168 A . J . Whelan et al .: J Extra Corpor Technol 2024 , 56 , 167 – 173
milrinone disposition in ECLS , patients are at risk for either treatment failure from subtherapeutic dosing or toxicity from excessive exposure . Therefore , there is an urgent need to understand how milrinone interacts with ECLS circuits .
The interactions between ECLS circuits and individual drugs have been studied in ex vivo experiments where a blood-primed ECLS circuit is dosed with a drug and concentrations are measured over time [ 16 , 25 – 29 ]. The extent of drug adsorption to the circuit components , impact of surface coatings , and drug clearance by the hemofilter can be assessed by this method . We performed ex vivo CRRT and ECMO experiments in this study to quantify the interaction of milrinone with ECLS to determine the degree of extraction and inform optimal dosing . Since milrinone is mildly protein-bound , we hypothesize that it will likely be readily filtered by CRRT and adsorb minimally to ECMO circuits .
Materials and methods
We evaluated the extraction of milrinone from bloodprimed ECMO and CRRT circuits . ECMO and CRRT circuits were each set up in a closed loop and primed with a blood solution using previously published methods [ 16 , 30 , 31 ]. Circuits were then dosed with milrinone to achieve therapeutic concentrations ( 200 ng / mL ) [ 22 , 23 ], and concentrations were measured over time to quantify the extent of milrinone extraction by the circuit . A control sample containing the same prime solution was also dosed to achieve similar concentrations ( 200 ng / mL ) that were measured over time to understand milrinone degradation [ 19 , 20 ].
Extracorporeal membrane oxygenation ( ECMO ) setup
ECMO ex-vivo experiments were conducted as previously reported [ 31 ]. In short , ECMO circuits ( n = 3 ) were set up as illustrated in Figure 1A and included : a reservoir ( Viaflex 1000 mL , Baxter , Deerfield , IL ); a centrifugal pump ( Rotaflow Pump , Maquet ); a hollow-fiber oxygenator ( Quadrox-iD , Maquet , Hirrlingen , Germany ); and tubing ( Sorin Smart Perfusion Pack , LivaNova , London , UK ) ( Table 1 ). Circuits were primed with a mixture of Plasma-Lyte A crystalloid (~ 200 – 300 mL ; Baxter Healthcare , Deerfield , IL ), 1 unit of thawed human plasma frozen within 24 hours after phlebotomy ( 250-300 mL ), 2 units of human red blood cells ( adenine saline added leukocyte reduced [~ 600 mL ]), tromethamine ( 2 g ), heparin sulfate ( 500 units ), sodium bicarbonate ( 7 mEq ), calcium gluconate ( 650 mg ), and human serum albumin ( 12.5 g ) for a total priming fluid volume of about 1300 mL . To minimize the strain on hospital blood bank supply , expired blood products from the American Red Cross were used to prime the circuits . In order to maintain the physiologic pH ( 7.2 – 7.5 ), CO 2 and tromethamine were added to the priming solution and as necessary during the experiment . ECMO circuit flow was maintained at 1.0 L / min ( HT110 with H8XL flow sensor , Transonic , Davis , CA ). A Quadrox-iD integrated heat exchanger was connected to an ECMO water heater ( Cincinnati Sub-Zero , Cincinnati , OH ) to maintain the temperature of the priming solution at 37 ° C . The haemoglobin of this solution was 10 g / dL .
Continuous renal replacement therapy ( CRRT ) setup
Three CRRT ex-vivo experiments were run using the PrisMax system ( Baxter Healthcare , Deerfield , IL ) with HF1000 filters as previously reported [ 31 ]. In short , the HF1000 filter set was set up in a closed loop by connecting it to a reservoir ( EXACTAMIX EVA , Baxter Healthcare ). The system was primed with 0.4 units of fresh frozen plasma ( 125 mL ), 1 unit of packed red blood cells ( 300 mL ), Plasma-Lyte A crystalloid ( 150 mL ), tromethamine ( 1.5 g ), heparin sulfate ( 350 units ), sodium bicarbonate ( 7 mEq ), calcium gluconate ( 180 mg ), and human serum albumin ( 6.25 g ) for a total priming fluid volume of ~ 600 mL . To maintain physiologic pH additional tromethamine was added as needed . The blood mixture temperature was kept at 37 ° Cby a TherMax blood warmer . At our institution , continuous venovenous hemodiafiltration ( CVVHDF ) is the modality used for critically ill patients , thus this was the mode selected for these experiments . The following prescription was used : blood flow rate ( BFR ) 9000 mL / h , dialysis solution flow rate ( DIA ) 1000 mL / h , pre-blood pump fluid ( PBP ) 700 mL / h , total replacement solution flow rate ( REP ) 300 mL / h delivered after filtration , and patient fluid removal flow rate net 0 mL / h ( Table 1 ), for a total effluent rate of 2000 mL / h . PrismaSATE 4 / 2.5 Dialysis Solution ( Baxter Healthcare , Deerfield , IL ) was used for dialysis , a pre-blood pump , and all replacement fluids .
Control setup
To determine drug degradation in the blood mixture throughout the experiment , three control experiments were included . For each control , 50 mL of blood priming solution was drawn from the ECMO circuits before drug dosing and placed into polypropylene centrifuge tubes ( 229426 , CELL- TREAT , Pepperell , MA ). These samples were dosed to achieve similar concentrations as the circuits and incubated in a water bath heated to 37 ° C for the duration of the experiment .
Drug administration and sample collection
Milrinone was administered to the ECMO circuits via a three-way stopcock downstream to the sampling port on the arterial limb ( Figure 1A ). One dose of milrinone was administered to each circuit at time zero and blood samples were collected at 1 , 5 , 15 , 30 , 60 , 120 , 180 , 240 , and 360 min after drug administration .
Milrinone was administered to CRRT circuits via an access line downstream of the reservoir ( Figure 1B ), blood was sampled from the post-filter sampling port , and effluent was sampled from the post-filter effluent sampling port . Milrinone was administered at time zero and blood and effluent samples were collected at 1 , 5 , 15 , 30 , 60 , 120 , 180 , 240 , and 360 min . Blood samples were centrifuged at 3000g and 4 ° C for 10 minutes immediately after collection . Plasma and effluent samples were then frozen in cryovials ( Fisher Scientific , Pittsburgh , PA ) at �20 ° C for < 6 h and stored at �80 ° Cuntil analysis .