Knowledge Graph for Cardiovascular Drug Safety and Pharmacovigilance: A Scoping Review.

Area of Science:

• Biomedical Informatics
• Pharmacovigilance
• Cardiovascular Research

Background:

• Cardiovascular diseases (CVDs) are a leading global cause of death.
• Adverse drug reactions (ADRs) pose significant challenges in managing CVDs.
• Knowledge graphs (KGs) offer a novel approach to integrate complex biomedical data for enhanced pharmacovigilance (PV).

Purpose of the Study:

• To conduct a scoping review of current applications of KGs in cardiovascular drug safety.
• To identify methodologies, technologies, and existing evidence gaps in this field.
• To assess the potential of KGs in improving pharmacovigilance for cardiovascular drugs.

Main Methods:

• Systematic literature search across six databases following PRISMA-ScR guidelines.
• Inclusion of studies utilizing KGs or semantic technologies for cardiovascular disease (CVD)-related PV tasks.
• Analysis of identified studies focusing on methodologies, data integration, and application areas.

Main Results:

• Six eligible studies were identified, focusing on signal detection, mechanistic interpretation, and class-effect analysis.
• KGs successfully integrated clinical, molecular, and pharmacological data using standard ontologies (MedDRA, RxNorm, HPO, GO).
• KGs demonstrated potential in improving mechanistic reasoning and hypothesis generation for drug safety.

Conclusions:

• KGs show significant potential for advancing cardiovascular pharmacovigilance.
• Current limitations include the need for more extensive validation against standard PV methods.
• Enhancing data completeness and clinical generalizability, alongside integration with Electronic Health Records (EHRs), is crucial for realizing clinical impact.