- Email: [email protected]
Effect of vascular endothelial growth factor on laryngeal wound healing in rabbits
Otolaryngology–Head and Neck Surgery (2007) 137, 465-470
ORIGINAL RESEARCH
Effect of vascular endothelial growth factor on laryngeal wound healing in rabbits James W. Schroeder, Jr, MD, Jeffrey C. Rastatter, MD, and David L. Walner, MD, Chicago, IL OBJECTIVE: Study the effects of vascular endothelial growth factor (VEGF) on laryngeal wound healing in a rabbit model. STUDY DESIGN: Prospective, randomized, blinded. METHODS: The anterior cricoid cartilage of 10 rabbits was split and a VEGF-soaked collagen sponge was sewn between the cut edges. In 10 control animals, the collagen sponge was soaked with phosphate-buffered saline solution. The larynx was harvested on day 10. The degree of epithelial closure, the degree of soft tissue closure, and the presence of inflammatory cells was graded. RESULTS: There was complete epithelial closure in the control group. There was a slightly higher, but not statistically significant, grade of soft tissue closure in the experimental group. The experimental group had a lower but not statistically significant acute inflammatory response score. CONCLUSIONS: The topical application of VEGF through an implanted collagen sponge to an anterior, subglottic incision in a rabbit has no significant effect on tracheal luminal epithelial closure, acute inflammatory response, or soft tissue repair at postsurgical day 10. © 2007 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
P
roper wound healing is important to the success of laryngotracheal surgery. Failure of such surgery is often the result of an abundant inflammatory response that can result in granulation tissue, collagen deposition, and scar tissue formation that leads to restenosis. The direct application of growth factors to the reconstructed larynx to help regulate the inflammatory response has significant potential to effect wound healing within the airway. A surgical technique that could utilize a growth factor’s ability to enhance wound healing would have a clinical application for children with subglottic stenosis. Wound healing involves a series of biological events that involve inflammation, connective tissue deposition, epithelial covering, and wound remodeling. Each phase is interrelated, and the entire complex process is regulated at least in part by growth factors, biologically active substances that interact with specific cell surface receptors. One such factor is vascular endothelial growth factor (VEGF).
VEGF is produced or released locally in wounds during the time when new vessel growth is initiated.1 In healing wounds, newly formed vessels are first evident two to three days after injury with maximal evidence at one to two weeks.2 VEGF levels increase steadily through the first week after injury and when this process is blocked, new vessel formation is hampered.1 It has been demonstrated that topical treatment of free tracheal autografts with VEGF (5 g/mL), before their implantation, enhanced wound healing in a rabbit tracheal reconstruction model.3 VEGF accelerated autograft revascularization, reduced submucosal fibrosis and inflammation, and preserved the normal tracheal architecture by postoperative day 14 in 16 rabbits.3 This study was designed to determine if the application of VEGF to the tracheal wound itself through an implanted delivery system will enhance wound healing. VEGF was applied in a high concentration (10 g/mL) directly to the tracheal incision (not through a cartilage autograft as previously described) at the time of injury and held at that site by a collagen sponge. The incision site was examined at postinjury day 10, the expected time of maximal new vessel formation.
MATERIALS AND METHODS Twenty male New Zealand White Rabbits (weight range, 6 lbs 1 oz to 7 lbs 10 oz; mean weight, 6 lbs 8 oz) underwent an anterior cricoid split surgical procedure according to the model previously described.4 All animals received humane care in compliance with the “Guide for the Care and Use of Laboratory Animals” prepared by the Institute of Laboratory Animal Resources, National Research Council, and published by the National Academy Press, revised 1996. Our institution’s animal review committee approved the protocol. Surgical procedures were performed in a sterile operating suite, in a sterile manner. Induction anesthesia was
Received February 4, 2007; revised April 3, 2007; accepted April 26, 2007.
0194-5998/$32.00 © 2007 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved. doi:10.1016/j.otohns.2007.04.027
466
Otolaryngology–Head and Neck Surgery, Vol 137, No 3, September 2007
produced by intramuscular injection of 40 mg/kg ketamine HCL (total dose not to exceed 120 mg or 1.2 mL of 100 mg/mL solution), 1 mg/kg acepromazine and 5 mg/kg rompun (total not to exceed 15 mg or 0.75 mL of 20 mg/mL solution). Animals were placed in the supine position, shaved, and then sterilely prepped and draped. A 2 cm vertical, midline, skin incision was made over the larynx to expose the tracheal and cricoid cartilages (Fig 1A). The anterior cricoid cartilage and the first tracheal ring were split in the midline using a number 11 blade and a 7 ⫻ 2 mm collagen sponge was placed in the wound (Fig 1B). The sponge was soaked with 20 L of a 10 g/mL solution of VEGF (n ⫽ 10) in the experimental group and 20 L of sterile PBS solution (n ⫽ 10) in the control group. The supratheraputic dose of VEGF was chosen because the implanted collagen sponge will bathe the wound over a long
period of time. The animals were randomized to each group based on their ID tag number. Even numbers were placed in the experimental group. Sterile PBS solution (pH 7.2) was used to reconstitute the lyophilized human recombinant VEGF (Oncogene Research Products). The sponges were then sewn into place between the cut edges of the cricoid cartilage with a 4-0 prolene stitch (Fig 1C). The wound was then closed in two layers with buried absorbable suture. All animals were monitored for pain and distress postoperatively. Pain medication was administered as needed in the form of buprenorphine (0.05 mg/kg). The staff at the Rush Comparative Research Center provided veterinary care. The Comparative Research Center (CRC) was involved in the administering of anesthesia as well as in the monitoring of the animals intraoperatively and postoperatively.
Figure 1 (A) Midline cervical incision exposes the cricoid and laryngeal cartilages. (B) Transcricoid incision. (C) Collagen sponge sewn into place.
Schroeder et al
Effect of vascular endothelial growth factor . . .
Table 1 Standardized scoring system for hematoxylin and eosin stained slides of tracheal specimens Score Inflammation Few inflammatory cells within 1 mm of anterior midline incision Moderate number of inflammatory cells within 1 mm of anterior midline incision Abundant inflammatory cells within 1 mm of anterior midline incision Epithelial Closure Grossly incomplete closure on airway luminal surface Near complete closure on airway luminal surface Complete closure on airway luminal surface Soft Tissue Closure Absence of collagen deposition just deep to epithelium Moderate collagen deposition with presence of granulation tissue just deep to epithelium Near complete or complete closure
0 1 2
0
467
was summarized both as categorical with the use of cross tabs, and continuous with the use of means, standard deviations, medians, and ranges. Based on this scoring system, 10 animals per group for a total of 20 animals provides 95% power to detect true difference using a 2-sided Mann-Whitney (Wilcoxon ranksum) test assuming a true elevation in the mean of 0.80 and a standard deviation of 0.46. Level of significance (alpha) is 0.05. Statistical significance of the difference between control and experimental groups on the ratings of the three measures was determined by Wilcoxon rank-sum 2-sample test (Mann-Whitney U test). All tests performed were 2-sided. Level of significance used was alpha ⫽ 0.05. All analyses were conducted in SAS 9.1 (SAS Institute, Cary, NC).
1 2
0
1 2
Ten days after the procedure, the animals in each group were sacrificed. Euthanasia was achieved by intramuscular injection of 40 mg/kg of ketamine HCL and % mg/kg rompun. When adequately sedated, an intravascular dose of pentobarbital sodium 100 mg/kg was administered to effect. The animal’s heartbeat was assessed for death. This method was developed based on the recommendations of the Panel of Euthanasia of the American Veterinary Medical Association. After euthanasia, all larynges were harvested, fixed in formaldehyde and embedded in paraffin. Sections were taken from the level of the cricoid split and sponge. They were cut to a thickness of 4 m, placed on microscope slides, and then stained with hematoxylin and eosin. Analysis consisted of evaluation of the degree of inflammation present, the degree of epithelial closure within the airway lumen, and the degree of connective tissue closure deep to the airway epithelium. The scoring system used to evaluate the tissue is outlined in Table 1. The score was determined by a pathologist who was blinded as to whether the specimen was harvested from the control group or the experimental group. Two pathologists independently scored the specimens. If their score differed, it was averaged. Weighted Kappa statistic was computed to determine agreement between the two raters on each of the three measures (epi closure, soft tissue, inflammation). The two raters’ scores were averaged to determine a score for analysis. The score for analysis is then a 5-point ordinal scale (0, 0.5, 1, 1.5, 2) in order to accommodate an average score between the two ordinal scores. Descriptive statistics were calculated as follows: the 5-point ordinal rating scale
RESULTS There were 10 animals in the control group. The average weight of these 10 animals before surgery was 6 lbs 9 oz. The average weight of these animals on the day of harvest (10 days after surgery) was 6 lbs 11 oz. None of the animals in this group lost weight and one animal had no change in weight. All of the animals survived until harvest. There were 10 animals in the experimental group. The average weight of these animals before surgery was 6 lbs 7 oz. The average weight of this group on the day of harvest was 6 lbs 9 oz. These animals gained, on average, 2 oz during the 10 days before harvest. This is the same average gain seen in the control group. However, in contrast to the control group, three rabbits in the experimental group lost weight. In addition, one rabbit in the experimental group became sick and experienced a significant amount of vomiting and diarrhea. All the animals in this group survived until harvest. The histologic findings for both control and experimental groups are summarized in Table 2. In the control group, all 10 animals received a score of 2 with respect to epithelial closure. The luminal epithelium had completely closed in all animals. Nine of the 10 animals in the experimental group had complete closure; one showed a completely disrupted epithelium. The average score was 1.8. The animal with a completely disrupted epithelium was the only animal that lost weight. This was the animal with emesis and
Table 2 Average scores of experimental and control tissues based on the standardized scoring system Results (mean [SD])
Controls
Experimental
P value
Epithelial closure Inflammation Soft tissue closure
2.0 (0.00) 1.3 (0.49) 1.1 (0.46)
1.8 (0.63) 0.9 (0.66) 1.4 (0.24)
1.0 0.15 0.14
468
Otolaryngology–Head and Neck Surgery, Vol 137, No 3, September 2007
persistent diarrhea. The Kappa statistic for agreement between pathologist raters is 1.00, 95% CI (1.00 to 1.00). There is no significant difference in epithelial closure between the experimental and control group (P ⫽ 0.38). The average score for inflammation in the control group was 1.3 and in the experimental group the inflammation score was only 0.9. The inflammatory reaction seen in both groups was similar with respect to the type of inflammatory cells present. There were signs of acute and chronic inflammation as evidenced by lymphocytes and monocytes. There was no foreign body reaction noted in either group. The Kappa statistic for rater agreement is 0.50, 95% CI (0.23 to 0.77). There is no significant difference between the experimental and control groups (P ⫽ 0.15). The average score for the control group with respect to soft tissue closure was 1.1, in contrast to a slightly higher score of 1.4 in the experimental group. Four of the cases in the experimental group were scored as a 2. This indicated almost complete wound repair. No cases in the control group received a grade of 2 with respect to soft tissue closure. The Kappa statistic for agreement between raters is 0.15, 95% CI (0 to 0.39). There is no significant difference in soft tissue between the experimental and control groups (P ⫽ 0.18). Figure 2 demonstrates the histological findings with the use of hematoxylin and eosin stain. The control group (A
and B) demonstrate abundant inflammatory cells in the surgical site at low (⫻4) and high (⫻20) power magnification, respectively. In the experimental group (C and D) there is less inflammation. Both examples show complete closure of the epithelium on the airway luminal surface. There is no foreign body reaction to the collagen sponge. The collagen has been absorbed.
DISCUSSION Subglottic stenosis is an important cause of life-threatening respiratory distress in children. Severe stenosis can be improved by performing one of several procedures that can limit the need for a long-term tracheostomy tube. Anterior and/or posterior cricoid split with luminal augmentation is a common surgical option for severe subglottic stenosis. The success of such a procedure depends greatly on the local inflammatory response at the surgical incision. Excessive tissue response to injury could lead to undesirable results such as restenosis. A technique that would lead to enhanced wound healing could lead to more desirable surgical results and, therefore, have a significant clinical application for children with subglottic stenosis. Wound healing involves a series of biological events that begins with hemostasis. This is followed by the inflamma-
Figure 2 Control group with hematoxylin and eosin stain: (A) ⫻4; (B) ⫻20. Experimental group with hematoxylin and eosin stain: (C) ⫻4; (D) ⫻20.
Schroeder et al
Effect of vascular endothelial growth factor . . .
tory response, the formation of connective tissue, the covering of the wound with epithelium and, finally, the remodeling of the wound.5 Each of these phases is controlled and regulated by biologically active substances called growth factors. These growth factors are hormone-like molecules that interact with specific cell surface receptors to control the process of tissue repair. Within the wound space, vasoconstriction and vasodilation both occur. Vasoconstriction occurs to augment the hemostatic response and vasodilation occurs to allow other healing factors to be brought into the wound. There is also a local increase in vascular permeability that allows the healing factors to enter the wound from the blood.6 An angiogenic response is also characteristic of the early phases of wound repair and provides the vasculature that will supply the newly formed granulation tissue. The discovery of increased VEGF protein production7 and message expression8 in skin wounds suggested that VEGF might regulate these key wound-healing events. VEGF was originally discovered as a tumor-secreted protein that rendered venules and small veins hyperpermeable to macromolecules.9 Subsequently, it was discovered that VEGF is also a potent angiogenic factor in vivo.10 VEGF has also been implicated in enhanced wound healing, increased vascular permeability, angiogenesis, and the stimulation of nitric oxide release from vascular endothelial cells.1,11 It has also recently been shown that the topical treatment of free tracheal autografts with VEGF in a rabbit tracheal reconstruction model enhanced healing.3 There was evidence that the VEGF accelerated autograft revascularization, reduced submucosal fibrosis and inflammation and preservation of the normal tracheal architecture.3 In this study, VEGF was applied directly to the tracheal wound at the time of injury through an implanted collagen sponge. The application of VEGF appeared to have no influence with respect to epithelial closure of the tracheal lumen at the surgical site. There was complete closure of the epithelial lining in all animals except in the one animal in the experimental group that developed a systemic illness and likely had poor nutrition as a result. It is unclear if the VEGF played a role in this illness. It was the opinion of the veterinarian caring for the animals that the sick animal had contracted a systemic viral illness. The presence of the collagen sponge in the wound did not appear to hinder wound closure. However, another control group not treated with an implanted sponge would help answer this question. There was a less robust acute inflammatory response and a more advanced stage of tissue closure in several animals in the experimental group. This is demonstrated in Figure 2. These differences were not statistically significant. It is unclear if the direct application of VEGF to the tracheal wound through a collegen sponge affected the healing process. We hypothesized that by altering the intensity of the inflammatory response or by altering its time course by enhancing angiogenesis through the manipulation of local VEGF concentrations, surgical wound healing would
469
be improved. We did not demonstrate this. Perhaps the power of the next study could be improved by increasing the number of animals. Also, the interobserver availability could be decreased with a change in the grading system and an increase in the number of pathologists used to score each variable. The clinical benefits of topical application of VEGF in tracheal wound healing are still unclear. Both positive and negative outcomes have been reported. In an immunohistochemical analysis of cartilage harvested from children with poor healing capability after laryngotracheal reconstruction, there was a positive correlation with elevated levels of stromal VEGF.12 This indicates that the inflammatory response caused by excess VEGF at the injury site may have a negative long-term effect on local wound healing. We report neither a positive nor negative effect in this study.
CONCLUSION The topical application of VEGF (10 g/mL) through an implanted collagen sponge to a complete, anterior, subglottic incision in a rabbit has no significant effect on luminal epithelial closure, the inflammatory response, or soft tissue repair at postsurgical day 10. The angiogenic properties of VEGF and how they regulate inflammation and soft tissue repair in the larynx are important when studying wound repair. Both positive and negative clinical effects have been reported. The role of topical VEGF in laryngeal surgery requires further study.
ACKNOWLEDGMENT The authors thank Juan-Miguel Mosquera, MD, of the Department of Pathology, Rush University Medical Center, Chicago, IL.
AUTHOR INFORMATION From the Children’s Memorial Hospital (Dr Schroeder), Chicago IL; Department of Otolaryngology (Dr Schroeder), Northwestern University, Chicago, IL; and Department of Otolaryngology (Drs Rastatter and Walner), Rush University Medical Center, Chicago, IL. Corresponding author: James W. Schroeder, Jr, MD, Division of Pediatric Otolaryngology, Children’s Memorial Hospital, 2300 Children’s Plaza, Box 25, Chicago, IL 60614. E-mail address: [email protected].
FINANCIAL DISCLOSURE None.
470
Otolaryngology–Head and Neck Surgery, Vol 137, No 3, September 2007
REFERENCES 1. Nissen NH, Polverini PJ, Koch AE, et al. Vascular endothelial growth factor mediates angiogenic activity during the proliferative phase of wound healing. Am J Pathol 1998;152:1445–52. 2. Ross R, Benditt E. Wound healing and collagen production: fine structure in experimental scurvy. J Cell Biol 1962;12:533–54. 3. Dodge-Khatami A, Backer C, Holinger L, et al. Healing of a free tracheal autograft is enhanced by topical vascular endothelial growth factor in an experimental rabbit model. Gen Thor Sur 2001;122:554 – 61. 4. Walner D, Cotton R, Willging J, et al. Model for evaluating the effect of growth factors on the larynx. Otolaryngol Head Neck Surg 1999; 120:78 – 83. 5. Falanga V, Zitelli JA, Eaglstein WH. Wound healing. J Am Acad Dermatol 1988;19:559 – 63. 6. Steed D. The role of growth factors in wound healing. Surg Clin North Am 1997;77:575– 86.
7. Howdieshell TR, Riegner C, Gupta V, et al. Normoxic wound fluid contains high levels of vascular endothelial growth factor. Ann Surg 1998;228:707. 8. Brown LF, Yeo KT, Berse B, et al. Expression of vascular permeability factor (vascular endothelial growth factor) by epidermal keratinocytes during wound healing. J Exp Med 1992;176:1375. 9. Dvorak HF, Dvorak AM, Manseau EJ, et al. Fibrin gel investment associated with line 1 and line 10 solid tumor growth, angiogenesis and fibroplasias in guinea pigs. J Natl Cancer Inst 1979;62:1459. 10. Senger DR, Peruzzi CA, Feder J, et al. A highly conserved vascular permeability factor secreted by a variety of human and rodent tumor cell lines. Cancer Res 1986;46:5629. 11. Ferrara N, Alitalo K. Clinical applications of angiogenic growth factors and their inhibitors. Nat Med 1999;5:1359 – 64. 12. Walner D, Heffelfinger S, Stern Y, et al. Potential role of growth factors and extracellular matrix in wound healing after laryngotracheal reconstruction. Otolaryngol Head Neck Surg 2000;122;363– 6.
ORIGINAL RESEARCH
Effect of vascular endothelial growth factor on laryngeal wound healing in rabbits James W. Schroeder, Jr, MD, Jeffrey C. Rastatter, MD, and David L. Walner, MD, Chicago, IL OBJECTIVE: Study the effects of vascular endothelial growth factor (VEGF) on laryngeal wound healing in a rabbit model. STUDY DESIGN: Prospective, randomized, blinded. METHODS: The anterior cricoid cartilage of 10 rabbits was split and a VEGF-soaked collagen sponge was sewn between the cut edges. In 10 control animals, the collagen sponge was soaked with phosphate-buffered saline solution. The larynx was harvested on day 10. The degree of epithelial closure, the degree of soft tissue closure, and the presence of inflammatory cells was graded. RESULTS: There was complete epithelial closure in the control group. There was a slightly higher, but not statistically significant, grade of soft tissue closure in the experimental group. The experimental group had a lower but not statistically significant acute inflammatory response score. CONCLUSIONS: The topical application of VEGF through an implanted collagen sponge to an anterior, subglottic incision in a rabbit has no significant effect on tracheal luminal epithelial closure, acute inflammatory response, or soft tissue repair at postsurgical day 10. © 2007 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
P
roper wound healing is important to the success of laryngotracheal surgery. Failure of such surgery is often the result of an abundant inflammatory response that can result in granulation tissue, collagen deposition, and scar tissue formation that leads to restenosis. The direct application of growth factors to the reconstructed larynx to help regulate the inflammatory response has significant potential to effect wound healing within the airway. A surgical technique that could utilize a growth factor’s ability to enhance wound healing would have a clinical application for children with subglottic stenosis. Wound healing involves a series of biological events that involve inflammation, connective tissue deposition, epithelial covering, and wound remodeling. Each phase is interrelated, and the entire complex process is regulated at least in part by growth factors, biologically active substances that interact with specific cell surface receptors. One such factor is vascular endothelial growth factor (VEGF).
VEGF is produced or released locally in wounds during the time when new vessel growth is initiated.1 In healing wounds, newly formed vessels are first evident two to three days after injury with maximal evidence at one to two weeks.2 VEGF levels increase steadily through the first week after injury and when this process is blocked, new vessel formation is hampered.1 It has been demonstrated that topical treatment of free tracheal autografts with VEGF (5 g/mL), before their implantation, enhanced wound healing in a rabbit tracheal reconstruction model.3 VEGF accelerated autograft revascularization, reduced submucosal fibrosis and inflammation, and preserved the normal tracheal architecture by postoperative day 14 in 16 rabbits.3 This study was designed to determine if the application of VEGF to the tracheal wound itself through an implanted delivery system will enhance wound healing. VEGF was applied in a high concentration (10 g/mL) directly to the tracheal incision (not through a cartilage autograft as previously described) at the time of injury and held at that site by a collagen sponge. The incision site was examined at postinjury day 10, the expected time of maximal new vessel formation.
MATERIALS AND METHODS Twenty male New Zealand White Rabbits (weight range, 6 lbs 1 oz to 7 lbs 10 oz; mean weight, 6 lbs 8 oz) underwent an anterior cricoid split surgical procedure according to the model previously described.4 All animals received humane care in compliance with the “Guide for the Care and Use of Laboratory Animals” prepared by the Institute of Laboratory Animal Resources, National Research Council, and published by the National Academy Press, revised 1996. Our institution’s animal review committee approved the protocol. Surgical procedures were performed in a sterile operating suite, in a sterile manner. Induction anesthesia was
Received February 4, 2007; revised April 3, 2007; accepted April 26, 2007.
0194-5998/$32.00 © 2007 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved. doi:10.1016/j.otohns.2007.04.027
466
Otolaryngology–Head and Neck Surgery, Vol 137, No 3, September 2007
produced by intramuscular injection of 40 mg/kg ketamine HCL (total dose not to exceed 120 mg or 1.2 mL of 100 mg/mL solution), 1 mg/kg acepromazine and 5 mg/kg rompun (total not to exceed 15 mg or 0.75 mL of 20 mg/mL solution). Animals were placed in the supine position, shaved, and then sterilely prepped and draped. A 2 cm vertical, midline, skin incision was made over the larynx to expose the tracheal and cricoid cartilages (Fig 1A). The anterior cricoid cartilage and the first tracheal ring were split in the midline using a number 11 blade and a 7 ⫻ 2 mm collagen sponge was placed in the wound (Fig 1B). The sponge was soaked with 20 L of a 10 g/mL solution of VEGF (n ⫽ 10) in the experimental group and 20 L of sterile PBS solution (n ⫽ 10) in the control group. The supratheraputic dose of VEGF was chosen because the implanted collagen sponge will bathe the wound over a long
period of time. The animals were randomized to each group based on their ID tag number. Even numbers were placed in the experimental group. Sterile PBS solution (pH 7.2) was used to reconstitute the lyophilized human recombinant VEGF (Oncogene Research Products). The sponges were then sewn into place between the cut edges of the cricoid cartilage with a 4-0 prolene stitch (Fig 1C). The wound was then closed in two layers with buried absorbable suture. All animals were monitored for pain and distress postoperatively. Pain medication was administered as needed in the form of buprenorphine (0.05 mg/kg). The staff at the Rush Comparative Research Center provided veterinary care. The Comparative Research Center (CRC) was involved in the administering of anesthesia as well as in the monitoring of the animals intraoperatively and postoperatively.
Figure 1 (A) Midline cervical incision exposes the cricoid and laryngeal cartilages. (B) Transcricoid incision. (C) Collagen sponge sewn into place.
Schroeder et al
Effect of vascular endothelial growth factor . . .
Table 1 Standardized scoring system for hematoxylin and eosin stained slides of tracheal specimens Score Inflammation Few inflammatory cells within 1 mm of anterior midline incision Moderate number of inflammatory cells within 1 mm of anterior midline incision Abundant inflammatory cells within 1 mm of anterior midline incision Epithelial Closure Grossly incomplete closure on airway luminal surface Near complete closure on airway luminal surface Complete closure on airway luminal surface Soft Tissue Closure Absence of collagen deposition just deep to epithelium Moderate collagen deposition with presence of granulation tissue just deep to epithelium Near complete or complete closure
0 1 2
0
467
was summarized both as categorical with the use of cross tabs, and continuous with the use of means, standard deviations, medians, and ranges. Based on this scoring system, 10 animals per group for a total of 20 animals provides 95% power to detect true difference using a 2-sided Mann-Whitney (Wilcoxon ranksum) test assuming a true elevation in the mean of 0.80 and a standard deviation of 0.46. Level of significance (alpha) is 0.05. Statistical significance of the difference between control and experimental groups on the ratings of the three measures was determined by Wilcoxon rank-sum 2-sample test (Mann-Whitney U test). All tests performed were 2-sided. Level of significance used was alpha ⫽ 0.05. All analyses were conducted in SAS 9.1 (SAS Institute, Cary, NC).
1 2
0
1 2
Ten days after the procedure, the animals in each group were sacrificed. Euthanasia was achieved by intramuscular injection of 40 mg/kg of ketamine HCL and % mg/kg rompun. When adequately sedated, an intravascular dose of pentobarbital sodium 100 mg/kg was administered to effect. The animal’s heartbeat was assessed for death. This method was developed based on the recommendations of the Panel of Euthanasia of the American Veterinary Medical Association. After euthanasia, all larynges were harvested, fixed in formaldehyde and embedded in paraffin. Sections were taken from the level of the cricoid split and sponge. They were cut to a thickness of 4 m, placed on microscope slides, and then stained with hematoxylin and eosin. Analysis consisted of evaluation of the degree of inflammation present, the degree of epithelial closure within the airway lumen, and the degree of connective tissue closure deep to the airway epithelium. The scoring system used to evaluate the tissue is outlined in Table 1. The score was determined by a pathologist who was blinded as to whether the specimen was harvested from the control group or the experimental group. Two pathologists independently scored the specimens. If their score differed, it was averaged. Weighted Kappa statistic was computed to determine agreement between the two raters on each of the three measures (epi closure, soft tissue, inflammation). The two raters’ scores were averaged to determine a score for analysis. The score for analysis is then a 5-point ordinal scale (0, 0.5, 1, 1.5, 2) in order to accommodate an average score between the two ordinal scores. Descriptive statistics were calculated as follows: the 5-point ordinal rating scale
RESULTS There were 10 animals in the control group. The average weight of these 10 animals before surgery was 6 lbs 9 oz. The average weight of these animals on the day of harvest (10 days after surgery) was 6 lbs 11 oz. None of the animals in this group lost weight and one animal had no change in weight. All of the animals survived until harvest. There were 10 animals in the experimental group. The average weight of these animals before surgery was 6 lbs 7 oz. The average weight of this group on the day of harvest was 6 lbs 9 oz. These animals gained, on average, 2 oz during the 10 days before harvest. This is the same average gain seen in the control group. However, in contrast to the control group, three rabbits in the experimental group lost weight. In addition, one rabbit in the experimental group became sick and experienced a significant amount of vomiting and diarrhea. All the animals in this group survived until harvest. The histologic findings for both control and experimental groups are summarized in Table 2. In the control group, all 10 animals received a score of 2 with respect to epithelial closure. The luminal epithelium had completely closed in all animals. Nine of the 10 animals in the experimental group had complete closure; one showed a completely disrupted epithelium. The average score was 1.8. The animal with a completely disrupted epithelium was the only animal that lost weight. This was the animal with emesis and
Table 2 Average scores of experimental and control tissues based on the standardized scoring system Results (mean [SD])
Controls
Experimental
P value
Epithelial closure Inflammation Soft tissue closure
2.0 (0.00) 1.3 (0.49) 1.1 (0.46)
1.8 (0.63) 0.9 (0.66) 1.4 (0.24)
1.0 0.15 0.14
468
Otolaryngology–Head and Neck Surgery, Vol 137, No 3, September 2007
persistent diarrhea. The Kappa statistic for agreement between pathologist raters is 1.00, 95% CI (1.00 to 1.00). There is no significant difference in epithelial closure between the experimental and control group (P ⫽ 0.38). The average score for inflammation in the control group was 1.3 and in the experimental group the inflammation score was only 0.9. The inflammatory reaction seen in both groups was similar with respect to the type of inflammatory cells present. There were signs of acute and chronic inflammation as evidenced by lymphocytes and monocytes. There was no foreign body reaction noted in either group. The Kappa statistic for rater agreement is 0.50, 95% CI (0.23 to 0.77). There is no significant difference between the experimental and control groups (P ⫽ 0.15). The average score for the control group with respect to soft tissue closure was 1.1, in contrast to a slightly higher score of 1.4 in the experimental group. Four of the cases in the experimental group were scored as a 2. This indicated almost complete wound repair. No cases in the control group received a grade of 2 with respect to soft tissue closure. The Kappa statistic for agreement between raters is 0.15, 95% CI (0 to 0.39). There is no significant difference in soft tissue between the experimental and control groups (P ⫽ 0.18). Figure 2 demonstrates the histological findings with the use of hematoxylin and eosin stain. The control group (A
and B) demonstrate abundant inflammatory cells in the surgical site at low (⫻4) and high (⫻20) power magnification, respectively. In the experimental group (C and D) there is less inflammation. Both examples show complete closure of the epithelium on the airway luminal surface. There is no foreign body reaction to the collagen sponge. The collagen has been absorbed.
DISCUSSION Subglottic stenosis is an important cause of life-threatening respiratory distress in children. Severe stenosis can be improved by performing one of several procedures that can limit the need for a long-term tracheostomy tube. Anterior and/or posterior cricoid split with luminal augmentation is a common surgical option for severe subglottic stenosis. The success of such a procedure depends greatly on the local inflammatory response at the surgical incision. Excessive tissue response to injury could lead to undesirable results such as restenosis. A technique that would lead to enhanced wound healing could lead to more desirable surgical results and, therefore, have a significant clinical application for children with subglottic stenosis. Wound healing involves a series of biological events that begins with hemostasis. This is followed by the inflamma-
Figure 2 Control group with hematoxylin and eosin stain: (A) ⫻4; (B) ⫻20. Experimental group with hematoxylin and eosin stain: (C) ⫻4; (D) ⫻20.
Schroeder et al
Effect of vascular endothelial growth factor . . .
tory response, the formation of connective tissue, the covering of the wound with epithelium and, finally, the remodeling of the wound.5 Each of these phases is controlled and regulated by biologically active substances called growth factors. These growth factors are hormone-like molecules that interact with specific cell surface receptors to control the process of tissue repair. Within the wound space, vasoconstriction and vasodilation both occur. Vasoconstriction occurs to augment the hemostatic response and vasodilation occurs to allow other healing factors to be brought into the wound. There is also a local increase in vascular permeability that allows the healing factors to enter the wound from the blood.6 An angiogenic response is also characteristic of the early phases of wound repair and provides the vasculature that will supply the newly formed granulation tissue. The discovery of increased VEGF protein production7 and message expression8 in skin wounds suggested that VEGF might regulate these key wound-healing events. VEGF was originally discovered as a tumor-secreted protein that rendered venules and small veins hyperpermeable to macromolecules.9 Subsequently, it was discovered that VEGF is also a potent angiogenic factor in vivo.10 VEGF has also been implicated in enhanced wound healing, increased vascular permeability, angiogenesis, and the stimulation of nitric oxide release from vascular endothelial cells.1,11 It has also recently been shown that the topical treatment of free tracheal autografts with VEGF in a rabbit tracheal reconstruction model enhanced healing.3 There was evidence that the VEGF accelerated autograft revascularization, reduced submucosal fibrosis and inflammation and preservation of the normal tracheal architecture.3 In this study, VEGF was applied directly to the tracheal wound at the time of injury through an implanted collagen sponge. The application of VEGF appeared to have no influence with respect to epithelial closure of the tracheal lumen at the surgical site. There was complete closure of the epithelial lining in all animals except in the one animal in the experimental group that developed a systemic illness and likely had poor nutrition as a result. It is unclear if the VEGF played a role in this illness. It was the opinion of the veterinarian caring for the animals that the sick animal had contracted a systemic viral illness. The presence of the collagen sponge in the wound did not appear to hinder wound closure. However, another control group not treated with an implanted sponge would help answer this question. There was a less robust acute inflammatory response and a more advanced stage of tissue closure in several animals in the experimental group. This is demonstrated in Figure 2. These differences were not statistically significant. It is unclear if the direct application of VEGF to the tracheal wound through a collegen sponge affected the healing process. We hypothesized that by altering the intensity of the inflammatory response or by altering its time course by enhancing angiogenesis through the manipulation of local VEGF concentrations, surgical wound healing would
469
be improved. We did not demonstrate this. Perhaps the power of the next study could be improved by increasing the number of animals. Also, the interobserver availability could be decreased with a change in the grading system and an increase in the number of pathologists used to score each variable. The clinical benefits of topical application of VEGF in tracheal wound healing are still unclear. Both positive and negative outcomes have been reported. In an immunohistochemical analysis of cartilage harvested from children with poor healing capability after laryngotracheal reconstruction, there was a positive correlation with elevated levels of stromal VEGF.12 This indicates that the inflammatory response caused by excess VEGF at the injury site may have a negative long-term effect on local wound healing. We report neither a positive nor negative effect in this study.
CONCLUSION The topical application of VEGF (10 g/mL) through an implanted collagen sponge to a complete, anterior, subglottic incision in a rabbit has no significant effect on luminal epithelial closure, the inflammatory response, or soft tissue repair at postsurgical day 10. The angiogenic properties of VEGF and how they regulate inflammation and soft tissue repair in the larynx are important when studying wound repair. Both positive and negative clinical effects have been reported. The role of topical VEGF in laryngeal surgery requires further study.
ACKNOWLEDGMENT The authors thank Juan-Miguel Mosquera, MD, of the Department of Pathology, Rush University Medical Center, Chicago, IL.
AUTHOR INFORMATION From the Children’s Memorial Hospital (Dr Schroeder), Chicago IL; Department of Otolaryngology (Dr Schroeder), Northwestern University, Chicago, IL; and Department of Otolaryngology (Drs Rastatter and Walner), Rush University Medical Center, Chicago, IL. Corresponding author: James W. Schroeder, Jr, MD, Division of Pediatric Otolaryngology, Children’s Memorial Hospital, 2300 Children’s Plaza, Box 25, Chicago, IL 60614. E-mail address: [email protected].
FINANCIAL DISCLOSURE None.
470
Otolaryngology–Head and Neck Surgery, Vol 137, No 3, September 2007
REFERENCES 1. Nissen NH, Polverini PJ, Koch AE, et al. Vascular endothelial growth factor mediates angiogenic activity during the proliferative phase of wound healing. Am J Pathol 1998;152:1445–52. 2. Ross R, Benditt E. Wound healing and collagen production: fine structure in experimental scurvy. J Cell Biol 1962;12:533–54. 3. Dodge-Khatami A, Backer C, Holinger L, et al. Healing of a free tracheal autograft is enhanced by topical vascular endothelial growth factor in an experimental rabbit model. Gen Thor Sur 2001;122:554 – 61. 4. Walner D, Cotton R, Willging J, et al. Model for evaluating the effect of growth factors on the larynx. Otolaryngol Head Neck Surg 1999; 120:78 – 83. 5. Falanga V, Zitelli JA, Eaglstein WH. Wound healing. J Am Acad Dermatol 1988;19:559 – 63. 6. Steed D. The role of growth factors in wound healing. Surg Clin North Am 1997;77:575– 86.
7. Howdieshell TR, Riegner C, Gupta V, et al. Normoxic wound fluid contains high levels of vascular endothelial growth factor. Ann Surg 1998;228:707. 8. Brown LF, Yeo KT, Berse B, et al. Expression of vascular permeability factor (vascular endothelial growth factor) by epidermal keratinocytes during wound healing. J Exp Med 1992;176:1375. 9. Dvorak HF, Dvorak AM, Manseau EJ, et al. Fibrin gel investment associated with line 1 and line 10 solid tumor growth, angiogenesis and fibroplasias in guinea pigs. J Natl Cancer Inst 1979;62:1459. 10. Senger DR, Peruzzi CA, Feder J, et al. A highly conserved vascular permeability factor secreted by a variety of human and rodent tumor cell lines. Cancer Res 1986;46:5629. 11. Ferrara N, Alitalo K. Clinical applications of angiogenic growth factors and their inhibitors. Nat Med 1999;5:1359 – 64. 12. Walner D, Heffelfinger S, Stern Y, et al. Potential role of growth factors and extracellular matrix in wound healing after laryngotracheal reconstruction. Otolaryngol Head Neck Surg 2000;122;363– 6.