Transactions on Biomedicine and Health vol 2, © 1995 WIT Press, www.witpress.com, ISSN 1743-3525
The use of perfluorochemical liquid as an alternative method to traditional mechanical ventilation M.L. Constantino Department of Bioengineering, Politecnico di Milano,
Abstract Over the last years, results of physiological studies in both adult and premature animals have demonstrated that ventilation with an oxygenated perfluorochemical (PFC) liquid provides effective gas exchange and acid-base balance and improves lung function and recovery. Perfluorochemical liquids are biologically inert and non-biotransformable substances. The low surface tension and high respiratory gas solubility of liquid PFC enable adequate oxygenation and carbon dioxide removal at low insufflation pressure. This work is aimed to determine by means of a mathematical simulation model the distribution of PFC pC^ and pCOs content in the lung and proposes a circuit set up for neonatal application. The results point out the possibility of using this new technique of liquid ventilation as an alternative to traditional gas ventilation particularly when immature neonates with insufficient or absent production of surfactant are concerned. 1 Introduction Clinical statistics show that extrauterine survival of preterm born babies is strictly linked to gestational age or more specifically to pulmonary system immaturity . The main causes of pulmonary distress are the deficiency in production of surfactant, the high surface tension in the alveoli and thus the ensuing inefficiency in gas exchange capabilities. Mechanical ventilation has developed new methods of ventilatory treatment that can be applied to the patient in relation to his/her pathology from high frequency ventilation to positive end expiratory pressure to extracorporeal
Transactions on Biomedicine and Health vol 2, © 1995 WIT Press, www.witpress.com, ISSN 1743-3525
634
Computer Simulations in Biomedicine
membrane oxygenation (ECMO) or extracorporeal CC^ removal (ECCO2R) just to mention the most known techniques. Though many limitations still exist related to the gestational age at which these techniques can be applied successfully. This is why since the seventies the concept of "liquid ventilation" has been spreading out and many studies have been demonstrating the many advantages given by this application through animal experimentation. At the beginning the thought was to perfuse the airways of the lungs with oxygenated saline solution, but to reach the desired oxygen concentration in the solution it was necessary to oxygenate the fluid hyperbarically (up to 5 atm), that was somehow indaginous and complex. Moreover the conditions of the animals at the end of experiments was not so close to the physiological one. For these reasons other fluids capable to transport oxygen were seeked for and Clark and Gollan proposed the use of perfluorocarbons (PFC) normobarically oxygenated as the liquid to support ventilation of the lung, and from then on many other researchers studied this technique applying it on different animals ' ' ' . Moreover, the elimination of air-liquid interfacial surface tension, and the improvement in lung function at early developmental ages in premature lambs ' ' ' , due to liquid PCF ventilation, has recently given rise to this method as an investigational therapy for severe respiratory distress in human infants . The first modalities of delivering PFC that were adopted consisted in gravitational supplying die PFC to the lung through a cannula placed into the trachea of the animal, but this method was too closely function of the position of the animal and of the airway's resistance. This led Moskowitz and later Shaffer to adapt a mechanical ventilator as to process PFC instead of air. They demonstrated that CO% could be well removed and that the animals could be weaned from liquid ventilation to normal ventilation with no ensuing respiratory distress or affection of the alveoli or of the whole lung. This work is aimed to determine by means of a mathematical simulation model the distribution of PFC pOi and pCC^ content in the lung and proposes a circuit set up for neonatal application. 2 Materials and methods The goal that should reach liquid ventilation is to oxygenate the blood of the patient and to remove carbon dioxide from it without any contact between air and alveolar surface area because of the deficiency of surfactant production. In order to do so it is necessary to prime the lung of the patient with oxygenated and warmed liquid PFC until the Functional Residual Capacity (FRC) of the lung is filled. Then it is necessary that a Tidal Volume (TV) is delivered and subsequently aspirated from the lung to supplement the FRC with fresh oxygenated PFC and to remove the carbon dioxide that has been extracted from trie venous blood flowing inside pulmonary capillaries. The circuit The circuit set up consisting in two roller pumps, one continuous paediatric membrane oxygenator, one reservoir with filter, one emergency reservoir,
Transactions on Biomedicine and Health vol 2, © 1995 WIT Press, www.witpress.com, ISSN 1743-3525
Computer Simulations in Biomedicine
635
transducers for temperature, pressure and patient's weight monitoring is shown in Figure 1.
Figure 1: Circuit set up for liquid ventilation.
The frequency of supplementation of the TV is chosen, following the protocol used by Shaffer et al. , at 4 respiratory acts per minute; more specifically 5 s are allowed for inspiration (duration of the admission of PFC) and 10 s are allowed for expiration. The amount of TV is related to the patient's weight and is ranging among 15-25 ml/kg. The roller pumps are synchronised as to function alternatively during insufflation or aspiration. PFC flow rates range among 180 to 400 ml/(min x kg) for the inspiration phase and among 360 and 800 ml/(min x kg) during the expiration period. The gas flow rate in the oxygenator is chosen as to maintain the recommended ratio between PFC flow rate and gas flow rate. The patient is connected to the circuit trough a tracheal bilumen cannula whose dimensions depend on the patient size. The patient weight is monitored to control the constancy of PFC volume inside the lungs or that is the same that FRC do not vary during therapy that would imply the necessity of adding supplementary volume of PFC inside the lungs. Possible reason of this variation can be a leakage of PFC due to evaporation. The model The simulation model deals about the distribution of oxygen and carbon dioxide partial pressure inside the alveolar volume of the lung that is the zone in which gas exchange normally takes place. When preterm neonates are concerned it is no longer strictly exact to speak about alveolar volume as it is not yet totally developed but in any case with this word the surface area taking part to gas
Transactions on Biomedicine and Health vol 2, © 1995 WIT Press, www.witpress.com, ISSN 1743-3525
636
Computer Simulations in Biomedicine
exchange is addressed. When PFC exits the oxygenator oxygen concentration is known but when the TV is administered to the lungs there is diffusion of oxygen from fresh PFC towards the FRC of the lung and backward diffusion of carbon dioxide from FRO to the fresh PFC. The distribution of die partial pressure of the gases thus is different in any layer in which the lung can be divided into. The lung was divided into subsequent layers following the anatomic subdivision and considering that in the upper airways mass transfer takes place just with the adjacent layers; while in the alveolar volume gas transfer is of the diffusive kind between PFC gases and venous blood that is flowing in the pulmonary capillaries. The mass transfer among adjacent layers follows the Pick's law while the gas exchange to and from the pulmonary capillaries depends also on the rate of production and consumption of CO2 and O2 respectively of the body and on the ratio between the volume of each alveolus considered and the total lung alveolar volume. Once the total surface area of the alveoli, the coefficient of diffusion of the alveolar membrane, the pulmonary blood flow rate of the patient and the partial pressure difference between gas and blood are known, then it is possible to determine the concentration of oxygen and carbon dioxide in PFC or blood necessary to reach a efficient gas exchange. The Tidal Volume of fresh oxygenated and warm PFC is delivered to the lung at the beginning of each respiratory cycle and it is simulated resetting the initial partial pressure conditions in a tract of the airways whose volume corresponds to the TV. The diffusion of the gases inside the FRC is thus a dynamic process. The main hypothesis of the simulation model is that the diffusion of the gases in the PFC takes place at the end of the inspiration so that the volume of fluid that separates the alveolar volume from the TV always equals FRC. For this reason pO2 and pCC*2 partial pressure are computed as an average on the volume that at any instant takes part to the exchange process and subsequently balanced on the duration of each complete respiratory cycle. 3 Results Figure 2 shows the mean pO2 variation as a function of the number of respiratory cycles in the PFC contained in the alveolar volume of the lung. It can be noticed that during the first respiratory cycles PFC pO2 decreases rapidly and then it maintains constant reaching stability. The maximum value on the ordinates is close to the PFC pO2 at the outlet of the oxygenator. The stabilisation of oxygen concentration is still sufficient to assure an efficient gas exchange with the venous blood. Figure 3 shows the variation in pO2 during one ventilatory cycle calculated after the stabilisation of the oxygen concentration is reached as a function of the duration of the respiratory cycle itself. It can be seen that the moment at which oxygen starts to diffuse is the end of the inspiration. The supplementation of fresh oxygenated PFC as the TV induces an abrupt rise in oxygen concentration in the alveolar volume that then decreases due to the exchange process with the venous blood.
Transactions on Biomedicine and Health vol 2, © 1995 WIT Press, www.witpress.com, ISSN 1743-3525
Computer Simulations in Biomedicine
637
400 T
360
6 63%o 280
240
%
180
270
360
450
time [s] Figure 2: PFC mean pC^ variation during 30 ventilatory cycles inside the alveolar volume.
320;
300
^280 j
8 [ & 260 f
240
0
6
9
12
15
time [s] Figure 3: pO, variation inside alveolar volume during one ventilatory cycle, once the mean pO% in PFC has reached stability.
Transactions on Biomedicine and Health vol 2, © 1995 WIT Press, www.witpress.com, ISSN 1743-3525
638
Computer Simulations in Biomedicirte
Figure 4 shows the course of the variation of pOO? in PFC from the beginning of the ventilation therapy to the 30th ventilatory cycle. Also in this case there is a rise in CC>2 concentration during the first ventilatory cycles that then stabilise at a value still sufficient to perform CC^ removal from the venous blood.
H 0
90
1—
180
270
360
450
time [s] Figure 4: PFC mean pCC^ variation during 30 ventilatory cycles inside the alveolar volume.
50 y 46
8 34 | 30 | 26
0
6
9
12
15
time [s] Figure 5: pCO^ variation inside alveolar volume during one ventilatory cycle, once the mean pCC^ in PFC has reached stability.
Transactions on Biomedicine and Health vol 2, © 1995 WIT Press, www.witpress.com, ISSN 1743-3525
Computer Simulations in Biomedicine
639
Figure 5 shows the course of pOO^ variation during one ventilatory cycle after the stabilisation of GO^ concentration in the alveolar volume has become stable. Also in this case the delivery of fresh oxygenated PFC induces a rapid reduction in pOO^ that then slowly increases as gas exchange takes place. 4 Discussion The results obtained show how the gas concentration in PFC varies in the alveolar volume that is in the tract that is concerned with gas exchange in the lung. The hypothesis made about the instant at which the diffusion of oxygen takes places is acceptable as the fluid moves in the airways with negligible velocity, but on the other hand makes the diffusion of gases to appear discontinuous inside the alveolar volume. If the discontinuity was not present the path of the curves representing oxygen or carbon dioxide concentrations in the alveolar volume would have been more smooth and a gradual variation would have appeared between die periods of inspiration and expiration, but no difference would be present in the figures related to gas exchange. The circuit proposed can be easily assembled, according to the size of the patient, and could be for instance set up using two conventional roller pumps used for single needle dialysis. The efficacy of the circuit set up has been tested in vitro and the different control parameters have been optimised. On the contrary a validation of the model requires an in vivo human experimentation because of the quite indaginous reproduction of lung compliance in vitro. In any case liquid ventilation therapy appears to be an attractive and efficient way to treat respiratory distress in crucial patient that cannot be otherwise treated with conventional gaseous ventilation.
References 1. Clements JA, Tierney DF. Alveolar stability associated with altered surface tension. In Handbook of Physiology. Respiration. American Physiology Society, ppl565-1583, Betesda, 1964. 2. Greenspan JS, Wolfs on MR, Rubenstein SD, Shaffer TH. Liquid ventilation of human preterm neonates, J Ped, 117, 1990, 106-111. 3. Clark LC, G oil an F. Survival of mammals breathing organic liquids equilibrated with oxygen at atmospheric pressure, Science, 152, 1966, 1755-1 756. 4. Saga S, Modell JH, Calderwood HW, Lucas AJ, Tham MK, Swenson MK. Pulmonary function after ventilation with fluorocarbon liquid PI 2-F (Caroxin-F), J AppI PKW, 34,1973,160-164. 5. Tuazon JG, Modell JH, Hood CI, Swenson FH. Pulmonary functions after ventilation with fluorocarbon liquid (Caroxin-D), AnestHesioiog}, 38, 1973, 134140.
Transactions on Biomedicine and Health vol 2, © 1995 WIT Press, www.witpress.com, ISSN 1743-3525
640
Computer Simulations in Biomedicine
6. Modell JH, Hood CI, Kuck EJ, Ruiz BC. Oxygenation by ventilation with fluorocarbon liquid (FX-80), AnescAesWog}, 34, 1971, 312-320. 7. Modell JH, Calderwood HW, Ruiz BC, Tham MK, Hood CL Liquid ventilation of primates, CKesc, 69, 1976, 79-81. 8. Rufer R, Spitzer HL Liquid ventilation in the respiratory distress syndrome, Ckest, 66, 1974, 298. 9. Schwieler GH, Robertson B. Liquid ventilation in immature newborn rabbits, BW Neonate, 29, 1976, 343-353. 10. Shaffer TH, Rubenstein D, Moskowitz GD, Delivoria-Papadopoulos M. Gaseous exchange and acid base balance in premature lambs during liquid ventilation since birth, Pediatr Res, 10, 1976, 227-231. 11. Shaffer TH, Tran N, Bhutan! VK, Sivieri EM. Cardiopulmonary function in very preterm lambs during liquid ventilation. Pediatr Res, 17, 1983, 680-684. 12. Moskowitz GD. A mechanical respirator for control of liquid breathing, Fed Proc, 29,1970, 1751-1752. 13. Shaffer TH, Moskowitz GD. Demand-controlled liquid ventilation of the lungs, J A%)f PkW, 36, 1974, 208-215. 14. Shaffer TH, Douglas PR, Lowe CA, Bhutani VK. Liquid ventilation: improved gas exchange and lung compliance in preterm lambs, Pediatr Res, 17, 1983, 303306.
