ABSTRACT
Objective: Currently, limited information is available on the milk transfer properties of drugs when consumed by lactating women. Therefore, we aim to construct a prediction model of milk transfer of drugs using machine learning methods.Methods: We obtained data from Hale’s Medications & Mothers’ Milk (MMM) and SciFinder®, and then constructed the datasets. The physicochemical and pharmacokinetic data were used as feature variables with M/P ratio ≥ 1 and M/P ratio < 1 as the objective variables, classified into two groups as the classification of milk transferability. In this study, analyses were conducted using machine learning methods: logistic regression, linear support vector machine (linear SVM), kernel method support vector machine (kernel SVM), random forest, and k-nearest neighbor classification. The results were compared to those obtained with the linear regression equation of Yamauchi et al. from a previous study. The analysis was performed using scikit-learn (version 0.24.2) with python (version 3.8.10).Results: Model construction and validation were performed on the training data comprising 159 drugs. The results revealed that the random forest had the highest accuracy, area under the receiver operating characteristic curve (AUC), and F value. Additionally, the results with test data A and B (n = 36, 31), which were not used for training, showed that both F value and accuracy for the random forest and the kernel method SVM exceeded those with the linear regression equation of Yamauchi et al. Conclusion: We were able to construct a predictive model of milk transferability with relatively high performance using a machine learning method capable of nonlinear separation. The predictive model in this study can be applied to drugs with unknown M/P ratios for providing a new source of information on milk transfer.
ABSTRACT
Studies on milk transfer of drugs in non-human primates (NHPs) are among the crucial components in the assessment of peri- and postnatal toxicity because of the similarity between NHPs and humans. To evaluate the milk transfer of valproic acid (VPA) in NHPs, the toxicokinetics of VPA, an antiepileptic drug, were studied in pregnant cynomolgus monkeys. VPA was administered once daily to pregnant cynomolgus monkeys at doses of 0, 30, 90, and 270 mg/kg by oral gavage from Day 100 of gestation (GD 100) to Day 31 of lactation (LD 31). Concentrations of VPA and its metabolite, 4-ene-VPA, in the maternal plasma on GD 100, GD 140, and LD 30, and concentrations of VPA and 4-ene-VPA in the offspring plasma and milk on LDs 30 and 31, respectively, were quantified using liquid chromatography tandem mass spectrometry (LC/MS/MS). After administration of a single oral dose of VPA to pregnant monkeys on GD 100, the concentrations of VPA and 4-ene-VPA were generally quantifiable in the plasma of all treatment groups up to 24 hr after administration, which showed that VPA was absorbed and that the monkeys were systemically exposed to VPA and 4-ene-VPA. After administration of multiple doses of VPA to the monkeys, VPA was detected in the pup's plasma and in milk taken on LD 30 and LD 31, respectively, which showed that VPA was transferred via milk, and the pup was exposed to VPA. Further, the concentration of VPA in the milk increased with an increase in the dose. Extremely low concentrations of 4-ene VPA were detected in the milk and in the pup plasma. In conclusion, pregnant monkeys were exposed to VPA and 4-ene-VPA after oral administration of VPA at doses of 30, 90, and 270 mg/kg/day from GD 100 to LD 31. VPA was transferred via milk, and the VPA exposure to the pup increased with an increase in the dose of VPA. The metabolite, 4-ene VPA, was present in extremely low concentrations (< 0.5 microg/ml) in the milk and in the pup plasma. In this study, we established methods to confirm milk transfer in NHPs, such as mating and diagnosis of pregnancy by examining gestational sac with ultrasonography, collection of milk and pup plasma and determination of toxicokinetics, using cynomolgus monkeys.