Physiologically Based Pharmacokinetic (PBPK) Modelling to Predict Human Milk Exposure to medicines: a contribution from the ConcePTION Project
Nina Nauwelaerts (1) , Julia Macente (1) , Neel Deferm (1,2) , Rodolfo Hernandes Bonan (3) , Miao-Chan Huang (1) , Martje Van Neste (4), David Bibi (5), Justine Badee (6), Frederico S. Martins (1), Anne Smits (7, 8, 9) Karel Allegaert (4, 7, 8, 10), Thomas Bouillon (3), Pieter Annaert (1, 3)
(1) Drug Delivery and Disposition, Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Leuven, Belgium; (2) Certara UK Ltd. (Simcyp Division), Sheffield, United Kingdom; (3) BioNotus GCV, Niel, Belgium; (4) Clinical Pharmacology and Pharmacotherapy, Department of Pharmaceutical and Pharmacological Sciences, KU Leuven, Leuven, Belgium; (5) Global Research and Development, Teva Pharmaceutical Industries Ltd., Netanya, Israel.; (6) Novartis Institutes for BioMedical Research, Novartis, Basel Switzerland; (7) Department of Development and Regeneration, KU Leuven, Leuven, Belgium; (8) L-C&Y, KU Leuven Child & Youth Institute, Leuven, Belgium; (9) Neonatal Intensive Care Unit, University Hospitals Leuven, Leuven, Belgium; (10) Department of Hospital Pharmacy, Erasmus University Medical Center, Rotterdam, Netherlands
Objectives:
Breastfeeding is associated with positive health effects for mother and infant [1]. However, postpartum women often need medicines [2]. There is little evidence-based information about the exposure to maternal medicines via human milk. As a result, women might delay the initiation of their much-needed pharmacotherapy, or discontinue breastfeeding.
The aim of this study was to evaluate the predictive performance of PBPK modeling to generate expertise regarding the exposure to maternal medicines via breastfeeding for ten physiochemically diverse medicines.
Methods:
Results are illustrated for amoxicillin, caffeine, cetirizine and levetiracetam. The PBPK models were first developed for “non-lactating” adult individuals in PK-Sim/MoBi v9.1 (Open Systems Pharmacology). The models were accepted when the geometric mean fold error (GMFE) for area-under-the-curve (AUC) and maximum plasma concentration (Cmax) was within 2-fold.
Next, the PBPK models were extended with a breast compartment [3]. In addition, a compartment representing the human milk (volume: 0.5 L) was connected to plasma. The transfer between plasma and human milk was parametrized by a secretion clearance (CLsec) and reuptake clearance (CLre) [4]. The unbound fractions of the medicines in human milk were implemented based on the equations described by Atkinson and Begg (1990) [5].
The structural model was imported into PK-Sim for population simulations with a three months postpartum population (n=1000) [6]. The median breast volume and blood flow were 1 L and 27 mL/min/100 g, respectively. A geometric standard deviation of 1.16 was assumed for the human milk volume. The plasma and human milk concentration-time profiles were simulated, and compared to observed data. The milk-to-plasma ratio (M/P ratio) was calculated based on the predicted median AUC in plasma and human milk for a dosing interval at steady-state.
Finally, the daily infant dosage (DID, mg/kg/day) was calculated by multiplying the average concentration in human milk with the daily milk intake (150 mL/kg/day) [7]. The DID was compared to the maternal dosage (relative infant dose, RID, %) and to the recommended dose prescribed to infants for therapeutic use (50 mg/kg/day for amoxicillin, 5 mg/kg/day for caffeine, 0.5 mg/kg/day for cetirizine and 40 mg/kg/day for levetiracetam).
Results:
The PBPK models for “non-lactating” adult individuals were able to capture the pharmacokinetic profile of the selected medicines, with GMFE for AUC and Cmax within 2-fold prediction error.
The lactation PBPK models resulted in a reasonable prediction of the plasma and human milk concentrations, with most datapoints within the 5th-95th prediction interval. The predicted M/P ratio was 0.15 for amoxicillin. Unfortunately, there is only a single study available for amoxicillin [9]. Applying non-compartmental analysis, and assuming the elimination slope in milk is identical to plasma, a M/P ratio of 0.04 was calculated (4-fold prediction error). For cetirizine, the predicted M/P ratio was 0.12. The observed M/P ratio was calculated using the steady-state AUC in human milk (0.50 mg*h/L) [10], and the plasma AUC in non-lactating adults receiving the same dosing regimen (2.50 mg*h/L). This results in a M/P ratio of 0.12 (1.67 prediction error). For caffeine, the M/P ratio was 0.95 (observed range: 0.52-1.16) [11–13]. For levetiracetam a M/P ratio of 1.11 was predicted (observed range: 0.46-1.79) [14–17].
The infant doses received via breastfeeding were low for amoxicillin, cetirizine and caffeine. The DID (RID) were respectively 0.11 mg/kg/day (0.24 %), 0.002 mg/kg/day (1.24 %), and 0.30 mg/kg/day (5.98 %). The DID of levetiracetam was 6.16 mg/kg/day (12 %), which is still relatively low. These doses were less than 1 % of the therapeutic dose administered to infants for amoxicillin and cetirizine. For caffeine and levetiracetam, the dose was 6 % and 15 % of the therapeutic dose.
Conclusions:
This workflow was applied to ten physiochemically diverse medicines, and resulted in a reasonable prediction for eight of the ten medicines. Similar results were obtained in SimCYP (Certara, Sheffield, UK). Ongoing efforts aim to implement in vitro permeability coefficients across the blood-milk barrier in the PBPK models. A particular strength of this generic workflow is that unique quantitative information on infant exposure to maternal medicines can be generated in an early drug development stage.
Acknowledgement:
This work has received support from the EU/EFPIA Innovative Medicines Initiative [2] Joint Undertaking ConcePTION grant No. 821520. The research leading to these Results was conducted as part of the ConcePTION consortium. This abstract only reflects the personal views of the stated authors. Nina Nauwelaerts also received a fellowship from the Research Foundation—Flanders (1S50721N).
References:
[1] Binns, C.; Lee, M.; Low, W.Y. The Long-Term Public Health Benefits of Breastfeeding. Asia-Pacific J. Public Heal. 2016, 28, 7–14, doi:10.1177/1010539515624964.
[2] Saha, M.R.; Ryan, K.; Amir, L.H. Postpartum Women’s Use of Medicines and Breastfeeding Practices: A Systematic Review. Int. Breastfeed. J. 2015, 10, 1–10, doi:10.1186/s13006-015-0053-6.
[3] Dallmann, A.; Solodenko, J.; Ince, I.; Eissing, T. Applied Concepts in PBPK Modeling: How to Extend an Open Systems Pharmacology Model to the Special Population of Pregnant Women. CPT pharmacometrics Syst. Pharmacol. 2018, 7, 419–431, doi:10.1002/PSP4.12300.
[4] Koshimichi, H.; Ito, K.; Hisaka, A.; Honma, M.; Suzuki, H. Analysis and Prediction of Drug Transfer into Human Milk Taking into Consideration Secretion and Reuptake Clearances across the Mammary Epithelia. Drug Metab. Dispos. 2011, 39, 2370–2380, doi:10.1124/dmd.111.040972.
[5] Atkinson, U.C.; Begg, E.J.E.J.; Atkinson, H.C.; Begg, E.J.E.J. Prediction of Drug Distribution into Human Milk from Physicochemical Characteristics. Clin. Pharmacokinet. 1990, 18, 151–167, doi:10.2165/00003088-199018020-00005.
[6] Job, K.M.; Dallmann, A.; Parry, S.; Saade, G.; Haas, D.M.; Hughes, B.; Berens, P.; Chen, J.; Fu, C.; Humphrey, K.; et al. Development of a Generic Physiologically-Based Pharmacokinetic Model for Lactation and Prediction of Maternal and Infant Exposure to Ondansetron via Breast Milk. Clin. Pharmacol. Ther. 2022, 111, 1111–1120, doi:10.1002/cpt.2530.
[7] FDA Clinical Lactation Studies: Considerations for Study Design Guidance for Industry. Draft Guid. 2019, May, 1–4.
[8] Dallmann, A.; Himstedt, A.; Solodenko, J.; Ince, I.; Hempel, G.; Eissing, T. Integration of Physiological Changes during the Postpartum Period into a PBPK Framework and Prediction of Amoxicillin Disposition before and Shortly after Delivery. J. Pharmacokinet. Pharmacodyn. 2020, 47, 341–359, doi:10.1007/s10928-020-09706-z.
[9] Kafetzis, A.; Siafas, C.A.; Georgakopoulos, P.A.; Papadatos, C.J. Cephalosporins Represent a Group of Antibiot- Ics of Increasing Interest . This Is Due to Their High Antibacterial Activity , Low Toxicity , and Easy Modification of the Molecule Resulting in the Production of New Compounds with Desir- Able Properties . T. Acta Pa’diatr Scand 1981, 70, 285–288.
[10] Wilkerson, H.; Datta, P.; Rewers-Felkins, K.; Baker, T.; Hale, T.W. Maternal Transfer of Cetirizine Into Human Milk. J. Hum. Lact. 2021, 37, 135–138, doi:10.1177/0890334420949847.
[11] Findlay, J.W.A.; DeAngelis, R.L.; Kearney, M.F.; Welch, R.M.; Findlay, J.M. Analgesie Drugs in Breast Milk and Plasma*. Clin. Pharmacol. Ther. 1981, 29, 625–633, doi:10.1038/clpt.1981.87.
[12] Stavchansky, S.; Delgado, M.; Joshi, A.; Combs, A.; Sagraves, R. Pharmacokinetics of Caffeine in Breast Milk and Plasma after Single Oral Administration of Caffeine to Lactating Mothers. Biopharm. Drug Dispos. 1988, 9, 285–299, doi:10.1002/bod.2510090307.
[13] Tyrala, E.E.; Dodson, W.E. Caffeine Secretion into Breast Milk. Arch. Dis. Child. 1979, 54, 787–800, doi:10.1136/adc.54.10.787.
[14] Dinavitser, N.; Kohn, E.; Berlin, M.; Brandriss, N.; Bar-Chaim, A.; Keidar, R.; Livne, A.; Stepensky, D.; Berkovitch, M.; Sheinberg, R. Levetiracetam in Lactation: How Much Is Excreted into Human Breast Milk? Br. J. Clin. Pharmacol. 2022, 88, 199–205, doi:10.1111/bcp.14940.
[15] Johannessen, S.I.; Helde, G.; Brodtkorb, E. Levetiracetam Concentrations in Serum and in Breast Milk at Birth and during Lactation. Epilepsia 2005, 46, 775–777, doi:10.1111/j.1528-1167.2005.54804.x.
[16] Kacirova, I.; Grundmann, M.; Brozmanova, H. Umbilical Cord, Maternal Milk and Breastfed Infant Levetiracetam Concentrations Monitoring at Delivery and during Early Postpartum Period. Pharmaceutics 2021, 13, doi:10.3390/pharmaceutics13030398.
[17] Tomson, T.; Palm, R.; Källén, K.; Ben-Menachem, E.; Söderfeldt, B.; Danielsson, B.; Johansson, R.; Luef, G.; Öhman, I. Pharmacokinetics of Levetiracetam during Pregnancy, Delivery, in the Neonatal Period, and Lactation. Epilepsia 2007, 48, 1111–1116, doi:10.1111/j.1528-1167.2007.01032.x.
