- Email: [email protected]
Tympanic membrane perforations in the diabetic rat: A model of impaired wound healing
Tympanic membrane perforations in the diabetic rat: A model of impaired wound healing JEFFREY T. VRABEC, MD, Galveston, Texas
Animal models of type I and type II diabetes mellitus have been studied intensively in an effort to define the pathophysiology of the diabetic condition. An often-observed clinical manifestation of diabetes is poor wound repair. Thus diabetic animals have emerged as useful models for the study of impaired wound healing. The healing of acute tympanic membrane (TM) perforations in diabetic animals has not been reported. This investigation compares time to closure of a standardized TM perforation in rats with streptozotocininduced diabetes, Zucker diabetic fatty rats, and normal control rats. The Zucker diabetic fatty rats demonstrate a significantly prolonged time to closure compared with the other two groups. This animal model may be useful for future study of TM wound repair. (Otolaryngol Head Neck Surg 1998;118:304-8.)
The tympanic membrane (TM) displays a remarkable capacity for spontaneous repair under a wide variety of clinical and experimental conditions. This tendency makes it challenging to ascertain factors reliably that prevent or delay repair. Methods for determining the causes of impaired healing differ in the clinical and experimental settings. In the former, retrospective study may reveal the common characteristics of patients with poor outcomes. In the latter, a hypothetical adverse factor is tested against a control condition. A combination of these approaches is necessary to define clearly the reasons for impaired healing of TM defects. The process of spontaneous repair of TM defects in animals has been described.1 Although the sequence of wound repair is known, the rate-limiting step is not clearly defined. Clinical problems such as chronic perforation, atrophy, and delayed perforation after tympanoplasty are likely a result of arrest of the repair process at different stages. Most experimental work addressing abnormal healing of the TM has focused on the production of a chronic perforation. In attempting to reproduce this clinical problem, factors that significantly impair healing can be identified. From the Department of Otolaryngology, University of Texas Medical Branch. Supported in part by a grant from the University of Texas Medical Branch Small Grant Program (No. 93-10-066). Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, New Orleans, La., Sept. 17-20, 1995. Reprint requests: Jeffrey T. Vrabec, MD, Department of Otolaryngology, University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555-0521. Copyright © 1998 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/98/$5.00 + 0 23/1/79302 304
Rogers and Snow2 treated the edges of acute TM perforations with electrocautery, phenol, thioglycolic acid, and nitrogen mustard. Only the latter two agents produced a significant delay in spontaneous healing. No chronic perforations were created. Another effort of note was that of Goycoolea et al.3 in their study of the pathogenesis of chronic otitis media. In this experiment the eustachian tube was obliterated and a perforation was created. Each ear developed an effusion; however, the TM healed in all animals observed for greater than 1 month. More recent, successful efforts have employed chronic steroid treatment, application of glutaraldehyde, and electrocautery combined with infolding of the edges of the TM.4-6 Direct clinical correlation of the experimental information is more difficult and at times contradictory. For example, electrocautery-induced perforations demonstrate delayed healing in clinical subjects, but creation of chronic perforations is less successful.7,8 The use of steroid-containing drops after otologic procedures receives little attention. Two reports of steroid use in traumatic perforations are noted, one anecdotally condemning their use, and the other showing no adverse effect.9,10 One patient population that is well known to manifest impaired wound healing are individuals with diabetes. Research into the pathogenesis of diabetes has fostered the development of numerous animal models of both type I and type II diabetes.11,12 The plethora of animal models emphasizes further that multiple disorders can produce the hyperglycemic state. The alterations in wound repair in diabetes have been investigated in several animal models. The mechanisms for this response are multifactorial and include impaired platelet and neutrophil function, reduced col-
Otolaryngology– Head and Neck Surgery Volume 118 Number 3 Part 1
lagen synthesis by fibroblasts, increased thickening of capillary basement membranes leading to increased permeability, altered microcirculation as a result of increased blood viscosity, and possible distortion of vasomotor responses.13,14 Not one of these mechanisms is predominant in its ability to alter wound repair, although improved control of glucose metabolism is clearly beneficial. The purpose of this study was to compare healing of TM perforations in diabetic animals with that of nondiabetic control animals. Two animal models were chosen for study: streptozotocin-induced diabetic rats (STZ) and the congenitally diabetic Zucker diabetic fatty rat (ZDF). Streptozotocin is highly toxic to pancreatic beta cells, leading to cell death and insulinopenia within 48 hours, thus mimicking type I diabetes. The ZDF is a model of type II diabetes and demonstrates both hyperglycemia and insulin resistance.11 Objectives included defining the rate of wound repair and attempting to identify a quantitative or qualitative difference in wound architecture between groups. METHODS
Three different groups of rats were evaluated. Each group contained eight animals. All 24 animals were male and were 12 to 13 weeks old when the perforations were created. Both ears are included in the data analysis, yielding 16 ears in each group. All animals received a standard diet and were housed under identical conditions. The experimental protocol was approved by the University of Texas Medical Branch Animal Care and Use Committee. The first group consisted of Sprague-Dawley rats and served as the control group. These animals were obtained from the vendor at 9 weeks of age. These animals received an intraperitoneal injection of saline solution after arrival to control for the effects of the injection given to group 2. Mean weight and mean serum glucose levels were recorded on the day the perforations were created for all groups. Serum glucose levels were determined according to the Accu-chek (Boehringer Mannheim Corp., Indianapolis, Ind.) system for all animals, after obtaining a drop of blood from the tail vein. The second group (STZ) of Sprague-Dawley rats was also obtained from the vendor at 9 weeks of age. These rats were given an intraperitoneal injection of streptozotocin (45 mg/kg) on arrival. Serum glucose levels were obtained before injection and then monitored every 1 to 2 weeks to ensure maintenance of the hyperglycemic state. The third group were the ZDFs. These animals were obtained at approximately 12 weeks of age. Measure-
VRABEC 305
ments of serum glucose and weight were obtained before creation of the perforations as in the other groups. Perforations were created in an identical manner in all animals. Each animal was anesthetized with an intraperitoneal injection of ketamine (30 mg/kg) and xylazine (3 mg/kg). The posterior one half of the pars tensa of each TM was excised with standard otologic instruments and an operating microscope. The operative defect was observed daily after day 6. No perforations healed in less than 7 days. Manipulation of the TM was minimal, consisting solely of gentle removal of exudate so that the defect size could be recorded accurately. Ears that developed otorrhea were suctioned to remove purulent drainage. The time to complete closure was recorded for each TM. The animals were killed at varied intervals after closure of the TM perforations. The maximal period of observation was 2 months. After an overdose of pentobarbital, the temporal bones were removed and placed in formalin for 48 hours. Next, the bones were decalcified in buffered ethylenediamine tetraacetic acid. The TM was then removed and prepared for histologic study. The TM was sectioned perpendicularly to the long process of the malleus in 20 µm thick slices. Hematoxylin and eosin stain was used. Five specimens from each of the three groups were evaluated quantitatively. Digitized images were captured and analyzed with the software package Optimas 5.1 (Bioscan, Inc., Washington, D.C.). The mean thickness of the TM was obtained for each specimen by measuring the total cross-sectional area of the TM from malleus to anulus on both the operated and nonoperated (control) sides of the TM. This measurement was then divided by the measured length of the section. All measurements were taken at 40× magnification. RESULTS
The control and STZ animals were the same weight at 9 weeks of age. After 3 weeks of observation, the control animals had gained an average of 75 gm. The control group had a mean weight of 345 gm and a mean glucose level of 121 gm/dl. By 12 weeks of age, the STZ animals had a mean weight of 246 gm and a mean glucose level of 402 gm/dl. Throughout the remainder of the study, STZ animals continued to gain weight slowly, although never approaching that of the control group. This observation is in accordance with other reports of weight gain in streptozocin-treated animals.15 At 12 weeks of age the ZDF animals had a mean weight of 473 gm and a mean glucose level of 387 gm/dl. All animals demonstrated spontaneous closure of the surgical defect. The mean time required for complete closure of the perforation was 8.5 ± 1.2 days in the con-
Otolaryngology– Head and Neck Surgery March 1998
306 VRABEC
Table 1. Mean thickness of the TM in the representative specimens Mean thickness of TM (µm)
Control STZ ZDF
OP
NON
39.83 ± 7.58 52.83 ± 9.36 29.93 ± 7.45
18.90 ± 4.52 20.33 ± 4.32 15.96 ± 1.76
OP, Operated half of TM; NON, control half of TM.
Fig. 1. Operated side of TM, control group. (Original magnification 3400.)
Fig. 2. Operated side of TM, STZ group. (Original magnification 3400.)
trol rats, 8.9 ± 1.3 days in the STZ rats, and 11.4 ± 4.1 days in the ZDF rats. The time to closure was compared by a t test revealing a significant delay in closure by the ZDF animals compared with the control (p = 0.002) and STZ (p = 0.005) groups. Spontaneous otorrhea developed in zero of 16 control ears, seven of 16 STZ ears and one of 16 ZDF ears. The increased incidence of infection in the STZ group compared with control rats was statistically significant according to the signed-rank test (p = 0.01). The otorrhea was treated by suctioning of the discharge on each of the daily examinations. No antibiotics were used for prophylaxis or treatment of ongoing drainage. A single culture was obtained from one of the discharging ears, revealing Staphylococcus aureus. Repeated suctioning was sufficient to eliminate the otorrhea in most of the animals. Two STZ animals and one ZDF animal had middle ear effusion and mucosal inflammation at the time of death. The mean thickness of the TM in the representative specimens is presented in Table 1. The thickness of the nonoperated side is in close agreement. On the operated side, the STZ group had significantly thicker TMS than the other groups. However, four of the five specimens were obtained less than 1 month after wounding, whereas specimens in all other animals were obtained 1 to 2 months after wounding. Representative sections of the control, STZ, and ZDF groups are presented in Figs. 1, 2, and 3, respectively. All photos are at equivalent magnification. The structure of the TM appears similar in each of the groups. The thickness of the healed portion of the TM was not uniform in any of the specimens. Organization within the middle fibrous layer is distorted compared with the nonoperated side of the TM in all groups. Hypertrophy of the squamous epithelium or inner mucosal layer was not evident in any of the specimens. DISCUSSION
Fig. 3. Operated side of TM, ZDF group. (Original magnification 3400.)
The differences in TM thickness observed reveal some interesting trends. On the nonoperated half of the TM, the thickness of the control and STZ groups are in
Otolaryngology– Head and Neck Surgery Volume 118 Number 3 Part 1
close agreement, as expected. The TMs of the ZDF group are somewhat thinner, but this finding did not achieve statistical significance in the limited sample size. On the operated side of the TM, the STZ group has much thicker TMs than the other two groups; however, this finding may be explained by a shorter time to death in the STZ group. The TM would be expected to be thinner in the more mature state as noted in earlier work.16 Again, a trend toward a thinner TM in the ZDF group compared with control rats is noted. The diabetic patient has not often been investigated within the context of otologic wound healing. General statements concerning the increased risk of tympanoplasty in diabetic patients are often voiced, but there are few, if any, studies that directly implicate diabetes as a cause of surgical failure.17 The primary reason to explain the lack of data is that diabetes has not been implicated as a cause of otitis media or cholesteatoma. If diabetes and chronic ear disease are independent variables, the incidence of diabetes in patients undergoing otologic surgery would be equal to the prevalence of diabetes in the general population, which is approximately 1%. Clinical diabetes mellitus is a heterogeneous disease. A study of all patients with diabetes undergoing chronic ear operations would, by definition, group individuals whose common variable, the diabetic state, was produced by numerous underlying causes. In addition, the severity of illness and response of the hyperglycemia to therapy would also differ. Thus it would be difficult to assume that the overall effect on wound healing is the same in all of the patients. The same reasoning applies to the STZ and ZDF animal models. Even though the serum glucose levels are equivalent, the effects of the hyperglycemia on wound repair would not necessarily be expected to be the same. This difference in animal models is illustrated in a study of incisional wounds by Greenwald et al.18 Genetically diabetic animals had greater impairment of wound repair than had animals with streptozotocininduced diabetes. Results in these animal models may not be extrapolated directly to all patients with diabetes. The possibility that diabetes increases complications in otologic surgery must be confirmed by clinical studies. In this study the incidence of infection was greatest in the STZ animals, yet the overall rate of repair was intermediate among the three groups. These results seem to contradict the often-expressed belief that infection is the primary cause of failure in tympanoplasty. Although this clinical dictum may be true, experimental evidence is lacking. Quantitative bacterial studies in surgical wounds have demonstrated that when the concentration of bacteria at the wound site exceeds a criti-
VRABEC 307
cal concentration, usually 105 organisms/gm tissue, healing is retarded. Efforts to close contaminated wounds with skin grafts or pedicled flaps are not likely to succeed.19 However, low levels of bacterial contamination do not impair healing and, given the increased inflammatory response, wound healing may even be accelerated.20 Quantitative study of bacterial counts were not performed in this study. Low levels of bacteria might be suspected given the improvement in otorrhea despite lack of any antibiotic treatment. Both of the animal models presented show promise for use in future studies of otologic wound healing. The STZ animals might be useful because of the increased susceptibility to infection. The increased susceptibility of the STZ animals to infection, especially staphylococcal, has been noted.14 It is theorized that this results from a direct effect of the drug on macrophage and neutrophil function. A natural application might be to attempt to define optimal antibiotic agents for postoperative prophylaxis. The ZDF animals show evidence of delayed healing. As a model of acute perforation, it may be valuable in identification of agents useful in promoting the healing process. Further attempts to produce chronic perforations in this model might also be considered. Another limitation of the model is the expense of obtaining this species. The ultimate definition of factors responsible for delay in TM healing results from a combination of clinical and experimental observations. Factors that impair wound repair may lead to increased complications as a result of the delay. In this study an animal model of type II diabetes demonstrated prolonged time to closure of an acute TM perforation. Although this animal model may prove to be valuable in experimental studies, extrapolation of this finding to diabetic patients undergoing otologic surgery is not possible without a carefully controlled analysis of results in the clinical setting. I thank James Hokanson, PhD, for assistance with statistical analysis. REFERENCES 1. Reijnen CJH, Kuijpers W. The healing pattern of the drum membrane. Acta Otolaryngol Suppl (Stockh) 1971;287:1-74. 2. Rogers KA Jr, Snow JB Jr. Closure of experimental tympanic membrane perforations. Ann Otol Rhinol Laryngol 1968;77: 66-71. 3. Goycoolea MV, Paparella MM, Juhn SK, Carpenter AM. Otitis media with perforation of the tympanic membrane: a longitudinal experimental study. Laryngoscope 1980;90:2037-45. 4. Spandow O, Hellstrom S. Animal model for persistent tympanic membrane perforations. Ann Otol Rhinol Laryngol 1993;102: 467-72. 5. Truy E, Disant F, Morgan A. Chronic tympanic membrane perforation: an animal model. Am J Otol 1995;16:222-5. 6. Amoils CP, Jackler RK, Milczuk H, Kelly KE, Cao K. An ani-
Otolaryngology– Head and Neck Surgery March 1998
308 VRABEC
7. 8. 9. 10. 11. 12. 13.
mal model of chronic tympanic membrane perforation. Otolaryngol Head Neck Surg 1992;106:47-55. Saito H, Miyamoto K, Kishimoto S, Higashitsuji H, Kitamura H. Burn perforation as a method of middle ear ventilation. Arch Otolaryngol 1978;104:79-81. Lau P, Shelton C, Goode RL. Heat myringotomy. Laryngoscope 1985;95:38-42. Armstrong BW. Traumatic perforations of the tympanic membrane: observe or repair. Laryngoscope 1972;82:1822-30. Griffin WL. A retrospective study of traumatic tympanic membrane perforations in a clinical practice. Laryngoscope 1979;89:261-82. Shafrir E. Animal models of non-insulin dependent diabetes. Diabetes Metab Rev 1992;8:179-208. Bell RH, Hye RJ. Animal models of diabetes mellitus: physiology and pathology. J Surg Res 1983;35:433-60. Goodson WH. Wound healing in diabetes. In: Bergman M, Sicard GA, editors. Surgical management of the diabetic patient. New York: Raven Press, 1991:39-50.
14. Morain WD, Colen LB. Wound healing in diabetes mellitus. Clin Plast Surg 1990;17:493-501. 15. Goodson WH III, Hunt TK. Studies of wound healing in experimental diabetes mellitus. J Surg Res 1977;22:221-7. 16. Vrabec JT, Schwaber MK, Davidson JM, Clymer MA. Evaluation of basic fibroblast growth factor in tympanic membrane repair. Laryngoscope 1994;104:1059-64. 17. Taylor GD. Contraindications to tympanoplasty. Laryngoscope 1975;85:64-83. 18. Greenwald DP, Shumway S, Zachary LS, et al. Endogenous versus toxin-induced diabetes in rats: a mechanical comparison of two skin wound-healing models. Plast Reconstr Surg 1993;91: 1087-92. 19. Robson MC, Stenberg BD, Heggers JP. Wound healing alterations caused by infection. Clin Plast Surg 1990;17:485-92. 20. Laato M, Niinikoski J. Lundberg C. Inflammatory reaction and blood flow in experimental wounds inoculated with Staphylococcus aureus. Eur Surg Res 1988;20:33-8.
European Consensus Development Conference on Neonatal Hearing Screening
This conference will be held May 15-16, 1998, in Milan, Italy. Large neonatal hearing screening programs have started in several European countries over the last 2 to 3 years, but many who work in clinics may not be fully aware of the results achievable in this rapidly growing area. The European Project AHEAD from the Biomedical and Health Programme of the European Commission (1996-1999) is showing that there is an urgent need to evaluate effectiveness and to establish standards of best practice. This conference is an ideal forum for discussion of this important issue. Specialists in related fields will review the most up-to-date evidence in their own field, thus providing the evidence to a panel of leading experts in an open forum. Full deliberations of the panel will be published at a later date. For further information, contact Dr. F. Grandori, Center of Biomedical Engineering, Polytechnic of Milan, Piazza Leonardo da Vinci 32, 20133 Milan, Italy; fax 39-2-23993360; e-mail, [email protected].
Animal models of type I and type II diabetes mellitus have been studied intensively in an effort to define the pathophysiology of the diabetic condition. An often-observed clinical manifestation of diabetes is poor wound repair. Thus diabetic animals have emerged as useful models for the study of impaired wound healing. The healing of acute tympanic membrane (TM) perforations in diabetic animals has not been reported. This investigation compares time to closure of a standardized TM perforation in rats with streptozotocininduced diabetes, Zucker diabetic fatty rats, and normal control rats. The Zucker diabetic fatty rats demonstrate a significantly prolonged time to closure compared with the other two groups. This animal model may be useful for future study of TM wound repair. (Otolaryngol Head Neck Surg 1998;118:304-8.)
The tympanic membrane (TM) displays a remarkable capacity for spontaneous repair under a wide variety of clinical and experimental conditions. This tendency makes it challenging to ascertain factors reliably that prevent or delay repair. Methods for determining the causes of impaired healing differ in the clinical and experimental settings. In the former, retrospective study may reveal the common characteristics of patients with poor outcomes. In the latter, a hypothetical adverse factor is tested against a control condition. A combination of these approaches is necessary to define clearly the reasons for impaired healing of TM defects. The process of spontaneous repair of TM defects in animals has been described.1 Although the sequence of wound repair is known, the rate-limiting step is not clearly defined. Clinical problems such as chronic perforation, atrophy, and delayed perforation after tympanoplasty are likely a result of arrest of the repair process at different stages. Most experimental work addressing abnormal healing of the TM has focused on the production of a chronic perforation. In attempting to reproduce this clinical problem, factors that significantly impair healing can be identified. From the Department of Otolaryngology, University of Texas Medical Branch. Supported in part by a grant from the University of Texas Medical Branch Small Grant Program (No. 93-10-066). Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, New Orleans, La., Sept. 17-20, 1995. Reprint requests: Jeffrey T. Vrabec, MD, Department of Otolaryngology, University of Texas Medical Branch, 301 University Blvd., Galveston, TX 77555-0521. Copyright © 1998 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/98/$5.00 + 0 23/1/79302 304
Rogers and Snow2 treated the edges of acute TM perforations with electrocautery, phenol, thioglycolic acid, and nitrogen mustard. Only the latter two agents produced a significant delay in spontaneous healing. No chronic perforations were created. Another effort of note was that of Goycoolea et al.3 in their study of the pathogenesis of chronic otitis media. In this experiment the eustachian tube was obliterated and a perforation was created. Each ear developed an effusion; however, the TM healed in all animals observed for greater than 1 month. More recent, successful efforts have employed chronic steroid treatment, application of glutaraldehyde, and electrocautery combined with infolding of the edges of the TM.4-6 Direct clinical correlation of the experimental information is more difficult and at times contradictory. For example, electrocautery-induced perforations demonstrate delayed healing in clinical subjects, but creation of chronic perforations is less successful.7,8 The use of steroid-containing drops after otologic procedures receives little attention. Two reports of steroid use in traumatic perforations are noted, one anecdotally condemning their use, and the other showing no adverse effect.9,10 One patient population that is well known to manifest impaired wound healing are individuals with diabetes. Research into the pathogenesis of diabetes has fostered the development of numerous animal models of both type I and type II diabetes.11,12 The plethora of animal models emphasizes further that multiple disorders can produce the hyperglycemic state. The alterations in wound repair in diabetes have been investigated in several animal models. The mechanisms for this response are multifactorial and include impaired platelet and neutrophil function, reduced col-
Otolaryngology– Head and Neck Surgery Volume 118 Number 3 Part 1
lagen synthesis by fibroblasts, increased thickening of capillary basement membranes leading to increased permeability, altered microcirculation as a result of increased blood viscosity, and possible distortion of vasomotor responses.13,14 Not one of these mechanisms is predominant in its ability to alter wound repair, although improved control of glucose metabolism is clearly beneficial. The purpose of this study was to compare healing of TM perforations in diabetic animals with that of nondiabetic control animals. Two animal models were chosen for study: streptozotocin-induced diabetic rats (STZ) and the congenitally diabetic Zucker diabetic fatty rat (ZDF). Streptozotocin is highly toxic to pancreatic beta cells, leading to cell death and insulinopenia within 48 hours, thus mimicking type I diabetes. The ZDF is a model of type II diabetes and demonstrates both hyperglycemia and insulin resistance.11 Objectives included defining the rate of wound repair and attempting to identify a quantitative or qualitative difference in wound architecture between groups. METHODS
Three different groups of rats were evaluated. Each group contained eight animals. All 24 animals were male and were 12 to 13 weeks old when the perforations were created. Both ears are included in the data analysis, yielding 16 ears in each group. All animals received a standard diet and were housed under identical conditions. The experimental protocol was approved by the University of Texas Medical Branch Animal Care and Use Committee. The first group consisted of Sprague-Dawley rats and served as the control group. These animals were obtained from the vendor at 9 weeks of age. These animals received an intraperitoneal injection of saline solution after arrival to control for the effects of the injection given to group 2. Mean weight and mean serum glucose levels were recorded on the day the perforations were created for all groups. Serum glucose levels were determined according to the Accu-chek (Boehringer Mannheim Corp., Indianapolis, Ind.) system for all animals, after obtaining a drop of blood from the tail vein. The second group (STZ) of Sprague-Dawley rats was also obtained from the vendor at 9 weeks of age. These rats were given an intraperitoneal injection of streptozotocin (45 mg/kg) on arrival. Serum glucose levels were obtained before injection and then monitored every 1 to 2 weeks to ensure maintenance of the hyperglycemic state. The third group were the ZDFs. These animals were obtained at approximately 12 weeks of age. Measure-
VRABEC 305
ments of serum glucose and weight were obtained before creation of the perforations as in the other groups. Perforations were created in an identical manner in all animals. Each animal was anesthetized with an intraperitoneal injection of ketamine (30 mg/kg) and xylazine (3 mg/kg). The posterior one half of the pars tensa of each TM was excised with standard otologic instruments and an operating microscope. The operative defect was observed daily after day 6. No perforations healed in less than 7 days. Manipulation of the TM was minimal, consisting solely of gentle removal of exudate so that the defect size could be recorded accurately. Ears that developed otorrhea were suctioned to remove purulent drainage. The time to complete closure was recorded for each TM. The animals were killed at varied intervals after closure of the TM perforations. The maximal period of observation was 2 months. After an overdose of pentobarbital, the temporal bones were removed and placed in formalin for 48 hours. Next, the bones were decalcified in buffered ethylenediamine tetraacetic acid. The TM was then removed and prepared for histologic study. The TM was sectioned perpendicularly to the long process of the malleus in 20 µm thick slices. Hematoxylin and eosin stain was used. Five specimens from each of the three groups were evaluated quantitatively. Digitized images were captured and analyzed with the software package Optimas 5.1 (Bioscan, Inc., Washington, D.C.). The mean thickness of the TM was obtained for each specimen by measuring the total cross-sectional area of the TM from malleus to anulus on both the operated and nonoperated (control) sides of the TM. This measurement was then divided by the measured length of the section. All measurements were taken at 40× magnification. RESULTS
The control and STZ animals were the same weight at 9 weeks of age. After 3 weeks of observation, the control animals had gained an average of 75 gm. The control group had a mean weight of 345 gm and a mean glucose level of 121 gm/dl. By 12 weeks of age, the STZ animals had a mean weight of 246 gm and a mean glucose level of 402 gm/dl. Throughout the remainder of the study, STZ animals continued to gain weight slowly, although never approaching that of the control group. This observation is in accordance with other reports of weight gain in streptozocin-treated animals.15 At 12 weeks of age the ZDF animals had a mean weight of 473 gm and a mean glucose level of 387 gm/dl. All animals demonstrated spontaneous closure of the surgical defect. The mean time required for complete closure of the perforation was 8.5 ± 1.2 days in the con-
Otolaryngology– Head and Neck Surgery March 1998
306 VRABEC
Table 1. Mean thickness of the TM in the representative specimens Mean thickness of TM (µm)
Control STZ ZDF
OP
NON
39.83 ± 7.58 52.83 ± 9.36 29.93 ± 7.45
18.90 ± 4.52 20.33 ± 4.32 15.96 ± 1.76
OP, Operated half of TM; NON, control half of TM.
Fig. 1. Operated side of TM, control group. (Original magnification 3400.)
Fig. 2. Operated side of TM, STZ group. (Original magnification 3400.)
trol rats, 8.9 ± 1.3 days in the STZ rats, and 11.4 ± 4.1 days in the ZDF rats. The time to closure was compared by a t test revealing a significant delay in closure by the ZDF animals compared with the control (p = 0.002) and STZ (p = 0.005) groups. Spontaneous otorrhea developed in zero of 16 control ears, seven of 16 STZ ears and one of 16 ZDF ears. The increased incidence of infection in the STZ group compared with control rats was statistically significant according to the signed-rank test (p = 0.01). The otorrhea was treated by suctioning of the discharge on each of the daily examinations. No antibiotics were used for prophylaxis or treatment of ongoing drainage. A single culture was obtained from one of the discharging ears, revealing Staphylococcus aureus. Repeated suctioning was sufficient to eliminate the otorrhea in most of the animals. Two STZ animals and one ZDF animal had middle ear effusion and mucosal inflammation at the time of death. The mean thickness of the TM in the representative specimens is presented in Table 1. The thickness of the nonoperated side is in close agreement. On the operated side, the STZ group had significantly thicker TMS than the other groups. However, four of the five specimens were obtained less than 1 month after wounding, whereas specimens in all other animals were obtained 1 to 2 months after wounding. Representative sections of the control, STZ, and ZDF groups are presented in Figs. 1, 2, and 3, respectively. All photos are at equivalent magnification. The structure of the TM appears similar in each of the groups. The thickness of the healed portion of the TM was not uniform in any of the specimens. Organization within the middle fibrous layer is distorted compared with the nonoperated side of the TM in all groups. Hypertrophy of the squamous epithelium or inner mucosal layer was not evident in any of the specimens. DISCUSSION
Fig. 3. Operated side of TM, ZDF group. (Original magnification 3400.)
The differences in TM thickness observed reveal some interesting trends. On the nonoperated half of the TM, the thickness of the control and STZ groups are in
Otolaryngology– Head and Neck Surgery Volume 118 Number 3 Part 1
close agreement, as expected. The TMs of the ZDF group are somewhat thinner, but this finding did not achieve statistical significance in the limited sample size. On the operated side of the TM, the STZ group has much thicker TMs than the other two groups; however, this finding may be explained by a shorter time to death in the STZ group. The TM would be expected to be thinner in the more mature state as noted in earlier work.16 Again, a trend toward a thinner TM in the ZDF group compared with control rats is noted. The diabetic patient has not often been investigated within the context of otologic wound healing. General statements concerning the increased risk of tympanoplasty in diabetic patients are often voiced, but there are few, if any, studies that directly implicate diabetes as a cause of surgical failure.17 The primary reason to explain the lack of data is that diabetes has not been implicated as a cause of otitis media or cholesteatoma. If diabetes and chronic ear disease are independent variables, the incidence of diabetes in patients undergoing otologic surgery would be equal to the prevalence of diabetes in the general population, which is approximately 1%. Clinical diabetes mellitus is a heterogeneous disease. A study of all patients with diabetes undergoing chronic ear operations would, by definition, group individuals whose common variable, the diabetic state, was produced by numerous underlying causes. In addition, the severity of illness and response of the hyperglycemia to therapy would also differ. Thus it would be difficult to assume that the overall effect on wound healing is the same in all of the patients. The same reasoning applies to the STZ and ZDF animal models. Even though the serum glucose levels are equivalent, the effects of the hyperglycemia on wound repair would not necessarily be expected to be the same. This difference in animal models is illustrated in a study of incisional wounds by Greenwald et al.18 Genetically diabetic animals had greater impairment of wound repair than had animals with streptozotocininduced diabetes. Results in these animal models may not be extrapolated directly to all patients with diabetes. The possibility that diabetes increases complications in otologic surgery must be confirmed by clinical studies. In this study the incidence of infection was greatest in the STZ animals, yet the overall rate of repair was intermediate among the three groups. These results seem to contradict the often-expressed belief that infection is the primary cause of failure in tympanoplasty. Although this clinical dictum may be true, experimental evidence is lacking. Quantitative bacterial studies in surgical wounds have demonstrated that when the concentration of bacteria at the wound site exceeds a criti-
VRABEC 307
cal concentration, usually 105 organisms/gm tissue, healing is retarded. Efforts to close contaminated wounds with skin grafts or pedicled flaps are not likely to succeed.19 However, low levels of bacterial contamination do not impair healing and, given the increased inflammatory response, wound healing may even be accelerated.20 Quantitative study of bacterial counts were not performed in this study. Low levels of bacteria might be suspected given the improvement in otorrhea despite lack of any antibiotic treatment. Both of the animal models presented show promise for use in future studies of otologic wound healing. The STZ animals might be useful because of the increased susceptibility to infection. The increased susceptibility of the STZ animals to infection, especially staphylococcal, has been noted.14 It is theorized that this results from a direct effect of the drug on macrophage and neutrophil function. A natural application might be to attempt to define optimal antibiotic agents for postoperative prophylaxis. The ZDF animals show evidence of delayed healing. As a model of acute perforation, it may be valuable in identification of agents useful in promoting the healing process. Further attempts to produce chronic perforations in this model might also be considered. Another limitation of the model is the expense of obtaining this species. The ultimate definition of factors responsible for delay in TM healing results from a combination of clinical and experimental observations. Factors that impair wound repair may lead to increased complications as a result of the delay. In this study an animal model of type II diabetes demonstrated prolonged time to closure of an acute TM perforation. Although this animal model may prove to be valuable in experimental studies, extrapolation of this finding to diabetic patients undergoing otologic surgery is not possible without a carefully controlled analysis of results in the clinical setting. I thank James Hokanson, PhD, for assistance with statistical analysis. REFERENCES 1. Reijnen CJH, Kuijpers W. The healing pattern of the drum membrane. Acta Otolaryngol Suppl (Stockh) 1971;287:1-74. 2. Rogers KA Jr, Snow JB Jr. Closure of experimental tympanic membrane perforations. Ann Otol Rhinol Laryngol 1968;77: 66-71. 3. Goycoolea MV, Paparella MM, Juhn SK, Carpenter AM. Otitis media with perforation of the tympanic membrane: a longitudinal experimental study. Laryngoscope 1980;90:2037-45. 4. Spandow O, Hellstrom S. Animal model for persistent tympanic membrane perforations. Ann Otol Rhinol Laryngol 1993;102: 467-72. 5. Truy E, Disant F, Morgan A. Chronic tympanic membrane perforation: an animal model. Am J Otol 1995;16:222-5. 6. Amoils CP, Jackler RK, Milczuk H, Kelly KE, Cao K. An ani-
Otolaryngology– Head and Neck Surgery March 1998
308 VRABEC
7. 8. 9. 10. 11. 12. 13.
mal model of chronic tympanic membrane perforation. Otolaryngol Head Neck Surg 1992;106:47-55. Saito H, Miyamoto K, Kishimoto S, Higashitsuji H, Kitamura H. Burn perforation as a method of middle ear ventilation. Arch Otolaryngol 1978;104:79-81. Lau P, Shelton C, Goode RL. Heat myringotomy. Laryngoscope 1985;95:38-42. Armstrong BW. Traumatic perforations of the tympanic membrane: observe or repair. Laryngoscope 1972;82:1822-30. Griffin WL. A retrospective study of traumatic tympanic membrane perforations in a clinical practice. Laryngoscope 1979;89:261-82. Shafrir E. Animal models of non-insulin dependent diabetes. Diabetes Metab Rev 1992;8:179-208. Bell RH, Hye RJ. Animal models of diabetes mellitus: physiology and pathology. J Surg Res 1983;35:433-60. Goodson WH. Wound healing in diabetes. In: Bergman M, Sicard GA, editors. Surgical management of the diabetic patient. New York: Raven Press, 1991:39-50.
14. Morain WD, Colen LB. Wound healing in diabetes mellitus. Clin Plast Surg 1990;17:493-501. 15. Goodson WH III, Hunt TK. Studies of wound healing in experimental diabetes mellitus. J Surg Res 1977;22:221-7. 16. Vrabec JT, Schwaber MK, Davidson JM, Clymer MA. Evaluation of basic fibroblast growth factor in tympanic membrane repair. Laryngoscope 1994;104:1059-64. 17. Taylor GD. Contraindications to tympanoplasty. Laryngoscope 1975;85:64-83. 18. Greenwald DP, Shumway S, Zachary LS, et al. Endogenous versus toxin-induced diabetes in rats: a mechanical comparison of two skin wound-healing models. Plast Reconstr Surg 1993;91: 1087-92. 19. Robson MC, Stenberg BD, Heggers JP. Wound healing alterations caused by infection. Clin Plast Surg 1990;17:485-92. 20. Laato M, Niinikoski J. Lundberg C. Inflammatory reaction and blood flow in experimental wounds inoculated with Staphylococcus aureus. Eur Surg Res 1988;20:33-8.
European Consensus Development Conference on Neonatal Hearing Screening
This conference will be held May 15-16, 1998, in Milan, Italy. Large neonatal hearing screening programs have started in several European countries over the last 2 to 3 years, but many who work in clinics may not be fully aware of the results achievable in this rapidly growing area. The European Project AHEAD from the Biomedical and Health Programme of the European Commission (1996-1999) is showing that there is an urgent need to evaluate effectiveness and to establish standards of best practice. This conference is an ideal forum for discussion of this important issue. Specialists in related fields will review the most up-to-date evidence in their own field, thus providing the evidence to a panel of leading experts in an open forum. Full deliberations of the panel will be published at a later date. For further information, contact Dr. F. Grandori, Center of Biomedical Engineering, Polytechnic of Milan, Piazza Leonardo da Vinci 32, 20133 Milan, Italy; fax 39-2-23993360; e-mail, [email protected].