- Email: [email protected]
Normal bronchial healing without bronchial wrapping in canine lung transplantation
Normal Bronchial Healing Without Bronchial Wrapping in Canine Lung Transplantation Joseph S. Auteri, MD, Valluvan Jeevanandam, MD, Juan A. Sanchez, MD, Charles C. Marboe, MD, Thomas J. Kirby, MD, and Craig R. Smith, MD Departments of Surgery and Pathology, Columbia Presbyterian Medical Center, New York, New York
-
The deleterious effect of steroids on bronchial healing in lung transplantation has led to the development of techniques to protect the anastomosis and to the exclusion of steroid-dependent patients from transplantation. The effect of steroids on bronchial healing was tested in a canine single-lung allotransplantation model. Twenty size-matched mongrel dogs (20 to 30 kg) underwent left lung transplantation without anastomotic wrap or direct revascularization. Postoperatively, all received daily doses of cyclosporine (15 mg/kg) and azathioprine (1 mglkg) and were subdivided into three steroid dosage groups. Group A (n = 10) animals received 1.5 mgkg of prednisone per day whereas groups B (n = 5) and C (n = 5) received 5.0 mg/kg of prednisone per day for 28 postoperative days. In addition, group C received pred-
nisone (5.0 mg kg-' * day-') for 1 month preoperatively. In group A, 8 of 10 dogs survived 28 days without evidence of respiratory compromise, with anastomotic bursting pressure greater than 510 mm Hg. In group B, all 5 dogs survived to 28 days without evidence of respiratory compromise and with intact bronchial anastomoses (bursting pressures greater than 510 mm Hg). In group C, 3 of 5 animals survived to 28 days with intact anastomoses. Histological examination demonstrated normal bronchial healing in all anastomoses. These data suggest that preoperative steroid dependence should not be a contraindication to lung transplantation and that bronchial anastomotic wrapping with vascular tissue may not be essential. (Ann Thorac Surg 1992;53:804)
I
we reexamined bronchial anastomotic healing in the presence of conventional as well as suprapharmacologic doses of corticosteroids.
n contrast to transplantation of other solid organs, single- and double-lung transplantation continue to be plagued with technical as well as immunologic problems. Single-lung transplantation in humans has been hampered by bronchial anastomotic complications, primarily considered a consequence of donor bronchus devascularization by the interruption of the bronchial arterial supply. Attempts at increasing bronchial blood flow by direct revascularization have met with limited success [l]. Wrapping the anastomosis with omentum has been shown to recruit arterial supply as early as the fourth postoperative
For editorial comment, see page 5. day in experimental models but increases operative time and makes the procedure technically more complex [2]. The deleterious effects of steroids and other immunosuppressive agents on wound healing have been implicated as a cofactor in dehiscence. This has led to empiric exclusion of steroid-dependent patients as candidates for lung transplantation and has led to a modification in early maintenance immunosuppression, substituting antithymocyte globulins for steroids. Experimental data to support these views are scant. Therefore, in a canine singlelung allotransplant model, without bronchial wrapping, Presented at the Twenty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Feb E-20, 1991. Address reprint requests to Dr Auteri, Department of Surgery, Box 91, Columbia-Presbyterian Medical Center, 622 West 168th St, New York, NY 10032.
0 1992 by The Society of Thoracic Surgeons
Material and Methods Under general endotracheal anesthesia, through a left thoracotomy, 20 mongrel dogs (weight, 20 to 25 kg) underwent left single-lung allotransplantation from sizematched donors. All animals received humane care in compliance with the "Guide for the Care and Use of Laboratory Animals" (NIH publication No. 85-23, revised 1985). All transplants were standardized to a 45-minute ischemic time in 4°C normal saline solution. Donor lungs were neither heparinized nor perfused. Lungs were preserved by submersion in normal saline solution at 4°C. The atrial anastomosis, an inverting continuous horizontal mattress suture, was performed first, followed by the pulmonary artery and bronchus anastomoses. All three anastomoses were performed using running 4-0 polypropylene suture. No attempt was made to cover, wrap, or revascularize the bronchial anastomosis. All bronchial anastomoses were immediately submerged in saline solution and tested for competence by forced ventilation through a cuffed endotracheal tube. The lung was then reexpanded, the chest incision was closed, and the dogs recovered. Postoperatively, all dogs received daily immunosuppression with cyclosporine (15 mg/kg) and azathioprine (1 mgkg) and were subdivided into three steroid dosage groups. Group A (n = 10) received 1.5 mg/kg of prednisone daily. Groups B (n = 5) and C (n = 5) received 0003-4975/92/$3.50
AUTERIETAL BRONCHIAL HEALING IN LUNG TRANSPLANTATION
Ann Thorac Surg 1992;53:W
Fig I . Preparation for measurement of bursting pressure.
5.0 mg/kg of prednisone daily. In addition, group C animals received prednisone (5.0 mg * kg-’ * day-’) for 1 month preoperatively. All survivors were sacrificed at 28 days. At autopsy, the bronchus was excised from 2 cm proximal to 2 cm distal to the anastomosis. The ends were cross-clamped and an 18-gauge Silastic (Dow Corning, Midland, MI) catheter was introduced into the lumen (Fig 1). Normal saline solution stained with methylene blue dye was injected into the lumen of the bronchus as intraluminal pressure was transduced to an oscilloscope. The wave form was quantitated in millimeters of mercury and calibrated before each analysis. The upper limit of measurement in this system was 510 mm Hg. The bursting pressure was the point at which methylene blue dye could be identified externally. Those that achieved the upper limit without bursting were recorded as bursting pressure greater than 510 mm Hg. To determine the percent stenosis, the intraluminal radius was measured at the anastomosis (Anast) and at 2 cm proximal to the anastomosis (Prox). Percent stenosis was calculated according to the following formula: % Stenosis = [ l - (Anast%’rox2)]
x 100.
Histologic analysis of anastomoses was done after hematoxylin, phloxine, and saffron staining of paraffin section. Lung parenchyma was sectioned, and histologic characteristics were classified according to the International Society for Heart Transplantation Standardization of Nomenclature in the Diagnosis of Lung Rejection [31.
Results Intraoperatively, all anastomoses were airtight during a sustained inflation to 40 mm Hg. Higher pressures were
81
avoided to prevent parenchymal lung injury. In group A, 8 of 10 dogs survived 28 days without evidence of respiratory compromise. These animals had bursting pressures greater than 510 mm Hg. Two dogs died 6 and 9 days postoperatively as a result of left atrial thrombosis. These animals demonstrated intact bronchial anastomoses at autopsy (bursting pressures of 303 and 475 mm Hg, respectively). In group B, all 5 dogs survived to 28 days without evidence of respiratory compromise and with intact bronchial anastomoses (bursting pressures greater than 510 mm Hg). In group C, 3 of five animals survived 28 days. The 3 survivors had intact anastomoses at autopsy (bursting pressures greater than 510 mm Hg). One dog exsanguinated from a gastroduodenal ulcer on the first postoperative day and bursting pressure was not obtained. The other died on the thirteenth postoperative day of congestive heart failure resulting from an intraoperative myocardial infarction sustained during pulmonary artery injury. This animal had a bronchial anastomotic bursting pressur of 185 mm Hg. Anastomotic histology was considered normal if regenerated epithelium was observed lining the lumen with normal scar formation and healing fibrosis deep to the anastornotic site. This was seen in all anastomoses, including all three with bursting pressures less than 510 mm Hg, and in 3 animals with mild, moderate, and severe rejection as graded by the International Society for Heart Transplantation Standard of Nomenclature [3]. The animal with severe (A4a) rejection, despite extensive bronchiolitis obliterans, had a bursting pressure greater than 510 mm Hg with no evidence of bronchial dehiscence. No manifestations of clinically significant stenosis, such as stridor, respiratory distress, or atelectasis, were detected in any animal. In groups B and C .he average stenosis, as calculated by the formula in the Material and Methods section, was 19.1% (range, 14.3%to 20.3%).
Comment Concern over the fate of the healing bronchial anastomosis in lung transplantation has existed since investigation in this arena began [4, 51. Of the first 38 clinical lung transplantations done in the world, only 16 patients were alive beyond 10 days. Of these 16 patients, 10 died of bronchial complications within 7 months [5]. This experience prompted extensive research into the etiologic factors associated with bronchial dehiscence in lung transplantation. Early contributions to improve bronchial healing included wrapping the anastomosis with perihilar connective tissue [6], telescoping the bronchial ends [7], shortening the length of the donor bronchus [8], and bronchial artery reanastomosis [l,91. Particularly for the latter, which remains a difficult technical problem, application of these advances produced marginal improvement in results. This led to the landmark work of Cooper and his colleagues in Toronto, who demonstrated in a canine autotransplant model that omentum can bring saturated blood to the anastomosis in the early postoperative period [2, 101. Related studies suggested a deleterious effect of
82
AUTERIETAL BRONCHIAL HEALING IN LUNG TRANSPLANTATION
methylprednisolone on bronchial healing when compared with a cyclosporine group or with nonimmunosuppressed controls [ll, 121. This experimental work led directly to the first successful series of single-lung transplants [13] and established three principles of clinical practice: (1)to wrap all bronchial anastomoses with omentum, (2) to avoid postoperative steroids, and (3) to deny transplantation to patients dependent on steroids for treatment of their underlying lung disease. There are serious disadvantages to these principles. Mobilizing omentum requires a laparotomy, which adds the risk of pulmonary and other complications in an immunocompromised host. Omentum must be passed through a diaphragmatic defect, which can become a clinically significant hernia [2]. When maintenance immunosuppression is modified by substitution of antithymocyte globulins for steroids, the risk of viral and fungal infection is increased. Finally, many patients who might otherwise be good candidates for lung transplantation are steroid dependent and are denied lung transplantation until withdrawn from steroid treatment. For these reasons, this study was undertaken to reexamine these concepts. This study differs in several respects from the important work by Cooper and his colleagues in Toronto. In one series of experiments they compared anastomoses with omental wrapping to those without and reported that wrapped anastomoses had improved gross and histologic appearance, reduced bronchial stenosis, and neovascularity arising from the omentum demonstrated on celiac arteriography [2]. However, because arteriography was not performed in the unwrapped animals, the amount of neovascularity achieved without omental wrap cannot be determined. The present study was not designed to compare wrapped with unwrapped animals, but the histologic appearance and percent stenosis of the unwrapped anastomoses in all three steroid dosage groups was consistent with normal healing. In the present study, bursting pressure was measured to quantitate the strength of the anastomosis, whereas tensile strength was measured in the Cooper studies. Tensile strength reflects resistance to disruption by linear forces, as if the lung were simply a weight hanging from the anastomosis, and was found to be decreased by steroids [ll]. Linear forces are probably minimal in a lung that fills the hemithorax. Our use of bursting pressure for measurement of the integrity of the anastomosis reflects our dissatisfaction with the tensile strength method, which measures only linear forces. Bursting pressure reflects resistance to disruption by radial forces generated within the lumen and is analogous to the airway pressure generated during cough or positive pressure ventilation, perhaps a more clinically relevant measurement. Data to support this contention are not currently available. In the present study the maximum intraluminal pressure applied (>510 mm Hg) is far in excess of the maximum physiologic pressure of 100 to 120 mm Hg [14] and was accepted as evidence of excellent anastomotic strength. The 3 animals with bursting pressures less than 510 mm Hg still had bursting pressures well above the usual physio-
Ann Thorac Surg 19!%!;538C4
logic range (185, 303, and 475 mm Hg). The animal with the lowest bursting pressure measured (185 mm Hg) was sacrificed after 13 days of low cardiac output following intraoperative myocardial infarction. In a limited cross-comparison between bursting pressure and tensometry, 3 animals with bursting pressures greater than 510 mm Hg were found to have a bronchial tensile strength (1,064 f 282 g) that correlated with those considered to be in the satisfactory range in the Cooper study (1,232 f 101 g) [12]. Finally, it is particularly noteworthy that bronchial dehiscence was not seen in any animal in this allograft study or in the Cooper autograft studies, and all animals in the latter survived without incident to sacrifice at 23 days, regardless of histologic or arteriographic findings [2]. None of the three deaths in this study was related to bronchial healing. Satisfactory healing was seen even in the presence of severe pulmonary rejection and bronchiolitis obliterans. In summary, the data in this canine allograft model suggest that steroid immunosuppression is compatible with excellent bronchial healing, even at higher than conventional dosage. Preoperative steroid administration, at sufficiently high dosage levels to produce gastrointestinal ulceration and bleeding in more than 1experimental animal, did not have deleterious effects on bronchial healing. Finally, excellent healing was obtained without wrapping the anastomoses with omentum or with any other material. It would be hazardous to conclude that bronchial healing is no longer a problem in lung transplantation, as recent experience has demonstrated [15]. Although previous studies may have overemphasized the detrimental effects of steroid, or the importance of omental wrapping, the bronchial anastomosis is undeniably hypovascular unless bronchial arteries are rejoined [16]. In a clinical allograft, lung preservation, ischemic time, rejection, infection, and the hemodynamic stability of the recipient undoubtedly play a role in bronchial healing, and it will always be difficult to isolate the contribution of each variable. Vascular wraps are still receiving appropriate experimental attention [17], and attempts are being made to simplify bronchial artery anastomosis [18]. In addition, the advantages of a telescoping anastomosis have been reported especially in patients receiving steroids [191. Continued evolution of the basic principles of bronchial healing can be expected.
References 1. Pinsker KL, Montefusco CM, Kamholz SL, Hagtstrom JWC, Gliedman ML, Veith FJ. Improved bronchial anastomotic healing secondary to maintenance or restoration of bronchial arterial circulation by microsurgical techniques. Surg Forum 1980;31:23&1. 2. Morgan E, 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;85:134-9. 3. Yousem SA, Berry GJ, Brunt EM, et al. Working formulation for the standardization of nomenclature in the diagnosis of
Ann Thorac Surg
1992:53804
heart and lung rejection: lung rejection study group. J Heart Transplant 1990;659%4. 4. Hardy JD, Webb WR, Dalton ML, Walker GR. Lung homotransplantation in man, report of the initial case. JAMA 1963;121065-74. 5. Veith FJ, Kamholz SL, Mollenkopf FP, Montefusco CM. Lung transplantation. Transplantation 1983;35:271-8. 6. Blumenstock DA, Kahn DR. Replantation and transplantation of the canine lung. J Surg Res 1961;l:U. 7. Sinha SBP, Doughterty JC, Boley SJ, Veith FJ. Elimination of bronchial complications in lung transplantation. Surg Forum 1971;22.22532. 8. Pinsker KL, Koerner SK, Kamhotz SL, Hagstrom JWC, Veith FJ. Effect of donor bronchial length on healing. J Thorac Cardiovasc Surg 1979;77:66%77. 9. Veith F, Blumenstock DA. Lung transplantation. J Surg Res 1971;11:3>9. 10. Lima 0,Goldberg M, Peters WJ, Ayabe H, Townsend E, Cooper JD. Bronchial omentopexy in canine lung transplantation. J Thorac Cardiovasc Surg 1982;83:418-27. 11. Lima 0,Cooper JD, Peters WJ, et al. Effects of methylprednisolone and azathioprine on bronchial healing following lung autotransplantation. J Thorac Cardiovasc Surg 1981;82 211-5.
AUTERIETAL BRONCHIAL HEALING IN LUNG TRANSPLANTATION
83
12. Goldberg M, Lima 0, Morgan E, et al. A comparison between cyclosporin A and methylprednisolone plus azathioprine on bronchial healing following canine lung autotransplantation. J Thorac Cardiovasc Surg 1983;85:821-6. 13. Toronto Lung Transplant Group. Unilateral lung transplantation for pulmonary fibrosis. N Engl J Med 1986;3141140-5. 14. Rahn H, Otis AB, Chadwick LE, Fenn WO. Am J Physiol 1946;146:161-6. 15. Cooper JD, Pearson FG, Patterson GA, et al. Technique of successful lung transplantation in humans. J Thorac Cardiovasc Surg 1987;93173-81. 16. Ladowski JS, Hardesty RL, Griffith BP. The pulmonary artery blood supply to the supracarinal trachea. J Heart Transplant 19&4;4:4&2. 17. Haverich A, Frimpong-Boateng K, Wahlers T, Schafers HJ. Pericardial flap-plasty for protection of the tracheal anastomosis in heart-lung transplantation. J Cardiac Surg 1989;4 136-9. 18. Schreinemakers HHJ, Weder W, Miyoshi S, et al. Direct revascularization of bronchial arteries for lung transplantation: an anatomic study. Ann Thorac Surg 1990;49:4&54. 19. Calhoon JH, Grover FL, Gibbons WJ, et al. Single lung transplantation: alternative indications and technique. J Thorac Cardiovasc Surg 1991;101:816-25.
DISCUSSION DR HANI SHENNIB (Montreal, Que, Canada): I enjoyed your paper very much. This is an elegant experimental model to show that steroids are not necessarily detrimental to the healing of the airway. However, we have looked at our own patients in 20 lung transplantations that were done in the last year and a few months, and the average steroid dose that we have given was about 6.8 mgkg body weight every day over the first 2 weeks. So even clinically, and at least from the data from Trinkle in San Antonio, the effect of steroids does not seem to be that bad. I, however, have a question in regard to the bursting pressure. I noticed that in your animal model you used Prolene. We do not use Prolene, we use Maxon and Vicryl, and I am just wondering whether the bursting pressure actually tests the actual healing of tissue or tests the actual strength of the suture material? DR AUTERI: We did not take out the suture material. In those animals where it did burst at lower than 510 mm Hg-303, 475, and 185 mm Hg-it burst between sutures, and we think, given this model, we can see a small pinhole burst that could occur between sutures. I should also note that we listed some anastomoses as “greater than 510 mm Hg” because that was the limit of our apparatus. However, in some of those the burst occurred at sites other than the anastomosis, indicating a very strong anastomosis. DR STEPHAN W.HIRT (Hannover, Germany): I would like to congratulate Dr Auteri and his associates for their study and would like to comment on our clinical lung transplant experience. In the last 2 years we have accepted recipients with a regular daily dose of prednisone up to 20 mg, and recently we use no bronchial wrapping. Reviewing the last 27 single-lung and 6 double-lung transplants, we have observed only one limited dehiscence of a bronchial anastomosis in a patient with a doublelung transplant who was not receiving corticosteroids before transplantation and did not receive postoperative steroids for 6 weeks.
DR J KENT TRINKLE (San Antonio, TX): I enjoyed your paper because it parallels our clinical experience. When I first started doing transplantation in patients who were receiving prednisone and giving them large doses of methylprednisolone postoperatively, everybody thought I was crazy. We have done 37 single-lung transplantations with no bronchial complications, neither stenosis nor dehiscence. I think it is a multifactorial thing. About two thirds of OUT patients are receiving preoperative prednisone and many of them would have died if we had weaned them off of prednisone. So I think that is an important point. We have also kept all of our patients on largedose methylprednisolone regimens for the first 3 days, and then they receive 1 mgikg of prednisone for a week, which is tapered to 0.75, 0.5 mg kg-’ day-’, and so forth at weekly intervals. I think the telescoping technique is important, as is keeping the bronchial stump short. Prolene sutures are also important, and I think steroids may improve bronchial healing. Now, in the patients in whom we perform bronchoscopy we see an unusual phenomena: the donor bronchial mucosa is not pale like you sometimes see in trauma cases or in sleeve resections for carcinoma. Instead, the distal bronchus tends to be a little pinker or more hyperemic. And I think-the key word there is think-that the steroids preventing rejection may maintain that vascularity. DR JOHN W.FEHRENBACHER (Indianapolis, IN): Just a concern. I may be personally brainwashed by Dr Cooper teaching us. I am still concerned that the dogs were receiving steroids for only a month preoperatively. It has been my clinical impression with some of these patients who receive steroids for a year or 2 years that you can just tell by looking at their skin changes and such, and you really worry about healing. Do you have any comments on that? DR AUTERI: I am unable to quantitate exactly how much steroid was on board. We did not have steroid dosage levels. And a fault of this study compared with Dr Cooper’s study was that skin
84
Ann Thorac Surg 1992;53:80-4
AUTERI ET AL BRONCHIAL HEALING IN LUNG TRANSPLANTATION
incisions were made in the latter to test if steroid levels were in fact adequate. I can say that, at least subjectively looking at the tissues, the tissues were what I would call steroid tissues, very friable, very much unable to hold a stitch, clearly steroid tissues. In addition, I should note that in group C, the group that received preoperative 1-month steroid treatment, all 5 of the animals, 1 of which died of a gastrointestinal bleed, had some evidence of hemorrhage from the gastrointestinal tract and needed H, blockers and antacids. DR JOEL D. COOPER (St. Louis, MO): But in fairness, you were using a much greater than clinical dose. DR AUTERI: Yes. DR COOPER So you were really stressing your model, because the studies we did used a dose of 2 mgkg, and nobody uses that much anymore; it is usually 0.5 mgkg. So in fairness to your model, you were really using a very high-dose steroid. DR PAUL F. WATERS (Los Angeles, CA): Did I hear you say that the ischemic time on your transplants was 45 minutes? DR AUTERI: That is correct. DR WATERS: I wonder if that is not a little short considering that I do not think many of us can do a lung transplantation in 45 minutes. I wonder if it might have been more clinically relevant to employ a longer ischemic time. DR AUTERI:I can say these were done side by side in the same room, and the recipient Mum was dissected-most of the work was done on the recipient Mum-and so it was a matter of sewing it in in 45 minutes. It was out and back in again in 45 minutes. We chose 45 minutes to eliminate any effect preservation would have on bronchial healing. We wished only to test the effect of steroids on bronchial healing.
DR COOPER: Let met ask you. When you just eyeballed your anastomoses-forgetting about the strength, et cetera4id they look normal and healed? DR AUTERI. I think the strongest point here is that yes, in fact they looked very strong, and I was on the other end of that catheter and pushing quite hard in that study. DR COOPER But when you opened up the bronchus and looked at the anastomosis, you were satisfied? DR AUTERI: Yes. DR COOPER I think,frankly, that is probably more important than anything else. My general feeling is that in the early days of transplantation, as you have pointed out, things were multifactorialpreservation was not the best, patients were on ventilators for long periods of time, infection was more common-and I think that in the early days, to get things going, there were a lot of extra security measures that maybe we and others took. I think what your work points out, and what the work that Dr Trinkle and the Hannover group and Toronto and others have done has pointed out, is that probably these days the need for the omentum and the avoidance of early steroid treatment are no longer as important as perhaps they were to get things going. We still use the omentum, but certainly we do cases without it if it is not available, and we will take people receiving low-dose steroids. We have not started as early as other people have suggested. But I think the scene has changed. We have an international registry of about 510 cases in it that we follow up. I just looked the other day, and the death rate from airway complicationsis less than 5% in that group. And so I think that the situation has changed considerably, and I think your paper is a very important contribution. Thank you for it.
-
The deleterious effect of steroids on bronchial healing in lung transplantation has led to the development of techniques to protect the anastomosis and to the exclusion of steroid-dependent patients from transplantation. The effect of steroids on bronchial healing was tested in a canine single-lung allotransplantation model. Twenty size-matched mongrel dogs (20 to 30 kg) underwent left lung transplantation without anastomotic wrap or direct revascularization. Postoperatively, all received daily doses of cyclosporine (15 mg/kg) and azathioprine (1 mglkg) and were subdivided into three steroid dosage groups. Group A (n = 10) animals received 1.5 mgkg of prednisone per day whereas groups B (n = 5) and C (n = 5) received 5.0 mg/kg of prednisone per day for 28 postoperative days. In addition, group C received pred-
nisone (5.0 mg kg-' * day-') for 1 month preoperatively. In group A, 8 of 10 dogs survived 28 days without evidence of respiratory compromise, with anastomotic bursting pressure greater than 510 mm Hg. In group B, all 5 dogs survived to 28 days without evidence of respiratory compromise and with intact bronchial anastomoses (bursting pressures greater than 510 mm Hg). In group C, 3 of 5 animals survived to 28 days with intact anastomoses. Histological examination demonstrated normal bronchial healing in all anastomoses. These data suggest that preoperative steroid dependence should not be a contraindication to lung transplantation and that bronchial anastomotic wrapping with vascular tissue may not be essential. (Ann Thorac Surg 1992;53:804)
I
we reexamined bronchial anastomotic healing in the presence of conventional as well as suprapharmacologic doses of corticosteroids.
n contrast to transplantation of other solid organs, single- and double-lung transplantation continue to be plagued with technical as well as immunologic problems. Single-lung transplantation in humans has been hampered by bronchial anastomotic complications, primarily considered a consequence of donor bronchus devascularization by the interruption of the bronchial arterial supply. Attempts at increasing bronchial blood flow by direct revascularization have met with limited success [l]. Wrapping the anastomosis with omentum has been shown to recruit arterial supply as early as the fourth postoperative
For editorial comment, see page 5. day in experimental models but increases operative time and makes the procedure technically more complex [2]. The deleterious effects of steroids and other immunosuppressive agents on wound healing have been implicated as a cofactor in dehiscence. This has led to empiric exclusion of steroid-dependent patients as candidates for lung transplantation and has led to a modification in early maintenance immunosuppression, substituting antithymocyte globulins for steroids. Experimental data to support these views are scant. Therefore, in a canine singlelung allotransplant model, without bronchial wrapping, Presented at the Twenty-seventh Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Feb E-20, 1991. Address reprint requests to Dr Auteri, Department of Surgery, Box 91, Columbia-Presbyterian Medical Center, 622 West 168th St, New York, NY 10032.
0 1992 by The Society of Thoracic Surgeons
Material and Methods Under general endotracheal anesthesia, through a left thoracotomy, 20 mongrel dogs (weight, 20 to 25 kg) underwent left single-lung allotransplantation from sizematched donors. All animals received humane care in compliance with the "Guide for the Care and Use of Laboratory Animals" (NIH publication No. 85-23, revised 1985). All transplants were standardized to a 45-minute ischemic time in 4°C normal saline solution. Donor lungs were neither heparinized nor perfused. Lungs were preserved by submersion in normal saline solution at 4°C. The atrial anastomosis, an inverting continuous horizontal mattress suture, was performed first, followed by the pulmonary artery and bronchus anastomoses. All three anastomoses were performed using running 4-0 polypropylene suture. No attempt was made to cover, wrap, or revascularize the bronchial anastomosis. All bronchial anastomoses were immediately submerged in saline solution and tested for competence by forced ventilation through a cuffed endotracheal tube. The lung was then reexpanded, the chest incision was closed, and the dogs recovered. Postoperatively, all dogs received daily immunosuppression with cyclosporine (15 mg/kg) and azathioprine (1 mgkg) and were subdivided into three steroid dosage groups. Group A (n = 10) received 1.5 mg/kg of prednisone daily. Groups B (n = 5) and C (n = 5) received 0003-4975/92/$3.50
AUTERIETAL BRONCHIAL HEALING IN LUNG TRANSPLANTATION
Ann Thorac Surg 1992;53:W
Fig I . Preparation for measurement of bursting pressure.
5.0 mg/kg of prednisone daily. In addition, group C animals received prednisone (5.0 mg * kg-’ * day-’) for 1 month preoperatively. All survivors were sacrificed at 28 days. At autopsy, the bronchus was excised from 2 cm proximal to 2 cm distal to the anastomosis. The ends were cross-clamped and an 18-gauge Silastic (Dow Corning, Midland, MI) catheter was introduced into the lumen (Fig 1). Normal saline solution stained with methylene blue dye was injected into the lumen of the bronchus as intraluminal pressure was transduced to an oscilloscope. The wave form was quantitated in millimeters of mercury and calibrated before each analysis. The upper limit of measurement in this system was 510 mm Hg. The bursting pressure was the point at which methylene blue dye could be identified externally. Those that achieved the upper limit without bursting were recorded as bursting pressure greater than 510 mm Hg. To determine the percent stenosis, the intraluminal radius was measured at the anastomosis (Anast) and at 2 cm proximal to the anastomosis (Prox). Percent stenosis was calculated according to the following formula: % Stenosis = [ l - (Anast%’rox2)]
x 100.
Histologic analysis of anastomoses was done after hematoxylin, phloxine, and saffron staining of paraffin section. Lung parenchyma was sectioned, and histologic characteristics were classified according to the International Society for Heart Transplantation Standardization of Nomenclature in the Diagnosis of Lung Rejection [31.
Results Intraoperatively, all anastomoses were airtight during a sustained inflation to 40 mm Hg. Higher pressures were
81
avoided to prevent parenchymal lung injury. In group A, 8 of 10 dogs survived 28 days without evidence of respiratory compromise. These animals had bursting pressures greater than 510 mm Hg. Two dogs died 6 and 9 days postoperatively as a result of left atrial thrombosis. These animals demonstrated intact bronchial anastomoses at autopsy (bursting pressures of 303 and 475 mm Hg, respectively). In group B, all 5 dogs survived to 28 days without evidence of respiratory compromise and with intact bronchial anastomoses (bursting pressures greater than 510 mm Hg). In group C, 3 of five animals survived 28 days. The 3 survivors had intact anastomoses at autopsy (bursting pressures greater than 510 mm Hg). One dog exsanguinated from a gastroduodenal ulcer on the first postoperative day and bursting pressure was not obtained. The other died on the thirteenth postoperative day of congestive heart failure resulting from an intraoperative myocardial infarction sustained during pulmonary artery injury. This animal had a bronchial anastomotic bursting pressur of 185 mm Hg. Anastomotic histology was considered normal if regenerated epithelium was observed lining the lumen with normal scar formation and healing fibrosis deep to the anastornotic site. This was seen in all anastomoses, including all three with bursting pressures less than 510 mm Hg, and in 3 animals with mild, moderate, and severe rejection as graded by the International Society for Heart Transplantation Standard of Nomenclature [3]. The animal with severe (A4a) rejection, despite extensive bronchiolitis obliterans, had a bursting pressure greater than 510 mm Hg with no evidence of bronchial dehiscence. No manifestations of clinically significant stenosis, such as stridor, respiratory distress, or atelectasis, were detected in any animal. In groups B and C .he average stenosis, as calculated by the formula in the Material and Methods section, was 19.1% (range, 14.3%to 20.3%).
Comment Concern over the fate of the healing bronchial anastomosis in lung transplantation has existed since investigation in this arena began [4, 51. Of the first 38 clinical lung transplantations done in the world, only 16 patients were alive beyond 10 days. Of these 16 patients, 10 died of bronchial complications within 7 months [5]. This experience prompted extensive research into the etiologic factors associated with bronchial dehiscence in lung transplantation. Early contributions to improve bronchial healing included wrapping the anastomosis with perihilar connective tissue [6], telescoping the bronchial ends [7], shortening the length of the donor bronchus [8], and bronchial artery reanastomosis [l,91. Particularly for the latter, which remains a difficult technical problem, application of these advances produced marginal improvement in results. This led to the landmark work of Cooper and his colleagues in Toronto, who demonstrated in a canine autotransplant model that omentum can bring saturated blood to the anastomosis in the early postoperative period [2, 101. Related studies suggested a deleterious effect of
82
AUTERIETAL BRONCHIAL HEALING IN LUNG TRANSPLANTATION
methylprednisolone on bronchial healing when compared with a cyclosporine group or with nonimmunosuppressed controls [ll, 121. This experimental work led directly to the first successful series of single-lung transplants [13] and established three principles of clinical practice: (1)to wrap all bronchial anastomoses with omentum, (2) to avoid postoperative steroids, and (3) to deny transplantation to patients dependent on steroids for treatment of their underlying lung disease. There are serious disadvantages to these principles. Mobilizing omentum requires a laparotomy, which adds the risk of pulmonary and other complications in an immunocompromised host. Omentum must be passed through a diaphragmatic defect, which can become a clinically significant hernia [2]. When maintenance immunosuppression is modified by substitution of antithymocyte globulins for steroids, the risk of viral and fungal infection is increased. Finally, many patients who might otherwise be good candidates for lung transplantation are steroid dependent and are denied lung transplantation until withdrawn from steroid treatment. For these reasons, this study was undertaken to reexamine these concepts. This study differs in several respects from the important work by Cooper and his colleagues in Toronto. In one series of experiments they compared anastomoses with omental wrapping to those without and reported that wrapped anastomoses had improved gross and histologic appearance, reduced bronchial stenosis, and neovascularity arising from the omentum demonstrated on celiac arteriography [2]. However, because arteriography was not performed in the unwrapped animals, the amount of neovascularity achieved without omental wrap cannot be determined. The present study was not designed to compare wrapped with unwrapped animals, but the histologic appearance and percent stenosis of the unwrapped anastomoses in all three steroid dosage groups was consistent with normal healing. In the present study, bursting pressure was measured to quantitate the strength of the anastomosis, whereas tensile strength was measured in the Cooper studies. Tensile strength reflects resistance to disruption by linear forces, as if the lung were simply a weight hanging from the anastomosis, and was found to be decreased by steroids [ll]. Linear forces are probably minimal in a lung that fills the hemithorax. Our use of bursting pressure for measurement of the integrity of the anastomosis reflects our dissatisfaction with the tensile strength method, which measures only linear forces. Bursting pressure reflects resistance to disruption by radial forces generated within the lumen and is analogous to the airway pressure generated during cough or positive pressure ventilation, perhaps a more clinically relevant measurement. Data to support this contention are not currently available. In the present study the maximum intraluminal pressure applied (>510 mm Hg) is far in excess of the maximum physiologic pressure of 100 to 120 mm Hg [14] and was accepted as evidence of excellent anastomotic strength. The 3 animals with bursting pressures less than 510 mm Hg still had bursting pressures well above the usual physio-
Ann Thorac Surg 19!%!;538C4
logic range (185, 303, and 475 mm Hg). The animal with the lowest bursting pressure measured (185 mm Hg) was sacrificed after 13 days of low cardiac output following intraoperative myocardial infarction. In a limited cross-comparison between bursting pressure and tensometry, 3 animals with bursting pressures greater than 510 mm Hg were found to have a bronchial tensile strength (1,064 f 282 g) that correlated with those considered to be in the satisfactory range in the Cooper study (1,232 f 101 g) [12]. Finally, it is particularly noteworthy that bronchial dehiscence was not seen in any animal in this allograft study or in the Cooper autograft studies, and all animals in the latter survived without incident to sacrifice at 23 days, regardless of histologic or arteriographic findings [2]. None of the three deaths in this study was related to bronchial healing. Satisfactory healing was seen even in the presence of severe pulmonary rejection and bronchiolitis obliterans. In summary, the data in this canine allograft model suggest that steroid immunosuppression is compatible with excellent bronchial healing, even at higher than conventional dosage. Preoperative steroid administration, at sufficiently high dosage levels to produce gastrointestinal ulceration and bleeding in more than 1experimental animal, did not have deleterious effects on bronchial healing. Finally, excellent healing was obtained without wrapping the anastomoses with omentum or with any other material. It would be hazardous to conclude that bronchial healing is no longer a problem in lung transplantation, as recent experience has demonstrated [15]. Although previous studies may have overemphasized the detrimental effects of steroid, or the importance of omental wrapping, the bronchial anastomosis is undeniably hypovascular unless bronchial arteries are rejoined [16]. In a clinical allograft, lung preservation, ischemic time, rejection, infection, and the hemodynamic stability of the recipient undoubtedly play a role in bronchial healing, and it will always be difficult to isolate the contribution of each variable. Vascular wraps are still receiving appropriate experimental attention [17], and attempts are being made to simplify bronchial artery anastomosis [18]. In addition, the advantages of a telescoping anastomosis have been reported especially in patients receiving steroids [191. Continued evolution of the basic principles of bronchial healing can be expected.
References 1. Pinsker KL, Montefusco CM, Kamholz SL, Hagtstrom JWC, Gliedman ML, Veith FJ. Improved bronchial anastomotic healing secondary to maintenance or restoration of bronchial arterial circulation by microsurgical techniques. Surg Forum 1980;31:23&1. 2. Morgan E, 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;85:134-9. 3. Yousem SA, Berry GJ, Brunt EM, et al. Working formulation for the standardization of nomenclature in the diagnosis of
Ann Thorac Surg
1992:53804
heart and lung rejection: lung rejection study group. J Heart Transplant 1990;659%4. 4. Hardy JD, Webb WR, Dalton ML, Walker GR. Lung homotransplantation in man, report of the initial case. JAMA 1963;121065-74. 5. Veith FJ, Kamholz SL, Mollenkopf FP, Montefusco CM. Lung transplantation. Transplantation 1983;35:271-8. 6. Blumenstock DA, Kahn DR. Replantation and transplantation of the canine lung. J Surg Res 1961;l:U. 7. Sinha SBP, Doughterty JC, Boley SJ, Veith FJ. Elimination of bronchial complications in lung transplantation. Surg Forum 1971;22.22532. 8. Pinsker KL, Koerner SK, Kamhotz SL, Hagstrom JWC, Veith FJ. Effect of donor bronchial length on healing. J Thorac Cardiovasc Surg 1979;77:66%77. 9. Veith F, Blumenstock DA. Lung transplantation. J Surg Res 1971;11:3>9. 10. Lima 0,Goldberg M, Peters WJ, Ayabe H, Townsend E, Cooper JD. Bronchial omentopexy in canine lung transplantation. J Thorac Cardiovasc Surg 1982;83:418-27. 11. Lima 0,Cooper JD, Peters WJ, et al. Effects of methylprednisolone and azathioprine on bronchial healing following lung autotransplantation. J Thorac Cardiovasc Surg 1981;82 211-5.
AUTERIETAL BRONCHIAL HEALING IN LUNG TRANSPLANTATION
83
12. Goldberg M, Lima 0, Morgan E, et al. A comparison between cyclosporin A and methylprednisolone plus azathioprine on bronchial healing following canine lung autotransplantation. J Thorac Cardiovasc Surg 1983;85:821-6. 13. Toronto Lung Transplant Group. Unilateral lung transplantation for pulmonary fibrosis. N Engl J Med 1986;3141140-5. 14. Rahn H, Otis AB, Chadwick LE, Fenn WO. Am J Physiol 1946;146:161-6. 15. Cooper JD, Pearson FG, Patterson GA, et al. Technique of successful lung transplantation in humans. J Thorac Cardiovasc Surg 1987;93173-81. 16. Ladowski JS, Hardesty RL, Griffith BP. The pulmonary artery blood supply to the supracarinal trachea. J Heart Transplant 19&4;4:4&2. 17. Haverich A, Frimpong-Boateng K, Wahlers T, Schafers HJ. Pericardial flap-plasty for protection of the tracheal anastomosis in heart-lung transplantation. J Cardiac Surg 1989;4 136-9. 18. Schreinemakers HHJ, Weder W, Miyoshi S, et al. Direct revascularization of bronchial arteries for lung transplantation: an anatomic study. Ann Thorac Surg 1990;49:4&54. 19. Calhoon JH, Grover FL, Gibbons WJ, et al. Single lung transplantation: alternative indications and technique. J Thorac Cardiovasc Surg 1991;101:816-25.
DISCUSSION DR HANI SHENNIB (Montreal, Que, Canada): I enjoyed your paper very much. This is an elegant experimental model to show that steroids are not necessarily detrimental to the healing of the airway. However, we have looked at our own patients in 20 lung transplantations that were done in the last year and a few months, and the average steroid dose that we have given was about 6.8 mgkg body weight every day over the first 2 weeks. So even clinically, and at least from the data from Trinkle in San Antonio, the effect of steroids does not seem to be that bad. I, however, have a question in regard to the bursting pressure. I noticed that in your animal model you used Prolene. We do not use Prolene, we use Maxon and Vicryl, and I am just wondering whether the bursting pressure actually tests the actual healing of tissue or tests the actual strength of the suture material? DR AUTERI: We did not take out the suture material. In those animals where it did burst at lower than 510 mm Hg-303, 475, and 185 mm Hg-it burst between sutures, and we think, given this model, we can see a small pinhole burst that could occur between sutures. I should also note that we listed some anastomoses as “greater than 510 mm Hg” because that was the limit of our apparatus. However, in some of those the burst occurred at sites other than the anastomosis, indicating a very strong anastomosis. DR STEPHAN W.HIRT (Hannover, Germany): I would like to congratulate Dr Auteri and his associates for their study and would like to comment on our clinical lung transplant experience. In the last 2 years we have accepted recipients with a regular daily dose of prednisone up to 20 mg, and recently we use no bronchial wrapping. Reviewing the last 27 single-lung and 6 double-lung transplants, we have observed only one limited dehiscence of a bronchial anastomosis in a patient with a doublelung transplant who was not receiving corticosteroids before transplantation and did not receive postoperative steroids for 6 weeks.
DR J KENT TRINKLE (San Antonio, TX): I enjoyed your paper because it parallels our clinical experience. When I first started doing transplantation in patients who were receiving prednisone and giving them large doses of methylprednisolone postoperatively, everybody thought I was crazy. We have done 37 single-lung transplantations with no bronchial complications, neither stenosis nor dehiscence. I think it is a multifactorial thing. About two thirds of OUT patients are receiving preoperative prednisone and many of them would have died if we had weaned them off of prednisone. So I think that is an important point. We have also kept all of our patients on largedose methylprednisolone regimens for the first 3 days, and then they receive 1 mgikg of prednisone for a week, which is tapered to 0.75, 0.5 mg kg-’ day-’, and so forth at weekly intervals. I think the telescoping technique is important, as is keeping the bronchial stump short. Prolene sutures are also important, and I think steroids may improve bronchial healing. Now, in the patients in whom we perform bronchoscopy we see an unusual phenomena: the donor bronchial mucosa is not pale like you sometimes see in trauma cases or in sleeve resections for carcinoma. Instead, the distal bronchus tends to be a little pinker or more hyperemic. And I think-the key word there is think-that the steroids preventing rejection may maintain that vascularity. DR JOHN W.FEHRENBACHER (Indianapolis, IN): Just a concern. I may be personally brainwashed by Dr Cooper teaching us. I am still concerned that the dogs were receiving steroids for only a month preoperatively. It has been my clinical impression with some of these patients who receive steroids for a year or 2 years that you can just tell by looking at their skin changes and such, and you really worry about healing. Do you have any comments on that? DR AUTERI: I am unable to quantitate exactly how much steroid was on board. We did not have steroid dosage levels. And a fault of this study compared with Dr Cooper’s study was that skin
84
Ann Thorac Surg 1992;53:80-4
AUTERI ET AL BRONCHIAL HEALING IN LUNG TRANSPLANTATION
incisions were made in the latter to test if steroid levels were in fact adequate. I can say that, at least subjectively looking at the tissues, the tissues were what I would call steroid tissues, very friable, very much unable to hold a stitch, clearly steroid tissues. In addition, I should note that in group C, the group that received preoperative 1-month steroid treatment, all 5 of the animals, 1 of which died of a gastrointestinal bleed, had some evidence of hemorrhage from the gastrointestinal tract and needed H, blockers and antacids. DR JOEL D. COOPER (St. Louis, MO): But in fairness, you were using a much greater than clinical dose. DR AUTERI: Yes. DR COOPER So you were really stressing your model, because the studies we did used a dose of 2 mgkg, and nobody uses that much anymore; it is usually 0.5 mgkg. So in fairness to your model, you were really using a very high-dose steroid. DR PAUL F. WATERS (Los Angeles, CA): Did I hear you say that the ischemic time on your transplants was 45 minutes? DR AUTERI: That is correct. DR WATERS: I wonder if that is not a little short considering that I do not think many of us can do a lung transplantation in 45 minutes. I wonder if it might have been more clinically relevant to employ a longer ischemic time. DR AUTERI:I can say these were done side by side in the same room, and the recipient Mum was dissected-most of the work was done on the recipient Mum-and so it was a matter of sewing it in in 45 minutes. It was out and back in again in 45 minutes. We chose 45 minutes to eliminate any effect preservation would have on bronchial healing. We wished only to test the effect of steroids on bronchial healing.
DR COOPER: Let met ask you. When you just eyeballed your anastomoses-forgetting about the strength, et cetera4id they look normal and healed? DR AUTERI. I think the strongest point here is that yes, in fact they looked very strong, and I was on the other end of that catheter and pushing quite hard in that study. DR COOPER But when you opened up the bronchus and looked at the anastomosis, you were satisfied? DR AUTERI: Yes. DR COOPER I think,frankly, that is probably more important than anything else. My general feeling is that in the early days of transplantation, as you have pointed out, things were multifactorialpreservation was not the best, patients were on ventilators for long periods of time, infection was more common-and I think that in the early days, to get things going, there were a lot of extra security measures that maybe we and others took. I think what your work points out, and what the work that Dr Trinkle and the Hannover group and Toronto and others have done has pointed out, is that probably these days the need for the omentum and the avoidance of early steroid treatment are no longer as important as perhaps they were to get things going. We still use the omentum, but certainly we do cases without it if it is not available, and we will take people receiving low-dose steroids. We have not started as early as other people have suggested. But I think the scene has changed. We have an international registry of about 510 cases in it that we follow up. I just looked the other day, and the death rate from airway complicationsis less than 5% in that group. And so I think that the situation has changed considerably, and I think your paper is a very important contribution. Thank you for it.