Illustration by Rob Hadley for Cure
Overview
Most clinical trials fail, and the reasons extend far beyond the science.
More than 90 percent of clinical trials fail. According to Orr Inbarr, CEO of QuantHealth, an AI-powered clinical trial simulation platform, about 75 percent of that failure comes down to efficacy or safety not translating as expected into human trials. "The way that we think the drug works, and the way that we position it against the human body and human disease is somewhat incorrect, and then we either miss out on efficacy or on safety," he said.
That translation gap is the starting point but not the whole story. The reasons promising biomedical discoveries stall before reaching patients extend well beyond the science, into funding, regulation, market dynamics, and an increasingly unstable policy environment.
The Funding Squeeze
Katarina Klett, PhD, Director of Research Translation at QB3, a biotech entrepreneurship hub at UC Berkeley, says a core problem is that even effective therapies need a business case to survive. "There's a lack of funding for early stage and scaling a therapy," she said. "The market that you're addressing has to be a big enough business to matter to pharma. It has to be at least a billion dollar-a-year drug. Not everything is. Even if your drug can cure the disease, if it doesn't have a big enough market, it won't get investor money, and there's no other source of money to make it happen."
That dynamic hits hardest in areas like rare diseases and neurodegenerative conditions, where the patient populations are small and the clinical track record is littered with failures. These are the programs most likely to be overlooked, regardless of their scientific promise.
The broader investment climate has made things worse. Inflation has pushed investors toward safer bets, and biotech is not where risk-averse capital wants to be. "There actually is biotech money getting invested, but it's shifted toward later stage companies that have Phase 2 trial data," Klett said. Inbarr sees the same pattern: fewer rounds, but much larger ones, concentrated in programs with enough data to reduce the risk.
Meanwhile, federal funding is contracting. John Yu, MD, an NIH-funded investigator at Cedars-Sinai Medical Center and CEO of Kairos Pharma, says proposed cuts to indirect funding threaten the foundational discovery work that feeds the entire pipeline. "The Department of Defense has already discontinued funding for lung and prostate cancer research, a devastating blow to institutions and companies, like ours, that have long relied on these grants to explore promising new therapies," he said.
The shifts extend beyond budgets. Support for vaccine development, women's health, and studies affected by DEI-related language restrictions has declined. Some leading academic institutions have had research portfolios frozen or placed under review. Restrictions on graduate and postdoctoral visas are shrinking the scientific workforce.
"A lot of these sort of government efforts to support biomedical research have really diminished and are putting more stress on small biotech companies to try to make it on their own," Yu said.
What's Changing in How Therapies Get Tested
Some parts of the system are starting to adapt to these new realities, particularly in how therapies get tested before they reach human trials.
One shift is in preclinical testing. Yu says there is growing momentum toward human organoid models as an alternative to animal testing. Organoids are micro-organs containing all of the cell types of a particular organ, organized in a similar fashion to the human organ. "You can test disease models in those organoids and test the safety," he said. "For instance, you can test liver organoids and look for liver toxicities of certain drugs through that kind of model."
Yu sees this as a way to reduce the cost of preclinical work while improving how well results predict what will happen in humans, which is where the majority of clinical trials currently fall apart.
On the trial design side, Inbarr's company is using AI to simulate clinical trials before they run. QuantHealth has simulated over 350 clinical trials with 85 percent accuracy. "We can basically walk into any clinical trial design team or process and immediately help them validate and test out different clinical hypotheses and make the best decision for their trial," he said.
Competing Globally
Biotech companies are also contending with the fact that clinical trials are increasingly moving to other countries. China in particular has become a hub for early human trials, offering access to large patient populations at lower cost.
"There are a lot of drugs coming out of China that have gone through the first human trials, and that have generated enough excitement to be taken up by large pharma companies in the U.S.," Yu said. "And that's sort of been one of the models of low cost and efficient generation of human data."
To compete, Yu is helping build something new: a consortium of academic medical centers, co-led by Cedars-Sinai and the Mayo Clinic, that includes Stanford, Cleveland Clinic, Rutgers, and Vanderbilt. The model centralizes IRB approval, pools patient volume across institutions, and runs investigator-initiated trials at lower cost.
"It will take this novel type of a collaboration between academic centers and biotech companies to come up with these types of strategies to increase the scale of clinical trials and decrease the cost so that we're competitive with out-of-country networks," he said.
Klett adds that where trials run is only part of the problem. Clinical trials measure success against specific biological markers or patient outcomes that determine whether a drug actually works. Choose the wrong ones and a promising therapy can look like a failure. "It's hard to prove a drug works if you're measuring the wrong thing and it's not predictive. Picking the right endpoints is crucial."
As U.S. funding contracts, other countries are building the clinical infrastructure to absorb the work that American labs can no longer afford to do. The stakes extend beyond any single therapy or company. "How can a country that empowered small biotech companies to pioneer CAR T cell therapies, checkpoint inhibitors, gene therapies, and the COVID-19 vaccine remain at the forefront of biomedical innovation if we continue down this path?" Yu said. "American science has never thrived on certainty; it thrives on resilience, grit, and vision. We must preserve the ecosystem that enables innovation before we lose the momentum that once made us a global leader in medical breakthroughs."
