Assessing the Clinical Impact of Risk Prediction Models With Decision Curves: Guidance for Correct Interpretation and Appropriate Use

Prediction of the risk of 3-year chronic kidney disease among elderly people: a community-based cohort study

OBJECTIVE

We conducted a community-based cohort study to predict the 3-year occurrence of chronic kidney disease (CKD) among population aged ≥60 years.

METHOD

Participants were selected from two communities through randomized cluster sampling in Jiading District of Shanghai, China. The two communities were randomly divided into a development cohort (n = 12012) and a validation cohort (n = 6248) with a 3-year follow-up. Logistic regression analysis was used to determine the independent predictors. A nomogram was established to predict the occurrence of CKD within 3 years. The area under the curve (AUC), the calibration curve and decision curve analysis (DCA) curve were used to evaluate the model.

RESULT

At baseline, participants in development cohort and validation cohort were with the mean age of 68.24 ± 5.87 and 67.68 ± 5.26 years old, respectively. During 3 years, 1516 (12.6%) and 544 (8.9%) new cases developed CKD in the development and validation cohorts, respectively. Nine variables (age, systolic blood pressure, body mass index, exercise, previous hypertension, triglycerides, fasting plasma glucose, glycated hemoglobin and serum creatinine) were included in the prediction model. The AUC value was 0.742 [95% confidence interval (CI), 0.728-0.756] in the development cohort and 0.881(95%CI, 0.867-0.895) in the validation cohort, respectively. The calibration curves and DCA curves demonstrate an effective predictive model.

CONCLUSION

Our nomogram model is a simple, reasonable and reliable tool for predicting the risk of 3-year CKD in community-dwelling elderly people, which is helpful for timely intervention and reducing the incidence of CKD.

A nomogram model for predicting the risk of axillary lymph node metastasis in patients with early breast cancer and cN0 status

Axillary staging is commonly performed via sentinel lymph node biopsy for patients with early breast cancer (EBC) presenting with clinically negative axillary lymph nodes (cN0). The present study aimed to investigate the association between axillary lymph node metastasis (ALNM), clinicopathological characteristics of tumors and results from axillary ultrasound (US) scanning. Moreover, a nomogram model was developed to predict the risk for ALNM based on relevant factors. Data from 998 patients who met the inclusion criteria were retrospectively reviewed. These patients were then randomly divided into a training and validation group in a 7:3 ratio. In the training group, receiver operating characteristic curve analysis was used to identify the cutoff values for continuous measurement data. R software was used to identify independent ALNM risk variables in the training group using univariate and multivariate logistic regression analysis. The selected independent risk factors were incorporated into a nomogram. The model differentiation was assessed using the area under the curve (AUC), while calibration was evaluated through calibration charts and the Hosmer-Lemeshow test. To assess clinical applicability, a decision curve analysis (DCA) was conducted. Internal verification was performed via 1000 rounds of bootstrap resampling. Among the 998 patients with EBC, 228 (22.84%) developed ALNM. Multivariate logistic analysis identified lymphovascular invasion, axillary US findings, maximum diameter and molecular subtype as independent risk factors for ALNM. The Akaike Information Criterion served as the basis for both nomogram development and model selection. Robust differentiation was shown by the AUC values of 0.855 (95% CI, 0.817-0.892) and 0.793 (95% CI, 0.725-0.857) for the training and validation groups, respectively. The Hosmer-Lemeshow test yielded P-values of 0.869 and 0.847 for the training and validation groups, respectively, and the calibration chart aligned closely with the ideal curve, affirming excellent calibration. DCA showed that the net benefit from the nomogram significantly outweighed both the 'no intervention' and the 'full intervention' approaches, falling within the threshold probability interval of 12-97% for the training group and 17-82% for the validation group. This underscores the robust clinical utility of the model. A nomogram model was successfully constructed and validated to predict the risk of ALNM in patients with EBC and cN0 status. The model demonstrated favorable differentiation, calibration and clinical applicability, offering valuable guidance for assessing axillary lymph node status in this population.

MRI radiomics predicts the efficacy of EGFR-TKI in EGFR-mutant non-small-cell lung cancer with brain metastasis

AIM

To develop and validate models based on magnetic resonance imaging (MRI) radiomics for predicting the efficacy of epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) in EGFR-mutant non-small-cell lung cancer (NSCLC) patients with brain metastases.

MATERIALS AND METHODS

117 EGFR-mutant NSCLC patients with brain metastases who received EGFR-TKI treatment were included in this study from January 1, 2014 to December 31, 2021. Patients were randomly divided into training and validation cohorts in a ratio of 2:1. Radiomics features extracted from brain MRI were screened by least absolute shrinkage and selection operator (LASSO) algorithm. Logistic regression analysis and Cox proportional hazard regression analysis were used to screen clinical risk factors. Clinical (C), radiomics (R), and combined (C + R) nomograms were constructed in models predicting short-term efficacy and intracranial progression-free survival (iPFS), respectively. Calibration curves, Harrell's concordance index (C-index), and decision curve analysis (DCA) were used to evaluate the performance of models.

RESULTS

Overall response rate (ORR) was 57.3% and median iPFS was 12.67 months. The C + R nomograms were more effective. In the short-term efficacy model, the C-indexes of C + R nomograms in training cohort and validation cohort were 0.860 (0.820-0.901, 95%CI) and 0.843 (0.783-0.904, 95%CI). In iPFS model, the C-indexes of C + R nomograms in training cohort and validation cohort were 0.837 (0.751-0.923, 95%CI) and 0.850 (0.763-0.937, 95%CI).

CONCLUSION

The C + R nomograms were more effective in predicting EGFR-TKI efficacy of EGFR-mutant NSCLC patients with brain metastases than single clinical or radiomics nomograms.

Multicenter validation of an artificial intelligence (AI)-based platform for the diagnosis of acute appendicitis

BACKGROUND

The current scores used to help diagnose acute appendicitis have a "gray" zone in which the diagnosis is usually inconclusive. Furthermore, the universal use of CT scanning is limited because of the radiation hazards and/or limited resources. Hence, it is imperative to have an accurate diagnostic tool to avoid unnecessary, negative appendectomies.

METHODS

This was an international, multicenter, retrospective cohort study. The diagnostic accuracy of the artificial intelligence platform was assessed by sensitivity, specificity, negative predictive value, the area under the receiver curve, precision curve, F1 score, and Matthews correlation coefficient. Moreover, calibration curve, decision curve analysis, and clinical impact curve analysis were used to assess the clinical utility of the artificial intelligence platform. The accuracy of the artificial intelligence platform was also compared to that of CT scanning.

RESULTS

Two data sets were used to assess the artificial intelligence platform: a multicenter real data set (n = 2,579) and a well-qualified synthetic data set (n = 9736). The platform showed a sensitivity of 92.2%, specificity of 97.2%, and negative predictive value of 98.7%. The artificial intelligence had good area under the receiver curve, precision, F1 score, and Matthews correlation coefficient (0.97, 86.7, 0.89, 0.88, respectively). Compared to CT scanning, the artificial intelligence platform had a better area under the receiver curve (0.92 vs 0.76), specificity (90.9 vs 53.3), precision (99.8 vs 98.9), and Matthews correlation coefficient (0.77 vs 0.72), comparable sensitivity (99.2 vs 100), and lower negative predictive value (67.6 vs 99.5). Decision curve analysis and clinical impact curve analysis intuitively revealed that the platform had a substantial net benefit within a realistic probability range from 6% to 96%.

CONCLUSION

The current artificial intelligence platform had excellent sensitivity, specificity, and accuracy exceeding 90% and may help clinicians in decision making on patients with suspected acute appendicitis, particularly when access to CT scanning is limited.

A prognostic binary classifier comprised of five critical mRNAs stratified pancreatic cancer patients following resection

Background

Pancreatic cancer is characterized by an extremely poor prognosis, even following potentially curative resection. Classical prognostic markers such as histopathological or clinical parameters have limited predictive power. The present study aimed to establish a prognostic model combining mRNA expression data with histopathological and clinical data to better predict survival and stratify pancreatic cancer patients following resection. We pioneered three models in one study and systematically evaluated the clinical benefits of all three models.

Methods

To identify differentially expressed genes in pancreatic cancer, mRNA data from normal (GTEx database) and pancreatic cancer (TCGA database) tissues were used. Survival analysis was carried out to identify prognosis-relevant genes from the identified differentially expressed genes and LASSO regression was used to filter out hub genes. The risk score of several hub genes was calculated according to gene expression and coefficients. Validation was carried out using an independent set of GEO microarray data. Multivariate COX regression was used for identifying independent clinical and pathological risk factors related to patient's survival in the TCGA database and a prognostic model combining mRNA expression data with histopathological and clinical data was established. Another prognostic model using clinicopathological factors from the SEER database was conceived based on multivariate COX regression. NRI (net reclassification improvement) and IDI (integrated discrimination index) were used to compare the predictive capabilities of the different models.

Results

We identified 1589 differentially expressed genes (DEGs) through the comparison of normal and pancreatic cancer tissues, of whom 317 were associated with prognosis(p < 0.05). LASSO regression identified five hub genes, MYEOV, ANXA2P2, MET, CEP55, and KRT7, that were used for the five-mRNA-classifier prognostic model. The classifier could stratify patients into a short and long survival group: 5-year overall survival in the training set (TCGA, 6 % vs 52 %, p < 0.001), test set (TCGA, 18 % vs 55 %,p < 0.01) and external validation set (GEO, 0 % vs 25 %, p < 0.05). Sensitivity analysis showed that the mRNA model (model 1) was better than the clinicopathological no-mRNA model (model 2) in predicting 5-year survival in the TCGA database (AUC: 0.877 vs 0.718, z = 3.165, p < 0.01) and better than the multi-factor prognostic model (model 3) from the SEER database (AUC: 0.754, z = 2.637, p < 0.01). On predictive performance, model 1 improved model 2 (NRI = 0.084, z = 1.288, p = 0.198; IDI = 0.055, z = 1.041,p = 0.298) and model 3 (NRI = 0.167,z = 1.961,p = 0.05; IDI = 0.086, z = 1.427, p = 0.154).

Conclusion

The five-mRNA-classifier is a reliable and feasible instrument to predict the prognosis of pancreatic cancer patients following resection. It might help in patiens counseling and assist clinicians in providing individualized treatment for patients in different risk groups.

Evaluating the Shock Index, Revised Assessment of Bleeding and Transfusion (RABT), Assessment of Blood Consumption (ABC) and novel PTTrauma score to predict critical transfusion threshold (CAT) in penetrating thoracic trauma

The shock index (SI) has been associated with predicting transfusion needs in trauma patients. However, its utility in penetrating thoracic trauma (PTTrauma) for predicting the Critical Administration Threshold (CAT) has not been well-studied. This study aimed to evaluate the prognostic value of SI in predicting CAT in PTTrauma patients and compare its performance with the Assessment of Blood Consumption (ABC) and Revised Assessment of Bleeding and Transfusion (RABT) scores. We conducted a prognostic type 2, single-center retrospective observational cohort study on patients with PTTrauma and an Injury Severity Score (ISS) > 9. The primary exposure was SI at admission, and the primary outcome was CAT. Logistic regression and decision curve analysis were used to assess the predictive performance of SI and the PTTrauma score, a novel model incorporating clinical variables. Of the 620 participants, 53 (8.5%) had more than one CAT. An SI > 0.9 was associated with CAT (adjusted OR 4.89, 95% CI 1.64-14.60). The PTTrauma score outperformed SI, ABC, and RABT scores in predicting CAT (AUC 0.867, 95% CI 0.826-0.908). SI is a valuable predictor of CAT in PTTrauma patients. The novel PTTrauma score demonstrates superior performance compared to existing scores, highlighting the importance of developing targeted predictive models for specific injury patterns. These findings can guide clinical decision-making and resource allocation in the management of PTTrauma.

Prediction of sustained opioid use in children and adolescents using machine learning

BACKGROUND

Opioid misuse in the paediatric population is understudied. This study aimed to develop a machine learning classifier to differentiate between occasional and sustained opioid users among children and adolescents in outpatient settings.

METHODS

Data for 29,335 patients under 19 yr with recorded opioid purchases were collected from medical records. Machine learning methods were applied to predict sustained opioid use within 1, 2, or 3 yr after first opioid use, using sociodemographic information, medical history, and healthcare usage variables collected near the time of first prescription fulfilment. The models' performance was evaluated with classification and calibration metrics, and a decision curve analysis. An online tool was deployed for model self-exploration and visualisation.

RESULTS

The models demonstrated good performance, with a 1-yr follow-up model achieving a sensitivity of 0.772, a specificity of 0.703, and an ROC-AUC of 0.792 on an independent test set, with calibration intercept and slope of 0.00 and 1.02, respectively. Decision curve analysis revealed the clinical benefit of using the model relative to other strategies. SHAP analysis (SHapley Additive exPlanations) identified influential variables, including the number of diagnoses, medical images, laboratory tests, and type of opioid used.

CONCLUSIONS

Our model showed promising performance in predicting sustained opioid use among paediatric patients. The online risk prediction tool can facilitate compliance to such tools by clinicians. This study presents the potential of machine learning in identifying at-risk paediatric populations for sustained opioid use, potentially contributing to secondary prevention of opioid abuse.

Predicting the risk of 1-year mortality among patients hospitalized for acute heart failure in China

BACKGROUND

We aimed to develop and validate a model to predict 1-year mortality risk among patients hospitalized for acute heart failure (AHF), build a risk score and interpret its application in clinical decision making.

METHODS

By using data from China Patient-Centred Evaluative Assessment of Cardiac Events Prospective Heart Failure Study, which prospectively enrolled patients hospitalized for AHF in 52 hospitals across 20 provinces, we used multivariate Cox proportional hazard model to develop and validate a model to predict 1-year mortality.

RESULTS

There were 4,875 patients included in the study, 857 (17.58%) of them died within 1-year following discharge of index hospitalization. A total of 13 predictors were selected to establish the prediction model, including age, medical history of chronic obstructive pulmonary disease and hypertension, systolic blood pressure, Kansas City Cardiomyopathy Questionnaire-12 score, angiotensin converting enzyme inhibitor or angiotensin receptor blocker at discharge, discharge symptom, N-terminal pro-brain natriuretic peptide, high-sensitivity troponin T, serum creatine, albumin, blood urea nitrogen, and highly sensitive C-reactive protein. The model showed a high performance on discrimination (C-index was 0.759 [95% confidence interval: 0.739, 0.778] in development cohort and 0.761 [95% confidence interval: 0.731, 0.791] in validation cohort), accuracy, calibration, and outperformed than several existed risk scores. A point-based risk score was built to stratify low- (0-12), intermediate- (13-16), and high-risk group (≥17) among patients.

CONCLUSIONS

A prediction model using readily available predictors was developed and internal validated to predict 1-year mortality risk among patients hospitalized for AHF. It may serve as a useful tool for individual risk stratification and informing decision making to improve clinical care.

A novel cardiac magnetic resonance-based personalized risk stratification model in dilated cardiomyopathy: a prospective study

OBJECTIVES

To explore individual weight of cardiac magnetic resonance (CMR) metrics to predict mid-term outcomes in patients with dilated cardiomyopathy (DCM), and develop a risk algorithm for mid-term outcome based on CMR biomarkers.

MATERIALS AND METHODS

Patients with DCM who underwent CMR imaging were prospectively enrolled in this study. The primary endpoint was a composite of heart failure (HF) death, sudden cardiac death (SCD), aborted SCD, and heart transplantation.

RESULTS

A total of 407 patients (age 48.1 ± 13.8 years, 331 men) were included in the final analysis. During a median follow-up of 21.7 months, 63 patients reached the primary endpoint. NYHA class III/IV (HR = 2.347 [1.073-5.133], p = 0.033), left ventricular ejection fraction (HR = 0.940 [0.909-0.973], p < 0.001), late gadolinium enhancement (LGE) > 0.9% and ≤ 6.6% (HR = 3.559 [1.020-12.412], p = 0.046), LGE > 6.6% (HR = 6.028 [1.814-20.038], p = 0.003), and mean extracellular volume (ECV) fraction ≥ 32.8% (HR = 5.922 [2.566-13.665], p < 0.001) had a significant prognostic association with the primary endpoints (C-statistic: 0.853 [0.810-0.896]). Competing risk regression analyses showed that patients with mean ECV fraction ≥ 32.8%, LGE ≥ 5.9%, global circumferential strain ≥ - 5.6%, or global longitudinal strain ≥ - 7.3% had significantly shorter event-free survival due to HF death and heart transplantation. Patients with mean ECV fraction ≥ 32.8% and LGE ≥ 5.9% had significantly shorter event-free survival due to SCD or aborted SCD.

CONCLUSION

ECV fraction may be the best independently risk factor for the mid-term outcomes in patients with DCM, surpassing LVEF and LGE. LGE has a better prognostic value than other CMR metrics for SCD and aborted SCD. The risk stratification model we developed may be a promising non-invasive tool for decision-making and prognosis.

CLINICAL RELEVANCE STATEMENT

"One-stop" assessment of cardiac function and myocardial characterization using cardiac magnetic resonance might improve risk stratification of patients with DCM. In this prospective study, we propose a novel risk algorithm in DCM including NYHA functional class, LVEF, LGE, and ECV.

KEY POINTS

• The present study explores individual weight of CMR metrics for predicting mid-term outcomes in dilated cardiomyopathy. • We have developed a novel risk algorithm for dilated cardiomyopathy that includes cardiac functional class, ejection fraction, late gadolinium enhancement, and extracellular volume fraction. • Personalized risk model derived by CMR contributes to clinical assessment and individual decision-making.

Prediction of clinical risk assessment and survival in chronic obstructive pulmonary disease with pulmonary hypertension

BACKGROUND

Patients with pulmonary hypertension (PH) and chronic obstructive pulmonary disease (COPD) have an increased risk of disease exacerbation and decreased survival. We aimed to develop and validate a non-invasive nomogram for predicting COPD associated with severe PH and a prognostic nomogram for patients with COPD and concurrent PH (COPD-PH).

METHODS

This study included 535 patients with COPD-PH from six hospitals. A multivariate logistic regression analysis was used to analyse the risk factors for severe PH in patients with COPD and a multivariate Cox regression was used for the prognostic factors of COPD-PH. Performance was assessed using calibration, the area under the receiver operating characteristic curve and decision analysis curves. Kaplan-Meier curves were used for a survival analysis. The nomograms were developed as online network software.

RESULTS

Tricuspid regurgitation velocity, right ventricular diameter, N-terminal pro-brain natriuretic peptide (NT-proBNP), the red blood cell count, New York Heart Association functional class and sex were non-invasive independent variables of severe PH in patients with COPD. These variables were used to construct a risk assessment nomogram with good discrimination. NT-proBNP, mean pulmonary arterial pressure, partial pressure of arterial oxygen, the platelet count and albumin were independent prognostic factors for COPD-PH and were used to create a predictive nomogram of overall survival rates.

CONCLUSIONS

The proposed nomograms based on a large sample size of patients with COPD-PH could be used as non-invasive clinical tools to enhance the risk assessment of severe PH in patients with COPD and for the prognosis of COPD-PH. Additionally, the online network has the potential to provide artificial intelligence-assisted diagnosis and treatment.

HIGHLIGHTS

A multicentre study with a large sample of chronic obstructive pulmonary disease (COPD) patients diagnosed with PH through right heart catheterisation. A non-invasive online clinical tool for assessing severe pulmonary hypertension (PH) in COPD. The first risk assessment tool was established for Chinese patients with COPD-PH.

Deep learning-based multi-modal data integration enhancing breast cancer disease-free survival prediction

Background

The prognosis of breast cancer is often unfavorable, emphasizing the need for early metastasis risk detection and accurate treatment predictions. This study aimed to develop a novel multi-modal deep learning model using preoperative data to predict disease-free survival (DFS).

Methods

We retrospectively collected pathology imaging, molecular and clinical data from The Cancer Genome Atlas and one independent institution in China. We developed a novel Deep Learning Clinical Medicine Based Pathological Gene Multi-modal (DeepClinMed-PGM) model for DFS prediction, integrating clinicopathological data with molecular insights. The patients included the training cohort (n = 741), internal validation cohort (n = 184), and external testing cohort (n = 95).

Result

Integrating multi-modal data into the DeepClinMed-PGM model significantly improved area under the receiver operating characteristic curve (AUC) values. In the training cohort, AUC values for 1-, 3-, and 5-year DFS predictions increased to 0.979, 0.957, and 0.871, while in the external testing cohort, the values reached 0.851, 0.878, and 0.938 for 1-, 2-, and 3-year DFS predictions, respectively. The DeepClinMed-PGM's robust discriminative capabilities were consistently evident across various cohorts, including the training cohort [hazard ratio (HR) 0.027, 95% confidence interval (CI) 0.0016-0.046, P < 0.0001], the internal validation cohort (HR 0.117, 95% CI 0.041-0.334, P < 0.0001), and the external cohort (HR 0.061, 95% CI 0.017-0.218, P < 0.0001). Additionally, the DeepClinMed-PGM model demonstrated C-index values of 0.925, 0.823, and 0.864 within the three cohorts, respectively.

Conclusion

This study introduces an approach to breast cancer prognosis, integrating imaging and molecular and clinical data for enhanced predictive accuracy, offering promise for personalized treatment strategies.

Development and Validation of a Prognostic Model for Postoperative Anastomotic Recurrence in Siewert II or III Adenocarcinomas Without Neoadjuvant Therapy in an East Asian Population

PURPOSE

Anastomotic recurrence leads to poor prognosis in patients with Siewert II or III adenocarcinoma who undergo radical gastrectomy and do not receive neoadjuvant therapy. We aimed to establish a prognostic model to evaluate the risk of postoperative anastomotic recurrence in patients with Siewert II or III adenocarcinoma who did not receive neoadjuvant therapy.

METHODS

We included 366 patients with Siewert II or III adenocarcinoma who were treated with radical gastrectomy without neoadjuvant therapy at Fujian Provincial Hospital (FPH) between 2012 and 2018 as the development cohort. Cox regression was used to verify prognostic factors for anastomotic recurrence, and a nomogram was established. The nomogram was externally validated using a combined cohort of two external centers. Patients were classified into high- or low-risk groups according to the diagnostic threshold and nomogram scores, and recurrence-related survival analysis was analyzed.

RESULTS

The average age was 64.6 years, and 285 patients were male. All surgeries were successfully performed (185 open vs 181 laparoscopic). The 3-year anastomotic recurrence rate was significantly lower in the low-risk group (3.5% vs 18.8%, P < 0.001). The predictive performance was verified in the external validation cohort. This model better stratified patient survival than the American Joint Committee on Cancer (AJCC) TNM staging system.

CONCLUSIONS

This novel nomogram with surgical margin, postoperative tumor node metastasis (pTNM) stage, and neural invasion as prognostic factors has a significant predictive performance for the risk of anastomotic recurrence after radical gastrectomy in patients with Siewert II or III adenocarcinoma.

Prognostic prediction model of colorectal cancer based on preoperative serum tumor markers

BACKGROUND

Preoperative serum tumor markers not only play a role in the auxiliary diagnosis and postoperative monitoring in colorectal cancer (CRC), but also have been found to have potential prognostic value.

AIM

To analyze whether preoperative serum tumor markers, including carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9), affect the prognosis of CRC.

METHODS

This was a retrospective study conducted in a single center. Patients with nonmetastatic CRC who underwent initial surgery between January 2011 and January 2020 were enrolled and divided into development site and validation site groups at a ratio of 7:3. The independent prognostic factors were screened by Cox regression analysis, and finally, a prognostic nomogram model was established. The newly developed model was tested by internal validation.

RESULTS

Eventually, 3526 postoperative patients with nonmetastatic CRC were included in the study. There were 2473 patients at the development site and 1056 patients at the validation site. Age (P < 0.01, HR = 1.042, 95%CI = 1.033-1.051), tumor node metastasis (TNM) classification (P < 0.01, HR = 1.938, 95%CI = 1.665-2.255), preoperative CEA (P = 0.001, HR = 1.393, 95%CI = 1.137-1.707) and CA19-9 (P < 0.01, HR = 1.948, 95%CI = 1.614-2.438) levels were considered independent prognostic factors for patients with nonmetastatic CRC and were used as variables in the nomogram model. The areas under the curve of the development and validation sites were 0.655 and 0.658, respectively. The calibration plot also showed the significant performance of the newly established nomogram.

CONCLUSION

We successfully constructed a nomogram model based on age, TNM stage, preoperative CEA, and CA19-9 levels to evaluate the overall survival of patients with nonmetastatic CRC.

Developing targeted therapies for neuroblastoma by dissecting the effects of metabolic reprogramming on tumor microenvironments and progression

Rationale: Synergic reprogramming of metabolic dominates neuroblastoma (NB) progression. It is of great clinical implications to develop an individualized risk prognostication approach with stratification-guided therapeutic options for NB based on elucidating molecular mechanisms of metabolic reprogramming. Methods: With a machine learning-based multi-step program, the synergic mechanisms of metabolic reprogramming-driven malignant progression of NB were elucidated at single-cell and metabolite flux dimensions. Subsequently, a promising metabolic reprogramming-associated prognostic signature (MPS) and individualized therapeutic approaches based on MPS-stratification were developed and further validated independently using pre-clinical models. Results: MPS-identified MPS-I NB showed significantly higher activity of metabolic reprogramming than MPS-II counterparts. MPS demonstrated improved accuracy compared to current clinical characteristics [AUC: 0.915 vs. 0.657 (MYCN), 0.713 (INSS-stage), and 0.808 (INRG-stratification)] in predicting prognosis. AZD7762 and etoposide were identified as potent therapeutics against MPS-I and II NB, respectively. Subsequent biological tests revealed AZD7762 substantially inhibited growth, migration, and invasion of MPS-I NB cells, more effectively than that of MPS-II cells. Conversely, etoposide had better therapeutic effects on MPS-II NB cells. More encouragingly, AZD7762 and etoposide significantly inhibited in-vivo subcutaneous tumorigenesis, proliferation, and pulmonary metastasis in MPS-I and MPS-II samples, respectively; thereby prolonging survival of tumor-bearing mice. Mechanistically, AZD7762 and etoposide-induced apoptosis of the MPS-I and MPS-II cells, respectively, through mitochondria-dependent pathways; and MPS-I NB resisted etoposide-induced apoptosis by addiction of glutamate metabolism and acetyl coenzyme A. MPS-I NB progression was fueled by multiple metabolic reprogramming-driven factors including multidrug resistance, immunosuppressive and tumor-promoting inflammatory microenvironments. Immunologically, MPS-I NB suppressed immune cells via MIF and THBS signaling pathways. Metabolically, the malignant proliferation of MPS-I NB cells was remarkably supported by reprogrammed glutamate metabolism, tricarboxylic acid cycle, urea cycle, etc. Furthermore, MPS-I NB cells manifested a distinct tumor-promoting developmental lineage and self-communication patterns, as evidenced by enhanced oncogenic signaling pathways activated with development and self-communications. Conclusions: This study provides deep insights into the molecular mechanisms underlying metabolic reprogramming-mediated malignant progression of NB. It also sheds light on developing targeted medications guided by the novel precise risk prognostication approaches, which could contribute to a significantly improved therapeutic strategy for NB.

An exosomes-related lncRNA prognostic model correlates with the immune microenvironment and therapy response in lung adenocarcinoma

Recent research highlights the significance of exosomes and long noncoding RNAs (lncRNAs) in cancer progression and drug resistance, but their role in lung adenocarcinoma (LUAD) is not fully understood. We analyzed 121 exosome-related (ER) mRNAs from the ExoBCD database, along with mRNA and lncRNA expression profiles of TCGA-LUAD using "DESeq2", "survival," "ConsensusClusterPlus," "GSVA," "estimate," "glmnet," "clusterProfiler," "rms," and "pRRophetic" R packages. This comprehensive approach included univariate cox regression, unsupervised consensus clustering, GSEA, functional enrichment analysis, and prognostic model construction. Our study identified 134 differentially expressed ER-lncRNAs, with 19 linked to LUAD prognosis. These ER-lncRNAs delineated two patient subtypes, one with poorer outcomes. Additionally, 286 differentially expressed genes were related to these ER-lncRNAs, 261 of which also correlated with LUAD prognosis. We constructed an ER-lncRNA-related prognostic model and calculated an ER-lncRNA-related risk score (ERS), revealing that a higher ERS correlates with poor overall survival in both the Meta cohort and two validation cohorts. The ERS potentially serves as an independent prognostic factor, and the prognostic model demonstrates superior predictive power. Notably, significant differences in the immune landscape were observed between the high- and low-ERS groups. Drug sensitivity analysis indicated varying responses to common chemotherapy drugs based on ERS stratification, with the high-ERS group showing greater sensitivity, except to rapamycin and erlotinib. Experimental validation confirmed that thymidine kinase 1 enhances lung cancer invasion, metastasis, and cell cycle progression. Our study pioneers an ER-lncRNA-related prognostic model for LUAD, proposing that ERS-based risk stratification could inform personalized treatment strategies to improve patient outcomes.

A nomogram for predicting liver metastasis in patients with gastric gastrointestinal stromal tumor

BACKGROUND

Liver metastasis (LIM) is an important factor in the diagnosis, treatment, follow-up, and prognosis of patients with gastric gastrointestinal stromal tumor (GIST). There is no simple tool to assess the risk of LIM in patients with gastric GIST. Our aim was to develop and validate a nomogram to identify patients with gastric GIST at high risk of LIM.

METHODS

Patient data diagnosed as having gastric GIST between 2010 and 2019 were extracted from the Surveillance, Epidemiology, and End Results (SEER) database and randomly divided into training cohort and internal validation cohort in a 7:3 ratio. For external validation, retrospective data collection was performed on patients diagnosed as having gastric GIST at Yunnan Cancer Center (YNCC) between January 2015 and May 2023. Univariate and multivariate logistic regression analyses were used to identify independent risk factors associated with LIM in patients with gastric GIST. An individualized LIM nomogram specific for gastric GIST was formulated based on the multivariate logistic model; its discriminative performance, calibration, and clinical utility were evaluated.

RESULTS

In the SEER database, a cohort of 2341 patients with gastric GIST was analyzed, of which 173 cases (7.39%) were found to have LIM; 239 patients with gastric GIST from the YNCC database were included, of which 25 (10.46%) had LIM. Multivariate analysis showed tumor size, tumor site, and sex were independent risk factors for LIM (P < .05). The nomogram based on the basic clinical characteristics of tumor size, tumor site, sex, and age demonstrated significant discrimination, with an area under the curve of 0.753 (95% CI, 0.692-0.814) and 0.836 (95% CI, 0.743-0.930) in the internal and external validation cohort, respectively. The Hosmer-Lemeshow test showed that the nomogram was well calibrated, whereas the decision curve analysis and the clinical impact plot demonstrated its clinical utility.

CONCLUSION

Tumor size, tumor subsite, and sex were significantly correlated with the risk of LIM in gastric GIST. The nomogram for patients with GIST can effectively predict the individualized risk of LIM and contribute to the planning and decision making related to metastasis management in clinical practice.

A novel anoikis-related gene signature can predict the prognosis of hepatocarcinoma patients

Background

Hepatocellular carcinoma (HCC) is a major health problem with more than 850,000 cases per year worldwide. This cancer is now the third leading cause of cancer-related deaths worldwide, and the number is rising. Cancer cells develop anoikis resistance which is a vital step during cancer progression and metastatic colonization. However, there is not much research that specifically addresses the role of anoikis in HCC, especially in terms of prognosis.

Methods

This study obtained gene expression data and clinical information from 371 HCC patients through The Cancer Genome Atlas (TCGA) Program and The Gene Expression Omnibus (GEO) databases. A total of 516 anoikis-related genes (ANRGs) were retrieved from GeneCard database and Harmonizome portal. Differential expression analysis identified 219 differentially expressed genes (DEGs), and univariate Cox regression analysis was utilized to select 99 ANRGs associated with the prognosis of HCC patients. A risk scoring model with seven genes was established using the least absolute shrinkage and selection operator (LASSO) regression model, and internal validation of the model was performed.

Results

The identified 99 ANRGs are closely associated with the prognosis of HCC patients. The risk scoring model based on seven characteristic genes demonstrates excellent predictive performance, further validated by receiver operating characteristic (ROC) curves and Kaplan-Meier survival curves. The study reveals significant differences in immune cell infiltration, gene expression, and survival status among different risk groups.

Conclusions

The prognosis of HCC patients can be predicted using a unique prognostic model built on ANRGs in HCC.

Clinical Decision Analysis of Genetic Evaluation and Testing in 1013 Intensive Care Unit Infants with Congenital Heart Defects Supports Universal Genetic Testing

Extracardiac anomalies (ECAs) are strong predictors of genetic disorders in infants with congenital heart disease (CHD), but there are no prior studies assessing performance of ECA status as a screen for genetic diagnoses in CHD patients. This retrospective cohort study assessed this in our comprehensive inpatient CHD genetics service focusing on neonates and infants admitted to the intensive care unit (ICU). The performance and diagnostic utility of using ECA status to screen for genetic disorders was assessed using decision curve analysis, a statistical tool to assess clinical utility, determining the threshold of phenotypic screening by ECA versus a Test-All approach. Over 24% of infants had genetic diagnoses identified (n = 244/1013), and ECA-positive status indicated a 4-fold increased risk of having a genetic disorder. However, ECA status had low-moderate screening performance based on predictive summary index, a compositive measure of positive and negative predictive values. For those with genetic diagnoses, nearly one-third (32%, 78/244) were ECA-negative but had cytogenetic and/or monogenic disorders identified by genetic testing. Thus, if the presence of multiple congenital anomalies is the phenotypic driver to initiate genetic testing, 13.4% (78/580) of infants with isolated CHD with identifiable genetic causes will be missed. Given the prevalence of genetic disorders and limited screening performance of ECA status, this analysis supports genetic testing in all CHD infants in intensive care settings rather than screening based on ECA.

Deciphering the prognostic features of bladder cancer through gemcitabine resistance and immune-related gene analysis and identifying potential small molecular drug PIK-75

BACKGROUND

Bladder cancer (BCa) stands out as a prevalent and highly lethal malignancy worldwide. Chemoresistance significantly contributes to cancer recurrence and progression. Traditional Tumor Node Metastasis (TNM) stage and molecular subtypes often fail to promptly identify treatment preferences based on sensitivity.

METHODS

In this study, we developed a prognostic signature for BCa with uni-Cox + LASSO + multi-Cox survival analysis in multiple independent cohorts. Six machine learning algorithms were adopted to screen out the hub gene, RAC3. IHC staining was used to validate the expression of RAC3 in BCa tumor tissue. RT-qPCR and Western blot were performed to detect and quantify the mRNA and protein levels of RAC3. CCK8, colony formation, wound healing, and flow cytometry analysis of apoptosis were employed to determine cell proliferation, migration, and apoptosis. Molecular docking was used to find small target drugs, PIK-75. 3D cell viability assay was applied to evaluate the ATP viability of bladder cancer organoids before and after PIK-75 treated.

RESULTS

The established clinical prognostic model, GIRS, comprises 13 genes associated with gemcitabine resistance and immunology. This model has demonstrated robust predictive capabilities for survival outcomes across various independent public cohorts. Additionally, the GIRS signature shows significant correlations with responses to both immunotherapy and chemotherapy. Leveraging machine learning algorithms, the hub gene, RAC3, was identified, and potential upstream transcription factors were screened through database analysis. IHC results showed that RAC3 was higher expressed in GEM-resistant BCa patients. Employing molecular docking, the small molecule drug PIK-75, as binding to RAC3, was identified. Experiments on cell lines, organoids and animals validated the biological effects of PIK-75 in bladder cancer.

CONCLUSIONS

The GIRS signature offers a valuable complement to the conventional anatomic TNM staging system and molecular subtype stratification in bladder cancer. The hub gene, RAC3, plays a crucial role in BCa and is significantly associated with resistance to gemcitabine. The small molecular drug, PIK-75 having the potential as a therapeutic agent in the context of gemcitabine-resistant and immune-related pathways.

Non-Invasive Diagnosis of Prostate Cancer and High-Grade Prostate Cancer Using Multiparametric Ultrasonography and Serological Examination

OBJECTIVES

This study aimed to assess the efficacy of multiparametric ultrasonography (mpUS) combined with serological examination, as a non-invasive method, in detecting prostate cancer (PCa) or high-grade prostate cancer (HGPCa) respectively.

METHODS

A cohort of 245 individuals with clinically suspected PCa were enrolled. All subjects underwent a comprehensive evaluation, including basic data collection, serological testing, mpUS and prostate biopsy. Random Forest (RF) models were developed, and the mean area under the curve (AUC) in 100 cross-validations was used to assess the performance in distinguishing PCa from HGPCa.

RESULTS

mpUS features showed significant differences (p < 0.001) between the PCa and non-PCa groups, as well as between the HGPCa and low-grade prostate cancer (LGPCa) groups including prostate-specific antigen density (PSAD), transrectal real-time elastography (TRTE) and intensity difference (ID). The RF model, based on these features, demonstrated an excellent discriminative ability for PCa with a mean area under the curve (AUC) of 0.896. Additionally, another model incorporating free prostate-specific antigen (FPSA) and color Doppler flow imaging (CDFI) achieved a high accuracy in predicting HGPCa with a mean AUC of 0.830. The nomogram derived from these models exhibited excellent individualized prediction of PCa and HGPCa.

CONCLUSION

The RF models incorporating mpUS and serological variables achieved satisfactory accuracies in predicting PCa and HGPCa.

Clinical characteristics and MRI based radiomics nomograms can predict iPFS and short-term efficacy of third-generation EGFR-TKI in EGFR-mutated lung adenocarcinoma with brain metastases

BACKGROUND

Predicting short-term efficacy and intracranial progression-free survival (iPFS) in epidermal growth factor receptor gene mutated (EGFR-mutated) lung adenocarcinoma patients with brain metastases who receive third-generation epidermal growth factor receptor tyrosine kinase inhibitor (EGFR-TKI) therapy was of great significance for individualized treatment. We aimed to construct and validate nomograms based on clinical characteristics and magnetic resonance imaging (MRI) radiomics for predicting short-term efficacy and intracranial progression free survival (iPFS) of third-generation EGFR-TKI in EGFR-mutated lung adenocarcinoma patients with brain metastases.

METHODS

One hundred ninety-four EGFR-mutated lung adenocarcinoma patients with brain metastases who received third-generation EGFR-TKI treatment were included in this study from January 1, 2017 to March 1, 2023. Patients were randomly divided into training cohort and validation cohort in a ratio of 5:3. Radiomics features extracted from brain MRI were screened by least absolute shrinkage and selection operator (LASSO) regression. Logistic regression analysis and Cox proportional hazards regression analysis were used to screen clinical risk factors. Single clinical (C), single radiomics (R), and combined (C + R) nomograms were constructed in short-term efficacy predicting model and iPFS predicting model, respectively. Prediction effectiveness of nomograms were evaluated by calibration curves, Harrell's concordance index (C-index), receiver operating characteristic (ROC) curves and decision curve analysis (DCA). Kaplan-Meier analysis was used to compare the iPFS of high and low iPFS rad-score patients in the predictive iPFS R model and to compare the iPFS of high-risk and low-risk patients in the predictive iPFS C + R model.

RESULTS

Overall response rate (ORR) was 71.1%, disease control rate (DCR) was 91.8% and median iPFS was 12.67 months (7.88-20.26, interquartile range [IQR]). There were significant differences in iPFS between patients with high and low iPFS rad-scores, as well as between high-risk and low-risk patients. In short-term efficacy model, the C-indexes of C + R nomograms in training cohort and validation cohort were 0.867 (0.835-0.900, 95%CI) and 0.803 (0.753-0.854, 95%CI), while in iPFS model, the C-indexes were 0.901 (0.874-0.929, 95%CI) and 0.753 (0.713-0.793, 95%CI).

CONCLUSIONS

The third-generation EGFR-TKI showed significant efficacy in EGFR-mutated lung adenocarcinoma patients with brain metastases, and the combined line plot of C + R can be utilized to predict short-term efficacy and iPFS.

Machine learning-based radiomics nomograms to predict number of fields in postoperative IMRT for breast cancer

BACKGROUND

Breast cancer is now the most commonly diagnosed cancer in women worldwide. Radiotherapy is an important part of the treatment for breast cancer, while setting proper number of fields dramatically affects the benefits one can receive. Machine learning and radiomics have been widely investigated in the management of breast cancer. This study aims to provide models to predict the best number of fields based on machine learning and improve the prediction performance by adding clinical factors.

METHODS

Two-hundred forty-two breast cancer patients were retrospectively enrolled for this study, all of whom received postoperative intensity modulated radiation therapy. The patients were randomized into a training set and a validation set at a ratio of 7:3. Radiomics shape features were extracted for eight machine learning algorithms to predict the number of fields. Univariate and multivariable logistic regression were implemented to screen clinical factors. A combined model of rad-score and clinical factors were finally constructed. The area under receiver operating characteristic curve, precision, recall, F1 measure and accuracy were used to evaluate the model.

RESULTS

Random Forest outperformed from eight machine learning algorithms while predicting the number of fields. Prediction performance of the radiomics model was better than the clinical model, while the predictive nomogram combining the rad-score and clinical factors performed the best.

CONCLUSIONS

The model combining rad-score and clinical factors performed the best. Nomograms constructed from the combined models can be of reliable references for medical dosimetrists.

A nomogram based on clinical factors and CT radiomics for predicting anti-MDA5+ DM complicated by RP-ILD

OBJECTIVES

Anti-melanoma differentiation-associated gene 5 antibody-positive (anti-MDA5+) DM complicated by rapidly progressive interstitial lung disease (RP-ILD) has a high incidence and poor prognosis. The objective of this study was to establish a model for the prediction and early diagnosis of anti-MDA5+ DM-associated RP-ILD based on clinical manifestations and imaging features.

METHODS

A total of 103 patients with anti-MDA5+ DM were included. The patients were randomly split into training and testing sets of 72 and 31 patients, respectively. After image analysis, we collected clinical, imaging and radiomics features from each patient. Feature selection was performed first with the minimum redundancy and maximum relevance algorithm and then with the best subset selection method. The final remaining features comprised the radscore. A clinical model and imaging model were then constructed with the selected independent risk factors for the prediction of non-RP-ILD and RP-ILD. We also combined these models in different ways and compared their predictive abilities. A nomogram was also established. The predictive performances of the models were assessed based on receiver operating characteristics curves, calibration curves, discriminability and clinical utility.

RESULTS

The analyses showed that two clinical factors, dyspnoea (P = 0.000) and duration of illness in months (P = 0.001), and three radiomics features (P = 0.001, 0.044 and 0.008, separately) were independent predictors of non-RP-ILD and RP-ILD. However, no imaging features were significantly different between the two groups. The radiomics model built with the three radiomics features performed worse than the clinical model and showed areas under the curve (AUCs) of 0.805 and 0.754 in the training and test sets, respectively. The clinical model demonstrated a good predictive ability for RP-ILD in MDA5+ DM patients, with an AUC, sensitivity, specificity and accuracy of 0.954, 0.931, 0.837 and 0.847 in the training set and 0.890, 0.875, 0.800 and 0.774 in the testing set, respectively. The combination model built with clinical and radiomics features performed slightly better than the clinical model, with an AUC, sensitivity, specificity and accuracy of 0.994, 0.966, 0.977 and 0.931 in the training set and 0.890, 0.812, 1.000 and 0.839 in the testing set, respectively. The calibration curve and decision curve analyses showed satisfactory consistency and clinical utility of the nomogram.

CONCLUSION

Our results suggest that the combination model built with clinical and radiomics features could reliably predict the occurrence of RP-ILD in MDA5+ DM patients.

Nomogram based on CMTM6 expression and clinical characteristics to predict postoperative overall survival in patients with hepatocellular carcinoma

BACKGROUND

The purpose of this study was to investigate the expression of CMTM6 in HCC tissues and its prognostic value, and to try to develop a nomogram prognostic model based on CMTM6.

METHODS

In this retrospective study, immunohistochemical (IHC) staining was performed in 178 patients who underwent radical hepatectomy in the same surgical team. R software was used to construct the nomogram model. The Bootstrap sampling method was used for internal validation.

RESULTS

CMTM6 is significantly expressed in HCC tissues and is closely associated with decreased overall survival (OS). PVTT (HR = 6.2, 95% CI: 3.06 12.6, P<0.001), CMTM6 (HR=2.30, 95% CI: 1.27 4.0, P=0.006) and MVI (HR=10.8, 95% CI: 4.19-27.6, P<0.001) were independent predictors of OS. The nomogram combined with CMTM6, PVTT and MVI was more predictive than the traditional TNM scoring system, and the prediction effects of 1-year and 3-year OS were accurate.

CONCLUSIONS

The prognosis of a patient may be predicted using high levels of CMTM6 expression in HCC tissues, and the nomogram model including CMTM6 expression has the best predictive ability.

Infiltrating characteristics and prognostic value of tertiary lymphoid structures in resected gastric neuroendocrine neoplasm patients

Objectives

Tertiary lymphoid structures (TLSs) are lymphocyte aggregates that play an anti-tumor role in most solid tumors. However, the functions of TLS in gastric neuroendocrine neoplasms (GNENs) remain unknown. This study aimed to determine the characteristics and prognostic values of TLS in resected GNEN patients.

Methods

Haematoxylin-eosin, immunohistochemistry (IHC) and multiple fluorescent IHC staining were used to assess TLS to investigate the correlation between TLSs and clinicopathological characteristics and its prognostic value.

Results

Tertiary lymphoid structures were identified in 84.3% of patients with GNEN. They were located in the stromal area or outside the tumor tissue and mainly composed of B and T cells. A high density of TLSs promoted an anti-tumor immune response in GNEN. CD15+ TANs and FOXP3+ Tregs in TLSs inhibited the formation of TLSs. High TLS density was significantly associated with prolonged recurrence-free survival (RFS) and overall survival (OS) of GNENs. Univariate and multivariate Cox regression analyses revealed that TLS density, tumor size, tumor-node-metastasis (TNM) stage and World Health Organisation (WHO) classification were independent prognostic factors for OS, whereas TLS density, tumor size and TNM stage were independent prognostic factors for RFS. Finally, OS and RFS nomograms were developed and validated, which were superior to the WHO classification and the TNM stage.

Conclusion

Tertiary lymphoid structures were mainly located in the stromal area or outside the tumor area, and high TLS density was significantly associated with the good prognosis of patients with GNEN. Incorporating TLS density into a nomogram may improve survival prediction in patients with resected GNEN.

Development and Evaluation of a Nomogram for Predicting the Outcome of Immune Reconstitution Among HIV/AIDS Patients Receiving Antiretroviral Therapy in China

This study aims to develop and evaluate a model to predict the immune reconstitution among HIV/AIDS patients after antiretroviral therapy (ART). A total of 502 HIV/AIDS patients are randomized to the training cohort and evaluation cohort. Least absolute shrinkage and selection operator (LASSO) regression and multivariate logistic regression analysis are performed to identify the indicators and establish the nomogram for predicting the immune reconstitution. Decision curve analysis (DCA) and clinical impact curve (CIC) are used to evaluate the clinical effectiveness of the nomogram. Predictive factors included white blood cells (WBC), baseline CD4+ T-cell counts (baseline CD4), ratio of effector regulatory T cells to resting regulatory T cells (eTreg/rTreg) and low-density lipoprotein cholesterol (LDL-C) and are incorporated into the nomogram. The area under the curve (AUC) is 0.812 (95% CI, 0.767∼0.851) and 0.794 (95%CI, 0.719∼0.857) in the training cohort and evaluation cohort, respectively. The calibration curve shows a high consistency between the predicted and actual observations. Moreover, DCA and CIC indicate that the nomogram has a superior net benefit in predicting poor immune reconstitution. A simple-to-use nomogram containing four routinely collected variables is developed and internally evaluated and can be used to predict the poor immune reconstitution in HIV/AIDS patients after ART.

An online-predictive model of acute kidney injury after pancreatic surgery

OBJECTIVE

Acute kidney injury(AKI) after pancreatic surgery is associated with increased mortality, longer hospital stays and poor prognosis. This study aims to identify the risk factors and establish an easy-to-use prediction calculator by the nomogram to predict the risk of AKI after pancreatic surgery.

METHODS

From January 2016 to June 2018, 1504 patients who underwent pancreatic surgery in our center were included in this retrospective analysis and randomly assigned to primary (1054 patients) and validation (450 patients) cohorts. The independent risk factors of AKI were identified using univariate and multivariate analyses. A risk-predicted nomogram for AKI was developed through multivariate logistic regression analysis in the primary cohort while the nomogram was evaluated in the validation cohort. Nomogram discrimination and calibration were assessed using C-index and calibration curves in the primary and validation cohorts. The clinical utility of the final nomogram was evaluated using decision curve analysis.

RESULTS

The overall incidence of AKI after pancreatic surgery was 5.3% (79/1504). Independent risk factors including smoking history, cardiovascular disease, ASA score, baseline eGFR, bilirubin＞2 ​mg/dL, undergoing pancreaticoduodenectomy, and intraoperative blood loss＞400 ​mL were identified by multivariate analysis. Nomogram revealed moderate discrimination and calibration in estimating the risk of AKI, with an unadjusted C-index of 0.79 (95 %CI, 0.73-0.85). Application of the nomogram in the validation cohort provided moderate discrimination (C-index,0.80 [95% CI, 0.72-0.88]) and good calibration. Besides, the decision curve analysis (DCA) confirmed the clinical usefulness of the nomogram.

CONCLUSIONS

An easy-to-use online prediction calculator comprised of preoperative and intraoperative factors was able to individually predict the occurrence risk of AKI among patients with pancreatic surgery, which may help identify reasonable risk judgments and develop proper treatment strategies to a certain extent.

Development of a machine learning-based model to predict hepatic inflammation in chronic hepatitis B patients with concurrent hepatic steatosis: a cohort study

Background

With increasingly prevalent coexistence of chronic hepatitis B (CHB) and hepatic steatosis (HS), simple, non-invasive diagnostic methods to accurately assess the severity of hepatic inflammation are needed. We aimed to build a machine learning (ML) based model to detect hepatic inflammation in patients with CHB and concurrent HS.

Methods

We conducted a multicenter, retrospective cohort study in China. Treatment-naive CHB patients with biopsy-proven HS between April 2004 and September 2022 were included. The optimal features for model development were selected by SHapley Additive explanations, and an ML algorithm with the best accuracy to diagnose moderate to severe hepatic inflammation (Scheuer's system ≥ G3) was determined and assessed by decision curve analysis (DCA) and calibration curve. This study is registered with ClinicalTrials.gov (NCT05766449).

Findings

From a pool of 1,787 treatment-naive patients with CHB and HS across eleven hospitals, 689 patients from nine of these hospitals were chosen for the development of the diagnostic model. The remaining two hospitals contributed to two independent external validation cohorts, comprising 509 patients in validation cohort 1 and 589 in validation cohort 2. Eleven features regarding inflammation, hepatic and metabolic functions were identified. The gradient boosting classifier (GBC) model showed the best performance in predicting moderate to severe hepatic inflammation, with an area under the receiver operating characteristic curve (AUROC) of 0.86 (95% CI 0.83-0.88) in the training cohort, and 0.89 (95% CI 0.86-0.92), 0.76 (95% CI 0.73-0.80) in the first and second external validation cohorts, respectively. A publicly accessible web tool was generated for the model.

Interpretation

Using simple parameters, the GBC model predicted hepatic inflammation in CHB patients with concurrent HS. It holds promise for guiding clinical management and improving patient outcomes.

Funding

This research was supported by the National Natural Science Foundation of China (No. 82170609, 81970545), Natural Science Foundation of Shandong Province (Major Project) (No. ZR2020KH006), Natural Science Foundation of Jiangsu Province (No.BK20231118), Tianjin Key Medical Discipline (Specialty), Construction Project, TJYXZDXK-059B, Tianjin Health Science and Technology Project key discipline special, TJWJ2022XK034, and Research project of Chinese traditional medicine and Chinese traditional medicine combined with Western medicine of Tianjin municipal health and Family Planning Commission (2021022).

Brain connectomics improve prediction of 1-year decreased quality of life in breast cancer: A multi-voxel pattern analysis

PURPOSE

Whether brain connectomics can predict 1-year decreased Quality of Life (QoL) in patients with breast cancer are unclear. A longitudinal study was utilized to explore their prediction abilities with a multi-center sample.

METHODS

232 breast cancer patients were consecutively enrolled and 214 completed the 1-year QoL assessment (92.2%). Resting state functional magnetic resonance imaging was collected before the treatment and a multivoxel pattern analysis (MVPA) was performed to differentiate whole-brain resting-state connectivity patterns. Net Reclassification Improvement (NRI) as well as Integrated Discrimination Improvement (IDI) were calculated to estimate the incremental value of brain connectomics over conventional risk factors.

RESULTS

Paracingulate Gyrus, Superior Frontal Gyrus and Frontal Pole were three significant brain areas. Brain connectomics yielded 7.8-17.2% of AUC improvement in predicting 1-year decreased QoL. The NRI and IDI ranged from 20.27 to 54.05%, 13.21-33.34% respectively.

CONCLUSION

Brain connectomics contribute to a more accurate prediction of 1-year decreased QoL in breast cancer. Significant brain areas in the prefrontal lobe could be used as potential intervention targets (i.e., Cognitive Behavioral Group Therapy) to improve long-term QoL outcomes in breast cancer.

Prediction of Coronary Artery Lesions in Patients With Recurrent Kawasaki Disease

BACKGROUND

A subset of patients with Kawasaki disease (KD) will suffer recurrence. However, there is still a lack of accurate prediction models for coronary artery lesions (CAL) in recurrent KD patients. It is necessary to establish a new nomogram model for predicting CAL in patients with recurrent KD.

METHODS

Data from patients with recurrent KD between 2015 and 2021 were retrospectively reviewed. After splitting the patients into training and validation cohorts, the least absolute shrinkage and selection operator was used to select the predictors of CAL and multivariate logistic regression was used to construct a nomogram based on the selected predictors. The application of area under the receiver operating characteristic curve (AUC), calibration curves, Hosmer-Lemeshow test, Brier score and decision curve analysis were used to assess the model performance.

RESULTS

A total of 159 recurrent KD patients were enrolled, 66 (41.5%) of whom had CAL. Hemoglobin levels, CAL at the first episode, and intravenous immunoglobulin resistance at recurrence were identified by the least absolute shrinkage and selection operator regression analysis as significant predictors. The model incorporating these predictors showed good discrimination (AUC, 0.777) and calibration capacities (Hosmer-Lemeshow P value, 0.418; Brier score, 0.190) in the training cohort. Application of the model to the validation cohort yielded an AUC of 0.741, a Hosmer-Lemeshow P value of 0.623 and a Brier score of 0.190. The decision curve analysis demonstrated that the nomogram model was clinically useful.

CONCLUSIONS

The proposed nomogram model could help clinicians assess the risk of CAL in patients with recurrent KD.

Development and validation of a risk model for intracardiac thrombosis in patients with dilated cardiomyopathy: a retrospective study

Intracardiac thrombosis is a severe complication in patients with non-ischemic dilated cardiomyopathy. This study aims to develop and validate an individualized nomogram to evaluate the risk of intracardiac thrombosis in patients with non-ischemic dilated cardiomyopathy. This retrospective study included patients diagnosed with dilated cardiomyopathy at first admission. Clinical baseline characteristics were acquired from electronic medical record systems. Multiple methods were applied to screen the key variables and generate multiple different variable combinations. Multivariable logistic regression was used to build the models, and the optimal model was chosen by comparing the discrimination. Then we checked the performance of the model in different thrombus subgroups. Finally, the model was presented using a nomogram and evaluated from the perspectives of discrimination, calibration, and clinical usefulness. Internal validation was performed by extracting different proportions of data for Bootstrapping. Ultimately, 564 eligible patients were enrolled, 67 of whom developed an intracardiac thrombosis. Risk factors included d-dimer, white blood cell count, high-sensitivity C-reactive protein, pulse pressure, history of stroke, hematocrit, and NT-proBNP in the optimal model. The model had good discrimination and calibration, and the area under the curve (AUC) was 0.833 (0.782-0.884), and the model's performance in each subgroup was stable. Clinical decision curve analysis showed that the model had clinical application value when the high-risk threshold was between 2% and 78%. The AUC of interval validation (30% and 70% data resampling) was 0.844 (0.765-0.924) and 0.833 (0.775-0.891), respectively. This novel intracardiac thrombosis nomogram could be conveniently applied to facilitate the individual intracardiac thrombosis risk assessment in patients with non-ischemic dilated cardiomyopathy.

Development and validation of a novel combinatorial nomogram model to predict in-hospital deaths in heart failure patients

BACKGROUND

The purpose of this study was to develop a Nomogram model to identify the risk of all-cause mortality during hospitalization in patients with heart failure (HF).

METHODS

HF patients who had been registered in the Medical Information Mart for Intensive Care (MIMIC) III and IV databases were included. The primary outcome was the occurrence of all-cause mortality during hospitalization. Two Logistic Regression models (LR1 and LR2) were developed to predict in-hospital death for HF patients from the MIMIC-IV database. The MIMIC-III database were used for model validation. The area under the receiver operating characteristic curve (AUC) was used to compare the discrimination of each model. Calibration curve was used to assess the fit of each developed models. Decision curve analysis (DCA) was used to estimate the net benefit of the predictive model.

RESULTS

A total of 16,908 HF patients were finally enrolled through screening, of whom 2,283 (13.5%) presented with in-hospital death. Totally, 48 variables were included and analyzed in the univariate and multifactorial regression analysis. The AUCs for the LR1 and LR2 models in the test cohort were 0.751 (95% CI: 0.735∼0.767) and 0.766 (95% CI: 0.751-0.781), respectively. Both LR models performed well in the calibration curve and DCA process. Nomogram and online risk assessment system were used as visualization of predictive models.

CONCLUSION

A new risk prediction tool and an online risk assessment system were developed to predict mortality in HF patients, which performed well and might be used to guide clinical practice.

Pleural fluid carbohydrate antigen 72-4 and malignant pleural effusion: a diagnostic test accuracy study

BACKGROUND

The prognosis of malignant pleural effusion (MPE) is poor. A timely and accurate diagnosis is the prerequisite for managing MPE patients. Carbohydrate antigen 72-4 (CA72-4) is a diagnostic tool for MPE.

OBJECTIVE

We aimed to evaluate the diagnostic accuracy of pleural fluid CA72-4 for MPE.

DESIGN

A prospective, preregistered, and double-blind diagnostic test accuracy study.

METHODS

We prospectively enrolled participants with undiagnosed pleural effusions from two centers in China (Hohhot and Changshu). CA72-4 concentration in pleural fluid was measured by electrochemiluminescence. Its diagnostic accuracy for MPE was evaluated by a receiver operating characteristic (ROC) curve. The net benefit of CA72-4 was determined by a decision curve analysis (DCA).

RESULTS

In all, 153 participants were enrolled in the Hohhot cohort, and 58 were enrolled in the Changshu cohort. In both cohorts, MPE patients had significantly higher CA72-4 levels than benign pleural effusion (BPE) patients. At a cutoff value of 8 U/mL, pleural fluid CA72-4 had a sensitivity, specificity, and area under the ROC curve (AUC) of 0.46, 1.00, and 0.79, respectively, in the Hohhot cohort. In the Changshu cohort, CA72-4 had a sensitivity, specificity, and AUC of 0.27, 0.94, and 0.86, respectively. DCA revealed the relatively high net benefit of CA72-4 determination. In patients with negative cytology, the AUC of CA72-4 was 0.67.

CONCLUSION

Pleural fluid CA72-4 helps differentiate MPE and BPE in patients with undiagnosed pleural effusions.

Development of a model to predict the risk of early graft failure after adult-to-adult living donor liver transplantation: An ELTR study

Graft survival is a critical end point in adult-to-adult living donor liver transplantation (ALDLT), where graft procurement endangers the lives of healthy individuals. Therefore, ALDLT must be responsibly performed in the perspective of a positive harm-to-benefit ratio. This study aimed to develop a risk prediction model for early (3 months) graft failure (EGF) following ALDLT. Donor and recipient factors associated with EGF in ALDLT were studied using data from the European Liver Transplant Registry. An artificial neural network classification algorithm was trained on a set of 2073 ALDLTs, validated using cross-validation, tested on an independent random-split sample (n=518), and externally validated on United Network for Organ Sharing Standard Transplant Analysis and Research data. Model performance was assessed using the AUC, calibration plots, and decision curve analysis. Graft type, graft weight, level of hospitalization, and the severity of liver disease were associated with EGF. The model ( http://ldlt.shinyapps.io/eltr_app ) presented AUC values at cross-validation, in the independent test set, and at external validation of 0.69, 0.70, and 0.68, respectively. Model calibration was fair. The decision curve analysis indicated a positive net benefit of the model, with an estimated net reduction of 5-15 EGF per 100 ALDLTs. Estimated risks>40% and<5% had a specificity of 0.96 and sensitivity of 0.99 in predicting and excluding EGF, respectively. The model also stratified long-term graft survival ( p <0.001), which ranged from 87% in the low-risk group to 60% in the high-risk group. In conclusion, based on a panel of donor and recipient variables, an artificial neural network can contribute to decision-making in ALDLT by predicting EGF risk.

The role of carotid artery stenosis in predicting stroke after coronary artery bypass grafting in a Chinese cohort study

Current guidelines give priority to surgical treatment of carotid artery stenosis (CAS) before coronary artery bypass grafting (CABG), especially in symptomatic patients. Carotid artery stenting is an alternative treatment for narrowing of the carotid arteries. This study sought to demonstrate the role of severe CAS in predicting stroke after CABG and assess the efficacy of carotid artery stenting in preventing postoperative stroke in a Chinese cohort. From 2015 to 2021, 1799 consecutive patients undergoing isolated CABG surgery were retrospectively recruited in a Chinese cohort. The predictive value of severe CAS in postoperative stroke and carotid stenting in preventing postoperative stroke was statistically analyzed. The incidence of postoperative stroke was 1.67%. The incidence of CAS with stenosis ≥ 50% and ≥ 70% was 19.2% and 6.9%. After propensity matching, the incidence of stroke was 8.0% in the severe CAS group and 0% in the non-severe CAS group. We successfully established an optimal predictive nomogram for predicting severe CAS in patients undergoing CABG. Carotid artery stenting was found ineffective in preventing postoperative stroke. The present study provides the incidence of CAS and postoperative stroke in a Chinese cohort, identifies severe CAS as an independent risk factor for postoperative stroke after CABG, constructs a nomogram predicting the incidence of severe CAS, and evaluates the effectiveness of carotid artery stenting in preventing postoperative stroke after CABG.

Estimating surgical probability: Development and validation of a prognostic model for patients with lumbar disc herniation treated with acupuncture

Lumbar disc herniation (LDH) is a common cause of pain in the lumbar spine and legs. While acupuncture has become the primary conservative treatment for LDH, some patients experience treatment failure and require surgery, causing substantial concern for clinicians. We developed an effective personalized clinical prediction model to identify the independent risk factors associated with acupuncture failure in patients with LDH. Our model aimed to predict the probability of surgery within 6 months of acupuncture failure in patients with LDH. A total of 738 patients with LDH who underwent acupuncture at 4 Chinese hospitals between January 2019 and October 2021 were selected. The patients were divided into training (n = 496) and validation (n = 242) cohorts. Seven predictive variables, including smoking, Oswestry Disability Index (ODI) score, lower-limb herniation, disc herniation type, lumbar spinal stenosis, lumbar lateral recess stenosis, and acupuncture frequency, were selected as risk factors using least absolute shrinkage and selection operato (LASSO) regression. A prediction model was developed using multivariate logistic regression analysis and a nomogram was constructed. The model exhibited good discrimination, with an area under the ROC curve (AUC) of 0.903 for the development cohort and 0.899 for the validation cohort. The Hosmer-Lemeshow goodness-of-fit test was a good fit for both cohorts (P = .956 for the development cohort; P = .513 for the validation cohort). Decision curve analysis (DCA) demonstrated that the threshold probabilities for the 2 cohorts ranged from > 4% and 5-95%, respectively. Therefore, the prediction model had a good net benefit. The nomogram established in this study, incorporating 7 risk factors, demonstrated a good predictive ability. It could predict acupuncture failure in LDH patients and the risk of surgery within 6 months, enabling physicians to conduct individualized treatment measures.

In-Hospital Fall Risk Prediction by Objective Measurement of Lower Extremity Function in a High-Risk Population

OBJECTIVES

Limited data exist regarding association between physical performance and in-hospital falls. This study was performed to investigate the association between physical performance and in-hospital falls in a high-risk population.

DESIGN

Retrospective cohort study.

SETTING AND PARTICIPANTS

The study population consisted of 1200 consecutive patients with a median age of 74 years (50.8% men) admitted to a ward with high incidence rates of falls, primarily in the departments of geriatrics and neurology, in a university hospital between January 2019 and December 2021.

METHODS

Short Physical Performance Battery (SPPB) was measured after treatment in the acute phase. As the primary end point of the study, the incidence of in-hospital falls was examined prospectively based on data from mandatory standardized incident report forms and electronic patient records.

RESULTS

SPPB assessment was performed at a median of 3 days after admission, and the study population had a median SPPB score of 3 points. Falls occurred in 101 patients (8.4%) over a median hospital stay of 15 days. SPPB score showed a significant inverse association with the incidence of in-hospital falls after adjusting for possible confounders (adjusted odds ratio for each 1-point decrease in SPPB: 1.19, 95% CI 1.10-1.28; P < .001), and an SPPB score ≤6 was significantly associated with increased risk of in-hospital falls. Inclusion of SPPB with previously identified risk factors significantly increased the area under the curve for in-hospital falls (0.683 vs. 0.740, P = .003).

CONCLUSION AND IMPLICATIONS

This study demonstrated an inverse association of SPPB score with risk of in-hospital falls in a high-risk population and showed that SPPB assessment is useful for accurate risk stratification in a hospital setting.

Comparison of Prognostic Models for Functional Outcome in Aneurysmal Subarachnoid Hemorrhage Based on Machine Learning

BACKGROUND

Controversy exists regarding the superiority of the performance of prognostic tools based on advanced machine learning (ML) algorithms for patients with aneurysmal subarachnoid hemorrhage (aSAH). However, it is unclear whether ML prognostic models will benefit patients due to the lack of a comprehensive assessment. We aimed to develop and evaluate ML models for predicting unfavorable functional outcomes for aSAH patients and identify the model with the greatest performance.

METHODS

In this retrospective study, a dataset of 955 patients with aSAH was used to construct and validate prognostic models for functional outcomes assessed using the modified Rankin scale during a follow-up period of 3-6 months. Clinical scores and clinical and radiological features on admission and secondary complications were used to construct models based on 5 ML algorithms (i.e., logistic regression [LR], k-nearest neighbor, extreme gradient boosting, random forest, and artificial neural network). For evaluation among the models, the area under the receiver operating characteristic curve, area under the precision-recall curve, calibration curve, and decision curve analysis were used.

RESULTS

Composite models had significantly higher area under the receiver operating characteristic curves than did simple models in predicting unfavorable functional outcomes. Compared with other composite models (random forest and extreme gradient boosting) with good calibration, LR had the highest area under the precision-recall score and showed the greatest benefit in decision curve analysis.

CONCLUSIONS

Of the 5 studied ML models, the conventional LR model outperformed the advanced algorithms in predicting the prognosis and could be a useful tool for health care professionals.

Non-alcoholic fatty liver disease risk prediction model and health management strategies for older Chinese adults: a cross-sectional study

BACKGROUND

Non-alcoholic fatty liver disease (NAFLD) is a common chronic liver condition that affects a quarter of the global adult population. To date, only a few NAFLD risk prediction models have been developed for Chinese older adults aged ≥ 60 years. This study presented the development of a risk prediction model for NAFLD in Chinese individuals aged ≥ 60 years and proposed personalised health interventions based on key risk factors to reduce NAFLD incidence among the population.

METHODS

A cross-sectional survey was carried out among 9,041 community residents in Shanghai. Three NAFLD risk prediction models (I, II, and III) were constructed using multivariate logistic regression analysis based on the least absolute shrinkage and selection operator regression analysis, and random forest model to select individual characteristics, respectively. To determine the optimal model, the three models' discrimination, calibration, clinical application, and prediction capability were evaluated using the receiver operating characteristic (ROC) curve, calibration plot, decision curve analysis, and net reclassification index (NRI), respectively. To evaluate the optimal model's effectiveness, the previously published NAFLD risk prediction models (Hepatic steatosis index [HSI] and ZJU index) were evaluated using the following five indicators: accuracy, precision, recall, F1-score, and balanced accuracy. A dynamic nomogram was constructed for the optimal model, and a Bayesian network model for predicting NAFLD risk in older adults was visually displayed using Netica software.

RESULTS

The area under the ROC curve of Models I, II, and III in the training dataset was 0.810, 0.826, and 0.825, respectively, and that of the testing data was 0.777, 0.797, and 0.790, respectively. No significant difference was found in the accuracy or NRI between the models; therefore, Model III with the fewest variables was determined as the optimal model. Compared with the HSI and ZJU index, Model III had the highest accuracy (0.716), precision (0.808), recall (0.605), F1 score (0.692), and balanced accuracy (0.723). The risk threshold for Model III was 20%-80%. Model III included body mass index, alanine aminotransferase level, triglyceride level, and lymphocyte count.

CONCLUSIONS

A dynamic nomogram and Bayesian network model were developed to identify NAFLD risk in older Chinese adults, providing personalized health management strategies and reducing NAFLD incidence.

Noninvasive Staging of Lymph Node Status in Breast Cancer Using Machine Learning: External Validation and Further Model Development

BACKGROUND

Most patients diagnosed with breast cancer present with a node-negative disease. Sentinel lymph node biopsy (SLNB) is routinely used for axillary staging, leaving patients with healthy axillary lymph nodes without therapeutic effects but at risk of morbidities from the intervention. Numerous studies have developed nodal status prediction models for noninvasive axillary staging using postoperative data or imaging features that are not part of the diagnostic workup. Lymphovascular invasion (LVI) is a top-ranked predictor of nodal metastasis; however, its preoperative assessment is challenging.

OBJECTIVE

This paper aimed to externally validate a multilayer perceptron (MLP) model for noninvasive lymph node staging (NILS) in a large population-based cohort (n=18,633) and develop a new MLP in the same cohort. Data were extracted from the Swedish National Quality Register for Breast Cancer (NKBC, 2014-2017), comprising only routinely and preoperatively available documented clinicopathological variables. A secondary aim was to develop and validate an LVI MLP for imputation of missing LVI status to increase the preoperative feasibility of the original NILS model.

METHODS

Three nonoverlapping cohorts were used for model development and validation. A total of 4 MLPs for nodal status and 1 LVI MLP were developed using 11 to 12 routinely available predictors. Three nodal status models were used to account for the different availabilities of LVI status in the cohorts and external validation in NKBC. The fourth nodal status model was developed for 80% (14,906/18,663) of NKBC cases and validated in the remaining 20% (3727/18,663). Three alternatives for imputation of LVI status were compared. The discriminatory capacity was evaluated using the validation area under the receiver operating characteristics curve (AUC) in 3 of the nodal status models. The clinical feasibility of the models was evaluated using calibration and decision curve analyses.

RESULTS

External validation of the original NILS model was performed in NKBC (AUC 0.699, 95% CI 0.690-0.708) with good calibration and the potential of sparing 16% of patients with node-negative disease from SLNB. The LVI model was externally validated (AUC 0.747, 95% CI 0.694-0.799) with good calibration but did not improve the discriminatory performance of the nodal status models. A new nodal status model was developed in NKBC without information on LVI (AUC 0.709, 95% CI: 0.688-0.729), with excellent calibration in the holdout internal validation cohort, resulting in the potential omission of 24% of patients from unnecessary SLNBs.

CONCLUSIONS

The NILS model was externally validated in NKBC, where the imputation of LVI status did not improve the model's discriminatory performance. A new nodal status model demonstrated the feasibility of using register data comprising only the variables available in the preoperative setting for NILS using machine learning. Future steps include ongoing preoperative validation of the NILS model and extending the model with, for example, mammography images.

Multisequence MRI-based radiomics model for predicting POLE mutation status in patients with endometrial cancer

OBJECTIVES

Preoperative identification of POLE mutation status would help tailor the surgical procedure and adjuvant treatment strategy. This study aimed to explore the feasibility of developing a radiomics model to pre-operatively predict the pathogenic POLE mutation status in patients with EC.

METHODS

The retrospective study involved 138 patients with histopathologically confirmed EC (35 POLE-mutant vs 103 non-POLE-mutant). After selecting relevant features with a series of steps, three radiomics signatures were built based on axial fat-saturation T2WI, DWI, and CE-T1WI images, respectively. Then, two radiomics models which integrated features from T2WI + DWI and T2WI + DWI+CE-T1WI were further developed using multivariate logistic regression. The performance of the radiomics model was evaluated from discrimination, calibration, and clinical utility aspects.

RESULTS

Among all the models, radiomics model2 (RM2), which integrated features from all three sequences, showed the best performance, with AUCs of 0.885 (95%CI: 0.828-0.942) and 0.810 (95%CI: 0.653-0.967) in the training and validation cohorts, respectively. The net reclassification index (NRI) and integrated discrimination improvement (IDI) analyses indicated that RM2 had improvement in predicting POLE mutation status when compared with the single-sequence-based signatures and the radiomics model1 (RM1). The calibration curve, decision curve analysis, and clinical impact curve suggested favourable calibration and clinical utility of RM2.

CONCLUSIONS

The RM2, fusing features from three sequences, could be a potential tool for the non-invasive preoperative identification of patients with POLE-mutant EC, which is helpful for developing individualized therapeutic strategies.

ADVANCES IN KNOWLEDGE

This study developed a potential surrogate of POLE sequencing, which is cost-efficient and non-invasive.

Prediction models in prehospital and emergency medicine research: How to derive and internally validate a clinical prediction model

Clinical prediction models are created to help clinicians with medical decision making, aid in risk stratification, and improve diagnosis and/or prognosis. With growing availability of both prehospital and in-hospital observational registries and electronic health records, there is an opportunity to develop, validate, and incorporate prediction models into clinical practice. However, many prediction models have high risk of bias due to poor methodology. Given that there are no methodological standards aimed at developing prediction models specifically in the prehospital setting, the objective of this paper is to describe the appropriate methodology for the derivation and validation of clinical prediction models in this setting. What follows can also be applied to the emergency medicine (EM) setting. There are eight steps that should be followed when developing and internally validating a prediction model: (1) problem definition, (2) coding of predictors, (3) addressing missing data, (4) ensuring adequate sample size, (5) variable selection, (6) evaluating model performance, (7) internal validation, and (8) model presentation. Subsequent steps include external validation, assessment of impact, and cost-effectiveness. By following these steps, researchers can develop a prediction model with the methodological rigor and quality required for prehospital and EM research.

Comprehensive analysis of m6A subtype classification for immune microenvironment of pituitary adenomas

BACKGROUND

N6-methyladenosine (m6A) RNA methylation and tumor immune microenvironment (IME) have an essential role in tumor development. However, their relationships in pituitary adenomas (PAs) remains unclear.

METHODS

PA datasets from the Gene Expression Omnibus (GEO) and European Bioinformatics Institute (EMBL-EBI) were used. We utilized hierarchical clustering algorithms based on the m6A regulator gene set to identify m6A subtypes. ESTIMATE and CIBERSORT algorithms were applied to explore the compositions of stromal and immune cells. A nomogram model was constructed for the prediction of m6A subtypes in PAs. Immunohistochemistry and multiplex immunofluorescence staining were used to analyze the expression level of m6A regulator YTHDF2 in relation to M2 macrophages and immune checkpoints in PAs.

RESULTS

We concluded the IME landscape of m6A subtype classification and characterized two emerging m6A subtypes. Different IME between these two m6A subtypes were identified. Simultaneously, a polygenic nomogram model was constructed for predicting m6A subtype classification, with excellent predictive performance (training set, AUC = 0.984; validation set, AUC = 0.986). YTHDF2 was highly expressed in PAs and accompanied by upregulated M2 macrophages and expression of PD-L1.

CONCLUSIONS

We proposed two novel m6A subtypes in PAs for the first time and constructed a reliable and clinically accessible nomogram model for them. Meanwhile, YTHDF2 was first identified as a promising biomarker for immunotherapy and potential molecular target in PAs.

Construction and validation of infection risk model for patients with external ventricular drainage: a multicenter retrospective study

PURPOSE

External ventricular drainage (EVD) is a life-saving neurosurgical procedure, of which the most concerning complication is EVD-related infection (ERI). We aimed to construct and validate an ERI risk model and establish a monographic chart.

METHODS

We retrospectively analyzed the adult EVD patients in four medical centers and split the data into a training and a validation set. We selected features via single-factor logistic regression and trained the ERI risk model using multi-factor logistic regression. We further evaluated the model discrimination, calibration, and clinical usefulness, with internal and external validation to assess the reproducibility and generalizability. We finally visualized the model as a nomogram and created an online calculator (dynamic nomogram).

RESULTS

Our research enrolled 439 EVD patients and found 75 cases (17.1%) had ERI. Diabetes, drainage duration, site leakage, and other infections were independent risk factors that we used to fit the ERI risk model. The area under the receiver operating characteristic curve (AUC) and the Brier score of the model were 0.758 and 0.118, and these indicators' values were similar when internally validated. In external validation, the model discrimination had a moderate decline, of which the AUC was 0.720. However, the Brier score was 0.114, suggesting no degradation in overall performance. Spiegelhalter's Z-test indicated that the model had adequate calibration when validated internally or externally (P = 0.464 vs. P = 0.612). The model was transformed into a nomogram with an online calculator built, which is available through the website: https://wang-cdutcm.shinyapps.io/DynNomapp/ .

CONCLUSIONS

The present study developed an infection risk model for EVD patients, which is freely accessible and may serve as a simple decision tool in the clinic.

Development of a Nomogram to Predict Clinically Relevant Postoperative Pancreatic Fistula After Pancreaticoduodenectomy on the Basis of Visceral Fat Area and Magnetic Resonance Imaging

BACKGROUND

The aim of this study was to develop a nomogram to predict the risk of developing clinically relevant postoperative pancreatic fistula (CR-POPF) after pancreaticoduodenectomy (PD) using preoperative clinical and imaging data.

METHODS

The data of 205 patients were retrospectively analyzed, randomly divided into training (n = 125) and testing groups (n = 80). The patients' preoperative laboratory indicators, preoperative clinical baseline data, and preoperative imaging data [enhanced computed tomography (CT), enhanced magnetic resonance imaging (MRI)] were collected. Univariate analyses combined with multivariate logistic regression were used to identify the independent risk factors for CR-POPF. These factors were used to train and validate the model and to develop the risk nomogram. The area under the curve (AUC) was used to measure the predictive ability of the models. The integrated discrimination improvement index (IDI) and decision curve analysis (DCA) were used to assess the clinical feasibility of the nomogram in relation to five other models established in literature.

RESULTS

CT visceral fat area (P = 0.014), the pancreatic spleen signal ratio on T1 fat-suppressed MRI sequences (P < 0.001), and CT main pancreatic duct diameter (P = 0.001) were identified as independent prognostic factors and used to develop the model. The final nomogram achieved an AUC of 0.903. The IDI and DCA showed that the nomogram outperformed the other five CR-POPF models in the training and testing cohorts.

CONCLUSION

The nomogram achieved a superior predictive ability for CR-POPF following PD than other models described in literature. Clinicians can use this simple model to optimize perioperative planning according to the patient's risk of developing CR-POPF.

Prognostic analysis of cutaneous Kaposi sarcoma based on a competing risk model

The data regarding the prognosis of cutaneous Kaposi sarcoma (KS) was limited. The current study aimed to explore the risk factors and develop a predictive model for the prognosis of cutaneous KS patients. Data were extracted from Surveillance, Epidemiology, and End Results database from 2000 to 2018 and randomly divided into training and validation cohort. The Kaplan-Meier analysis, cumulative incidence function based on the competing risk model and Fine-Gray multivariable regression model was used to identify the prognostic factors and then construct a 5-, 10-, and 15-year KS-specific death (KSSD) nomogram for patients. The concordance index (C-index), area under the curve (AUC) of operating characteristics and calibration plots were used to evaluate the performance of the model. The clinical utility of the model was measured by decision curve analysis (DCA). In 2257 cutaneous KS patients identified from database, the overall median survival time was about 13 years. Radiotherapy (p = 0.013) and surgery (p < 0.001) could lower the KSSD, while chemotherapy (p = 0.042) and surgery (p < 0.001) could increase the overall survival (OS) of patients with metastatic and localized lesions, respectively. Race, number of lesions, surgery, extent of disease, year of diagnosis and age were identified as risk factors associated with cutaneous KS-specific survival. Performance of the nomogram was validated by calibration and discrimination, with C-index values of 0.709 and AUC for 5-, 10-, and 15-year-KSSD of 0.739, 0.728 and 0.725 respectively. DCA indicated that the nomogram had good net benefits in clinical scenarios. Using a competing-risk model, this study firstly identified the prognostic factors, and constructed a validated nomogram to provide individualized assessment and reliable prognostic prediction for cutaneous KS patients.

Nomogram for Predicting Risk of Cancer Therapy-Related Cardiac Dysfunction in Patients With Human Epidermal Growth Factor Receptor 2-Positive Breast Cancer

BACKGROUND

Cancer therapy-related cardiac dysfunction (CTRCD) is an important treatment-limiting toxicity for patients with human epidermal growth factor receptor 2 (HER2)-positive breast cancer that adversely affects cancer and cardiovascular outcomes. Easy-to-use tools that incorporate readily accessible clinical variables for individual estimation of CTRCD risk are needed.

METHODS AND RESULTS

From 2004 to 2013, 1440 patients with stage I to III HER2-positive breast cancer treated with trastuzumab-based therapy were identified. A multivariable Cox proportional hazards model was constructed to identify risk factors for CTRCD and included the 1377 patients in whom data were complete. Nine clinical variables, including age, race, body mass index, left ventricular ejection fraction, systolic blood pressure, coronary artery disease, diabetes, arrhythmia, and anthracycline exposure were built into a nomogram estimating risk of CTRCD at 1 year. The nomogram was validated for calibration and discrimination using bootstrap resampling. A total of 177 CTRCD events occurred within 1 year of HER2-targeted treatment. The nomogram for prediction of 1-year CTRCD probability demonstrated good discrimination, with a concordance index of 0.687. The predicted and observed probabilities of CTRCD were similar, demonstrating good model calibration.

CONCLUSIONS

A nomogram composed of 9 readily accessible clinical variables provides an individualized 1-year risk estimate of CTRCD among women with HER2-positive breast cancer receiving HER2-targeted therapy. This nomogram represents a simple-to-use tool for clinicians and patients that can inform clinical decision-making on breast cancer treatment options, optimal frequency of cardiac surveillance, and role of cardioprotective strategies.

Combining Asian and European genome-wide association studies of colorectal cancer improves risk prediction across racial and ethnic populations

Polygenic risk scores (PRS) have great potential to guide precision colorectal cancer (CRC) prevention by identifying those at higher risk to undertake targeted screening. However, current PRS using European ancestry data have sub-optimal performance in non-European ancestry populations, limiting their utility among these populations. Towards addressing this deficiency, we expand PRS development for CRC by incorporating Asian ancestry data (21,731 cases; 47,444 controls) into European ancestry training datasets (78,473 cases; 107,143 controls). The AUC estimates (95% CI) of PRS are 0.63(0.62-0.64), 0.59(0.57-0.61), 0.62(0.60-0.63), and 0.65(0.63-0.66) in independent datasets including 1681-3651 cases and 8696-115,105 controls of Asian, Black/African American, Latinx/Hispanic, and non-Hispanic White, respectively. They are significantly better than the European-centric PRS in all four major US racial and ethnic groups (p-values < 0.05). Further inclusion of non-European ancestry populations, especially Black/African American and Latinx/Hispanic, is needed to improve the risk prediction and enhance equity in applying PRS in clinical practice.

Developing a nomogram model and prognostic analysis of nasopharyngeal squamous cell carcinoma patients: a population-based study

BACKGROUND

Nasopharyngeal squamous cell carcinoma (NPSCC) is a common histo-logical subtype of nasopharyngeal cancer with a generally poor prognosis. The aim of this study is to identify factors affecting the survival prognosis of NPSCC patients and develop a specialized nomogram model.

METHODS

We extracted clinical data of 1235 diagnosed cases of NPSCC from the SEER database using SEER*Stat software. Univariate and multivariate Cox proportional hazards regression analyses were conducted to explore clinical factors that impact the prognosis of NPSCC patients. Based on significant independent factors, we developed a nomogram to predict the 1, 3, and 5 years overall survival rates. The discriminative and predictive abilities of the nomogram were evaluated using C-index, calibration curve, area under the curve (AUC), and receiver operating characteristic curve. We evaluated the clinical value of the nomogram using decision curve analysis (DCA) and clinical impact curve (CIC).

RESULTS

We performed a cohort analysis on 846 patients with nasopharyngeal cancer in the training cohort. Multivariate Cox regression analysis revealed age, race, marital status, primary tumor, radiation therapy, chemotherapy, SJCC stage, primary tumor size, Lung metastasis and brain metastasis as independent prognostic factors for NPSCC patients, which we used to construct the nomogram prediction model. The C-index of the training cohort was 0.737. The ROC curve analysis indicated that the AUC of the OS rate at 1, 3, and 5 years in the training cohort was > 0.75. The calibration curves of the two cohorts showed good consistency between the predicted and observed results. DCA and CIC demonstrated that the nomogram prediction model had good clinical benefits.

CONCLUSIONS

The nomogram risk prediction model for NPSCC patient survival prognosis, constructed in this study, has exhibited excellent predictive capability. This model can be employed for swift and precise assessment of individualized survival prognosis. It can offer valuable guidance to clinical physicians in diagnosing and treating NPSCC patients.

Non-invasive diagnosis of primary Sjögren's syndrome using ultrasonography and transcriptome biomarkers

Diagnosing primary Sjögren's syndrome (pSS) is difficult due to clinical heterogeneity and the absence of non-invasive specific biomarkers. To develop non-invasive pSS diagnosis methods that integrate classic clinical indexes, major salivary gland ultrasonography (SGUS), and gene expression profiles shared by labial gland and peripheral blood, we conducted a study on a cohort of 358 subjects. We identified differentially expressed genes (DEGs) in glands and blood that were enriched in defense response to virus and type I interferon production pathways. Four upregulated DEGs common in glands and blood were identified as hub genes based on the protein-protein interaction networks. A random forest model was trained using features, including SGUS, anti-SSA/Ro60, keratoconjunctivitis sicca tests, and gene expression levels of MX1 and RSAD2. The model achieved comparable pSS diagnosis accuracy to the golden standard method based on labial gland biopsy. Our findings implicate this novel model as a promising diagnosis technique of pSS.