Pulmonary venous anastomosis in lung transplantation without donor left atrium

J THoRAc CARDIOVASC SURG 1989;97:582-6

Pulmonary venous anastomosis in lung transplantation without donor left atrium Experimental and clinical results Single lung transplantation now is a therapeutic option for some patients with end-stage lung disease. Cyclosporine immunosuppression and refinements in bronchial anastomosis have been responsible for recent successes. Since 1953, the usual pulmonary venous anastomosis, both in experimental animals and in humans, has been an atrium-to-atrium cormection. This technique may limit the availability of usable donor lungs, since the donor heart, along with the atrium, is usually harvested for another recipient. Although techniques can he developed to aUow both transplant teams to harvest atrial tissue, this study was undertaken to determine if, in fact, anastomosis with donor left atrium is necessary. Twenty-four dogs were anesthetized and a left tboracotomy performed. After heparinization (3 mgjkg), the pulmonary artery and left atrium were occluded. One of four different pulmonary venous anastomoses was performed at 3.5x magnification: superior pulmonary vein end to end (group I), inferior pulmonary vein end to end (group II), superior pulmonary vein implantation into left atrium (group III), and left atrium-to-Ieft atrium anastomosis as control (group IV). Everting mattress sutures of 7-0 polypropylene were used in groups I, II, and ill and 6-0 in group IV. Average crossclamp time for group I, group II, and group IV was 20 minutes. The average crossclamp time for group ill was 10 minutes. All anastomoses were patent at the time of I-week reevaluation. Gross and microscopic examination demonstrated establishment of an intimal lining; organized nonocclusive thrombus was present in only one anastomosis. We conclude that atrium-to-atrium anastomosis is not necessary for a successful single lung transplantation, and that transplantation of a single lobe is feasible. The best alternative is implantation of the pulmonary vein into the left atrium, which will easily aUow use of the heart and both lungs from a single donor to different recipients. We have used this anastomosis in one patient without difficulty.

Joseph LoCicero III, MD, Shou-Ren Shih, MD, Fengrui Zhao, MD, James W. Frederiksen, MD, Renee S. Hartz, MD, and Lawrence L. Michaelis, MD,

Chicago, Ill.

Lung transplantation was first described in the dog model in 1950. 1 Since then, many advances have occurred primarily related to the use of cyclosporine for rejection and the wrapping of the bronchial anastomosis with omentum to improve bronchial healing." 3 These two major steps have allowed successful clinical lung transplantation.' However, there are steps in the proceFrom the Department of Surgery, Northwestern University Medical School, Chicago, Ill. Received for publication April 6, 1988. Accepted for publication Oct. 13, 1988. Address for reprints: Joseph LoCicero III, MD, 303 East Chicago Ave., Ward Building 9-105, Chicago, IL 60611.

582

dure of lung transplantation that have not been reevaluated since their original development in the early 1950s. One of these is the pulmonary venous anastomosis. It has been accepted that vein-to-vein anastomosis cannot be performed in clinical lung transplantation because of the threat of thrombosis.' These experiments were done on dogs by means of the surgical technique of the day. Since the establishment of this method, major improvements have been made in suture material and routine use of microsurgical techniques has been adopted. Still, to the present day, an atrial cuff has been used in all clinical single lung transplantations of which we are aware. Although details of combined heart/lung harvest from the same donor with adequate atrial cuff

Volume 97 Number 4

Lung transplantation 5 8 3

April 1989

Fig. 1. Vein-to-vein anastomosis at necropsy. Specimen divided longitudinally with anastomosis exposed. No evidence of thrombosis.

for both organs can be developed, the requirement for atrium in single lung transplantation may be based on outdated surgical technique. The specific aim of this endeavor was to determine if the pulmonary venous anastomosis can be performed without an atrial cuff with the use of monofilament vascular suture and microsurgical technique. This would allow both lungs, as well as the heart, to be harvested from a single donor. Methods and materials Twenty-four dogs weighing between 20 and 40 kg (average 28 kg) underwent left thoracotomy after an overnight fast. Anesthesia was induced and maintained with pentobarbital infusions. Conventional ventilation (tidal volume of IO nil/kg with a positive end-expiratory pressure of 5 em H 20 ) was delivered through a standard endotracheal tube by a volume ventilator. The pericardium was incised to expose the left atrial appendage and the insertion of both the superior and inferior pulmonary veins. Minimal dissection was performed on the bronchus and pulmonary artery. After intravenous heparinization (3 rug/kg), the pulmonary artery and then the left atrium were clamped. At this point, one of four procedures was performed. In group I (six animals) the superior pulmonary vein (average diameter 3.1 mm) was divided and reconstructed end to end. In group II (six dogs) the inferior pulmonary vein (average diameter 4.8 mm) was divided and reconstructed end to end. In both groups a running everting mattress suture of 7-0 polypropylene was inserted under 3.5X magnification. In group III (six dogs) the superior pulmonary vein (average diameter 2.9 mm) was divided and the atrial side of the vein closed with 6-0 polypropylene. A 3 mm incision was made in the left atrial appendage and the vein reimplanted into the appendage with a running everting mattress suture of 7-0 polypropylene. In the control group

(group IV) a cuff of left atrium containing both pulmonary veins was divided from the body of the left atrium and then resewn with an everting 6-0 polypropylene suture. At the completion of the anastomosis, the atrial clamp was removed followed by the pulmonary artery clamp. In one atrium-to-atrium anastomosis, the clamps were reapplied to oversew a 2 mm gap in the suture line. Visual and photographic assessment was then made of each anastomosis. Patency of the anastomosis was assessed by alternate filling and emptying of the pulmonary veins during the cardiac cycle and by "milking" of blood through the anastomosis. After patency was established, the thoracotomy incision was closed in layers over a temporary chest tube, which was removed at the conclusion of the procedure. The heparin was not reversed. The dogs were allowed to slowly recover. When breathing spontaneously, they were extubated. When fully awake, they were returned to standard housing and observed for I week. All animal care was performed in accordance with the standards established by the National Institutes of Health and the Department of Agriculture on the Care and Use of Experimental Animals. At the conclusion of 1 week, necropsies were performed. patency was again assessed at this time. The entire specimen was then excised and formalin fixed. Microscopic sections of the endothelial surface stained with hematoxylin and eosin were examined.

Results

All animals survived the operation and the observation period. One animal in each group, at the time of initial operation, had a small leak that required the placement of a hemostatic stitch. One additional animal in group II had a postanastomotic leak large enough to necessitate reclamping of the atrium and vein to effect

The Jou rnal of

584

loCicero et al.

Thor acic and Cardiovascula r Surgery

Fig . 2. Vein-to-atr ium anastomosis at necropsy. Vein is widely paten t with anastomosis well inside atrial cavity. No thrombosi s present.

Fig. 3. Atrium-to-atrium anastomosis at necropsy. Suture line is intact without thrombus forma tion.

the repair. At necropsy, this animal had a near totally obstructing thrombus that originated at the site of the atrial clamping. In all dogs, the postmortem measurement of the anastomosis showed no narrowing. In group I the vein that required repair had a 1 mm thrombus at the site of repair. All others were widely patent (Fig. 1). Similar results were noted in group II. Histologic evaluation of both groups showed eversion of the cut edges of the vein. Small deposits of platelets and white cells were noted partially filling these everted spaces and reconstituting the endothelial surface . All anastomoses in group III were widely patent.

Grossly, the vein edge appeared to be implanted well inside the atrium (Fig. 2). The healing suture line was free of thrombus and smooth. Microscopic evaluation showed a more organized endothelial surface than those of groups I and II. Control anastomoses (left atrium to left atrium) also were free of thrombus and showed a well-healing suture line (Fig. 3). The endothelial surface was histologically similar to that in group III , with a well-organized platelet plug and pseudoendothelial lining. After completion of this work, there was an opportunity to perform a human lung tran splant operation in our ongoing program. The technique was the same as

Volume 97 Number 4 April 1989

Lung transplantation

that used by Cooper and co-workers" with the exception of the pulmonary venous anastomosis. A summary of that patient's course follows.

585

2 3

Case report A 62-year old man with a 4-year history of familial pulmonary fibrosis had increasing shortness of breath. His mother had died of the same condition several years previously. He could still conduct his activities at a decreased capacity while receiving 2 L of nasal oxygen. A previous trial of corticosteroids was of no benefit. Results of pulmonary function tests were as follows: forced vital capacity, 3.551 (60% of predicted); l-second forced expiratory volume, 2.491; diffusion capacity, 4% of the predicted value. Evaluation of the heart, including right heart catheterization and nuclear gated studies, showed no impairment. He was accepted as a single lung transplant candidate and placed on the national and local waiting lists. During the waiting period, the patient had complications of a peptic ulcer necessitating a transfusion of 4 units of blood. After this episode, he required increased nasal oxygen to 6 L/min to maintain function around the home. Eleven months after he was accepted as a potential candidate, an appropriate donor 120 miles away was identified. The left lung had potential contusions and thus the right lung was chosen. The organs procured from the donor included the lungs, heart, liver, kidneys, and cornea. Because the harvesting cardiac team was reluctant to share the left atrium on the right side, the pulmonary veins were divided intrapericardially as close to the atrium as possible. The lung was preserved by cold storage and transplanted into the recipient. The pulmonary veins were implanted separately into the left atrium as described herein. The artery and the bronchus were anastomosed end to end. Omentum was used to wrap the bronchial anastomosis. The operation time was 6 hours with an ischemic time of 4 hours 45 minutes. Initial blood flow to the transplant lung, as assessed by nuclear perfusion scan on the day of the operation, was 50%. At I week the transplant perfusion as assessed by nuclear perfusion scan (Fig. 4) was 74%, and by 3 weeks it was 85%. Multiple episodes of sepsis prevented the patient from being extubated, and ultimately he died at 6 weeks of systemic candidiasis. Perfusion scans showed excellent function throughout his entire posttransplant course. Although many attempts were made, autopsy permission was refused.

Discussion Although Demikhov claimed post facto to have performed the first lung transplantation,' Metras' is usually credited with demonstrating in the dog model in 1950 that single lung transplantation is feasible. He anastomosed the left atrium to the left atrium as it is performed clinically today. Juvenelle and colleagues" confirmed this work in 1951 by sequentially dividing the veins and anastomosing each in turn. However, in 1952 Hardin, Kittle, and Schafer? noted a high prevalence of intravascular thrombosis with this method. By 1953

REG I ON

1

2 3

PI XELS

COU N TS

2 59 .. 9 0 1001 7 9

381

56 9 7 0 .. 15

Fig. 4. Quantitative anterior perfusion lung scan I week after right lung transplant. Most of perfusion (74%) is seen in right lung.

Neptune, Wheller, and Bailey' proposed that anastomosis of the left atrium instead of direct venous connections would result in the greatest success because of larger size, tougher, and easier to handle tissue. This established the left atrial cuff anastomosis as the experimental tissue standard, and it is the technique currently used in the clinical successes today. Many factors led to this change in surgical technique. The common suture technique of the day was interrupted mattress, 4-0 silk, or Dacron polyester suture. These materials were intrinsically more thrombogenic than the present finer 6-0 monofilament polypropylene now used for small-vessel anastomosis. The canine pulmonary veins were indeed thin and friable. The left atrium afforded a much tougher tissue for the techniques of that time. Magnification was not used, which may have promoted stenosis of the anastomoses occasionally seen even in skilled hands.'? Heparinization also was not routinely used to prevent early clot formation and appeared, in some cases, to cause more harm than good." However, Blumenstock and Kahn" raised the issue of heparinization for the recipient, which remains unsettled today. Some centers performing human transplants do not heparinize the recipient, whereas we have routinely employed this technique. In addition, Huggins" emphasized the importance of everting the edges of all canine vascular anastomoses to prevent thrombus formation. This effectively decreases the lumen size of any anastomosis and prompted Benfield and Coon! 3 to suggest that many of the problems encountered in pulmonary flow were due solely to the size of the animals (10 to 20 kg dogs) and that such problems might be minimized with larger animals.

The Journal of Thoracic and Cardiovascular Surgery

5 8 6 LoCicero et al.

Veith" has pointed out that appropriate lung donors are scarce. The major reasons for this scarcity include the fact that lungs are one of the more commonly injured organs in blunt trauma. Colonization with bacteria in patients with prolonged intubation and ventilation precludes donation. In addition, size match is of great importance so that pleural space problems do not lead to future restrictive disease." This limit in the number of available donor organs is further diminished by the reluctance of many cardiac transplant teams to share the left atrium. Although Veith has described the surgical technique for sharing the atrium, preservation of both organs may be affected by the increased delicate dissection. We have shown that pulmonary vein-left atrium anastomosis in humans is technically successful with acceptably short ischemic times. Finally, as suggested by Huggins," successful separate pulmonary venous anastomoses will pave the way for transplantation of less than a total lung. Now that other serious problems have been managed, donation from a living, related subject may not only be feasible but desirable. There now appears to be no roadblock to such an endeavor. REFERENCES 1. Metras H. Note preliminary sur la graffe totale du pumon chez Ie chien. CR Acad Sci (Paris) 1950;231:1176-7. 2. Goldberg M, Cooper JD, Lima 0, et aI. A comparison between Cyclosporine-A and methylprednisone plus azathioprine on bronchial healing following canine lung

autotransplantation. J THORAC CARDIOVASC SURG 1983; 85:821-6.

3. Morgan WE, Lima 0, Goldberg M, Ayabe H, Ferdman A, Cooper JD. Improved bronchial healing in canine left lung reimplantation using omental pedicle wrap. J THORAC CARDIOVASC SURG 1983;85:134-49. 4. Toronto Lung Transplant Group. Unilateral lung transplantation for pulmonary fibrosis. N Engl J Moo 1986;314:1140-5.

5. Neptune WB, Weller R, Bailey CPo Experimental lung transplantation. J THORAC SURG 1953;26:275-89. 6. Cooper JD, Pearson FG, Patterson GA, et aI. Technique of successful lung transplantation in humans. J THORAC CARDIOVASC SURG 1987;93:173-81. 7. Perelman MI, Rabinovich 11. Methods and technique of experimental autotransplantation of the lung. J THORAC CARDIOVASC SURG 1970;59:275-82. 8. Juvenelle AA, Citret C, Wiles CE, Stewart JD. Pneumonectomy with replantation of the lung in the dog for physiologic study. J THORAC SURG 1951;21:111-5. 9. Hardin CA, Kittle CF, Schafer PW. Preliminary observations on homologous lung transplants in dogs. Surg Forum 1952;3:374-84. 10. Trummer MJ. Experimental transplantation of the lung. Ann Thorac Surg 1965;1:203-19. 11. Blumenstock DA, Kahn DR. Replantation and transplantation of the canine lung. J Surg Res 1961;1:40-7. 12. Huggins CEo Reimplantation of lobes of the lung: an experimental technique. Lancet 1959;2:1059-62. 13. Benfield JR, Coon R. The role of the left atrial anastomosis in pulmonary reimplantation. J THORAC CARDIO· VASC SURG 1967;53:676-84. 14. Veith FJ. Discussion of Cooper et aI.7