Industry experts navigate clinical trial safety at NCLifeSci forum
Drug development leaders gathered Oct. 30 at an NCLifeSci forum to discuss the growing complexities of ensuring patient safety in clinical trials, from managing data safety monitoring boards to navigating FDA regulations and exploring the potential of artificial intelligence.
The panel discussion, moderated by Gerry Klein, M.D., principal of MedSurgPI, highlighted how modern drug development requires balancing innovation with rigorous safety oversight, particularly as gene and cell therapies introduce new challenges. The panel comprised
- June Almenoff, M.D., Ph.D., consultant, BioBridges;
- Christopher Crean, chief executive officer, Xyzagen;
- Jamie Lazar, principal safety committee specialist, PPD clinical research business of Thermo Fisher Scientific; and
- Rich Shea, president and chief operating officer, Basking Biosciences.
The event was sponsored by Hodess Cleanrooms and Uniphar Clinical.
AI shows promise but faces limitations
The role of AI in clinical trial safety emerged as a central theme, with panelists offering cautious optimism tempered by practical concerns.
Rich Shea of Basking Biosciences said his company will begin working with a Boston-based firm using machine learning algorithms to analyze safety data and trial design in early 2025.
"We have very experienced development team, my team and I averaged 30 years in the industry, and we also have some of the leading stroke clinical trialists in the world advising the company," Shea said. "Now, in three months, we may be amending the protocol, because they may come back and say, Well, look, we found these things."
Crean of Xyzagen said he was skeptical about AI replacing human expertise.
"AI is as only good as the large language model that's put into it," he said. "I think it's going to be useful in writing NDAs and essentially summarizing documents and making our life easier and generating reports, but actually trusting in AI? I wouldn't trust the computers."
June Almenoff, an accomplished biopharma executive with over 25 years of experience, shared her experience leading initiatives at GlaxoSmithKline to build signal-detection systems using Bayesian statistics.
"What we found were a lot of false positives," Almenoff said. "My hope is that for AI, it brings those statistical methods to a different level, where you're not having so many false positives. It actually sorts through the noise for you."
Safety as foundation of drug development
Almenoff emphasized the fundamental importance of safety throughout drug development, noting that monitoring and managing safety is "how we keep the promise of our intent to do no harm."
She described safety as more than just tracking adverse events.
"If you are astute at looking at safety, you learn a lot about whether a drug is likely to be effective," Almenoff said. "It informs you whether the drug is hitting the biological target."
The discussion highlighted how safety profiles affect a product's approvability, label and market competitiveness. Almenoff noted that a gene therapy company's stock fell 42% after pausing studies due to a single hospitalization, demonstrating the outsized impact safety concerns have on drug development.
Data safety monitoring boards provide critical oversight
PPD’s Jamie Lazar explained the role of data safety monitoring boards in protecting patient safety during clinical trials. These independent committees typically comprise three to five experts in the relevant indication plus a biostatistician.
"The purpose of those meetings are to provide the DSMB with typically unblinded outputs," Lazar said. "They meet and formulate their recommendations. The recommendations are then passed on to the pharmaceutical company, usually saying no safety concerns. Study continues."
Lazar, who manages 28 different committees and leads training for PPD's safety committee group, emphasized the importance of independence. All communications must be filtered through the coordinator to maintain separation between the pharmaceutical company and board members.
"The biggest thing to point out about those committees is that they are independent," Lazar said. "That goal is to preserve the integrity of the study, the study data, as well as allowing them to make unbiased recommendations of their review."
Global development presents regulatory challenges
The panel explored how companies navigate international clinical trials while maintaining safety standards.
Shea said Basking operates in Australia, the U.S., Canada and Europe for its reversible thrombolytic therapy development. The company chose not to file an investigational new drug application with FDA for its reversal agent, instead going to Australia first.
"FDA right now, there's a lot of uncertainty. There's a lot of turnover," Shea said. "It's a risk management decision. We elected we're not going to put the program at risk, because what we're doing is fundamentally sound while they work through whatever transition they're going through."
While international trials follow International Council for Harmonisation guidelines ensuring consistent safety standards, Crean noted that different countries have different thresholds for preclinical toxicology studies required before starting clinical trials. Australia and Canada allow companies to begin with more limited toxicology packages compared with the U.S. and Europe.
"The requirements for the tox studies are essentially the same," Crean said. "Certain countries might have a different threshold for the basket of toxicology studies that may need to be conducted."
Preclinical safety requirements under scrutiny
The panel discussed potential changes to preclinical safety requirements, particularly for biologics. Crean mentioned recent FDA signals about potentially reducing nonclinical testing requirements but remained cautious.
"I still don't see the FDA getting away from eliminating some sort of general tox," Crean said. He suggested that reproductive toxicology studies might be eliminated for biologics with very specific mechanisms of action that would not impact fertility.
The discussion touched on in vitro alternatives to animal testing, but panelists expressed doubt about replacing animal studies entirely.
"The biggest problem in particular — why pharmacokinetics is done in vivo — is that it's the most complex system," Crean said. "These approaches of trying to integrate AI and tox on a chip, they are a simplistic model of the most complex system."
Real-world evidence gains importance
The panel talked about how real-world evidence and postmarket surveillance complement clinical trial data. Crean noted that drug monitoring never stops after approval, with labels evolving over time based on new information.
"The clinical trial was approved based upon risk benefit within that very finite period of time, and the benefit outweighed the risk," Crean said. "The drug cycle and the monitoring never stops for that drug."
However, the panel acknowledged challenges with postmarket safety reporting. Klein noted that only 1% of side effects after drugs reach market get reported, and that associations found in large databases through meta-analysis don't necessarily indicate cause and effect.
Investment climate remains challenging
The forum concluded with discussion of the difficult investment environment for early-stage drug development companies.
Laura Gunter of NCLifeSci said money remains on the sidelines due to uncertainty, though Small Business Innovation Research grants offer one pathway for early-stage funding to generate data that attracts investors.
Shea said his company recently closed a second tranche of financing led by ARCH Venture Partners.
"Deals are getting done, but they're kind of either later clinical or they're shoring up portfolio companies that they believe in," he said. "There is a ton of money out there, but it's being deployed very cautiously."
The panelists agreed that uncertainty remains the key challenge, affecting both regulatory processes and investment decisions across the drug development landscape.