Our investment in Vivodyne and its vision to make fewer drugs fail
Recent years have seen breakthrough therapeutic success ushered in by new and emergent drug modalities. Since 2017 eight fundamentally new drug modalities have embarked on first clinical trials or received approvals. Just last week, the first CRISPR drug was approved in the UK. We’ve seen an explosion in cell therapy blockbusters with hundreds more in the pipeline. PORTACs have stormed into fashion and ADCs are seeing deep clinical success. The drug hunter’s toolkit is as full as it has ever been.
Despite these leaps forward, drug discovery is becoming less efficient, not more. A recent Deloitte report showed dwindling ROI on pharma R&D dollars, longer timelines to approval, and a 15% YoY increase in the cost to bring a new drug to market. If we expect to make significant progress against areas of extreme patient need, this paradigm must shift. Much of this declining productivity is driven by late-stage failure: the inability of the industry to accurately predict which drug candidates will find success out of the lab and in patients.
90% of drugs that enter clinical trials ultimately fail. For a drug to reach human testing, the smartest scientists at each company have assessed the candidate and its associated data package and decided it is worth the expense and risk of testing in humans, and the FDA has given their stamp of approval. Up to this point, every computational model, cell-based assay, and animal model has suggested the candidate will have clinical efficacy with enough conviction to justify the exorbitant cost of a clinical trial.
If we hope to have a shot at curing many of the world’s worst diseases, this 90% clinical failure rate must improve. We believe this is the most critical problem in drug discovery.
We are very excited to announce our investment in Vivodyne which is running headfirst at this problem. Vivodyne has built a high-throughput platform to test, discover, and develop new drugs on lifelike, functional, lab-grown human organ tissues. Their platform brings translatable human data to the earliest stages of the drug discovery process. Vivodyne’s fully automated robotics and software-driven platform allows preclinical testing on complex, organ-scale human tissues at the earliest stages of a drug discovery program. Their throughput allows them to generate AI-scale data, enabling computational refinement of the therapeutics alongside direct, clinically equivalent data from their experiments. Vivodyne partners with drug developers to peer into the future and understand how their new druggable targets and assets might succeed (or fail) in humans years earlier and at a fraction of the cost of a clinical trial.
It’s no secret that 2D cell models are woefully inadequate at recapitulating the immense complexity of human biology. Prior to Vivodyne, if you sought to understand the translatability of a new drug candidate to higher order systems, you were largely left to rely on animal models of disease (mice, ferrets, chimps, etc.). Moral implications aside, animal models are prohibitively expensive (a single monkey can cost well over $100k), are in extremely short supply due to rising geopolitical tension and facilities to house the animals, known as vivariums, consume 10x the energy of a similarly sized office building. As a result, drug discovery campaigns rely on basic cellular models until the last stages of development where a select few compounds will be tested in animals. It’s no wonder few drugs find success in humans.
Vivodyne is founded by Dr. Andrei Georgescu and Dr. Dan Huh based on work the two completed in Dan’s lab at the University of Pennsylvania. Dan kickstarted the field of research into organ-on-a-chip models with his landmark Science paper in 2010. His lab has been systematically expanding into new organ systems and increasing the physiological relevance of such models since. Yet, producing these systems remained a highly manual and artisanal task — requiring highly trained teams to run each experiment and thus limiting their adoption beyond the lab.
Setting out to overcome these limits, Andrei and Dan built a fully automated platform that drastically scales the throughput of these models while ensuring tight variance and repeatability between replicates. Even more importantly, Vivodyne’s automation allows them to rapidly improve their tissue models with AI-driven refinement using their high-throughput infrastructure. As a result, Vivodyne conducts experiments with up to 10,000 unique, complex organ tissues, while the next closest system supports no more than 12 samples at a time (and even less human realism). Couple this with an AI stack developed by Vivodyne to take advantage of their ability to generate human data on-demand — fed by in-line microscopy, multi-omics, and a litany of other assays used on human biopsies — and you have the recipe to fundamentally alter the success of drug development. It’s rare that we come across early-stage teams with technology as robust and functional as Vivodyne’s.
The potential of Vivodyne’s platform has not gone unnoticed by pharma — Vivodyne currently partners with five of the top ten pharma to improve their prioritization of pre-clinical assets, and is developing a strong, wholly owned internal pipeline of drugs as well. Today, pharma sales in the U.S. alone top $500B. If Vivodyne could decrease the probability of clinical failure from 90% to 80%, that would yield a doubling in new drugs entering the market. This bold vision could create an enormous amount of economic value in addition to a dramatic impact on saving and/or improving the quality of life for countless people around the world.
We find the most exciting investments are at the intersection of a number of emergent technologies and Vivodyne is a perfect example of this bringing together deep expertise in robotics, automation, cellular biology, physiology, and machine learning. From our first conversations with Andrei at the beginning of 2023, it’s been clear that he and the team at Vivodyne are building something truly transformative. Andrei’s drive and vision for what Vivodyne will become is compelling and has allowed him to recruit a deep bench of skilled scientists, engineers, and drug hunters to work alongside him. We are humbled to be a small part in this story and are excited to roll up our sleeves alongside the Vivodyne team and fellow investors Khosla Ventures, MBX, CS Ventures, and Kairos.
For further information, you can read Vivodyne’s funding announcement here.