Personalizing mycophenolate therapy: genetic determinants and practical limited sampling strategy for pediatric lupus nephritis.

OBJECTIVE

To identify determinants of mycophenolic acid (MPA) pharmacokinetics (PK) in pediatric patients with lupus nephritis (LN), and to develop a limited sampling strategy (LSS) for estimating MPA-AUC0-12 h.

METHODS

A single-center observational study enrolled biopsy-proven LN receiving mycophenolate mofetil (MMF) at least 7 consecutive days. Univariate analysis were performed to assess associations between MPA PK parameters and demographics, genetic polymorphisms, clinical variables. LSS models for predicting MPA-AUC0-12 h were developed via multivariate stepwise regression analysis.

RESULTS

Statistically significant correlations were observed between MPA-AUC0-12 h and MMF dose, plasma albumin level, estimated glomerular filtration rate, and genetic polymorphisms in HNF1A (rs56097722) and CYP2C8 (rs1058932). CES2 (rs11075646) was significantly associated with MPA half-life (T₁/₂). Single timepoint MPA concentrations demonstrated poor predictive performance for AUC0-12 h(R2 = 0.194-0.553). The optimal LSS model was achieved with four timepoints: MPA-AUC0-12 h = 2.935 + 0.716·C0.5 + 1.305·C1.5 + 3.176·C4 + 4.705·C9 (R2 = 0.945). This model demonstrated good predictive performance (MPE = 1.84%, MAPE = 9.47%, RMSE = 12.53%). The LSS model was validated internally and externally for stability and applicability.

CONCLUSION

Reduced MPA exposure was associated with MMF doses < 20 mg·kg⁻1·day⁻1, hypoproteinemia, preserved or mildly impaired renal function, and HNF1A rs56097722 C/C or CYP2C8 rs1058932 G/G genotypes. CES2 rs11075646 C/C tended to correlate with a shorter MPA T₁/₂. The validated LSS models are recommended for estimating MPA-AUC0-12 h in pediatric LN, which are expected to simplify the therapeutic drug monitoring process.