A nomogram based on serum cystatin C for predicting acute kidney injury in patients with traumatic brain injury

Classification and Regression Tree Predictive Model for Acute Kidney Injury in Traumatic Brain Injury Patients

Background

Acute kidney injury (AKI) is prevalent in hospitalized patients with traumatic brain injury (TBI), and increases the risk of poor outcomes. We designed this study to develop a visual and convenient decision-tree-based model for predicting AKI in TBI patients.

Methods

A total of 376 patients admitted to the emergency department of the West China Hospital for TBI between January 2015 and June 2019 were included. Demographic information, vital signs on admission, laboratory test results, radiological signs, surgical options, and medications were recorded as variables. AKI was confirmed since the second day after admission, based on the Kidney Disease Improving Global Outcomes criteria. We constructed two predictive models for AKI using least absolute shrinkage and selection operator (LASSO) regression and classification and regression tree (CART), respectively. Receiver operating characteristic (ROC) curves of these two predictive models were drawn, and the area under the ROC curve (AUC) was calculated to compare their predictive accuracy.

Results

The incidence of AKI on the second day after admission was 10.4% among patients with TBI. Lasso regression identified five potent predictive factors for AKI: glucose, serum creatinine, cystatin C, serum uric acid, and fresh frozen plasma transfusions. The CART analysis showed that glucose, serum uric acid, and cystatin C ranked among the top three in terms of the feature importance of the decision tree model. The AUC value of the decision-tree predictive model was 0.892, which was higher than the 0.854 of the LASSO regression model, although the difference was not statistically significant.

Conclusion

The decision tree model is valuable for predicting AKI among patients with TBI. This tree-based flowchart is convenient for physicians to identify patients with TBI who are at high risk of AKI and prompts them to develop suitable therapeutic strategies.

Serum cystatin C is correlated with mortality of traumatic brain injury patients partially mediated by acute kidney injury

BACKGROUND

Evaluating risk of poor outcome for Traumatic Brain Injury (TBI) in early stage is necessary to make treatment strategies and decide the need for intensive care. This study is designed to verify the prognostic value of serum cystatin C in TBI patients.

METHODS

415 TBI patients admitted to West China hospital were included. Logistic regression was performed to explore risk factors of mortality and testify the correlation between cystatin C and mortality. Mediation analysis was conducted to test whether Acute Kidney Injury (AKI) and brain injury severity mediate the relationship between cystatin C level and mortality. Area under the receiver operating characteristic curve (AUC) was used to evaluate the prognostic value of cystatin C and the constructed model incorporating cystatin C.

RESULTS

The mortality rate of 415 TBI patients was 48.9%. Non-survivors had lower GCS (5 vs 8, p < 0.001) and higher cystatin C (0.92 vs 0.71, p < 0.001) than survivors. After adjusting confounding effects, multivariate logistic regression indicated GCS (p < 0.001), glucose (p < 0.001), albumin (p = 0.009), cystatin C (p < 0.001) and subdural hematoma (p = 0.042) were independent risk factors of mortality. Mediation analysis showed both AKI and brain injury severity exerted mediating effects on relationship between cystatin C and mortality of included TBI patients. The AUC of combining GCS with cystatin C was 0.862, which was higher than that of GCS alone (Z = 1.7354, p < 0.05).

CONCLUSION

Both AKI and brain injury severity are mediating variables influencing the relationship between cystatin C and mortality of TBI patients. Serum cystatin C is an effective prognostic marker for TBI patients.

Association of hyperchloremia and acute kidney injury in pediatric patients with moderate and severe traumatic brain injury

PURPOSE

Acute kidney injury (AKI) is an established complication of adult traumatic brain injury (TBI) and known risk factor for mortality. Evidence demonstrates an association between hyperchloremia and AKI in critically ill adults but studies in children are scarce. Given frequent use of hypertonic saline in the management of pediatric TBI, we believe the incidence of hyperchloremia will be high and hypothesize that it will be associated with development of AKI.

METHODS

Single-center retrospective cohort study was completed at an urban, level 1 pediatric trauma center. Children > 40 weeks corrected gestational age and < 21 years of age with moderate or severe TBI (presenting GCS < 13) admitted between January 2016 and December 2021 were included. Primary study outcome was presence of AKI (defined by pediatric Kidney Disease: Improving Global Outcomes criteria) within 7 days of hospitalization and compared between patients with and without hyperchloremia (serum chloride ≥ 110 mEq/L).

RESULTS

Fifty-two children were included. Mean age was 5.75 (S.D. 5.4) years; 60% were male (31/52); and mean presenting GCS was 6 (S.D. 2.9). Thirty-seven patients (71%) developed hyperchloremia with a mean peak chloride of 125 (S.D. 12.0) mEq/L and mean difference between peak and presenting chloride of 16 (S.D. 12.7) mEq/L. Twenty-three patients (44%) developed AKI; of those with hyperchloremia, 62% (23/37) developed AKI, while among those without hyperchloremia, 0% (0/15) developed AKI (difference 62%, 95% CI 42-82%, p < 0.001). Attributable risk of hyperchloremia leading to AKI was 62.2 (95% CI 46.5-77.8, p = 0.0015).

CONCLUSION

Hyperchloremia is common in the management of pediatric TBI and is associated with development of AKI. Risk appears to be associated with both the height of serum chloride and duration of hyperchloremia.

A Machine Learning Approach for the Prediction of Severe Acute Kidney Injury Following Traumatic Brain Injury

BACKGROUND

Acute kidney injury (AKI), a prevalent non-neurological complication following traumatic brain injury (TBI), is a major clinical issue with an unfavorable prognosis. This study aimed to develop and validate machine learning models to predict severe AKI (stage 3 or greater) incidence in patients with TBI.

METHODS

A retrospective cohort study was conducted by using two public databases: the Medical Information Mart for Intensive Care IV (MIMIC)-IV and the eICU Collaborative Research Database (eICU-CRD). Recursive feature elimination was used to select candidate predictors obtained within 24 h of intensive care unit admission. The area under the curve and decision curve analysis curves were used to determine the discriminatory ability. On the other hand, the calibration curve was employed to evaluate the calibrated performance of the newly developed machine learning models.

RESULTS

In the MIMIC-IV database, there were 808 patients diagnosed with moderate and severe TBI (msTBI) (msTBI is defined as Glasgow Coma Score < 12). Of these, 60 (7.43%) patients experienced severe AKI. External validation in the eICU-CRD indicated that the random forest (RF) model had the highest area under the curve of 0.819 (95% confidence interval 0.783-0.851). Furthermore, in the calibration curve, the RF model was well calibrated (P = 0.795).

CONCLUSIONS

In this study, the RF model demonstrated better discrimination in predicting severe AKI than other models. An online calculator could facilitate its application, potentially improving the early detection of severe AKI and subsequently improving the clinical outcomes among patients with msTBI.

A Prediction Model for Acute Kidney Injury in Adult Patients With Minimal Change Disease

Background

Early prediction of acute kidney injury (AKI) can allow for timely interventions, but there are still few methods that are easy and convenient to apply in predicting AKI, specially targeted at patients with minimal change disease (MCD). Motivated by this, we aimed to develop a predicting model for AKI in patients with MCD within the KDIGO criteria.

Methods

Data on 401 hospitalized adult patients, whose biopsy was diagnosed as MCD from 12/31/2010 to 15/7/2021, were retrospectively collected. Among these data, patients underwent biopsy earlier formed the training set (n = 283), while the remaining patients formed the validation set (n = 118). Independent risk factors associated with AKI were analyzed. From this, the prediction model was developed and nomogram was plotted.

Results

AKI was found in 55 of 283 patients (19%) and 15 of 118 patients (13%) in the training and validation cohorts, respectively. According to the results from lasso regression and logistic regression, it was found that four factors, including mean arterial pressure, serum albumin, uric acid, and lymphocyte counts, were independent of the onset of AKI. Incorporating these factors, the nomogram achieved a reasonably good concordance index of 0.84 (95%CI 0.77–0.90) and 0.75 (95%CI 0.62–0.87) in predicting AKI in the training and validation cohorts, respectively. Decision curve analysis suggested clinical benefit of the prediction models.

Conclusions

Our predictive nomogram provides a feasible approach to identify high risk MCD patients who might develop AKI, which might facilitate the timely treatment.

A new model to predict acute kidney injury after cardiac surgery in patients with renal insufficiency

Objective

To establish a simple model for predicting postoperative acute kidney injury (AKI) requiring renal replacement therapy (RRT) in patients with renal insufficiency (CKD stages 3–4) who underwent cardiac surgery.

Methods

A total of 330 patients were enrolled. Among them, 226 were randomly selected for the development group and the remaining 104 for the validation group. The primary outcome was AKI requiring RRT. A nomogram was constructed based on the multivariate analysis with variables selected by the application of the least absolute shrinkage and selection operator. Meanwhile, the discrimination, calibration, and clinical power of the new model were assessed and compared with those of the Cleveland Clinic score and Simplified Renal Index (SRI) score in the validation group. Results: The rate of RRT in the development group was 10.6% (n = 24), while the rate in the validation group was 14.4% (n = 15). The new model included four variables such as postoperative creatinine, aortic cross‐clamping time, emergency, and preoperative cystatin C, with a C-index of 0.851 (95% CI, 0.779–0.924). In the validation group, the areas under the receiver operating characteristic curves for the new model, SRI score, and Cleveland Clinic score were 0.813, 0.791, and 0.786, respectively. Furthermore, the new model demonstrated greater clinical net benefits compared with the Cleveland Clinic score or SRI score.

Conclusions

We developed and validated a powerful predictive model for predicting severe AKI after cardiac surgery in patients with renal insufficiency, which would be helpful to assess the risk for severe AKI requiring RRT.

Machine learning for the prediction of acute kidney injury in critical care patients with acute cerebrovascular disease

Purpose

Acute kidney injury (AKI) is a common complication and associated with a poor clinical outcome. In this study, we developed and validated a model for predicting the risk of AKI through machine learning methods in critical care patients with acute cerebrovascular disease.

Methods

This study was a retrospective study based on two different cohorts. Five machine learning methods were used to develop AKI risk prediction models. We used six popular metrics (AUROC, F2-Score, accuracy, sensitivity, specificity and precision) to evaluate the performance of these models.

Results

We identified 2935 patients in the MIMIC-III database and 499 patients in our local database to develop and validate the AKI risk prediction model. The incidence of AKI in these two different cohorts was 18.3% and 61.7%, respectively. Analysis showed that several laboratory parameters (serum creatinine, hemoglobin, white blood cell count, bicarbonate, blood urea nitrogen, sodium, albumin, and platelet count), age, and length of hospital stay, were the top ten important factors associated with AKI. The analysis demonstrated that the XGBoost had higher AUROC (0.880, 95%CI: 0.831–0.929), indicating that the XGBoost model was better at predicting AKI risk in patients with acute cerebrovascular disease than other models.

Conclusions

This study developed machine learning methods to identify critically ill patients with acute cerebrovascular disease who are at a high risk of developing AKI. This result suggested that machine learning techniques had the potential to improve the prediction of AKI risk models in critical care.

Association Cystatin C and Risk of Stroke in Elderly Patients With Obstructive Sleep Apnea: A Prospective Cohort Study

Background: Few prospective cohort studies have assessed the relationship between Cystatin C (Cys-C) and risk of stroke in elderly patients with obstructive sleep apnea (OSA). The study sought to examine the association between baseline serum Cys-C and long-term risk of stroke among elderly OSA patients.

Methods: A total of 932 patients with OSA, no history of stroke, ≥60 years of age, and complete serum Cys-C records were included in this study. All patients had completed polysomnography (PSG). OSA was defined as an apnea-hypopnea index (AHI) of ≥5 events per hour. Participants were categorized into four groups according to baseline serum Cys-C concentration, split into quartiles. Multivariate Cox regression were used to evaluate the association between Cys-C and the incidence of new-onset stroke.

Results: Stroke occurred in 61 patients during the median 42-month follow-up period. The cumulative incidence rate of stroke was 6.5%, which included 54 patients with ischemic stroke and 7 patients with hemorrhagic stroke. The cumulative incidence of stroke was higher among patients with baseline serum Cys-C concentration of ≥1.15 mg/L when compared with other groups (P_Log–rank < 0.001). After adjusting for potential confounding factors in the Cox regression model, patients with a serum Cys-C concentration of ≥1.15 mg/L had a 2.16-fold higher risk of developing stroke compared with patients with serum Cys-C ≤ 0.81 mg/L (HR, 2.16, 95%CI, 1.09–6.60; P = 0.017). Additionally, there was a higher risk in those of age ≥70 years (HR, 3.23, 95%CI, 1.05–9.24; P = 0.010). The receiver-operating characteristic curves showed that the capability of Cys-C to identify elderly patients with OSA who had a long-time risk of stroke was moderate (AUC = 0.731, 95% CI: 0.683–0.779, P = 0.001).

Conclusion: Increased Cys-C concentration was identified as a risk factor in the incidence of stroke in elderly patients with OSA, independent of gender, BMI, hypertension and other risk factors. Additionally, it conferred a higher risk in patients of age ≥70 years.

Identification of CDK2-Related Immune Forecast Model and ceRNA in Lung Adenocarcinoma, a Pan-Cancer Analysis

Background

Tumor microenvironment (TME) plays important roles in different cancers. Our study aimed to identify molecules with significant prognostic values and construct a relevant Nomogram, immune model, competing endogenous RNA (ceRNA) in lung adenocarcinoma (LUAD).

Methods

“GEO2R,” “limma” R packages were used to identify all differentially expressed mRNAs from Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases. Genes with P-value <0.01, LogFC>2 or <-2 were included for further analyses. The function analysis of 250 overlapping mRNAs was shown by DAVID and Metascape software. By UALCAN, Oncomine and R packages, we explored the expression levels, survival analyses of CDK2 in 33 cancers. “Survival,” “survminer,” “rms” R packages were used to construct a Nomogram model of age, gender, stage, T, M, N. Univariate and multivariate Cox regression were used to establish prognosis-related immune forecast model in LUAD. CeRNA network was constructed by various online databases. The Genomics of Drug Sensitivity in Cancer (GDSC) database was used to explore correlations between CDK2 expression and IC50 of anti-tumor drugs.

Results

A total of 250 differentially expressed genes (DEGs) were identified to participate in many cancer-related pathways, such as activation of immune response, cell adhesion, migration, P13K-AKT signaling pathway. The target molecule CDK2 had prognostic value for the survival of patients in LUAD (P = 5.8e-15). Through Oncomine, TIMER, UALCAN, PrognoScan databases, the expression level of CDK2 in LUAD was higher than normal tissues. Pan-cancer analysis revealed that the expression, stage and survival of CDK2 in 33 cancers, which were statistically significant. Through TISIDB database, we selected 13 immunodepressants, 21 immunostimulants associated with CDK2 and explored 48 genes related to these 34 immunomodulators in cBioProtal database (P < 0.05). Gene Set Enrichment Analysis (GSEA) and Metascape indicated that 49 mRNAs were involved in PUJANA ATM PCC NETWORK (ES = 0.557, P = 0, FDR = 0), SIGNAL TRANSDUCTION (ES = –0.459, P = 0, FDR = 0), immune system process, cell proliferation. Forest map and Nomogram model showed the prognosis of patients with LUAD (Log-Rank = 1.399e-08, Concordance Index = 0.7). Cox regression showed that four mRNAs (SIT1, SNAI3, ASB2, and CDK2) were used to construct the forecast model to predict the prognosis of patients (P < 0.05). LUAD patients were divided into two different risk groups (low and high) had a statistical significance (P = 6.223e-04). By “survival ROC” R package, the total risk score of this prognostic model was AUC = 0.729 (SIT1 = 0.484, SNAI3 = 0.485, ASB2 = 0.267, CDK2 = 0.579). CytoHubba selected ceRNA mechanism medicated by potential biomarkers, 6 lncRNAs-7miRNAs-CDK2. The expression of CDK2 was associated with IC50 of 89 antitumor drugs, and we showed the top 20 drugs with P < 0.05.

Conclusion

In conclusion, our study identified CDK2 related immune forecast model, Nomogram model, forest map, ceRNA network, IC50 of anti-tumor drugs, to predict the prognosis and guide targeted therapy for LUAD patients.

Elevated Ras related GTP binding B (RRAGB) expression predicts poor overall survival and constructs a prognostic nomogram for colon adenocarcinoma

Currently, no articles have explored the roles of RRAGB gene in the occurrence and development of cancer. By means of The Cancer Genome Atlas (TCGA) data mining, we found that this gene might be a novel prognostic predictor for colon adenocarcinoma (COAD). Hence, this article was carried out to explore its roles in COAD and associations with immunity. RRAGB single-gene expression matrix and corresponding clinical information were extracted from TCGA database. Univariate/multivariate cox regression analyses and gene set enrichment analysis (GSEA) were utilized to identify independent prognostic factors and RRAGB related pathways, respectively. Relationships between RRAGB and immunity were also analyzed. Boxplot and K-M survival analysis indicated that RRAGB was not only differently expressed in COAD (P < 0.05), but also significantly associated with overall survival (OS; P < 0.05). Univariate and multivariate Cox hazard regression analyses indicated that RRAGB could serve as an independent prognostic factor for COAD (both P < 0.05). GSEA identified five signaling pathways significantly enriched in the high-RRAGB expression phenotype. Moreover, a RRAGB-based nomogram was successfully constructed and displayed a satisfactory performance. In addition, RRAGB expression was found to be significantly associated with microsatellite instability (MSI), tumor mutational burden (TMB) and immunity. Our results revealed that RRAGB could be a prognostic biomarker for COAD in terms of OS and markedly related to MSI, TMB, and immunity. We also constructed an RRAGB-based nomogram with a satisfactory performance. Further researches should be carried out to validate our findings.