- Email: [email protected]
Randomized Trial on the Effect of Sevoflurane on Polypropylene Membrane Oxygenator Performance
Randomized Trial on the Effect of Sevoflurane on Polypropylene Membrane Oxygenator Performance Caetano Nigro Neto, MD, PhD,* Renato Arnoni, MD, PhD,* Bilal Smaili Rida, MD,* Giovanni Landoni, MD,†,z and Maria Angela Tardelli, MD, PhD y Objectives: Volatile anesthetics have cardioprotective properties that improve clinically relevant outcomes in cardiac surgery, and can be used during cardiopulmonary bypass (CPB) through adapted calibrated vaporizers together with air and oxygen (O2). The effect of volatile agents on the membrane oxygenator is unknown. The aim of this study was to evaluate, for the first time, the performance of semiporous polypropylene membrane oxygenators after the use of sevoflurane vaporized during CPB in cardiac surgery. Design: A prospective, randomized, controlled trial. Setting: Teaching hospital. Participants: Thirty-two consecutive patients scheduled to undergo coronary artery bypass graft with CPB. Interventions: Patients were allocated randomly to receive either a volatile anesthetic (sevoflurane 1%-3%, 16 patients) or an intravenous hypnotic (midazolam, 16
patients) during CPB. After surgery, the membrane oxygenators used during CPB were tested with regard to O2 transfer, carbon dioxide transfer, and pressure drop. Measurements and Main Results: The authors observed no protocol deviation or crossover. The performance of the membrane oxygenator was similar between the 2 groups, as documented by O2 transfer (55 ⫾ 6.4 mL/min/L in the sevoflurane group versus 57 ⫾ 4.7 mL/min/L in the midazolam group, p ¼ 0.4), carbon dioxide transfer, and pressure drop. Conclusions: The use of sevoflurane during CPB in cardiac surgery does not affect membrane oxygenator performance. & 2013 Elsevier Inc. All rights reserved.
V
body mass index above 32 kg/m2, and New York Heart Association class III or IV. Patients were allocated randomly to receive either sevoflurane— expired fraction 1% to 3% (sevoflurane group—16 patients), or midazolam—0.05 mg/kg boluses (control group—16 patients), as the main hypnotic drug during CPB. Randomization took place using sealed, unmarked, opaque envelopes that were allocated to participants in an area separate from the research and treatment rooms. Anesthesiologists and perfusionists were not blinded to the anesthetic regimen. All the other staff members, including the clinicians in the intensive care unit (ICU) and the engineers in the laboratory who managed the membrane oxygenator after surgery, were blinded to the anesthetic regimen. Patients received their usual medications until the day of surgery, with the exception of agonists and antagonists of ATP-dependent potassium channels (glibenclamide, theophiline, allopurinol). According to a local standard protocol, designed to enhance the cardioprotective properties of volatile agents, these drugs were discontinued; with the only exception being glibenclamide, which was replaced by insulin at least 3 days before surgery. Furthermore, aspirin was interrupted 2 days before surgery in all patients.
OLATILE ANESTHETICS are used widely in cardiac anesthesia and have cardioprotective effects, resulting in reduced cardiac troponin release when compared to total intravenous anesthesia.1,2 This effect may translate into an improved survival rate.1,3-6 Sevoflurane is used during cardiopulmonary bypass (CPB) because it can be delivered from adapted calibrated vaporizers into the membrane oxygenator together with air and oxygen (O2).7 Currently, there are 2 groups of hollow-fiber membrane oxygenators used in clinical practice: diffusion and hollowfiber polypropylene (PPL) membranes.8 Diffusion plasmaresistant oxygenators, the basic membrane compound of which is poly-(4-methyl-1-pentene), have been used increasingly for extracorporeal life support or extracorporeal membrane oxygenation. This type of membrane oxygenator can be used for more than 6 hours, but increases the risk of intraoperative awareness during CPB when the volatile agent used is isoflurane.9-11 The hollow-fiber membranes, the basic compound of which is semi-porous PPL, are the standard oxygenators used for cardiac surgery with CPB. They offer excellent O2 exchange and carbon dioxide (CO2) removal up to 6 hours. They are very fragile, and their performance has never been formally tested in the presence of volatile agents.8 The aim of this study was to evaluate, for the first time, the performance of a semi-porous PPL membrane oxygenator after the use of sevoflurane during CPB in cardiac surgery.
METHODS After approval by the local institutional ethics committee and after obtaining written informed consent, 32 consecutive patients scheduled to undergo coronary artery bypass grafting with CPB were included in this study. Exclusion criteria were emergency procedures, reoperations,
KEY WORDS: cardiopulmonary bypass, anesthesia, drugs, evaluation, membrane oxygenator, gas transfer, oxygen, sevoflurane
From the *Dante Pazzanese Institute of Cardiology, Brazil; yAnesthesia and Intensive Care Department, San Raffaele Scientific Institute, Italy; zMaieutics Foundation, Milan, Italy; and yAnesthesia and Intensive Care Department, Federal University of Sao Paulo, Brazil. This study was supported entirely by departmental funds. Sevoflurane (Sevorane, Abbott Laboratories, Argentina) was provided free of charge by the distributor (Abbott Brazil). The laboratory analyses of all membrane oxygenators were provided free of charge by Nipro Medical LTDA, Sorocaba, SP, Brazil. Address reprint requests to Caetano Nigro Neto, MD, Dante Pazzanese Institute of Cardiology and Federal University of S~ao Paulo (UNIFESP), Rua Martiniano de Carvalho, N1 864 – 6a andar – Cj. 603/604 – Bela Vista, S~ao Paulo – SP – Brasil. E-mail: [email protected] & 2013 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2012.12.012
Journal of Cardiothoracic and Vascular Anesthesia, Vol 27, No 5 (October), 2013: pp 903–907
903
904
All patients were premedicated with midazolam (7.5 mg intramuscular) 30 minutes before the induction of anesthesia. Patients received spinal analgesia with 5 mL of saline solution, 0.9%, containing 1 mg/kg of sufentanil, through a 25-gauge needle. Monitoring included bispectral index (BIS) positioned before induction of general anesthesia. General anesthesia was induced in all patients with sevoflurane in O2 by mask, 0.5 mg/kg of sufentanil, and 0.1 mg/kg of pancuronium. Maintenance in the periods before and after CPB was performed with sevoflurane (expired fraction 1% to 3%) to maintain the BIS range between 40 and 65. Sufentanil bolus (0.5 mg/ kg) was administered to maintain mean arterial pressure levels between 60 and 80 mmHg. The only difference between groups was that sevoflurane was given to 1 group only during CPB. Both groups could receive midazolam. During CPB, patients from the sevoflurane group received sevoflurane, 1% to 3% (Abbott Laboratories Argentina), through a calibrated vaporizer with a mixture of O2 50% and compressed air in the oxygenator circuit to maintain the BIS value between 40 and 65. When the BIS value was above these limits, patients received an additional dose of midazolam (0.05 mg/kg boluses). In the control group, all patients received only midazolam (0.05 mg/kg boluses) to maintain the BIS value between 40 and 65. Hypertension (mean arterial pressure 470 mmHg) was treated first with intravenous sufentanil (0.5 mg/kg, maximum 2 boluses) and then with nitroprusside. Aortic clamping during the surgical procedure produced ischemia with a maximum duration of 15 minutes and an episode interval of at least 2 minutes. The circuit was primed with 1,000 mL of saline solution, 200 mL of mannitol 20%, 50 mL of sodium bicarbonate 8.4%, 3 g of magnesium sulfate, and 10,000 IU of heparin. A BPX-80 Bio-Pump Plus centrifugal blood pump (Medtronic Perfusion Systems, Minneapolis, MN) was used in all procedures. The membrane oxygenator, Vital, a tubing screen pack, and 1 venous reservoir (Nipro, Osaka, Japan) were used in all procedures. Continuous blood flow during CPB was kept at 50 mL/min/ m2. The arterial line pressures in the CPB apparatus were kept between 110 and 120 mmHg (maximum). Anticoagulation was carried out initially with heparin (350 U/kg), and additional doses were administered to keep the activated coagulation time above 500 seconds during the entire CPB time. After surgery, patients were transferred to the ICU without sedation. Extubation and ICU discharge were performed according to standard local protocols. The primary endpoint was O2 gas transfer across the membrane oxygenator. The second and third endpoints were CO2 gas transfer and pressure drop across the membrane oxygenator. The performance of the membrane oxygenator was measured in the laboratories of Nipro Medical LTDA, Sorocaba, Sao Paulo, Brazil and included O2 transfer, CO2 transfer, and pressure drop (DP). The procedure is shown in Fig 1 and described in detail as follows. After surgery, all membrane oxygenators used during CPB were cleaned up with water, sterilized with 2% peracetic acid solution, and tested using an in vitro oxygenator-deoxygenator single-pass circuit. The circuit had the following 2 interconnected flow paths: an arterial flow path to move the conditioned venous blood into the test oxygenator at preselected blood flow rates, and a venous condition flow path to establish the standard inlet conditions of the venous blood prior to entry in the test oxygenator. Inlet conditions were achieved with the aid of a deoxygenator connected to the venous flow path and were the same as those established by the Association for the Advancement of Medical Instrumentation (AAMI) for standardized testing (American National Standards Institute(ANSI)/AAMI/International Organization for Standardization(ISO) 7199:2009/A1:2012—Cardiovascular implants and artificial organs—Blood-gas exchangers. Amendment1) with oxyhemoglobin saturation ¼ 65 ⫾ 5%, hemoglobin content ¼ 12 g/dL,
NETO ET AL
pCO2 ¼ 45 ⫾ 5 mmHg, and base excess ¼ 0 ⫾ 5 meq/L. For these tests, a pool of bovine fresh blood was used, the anticoagulation was carried out with ETDA (Ethylenediaminetetraacetic acid) 15%, and the inlet condition values were measured systematically and kept constant throughout the test. Two roller pumps were used to pump the blood through the flow paths of the test circuit. Two large 20-L polycarbonate reservoirs were used to hold the conditioned blood and to receive the oxygenated blood pumped through the membrane oxygenator. The %CO2 in the outlet gas from the oxygenator was determined with an appropriate CO2 analyzer. The inlet and outlet pressures of the oxygenator were monitored with appropriately calibrated transducers connected to a pressure monitor. The blood was maintained at 37 ⫾ 11C throughout the test by supplying water to the deoxygenator heat exchangers from a heat-cooler, and blood temperature was monitored with probes. Blood sampling ports were used to collect blood gases and were placed directly before the inlet and directly after the outlet of the oxygenator. Once the AAMI standards were verified, the conditioned blood was pumped through the membrane oxygenator, with the predetermined combinations of test variables (blood flow rate ¼ 6 L/min; FIO2 ¼ 100%; gas-to-blood flow rates ¼ 1) for 6 hours. In this situation, the normal values of oxygenator gas transfer for the Vital O2 membrane are O2 transfer 450 mL/min/L and CO2 transfer 442 mL/min/L. Likewise, the normal pressure drop ranges from a minimum of 160 mmHg to a maximum of 250 mmHg. At the end of the test, exhausted gas %CO2 was noted, and 3 blood samples were drawn from the venous inlet and arterial outlet of the tested oxygenator and immediately analyzed for hemoglobin content, arterial and venous hemoglobin saturation, and arterial and venous partial pressure of O2 (PaO2). O2 gas transfer was calculated by a mean of the 3 samples, as follows: O2 transfer (mL/min) ¼ ([{SaO2SvO2} hgb 1.34]þ [{PaO2PvO2} 0.003]) Qb 10, where SaO2 ¼ arterial hemoglobin saturation (decimal form), SvO2 ¼ venous hemoglobin saturation (decimal form), hgb ¼ hemoglobin contend (g/dL), PaO2 ¼ arterial partial pressure of O2 (mmHg), PvO2 ¼ venous partial pressure of O2 (mmHg), Qb ¼ blood flow rate (L/min), 1.34 ¼ constant (mL O2/ g hgb), 0.003 ¼ constant (mL O2/dL) O, and 10 ¼ conversion factor to change mL O2/dL to mL O2/L. CO2 gas transfer was calculated by a mean of the 3 samples, as follows: CO2 transfer (mL/min) ¼ (Qg 1,000) %CO2, where Qg ¼ gas flow rate (L/min) and %CO2 ¼ exhaust %CO2 (decimal form). Pressure drop across the membrane oxygenator (between the membrane and blood flow) were calculated during gas exchange evaluation by a mean of 3 measures, as follows: DP ¼ PiPo, where DP ¼ pressure drop (mmHg), Pi ¼ inlet pressure (mmHg), and Po ¼ outlet pressure (mmHg). The standard deviation (SD) of the primary end-point (O2 transfer) was 5.0 in the study by Griffith et al (the authors had a mean ⫾ SD of 64 ⫾ 5.0 mL/min/L in their experience).12 Therefore, the present authors calculated that a sample size of 14 subjects per group was required to detect a difference of 6 mL/min between groups with a 2-sided, 0.05 significance level test with 90% statistical power, and they included 16 patients per group to account for a 15% dropout rate because of possible protocol deviations. All data were analyzed according to the intention-to-treat principle. They were stored electronically and analyzed with STATA software, version 10.0 (TX). All data analysis was performed according to a pre-established analysis plan. Dichotomous data were compared using a 2-sided test with the Yates correction or Fisher exact test when appropriate; 95% confidence interval estimation for the differences between independent proportions was performed with
SEVOFLURANE ON THE POLYPROPYLENE MEMBRANE OXYGENATOR PERFORMANCE
Fig 1.
905
Diagram of the in vitro oxygenator-deoxygenator single-pass circuit.
methods based on the Wilson score. Continuous measures were compared by Student t-test or the Mann-Whitney U test when appropriate; 95% confidence interval estimate for the mean/median difference was performed. Two-sided significance tests were used throughout. A p value o0.05 was considered significant. Descriptive data are shown as median and interquartile range, mean ⫾ SD, or number and percentage. RESULTS
No protocol deviation and no crossovers were observed. All 32 oxygenators were sent to the laboratory for testing (Fig 2). There were no differences between the groups with respect to patient characteristics and prerandomization data, while the sevoflurane group required less midazolam and less nitroprusside than the control group during CPB (Table 1). The performance of the membrane oxygenator was similar between the 2 groups at 6 L/min flow and FIO2 ¼ 1.0, as documented by O2 transfer (55 ⫾ 6.4 mL/min/L in the sevoflurane group v 57 ⫾ 4.7 mL/min/L in the control group), CO2 transfer, and pressure drop (Table 2). DISCUSSION
This is the first study demonstrating that sevoflurane does not affect the performance of a semiporous PPL hollow-fiber membrane oxygenator and can be used during CPB to enhance cardiac protection in cardiac surgery. It should be remembered that there are 2 groups of membrane oxygenators and that the authors only tested the hollow-fiber membranes made of semiporous PPL, because they are used widely for standard CPB, offering excellent O2 exchange and CO2 removal for up to 6 hours.8 The other types of membrane oxygenators have not been tested in this study. They include diffusion and plasma-resistant oxygenators,
which are used increasingly for extracorporeal life support or extracorporeal membrane oxygenation in several clinical settings; for example, in patients who can no longer be supported by mechanical ventilation. There are some basic objectives in the design of a membrane oxygenator. The principal aim is to provide optimal gas exchange (O2 and CO2 transfer) during CPB. Many factors influence the performance of membrane oxygenators during cardiac surgery. Among these, patient characteristics and surgery-related events are the most common reasons for possible complications.7,13 In the present randomized study, patient characteristics, type of cardioprotection, duration of aortic cross-clamp, and CPB time were similar in the 2 groups, and, therefore, the authors demonstrated that the 2 anesthetic strategies have similar effects on the membrane oxygenator. Knowing that the performance of the oxygenator is not affected by volatile agents during CPB allows the anesthesiologist to use and titrate volatile anesthetics more accurately.14,15 R¨odig et al16 described the effects of sevoflurane in different concentrations on systemic vascular resistance during CPB. They found that an inspiratory concentration of 3% sevoflurane significantly reduced systemic vascular resistance index when compared with baseline values. In the present study, the authors confirmed that relatively high concentrations of sevoflurane (up to 3%) can be used during CPB to maintain anesthesia and also to control blood pressure while using less antihypertensive drugs (nitroprusside) and without using high doses of intravenous hypnotics (midazolam). Indeed, the authors used nitroprusside in 9 patients (56%) in the sevoflurane group and in 15 patients (94%) in the control group. Knowledge and predictability of oxygenator performance are crucial for effective conduct of CPB.8 The ability to adequately oxygenate a patient on CPB depends on various physiologic
906
NETO ET AL
32 patients
Sevoflurane Group(n=16)
Control Group(n=16)
16 membrane
16 membrane
oxygenators
oxygenators
Fig 2. Study flow chart.
factors (hemoglobin concentration, the percentage of hemoglobin saturated with O2 in the arterial blood, cardiac output, etc) and on the physical status of the membrane oxygenator.8 With respect to oxygenation, the function of an artificial lung is to transfer at least 4.5 mL O2/100 mL of blood at its maximum rated blood flow in accordance with AAMI/ISO standards. In the present study, O2 transfer testing results were in line with the manufacturer’s recommended normal values in both groups. This is an extremely important finding because volatile agents may improve survival in cardiac surgery1,3-6 and their effect seems to be increased if used during CPB.4,17 Notably, to date there are still no CPB machines manufactured, equipped, and ready to permit the use of halogenated agents, and each cardiac surgical center must find the best way to connect the vaporizer to the oxygenator. This is surprising because volatile agents are the only anesthetic drugs that might have an influence on survival, as suggested by the first international consensus conference6 on this topic.
Determination of the oxygenator performance in the laboratory was carried out under a strict set of conditions established by the AAMI. The authors did not study the pharmacokinetics of volatile agents during CPB, but they are aware that other authors11 have described the pharmacokinetics of these drugs with different membrane oxygenators and evaluated the elimination of volatile agents from blood. They demonstrated that PPL oxygenators had better uptake and elimination of isoflurane compared to polydimethylsiloxane oxygenators. All the studies mentioned 9-11 and the present findings suggest that semiporous PPL oxygenators (the standard oxygenator for cardiac surgery with CPB) offer excellent performance in the presence of volatile agents. In contrast, previous findings indicated that the diffusion oxygenators used for extracorporeal life support or extracorporeal membrane oxygenation increased the risk of awareness during CPB when isoflurane was used.9-11 Moreover, liquid sevoflurane, enflurane, and isoflurane were described as not causing damage to the PPL fibers of the membrane oxygenators according to light-microscopic examination, even after saturating concentrations for 3 hours.18 Limitations of the study: one study12 suggested that comparison between laboratory and clinical data are difficult, if not impossible, without appropriate statistical or computed modeling, suggesting a limited value of performance tests in the clinical setting. The perfusionist generally operates outside this standardized set of parameters and it is not common practice to measure the amount of O2 and CO2 transferred across the artificial lung after 6 hours of CPB. However, this is the gold standard to refer to, and there are situations during CPB when the determination of oxygenator gas exchange is critical and perhaps life saving. In the case of oxygenator dysfunction, for example, it is vital to distinguish between primary oxygenator dysfunction and a hypermetabolic state of the patient.8 The authors acknowledge that they tested only one type of membrane oxygenator and one inhalation agent, the reason being
Table 1. Demographic Data and Necessity of Drugs During Cardiopulmonary Bypass Sevoflurane Group (n ¼ 16)
Control Group (n ¼ 16)
Preoperative Data Sex (female), n 4 (25%) 6 (37%) Age, y 60 (57-62) 59.5 (53-69) ASA II, n 3 (19%) 2 (12%) ASA III, n 13 (81%) 14 (87%) Smoking, n 13 (81%) 9 (56%) Obesity, n 7 (44%) 8 (50%) Hypertension, n 15 (94%) 16 (100%) Diabetes mellitus, n 10 (62%) 7 (44%) Dyslipidemia, n 16 (100%) 15 (94%) Intraoperative Data Anesthesia duration, min 315 (300-341) 292 (258-313) Surgery duration, min 287 (280-315) 275 (242-300) Cross-clamping, min 55 (46-64) 63 (40-90) CPB duration, min 85 (72-101) 70 (59-91) Patients needing additional intravenous anesthetics or vasodilators during cardiopulmonary bypass Nitroprusside, n 9 (56%) 15 (94%) Midazolam bolus, n 1 (6%) 16 (100%) NOTE. Continuous data are expressed as median (IQR). Count data are shown as number and percentage. Abbreviations: ASA, American Society of Anesthesiologists; IQR, interquartile range.
p
0.4 0.4 0.9 0.9 0.1 0.7 0.9 0.3 0.9 0.09 0.17 0.4 0.07 0.03 o0.001
907
SEVOFLURANE ON THE POLYPROPYLENE MEMBRANE OXYGENATOR PERFORMANCE
CONCLUSION
Table 2. Performance of the Membrane Oxygenators
O2 Transfer (mL/min/L) Normal values: 450 mL/min/L CO2 Transfer (mL/min/L) Normal values: 442 mL/min/L Pressure Drop (mmHg) Normal range values: 4160 mmHg o250 mmHg
Sevoflurane
Control Group
Group (n ¼ 16)
(n ¼ 16)
p
55 ⫾ 6.4
57 ⫾ 4.7
0.4
54 ⫾ 10.0
53 ⫾ 5.0
0.8
184 ⫾ 9.1
196 ⫾ 25.0
The authors demonstrated, for the first time, that the use of sevoflurane as the main hypnotic drug during CPB for cardiac surgery did not affect the performance of the semi-porous PPL hollow-fiber membrane oxygenator. This could lead to a more widespread use of volatile agents during CPB to enhance organ protection and improve clinically relevant outcomes in patients undergoing cardiac surgery. ACKNOWLEDGMENTS
NOTE. Data are expressed as mean ⫾ SD.
that they are used most commonly in the authors’ institution and that the Vital membrane oxygenator was the only company to carry out the performance tests without any cost.
The authors thank all surgeons who helped and contributed to the study. Also, they are indebted to all the perfusionists who helped provide care to these patients during cardiopulmonary bypass.
REFERENCES 1. Landoni G, Biondi-Zoccai GG, Zangrillo A, et al: Desflurane and sevoflurane in cardiac surgery: A meta-analysis of randomized clinical trials. J Cardiothorac Vasc Anesth 21:502-511, 2007 2. Belhomme D, Peynet J, Louzy M, et al: Evidence for preconditioning by isoflurane in coronary artery bypass graft surgery. Circulation 100:II-340-II-344, 1999 (suppl II) 3. Jakobsen CJ, Berg H, Hindsholm KB, et al: The influence of propofol versus sevoflurane anesthesia on outcome in 10,535 cardiac surgical procedures. J Cardiothorac Vasc Anesth 21:664-671, 2007 4. Bignami E, Biondi-Zoccai G, Landoni G, et al: Volatile anesthetics reduce mortality in cardiac surgery. J Cardiothorac Vasc Anesth 23:594-599, 2009 5. De Hert S, Vlasselaers D, Barbe´ R, et al: A comparison of volatile and nonvolatile agents for cardioprotection during on-pump coronary surgery. Anaesthesia 64:953-960, 2009 6. Landoni G, Rodseth RN, Santini F, et al: Randomized evidence for reduction of perioperative mortality. J Cardiothorac Vasc Anesth 26:764-772, 2012 7. Nigro Neto C, Tardelli MA, Paulista PH: Use of volatile anesthetics in extracorporeal circulation. Rev Bras Anestesiol 62:346-355, 2012 8. Gleen PG: Cardiopulmonary Bypass Principles and Practice, (ed 3). Philadelphia: Lippincott Williams and Wilkins; 2008 pp 47-49, 2008 9. Philipp A, Wiesenack C, Behr R, et al: High risk of intraoperative awareness during cardiopulmonary bypass with isoflurane administration via diffusion membrane oxygenators. Perfusion 17:175-178, 2002
10. Wiesenack C, Wiesner G, Keyl C, et al: In vivo uptake and elimination of isoflurane by different membrane oxygenators during cardiopulmonary bypass. Anesthesiology 97:133-138, 2002 11. Hickey S, Gaylor JD, Kenny GN: In vitro uptake and elimination of isoflurane by different membrane oxygenators. J Cardiothorac Vasc Anesth 10:352-355, 1996 12. Griffith KE, Vasquez MR, Beckley PD, et al: Predicting oxygenator clinical performance from laboratory in-vitro testing. J Extra Corpor Technol 26:114-120, 1994 13. Berdajs DA, de Stefano E, Delay D, et al: The new advanced membrane gas exchanger. Interact Cardiovasc Thorac Surg 13:591-596, 2011 14. Blokker-Veldhuis MJ, Rutten PM, De Hert SG: Occupational exposure to sevoflurane during cardiopulmonary bypass. Perfusion 26:383-389, 2011 15. Reinsfelt B, Westerlind A, Ricksten SE: The effects of sevoflurane on cerebral blood flow autoregulation and flow-metabolism coupling during cardiopulmonary bypass. Acta Anaesthesiol Scand 55:118-123, 2011 16. R¨odig G, Keyl C, Wiesner G, et al: Effects of sevoflurane and isoflurane on systemic vascular resistance: Use of cardiopulmonary bypass as a study model. Br J Anaesth 76:9-12, 1996 17. De Hert SG, Van der Linden PJ, Cromheecke S, et al: Cardioprotective properties of sevoflurane in patients undergoing coronary surgery with cardiopulmonary bypass are related to the modalities of its administration. Anesthesiology 101:299-310, 2004 18. Crosbie AE, Vuylsteke A, Latimer RD: Inhalation anaesthetics and the Medtronic Maxima Plus membrane oxygenator. Br J Anaesth 80:878, 1998
patients) during CPB. After surgery, the membrane oxygenators used during CPB were tested with regard to O2 transfer, carbon dioxide transfer, and pressure drop. Measurements and Main Results: The authors observed no protocol deviation or crossover. The performance of the membrane oxygenator was similar between the 2 groups, as documented by O2 transfer (55 ⫾ 6.4 mL/min/L in the sevoflurane group versus 57 ⫾ 4.7 mL/min/L in the midazolam group, p ¼ 0.4), carbon dioxide transfer, and pressure drop. Conclusions: The use of sevoflurane during CPB in cardiac surgery does not affect membrane oxygenator performance. & 2013 Elsevier Inc. All rights reserved.
V
body mass index above 32 kg/m2, and New York Heart Association class III or IV. Patients were allocated randomly to receive either sevoflurane— expired fraction 1% to 3% (sevoflurane group—16 patients), or midazolam—0.05 mg/kg boluses (control group—16 patients), as the main hypnotic drug during CPB. Randomization took place using sealed, unmarked, opaque envelopes that were allocated to participants in an area separate from the research and treatment rooms. Anesthesiologists and perfusionists were not blinded to the anesthetic regimen. All the other staff members, including the clinicians in the intensive care unit (ICU) and the engineers in the laboratory who managed the membrane oxygenator after surgery, were blinded to the anesthetic regimen. Patients received their usual medications until the day of surgery, with the exception of agonists and antagonists of ATP-dependent potassium channels (glibenclamide, theophiline, allopurinol). According to a local standard protocol, designed to enhance the cardioprotective properties of volatile agents, these drugs were discontinued; with the only exception being glibenclamide, which was replaced by insulin at least 3 days before surgery. Furthermore, aspirin was interrupted 2 days before surgery in all patients.
OLATILE ANESTHETICS are used widely in cardiac anesthesia and have cardioprotective effects, resulting in reduced cardiac troponin release when compared to total intravenous anesthesia.1,2 This effect may translate into an improved survival rate.1,3-6 Sevoflurane is used during cardiopulmonary bypass (CPB) because it can be delivered from adapted calibrated vaporizers into the membrane oxygenator together with air and oxygen (O2).7 Currently, there are 2 groups of hollow-fiber membrane oxygenators used in clinical practice: diffusion and hollowfiber polypropylene (PPL) membranes.8 Diffusion plasmaresistant oxygenators, the basic membrane compound of which is poly-(4-methyl-1-pentene), have been used increasingly for extracorporeal life support or extracorporeal membrane oxygenation. This type of membrane oxygenator can be used for more than 6 hours, but increases the risk of intraoperative awareness during CPB when the volatile agent used is isoflurane.9-11 The hollow-fiber membranes, the basic compound of which is semi-porous PPL, are the standard oxygenators used for cardiac surgery with CPB. They offer excellent O2 exchange and carbon dioxide (CO2) removal up to 6 hours. They are very fragile, and their performance has never been formally tested in the presence of volatile agents.8 The aim of this study was to evaluate, for the first time, the performance of a semi-porous PPL membrane oxygenator after the use of sevoflurane during CPB in cardiac surgery.
METHODS After approval by the local institutional ethics committee and after obtaining written informed consent, 32 consecutive patients scheduled to undergo coronary artery bypass grafting with CPB were included in this study. Exclusion criteria were emergency procedures, reoperations,
KEY WORDS: cardiopulmonary bypass, anesthesia, drugs, evaluation, membrane oxygenator, gas transfer, oxygen, sevoflurane
From the *Dante Pazzanese Institute of Cardiology, Brazil; yAnesthesia and Intensive Care Department, San Raffaele Scientific Institute, Italy; zMaieutics Foundation, Milan, Italy; and yAnesthesia and Intensive Care Department, Federal University of Sao Paulo, Brazil. This study was supported entirely by departmental funds. Sevoflurane (Sevorane, Abbott Laboratories, Argentina) was provided free of charge by the distributor (Abbott Brazil). The laboratory analyses of all membrane oxygenators were provided free of charge by Nipro Medical LTDA, Sorocaba, SP, Brazil. Address reprint requests to Caetano Nigro Neto, MD, Dante Pazzanese Institute of Cardiology and Federal University of S~ao Paulo (UNIFESP), Rua Martiniano de Carvalho, N1 864 – 6a andar – Cj. 603/604 – Bela Vista, S~ao Paulo – SP – Brasil. E-mail: [email protected] & 2013 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2012.12.012
Journal of Cardiothoracic and Vascular Anesthesia, Vol 27, No 5 (October), 2013: pp 903–907
903
904
All patients were premedicated with midazolam (7.5 mg intramuscular) 30 minutes before the induction of anesthesia. Patients received spinal analgesia with 5 mL of saline solution, 0.9%, containing 1 mg/kg of sufentanil, through a 25-gauge needle. Monitoring included bispectral index (BIS) positioned before induction of general anesthesia. General anesthesia was induced in all patients with sevoflurane in O2 by mask, 0.5 mg/kg of sufentanil, and 0.1 mg/kg of pancuronium. Maintenance in the periods before and after CPB was performed with sevoflurane (expired fraction 1% to 3%) to maintain the BIS range between 40 and 65. Sufentanil bolus (0.5 mg/ kg) was administered to maintain mean arterial pressure levels between 60 and 80 mmHg. The only difference between groups was that sevoflurane was given to 1 group only during CPB. Both groups could receive midazolam. During CPB, patients from the sevoflurane group received sevoflurane, 1% to 3% (Abbott Laboratories Argentina), through a calibrated vaporizer with a mixture of O2 50% and compressed air in the oxygenator circuit to maintain the BIS value between 40 and 65. When the BIS value was above these limits, patients received an additional dose of midazolam (0.05 mg/kg boluses). In the control group, all patients received only midazolam (0.05 mg/kg boluses) to maintain the BIS value between 40 and 65. Hypertension (mean arterial pressure 470 mmHg) was treated first with intravenous sufentanil (0.5 mg/kg, maximum 2 boluses) and then with nitroprusside. Aortic clamping during the surgical procedure produced ischemia with a maximum duration of 15 minutes and an episode interval of at least 2 minutes. The circuit was primed with 1,000 mL of saline solution, 200 mL of mannitol 20%, 50 mL of sodium bicarbonate 8.4%, 3 g of magnesium sulfate, and 10,000 IU of heparin. A BPX-80 Bio-Pump Plus centrifugal blood pump (Medtronic Perfusion Systems, Minneapolis, MN) was used in all procedures. The membrane oxygenator, Vital, a tubing screen pack, and 1 venous reservoir (Nipro, Osaka, Japan) were used in all procedures. Continuous blood flow during CPB was kept at 50 mL/min/ m2. The arterial line pressures in the CPB apparatus were kept between 110 and 120 mmHg (maximum). Anticoagulation was carried out initially with heparin (350 U/kg), and additional doses were administered to keep the activated coagulation time above 500 seconds during the entire CPB time. After surgery, patients were transferred to the ICU without sedation. Extubation and ICU discharge were performed according to standard local protocols. The primary endpoint was O2 gas transfer across the membrane oxygenator. The second and third endpoints were CO2 gas transfer and pressure drop across the membrane oxygenator. The performance of the membrane oxygenator was measured in the laboratories of Nipro Medical LTDA, Sorocaba, Sao Paulo, Brazil and included O2 transfer, CO2 transfer, and pressure drop (DP). The procedure is shown in Fig 1 and described in detail as follows. After surgery, all membrane oxygenators used during CPB were cleaned up with water, sterilized with 2% peracetic acid solution, and tested using an in vitro oxygenator-deoxygenator single-pass circuit. The circuit had the following 2 interconnected flow paths: an arterial flow path to move the conditioned venous blood into the test oxygenator at preselected blood flow rates, and a venous condition flow path to establish the standard inlet conditions of the venous blood prior to entry in the test oxygenator. Inlet conditions were achieved with the aid of a deoxygenator connected to the venous flow path and were the same as those established by the Association for the Advancement of Medical Instrumentation (AAMI) for standardized testing (American National Standards Institute(ANSI)/AAMI/International Organization for Standardization(ISO) 7199:2009/A1:2012—Cardiovascular implants and artificial organs—Blood-gas exchangers. Amendment1) with oxyhemoglobin saturation ¼ 65 ⫾ 5%, hemoglobin content ¼ 12 g/dL,
NETO ET AL
pCO2 ¼ 45 ⫾ 5 mmHg, and base excess ¼ 0 ⫾ 5 meq/L. For these tests, a pool of bovine fresh blood was used, the anticoagulation was carried out with ETDA (Ethylenediaminetetraacetic acid) 15%, and the inlet condition values were measured systematically and kept constant throughout the test. Two roller pumps were used to pump the blood through the flow paths of the test circuit. Two large 20-L polycarbonate reservoirs were used to hold the conditioned blood and to receive the oxygenated blood pumped through the membrane oxygenator. The %CO2 in the outlet gas from the oxygenator was determined with an appropriate CO2 analyzer. The inlet and outlet pressures of the oxygenator were monitored with appropriately calibrated transducers connected to a pressure monitor. The blood was maintained at 37 ⫾ 11C throughout the test by supplying water to the deoxygenator heat exchangers from a heat-cooler, and blood temperature was monitored with probes. Blood sampling ports were used to collect blood gases and were placed directly before the inlet and directly after the outlet of the oxygenator. Once the AAMI standards were verified, the conditioned blood was pumped through the membrane oxygenator, with the predetermined combinations of test variables (blood flow rate ¼ 6 L/min; FIO2 ¼ 100%; gas-to-blood flow rates ¼ 1) for 6 hours. In this situation, the normal values of oxygenator gas transfer for the Vital O2 membrane are O2 transfer 450 mL/min/L and CO2 transfer 442 mL/min/L. Likewise, the normal pressure drop ranges from a minimum of 160 mmHg to a maximum of 250 mmHg. At the end of the test, exhausted gas %CO2 was noted, and 3 blood samples were drawn from the venous inlet and arterial outlet of the tested oxygenator and immediately analyzed for hemoglobin content, arterial and venous hemoglobin saturation, and arterial and venous partial pressure of O2 (PaO2). O2 gas transfer was calculated by a mean of the 3 samples, as follows: O2 transfer (mL/min) ¼ ([{SaO2SvO2} hgb 1.34]þ [{PaO2PvO2} 0.003]) Qb 10, where SaO2 ¼ arterial hemoglobin saturation (decimal form), SvO2 ¼ venous hemoglobin saturation (decimal form), hgb ¼ hemoglobin contend (g/dL), PaO2 ¼ arterial partial pressure of O2 (mmHg), PvO2 ¼ venous partial pressure of O2 (mmHg), Qb ¼ blood flow rate (L/min), 1.34 ¼ constant (mL O2/ g hgb), 0.003 ¼ constant (mL O2/dL) O, and 10 ¼ conversion factor to change mL O2/dL to mL O2/L. CO2 gas transfer was calculated by a mean of the 3 samples, as follows: CO2 transfer (mL/min) ¼ (Qg 1,000) %CO2, where Qg ¼ gas flow rate (L/min) and %CO2 ¼ exhaust %CO2 (decimal form). Pressure drop across the membrane oxygenator (between the membrane and blood flow) were calculated during gas exchange evaluation by a mean of 3 measures, as follows: DP ¼ PiPo, where DP ¼ pressure drop (mmHg), Pi ¼ inlet pressure (mmHg), and Po ¼ outlet pressure (mmHg). The standard deviation (SD) of the primary end-point (O2 transfer) was 5.0 in the study by Griffith et al (the authors had a mean ⫾ SD of 64 ⫾ 5.0 mL/min/L in their experience).12 Therefore, the present authors calculated that a sample size of 14 subjects per group was required to detect a difference of 6 mL/min between groups with a 2-sided, 0.05 significance level test with 90% statistical power, and they included 16 patients per group to account for a 15% dropout rate because of possible protocol deviations. All data were analyzed according to the intention-to-treat principle. They were stored electronically and analyzed with STATA software, version 10.0 (TX). All data analysis was performed according to a pre-established analysis plan. Dichotomous data were compared using a 2-sided test with the Yates correction or Fisher exact test when appropriate; 95% confidence interval estimation for the differences between independent proportions was performed with
SEVOFLURANE ON THE POLYPROPYLENE MEMBRANE OXYGENATOR PERFORMANCE
Fig 1.
905
Diagram of the in vitro oxygenator-deoxygenator single-pass circuit.
methods based on the Wilson score. Continuous measures were compared by Student t-test or the Mann-Whitney U test when appropriate; 95% confidence interval estimate for the mean/median difference was performed. Two-sided significance tests were used throughout. A p value o0.05 was considered significant. Descriptive data are shown as median and interquartile range, mean ⫾ SD, or number and percentage. RESULTS
No protocol deviation and no crossovers were observed. All 32 oxygenators were sent to the laboratory for testing (Fig 2). There were no differences between the groups with respect to patient characteristics and prerandomization data, while the sevoflurane group required less midazolam and less nitroprusside than the control group during CPB (Table 1). The performance of the membrane oxygenator was similar between the 2 groups at 6 L/min flow and FIO2 ¼ 1.0, as documented by O2 transfer (55 ⫾ 6.4 mL/min/L in the sevoflurane group v 57 ⫾ 4.7 mL/min/L in the control group), CO2 transfer, and pressure drop (Table 2). DISCUSSION
This is the first study demonstrating that sevoflurane does not affect the performance of a semiporous PPL hollow-fiber membrane oxygenator and can be used during CPB to enhance cardiac protection in cardiac surgery. It should be remembered that there are 2 groups of membrane oxygenators and that the authors only tested the hollow-fiber membranes made of semiporous PPL, because they are used widely for standard CPB, offering excellent O2 exchange and CO2 removal for up to 6 hours.8 The other types of membrane oxygenators have not been tested in this study. They include diffusion and plasma-resistant oxygenators,
which are used increasingly for extracorporeal life support or extracorporeal membrane oxygenation in several clinical settings; for example, in patients who can no longer be supported by mechanical ventilation. There are some basic objectives in the design of a membrane oxygenator. The principal aim is to provide optimal gas exchange (O2 and CO2 transfer) during CPB. Many factors influence the performance of membrane oxygenators during cardiac surgery. Among these, patient characteristics and surgery-related events are the most common reasons for possible complications.7,13 In the present randomized study, patient characteristics, type of cardioprotection, duration of aortic cross-clamp, and CPB time were similar in the 2 groups, and, therefore, the authors demonstrated that the 2 anesthetic strategies have similar effects on the membrane oxygenator. Knowing that the performance of the oxygenator is not affected by volatile agents during CPB allows the anesthesiologist to use and titrate volatile anesthetics more accurately.14,15 R¨odig et al16 described the effects of sevoflurane in different concentrations on systemic vascular resistance during CPB. They found that an inspiratory concentration of 3% sevoflurane significantly reduced systemic vascular resistance index when compared with baseline values. In the present study, the authors confirmed that relatively high concentrations of sevoflurane (up to 3%) can be used during CPB to maintain anesthesia and also to control blood pressure while using less antihypertensive drugs (nitroprusside) and without using high doses of intravenous hypnotics (midazolam). Indeed, the authors used nitroprusside in 9 patients (56%) in the sevoflurane group and in 15 patients (94%) in the control group. Knowledge and predictability of oxygenator performance are crucial for effective conduct of CPB.8 The ability to adequately oxygenate a patient on CPB depends on various physiologic
906
NETO ET AL
32 patients
Sevoflurane Group(n=16)
Control Group(n=16)
16 membrane
16 membrane
oxygenators
oxygenators
Fig 2. Study flow chart.
factors (hemoglobin concentration, the percentage of hemoglobin saturated with O2 in the arterial blood, cardiac output, etc) and on the physical status of the membrane oxygenator.8 With respect to oxygenation, the function of an artificial lung is to transfer at least 4.5 mL O2/100 mL of blood at its maximum rated blood flow in accordance with AAMI/ISO standards. In the present study, O2 transfer testing results were in line with the manufacturer’s recommended normal values in both groups. This is an extremely important finding because volatile agents may improve survival in cardiac surgery1,3-6 and their effect seems to be increased if used during CPB.4,17 Notably, to date there are still no CPB machines manufactured, equipped, and ready to permit the use of halogenated agents, and each cardiac surgical center must find the best way to connect the vaporizer to the oxygenator. This is surprising because volatile agents are the only anesthetic drugs that might have an influence on survival, as suggested by the first international consensus conference6 on this topic.
Determination of the oxygenator performance in the laboratory was carried out under a strict set of conditions established by the AAMI. The authors did not study the pharmacokinetics of volatile agents during CPB, but they are aware that other authors11 have described the pharmacokinetics of these drugs with different membrane oxygenators and evaluated the elimination of volatile agents from blood. They demonstrated that PPL oxygenators had better uptake and elimination of isoflurane compared to polydimethylsiloxane oxygenators. All the studies mentioned 9-11 and the present findings suggest that semiporous PPL oxygenators (the standard oxygenator for cardiac surgery with CPB) offer excellent performance in the presence of volatile agents. In contrast, previous findings indicated that the diffusion oxygenators used for extracorporeal life support or extracorporeal membrane oxygenation increased the risk of awareness during CPB when isoflurane was used.9-11 Moreover, liquid sevoflurane, enflurane, and isoflurane were described as not causing damage to the PPL fibers of the membrane oxygenators according to light-microscopic examination, even after saturating concentrations for 3 hours.18 Limitations of the study: one study12 suggested that comparison between laboratory and clinical data are difficult, if not impossible, without appropriate statistical or computed modeling, suggesting a limited value of performance tests in the clinical setting. The perfusionist generally operates outside this standardized set of parameters and it is not common practice to measure the amount of O2 and CO2 transferred across the artificial lung after 6 hours of CPB. However, this is the gold standard to refer to, and there are situations during CPB when the determination of oxygenator gas exchange is critical and perhaps life saving. In the case of oxygenator dysfunction, for example, it is vital to distinguish between primary oxygenator dysfunction and a hypermetabolic state of the patient.8 The authors acknowledge that they tested only one type of membrane oxygenator and one inhalation agent, the reason being
Table 1. Demographic Data and Necessity of Drugs During Cardiopulmonary Bypass Sevoflurane Group (n ¼ 16)
Control Group (n ¼ 16)
Preoperative Data Sex (female), n 4 (25%) 6 (37%) Age, y 60 (57-62) 59.5 (53-69) ASA II, n 3 (19%) 2 (12%) ASA III, n 13 (81%) 14 (87%) Smoking, n 13 (81%) 9 (56%) Obesity, n 7 (44%) 8 (50%) Hypertension, n 15 (94%) 16 (100%) Diabetes mellitus, n 10 (62%) 7 (44%) Dyslipidemia, n 16 (100%) 15 (94%) Intraoperative Data Anesthesia duration, min 315 (300-341) 292 (258-313) Surgery duration, min 287 (280-315) 275 (242-300) Cross-clamping, min 55 (46-64) 63 (40-90) CPB duration, min 85 (72-101) 70 (59-91) Patients needing additional intravenous anesthetics or vasodilators during cardiopulmonary bypass Nitroprusside, n 9 (56%) 15 (94%) Midazolam bolus, n 1 (6%) 16 (100%) NOTE. Continuous data are expressed as median (IQR). Count data are shown as number and percentage. Abbreviations: ASA, American Society of Anesthesiologists; IQR, interquartile range.
p
0.4 0.4 0.9 0.9 0.1 0.7 0.9 0.3 0.9 0.09 0.17 0.4 0.07 0.03 o0.001
907
SEVOFLURANE ON THE POLYPROPYLENE MEMBRANE OXYGENATOR PERFORMANCE
CONCLUSION
Table 2. Performance of the Membrane Oxygenators
O2 Transfer (mL/min/L) Normal values: 450 mL/min/L CO2 Transfer (mL/min/L) Normal values: 442 mL/min/L Pressure Drop (mmHg) Normal range values: 4160 mmHg o250 mmHg
Sevoflurane
Control Group
Group (n ¼ 16)
(n ¼ 16)
p
55 ⫾ 6.4
57 ⫾ 4.7
0.4
54 ⫾ 10.0
53 ⫾ 5.0
0.8
184 ⫾ 9.1
196 ⫾ 25.0
The authors demonstrated, for the first time, that the use of sevoflurane as the main hypnotic drug during CPB for cardiac surgery did not affect the performance of the semi-porous PPL hollow-fiber membrane oxygenator. This could lead to a more widespread use of volatile agents during CPB to enhance organ protection and improve clinically relevant outcomes in patients undergoing cardiac surgery. ACKNOWLEDGMENTS
NOTE. Data are expressed as mean ⫾ SD.
that they are used most commonly in the authors’ institution and that the Vital membrane oxygenator was the only company to carry out the performance tests without any cost.
The authors thank all surgeons who helped and contributed to the study. Also, they are indebted to all the perfusionists who helped provide care to these patients during cardiopulmonary bypass.
REFERENCES 1. Landoni G, Biondi-Zoccai GG, Zangrillo A, et al: Desflurane and sevoflurane in cardiac surgery: A meta-analysis of randomized clinical trials. J Cardiothorac Vasc Anesth 21:502-511, 2007 2. Belhomme D, Peynet J, Louzy M, et al: Evidence for preconditioning by isoflurane in coronary artery bypass graft surgery. Circulation 100:II-340-II-344, 1999 (suppl II) 3. Jakobsen CJ, Berg H, Hindsholm KB, et al: The influence of propofol versus sevoflurane anesthesia on outcome in 10,535 cardiac surgical procedures. J Cardiothorac Vasc Anesth 21:664-671, 2007 4. Bignami E, Biondi-Zoccai G, Landoni G, et al: Volatile anesthetics reduce mortality in cardiac surgery. J Cardiothorac Vasc Anesth 23:594-599, 2009 5. De Hert S, Vlasselaers D, Barbe´ R, et al: A comparison of volatile and nonvolatile agents for cardioprotection during on-pump coronary surgery. Anaesthesia 64:953-960, 2009 6. Landoni G, Rodseth RN, Santini F, et al: Randomized evidence for reduction of perioperative mortality. J Cardiothorac Vasc Anesth 26:764-772, 2012 7. Nigro Neto C, Tardelli MA, Paulista PH: Use of volatile anesthetics in extracorporeal circulation. Rev Bras Anestesiol 62:346-355, 2012 8. Gleen PG: Cardiopulmonary Bypass Principles and Practice, (ed 3). Philadelphia: Lippincott Williams and Wilkins; 2008 pp 47-49, 2008 9. Philipp A, Wiesenack C, Behr R, et al: High risk of intraoperative awareness during cardiopulmonary bypass with isoflurane administration via diffusion membrane oxygenators. Perfusion 17:175-178, 2002
10. Wiesenack C, Wiesner G, Keyl C, et al: In vivo uptake and elimination of isoflurane by different membrane oxygenators during cardiopulmonary bypass. Anesthesiology 97:133-138, 2002 11. Hickey S, Gaylor JD, Kenny GN: In vitro uptake and elimination of isoflurane by different membrane oxygenators. J Cardiothorac Vasc Anesth 10:352-355, 1996 12. Griffith KE, Vasquez MR, Beckley PD, et al: Predicting oxygenator clinical performance from laboratory in-vitro testing. J Extra Corpor Technol 26:114-120, 1994 13. Berdajs DA, de Stefano E, Delay D, et al: The new advanced membrane gas exchanger. Interact Cardiovasc Thorac Surg 13:591-596, 2011 14. Blokker-Veldhuis MJ, Rutten PM, De Hert SG: Occupational exposure to sevoflurane during cardiopulmonary bypass. Perfusion 26:383-389, 2011 15. Reinsfelt B, Westerlind A, Ricksten SE: The effects of sevoflurane on cerebral blood flow autoregulation and flow-metabolism coupling during cardiopulmonary bypass. Acta Anaesthesiol Scand 55:118-123, 2011 16. R¨odig G, Keyl C, Wiesner G, et al: Effects of sevoflurane and isoflurane on systemic vascular resistance: Use of cardiopulmonary bypass as a study model. Br J Anaesth 76:9-12, 1996 17. De Hert SG, Van der Linden PJ, Cromheecke S, et al: Cardioprotective properties of sevoflurane in patients undergoing coronary surgery with cardiopulmonary bypass are related to the modalities of its administration. Anesthesiology 101:299-310, 2004 18. Crosbie AE, Vuylsteke A, Latimer RD: Inhalation anaesthetics and the Medtronic Maxima Plus membrane oxygenator. Br J Anaesth 80:878, 1998