Mantis Biotech Wants to Replace Your $2.6 Billion Clinical Trial

Mantis Biotech Wants to Replace Your $2.6 Billion Clinical Trial With a Simulation

The average approved drug costs $2.6 billion to bring to market. Most of that money burns in clinical trials where patient recruitment alone can stall timelines by 6 to 18 months. Mantis Biotech just told the industry it plans to build synthetic humans that could make a significant chunk of that spend obsolete.

The Signal

Mantis Biotech is building digital twins of human physiology by stitching together genomics, medical imaging, electronic health records, and physiological data into synthetic patient populations. The idea is straightforward. If you can simulate how a thousand patients respond to a compound before enrolling a single real person, you compress Phase II timelines, cut failure rates, and redirect billions in R&D capital away from physical trial infrastructure. Large language models trained on medical datasets sit at the core, accelerating everything from drug candidate screening to diagnostic decision making.

This is not a research curiosity. It is a capital allocation event. Every dollar pharma currently spends on recruiting patients, managing trial sites, and running data monitoring boards becomes a line item up for renegotiation. Contract research organizations, hospital systems with research arms, and diagnostic equipment manufacturers all sit downstream of this shift. The question is not whether digital twins will reshape clinical development. The question is who captures the economics when they do.

That trajectory is the context for every decision below. BLS data shows the Medical Care CPI climbed from 559.27 in March 2024 to 592.55 by February 2026. That is a 6% increase in under two years, accelerating upward. Every month that clinical development stays expensive, the cost pressure compounds. Digital twins are not arriving into a stable cost environment. They are arriving into one that is actively punishing the old model.

Capital Reallocation Is Already the Board Conversation

The math is not subtle. If digital twin platforms can reduce Phase II failure rates by even 10 to 15%, the downstream savings cascade through the entire development pipeline. Phase III trials, which account for roughly 60% of total clinical development costs, get smaller or get restructured around hybrid designs that blend simulated and real patient data. That means CFOs at midsized pharma companies and contract manufacturers need to start modeling scenarios where R&D budgets shift from physical trial site contracts toward computational infrastructure and synthetic data licensing.

The decision is not whether to invest. It is how fast to reallocate. Companies that wait for regulatory clarity before moving capital will find themselves licensing synthetic data from the same tech platforms that locked up the training datasets two years earlier. The framework here is simple. Identify which Phase II programs in your current pipeline have the longest recruitment timelines. Those are your candidates for hybrid trial design. Run the cost comparison now, not after Mantis or a competitor signs an exclusive data partnership with a major health system. Medical care costs are not coming down. According to Federal Reserve economic data, the trend line has barely paused since mid 2024. Every month of delay in adopting cost compressing technology is a month of compounding exposure to a cost structure that only moves in one direction.

Hospital Systems Are Sitting on a Data Gold Mine They Have Not Priced

Here is the part most health system executives have not internalized. The deidentified EHR data sitting in your servers is the raw material Mantis Biotech and every competitor in this space needs to build accurate digital twins. That data has a market value that did not exist three years ago. Hospital systems with large, diverse patient populations and robust electronic health records are suddenly holding a strategic asset that synthetic data companies will pay to access.

The decision for a Chief Medical Officer or VP of Research Partnerships is whether to monetize that asset proactively or let it get commoditized. The framework starts with an inventory. How clean is your EHR data? How diverse is your patient population across age, ethnicity, comorbidity profiles? Those variables determine your negotiating leverage. Health systems that can offer longitudinal data across multiple disease states will command premium licensing agreements. Systems with fragmented or poorly coded records will get table scraps. The window matters because once a platform like Mantis assembles a sufficiently representative synthetic population, the marginal value of additional real world data drops. First movers in data licensing will capture outsized economics. Laggards will find the market has already been built without them.

Contract Research Organizations Face an Existential Pivot

CROs have built their entire business model around the operational complexity of running physical trials. Site management. Patient recruitment. Data monitoring. Regulatory submissions. Digital twins threaten the core revenue engine by reducing the number of patients needed in early stage trials and by enabling sponsors to run simulated control arms instead of placebo groups.

The decision is existential but not binary. CROs that develop inhouse synthetic data capabilities or form strategic partnerships with digital twin platforms will evolve into hybrid operators. Those that cling to the traditional full service trial model will watch their margins erode as sponsors demand smaller, faster, cheaper trials. The framework for CRO leadership is to audit current revenue by trial phase. If more than 40% of your revenue comes from Phase I and Phase II site management, you have significant exposure. Start building computational trial design capabilities now. Hire biostatisticians who understand synthetic data validation. Form partnerships with companies like Mantis before your competitors lock up exclusive arrangements. The Medical Care CPI hitting 592.55 in February 2026 tells you that your pharma clients are under relentless cost pressure. They will cut your scope before they cut their pipeline. Give them a reason to keep you in the room by offering the capability they need next, not the one they needed five years ago.

Regulatory Risk Is Real but Navigable

The FDA has not issued comprehensive guidance on synthetic control arms or digital twin derived endpoints. That regulatory ambiguity is the single biggest brake on adoption. But it is a timing risk, not a structural one. The agency has already accepted real world evidence in supplemental approvals and has signaled openness to adaptive trial designs that incorporate computational modeling.

The decision for regulatory affairs leaders is how aggressively to engage the FDA now versus waiting for formal guidance. The framework favors engagement. Companies that submit pre IND meeting packages incorporating digital twin methodology will help shape the regulatory framework rather than react to it. That is a competitive advantage measured in years. File your questions early. Propose pilot programs. Reference the growing body of published validation studies showing synthetic patient populations can replicate real world trial outcomes within acceptable confidence intervals. The worst outcome is not that the FDA says no. The worst outcome is that a competitor gets to yes first because they started the conversation 18 months before you did. Every quarter of regulatory delay costs the industry real money. With medical care costs climbing 6% in under two years according to BLS figures, the economic case for accelerating regulatory engagement is not theoretical. It is arithmetic.

The operators who win the next decade of pharma development will not be the ones with the biggest trial budgets. They will be the ones who figured out how to make the trial budget irrelevant. If your capital allocation model still assumes physical trials are the only path to approval, the market is about to educate you on the cost of that assumption.

This article is part of the Industry Intelligence series on NeuralPress. New analysis published daily.