Improvement of tracheal autograft revascularization by means of fibroblast growth factor

Improvement of Tracheal Autograft Revascularization by Means of Fibroblast Growth Factor Johannes M. Albes, MD, Thomas Klenzner, MD, Jorg Kotzerke, MD, Klaus U. Thiedemann, PhD, Hans-Joachim Schafers, MD, and Hans-Georg Borst, MD Division of Thoracic and Cardiovascular Surgery, Surgical Center, and Department of Nuclear Medicine, Hannover Medical School; and Department of Ultrastructure Research, Fraunhofer Institute of Toxicology and Aerosol Research, Hannover, Germany

Ischemic airway complications after lung transplantation remain a significant problem despite the use of bronchial omentopexy. Clinical observations suggest that enhancement of vascular ingrowth could possibly increase the efficacy of a bronchial omental flap. This study was therefore designed to investigate whether basic fibroblast growth factor can enhance blood supply of an ischemic airway by acceleration of vascular ingrowth in a rabbit autotransplant model. Segments of the trachea were harvested and transplanted into a subcutaneous pouch. The animals were randomly assigned to one of four groups: group I, no omentopexy; group 11, omentopexy; group 111, omentopexy and fibrin glue; or group IV, omentopexy and fibrin glue enriched with 2.5 pg basic fibroblast growth factor. After 14 days the animals were sacrificed. The extent of perfusion was investigated by means of radioactive microspheres. The morphology of

the tracheal segments was investigated in a blinded fashion macroscopically, by means of light microscopy, and by means of scanning electron microscopy. The radioactivity measurements revealed a significantly increased perfusion of group IV (77% f 42%) as compared with groups I (17% 2 13%) and 111 (20% k 16%). By macroscopic and light microscopic assessment, the epithelial integrity of group IV was significantly improved compared with groups I and 11. At electron microscopy the integrity of group IV was significantly superior to all remaining groups. We conclude that a deposit of basic fibroblast growth factor and fibrin glue appears to increase revascularization of an ischemic airway from omentum and thus results in improved epithelial preservation of a tracheal autograft.

I

angiogenesis [5], seemed to be the appropriate substance for an experimental trial. The objective of this study, therefore, was to evaluate whether vascular ingrowth from omentum into the devascularized donor bronchus could be accelerated using basic fibroblast growth factor (bFGF), resulting in improved structural preservation.

schemic airway complications remain a major source of morbidity and mortality after lung transplantation despite advances in operative technique and graft preservation. The total deprivation of the physiologic blood supply of the donor bronchus has been acknowledged as the main reason for these problems [l, 21. Extensive laboratory initiatives seeking an improvement of the blood supply finally resulted in the use of omental wrapping of the bronchus. By means of this technique, the prevalence of bronchial complications was significantly reduced; bronchial ischemia and consequent healing problems, however, were still encountered [3]. Clinical observations showed that in these instances ischemia was often observed within the first postoperative week, ie, before the time when the omentum could be expected to have provided additional blood supply to the donor airway [4]. We therefore hypothesized that initial enhancement of vascular ingrowth from the omentum could possibly ameliorate bronchial ischemia. Fibroblast growth factor, one of the most potent promoters of Accepted for publication April 30, 1993. Address reprint requests to Dr Schafers, Klinik fur Thorax-, Herz- und GefaDchirurgie, Medizinische Hochschule Hannover, KonstantyGutschow-Str 8, 3000 Hannover 61, Germany.

0 1994 by The Society of Thoracic Surgeons

(Ann Thoruc Surg 1994;57:444-9)

Material a n d Methods A model of heterotopic autotransplantation of tracheal segments in rabbits was chosen. Adult male inbred rabbits (3.5 to 4.5 kg) were premedicated intramuscularly with 25 mgkg body weight ketamine. Anesthesia was induced with pentobarbital (30 mg) injected intravenously and maintained subsequently with additional doses of pentobarbital (12 mg). The animals breathed spontaneously during the entire operation. A neck incision was performed under sterile conditions, and the trachea was exposed. An eight-ring segment of the trachea was excised and an end-to-end anastomosis of the remaining trachea was performed using interrupted sutures (Vicryl 5-0; Ethicon, Hamburg, Germany). The excised segment was rinsed with saline solution, its ends were closed with interrupted sutures (Vicryl 5-0), and the segment was 0003-4975/94/$7.00

Ann Thorac Surg 1994;57444-9

A

Fig I . Excision of an eight-ring segment of the trachea, closure of both ends of the segment, and wrapping of the segment in the omental pedicle.

stored in saline solution (Fig 1). The neck incision was then closed in layers. A median incision was performed on the upper abdomen, and a subcutaneous pocket of sufficient size was created. Animals were randomly assigned to one of four treatment groups (n = 6 for each group). In group I (controls) the tracheal autograft was simply placed in the subcutaneous pocket. In group I1 the tracheal segment was wrapped in omentum and then placed in the subcutaneous pocket. In group I11 omental wrapping was performed with 2 mL of fibrin glue (Tissucol, Immuno GmbH, Heidelberg, Germany) (1 mL proteidaprotinin; 1 mL thrombin) before the segment was placed in the pocket. In group IV omental wrapping was performed with 2 mL of fibrin glue mixed with 2.5 pg bFGF (CR-fibroblast-growthfactor-basic; Biomol Feinchemikalien, Hamburg, Germany) and the autograft was placed in the subcutaneous pocket. In the study groups the omentum was mobilized through a 1- to 2-cm midline abdominal incision and the tracheal segment was wrapped in the omentum (see Fig 1). The edges of the omentum were approximated with resorbable sutures (Vicryl 5-0), ensuring that the blood supply was not compromised. The tracheal segment wrapped in omentum was then placed in the pouch. In groups 111 and IV, the omentum was wrapped around the autograft placed in the pouch with 2 mL of fibrin glue or with 2 mL of fibrin glue enriched with 2.5 pg bFGF (diluted in 0.5 mL saline solution) added to the protein component before application. The abdominal incision was closed in layers. The ani-

ALBESETAL IMPROVED TRACHEAL REVASCULARIZATION WITH bFGF

445

mals were allowed to recover and kept in single boxes with free access to food and water. Antibiotic prophylaxis was used with 400 mg of chloramphenicol given intramuscularly immediately after the operation and on the first postoperative day, followed by daily doses of 70 mglkg body weight added to the diet for the entire study period. Fourteen days postoperatively, the animals were anesthetized in an identical fashion. A tracheostomy was performed, and the animals were intubated with a 5-mm silicone tube and ventilated (tidal volume of 60 mL, 35 breathdmin, inspired oxygen fraction of 0.30). A median sternotomy was performed and the left atrial appendage exposed, and 5,000 IE heparin was given via the appendage; thereafter, 2 mL of microspheres (7 to 25 pm) (Human Albumin Microsphere Kit [TCK-5-S]; Sorin Biomedics, Divisione Diagnostici, Saluggia, Italy) labeled with technetium 99m (370 mBq activity), diluted in saline solution, was injected into the left atrium. After 1 minute the animals were sacrificed, the abdominal pouch was opened, and the omental tissue was carefully removed. The autograft was incised longitudinally and inspected macroscopically. A segment of the normal trachea was excised, and radioactivity of the normal trachea and the autograft was measured with a scintillation counter (Automatischer Gamma-Proben-Wechsler; LB MAG 312). Measurements were corrected for weight (counts per minute per gram). Activity of the autotransplant was computed as percentage of activity of the normal tracheal segment. The segments of the normal trachea and the autograft were separated into two parts. One part was fixed with 2% glutaraldehyde in 0.15 m o m Na-cacodylate buffer for scanning electron microscopic assessment. The other part was processed for light microscopic evaluation. For scanning electron microscopic processing, specimens were subsequently rinsed in several changes of cold 0.15 m o m Na-cacodylate buffer and postfixed in cold 1% osmiumtetroxide in 0.15 m o m Na-cacodylate buffer, rinsed in buffer solution, and dehydrated using a graded series of acetone under continuous agitation. The dehydrated tissue samples were critical-point dried from liquid CO, according to the method of Anderson [6] using a criticalpoint drying apparatus (E 3100; Polaron, Bio-Rad Lab GmbH, Miinchen, Germany), mounted epithelial side up on aluminium stubs with a conductive adhesive (Leit C nach Gocke; Neubauer, Miinster, Germany), sputtered with about 30 nm of gold (2.5 kV accelerating voltage, 20 mA current, 50 mm working distance) in a cool-sputter coater (Polaron) and kept under vacuum until scanning electron microscopic study. Scanning electron microscopic evaluation of the entire epithelial surface was performed using a scanning electron microscope (Stereoscan S360; Fa. Leica, Bensheim, Germany) at 15 kV accelerating voltage. The scanning electron microscopic observations were documented by recording micrographs at appropriate primary magnification on 35-mm film. Processing for light microscopy was performed by fixation with Bouin's solution, dehydration by means of increasing concentrations of alcohol, embedding in paraffin,

446

ALBESETAL IMPROVED TRACHEAL REVASCULAREATION WITH bFGF

cutting of 5- to 7-pm-thick cross-sections, and staining with hematoxylin and eosin. All morphometric examinations were performed in a blinded fashion by an independent investigator. For the macroscopic, light microscopic, and scanning electron microscopic examinations, a semiquantitative assessment was chosen. The specimens were assigned to one of three grades (poor, intermediate, good) for the macroscopic and light microscopic aspects and to one of four grades (poor, intermediate, good, excellent) for the scanning electron microscopic assessment. Macroscopically, specimens were good if cartilage and mucosal blood vessels appeared normal. Specimens were intermediate if the structure of the cartilage was normal but the mucosa as pale and the mucosal vessels were inhomogeneous or not visible. Specimens were poor if cartilage was not identifiable and the mucosa was not vital, or was replaced by apparent granulation tissue. At light microscopy, criteria for a good specimen were presence of the respiratory epithelium and normal structure of cartilage. Intermediate specimens showed destruction of the epithelium and impaired cartilaginous structure. Poor specimens exhibited absence of epithelium and necrotic changes in the cartilage. At scanning electron microscopy, the quality of the epithelium was investigated primarily with regard to percentage of normal-appearing epithelial lining of the basal membrane. Specimens were assigned one of four ratings: poor, less than 25% coverage of the total surface; intermediate, 25% to 50% coverage; good, 50% to 75% epithelial coverage; and excellent, more than 75% coverage. Statistical analysis was performed with unpaired t test for parametric data and Mann-Whipey U test for nonparametric data. Parametric data are expressed as mean standard deviation. All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Science and published by the National Institutes of Health (NIH publication 85-23, revised 1985).

*

Results Perfusion Transplanted segments from group I showed an average of 17% 13% of normal tracheal perfusion by analysis of radioactive microsphere activity. Group I1 had an activity of 40% 5 28% of the normal trachea; group 111, 20% 2 16%;and group IV, 77% 5 42%. Statistically, a significant difference was found between groups I and IV and between groups I11 and IV ( p < 0.05) (Fig 2).

*

Macroscopic Aspect In group 1, four specimens had poor cartilage and mucosal blood vessels and 2 specimens showed average findings (Table 1). In group I1 3 autografts were poor and 3

Ann Thorac Surg 1994;574449

125

i

'Oat 75

s

50 25 0

I (controls)

I1 (OW)

Ill (FIB)

IV

(bFGF) *=P<0.05

Fig 2 . Extent of perfusion assessed by measurement of radioactively labeled microspheres (n = 6 in each group). (bFGF = basic fibroblast growth factor and fibrin glue; FIB = fibrin glue; OW = omental wrapping.)

autografts were intermediate. In group 111, 2 specimens were classified as poor, 3 as intermediate, and 1 as good. In group IV 1 tracheal segment exhibited a poor aspect, 1 segment an intermediate aspect, and 4 segments a good aspect. No statistically significant differences were observed between group I and groups I1 and 111. The results in group IV were significantly better as compared with group I and group I1 ( p < 0.05).

Light Microscopic Assessment In group I, some epithelial insula were noticed, whereas uncovered cartilage was often found. Five specimens were therefore graded as poor and 1 specimen as intermediate. In group I1 and group 111, several specimens had an epithelial layer with variable degrees of destructive changes being present in the mucosa, submucosa, and cartilage. In group II,4 specimens presented a poor result and 2 specimens showed an intermediate aspect. Three group I11 specimens had poor aspects and 3 had intermediate aspects. The best histologic results were obtained in group IV. The epithelial layer was always present and in portions exhibited normal cilia. Submucosa and cartilage were almost normal. One specimen was classified as poor, 3 as intermediate, and 2 as good. Between groups I, 11, and 111 no statistically significant differences were found, whereas results in group IV were significantly superior when compared with groups I and I1 ( p < 0.05) (Table 2).

Table 1. Macroscopic Aspect Group" I (controls) I1 (OW) 111 (FIB) IV (bFGF) a

n

=

bFGF glue;

Poor

Intermediate

Good

4 3

2 3 3 1

... ...

2 1

1 4

6 in each group =

basic fibroblast growth factor and fibrin glue; OW = omental wrapping.

FIB

=

fibrin

ALBESETAL IMPROVED TRACHEAL REVASCULARIZATION WITH bFGF

Ann Thorac Surg 1994;5744&9

447

Table 2. Light microscon Group” I (controls) I1 (OW) 111 (FIB) IV (bFGF) a

n

=

bFGF glue;

Poor

Intermediate

Good

5

1 2 3 3

... ... ...

4

3 1

2

6 in each group. =

basic fibroblast growth factor and fibrin glue; OW = omental wrapping.

FIB

=

fibrin

Scanning Electron Microscopic Assessment Almost all specimens in group I revealed an absence of epithelium over up to 90% of the surface (Fig 3). Only 2 specimens showed sparse intact epithelial insula. Four specimens were therefore classified as poor and 2 specimens as intermediate. In group I1 we found 2 poor specimens, 3 intermediate specimens, and 1 good specimen. In group I11 2 specimens showed a poor aspect, 2 specimens exhibited an intermediate aspect, and 2 specimens were classified as good. In group IV all but 1 specimen were classified as good (n = 3) or even excellent (n = 2). The epithelial lining of these specimens consisted of variable proportions of flat, juvenile cells (comparable with those shown in Figure 4) and normal-appearing ciliated epithelium (Fig 5 ) . The statistical evaluation revealed no significant differences between groups I through 111, whereas a significant difference between group IV and the remaining groups was present ( p < 0.05) (Table 3).

Fig 4. Scanning electron photomicrograph of a specimen from group IV showing intact epithelial layer with two solitary ciliated cells. (Original magnification, ~3,500.)

In lung transplantation, ischemia of the donor bronchus and its sequelae continue to be a significant problem [7]. Ischemia often results in bronchial ulceration, which can subsequently lead to stenosis or dehiscence of the anastomosis [8]. Attempts to enhance blood supply led to use

of omental wrapping, an idea that dates back to 1906, when Morrison [9] first saw the capability of the omentum to maintain viable tissue without apparent blood supply. Based on promising experimental data, this technique entered lung transplantation with its growing demand for secure coverage of bronchial anastomoses [ 101. However, with omental wrapping alone the complications could be lowered but not completely eradicated [4]. In recent years, alternative approaches to improve the blood supply have been evaluated. Revascularization of the bronchial arterial tree restored the physiologic blood supply and resulted in promising data in animal experiments [ll].However, this approach is technically demanding and increases the ischemic time of the donor organ; thus it has not found wide acceptance in the surgical routine. Improvement of the microcirculation with drugs such as heparin or prostaglandin I, has exhibited certain positive effects [12]. At the present time an acceptable but by no means satisfactory state of the art has been achieved with the incidence of airway complication still exceeding 10% in the experience of several centers [4].

Fig 3. Scanning electron photomicrograph of a specimen from group I showing uncovered basal membrane and cell debris. (Original magnification, ~ 6 0 0 . )

Fig 5. Scanning electron photomicrograph of a specimen from group IV showing intact epithelial layer exhibiting a high percentage of ciliated cells. (Original magnification, x1,250.)

Comment

448

Table 3. Electron Microscorn Groupa I (controls) I1 (OW)

111 (FIB) IV @FGF) a

n

=

Ann Thorac Surg 1994;574.44-9

ALBESETAL IMPROVED TRACHEAL REVASCULARIZATION WITH bFGF

Poor

Intermediate (25%-50%)

(50%-75%)

Good

Excellent

4 2 2

2 3 2 1

...

...

1 2 3

...

(<25%)

...

(>75%)

... 2

6 in each group.

bFGF = basic fibroblast growth factor and fibrin glue; glue; OW = omental wrapping.

FIB

=

fibrin

The clinical observation of ischemic problems occurring within the first 2 weeks postoperatively suggests that vascular ingrowth from the omentum might come too late to prevent ischemia. This theory is supported by good clinical results of recent series in which omental wrapping was omitted [13]. Tracheal healing problems have been a focus of intense surgical interest for decades. First attempts of tracheal replacement in the mid-1950s resulted in poor healing of tracheal autografts, homografts, and allografts [14-171. In addition, it was shown that revascularization of an autograft was greatly dependent on the length of the graft, supporting the theory that these segments are revascularized from their ends rather than from their midportion. In a canine autograft model, Balderman and Weinblatt [18] demonstrated the limitations of tracheal healing despite the use of omentum. However, a new approach to this issue has arisen in the last years. In 1986 Goldsmith and associates [19] proposed the existence of a growth factor that increases revascularization in different tissues in a lipid fraction of the omentum. Shortly thereafter, fibroblast growth factors were shown to be potent promoters of angiogenesis [20]. Various growth factors have been isolated from almost every human tissue [21]. At present, acidic and basic fibroblast growth factor are most commonly investigated [22]. In 1989 Olech and associates [23] started experiments with bFGF in a rabbit autotransplant model to investigate the assumed effect on tracheal healing by applying bFGF in a dosage of 10 ng. Unfortunately, no increase in revascularization was achieved [23]. The lack of the desired effect possibly was due to inadequate dosage and the limited application time. Enhanced revascularization was later realized, however, by Mayer and colleagues [24] with an increased dosage of 400 ng/mL and prolonged application. In their experimental setting, they used a pump for continuous supply of bFGF. In our experimental design, we therefore employed a deposit of bFGF in fibrin glue to combine a prolonged application time with a clinically practicable mode of application. Furthermore we used a dosage 250 times higher than that applied by Olech and associates. A reasonable consistency of the four different methods used was found in our experiments. In each morphologic examination, the controls exhibited the poorest results. Tracheal segments covered with omentum enriched with bFGF presented a significantly better morphologic integ-

rity of the graft than controls, omentum alone, and omentum plus fibrin glue. The electron microscopy in particular revealed the superiority of the bFGF group in terms of epithelial protection. This is of special interest, because reepithelialization of exposed cartilage is known to be poor. Therefore, an initial loss of epithelium should be avoided [25]. In the literature fibrin glue alone [26] as well as omentum alone can enhance angiogenesis. In our investigation this could not be proved statistically. However, a trend toward better healing of the autotransplants covered with omentum and omentum plus fibrin glue was observed. This underlines the assumption that omentum alone or omentum in combination with fibrin glue is not sufficient for improved healing. Study of the microcirculation by means of radioactively labeled microspheres clearly supported the morphologic results and revealed an almost normal blood supply in the bFGF group, significantly exceeding the effect attained in the remaining groups. These findings also support the hypothesis that the positive effect on the epithelium of bFGF is due to an improved revascularization rather than a proposed mitogenic effect [24]. We conclude from our investigation that bFGF can improve revascularization and thus preserve the integrity of the epithelium of a rabbit tracheal autotransplant in the critical initial postoperative period if the drug is generously applied in a deposit of fibrin glue using omentopexy. In contrast, omentopexy alone or in combination with fibrin glue is significantly less effective. Our results encourage us to use fibroblast growth factor in a clinical trial for lung transplantation, especially as toxic side effects in several clinical investigations have not been observed [27].

References 1. Veith FJ. Lung transplantation. Surg Clin North Am 1978;58: 357-64. 2. Mills NL, Boyd AD, Gheranpong C. The sigruficance of bronchial circulation in lung transplantation. J Thorac Cardiovasc Surg 1970;60866-78. 3. Cooper JD, Pearson FG, Patterson GA, et al. Technique of successful lung transplantation in humans. J Thorac Cardiovasc Surg 1987;93:173-81. 4. Schafers HJ, Haydock DA, Cooper JD. The prevalence of and management of bronchial anastomotic complications in lung transplantation. J Thorac Cardiovasc Surg 1991;101:1044-52. 5. Schweigerer L. FGF and angiogenesis. Z Kardiol 1989; 78:12-5. 6. Anderson TF. Techniques for the preservation of three dimensional structures in preparing specimens for the electron microscope. Acta Clin Belg 1951;13:321-8. 7. Schafers HJ, Cooper JD, Patterson GA, Toronto Lung Trans8.

9. 10. 11

12.

plantation Group. Airway complications following lung transplantation. Am Rev Respir Dis 1988;137214. Griffth BP, Hardesty RL, Trento A, et al. Heart lung transplantation: lessons learned and future hopes. Ann Thorac Surg 1987;43:6-16. Momson R. Remarks on some function of the omentum. Br Med J 1906;1:76-9. Lima 0, Goldbere M, Peters WT, Avabe H, Townsend E, Cooper JD. Bronckal omentope+ in’canine lung transplantation. J Thorac Cardiovasc Surg 1982;83:418-21. Siegelman SS, Hagstrom JWC, Koemer SK, Veith FJ. Restoration of bronchial artery circulation after canine lung allotransdantation. 1 Thorac Cardiovasc Sure 1977;73:792-5. Inui k, Schafers’ HJ, Aoki M, et al. EffGct of methylpred-

Ann Thorac Surg 1994;574.4&9

13. 14. 15. 16. 17.

18. 19. 20.

nisolone and prostacyclin on bronchial perfusion in lung transplantation. Ann Thorac Surg 1993;55:364-9. Calhoon JH, Grover FL, Gibbons WJ, et al. Single lung transplantation: alternative indications and technique. J Thorac Cardiovasc Surg 1991;101:816-25. Belsley R. Resection and reconstruction of intrathoracic trachea. Br J Surg 1950;38:200-5. Craig RL, Holmes GU, Shabart EJ. Tracheal resection with replacement with a prosthesis. J Thorac Surg 1953;25:38&96. Bryant LR. Replacement of tracheobronchial defects with autogenous pericardium. J Thorac Cardiovasc Surg 1964;48: 733-40. Pearson FG, Henderson RD, Gross AE, et al. The reconstruction of tracheal defects with a porous prosthesis: an experimental and clinical study using heavy Marlex mesh. J Thorac Cardiovasc Surg 1968;55:60516. Balderman SC, Weinblatt G. Tracheal autograft revascularization. J Thorac Cardiovasc Surg 1987;94:434-41. Goldsmith HS, Griffith AL, Catsimpoolas N. Increased vascular perfusion after administration of an omental lipid fraction. Surg Gynecol Obstet 1986;162:579-83. Folkman J, Klagsbrun M. Angiogenic factors. Science 1987; 235:4424.

ALBESETAL IMPROVED TRACHEAL REVASCULARIZATION WITH bFGF

449

21. Montesano R, Vassalli JD, Baird A, Guillemin R, Orci L. Basic fibroblast growth factor induces angiogenesis in vitro. Proc Natl Acad Sci 1986;83:7297-301. 22. Risau W, Zweres HG. Role of growth factors in the formation of blood vessels. Z Kardiol 1989;78:9-11. 23. Olech VM, Keshavjes SH, Chamberlain D, Slutsky AS, Patterson GA. Role of basic fibroblast growth factor in revascularization of rabbit tracheal autografts. Ann Thorac Surg 1991;52:25844. 24. Mayer E, Cardoso PFG, Puskas JD, et al. The effect of basic fibroblast growth factor and omentopexy on revascularkation and epithelial regeneration of heterotopic rat tracheal isografts. J Thorac Cardiovasc Surg 1992;104:180-8. 25. Eisenstein R, Kuettner WE, Neapolitan C, et al. Cartilage extracts inhibit the growth of fibroblasts and endothelial cells in culture. Am J Pathol 1975;81:33743. 26. Dvorak HF, Harvey VS, Esterella P, Brown LF, McDonagh J, Dvorak AM. Fibrin containing gels induce angiogenesis. Lab Invest 1987;5767=6. 27. Brown GL, Nanny LB, Griffen J, et al. Enhancement of wound healing by topical treatment with epidermal growth factor. N Engl J Med 1989;321:7&9.