Efficacy and Safety of ab Interno Trabeculectomy With the Kahook Dual Blade

This new device, the Kahook Dual Blade, has been specially designed to create a gentle cut compared to other cutting-devices that also remove the trabecular meshwork. Preclinical investigation showed no harm in the surrounding tissues. Only two prospective studies have been published, showing good results in terms of IOP reduction and a significant cut-off in the glaucoma medications. However, to better assess the real effect, a prospective comparison with cataract surgery or any other trabecular technique should be performed. We also aimed to determine the safety of the procedure.

Study design A prospective, randomized controlled, interventional study was conducted to determine the efficacy and safety of the Kahook Dual Blade. Patients with cataract and open-angle glaucoma or ocular hypertension (OHT) were included.

Subjects Patients were consecutively recruited from the Department of Glaucoma. Eligible patients were asked to participate if cataract surgery was indicated due to best-corrected visual acuity (BCVA) bellow 0.5 (in decimal scale), and moderate IOP reduction and a cut-off of the glaucoma medications was also aimed. Patients included in the study had been previously diagnosed with primary open-angle (POAG), pseudoexfoliative glaucoma (PSX) or OHT, and mild-to-moderate glaucoma according to the Hodapp-Parrish-Anderson classification, preoperative IOP ≤ 24 mmHg on at least one hypotensive medication, and coexistence of cataract without narrowing the angle (visible scleral spur at least in 2 quadrants). Both eyes from the same patient could be eligible if they both met the inclusion criteria.

Subjects who did not fulfill the inclusion criteria were excluded: patients with other glaucoma subtypes (elevated episcleral venous pressure, orbital occupancy, Sturge-Weber syndrome or any type of angle-closure glaucoma). Patients who underwent any intraocular surgery in the previous six months, including any laser or surgical intervention for glaucoma, or patients with any ophthalmic diseases that may interfere with tests were excluded. Severe or end-stage glaucoma or those with an indication of filtering or glaucoma drainage implant surgeries due to a high IOP with the maximum tolerated medication or glaucoma progression were also exclude.

All subjects were thoroughly examined before clinical intervention. This included a medical history review, slit-lamp examination, IOP determination with hand-held applanation tonometer, gonioscopy, dilated fundus examination, endothelial cell count (ECC) including cell and size automated detection with specular microscopy (EM-3000, Tomey Corporation, Nagoya, Japan), simulated keratometry (simK) determined by topography (Pentacam®, Oculus, Wetzlar, Germany), and standard automated perimetry (VF) (G1-Tendency Oriented Perimeter, Octopus® 1-2-3, Haag-Streit, USA).

Assignation and surgical technique For safety reasons, patients did not discontinue any topical hypotensive medications before surgery.

A randomized, controlled study was designed. Eyes were randomly assigned to each treatment group (www.random.org). If both eyes of the same patient were eligible to participate in the study, they could be allocated in different treatment groups due to randomization. All surgeries were performed by two experienced glaucoma surgeons. Standard cataract surgery with phacoemulsification using a 1.8 mm clear corneal incision (CCI) and intraocular lens (IOL) capsular bag placement was performed under topical anesthesia and 1% intracameral lidocaine. The surgeons were blinded to randomization until the end of the cataract surgery. At this point, if the patient was assigned to the treatment group the surgeon instilled acetylcholine 1% to reduce the pupillary size and facilitate visualization of the angle structures. To help maintain width, the anterior chamber was filled with a viscoelastic agent. Since the ab interno trabeculectomy is performed in the nasal quadrant, the surgeons then situated temporal, the patient's head was rotated 45 degrees in the opposite direction of the surgical site, and the microscope was tilted 30 to 45 degrees towards the surgeon. To guarantee better access to the nasal area of the angle, a 1.8 mm keratome was employed to widen the paracentesis. Further details of the surgical technique are described elsewhere.

Postoperative medications and follow-up Hypotensive medication was not administered intra- or post-operatively unless an IOP spike was detected. Postoperative topical treatment included a fixed combination of antibiotic plus steroid (tobramycin 0.3% and dexamethasone 0.1%) six times a day for seven days; this was tapered over the next four weeks.

Visits were undertaken at 1 day, 1 week, 1 month, 2 months, 3 months, 6 months, and 12 months. During follow-up, results of the slit-lamp examination, applanation tonometer IOP, and the number of medications were recorded at all visits. The VF and the ECC were also re-tested at 6 and 12 months. All the postoperative visits were conducted by the same examiner (NVA). Topical glaucoma medication was re-introduced if the target IOP according to the severity of the visual field was not achieved, o a higher than preoperative IOP was found.

Outcome measures - Data analysis Data obtained from each group were evaluated to determine the efficacy and safety of both interventions. The primary study outcomes were IOP and the number of glaucoma treatments. VF, ECC, BCVA and simK were secondary outcomes used to define the safety of both procedures. Success was defined as an IOP reduction of ≥ 20% from baseline; this cut-off was based on that reported in other glaucoma surgeries and is used for evaluation of this trabecular technique. The Saphiro-Wilk test was used to assess the sample distribution. Independent and dependent sample t-tests, when possible, were used to compare the results between groups and over time for each group. As both eyes of the same patient could be included in any of the treatment groups, generalized estimating equations (GEE) for repeated measures (IOP and number of medications) were also performed. A random-effect logistic regression was employed to analyse baseline characteristics associated with the success endpoint (IOP reduction of ≥20% from baseline). Statistical significance was set at a two-sided p-value ≤ 0.05. Statistical analyses were performed using Stata® version 14.1 (StataCorp, College Station, Texas, USA).