IntroductionAtrial fibrillation (AF) increases cardiovascular morbidity and mortality. To reduce thrombosis and bleeding risks, and due to high variability of unfractionated heparin (UFH) effect, activated clotting time (ACT) is used during radiofrequency catheter ablation (RFCA) of AF to guide UFH dose. This study aimed to develop a population PK-PD model and perform multivariable analysis in order to identify the most significant covariates associated with interindividual variability of UFH. MethodsElectronic medical records from 668 patients undergoing RFCA were analyzed, including relevant covariates. The relationship between UFH dose and ACT and the impact of the main covariates were characterized using a mixed-effect PK-PD model. Multivariable analysis was then used to identify predictors of ACT 15 minutes after UFH administration (ACT15). ResultsA two-compartment PK model with linear elimination and a direct Emax PD model with a baseline and sigmoidicity best described the observed ACT values. Pretreatment with dabigatran, warfarin, or fluindione significantly influenced baseline ACT. Pretreatment with vitamin K antagonists or low molecular weight heparin explained Emax variability. The multivariable model identified baseline ACT, initial UFH dose, and previous anticoagulant as the main predictors of ACT15. Model evaluation through resampling and external validation showed accurate ACT15 predictions. ConclusionThis study presents the first population PK-PD model characterizing the relationship between UFH doses and ACT during RFCA, along with multivariable analysis. Additionally, predictive calculators for ACT15 and UFH dose based on patient and procedural characteristics were developed, enhancing personalized anticoagulation management during RFCA.
Read full abstract