Extracorporeal life support systems (ECLSs) have been widely used for cardiopulmonary support, especially under critical situations such as cardiac arrest and acute respiratory distress syndrome. A conventional membrane oxygenator in the ECLS circuit requires a driving force, either static or pulsatile, for the blood to pass through pressure-resisting hollow fiber membranes. The devised pulsatile ECLS has twin blood sacs that are alternately compressed by a single moving actuator. One sac feeds the blood into the membrane oxygenator, and the other sac, which is located at the outlet port of the membrane oxygenator, generates pulsatile blood flow in the systemic circulation (Figure 1).
We have found that the pulsatile ECLS achieves higher pulse power index (PPI) levels and better kidney tissue perfusion than a nonpulsatile centrifugal pump in a porcine cardiac arrest model.1 The pulsatile ECLS also showed beneficial effects on coronary perfusion in terms of blood flow, flow velocity, and resistance when compared with a nonpulsatile centrifugal pump in a porcine cardiac arrest model.2 The twin-pulse mechanism of the devised pulsatile ECLS has proven safe and effective. One theoretical limitation may be low pump-setting rates, which result in relatively low pulse rates and pump output because of the characteristics of the pump driving mechanism.
We hypothesized that if a gravity-flow hollow fiber membrane oxygenator is installed in the circuit, the twin blood sacs of the devised pulsatile ECLS can be placed downstream of the membrane oxygenator. This would increase pump output by doubling the pulse rate at a given pump-setting rate while maintaining effective pulsatility. The purpose of this study was to evaluate the circuit configurations for the devised pulsatile ECLS in an in vivo experiment with respect to energy equivalent pressure (EEP), surplus hemodynamic energy (SHE), and pump output.3–8
Materials and Methods
All animals were treated humanely as described in the Guide for the Care and Use of Laboratory Animals issued by the Korea University School of Medicine. Twelve Yorkshire swine weighing 35–45 kg were randomly assigned into two groups according to the different circuit configurations of the devised pulsatile ECLS (T-PLS; Twin-pulse Life Support, Newheartbio Co., Seoul, Korea). In the serial group (n = 6), a conventional membrane oxygenator (Capiox SX10, Terumo Co., Tokyo, Japan) was placed between the twin blood sacs serially (Figure 2a). In the parallel group (n = 6), the twin blood sacs of the T-PLS were placed (parallel) downstream of a gravity flow membrane oxygenator (Capiox CX230, Terumo Co.) (Figure 2b).
Animals were premedicated with intramuscular ketamine (10 mg/kg) and placed on the surgery table after weighing. Electrocardiography electrodes were attached. A tracheostomy was performed, and a 6–7 endotracheal tube was inserted for mechanical ventilation. An intravenous fluid route was established at the right external jugular vein. General anesthesia was induced with thiopental sodium (5–10 mg/kg) and vecuronium bromide (0.1 mg/kg), and was maintained by the continuous infusion of propofol (6 mg·kg–1·h–1) together with N2O/O2 mixed gas (each 2 l/min). Mechanical ventilation was maintained with a tidal volume of 10–15 ml/kg and a respiratory rate of 20–25 breaths per minute. A 20G catheter was inserted into the right carotid artery for continuous monitoring of arterial pressure during the operation.
After making a clamshell incision through the third intercostal space, both internal thoracic arteries were dissected and ligated. The pericardium was incised and secured, and then 3 mg/kg heparin was injected. An extracorporeal circulation was constructed by inserting an 18Fr arterial cannulae into the ascending aorta and bicaval 22Fr venous cannulae through the right atrium, which were then connected to the T-PLS circuit.
The extracorporeal circuit was primed with lactated Ringer's solution. The total priming volume was approximately 1,300 ml. Gas flow to the membrane oxygenator was fixed at 2 l/min (FiO2 0.6) during the entire extracorporeal circulation. Ventricular fibrillation was induced by placing a 9V DC battery onto the right ventricular outflow tract, and then the total extracorporeal circulation was started. The pump-setting rates were varied as 30, 40, and 50 bpm, and the parameters were collected as follows.
Flow rates were measured at the arterial cannula using an ultrasonic flow meter (TS410 flow meter, ME11PXL probe, Transonic, Ithaca, NY). Mean arterial pressure (MAP) was measured in the descending aorta. At each pump-setting rate, pressure and flow waveforms were measured for 30 seconds continuously as well as simultaneously by using a homemade data-acquisition board. Each measurement was repeated six times, and the waveforms were analyzed with the MATLAB software (Mathworks, Natick, MA). At the end of the experiment, all animals were euthanized in the anesthetic state following our institutional guidelines.
Quantification of Pressure-Flow Waveforms
The following formula defines the EEP as described elsewhere3–8:
where f is the pump flow rate (l/min), p the arterial pressure (mm Hg), and time integrals were over one pulse cycle. EEP, expressed in millimeters of mercury, is the ratio of the areas beneath the hemodynamic power curve and the pump flow-rate curve at the end of the flow and pressure cycles.
Surplus hemodynamic energy is defined as follows5,8:
where MAP is the mean aortic pressure (mm Hg). The constant (i.e., 1,332) converts pressure from millimeters of mercury to dynes per square centimeter. SHE is the “extra energy” that exists only if there is some degree of pulsatility in pressure or flow.
The Wilcoxon rank sum test was used to compare the percent changes of MAP to EEP and pump outputs between the serial and parallel groups. The Kruskal-Wallis test was used to compare parameters within the same group at different pump rates. For all tests, a p value of less than 0.05 was considered statistically significant. All results are expressed as means and standard error of the means.
Preoperative body weight was not different between the groups (40 ± 1.5 kg in the parallel group vs. 41 ± 1.3 kg in the serial group; p = NS).
In the serial group, the pressure and flow waveforms showed a large pulse followed by a small pulse, i.e., twin pulses. The large pulse was generated by the T-PLS actuator directly pushing the blood sac connected to the aortic cannula. The small pulse was generated by the actuator pushing the other blood sac connected to membrane oxygenator, which indirectly conducted small amounts of blood flow to the aortic cannula (Figure 3). In the parallel group, the pressure and flow waveforms showed “doubled pulse rate” of similar amplitude at a given pump-setting rate. The twin blood sacs were connected to the aortic cannula in the parallel fashion, and thus one reciprocating movement of the actuator compressed each sac alternately, i.e., the blood was pumped twice per beat (Figure 4).
In the parallel group, the percent changes of MAP to EEP were 13.0% ± 1.7%, 12.0% ± 1.9%, and 7.6% ± 0.9% at pump-setting rates of 30, 40, and 50 bpm, respectively. The difference within the parallel group was significant (p = 0.035). In the serial group, the percent changes of MAP to EEP were 22.5% ± 2.4%, 23.2% ± 1.9%, and 21.8% ± 1.4% at pump-setting rates of 30, 40, and 50 bpm, respectively. The difference within the serial group was insignificant (p = NS). The serial group had higher percent changes of MAP to EEP than the parallel group at the pump-setting rates of 30, 40, and 50 bpm, and the differences between the groups were significant (p = 0.030) (Table 1, Figure 5).
The SHE values were 20,131 ± 1,408 erg/cm3, 21,739 ± 2,470 erg/cm3, and 15,048 ± 2,108 erg/cm3 at pump-setting rates of 30, 40, and 50 bpm, respectively, in the parallel group. These differences were significant (p = 0.04). In the serial group, SHE values were 33,968 ± 3,001 erg/cm3, 38,232 ± 3,281 erg/cm3, and 37,964 ± 2,693 erg/cm3 at pump-setting rates of 30, 40 and 50 bpm, respectively. However, these differences were insignificant (p = NS) (Table 1).
Pump outputs were 2.3 ± 0.2 l/min, 3.1 ± 0.2 l/min, and 3.7 ± 0.2 l/min at pump-setting rates of 30, 40, and 50 bpm, respectively, in the parallel group. These differences were significant (p = 0.003). In the serial group, pump outputs were 1.9 ± 0.1 l/min, 2.2 ± 0.1 l/min, and 2.5 ± 0.1 l/min at pump-setting rates of 30, 40 and 50 bpm, respectively. These differences were also significant (p = 0.001). Both groups showed increased pump output with increased pump-setting rates, though the parallel group had higher pump outputs than the serial group at pump-setting rates of 40 and 50 bpm. (p = 0.010) (Table 1, Figure 6).
A single pulsatile ECLS system was previously developed but failed because a high level of circuit pressure was generated at the membrane inlet with resulting blood cell trauma.9,10 To minimize the pressure plateau in the circuit, a novel pulsatile ECLS system (T-PLS) was developed by adapting the Korean artificial heart (AnyHeart) technology.11,12 The twin-pulse mechanism of the T-PLS can effectively reduce the circuit pressure at the membrane inlet and thus hemolysis.10,13 In pulsatile ECLS, however, the pressure impact on the membrane oxygenator remains a concern.3,10 It has been suggested that an oxygenator with low pressure drop across the membrane, such as the Capiox SX10 (Terumo Co.), is beneficial to pulsatility and cell trauma.14,15 Ündar et al. also noted that a gravity-flow hollow fiber membrane oxygenator such as the Capiox CX230 is superior to other types of membrane oxygenators for pulsatile flow, because it is placed on the venous side of the extracorporeal circuit (which is a parallel configuration in this study) and therefore does not dampen the pulsatility.16 It has also been reported that a higher dP/dtmax causes more hemolysis and that the twin-pulse mechanism can effectively reduce dP/dtmax compared with a single-pulse mechanism (216 vs. 744 mm Hg/sec).13 We have adopted the above ideas to optimize the T-PLS circuit by comparing the serial vs. parallel configuration (Figure 2b).
We hypothesized that by doubling the pulse rate at a given pump-setting rate while maintaining effective pulsatility, a parallel circuit configuration incorporating a gravity-flow membrane oxygenator would have higher pump output than a serial configuration with a conventional membrane oxygenator. Our data show that the parallel circuit configuration has higher pump outputs than the serial configuration while generating pulsatile energy that is as effective as the serial configuration.
The parallel circuit configuration shows a higher pump output than the serial configuration. Twin blood sacs are connected to an aortic cannula in parallel, and thus one reciprocating movement of the actuator generates two pulses by compressing the twin blood sacs alternately at a given pump-setting rate. For example, the pump-setting rate of 30 bpm in the parallel circuit configuration produces a pulse rate of 60 bpm, whereas same pump-setting in the serial circuit configuration produces a pulse rate of 30 bpm. Thus, the parallel circuit configuration can generate a higher pump output than a serial configuration at any given pump-setting rate.
In comparing the pulsatility of the two-circuit configuration, both EEP and SHE are thought to be useful tools.3–8 As discussed in a previous study,17 in the evaluation of pulsatility, SHE may represent a more constant value than the percent changes in EEP and MAP. We are currently investigating this phenomenon.
One of the limitations of this study stems from not including the circuit pressure and blood profiles (e.g., hemolysis or hemoglobin), which are important issues and potential advantages of the pulsatile ECLS. Although this study focused on pulsatility related to the circuit configuration, those issues should be studied in the future. Another limitation is that, in the serial circuit configuration, the pump -rates of 30, 40, and 50 bpm (same as the pulse rates in the serial group) are too low for 35–45 kg medium-sized animals. The low pump-setting rate is mainly derived from the characteristics of the driving mechanism of the T-PLS. However, it should not be a concern in the parallel circuit configuration, because the pulse rate is doubled to 60, 80, and 100 bpm at pump-setting rates of 30, 40, and 50 bpm, respectively.
In conclusion, both the parallel and serial circuit configurations of the pulsatile ECLS (T-PLS) generate effective pulsatility. As for the pump output, the parallel circuit configuration provides higher flow than the serial circuit configuration by doubling the pulse rate at a given pump-setting rate.
This study was supported by the Korean Health 21 R&D Project, Korean Ministry of Health & Welfare, (grant 02-PJ3-PG6-EV09-0001), and by the Brain Korea 21 Project of the Korean Ministry of Education and Human Resource Development. The authors thank Dr. Akif Ündar of Penn State Children's Hospital for the invaluable advice that he provided during the planning of this study and during manuscript revision.
1.Kim HK, Son HS, Fang YH et al: The effects of pulsatile flow on renal tissue perfusion during cardiopulmonary bypass: A comparative study of pulsatile and nonpulsatile flow. ASAIO J
51: 30–36, 2005.
2.Son HS, Sun K, Fang YH, et al: The effects of pulsatile versus non-pulsatile extracorporeal circulation on the pattern of coronary artery blood flow during cardiac arrest. Int J Artif Organs
28: 609–616, 2005.
3.Ündar A:Myths and truths of pulsatile and non-pulsatile perfusion during acute and chronic cardiac support. Artif Organs
28: 439–443, 2004.
4.Shepard RB, Simpson DC, Sharp JF: Energy equivalent pressure. Arch Surg
93: 730–740, 1966.
5.Ündar A, Zapanta CM, Reibson JD et al: Precise quantification of pressure flow waveforms of a pulsatile ventricular assist device. ASAIO J
51: 56–59, 2005.
6.Wright G: Hemodynamic analysis could resolve the pulsatile blood flow controversy. Ann Thorac Surg
58: 1199–1204, 1994.
7.Ündar A: Energy equivalent pressure formula is for precise quantification of different perfusion modes. Ann Thorac Surg
8.Ündar A, Rosenberg G, Myers JL: Major factors in the controversy of pulsatile versus nonpulsatile flow during acute and chronic cardiac support. ASAIO J
51: 173–175, 2005.
9.Kim TS, Sun K, Lee KB et al: Application of a pressure-relieving air compliance chamber in a single-pulsatile extracorporeal life support system: An experimental study. Artif Organs
28: 1106–1109, 2004.
10.Lee HS, Rho YR, Lee HS et al: In vivo evaluation of the pulsatile ECLS system. Int J Artif Organs
6: 25–29, 2003.
11.Sun K, Son HS, Jung JS et al: Korean artificial heart (AnyHeart):An experimental study and the first human application. Artif Organs
27: 8–13, 2003.
12.Shin JS, Sun K, Son HS et al: A preclinical cadaver fitting study of implantable biventricular assist device—AnyHeart. Int J Artif Organs
27: 495–500, 2004.
13.Rho YR, Choi H, Lee JC et al: Applications of the pulsatile flow versatile ECLS: In vivo studies. Int J Artif Organs
26: 428–435, 2003.
14.Ündar A, Koenig KM, Frazier OH, Fraser Jr CD: Impact of membrane oxygenators on pulsatile versus nonpulsatile perfusion in a neonatal model. Perfusion
15: 111–120, 2000.
15.Ündar A, Owens WR, McGarry MC, et al: Comparison of hollow-fiber membrane oxygenators in terms of pressure drop of the membranes during normothermic and hypothermic cardiopulmonary bypass in neonates. Perfusion
20: 1–4, 2005.
16.Ündar A, Frazier OH, Fraser Jr CD: Defining pulsatile perfusion: Quantification in terms of energy equivalent pressure. Artif Organs
23: 712–716, 1999.
17.Lee JJ, Lim CH, Son HS, et al: In vitro evaluation of the performance of Korean pulsatile ECLS(T-PLS) using precise quantification of pressure-flow waveforms. ASAIO J
51: xxx–xxx, 2005.