Extracorporeal membrane oxygenation (ECMO) is an essential therapy for patients experiencing acute respiratory or cardiac failure, which supports blood oxygenation and circulation for a long period.1,2 Therefore, the constituents of the ECMO circuit, including the tubing, blood pump, and oxygenator, are a vital part of patients’ body during the treatment by engaging in their complex physiologic processes. A major complication of ECMO treatment is significant changes in pharmacokinetics of drugs injected into the patients for antibiosis, sedation, analgesia, or anticoagulation, which frequently leads to therapeutic failure because of drug sequestration in the ECMO circuit.3
Because the physiologic conditions, materials of the ECMO circuit, and drug characteristics differ among patients, products, and drugs, respectively, such drug sequestration is likely associated with several conditioning factors. Potential determinants of the altered pharmacokinetics revealed by previous studies include molecular size, degree of ionization,4 lipophilicity,5 plasma protein binding,6 type of priming solution,7 and type of pump or oxygenator.8 However, the responsible factor is unclear, likely because of improperly controlled experiments.
For example, although materials used to manufacture each component of the ECMO circuit and those for surface coating vary among products, most previous studies involved ex vivo experiments using an entire ECMO circuit5,6,9 rather than focusing on individual components of the circuit, having failed to identify out the constituents with the greatest influence on drug sequestration. Indeed, a previous study has shown the necessity of considering the effect of coatings on drug sequestration by demonstrating that different surface coatings on extracorporeal life support (ECLS) circuits affect fentanyl and morphine sulfate adsorption.10
The purpose of the current study is therefore to quantify the influences of each circuit element with different coating materials on drug sequestration by focusing on the interactions between materials and drugs. We used deionized (DI) water as the base solution for dissolving each drug to minimize other influential sources related to ionic reactions. The drugs used in this in vitro study were the antibiotic meropenem, analgesic dexmedetomidine, and anticoagulant heparin.
Materials and Methods
Extracorporeal Membrane Oxygenation Circuit Materials and Components
The ECMO circuit components that are in contact with circulating blood are the tubing, pump housing, impeller, connector, gas exchanger, heat exchanger, and oxygenator housing. Because the pump housing, impeller, connector, and oxygenator housing are made of polycarbonate (PC), the materials that need to be assessed for drug loss are the tubing, gas exchanger, heat exchanger, and oxygenator housing. Accordingly, a total of 132 samples including the controls for drug loss at 22°C (R) and at 37°C (C) were analyzed for each drug at 1, 6, 12, and 24 h. Regarding tubing, Tygon S-50-HL Medical tubing (T), Maquet Bioline-coated tubing (B) (Maquet Cardiopulmonary, AG Hirrlingen, Germany), and Terumo X-coated tubing (X) (Terumo Cardiovascular Systems Corp., Ann Arbor, MI) were evaluated. The Quadrox PLS and Capiox RX oxygenators were disassembled into a gas exchanger, heat exchanger, and housing. The Maquet Quadrox PLS oxygenator is composed of polymethylpentene (PMP) hollow fibers for gas exchange (MaG), polyurethane (PU) fibers for heat exchange by water flow (MaW), and a PC housing (MaH), all of which have a Bioline coating. The Terumo Capiox RX oxygenator is composed of polypropylene microporous fibers for gas exchange (TeG), a stainless steel heat exchanger (TeW), and PC housing (TeH), all of which have an Xcoating. The Maquet and Capiox ECMO circuits are adult sized (Figure 1).
Circuit sample preparation.
Disassembled components of oxygenators (MaG, MaW, MaH, TeG, TeW, and TeH) were dissected into approximately 1 g sections. Because circulating blood contacts only the inside of the tubing, the tubing was cut into approximately 7 cm sections to enclose approximately 3.5 ml of drug solution by clamping both sides. All ECMO circuit components (T, B, X, MaG, MaW, MaH, TeG, TeW, and TeH) were then sterilized by ethylene oxide gas (Figure 2A).
Drug Solution Preparation
Drugs were dissolved in DI water to minimize ionic effects on the drug loss. Drug solutions with initial concentrations of 0.0075 µg/ml, 0.015 mg/ml, and 1 U/ml for dexmedetomidine (Cat. No. SML0956; Sigma-Aldrich, St Louis, MO), meropenem (Cat. No. M2574; Sigma-Aldrich, St Louis, MO), and heparin (Cat. No. H5515; Sigma-Aldrich, St Louis, MO), respectively, were prepared in a 1000 ml volume. Initial concentrations of all drugs were chosen based on the actual dose used for ECMO treatment of adults.5 Because that meropenem is soluble in methanol, we first prepared high-concentrated meropenem solutions in <1 ml methanol and then diluted with DI water to 1000 ml. A heparin concentration of 1 U/ml was used based on the amount of initial bolus injection (50–100 U/kg body) specified in the extracorporeal life support organization criteria assuming that the average weight and blood volume of an adult human male are 70 kg and 5 L, respectively.
Drug solutions with defined initial concentrations were prepared just before the experiment, and aliquots of the solution were frozen at −80°C for use as the baseline. For each drug, four groups of 11 samples (R, C, T, B, X, MaG, MaW, MaH, TeG, TeW, and TeH) were prepared for evaluation of drug loss at four time points. Each sterilized oxygenator component was placed in a 30 ml amber glass bottle, to which was added 25 ml of the drug solution. Two control groups, R and C, were also prepared by introducing 25 ml of the drug solution into 30 ml amber glass bottles that did not contain materials. All the glass bottles were covered with screw lid and tightly sealed with parafilm. Each sterilized section of tubing was filled with 3.5 ml of the drug solution by clamping both sides. Tubings were also tightly clamped to prevent drug solutions from exposure to the atmosphere. Four groups of 10 samples, excluding the control group R, were placed in a water bath maintained at 37°C with agitation at 30 RPM (Figure 2B), while the four control groups (R) remained at room temperature (22°C).
At 1, 6, 12, and 24 h, samples were withdrawn from the water bath and aliquots of the drug solutions were collected into screw-cap freeze vials and stored at −80°C before analysis. Three identical experiments were carried out for each drug.
Measurement of Drug Concentrations
Concentrations of dexmedetomidine and meropenem in samples were determined by liquid chromatography-tandem mass spectrometry, and those of heparin in samples were determined using an analytical procedure described previously.11 A UV-1800 UV-VIS spectrophotometer (Shimadzu Europa GmbH, Germany) was used to measure absorbance at 631 nm.
The amount of remaining drugs is described as a percentage relative to the baseline. Student’s two sample t-test, which is appropriate for use with small sample sizes (N ≤ 5) so long as the effect size is large (D > 0.8),12 was conducted to assess differences between the control group C and other samples at each time point using MATLAB (The MathWorks, Inc., Natick, MA) “ttest2” function. To apply variance type for the two-sample t-test, homogeneity of variances was tested using SPSS 17.0 for Windows (SPSS, Inc., Chicago, IL). Statistical analyses were performed at the 0.05 significance level.
Concentration of dexmedetomidine was significantly reduced in all three types of tubing (T, B, and X) compared with the control group C. Among the three, B exerted the greatest influence on drug adsorption and exhibited a significant 20% loss at 1 h to 30% loss at 24 h. T also showed significant loss from 12 h and X from 6 h (Figure 3). Among the oxygenator samples, MaW showed significant loss from 6 h and 42% was lost at 24 h. TeG, which is stainless steel, also showed significant loss at 24 h (Figure 3 and Table 1).
Meropenem exhibited no significant loss compared with the control group C, which also showed decreasing concentration over time (12% loss at 24 h) (Figure 3). However, the meropenem solution that remained at room temperature (R) did not exhibit such instability, suggesting temperature to be the cause of the drug loss. A previous study in ECMO circuits also demonstrated chemically unstable characteristics of meropenem in blood, showing an 80% loss in the experimental group and 58% loss in the control group, which remained at 37°C.5 Compared with the control group R, however, all three types of tubing (20% loss for T, 9% for B and X), MaW (9% loss), MaH (11% loss), and TeW (15% loss), showed significant losses at 24 h (Table 1).
No significant loss of heparin occurred for any material (Table 1). However, the concentration within MaG was greater than the control group at all time points, possibly because of the heparin-based Bioline surface coating, whereas other samples were similar to the control groups R and C (Figure 3).
Compared with meropenem and heparin, dexmedetomidine exhibited relatively early loss, as expected from a previous report that dexmedetomidine in a blood-primed ECMO circuit was absorbed, with a maximum 73% loss after 1 h.13 Unlike meropenem, dexmedetomidine and heparin were chemically stable at 37°C; therefore, both R and C showed similar results, with no significant losses (Figure 3).
Most previous studies related to drug sequestration in ECMO circuits used the entire circuit primed with either saline or blood, in which interactions among all combinations of materials composing the circuit and those of ions included in the base solution are involved. Mehta et al.7 demonstrated that drug losses and their stability can differ not only by drug type but also by the base solution. For example, 71.8% and 15.4% of ampicillin was lost in a crystalloid-primed and blood-primed circuit after 24 h of circulation, respectively. Ampicillin samples stored in glass jars for the control group, one in crystalloid and the other in blood, also exhibited different concentrations at 24 h; that is, 0% vs. 38.6%, demonstrating that complex chemical mechanisms are involved in drug loss. Accordingly, we used a base solution of DI water to minimize ionic influences and investigated concentration changes of each drug by individual components, which were chosen based on identifying the blood path in the ECMO circuit and confirmed that they would be in contact with blood during the actual ECMO treatment.
Besides chemical factors, drug loss in the ECMO circuit could be associated with mechanical factors, as blood in the circuit moves with varying flow rates or pressures, being continuously exposed to friction with component surfaces. The key purpose of our study is to evaluate interactions between drug and materials by minimizing other interruptions, and such dynamic conditions could certainly be one of the influential factors. Assuming that the flow condition can be possibly another influential factor affecting drug absorption, the experiment presented here was performed under relatively static environment. Systematic experiments could be important in follow-up studies by including a number of influential factors one by one (e.g., dynamic flow, different properties of temperature of blood) to quantify their individual effects in drug sequestration in ECMO circuit.
The greatest loss of dexmedetomidine occurred with B and MaW, all of which are from Maquet Inc. Because there was no significant difference in concentration from MaH and TeH over time, the only difference between which is the surface coating materials, the coating material might not be the cause of drug loss. Moreover, although B showed the greatest decrease in dexmedetomidine concentration, considering that there was no significant difference in the concentrations among the three tubing types, the tubing material could be the source of the cause rather than the coating material used for either Maquet or Terumo circuit. Indeed, Preston et al.14 demonstrated that uncoated polyvinyl chloride (PVC) tubing was the primary cause of fentanyl and morphine sulfate loss, and both Bioline-coated Quadrox D (Maquet Cardiopulmonary) and X-coated Terumo Baby Rx (Terumo Cardiovascular Systems Corp.) had very little effect. Therefore, PVC itself may have greater influence on drug loss than these surface coating materials. However, their follow-up study with modified surface coating demonstrated that different coatings could affect drug absorption of fentanyl and morphine,10 suggesting that coating material cannot be neglected for evaluating drug sequestration in ECMO circuit.
Dexmedetomidine is emerging as a promising sedative-analgesic agent for intensive care unit (ICU) patients.15 Dexmedetomidine has gained attention because classic sedatives including midazolam and propofol have adverse side effects such as prolonged mechanical ventilation and propofol infusion syndrome.16 Moreover, particularly for delirium which generally comes with deep sedation and is one of the most important issues in the ICU,17–19 dexmedetomidine proved to be more effective in reducing such mental disability than either midazolam or propofol.16,20 Accordingly, the pharmacokinetic characteristics of such novel sedative-analgesic agent in ECMO circuit are of interest,13 and our in vitro study thus provides an important contribution to related fields.
Shekar et al.5 demonstrated the instability of meropenem in a blood-primed circuit maintained at 37°C, which can be associated with both blood components and temperature. Berthoin et al.21 evaluated meropenem stability in aqueous solution and reported that meropenem degradation is dependent on time, temperature, and concentration. Concomitant with these previous studies, meropenem at 0.015 mg/ml used in our study maintained stability over time at room temperature, but degrades at physiologic temperature (37°C). Considering such degradation, the initial bolus of meropenem for clinical use should be optimized.
To quantify heparin concentration based on the colorimetric method of Smith et al.,11 a great number of pipetting procedures are necessary, which could result in technical error. To minimize such error, we used dispensers (VITLAB GmbH, Grossostheim, Germany) and Eppendorf Multipette M4 (Eppendorf, Hamburg, Germany), which improved reproducibility compared with preliminary experiments. Interestingly, discovery of the heparin concentration increased with MaG, which was also seen in the preliminary experiments in which experimental procedures were exactly the same but both dispensers and Eppendorf Mutipette (N = 3, data not shown) were not used. According to Maquet Inc., Bioline is an albumin-heparin coating in which heparin is covalently bonded to albumin immobilized on the surface. The elevated heparin concentration level within MaG might be associated with the bonding strength between the coating material and the PMP hollow fibers of MaG.
Because of cost and limited amount of materials available from a single circuit, we used a small sample size (N = 3). Although the larger the sample size, the greater the statistical power; Student’s t-test facilitates statistical analysis of small sample sizes (N < 5), as confirmed by de Winter.12
We quantitated drug loss because of individual components of the ECMO circuit, of which PVC tubing and MaW were critical for dexmedetomidine, while no significant losses were found for meropenem and heparin. Considering that there was no significant difference in drug loss because of MaH and TeH, both of which are made of PC, or of PVC-based tubing, the only difference in which is the coating material, coating materials may have marginal effects on drug loss. Instead, the materials composing tubing and MaW, PVC and PU fibers, respectively, are likely the primary cause of drug loss, suggesting that material selection is vital in ECMO development. Investigations of chemical responses depending on drug properties to develop an ECMO-specific drug dosing protocol and to resolve the altered pharmacokinetics are important, but use of different materials could also minimize pharmacokinetic changes during ECMO treatment.
1. Guttendorf J, Boujoukos AJ, Ren D, Rosenzweig MQ, Hravnak M. Discharge outcome in adults treated with extracorporeal membrane oxygenation
. Am J Crit Care 2014.23: 365–377.
2. Kays DW, Islam S, Richards DS, Larson SD, Perkins JM, Talbert JL. Extracorporeal life support in patients with congenital diaphragmatic hernia: How long should we treat? J Am Coll Surg 2014.218: 808–817.
3. Shekar K, Fraser JF, Smith MT, Roberts JA. Pharmacokinetic changes in patients receiving extracorporeal membrane oxygenation
. J Crit Care 2012.27: 741.e9–741.18.
4. Buck ML. Pharmacokinetic changes during extracorporeal membrane oxygenation
: Implications for drug therapy of neonates. Clin Pharmacokinet 2003.42: 403–417.
5. Shekar K, Roberts JA, Mcdonald CI, et al. Sequestration of drugs in the circuit may lead to therapeutic failure during extracorporeal membrane oxygenation
. Crit Care 2012.16: R194.
6. Shekar K, Roberts JA, Mcdonald CI, et al. Protein-bound drugs are prone to sequestration in the extracorporeal membrane oxygenation
circuit: Results from an ex vivo study. Crit Care 2015.19: 164.
7. Mehta NM, Halwick DR, Dodson BL, Thompson JE, Arnold JH. Potential drug sequestration
during extracorporeal membrane oxygenation
: Results from an ex vivo experiment. Intensive Care Med 2007.33: 1018–1024.
8. Wildschut ED, Ahsman MJ, Allegaert K, Mathot RAA, Tibboel D. Determinants of drug absorption in different ECMO circuits. Intensive Care Med 2010.36: 2109–2116.
9. Shekar K, Roberts JA, Smith MT, Fung YL, Fraser JF. The ECMO PK Project: An incremental research approach to advance understanding of the pharmacokinetic alterations and improve patient outcomes during extracorporeal membrane oxygenation
. BMC Anesthesiol 2013.13: 7.
10. Preston TJ, Ratliff TM, Gomez D, et al. Modified surface coatings and their effect on drug adsorption within the extracorporeal life support circuit. J Extra Corpor Technol 2010.42: 199–202.
11. Smith PK, Mallia AK, Hermanson GT. Colorimetric method for the assay of heparin content in immobilized heparin preparations. Anal Biochem 1980.109: 466–473.
12. de Winter JC. Using the Student’s t-test with extremely small sample sizes. Pract Assess Res Eval 2013.18: 1–12.
13. Wagner D, Pasko D, Phillips K, Waldvogel J, Annich G. In vitro
clearance of dexmedetomidine in extracorporeal membrane oxygenation
. Perfusion 2013.28: 40–46.
14. Preston TJ, Hodge AB, Riley JB, Leib-Sargel C, Nicol KK. In vitro
drug adsorption and plasma free hemoglobin levels associated with hollow fiber oxygenators in the extracorporeal life support (ECLS) circuit. J Extra Corpor Technol 2007.39: 234–237.
15. Wujtewicz M, Maciejewski D, Misiołek H, et al. Use of dexmedetomidine in the adult intensive care unit. Anaesthesiol Intensive Ther 2013.45: 235–240.
16. Jakob SM, Ruokonen E, Grounds RM, et al; Dexmedetomidine for Long-Term Sedation Investigators: Dexmedetomidine vs midazolam or propofol for sedation during prolonged mechanical ventilation: Two randomized controlled trials. JAMA 2012.307: 1151–1160.
17. Ouimet S, Kavanagh BP, Gottfried S, Skrobik Y. Incidence, risk factors and consequences of ICU delirium. Intensive Care Med 2007.33: 66–73.
18. Patel RP, Gambrell M, Speroff T, et al. Delirium and sedation in the intensive care unit: Survey of behaviors and attitudes of 1384 healthcare professionals. Crit Care Med 2009.37: 825–832.
19. Shehabi Y, Botha JA, Boyle MS, et al. Sedation and delirium in the intensive care unit: An Australian and New Zealand perspective. Anaesth Intensive Care 2008.36: 570–578.
20. Riker RR, Shehabi Y, Bokesch PM, et al; SEDCOM (Safety and Efficacy of Dexmedetomidine Compared With Midazolam) Study Group: Dexmedetomidine vs midazolam for sedation of critically ill patients: A randomized trial. JAMA 2009.301: 489–499.
21. Berthoin K, Le Duff CS, Marchand-Brynaert J, Carryn S, Tulkens PM. Stability of meropenem and doripenem solutions for administration by continuous infusion. J Antimicrob Chemother 2010.65: 1073–1075.
Keywords:Copyright © 2017 by the American Society for Artificial Internal Organs
extracorporeal membrane oxygenation; circuit constituents; drug sequestration; material selection