Impact of Milk pH and Fat Content on the Prediction of Milk-to-Plasma Ratio: Knowledge Gap and Considerations for Lactation Study Design and Interpretation.

Background and objective: Different empirical lactation models have been published to predict the milk-to-plasma (M/P) ratio of drugs to gain knowledge on the extent of drug distribution to the breastmilk. M/P ratios will likely vary across the lactation period due to differences in physiological milk pH and fat content, which are not routinely reported in clinical lactation pharmacokinetic studies. This work aims to evaluate the sensitivity of two (a theory-based phase distribution and a log-transformed regression) lactation models for M/P prediction at different physiological milk pH and fat content.

Methods: A literature search was conducted to collate reported M/P ratios for different drugs and their physicochemical parameters required for the prediction of the M/P ratio. Two distribution models were used for M/P ratio predictions. The M/P ratio of drugs was predicted under the physiological milk pHs of 6.8, 7.0, 7.2, and 7.4 and at of 1%, 3%, and 6% fat content. Calculated M/P ratios were compared with the observed M/P ratios.

Results: A total of 200 M/P ratios for 130 compounds (40 acids and 90 bases) were collected from clinical studies and included in the analysis. For both model, precision decreases and bias increases outside the milk pH range 7.0-7.2 and fat contents more than 3%. Significant variability exists in the observed M/P ratios. Both milk pH and fat content are important parameters for model prediction.

Conclusion: Calculated M/P ratios are influenced by multiple covariates, including milk pH and fat content. The phase distribution model is less sensitive to these covariates than the log-transformed model, especially for acidic compounds. For complex matrices such as breastmilk, the actual physiological parameters of the sampled milk, at least milk fat and pH, and their distributions are required covariates to improve the prediction outcomes, design lactation pharmacokinetic studies, and inform the potential breastfed infant dose.