scholarly journals Investigating ADR mechanisms with Explainable AI: a feasibility study with knowledge graph mining

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Emmanuel Bresso ◽  
Pierre Monnin ◽  
Cédric Bousquet ◽  
François-Elie Calvier ◽  
Ndeye-Coumba Ndiaye ◽  
...  
Keyword(s):  
Clinical Trials ◽  
Drug Targets ◽  
Graph Mining ◽  
Molecular Mechanisms ◽  
Randomized Clinical Trials ◽  
Knowledge Graph ◽  
Drug Induced ◽  
Liver Injuries ◽  
Explainable Ai ◽  
Knowledge Graphs

Abstract Background Adverse drug reactions (ADRs) are statistically characterized within randomized clinical trials and postmarketing pharmacovigilance, but their molecular mechanism remains unknown in most cases. This is true even for hepatic or skin toxicities, which are classically monitored during drug design. Aside from clinical trials, many elements of knowledge about drug ingredients are available in open-access knowledge graphs, such as their properties, interactions, or involvements in pathways. In addition, drug classifications that label drugs as either causative or not for several ADRs, have been established. Methods We propose in this paper to mine knowledge graphs for identifying biomolecular features that may enable automatically reproducing expert classifications that distinguish drugs causative or not for a given type of ADR. In an Explainable AI perspective, we explore simple classification techniques such as Decision Trees and Classification Rules because they provide human-readable models, which explain the classification itself, but may also provide elements of explanation for molecular mechanisms behind ADRs. In summary, (1) we mine a knowledge graph for features; (2) we train classifiers at distinguishing, on the basis of extracted features, drugs associated or not with two commonly monitored ADRs: drug-induced liver injuries (DILI) and severe cutaneous adverse reactions (SCAR); (3) we isolate features that are both efficient in reproducing expert classifications and interpretable by experts (i.e., Gene Ontology terms, drug targets, or pathway names); and (4) we manually evaluate in a mini-study how they may be explanatory. Results Extracted features reproduce with a good fidelity classifications of drugs causative or not for DILI and SCAR (Accuracy = 0.74 and 0.81, respectively). Experts fully agreed that 73% and 38% of the most discriminative features are possibly explanatory for DILI and SCAR, respectively; and partially agreed (2/3) for 90% and 77% of them. Conclusion Knowledge graphs provide sufficiently diverse features to enable simple and explainable models to distinguish between drugs that are causative or not for ADRs. In addition to explaining classifications, most discriminative features appear to be good candidates for investigating ADR mechanisms further.

scholarly journals Peripheral Neuropathy (PN) with Immunomodulatory Drugs in Patients with Multiple Myeloma (MM)

Blood ◽  
2016 ◽  
Vol 128 (22) ◽  
pp. 5677-5677
Author(s):  
Melinda Mezo ◽  
Carmen Castaneda ◽  
Lilia Weiss ◽  
Neil Minton ◽  
Damien Hirsch ◽  
...  
Keyword(s):  
Clinical Trials ◽  
Peripheral Neuropathy ◽  
Dose Reduction ◽  
Randomized Clinical Trials ◽  
Employment Equity ◽  
Therapy Discontinuation ◽  
Drug Induced ◽  
Narrow Scope ◽  
Therapeutic Class ◽  
Equity Ownership

Abstract Background: PN is a recognized adverse event (AE) with thalidomide (THAL). Despite similarities between THAL and its analogues (lenalidomide [LEN] and pomalidomide [POM]) that include structural, binding target, and common drug-induced substrates, the result of unshared downstream molecular, cellular, and microenvironment effects among the compounds is a diverse array of biological responses and different efficacy and safety outcomes (Bjorklund CC, et al. Blood Cancer J. 2015;5:e354). This underscores the mechanistic diversity of a family of agents rather than the mechanistic uniformity of a therapeutic class. In untreated patients with MM, low incidence (3%-13%) of symptomatic PN has been reported, while the estimated rate was much higher (40%-60%) for subclinical PN (Kwan JY. Neurol Clin. 2007;25:47-69). Objective: To describe the frequency, severity, and tolerability of PN in patients with MM treated with THAL-, LEN-, and POM-based therapy from Celgene-sponsored pivotal and other registrational studies. Methods: AEs of new and worsening PN were analyzed for THAL-, LEN-, and POM-based therapy in 9 randomized clinical trials in 3518 patients with MM. Eight of the 9 clinical trials included patients with baseline grade 1 PN, while 1 clinical trial excluded all patients with PN. The search for PN used narrow-scope Standardised MedDRA Query "peripheral neuropathy." The AEs were rated per NCI-CTCAE. Results: PN was reported more frequently in patients in the THAL-based pool (36.2%; 460/1272) compared with the LEN-based pool (23.2%; 400/1727) and POM-based pool (12.3%; 64/519). Generally, PN was grade 1/2: 27.2% in the THAL-, 19.9% in the LEN-, and 11.4% in the POM-based pool. The frequency of patients with grade 3/4 PN events was 9.0% in the THAL-, 3.2% in the LEN-, and 1.0% in the POM-based pool. The frequencies of patients with serious AEs (SAEs) of PN were low (≤ 1%) for both THAL- and LEN-based pools; no patient had an SAE of PN in the POM-based pool. PN led to THAL-based therapy discontinuation in 7.0% and dose adjustments (reduction, modification, or interruption) in 16.7% of the patients. PN led to LEN-based therapy discontinuation in 0.6%, dose reduction in 2.5%, and interruption in 1.5% of the patients. PN led to POM-based therapy discontinuation in 0.4%, dose reduction in 0.6%, and interruption/reduction in 0.2% of the patients. Conclusions: The frequency and severity of PN and therapy discontinuation or modification is highest in THAL-treated patients with MM compared with LEN- and POM-treated patients. LEN and POM demonstrate improved safety for PN compared with THAL. Disclosures Mezo: Celgene: Employment, Equity Ownership. Castaneda:Celgene: Employment. Weiss:Celgene: Consultancy, Employment, Equity Ownership. Minton:Celgene: Employment, Equity Ownership. Hirsch:Celgene: Employment, Equity Ownership. Renz:Celgene: Employment. Arunagiri:Celgene: Employment. Brown:Celgene: Employment, Equity Ownership. Gambini:Celgene: Employment, Equity Ownership. Peng:Celgene: Employment. Yu:Celgene: Employment, Equity Ownership. Yu:Celgene: Employment. Freeman:Celgene: Employment.

scholarly journals Drug-induced cutaneous vasculitis developing during Etanercept treatment for rheumatoid arthritis

2021 ◽  
Vol 97 (4) ◽  
pp. 113-119
Author(s):  
Maria N. Chamurlieva ◽  
Yulia L. Korsakova ◽  
Stefka G. Radenska-Lopovok ◽  
Tatiana V. Korotaeva
Keyword(s):  
Rheumatoid Arthritis ◽  
Clinical Trials ◽  
Randomized Clinical Trials ◽  
Tnf Inhibitors ◽  
Cutaneous Vasculitis ◽  
Inflammatory Rheumatic Diseases ◽  
Tnf Inhibitor ◽  
Drug Induced ◽  
Etanercept Treatment ◽  
Increased Risk

Biological disease-modifying anti-rheumatic drugs (bDMARDs) are widely used for the treatment of chronic inflammatory rheumatic diseases. Since the introduction of tumor necrosis factor alpha (TNF-) inhibitors, the treatment of rheumatoid arthritis has been revolutionized. The approach of targeting TNF- has considerably improved the success of the treatment of rheumatoid arthritis. Their effectiveness has been extensively proven in randomized clinical trials and in clinical practice. Randomized clinical trials and post-marketing studies proved that patients undergoing TNF- inhibitors therapy are at increased risk of infectious disease, bacterial, viral, fungal, opportunistic, oncology and skin adverse effects such as psoriasis and angiitis of the skin. In this case report drug-induced cutaneous vasculitis developing during TNF- inhibitor (Etanercept) treatment for rheumatoid arthritis is described.

scholarly journals Syn-Lethality: An Integrative Knowledge Base of Synthetic Lethality towards Discovery of Selective Anticancer Therapies

2014 ◽  
Vol 2014 ◽  
pp. 1-7 ◽  
Author(s):  
Xue-juan Li ◽  
Shital K. Mishra ◽  
Min Wu ◽  
Fan Zhang ◽  
Jie Zheng
Keyword(s):  
Knowledge Base ◽  
Drug Targets ◽  
Molecular Mechanisms ◽  
Synthetic Lethality ◽  
Human Cancer ◽  
Single Gene ◽  
Drug Induced ◽  
Human Genes ◽  
A Cell ◽  
Novel Strategy

Synthetic lethality (SL) is a novel strategy for anticancer therapies, whereby mutations of two genes will kill a cell but mutation of a single gene will not. Therefore, a cancer-specific mutation combined with a drug-induced mutation, if they have SL interactions, will selectively kill cancer cells. While numerous SL interactions have been identified in yeast, only a few have been known in human. There is a pressing need to systematically discover and understand SL interactions specific to human cancer. In this paper, we present Syn-Lethality, the first integrative knowledge base of SL that is dedicated to human cancer. It integrates experimentally discovered and verified human SL gene pairs into a network, associated with annotations of gene function, pathway, and molecular mechanisms. It also includes yeast SL genes from high-throughput screenings which are mapped to orthologous human genes. Such an integrative knowledge base, organized as a relational database with user interface for searching and network visualization, will greatly expedite the discovery of novel anticancer drug targets based on synthetic lethality interactions. The database can be downloaded as a stand-alone Java application.

scholarly journals Clinical Trials of Cannabidiol for Substance Use Disorders: Outcome Measures, Surrogate Endpoints, and Biomarkers

2021 ◽  
Vol 12 ◽  
Author(s):  
Alix Morel ◽  
Pierre Lebard ◽  
Alexandra Dereux ◽  
Julien Azuar ◽  
Frank Questel ◽  
...  
Keyword(s):  
Clinical Trials ◽  
Substance Use ◽  
Substance Use Disorders ◽  
Outcome Measures ◽  
Randomized Clinical Trials ◽  
Cochrane Library ◽  
Surrogate Endpoints ◽  
Opioid Use ◽  
Short Term ◽  
Drug Induced

Background: Cannabidiol (CBD) is a cannabinoid of potential interest for the treatment of substance use disorders. Our aim was to review the outcome measures, surrogate endpoints, and biomarkers in published and ongoing randomized clinical trials.Methods: We conducted a search in PubMed, Web of Science, PMC, PsycINFO, EMBASE, CENTRAL Cochrane Library, “clinicalTrials.gov,” “clinicaltrialsregister.eu,” and “anzctr.org.au” for published and ongoing studies. Inclusion criteria were randomized clinical trials (RCTs) examining the use of CBD alone or in association with other cannabinoids, in all substance use disorders. The included studies were analyzed in detail and their qualities assessed by a standardized tool (CONSORT 2010). A short description of excluded studies, consisting in controlled short-term or single administration in non-treatment-seeking drug users, is provided.Findings: The screening retrieved 207 published studies, including only 3 RCTs in cannabis use disorder. Furthermore, 12 excluded studies in cannabis, tobacco, and opioid use disorders are described.Interpretation: Primary outcomes were validated withdrawal symptoms scales and drug use reduction in the three RCTs. In the short-term or crossover studies, the outcome measures were visual analog scales for subjective states; self-rated scales for withdrawal, craving, anxiety, or psychotomimetic symptoms; and laboratory tasks of drug-induced craving, effort expenditure, attentional bias for substance, impulsivity, or anxiety to serve as surrogate endpoints for treatment efficacy. Of note, ongoing studies are now adding peripheral biomarkers of the endocannabinoid system status to predict treatment response.Conclusion: The outcome measures and biomarkers assessed in the ongoing CBD trials for substance use disorders are improving.

scholarly journals Towards a Knowledge Graph-Based Explainable Decision Support System in Healthcare

2021 ◽  
Author(s):  
Enayat Rajabi ◽  
Kobra Etminani
Keyword(s):  
Decision Support ◽  
Decision Support Systems ◽  
Clinical Decision Support ◽  
Support System ◽  
Support Systems ◽  
Clinical Decision Support Systems ◽  
Clinical Decision ◽  
Knowledge Graph ◽  
Explainable Ai ◽  
Knowledge Graphs

The decisions derived from AI-based clinical decision support systems should be explainable and transparent so that the healthcare professionals can understand the rationale behind the predictions. To improve the explanations, knowledge graphs are a well-suited choice to be integrated into eXplainable AI. In this paper, we introduce a knowledge graph-based explainable framework for AI-based clinical decision support systems to increase their level of explainability.

L-arginine effect on inflammatory mediators: A systematic review of randomized controlled clinical trials

2019 ◽  
pp. 1-9 ◽  
Author(s):  
Seyed Reza Mirhafez ◽  
Mitra Hariri
Keyword(s):  
Systematic Review ◽  
Clinical Trials ◽  
Inflammatory Mediators ◽  
Randomized Clinical Trials ◽  
Randomized Controlled Clinical Trials ◽  
Patients With Cancer ◽  
Reactive Protein ◽  
Randomized Controlled ◽  
Human Adults ◽  
Factor Α

Abstract. L-arginine is an important factor in several physiological and biochemical processes. Recently, scientists studied L-arginine effect on inflammatory mediators such as C-reactive protein (CRP), tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6). We conducted a systematic review on randomized controlled trials assessing L-arginine effect on inflammatory mediators. We searched data bases including Google scholar, ISI web of science, SCOPUS, and PubMed/Medline up to April 2019. Randomized clinical trials assessing the effect of L-arginine on inflammatory mediators in human adults were included. Our search retrieved eleven articles with 387 participants. Five articles were on patients with cancer and 6 articles were on adults without cancer. L-arginine was applied in enteral form in 5 articles and in oral form in 6 articles. Eight articles were on both genders, two articles were on women, and one article was on men. L-arginine could not reduce inflammatory mediators among patients with and without cancer except one article which indicated that taking L-arginine for 6 months decreased IL-6 among cardiopathic nondiabetic patients. Our results indicated that L-arginine might not be able to reduce selected inflammatory mediators, but for making a firm decision more studies are needed to be conducted with longer intervention duration, separately on male and female and with different doses of L-arginine.

Ongoing randomized clinical trials for the treatment of thrombosis in the antiphospholipid syndrome

2001 ◽  
Vol 21 (02) ◽  
pp. 77-81 ◽  
Author(s):  
G. Finazzi
Keyword(s):  
Clinical Trials ◽  
Large Scale ◽  
Low Dose ◽  
Randomized Clinical Trials ◽  
Oral Anticoagulants ◽  
Collaborative Study ◽  
Clinical Problem ◽  
Vascular Events ◽  
Dose Oral ◽  
Adverse Pregnancy

SummaryThrombotic events are a major clinical problem for patients with antiphospholipid antibodies (APA). However, current recommendations for their prevention and treatment are still based on retrospective studies. Data from large scale, prospective clinical trials are required to ultimately identify the optimal management of these patients. To date, at least four randomized studies are underway. The WAPS and PAPRE clinical trials are aimed to establish the correct duration and intensity of oral anticoagulation in APA patients with major arterial or venous thrombosis. The WARSS-APASS is a collaborative study to evaluate the efficacy and safety of aspirin or low-dose oral anticoagulants in preventing the recurrence of ischemic stroke. The recently announced UK Trial compares low-dose aspirin with or without low-intensity anticoagulation for the primary prevention of vascular events in APA-positive patients with SLE or adverse pregnancy history, but still thrombosis-free. It is hoped that the results of these trials will be available soon since clinicians urgently need more powerful data to treat their patients with the APA syndrome.

Sign in / Sign up

Export Citation Format

Share Document