Repair of chronic tympanic membrane perforations with fibroblast growth factor.

A number of angiogenic growth factors have been shown to accelerate wound healing. Previous work has demonstrated that topical application of epidermal growth factor is effective in healing chronic tympanic membrane perforations in an animal model. Theoretically, fibroblast growth factor may result in a superior healed membrane through preferential stimulation of the fibroblasts within the middle layer of the tympanic membrane. To test this hypothesis, the effects of exogenously applied fibroblast growth factor on the chronically perforated tympanic membrane were evaluated. A buffered solution of fibroblast growth factor (25 microliters of fibroblast growth factor, 0.2 mg/ml) was administered to a Gelfoam pledget placed over chronic tympanic membrane perforations in chinchillas. Control ears were treated with Gelfoam and the buffer solution only. Complete closure of the tympanic membrane perforation was observed in 81% (13 of 16) of the fibroblast growth factor-treated ears, but in only 41% (7 of 17) of the controls (p = 0.05). Heading took place gradually, requiring an average of 4 weeks for the fibroblast growth factor-treated and 6.5 weeks for the control ears that healed. The relatively high healing rate for the control group does not imply that the pretreatment perforations were not chronic, rather there appears to be some efficacy to the control protocol of repeated applications of Gelfoam and buffer. A histologic analysis of the fibroblast growth factor-healed eardrums immediately after closure demonstrated hypertrophy of the squamous and fibrous layers of the tympanic membrane. Over time, the eardrum thinned to reach proportions similar to those of the normal tympanic membrane, including the presence of a substantial middle fibrous layer. A screening ototoxicity study revealed no structural damage to the organ of Corti after growth factor treatment. To assess the potential for systemic toxicity, blood and peripheral tissues were analyzed for radioactivity at time points during a 48-hour period after application of 25 microliters of 125I-fibroblast growth factor to the perforated tympanic membrane. More than 78% of the radioactivity remained at the application site. Given the tiny original dosage, the small fraction absorbed systemically is minuscule and highly unlikely to induce adverse effects in light of published toxicity data. On the basis of these promising safety and efficacy data in the chinchilla model, clinical trials of fibroblast growth factor in repair of chronic tympanic membrane perforations in human beings are being initiated.