Integration Of Clinical Data Research Articles

Abstract 4421: Development of a clinical-genomic database to study tumor evolution and molecular biomarkers of drug resistance in a real-world setting

Abstract Background: Real-world evidence (RWE), specifically the integration of clinical and molecular data, plays an increasingly important role in elucidating the molecular landscape of advanced solid tumors, at diagnosis and at progression, and furthering precision oncology research. However, reliance on tissue genomic testing results within these RWE databases significantly limits our understanding of tumor evolution as tissue profiles are typically limited to a single time point, generally at diagnosis, and biopsies are not routinely completed at progression. Circulating tumor DNA (ctDNA) analysis at diagnosis and progression is an alternative to serial tissue testing. Studying the real-world data derived from ctDNA testing results can provide additional information, specifically around tumor evolution and mechanisms of innate and acquired resistance in response to intervening therapies (targeted and non-targeted). Methods: Anonymized results from patients consecutively tested with Guardant360®, a CLIA-certified, CAP-accredited, NYSDOH-approved ctDNA test indicated for patients with advanced and metastatic solid tumor cancers (Guardant Health, Redwood City, California) were analyzed. Using a secure and HIPAA-compliant methodology, these results were then linked to a de-identified encounters database containing medical and pharmacy claims to provide a longitudinal view of patients' journeys, including diagnosis, treatments, and real-world outcomes. Results: Of 85,874 patients, the disease groups most highly represented were lung, breast, colorectal, prostate, and pancreas cancers. 48,317 patients (56.3%) had a record of treatment with an antineoplastic treatment (i.e. chemo, hormonal, small molecule, or immuno-therapy). Of patients with a record of therapy, 14,608 (30.2%) had a record of therapy following, but none prior to, the Guardant360 test date, 14,493 (30.0%) had a record of therapy prior to, but not following theGuardant360 test date, and 19,216 (39.8%) had a record of therapy both before and after Guardant360 test date. Conclusions: In a cohort of patients who underwent ctDNA testing and for whom treatment information was available, post-therapy molecular results were available in 69.8%. The availability of genomic information following systemic therapy suggest that this dataset is a viable resource to study drug resistance and tumor evolution in the real-world setting. Citation Format: Naveen Kumar, Rajesh Kucharlapati, Aaron Hardin, Victoria M. Raymond, Stephen Fairclough, Kathryn Lang. Development of a clinical-genomic database to study tumor evolution and molecular biomarkers of drug resistance in a real-world setting [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4421.

Predicting rate of cognitive decline at baseline using a deep neural network with multidata analysis.

Purpose: Our study investigates whether a machine-learning-based system can predict the rate of cognitive decline in mildly cognitively impaired patients by processing only the clinical and imaging data collected at the initial visit. Approach: We built a predictive model based on a supervised hybrid neural network utilizing a three-dimensional convolutional neural network to perform volume analysis of magnetic resonance imaging (MRI) and integration of nonimaging clinical data at the fully connected layer of the architecture. The experiments are conducted on the Alzheimer's Disease Neuroimaging Initiative dataset. Results: Experimental results confirm that there is a correlation between cognitive decline and the data obtained at the first visit. The system achieved an area under the receiver operator curve of 0.70 for cognitive decline class prediction. Conclusion: To our knowledge, this is the first study that predicts "slowly deteriorating/stable" or "rapidly deteriorating" classes by processing routinely collected baseline clinical and demographic data [baseline MRI, baseline mini-mental state examination (MMSE), scalar volumetric data, age, gender, education, ethnicity, and race]. The training data are built based on MMSE-rate values. Unlike the studies in the literature that focus on predicting mild cognitive impairment (MCI)-to-Alzheimer's disease conversion and disease classification, we approach the problem as an early prediction of cognitive decline rate in MCI patients.

Novel prognostic prediction model constructed through machine learning on the basis of methylation-driven genes in kidney renal clear cell carcinoma.

Kidney renal clear cell carcinoma (KIRC) is a common tumor with poor prognosis and is closely related to many aberrant gene expressions. DNA methylation is an important epigenetic modification mechanism and a novel research target. Thus, exploring the relationship between methylation-driven genes and KIRC prognosis is important. The methylation profile, methylation-driven genes, and methylation characteristics in KIRC was revealed through the integration of KIRC methylation, RNA-seq, and clinical information data from The Cancer Genome Atlas. The Lasso regression was used to establish a prognosis model on the basis of methylation-driven genes. Then, a trans-omics prognostic nomogram was constructed and evaluated by combining clinical information and methylated prognosis model. A total of 242 methylation-driven genes were identified. The Gene Ontology terms of these methylation-driven genes mainly clustered in the activation, adhesion, and proliferation of immune cells. The methylation prognosis prediction model that was established using the Lasso regression included four genes in the methylation data, namely, FOXI2, USP44, EVI2A, and TRIP13. The areas under the receiver operating characteristic curve of 1-, 3-, and 5-year survival rates were 0.810, 0.824, and 0.799, respectively, in the training group and 0.794, 0.752, and 0.731, respectively, in the testing group. An easy trans-omics nomogram was successfully established. The C-indices of the nomogram in the training and the testing groups were 0.8015 and 0.8389, respectively. The present study revealed the overall perspective of methylation-driven genes in KIRC and can help in the evaluation of the prognosis of KIRC patients and provide new clues for further study.

Abstract IA04: Clinical challenges in oncologic imaging: AI support from image analysis to integrated diagnostics

Abstract The automated analysis of large-scale clinical datasets poses new compelling challenges for data-driven and model-based computational methods in medical imaging. As a matter of fact, the amount of heterogeneous biomedical data is considerably increasing due to the advances in medical imaging acquisition modalities and high-throughput technologies for multi-omics. In addition, electronic health records can be properly integrated to support personalized screening and diagnosis. In such a context, artificial intelligence (AI) is revolutionizing cancer image analysis by relying on sophisticated machine learning and computational intelligence techniques. Therefore, cutting-edge AI methods can enable the shift from organization-centric to patient-centric models, leading to effective multi-institutional health care services in terms of both clinical outcomes and costs. Computerized oncologic image analysis is encouraging the transition from largely qualitative image interpretation to quantitative assessment through automated methods aiming at early detection as well as lesion characterization. Nevertheless, several challenges and opportunities exist: • Reproducible and reliable segmentation methods are required to cope with time-consuming and error-prone manual delineation in laborious human decision-making tasks. As a matter of fact, accurate operator-independent segmentation procedures can improve the robustness of radiomics analyses. • Accurate computer-assisted diagnosis, associated with proper data curation, can reduce the risk of overlooking the diagnosis in a clinical environment. • Prognostic/predictive biomarker discovery by capturing the underlying tumor heterogeneity. • Quantification and monitoring of intra-/intertumoral heterogeneity during the course of the disease, by a proper integration of imaging, clinical, and molecular data on a patient-by-patient level • From a health-economics perspective, increasing costs of highly specific oncologic treatments (including immunotherapies) require accurate patient selection strategies, as well as better biomarkers of treatment response. Currently, many state-of-the-art methods based on deep learning have been achieving outstanding performance. The key to their success is the strong learning ability of fully supervised ML models and the availability of large-scale labeled datasets that include precise annotations. Unfortunately, in biomedical research, collecting such accurate annotations is an expensive process due to the need for domain experts’ knowledge. The generation of large mineable imaging datasets might mitigate this challenge by overcoming data paucity and heterogeneity issues. However, along with the availability of samples, data quality and diversity should be considered by collecting and preparing harmonized datasets. Therefore, the generalization abilities in multi-institutional studies can be improved by exploiting transfer learning and domain adaptation techniques. Finally, the explainability and safety of clinical decision support systems should be guaranteed, by avoiding “black-box” AI models exhibiting unpredictable behaviors. In conclusion, precision oncology should be directed towards the development of integrated radiogenomics paradigms that provide robust computational tools for investigating cancer biology as well as its implications for predicting treatment response. This solution allows for large-scale data collection (from multiple institutions) and continuous learning, by dealing with cyber-security and privacy issues. At present, the ultimate challenge is bridging the gap between AI and clinical practice, by firstly performing well-validated clinical research studies. This step is vital for the final translation and deployment of these advanced AI approaches in precision oncology. Citation Format: Evis Sala. Clinical challenges in oncologic imaging: AI support from image analysis to integrated diagnostics [abstract]. In: Proceedings of the AACR Special Conference on Advancing Precision Medicine Drug Development: Incorporation of Real-World Data and Other Novel Strategies; Jan 9-12, 2020; San Diego, CA. Philadelphia (PA): AACR; Clin Cancer Res 2020;26(12_Suppl_1):Abstract nr IA04.

P0502BIOBANKING FOR GLOMERULAR DISEASES: A STUDY DESIGN AND PROTOCOL FOR KOREA RENAL BIOBANK NETWORK SYSTEM TOWARD NEXT-GENERATION ANALYSIS (KORNERSTONE)

Abstract Background and Aims Glomerular diseases, a set of debilitating and complex disease entities, are related to mortality and morbidity. To gain insight into pathophysiology and novel treatment targets of glomerular disease, various types of biospecimens linked to deep clinical phenotyping including clinical information, digital pathology, and well-defined outcomes are required. We provide the rationale and design of the KOrea Renal biobank NEtwoRk System TOward Next-generation analysis (KORNERSTONE). Method The KORNERSTONE, which has been initiated by Korea Centres for Disease Control and Prevention, is designed as a multi-centre, prospective cohort study and biobank for glomerular diseases. Clinical data, questionnaires will be collected at the time of kidney biopsy and subsequently every one year after kidney biopsy. All of the clinical data will be extracted from the electrical health record and automatically uploaded to the web-based database. High-quality digital pathologies are obtained and connected in the database. Various types of biospecimens are collected at baseline and during follow-up: serum, urine, buffy coat, stool, glomerular complementary DNA (cDNA), tubulointerstitial cDNA. All data and biospecimens are processed and stored in a standardised manner. The primary outcomes are mortality and end-stage renal disease. The secondary outcomes will be deterioration renal function, remission of proteinuria, cardiovascular events and quality of life. Disussion Ethical approval has been obtained from the institutional review board of each participating centre and ethics oversight committee. The KORNERSTONE is designed to deliver pioneer insights into glomerular diseases. The study design allows comprehensive, integrated and high-quality data collection on baseline laboratory findings, clinical outcomes including administrative data and digital pathologic images. This may provide various biospecimens and information to many researchers, establish the rationale for future more individualised treatment strategies for glomerular diseases. Conclusion In conclusion, we describe the objectives and clinical protocol for the KORNERSTONE. As the first large-scale glomerulonephropathy cohort study with the integration of clinical data, biospecimens and digital pathologic images in Korea, the KORNERSTONE will help to clarify the natural course, complication profiles, and novel treatment targets of the Asian population with glomerular disease.

Initial correlative studies from a trial of cetuximab and pembrolizumab in metastatic colorectal cancer (mCRC).

4062 Background: Cetuximab is an EGFR-targeting IgG1 mAb. Pre-clinical data suggests cetuximab induces CD8+ cytotoxic T-cell (CTL) infiltration of tumors. We hypothesized that augmentation of CTLs in the tumor microenvironment (TME) may provide the proper milieu for effective PD-1 inhibition in metastatic colorectal cancer (mCRC). We conducted a phase Ib/II study of cetuximab with the PD-1 antibody, pembrolizumab, in mCRC. Correlative blood and tissue samples were collected to assess the impact of this treatment on CTLs, as well as potential compensatory alterations in regulatory T-cells (Tregs) and suppressive MDSCs (NCT02713373). Methods: 3 week treatment cycles included pembrolizumab at 200 mg on day 1 and cetuximab 250 mg/m2 following the 400 mg/m2 loading dose. Tumor biopsies were obtained at baseline and at c4d1 (Day 64). Peripheral blood (PB) was drawn at baseline, c2d1 (day 22) and c4d1 (Day 64). Flow cytometry was performed within 24 hours with additional samples stored for future analysis. In the present analysis, we assessed changes in levels of the cellular populations of interest between cycle 4 and cycle 1. Results: Forty-two RAS-wt patients were enrolled through October 2019. Paired tumor tissues were successfully analyzed for 16 patients and PB for 38. Intratumoral CTLs (CD3+CD8+) increased significantly (+47%, p < .05). In PB, there was a slight overall decrease in PB CTLs (-5%, p = NS) and a significant decrease in CD8+CD45RO+PD1+ cells (-42%, p < .05). We saw simultaneous decreases in PD-1+ CTLs in the tumor and PB. There was a trend for increase in Tregs (CD4+ Cd25+ FoxP3+) in PB (+11%, p = NS), but an overall increase in the Teff:Treg ratio (+30%, p = NS). CD4+CTLA4+ cells significantly increased (+37%, p < .05). Granulocytic MDSCs (CD11b+CD14−CD33+HLADR−) in PB decreased significantly on-treatment (-30%, p < 0.5). The sample size and tissue limitations prohibited meaningful evaluation of tissue Tregs and MDSCs via the present methods. Conclusions: Cetuximab and pembrolizumab induced dynamic changes in the TME and PB. The treatment associated increase in intratumoral CTLs was particularly pronounced, consistent with their local expansion and/or influx. This was accompanied by a decrease in PB CTLs. Decreases in multiple PD-1+ lymphoid populations were observed in both tumor and PB, notably PD-1+ CTLs. Of note, we saw a synchronous increase in immunosuppressive CD4+CTLA4+ T-cells in PB. Patient outcomes are pending maturation. Further analyses are planned, coupled with integration of clinical data. Clinical trial information: NCT02713373 .

Error-free, automated data integration of exosome cargo protein data with extensive clinical data in an ongoing, multi-omic translational research study.

e16743 Background: Major advances in understanding the biology of cancer have come from genomic analysis of tumor and normal tissue. Integrating extensive patient-related data with deep analysis of omic data is crucial to informing omic data interpretation. Currently, such integrations are a highly manual, asynchronous, and costly process as well as error-prone and time-consuming. To develop new blood assays that may detect very early stage PDAC, a multi-omic investigation with deep clinical annotation is needed. Using pilot data from an on-going study, we test a new platform allowing automated error-free integration of an extensive clinical database with extensive omic data. Methods: Demographic, clinical, family pedigree and pathology data were collected on the Rave EDC platform. Exosomes were purified from 46 plasma samples from 14 controls and 24 PDAC patients and cargo proteins were quantified via SILAC. The Rave Omics platform was used to ingest and integrate clinical and omic data, run quality checks and generate integrated clinical-omic datasets. Data fidelity was validated by systematically computing differences between corresponding values in the source flies with those present in the extracted data object (integrated data). The root mean squared error (RMSE) was calculated for numeric values in each sample. Additional validation was conducted by manual inspection to ascertain data integrity. Results: We demonstrated automatic integration, without human intervention, of a subset of the clinical data and all available SILAC data into an analysis-ready data object. Data transfer was completely faithful, with 100% concordance between the source and the integrated data without loss of features. All proteins (n = 1515) and clinical variables (n = 64) were imported. Their nomenclature and corresponding sample values (n = 69690) and clinical values (n = 2432) matched exactly between datasets. In all samples, the RMSE was exactly zero, indicating no deviation between data sources. Conclusions: We demonstrated that automatic, efficient, and reliable integration of clinical-omic data is achievable during an in-flight PDAC trial. Automatic exploratory analytics supporting biomarker discovery are currently being used to uncover associations between omic and clinical features. The Rave Omics platform is disease-agnostic and we plan to expand to trials of varying size, indication, and completion status where systematic, automated integration of clinical and (multi)omic data is needed.

Bases para a Reforma da Atenção Primária à Saúde no Brasil em 2019: mudanças estruturantes após 25 anos do Programa de Saúde da Família

O ano de 2019 marcou a estruturação das bases para uma profunda reforma na Atenção Primária à Saúde (APS) do Brasil. Os desafios enfrentados através dessa reforma foram a falta de priorização política real da APS, o financiamento insuficiente da APS e focado na estrutura de serviços, os obstáculos ao acesso de primeiro contato, a escassez de profissionais qualificados, a necessidade de maior informatização da APS e a ausência de integração de dados clínicos, a fragilidade clínica e necessidade de ampliação do escopo profissional e a falta de informação de qualidade para tomada de decisão clínica e gerencial. Com ações direcionadas a cada um desses desafios, a Secretaria de Atenção Primária à Saúde do Ministério da Saúde criou estratégias sistêmicas e integradas entre si que representam mudanças estruturantes e investimentos em governança clínica para transformar a APS brasileira, garantindo mais e melhor saúde para a população, com mecanismos transparentes e técnicos para seu financiamento, organização, oferta, monitoramento e avaliação.

Advancing clinical cohort selection with genomics analysis on a distributed platform.

The affordability of next-generation genomic sequencing and the improvement of medical data management have contributed largely to the evolution of biological analysis from both a clinical and research perspective. Precision medicine is a response to these advancements that places individuals into better-defined subsets based on shared clinical and genetic features. The identification of personalized diagnosis and treatment options is dependent on the ability to draw insights from large-scale, multi-modal analysis of biomedical datasets. Driven by a real use case, we premise that platforms that support precision medicine analysis should maintain data in their optimal data stores, should support distributed storage and query mechanisms, and should scale as more samples are added to the system. We extended a genomics-based columnar data store, GenomicsDB, for ease of use within a distributed analytics platform for clinical and genomic data integration, known as the ODA framework. The framework supports interaction from an i2b2 plugin as well as a notebook environment. We show that the ODA framework exhibits worst-case linear scaling for array size (storage), import time (data construction), and query time for an increasing number of samples. We go on to show worst-case linear time for both import of clinical data and aggregate query execution time within a distributed environment. This work highlights the integration of a distributed genomic database with a distributed compute environment to support scalable and efficient precision medicine queries from a HIPAA-compliant, cohort system in a real-world setting. The ODA framework is currently deployed in production to support precision medicine exploration and analysis from clinicians and researchers at UCLA David Geffen School of Medicine.

A systematic strategy for estimating hERG block potency and its implications in a new cardiac safety paradigm

IntroductionhERG block potency is widely used to calculate a drug's safety margin against its torsadogenic potential. Previous studies are confounded by use of different patch clamp electrophysiology protocols and a lack of statistical quantification of experimental variability. Since the new cardiac safety paradigm being discussed by the International Council for Harmonisation promotes a tighter integration of nonclinical and clinical data for torsadogenic risk assessment, a more systematic approach to estimate the hERG block potency and safety margin is needed. MethodsA cross-industry study was performed to collect hERG data on 28 drugs with known torsadogenic risk using a standardized experimental protocol. A Bayesian hierarchical modeling (BHM) approach was used to assess the hERG block potency of these drugs by quantifying both the inter-site and intra-site variability. A modeling and simulation study was also done to evaluate protocol-dependent changes in hERG potency estimates. ResultsA systematic approach to estimate hERG block potency is established. The impact of choosing a safety margin threshold on torsadogenic risk evaluation is explored based on the posterior distributions of hERG potency estimated by this method. The modeling and simulation results suggest any potency estimate is specific to the protocol used. DiscussionThis methodology can estimate hERG block potency specific to a given voltage protocol. The relationship between safety margin thresholds and torsadogenic risk predictivity suggests the threshold should be tailored to each specific context of use, and safety margin evaluation may need to be integrated with other information to form a more comprehensive risk assessment.

FHIR PIT: an open software application for spatiotemporal integration of clinical data and environmental exposures data

BackgroundInformatics tools to support the integration and subsequent interrogation of spatiotemporal data such as clinical data and environmental exposures data are lacking. Such tools are needed to support research in environmental health and any biomedical field that is challenged by the need for integrated spatiotemporal data to examine individual-level determinants of health and disease.ResultsWe have developed an open-source software application—FHIR PIT (Health Level 7 Fast Healthcare Interoperability Resources Patient data Integration Tool)—to enable studies on the impact of individual-level environmental exposures on health and disease. FHIR PIT was motivated by the need to integrate patient data derived from our institution’s clinical warehouse with a variety of public data sources on environmental exposures and then openly expose the data via ICEES (Integrated Clinical and Environmental Exposures Service). FHIR PIT consists of transformation steps or building blocks that can be chained together to form a transformation and integration workflow. Several transformation steps are generic and thus can be reused. As such, new types of data can be incorporated into the modular FHIR PIT pipeline by simply reusing generic steps or adding new ones. We validated FHIR PIT in the context of a driving use case designed to investigate the impact of airborne pollutant exposures on asthma. Specifically, we replicated published findings demonstrating racial disparities in the impact of airborne pollutants on asthma exacerbations.ConclusionsWhile FHIR PIT was developed to support our driving use case on asthma, the software can be used to integrate any type and number of spatiotemporal data sources at a level of granularity that enables individual-level study. We expect FHIR PIT to facilitate research in environmental health and numerous other biomedical disciplines.

A continuous-time Markov model approach for modeling myelodysplastic syndromes progression from cross-sectional data.

The integration of both genomics and clinical data to model disease progression is now possible, thanks to the increasing availability of molecular patients' profiles. This may lead to the definition of novel decision support tools, able to tailor therapeutic interventions on the basis of a "precise" patients' risk stratification, given their health status evolution. However, longitudinal analysis requires long-term data collection and curation, which can be time demanding, expensive and sometimes unfeasible. Here we present a clinical decision support framework that combines the simulation of disease progression from cross-sectional data with a Markov model that exploits continuous-time transition probabilities derived from Cox regression. Trajectories between patients at different disease stages are stochastically built according to a measure of patient similarity, computed with a matrix tri-factorization technique. Such trajectories are seen as realizations drawn from the stochastic process driving the transitions between the disease stages. Eventually, Markov models applied to the resulting longitudinal dataset highlight potentially relevant clinical information. We applied our method to cross-sectional genomic and clinical data from a cohort of Myelodysplastic syndromes (MDS) patients. MDS are heterogeneous clonal hematopoietic disorders whose patients are characterized by different risks of Acute Myeloid Leukemia (AML) development, defined by an international score. We computed patients' trajectories across increasing and subsequent levels of risk of developing AML, and we applied a Cox model to the simulated longitudinal dataset to assess whether genomic characteristics could be associated with a higher or lower probability of disease progression. We then used the learned parameters of such Cox model to calculate the transition probabilities of a continuous-time Markov model that describes the patients' evolution across stages. Our results are in most cases confirmed by previous studies, thus demonstrating that simulated longitudinal data represent a valuable resource to investigate disease progression of MDS patients.

Applying Analytics from the Edge to the Cloud in an Unified Scalable Healthcare Systems

Healthcare research is becoming increasingly large-scale and international. Cloud computing enables the comprehensive integration of research and clinical data, and the global sharing and collaborative processing of these data within a flexibly scalable infrastructure. Clouds offer novel research opportunities in research, as they facilitate cohort studies to be carried out at unprecedented scale, and they enable computer processing with superior pace and throughput, allowing researchers to address questions that could not be addressed by studies using limited cohorts. A well-developed example of such research is the arrhythmias Analysis of heart disease (Coronary artery disease), which involves the analysis of petabyte-scale electronic health records datasets from research centers in different locations or countries and different jurisdictions. Predictive analytics in healthcare plays a ma or role in finding the presence of disease and its severity using observations and symptoms associated with patients Health records. Clustering is a process of discovering relationships among data in datasets. It is also used for predicting relationships in the results discovered. Machine Learning is widely used in various applications such as business organizations, e-commerce, and health care industry, scientific and engineering for predicting and discovering relationships among data. In the health care industry, the predictive analytics in Machine Learning is mainly used for Disease Prediction. The ob ective of work is to predict the heart disease with reduced number of features from Electro Cardio Images. Early detection of disease and by mapping the drug side effects with patient histories is the needed approach. The real time applications of knowledge acquisition of health care data research require unified deep learning healthcare systems. This paper presents deep learning service share model architecture for the generalizations of knowledge processing which is available in form of cloud and by using various parameters which enhances the assistive intelligence.

Improving prediction performance of colon cancer prognosis based on the integration of clinical and multi-omics data

BackgroundColon cancer is common worldwide and is the leading cause of cancer-related death. Multiple levels of omics data are available due to the development of sequencing technologies. In this study, we proposed an integrative prognostic model for colon cancer based on the integration of clinical and multi-omics data.MethodsIn total, 344 patients were included in this study. Clinical, gene expression, DNA methylation and miRNA expression data were retrieved from The Cancer Genome Atlas (TCGA). To accommodate the high dimensionality of omics data, unsupervised clustering was used as dimension reduction method. The bias-corrected Harrell’s concordance index was used to verify which clustering result provided the best prognostic performance. Finally, we proposed a prognostic prediction model based on the integration of clinical data and multi-omics data. Uno’s concordance index with cross-validation was used to compare the discriminative performance of the prognostic model constructed with different covariates.ResultsCombinations of clinical and multi-omics data can improve prognostic performance, as shown by the increase of the bias-corrected Harrell’s concordance of the prognostic model from 0.7424 (clinical features only) to 0.7604 (clinical features and three types of omics features). Additionally, 2-year, 3-year and 5-year Uno’s concordance statistics increased from 0.7329, 0.7043, and 0.7002 (clinical features only) to 0.7639, 0.7474 and 0.7597 (clinical features and three types of omics features), respectively.ConclusionIn conclusion, this study successfully combined clinical and multi-omics data for better prediction of colon cancer prognosis.

Paediatric Malignant Blue Cell Tumours- A Practical Pathological and Immunohistochemical Study in Duhok, Iraq

Introduction: Malignant Blue Cell Tumours (MBCTs) are common heterogeneous tumours among paediatric age group. Their heterogeneity is reflected in their therapeutic and prognostic diversity. Mere morphology is not enough to differentiate them. Immunohistochemistry is a key tool for identification of small blue cell tumours that lack evidence of lineage differentiation on the ground of light microscopic morphology. Aim: To identify the immunohistochemical identity of paediatric MBCTs in Duhok, Iraq. Materials and Methods: This was a cross-sectional study performed on 120 cancers reported morphologically as MBCTs over 11-year period, from January 2009 to December 2019. Clinical data and histomorphologic acumen were considered and integrated with the immunohistochemical findings. Applying autostainer, immunohistochemistry was performed using monoclonal or polyclonal antibodies. Results: Lymphoma/leukaemia topped the diagnostic list 36 (30%) formed the commonest category, followed by Ewing’s/Primitive Neuro Ectodermal Tumour (PNET) 29 (24.2%), Neuroblastoma 16 (13.3%), Wilm’s tumour 11 (9.2%), Rhabdomyosarcoma 9 (7.5%) and Medulloblastoma 7 (5.8%). The remainders comprised Retinoblastoma (3.4%), Glioblastoma and Ependymoma (2.5% each). Hepatoblastoma and Astroblastoma formed the least frequent tumours (0.8% each). These tumours were more frequently located in the soft tissue (30%) followed by brain (14.2%), bone (10%), Lymph Node (LN) (9.2%) and kidney (10%). Conclusion: Integration of clinical data and histomorphologic acumen helped much for categorisation of MBCTs. Application of immunohistochemistry in this study has shown a significant improvement in the diagnostic accuracy of paediatric MBCTs. Loss of some differentiation antigens and aberrant expression of some markers often necessitate the use of panels of antibodies and, even molecular testing to target the task.

Assessment of the prognostic significance of low gradient severe aortic stenosis and preserved left ventricular function requires the integration of the consistency of stroke volume calculation and clinical data.

This study was to evaluate the prognostic significance of low gradient severe aortic stenosis (LG SAS) and preserved left ventricular ejection fraction (LVEF) with the integration of echocardiographic and clinical data. The study included 172 patients with LG SAS (AVAi≤0.6cm2 /m2 , mean aortic pressure gradient<40mmHg) and LVEF (≥ 50%). LV outflow tract diameters were measured at both the aortic valve annulus and 5mm below the annulus for the measurement consistency. Patients were divided into the low flow LG SAS (LF/LG SAS: SVi<35mL/m2 and AVAi≤0.6cm2 /m2 ) and normal-flow LG SAS groups (NF/LG SAS: SVi≥35mL/m2 and AVAi≤0.6cm2 /m2 ). Echocardiographic findings and clinical data were systematically analyzed with mean follow-up of 3.0±1.6years. LF/LG SAS had significantly smaller AVAi, lower SVi, a higher prevalence of atrial fibrillation (28% vs 12% P=.01) and diabetes (47% vs 27% P=.007) and lower 3-year cumulative survival than NF/LG SAS. Multivariable analysis showed that dyspnea, renal dysfunction (CI 1.42-3.99, P<.01), left atrial diameter, and SVi were independently associated with an increased risk for all-cause mortality. Aortic valve intervention (AVI) improved survival in LF/LG SAS (68% vs 48%, P<.05) in comparison with medical management (HR: 4.20, CI: 1.12-15.76, P=.03), but only modestly in NF/LG SAS (75% vs 65% P>.05). Outcome of LG SAS was independently associated with clinical characteristics. AVI likely improved outcome of LF/LG SAS who had high-risk clinical characteristics and unfavorable echocardiographic findings.

Multimodality Imaging of Dementia: Clinical Importance and Role of Integrated Anatomic and Molecular Imaging.

Neurodegenerative diseases are a devastating group of disorders that can be difficult to accurately diagnose. Although these disorders are difficult to manage owing to relatively limited treatment options, an early and correct diagnosis can help with managing symptoms and coping with the later stages of these disease processes. Both anatomic structural imaging and physiologic molecular imaging have evolved to a state in which these neurodegenerative processes can be identified relatively early with high accuracy. To determine the underlying disease, the radiologist should understand the different distributions and pathophysiologic processes involved. High-spatial-resolution MRI allows detection of subtle morphologic changes, as well as potential complications and alternate diagnoses, while molecular imaging allows visualization of altered function or abnormal increased or decreased concentration of disease-specific markers. These methodologies are complementary. Appropriate workup and interpretation of diagnostic studies require an integrated, multimodality, multidisciplinary approach. This article reviews the protocols and findings at MRI and nuclear medicine imaging, including with the use of flurodeoxyglucose, amyloid tracers, and dopaminergic transporter imaging (ioflupane). The pathophysiology of some of the major neurodegenerative processes and their clinical presentations are also reviewed; this information is critical to understand how these imaging modalities work, and it aids in the integration of clinical data to help synthesize a final diagnosis. Radiologists and nuclear medicine physicians aiming to include the evaluation of neurodegenerative diseases in their practice should be aware of and familiar with the multiple imaging modalities available and how using these modalities is essential in the multidisciplinary management of patients with neurodegenerative diseases.©RSNA, 2020.

A novel framework for horizontal and vertical data integration in cancer studies with application to survival time prediction models

BackgroundRecently high-throughput technologies have been massively used alongside clinical tests to study various types of cancer. Data generated in such large-scale studies are heterogeneous, of different types and formats. With lack of effective integration strategies novel models are necessary for efficient and operative data integration, where both clinical and molecular information can be effectively joined for storage, access and ease of use. Such models, combined with machine learning methods for accurate prediction of survival time in cancer studies, can yield novel insights into disease development and lead to precise personalized therapies.ResultsWe developed an approach for intelligent data integration of two cancer datasets (breast cancer and neuroblastoma) − provided in the CAMDA 2018 ‘Cancer Data Integration Challenge’, and compared models for prediction of survival time. We developed a novel semantic network-based data integration framework that utilizes NoSQL databases, where we combined clinical and expression profile data, using both raw data records and external knowledge sources. Utilizing the integrated data we introduced Tumor Integrated Clinical Feature (TICF) − a new feature for accurate prediction of patient survival time. Finally, we applied and validated several machine learning models for survival time prediction.ConclusionWe developed a framework for semantic integration of clinical and omics data that can borrow information across multiple cancer studies. By linking data with external domain knowledge sources our approach facilitates enrichment of the studied data by discovery of internal relations. The proposed and validated machine learning models for survival time prediction yielded accurate results.ReviewersThis article was reviewed by Eran Elhaik, Wenzhong Xiao and Carlos Loucera.

Deep learning, reusable and problem-based architectures for detection of consolidation on chest X-ray images

Background and objectiveIn most patients presenting with respiratory symptoms, the findings of chest radiography play a key role in the diagnosis, management, and follow-up of the disease. Consolidation is a common term in radiology, which indicates focally increased lung density. When the alveolar structures become filled with pus, fluid, blood cells or protein subsequent to a pulmonary pathological process, it may result in different types of lung opacity in chest radiograph. This study aims at detecting consolidations in chest x-ray radiographs, with a certain precision, using artificial intelligence and especially Deep Convolutional Neural Networks to assist radiologist for better diagnosis. MethodsMedical image datasets usually are relatively small to be used for training a Deep Convolutional Neural Network (DCNN), so transfer learning technique with well-known DCNNs pre-trained with ImageNet dataset are used to improve the accuracy of the models. ImageNet feature space is different from medical images and in the other side, the well-known DCNNs are designed to achieve the best performance on ImageNet. Therefore, they cannot show their best performance on medical images. To overcome this problem, we designed a problem-based architecture which preserves the information of images for detecting consolidation in Pediatric Chest X-ray dataset. We proposed a three-step pre-processing approach to enhance generalization of the models. To demonstrate the correctness of numerical results, an occlusion test is applied to visualize outputs of the model and localize the detected appropriate area. A different dataset as an extra validation is used in order to investigate the generalization of the proposed model. ResultsThe best accuracy to detect consolidation is 94.67% obtained by our problem based architecture for the understudy dataset which outperforms the previous works and the other architectures. ConclusionsThe designed models can be employed as computer aided diagnosis tools in real practice. We critically discussed the datasets and the previous works based on them and show that without some considerations the results of them may be misleading. We believe, the output of AI should be only interpreted as focal consolidation. The clinical significance of the finding can not be interpreted without integration of clinical data.

Understanding temporal and spatial variations of viral disease in the US: The need for a one-health-based data collection and analysis approach

Viral diseases exhibit spatial and temporal variation, and there are many factors that can affect their occurrence. The identification of these factors is critical in the efforts to predict and lessen viral disease burden. Because viral infection is able to spread to humans from the environment, animals, and other humans, the One-Health framework can be used to investigate the critical pathways through which viruses are transported and transmitted. A holistic approach, incorporating publicly available clinical data for human, livestock, and wildlife disease occurrence, together with environmental data reported in federal and state databases such as parameters related to land use, environmental quality, and weather, can enhance the understanding of variations in disease patterns, leading to the design and implementation of surveillance systems. An example analysis approach is presented for Michigan, United States, which is a state with large urban centers as well as a sizeable rural and agricultural population. Analysis of publicly available data from 2017 indicates that gastrointestinal (GI) and influenza-associated illnesses in Michigan may have been related with agricultural land use to a higher extent than with developed land use during that year. Meanwhile, hepatitis A virus appears to be most closely related with developed land use in dense population areas. GI illnesses may be related to precipitation, and this relationship is strongest in the springtime, although GI illnesses are most common in the winter months. Integration of human-related clinical data, animal disease data, and environmental data can ultimately be used for prioritization of the most critical locations and times for viral outbreaks in both urban and rural environments.