170 A . J . Whelan et al .: J Extra Corpor Technol 2024 , 56 , 167 – 173
Milrinone circuit clearance
Milrinone ex vivo clearance from the CRRT circuit was calculated using the equation :
CL ¼ Dose ð5Þ
AUC
where the dose is 0.1 mg , the amount of milrinone is given to the circuit , and AUC is the area under the curve for the measured milrinone concentrations over time in ex vivo CRRT circuits . AUC was calculated using R Version 4.2 using MESS and tidyverse .
Statistics
We compared milrinone recovery between ECLS circuits and controls at time = 360 minutes using a two-sample t-test with significance defined as p < 0.05 . The lower limit of quantitation ( LLOQ ) for this assay was 2 ng / mL . Thus , a value of 2 ng / mL was used in our calculations for samples that returned as < LLOQ .
Figure 2 . Plasma percent recovery from control and ECMO experiments for milrinone . Error bars represent one standard deviation for n = 3 for control and n = 3 for ECMO experiments .
Results Milrinone in ECMO circuits
Milrinone was minimally extracted by the ECMO circuit ( Figure 2 ). The mean ( standard deviation ) recovery of milrinone in the ECMO circuit at t = 360 min was 100 % ( 10.1 ). Recovery in the ECMO circuit was not significantly different compared to mean ( standard deviation , SD ) recovery in the control ( 103 % [ 4.5 ]) at t = 360min ( p = 0.7 ). Control and ECMO milrinone plasma concentration measurements are reported in Supplementary Tables 1 and 2 .
Milrinone in CRRT circuits
Milrinone was rapidly extracted by the CRRT circuit ( Figure 3 ) with milrinone concentrations below the lower limit of quantitation by 3 h . The mean ( SD ) recovery of milrinone in the CRRT circuit at t = 360 min was 0.73 % ( 0.09 ). Recovery in the CRRT circuit was significantly different compared to mean ( standard deviation ) recovery in the control ( 103 % [ 4.5 ]) at t = 360min ( p 0.001 ).
Because milrinone concentrations in plasma and hemofiltration were below the lower limit of quantitation after 2 h , the sieving coefficient was calculated using all time points up to 2 hours . The mean ( SD ) sieving coefficient across all CRRT circuits was 0.71 ( 0.23 ). This resulted in a mean ( SD ) milrinone transmembrane clearance of 119 mL / min ( 39.3 ). Control and CRRT milrinone plasma and effluent concentration measurements are reported in Supplementary Tables 1 and 3 .
Milrinone CRRT circuit clearance
Milrinone ex vivo AUC from minute 1 to hour 6 , and clearance from the CRRT circuit by run is reported in Table 2 and Supplementary Table 2 . The average clearance is 1350 mL / h ( 9.45 L / h / 70 kg ).
Figure
3 . Plasma percent recovery from CRRT circuit experiments with milrinone . Error bars represent one standard deviation for n = 3 for control and n = 3 for CRRT ( plasma ).
Discussion
In this study , we performed ex vivo ECLS experiments with milrinone to assess its interaction with ECMO and CRRT circuits . Milrinone recovery was not significantly different in ECMO experiments compared to the control ( Figure 2 ) which suggests no adsorption to ECMO circuit components . Both the control and ECMO experiments showed steady concentrations of milrinone with no degradation over time . This contrasts with the CRRT experiments where milrinone was rapidly cleared .
To our knowledge , this is the first evaluation of milrinone interactions with ECMO circuits . Our ECMO experiments showed essentially no interaction of milrinone with the ECMO circuit with 100 % recovery of milrinone over the 6-hour experiments . Prior studies with other drugs and ECMO showed adsorption to circuit components was more likely with highly lipophilic and protein-bound drugs [ 26 , 32 – 37 ]. Milrinone is