Min Hai

Background: High-dose melphalan 140–200 mg/m² (HDM) with autologous stem cell transplant (ASCT) is standard first-line treatment in multiple myeloma (MM), yet standard BSA-based dosing results in wide variation in systemic exposure (AUC). We developed a pharmacokinetic (PK)-guided dosing strategy using a 100 mg/m² first dose, enabling real-time PK assessment and individualized adjustment for the second dose. We report final results of Phase A of our multi-center Phase 1 trial (NCT04483206, MyMel) evaluating feasibility and accuracy of this personalized approach. Methods: Patients received melphalan 100 mg/m² on Day −3, and seven PK samples were collected and shipped overnight for LC-MS/MS analysis. Real-time AUC estimation using noncompartmental analysis (NCA) guided Day −1 dosing to achieve pre-specified AUC targets (13.5 or 14.5 mg × h/L). For comparison, post hoc Bayesian estimation using a nonlinear mixed effects (NLME) model was performed. Sparse sampling designs were evaluated using NONMEM. Results: All 20 patients successfully received PK-guided dosing, with Day −1 doses determined within 48 h. PK-guided dosing reduced AUC variability (CV 4.20–5.62%), with 19 of 20 achieving AUCs within ±10% of the target, compared to what would have been achieved by BSA dosing (CV 9.74–15.34%). NLME improved accuracy, particularly in patients with missing samples, and maintained performance using only four PK time points. Conclusions: This study demonstrates that PK-guided dosing is accurate and feasible with HDM-ASCT. NLME enhances accuracy and enables simplified sampling. Phase B will identify maximum tolerated systemic exposure of seven additional AUC cohorts using the NLME model and a four-sample design.

Busulfan is commonly used in hematopoietic stem cell transplantation conditioning, with therapeutic drug monitoring (TDM) guiding dose adjustments to promote target exposure attainment. Published busulfan population pharmacokinetic (popPK) models vary in structure and covariates, and it is unclear how broadly applicable one popPK model is across different datasets and patient populations. A popPK model was developed from a retrospective busulfan PK dataset (N = 109 patients). Individual PK parameters were derived by maximum a posteriori (MAP) estimation in four additional datasets. Model-based and TDM-based dosing performance was assessed and compared with noncompartmental analysis and standard weight-based dosing across the five datasets. The model was subsequently used to evaluate minimal sampling schemes. A two-compartment popPK model with sex and body surface area as covariates reliably estimated PK parameters and AUCs across five datasets (N = 464), closely agreeing with NCA. In the original dataset, weight-based dosing would have achieved target AUC in 34% of patients, compared with 40% and 80% using model-based and TDM-based dosing, respectively. Across remaining datasets, model-based dosing performed comparably to weight-based dosing. Sampling could be reduced to four time points, and even a single late sample enabled accurate AUC estimation. The single busulfan popPK model demonstrated adequate generalizability and performed equally with weight-based dosing for achieving specified AUCs. The model could also be used for AUC determination using minimal sampling, though its performance declined in datasets where sampling and dosing schemes differed from the original dataset, highlighting limitations in applying one popPK model across diverse scenarios.

Background/Objectives: Hepatocellular carcinoma (HCC) is the predominant form of liver cancer and currently is the second-leading cause of cancer-related mortality globally. Current front-line systemic therapies for advanced HCC offer only modest improvements in patient overall survival. HCC is a sexually dimorphic disease, and cancer progression is driven in part by AR activity. Here, we present novel niclosamide pro-drugs for use in advanced HCC based upon niclosamide’s known anti-AR activity and additional anti-cancer pathway efficacy. Methods: Niclosamide analogs were evaluated for their impacts on the AR protein in two HCC cell lines with different AR phenotypes. Amino acid conjugates of niclosamide were developed, and pharmacokinetic (PK) analyses were conducted to determine improvements in clearance and oral exposure. Finally, niclosamide analogs and amino acid conjugates were evaluated in an in vivo model of HCC. Results: Niclosamide analogs maintained anti-AR properties in HCC. Valine-conjugated niclosamide showed improved oral exposure, positioning it as a potential therapeutic in advanced HCC. Conclusions: Valine–niclosamide improves upon niclosamide’s poor solubility and oral bioavailability, increasing its utility for a variety of therapeutic uses. Further study of valine–niclosamide in advanced HCC and in other cancers or diseases is warranted.

Background: HOSU-53 (JBZ-001), an orally bioavailable new chemical entity, represents a highly potent dihydroorotate dehydrogenase (DHODH) inhibitor in late preclinical development for application in cancer therapy. Methods: Multiple Good Laboratory Practice (GLP) and non-GLP preclinical studies were conducted in mice, rats, and dogs. Plasma samples of HOSU-53 and dihydroorotate (DHO), the substrate of DHODH, were collected for pharmacokinetic (PK) and pharmacodynamic (PD) assessment and modeling. Two modeling approaches were utilized to understand the PK/PD properties of HOSU-53 and to recommend a first-in-human (FIH) dose. Results: A population PK/PD model was developed using a stochastic approximation of the expectation-maximization method and evaluated using graphical and numerical methods. The PK of HOSU-53 was well described by a two-compartment model with a first-order absorption and linear elimination, and the PD was described by a turnover model. No covariates were considered significant on PK/PD parameters. This model was subsequently used to predict DHO exposures in humans across a range of doses. Additionally, predicted human hepatocellular HOSU-53 concentrations were obtained from a physiologically based PK model constructed in PK-Sim. Conclusions: A first-in-human starting dose of 5 mg once daily was established from the model approaches and will be utilized in the upcoming FIH clinical study.

The exposure to ticagrelor (BRILINTA) is higher in the East Asian population compared with the White population, thus, East Asians have an increased risk of bleeding. We developed a population pharmacokinetic (PopPK) model of ticagrelor based on a randomized 3 × 3 crossover study in healthy subjects. The area under the concentration‐time curve (AUC) of Chinese patients with acute coronary syndrome was simulated based on this model. Following this, eight machine learning (ML) methods were used to construct bleeding risk models. Variables included in the final bleeding risk model were age, hypertension, body weight, AUC, drinking status, calcium channel blockers, antidiabetic medications, β‐blockers, peripheral vascular disease, diabetes, transient ischemic attack, sex, and proton pump inhibitor. In terms of F1 scores and area under the curve of receiver operating characteristic curve (ROC‐AUC), the Random Forest model performed best among all models, with an F1 score of 0.73 and ROC‐AUC of 0.81. Moreover, the PopPK model and ML algorithm were used to bridge the real‐world data to build a bleeding risk prediction model based on drug exposure and clinical information. Using this model, a ticagrelor regimen that is associated with a lower risk of bleeding in individuals can be obtained. This model should be further validated prospectively in clinical settings.