Dataset Information

Neural-ODE for pharmacokinetics modeling and its advantage to alternative machine learning models in predicting new dosing regimens.

ABSTRACT: Forecasting pharmacokinetics (PK) for individual patients is a fundamental problem in clinical pharmacology. One key challenge is that PK models constructed using data from one dosing regimen must predict PK data for different dosing regimen(s). We propose a deep learning approach based on neural ordinary differential equations (neural-ODE) and tested its generalizability against a variety of alternative models. Specifically, we used the PK data from two different treatment regimens of trastuzumab emtansine. The models performed similarly when the training and the test sets come from the same dosing regimen. However, for predicting a new treatment regimen, the neural-ODE model showed substantially better performance. To date, neural-ODE is the most accurate PK model in predicting untested treatment regimens. This study represents the first time neural-ODE has been applied to PK modeling and the results suggest it is a widely applicable algorithm with the potential to impact future studies.

SUBMITTER: Lu J

PROVIDER: S-EPMC8283337 | biostudies-literature | 2021 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Neural-ODE for pharmacokinetics modeling and its advantage to alternative machine learning models in predicting new dosing regimens.

Lu James J Deng Kaiwen K Zhang Xinyuan X Liu Gengbo G Guan Yuanfang Y

iScience 20210630 7

Forecasting pharmacokinetics (PK) for individual patients is a fundamental problem in clinical pharmacology. One key challenge is that PK models constructed using data from one dosing regimen must predict PK data for different dosing regimen(s). We propose a deep learning approach based on neural ordinary differential equations (neural-ODE) and tested its generalizability against a variety of alternative models. Specifically, we used the PK data from two different treatment regimens of trastuzum ...[more]

PMID: 34308294

Similar Datasets

Project description:Dolutegravir-based regimens are recommended as first-line therapy for HIV in low- and middle-income countries where tuberculosis is the most common opportunistic infection. Concurrent HIV/tuberculosis treatment is challenging because of drug-drug interactions. Our analysis aimed to characterize dolutegravir's population pharmacokinetics when coadministered with rifampicin and assess alternative dolutegravir dosing regimens. We developed a population pharmacokinetic model of dolutegravir in NONMEM with data from two healthy-volunteer studies (RADIO and ClinicalTrials.gov identifier NCT01231542) and validated it with data from the INSPIRING study, which consisted of participants living with HIV. The model was developed with 817 dolutegravir plasma concentrations from 41 participants. A 2-compartment model with first-order elimination and lagged absorption best described dolutegravir's pharmacokinetics. For a typical 70-kg individual, we estimated a clearance, absorption rate constant, central volume, and peripheral volume of 1.03 L/h, 1.61 h-1, 12.7 L, and 3.85 L, respectively. Rifampicin coadministration increased dolutegravir clearance by 144% (95% confidence interval [CI], 126 to 161%). Simulations showed that when 50 or 100 mg once-daily dolutegravir is coadministered with rifampicin in 70-kg individuals, 71.7% and 91.5% attain trough concentrations above 0.064 mg/L, the protein-adjusted 90% inhibitory concentration (PA-IC90), respectively. The model developed from healthy-volunteer data describes patient data reasonably well but underpredicts trough concentrations. Although 50 mg of dolutegravir given twice daily achieves target concentrations in more than 99% of individuals cotreated with rifampicin, 100 mg of dolutegravir, once daily, in the same population is predicted to achieve satisfactory pharmacokinetic target attainment. The efficacy of this regimen should be investigated since it presents an opportunity for treatment simplification.

Dataset Information

Neural-ODE for pharmacokinetics modeling and its advantage to alternative machine learning models in predicting new dosing regimens.

Publications

Neural-ODE for pharmacokinetics modeling and its advantage to alternative machine learning models in predicting new dosing regimens.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets