Revascularization of ischemic bronchial anastomoses by an intercostal pedicle flap

J

THoRAc CARDIOV ASC SURG

90: 172-178, 1985

Revascularization of ischemic bronchial anastomoses by an intercostal pedicle flap Ischemia of the donor bronchus, perfused solely by retrograde coUaterais from the pulmonary circulation, is an important factor in the impaired healing of the bronchial anastomosis of transplanted lungs. The healing of two experimental models of bronchial anastomotic ischemia, the bronchial segmental autograft and the postpneumonectomy bronchial autograft, was assessed in dogs. The application of a polytetrafluoroethylene wrap to the bronchial segmental autograft and the application of an intercostal pedicle flap to the postpneumonectomy bronchial autograft, with and without concomitant administration of corticosteroids, were also studied to elucidate factors that affect bronchial anastomotic healing. The bronchial segmental autograft healed normaUy without stricture, but isolation of this autograft from the mediastinum and lung by the polytetrafluoroethylene wrap resulted in necrosis of the autograft. AU dogs that had a postpneumonectomy bronchial autograft died of bronchopleural fistulas due to autograft necrosis. Application of an intercostal pedicle flap to the autograft resulted in healing in aU animals. Arteriography and Microfd injection demonstrated revascuiarization of the postpneumonectomy bronchial autograft by the pedicled intercostal artery. Several conclusions can be drawn: With the lung in situ the bronchial segmental autograft survives, probably as a free composite graft. In contrast, the postpneumonectomy bronchial autograft is an exceUent model of bronchial anastomotic ischemia. The intercostal pedicle flap is a reliable method for providing neovascuiarity and mechanical reinforcement to an ischemic bronchial anastomosis. Its effect on bronchial anastomotic healing was not diminished by administration of corticosteroids. The intercostal pedicle flap may be useful in preventing bronchial anastomotic complications in clinical lung transplantation.

Stanley C. Fell, M.D., Fred P. Mollenkopf, P.A.-C, Cheryl M. Montefusco, Ph.D., Sumi Mitsudo, M.D., Stephan L. Kamholz, M.D., Jamie Goldsmith, R.N., and Frank J. Veith, M.D., New York. N. Y.

Impaired healing of the bronchial anastomosis has been a constant impediment to successful human lung transplantation. This defective healing is thought to be due to ischemia. The purposes of this investigation were to assess the healing of two previously described devascularized bronchial autografts in dogs, the bronchial segmental autoFrom the Departments of Surgery, Pathology, and Medicine, Montefiore Medical Center-Albert Einstein College of Medicine, New York, N. Y. Supported in part by grants from the U.S. Public Health Service (H L17417) and The James Hilton Manning and Emma Austin Manning Foundation. Received for publication April 30, 1n4. Accepted for publication Oct. 17, 1984. Address for reprints: Stanley C. Fell, M.D., Department of Cardiothoracic Surgery, Montefiore Medical Center, III East 210th St., New York, N. Y. 10467.

172

graft' and the postpneumonectomy bronchial autograft.' The effects of application of an intercostal pedicle flap to the postpneumonectomy bronchial autograft, with and without concomitant administration of corticosteroids, were also studied. Methods All animals received humane care in compliance with guidelines formulated by the National Society for Medical Research and the National Academy of Sciences.' All animals were anesthetized with intravenous pentobarbital sodium (25 to 30 rug/kg) and intravenous pancuronium bromide (4 mg). During the operation, the animals were intubated and mechanically ventilated. Thirty greyhounds weighing 35 to 45 kg were prepared as described and were divided into the following groups: Group I: Bronchial segmental autograft (n = 6).

Volume 90 Number 2

Revascularization of ischemic bronchial anastomoses 1 7 3

August, 1985

Table IA. Bronchoscopic findings immediately before death in Group I (bronchial segmental autograft) Animal No.

Postop. day killed

383

60

398

48

396

59

403

52

429 48975

33 49

Bronchoscopic findings Healed anastomosis, Intact mucosa, no stricture Healed anastomosis, small area of exposed cartilage, no stricture No stricture, 2 mm area of exposed cartilage Healed anastomosis, two areas of exposed cartilage at proximal anastomosis, no stricture Mucosa intact, no stricture Anastomosis patent, small area of exposed cartilage, no stricture

After the peribronchial tissue was stripped and the bronchial vessels were ligated, the left main bronchus was occluded with a noncrushing vascular clamp and transected twice. A 1.5 em bronchial ring was thus produced (Fig. 1, A). The segment was reanastomosed (5-0 polypropylene) in its original position and the left lung was reinflated. Ischemic time for the bronchial segment ranged from 50 to 70 minutes. Group I dogs were subjected to bronchoscopy at the conclusion of the procedure and periodically thereafter. The dogs were put to death 27 to 60 days (mean 44) after operation (Table IA). Group II: Bronchial segmental autograft plus PTFE wrap (n = 4). The bronchial segmental autograft was isolated from adjacent pulmonary and mediastinal tissues with polytetrafluoroethylene* (PTFE), which was wrapped loosely around the autograft and sutured in palce (Fig. 1, B). Bronchoscopy was done before the animals were put to death 4 to 12 days after operation (Table IIA). Group III: Postpneumonectomy bronchial autograft (n = 5). A modification of the bronchial autograft technique described by Morgan and associates' was employed (Fig. 1, C). Following thoracotomy via the left fifth intercostal space, the pulmonary vessels were divided. The left bronchus was isolated and stripped of peribronchial tissue and the lobar bronchi were stapled "Gorc-Tcx peripheral vascular patch, W. L. Gore & Associates, Inc., Elkton, Md.

m.

Table Postmortem findings in Group I (bronchial segmental autograft) Animal No.

Gross findings

Histologic findings

383

Well healed, no stricture

398

Well healed, no stricture

396

Well healed, no stricture, 2 mm area of exposed cartilage Well healed, no stricture

Mucosal fibrosis, cuboidal epithelium, absent glands, fibrous scarring, foreign body reaction in viable cartilage Occasional areas of devitalized cartilage, subepithelial fibrosis Reepithelialization of autograft, a few areas of incompletely devitalized cartilage

403

429

Well healed, no stricture

48975

Well healed, no stricture

Cuboidal reepithelialization, subepithelial fibrosis, foreign body reaction about sutures, viable cartilage Cuboidal epithelium, subepithelial fibrosis, absent glands, viable cartilage with fibrosis Cuboidal epithelium, regenerating cartilage

closed. The bronchi were divided distal to the staple line and the lung was discarded. The left main bronchus was occluded as described earlier, completely transected 1.5 cm proximal to the staple line, and then reanastomosed (5-0 polypropylene). This procedure resulted in a devascularized postpneumonectomy bronchial autograft. Necropsy was performed at the time of death (Table III). Group IV: Postpneumonectomy bronchial autograft plus intercostal pedicle flap (n = 9). Following creation of the postpneumonectomy bronchial autograft, the fifth rib was excised subperiosteally. A dorsally based intercostal pedicle consisting of parietal pleura, periosteum, vessels, nerve, and muscle was constructed (Fig. 1, D). The orifice of the distal end of the intercostal artery bled freely and required ligation in all cases. In four dogs the periosteum was cauterized with 20% silver nitrate to inhibit osteogenesis.' The pedicle was wrapped around the autograft and held in place with a few 5-0 polypropylene sutures. The chest was closed without drainage. The dogs were put to death 8 to 66 days (mean 33) after operation (Table IV). Before two of the dogs were put to death, selective intercostal arteriography was performed via a radiopaque catheter introduced into the right carotid artery and advanced under fluoro-

174

The Journal of Thoracic and Cardiovascular Surgery

Fell et al.

Table lIB. Postmortem findings in Group II (bronchial segmental autograft plus PTFE wrap) Animal No.

502

Disrupted necrotic autograft, purulent exudate, pneumonia in left upper and lower lobes

521

Hemorrhagic, necrotic mucosa, exposed cartilage, PTFE adherent to bronchus, PTFE partially eroded into bronchial lumen

B

A

c

D

Fig. 1. A. The bronchial segmental autograft. B. The completed bronchial segmental autograft is wrapped with polytetrafluoroethylene. C. The postpneumonectomy bronchial autograft. D. The postpneumonectomy bronchial autograft is wrapped with the intercostal pedicle flap.

511

Table IIA. Bronchoscopic findings immediately before death in Group II (bronchial segmental autograft plus PTFE wrap)

528

day killed

502

9

521

12

511

6

528

4

Mucosal slough, exposed cartilage, partial disru ption of anastomosis, PTFE partially in bronchial lumen Mucosal slough and hemorrhage, exposed cartilage

Histologic findings Acute inflammatory reaction about PTFE wrap. marked foreign body reaction, dissolution of cartilage Mucosal slough, acute and chronic

inflarnrnation, degenerated cartilage, PTFE adherent to bronchus by underlying granulation Areas of transmural necrosis of autograft, acute inflammation and hemorrhage Partial mucosal slough and hemorrhage, acu te

inflammatory

POS{Op.

Animal No.

Gross findings

reaction about cartilage, granulation between PTFE and bronchial adventitia

Bronchoscopic findings Disrupted anastomosis, pus in bronchial lumen Hemorrhagic mucosa, purulent exudate, suture line appeared necrotic Edema of anastomotic suture lines, parital disruption of distal anastomosis Mucosal slough and hemorrhagic mucosa

Legend: PTFE. Polytctrafluorocthylcnc.

scopic guidance until its tip entered the orifice of the pedicled intercostal artery. In four of the remaining animals, Microfil* was injected into a 2 to 3 ern segment of the descending thoracic aorta containing the orifice of the pedicled intercostal artery. After the Microfil cured, the tissue containing the intercostal pedicle, trachea, carina, and main-stem bronchus was excised and photographed prior to formalin fixation. Histologic sections of all anastomoses were stained with hematoxylin and eosin. Group V: Postpneumonectomy bronchial autograft, intercostal pedicle flap, and prednisone (n = 6). In *Canton Bioproducts, Boulder, Colo.

addition to the postpneumonectomy bronchial autograft and intercostal pedicle flap, these dogs received prednisone 0.5 mg/kg/day commencing intraoperatively and continuing until they were put to death 21 to 44 days (mean 28) after operation (Table V). Gross and microscopic examinations were performed as in Group IV. Results Group I: Bronchial segmental autograft. Bronchoscopy performed immediately postoperatively demonstrated slight inversion of an intact suture line. Subsequent bronchoscopic studies revealed mucosal edema and scattered areas of cartilage denuded of mucosa. Bronchoscopic and gross examination when the animals were put to death demonstrated intact bronchial mucosa with no evidence of stricture (Tables IA and IB and Fig. 2). Microscopic examination revealed cuboidal bronchial mucosa suggestive of regeneration, moderate submucosal fibrosis, and viable cartilage (Table IB and Fig. 3).

Volume 90 Number 2

Revascularization of ischemic bronchial anastomoses

August, 1985

Fig. 2. Gross specimen of a bronchial segmental autograft 50 days postoperatively, demonstrating healed suture lines, intact mucosa and no evidence of structure. Arrow indicates the bronchial segmental autograft.

17 5

Fig. 3. Suture line for the bronchial segmental autograft 33 days postoperatively, showing healing with reepitheliazation, underlying fibrosis, and viable cartilage. (Original magnification X25.)

Table m. Group III: Postpneumonectomy bronchial autograft Postop. day

Animal No.

of death

1983

6*

373

7

366

7

434

7

356

7

Gross findings

Histologic findings

Bronchus necrotic distal to suture line Bronchopleural fistula. empyema Bronchopleural fistula, mediastinal emphysema, necrotic autograft Bronchopleural fistula, necrotic autograft, suture line disruption Bronchopleural fistula, massive empyema

Necrotic autograft. inflammatory reaction Necrotic autograft, exposed cartilage, inflammatory reaction Necrotic autograft, acute inflammatory reaction

Necrotic autograft, exposed cartilage, acute inflammatory reaction Necrotic autograft, acute inflammatory reaction

'The first dog was put to death. and the other four died spontaneously.

Group II: Bronchial segmental autograft plus PTFE wrap. At bronchoscopy performed immediately before the animals were killed (Table II), the bronchial lumen was seen to be occluded by purulent exudate, In two dogs, the necrotic bronchial suture lines had disrupted and allowed the PTFE wrap to protrude into the bronchial lumen. When the dogs were killed (4 to 12 days after operation) the autograft was found to be grossly and microscopicallynecrotic (Table lIB). Suture line disruption was noted in two animals. Peribronchial abscesses were walled off by contiguous lung and mediastinal tissues. Microscopic examination of all

autografts showed sloughing of the mucosa, necrosis of the submucosa, transmural hemorrhage, and acute inflammatory reaction. Group III: Postpneumonectomy bronchial autograft. All dogs died of bronchopleural fistulas 5 to 7 days after operation (Table III). Histologic examination of postmortem specimens revealed disruption of the bronchial suture line and necrosis of the bronchial autograft (Table III). Group IV: Postpneumonectomy bronchial autograft plus intercostal pedicle ftap. Premortem bronchoscopy in all dogs revealed well-healed anastomoses

The Journal of Thoracic and Cardiovascular Surgery

I 7 6 Fell et al.

Table IV. Group IV: Postpneumonectomy bronchial autograft plus intercostal pedicle flap Animal No.

Postop. Gross findings

Histologic findings

339

23

No

Healed anastomosis, adherent pedicle

44976

42

No

Autograft healed, nonconstricting ossification of pedicle

345

22

Yes

Healed anastomosis, Microfil visible across suture lines

357

46

No

371

68

No

382

34

Yes

Pedicle adherent, autograft healed, nonconstricting ossification, Microtil across suture lines Autograft healed, extensive nonconstricting ossification of pedicle wrap Autograft healed, pedicle adherent

407

6

Yes

471

10

Yes

523

7

No

A trophy of pedicle muscle. viable autograft, intact mucosa Some bone deposition, atrophic skeletal muscle, viable mucosa, organizing fibrous tissue Healed mucosa, viable skeletal muscle, foreign body reaction about sutures Skeletal muscle atrophy, intact mucosa, extensive bone deposition Skeletal muscle atrophy, viable mucosa, extensive bone deposition Viable autograft, intact mucosa, beginning atrophy of skeletal muscle Viable mucosa, cartilage formation skeletal muscle viable Viable autograft, intact mucosa, viable skeletal muscle Mucosa and skeletal muscle viable, viable autograft

Table V. Group

day killed

Viable pedicle, healed autograft, Microfil visible in autograft and across suture lines Healed autograft, adherent pedicle Viable pedicle, Microfil visible across suture lines

v.- Postpneumonectomy bronchial autograft,

intercostal pedicle flap, and prednisone

Postop.

Animal No.

day killed

Gross findings

Histologic findings

442

22

Viable mucosa, skeletal muscle atrophy

426

22

Yes

Viable pedicle, no stricture of autograft, anastomosis healed Healed autograft, adherent pedicle

439

24

Yes

Healed autograft, adherent pedicle

455

21

No

Healed autograft, adherent pedicle

454

35

No

Healed autograft, minimal nonconstrictive bone formation

448

44

Yes

Healed autograft, adherent pedicle

and staple lines. Arteriograms demonstrated patency of the pedicled intercostal arteries and a vascular network surrounding the autografts. Gross examination of the intercostal pedicle flaps showed them to be adherent to the underlying bronchus (Table IV). Microfil injection of the aortic segment containing the orifice of the pedicled intercostal artery demonstrated filling of submucosal arteries of the bronchial autograft (Fig. 4). Histologic examination revealed healed suture lines of the bronchial autografts with normal mucosa and viable bronchial cartilage (Table IV and Fig. 5). The pedicles

Intact mucosa, skeletal muscle atrophy, organizing fibrous tissue Viable mucosa, skeletal muscle atrophy, organizing fibrous tissue Viable mucosa, minimal new bone formation, skeletal muscle atrophy Intact mucosa, scattered areas of new bone formation, atrophic skeletal muscle Viable mucosa, skeletal muscle atrophy, organizing fibrous tissue

uniformly showed reparative granulation tissue, fibrosis, and skeletal muscle atrophy. In the four animals in which theperiosteum wascauterized, there was minimal evidence of bone formation. The remaining five dogs showed extensive ossification of the pedicle without constriction of the bronchial autograft. Group V:Postpneumonectomy bronchial autograft, intercostal pedicle flap, and prednisone. Healing of the bronchial autograft and adherence of the intercostal pedicle flap were similar to that noted in Group IV animals (Table V).

Volume 90 Number 2 August, 1985

Revascularization of ischemic bronchial anastomoses

177

Fig. 4. Mucosal surface of a postpneumonectomy bronchial autograft wrapped with an intercostal pedicle flap, 40 days postoperatively. Microfil injection of isolated aortic segment containing the orifice of the pedicled intercostal artery demonstrates dye-filled vessels in the muscosa of the viable autograft. Arrows indicate the postpneumonectomy bronchial autograft.

Discussion

Ischemia of the donor bronchus is believed to be an important cause of impaired healing of the bronchial anastomosis in clinical lung transplantation. After transplantation, the donor bronchus is perfused solely with desaturated blood, via low-pressure pulmonary artery collaterals, and is thus rendered ischemic until the bronchial artery circulation is reestablished 2 to 4 weeks later.v' For this reason, direct revascularization of the bronchus with a Carrel patch of donor aorta containing the bronchial arteries has been performed in dogs and in two human lung transplant recipients. 8. 9 This technique has not been employed further, as it increased the complexity of both the harvesting and insertion of the donor lung and it may not be technically possible in cases of right lung transplantation. Invaginating the donor bronchus into that of the recipient has produced successful healing of experimental bronchial anastomoses,'? probably because the recipient bronchus supplies the vascularized bed of granulation that is required for bronchial healing. I I However, this method has not precluded bronchial healing defects in clinical lung transplants." Kiriluk and Merendino I noted necrosis or stricture of bronchial segmental autografts that were anastomosed with silk or chromic sutures. In contrast, the bronchial segmental autografts in the present study healed satisfactorily, perhaps because monofilament suture material was used. However, the bronchial segmental autograft became necrotic when it was isolated from the medias-

Fig. 5. Overall view of a postpneumonectomy bronchial autograft with an intercostal pedicle flap (I.P.F.) 34 days postoperatively, demonstrating a viable autograft with intact mucosa. (Original magnification x2.5.)

1 78

Fell et al.

tinal tissues by a PIFE wrap. This observation suggests that the bronchial segmental autograft (without the PIFE wrap) maintains viability as a free composite graft in the mediastinum until bronchial arterial circulation is reestablished. The bronchial segmental autograft alone is therefore not a reliable model of bronchial anastomotic ischemia and was not studied further. Although Milhiet and colleagues' reported satisfactory healing of the postpneumonectomy bronchial autograft, this study confirms the work of Morgan and associates,' who found the unsupported postpneumonectomy bronchial allograft to be a lethal model of bronchial ischemia. None of the animals in the present study that received postpneumonectomy bronchial autografts alone survived beyond 7 days; all died of bronchopleural fistulas resulting from ischemia of the autograft. These results indicate that the proximal bronchial vascular supply is inadequate to support healing of a bronchial autograft. In those animals with intercostal pedicles applied to the postpneumonectomy bronchial allograft, arteriography and Microfil injection of the pedicled intercostal artery demonstrated perfusion of the mucosal vessels of the autograft. The strong adherence of the pedicles to the autografts suggests that, in addition to supplying neovascularity, the pedicles also provide mechanical support of the bronchial anastomoses. The efficacy of the intercostal pedicle flap was not diminished by the administration of corticosteroids in a dose range that is known to impede wound healing. I J In addition, extensive clinical experience has demonstrated that the intercostal pedicle flap is a durable structure for the intrathoracic repair of tracheal and esophageal defects" and bronchopleural fistulas, despite the presence of local infection. IS• 16 An omental pedicle flap has been used successfully to revascularize the postpneumonectomy bronchial autograft in laboratory investigations similar to the present study.' However, the intercostal pedicle flap is readily available at thoracotomy for lung transplantation and, unlike the omental pedicle flap, does not necessitate a laparotomy. This advantage may make it the optimal choice for revascularizing the bronchial anastomosis of transplanted lungs. REFERENCES Kiriluk LB, Merendino KA: An experimental evaluation of bronchial anastomosis and healing with special consid-

·The Journal of Thoracic and Cardiovascular Surgery

eration of the plane of transection. Surg Gynecol Obstet 96:175-182,1983 2 Milhiet H, Renault P, Drutel P, Chartier P, Jager P, Tismit G: Contribution a la greffe pulmonaire experimentale explication de la revascularization bronchique du greffon par l'etude d'auto-greffons bronchiques libres. Ann Chir Thorac Cardiovasc 7:563-568, 1968 3 Guide for the Care and Use of Laboratory Animals, DHEW Publication No. (NIH) 80-23, Revised 1978, Reprinted 1980, Office of Science and Health Reports, ORR/NIH, Bethesda, Md. 20205. 4 Morgan E, Lima 0, Goldberg M, Ferdman A, Luk SK, Cooper JD: Successful revascularization of totally ischemic bronchial autografts with omental pedicle flap in dogs. J THoRAc CARDIOVASC SURG 84:204-210, 1982 5 Bryant L: Experimental evaluation of intercostal pedicle grafts in esophageal repair. JTHORAC CARDIOVASC SURG 50:626-633, 1965 6 Stone RM, Ginsburg RJ, Colapinto RF, Pearson FG: Bronchial artery regeneration after radical hilar stripping. Surg Forum 17:109-110, 1966 7 Siegelman SS, Hagstrom JWC, Koerner SK, Veith FJ: Restoration of bronchial artery circulation after canine lung allotransplantation. J THoRAc CARDIOVASC SURG 73:792-795, 1977 8 Haglin JJ, Ruiz E, Baker RC, Anderson WR: Histologic studies of human lung allotransplantation, Morphology in Lung Transplantation, CRH Wildevuur, ed., Basel, Switzerland, 1973, S Karger AG, pp 13-22 9 Mills NL, Boyd AD, Gheranpong C: The significance of bronchial circulation in lung transplantation. J THORAC CARDIOVASC SURG 60:866-878, 1970 10 Veith FJ, Richards K: Improved technic for canine lung transplantation. Ann Surg 171:553-558, 1970 II Rienhoff WF Jr, Gannon J Jr, Sherman I: Closure of the bronchus following total pneumonectomy. Experimental and clinical observations. Ann Surg 116:481-531, 1942 12 Veith FJ, Koerner SK, Siegelman SS, Torres M, Attai LA, Takaro T, Gliedman ML: Single lung transplantation in experimental and human observations. Ann Surg 116:481-531,1942 13 Wrenn RN, Goldner JL, Markee JL: An experimental study of the effect of cortisone on the healing process and tensile strength of tendons. J Bone Joint Surg 36:588-60 I, 1954 14 Ross RR: Repair of tracheal and esophageal defect by use of a pedicle graft. Surgery 39:655-662, 1955 15 Shenstone NS: The use of intercostal muscle in the closure of bronchial fistulae. Ann Surg 104:560-571, 1936 16 Demons NJ, Timmes JJ: Myoplasty for closure oftracheobronchial fistula. Ann Thorac Surg 15:88-93, 1973