(1164) A Pharmacogenetic-Based Integrated Limited Sampling Strategy for Mycophenolic Acid in Lung Transplant Recipients

PurposeMycophenolic Acid (MPA) is the most used anti-proliferative agent after lung transplantation, but its pharmacokinetic (PK) variability has limited therapeutic drug monitoring. The purpose of this study was to integrate genetic and clinical factors associated with MPA PK variability with traditional PK measurements to create a clinically feasible, more accurate exposure estimation tool.MethodsWe performed PK analysis (AUC0-12) on 60 adult lung transplant recipients receiving MPA. We genotyped polymorphisms previously associated with MPA metabolism and collected relevant clinical data. We calculated Pearson correlation and performed univariate linear regression to evaluate the association of AUC0-12 with each pharmacokinetic, genetic, and clinical variable. Limiting PK timespan to 2 hours for clinical feasibility, we performed lasso regression analysis to create a final integrated estimation tool.ResultsOf individual PK measurements, C0, C0.25, C1, and C2 were most strongly associated with MPA AUC0-12 (r 2=0.44, 0.50, 0.52, and 0.67). SLCO1B3 SNP rs7311358 was also significantly associated with MPA AUC0-12 (p=0.049). Inclusion of clinical and genetic variables significantly improved predictive performance of all single PK measures. A final model including C0, C0.25, C2, rs7311358, creatinine clearance, albumin, and valcyte use predicted MPA AUC0-12 with an R2 of 0.90 and was superior to a PK-only model including the same time points (p=0.016).ConclusionAn integrated AUC estimation tool is superior to PK only estimation in MPA therapeutic monitoring. Validation of this tool can streamline future MPA pharmacokinetic and pharmacogenetic research and would allow for clinically feasible MPA therapeutic drug monitoring. Mycophenolic Acid (MPA) is the most used anti-proliferative agent after lung transplantation, but its pharmacokinetic (PK) variability has limited therapeutic drug monitoring. The purpose of this study was to integrate genetic and clinical factors associated with MPA PK variability with traditional PK measurements to create a clinically feasible, more accurate exposure estimation tool. We performed PK analysis (AUC0-12) on 60 adult lung transplant recipients receiving MPA. We genotyped polymorphisms previously associated with MPA metabolism and collected relevant clinical data. We calculated Pearson correlation and performed univariate linear regression to evaluate the association of AUC0-12 with each pharmacokinetic, genetic, and clinical variable. Limiting PK timespan to 2 hours for clinical feasibility, we performed lasso regression analysis to create a final integrated estimation tool. Of individual PK measurements, C0, C0.25, C1, and C2 were most strongly associated with MPA AUC0-12 (r 2=0.44, 0.50, 0.52, and 0.67). SLCO1B3 SNP rs7311358 was also significantly associated with MPA AUC0-12 (p=0.049). Inclusion of clinical and genetic variables significantly improved predictive performance of all single PK measures. A final model including C0, C0.25, C2, rs7311358, creatinine clearance, albumin, and valcyte use predicted MPA AUC0-12 with an R2 of 0.90 and was superior to a PK-only model including the same time points (p=0.016). An integrated AUC estimation tool is superior to PK only estimation in MPA therapeutic monitoring. Validation of this tool can streamline future MPA pharmacokinetic and pharmacogenetic research and would allow for clinically feasible MPA therapeutic drug monitoring.