DART Demonstrates Human-Relevant Cardiotoxicity Prediction with AI
Executive Summary
A U.S.-based biopharmaceutical company specializing in small molecule therapeutics sought to enhance its cardiac safety screening process across a growing library of drug candidates. Traditional animal-based models proved slow, costly, and biologically limited in predicting human cardiotoxicity.
By implementing Quantiphi’s DART platform—integrating human cardiac microphysiological systems and AI-powered predictive analytics through the CardioSIGHT module—the company successfully decoded mechanistic toxicity insights at the cellular, molecular, and biochemical levels. This enabled faster, more confident go/no-go decisions and accelerated preclinical discovery timelines while maintaining regulatory readiness.
About the Client
A next-generation biopharma company focused on scaffold-driven small molecule design, developing novel therapeutics with improved safety profiles and faster lead optimization cycles.
- Industry: Life Sciences (Small molecule drug discovery)
- Size: ~100 employees
- Country: USA
Problem Statement
The client, a U.S.-based small molecule innovator, needed to evaluate cardiac safety risks across a diverse library of drug candidates modeled after known therapeutic scaffolds. Conventional animal-based testing methods lacked the speed, scalability, and human relevance necessary to support modern drug discovery pipelines. The company required a faster, AI-enabled, and regulatory-aligned solution capable of identifying cardiotoxicity risks early—allowing confident go/no-go decisions and more efficient candidate prioritization within preclinical development.
Challenges
- Animal models lacked scalability for high-throughput screening.
- Non-human systems failed to capture cardiac-specific human biology.
- Ethical limitations restricted reliance on in vivo testing.
- High screening costs and tight timelines slowed regulatory submissions.
The Solution
The DART platform provided a human-relevant, AI-powered testing solution that combined cardiac microphysiological systems with predictive analytics through its CardioSIGHT module. This integration allowed decoding of toxicity signals across cellular, molecular, and biochemical levels with human-level accuracy and throughput efficiency.
Key solution highlights:
- Human cardiac modeling: Leveraged microphysiological systems to simulate authentic cardiac function and response.
- AI-driven analytics: Used CardioSIGHT to identify mechanistic toxicity pathways through advanced phenotype-to-pathway mapping.
- High-throughput screening: Enabled early-stage assessment of compound libraries to prioritize safe candidates rapidly.
- Regulatory alignment: Provided validated, mechanistic insights consistent with preclinical safety pharmacology standards.
By integrating AI and human cardiac models, DART empowered the client to detect early cardiotoxicity indicators, reduce development costs, and make data-driven decisions without reliance on animal models.
Customer’s Testimonial
“We are satisfied with the insights provided regarding the risks of human cardiotoxicity. With the data now available, we are prepared to select the IND for further investment in our lead program.”
Clinical Development Scientist
Impacts
- Delivered human-relevant cardiotoxicity insights for prioritized molecules
- Supported confident internal go/no-go decisions with mechanistic evidence
- Reduced screening costs while improving speed and throughput
- Strengthened the basis for regulator-aligned cardiac safety review
Technologies Used
DART
CardioSIGHT
Human cardiac microphysiological systems (hMPS)
AI/ML-driven image and data analysis
Results
- Decoded cardiotoxicity risks with 100% relevance to human cardiac biology.
- Uncovered mechanistic toxicity insights using phenotype-to-pathway analysis.
- Enabled confident lead candidate selection for downstream discovery.
- Published human-relevant findings supporting internal and regulatory review.
- Accelerated discovery timelines with regulator-aligned cardiac safety data.