Artificial Neural Network vs. Pharmacometric Model for Population Prediction of Plasma Concentration in Real‐World Data: A Case Study on Valproic Acid

Abstract

We compared the predictive performance of an artificial neural network to traditional pharmacometric modeling for population prediction of plasma concentrations of valproate in real‐world data. We included individuals aged 65 years or older with epilepsy who redeemed their first prescription of valproate after the diagnosis of epilepsy and had at least one valproate plasma concentration measured. A long short‐term memory neural network (LSTM) was developed using the training data set to fit the LSTM and the test data set to validate the model. Predictions from the LSTM were compared with those obtained from the population predictions from a pharmacometric model by Birnbaum et al. which had the best predictive performance for population predictions of valproate concentrations in Danish databases. We used the cutoff of ± 20 mg/L of prediction error to define good predictions. A total of 1,252 individuals were included in the study. The LSTM fitted using the training data set had poor predictive performance in the test data set, but better than that of the pharmacometric model. The proportion of individuals with at least one predicted concentration within ± 20 mg/L of observed concentration was largest in case of the LSTM (64.4%, 95% confidence interval (CI): 58.4–70.2%) compared with the pharmacometric model by Birnbaum et al. (49.8%, 95% CI: 47.0–52.6%). LSTM shows better predictive performance to predict valproate plasma concentrations compared with a traditional pharmacometric model in the investigated setting with real‐world data in older patients with epilepsy where information on exact timepoints for both dosing and plasma concentration measurement are missing.

Artificial Intelligence and Machine Learning Approaches to Facilitate Therapeutic Drug Management and Model-Informed Precision Dosing

“…An additional study used a long short-term memory neural network, a type of artificial neural network, to predict plasma concentrations of valproic acid in older adults and compared this with predicted concentrations from a previously published population PK model. 55 Model performance was assessed by comparing the proportion of individuals with at least one predicted concentration within 620 mg/L of the observed concentrations between the two models. The results showed that the DL model outperformed the population PK model, but further work to improve the predictive performance of this approach was noted.…”

Deep Learning Methods Applied to Drug Concentration Prediction of Olanzapine

“…It was concluded that the regression tree performed the best among the machine learning algorithms in both the derivation and testing cohorts. A long short-term memory (LSTM) model was developed to assess its ability in predicting valproate concentrations in older patients with epilepsy [18]. When compared with the population pharmacometrics model of valproate, the developed LSTM model had a better predictive performance in the external evaluation study.…”