- Email: [email protected]
Extracorporeal membrane oxygenator support for human lung transplantation
Extracorporeal membrane oxygenator support for human lung transplantation Extracorporeal membrane oxygenator (ECMO) support was provided for a 19-year-old boy undergoing right lung transplantation. Perfusion was begun several hours prior to transplant, to correct profound hypercapnia. After the operation, ECMO was required because of inadequate gas exchange by the transplanted lung. Perfusion was continued for a total of 96 hours. During this time, the temporarymalfunction of the transplanted lung owing to the reimplantation response reversed, and the patient was successfully removed from the oxygenator and subsequently weaned from the ventilator. He died on the eighteenth postoperative day of bronchial dehiscence. ECMO support appears to be a feasible means of supporting patients during lung transplantation and during the period of reversible lung malfunction that may occur in the early postoperative period.
J. M. Nelems, M.D., J. Duffin, Ph.D., M. F. X. Glynn, M.D., J. Brebner, M.D., A. A. Scott, M . D . , and J. D. Cooper, M.D., Toronto, Ontario, Canada
A A u m a n lung transplantation remains in the developmental stage, with several basic problems yet to be solved. One such problem is the temporary malfunction of the transplanted lung during the initial few days after transplantation. This malfunction, termed the reimplantation response, may be refractory to maximal ventilatory support and result in the death of the recipient in the first week. A review of the previous clinical experience with lung transplantation 1 revealed that 43 percent of the recipients died within the first week, usually of respiratory failure. Although intraoperative cardiopulmonary support with a bubble oxygenator has been used for several lung transplants (F. J. Veith, personal communication), no such extracorporeal support has been attempted during the postoperative period because of the coagulopathy which inevitably develops after only several hours with a conventional bubble oxygenator. 2 However, recent improvements in the design and utilization of membrane oxygenators have permitted perfusion to be conducted safely for days. It has therefore been anticipated 1 that extracorporeal membrane oxyFrom Toronto General Hospital, Toronto, Ontario, Canada. This work supported in part by the Ontario Heart Foundation, Crant No. 1-27, the Medical Research Council of Canada, Grano No. Ma 5953, and the Toronto General Hospital Foundation. Received for publication Dec. 22, 1977. Accepted for publication March 10, 1978. Address for reprints: J. D. Cooper, M.D., 1-131 University Wing, Toronto General Hospital, Toronto, Ontario, Canada M5G 1L7. 28
genator (ECMO) perfusion might provide temporary postoperative support for lung transplant recipients during the initial period of transplant malfunction. We recently provided membrane oxygenator support for a lung transplant recipient for a period of 96 hours. This support enabled the patient to survive a severe reimplantation response. The patient was subsequently weaned successfully from all ventilatory assistance and became ambulatory. He died on the eighteenth postoperative day of bronchial dehiscence. This report outlines the technical aspects of membrane oxygenator perfusion for lung transplantation. Case report A 19-year-old man sustained burns over 30 percent of his body in a house fire. On admission to the hospital there was evidence of an airway burn. The patient remained on a ventilator for the next 5 months, during which a progressive deterioration of lung function and increasing hypercapnia developed. The PaCo2 rose to 120 mm. Hg despite a minute ventilation of 24 L.; oxygenation was less of a problem, with aPa 0;! of 85 mm. Hg on an Fi 0 , of 0.5. When the hypercapnia continued to worsen despite all therapy, it was decided to proceed with right lung transplantation. The patient was connected to a membrane oxygenator 5 hours prior to the actual operation to permit gradual correction of the hypercapnia and the respiratory acidosis. The operation was performed without difficulty, but despite an ischemic time of only 2 hours, 15 minutes, the transplanted lung failed to function satisfactorily immediately following the transplantation procedure. ECMO was continued for a total of 96 hours, by which time the transplanted lung had recovered sufficiently to support the patient. Mechanical ven0022-5223/78/0176-0028S00.50/0 © 1978 The C. V. Mosby Co.
Volume 76
ECMO support for lung transplantation
Number 1 July, 1978
29
tilatory support was required for thefirst7 postoperative days. On the tenth day, the patient was ambulatory and was weaned from all supplementary oxygen. He died on the eighteenth postoperative day of bronchial dehiscence associated with localized necrosis. No parenchymal lung infection was present. Details of perfusion A 3.5 sq. M. Sci-Med Kolobow lung* with a fillerfree silicone rubber surface was used for the perfusion. The circuit consisted of silicone rubber tubing throughout, with polycarbonate tubing connectors. The details of the control, monitor, and alarm systems have previously been reported.3 The cannulation scheme was designed to allow rapid alternation between venovenous and venoarterial perfusion modes as might be required. Fig. 1 illustrates the circuit in schematic form. Thin-walled, spring-wirereinforced polyurethane cannulas, as initially described by Kolobow and Zapol,4 were used. The inferior vena cava was cannulated to the level of the diaphragm via the right common femoral vein, and the distal femoral vein was drained with a separate, smaller catheter. The superior vena cava was cannulated via the right internal jugular vein. The aortic arch was cannulated retrograde via the right common femoral artery with a cannula 75 cm. in length and 6.35 mm. in diameter.5 A separate, smaller arterial cannula was inserted into the distal femoral artery for perfusion of the right leg. Venovenous bypass was begun and the flow rate was gradually increased to 3 L. per minute. The membrane lung was ventilated with 5 percent carbon dioxide in oxygen to prevent too rapid a removal of carbon dioxide from the patient, and the carbon dioxide content of the effluent gas from the membrane lung was continuously monitored with a rapid carbon dioxide analyzer to measure the rate of carbon dioxide removal. After five hours of perfusion, the PaCo2 was 80 mm. Hg and the pH was 7.35. During this time 50 mEq. of potassium chloride was given intravenously. The venoarterial option for perfusion was provided in case venovenous perfusion proved inadequate during the transplantation operation. The venoarterial perfusion circuit was temporarily tested prior to removal of the right lung, and it performed well. With the tip of the aortic cannula near the left subclavian artery, oxygen saturation was measured in the right ear with an ear oximeter. Since the saturation rose to above 95 percent when venoarterial perfusion was begun, it was not believed necessary to advance the cannula into the ascending aorta. After several minutes of venoarterial perfusion, the circuit was switched back to the *Sci-Med Life Systems, Inc., Minneapolis, Minn.
i
t
Fig. 1. Schematic diagram of the perfusion circuit. For venovenous perfusion, occluding clamps were placed at points v. For venoarterial perfusion, the clamps at points v were removed, and an occluding clamp was placed at point a. venovenous mode and the operation was begun. Despite a mean pulmonary artery pressure in excess of 50 mm. Hg, the patient's condition remained stable. This permitted the venovenous mode of perfusion to be used exclusively. During the transplantation procedure, there was no indication of unusual bleeding, and the amount of blood transfused (4 units of whole blood and 1 unit of packed cells) was appropriate for the nature of the surgery performed. Following completion of the transplantation and closure of the thoracotomy, an unsuccessful attempt was made to wean the patient from ECMO. The Po 2 fell to
3 0 Nelems et al.
38 mm. Hg at this point when flow was reduced to 1 L. per minute. ECMO via the venovenous route was therefore continued postoperatively. As venoarterial perfusion had not been required, the arterial cannulas were removed at the end of the transplantation procedure. Postoperatively, an average flow rate of 2 L. per minute was used. Since the patient had an estimated surface area of 1.75 sq. M., the flow rate was approximately 1.15 L. per minute per square meter. On the fourth day following transplantation, ECMO support was discontinued, the venous cannulas were removed, the venotomy sites were closed, and flow was restored in the femoral and jugular veins. Detailed examination of the membrane oxygenator surface indicated that less than 4 percent of its surface was covered with adherent thrombus. The remaining surface appeared entirely clean. Coagulation during ECMO was managed as follows: Prior to cannulation, a loading dose of 5,000 units of heparin was given intravenously, after which heparin was administered by continuous infusion. Hourly measurements of the activated clotting time were used to determine the heparin requirement. The activated clotting time, measured with the Hemochron apparatus,* was maintained between 120 and 150 seconds. The average hourly dose of heparin during the first 8 hours of perfusion (the preoperative and intraoperative periods) was 212 units per hour. For the entire 96 hours of perfusion, the average rate of heparin administration was 620 units per hour. Fig. 2 gives the gas exchange data and the platelet counts for the entire period of ECMO support. The decline in the platelet count during the initial hours of ECMO resulted almost entirely from hemodilution with the 2 L. of crystalloid prime. Thereafter, there was a gradual but continuing platelet loss, resulting in a platelet count of 50,000 per cubic millimeter at the time of ECMO termination. Platelet transfusion was not required. The fibrinogen titer and the thrombin time (with heparin neutralization) remained within normal limits, as did the reptilase time; thus there was no disturbance in the normal hemostatic mechanism. Discussion Although cardiopulmonary bypass with a bubble oxygenator has been used intraoperatively in previous lung transplantations, post-transplantation extracorporeal support has not been previously available because of the major hemostatic derangements resulting after several hours of perfusion with a bubble oxygen*InternationaI Technidyne Company, Edison, N. J.
The Journal of Thoracic and Cardiovascular Surgery
ator. However, improvements in the design and ultilization of membrane oyygenators have permitted safe perfusions for more than a week without significant alteration of hemostatic function.6, 7 Because of this experience, we were encouraged to believe that prolonged ECMO support might safely be used as an adjunct to lung transplantation. The reimplantation response is defined as the morphologic, radiographic, and functional changes which occur in a transplanted lung as the result of the surgical procedure itself, exclusive of any effects owing to graft rejection. The changes include the development of alveolar infiltrates evident on chest x-ray films and the impairment of alveolar ventilation and gas exchange. The changes reach a peak within 3 days and regress over the next 3 weeks.1' 8 - 9 On review of the 37 lung transplant operations previously performed,1 it was noted that 16 patients (43 percent) had died within the first week, primarily of respiratory insufficiency. Of this number, eight (21 per cent) failed to survive the first 24 hours. It seems likely that many of these early postoperative failures were due more to the reimplantation response than to the effects of rejection. In our own case, the patient probably would have died during the initial 24 hour posttransplant period without the availability of ECMO support. Prior to the start of ECMO, the patient was given 2 Gm. of sulfinpyrazone (Anturan) intravenously. This dose was based upon evidence from our laboratory10 showing a marked improvement in early platelet sparing when sheep were pretreated with intravenous sulfinpyrazone prior to the onset of membrane oxygenator perfusion. The membrane lung was flushed with carbon dioxide for 2 hours before being primed with lactated Ringer's solution, which had been degassed for 1 hour in a vacuum chamber to lower its gas content. This priming technique is based upon experiments from our laboratory11 which have demonstrated improved platelet sparing when gas nuclei, trapped in the surface roughness of the membrane, are removed in the priming process. The bypass circuit was designed to permit conversion between the venovenous and venoarterial modes of perfusion (Fig. 1). In general, we prefer the venovenous mode, because it is safer, technically simpler to control, does not interfere with the patient's cardiovascular function, and assures that the oxygenated blood is distributed equally to the entire body. The recipient had excellent cardiac function, so that left ventricular support with the venoarterial bypass mode was not required. However, the mean pulmonary artery pressure
Volume 76 Number 1 July, 1978
ECMO support for lung transplantation
3 I
200,000 PLATELET 150,000 COUNT
100,000
\
50,000 h 0 PEEP cm H,0
2010 0
.75'T r1 L_r* r
I02% l/min
Pa 02 mm Hg
FLOW l/min
.50 .25 0 200 150 100 50 0 3.0 2.0 1.0 0
'1-
'*\Y^**\^**-*
'-•-•-•••••y-*^
V
,...-/\A..^V'
v
JUL
4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96
ON ECMO
HOURS
F F E°C M 0
Fig. 2. Platelet count, Pa,l2, ventilator settings, and extracorporeal flow rate during extracorporeal membrane oxygenation (ECMO). The initial decline in the platelet count can be accounted for by hemodilution with the priming solution. PEEP, Positive end-expiratory pressure. was greater than 50 mm. Hg preoperatively, and it was anticipated that the occlusion of the right pulmonary artery for the transplantation procedure might further elevate pulmonary artery pressue, overload the right ventricle, and result in cardiac failure and a fall in cardiac output. It was for this reason that the provision for rapid conversion to venoarterial perfusion was made. In actual fact, clamping of the right pulmonary artery was well tolerated and did not result in the further elevation of the mean pulmonary artery pressure. Thus the venovenous mode was used exclusively. No attempt was made to re-establish the bronchial artery blood supply as part of the transplantation procedure, and thus the blood supply to the transplanted bronchus came from collateral flow from the pulmonary artery. The venovenous mode of perfusion directly increased the oxygen content in the pulmonary artery blood. This may have been of some value in improving the oxygenation of the transplanted bronchus. Although lung transplantation remains an experimental procedure with major problems yet to be solved, our experience leads us to conclude that postoperative support with a membrane oxygenator can play a useful role in the management of future transplant recipients.
We wish to thank Mrs. Bridget Martin, Mr. Stanley Gregory, and Mr. Ted Kelley for their invaluable technical assistance.
REFERENCES 1 Veith FJ, Koerner SK: The present status of lung transplantation. Arch Surg 109:734-740, 1974 2 Liddicoat JE, Bekassy SM, Beall AC, Glaeser DH, De Bakey ME: Membrane versus bubble oxygenator. Clinical comparison. Ann Surg 181:747-753, 1975 3 Duffin J, Martin B, Cooper JD: Control, monitor and alarm system for clinically applied membrane oxygenator. Can Anaesth Soc J 23:143, 1976 4 Kolobow T, Zapol WM: A new thin-walled non-kinking catheter for peripheral vascular cannulation. Surgery 68:625, 1970 5 Cooper JD, Duffin J, Zapol WM: Cannulation of ascending aorta for long-term membrane oxygenator support. J THORAC CARDIOVASC SURG 69:905-908, 1975
6 Bartlett RH, Gazzaniga AB, Fong SW, Jefferies MR, Roohk HV, Haiduc N: Extracoproreal membrane oxygenator support for cardiopulmonary failure. Experience in 28 cases. J THORAC CARDIOVASC SURG 73:375-386,
1977 7 Kolobow T, Stool EW, Sacks KL, Vurek GG: Acute respiratory failure. Survival following ten days' support
32
Nelems et al.
with a membrane lung. J THORAC CARDIOVASC SURG
69:947-953, 1975 8 Siegelman SS, Sinha SBP, Veith FJ: Pulmonary reimplantation response. Ann Surg 117:30-36, 1973 9 Veith FJ, Sinha SBP, Dougherty JC: Nature and evolution of lung allograft rejection with and without immunosuppression. J THORAC CARDIOVASC SURG 63:509-
520, 1972
The Journal of Thoracic and Cardiovascular Surgery
10 Birek A, Duffin J, Glynn MFX, Cooper JD: The effect of sulfinpyrazone on platelet and pulmonary responses to onset of membrane oxygenator perfusion. Trans Am Soc Artif Intern Organs 22:94-101, 1976 11 Osada H, Ward CA, Duffin J, Nelems JM, Cooper JD: Micro-bubble elimination during priming improves biocompatibility of membrane oxygenators. Am J Physiol: Heart and Circulatory Physiology (in press)
Information for authors Most of the provisions of the Copyright Act of 1976 became effective on January 1, 1978. Therefore, all manuscripts must be accompanied by the following statement, signed by each author: " T h e undersigned author(s) transfers all copyright ownership of the manuscript entitled (title of article) to The C. V. Mosby Company in the event the work is published. The author(s) warrants that the article is original, is not under consideration by another journal, and has not been previously published." Authors will be consulted, when possible, regarding republication of their material.
J. M. Nelems, M.D., J. Duffin, Ph.D., M. F. X. Glynn, M.D., J. Brebner, M.D., A. A. Scott, M . D . , and J. D. Cooper, M.D., Toronto, Ontario, Canada
A A u m a n lung transplantation remains in the developmental stage, with several basic problems yet to be solved. One such problem is the temporary malfunction of the transplanted lung during the initial few days after transplantation. This malfunction, termed the reimplantation response, may be refractory to maximal ventilatory support and result in the death of the recipient in the first week. A review of the previous clinical experience with lung transplantation 1 revealed that 43 percent of the recipients died within the first week, usually of respiratory failure. Although intraoperative cardiopulmonary support with a bubble oxygenator has been used for several lung transplants (F. J. Veith, personal communication), no such extracorporeal support has been attempted during the postoperative period because of the coagulopathy which inevitably develops after only several hours with a conventional bubble oxygenator. 2 However, recent improvements in the design and utilization of membrane oxygenators have permitted perfusion to be conducted safely for days. It has therefore been anticipated 1 that extracorporeal membrane oxyFrom Toronto General Hospital, Toronto, Ontario, Canada. This work supported in part by the Ontario Heart Foundation, Crant No. 1-27, the Medical Research Council of Canada, Grano No. Ma 5953, and the Toronto General Hospital Foundation. Received for publication Dec. 22, 1977. Accepted for publication March 10, 1978. Address for reprints: J. D. Cooper, M.D., 1-131 University Wing, Toronto General Hospital, Toronto, Ontario, Canada M5G 1L7. 28
genator (ECMO) perfusion might provide temporary postoperative support for lung transplant recipients during the initial period of transplant malfunction. We recently provided membrane oxygenator support for a lung transplant recipient for a period of 96 hours. This support enabled the patient to survive a severe reimplantation response. The patient was subsequently weaned successfully from all ventilatory assistance and became ambulatory. He died on the eighteenth postoperative day of bronchial dehiscence. This report outlines the technical aspects of membrane oxygenator perfusion for lung transplantation. Case report A 19-year-old man sustained burns over 30 percent of his body in a house fire. On admission to the hospital there was evidence of an airway burn. The patient remained on a ventilator for the next 5 months, during which a progressive deterioration of lung function and increasing hypercapnia developed. The PaCo2 rose to 120 mm. Hg despite a minute ventilation of 24 L.; oxygenation was less of a problem, with aPa 0;! of 85 mm. Hg on an Fi 0 , of 0.5. When the hypercapnia continued to worsen despite all therapy, it was decided to proceed with right lung transplantation. The patient was connected to a membrane oxygenator 5 hours prior to the actual operation to permit gradual correction of the hypercapnia and the respiratory acidosis. The operation was performed without difficulty, but despite an ischemic time of only 2 hours, 15 minutes, the transplanted lung failed to function satisfactorily immediately following the transplantation procedure. ECMO was continued for a total of 96 hours, by which time the transplanted lung had recovered sufficiently to support the patient. Mechanical ven0022-5223/78/0176-0028S00.50/0 © 1978 The C. V. Mosby Co.
Volume 76
ECMO support for lung transplantation
Number 1 July, 1978
29
tilatory support was required for thefirst7 postoperative days. On the tenth day, the patient was ambulatory and was weaned from all supplementary oxygen. He died on the eighteenth postoperative day of bronchial dehiscence associated with localized necrosis. No parenchymal lung infection was present. Details of perfusion A 3.5 sq. M. Sci-Med Kolobow lung* with a fillerfree silicone rubber surface was used for the perfusion. The circuit consisted of silicone rubber tubing throughout, with polycarbonate tubing connectors. The details of the control, monitor, and alarm systems have previously been reported.3 The cannulation scheme was designed to allow rapid alternation between venovenous and venoarterial perfusion modes as might be required. Fig. 1 illustrates the circuit in schematic form. Thin-walled, spring-wirereinforced polyurethane cannulas, as initially described by Kolobow and Zapol,4 were used. The inferior vena cava was cannulated to the level of the diaphragm via the right common femoral vein, and the distal femoral vein was drained with a separate, smaller catheter. The superior vena cava was cannulated via the right internal jugular vein. The aortic arch was cannulated retrograde via the right common femoral artery with a cannula 75 cm. in length and 6.35 mm. in diameter.5 A separate, smaller arterial cannula was inserted into the distal femoral artery for perfusion of the right leg. Venovenous bypass was begun and the flow rate was gradually increased to 3 L. per minute. The membrane lung was ventilated with 5 percent carbon dioxide in oxygen to prevent too rapid a removal of carbon dioxide from the patient, and the carbon dioxide content of the effluent gas from the membrane lung was continuously monitored with a rapid carbon dioxide analyzer to measure the rate of carbon dioxide removal. After five hours of perfusion, the PaCo2 was 80 mm. Hg and the pH was 7.35. During this time 50 mEq. of potassium chloride was given intravenously. The venoarterial option for perfusion was provided in case venovenous perfusion proved inadequate during the transplantation operation. The venoarterial perfusion circuit was temporarily tested prior to removal of the right lung, and it performed well. With the tip of the aortic cannula near the left subclavian artery, oxygen saturation was measured in the right ear with an ear oximeter. Since the saturation rose to above 95 percent when venoarterial perfusion was begun, it was not believed necessary to advance the cannula into the ascending aorta. After several minutes of venoarterial perfusion, the circuit was switched back to the *Sci-Med Life Systems, Inc., Minneapolis, Minn.
i
t
Fig. 1. Schematic diagram of the perfusion circuit. For venovenous perfusion, occluding clamps were placed at points v. For venoarterial perfusion, the clamps at points v were removed, and an occluding clamp was placed at point a. venovenous mode and the operation was begun. Despite a mean pulmonary artery pressure in excess of 50 mm. Hg, the patient's condition remained stable. This permitted the venovenous mode of perfusion to be used exclusively. During the transplantation procedure, there was no indication of unusual bleeding, and the amount of blood transfused (4 units of whole blood and 1 unit of packed cells) was appropriate for the nature of the surgery performed. Following completion of the transplantation and closure of the thoracotomy, an unsuccessful attempt was made to wean the patient from ECMO. The Po 2 fell to
3 0 Nelems et al.
38 mm. Hg at this point when flow was reduced to 1 L. per minute. ECMO via the venovenous route was therefore continued postoperatively. As venoarterial perfusion had not been required, the arterial cannulas were removed at the end of the transplantation procedure. Postoperatively, an average flow rate of 2 L. per minute was used. Since the patient had an estimated surface area of 1.75 sq. M., the flow rate was approximately 1.15 L. per minute per square meter. On the fourth day following transplantation, ECMO support was discontinued, the venous cannulas were removed, the venotomy sites were closed, and flow was restored in the femoral and jugular veins. Detailed examination of the membrane oxygenator surface indicated that less than 4 percent of its surface was covered with adherent thrombus. The remaining surface appeared entirely clean. Coagulation during ECMO was managed as follows: Prior to cannulation, a loading dose of 5,000 units of heparin was given intravenously, after which heparin was administered by continuous infusion. Hourly measurements of the activated clotting time were used to determine the heparin requirement. The activated clotting time, measured with the Hemochron apparatus,* was maintained between 120 and 150 seconds. The average hourly dose of heparin during the first 8 hours of perfusion (the preoperative and intraoperative periods) was 212 units per hour. For the entire 96 hours of perfusion, the average rate of heparin administration was 620 units per hour. Fig. 2 gives the gas exchange data and the platelet counts for the entire period of ECMO support. The decline in the platelet count during the initial hours of ECMO resulted almost entirely from hemodilution with the 2 L. of crystalloid prime. Thereafter, there was a gradual but continuing platelet loss, resulting in a platelet count of 50,000 per cubic millimeter at the time of ECMO termination. Platelet transfusion was not required. The fibrinogen titer and the thrombin time (with heparin neutralization) remained within normal limits, as did the reptilase time; thus there was no disturbance in the normal hemostatic mechanism. Discussion Although cardiopulmonary bypass with a bubble oxygenator has been used intraoperatively in previous lung transplantations, post-transplantation extracorporeal support has not been previously available because of the major hemostatic derangements resulting after several hours of perfusion with a bubble oxygen*InternationaI Technidyne Company, Edison, N. J.
The Journal of Thoracic and Cardiovascular Surgery
ator. However, improvements in the design and ultilization of membrane oyygenators have permitted safe perfusions for more than a week without significant alteration of hemostatic function.6, 7 Because of this experience, we were encouraged to believe that prolonged ECMO support might safely be used as an adjunct to lung transplantation. The reimplantation response is defined as the morphologic, radiographic, and functional changes which occur in a transplanted lung as the result of the surgical procedure itself, exclusive of any effects owing to graft rejection. The changes include the development of alveolar infiltrates evident on chest x-ray films and the impairment of alveolar ventilation and gas exchange. The changes reach a peak within 3 days and regress over the next 3 weeks.1' 8 - 9 On review of the 37 lung transplant operations previously performed,1 it was noted that 16 patients (43 percent) had died within the first week, primarily of respiratory insufficiency. Of this number, eight (21 per cent) failed to survive the first 24 hours. It seems likely that many of these early postoperative failures were due more to the reimplantation response than to the effects of rejection. In our own case, the patient probably would have died during the initial 24 hour posttransplant period without the availability of ECMO support. Prior to the start of ECMO, the patient was given 2 Gm. of sulfinpyrazone (Anturan) intravenously. This dose was based upon evidence from our laboratory10 showing a marked improvement in early platelet sparing when sheep were pretreated with intravenous sulfinpyrazone prior to the onset of membrane oxygenator perfusion. The membrane lung was flushed with carbon dioxide for 2 hours before being primed with lactated Ringer's solution, which had been degassed for 1 hour in a vacuum chamber to lower its gas content. This priming technique is based upon experiments from our laboratory11 which have demonstrated improved platelet sparing when gas nuclei, trapped in the surface roughness of the membrane, are removed in the priming process. The bypass circuit was designed to permit conversion between the venovenous and venoarterial modes of perfusion (Fig. 1). In general, we prefer the venovenous mode, because it is safer, technically simpler to control, does not interfere with the patient's cardiovascular function, and assures that the oxygenated blood is distributed equally to the entire body. The recipient had excellent cardiac function, so that left ventricular support with the venoarterial bypass mode was not required. However, the mean pulmonary artery pressure
Volume 76 Number 1 July, 1978
ECMO support for lung transplantation
3 I
200,000 PLATELET 150,000 COUNT
100,000
\
50,000 h 0 PEEP cm H,0
2010 0
.75'T r1 L_r* r
I02% l/min
Pa 02 mm Hg
FLOW l/min
.50 .25 0 200 150 100 50 0 3.0 2.0 1.0 0
'1-
'*\Y^**\^**-*
'-•-•-•••••y-*^
V
,...-/\A..^V'
v
JUL
4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96
ON ECMO
HOURS
F F E°C M 0
Fig. 2. Platelet count, Pa,l2, ventilator settings, and extracorporeal flow rate during extracorporeal membrane oxygenation (ECMO). The initial decline in the platelet count can be accounted for by hemodilution with the priming solution. PEEP, Positive end-expiratory pressure. was greater than 50 mm. Hg preoperatively, and it was anticipated that the occlusion of the right pulmonary artery for the transplantation procedure might further elevate pulmonary artery pressue, overload the right ventricle, and result in cardiac failure and a fall in cardiac output. It was for this reason that the provision for rapid conversion to venoarterial perfusion was made. In actual fact, clamping of the right pulmonary artery was well tolerated and did not result in the further elevation of the mean pulmonary artery pressure. Thus the venovenous mode was used exclusively. No attempt was made to re-establish the bronchial artery blood supply as part of the transplantation procedure, and thus the blood supply to the transplanted bronchus came from collateral flow from the pulmonary artery. The venovenous mode of perfusion directly increased the oxygen content in the pulmonary artery blood. This may have been of some value in improving the oxygenation of the transplanted bronchus. Although lung transplantation remains an experimental procedure with major problems yet to be solved, our experience leads us to conclude that postoperative support with a membrane oxygenator can play a useful role in the management of future transplant recipients.
We wish to thank Mrs. Bridget Martin, Mr. Stanley Gregory, and Mr. Ted Kelley for their invaluable technical assistance.
REFERENCES 1 Veith FJ, Koerner SK: The present status of lung transplantation. Arch Surg 109:734-740, 1974 2 Liddicoat JE, Bekassy SM, Beall AC, Glaeser DH, De Bakey ME: Membrane versus bubble oxygenator. Clinical comparison. Ann Surg 181:747-753, 1975 3 Duffin J, Martin B, Cooper JD: Control, monitor and alarm system for clinically applied membrane oxygenator. Can Anaesth Soc J 23:143, 1976 4 Kolobow T, Zapol WM: A new thin-walled non-kinking catheter for peripheral vascular cannulation. Surgery 68:625, 1970 5 Cooper JD, Duffin J, Zapol WM: Cannulation of ascending aorta for long-term membrane oxygenator support. J THORAC CARDIOVASC SURG 69:905-908, 1975
6 Bartlett RH, Gazzaniga AB, Fong SW, Jefferies MR, Roohk HV, Haiduc N: Extracoproreal membrane oxygenator support for cardiopulmonary failure. Experience in 28 cases. J THORAC CARDIOVASC SURG 73:375-386,
1977 7 Kolobow T, Stool EW, Sacks KL, Vurek GG: Acute respiratory failure. Survival following ten days' support
32
Nelems et al.
with a membrane lung. J THORAC CARDIOVASC SURG
69:947-953, 1975 8 Siegelman SS, Sinha SBP, Veith FJ: Pulmonary reimplantation response. Ann Surg 117:30-36, 1973 9 Veith FJ, Sinha SBP, Dougherty JC: Nature and evolution of lung allograft rejection with and without immunosuppression. J THORAC CARDIOVASC SURG 63:509-
520, 1972
The Journal of Thoracic and Cardiovascular Surgery
10 Birek A, Duffin J, Glynn MFX, Cooper JD: The effect of sulfinpyrazone on platelet and pulmonary responses to onset of membrane oxygenator perfusion. Trans Am Soc Artif Intern Organs 22:94-101, 1976 11 Osada H, Ward CA, Duffin J, Nelems JM, Cooper JD: Micro-bubble elimination during priming improves biocompatibility of membrane oxygenators. Am J Physiol: Heart and Circulatory Physiology (in press)
Information for authors Most of the provisions of the Copyright Act of 1976 became effective on January 1, 1978. Therefore, all manuscripts must be accompanied by the following statement, signed by each author: " T h e undersigned author(s) transfers all copyright ownership of the manuscript entitled (title of article) to The C. V. Mosby Company in the event the work is published. The author(s) warrants that the article is original, is not under consideration by another journal, and has not been previously published." Authors will be consulted, when possible, regarding republication of their material.