Mizuno Lab

Projects

Supporting Publications

1. Clinical implementation of pharmacogenetics and model-informed precision dosing to improve patient care.
2. Model-informed Precision Dosing for Biologics Is Now Available at the Bedside for Patients With Inflammatory Bowel Disease.
3. Developmental Pharmacokinetics of Sirolimus: Implications for Precision Dosing in Neonates and Infants with Complicated Vascular Anomalies.

Advancing Next-Generation Precision Dosing through AI and Pharmacometrics

Our laboratory integrates artificial intelligence (AI) and machine learning (ML) with mechanistic pharmacometrics to revolutionize pediatric precision dosing. By bridging the gap between data-driven algorithms and physiological principles, we are developing scalable tools to optimize drug therapy.

Our core research pillars include:

• Generative AI for Virtual Populations: To address the scarcity of pediatric clinical data, we developed a physiology-informed Conditional Variational Autoencoder (cVAE). By integrating real-world measurements with mechanistic model-based simulations and constraints, this model generates realistic virtual patients with biologically coherent organ weights and blood flows, facilitating robust Physiologically Based Pharmacokinetic (PBPK) simulations.
• Hybrid PK-ML Modeling: We are enhancing predictive accuracy by combining population PK models with ML algorithms like XGBoost. Our hybrid frameworks leverage ensemble modeling with synthetic data and error-correction techniques to significantly reduce prediction errors for drugs such as infliximab compared to conventional Bayesian estimation alone.
• Automated Decision-Making via Reinforcement Learning: Moving beyond prediction to automated decision support, we employ Deep Q-Networks (DQN) trained in virtual simulation environments. These RL agents learn optimal, adaptive dosing strategies that maximize therapeutic target attainment while adhering to clinical constraints, offering a standardized approach to precision dosing.

Supporting Publications

1. Physiology-Informed Conditional Variational Autoencoder for Generating Pediatric Virtual Patients.
2. Model-Informed Deep Q-Networks to Guide Infliximab Dosing in Pediatric Crohn's Disease.
3. Hybrid Population Pharmacokinetic-Machine Learning Modeling to Predict Infliximab Pharmacokinetics in Pediatric and Young Adult Patients with Crohn's Disease.
4. Machine Learning Modeling for Predicting Infliximab Pharmacokinetics in Pediatric and Young Adult Patients With Crohn Disease: Leveraging Ensemble Modeling With Synthetic and Real-World Data.
5. Artificial Intelligence and Machine Learning Approaches to Facilitate Therapeutic Drug Management and Model-Informed Precision Dosing.

Development of EHR-Integrated Clinical Decision Support Dashboards

The process of gathering clinical information, performing model-informed analyses, and translating results into actionable recommendations is complex and time-consuming. To enable broad adoption of model-informed precision dosing (MIPD) at the bedside, EHR-integrated, user-friendly clinical decision support (CDS) tools are essential.

As part of multidisciplinary teams at Cincinnati Children’s, we work to develop CDS platforms that bring MIPD directly into clinical workflows. One prominent example is RoadMAB™, an EHR-integrated precision dosing dashboard for monoclonal antibodies led by Phillip Minar, MD, MS. RoadMAB provides real-time visualization of drug concentrations, disease activity markers, and individualized dosing recommendations to support proactive therapeutic decision making. More information about this project can be found on the Minar Lab.

Supporting Publications

1. Electronic Health Record (EHR)-embedded Decision Support Platform for Morphine Precision Dosing in Neonates.
2. Real-world Infliximab Pharmacokinetic Study Informs an Electronic Health Record-Embedded Dashboard to Guide Precision Dosing in Children with Crohn's Disease.
3. Precise infliximab exposure and pharmacodynamic control to achieve deep remission in paediatric Crohn's disease (REMODEL-CD): study protocol for a multicentre, open-label, pragmatic clinical trial in the USA.
4. Model-informed Precision Dosing for Biologics Is Now Available at the Bedside for Patients With Inflammatory Bowel Disease. 

Physiologically-Based Pharmacokinetics (PBPK) Modeling to Predict Drug Exposure in Pediatric and Maternal-Fetal Populations

PBPK is a mathematical modeling technique to describe drug behavior in the body (i.e., oral absorption, distribution to tissues, metabolism, and elimination) based on drug-specific physicochemical properties and the patient’s physiology. Our lab utilizes PBPK technologies to predict drug disposition and exposure in support of optimal dose selections in special populations such as neonates and infants, children, pregnant women, and their fetuses.

Supporting Publications

1. A review of pregnancy-induced changes in opioid pharmacokinetics, placental transfer, and fetal exposure: Towards fetomaternal physiologically-based pharmacokinetic modeling to improve the treatment of neonatal opioid withdrawal syndrome.
2. Physiologically-Based Pharmacokinetic Modeling to Investigate the Effect of Maturation on Buprenorphine Pharmacokinetics in Newborns with Neonatal Opioid Withdrawal Syndrome.
3. Forecasting Fetal Buprenorphine Exposure through Maternal-Fetal Physiologically Based Pharmacokinetic Modeling

Clinical Pharmacokinetics Consultation

We offer a model-informed clinical pharmacokinetics (PK) consultation service, providing real-time PK assessment and dose recommendations to achieve the predefined goal exposure as part of routine clinical care or prospective concentration-control clinical trials. We provide consultation for various drugs, including mTOR inhibitors, anti-cancer drugs, immunosuppressive agents, anti-inflammatory biologics, psychiatric drugs, and anti-infective drugs.

Supporting Publications

1. Model-based precision dosing of sirolimus in pediatric patients with vascular anomalies.
2. Significant effect of infection and food intake on sirolimus pharmacokinetics and exposure in pediatric patients with acute lymphoblastic leukemia.
3. Model-Informed Estimation of Acutely Decreased Tacrolimus Clearance and Subsequent Dose Individualization in a Pediatric Renal Transplant Patient With Posterior Reversible Encephalopathy Syndrome.
4. Preventative treatment of tuberous sclerosis complex with sirolimus: Phase I safety and efficacy results.

Model-Informed Pediatric Drug Development (MIPDD)

As part of Pharmacometrics Services, we provide strategic clinical pharmacology consulting, pediatric study design, and data analysis support, leveraging our pharmacometrics and systems pharmacology research expertise. We utilize quantitative modeling and clinical trial simulations to evaluate potential clinical trial designs, extrapolate from adult clinical data to pediatric patients, identify age-appropriate pediatric dosing regimens, determine optimal sampling designs, and estimate required sample sizes. We have a track record of a portfolio of MIPDD projects in collaboration with bio-pharmaceutical industries and academic collaborators. Learn more about our Pharmacometrics Services at Cincinnati Children's.

Supporting Publication

1. Model-Informed Pediatric Drug Development: Application of Pharmacometrics to Define the Right Dose for Children.