Most new drugs never make it to a pharmacy shelf. According to industry-wide data tracked by pharmaceutical trade groups, roughly 92 percent of drugs that enter clinical trials fail before reaching approval, a number that has actually gotten worse over the last decade, dropping from a 9 percent success rate to about 8 percent. VivoSim, a San Diego biotech company based in the Sorrento Valley research corridor, is building lab-grown human tissue models designed to catch failures earlier, before a drug ever reaches a human trial.
What VivoSim Actually Builds
VivoSim’s core technology is a form of engineered tissue sometimes described as “3D biology.” Instead of testing drugs on flat layers of cells in a petri dish or relying entirely on animal studies, the company grows small, three-dimensional tissue structures that behave more like real human organs.
Two examples came up during a recent conversation with company leadership: a liver model and an intestine model.
The liver model is built from tiny tissue structures, sometimes described as being about the width of a dime, that combine the four major cell types found in a real human liver. That detail matters because those cell types make up roughly 98 percent of what the liver actually does. Some of the less common cell types included, such as Kupffer cells, appear to play a protective role, reducing the liver’s own toxic response to certain compounds. Leaving them out, in other words, would give a less accurate picture of how a drug behaves.
The intestine model works differently. Researchers grow cells that form a thin barrier, much like the lining of a human intestine, on a small membrane. That barrier needs a very particular level of permeability, absorbing nutrients while still blocking other substances. When a drug is toxic to the intestine, that barrier tends to break down first, which is often an early physical sign of gastrointestinal side effects that show up later in human trials.
Both tissue types are grown in a format roughly the size of a postcard, packed with 96 individual samples at once, allowing many drug concentrations and conditions to be tested in parallel.
How This Affects Drug Costs
The commonly cited figure that it costs about $2 billion to bring a single drug to market is not the cost of developing one drug. It reflects the average cost of every failed attempt along the way. If a company spends around $150 to $180 million developing each drug candidate and only 1 in 12 succeeds, the failures get baked into that final number.
VivoSim’s pitch is that improving the success rate, even modestly, changes the cost significantly. Moving from an 8 percent success rate to something closer to 30 percent would mean fewer failed candidates absorbing costs before a single approved drug reaches the finish line, according to estimates the company shared. That kind of shift could, in theory, cut the effective cost of bringing a drug to market by more than half.
There’s also a narrower, more immediate use case. When a drug shows signs of liver toxicity in a standard six-month animal study, a pharmaceutical company sometimes has reason to believe the effect is specific to that animal species and would not occur in humans. Testing that hypothesis, known as “investigative toxicology”, using rodent-based tissue models built the same way as the human ones, according to VivoSim, can cost many times less than using animal studies to reach the same conclusion.
Image: https://vivosim.ai/ – Liver Toxicity Test: NAMKIND™ Tissue Model
A Drug Candidate Already in Human Trials
One outcome tied to this approach, developed before VivoSim, involved using its intestinal tissue model with diseased cells to identify a drug candidate for Crohn’s disease. That drug was later sold to Eli Lilly and, according to the company, is now in Phase 2 clinical trials.
Image: https://vivosim.ai/ – Medical Testing Lab
The Opportunity for Pharma and Biotech
Today, pharma executives often focus less on their early stage Research and Development efforts and more on their commercial drugs and late stage clinical trial drugs that will have the biggest impact the soonest. But this neglect of early pipelines is what VivoSim is trying to change. By messaging directly to pharma executives about the ability to screen out liver tox failures before preclinical and clinical studies, VivoSim expects C-level pharma executives to see the massive strategic advantage they gain. They can spend less time talking to their investors about surprising bad results and more time talking about how fast the patient numbers are growing. VivoSim also talks directly to Venture Capitalists about the strategic opportunity, suggesting that it makes the most sense to derisk a drug using insights from 3D biology before making a major investment to put it into clinical trials.
The Bigger Picture
VivoSim’s approach fits into a broader shift in early-stage drug development, sometimes referred to as new approach methodologies, or NAMs, which the FDA has outlined in its own roadmap in recent years. The agency has signaled increasing interest in reducing reliance on animal data where alternative models can demonstrate comparable or better predictive value.
Tissue-based models like VivoSim’s are specified as a standard part of that roadmap, and can get broader use by demonstrating consistent results across a wide range of drug types and disease areas. As a model is validated, its use grows.
At its core, VivoSim is a biotech company that is among the pioneers developing lab-grown, three-dimensional human tissue models, including liver and intestine tissue, used to help predict how experimental drugs will behave in the human body before clinical trials begin. Rather than relying solely on animal biology, which can respond differently to drugs than human tissue does, VivoSim grows tissue built from human cells arranged to mimic the structure and function of real organs, aiming to produce results that translate more directly to human outcomes. For now, the company’s public track record includes one drug candidate, the Crohn’s disease treatment sold to Eli Lilly, that has advanced into Phase 2 human trials after being shaped, in part, by this kind of tissue modeling technology, along with a growing body of internal data the company says supports its case for wider industry use.
