R atric patients for congenital stenosis, tracheotomy

Transcription

1 Growth of Tracheal Anastomoses: Advantage of Absorbable Interrupted Sutures Peter P. McKeown, FRCS(C), FRACS, Hidetoshi Tsuboi, MD, Takao Togo, MD, Robert Thomas, BA, Richard Tuck, BA, and David Gordon, MD Division of Cardiothoracic Surgery, Department of Surgery, and Department of Pathology, University of Washington, Seattle, Washington Growth of the trachea after complete transection and anastomosis was studied in four groups of 1-month-old New Zealand white rabbits. The trachea was transected at the fifth cartilaginous ring and then anastomosed with continuous 6-0 polypropylene (Prolene) (group 11, interrupted 6-0 polypropylene (group 21, continuous 6-0 polydioxanone (PDS) (group 3), or interrupted 6-0 PDS (group 4). The animals were followed up for 90 to 103 days (mean follow-up, 95 days). At the time the animals were killed, body weight had increased 125% (1.2 to 2.7 f 0.18 kg). Growth of the trachea was assessed at the time of death. Results from this study suggest that growth of a tracheal anastomosis is retarded in a growing animal model. The degree of resultant stenosis was significantly less when an absorbable suture material (PDS) and an interrupted suturing technique were used. (Ann Thorac Surg 1991;51:636-41) econstruction of the trachea may be required in pedi- R atric patients for congenital stenosis, tracheotomy and postintubation injuries, or tracheomalacia secondary to vascular compression. Tracheal anastomosis in these instances and in pediatric heart and lung transplantation carries the risk of stenosis at the site of anastomosis. Early experimental work by Maeda and Grillo [l] and Mendez- Picon and colleagues [2] demonstrated that satisfactory growth of the trachea was possible in puppies, although narrowing did occur. Controversy exists as to the best method of anastomosis and the choice of suture material [3-71. Clinically, Grillo [8] found absorbable suture (Vicryl; Ethicon, Johnson & Johnson, Somerville, NJ) to be advantageous for reconstruction of the trachea because of a decrease in granulation tissue at the anastomosis. In a series of pediatric patients, Alstrup and Sorensen [9] achieved satisfactory results using interrupted Dexon (Davis + Geck, Danbury, CT) or chromic catgut sutures. Polydioxanone (PDS; Ethicon) is an absorbable suture with some theoretical advantages, including monofilament structure and predictable absorption rates, that should make it even more suitable for tracheal anastomoses [lo-131. This study was designed to compare the use of interrupted versus continuous suturing techniques and absorbable (PDS) versus nonabsorbable (polypropylene [Prolene]; Ethicon) suture material by evaluating the degree of postsurgical stenosis of the trachea in a growing animal model. Accepted for publication Dec 28, Address reprint requests to Dr McKeown, Division of Cardiothoracic Surgery, Department of Surgery, University of South Florida, 4 Columbia Dr, Suite 730, Tampa, FL Material and Methods One-month-old New Zealand white rabbits (mean weight, 1.2 kg) were randomized to four equal study groups (n = 7). The rabbits were anesthetized with intramuscular administration of xylazine (5 mg/kg) and cyclohexylamine (35 mg/kg) and allowed to breathe spontaneously without intubation. Lidocaine hydrochloride (2%) (3 ml) was injected into the tissues in the region of the neck incision to decrease discomfort. Under sterile conditions, a vertical incision was made over the cervical trachea, the strap muscles were retracted, and the trachea was transected at the fifth cartilaginous ring. Mobilization of the trachea was limited to prevent interference with the blood supply. Internal and external dimensions in the anteroposterior and lateral planes were measured with fine calipers. Measurements were taken at the site of anastomosis (A) and at 5 mm above and 5 mm below each anastomosis. Normal (N) values were obtained as an average of those values from above and below the anastomosis. The cross-sectional area (CSA) was calculated using the formula CSA = (a121 (b/2) * T, where a = lateral diameter and b = anteroposterior diameter [14]. The tracheas were then anastomosed with continuous 6-0 polypropylene (group l), interrupted 6-0 polypropylene (group 2), continuous 6-0 PDS (group 3), or interrupted 6-0 PDS (group 4). The animals were electively killed after 90 to 103 days (mean follow-up, 95 days) when they had reached full size. During this period, there was a 125% increase in body weight to 2.7 * 0.18 kg. At the time of death, tracheal measurements were taken with fine calipers. The trachea was excised and fixed in 10% buffered formalin. Silastic (Dow Corning) casts were made of each trachea, by The Society of Thoracic Surgeons /91/$:3.50

2 Ann Thorac Surg McKEOWN ET AL 637 Table 1. lnternal Diameters Measured in the Four Groups Group I A N Lateral dimension (mm) 1: Continuous Prolene 2: Interrupted Prolene 3: Continuous PDS 4: Interrupted PDS Anteroposterior dimension (mm) 1: Continuous Prolene 2 Interrupted Prolene 3: Continuous PDS 4: Interrupted PDS 4.4 f f ? ? f f f f f f f 0.65b 3.9 f f f f f f f f f 0.53 A/N 0.73 f f f f O.llb 0.72 t f f f O.Wb a Data are shown as the mean t one standard deviation. A = anastomosis at death; below the anastomosis). Significance: p < 0.05, one-way analysis of variance. AIN = stenosis ratio; I = initial; N = normal diameter at death (average of the diameter 5 mm above and 5 mm and caliper measurements were again recorded at the site of anastomosis, 5 mm above and 5 mm below the anastomosis. Measurements from the casts were found to be more reproducible and consistent; therefore these data rather than measurements from the freshly killed specimens were used for the analysis. Statistical analysis of the data was performed using one-way analysis of variance and the unpaired Student s t test. Statistical significance was defined as a probability value of 0.05 or less. All experiments were performed in accordance 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 (NIH publication No , revised 1985). Results Four animals died intraoperatively of anesthesia-related problems. Three early deaths occurred on days 4, 8, and 13 and were due to aspiration pneumonia and infection, which could have been related to technical factors. There appeared to be no predilection for any one group and no relation to suture material. Therefore these animals were excluded from the analysis, and additional animals were used to complete the study groups. All surviving animals were killed an average of 95 days after transection and anastomosis of the trachea. Measurements There were no significant differences between the groups in body weight at the time of operation or death. The lateral and anteroposterior internal diameters and crosssectional areas of the trachea were also similar at the time of transection of the trachea. Lateral diameters were uniformly greater than anteroposterior diameters both initially and at death, a finding consistent with the recognized cross-sectional profile of the trachea. Internal diameters of both the lateral and anteroposterior dimensions for all four groups are recorded in Table 1, and calculated cross-sectional areas for all groups are shown in Table 2. Representative specimens of Silastic casts are depicted in Figure 1 (continuous polypropylene, group 1) and Figure 2 (interrupted PDS, group 4). The A/N ratio for both diameters and the cross-sectional area of the anastomoses showed that group 4 (interrupted PDS) was significantly superior to any other group (see Tables 1, 2). A separate analysis was performed using the AIN area ratio data for continuous (groups 1 and 3) versus interrupted technique (groups 2 and 4) and Prolene (groups 1 Table 2. Calculated Cross-Sectional Areas for the Four Groups Area (mm2) Group I A N A/N 1: Continuous Prolene 13.1? f f f : Interrupted Prolene 12.4 f t f f : Continuous PDS 12.0 f f f f Interrupted PDS 11.9 f f 3.30b f 0.08b a Data are shown as the mean t one standard deviation. Abbreviations are the same as in Table 1. Significance: p , one-way analysis of variance

3 638 McKEOWN ET AL Ann Thorirc Surg Table 3. AIN Area", Variable Continuous versus interrupted suture Continuous Interrupted p Value Prolene versus PDS Prolene PDS Value t 0.14 C :k :t 0.14 p Value C0.025 Fig 1. Silastic cast of tracheal anastomosis in an animal in group 1 (continuous Prolene). This was an extreme example of the narrowing that may occur. and 2) versus PDS (groups 3 and 4). Using the one-tailed unpaired t test, the differences were found to significantly favor interrupted over continuous technique (p < 0.005) and PDS over Prolene (p < 0.025) (Table 3). Histology Selected samples from each group were submitted for histological study. Histological sections of the tracheal anastomoses revealed persistent suture material in all four groups with an associated foreign-body giant cell reaction and a small amount of surrounding fibrosis (Figs 3, 4). PDS suture material revealed evidence of advanced hydrolysis and degradation (see Fig 4b), whereas the polypropylene material showed no apparent degeneration. Within the submucosa at the site of anastomosis, occasional small collections of mononuclear inflammatory cells (primarily lymphocytes) were seen (see Fig 4a). Focal areas of mucosal fibrosis, forming a small ridge that protruded into the lumen, were frequently present (see Fig 3a). Such protrusions in all anastomoses were always Fig 2. A fairly representative example of Silastic cast of tracheal anastomosis in an animal in group 4 (interrupted PDS). The region of the anastomosis is approximately at the midpoint of the specimen and shows only minimal deformity. a Data are shown as the mean f the standard deviation for A/N ratio for area. The p values were obtained using the one-tailed unpaired t test. A/N = ratio of the measurement of the anastomosis at death to the normal measurement at death (average of the value 5 mm above and 5 mm below the anastomosis). covered by respiratory epithelium. The mucosal elastica was disrupted and absent in the region of the anastomosis. Given the limited sampling, we were unable to quantitate the relative amounts of inflammation between the studied groups. Comment Several experimental studies [l-3, 151 demonstrating anastomotic growth of the trachea have been previously published. In 1972, Maeda and Grillo [l] reported their canine studies, which showed that growth of the trachea was possible after transection. The average growth rate was 82% of normal in the sagittal plane and 75% of normal in the coronal plane. Both nonabsorbable (4-0 Tevdek; Deknatel, Fall River, MA) and absorbable (4-0 chromic catgut) suture materials were used, but comparisons were not reported. Overlapping of the cartilaginous rings and scar formation were thought to contribute to stenosis of the anastomosis. Maeda and Grillo [l] demonstrated in a subsequent study the negative effects of tension on tracheal anastomoses after resection of part of the trachea. Mendez-Picon and associates [2] used interrupted 4-0 silk sutures after tracheal resection in puppies, and although there were 5 long-term survivors with satisfactory results, 3 animals died early with fibrotic stenosis. Over- Lapping of the cartilaginous rings again was considered to be a prime factor in the development of stricture at the anastomosis. In a study using pigs, Murphy and co-,workers [3] reported that overlapping of the cartilages :nay be disadvantageous and that a two-layer technique was useful. They used running and interrupted Dexon on the mucosa and interrupted silk on the tracheal rings. After simple transection, the edges of the trachea in our animals were approximated so as to avoid telescoping or overlapping of the rings, although some microscopic overlapping was seen histologically. Maeda and Grillo [l] suggested that relative elongation of the membranous portion of the trachea leads to an

4 Ann Thorac Surg McKEOWN ET AL 639 Fig 3. (a) Cross section of tracheal anastomotic site with polypropylene suture material (arrow). An intact respiratory epithelium (E) covers the luminul surface, and a focus of subepithelial fibrosis (F) is present. The cartilaginous rings (C) appear unremarkable except for focal calcification. (b) Higher-magnification photomicrograph of the suture-containing region in Figure 3a indicated by the arrow. It shows suture material (S) surrounded by multinuclear giant cells. The empty hole indicates the spot where another piece of suture material has fallen out of the tissue section. (a, X50 before 32% reduction, b; x.500 before 32% reduction.) (L = lumen.) alteration in the cross-sectional shape of the trachea with growth and that the membranous area is an important region of growth. Burrington [ 171 demonstrated fairly uniform growth on the convex surface in the cartilaginous portion. Remodeling progressed on the concave surface. A linear relationship exists between body weight and tracheal cross-sectional area [17]. This suggests that an interrupted method of suturing or absorbable sutures would be less likely to interfere with the pattern of growth. Our study confirms this. Nordin and Ohlsen [7], in an effort to avoid stenosis at the anastomotic site, used a Z-plasty technique with interrupted sutures. Rather than perform Z-plasty, we chose to look at the best method and the best suture material for a simple end-to-end anastomosis. Resection of the cartilaginous rings was not performed because of the risk of interrupting the blood supply of the trachea [18] and because we wanted to avoid the detrimental effects of tension on the anastomoses [16]. Lateral dissection was kept to a minimum. Previous studies presented conflicting results between absorbable and nonabsorbable sutures. Sezeur and colleagues [6] suggested that polyglycolic acid (Dexon) and braided polyglactin (Vicryl) were not ideal and had little advantage over nylon or polypropylene. Nordin and Ohlsen [7] preferred nonabsorbable suture to polyglycolic acid and suggested there was an increased chance of separation and stenosis with absorbable suture. Their study was limited by the number of study animals. Hsieh and co-workers [15] found in a canine model and in a study similar to ours that absorbable suture (Dexon-S or Vicryl) was superior to nonabsorbable suture (nylon or Prolene) for left bronchial anastomosis. Granulomas were not found with absorbable sutures. However, the numbers in each group were insufficient for a satisfactory comparison between continuous and interrupted sutures.

5 640 McKEOWN ET AL Ann Thorac Surg 1991;51: Fig 4. (a) Cross section of tracheal anastomotic site with PDS suture material (arrow). Microscopic overlapping of the cartilaginous rings (C) is present (arrowhead) but causes litfle to no bulging into the lumen 6). One small focus of mononuclear inflammatory cell infiltrate (M) is present just under the respiratory epithelium. (b) Higher-magnification photomicrograph of the suture-containing region in Figure 4a indicated by the arrow. It shows a surrounding macrophage and multinuclear giant cell reaction. The enclosed suture material 6) is fragmented. (a, x80 before 32% reduction, b, x500 before 32% reduction.) In a study involving rabbits and dogs, Scheele and colleagues [4] found polypropylene and the absorbable sutures Dexon and Vicryl caused less reaction and less granulation and abscess formation than Mersilene (Ethicon). Dexon was successfully used in another canine study [5], which showed that by 4 months, the suture material had disappeared from the tracheal and bronchial anastomoses. Based on his extensive clinical experience, Grillo [8] thought that absorbable sutures (Vicryl) were less likely to cause granulations. Polydioxanone (PDS) theoretically has some advantages over polyglactin (Vicryl) and polyglycolic acid (Dexon). It is a monofilament, is more flexible, retains tensile strength longer, and may cause less inflammation [lo-121. Complete hydrolytic absorption of PDS was shown to occur within 26 weeks in muscle and subcuticular tissue of rats (121. Catgut was thought to be inferior to these newer absorbable sutures [ll]. PDS retains 58% of its strength at 4 weeks. If the suture material persisted for a protracted period, it could conceivably affect growth at the anastomosis. In our study, there was still some residual suture in the anastomoses in the four PDS tracheas studied histologically, but the suture material appeared clear with moderate degradation. Hydrolytic absorption was well advanced (see Fig 4b). The inflammatory reaction was mild in all 4 PDS animals, and only 1 of them had moderate fibrosis at the suture line. Incursion into the lumen by a ridge of tissue was noted in both of the polypropylene animals (see Fig 3a). Some of this appeared to be associated with formation of scar tissue. Overlapping of the cartilages was observed in 1 of them. The number of anastomoses studied histologically is insufficient to make a comparative statement regarding the inflammatory reaction of PDS versus polypropylene suture material. No gross evidence of abscess formation, infection, or dehiscence was seen in any of the four groups. The results

6 Ann Thorac Surg McKEOWN ET AL 641 in the interrupted PDS group (group 4) were significantly better than in the other three groups. Nonetheless, a satisfactory functional outcome was achieved in the majority of animals studied. Previous studies [19,20] suggest that narrowing of the tracheal lumen by as much as 50% to 75% is required before symptomatic evidence of respiratory compromise occurs. Formalin may have induced some postfixation shrinkage of the tracheal segments, but difficulty in measuring such small specimens in the fresh state with consistent accuracy led to the use of the Silastic cast method. The measurements of the casts were technically easier, more reproducible, and less variable and were thought to be more precise than those taken in the fresh state. Diameters measured in the fresh state at harvest and measurements from the Silastic casts after the death of the animal are noted in Table 1. Calculated areas from these data are listed in Tables 2 and 3. Because each specimen served as its own control with measurements taken above and below the anastomosis, the comparative analysis remains valid. The mean area of normal to area of anastomosis (A/N area) ratio for all groups was 0.63 with no group being less than 50% (see Table 3). Considering that all of the animals had tracheal diameters of 5 mm or less at the time of anastomosis, these results appear satisfactory. Although there was little clinical evidence of functional impairment, growth at the anastornotic site was nonetheless inhibited in all groups (see Tables 1, 2). Conclusion In a comparison of polydioxanone versus polypropylene and interrupted versus continuous suturing techniques, we reached the following conclusions: (1) in a growing animal model, transection of the trachea with anastomosis resulted in stenosis and retardation of growth at the anastomotic site, and (2) the use of interrupted polydioxanone sutures in tracheal anastomoses proved to be significantly superior to the other methods. We recognize the contributions of Chris Larsen and the work of Jaci Wilkinson, Pat Conant, ARNP, and Bonnie Heath in the preparation and typing of the manuscript, and the Robert Wood Johnson Foundation for its support of Dr Gordon. References 1. Maeda M, Grillo HC. Tracheal growth following anastomosis in puppies. J Thorac Cardiovasc Surg 1972;64:30& Mendez-Picon G, Hutcher NE, Neifeld J, Salzberg AM. Long-term study of tracheal growth after segmental resection in puppies. J Pediatr Surg 1974;9: Murphy DA, Dunn GL, Poirier N, Martin M. Growth of tracheal anastomoses: an experimental study in weanling pigs. Ann Thorac Surg 1973;16: Scheele J, Gentsch HH, Hoffmann W, Pesch HJ. Anastomosentechnik an der Trachea. Laryngol Rhino1 Otol (Stuttg) 1982;61: Ohata M, Idia M, Omori K, Hashimoto N, Sezai Y. Experimental study on ultrastructure of tracheobronchial anastomosis. Risho Kyobu Geka 1981;1:56> Sezeur A, Leandri J, Rey P, Daumet P, Vouron J. etude experimentale du comportement des fils synthetiques A resorption lente dans les sutures tracheales. Ann Chir 1982;36: Nordin U, Ohlsen L. Prevention of tracheal stricture in end-to-end anastomosis. Arch Otolaryngol 1982;108: Grillo HC. Tracheal surgery. Scand J Thorac Cardiovasc Surg 1983; Alstrup P, Sorensen HR. Resection of acquired tracheal stenosis in childhood. J Thorac Cardiovasc Surg 1984; Berry AR, Wilson MC, Thomson JWW, McNair TJ. Polydioxanone: a new synthetic absorbable suture. J R Coll Surg Edinb 1981;26: Lerwick E. Studies on the efficacy and safety of polydioxanone monofilament absorbable suture. Surg Gynecol Obstet 1983;156: Ray JA, Doddi N, Regula D, Williams JA, Melveger A. Polydioxanone (PDS), a novel monofilament synthetic absorbable suture. Surg Gynecol Obstet 1981;153: Artandi C. A revolution in sutures. Surg Gynecol Obstet 1980;150: LeGal YM, Chittal SM, Wright ES. Early bronchial revascularization with an intercostal pedicle graft following canine lung autotransplantation. Can J Surg 1985;28: Hsieh C, Tomita M, Ayabe H, et al. Influence of suture on bronchial anastomosis in growing puppies. J Thorac Cardiovasc Surg 1988; Maeda M, Grillo HC. Effect of tension on tracheal growth after resection and anastomosis in puppies. J Thorac Cardiovasc Surg 1973;65:65% Burrington JD. Tracheal growth and healing. J Thorac Cardiovasc Surg 1978;76: Salassa JR, Pearson BW, Payne WS. Gross and microscopical blood supply of the trachea. Ann Thorac Surg 1977;24:10& Lindholm CE. Prolonged endotracheal intubation. Acta Anaesthesiol Scand [Suppl] 1970;33: Ferguson DJ, Wild SJ, Wagensteen OH. Experimental resection of the trachea. Surgery 1950;28: