Validation of the pharmacokinetic model for anti- TNFα clearance in infants exposed to anti- TNFα during pregnancy.

BACKGROUND AND AIMS: ECCO guideline recommend postponing live attenuated vaccines in infants exposed to anti-Tumor Necrosis Factor alpha (anti-TNFα) in utero until drug clearance. The aim was to validate the predictive performance of the anti-TNFα clearance model.

METHODS: Newborns and anti-TNFα concentrations from the prospective PETIT cohort were included. The anti-TNFα clearance model was used to predict all measured concentrations in the PETIT cohort, based on the measured cord blood concentration and the mean population clearance described in the model. Bayesian maximum a posteriori optimization was used to estimate the value of drug monitoring. Predictive capability and drug monitoring were assessed through Mean Absolute Error (MAE), Root mean Squared Prediction Error and Limits of Agreement according to Bland and Altman.

RESULTS: Observed drug concentrations after birth were within the 80% prediction interval in 94% of adalimumab samples and 93% of infliximab samples. The anti-TNFα clearance model accurately predicted the concentration at six months after birth with an MAE of 0.03 (SD 0.03) µg/mL for adalimumab and 0.11 (SD 0.18) µg/mL for infliximab based on cord blood concentrations. Addition of an additional sample between 1 and 4 months after birth improved the predictive accuracy for infliximab (MAE 0.05 (SD 0.09)) but not for adalimumab. Guidance for use in clinical practice was formulated.

CONCLUSIONS: The validity of the anti-TNFα clearance model is high, and hence can be used to guide clinicians regarding timing of live vaccines in infants exposed to adalimumab or infliximab in utero.