(W-078) Effect of Convenience Sampling on Overall Drug Exposure during Pregnancy

Jessica Barry, BS – PhD Student, University of Minnesota; Xintian Lyu, BS – PhD Candidate, University of Minnesota; Catherine Sherwin, PhD – Vice President of Clinical Pharmacology and Pharmacometrics, Differentia Bio; Angela Birnbaum, PhD, FAES, MLS(ASCP) – Professor, Experimental and Clinical Pharmacology, University of Minnesota

Objectives: Measuring drug exposure during pregnancy is important for understanding child postpartum outcomes. Researchers and clinicians often work with convenience samples collected during clinic visits with random time after dose (TAD). We characterized how different noncompartmental measures of drug exposure were affected by convenience sampling for different pharmacokinetic parameters.

Methods: Predicted drug concentrations were simulated using a one-compartment model with first order absorption and 1000mg twice daily dosing with linearly increasing clearance (0.032 L/hr/wk) and volume of distribution (0.12 L/wk) for the 40 weeks of pregnancy (Julia v.1.9.2). Two preconception clearances (0.5 and 4 L/h) and two preconception volumes (8 and 128 L) were simulated. Ten mid-monthly samples were taken with uniformly distributed TAD (0-12 h). Four methods were used to calculate drug exposure during pregnancy: 1) lowest concentration, 2) average concentration, 3) highest concentration, 4) trapezoidal AUC0-40wks. To account for gestational age, the methods 1, 2, and 3 were time-weighted by 40 weeks, and for method 4, the first and last concentrations were extended to the beginning and end of pregnancy. Sensitivity analyses on rate of change for volume and clearance were performed.

Results: Drug exposure by method 4 stayed within 3% of exposure calculated by method 3. At high volumes, the range of exposures was smaller (~44,000-46,000 mg/L*wk) and more uniformly distributed than exposures at low volumes (~500,000-600,000 mg/L*wk and heavily skewed right). As expected, the mean of calculated exposure values for each method were lower (~5x) in lower clearance than higher clearance due to decreased drug concentrations. Faster rate of volume increases during pregnancy resulted in a lower the range of possible exposures, which affected methods 1 and 2 more than method 3. For faster increases in clearance, the range in exposure due to TAD stayed consistent until the clearance was high enough to completely eliminate the drug between doses, at which point the range in exposure decreased.

Conclusions: Range of exposures due to random TAD is closely related to volume of distribution, while mean exposure level is more closely related to clearance. For drugs with higher volumes, use of convenience sampling minimally influences exposure measures, but drugs with lower volume may need more consideration of TAD and sampling design. With the physiological volume increases experienced in pregnancy, TAD may be of less concern in data analysis. Clinical studies for postpartum outcomes should consider varying TAD to prevent biasing results due to convenience sampling for overall exposure.

Citations: none