- Email: [email protected]
Implantable artificial lung
Implantable artificial lung Preliminary report The ultimate treatment of chronic respiratory insufficiency is pulmonary replacement by an artificial organ, homologous lung transplantation, or chronic paracorporeal respiratory supplementation. The woven capillary membrane oxygenator appears to be a major development toward implantable artificial organs. The four units tested are made up of screens 3.5 by 4.0 em. of capillary tubing 0.3 mm. I.D. by 0.64 mm. O.D. assembled into rectangular blocks. Units made up offive, ten, twenty, and forty screens have been assembled and tested according to the protocol suggested by Galletti. The maximum oxygen transfer rate with blood was 48 mi. per minute per square meter. Water carbon dioxide transfer rate was 23.1 mi. per minute per square meter. The pressure drops in the liquid phase were 8.5, 15.3, 13.8, 17.6 mm. Hg at 1 L. per minute flow. These results indicate that the woven capillary membrane lung is an acceptably efficient oxygenator. The characteristics of design and performance suggest that this oxygenator can be made to be implanted into the chest or used as a paracorporeal respiratory assistance device.
P. 1. Morin, Ph.D., C. Gosselin, R. Picard, M. Vincent, M.Sc., R. Guidoin, D.Sc., and C. I. H. Nicholl, Ph.D., Ste-Foy, Quebec, Canada
Chronic respiratory insufficiency remains one of the unresolved problems of modern medicine. An increasing percentage of the population has various aspects of the disease. This situation is due to increasing atmospheric pollution in our cities, to increases in industrial diseases such as silicosis and asbestosis, and to individual smoking. Medical therapeutics does not offer much hope for the future, and it is quite probable that a more determined approach will have to be envisaged. This could be homologous lung transplantation, artificial lung implantation, or artificial lung paracorporeal supplementation. A number of attempts at homologous lung transplantation have been made.;'"" but few successes have been recorded." Artificial lung implantation has yet to begin, From Laval Hospital and Laval University, Ste-Foy, Quebec, Canada. This research project is supported by the Quebec Thoracic Society, the Medical Research Council of Canada, and the Conseil de la recherche en Sante du Quebec. Received for publication Dec. 6, 1976. Accepted for publication Feb. 7, 1977. Address for reprints: Pierre J. Morin, Ph.D., Director, Research Laboratories, Laval Hospital, 2725 Ste-Foy Rd., Ste-Foy, Quebec, Canada.
130
and short-term paracorporeal supplementation is still in its infancy. The work of Hill," Bartlett," Peirce," White," Awad;'" Sadoul.!' Kolobow.P Lande.I" and many others in short-term paracorporeal respiratory assistance is providing much needed answers concerning short-term to medium-term use of artificial lungs in man. This paper deals with a successful attempt at defining a geometry which would be suitable for an implantable artificial lung. The results obtained with the four prototypes manufactured are herein presented.
Construction of the artificial lung The geometry proposed attempts to recreate in an artificial organ the spongelike texture 'of the normal lung. The possibility of matching homologous lung elasticity, pulse absorption, and resistance to flow was deemed essential in the artificial lung. The four implantable artificial lung prototypes were constructed with silicone rubber capillary tubing* with an inner diameter of 0.3 mm. and an outer diameter of 0.6 mm. The capillary tubing was woven by hand into a tight rectangular weave six strands in one direction by one strand in the other. *Dow Corning Corporation, Midland, Mich.
Volume 74 Number 1 July, 1977
Implantable artificial lung
I3 I
Fig. 1. A , The three components used to make up a unit. Silicone inlet and outlet manifolds are also used to make up the gas inlet and outlet chambers. The small block in the center is made from woven silicone tubing grids and comprises the oxygenating surface of the units tested. B , The capillary tubes permeate the plastic wall and open into the gas chamber. Squares, 5.5 by 5 .5 cm., of the woven material are cut and attached to rigid plexiglass frames 5 by 5 em. with a 4 by 4 cm. cutout in the center. Five, ten, twenty, and forty such grids are mounted one behind the other to form blocks. The open ends of the capillary tubes are closed by dipping the tubes into liquid 734 RTV Dow Corning* adhesive . The blocks are then inserted into molds and pressure injected by centrifugal force. The ends of the capillary tubes projecting beyond the plexiglass frames are trimmed off, so that the tubes are reopened at the wall of the block. (Fig. I, B). The block is then inserted into specially designed gas and liquid inlet manifolds and secured in place with silicone rubber adhesive (Figs. I, A, and 2) . Blood bathes the outside of the capillary tubes while gases are blown through the short tubes oriented transversely to the blood path. Testing the units. The five, ten , twenty , and forty grid units were tested by the protocol suggested by Kolobow and Zapol. 12 Blood and water flows during the gas transfer tests were =200 , 500, 1,000, and 1,800 ml. per minute , and the gas flow rates were 2, 4, 6, and 8 L. per minute for the five , ten, twenty , and forty grid units , respectively. Later, the effect of gas flow on fluid pressure drop and the effect of liqu id flow on gas pressure drop were tested . The blood presssure drops across the device and gas transfer values in blood were
obtained via a venoarterial bypass with a pump in a large dog. Blood Po2 , PC02, and pH were measured with an ABL I made by Radiometer, * and pressure measurements were obtained with a physiograph (Electronics for Medicine) and Statham pressure transduc-
*Dow Corning Corporation, Midland. Mich.
*Acid·Base Laboratory. Radiometer A/S, Copenhagen, Denmark.
Fig. 2. Five, ten, and twenty grid units assembled and ready to be tested. Note that the opposite comers of the oxygenator block are glued to the gas chamber manifold. One of the chambers thus formed introduces the gases into one half of the unit, while the corresponding opposite chamber collects gases having passed through the tubes.
The Journalof Thoracic and Cardiovascular Surgery
132 Morin et al.
TO, A
(min~~2
) 50
40
30
N= 40
20
10"------------==---------:-:::=--------"':":'':'':'"-_. 500 1000 1500
Q. (ml/min·l
Fig. 3. Oxygen transfer into blood for various blood flows. Oxygen transfer values are reported as milliliters of oxygen per square meter of oxygenator surface area when blood is kept at 37° C. Values for only the twenty and the forty grid units are reported here.
AP
(mm
Hg)
100
80
60
N=40
40
20
500
1000
1500
O. (ml/min.)
Fig. 4. Pressure drop across the oxygenator for various blood flows in the twenty and forty grid units. Values are reported in millimeters of mercury.
Volume 74 Number 1 July, 1977
Implantable artificial lung
6 Pc, (mmHg) 30
m_.
0-0
N = 5 N
e-e N 0--0
133
= 10
= 20
N = 40
N=5
20
N = 20 10
. . . _---r-__
L_L~~;;:.",:::.:t::::::::::_
2
- - N = 40
4
Co (L/min)
6
8
10
Fig. 5. Gas pressure drop across the oxygenator for gas flows of 2, 4, 6, 8, and 10 L. Note that there is no significant increase in pressure drop across the forty grid unit even at 10 L. gas flows. Values are reported as milimeters of mercury. ers. For oxygen transfer values, a minimum of six inlet to outlet P0 2 values were measured. In some instances twelve determinations were carried out. The animal's blood was anticoagulated with heparin to keep the accelerated blood clotting time between 160 and 200 seconds. Upon termination of the tests with blood, the artificial lungs were washed with 5 percent glucose and lactated Ringer's solution at a 50: 50 concentration. The units were then filled with an isotonic 10 percent glutaraldehyde solution and prepared for examination with a scanning electron microscope by routine laboratory methods. Results Oxygen transfer values in blood. As demonstrated in Fig. 3, in the twenty grid unit at a blood flow of 360 mI. per minute per square meter, oxygen transfer rate was 22 ml. per minute. At a flow of 1,060 ml. per minute, oxygen transfer was increased to 48 mi. per minute per square meter. Further increases in flow did not give greater transfer values, and at a blood flow of 1,440 ml. per minute, the oxygen transfer was 37 mI. per minute per square meter. In the forty grid unit, oxygen transfer was 34 ml. per minute per square meter at a 360 mi. per minute blood flow. An increase in flow rate to 1,060 ml. per minute produced an increase in oxygen transfer values to 43 mI. per minute per square meter. In this unit also, a
further increase in blood flow to 1,440 ml, per minute decreased oxygen transfer values to 34 ml. per minute per square meter. The five and ten grid units were not usable after the tests with water. Blood pressure drop. From Fig. 4, the blood pressure drop rose from 25 mm. Hg at a flow of 360 ml. per minute to 90 mm. Hg at a 1,440 ml. per minute flow in the twenty grid unit. In the forty grid unit, the pressure drop was 16 mm. Hg at a 360 mi. per minute flow. The pressure drop was 56 mm. Hg at 1,440 ml. per minute flow. The difference in resistance to flow found for the twenty and forty grid units is probably due to small differences in the tightness in the weaving of the strands of tubing of the material making up the artificial lungs. Oxygen pressure drop. As shown in Fig. 5, gas pressure drop was increased almost linearly in all units with increased gas flow. In the twenty grid unit, it was 5 mm. Hg at a 2 L per minute flow and II mm. Hg at an 8 L per minute flow. In the forty grid unit, the gas pressure drop was 6 mm. Hg at a 2 L. per minute flow and 8 mm. Hg at a 10 L. per minute flow. Scanning electron microscopy. The usual red cells, platelets, and fibrin deposits were found on the capillary tubes. These were located in the passage ways through the woven material, surrounding the capillary tubes, and in certain specific areas along the outer wall of the device. The injection of silicone adhesive gave a
The Journal of
1 34 Morin et at.
Thoracic and Cardiovascular Surgery
Fig. 6. Scanning electron micrograph of oxygenator blood path. A, Note the smooth junction between the oxygenator wall and the tubes. B, Note the smooth wall surface between rows of tubes. ,-
A ~----------H
.-----
,-
.-1-.-...
G
c
r----- D ~=:::=_--E
'---
F
Fig. 7. The artificial lung will be made to occupy almost the entire thoracic volume on one side. A, Gas inlet and outlet port. C, Artificial lung. D, Chest wall. E, Heart. F, Posterior anchoring piece for the artificial lung. G, Blood inlet tube leading from the pulmonary artery to the artificial lung and outlet tube draining from the artificial lung into the left atrium. H, Artificial lung inlet and outlet manifolds. smooth wall (Fig. 6). The tubing-to-wall junction was well rounded and smooth. The same observations could be made in the corners of the device where the walls meet a right angle and the tubing becomes imbedded in two walls at a time. The gross overview of the device gives a generally satisfactory appearance under the scanning electron microscope. Discussion The artificial lung is the inner part of a more complex system which will comprise gas chambers and nonreturn valves. The whole system is to be built so that it will be implantable inside the thoracic cage and will use the normal respiratory movements to propel the
respiratory gases through the device. Blood circulation will be ensured by the right side of the heart (Fig. 7). The mechanical aspects of respiration in the device are illustrated in Fig. 8. On inspiration, the outlet valve (A) situated in the thoracic wall inlet-outlet port closes (arrow), and the inlet valve opens and allows air into the chamber (B) of the device. On expiration, the process reverses, and the air in the thoracic chamber is forced through the device into the outlet chamber and out into the artificial lung outlet port. The inner cavity of the chest wall is to be lined by a skin pedicle as described by Wosornu and Melrose.!" The skin pedicle will extend posteriorly along the wall of the thoracic cage to the beginning of the gas outlet chamber (E) situated at the posterior aspect of the chest. The criteria for acceptability set at the outset for the device were those suggested by Mortensen 15 and based on his work on paracorporeal artificial lung replacement: 1. The device must have low resistance to blood flow (not more than 30 mm. Hg pressure drop at a mean blood flow of 3 L. per minute with a hematocrit value of 40 percent. 2. The blood flow through the oxygenator must be patterned and controlled so that all red blood cells will have direct exposure to the oxygenating membrane on a single passage through the oxygenator. Flow pattern should not be significantly altered by positional or gravitational changes.
Volume 74
Implantable artificial lung
Number 1
July, 1977
13 5
__ A
~===~~=::;:::-
'-"<-'1------ 8
+-
'H===~
C
E
;,L------F
. t
Fig. 8. Left, The space B around the artificial lung fills on inspiration, and on expiration the respiratory gases are expelled through the oxygenator. A, Gas inlet and outlet port. B, Gas chamber. C, Artificial lung. D, Chest wall. E, Heart. F, Posterior anchoring piece. K, Outlet chamber. Right, Inlet port of the artificial lung. A, Skin and chest wall button. B, Skin pedicle. C, Artificial lung. J, Inlet and outlet valves. K, Gas chamber. I, Accordion structure separating the gas inlet and outlet ports.
3. There must be significant compliance in the system to both blood and gas flows so that pulsatile blood flow and cyclic gas flow can be accommodated without significant changes in resistance to flow. 4. Gas flow through the oxygenator must be accomplished at rates sufficient to fully oxygenate blood and with gas pressures which are less than blood pressures within the oxygenator. 5. Over-all size and configuration of the oxygenator must be adaptable to the interior of the thorax. 6. The device must be easy to sterilize. 7. The device must be durable enough to function a minimum of 10 days without service or adjustment. 8. The device must have acceptable gas transfer characteristics (for oxygen and carbon dioxide and be able to maintain these functions reliably and consistently. 9. Spacing, configuration, and size of blood and gas inlet and outlet connections must be adaptable to canine vascular and cardiac anatomic requirements. 10. All materials contacting blood should be blood compatible, smooth, and thromboresistant or easily coatable with thromboresistive coatings without altering gas transfer performance. The specific performance requirements of a suitable prosthetic lung for application in man are as follows: I. The prosthesis should have a total mean blood flow rate of 2 to 3 L. per minute (instantaneous flow of
o to 10 L. per minute), accepting an inlet blood pressure of 20-50/10-20 mm. Hg (mean of 15 to 40) and yielding an outlet pressure of 10 to 15 mm. Hg. 2. The prosthetic lung should have a total over-all gas transfer rate of 150 to 250 C.c. of oxygen per minute (hematocrit value of 30 to 40 percent and inlet P0 2 of 20 to 30 mm. Hg) and 120 to 180 C.c. of carbon dioxide per minute. 3. The total membrane surface area for oxygenation (depending on gas transfer characteristics of the material used) should be approximately 3 to 6 sq. M. 4. The over-all external dimensions should be not more than 8 to 10 em. in diameter (each, if a pair) and not more than 18 to 22 em. long. 5. Resistance to gas flow should be sufficiently low that adequate gas flow can be achieved to fully oxygenate blood (up to 3 L. per minute) with gas pressure which is less than blood pressure in the device « 10 mm. Hg). The results obtained show that the spacing in the woven material, the transverse surface area, and the number of grids can be adjusted to give the pressure drop characteristics deemed essential in matching the resistance to blood flow found in the remaining lung. The gas pressure drop across the device is only a few millimeters of mercury at flow rates up to 10 L. per minute. Therefore, it is presumed that this prerequisite has been fulfilled.
I 36 Morin et at.
Blood flow increases do not seem to increase gas pressure drop in the device by measurable amounts. This characteristic is important in an implant situation, since differences in blood flow could have resulted, as in other commercially available artificial lungs, in increased resistance and thereby jeopardize gas flow through the device. On inspiration the gases will come into one gas chamber, and on expiration pressure exerted by the chest will force the air through the device into the other chamber. It is expected that if proper provisions are made in the device, this route may be used to wash out the device at regular intervals during the period of implantation. The gas exchange values are somewhat disappointing if one compares this unit with commercially available extracorporeal lungs. It must be realized, however, that the tests carried out using blood were done with blood 75 percent saturated with oxygen. Also, blood flow through the device was continuous and not pulsatile. These two situations are not likely to increase gas transfer efficiency. We assume therefore that the results obtained are representative of adverse conditions and that under normal operating conditions efficiency will be increased somewhat. The weaving of the capillary tubes must be better controlled than is the case presently in order to prevent preferential flow paths from forming within the device. On the other hand, the smoothness of the tubing to artificial lung wall junction and of the artificial lung wall itself are encouraging and do not presage any major thrombotic problems. If the present square configuration and the size suggested by Mortensen were used, the oxygenating surface of an artificial lung would be in the range of 4.5 sq. M. for a 10 by 10 by 18 cm. unit. If a semilunar configuration and chambers such as those suggested by Wosomu'" were used, the oxygenator surface area could be nearly doubled to 9 sq. M. with an estimated oxygen transfer capacity of 180 ml. at low blood flows and 450 ml. at high blood flows. The gas transfer capacity is deemed sufficient, and implantable units are currently being developed. In conclusion, the results obtained with the prototypes tested are quire encouraging, and the woven capillary membrane oxygenator presently appears to be a major breakthrough toward an implantable artificial lung. REFERENCES Hardy, J. D., Webb, W. R., Dalton, M. L., Jr., and Milder, G. R.: Lung Homotransplantation in Man, J. A. M. A. 186: 1065, 1963.
The Journal of Thoracic and Cardiovascular Surgery
2 Magovern, G. J., and Yates, A. 1.: Human Homotransplantation of the Lung, Ann. N. Y. Acad. Sci. 120: 710, 1969. 3 White, 1. J., Tanser, P. H., Anthonisen, N. R., Wynands, 1. E., Pare, 1. A., Becklake, M. R., Monro, D. D., and MacLean, L. D.: Human Lung Homotransplantation, Can. Med. Assoc. 1. 94: 1199, 1966. 4 Shinoi, K., Hayata, Y., Aoki, H., Hozaki, M., Yoshioka, K., Shinoda, A., lwahashi, H., Ito, M., and Ando, M.: Pulmonary Lobe Homotransplantation in Human Subjects, Am. J. Surg. 111: 617, 1966. 5 Derom, F., Barbier, F., Ringoir, S., Versieck, 1., Rolly, G., Berzsenyi, G., Vermeire, P., and Vrints, G.: Ten Month Survival After Lung Homotransplantation in Man, 1. THoRAc. CARDIOVASC. SURG. 61: 835, 1971. 6 Hill, J. D., de Leval, M. R., Fallat, R. 1., Bramson, M. L., Eberhart, R. c., Schulte, H. D., Osborn, J. 1., Barber, R., and Gerbode, F.: Acute Respiratory Insufficiency: Treatment With Prolonged Extracorporeal Oxygenation, J. THORAc. CARDIOVASC. SURG. 64: 551, 1972. 7 Bartlett, R. H., Gazzaniga, A. B., Fong, S. W., and Burns, N. E.: Prolonged Extracorporeal Cardiopulmonary Support in Man, 1. THORAC. CARDIOVASC. SURG. 68: 918, 1974. 8 Peirce, E. c., II: Clinical Use of the Membrane Lung, in Cooper, P., and Nyhus, L. M., editors: Surgery Annual New York, 1972, Appleton-Century-Crofts, p. 191. 9 White, J. J., Andrews, H. G., Risemberg, H., Mazur, D., and Haller, 1. A., Jr.: Prolonged Respiratory Support in Newborn Infants With a Membrane Oxygenator, Surgery 70: 288, 197 I . 10 Awad, 1. A., Cloutier, R., Paradis, B., Dorval, J., Martin, L., and Morin P. 1.: Prolonged Pulmonary Assistance With the Membrane Gas Exchanger: A Case Report, J. Pediatr. Surg. 8: 871,1973. I I Sadoul, P., and Gille, J. P.: Les espoirs de I'assistance respiratoire extracorporelle, Anesth. Analg. (Paris) 29: 6, 1972. 12 Kolobow, T., and Zapol, W. M.: Partial and Total Extracorporeal Respiratory Gas Exchange With the Spiral Membrane Lung, in Bartlett, R. H., Drinker, P. A., and Galletti, P. M., editors: Advances in Cardiology: Basel, 1971, S. Karger, p. II2. 13 Lande, A. J., Lowell, E., Block, J. H., Carlson, R. G., Subramanian, V. A., Ascheim, R. S., Scheidt, S., Fillmore, S., Killip, T., and Lillehei, C. W.: Clinical Experience With Emergency Use of Prolonged Cardiopulmonary By-pass With a Membrane Pump Oxygenator, Ann. Thorac. Surg. 10: 409, 1970. 14 Wosornu, J. L., and Melrose, D. G.: The Feasibility of Scaling up the Space for Implantable Artificial Lungs, Br. J. Surg. 60: 910, 1973. 15 Mortensen, 1. D.: Personal communication.
P. 1. Morin, Ph.D., C. Gosselin, R. Picard, M. Vincent, M.Sc., R. Guidoin, D.Sc., and C. I. H. Nicholl, Ph.D., Ste-Foy, Quebec, Canada
Chronic respiratory insufficiency remains one of the unresolved problems of modern medicine. An increasing percentage of the population has various aspects of the disease. This situation is due to increasing atmospheric pollution in our cities, to increases in industrial diseases such as silicosis and asbestosis, and to individual smoking. Medical therapeutics does not offer much hope for the future, and it is quite probable that a more determined approach will have to be envisaged. This could be homologous lung transplantation, artificial lung implantation, or artificial lung paracorporeal supplementation. A number of attempts at homologous lung transplantation have been made.;'"" but few successes have been recorded." Artificial lung implantation has yet to begin, From Laval Hospital and Laval University, Ste-Foy, Quebec, Canada. This research project is supported by the Quebec Thoracic Society, the Medical Research Council of Canada, and the Conseil de la recherche en Sante du Quebec. Received for publication Dec. 6, 1976. Accepted for publication Feb. 7, 1977. Address for reprints: Pierre J. Morin, Ph.D., Director, Research Laboratories, Laval Hospital, 2725 Ste-Foy Rd., Ste-Foy, Quebec, Canada.
130
and short-term paracorporeal supplementation is still in its infancy. The work of Hill," Bartlett," Peirce," White," Awad;'" Sadoul.!' Kolobow.P Lande.I" and many others in short-term paracorporeal respiratory assistance is providing much needed answers concerning short-term to medium-term use of artificial lungs in man. This paper deals with a successful attempt at defining a geometry which would be suitable for an implantable artificial lung. The results obtained with the four prototypes manufactured are herein presented.
Construction of the artificial lung The geometry proposed attempts to recreate in an artificial organ the spongelike texture 'of the normal lung. The possibility of matching homologous lung elasticity, pulse absorption, and resistance to flow was deemed essential in the artificial lung. The four implantable artificial lung prototypes were constructed with silicone rubber capillary tubing* with an inner diameter of 0.3 mm. and an outer diameter of 0.6 mm. The capillary tubing was woven by hand into a tight rectangular weave six strands in one direction by one strand in the other. *Dow Corning Corporation, Midland, Mich.
Volume 74 Number 1 July, 1977
Implantable artificial lung
I3 I
Fig. 1. A , The three components used to make up a unit. Silicone inlet and outlet manifolds are also used to make up the gas inlet and outlet chambers. The small block in the center is made from woven silicone tubing grids and comprises the oxygenating surface of the units tested. B , The capillary tubes permeate the plastic wall and open into the gas chamber. Squares, 5.5 by 5 .5 cm., of the woven material are cut and attached to rigid plexiglass frames 5 by 5 em. with a 4 by 4 cm. cutout in the center. Five, ten, twenty, and forty such grids are mounted one behind the other to form blocks. The open ends of the capillary tubes are closed by dipping the tubes into liquid 734 RTV Dow Corning* adhesive . The blocks are then inserted into molds and pressure injected by centrifugal force. The ends of the capillary tubes projecting beyond the plexiglass frames are trimmed off, so that the tubes are reopened at the wall of the block. (Fig. I, B). The block is then inserted into specially designed gas and liquid inlet manifolds and secured in place with silicone rubber adhesive (Figs. I, A, and 2) . Blood bathes the outside of the capillary tubes while gases are blown through the short tubes oriented transversely to the blood path. Testing the units. The five, ten , twenty , and forty grid units were tested by the protocol suggested by Kolobow and Zapol. 12 Blood and water flows during the gas transfer tests were =200 , 500, 1,000, and 1,800 ml. per minute , and the gas flow rates were 2, 4, 6, and 8 L. per minute for the five , ten, twenty , and forty grid units , respectively. Later, the effect of gas flow on fluid pressure drop and the effect of liqu id flow on gas pressure drop were tested . The blood presssure drops across the device and gas transfer values in blood were
obtained via a venoarterial bypass with a pump in a large dog. Blood Po2 , PC02, and pH were measured with an ABL I made by Radiometer, * and pressure measurements were obtained with a physiograph (Electronics for Medicine) and Statham pressure transduc-
*Dow Corning Corporation, Midland. Mich.
*Acid·Base Laboratory. Radiometer A/S, Copenhagen, Denmark.
Fig. 2. Five, ten, and twenty grid units assembled and ready to be tested. Note that the opposite comers of the oxygenator block are glued to the gas chamber manifold. One of the chambers thus formed introduces the gases into one half of the unit, while the corresponding opposite chamber collects gases having passed through the tubes.
The Journalof Thoracic and Cardiovascular Surgery
132 Morin et al.
TO, A
(min~~2
) 50
40
30
N= 40
20
10"------------==---------:-:::=--------"':":'':'':'"-_. 500 1000 1500
Q. (ml/min·l
Fig. 3. Oxygen transfer into blood for various blood flows. Oxygen transfer values are reported as milliliters of oxygen per square meter of oxygenator surface area when blood is kept at 37° C. Values for only the twenty and the forty grid units are reported here.
AP
(mm
Hg)
100
80
60
N=40
40
20
500
1000
1500
O. (ml/min.)
Fig. 4. Pressure drop across the oxygenator for various blood flows in the twenty and forty grid units. Values are reported in millimeters of mercury.
Volume 74 Number 1 July, 1977
Implantable artificial lung
6 Pc, (mmHg) 30
m_.
0-0
N = 5 N
e-e N 0--0
133
= 10
= 20
N = 40
N=5
20
N = 20 10
. . . _---r-__
L_L~~;;:.",:::.:t::::::::::_
2
- - N = 40
4
Co (L/min)
6
8
10
Fig. 5. Gas pressure drop across the oxygenator for gas flows of 2, 4, 6, 8, and 10 L. Note that there is no significant increase in pressure drop across the forty grid unit even at 10 L. gas flows. Values are reported as milimeters of mercury. ers. For oxygen transfer values, a minimum of six inlet to outlet P0 2 values were measured. In some instances twelve determinations were carried out. The animal's blood was anticoagulated with heparin to keep the accelerated blood clotting time between 160 and 200 seconds. Upon termination of the tests with blood, the artificial lungs were washed with 5 percent glucose and lactated Ringer's solution at a 50: 50 concentration. The units were then filled with an isotonic 10 percent glutaraldehyde solution and prepared for examination with a scanning electron microscope by routine laboratory methods. Results Oxygen transfer values in blood. As demonstrated in Fig. 3, in the twenty grid unit at a blood flow of 360 mI. per minute per square meter, oxygen transfer rate was 22 ml. per minute. At a flow of 1,060 ml. per minute, oxygen transfer was increased to 48 mi. per minute per square meter. Further increases in flow did not give greater transfer values, and at a blood flow of 1,440 ml. per minute, the oxygen transfer was 37 mI. per minute per square meter. In the forty grid unit, oxygen transfer was 34 ml. per minute per square meter at a 360 mi. per minute blood flow. An increase in flow rate to 1,060 ml. per minute produced an increase in oxygen transfer values to 43 mI. per minute per square meter. In this unit also, a
further increase in blood flow to 1,440 ml, per minute decreased oxygen transfer values to 34 ml. per minute per square meter. The five and ten grid units were not usable after the tests with water. Blood pressure drop. From Fig. 4, the blood pressure drop rose from 25 mm. Hg at a flow of 360 ml. per minute to 90 mm. Hg at a 1,440 ml. per minute flow in the twenty grid unit. In the forty grid unit, the pressure drop was 16 mm. Hg at a 360 mi. per minute flow. The pressure drop was 56 mm. Hg at 1,440 ml. per minute flow. The difference in resistance to flow found for the twenty and forty grid units is probably due to small differences in the tightness in the weaving of the strands of tubing of the material making up the artificial lungs. Oxygen pressure drop. As shown in Fig. 5, gas pressure drop was increased almost linearly in all units with increased gas flow. In the twenty grid unit, it was 5 mm. Hg at a 2 L per minute flow and II mm. Hg at an 8 L per minute flow. In the forty grid unit, the gas pressure drop was 6 mm. Hg at a 2 L. per minute flow and 8 mm. Hg at a 10 L. per minute flow. Scanning electron microscopy. The usual red cells, platelets, and fibrin deposits were found on the capillary tubes. These were located in the passage ways through the woven material, surrounding the capillary tubes, and in certain specific areas along the outer wall of the device. The injection of silicone adhesive gave a
The Journal of
1 34 Morin et at.
Thoracic and Cardiovascular Surgery
Fig. 6. Scanning electron micrograph of oxygenator blood path. A, Note the smooth junction between the oxygenator wall and the tubes. B, Note the smooth wall surface between rows of tubes. ,-
A ~----------H
.-----
,-
.-1-.-...
G
c
r----- D ~=:::=_--E
'---
F
Fig. 7. The artificial lung will be made to occupy almost the entire thoracic volume on one side. A, Gas inlet and outlet port. C, Artificial lung. D, Chest wall. E, Heart. F, Posterior anchoring piece for the artificial lung. G, Blood inlet tube leading from the pulmonary artery to the artificial lung and outlet tube draining from the artificial lung into the left atrium. H, Artificial lung inlet and outlet manifolds. smooth wall (Fig. 6). The tubing-to-wall junction was well rounded and smooth. The same observations could be made in the corners of the device where the walls meet a right angle and the tubing becomes imbedded in two walls at a time. The gross overview of the device gives a generally satisfactory appearance under the scanning electron microscope. Discussion The artificial lung is the inner part of a more complex system which will comprise gas chambers and nonreturn valves. The whole system is to be built so that it will be implantable inside the thoracic cage and will use the normal respiratory movements to propel the
respiratory gases through the device. Blood circulation will be ensured by the right side of the heart (Fig. 7). The mechanical aspects of respiration in the device are illustrated in Fig. 8. On inspiration, the outlet valve (A) situated in the thoracic wall inlet-outlet port closes (arrow), and the inlet valve opens and allows air into the chamber (B) of the device. On expiration, the process reverses, and the air in the thoracic chamber is forced through the device into the outlet chamber and out into the artificial lung outlet port. The inner cavity of the chest wall is to be lined by a skin pedicle as described by Wosornu and Melrose.!" The skin pedicle will extend posteriorly along the wall of the thoracic cage to the beginning of the gas outlet chamber (E) situated at the posterior aspect of the chest. The criteria for acceptability set at the outset for the device were those suggested by Mortensen 15 and based on his work on paracorporeal artificial lung replacement: 1. The device must have low resistance to blood flow (not more than 30 mm. Hg pressure drop at a mean blood flow of 3 L. per minute with a hematocrit value of 40 percent. 2. The blood flow through the oxygenator must be patterned and controlled so that all red blood cells will have direct exposure to the oxygenating membrane on a single passage through the oxygenator. Flow pattern should not be significantly altered by positional or gravitational changes.
Volume 74
Implantable artificial lung
Number 1
July, 1977
13 5
__ A
~===~~=::;:::-
'-"<-'1------ 8
+-
'H===~
C
E
;,L------F
. t
Fig. 8. Left, The space B around the artificial lung fills on inspiration, and on expiration the respiratory gases are expelled through the oxygenator. A, Gas inlet and outlet port. B, Gas chamber. C, Artificial lung. D, Chest wall. E, Heart. F, Posterior anchoring piece. K, Outlet chamber. Right, Inlet port of the artificial lung. A, Skin and chest wall button. B, Skin pedicle. C, Artificial lung. J, Inlet and outlet valves. K, Gas chamber. I, Accordion structure separating the gas inlet and outlet ports.
3. There must be significant compliance in the system to both blood and gas flows so that pulsatile blood flow and cyclic gas flow can be accommodated without significant changes in resistance to flow. 4. Gas flow through the oxygenator must be accomplished at rates sufficient to fully oxygenate blood and with gas pressures which are less than blood pressures within the oxygenator. 5. Over-all size and configuration of the oxygenator must be adaptable to the interior of the thorax. 6. The device must be easy to sterilize. 7. The device must be durable enough to function a minimum of 10 days without service or adjustment. 8. The device must have acceptable gas transfer characteristics (for oxygen and carbon dioxide and be able to maintain these functions reliably and consistently. 9. Spacing, configuration, and size of blood and gas inlet and outlet connections must be adaptable to canine vascular and cardiac anatomic requirements. 10. All materials contacting blood should be blood compatible, smooth, and thromboresistant or easily coatable with thromboresistive coatings without altering gas transfer performance. The specific performance requirements of a suitable prosthetic lung for application in man are as follows: I. The prosthesis should have a total mean blood flow rate of 2 to 3 L. per minute (instantaneous flow of
o to 10 L. per minute), accepting an inlet blood pressure of 20-50/10-20 mm. Hg (mean of 15 to 40) and yielding an outlet pressure of 10 to 15 mm. Hg. 2. The prosthetic lung should have a total over-all gas transfer rate of 150 to 250 C.c. of oxygen per minute (hematocrit value of 30 to 40 percent and inlet P0 2 of 20 to 30 mm. Hg) and 120 to 180 C.c. of carbon dioxide per minute. 3. The total membrane surface area for oxygenation (depending on gas transfer characteristics of the material used) should be approximately 3 to 6 sq. M. 4. The over-all external dimensions should be not more than 8 to 10 em. in diameter (each, if a pair) and not more than 18 to 22 em. long. 5. Resistance to gas flow should be sufficiently low that adequate gas flow can be achieved to fully oxygenate blood (up to 3 L. per minute) with gas pressure which is less than blood pressure in the device « 10 mm. Hg). The results obtained show that the spacing in the woven material, the transverse surface area, and the number of grids can be adjusted to give the pressure drop characteristics deemed essential in matching the resistance to blood flow found in the remaining lung. The gas pressure drop across the device is only a few millimeters of mercury at flow rates up to 10 L. per minute. Therefore, it is presumed that this prerequisite has been fulfilled.
I 36 Morin et at.
Blood flow increases do not seem to increase gas pressure drop in the device by measurable amounts. This characteristic is important in an implant situation, since differences in blood flow could have resulted, as in other commercially available artificial lungs, in increased resistance and thereby jeopardize gas flow through the device. On inspiration the gases will come into one gas chamber, and on expiration pressure exerted by the chest will force the air through the device into the other chamber. It is expected that if proper provisions are made in the device, this route may be used to wash out the device at regular intervals during the period of implantation. The gas exchange values are somewhat disappointing if one compares this unit with commercially available extracorporeal lungs. It must be realized, however, that the tests carried out using blood were done with blood 75 percent saturated with oxygen. Also, blood flow through the device was continuous and not pulsatile. These two situations are not likely to increase gas transfer efficiency. We assume therefore that the results obtained are representative of adverse conditions and that under normal operating conditions efficiency will be increased somewhat. The weaving of the capillary tubes must be better controlled than is the case presently in order to prevent preferential flow paths from forming within the device. On the other hand, the smoothness of the tubing to artificial lung wall junction and of the artificial lung wall itself are encouraging and do not presage any major thrombotic problems. If the present square configuration and the size suggested by Mortensen were used, the oxygenating surface of an artificial lung would be in the range of 4.5 sq. M. for a 10 by 10 by 18 cm. unit. If a semilunar configuration and chambers such as those suggested by Wosomu'" were used, the oxygenator surface area could be nearly doubled to 9 sq. M. with an estimated oxygen transfer capacity of 180 ml. at low blood flows and 450 ml. at high blood flows. The gas transfer capacity is deemed sufficient, and implantable units are currently being developed. In conclusion, the results obtained with the prototypes tested are quire encouraging, and the woven capillary membrane oxygenator presently appears to be a major breakthrough toward an implantable artificial lung. REFERENCES Hardy, J. D., Webb, W. R., Dalton, M. L., Jr., and Milder, G. R.: Lung Homotransplantation in Man, J. A. M. A. 186: 1065, 1963.
The Journal of Thoracic and Cardiovascular Surgery
2 Magovern, G. J., and Yates, A. 1.: Human Homotransplantation of the Lung, Ann. N. Y. Acad. Sci. 120: 710, 1969. 3 White, 1. J., Tanser, P. H., Anthonisen, N. R., Wynands, 1. E., Pare, 1. A., Becklake, M. R., Monro, D. D., and MacLean, L. D.: Human Lung Homotransplantation, Can. Med. Assoc. 1. 94: 1199, 1966. 4 Shinoi, K., Hayata, Y., Aoki, H., Hozaki, M., Yoshioka, K., Shinoda, A., lwahashi, H., Ito, M., and Ando, M.: Pulmonary Lobe Homotransplantation in Human Subjects, Am. J. Surg. 111: 617, 1966. 5 Derom, F., Barbier, F., Ringoir, S., Versieck, 1., Rolly, G., Berzsenyi, G., Vermeire, P., and Vrints, G.: Ten Month Survival After Lung Homotransplantation in Man, 1. THoRAc. CARDIOVASC. SURG. 61: 835, 1971. 6 Hill, J. D., de Leval, M. R., Fallat, R. 1., Bramson, M. L., Eberhart, R. c., Schulte, H. D., Osborn, J. 1., Barber, R., and Gerbode, F.: Acute Respiratory Insufficiency: Treatment With Prolonged Extracorporeal Oxygenation, J. THORAc. CARDIOVASC. SURG. 64: 551, 1972. 7 Bartlett, R. H., Gazzaniga, A. B., Fong, S. W., and Burns, N. E.: Prolonged Extracorporeal Cardiopulmonary Support in Man, 1. THORAC. CARDIOVASC. SURG. 68: 918, 1974. 8 Peirce, E. c., II: Clinical Use of the Membrane Lung, in Cooper, P., and Nyhus, L. M., editors: Surgery Annual New York, 1972, Appleton-Century-Crofts, p. 191. 9 White, J. J., Andrews, H. G., Risemberg, H., Mazur, D., and Haller, 1. A., Jr.: Prolonged Respiratory Support in Newborn Infants With a Membrane Oxygenator, Surgery 70: 288, 197 I . 10 Awad, 1. A., Cloutier, R., Paradis, B., Dorval, J., Martin, L., and Morin P. 1.: Prolonged Pulmonary Assistance With the Membrane Gas Exchanger: A Case Report, J. Pediatr. Surg. 8: 871,1973. I I Sadoul, P., and Gille, J. P.: Les espoirs de I'assistance respiratoire extracorporelle, Anesth. Analg. (Paris) 29: 6, 1972. 12 Kolobow, T., and Zapol, W. M.: Partial and Total Extracorporeal Respiratory Gas Exchange With the Spiral Membrane Lung, in Bartlett, R. H., Drinker, P. A., and Galletti, P. M., editors: Advances in Cardiology: Basel, 1971, S. Karger, p. II2. 13 Lande, A. J., Lowell, E., Block, J. H., Carlson, R. G., Subramanian, V. A., Ascheim, R. S., Scheidt, S., Fillmore, S., Killip, T., and Lillehei, C. W.: Clinical Experience With Emergency Use of Prolonged Cardiopulmonary By-pass With a Membrane Pump Oxygenator, Ann. Thorac. Surg. 10: 409, 1970. 14 Wosornu, J. L., and Melrose, D. G.: The Feasibility of Scaling up the Space for Implantable Artificial Lungs, Br. J. Surg. 60: 910, 1973. 15 Mortensen, 1. D.: Personal communication.