Mirjam Trame is the VP and head of pharmacometrics at Certara, specializing in model-informed drug development and regulatory science.
Most industries are transformed by new technologies. Drug development, however, is being reshaped by something much quieter: a change in what regulators consider credible scientific evidence. The biggest development this year was not a breakthrough therapy or a new artificial intelligence (AI) platform. It was a series of regulatory decisions that, taken together, signal a fundamental shift in how medicines will be developed and evaluated over the coming decade.
Although each guidance addresses a different aspect of drug development, they point in the same direction. Regulators are placing greater emphasis on integrated, predictive and scientifically credible evidence that helps sponsors make better decisions before another study begins. I believe that shift will influence far more than regulatory submissions. It will affect how organizations design development programs, allocate research budgets and decide which therapies move forward.
One of the clearest examples is ICH M15, the first globally harmonized guideline establishing general principles for model-informed drug development (MIDD). Until recently, computational modeling was often viewed as a specialized scientific capability that supported development programs. M15 elevates those approaches by providing an internationally aligned framework for how model-informed evidence should be planned, generated, evaluated and applied during regulatory decision-making. Rather than treating modeling as simply a supporting activity, the guidance recognizes it as an important component of the evidence used to inform regulatory decisions.
At nearly the same time, regulators advanced ICH E20, creating a common framework for adaptive clinical trials. Unlike traditional studies that follow a fixed protocol from beginning to end, adaptive trials allow predefined modifications as evidence accumulates. When designed appropriately, these approaches can answer important scientific questions more efficiently while maintaining the level of rigor regulators expect.
The third development received far less attention outside regulatory circles but may ultimately prove just as significant. In 2025, the FDA announced a roadmap to reduce reliance on animal testing for monoclonal antibodies and other drugs by encouraging scientifically validated alternatives, including computational modeling and other human-relevant methods.
Although these initiatives address different aspects of development, together they point to a broader shift in regulatory thinking: generating stronger evidence earlier to support better decisions throughout development. In April 2026, the FDA announced that it had achieved its key first-year implementation goals, reinforcing its commitment to modernizing drug development.
The common thread is not that regulators are replacing clinical research or lowering evidentiary standards. Instead, they are encouraging sponsors to generate stronger scientific evidence earlier in development so they can make better-informed decisions before launching another study. Computational models, adaptive trial designs and alternative methods are helping answer questions that previously required additional experiments, allowing organizations to reduce uncertainty before committing more patients, more time and more resources.
This represents a meaningful change in how evidence itself is viewed. For decades, drug development largely followed a sequential process in which one study was completed before the next decision was made. Increasingly, regulators recognize that carefully developed models and multiple complementary sources of evidence can reduce uncertainty and strengthen drug development decisions earlier in development. The goal is not to replace clinical trials, but to make every trial more informative and every development decision more deliberate.
This shift reflects a broader reality across the industry. Clinical drug development is becoming increasingly complex, as the therapeutic landscape expands to include sophisticated modalities that present challenges traditional development approaches alone cannot always address efficiently. Antibody-drug conjugates, gene therapies, cell therapies and personalized medicines are creating new questions around dose, exposure, efficacy, safety and patient variability.
For example, physiologically based pharmacokinetic modeling has been used to evaluate drug-drug interaction risk for the antibody-drug conjugate trastuzumab deruxtecan, demonstrating how advanced quantitative approaches can help address development questions associated with complex therapies.
The implications extend well beyond regulatory affairs. Development leaders should resist the temptation to view these new regulations and guidelines as requirements for regulatory teams alone. They are better understood as signals that expectations for scientific decision-making are changing across the industry. Organizations that recognize this shift early will be in a much stronger position to design development programs that are both scientifically rigorous and strategically efficient.
What does this mean for organizations developing new medicines? Three practical changes stand out to me as a leader in this space.
First, organizations should bring modeling, statistics, clinical pharmacology, translational science and regulatory strategy together at the beginning of a development program rather than asking each function to contribute independently later—the strongest evidence packages are designed collaboratively from the outset instead of being assembled shortly before submission.
Second, companies should match the rigor of an analysis to the importance of the decision being made. Not every development question requires the same level of evidence, but every major investment and development milestone should be supported by a transparent and scientifically defensible rationale.
Finally, organizations should think in terms of evidence strategies rather than individual studies. Increasingly, regulators are evaluating how clinical data, computational models, biomarkers, real-world evidence and mechanistic understanding work together to support a development decision. The strength of the evidence package comes from how these pieces reinforce one another rather than from any single analysis.
Leaders must stop asking, “What study generated this evidence?” Instead, they should start asking, “Does the totality of the evidence provide a scientifically credible foundation for this decision?” Organizations that embrace that shift today may be better positioned to develop medicines more efficiently, make better scientific decisions and ultimately bring more effective therapies to patients.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?