Prediction Of Individual Outcome Research Articles

Delivering patient-centered care through shared decision making in overactive bladder.

IntroductionMen and women living with overactive bladder (OAB) face many treatment decisions as they progress through the treatment pathway. Decisions to pursue specific therapies are highly preference sensitive and ideal for shared decision making (SDM). The aim of this narrative review is to provide urologists with a practical summary of methods to elicit preferences and facilitate SDM to promote patient‐centered care for OAB.MethodsWe explore OAB as a preference sensitive condition through a review of treatment outcomes and present available data on prediction tools, patient preferences, and decision aids. We propose a paradigm for applying Everyday SDM to OAB care.ResultsClinical outcome data points to equipoise (balanced outcomes) between options for first‐, second‐, and third‐line OAB therapies, making OAB preference sensitive and appropriate for SDM. Methods to personalize care through individualized outcome prediction calculators and tools to elicit patient preferences are emerging. While patient information about OAB is readily available, we identified few OAB decision aids that facilitate patient preference elicitation and SDM.ConclusionsOAB is a preference sensitive condition, where treatment is largely based on the patient's preferences and values. SDM is an ideal approach to supporting patients through these treatment decisions. We propose the application of Everyday SDM, a personalized, clinically efficient methodology as a method to support patient‐centered OAB care.

Prediction models in first-episode psychosis: systematic review and critical appraisal.

People presenting with first-episode psychosis (FEP) have heterogenous outcomes. More than 40% fail to achieve symptomatic remission. Accurate prediction of individual outcome in FEP could facilitate early intervention to change the clinical trajectory and improve prognosis. We aim to systematically review evidence for prediction models developed for predicting poor outcome in FEP. A protocol for this study was published on the International Prospective Register of Systematic Reviews, registration number CRD42019156897. Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidance, we systematically searched six databases from inception to 28 January 2021. We used the Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies and the Prediction Model Risk of Bias Assessment Tool to extract and appraise the outcome prediction models. We considered study characteristics, methodology and model performance. Thirteen studies reporting 31 prediction models across a range of clinical outcomes met criteria for inclusion. Eleven studies used logistic regression with clinical and sociodemographic predictor variables. Just two studies were found to be at low risk of bias. Methodological limitations identified included a lack of appropriate validation, small sample sizes, poor handling of missing data and inadequate reporting of calibration and discrimination measures. To date, no model has been applied to clinical practice. Future prediction studies in psychosis should prioritise methodological rigour and external validation in larger samples. The potential for prediction modelling in FEP is yet to be realised.

Development and Validation of a Nomogram for Predicting Seizure Outcomes After Epilepsy Surgery for Children with Focal Cortical Dysplasia.

To construct a prediction nomogram model for the postoperative seizure outcomes in children with focal cortical dysplasia (FCD). We retrospectively reviewed the clinical data of 97 children with epilepsy secondary to FCD who had undergone resection surgery at Children's Hospital of Chongqing Medical University from June 2013 to September 2019. Univariate and multivariate Cox proportional hazards regression were used to explore the predictors of postoperative persistent seizure, and a nomogram prediction model for postoperative seizure outcome was developed. The C-index was chosen to evaluate the discriminability of the nomogram with internal validation. Calibration curves and decision curve analysis were used to evaluate consistency and clinical efficacy, respectively. The complete resection of epileptogenic focus and the pathological type of FCD were independent predictors of persistent seizure in children with epilepsy secondary to FCD after surgery. Based on multivariate Cox proportional hazard regression, a predictive nomogram for epilepsy outcome was established and validated via the bootstrap method with 1000 resamples. The nomogram showed superior prediction accuracy (C-index = 0.883); by drawing and reviewing the calibration curve and decision curve, the nomogram presented good consistency and clinical efficacy. A nomogram prediction model of postsurgery seizure outcome in children with epilepsy secondary to FCD was constructed based on four variables, providing a reliable and convenient tool for individual seizure outcome prediction.

Precision medicine in stroke: towards personalized outcome predictions using artificial intelligence.

Stroke ranks among the leading causes for morbidity and mortality worldwide. New and continuously improving treatment options such as thrombolysis and thrombectomy have revolutionized acute stroke treatment in recent years. Following modern rhythms, the next revolution might well be the strategic use of the steadily increasing amounts of patient-related data for generating models enabling individualized outcome predictions. Milestones have already been achieved in several health care domains, as big data and artificial intelligence have entered everyday life.The aim of this review is to synoptically illustrate and discuss how artificial intelligence approaches may help to compute single-patient predictions in stroke outcome research in the acute, subacute and chronic stage. We will present approaches considering demographic, clinical and electrophysiological data, as well as data originating from various imaging modalities and combinations thereof. We will outline their advantages, disadvantages, their potential pitfalls and the promises they hold with a special focus on a clinical audience. Throughout the review we will highlight methodological aspects of novel machine-learning approaches as they are particularly crucial to realize precision medicine. We will finally provide an outlook on how artificial intelligence approaches might contribute to enhancing favourable outcomes after stroke.

Comprehensive Analysis Identified ETV7 as a Potential Prognostic Biomarker in Bladder Cancer.

Background The tumor microenvironment (TME) plays a crucial role in the initiation and progression of cancer. Bladder cancer (BLCA) is a malignant tumor of the genitourinary system. Its heterogeneity results in significant differences in the prognosis of patients. To date, this is still a huge challenge for clinical treatment. In recent years, more and more evidence showed that dysregulation of transcription factors (TFs) plays an important role in tumor progression, invasion, and metastasis. Unfortunately, the role of TFs on the tumor microenvironment in bladder cancer is unclear. Methods The original data of BLCA and corresponding adjacent tissues were obtained from The Cancer Genome Atlas (TCGA) database. TFs were downloaded from the Animal Transcription Factor DataBase (Animal TFDB). Intersection analysis was used to obtain TFs that were differentially expressed between tumor and adjacent tissues. Gene Set Cancer Analysis (GSCALite) and CIBERSORT software were used to reveal the key differentially expressed TFs (DE-TFs). Subsequently, UALCAN and Human Protein Atlas (HPA) databases were used to disclose the expression of key DE-TFs in BLCA. The K-M curve divulged the relationship between the key DE-TFs and the patient's overall survival (OS), and the univariate and multivariate Cox regression analyses were conducted to explore independent prognostic factors. The cluster profiler package and Gene Set Enrichment Analysis (GSEA) were used for functional enrichment of genes related to the key DE-TFs. Finally, CIBERSORT software analyzed the immune landscape of BLCA. Results We obtained a total of 117 BLCA-related DE-TFs. Among them, ETV7 was identified as the key DE-TFs due to its association with the autophagy activation pathway and various immune cells in cancer. Online databases of UALCAN and HPA indicated that ETV7 was overexpressed in tumors and negatively correlated with tumor severity. The K-M curve showed that the OS of patients with high expression of ETV7 was poor, which indicated that it was an independent prognostic factor. Functional enrichment of 87 DEGs between ETV7-high and -low expression groups indicated that it was closely related to the immune response and the functions of a variety of immune cells. Finally, CIBERSORT results proved that the high and low expression of ETV7 also caused significant differences in the tumor immune microenvironment of patients. Conclusion Overall, we proved that the transcription factor ETV7 was a novel prognostic factor, which may improve the individualized outcome prediction in BLCA by regulating the tumor immune microenvironment.

Multicenter Validation of Individual Preoperative Motor Outcome Prediction for Deep Brain Stimulation in Parkinson’s Disease

Background: Subthalamic nucleus deep brain stimulation (STN DBS) is an established therapy for Parkinson’s disease (PD) patients suffering from motor response fluctuations despite optimal medical treatment, or severe dopaminergic side effects. Despite careful clinical selection and surgical procedures, some patients do not benefit from STN DBS. Preoperative prediction models are suggested to better predict individual motor response after STN DBS. We validate a preregistered model, DBS-PREDICT, in an external multicenter validation cohort. Methods: DBS-PREDICT considered eleven, solely preoperative, clinical characteristics and applied a logistic regression to differentiate between weak and strong motor responders. Weak motor response was defined as no clinically relevant improvement on the Unified Parkinson’s Disease Rating Scale (UPDRS) II, III, or IV, 1 year after surgery, defined as, respectively, 3, 5, and 3 points or more. Lower UPDRS III and IV scores and higher age at disease onset contributed most to weak response predictions. Individual predictions were compared with actual clinical outcomes. Results: 322 PD patients treated with STN DBS from 6 different centers were included. DBS-PREDICT differentiated between weak and strong motor responders with an area under the receiver operator curve of 0.76 and an accuracy up to 77%. Conclusion: Proving generalizability and feasibility of preoperative STN DBS outcome prediction in an external multicenter cohort is an important step in creating clinical impact in DBS with data-driven tools. Future prospective studies are required to overcome several inherent practical and statistical limitations of including clinical decision support systems in DBS care.

Individual outcome prediction models for patients with COVID-19 based on their first day of admission to the intensive care unit

BackgroundCurrently, good prognosis and management of critically ill patients with COVID-19 are crucial for developing disease management guidelines and providing a viable healthcare system. We aimed to propose individual outcome prediction models based on binary logistic regression (BLR) and artificial neural network (ANN) analyses of data collected in the first 24 h of intensive care unit (ICU) admission for patients with COVID-19 infection. We also analysed different variables for ICU patients who survived and those who died. MethodsData from 326 critically ill patients with COVID-19 were collected. Data were captured on laboratory variables, demographics, comorbidities, symptoms and hospital stay related information. These data were compared with patient outcomes (survivor and non-survivor patients). BLR was assessed using the Wald Forward Stepwise method, and the ANN model was constructed using multilayer perceptron architecture. ResultsThe area under the receiver operating characteristic curve of the ANN model was significantly larger than the BLR model (0.917 vs 0.810; p < 0.001) for predicting individual outcomes. In addition, ANN model presented similar negative predictive value than the BLR model (95.9% vs 94.8%).Variables such as age, pH, potassium ion, partial pressure of oxygen, and chloride were present in both models and they were significant predictors of death in COVID-19 patients. ConclusionsOur study could provide helpful information for other hospitals to develop their own individual outcome prediction models based, mainly, on laboratory variables. Furthermore, it offers valuable information on which variables could predict a fatal outcome for ICU patients with COVID-19.

Risk stratification model for telemedicine-based cardiac rehabilitation

Abstract Background Physical training (PhT) is highly cost-benefit maneuver in patients with heart disease, as a part of an integral Cardiac Rehabilitation Program. Exercise-related adverse outcomes are low, mainly due to an adequate cardiovascular risk stratification and the correct prescription and supervision of PhT. Exercise testing (ET) is the cornerstone of this process. Advances in telecommunication technologies have boosted the possibility to deliver cardiac rehabilitation via the internet, a low cost and non-face-to-face intervention. However, it is not advisable to perform a remote stress test through digital media. Purpose The aim of this study is to estimate the effect of excluding exercise test data from the traditional risk stratification model on the prediction of individual adverse outcomes. Methods A cohort of patients with heart disease who participated in an outpatient hospital-based cardiac rehabilitation program was studied. All patients underwent a clinical evaluation, along with an ET. The data obtained were used to stratify the risk of adverse events during PhT. The physical exercise program was prescribed on an individual basis. Each patient performed 30 minutes of cycle ergometry, five times a week, with a moderate effort perception (6–20 Borg scale). All sessions were supervised by a cardiologist using continuous ECG telemetry and blood pressure measurement. Resistance training was complemented with a gymnastic circuit (kinesiotherapy). A bivariate and a logistic regression multivariable analysis were performed with clinical or paraclinical variables that had been previously used in risk stratification models and showed a statistically association with the outcome (traditional model). In order to simulate the lack of an exercise stress test, these data were excluded from multivariate regression model (TELERISK), Figure 1. Results Six hundred and thirty-nine patients with cardiovascular disease were studied. No major adverse outcome was recorded. Patients presented several minor adverse outcomes, including mainly arrhythmias (n=485), such as sinus bradycardia, sinus pauses, premature supraventricular complexes, supraventricular tachycardia, atrial fibrillation, premature ventricular complexes, ventricular bigeminy, ventricular couplets and non-sustained ventricular tachycardia. Other minor outcomes were exercise-induced ischemia (n=31), exercise-induced hypotension (n=15), hypertensive response to exercise (n=31) and dizziness (n=7). The predictive capacity of the TELERISK model was significantly lower (AUC=0.661) than that observed for the traditional model (AUC=0.766), Figure 2. Conclusion The predictive capacity of the risk stratification model for adverse events during physical training in patients with cardiovascular disease decreases significantly when excluding data from the exercise test. Funding Acknowledgement Type of funding sources: None. Figure 1. Telerisk multivariable modelFigure 2. Outcome prediction ROC curve

Identification of a novel glycolysis-related signature to predict the prognosis of patients with breast cancer

BackgroundBreast cancer (BC) has a high incidence and mortality rate in females. Its conventional clinical characteristics are far from accurate for the prediction of individual outcomes. Therefore, we aimed to develop a novel signature to predict the survival of patients with BC.MethodsWe analyzed the data of a training cohort from the Cancer Genome Atlas (TCGA) database and a validation cohort from the Gene Expression Omnibus (GEO) database. After the applications of Gene Set Enrichment Analysis (GSEA) and Cox regression analyses, a glycolysis-related signature for predicting the survival of patients with BC was developed; the signature contained AK3, CACNA1H, IL13RA1, NUP43, PGK1, and SDC1. Furthermore, on the basis of expression levels of the six-gene signature, we constructed a risk score formula to classify the patients into high- and low-risk groups. The receiver operating characteristic (ROC) curve and the Kaplan-Meier curve were used to assess the predicted capacity of the model. Later, a nomogram was developed to predict the outcomes of patients with risk score and clinical features over a period of 1, 3, and 5 years. We further used Human Protein Atlas (HPA) database to validate the expressions of the six biomarkers in tumor and sample tissues, which were taken as control.ResultsWe constructed a six-gene signature to predict the outcomes of patients with BC. The patients in the high-risk group showed poor prognosis than those in the low-risk group. The area under the curve (AUC) values were 0.719 and 0.702, showing that the prediction performance of the signature is acceptable. Additionally, Cox regression analysis revealed that these biomarkers could independently predict the prognosis of BC patients with BC without being affected by clinical factors. The expression levels of all six biomarkers in BC tissues were higher than that in normal tissues; however, AK3 was an exception.ConclusionWe developed a six-gene signature to predict the prognosis of patients with BC. Our signature has been proved to have the ability to make an accurate prediction and might be useful in expanding the hypothesis in clinical research.

THE COMBINATORIAL ROLE OF IROQUOIS HOMEOBOX GENES 3 AND 4 IN THE COMPACTION OF THE VENTRICULAR MYOCARDIUM

BACKGROUND Left ventricular non-compaction cardiomyopathy (LVNC) is a congenital cardiomyopathy, characterized by excessive ventricular trabeculation, with a ‘spongy’ appearance of the myocardium, deep intertrabecular recesses, and a thin compacted myocardial layer. LVNC has recently gained more attention because it has been increasingly recognized as a cause of heart failure, thromboembolism, or ventricular arrhythmia, which can all lead to sudden cardiac death. However, due to its heterogeneous anatomical and clinical features, LVNC is still a poorly understood condition. In addition, LVNC is suggested to be due to a premature arrest in the myocardial compaction process, but its genetic basis and pathologic mechanism remain largely unknown. METHODS AND RESULTS Here we introduce novel genetic factors implicated in ventricular compaction. Specifically, combinatorial loss of two Iroquois homeobox (Irx) transcription factors, Irx3, critical for development and function of the ventricular conduction system (e.g., His-bundle, bundle branches and Purkinje fibers) and Irx4, essential for establishing ventricular identity, recapitulates the clinical hallmarks of LVNC in mice. While mice lacking either Irx3 or Irx4 do not exhibit gross congenital cardiac malformations, mice lacking both Irx3 and Irx4 (Irx3;Irx4DKO) show postnatal lethality with abnormal muscular growths that resemble the polypoid pattern of LVNC, characterized by multiple muscular nodes in the left ventricle. Optical projection tomography in postnatal day 14 (P14) mutant hearts unveiled increased trabeculation, abnormal bridging below the papillary muscles, and disorganized muscle structures (Fig. 1). During embryogenesis, Irx3 and Irx4 exhibit overlapping expression in the developing ventricles. Consistent with the hypothesis that LVNC can be caused by a defect in embryonic heart growth, Irx3;Irx4DKO embryonic hearts showed abnormal trabeculation with thinner ventricular walls as early as E14.5. RNA sequencing on E14.5 Irx3;Irx4DKO ventricles revealed altered ventricular identity with atrial gene activation and enriched Bmp2/Tbx2 pathway that is critical for endocardial cushion development by suppressing cardiomyocyte differentiation. Moreover, co-immunoprecipitation discovered that Irx4 physically interacts with Irx3 and Gata4, and promoter assay further showed that Irx3, Irx4, and Gata4 synergistically suppress Tbx2 promoter activity, suggesting the repressive role of this transcriptional regulatory complex in the developing myocardium. CONCLUSION Our study demonstrates a novel mechanistic interaction of Irx3 and Irx4, in cooperation with Gata4, during the ventricular compaction process, thereby ensuring proper ventricular development. This finding can promote improvement of genetic testing for LVNC and prediction of individual outcomes with LVNC. Left ventricular non-compaction cardiomyopathy (LVNC) is a congenital cardiomyopathy, characterized by excessive ventricular trabeculation, with a ‘spongy’ appearance of the myocardium, deep intertrabecular recesses, and a thin compacted myocardial layer. LVNC has recently gained more attention because it has been increasingly recognized as a cause of heart failure, thromboembolism, or ventricular arrhythmia, which can all lead to sudden cardiac death. However, due to its heterogeneous anatomical and clinical features, LVNC is still a poorly understood condition. In addition, LVNC is suggested to be due to a premature arrest in the myocardial compaction process, but its genetic basis and pathologic mechanism remain largely unknown. Here we introduce novel genetic factors implicated in ventricular compaction. Specifically, combinatorial loss of two Iroquois homeobox (Irx) transcription factors, Irx3, critical for development and function of the ventricular conduction system (e.g., His-bundle, bundle branches and Purkinje fibers) and Irx4, essential for establishing ventricular identity, recapitulates the clinical hallmarks of LVNC in mice. While mice lacking either Irx3 or Irx4 do not exhibit gross congenital cardiac malformations, mice lacking both Irx3 and Irx4 (Irx3;Irx4DKO) show postnatal lethality with abnormal muscular growths that resemble the polypoid pattern of LVNC, characterized by multiple muscular nodes in the left ventricle. Optical projection tomography in postnatal day 14 (P14) mutant hearts unveiled increased trabeculation, abnormal bridging below the papillary muscles, and disorganized muscle structures (Fig. 1). During embryogenesis, Irx3 and Irx4 exhibit overlapping expression in the developing ventricles. Consistent with the hypothesis that LVNC can be caused by a defect in embryonic heart growth, Irx3;Irx4DKO embryonic hearts showed abnormal trabeculation with thinner ventricular walls as early as E14.5. RNA sequencing on E14.5 Irx3;Irx4DKO ventricles revealed altered ventricular identity with atrial gene activation and enriched Bmp2/Tbx2 pathway that is critical for endocardial cushion development by suppressing cardiomyocyte differentiation. Moreover, co-immunoprecipitation discovered that Irx4 physically interacts with Irx3 and Gata4, and promoter assay further showed that Irx3, Irx4, and Gata4 synergistically suppress Tbx2 promoter activity, suggesting the repressive role of this transcriptional regulatory complex in the developing myocardium. Our study demonstrates a novel mechanistic interaction of Irx3 and Irx4, in cooperation with Gata4, during the ventricular compaction process, thereby ensuring proper ventricular development. This finding can promote improvement of genetic testing for LVNC and prediction of individual outcomes with LVNC.

Personalised predictive modelling with brain-inspired spiking neural networks of longitudinal MRI neuroimaging data and the case study of dementia

Background:Longitudinal neuroimaging provides spatiotemporal brain data (STBD) measurement that can be utilised to understand dynamic changes in brain structure and/or function underpinning cognitive activities. Making sense of such highly interactive information is challenging, given that the features manifest intricate temporal, causal relations between the spatially distributed neural sources in the brain. Methods:The current paper argues for the advancement of deep learning algorithms in brain-inspired spiking neural networks (SNN), capable of modelling structural data across time (longitudinal measurement) and space (anatomical components). The paper proposes a methodology and a computational architecture based on SNN for building personalised predictive models from longitudinal brain data to accurately detect, understand, and predict the dynamics of an individual’s functional brain state. The methodology includes finding clusters of similar data to each individual, data interpolation, deep learning in a 3-dimensional brain-template structured SNN model, classification and prediction of individual outcome, visualisation of structural brain changes related to the predicted outcomes, interpretation of results, and individual and group predictive marker discovery. Results:To demonstrate the functionality of the proposed methodology, the paper presents experimental results on a longitudinal magnetic resonance imaging (MRI) dataset derived from 175 older adults of the internationally recognised community-based cohort Sydney Memory and Ageing Study (MAS) spanning 6 years of follow-up. Significance:The models were able to accurately classify and predict 2 years ahead of cognitive decline, such as mild cognitive impairment (MCI) and dementia with 95% and 91% accuracy, respectively. The proposed methodology also offers a 3-dimensional visualisation of the MRI models reflecting the dynamic patterns of regional changes in white matter hyperintensity (WMH) and brain volume over 6 years. Conclusion:The method is efficient for personalised predictive modelling on a wide range of neuroimaging longitudinal data, including also demographic, genetic, and clinical data. As a case study, it resulted in finding predictive markers for MCI and dementia as dynamic brain patterns using MRI data.

Identification of an autophagy-related prognostic signature in head and neck squamous cell carcinoma.

Autophagy-related genes (ARGs) have been significantly implicated in tumorigenesis and served as promising prognostic biomarkers for human cancer. Hence, this study was aimed to develop an ARGs-based prognostic signature for Head and neck squamous cell carcinoma (HNSCC). Prognostic ARG candidates were identified by univariate and multivariate Cox regression analysis in the training dataset (TCGA-HNSC) and incorporated into a 3-ARGs (EGFR, FADD, and PARK2) prognostic signature which was further verified in two independent validation cohorts (GSE41613 and GSE42743). Kaplan-Meier plots, Cox regression analyses, and receiver operating characteristics curves (ROC) were employed to evaluate the prognostic prediction of 3-ARGs signature. Differential expression of these 3 ARG between cancer and normal counterparts as well as their associations with autophagy markers were assessed in 60 pairs of freshly collected HNSCC and adjacent non-tumor samples and datasets from Human Protein Atlas, respectively. Patients with high-risk score had significantly inferior overall survival. Multivariate regression analyses revealed that 3-ARGs signature could be an independent prognostic factor after adjusting various clinicopathological parameters. ROC analyses revealed high predictive accuracy and sensitivity of the 3-ARGs signature. Increased mRNA and protein expression of EGFR, FADD, and PARK2 were found in HNSCC samples, and their expression significantly correlated with the abundances of ATG5, Beclin1, and LC3. Our results reveal that 3-ARGs signature is a powerful prognostic biomarker for HNSCC, which could be integrated into the current prognostic regime to realize individualized outcome prediction. EGFR, FADD, and PARK2likely contributed to autophagy during HNSCC tumorigenesis.

Outcome Prediction in CLL Patients Based on a Disease-Specific Oligonucleotide Microarray.

Chronic Lymphocytic Leukemia (CLL) is characterized by a extreme variability in the clinical course; some patients survive only a few months whereas others have stable disease for more than 10 years. Individual outcome prediction in CLL is still a challenge, in spite of the existence of multiple newly described prognostic factors: cytogenetic alterations, mutational IgVH status, and expression of surrogate markers such as ZAP-70. The accuracy of outcome prediction in CLL may improve thanks to the inclusion of new additional biological variables identified by high-throughput molecular techniques. This more reliable risk-stratification of patients could allow to assign a risk-adapted individualized therapy. With this aim, we have analysed with oligonucleotide microarrays three different series of CLL patients in a three-step process.In a first-step, RNA extracted from tumour samples of a series of 26 CLL patients, was hybridised using commercial 22K oligonucleotide chips (Agilent Technologies). A total of 135 genes whose expression was associated with changes in survival probability were identified. An additional review of published data yielded a total number of 497 genes whose expression has been related with outcome in CLL. A CLL specific chip was then built in a second-step including, besides these 632 genes, control genes until a total of 1023 genes. This comprehensive CLL-chip was then hybridised with RNA extracted from peripheral-blood samples in a new series of consecutively diagnosed 100 CLL patients. Cox's univariate analysis was performed to look for molecular variables associated with Progression Free Survival (PFS) as endpoint. FDR significance levels were also used when appropriate, for adjusting the associated p-value. Expression of 94 genes was significantly associated with PFS. Profile patterns of the 94 resulting genes were clustered, and allow to identify 12 clusters. Based on biological function, we selected a representing gene of each cluster, to perform a multivariate Cox's analysis. This approach generated a model of 4 genes, implicated in BCR signalling, signal transduction and transcription regulation, that distinguishes four quartiles of CLL patients with PFS at 5 years of 95%, 70%, 44% and 8% respectively.These results have been validated in a new independent series of 60 consecutively diagnosed cases of CLL. If confirmed in additional studies, the availability of this CLL-specific chip could allow a more accurate prediction of the clinical course in CLL patients, this giving support to a stratification of patients into different risk-groups

MRI combined with early clinical variables are excellent outcome predictors for newborn infants undergoing therapeutic hypothermia after perinatal asphyxia.

BackgroundBinary prediction-models for outcome [death, cognition, presence and severity of cerebral palsy (CP)], using MRI and early clinical data applicable for individual outcome prediction have not been developed.MethodsFrom Dec 1st 2006 until Dec 31st 2013, we recruited 178 infants into a population-based cohort with moderate or severe hypoxic-ischaemic encephalopathy (HIE) including postnatal collapse (PNC, n = 12) and additional diagnoses (n = 12) using CoolCap/TOBY-trial entry-criteria including depressed amplitude-integrated EEG (aEEG). Early clinical/biochemical variables and MRI scans (median day 8) were obtained in 168 infants. Injury severity was scored for cortex, basal ganglia/thalami (BGT), white matter (WM) and posterior limb of the internal capsule, summating to a total injury score (TIS, range 0–11). Outcome was categorized as adverse or favourable at 18–24 months from Bayley-III domains (cut-off 85) and neurological examination including CP classification.FindingsHIE and entry-aEEG severity were stable throughout the study. Outcome was favourable in 133/178 infants and adverse in 45/178: 17 died, 28 had low Cognition/Language scores, (including 9 with severe CP and 6 mild); seven had mild CP with favourable cognitive outcome. WMxBGT product scores and TIS were strong outcome predictors, and prediction improved when clinical/biochemical variables were added in binary logistic regression. The Positive Predictive Value for adverse outcome was 88%, increasing to 95% after excluding infants with PNC and additional diagnoses. Using WMxBGT in the regression predicted 8 of the 9 children with severe CP.InterpretationBinary logistic regression with WMxBGT or TIS and clinical variables gave excellent outcome prediction being 12% better than single variable cross-tabulation. Our MRI scoring and regression models are readily accessible and deserve investigation in other cohorts for group and individual prediction.FundingWe thank the National Health Service (NHS) and our Universities and funders in UK and Norway: SPARKS, The Moulton Foundation, The Norwegian Research Council, The Lærdal Foundation for Acute Medicine and charitable donations for their support for cooling therapy.

Prediction of Tumor Cellularity in Resectable PDAC from Preoperative Computed Tomography Imaging.

Simple SummaryPancreatic ductal adenocarcinoma (PDAC) remains a devastating disease. However, variations in tumor biology influence individual patient outcomes greatly. We previously showed a strong association between magnetic resonance imaging-based tumor cell estimates and patient survival. In this study we aimed to transfer this finding to more broadly applied computed tomography (CT) imaging for non-invasive risk stratification. We correlated in vivo CT imaging with histopathological analyses and could show a strong association between regional Hounsfield Units (HU) and tumor cellularity. In conclusion, our study suggests CT-based tumor cell estimates as a widely applicable way of non-invasive tumor cellularity characterization in PDAC.Background: PDAC remains a tumor entity with poor prognosis and a 5-year survival rate below 10%. Recent research has revealed invasive biomarkers, such as distinct molecular subtypes, predictive for therapy response and patient survival. Non-invasive prediction of individual patient outcome however remains an unresolved task. Methods: Discrete cellularity regions of PDAC resection specimen (n = 43) were analyzed by routine histopathological work up. Regional tumor cellularity and CT-derived Hounsfield Units (HU, n = 66) as well as iodine concentrations were regionally matched. One-way ANOVA and pairwise t-tests were performed to assess the relationship between different cellularity level in conventional, virtual monoenergetic 40 keV (monoE 40 keV) and iodine map reconstructions. Results: A statistically significant negative correlation between regional tumor cellularity in histopathology and CT-derived HU from corresponding image regions was identified. Radiological differentiation was best possible in monoE 40 keV CT images. However, HU values differed significantly in conventional reconstructions as well, indicating the possibility of a broad clinical application of this finding. Conclusion: In this study we establish a novel method for CT-based prediction of tumor cellularity for in-vivo tumor characterization in PDAC patients.

Radiomics and Magnetic Resonance Imaging of Rectal Cancer: From Engineering to Clinical Practice.

While cross-sectional imaging has seen continuous progress and plays an undiscussed pivotal role in the diagnostic management and treatment planning of patients with rectal cancer, a largely unmet need remains for improved staging accuracy, assessment of treatment response and prediction of individual patient outcome. Moreover, the increasing availability of target therapies has called for developing reliable diagnostic tools for identifying potential responders and optimizing overall treatment strategy on a personalized basis. Radiomics has emerged as a promising, still fully evolving research topic, which could harness the power of modern computer technology to generate quantitative information from imaging datasets based on advanced data-driven biomathematical models, potentially providing an added value to conventional imaging for improved patient management. The present study aimed to illustrate the contribution that current radiomics methods applied to magnetic resonance imaging can offer to managing patients with rectal cancer.

Biological stratification of clinical disease courses in childhood immune thrombocytopenia

BackgroundIn childhood immune thrombocytopenia (ITP), an autoimmune bleeding disorder, there is a need for better prediction of individual disease courses and treatment outcomes. ObjectiveTo predict the response to intravenous immunoglobulins (IVIg) and ITP disease course using genetic and immune markers. MethodsChildren aged younger than 7 years with newly diagnosed ITP (N = 147) from the Treatment With or Without IVIG for Kids with ITP study were included, which randomized children to an IVIg or observation group. A total of 46 variables were available: clinical characteristics, targeted genotyping, lymphocyte immune phenotyping, and platelet autoantibodies. ResultsIn the treatment arm, 48/80 children (60%) showed a complete response (platelets ≥100 × 109/L) that lasted for at least 1 month (complete sustained response [CSR]) and 32 exhibited no or a temporary response (absence of a sustained response [ASR]). For a biological risk score, five variables were selected by regularized logistic regression that predicted ASR vs CSR: (1) hemoglobin; (2) platelet count; (3) genetic polymorphisms of Fc‐receptor (FcγR) IIc; (4) the presence of immunoglobulin G (IgG) anti‐platelet antibodies; and (5) preceding vaccination. The ASR sensitivity was 0.91 (95% confidence interval, 0.80‐1.00) and specificity was 0.67 (95% confidence interval, 0.53‐0.80). In the 67 patients of the observation arm, this biological score was also associated with recovery during 1 year of follow‐up. The addition of the biological score to a predefined clinical score further improved the discrimination of favorable ITP disease courses. ConclusionsThe prediction of disease courses and IVIg treatment responses in ITP is improved by using both clinical and biological stratification.

Machine Learning for Localizing Epileptogenic-Zone in the Temporal Lobe: Quantifying the Value of Multimodal Clinical-Semiology and Imaging Concordance.

Background: Epilepsy affects 50 million people worldwide and a third are refractory to medication. If a discrete cerebral focus or network can be identified, neurosurgical resection can be curative. Most excisions are in the temporal-lobe, and are more likely to result in seizure-freedom than extra-temporal resections. However, less than half of patients undergoing surgery become entirely seizure-free. Localizing the epileptogenic-zone and individualized outcome predictions are difficult, requiring detailed evaluations at specialist centers.Methods: We used bespoke natural language processing to text-mine 3,800 electronic health records, from 309 epilepsy surgery patients, evaluated over a decade, of whom 126 remained entirely seizure-free. We investigated the diagnostic performances of machine learning models using set-of-semiology (SoS) with and without hippocampal sclerosis (HS) on MRI as features, using STARD criteria.Findings: Support Vector Classifiers (SVC) and Gradient Boosted (GB) decision trees were the best performing algorithms for temporal-lobe epileptogenic zone localization (cross-validated Matthews correlation coefficient (MCC) SVC 0.73 ± 0.25, balanced accuracy 0.81 ± 0.14, AUC 0.95 ± 0.05). Models that only used seizure semiology were not always better than internal benchmarks. The combination of multimodal features, however, enhanced performance metrics including MCC and normalized mutual information (NMI) compared to either alone (p < 0.0001). This combination of semiology and HS on MRI increased both cross-validated MCC and NMI by over 25% (NMI, SVC SoS: 0.35 ± 0.28 vs. SVC SoS+HS: 0.61 ± 0.27).Interpretation: Machine learning models using only the set of seizure semiology (SoS) cannot unequivocally perform better than benchmarks in temporal epileptogenic-zone localization. However, the combination of SoS with an imaging feature (HS) enhance epileptogenic lobe localization. We quantified this added NMI value to be 25% in absolute terms. Despite good performance in localization, no model was able to predict seizure-freedom better than benchmarks. The methods used are widely applicable, and the performance enhancements by combining other clinical, imaging and neurophysiological features could be similarly quantified. Multicenter studies are required to confirm generalizability.Funding: Wellcome/EPSRC Center for Interventional and Surgical Sciences (WEISS) (203145Z/16/Z).

Healthcare Analytics: A Comprehensive Review

Big data have attracted significant attention in recent years, as their hidden potentials that can improve human life, especially when applied in healthcare. Big data is a reasonable collection of useful information allowing new breakthroughs or understandings. This paper reviews the use and effectiveness of data analytics in healthcare, examining secondary data sources such as books, journals, and other reputable publications between 2000 and 2020, utilizing a very strict strategy in keywords. Large scale data have been proven of great importance in healthcare, and therefore there is a need for advanced forms of data analytics, such as diagnostic data and descriptive analysis, for improving healthcare outcomes. The utilization of large-scale data can form the backbone of predictive analytics which is the baseline for future individual outcome prediction.

The prediction of survival in Gastric Cancer based on a Robust 13-Gene Signature

Gastric cancer represents a major public health problem. Owing to the great heterogeneity of GC, conventional clinical characteristics are limited in the accurate prediction of individual outcomes and survival. This study aimed to establish a robust gene signature to predict the prognosis of GC based on multiple datasets. Initially, we downloaded raw data from four independent datasets of The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO), and performed univariate Cox proportional hazards regression analysis to identify prognostic genes associated with overall survival (OS) from each dataset. Thirteen common genes from four datasets were screened as candidate prognostic signatures. Then, a risk score model was developed based on this 13‑gene signature and validated by four independent datasets and the entire cohort. Patients with a high-risk score had poorer OS and recurrence-free survival (RFS). Multivariate regression and stratified analysis revealed that the 13-gene signature was not only an independent predictive factor but also associated with recurrence when adjusting for other clinical factors. Furthermore, in the high-risk group, gene set enrichment analysis (GSEA) showed that the mTOR signaling pathway and MAPK signaling pathway were significantly enriched. The present study provided a robust and reliable gene signature for prognostic prediction of both OS and RFS of patients with GC, which may be useful for delivering individualized management of patients.