A . Chevalier et al .: J Extra Corpor Technol 2024 , 56 , 101 – 107 105
Figure
3 . ( A ) The percent recovery of ketamine from control ( red ) and ex-vivo CRRT circuits ( blue ) over time . Error bars represent one standard deviation for n = 3 control and n = 5 experimental circuits . ( B ) The concentration of ketamine in plasma ( blue ) and hemofiltrate ( green ) of ex-vivo CRRT circuits over time . Error bars represent one standard deviation for n = 3 control and n = 5 experimental circuits .
Figure
4 . ( A ) The percent recovery of dexmedetomidine from control ( red ) and ex-vivo CRRT circuits ( blue ) over time . Error bars represent one standard deviation for n = 9 control and n = 3 experimental circuits . ( B ) The concentration of dexmedetomidine in plasma ( blue ) and hemofiltrate ( green ) of ex-vivo CRRT circuits over time . Error bars represent one standard deviation for n = 9 control and n = 3 experimental circuits .
a saturation coefficient of 0.18 . Dexmedetomidine is as lipidsoluble as ketamine ( logP = 2.8 )[ 32 ], but with significantly higher protein binding ( 94 % vs . ~ 50 %) [ 32 ]. Therefore , the fraction of unbound dexmedetomidine is expected to be much lower than for ketamine , contributing to the lower clearance across the membrane . With the low degree of membrane clearance , the majority of dexmedetomidine extraction by the CRRT circuit is theoretically due to adsorption to circuit components . We did not identify prior studies of dexmedetomidine pharmacokinetics during CRRT [ 13 ]. However , our results were consistent with previous ex-vivo studies of dexmedetomidine pharmacokinetics in ECMO , which revealed a high degree of adsorption to circuit components [ 11 , 15 , 33 – 35 ]. These results suggest that , despite lower transmembrane clearance , dexmedetomidine dosing adjustments should be considered in patients supported by CRRT .
These experiments have several limitations . First , these experiments utilized only a single model for each circuit component to mimic the components most frequently used at our institution ; as a result , these experiments are not necessarily reflective of interactions between these medications and alternative oxygenators or dialysis filters . Additionally , all CRRT circuits used a CVVHDF modality ; thus , results are not necessarily reflective of alternative modalities such as CVVHD or CVVHF . However , prior ex-vivo experiments for other drugs have not demonstrated significant differences in medication clearance across dialysis modalities , with the total effluent dose being the impactful determinant of clearance [ 17 , 21 ]. Finally ,