XEN^® Gel Stent in Medically Refractory Open-Angle Glaucoma: Results and Observations After One Year of Use in the United States

This was an investigator-initiated, prospective, noncomparative study of glaucoma patients referred within Park Nicollet Clinic, a multispecialty, non-university medical practice in St. Louis Park, Minnesota, USA. Eligible patients enrolled in the study were 21 years or older with severe-stage OAG that was uncontrolled, refractory to medical therapy, and requiring either trabeculectomy or tube shunt surgery to halt glaucoma progression. Uncontrolled, medically refractory OAG was defined as uncontrolled IOP, documented visual field progression, and/or increasing optic nerve damage despite maximum medical therapy (four or more topical IOP-lowering medications, or fewer in the case of tolerability or efficacy issues) and laser treatment, when possible. Patients with mild or moderate-stage OAG were not included in the study, as other MIGS devices were available for eyes with these stages of disease. Of the FDA-approved MIGS devices, the XEN^® stent is the only one that utilizes the subconjunctival space and the only one that can achieve intraocular pressure lowering to the same level as standard trabeculectomy or tube shunt surgery. Unlike the other categories of MIGS, the FDA approved use of the XEN^® stent with or without cataract surgery.

All procedures performed in studies involving human participants were in accordance with the ethical standards of the HealthPartners Institute Institutional Review Boards and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all participants in the study for both the procedure itself and inclusion in the study. The HealthPartners Institutional Review Board approved this study.

Preoperatively, eyes underwent a complete ophthalmic examination, including slit-lamp and funduscopic examinations, Goldmann applanation tonometry (two separate measurements, with resetting of the dial to 10 mmHg between readings), gonioscopy, corneal pachymetry, standardized automated perimetry, and optical coherence tomography analysis. Patients were required to have healthy, free, and mobile conjunctiva superiorly in the area of planned XEN^® stent implantation and an open angle with a visible trabecular meshwork. Exclusion criteria included previous incisional glaucoma filtration or tube surgery, conjunctival pathology, aphakia, epithelial downgrowth, anterior chamber intraocular lens, presence of vitreous in the anterior chamber, and the following types of glaucoma: uveitic, neovascular, angle-closure, iridocorneal endothelial syndrome, and Axenfeld-Rieger syndrome.

At minimum, postoperative examinations were conducted at day 1, week 1, and months 1, 3, 6, and 12. Each visit included slit-lamp and posterior segment examinations, IOP by Goldmann applanation tonometry, and gonioscopy. Standardized automated perimetry was performed at 1 year at minimum. Endothelial cell counts were not measured pre- or postoperatively. Visual acuity, medications, and adverse events were also recorded.

Surgeries were performed from March 2017 to May 2018 by a single surgeon (PK). All surgeries were performed under topical anesthesia. After prepping and draping the eye in sterile ophthalmic fashion, the superior nasal conjunctiva was marked 3 mm from the limbus in the planned exit point of the stent. Using a 30G needle, 0.1 ml of mitomycin-C (0.2 mg/ml) was injected posteriorly into Tenon’s capsule. Using cellulose sponges, the mitomycin-C was massaged over the area of anticipated stent placement, with efforts to keep it away from the limbus. Main and side port incisions were created. A combination of cohesive and dispersive viscoelastic was used to fill the anterior chamber. In patients with cataract surgery, preservative-free lidocaine was injected before the viscoelastic. Phacoemulsification and lens implantation were performed before XEN^® stent placement. The angle was inspected before XEN^® stent insertion with a disposable gonioprism (Katena, Denville, NJ, USA). After obtaining the XEN^® stent inserter, the protective devices were removed and the inserter was checked to confirm that it contained a stent. The inserter was introduced into the eye through the main incision, and the needle was directed across the anterior chamber towards 12 o’clock. Countertraction was supplied by using a Vera hook (Katena) in the side port. The gonioprism was then reapplied to ensure that the inserter was positioned to engage the angle just anterior to the trabecular meshwork.

The inserter needle was then slowly pushed through the sclera until the needle was seen exiting into the subconjunctival space close to the preoperative markings and, ideally, above Tenon’s capsule. The needle was then rotated towards 12 o’clock. Engagement of the blue lever, which advances and releases the XEN^® stent, occurred next, followed by slow advancement. Subsequent slow retraction of the inserter needle resulted in stent placement into the subconjunctival space. The injector was then removed from the eye. Stent positioning in the subconjunctival and anterior chamber stent was confirmed. Tissue or stent manipulation was rarely needed after insertion. Automated irrigation/aspiration was used to remove the viscoelastic from the eye. While doing so, the presence of a bleb superiorly was confirmed. The incisions were hydrated with a balanced salt solution and, where needed, sutured with 10-0 nylon.

Glaucoma medications were continued until the day of surgery. A washout period was not included. After XEN^® gel stent implantation, all glaucoma medications were stopped in the operative eye. Postoperative examinations were performed on day 1, week 1, and months 1, 3, 6, and 12. Additional postoperative visits were scheduled in between these time points, as dictated by the clinical findings. Topical antibiotics were used four times a day for 1 week. Topical steroids were used four times a day for 1 month, then tapered over 4 weeks. Restarting of glaucoma medications, bleb needling, 5-fluorouracil (5-FU) injections, or other interventions were done at the discretion of the surgeon.

Clinical exam findings determined if and when needling and 5-FU were to be done. Most commonly, eyes had elevated IOP with signs of decreasing bleb function (i.e., lack of microcysts, vascularization, and Tenon’s encapsulation, scarring in and around the external tip of the stent). Needling was performed under microscopic guidance in a minor procedure room. Betadine, antibiotic, and 2% lidocaine were topically applied. A solution of 2% lidocaine, 0.1 ml, was then injected subconjunctivally temporally, and cotton-tipped applicators were used to disperse the lidocaine to the location of the external stent. Under sterile conditions, an angled 27G needle was directed towards the external end of the stent and surrounding fibrosis. Using the tip of the needle, the fibrous adhesions surrounding the external end of the stent were lysed, and an attempt was made to raise the stent above the episcleral and scleral surface. On completion of the procedure, 0.1 ml of 5-FU (50 mg/ml) was injected temporally, and additional antibiotic drops were instilled. Subconjunctival injections of mitomycin could not be used due to institutional regulatory policies. Both topical antibiotic and steroid drops were prescribed four times a day. The antibiotic was stopped at 7 days and the steroid tapered over 4 weeks.

Measurements included IOP, medications, and complications. Degree of success was defined by World Glaucoma Association guidelines, which were also used for data presentation [30]. Complete success was defined as an IOP reduction of ≥ 20% and an IOP of < 18 mmHg without medication or any secondary glaucoma intervention. Qualified success was defined as an IOP reduction of ≥ 20% and an IOP < 18 mmHg with and without medication or any secondary glaucoma intervention. Eyes with follow-up that were in neither of these groups were deemed failures.

Data analysis was performed using R 3.4.1 statistical software. The IOP was analyzed by calculating various descriptive statistics, including range, median, and mean. The standard deviation, standard error, and 95% CI were computed around the mean. The change in IOP from preoperative measurements to each of the subsequent time points was also examined. The resultant values were reported as both a numeral and a percentage. These analyses were replicated for various subsets of the data, including XEN^® stent standalone procedures versus XEN^® stent with cataract surgery. To visualize the data, the following plots were created: (1) combined graph showing mean IOP change, percentage change of mean IOP, and mean medication reduction values from baseline over time, and (2) scatterplots of preoperative and postoperative IOP for 6 months and 12 months.

The above descriptive statistics were also calculated for the raw observed medication use values and the change in medication from preoperative measurements at the study time points. These computations were carried out for the overall study group, the standalone XEN^® stent eyes, and the combined XEN^® stent and cataract surgery eyes. A Wilcoxon test was chosen to analyze our data. This nonparametric test was chosen over a t test because our data failed the normality assumption.

Intraoperatively, one eye (2.1%) had a small, nonexpulsive choroidal hemorrhage that self-resorbed without issue. At 1 month, two stents (4.3%) in different patients were noted to have eroded through the conjunctiva, which necessitated immediate stent removal. A detailed review of both eyes was conducted. In each eye, XEN^® stent insertion was uneventful, with desired placement above Tenon’s capsule and underneath the conjunctiva. Mitomycin dosing and tissue manipulation were per study protocol. In the early postoperative period, each eye had a single-digit IOP, diffuse bleb, slightly shallow anterior chamber, and shallow serous choroidals. At postoperative day 1, the external portion of the stent was clearly visible; it was straight rather than angulated. However, in each eye, the midsection of the stent bowed anteriorly and pressed against the outer wall of the conjunctival filtration bleb. This appearance persisted throughout the entire month, despite choroidal resolution and anterior chamber deepening. With the thinness of the conjunctiva overlying the stent and concern that manipulation might create a buttonhole, neither stent was repositioned or redirected. Medications followed study protocol. Erosion was noted unexpectedly at the routine 1-month visit. Each eye was white and quiet. Neither patient reported any symptoms or new issues. Intraoperative stent removal was uneventful. In each eye, a second stent was inserted in a new location, away from the erosion. The second stents functioned less optimally than the first, but neither eroded. Fortunately, neither eye developed infection or vision loss from the erosion.

Hypotony (IOP < 6 mmHg) was seen in 10 eyes (21.3%) at day 1, with resolution in all eyes by 2 weeks. No hypotony, maculopathy, hyphema, endophthalmitis, anterior chamber flattening, cystoid macular edema, or perceived mitomycin-related complications occurred. One eye (2.1%) required trabeculectomy. No eyes showed significant loss in visual acuity due to either the procedure or adverse events.

Fourteen of 47 eyes (29.8%) required needling and 5-FU injection 1–3 months postoperatively. Needling was considered part of procedure management and not an adverse event, in agreement with another study [26]. Our 29.8% needling rate is similar to that of previous studies by Hengerer et al. (27.7%) [26], Galal et al. (30.7%) [25], and Grover et al. (32.3%) [28]. Our rate was lower than that reported by Fea et al. (50%) [31], Sheybani et al. (47%) [22], Schlenker et al. (43.2%) [29], and Mansouri et al. (37%) [27]. Our study differed from Grover et al. [28], who inserted the XEN^® stent without conjunctiva disruption, but applied mitomycin with a sponge after opening the conjunctiva.

Optimal time for needling with trabeculectomy varies widely, depending on surgical technique, surgeon preference, and patient population [32, 33]. As with trabeculectomy, the reported average time of needling after XEN^® stent insertion also varied with Mansouri et al. (4.5 months) [27] and Hengerer et al. [26] (1 week–3 months). In our study, needling was performed at 1–3 months, with improvement in IOP and bleb function in 10 of 14 eyes (71.4%). A second needling was performed in one eye without benefit.

No eyes showed significant loss in visual acuity due to either the procedure or adverse events. No observable corneal decompensation or other anterior segment issues related to the procedure were noted during the study. Endothelial cell counts were not included in this study.

The study has several limitations. It was a prospective study, but it was also a noncomparative analysis with no control group. In addition, the number of eyes is small. No medication washout was done at baseline; therefore, the unmedicated IOP is unknown. A randomized controlled study with a larger sample size and medication washout may yield different results.

Because all but one patient was white, this study cannot speak to the effectiveness or potential issues with XEN^® stent use in other ethnic populations such as African Americans and Hispanics. Neither can this study draw any conclusions about XEN^® stent effectiveness based on age.

XEN^® stent insertion is relatively straightforward, but as with any surgical procedure, there is a learning curve. As we became more experienced, the stent functioned more favorably when positioned above Tenon’s capsule. Eyes with stent placement more intra-Tenon’s capsule tended towards higher IOPs and a higher incidence of needling. Later eyes may have benefited from more ideal stent positioning. Needling techniques and timing differ among glaucoma surgeons, making it difficult to reach standardized conclusions about this procedure. In this study, needling occurred when clinical exam results suggested restricted aqueous flow. Because no needling took place after 3 months, it is unclear whether this procedure would be beneficial at later stages. Because post-needling antifibrotic medication was limited to 5-FU injections, we cannot comment on the effect of subconjunctival mitomycin-C injections on filtration enhancement.

Despite these study constraints, there is limited long-term data on the use of the XEN^® gel stent in the United States due to its relatively recent approval by the FDA. Thus, in spite of its limitations, we hope that this study broadens awareness and knowledge of this new treatment. In 12 months, the procedure has shown effective IOP lowering, a significant reduction in glaucoma medications, and a relatively good safety profile. However, the potential for conjunctival erosion exists, and with the significant needling rate, surgeons performing the procedure need to be comfortable with and knowledgeable about bleb management. In summary, the data from this study support the XEN^® gel stent as an effective, relatively safe, and less invasive surgical filtration procedure for patients with medically uncontrolled glaucoma, either as a standalone procedure or combined with cataract surgery.