Aspirin pharmacokinetics and pharmacodynamics through gestation
Rupsa C Boelig,Gagan Kaushal,Ankit Rochani,Steven E McKenzie,Walter K Kraft
DOI: https://doi.org/10.1016/j.ajog.2023.12.028
Abstract:Background: Low dose aspirin is recommended for prevention of preeclampsia, however there is not consensus on the appropriate dose. Pregnancy specific changes have the potential to impact the pharmacology of aspirin in pregnancy, however there are very limited studies on aspirin pharmacokinetics in pregnancy and none linking pharmacokinetics (PK, drug dose and drug level) to pharmacodynamics (PD, drug dose and physiologic response) in pregnancy. As a result, we do not have a good understanding of the pharmacologic response to aspirin in pregnancy, which has important implications for clinical efficacy. We sought to describe the PK and PD of aspirin through pregnancy and to identify individual covariates that impacted aspirin PK/PD. Objective: We sought to describe the PK and PD of aspirin through pregnancy (first and third trimester), to identify covariates that significantly impact aspirin PK and to identify the relationship between aspirin PK and PD. Study design: This is a prospective study of patients at high risk for preeclampsia recommended to take 81 mg aspirin daily. This study involved 3 visits as follows: (1) baseline, first trimester (10-16 weeks of gestation) 6-hour PK visit, done before initiation of aspirin; (2) follow-up 1: 2 to 4 weeks after aspirin initiation; and (3) follow-up 2: third trimester 6-hour PK visit (28-32 weeks of gestation). The following were assessed at each visit: weight or body mass index, platelet function analysis-100 (Siemens), urinary thromboxane B2, serum thromboxane B2, and plasma salicylic acid. The PK visits consisted of blood work done at baseline (predose), administration of 81 mg nonenteric coated aspirin, and then plasma blood level of salicylic acid assessed at 30 minutes and then hourly 1 to 6 hours after dose. Pearson correlation and multivariable regression were used to identify associations between parameters and identify relevant covariates. Log-adjusted values were used for regression analysis. P<.05 was considered statistically significant. Results: Nineteen participants were included with first trimester data, and 16 with third trimester data. There was no statistically significant change in mean PK parameters between the first and third trimester, although there was a trend to lower peak concentration in the third than in the first trimester (P=.08). In multivariable regression, baseline obesity and current body mass index as a continuous measures were negatively associated with log-adjusted peak salicylic acid concentration (-0.28 [-0.46 to -0.11], P=.003 and -0.02 [-0.03 to -0.009], P=.001, respectively) and log-adjusted plasma salicylic acid area under the curve 0 to 6 hours postdose (-0.25 [-0.45 to 0.05], P=.02, -0.04 [-0.07 to -0.01], P=.008 respectively). There was a significant decrease in urinary thromboxane 2 to 4 weeks after aspirin initiation compared with baseline, which correlated with a concomitant increase in platelet function analysis-100 closure time. In multivariable regression model, there was a strong association between plasma salicylic acid concentration (area under the curve 0-6 hours postdose) and urinary thromboxane (B=-3.12 [-5.38 to -1.04], P=.006), and with urinary thromboxane suppression and platelet inhibition, platelet function analysis-100 (-0.23 [-0.31 to -0.14], P<.001). With progressive thromboxane suppression, platelet inhibition (platelet function analysis-100 closure time) increased. Individual comorbidities, including weight, hypertension, and pregestational diabetes (Type I or II), also impacted aspirin response. Conclusion: We have demonstrated the relationship between individual factors, plasma concentrations of salicylic acid, thromboxane suppression, and platelet inhibition at a single dose (81 mg) of aspirin taken through pregnancy. Our findings suggest that dose modification of aspirin in pregnancy may help to achieve the optimal response. Our results may be used to facilitate computational modeling to identify optimal dosing, taking into consideration individual factors.