Translating Preclinical Research into Clinical Success

Bringing a new drug to patients requires more than just identifying a promising candidate. Scientists must rigorously test the compound’s safety, biological activity, and pharmacokinetic properties using in vitro and in vivo experiments before the therapeutic advances to human studies. Ultimately, the goal of preclinical research is to reject drug candidates with concerning adverse effects and generate data to guide clinical trial design.¹

During clinical development, scientists evaluate drug candidates through a series of increasingly complex clinical trials that unfold across three phases: Phase 1, Phase 2, and Phase 3.² In Phase 1 trials, investigators enroll a small number of healthy volunteers to examine how the drug candidate interacts with the human body and identify early safety concerns. Phase 2 clinical trials build on these findings by assessing the drug candidate’s safety in patients with the disease or condition, elucidating the optimal dose, and gathering preliminary evidence of its efficacy.² Finally, Phase 3 trials enroll larger patient populations to compare the new drug with existing treatments or placebos and to detect less common adverse effects. These results enable regulatory agencies to determine whether to approve the drug for market use.

Improving Preclinical Predictability

Despite rigorous testing, more than 90 percent of drug candidates entering clinical trials fail to achieve final approval, primarily due to inadequate efficacy or unexpected toxicity.¹ While many factors contribute to these failures, a major challenge lies in the reliance on animal models during the preclinical stage. Animal models often cannot fully replicate human tissues and the intricate biological processes present in health or disease. Additionally, inbred animals lack the genetic diversity necessary to predict responses in heterogeneous or complex diseases, such as those arising from differences in ethnicity, sex, or other patient characteristics.³

For decades, regulatory agencies, including the United States Food and Drug Administration (FDA), required researchers to demonstrate a drug candidate’s safety and efficacy in at least two different animal models before beginning clinical trials.² In response to the limitations of these models, the US Congress passed the FDA Modernization Act 2.0 in 2022, and the FDA released a roadmap in 2025 outlining plans to phase out animal testing for new drug evaluations.³ These changes enable scientists to pursue alternative models, known as New Approach Methodologies (NAMs), which include 2D and 3D cell culture models, such as human induced pluripotent stem cells, organoids, and organ-on-a-chip devices, and computational models. By better reflecting human biology, NAMs have the potential to improve preclinical predictability and help investigators account for patient variability in study design.¹

Achieving Therapeutic Exposure Levels

Even when a drug candidate shows promise in preclinical research, translation to humans can fail if the compound cannot reach or maintain effective levels in the body.⁴ Therapeutics can display strong activity against a molecular target in controlled cell-based assays but fail to penetrate tissues efficiently, clear too quickly, or accumulate unpredictably at higher doses. Furthermore, animal models frequently misrepresent these behaviors because drug absorption, distribution, metabolism, and clearance differ across species.³

NAMs provide human-relevant insight into how a compound moves through and persists in the body, helping researchers identify exposure limitations earlier. By pairing NAMs with formulation strategies, such as altering delivery methods or modifying chemical properties, scientists can improve tissue penetration and stability in systems that better reflect human biology, enabling them to design dosing strategies that support consistent target exposure.⁵

From Target to Clinical Effect

Beyond exposure, a drug may engage with its intended target without producing meaningful changes in disease biology. Redundant pathways or compensatory mechanisms can prevent this interaction from translating into clinical benefit.⁶ Carefully chosen biomarkers help reveal whether a drug engages its target and produces the expected biological effect, as well as guide dose selection and patient stratification.⁷ NAMs and these translational biomarkers together enable scientists to identify potential failures earlier and refine study designs to improve the likelihood of clinical success, ultimately increasing the chances that effective therapies reach patients.