In-office Repair of Chronic Tympanic Membrane Perforations

Background

The use of topically applied growth factors to promote TM closure has been suggested as an attractive alternative to conventional tympanoplasty. Previous clinical studies have reported a 92% and 98% complete closure rate of chronic tympanic membrane perforations with application of topical FGF-22 (Hakuba et al., Kanemaru et al). The investigators recently published a study a double-blinded, randomized placebo-controlled phase 2 clinical trial of 54 patients with chronic TM perforations also applying topical FGF-2 with saline as the control. Unlike prior studies the investigators found no statistically significant difference between FGF-2 (57.5%) and saline control TM (71.4%) closure rates (P value = 0.36).

The current study is proposed to examine healing rates with topical application of FGF-2 with surgical modifications to our prior study in the attempt to mirror the Japanese studies excellent results. The long-term goal is to repair TM perforations with nonsurgical topical application of the appropriate factors to avoid surgery, general anesthesia, and hospitalization. A secondary aim is to maintain hearing and to prevent complications of TMP.

Surgical modifications in this limited study of 20 patients randomized to active (FGF-2) or placebo (saline) treatment include 1) ensuring that no middle ear discharge is present at the time of the procedure, 2) placing a cotton pledget in the middle ear prior to rimming the perforation, 3) removing any tympanosclerotic tympanic membrane that abuts the perforation, 4) placing a layer of test article in the middle ear medial to the perforation and lateral to the perforation prior to adding a fibrin glue, 5) using up to 4 applications of the test article, and 6) the procedure will be performed by one surgeon only to ensure complete adherence to prescribed techniques and reduce surgeon variability.

Specific Aims and Objectives

The primary outcome is complete the closure of chronic tympanic membrane perforations evaluated at six months post treatment by photo-documentation and tympanometry. Secondary measures will include a) changes in pure tone and speech audiometry and b) wideband TM acoustic absorption.

General Description of Study Design

Randomized double-blinded placebo-controlled treatment of 20 patients with chronic (>6 month), dry tympanic membrane perforations will be repaired with a gelatin sponge soaked with FGF2 or saline in a 3:1 ratio, then covered with a fibrin glue. Up to four applications will be administered at 3 week intervals or until complete healing is observed.

Subject Enrollment Patients who present to participating otologist's MEE clinics with a chronic (6 months or longer), dry tympanic membrane perforation will be invited to participate by the investigators.

Study Procedure

Consenting subjects will be placed in the supine position in the minor operating suite of the Otology Clinic at MEE. The ear with the chronic perforation will be visualized with binocular microscopy and image will be taken for measurement of the TM perforation (time point 0). Any cerumen or debris occluding vision of the complete circumference of the central perforation will be removed. If the complete circumference of the central perforation cannot be visualized, an endoscope may be used. Complete view around the circumference of the perforation is required to proceed. If the perforation is a marginal perforation, the procedure will not proceed. Likewise, if there is discharge from the perforation or excessively thickened diseased mucosa, the procedure will be terminated.

Anesthesia of the clean, dry, chronic TM perforation will be accomplished by placing 4% lidocaine gel impregnated cotton against the TM with care to cover the edges of the perforation completely. The gel is left in place for a minimum of 15 minutes measured with a room timer. Following removal of the lidocaine gel cotton pledget, the anesthesia is tested with a sharp fine straight pick by creating postage stamp perforations around the circumference of the perforation. To ensure sufficient anesthesia, an injection of 1% lidocaine directly into the external auditory canal may be given if necessary to reduce pain. The rim of the perforation is then completely removed with small cup forceps. If an area of hyalinized TM or tympanosclerosis is abutting the perforation, the tympanosclerosis will be removed also.

After rimming the perforation is complete, saline or FGF2 soaked gelatin foam is placed through the perforation into the middle ear to fill the space medial to the perforation and surrounding it. A second layer of test article soaked gelatin is placed lateral to the perforation

Next a thin layer of fibrin glue is sprayed over the perforation and gelatin foam and allowed to congeal. A cotton ball is placed in the external meatus and the procedure terminated.

The patient returns for follow up in 3 weeks (+/- 1 week) for re-inspection of the perforation. The procedure may be repeated up to 3 more times if the perforation has not completely healed at follow up out to 9 weeks. If the perforation is completely healed and confirmed by pneumatic otoscopy, the patient will return 6 months from the procedure for follow up audiogram and tympanometry.

Any risks of the medical procedure to be performed will be explained thoroughly to patients.

Statistical Analysis

The null hypothesis is that there is no difference between 40 previously treated FGF-2 subjects (original surgical group 1) and this study (adjusted surgical group 2). From our prior study, FGF-2 treated group 1 had an effective closure rate of 40% in 40 patients. To show a difference with 80% closure rate in group 2, the investigators need 14 subjects at alpha 0.05, beta at 0.2 and power of .80. If 20 subjects are enrolled, 5 will receive placebo and 15 FGF-2. Of the 15 subjects receiving FGF-2, if one withdraws (similar to our prior study), 14 FGF-2 treated subjects suffice for the analysis. If 80% or more FGF2 show TMP complete closure with durability to six months, the investigators will propose further phase II clinical trials.

Secondary outcomes of hearing and tympanic membrane mobility will be measured with audiometry and tympanometry. Pure tone averages will be analyzed using ordinary linear regression and speech discrimination will be analyzed using quantile regression. Tympanometry and wide band immittance will be descriptive only in this small cohort.

Monitoring and Quality Assurance

The quality, data, and safety will be monitored regularly by an independent medical monitor. Monitoring and oversight will be maintained and documented through monthly meetings that will include the PI and study staff. Adverse events if any will be reported according to MGB guidelines.