REB approval was never issued

The current standard treatment for chronic tympanic membrane perforations (TMP) involves having fat grafted from the patient and inserted into the ear, through the perforation, to promote healing using the fat graft myringoplasty (FGM) technique. Platelet rich plasma (PRP) has also been used to promote TMP healing and involves having the patient provide a blood sample, which is processed to produce PRP and applied to the perforation. This prospective, multi-centre study will evaluate whether combining both these techniques can improve the rate of closure in patients with chronic TMP involving <50% of the membrane.

Patients will be randomized to receive either the standard FGM treatment or FGM treatment with the addition of PRP. At 3 months postintervention a blinded observer will rate the degree of TMP closure. Differences in closure rates between the 2 groups will be compared.

This will be a multi-centre, single-blinded, randomised control trial. Patients will be randomised at each site to undergo FGM with or without PRP.

Step 1 in both groups is the same. Under LA (or GA if the patient is unable to tolerate this), a fat graft is taken from just behind the mastoid process, or more posteriorly just beneath the hairline if necessary. The graft is kept moist in 0.9% saline.The ear canal is injected with LA and the edges of the perforation are freshened. Gelfoam is placed into the middle ear and the fat graft placed on top of this until it touches the underside of the TM and slightly bulges through. In an attempt to achieve standardisation of surgical technique between sites, surgeons will be provided with an operative video to watch beforehand.

Step 2. The surgeon is informed of the randomisation outcome into Group 1 or 2. Group 1 (Non PRP): patients will have FGM alone, and this will simply be covered with a piece of saline-soaked Gelfoam cut to completely cover the perforation and graft.

Group 2 (PRP): as for Group 1, but the Gelfoam will instead be soaked in PRP derived from the patient's own whole blood. The generation of PRP is descibed below: -

• 10-20 mL of autologous blood collected from an antecubital vein is placed in an adenosine citrate dextrose-acid (ACD-A) collection tube to prevent premature activation
• Blood immediately placed in the centrifuge at 1100g for 10 minutes (once)
• Supernatant removed and collected into syringe
• Injected onto surface of fat graft
• Rest added to piece of gelfoam
• Place gelfoam + PRP on the TM perforation

Post op care: patients will be asked to keep their ear dry, and cover it with a cotton ball when taking a shower or bathing for the first week, but otherwise to leave it open. They will be given an eye dropper, and instructed to use 2 drops of white vinegar in the ear three times a day for 10 days, with the head tilted over to allow the vinegar to reach the fat graft. Vinegar drops are an effective treatment for Pseudomonas infection, the most common infection in the external ear canal, and are commonly used in external ear infections. If this causes any discomfort, the patients will be instructed to stop this, and this information will be recorded at first follow up. If there is infected discharge, the patient will be instructed to contact the surgeon, and will be treated with ciprofloxacin-containing eardrops for one week. The occurrence of such an event will be recorded, but will not be a censoring event for the purposes of the study.

Follow up schedule: Patients will be followed at 2 weeks, 6 weeks and 3 months post surgery.

A planned interim analysis will be performed at the half-way point of the study; each centre will be compared for heterogeneity with the other sites using ANOVA, for both group 1 and group 2. Data from sites with heterogeneous data may be discarded, and/or other sites recruited, and/or existing site enrollments increased as needed to allow adequate patient recruitment.

Our primary outcome measure will be tested with an unpaired t-test between the two groups. The investigators will also perform secondary analyses of the degree of closure in the non-complete closure populations between the two treatment arms using Wilcoxon non-parametric testing

Power analysis: the investigators have deemed a 30% difference in closure rate between the two groups as clinically significant. The investigators estimate the closure rate for the fat alone to be 50%, based on previous publications. To show a difference of closure rate of 70% in the blood treated group, with a power of 0.8 and an alpha error of 0.05 would require 45 subjects in each group. This is for a binary outcome of pass fail. The investigators anticipate some drop out, because of failure of follow up, in the order of 30% at 3 months, and so aim to recruit 46 subjects in each arm.

Inclusion Criteria:

• Tympanic membrane perforation measuring <50% of the area of the membrane
• Perforation present for at least 6 months (based on history or direct observation)
• All edges of perforation are visible

Exclusion Criteria:

• Active ear infection at the time of the procedure
• cholesteatoma present
• Patients on immunosuppressive therapy (including oral steroids) or chemotherapy
• Patients with previous failed attempt at perforation repair