Physiologically-based Pharmacokinetic Modeling of Belimumab for Systemic Lupus Erythematosus Patients to Predict the Withdrawal Time in Pregnancy and Vaccine Time in Infants

Belimumab (BEL) is a monoclonal antibody that targets B cell activating factor (BAFF) and has been approved for treating systemic lupus erythematosus (SLE). However, its optimal withdrawal time during pregnancy and the appropriate vaccine timing for infants remain unresolved. This study aimed to analyze variations in the plasma concentrations of BEL in pregnant SLE patients and infants using a physiologically-based pharmacokinetic (PBPK) model. Additionally, it provided evidence-based guidance on drug withdrawal timing and infant vaccination schedules. A PBPK model was developed and validated through comparison with drug concentrations in clinical data. The results showed that most observed values were within 0.5 to 2 times the predicted values, indicating that the model successfully predicted the concentration-time profiles of BEL in pregnant SLE patients, fetuses, and infants. The PBPK model shows that both the IV 10 mg/kg every 4 weeks and the subcutaneous 3 mg/kg weekly (each dose of approximately 200 mg) regimen during pregnancy can keep maternal and fetal exposures within the range of the assumed appropriate exposure level for fetal safety. Subsequently, the PBPK model predicts that discontinuing the intravenous 10 mg/kg regimen at least three weeks before delivery and stopping the subcutaneous 3 mg/kg weekly or 4 mg/kg bi-weekly regimen two weeks prior to delivery would minimize neonatal exposure. Furthermore, the model predicts that the clearance of belimumab will persist until 6.5 to 7 months after birth, when vaccination may be considered if the immune system has recovered by this time. This work offers a PBPK-based quantitative reference for BEL pharmacokinetics during pregnancy and infancy and may help optimize treatment planning in pregnant patients with SLE. The findings should be interpreted as model-informed and require further validation.