Identifying and Validating Networks of Oncology Biomarkers Mined From the Scientific Literature

Biomarkers, as measurements of defined biological characteristics, can play a pivotal role in estimations of disease risk, early detection, differential diagnosis, assessment of disease progression and outcomes prediction. Studies of cancer biomarkers are published daily; some are well characterized, while others are of growing interest. Managing this flow of information is challenging for scientists and clinicians. We sought to develop a novel text-mining method employing biomarker co-occurrence processing applied to a deeply indexed full-text database to generate time-interval–delimited biomarker co-occurrence networks. Biomarkers across 6 cancer sites and a cancer-agnostic network were successfully characterized in terms of their emergence in the published literature and the context in which they are described. Our approach, which enables us to find publications based on biomarker relationships, identified biomarker relationships not known to existing interaction networks. This search method finds relevant literature that could be missed with keyword searches, even if full text is available. It enables users to extract relevant biological information and may provide new biological insights that could not be achieved by individual review of papers.

Abstract 252: Navigating networks of oncology biomarkers mined from the scientific literature: A new open research tool

Introduction: Oncology biomarkers have a pivotal role in diagnosis, prognosis, assessment of disease progression and the prediction of treatment efficacy. There are > 5000 known biomarkers; some are well characterised, while others are emerging as biomarkers of potential interest. Biomarker panels, incorporating combinations of biomarkers, are increasingly used to enhance the power and precision of testing.

Objective: We sought to identify key emerging biomarkers and biomarker relationships by applying analytical techniques to the corpus of recent scientific publications using an artificial intelligence (AI)-driven platform.

Methods: Biomarker targets were obtained from the Early Detection Research Network. Recent (<5 years old) publications mentioning biomarkers in six cancer types (bladder, breast, colorectal, lung, prostate and renal) were identified using Dimensions (a linked research knowledge system); emerging biomarkers were identified by number and growth in publications. For selected biomarkers, full text proximity searching was used to identify biomarker co-occurrence, which was visualized and characterized using network analysis and co-occurrence heatmaps per cancer type. Relationships between biomarkers were also assessed in an indication-agnostic dataset. Key biomarkers were identified by a measure of their network importance.

Results: A total of 720 biomarkers were analyzed, and 41 clusters identified across the cancer types. Data visualizations are available at https://reports.dimensions.ai/mined-oncology-biomarkers/

Conclusion: Using large-scale analytics of published literature, biomarkers across six cancer types were successfully characterized in terms of their emergence in the published literature and the context in which they are described. This novel AI-driven approach could help identify biomarkers and biomarker panels for exploration in a clinical setting.

Example clusters - growth rate is compound annual growth rate (CAGR)Cancer typeBiomarkers (n)Publications (n)Cluster key biomarkersCluster publication growth rate (%)Bladder49168CAV1/CXCL8/THBS153.1Breast2651593AREG/CXCL10/CXCL832.6Colorectal193921CAV1/CDCP1/MTHFR31.6Lung2231061SFTPC/C9/GPI41.4Prostate166717FABP5/FLNA/CAV131.6Renal44185CXCL10/CXCL8/CSNK2A149.5

Citation Format: Kim Wager, Dheepa Chari, Steffan Ho, Tomas J. Rees, Robert J. Schijvenaars. Navigating networks of oncology biomarkers mined from the scientific literature: A new open research tool [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 252.

HR-pQCT Measures of Bone Microarchitecture Predict Fracture: Systematic Review and Meta-Analysis

High-resolution peripheral quantitative computed tomography (HR-pQCT) is a noninvasive imaging modality for assessing volumetric bone mineral density (vBMD) and microarchitecture of cancellous and cortical bone. The objective was to (1) assess fracture-associated differences in HR-pQCT bone parameters; and (2) to determine if HR-pQCT is sufficiently precise to reliably detect these differences in individuals. We systematically identified 40 studies that used HR-pQCT (39/40 used XtremeCT scanners) to assess 1291 to 3253 and 3389 to 10,687 individuals with and without fractures, respectively, ranging in age from 10.9 to 84.7 years with no comorbid conditions. Parameters describing radial and tibial bone density, microarchitecture, and strength were extracted and percentage differences between fracture and control subjects were estimated using a random effects meta-analysis. An additional meta-analysis of short-term in vivo reproducibility of bone parameters assessed by XtremeCT was conducted to determine whether fracture-associated differences exceeded the least significant change (LSC) required to discern measured differences from precision error. Radial and tibial HR-pQCT parameters, including failure load, were significantly altered in fracture subjects, with differences ranging from -2.6% (95% confidence interval [CI] -3.4 to -1.9) in radial cortical vBMD to -12.6% (95% CI -15.0 to -10.3) in radial trabecular vBMD. Fracture-associated differences reported by prospective studies were consistent with those from retrospective studies, indicating that HR-pQCT can predict incident fracture. Assessment of study quality, heterogeneity, and publication biases verified the validity of these findings. Finally, we demonstrated that fracture-associated deficits in total and trabecular vBMD and certain tibial cortical parameters can be reliably discerned from HR-pQCT-related precision error and can be used to detect fracture-associated differences in individual patients. Although differences in other HR-pQCT measures, including failure load, were significantly associated with fracture, improved reproducibility is needed to ensure reliable individual cross-sectional screening and longitudinal monitoring. In conclusion, our study supports the use of HR-pQCT in clinical fracture prediction. © 2019 American Society for Bone and Mineral Research.

Neurodegeneration and Epilepsy in a Zebrafish Model of CLN3 Disease (Batten Disease)

The neuronal ceroid lipofuscinoses are a group of lysosomal storage disorders that comprise the most common, genetically heterogeneous, fatal neurodegenerative disorders of children. They are characterised by childhood onset, visual failure, epileptic seizures, psychomotor retardation and dementia. CLN3 disease, also known as Batten disease, is caused by autosomal recessive mutations in the CLN3 gene, 80–85% of which are a ~1 kb deletion. Currently no treatments exist, and after much suffering, the disease inevitably results in premature death. The aim of this study was to generate a zebrafish model of CLN3 disease using antisense morpholino injection, and characterise the pathological and functional consequences of Cln3 deficiency, thereby providing a tool for future drug discovery. The model was shown to faithfully recapitulate the pathological signs of CLN3 disease, including reduced survival, neuronal loss, retinopathy, axonopathy, loss of motor function, lysosomal storage of subunit c of mitochondrial ATP synthase, and epileptic seizures, albeit with an earlier onset and faster progression than the human disease. Our study provides proof of principle that the advantages of the zebrafish over other model systems can be utilised to further our understanding of the pathogenesis of CLN3 disease and accelerate drug discovery.

The neuronal ceroid lipofuscinoses: Opportunities from model systems

The neuronal ceroid lipofuscinoses are a group of severe and progressive neurodegenerative disorders, generally with childhood onset. Despite the fact that these diseases remain fatal, significant breakthroughs have been made in our understanding of the genetics that underpin these conditions. This understanding has allowed the development of a broad range of models to study disease processes, and to develop new therapeutic approaches. Such models have contributed significantly to our knowledge of these conditions. In this review we will focus on the advantages of each individual model, describe some of the contributions the models have made to our understanding of the broader disease biology and highlight new techniques and approaches relevant to the study and potential treatment of the neuronal ceroid lipofuscinoses. This article is part of a Special Issue entitled: "Current Research on the Neuronal Ceroid Lipofuscinoses (Batten Disease)".

Modelling inborn errors of metabolism in zebrafish

The majority of human inborn errors of metabolism are fatal multisystem disorders that lack proper treatment and have a poorly understood mechanistic basis. Novel technologies are required to address this issue, and the use of zebrafish to model these diseases is an emerging field. Here we present the published zebrafish models of inborn metabolic diseases, discuss their validity, and review the novel mechanistic insights that they have provided. We also review the available methods for creating and studying zebrafish disease models, advantages and disadvantages of using this model organism, and successful examples of the use of zebrafish for drug discovery and development. Using a zebrafish to model inborn errors of metabolism in vivo, although still in its infancy, shows promise for a deeper understanding of disease pathomechanisms, onset, and progression, and also for the development of specific therapies.

HIM roads less traveled. Interview by Michele Hogan

Increasingly, HIM professionals are taking unique roads in their professional lives--into other care settings and into boardrooms, classrooms, and more. How have their work lives changed? Four who chose less traveled roads tell their stories here.