Comparison of Efficacy and Safety of XEN45 Implant Versus Trabeculectomy in Asian Eyes

Précis: Combined phacoemulsification and trabeculectomy is able to achieve greater reduction in intraocular pressure (IOP), higher rates of complete surgical success and fewer postoperative manipulations and visits compared with combined phacoemulsification and Xen implantation in glaucomatous eyes. Purpose: Our study aims to compare and understand the differences between the efficacy and safety of XEN45 implantation and trabeculectomy in Asian eyes with glaucoma. Materials and Methods: This was a retrospective, single-center, comparative study of consecutive patients who underwent combined phacoemulsification and trabeculectomy (Phaco-Trab) from January 2013 to June 2014 and combined phacoemulsification and XEN45 implantation (Phaco-Xen) from May 2017 to September 2018 in a tertiary Ophthalmology center in Singapore. Outcome measures included IOP, number of anti-glaucoma eyedrops, success rate, factors leading to success/failure, number of postoperative interventions and visits required, and surgical complications. Results: A total of 137 eyes (91 Phaco-Trab, 46 Phaco-Xen) were included. Phaco-Trab group had greater mean IOP reduction at all time points beyond postoperative month (POM) 1 (mean difference 2.9 to 3.8 mm Hg; P<0.05), and greater reduction in mean number of antiglaucoma eyedrops beyond POM3, thought this was not statistically significant. At POM12, complete success was achieved in 83.5% in Phaco-Trab and 52.2% of Phaco-Xen group, respectively (P<0.001). There was no significant factor associated with surgical failure, other than the difference in surgical procedure. Phaco-Trab group required fewer number of postoperative interventions (P=0.009), with only a mean of 0.1 bleb interventions required per patient, versus 1.5 in Phaco-Xen group (P<0.001). Safety profiles in both groups were comparable, with no statistically significant difference in intraoperative/postoperative complications. Conclusion: Phaco-Trab has significantly higher reduction in both IOP and number of antiglaucoma medications compared with Phaco-Xen group, with greater surgical success and fewer postoperative manipulations and visits. Safety profiles were comparable.

G laucoma is one of the top leading causes of blindness worldwide and can be managed with topical medications, laser interventions or surgeries. 1 Recently, there have been many surgical advancements in the field of glaucoma, especially with regard to minimally invasive glaucoma surgery, which provides a safe and less-invasive alternative surgical technique to traditional glaucoma filtration surgeries in lowering the intraocular pressure (IOP). 2 Of these, XEN45 gel implant (Aquesys/Allergan Inc., Irvine, CA), is a novel device that is injected ab-internally (hence, conferring minimal-to-no intraoperative conjunctival manipulation), providing an outflow pathway into the subconjunctival/sub-Tenon space. This provides a minimally invasive and a safe alternative to the traditional glaucoma filtration surgery. 3 Many studies have thus far described the efficacy of XEN45 gel implant as a standalone procedure or as combined cataract surgery with XEN45 gel implant insertion. However, only few studies have compared the efficacy and safety of XEN45 gel implantation with traditional glaucoma filtration surgery (trabeculectomy), [4][5][6][7][8] of which none of the studies have been performed solely on Asian eyes. Our study is novel in looking into Asian population, which has shown to have higher rates of fibrosis and failures following subconjunctival filtering surgery. [9][10][11] Therefore, our study aims to compare and understand the differences between the outcome of XEN45 implantation and trabeculectomy in Asian eyes, so as to allow us to make better decisions in the management of glaucoma patients in the future.
Our study included only patients who underwent combined phacoemulsification and glaucoma surgery (with either XEN45 implant or trabeculectomy). This is to adjust for the development of cataract which may hinder postoperative assessment and management of glaucoma, and also the effects that cataract surgery might have on glaucoma filtration surgery. In fact, there have been differing opinions in the literature whether efficacy of XEN-45 is the same when used as a stand-alone procedure or in combination, where some reported similar success rate 12-15 while others reported lower success [16][17][18] when used in conjunction with cataract surgery. Therefore, we standardized our recruited patients as those who underwent combined procedures for both groups.

MATERIALS AND METHODS
This was a retrospective, single-center, comparative study of consecutive patients who underwent combined phacoemulsification and trabeculectomy (Phaco-Trab) from January 2013 to June 2014 and combined phacoemulsification and XEN45 implantation (Phaco-Xen) from May 2017 to September 2018 in a tertiary Ophthalmology center in Singapore. The study protocol was approved by the Institutional Review Board and met the tenets of Declaration of Helsinki .   Patients above 40 years old with primary open-angle  glaucoma, pseudoexfoliation glaucoma, pigment dispersion  glaucoma, normal tension glaucoma and primary angle  closure glaucoma, on at least one antiglaucoma medication, and presence of cataract were included. Excluded patients were those with other secondary glaucoma, previous intraocular surgery or procedures (but not including laser glaucoma procedures), and those who had shorter than 6 months of follow-up postoperatively. Preoperative baseline patient and ocular characteristics can be found in Table 1.

Surgical Technique for Phaco-Xen
Our surgical technique for ab-interno XEN45 implantation was described in our recent study. 19 Standard phacoemulsification was performed first before implantation of XEN45. After the insertion of intraocular lens, 0.1 to 0.2 mL (at the discretion of attending surgeon) of 0.2 mg/mL of mitomycin-C (MMC) was injected in the superonasal quadrant of the conjunctiva, followed by a massage over the area of anticipated XEN45 implant insertion. The injector was introduced via the main incision into the anterior chamber, and the needle was directed across the anterior chamber toward the superonasal quadrant. Under the guidance of direct gonioscope, the needle was advanced above the pigmented trabecular meshwork, making a scleral tunnel and exiting~3.0 mm from the limbus in the subconjunctival space. A second instrument was used to provide countertraction. The injector was then withdrawn, and the placement of XEN45 implant in subconjunctival and anterior chamber was confirmed, and the viscoelastic was removed from anterior chamber via irrigation and aspiration. Subconjunctival, intraocular or topical antibiotic with steroid was administered at the end of surgery.

Surgical Technique for Phaco-Trab
Superior corneal traction suture was placed. A fornixbased conjunctival peritomy was performed in superonasal/ superotemporal quadrant and a partial thickness scleral flap was made. MMC of 0.2 to 0.4 mg/mL was injected in subconjunctival space, or sponges soaked in 0.2 to 0.4 mg/mL MMC was placed under the conjunctival flap for 2.5 to 5 minutes (at the discretion of attending surgeon) before copious irrigation with normal saline. Standard phacoemulsification was then performed. After insertion of intraocular lens, anterior chamber was entered under the scleral flap. A sclerostomy, followed by a peripheral iridectomy, was created. The scleral flap was sutured with interrupted 10-0 nylon sutures (fixed/releasable) and the conjunctiva was closed with 8-0 vicryl. Subconjunctival, intraocular or topical antibiotic with steroid was administered at the end of surgery.

Postoperative Management
Patients were told to stop their antiglaucoma medications immediately after operation. Frequency of postoperative follow-up visits was at the discretion of the overseeing clinician, but all patients had at least the following visits: postoperative day 1 (POD1), month 1 (POM1), month 3 (POM3), month 6 (POM6), and month 12 (POM12). Postoperative data that were collected included: number of days from surgery, best-corrected visual acuity, IOP, number of anti-glaucoma medications, number and type of postoperative interventions, and presence of complications.
Patients were given topical antibiotic prophylaxis for 4 weeks and topical steroids for a minimum duration of 12 weeks for both groups. Decisions on the duration of topical therapy, need for antiglaucoma medications, interventions and reoperations were made by the treating physician.
Postoperative interventions included: bleb massage, needling of bleb only, subconjunctival antimetabolites (5-fluorouracil or MMC) or steroid injection (dexamethasone) with or without needling, anterior chamber reformation for hypotony, and suturelysis (specifically for Phaco-Trab). The former 3 procedures and suturelysis would usually be performed in the presence of suboptimal IOP control with low, flat or encapsulated bleb.

Outcome Measures
Primary outcome of this study was IOP at each time point (POD1, POM1, POM3, POM6, POM12) compared with preoperative baseline. This study used similar definition of complete success, qualified success and failure as our previous study on efficacy of XEN45 in East Asian Eyes 19 and were adapted from the World Glaucoma Association guidelines 20 : Failure was defined as: Sustained readings of IOP > 18 mm Hg or < 20% reduction from baseline on at least 2 consecutive followup visits, despite additions of anti-glaucoma medications, from (and inclusive of) POM1 timepoint onwards. Sustained readings of IOP ≤ 5 mm Hg on 2 consecutive follow-up visits, from (and inclusive of) POM1 timepoint onwards.
No perception of light after surgery or presence of visionthreatening severe complications. Reoperations for glaucoma (such as other glaucomafiltering surgery, or cyclo-destructive procedures), but not including slit-lamp based interventions. Revision or removal of XEN45 implant.
Eyes that did not meet the above criteria for surgical failure and did not require the use of additional antiglaucoma medications were considered to have met the criteria for "complete success." Eyes that did not meet the criteria for surgical failure but required additional antiglaucoma medications were labelled to have achieved "qualified success." Any data collected after surgical failure was not included in the study analysis.
Secondary outcome measures included number of antiglaucoma eyedrops, factors leading to success or failure, number of postoperative interventions and visits required, and intraoperative and postoperative complications.

Statistical Analysis
Before the study, we used MedCalc software to determine that a sample of at least 44 patients in each group was required to detect a difference of 2.8 mm Hg in mean IOP, at a significance level of 0.05, power of 0.80, and assuming a SD of 4.5 mm Hg for XEN45 and 4.7 mm Hg for trabeculectomy. 6,19 Remaining statistical analysis was performed using IBM SPSS Statistics version 22 (IBM Corp, New York). Statistical significance is defined by P-value < 0.05. Data was expressed as number (%), mean (SD or 95% confidence interval), median (interquartile range). Where Kolomogorov-Smirnov and Shapiro-Wilk test showed significant P-value, indicating that the variable follows a non-normal distribution, a nonparametric test was used for analysis. Wilcoxon Signed Rank test was used to evaluate the postoperative changes in IOP and number of glaucoma medications from preoperative baseline in respective groups, and Wilcoxon Rank Sum test was used to compare between the 2 groups. Categorical variables were compared using χ 2 test and Fisher exact test, as needed. Kaplan-Meier method was used to analyze the probabilities of survival in our study population.
Binary logistic regression was performed to review the effect of potential factors that may affect the probability of surgical failure at POM12. The model included type of intervention (Phaco-Trab vs. Phaco-Xen), age, sex, underlying diabetes mellitus, preoperative IOP, severity of glaucoma, preoperative number of anti-glaucoma medications, and type of glaucoma.
included in this study. There was no statistically significant difference in the baseline characteristics between the 2 study groups ( Table 1). The mean postoperative follow-up period of the eyes included in our study was 347 days (range 83 to 471 d) for Phaco-Xen group (for which 1 eye had surgical failure at 83 days, before POM6 visit, and data from visits after failure were not included in the study) and 358 days (range 146 to 507 d) for Phaco-Trab group.

Intraoperative Pressure and Number of Antiglaucoma Medications
Both study groups had reduction in IOP postoperatively ( Table 2, Fig. 1). This difference between preoperative and postoperative IOP was statistically significant at all time points for the Phaco-Trab group, with greatest IOP reduction being achieved POM3 (−5.8 mm Hg) where there was 32.0% mean reduction of IOP from preoperative baseline. Such differences between and preoperative and postoperative IOP was also statistically significant at POD1, POM1, POM3, and POM12 for the Phaco-Xen group, with greatest reduction being achieved at early stages, in particular, POD1 (−8.5 mm Hg) where 47.0% mean reduction was achieved. The mean IOP reduction was significantly greater in the Phaco-Xen group in POD1 (P = 0.006) compared with Phaco-Trab group by 5.0 mm Hg, but at all time points after POM1, Phaco-Trab group was able to achieve better IOP reduction (mean difference 2.9 to 3.8 mm Hg) (P < 0.05).
Both groups saw a significant reduction in the number of anti-glaucoma medications postoperatively (mean reduction of 2.0 to 2.8 drops) at each time point (P < 0.001, Table 2, Fig. 2), with greater reduction (mean difference of 0.2 to 0.5 drops) seen in the Phaco-Trab group at later timepoints (POM3, POM6, and POM12) though this was not statistically significant.

Surgical Success
At POM12, complete success was achieved in 83.5% of patients in the Phaco-Trab group compared with 52.2% in the Phaco-Xen group (P < 0.001), with more requiring concomitant eyedrops to achieve optimal IOP control (ie, qualified success) in the Phaco-Xen group (21.7%) compared with the Phaco-Trab group (11.0%) though this was not statistically significant (P = 0.0924). Kaplan-Meier curve was used to depict the association between the type of surgical intervention and surgical outcome (complete and qualified success), which showed a statistically significant difference between the 2 groups (P = 0.00087) (Fig. 3). Of the 5 patients in the Phaco-Trab group who had surgical failure, 3 were due to suboptimal IOP, of which 1 required reoperation for IOP reduction and 2 were due to hypotony. On the other hand, 12 patients who had surgical failure in the Phaco-Xen group were all due to suboptimal IOP control, of which 10 required another operation.

Factors Affecting Outcome
Multivariable statistical model was used to find possible factors affecting outcome (Table 3). Only the differences in surgical procedure had significant impact on risk of failure at POM12 (P = 0.005), with the Phaco-Xen group having higher risk of failure than the Phaco-Trab group [odds ratio (OR) 6.210, 95% confidence interval (CI): 1.736-22.213]. There was higher likelihood of surgical failure in eyes with higher preoperative number of anti-glaucoma medications (OR 3.403; P = 0.096), whereas lower likelihood of surgical failure was observed in those with underlying diabetes mellitus (OR 0.282; P = 0.074) but these were not statistically significant (Table 3).
The majority of the eyes (82.6%) from the Phaco-Xen group required a form of bleb intervention (needling of bleb only, subconjunctival antimetabolites or steroid injection with or without needling; but not including bleb massage) postoperatively. Of this, the majority had undergone these postoperative interventions in POM1 (37.0%), POM3 (43.5%), and POM6 (37.0%).

Postoperative Visits
Patients were reviewed at each of the postoperative time points as mentioned above, but further need for postoperative visits were decided at the discretion of the overseeing physician. The Phaco-Xen group required a significantly greater number of postoperative visits within 1 year follow-up (mean 10.0) compared with the Phaco-Trab group (mean 8.3) (P = 0.002).

Safety Outcomes
There was no statistically significant difference in the number of intraoperative (P = 0.086) and postoperative (P = 0.359) complications between the Phaco-Xen and the Phaco-Trab group ( Table 5). The main intraoperative challenge was difficulty with XEN45 implantation, such as XEN45 gel stent becoming broken or spliced, requiring reimplantation, or requiring multiple attempts of implantation. Postoperative complications included shallowing of anterior chamber without hypotony (of which 2 required anterior chamber reformation in the Phaco-Xen group), numerical hypotony (of which 1 clinically had hypotonous maculopathy in the Phaco-Trab group), hyphema, persistent inflammation, malignant glaucoma and other mechanical complications ( Table 5). None of the patients from both groups had loss of light perception postoperatively.

DISCUSSION
This study suggests that both Phaco-Trab and Phaco-Xen can significantly reduce IOP postoperatively. However, the Phaco-Trab group had a greater reduction in IOP (POM1 and beyond) and number of antiglaucoma eyedrops than the Phaco-Xen group (POM3 and beyond). The Phaco-Xen group had the greatest reduction of IOP at POD1 likely due to peritube leak 21 which resolved by POM1 (Fig. 1), but also possibly due to progressive fibrosis after surgery resulting in subsequent IOP increase. By POM12, the mean IOP reduction was greater in the Phaco-Trab group, and this was in keeping with other comparative studies 6-8 which also showed that greater reduction in IOP was achieved with trabeculectomy compared with Xen implantation. Our percentage reduction in IOP, however, was generally less than those reported in other studies-where our reported mean 13.2% IOP reduction at POM12 for Phaco-Xen was lower compared with other studies which reported IOP reduction of 22.9% to 55.3%. 17,[22][23][24][25] Higher rates of fibrosis and failures following subconjunctival filtering surgery were shown in Asian eyes, [9][10][11] which may explain the more modest IOP reduction as compared with the published literature. 17,[22][23][24][25] Overall, the Phaco-Trab group (83.5%) also had greater complete success compared with the Phaco-Xen group (52.2%; P < 0.001). This was also observed in other comparative studies 6-8 with descriptively higher proportion of patients achieving complete success at POM12 for trabeculectomy group (range: 39% to 78.6%) compared with Xen group (range 33% to 66.2%) though these were all not statistically significant. Although our study had lower mean percentage reduction in IOP, the rate of complete success in Phaco-Xen group (52.2%) in our study was comparable to other reported studies on Xen implantation (28.6% to 66%), [15][16][17]22,26 showing that our surgical outcomes approximate those from other institutions. The differences may be explained by differences in study protocols, study population and definition of complete success.
Other studies have found that factors such as age, 27-29 race, 4 and diabetes mellitus 4,29 may affect surgical outcome of subconjunctival filtering surgery. However, our study did not find any association between success at POM12 and sex, age, diabetes mellitus, preoperative IOP, and preoperative glaucoma severity. Further large-scale studies may be needed to look into these factors and its impact on surgical outcome.
In all, 82.6% of eyes in the Phaco-Xen group required postoperative bleb manipulation versus 12.1% of eyes in the Phaco-Trab group. Other comparative studies showed conflicting results as well. While Theilig et al 8    due to a few factors: (1) lower threshold to perform bleb manipulation as we generally performed bleb-related procedures at the slit-lamp rather than in the operating theater. (2) Bleb manipulation rates may notably seem higher as we also included cases where subconjunctival anti-metabolite injection was performed alone without needling, in an early attempt to reduce bleb scarring response. (3) Asian eyes are known to have greater scarring response. 11 (4) Limited experience with XEN-45 blebs resulting in a more aggressive approach. (5) Inability to control the ideal siting of XEN-45 implant in the supra-Tenon's layer (free from Tenon's tissue) and the lumen of the XEN-45 being only 45microns could be more easily be obstructed by fibrous tissue as compared with trabeculectomy (6) 30.6% of eyes included in Phaco-Xen group was of "offlabel" use, where use in primary angle closure glaucoma or normal tension glaucoma have not been listed as standard clinical indications for use of XEN-45. 33 Safety profiles between the Phaco-Trab and Phaco-Xen group were similar in our study, with no significant difference in the intraoperative and postoperative complications in both groups. Most of the complications in the Phaco-Xen group were due to technical difficulties with Xen implantation requiring reimplantation intraoperatively (mainly due to breaking of XEN-45 implant), 34 likely due to the learning curve required to perform Xen implantation.
Our study is not without its limitations. This was a single-center retrospective study with a small sample size and short study period. This study included consecutive patients who underwent Phaco-Xen and Phaco-Trab in their respective time periods and those who defaulted follow up (those with < 6 mo of follow up postoperatively) or met our other exclusion criteria were excluded from the study. Further analysis between 18 excluded and 91 included eyes were performed in the Phaco-Trab group, which showed some ocular and demographic differences between the 2, particularly due to the fact that many patients who were excluded were due to them having secondary glaucoma, therefore having younger age with higher baseline IOP and number of antiglaucoma medications. This difference could potentially introduce bias into the study. There was no difference in the demographics of 1 excluded and 46 included eyes for the Phaco-Xen group.
Furthermore, results from several glaucoma specialists were included in this study, which may have resulted in surgical outcome being affected by the differences in their experience and learning curve. The technique and practice of XEN-45 implantation have been evolving over the years, and our study included the earlier cases of XEN-45 implantation when it was first introduced to our center, resulting in learning curve and limited experience with XEN-45 implant. Moreover, the time periods for data collection for Phaco-Trab and Phaco-Xen groups were different. But given that the technique for trabeculectomy has been largely unchanged over the number of years, the differences in the time period will unlikely have a significant impact on the outcome of trabeculectomy surgery. Lastly, there was no standardization of criteria for bleb manipulation, restarting or escalating antiglaucoma medications, and follow-up protocol and such postoperative decisions were made according to each surgeon's discretion.
To our knowledge, this is the first study comparing Phaco-Xen and Phaco-Trab in Asian eyes. Despite the retrospective and nonrandomized nature of our study, we report realworld and early experience with XEN-45 as compared with the gold-standard trabeculectomy surgery. This is important information for surgeons starting out with XEN-45 surgery. Our study shows that Phaco-Trab has higher rates of complete success compared with Phaco-Xen, with greater reduction in IOP and antiglaucoma eyedrop use postoperatively, and fewer bleb-manipulations and postoperative visits that are required. Safety profiles of both groups were comparable. Therefore, it remains debatable as to whether combined phacoemulsification with minimally invasive subconjunctival glaucoma surgeries such as XEN45 implantation will be a comparable alternative to traditional glaucoma filtration surgeries like trabeculectomy in Asian eyes. Further studies are needed to evaluate these differences between the 2 groups, factors affecting these outcome, cost-effectiveness, and patient-reported satisfaction, so as to make better clinical judgment as to when to consider one surgical procedure over another.