Acute perforations in the tympanic membranes (TM) of animals are not g
ood models for assessment of materials used as graft membranes or of s
ubstances with the potential to activate tympanic membrane healing mec
hanisms for closing perforations. Most acute TM perforations heal spon
taneously, in both animals and humans. Acute TM perforation models are
not analogous to the pathologic human problem of long-standing TM per
foration. Bilateral subtotal symmetric perforations, with each animal
serving as its own control provide a suitable model. Fourteen dogs wer
e operated. Subtotal perforation of TM in right ears were soaked with
2% glutaraldehyde to develop an animal model for persistent tympanic m
embrane perforation. Subtotal myringectomy was performed on left ears,
without chemical treatment of the rim. Perforations were standardized
in size. The healing pattern was evaluated weekly for a 15-week perio
d. Six of the right ears and 14 of the left ears completely healed wit
hin 15 weeks. Statistically, the mean delay of closure was higher in t
he treated group than in the non-treated group, and the two groups wer
e different according the closure rate criterion. Histologic analyses
of persistent TM perforations and of healed cases were performed.