Sutureless Cartilage Graft Laryngotracheal Reconstruction Using Fibrin Sealant FREE

THE MANAGEMENT of subglottic stenosis has been a difficult problem in the past, with most children requiring long-term tracheotomy, with high morbidity and mortality. During the last 20 years, there have been significant advancements in the surgical treatment of this problem. Presently, many children can be decannulated of their tracheotomy after successful surgical correction. Laryngotracheal reconstruction with cartilage grafting has gained recent popularity with a high rate of successful decannulation.^{1 - 3} The procedure is technically challenging because many of these infants are very small and the thickness of the cartilage graft and the trachea is only a few millimeters. Failures of the procedure can occur secondary to infection and dissolution of the cartilage that ultimately causes failure of the cartilage graft to heal. Animal studies have demonstrated that the infections occur primarily along the needle tracts of the suture in the cartilage graft.⁴

Fibrin sealant, formed by mixing fibrinogen and thrombin, was used originally as a hemostatic agent. More recently, it has been used to improve wound healing and as an adhesive agent.^{5 - 8} The American Red Cross J. Holland Laboratory, Rockville, Md, has developed a viral deactivated pooled fibrin sealant. More than 1 million patients during the last 12 years have been treated with similar products without a single case of disease transmission.

The potential advantages of using fibrin sealant instead of suture are (1) shorter operative time, (2) decreased manipulation and technical difficulty in securing the graft, and (3) avoidance of suture tracts, which have previously shown a predisposition to infection.

A pilot study with 2 animals determined that fibrin sealant without sutures can secure a free cartilage graft in the trachea. This study was undertaken to compare the results of laryngotracheal reconstruction using fibrin sealant alone vs sutures.

The Scientific Review Committee and Laboratory Animal Care and Use Committee approved the research project. The animals involved in this study were procured, maintained, and used in accordance with the Animal Welfare Act of 1966, as amended, and the Guide for the Care and Use of Laboratory Animals prepared by the Institute of Laboratory Animal Resources, National Academy of Sciences–National Research Council. The ferret animal model was used. Ferrets have been demonstrated to be a good model for the infant airway.⁹ All animals were vaccinated with canine distemper and given a diet of water and dry ferret food for 1 week after arrival at the facility to ensure acclimation. Forty ferrets were assigned (convenience sampling) initially to 2 different groups of 20 animals each as follows: (1) suture only (4 sutures were used to stabilize the cartilage graft); (2) fibrin sealant only (American Red Cross J. Holland Laboratory fibrin sealant was used to stabilize the graft). Within these groups were 2 subsets determined by time of examination after surgery (postoperative days 7 and 30).

Adequate anesthesia using 2% isoflurane via endotracheal intubation was obtained. Preoperative medications included the following: intramuscular atropine, 0.05 mg/kg (average weight, 112 kg); intramuscular ketamine hydrochloride, 40 mg/kg; intramuscular penicillin G benzathine, 200000 U; and lidocaine hydrochloride, 4 mg/kg. After securing the endotracheal tube, the head was then extended with a shoulder roll and the hair over the ventral surface of the neck as well as over the right costal margin on the chest was shaved and prepared with povidone-iodine scrub. Drapes were placed sterilely. The free costal margin on the chest was identified and an incision made down to the perichondrium. The soft tissue over the perichondrium was dissected free circumferentially. Approximately 2.5 cm of the cartilage was removed and meticulous hemostasis was obtained using bipolar electrocautery. Positive pressure to 30 cm H₂O was applied to check for pleural leaks. The wound was then closed in layers. The costal cartilage graft (CCG) with its perichondrium was kept moist with isotonic sodium chloride in sterile gauze while the recipient site was being prepared. The ferret's thorax and lower body were then covered with a warming blanket to help maintain homeostasis throughout the operative procedure. A vertical cervical skin incision at the level of the cricoid was made and subplatysmal skin flaps were elevated laterally. The midline raphe of the strap muscles was identified, separated in the midline, and dissection carried down to the trachea. If the thyroid isthmus was encountered, it was divided in the midline using bipolar cautery. Dissection was carried to the pretracheal fascia, and the thyroid, cricoid, and tracheal cartilages were identified. The length of the planned anterior laryngotracheal incision extending from the lower thyroid cartilage to the second tracheal ring was measured (approximately 2 cm). The wound was covered with moist gauze while the CCG was carved into a "boat" shape. Care was taken to leave perichondrium on the intraluminal surface. An anterior laryngotracheal incision was then made in the midline, the opening distracted with forceps. The boat-shaped CCG was placed in the incision and fixed in place in the selected fashion depending on the study group. Hemostasis was obtained and the incision closed in layers using 5-0 monofilament dissolvable sutures. The time needed to stabilize the cartilage graft was observed for comparison between the 2 study groups.

The fibrin sealant was prepared as follows: At the beginning of the day of the procedures, the lyophilized fibrinogen was reconstituted with warm sterile water to produce a fibrinogen concentration of 200 mg/mL, and the thrombin was resuspended with 40-mmol/L calcium chloride. When the carved CCG was ready to be placed in the laryngotracheal incision, 0.5 mL of each was mixed and immediately applied to the CCG to be placed in the trachea. A firm coagulum formed within 30 seconds, securing the CCG in place while forming an airtight seal.

Postoperatively, all animals were given a liquid diet of balanced salt solution and were advanced to solid foods as tolerated. The animals in each group were closely monitored for signs of aspiration, respiratory difficulty, dehydration, or weight loss. Analgesics were given as needed.

On the appropriate postoperative day (day 7 or 30), the designated animal was taken back to the operating suite and direct laryngoscopy and bronchoscopy with video documentation were performed after appropriate anesthesia. The animal was then humanely killed. The laryngotracheal complex was harvested and underwent histopathologic evaluation. The harvested tracheal specimen was fixed in 10% neutral buffered formalin. The grafted cricoid segment was removed, processed, and paraffin embedded in the standard fashion. Histologic sections from the midpoint of the graft perpendicular to the tracheal lumen were prepared. Hematoxylin-eosin– and Masson trichrome–stained slides were reviewed for each specimen by a single pathologist (R.F.), who was unaware of the method of graft stabilization. The sections were evaluated for the following criteria: (1) expansion of the cricoid cartilage at the widest part; (2) graft displacement (external or internal relative to the tracheal circumference); (3) complete or incomplete intraluminal mucosal coverage overlying the graft at the widest part of the graft; (4) the inflammatory response adjacent to the graft, including the presence or absence of a foreign body giant cell reaction (inflammatory reactions were graded from 1 to 4, with 1 representing minimal chronic inflammation and 4 indicating a florid reaction with abscess formation); (5) the histological pattern of graft integration (fibrous vs fibrocartilaginous); and (6) the viability of the graft expressed in terms of percentage of the cartilage graft judged viable by the pathologist.

Method of statistical analysis was as follows: For the continuous variables—ie, graft expansion, inflammatory response, and graft viability—the 40 subjects were analyzed by a 2×2 factorial analysis of variance with interaction, contrasting the 2 treatment groups, the 2 time groups, and the interaction of treatment by time. For the nominal data—ie, graft displacement, mucosal growth, and pattern of graft integration—the treatment groups and the time groups were contrasted using the Fisher exact test. Statistical significance was taken as P<.05. Required sample size cannot be estimated in advance, as there is no comparable study in the literature from which we can derive preliminary SDs. The sample size was as large as possible within the constraints of experimental practicality.

All animals survived until killed. The only postoperative complication was formation of a small seroma on 3 of the first 5 animals. No treatment was necessary, as the seroma resolved spontaneously after 1 week. The results separated into each subgroup are expressed in Table 1 and Table 2.

Fibrin sealant stabilized the CCG within 30 seconds of mixing the fibrinogen and thrombin components. These animals all achieved a complete seal at the surgical site with no air leak at 30 cm H₂O. Suturing of the CCG took an average of 10 minutes, with a range of 5 to 30 minutes. None of the suture specimens achieved a complete seal at 30 cm H₂O.

Endoscopic examination of the subglottis and trachea using a 30° endoscope immediately after the animals were killed revealed that all the subglottic regions appeared expanded and the luminal mucosa to be intact. There was no gross evidence of graft displacement at harvesting of the laryngotracheal complex (Figure 1 and Figure 2).

Subglottic expansion, measured in millimeters, was taken at the midpoint of the graft (Figure 3). The range varied from 0.5 mm to 3.4 mm for the fibrin sealant specimens, with a mean of 1.7 mm. The range varied from 0.4 mm to 3.4 mm, with a mean of 2.2 mm for the suture specimens. Analysis of variance was not significant for these 2 groups (P=.07). When comparing the day 7 postoperative groups (fibrin sealant and suture) with the day 30 postoperative groups (fibrin sealant and suture), the mean values were 2.2 mm and 1.8 mm, respectively. These differences were not significant (P=.10).

Graft displacement was evaluated as either internally or externally displaced as measured in relationship of the graft to the tracheal lumen (Figure 4). There were 7 internal and 5 external displacements in the fibrin group, and 12 internal and 3 external displacements in the suture group. Using the Fisher exact test, these values were not statistically significant (P=.25).

Histological evaluation of complete or incomplete mucosal growth was measured at the midpoint of the graft (Figure 5). The assumption was made that the midpoint would be the last portion to achieve mucosal coverage following grafting of the costal cartilage. Thirteen (65%) of the 20 fibrin sealant specimens demonstrated evidence of complete mucosal growth vs 3 (15%) of the 20 suture specimens (P=.003, Fisher exact test). There was no statistical difference when the day 7 postoperative group (fibrin sealant and suture) specimens were compared with the day 30 postoperative group (fibrin sealant and suture) specimens (P≈1.00).

Inflammation was judged histologically at the midportion of the graft and was graded as follows: 1 indicates minimal; 2, mild; 3, extensive with or without focal abscess formation; and 4, severe with abscesses. The mean value for the fibrin sealant group was 2.5 vs 2.3 for the suture group, which was not significant by analysis of variance (P=.41). The mean values for the day 7 postoperative groups (fibrin sealant and suture) and the day 30 postoperative groups (fibrin sealant and suture) were 2.8 and 2.0, respectively, a significant difference (P=.002). Thirteen (65%) of the 20 fibrin sealant group specimens vs 19 (95%) of the 20 suture group specimens had foreign body reactions, which was not significant (P=.18, Fisher exact test).

Graft integration was judged as being fibrous or fibrocartilaginous integration (Figure 6). There was no significant difference when comparing fibrocartilaginous integration of the fibrin sealant group specimens (2 of 20) vs the suture group specimens (6 of 20) (P=.24). There was a significant difference when comparing the day 7 postoperative groups and the day 30 postoperative groups (0/20 vs 8/20; P=.003).

Graft viability was measured as the percentage of the CCG that was viable. The range varied from 70% to 100% for the fibrin sealant group with a mean of 95%. For the suture group, the range varied from 20% to 100% with a mean of 86%. These differences were not significantly different by analysis of variance (P=.06). When comparing the day 7 postoperative groups with the day 30 postoperative groups, a significant difference was found (P=.004).

One of the well-established methods of surgical management of moderate to severe acquired subglottic stenosis in infants and children consists of laryngotracheal reconstruction using cartilage grafting. Since the introduction of the technique of costal cartilage grafting via an anterior laryngotracheal incision by Fearon and Cotton in 1974,¹⁰ various modifications have improved the results of the procedure. The original procedure resulted in the patient being left with an indwelling stent in the airway. These patients would have to breath through their tracheotomy tubes with the attendant risks. After waiting a few months to allow proper healing, a second procedure is required to remove the stent and decannulate the patient's tracheotomy. With the recent onset of "single-stage laryngotracheal reconstruction,"^{1 - 2} these problems have been obviated in that (1) a secondary procedure is not required, (2) the tracheotomized patient is decannulated at the time of surgery, and (3) long-term stenting with its attendant problems is not necessary. The reported success rate of this procedure is over 90%.^{1 - 3} The failure of laryngotracheal reconstruction is frequently from the cartilage graft not healing. During the performance of laryngotracheal reconstruction, surgical trauma results from the tracheal incision and placement of the sutures for the cartilage graft. Because the laryngotracheal incision is generally a single sharp cut, the major surgical trauma is from manipulation of the cartilage graft and the trachea during the placement of the multiple sutures. The size of the infant as well as the tiny cartilage graft, which has a thickness of only a few millimeters, makes the placement of sutures technically very difficult. Even the most experienced surgeons have seen sutures being pulled through the thin, soft cartilage graft. This results in a loss of cartilage integrity and subsequently poorer wound healing. Animal studies have demonstrated that resorption of implanted cartilage occurs particularly where surgical trauma or sutures allowed granulation tissue to penetrate into the cartilage matrix.⁴ Sometimes, there is a leak around the CCG at the end of suture stabilization of the graft. In these cases, the wound must be closed loosely. This may result in some difficulty with postoperative ventilation and delay in wound healing. By using fibrin sealant to stabilize the graft in lieu of sutures, there will be less surgical trauma, and there should be no air leak around the graft.

Fibrin sealant is a biological adhesive that imitates the effects of the final stages of coagulation. The American Red Cross J. Holland Laboratory fibrin sealant consists of concentrated fibrinogen and factor XIII as a single lyophilized preparation that is activated by the addition of thrombin and calcium chloride. The fibrin network formed has strong bonding power. This network provides a matrix for immediate tissue regrowth leading to natural tissue union.^{5 - 6} Wound healing is actually enhanced by the immediate stimulation of fibroblasts. It has been shown to achieve a good seal and adhesion in tissue grafting.⁶ Kram et al¹¹ concluded that in tracheal reconstruction experiments in dogs, the use of fibrin sealant resulted in a stable "leakless trachea" and significantly reduced the number of sutures needed. Successful use of fibrin tissue glue to seal pleural and bronchial air leaks has been reported.¹² Some otolaryngologists in Canada and Europe have successfully used commercially available Tisseel (human fibrinogen and bovine thrombin) in lieu of sutures (Robert Ballough, MD, oral communication, September 1993). Fibrin sealant has been shown to improve wound healing. Use of fibrin sealant can produce less postoperative ecchymosis and edema,¹³ and provide improvement in neovascularization during wound healing,⁶ with less flap necrosis and more collagenization.¹⁴ Ideally, if fibrin sealant could provide the mechanical stability equal to that of traditional suturing, it would simplify the surgery as well as eliminate the risks of graft resorption from suture trauma and infection along the suture tracts.

Unfortunately, Tisseel (Immunol AG, Vienna, Austria) is not approved by the Food and Drug Administration for use in the United States. The autologous fibrin glues available in the United States currently do not have the necessary characteristics for use in place of sutures. The American Red Cross J. Holland Laboratory uses a solvent-detergent process that deactivates encapsulated viruses and eliminates the potential risk of viral transmission. By increasing the fibrinogen concentration, there is a corresponding increase in the adhesive strength.⁶

Our study offers fibrin sealant as a promising alternative to sutures for securing a free cartilage graft in the trachea. Complete intraluminal mucosal growth occurred significantly more frequently in the fibrin sealant group (65%) than in the suture group (15%). This may be explained by the decreased amount of time necessary to secure the graft and the decreased amount of surgical manipulation of the graft using the fibrin sealant in comparison to sutures. Although not statistically significant, the greater graft viability for the fibrin sealant group may be explained by the decreased surgical trauma associated with use of the sealant, which eliminates the risk of granulation tissue penetration into the cartilage matrix along suture tracts, which has been shown to increase cartilage resorption.

Our data suggest that sutures, the standard with which we are comparing the fibrin sealant, provide a rigid fixation over a longer period of time than the fibrin sealant. Although none of these findings was statistically significant, in the suture group the average amount of subglottic expansion was greater, the number of graft extrusions was lower, and graft integration revealed a greater rate of fibrocartilaginous union. These observations suggest that sutures provide a longer fixation period, which is not surprising since the fibrin sealant dissipates after 2 to 3 weeks.

Our study was based on the use of a live animal model to assess wound healing after surgical intervention. When comparing the day 7 postoperative groups with the day 30 postoperative groups, the data were consistent, with more advanced healing at 30 days. The CCGs were more viable, with better fibrocartilaginous integration and less of an inflammatory response noted at 30 days. Interestingly, there was no difference in mucosal growth between the day 7 and day 30 groups. This may mean that in this experimental model, mucosal healing progressed slowly after the initial first week. Further long-term studies are planned to evaluate this.

These results support the feasibility of sutureless laryngotracheal reconstruction. Given data suggesting that the adhesive property of the fibrin sealant does not last beyond 2 to 3 weeks, further studies must be undertaken to evaluate whether the 2 to 3 weeks' duration of support is adequate for normal wound healing to fully stabilize the cartilage graft. Meanwhile, fibrin sealant can be used as an adjunct to decrease the number of sutures and thus surgical trauma during costal cartilage laryngotracheal reconstruction while allowing for an "leakless trachea." Future applications may include using the fibrin sealant as a sustained-release vehicle for tissue growth factors and antibiotics that could further enhance the success of the procedure. Pending current approval by the Food and Drug Administration, the American Red Cross J. Holland Laboratory fibrin sealant shows promise as an alternative to suture as a method for tissue graft fixation.

Accepted for publication March 13, 1998.

The Chief, Bureau of Medicine and Surgery, Navy Department, Washington, DC, sponsored this report 5-95-095, as required by HSETCINST 6000.41A.

The views expressed in this article are those of the authors and do not reflect the official policy or position of the Department of the Navy, Department of Defense, or the US government.

Presented at the American Society of Pediatric Otolaryngology Meeting, Scottsdale, Ariz, May 15, 1997.

Reprints: CAPT D. R. Kang, MC, USN, c/o Clinical Investigation Department, Naval Medical Center, 34800 Bob Wilson Dr, San Diego, CA 92134-5000.