model was evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).

RESULTS

A total of 7 radiomics features associated with LN status were selected using LASSO regression. The classification performances of the training set and validation set were consistent, with area under the operating characteristic curve (AUC) values of 0.928 and 0.890, respectively. The model also demonstrated good stability in subsequent validation.

CONCLUSION

In this study, we established and evaluated a radiomics-based prediction model for LN metastasis in patients with hepatic AE using CT imaging. Our findings may provide a valuable reference for clinicians to determine the occurrence of LN metastasis in hepatic AE patients preoperatively, and help guide the implementation of individualized surgical plans to improve patient prognosis.

Differentiation between viral and autoimmune limbic encephalitis: a prospective cohort study with development and validation of a diagnostic model

BACKGROUND

Distinguishing between viral encephalitis (VE) and autoimmune limbic encephalitis (ALE) presents a clinical challenge due to the overlap in symptoms. We aimed to develop and validate a diagnostic prediction model to differentiate VE and ALE.

METHODS

A prospective observational multicentre cohort study, which continuously enrolled patients diagnosed with either ALE or VE from October 2011 to April 2023. The demographic data, clinical features, and laboratory test results were collected and subjected to logistic regression analyses. The model was displayed as a web-based nomogram and then modified into a scored prediction tool. Model performance was assessed in both derivation and external validation cohorts.

RESULTS

A total of 2423 individuals were recruited, and 1001 (496 VE, 505 ALE) patients were included. Based on the derivation cohort (389 VE, 388 ALE), the model was developed with eight variables including age at onset, acuity, fever, headache, nausea/vomiting, psychiatric or memory complaints, status epilepticus, and CSF white blood cell count. The model showed good discrimination and calibration in both derivation (AUC 0.890; 0.868-0.913) and external validation (107 VE, 117 ALE, AUC 0.872; 0.827-0.917) cohorts. The scored prediction tool had a total point that ranged from - 4 to 10 also showing good discrimination and calibration in both derivation (AUC 0.885, 0.863-0.908) and external validation (AUC 0.868, 0.823-0.913) cohorts.

CONCLUSIONS

The prediction model provides a reliable and user-friendly tool for differentiating between the VE and ALE, which would benefit early diagnosis and appropriate treatment and alleviate economic burdens on both patients and society.

Comparing CHA2DS2-VA and CHA2DS2-VASc scores for stroke risk stratification in patients with atrial fibrillation: a temporal trends analysis from the retrospective Finnish AntiCoagulation in Atrial Fibrillation (FinACAF) cohort

Background

Contemporary data have shown a decrease in the ischaemic stroke risk associated with female sex in patients with atrial fibrillation (AF). We evaluated temporal trends in the predictive value of a non-sex CHA2DS2-VASc risk score (ie. CHA2DS2-VA).

Methods

The FinACAF study covers all patients with incident AF between 2007 and 2018 in Finland from all levels of care. The CHA2DS2-VA score was compared with the CHA2DS2-VASc using continuous and category-based net reclassification indices (NRIs), integrated discrimination improvement (IDI), c-statistics and decision curve analyses.

Findings

We identified 144,879 anticoagulant naïve patients with new-onset AF between 2007 and 2018 (49.9% women; mean age 72.1 years), of whom 3936 (2.7%) experienced ischaemic stroke during one-year follow-up. Based on both continuous and category-based NRIs, the CHA2DS2-VA score was inferior to the CHA2DS2-VASc in the early years (-0.333 (95% CI -0.411 to -0.261) and -0.118 (95% CI -0.137 to -0.099), respectively). However, the differences attenuated over time, and by the end of the study period, the continuous NRI became non-significant (-0.093 (95% CI -0.165 to 0.032)), whereas the category-based NRI reversed in favor of the CHA2DS2-VA (0.070 (95% CI 0.048-0.087)). The IDI was non-significant in early years (0.0009 (95% CI -0.0024 to 0.0037)), but over time became statistically significant in favor of the CHA2DS2-VA score (0.0022 (95% CI 0.0001-0.0044)). The Cox models fitted with the CHA2DS2-VA and the CHA2DS2-VASc scores exhibited comparable discriminative capability in the beginning of the study (p-value 0.63), but over time marginal differences in favor of the CHA2DS2-VA score emerged (p-value 0.0002).

Interpretation

In 2007-2008 (when females had higher AF-related stroke risks than males), the CHA2DS2-VASc score outperformed the CHA2DS2-VA score, but the initial differences between the scores attenuated over time. By the end of the study period in 2017-2018 (with limited/no sex differences in AF-related stroke), there was marginal superiority for the CHA2DS2-VA score.

Funding

This work was supported by the Aarne Koskelo Foundation, The Finnish Foundation for Cardiovascular Research, The Finnish State Research funding, and Helsinki and Uusimaa Hospital District research fund.

Network analysis of histopathological image features and genomics data improving prognosis performance in clear cell renal cell carcinoma

INTRODUCTION

Clear cell renal cell carcinoma is the most common type of kidney cancer, but the prediction of prognosis remains a challenge.

METHODS

We collected whole-slide histopathological images, corresponding clinical and genetic information from the The Cancer Imaging Archive and The Cancer Genome Atlas databases and randomly divided patients into training (n = 197) and validation (n = 84) cohorts. After feature extraction by CellProfiler, we used 2 different machine learning techniques (Least Absolute Shrinkage and Selector Operation-regularized Cox and Support Vector Machine-Recursive Feature Elimination) and weighted gene co-expression network analysis to select prognosis-related image features and genes, respectively. These features and genes were integrated into a joint model using random forest and used to create a nomogram that combines other predictive indicators.

RESULTS

A total of 4 overlapped features were identified, represented by the computed histopathological risk score in the random forest model, and showed predictive value for overall survival (test set: 1-year area under the curves (AUC) = 0.726, 3-year AUC = 0.727, and 5-year AUC = 0.764). The histopathological-genetic risk score (HGRS) integrating the genetic information computed performed better than the model that used image features only (test set: 1-year AUC = 0.682, 3-year AUC = 0.734, and 5-year AUC = 0.78). The nomogram (gender, stage, and HGRS) achieved the highest net benefit according to decision curve analysis compared to HGRS or clinical model.

CONCLUSION

This study developed a histopathological-genetic-related nomogram by combining histopathological features and clinical predictors, providing a more comprehensive prognostic assessment for clear cell renal cell carcinoma patients.

Development of a propionate metabolism-related gene-based molecular subtypes and scoring system for predicting prognosis in bladder cancer

PURPOSE

Bladder cancer (BLCA) is a prevalent malignancy. Dysregulated propionate metabolism, a key cancer factor, suggests a potential target for treating metastatic cancer. However, a complete understanding of the link between propionate metabolism-related genes (PMRGs) and bladder cancer is lacking.

METHODS

From the Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases, we gathered BLCA patient data, which was classified into distinct subgroups using non-negative matrix factorization (NMF). Survival and pathway analyses were conducted between these clusters. The PMRGs model, created through univariate Cox and least absolute shrinkage and selection operator (LASSO) analyses, was assessed for prognostic significance using Kaplan-Meier and receiver operating characteristic (ROC) curves. A comprehensive evaluation included clinical, tumor microenvironment (TME), drug sensitivity, and immunotherapy analyses. Finally, the expression of HSD17B1 essential genes was confirmed via quantitative real-time polymerase chain reaction (qRT-PCR), with further validation through Transwell, wound healing, colony-formation, and EDU assays.

RESULTS

We discovered two distinct subcategories (CA and CB) within BLCA using NMF analysis, with CA demonstrating significantly better overall survival compared to CB. Additionally, six PMRGs emerged as critical factors associated with propionate metabolism and prognosis. Kaplan-Meier analysis revealed that high-risk PMRGs were correlated with a poorer prognosis in BLCA patients. Moreover, significant differences were observed between the two groups in terms of infiltrated immune cells, immune checkpoint expression, TME scores, and drug sensitivity. Notably, we found that suppressing HSD17B1 gene expression inhibited the invasion of bladder cancer cells.

CONCLUSION

Our study proposes molecular subtypes and a PMRG-based score as promising prognostic indicators in BLCA. Additionally, cellular experiments underscore the pivotal role of HSD17B1 in bladder cancer metastasis and invasion, suggesting its potential as a novel therapeutic target.

Exploring the prognostic value and biological pathways of transcriptomics and radiomics patterns in glioblastoma multiforme

Objectives

To develop a multi-omics prognostic model integrating transcriptomics and radiomics for predicting overall survival in patients with glioblastoma multiforme (GBM), and investigate the biological pathways of radiomics patterns.

Materials and methods

Transcription profiles of GBM patients and normal controls were used to obtain differentially expressed mRNAs and long non-coding RNAs (lncRNAs). Radiomics features were extracted from magnetic resonance imaging (MRI). Least absolute shrinkage and selection operator (LASSO) Cox regression was employed to select survival-associated features for the construction of transcriptomics and radiomics signatures. Genes associated with GBM prognosis were identified through the analysis of lncRNA-mRNA co-expression networks and Weighted Gene Co-expression Network Analysis (WGCNA), and their biological pathways were investigated using Genomes enrichment analysis. Transcriptomics, radiomics, and clinical data were integrated to evaluate the multi-omics prognostic model's performance.

Results

LASSO Cox regression yielded 21 survival-related features, including 19 transcriptomics features and 2 radiomics features. Based on transcriptomics and radiomics signature, GBM patients were classified as high-risk or low-risk. The genes obtained from the co-expression network screen were associated with microtubule binding, while those from the WGCNA screen were associated with growth factor receptor binding. In the training set, the AUC values for the multi-omics model and clinical model were 0.964 and 0.830, respectively, while in the validation set, they were 0.907 and 0.787. The multi-omics prognostic model outperformed the clinical prognostic model.

Conclusions

The co-expression network and WGCNA methods revealed genes associated with multiple biological pathways in GBM. The multi-omics prognostic model demonstrated excellent performance and indicated significant potential for clinical application.

Risk Factors and Prediction of 28-Day-All Cause Mortality Among Critically Ill Patients with Acute Pancreatitis Using Machine Learning Techniques: A Retrospective Analysis of Multi-Institutions

Objective

This study aimed to identify the risk factors and construct a reliable prediction model of 28-day all-cause mortality in critically ill patients with acute pancreatitis (AP) using machine learning techniques.

Methods

A total of 534 patients from three different institutions were included. Thirty-eight possible variables were collected from the Intensive care unit (ICU) admission for investigation. Patients were split into a training cohort (n = 400) and test cohort (n = 134) according to their source of hospital. The synthetic minority oversampling technique (SMOTE) was introduced to handle the inherent class imbalance. Six machine learning algorithms were applied in this study. The optimal machine learning model was chosen after patients in the test cohort were selected to validate the models. SHapley Additive exPlanation (SHAP) analysis was performed to rank the importance of variable. The predictive performance of the models was evaluated by the calibration curve, area under the receiver operating characteristics curves (AUROC), and decision clinical analysis.

Results

About 13.5% (72/534) of all patients eventually died of all-cause within 28 days of ICU admission. Eight important variables were screened out, including white blood cell count, platelets, body temperature, age, blood urea nitrogen, red blood cell distribution width, SpO2, and hemoglobin. The support vector machine (SVM) algorithm performed best in predicting 28-d all-cause death. Its AUROC reached 0.877 (95% CI: 0.809 to 0.927, p < 0.001), the Youden index was 0.634 (95% CI: 0.459 to 0.717). Based on the risk stratification system, the difference between the high-risk and low-risk groups was significantly different.

Conclusion

In conclusion, this study developed and validated SVM model, which better predicted 28-d all-cause mortality in critically ill patients with AP. In the future, we will continue to include patients from more institutions to conduct validation in different contexts and countries.

Serum biomarkers REG1A and REG3A combined with the traditional CEA represent a novel nomogram for the screening and risk stratification of colorectal cancer

BACKGROUND

To develop and validate a serum protein nomogram for colorectal cancer (CRC) screening.

METHODS

The serum protein characteristics were extracted from an independent sample containing 30 colorectal cancer and 12 polyp tissues along with their paired samples, and different serum protein expression profiles were validated using RNA microarrays. The prediction model was developed in a training cohort that included 1345 patients clinicopathologically confirmed CRC and 518 normal participants, and data were gathered from November 2011 to January 2017. The lasso logistic regression model was employed for features selection and serum nomogram building. An internal validation cohort containing 576 CRC patients and 222 normal participants was assessed.

RESULTS

Serum signatures containing 27 secreted proteins were significantly differentially expressed in polyps and CRC compared to paired normal tissue, and REG family proteins were selected as potential predictors. The C-index of the nomogram1 (based on Lasso logistic regression model) which contains REG1A, REG3A, CEA and age was 0.913 (95% CI, 0.899 to 0.928) and was well calibrated. Addition of CA199 to the nomogram failed to show incremental prognostic value, as shown in nomogram2 (based on logistic regression model). Application of the nomogram1 in the independent validation cohort had similar discrimination (C-index, 0.912 [95% CI, 0.890 to 0.934]) and good calibration. The decision curve (DCA) and clinical impact curve (ICI) analysis demonstrated that nomogram1 was clinically useful.

CONCLUSIONS

This study presents a serum nomogram that included REG1A, REG3A, CEA and age, which can be convenient for screening of colorectal cancer.

Establishment and validation of a CT-based prediction model for the good dissolution of mild chronic subdural hematoma with atorvastatin treatment

PURPOSE

To develop and validate a prediction model based on imaging data for the prognosis of mild chronic subdural hematoma undergoing atorvastatin treatment.

METHODS

We developed the prediction model utilizing data from patients diagnosed with CSDH between February 2019 and November 2021. Demographic characteristics, medical history, and hematoma characteristics in non-contrast computed tomography (NCCT) were extracted upon admission to the hospital. To reduce data dimensionality, a backward stepwise regression model was implemented to build a prognostic prediction model. We calculated the area under the receiver operating characteristic curve (AUC) of the prognostic prediction model by a tenfold cross-validation procedure.

RESULTS

Maximum thickness, volume, mean density, morphology, and kurtosis of the hematoma were identified as the most significant predictors of good hematoma dissolution in mild CSDH patients undergoing atorvastatin treatment. The prediction model exhibited good discrimination, with an area under the curve (AUC) of 0.82 (95% confidence interval [CI], 0.74-0.90) and good calibration (p = 0.613). The validation analysis showed the AUC of the final prognostic prediction model is 0.80 (95% CI 0.71-0.86) and it has good prediction performance.

CONCLUSION

The imaging data-based prediction model has demonstrated great prediction accuracy for good hematoma dissolution in mild CSDH patients undergoing atorvastatin treatment. The study results emphasize the importance of imaging data evaluation in the management of CSDH patients.

Developing and validating clinical prediction models in hepatology - An overview for clinicians

Prediction models are everywhere in clinical medicine. We use them to assign a diagnosis or a prognosis, and there have been continuous efforts to develop better prediction models. It is important to understand the fundamentals of prediction modelling, thus, we herein describe nine steps to develop and validate a clinical prediction model with the intention of implementing it in clinical practice: Determine if there is a need for a new prediction model; define the purpose and intended use of the model; assess the quality and quantity of the data you wish to develop the model on; develop the model using sound statistical methods; generate risk predictions on the probability scale (0-100%); evaluate the performance of the model in terms of discrimination, calibration, and clinical utility; validate the model using bootstrapping to correct for the apparent optimism in performance; validate the model on external datasets to assess the generalisability and transportability of the model; and finally publish the model so that it can be implemented or validated by others.

A novel ARHGAP family gene signature for survival prediction in glioma patients

ARHGAP family genes are often used as glioma oncogenic factors, and their mechanism of action remains unexplained. Our research entailed a thorough examination of the immune microenvironment and enrichment pathways across various glioma subtypes. A distinctive 6-gene signature was developed employing the CGGA cohort, leading to insights into the disparities in clinical characteristics, mutation patterns, and immune cell infiltration among distinct risk categories. Additionally, a unique nomogram was established, grounded on ARHGAPs, with DCA curves illustrating the model's prospective clinical utility in guiding therapeutic strategies. Emphasizing the role of ARHGAP30, integral to our model, its impact on glioma severity and the credibility of our risk assessment model were substantiated through RT-qPCR, Western blot analysis, and cellular functional assays. We identified 6 ARHGAP family genes associated with glioma prognosis. Analysis using the Kaplan-Meier method indicated a correlation between elevated risk levels and adverse outcomes in glioma patients. The risk score, linked with tumour staging and IDH mutation status, emerged as an independent factor predicting prognosis. Patients in the high-risk category exhibited increased immune cell infiltration, enhanced tumour mutational burden, more pronounced expression of immune checkpoint genes, and a better response to ICB therapy. A nomogram, integrating the risk score with the pathological features of glioma patients, was developed. DCA analysis and cellular studies confirmed the model's potential to improve clinical treatment outcomes for patients. A novel ARHGAP family gene signature reveals the prognosis of glioma.

Fracture Risk Assessment Tool-Based Screening for Osteoporosis in Older Adults in Resource-Limited Settings

PURPOSE

Osteoporosis is a pressing public health concern among older adults, contributing to substantial mortality and morbidity rates. Low- to middle-income countries (LMICs) often grapple with limited access to dual-energy X-ray absorptiometry (DXA), the gold standard for early osteoporosis detection. This study aims to assess the performance of the FRAX® score as a population-wide screening tool for predicting osteoporosis risk, rather than fracture, in individuals aged 50 and above within an LMIC context.

METHODS

This retrospective cohort study (n=864) assessed the performance of the FRAX® score for predicting osteoporosis risk using comparative c-statistics from Receiver Operating Characteristic (ROC) curves. Hazard ratios (HR) and 95 % confidence intervals (CI) were calculated, with p-values <0.05 indicating statistically significant.

RESULTS

The 10-year FRAX® probability for hip fracture, calculated without bone mass density (BMD), exhibited significantly superior performance compared to the 10-year FRAX® probability for major fracture in predicting osteoporosis risk (AUROC: 0.71 versus 0.67, p<0.001). Within 2 to 10 years of follow-up, the 10-year FRAX® probability for hip fracture showed both greater predictive performance and net benefit in the decision curve compared to the FRAX® 10-year probability for major fracture. A newly established cutoff of 1.9 % yielded a negative predictive value of 92.9 % (95 %CI: 90.4-94.8 %) for the 10-year FRAX® probability for hip fracture.

CONCLUSION

The 10-year FRAX® probability for hip fracture estimated without BMD emerges as an effective 10-year screening tool for identifying osteoporosis risk in aged 50 and older, especially when confronted with limited access to DXA scans in LMICs.

MINI ABSTRACT

The Fracture Risk Assessment Tool score performance as an osteoporosis screening tool was assessed in areas with limited dual-energy X-ray access. The hip fracture probability showed better performance than major fracture probability within 2 to 10 years. The tool emerges as effective for screening osteoporosis risk in individuals over 50.

Understanding decision curve analysis in clinical prediction model research

BACKGROUND

Many medical graduate students lack a thorough understanding of decision curve analysis (DCA), a valuable tool in clinical research for evaluating diagnostic models.

METHODS

This article elucidates the concept and process of DCA through the lens of clinical research practices, exemplified by its application in diagnosing liver cancer using serum alpha-fetoprotein levels and radiomics indices. It covers the calculation of probability thresholds, computation of net benefits for each threshold, construction of decision curves, and comparison of decision curves from different models to identify the one offering the highest net benefit.

RESULTS

The paper provides a detailed explanation of DCA, including the creation and comparison of decision curves, and discusses the relationship and differences between decision curves and receiver operating characteristic curves. It highlights the superiority of decision curves in supporting clinical decision-making processes.

CONCLUSION

By clarifying the concept of DCA and highlighting its benefits in clinical decisionmaking, this article has improved researchers' comprehension of how DCA is applied and interpreted, thereby enhancing the quality of research in the medical field.

Elucidating prognosis in cervical squamous cell carcinoma and endocervical adenocarcinoma: a novel anoikis-related gene signature model

Background

Cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC) are among the most prevalent gynecologic malignancies globally. The prognosis is abysmal once cervical cancer progresses to lymphatic metastasis. Anoikis, a specialized form of apoptosis induced by loss of cell adhesion to the extracellular matrix, plays a critical role. The prediction model based on anoikis-related genes (ARGs) expression and clinical data could greatly aid clinical decision-making. However, the relationship between ARGs and CESC remains unclear.

Methods

ARGs curated from the GeneCards and Harmonizome portals were instrumental in delineating CESC subtypes and in developing a prognostic framework for patients afflicted with this condition. We further delved into the intricacies of the immune microenvironment and pathway enrichment across the identified subtypes. Finally, our efforts culminated in the creation of an innovative nomogram that integrates ARGs. The utility of this prognostic tool was underscored by Decision Curve Analysis (DCA), which illuminate its prospective benefits in guiding clinical interventions.

Results

In our study, We discerned a set of 17 survival-pertinent, anoikis-related differentially expressed genes (DEGs) in CESC, from which nine were meticulously selected for the construction of prognostic models. The derived prognostic risk score was subsequently validated as an autonomous prognostic determinant. Through comprehensive functional analyses, we observed distinct immune profiles and drug response patterns among divergent prognostic stratifications. Further, we integrated the risk scores with the clinicopathological characteristics of CESC to develop a robust nomogram. DCA corroborated the utility of our model, demonstrating its potential to enhance patient outcomes through tailored clinical treatment strategies.

Conclusion

The predictive signature, encompassing nine pivotal genes, alongside the meticulously constructed nomogram developed in this research, furnishes clinicians with a sophisticated tool for tailoring treatment strategies to individual patients diagnosed with CESC.

Stemness subtypes in lower-grade glioma with prognostic biomarkers, tumor microenvironment, and treatment response

Our research endeavors are directed towards unraveling the stem cell characteristics of lower-grade glioma patients, with the ultimate goal of formulating personalized treatment strategies. We computed enrichment stemness scores and performed consensus clustering to categorize phenotypes. Subsequently, we constructed a prognostic risk model using weighted gene correlation network analysis (WGCNA), random survival forest regression analysis as well as full subset regression analysis. To validate the expression differences of key genes, we employed experimental methods such as quantitative Polymerase Chain Reaction (qPCR) and assessed cell line proliferation, migration, and invasion. Three subtypes were assigned to patients diagnosed with LGG. Notably, Cluster 2 (C2), exhibiting the poorest survival outcomes, manifested characteristics indicative of the subtype characterized by immunosuppression. This was marked by elevated levels of M1 macrophages, activated mast cells, along with higher immune and stromal scores. Four hub genes-CDCA8, ORC1, DLGAP5, and SMC4-were identified and validated through cell experiments and qPCR. Subsequently, these validated genes were utilized to construct a stemness risk signature. Which revealed that Lower-Grade Glioma (LGG) patients with lower scores were more inclined to demonstrate favorable responses to immune therapy. Our study illuminates the stemness characteristics of gliomas, which lays the foundation for developing therapeutic approaches targeting CSCs and enhancing the efficacy of current immunotherapies. By identifying the stemness subtype and its correlation with prognosis and TME patterns in glioma patients, we aim to advance the development of personalized treatments, enhancing the ability to predict and improve overall patient prognosis.

CDKN2A copy number alteration in bladder cancer: Integrative analysis in patient-derived xenografts and cancer patients

Bladder cancer (BlCa) is an extensively heterogeneous disease that leads to great variability in tumor evolution scenarios and lifelong patient surveillance, emphasizing the need for modern, minimally invasive precision medicine. Here, we explored the clinical significance of copy number alterations (CNAs) in BlCa. CNA profiling was performed in 15 patient-derived xenografts (PDXs) and validated in The Cancer Genome Atlas BlCa (TCGA-BLCA; n = 408) and Lindgren et al. (n = 143) cohorts. CDKN2A copy number loss was identified as the most frequent CNA in bladder tumors, associated with reduced CDKN2A expression, tumors of a papillary phenotype, and prolonged PDX survival. The study's screening cohort consisted of 243 BlCa patients, and CDKN2A copy number was assessed in genomic DNA and cell-free DNA (cfDNA) from 217 tumors and 189 pre-treatment serum samples, respectively. CDKN2A copy number loss was correlated with superior disease-free and progression-free survival of non-muscle-invasive BlCa (NMIBC) patients. Moreover, a higher CDKN2A index (CDKN2A/LEP ratio) in pre-treatment cfDNA was associated with advanced tumor stage and grade and short-term NMIBC progression to invasive disease, while multivariate models fitted for CDKN2A index in pre-treatment cfDNA offered superior risk stratification of T1/high-grade and EORTC high-risk patients, enhancing prediction of treatment outcome. CDKN2A copy number status could serve as a minimally invasive tool to improve risk stratification and support personalized prognosis in BlCa.

Basement membrane-related MMP14 predicts poor prognosis and response to immunotherapy in bladder cancer

BACKGROUND

Basement membrane (BM) is an important component of the extracellular matrix, which plays an important role in the growth and metastasis of tumor cells. However, few biomarkers based on BM have been developed for prognostic assessment and prediction of immunotherapy in bladder cancer (BLCA).

METHODS

In this study, we used the BLCA public database to explore the relationship between BM-related genes (BMRGs) and prognosis. A novel molecular typing of BLCA was performed using consensus clustering. LASSO regression was used to construct a signature based on BMRGs, and its relationship with prognosis was explored using survival analysis. The pivotal BMRGs were further analyzed to assess its clinical characteristics and immune landscape. Finally, immunohistochemistry was used to detect the expression of the hub gene in BLCA patients who underwent surgery or received immune checkpoint inhibitor (ICI) immunotherapy in our hospital.

RESULTS

We comprehensively analyzed the relationship between BMRGs and BLCA, and established a prognostic-related signature which was an independent influence on the prognostic prediction of BLCA. We further screened and validated the pivotal gene-MMP14 in public database. In addition, we found that MMP14 expression in muscle invasive bladder cancer (MIBC) was significantly higher and high MMP14 expression had a poorer response to ICI treatment in our cohort.

CONCLUSIONS

Our findings highlighted the satisfactory value of BMRGs and suggested that MMP14 may be a potential biomarker in predicting prognosis and response to immunotherapy in BLCA.

A computed tomography-based nomogram for neoadjuvant chemotherapy plus immunotherapy response prediction in patients with advanced esophageal squamous cell carcinoma

Objective

To develop a CT-based nomogram to predict the response of advanced esophageal squamous cell carcinoma (ESCC) to neoadjuvant chemotherapy plus immunotherapy.

Methods

In this retrospective study, 158 consecutive patients with advanced ESCC receiving contrast-enhanced CT before neoadjuvant chemotherapy plus immunotherapy were randomized to a training cohort (TC, n = 121) and a validation cohort (VC, n = 37). Response to treatment was assessed with response evaluation criteria in solid tumors. Patients in the TC were divided into the responder (n = 69) and non-responder (n = 52) groups. For the TC, univariate analyses were performed to confirm factors associated with response prediction, and binary analyses were performed to identify independent variables to develop a nomogram. In both the TC and VC, the nomogram performance was assessed by area under the receiver operating characteristic curve (AUC), calibration slope, and decision curve analysis (DCA).

Results

In the TC, univariate analysis showed that cT stage, cN stage, gross tumor volume, gross volume of all enlarged lymph nodes, and tumor length were associated with the response (all P < 0.05). Binary analysis demonstrated that cT stage, cN stage, and tumor length were independent predictors. The independent factors were imported into the R software to construct a nomogram, showing the discriminatory ability with an AUC of 0.813 (95% confidence interval: 0.735-0.890), and the calibration curve and DCA showed that the predictive ability of the nomogram was in good agreement with the actual observation.

Conclusion

This study provides an accurate nomogram to predict the response of advanced ESCC to neoadjuvant chemotherapy plus immunotherapy.

Development and validation of a predictive model for severe white matter hyperintensity with obesity

Purpose

The purpose of the present study was to identify predictors of severe white matter hyperintensity (WMH) with obesity (SWO), and to build a prediction model for screening obese people with severe WMH without Nuclear Magnetic Resonance Imaging (MRI) examination.

Patients subjects and methods

From September 2020 to October 2021, 650 patients with WMH were recruited consecutively. The subjects were divided into two groups, SWO group and non-SWO group. Univariate and Logistic regression analysis were was applied to explore the potential predictors of SWO. The Youden index method was adopted to determine the best cut-off value in the establishment of the prediction model of SWO. Each parameter had two options, low and high. The score table of the prediction model and nomogram based on the logistic regression were constructed. Of the 650 subjects, 487 subjects (75%) were randomly assigned to the training group and 163 subjects (25%) to the validation group. By resampling the area under the curve (AUC) of the subject's operating characteristics and calibration curves 1,000 times, nomogram performance was verified. A decision curve analysis (DCA) was used to evaluate the nomogram's clinical usefulness. By resampling the area under the curve (AUC) of the subject's operating characteristics and calibration curves 1,000 times, nomogram performance was verified. A decision curve analysis (DCA) was used to evaluate the nomogram's clinical usefulness.

Results

Logistic regression demonstrated that hypertension, uric acid (UA), complement 3 (C3) and Interleukin 8 (IL-8) were independent risk factors for SWO. Hypertension, UA, C3, IL-8, folic acid (FA), fasting C-peptide (FCP) and eosinophil could be used to predict the occurrence of SWO in the prediction models, with a good diagnostic performance, Areas Under Curves (AUC) of Total score was 0.823 (95% CI: 0.760-0.885, p < 0.001), sensitivity of 60.0%, specificity of 91.4%. In the development group, the nomogram's AUC (C statistic) was 0.829 (95% CI: 0.760-0.899), while in the validation group, it was 0.835 (95% CI: 0.696, 0.975). In both the development and validation groups, the calibration curves following 1,000 bootstraps showed a satisfactory fit between the observed and predicted probabilities. DCA showed that the nomogram had great clinical utility.

Conclusion

Hypertension, UA, C3, IL-8, FA, FCP and eosinophil models had the potential to predict the incidence of SWO. When the total score of the model exceeded 9 points, the risk of SWO would increase significantly, and the nomogram enabled visualization of the patient's WMH risk. The application prospect of our models mainly lied in the convenient screening of SWO without MRI examination in order to detect SWO and control the WMH hazards early.

Constructing and validating a predictive nomogram for osteoporosis risk among Chinese single-center male population using the systemic immune-inflammation index

Osteoporosis (OP) is a bone metabolism disease that is associated with inflammatory pathological mechanism. Nonetheless, rare studies have investigated the diagnostic effectiveness of immune-inflammation index in the male population. Therefore, it is interesting to achieve early diagnosis of OP in male population based on the inflammatory makers from blood routine examination. We developed a prediction model based on a training dataset of 826 Chinese male patients through a retrospective study, and the data was collected from January 2022 to May 2023. All participants underwent the dual-energy X-ray absorptiometry (DXEA) and blood routine examination. Inflammatory markers such as systemic immune-inflammation index (SII) and platelet-to-lymphocyte ratio (PLR) was calculated and recorded. We utilized the least absolute shrinkage and selection operator (LASSO) regression model to optimize feature selection. Multivariable logistic regression analysis was applied to construct a predicting model incorporating the feature selected in the LASSO model. This predictive model was displayed as a nomogram. Receiver operating characteristic (ROC) curve, C-index, calibration curve, and clinical decision curve analysis (DCA) to evaluate model performance. Internal validation was test by the bootstrapping method. This study was approved by the Ethic Committee of the First Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine (Ethic No. JY2023012) and conducted in accordance with the relevant guidelines and regulations. The predictive factors included in the prediction model were age, BMI, cardiovascular diseases, cerebrovascular diseases, neuropathy, thyroid diseases, fracture history, SII, PLR, C-reactive protein (CRP). The model displayed well discrimination with a C-index of 0.822 (95% confidence interval: 0.798-0.846) and good calibration. Internal validation showed a high C-index value of 0.805. Decision curve analysis (DCA) showed that when the threshold probability was between 3 and 76%, the nomogram had a good clinical value. This nomogram can effectively predict the incidence of OP in male population based on SII and PLR, which would help clinicians rapidly and conveniently diagnose OP with men in the future.

Prediction of histopathologic grades of bladder cancer with radiomics based on MRI: Comparison with traditional MRI

PURPOSE

To compare biparametric magnetic resonance imaging (bp-MRI) radiomics signatures and traditional MRI model for the preoperative prediction of bladder cancer (BCa) grade.

MATERIALS AND METHODS

This retrospective study included 255 consecutive patients with pathologically confirmed 113 low-grade and 142 high-grade BCa. The traditional MRI nomogram model was developed using univariate and multivariate logistic regression by the mean apparent diffusion coefficient (ADC), vesical imaging reporting and data system, tumor size, and the number of tumors. Volumes of interest were manually drawn on T2-weighted imaging (T2WI) and ADC maps by 2 radiologists. Using one-way analysis of variance, correlation, and least absolute shrinkage and selection operator methods to select features. Then, a logistic regression classifier was used to develop the radiomics signatures. Receiver operating characteristic (ROC) analysis was used to compare the diagnostic abilities of the radiomics and traditional MRI models by the DeLong test. Finally, decision curve analysis was performed by estimating the clinical usefulness of the 2 models.

RESULTS

The area under the ROC curves (AUCs) of the traditional MRI model were 0.841 in the training cohort and 0.806 in the validation cohort. The AUCs of the 3 groups of radiomics model [ADC, T2WI, bp-MRI (ADC and T2WI)] were 0.888, 0.875, and 0.899 in the training cohort and 0.863, 0.805, and 0.867 in the validation cohort, respectively. The combined radiomics model achieved higher AUCs than the traditional MRI model. decision curve analysis indicated that the radiomics model had higher net benefits than the traditional MRI model.

CONCLUSION

The bp-MRI radiomics model may help distinguish high-grade and low-grade BCa and outperforming the traditional MRI model. Multicenter validation is needed to acquire high-level evidence for its clinical application.