Bevacizumab Versus Mitomycin C as Trabeculectomy Adjuvant in Uncontrolled Glaucoma: A Randomized Pilot Trial.

This is a prospective, parallel, randomized, comparative, interventional feasibility study. Eyes of Caucasian patients with medically uncontrolled glaucoma or intolerance to glaucoma medications, who are evaluated at the glaucoma service of the Athens Vision Eye Institute and are scheduled for filtration surgery within one calendar year, are enrolled. If both eyes of one patient are eligible one eye is randomly selected. Inclusion criteria include adult patients with primary or secondary open angle or angle closure glaucoma with preoperative intraocular pressure (IOP) > 21 mmHg on maximally tolerated medical therapy at least on 2 occasions prior to randomization with the ability to attend regular follow-up. Exclusion criteria include age (< 18 years), pregnancy, severe ocular surface disease, need for combined phacotrabeculectomy, uveitic or neovascular glaucoma, any prior intraocular surgery except for uncomplicated phacoemulsification and a history of a systemic thromboembolic event within 6 months before surgery. The research adheres to the tenets of the Declaration of Helsinki and is approved by the Human Research Ethics Committee of the Athens Vision Eye Institute. Informed consent will be granted from all patients preoperatively explaining indications and risks of the procedure.

All patients are recruited by the principal investigator and surgeon (GK). The study is unblinded to both patients and investigators. Eyes are randomly assigned to 2 treatment arms by an online random number generator (www.random.org): the "mitomycin C group" that undergoes standard guarded trabeculectomy supplemented with mitomycin C (Kyowa, Galabank Business Park, UK); the "bevacizumab group" that undergoes guarded trabeculectomy with a single 1.25 mg intracameral injection of bevacizumab (Roche, Welwyn Garden City, UK). The online random number generator is accessed for each individual patient by the nursing staff (circulating nurse), which is unaware of study design and patient's characteristics, immediately prior to the surgical procedure. Sample size analysis (www.powerandsamplesize.com) reveals that for a power of 80% assuming a 5% type I error, an average post-op intraocular pressure of 12mmHg and a similar sampling ratio, 13 eyes in each group are necessary to detect a 3mmHg difference in IOP at twelve months (superiority or non-inferiority design). Assuming a dropout rate of 30% at 5 years the investigators aim for 20 eyes in each study group. Furthermore, power analysis reveals that for a power of 80% assuming a 5% type I error, a surgical success rate of 90% at 1 year and a similar sampling ratio, 20 eyes in each group are necessary to detect a 24% difference in rate of surgical success at twelve months (superiority or non-inferiority design).

Before surgery, all patients undergo a comprehensive eye examination including slit lamp biomicroscopy, best corrected visual acuity (BCVA) determination, intraocular pressure measurement using Goldmann applanation tonometry (GAT) at least on 2 separate occasions performed one week apart, gonioscopy and dilated fundus examination. All patients should have at least one visual field test on file, performed not more than 6 months prior to randomization. The Humphrey Visual Field Analyzer (HFA II-i, Carl Zeiss Meditec Inc, Dublin, CA, USA) will be used and the 24/2 SITA-Standard protocol will be utilized in all instances. Preoperative data collected from the medical record of individual patients include age, race, sex, diagnosis, study eye, BCVA, number and type of glaucoma medications, intraocular pressure, global visual field indices, central corneal thickness (CCT), cup to disc ratio, number and type of previous intraocular surgeries and previous glaucoma laser procedures. All pseudophakic eyes should have a history of primary open angle or exfoliation glaucoma, and should have undergone uneventful cataract extraction with a posterior chamber intraocular lens implantation.

Patients are examined on post-operative day one, day three, week one and every week thereafter for the first month, and subsequently on post-operative months 1, 3, 6, 9, and 12. A window of + 1 week is allowed for the 1st and 3rd month post-operative visits, and a window of + 2 weeks is allowed for the 6th, 9th and 12th month post-operative visits. During each visit a comprehensive ophthalmic examination is carried out. This includes intraocular pressure measurement, glaucoma medication requirement update, BCVA, review of early and late complications related to surgery (i.e., blebitis, bleb leak) and additional surgical interventions. These data are recorded in the patients' medical record and are extracted and tabulated in Excel 2013 spreadsheets (Microsoft, Redmond, WA) by an individual who is not participating in patient's management decisions and is blinded with respect to treatment assignment. Subsequently, data are analyzed by a separate individual who reveals study group assignment and calculates means, surgical success rates and complication rates. All patients' data are dealt with confidentiality. Statistical comparisons are performed according to initial treatment group assignment (intention to treat analysis).

Additional post-operative visits and/or interventions at the slit lamp such as 5-fluorouracil (5-FU) injections, bleb needlings and laser suture lyses are at the discretion of the treating ophthalmologist and do not qualify as failures. BCVA is recorded by certified optometrists utilizing clear chart digital screens (Reichert, Depew, NY, USA) placed at 4 m. All patients will undergo 24/2 Swedish Interactive Threshold Algorithm (SITA) Standard Automated Perimetry (SAP) on an annual basis in the post-operative period. SAP will be repeated typically within one month if unreliable.