VivoSim's AI-Powered Models Aim to De-Risk Next-Gen Cancer Drugs

SAN DIEGO, CA – February 11, 2026 – As the pharmaceutical industry races to develop more potent and targeted cancer therapies, a San Diego-based company is poised to address one of the biggest hurdles in drug development: predicting toxic side effects. VivoSim Labs, Inc. has announced it will present groundbreaking data at the upcoming Society of Toxicology (SOT) meeting, demonstrating that its advanced human tissue models can accurately forecast the safety profiles of a powerful class of treatments known as Antibody Drug Conjugates (ADCs).

This development comes at a pivotal moment, as regulators actively push for a move away from traditional animal testing. The data, which VivoSim claims shows a “close correlation” with known clinical outcomes for marketed ADCs, could represent a significant step forward in making drug development faster, cheaper, and safer for patients.

A New Paradigm in Preclinical Safety

The landscape of preclinical drug testing is undergoing a seismic shift, driven by both ethical considerations and scientific limitations of traditional methods. For decades, animal models have been the gold standard, but their ability to predict human responses has often fallen short, contributing to high failure rates in clinical trials. This long-standing challenge prompted a landmark policy change in the United States.

On April 10, 2025, the U.S. Food and Drug Administration (FDA) unveiled its “Roadmap to Reducing Animal Testing in Preclinical Safety Studies.” Spurred by the FDA Modernization Act 2.0, the initiative explicitly authorized and encouraged the use of non-animal alternatives, or New Approach Methodologies (NAMs), for drug applications. The agency’s goal is to integrate methods like AI-based computational models, 3D organoids, and organ-on-a-chip systems across all areas of drug safety testing, with the long-term vision of making animal studies the exception rather than the rule.

VivoSim Labs is directly answering this call. The company specializes in creating sophisticated 3D models of human liver and intestinal tissues using primary human cells. These “NAMkind™” models are designed to replicate the complex biology of human organs in a controlled laboratory setting, offering a more relevant testing ground for new drug candidates. The industry has largely welcomed this regulatory pivot, viewing NAMs as a pathway to more informative data, accelerated development timelines, and ultimately, reduced R&D costs.

“VivoSim’s scientific leadership in the field of human-relevant NAM models with results that translate to the clinic is bolstered by these new data,” said Keith Murphy, VivoSim’s Executive Chairman, in a statement. “Our ability to demonstrate results that match known clinical results of cutting-edge drug modalities confirms that our 3D human cell models have a strong ability to faithfully reproduce the complex biology of human tissues.”

Targeting the Toxicity of Advanced Cancer Therapies

Nowhere is the need for better predictive models more urgent than in the field of oncology, particularly with the rapid expansion of Antibody Drug Conjugates. ADCs are a revolutionary class of drugs that act like guided missiles, using a monoclonal antibody to deliver a potent cytotoxic payload directly to cancer cells. The promise of this approach has ignited the market, with industry data showing well over 400 ADCs in various stages of development and a global market size projected to surpass $30 billion by 2035. The recent $43 billion acquisition of ADC pioneer Seagen by Pfizer underscores the immense value placed on this technology.

However, the very power of ADCs is also their greatest challenge. If the cytotoxic payload is released prematurely or the antibody binds to healthy cells, it can cause severe off-target toxicity, most commonly in the liver and gastrointestinal tract. Differentiating this unwanted cytotoxicity from the desired anti-cancer effect is a critical and difficult step in preclinical development.

This is the specific problem VivoSim’s new data aims to solve. By testing existing, marketed ADCs in their NAMkind™ liver and intestine models, the company reports it has been able to reproduce the known clinical side effects, such as liver toxicity and drug-induced diarrhea. According to the company, these models can reveal mechanistic details like premature linker cleavage and “bystander effects,” where the payload kills healthy cells near the tumor—insights that are difficult to obtain from conventional tests.

“These ADC toxicity results show a close correlation to clinical safety outcomes,” stated Amar Sethi, Chief Scientific Officer at VivoSim. “We think that our partners working with our testing models will be able to screen out toxicities during lead candidate optimization, resulting in greater success in the clinic at eliminating cancers using drugs with limited side effect profiles.”

While the full dataset will be unveiled at the SOT meeting in March, this validation builds on VivoSim's previous successes. In 2025, its liver model was recognized for high sensitivity and specificity in identifying liver-toxic drugs, and its intestine model received an award for its ability to predict gastrointestinal toxicity.

From Lab to Global Market

VivoSim is not just pursuing a scientific breakthrough; it is executing a clear commercial strategy to capitalize on the converging trends of regulatory change and market demand. The company is positioning its NAMkind™ toxicology services as a critical tool for pharmaceutical partners looking to de-risk their drug pipelines and avoid costly late-stage failures.

The competitive landscape for NAMs is expanding, but VivoSim is leveraging its deep expertise in 3D human tissue biology, with leadership that includes the founder of 3D bioprinting pioneer Organovo. By integrating its advanced cell models with artificial intelligence and machine learning, the company aims to enhance predictive accuracy and provide clients with clear, decision-ready insights early in the development process.

Reflecting the global nature of drug development, VivoSim has already made its services available not only in the United States and Europe but also through distributors across Korea and China. This international footprint positions the company to support a worldwide client base as it continues to scale its capacity to meet what it describes as “expanding global demand and urgent, real-world development needs.” The upcoming presentation in San Diego will be a crucial test, as the broader scientific community gets its first look at the data that could help shape the future of how safer, more effective medicines are made.