Pharmacokinetic Modeling Emerges as Key to Unlocking Nanoparticle Medicines

Nanoparticles have long promised to transform medicine, offering a way to ferry drugs through the bloodstream, shield fragile payloads from degradation, and deliver their cargo precisely where it is needed. Yet for all the enthusiasm in laboratories around the world, the journey from laboratory bench to patient bedside remains extraordinarily difficult. Fewer than one hundred nanoparticle formulations have ever made it through the gauntlet of preclinical testing and clinical trials to reach regulatory approval, a strikingly small number given the thousands of promising studies published each year. A new comprehensive review published in Bioengineering & Translational Medicine argues that a powerful but underused tool—pharmacokinetic modeling—could be the missing link that finally closes this stubborn translational gap.

The review, led by researchers at the University of Utah, examines why so many nanoparticle drug delivery systems stall before reaching patients. Only about twenty percent of nanoparticle-based delivery systems assessed in preclinical studies are ever considered for translational investigation, and the actual clinical translation rate is lower still, undermined by problems with stability, scalability, efficacy, and toxicity. Between 2001 and 2024, just 57 nanoparticle-based drugs advanced from Phase I and II trials into Phase III, and only 17 progressed through the first three phases into Phase IV. Most failures occur in Phase II, where lack of efficacy or unforeseen toxicity ends development. The authors contend that these repeated disappointments stem largely from the absence of integrative, predictive frameworks that systematically connect nanoparticle design, biological interactions, and clinical drug exposure.

The stakes of closing this gap were made vividly clear during the COVID-19 pandemic. Lipid nanoparticles carrying messenger RNA proved to be the linchpin of the Pfizer/BioNTech and Moderna vaccines, whose ionizable lipids protonate at acidic pH to complex and protect mRNA and then release it inside cells at endosomal pH. It has been estimated that the first year of COVID-19 vaccination alone prevented 14.4 million deaths. Yet biodistribution studies of these lipid nanoparticles revealed predominant accumulation in the liver, a finding that determines both efficacy and safety and underscores why quantitative pharmacokinetic analysis is essential for any nanoparticle system with organ-specific accumulation. Despite this landmark success, only two lipid nanoparticle products and no polymeric nanoparticles have been approved since Comirnaty, the first authorized mRNA vaccine.

The review catalogues the approved landscape to show both what has been achieved and how limited it remains. Of 75 nanomedicines approved by the FDA or the European Medicines Agency, 25 are lipid-based and 24 are polymer-based. As of February 2024, a search for nanoparticle therapeutics on ClinicalTrials.gov identified 164 clinical trials involving systemically administered nanoparticles, with most candidates languishing in early-phase testing. Lipid nanoparticles dominate systemic applications because they are weakly immunogenic, can encapsulate both hydrophilic and hydrophobic drugs, and are relatively easy to manufacture at scale. Polymeric nanoparticles follow, prized for biodegradability and controlled release but plagued by limited encapsulation of hydrophilic cargos and batch-to-batch variability. Protein-based nanoparticles, nanocrystals, and inorganic particles make up the remainder of the approved arsenal.

Central to the translation problem is a chronic shortage of high-quality pharmacokinetic data. Preclinical studies frequently measure only organ-level biodistribution at fixed time points, describing where nanoparticles travel without the mathematical rigor needed to characterize how fast they arrive, how long they persist, and how they are cleared. A meta-analysis compiling more than 2,000 nanoparticle datasets in mice found that 65 percent contained fewer than four time points—far too sparse for proper modeling. Moreover, most preclinical pharmacokinetic studies are conducted with the drug-loaded particle rather than the nanoparticle alone, making it impossible to decouple the contributions of carrier and cargo. Without tissue partition coefficients, organ-specific distribution data, and elimination parameters, even sophisticated models cannot predict clinical behavior accurately.

The physics and biology of nanoparticles make their behavior notoriously difficult to anticipate. Size, surface charge, shape, and composition all govern circulation time, uptake, and clearance, often in nonlinear ways. Particles of roughly 8 to 10 nanometers are typically removed by glomerular filtration in the kidney, yet mesoporous silica and polymeric particles as large as 200 nanometers have been detected in urine. PEG-coated particles up to 114 nanometers have crossed the blood-brain barrier despite intercellular spaces of only 20 to 60 nanometers. Particles larger than about one micrometer become trapped in capillaries and are swiftly engulfed by phagocytic cells. Surface chemistry adds another layer of complexity: PEGylation extends circulation by sterically shielding particles from opsonization, but anti-PEG antibodies can accelerate clearance, and pre-existing anti-PEG immunity has been shown to reduce tumor accumulation and circulating levels of lipid nanoparticles in mice.

Off-target effects compound these challenges. An estimated 95 percent of an administered nanoparticle dose is sequestered by the liver, spleen, and lungs and never reaches its intended target. Accumulation in filtration organs can trigger immune responses, and there is growing evidence that large doses can overwhelm the mononuclear phagocyte system, temporarily impairing its ability to respond to pathogens. Long-term biodistribution and clearance studies, essential for predicting chronic toxicity, remain sparse because they are expensive, time-consuming, and constrained by the short lifespans of laboratory animals. Pharmacokinetic modeling, the authors argue, can fill these voids by simulating multiple-dose regimens, predicting tissue accumulation, and extrapolating toxicity risk without requiring exhaustive animal studies.

To organize these efforts, the review introduces a framework called model-informed nanoparticle development, or MIND, which integrates prior knowledge of nanoparticle physicochemical properties, in vitro and in vivo biological data, and mechanistic modeling into a prospective decision-making pipeline. Within this framework, physiologically based pharmacokinetic (PBPK) models incorporate organ blood flows, tissue composition, and particle-specific processes such as size-dependent uptake, opsonization, and clearance by the mononuclear phagocyte system to predict human exposure from animal data—a critical capability for refining first-in-human dose selection. Population pharmacokinetic (PopPK) models, built on nonlinear mixed-effects approaches, capture variability across patients using sparse sampling, as demonstrated by a model of the nanoparticle-drug conjugate NLG207 developed from 27 patients in two Phase II trials, which revealed distinct kinetics between nanoparticle-bound and free camptothecin. Quantitative systems pharmacology models go further, embedding cellular pathways and disease mechanisms to explain how and why particles behave as they do. Bayesian methods and stochastic, Monte Carlo-based simulations add rigorous quantification of uncertainty and variability.

Real-world examples show what modeling can accomplish. A multiscale PBPK model of mesoporous silica nanoparticles in rats subdivided organs into vascular, extravascular, and phagocytic compartments and identified nanoparticle degradation rate, tumor blood viscosity, particle size, and vascular porosity as dominant determinants of tumor delivery. A gold nanoparticle model developed in mice was extrapolated to rats across 15 organ compartments with long-term simulations extending to 56 days, demonstrating the principle of interspecies scaling on which clinical translation depends. In neonates, where dosing errors are most dangerous, PBPK models built from adult intravenous data have been converted into neonatal models to generate dose recommendations where direct data are nearly impossible to collect. Machine learning is now being layered onto these approaches, optimizing manufacturing parameters and predicting quality attributes before production even begins.

Manufacturing remains a parallel hurdle, and one that modeling increasingly touches. Scale-up demands reproducible physicochemical properties with minimal batch variation, and intricate synthesis schemes often cannot survive industrial demands. The COVID-19 vaccines illustrated the alternative: a deliberately simple lipid nanoparticle design, built from well-characterized components, enabled global production of 19 million doses within months and more than 807 million doses distributed by Moderna alone in 2021. Quality-by-design procedures recommended by international harmonization guidelines, along with 12-month stability studies and roughly 95 percent purity targets, are raising the bar for nanoformulations, though long-term hydrolytic degradation of PEGylated and polymeric particles continues to erode shelf life.

The path forward, the authors conclude, requires culture change as much as technical advance. Regulators and industry have embraced model-informed drug development for small molecules and biologics, but have been slower to accept modeling strategies for nanomedicines, in part because standardized characterization and pharmacokinetic methods remain immature. The reviewers call for regulatory guidelines specifying simulation methods, validation criteria, and reporting standards for nanoparticle models, and for routine integration of real-time biodistribution data from PET imaging and omics datasets into model refinement. If the field adopts this model-informed mindset, the authors argue, the same quantitative rigor that carried mRNA vaccines to billions of arms can be brought to bear on the thousands of nanoparticle designs still waiting in the laboratory—accelerating the arrival of safer, more effective nanomedicines at the bedside.

Subject of Research: Use of pharmacokinetic modeling to overcome translational barriers in nanoparticle-based drug delivery

Article Title: From bench to bedside: Overcoming translational hurdles in nanoparticle research with pharmacokinetic modeling

Article References: Parrot, M., Xu, N., Adnan, M., Cave, J., Ghandehari, H., Nance, E., & Yellepeddi, V. (2026). From bench to bedside: Overcoming translational hurdles in nanoparticle research with pharmacokinetic modeling. Bioengineering & Translational Medicine, Article e70176. https://doi.org/10.1002/btm2.70176

Image Credits: AI Generated

DOI: 10.1002/btm2.70176

Keywords: nanoparticles, pharmacokinetic modeling, PBPK, PopPK, lipid nanoparticles, drug delivery, clinical translation, nanomedicine, biodistribution, MIND framework, toxicity, mRNA vaccines