Segmental Suture Gonioscopy-Assisted Transluminal Trabeculotomy: Comparison of Superior Versus Inferior Hemisphere Outcomes

Abstract

Précis: 

Analysis of efficacy, safety, and risk factors for failure of superior versus inferior 180-degree segmental gonioscopy-assisted transluminal trabeculectomy showed no significant difference between approaches, with novel risk factors for failure identified.

Purpose: 

Compare the efficacy, safety, and risk factors for failure of superior versus inferior 180-degree segmental suture gonioscopy-assisted transluminal trabeculotomy (GATT).

Design: 

Multicenter, retrospective interventional cohort study of 297 eyes of 243 glaucomatous patients, which underwent superior or inferior 180-degree suture hemi-GATT surgery combined with phacoemulsification at one of 3 Canadian ophthalmological surgical centres in Calgary, Alberta or Toronto, Ontario

Main Outcome Measures: 

The primary outcome measure was the hazard ratio (HR) of failure for the “primary success” criteria. “Primary success” was defined as an intraocular pressure (IOP) <18 mm Hg and either 1) IOP reduced by ≥20% from baseline on the same number of IOP-lowering medications or 2) IOP ≤ baseline on fewer medications. Secondary outcome measures included HRs of failure for alternative criteria (“complete success”, “qualified success” and “20% IOP reduction”), cross-sectional analysis, and Cox proportional hazard analysis for risk factors associated with increased failure for the complete cohort.

Results: 

Baseline characteristics were similar between groups. The crude and adjusted HR of failure for the “primary success” criteria for superior surgeries relative to inferior surgeries were 1.27 (95% CI = 0.86–1.88) and 1.50 (95% CI = 0.91–2.46), with no statistically significant difference between approaches. Of the secondary criteria, there was statistical significance in favor of inferior surgeries only for crude analysis of the “20% reduction” criteria (1.40/1.27 (95% CI = 1.01–1.92). Increased risk of failure by the “primary success” measure for either intervention was seen with primary open angle glaucoma, advanced disease, and age below 70 years. There were no significant differences in the frequency of postoperative complications between cohorts, which were present in 72 superior (44.4%) and 67 inferior (49.6%; P value = 0.41) eyes; mostly early postoperative hyphema, iritis, and corneal edema.

Conclusions: 

This retrospective study showed no difference in inferior versus superior 180 degrees of hemi-GATT/phacoemulsification cataract surgeries through the majority of analyses. Nonmodifiable factors including glaucoma type, advanced disease, and younger age were associated with a significantly higher risk of failure in this cohort. Further study is warranted.

Glaucoma is an ocular disease that afflicts more than 60 million people globally and is the leading cause of irreversible blindness worldwide.¹ Management of intraocular pressures (IOPs) is key in preserving vision in individuals living with glaucoma. When medical and laser therapies do not adequately control IOPs, surgical intervention is required.^2–4

Minimally invasive glaucoma surgeries (MIGS) are a relatively novel subset of surgeries, which utilize microincisions and an ab interno approach to minimize disruption of ocular anatomy with a resulting reduction in complications and time to recovery while maintaining efficacy.^5–7 An additional advantage of MIGS is their potential to be used in combination with cataract excision for IOP control. Multiple studies have shown that while cataract development is increased in eyes that have undergone surgery, phacoemulsification has an adverse effect on IOPs in patients with prior traditional filtering surgeries.^8–10 Trabecular bypass MIGS, by contrast, are neutral or synergistic in combination with phacoemulsification.^11–15 Thus, concomitant cataracts and MIGS offer an opportunity to excise the cataractous lens, improve the refractive status, and lower the glaucoma disease burden in a single surgical exposure.¹⁶

Gonioscopy-assisted transluminal trabeculotomy (GATT) is a MIGS originally described in 2014 by Grover et al¹⁷ that has subsequently been shown to be effective at managing IOP in glaucoma of multiple etiologies, has relatively low rates of sight-threatening complications compared with traditional filtering surgeries, and is not antagonistic with concurrent phacoemulsification.^11,18–25 A recent meta-analysis of published GATT studies showed a mean IOP reduction of 9.81 mm Hg with a mean reduction in glaucoma medications of 1.67 across the 9 included articles, while sight-threatening complications were exceedingly rare.²⁰ In addition to these common benefits of MIGS, GATT can be performed with a suture, and thus for a fraction of the cost of many other MIGS that can require expensive commercial devices.

Although the original method was described as unroofing 360 degrees of the Schlemm canal, emerging evidence suggests that one may not need to treat the entire 360 degrees to achieve maximum efficacy.^26,27 Alternative techniques disrupting fewer degrees have been used by some surgeons with the advantages of titratability to patient needs, adaptability to anatomically difficult cannulations, and a potential further improved safety profile.⁷ Hemi-GATT (performed on 180 degrees of Schlemm canal) spares half of the trabecular meshwork (TM) for potential future interventions and thus may be utilized in a stepwise approach to management, allowing subsequent repeat GATT or filtering surgery if necessary.

Given the emerging evidence in support of the 180 degrees GATT approach and the limitations of previous studies, further investigation is warranted into variations of the technique that may affect patient outcomes. This study is a retrospective cohort analysis of consecutive patients who underwent hemi-GATT of either the inferior or superior 180 degrees of the Schlemm canal as an adjunct to phacoemulsification cataract surgery. The objective of this study was to evaluate the hazard ratio (HR) of surgical failure for eyes receiving superior versus inferior 180 degrees suture hemi-GATT surgery combined with phacoemulsification, evaluate relative efficacy and safety, and identify factors that may be associated with increased risk of failure for 180 degrees GATT.

METHODS

Design

This study is a multicenter retrospective cohort study of consecutive patients who underwent either superior 180 degrees or inferior 180 degrees suture GATT with concurrent phacoemulsification with intraocular lens implantation between June 2016 and October 2018 by two ophthalmological surgeons (P. G. and M. B. S.) and their fellows in one of 3 surgical centres in Calgary or Toronto, Canada. This study adhered to the tenets of the Declarations of Helsinki and was approved by the Health Research Ethics Board of Alberta (CTC-17-0076) and the University of Toronto Research Ethics Board (#00037802). Cases were identified by a filter function within the electronic medical record utilized at each site. Data were obtained through manual chart review and included the following: (1) demographics (age, sex, eye, and systemic disease), (2) baseline ocular characteristics (glaucoma diagnosis, best-corrected visual acuity, relevant medical conditions, previous ocular surgeries/interventions, and cup-to-disc ratio), (3) preoperative data (IOP and glaucoma medications), (4) surgical data (site, surgeon, quadrants, degrees of trabeculotomy performed, and intraoperative complications), and (5) follow-up (IOPs, glaucoma medications, postoperative complications, and additional surgeries/interventions).

Inclusion/Exclusion Criteria

Inclusion criteria included any eye, which underwent superior or inferior 180 degrees GATT combined with phacoemulsification cataract surgery during the aforementioned period at the above centers for the management of glaucoma of any etiology. Cases were documented as 180 degrees GATT only if the performing surgeon observed between 150 degrees and 210 degrees of successful trabeculotomy intraoperatively. Exclusion criteria included eyes that had previously undergone angle-affecting glaucoma surgeries (including trabeculectomy, tube shunts, GATT, and other MIGS), and cases with postsurgical follow-up <1 month.

Surgical Procedure

A 2.2 mm temporal clear corneal wound and two 1.0 mm paracenteses with an Microvitreoretinal blade were made at the surgeon’s 4 or 8 o’clock positions for access to the anterior chamber (AC) in each case. Cataract excision was performed before GATT using standard capsulorhexis and phacoemulsification techniques. Upon completion, the microscope was repositioned to allow for visualization of the nasal TM and a cohesive oculoviscoelastic device was used to inflate the AC to minimize blood reflux. Low-temperature cautery was then used to round out the tip of the 5-0 polypropylene suture to facilitate the advancement of the suture through the canal and provide some anchoring after advancing 180 degrees. A small goniotomy was then created through an ab interno approach with a 25G needle or Microvitreoretinal blade. Goniosynechialysis of 1 to 2 hours was performed to access the angle for goniotomy if necessary. The leading tip of the suture was then advanced through the goniotomy site with 23G Ahmed Micro-Graspers (MST, Redmond, WA) into the Schlemm canal. After advancing the suture 180 degrees (~20×1 mm or ~10×2 mm advancements, and usually at the initial start of suture resistance), it was grasped through the goniotomy site and pulled centrally while advancing (“walk the dog” technique) for the first 90–100 degrees of the trabeculotomy. The remaining portion of the trabeculotomy is completed by grasping the suture and externalizing it through the temporal wound (“rip-cord” technique) to complete the 180-degree trabeculotomy. Any intraoperative hyphema was evacuated from the AC. A 20%–40% dispersive oculoviscoelastic device fill was left in the AC at the discretion of the surgeon, and the IOP was left at 25–35 mm Hg at the end of the case at the discretion of the surgeon. Postoperative prophylactic topical antibiotics were prescribed for 1 week after surgery, and topical steroid and Non-steroidal anti-inflammatory drug drops for a minimum of 4 weeks. Pilocarpine was used at the discretion of the treating surgeon.

Outcome Measures

The primary outcome measure was HR of failure for the “primary success” criteria for superior versus inferior 180 degrees hemi-GATT. “Primary success” was defined as an intraocular pressure <18 mm Hg and one of the following: (1) IOP reduced by ≥20% from baseline on the same number of IOP-lowering medications OR (2) IOP ≤ baseline on fewer medications. Baseline IOP and glaucoma medications were defined as the values measured at the visit when the surgical intervention was decided upon. Cases were given a 3-month postoperative grace period to allow for heme and steroid washout. Subsequently, consecutive IOPs above 17 mm Hg were given 2 months for optimization of medical management within criteria limits before being classified as a failure. Major interventions including cyclophotocoagulation, surgical revisions, or reoperations constituted failure at any point (including the first 3 months), as did the loss of light perception.

Multiple secondary outcomes were also evaluated for comparison of superior versus inferior cohorts. First, HRs of failure for alternative criteria including “20% IOP reduction”, “complete success”, and “qualified success” were determined (See Table 1). The “20% reduction” criteria required a 20% reduction in IOP compared with baseline after 3 months, with the same or fewer glaucoma medications and no interventions. “Complete success” was defined as postoperative IOPs between 6 and 17 after 3 months, requiring no glaucoma medications or interventions. “Qualified success” was defined as postoperative IOPs between 6 and 17 after 3 months, allowing postoperative medication use with the same or fewer glaucoma medications than baseline, and minor interventions (laser peripheral iridotomy/laser trabeculoplasties)

TABLE 1 - Intraocular Pressure, IOP-lowering Glaucoma Medication, and Intervention Allowances for Primary and Secondary Criteria in this Study

	Intraocular pressure	Glaucoma medications	Interventions
“Primary”	6–17	<Baseline with IOP ≤baseline IOP OR = baseline with 20% IOP reduction	No
“Complete”	6–17	None	No
“Qualified”	6–17	≤Baseline	ALT/SLT/MLT/LPI
“20% IOP reduction”	20% IOP reduction	≤Baseline	No

ALT indicates Argon laser trabeculoplasty; IOP, intraocular pressure; LPI, Laser peripheral iridotomy; MLT, micropulse laser trabeculoplasty; SLT, selective laser trabeculoplasty.

Cross-sectional analyses of median IOP and glaucoma medication use at sentinel postoperative time points with statistical comparisons between cohorts and the preoperative baseline were also undertaken. Finally, postoperative procedural safety data including early and late complications, as well as interventions and reoperations were recorded and compared between study groups. IOP spikes were defined as IOPs >30 mm Hg, or >10 mm Hg above baseline IOP in the first 3 postoperative months.

For the complete cohort, risk factors for surgical failure by the “primary success” criteria were identified using Cox regression analysis of preoperative characteristics including sex, age, operative eye, glaucoma type, glaucoma severity, preoperative vision, baseline IOP, and baseline medication burden.

Statistical Analyses

Categorical variables describing the study populations were expressed as proportions of the cohort and compared between groups using Fisher exact test. Continuous variables were described as medians with interquartile ranges (IQRs) and compared through the Mann-Whitney-Wilcoxon test. Follow-up was truncated to the first follow-up after 12 months to account for differences in follow-up after 12 months between cohorts. Kaplan-Meier survival analysis was performed to display the relative success of the interventions over time. Log-Rank tests were conducted to compare survival distributions between intervention groups. Cox proportional hazard models were utilized to determine HRs associated with intervention groups (superior vs inferior 180 degrees GATT), as well as various risk factors for failure for the complete cohort including sex, age, operative eye, glaucoma type, glaucoma severity, preoperative vision, baseline IOP, and baseline medication burden. Proportional hazard assumptions for each covariate were tested using multiple methods. First, stratified Kaplan-Meier and “log minus log” curves were visually inspected for proportionality. Covariates with crossing or equivocal curves were further analyzed through the Schoenfeld residuals method. HRs for each covariate were determined through crude (univariable; in which the effect of the variable is determined in isolation) and adjusted (multivariable; in which the effect of all covariables is determined simultaneously) analyses. Posthoc power analysis was conducted with actual rates of failure and loss to follow-up, using a 2-sided type I error of 0.05. Statistical analyses were performed and figures were generated using IBM SPSS Statistics 26 and GraphPad Prism 9.0.

RESULTS

Study Population and Demographics

At the surgical centres during the study window, 304 superior or inferior 180 degrees hemi-GATT surgeries with concurrent phacoemulsification cataract excision were performed. Seven cases were excluded from analysis due to <1-month postoperative follow-up (out-of-town patients). No cases were excluded for prior glaucoma surgery. Of 247 patients, 297 eyes remained in the analysis. Baseline demographics were similar for the 2 intervention groups, apart from the operative eye (Table 2). This disparity is attributable to the correlation between surgeon handedness and ease of surgical technique from each side, with most surgeries performed by a left-handed surgeon (for whom the inferior canal is more easily cannulated on the right side, and vice versa). There were no relevant statistical differences in characteristics for left versus right eyes. Average characteristics varied only slightly between superior and inferior populations, with a nonstatistically significant, but slightly higher median IOP (16.0 vs 15.0 mm Hg) in the inferior cohort. The majority (76.1%) of patients had either primary open angle (35.7%), or primary angle closure (40.4%) glaucoma, with the remainder made up of pseudoexfoliative, pigment dispersion, combined mechanism, uveitic, and normal tension glaucoma subtypes. In Calgary, 217 of the included surgeries (73.1%) were performed with the remaining 80 (26.9%) performed in Toronto. Of the 50 patients with both eyes included in the study, 30 (60%) of these had different surgical approaches on each eye (eg, superior for left, inferior for right, or vice versa).

TABLE 2 - Baseline Characteristics of Patients Included in the Study.

	Total (N=297 eyes); n (%)	Superior 180 (N=162 eyes); n (%)	Inferior 180 (N=135 eyes); n (%)	P
Demographics
Age
Median (IQR), years	71 (65–77)	71 (65–77)	70 (65–76)	0.47
>70 y	152 (51.2)	85 (52.5)	67 (49.6)	0.64
Left eye	144 (48.5)	90 (55.6)	54 (40.0)	0.01
Female	154 (51.9)	82 (50.6)	72 (53.3)	0.73
Diabetes	55 (18.5)	30 (18.5)	25 (18.5)	1.00
Vision
logMAR Preoperative BCVA, median (IQR)	0.30 (0.18–0.48)	0.30 (0.18–0.48)	0.30 (0.18–0.40)	0.55
logMAR BCVA ≥0.4	113 (38.0)	64 (39.5)	49 (36.3)	0.63
Glaucoma
Severity
Preoperative VF MD, median (IQR); dB	−6.95 (−14.02 to −3.305)	−6.61 (−12.90 to −3.79)	−7.07 (−15.3 to −3.07)	0.87
VF MD <−12	62 (20.9)	32 (19.8)	30 (22.2)	0.67
Cup-to-disc ratio, median (IQR)	0.7 (0.6–0.8)	0.75 (0.6–0.8)	0.7 (0.59–0.8)	0.53
Type
Primary open angle	106 (35.7)	57 (35.2)	49 (36.2)	0.90
Primary angle closure	120 (40.4)	66 (40.7)	54 (40.0)	1.00
Decision status
IOP, median (IQR); mm Hg	15.0 (12.0–19.0)	15.0 (12.0–19.0)	16.0 (12.0–19.0)	0.52
IOP, mean (SD); mm Hg	16.1 (5.7)	15.9 (5.4)	16.4 (6.0)	—
IOP >21	46 (15.5)	25 (15.4)	21 (15.6)	1.00
Medications, median (IQR)	2.0 (0.5–3.0)	2.0 (1.0–3.0)	2.0 (0.0–3.0)	0.37
Medications, mean (SD)	1.87 (1.48)	1.79 (1.43)	1.96 (1.54)	—
Prior interventions
Laser peripheral iridotomy	137 (46.1)	80 (49.4)	57 (42.2)	0.24
Laser trabeculoplasty	51 (17.2)	29 (17.9)	22 (16.3)	0.88

BCVA indicates best-corrected visual acuity; IOP, Intraocular pressure; IQR, interquartile range; MD, mean deviation; VF, visual field.

Survival Outcomes and Hazard Ratios of Failure

Kaplan-Meier survival curves for the 4 success criteria (“primary success”, “complete success”, “qualified success”, and “20% IOP reduction”) are shown in Figure 1. Crude (univariate) and adjusted (multivariate) HRs of failure comparing superior 180 to inferior 180 GATTs for each criterion are shown in Table 3. For “primary success”, the estimated time to failure was 17.8 months for inferior hemi-GATTs versus 14.8 months for eyes with superior hemi-GATTs (log-Rank P value = 0.22). There was a nonstatistically significant trend toward increased failure for the superior group for the primary outcome (crude HR = 1.27 (95% CI = 0.86–1.88); adjusted HR = 1.50 (95% CI = 0.91–2.46)). Reasons for failure are shown in Table 4. Of the secondary outcomes, analysis of the “qualified success” criteria also trended towards increased failure in the superior cohort (crude HR = 1.12, 95% CI = 0.70–1.80; adjusted HR = 1.36, 95% CI = 0.73–2.54). Analyses of “complete success” were similar between groups with an estimated time to failure of 7.6 months (95% CI = 4.2–11.1) for superior and 9.3 months (95% CI = 5.4–13.2) for inferior surgeries, and HRs nearing 1 (crude HR = 1.00, 95% CI = 0.73–1.37; adjusted HR = 1.04, 95% CI = 0.67–1.62). Notably, the “20% IOP reduction” criteria showed a statistically significant higher rate of failure in all but the adjusted HR analysis (log-Rank P value = 0.04; crude HR = 1.40, 95% CI = 1.01–1.92; adjusted HR = 1.27, 95% CI = 0.84–1.93).

FIGURE 1: Kaplan-Meier survival curves for surgical success as measured by 4 criteria. Estimated time to 25% or 50% failure as appropriate with 95% confidence intervals (if able to calculate) and Log-Rank test P values are included in insets on each.

TABLE 3 - Crude (Univariate) and Adjusted (Multivariate) HRs of Failure for Superior Relative to Inferior 180 Degrees Hemi-GATT Eyes

	HR (95% CI)
Criteria	Crude	Adjusted
Primary success	1.27 (0.86–1.88)	1.50 (0.91–2.46)
Complete success	1.00 (0.73–1.37)	1.04 (0.67–1.62)
Qualified success	1.12 (0.7–1.8)	1.36 (0.73–2.54)
20% IOP reduction, same or fewer medications	1.40 (1.01–1.92)	1.27 (0.84–1.93)

GATT indicates gonioscopy-assisted transluminal trabeculotomy; HR, hazard ratio; IOP, intraocular pressure.

TABLE 4 - Reasons for Failure of “Primary Success” in Superior 180 Degrees and Inferior 180 Degrees Eyes

	Total	Superior 180	Inferior 180
Failures	n=105; n (%)	n=62; n (%)	n=43; n (%)
Reoperation	4 (3.8)	3 (4.8)	1 (2.1)
Consecutive IOPs >18 mm Hg	12 (11.4)	9 (14.5)	3 (7.0)
Fewer medications than baseline, IOP above baseline	20 (19.0)	12 (19.3)	8 (18.6)
Medication use above baseline	37 (35.2)	17 (27.4)	20 (46.5)
Medications same as baseline, <20% reduction from baseline IOP	32 (30.5)	22 (33.9)	11 (25.6)

IOP indicates intraocular pressure.

Multiple subanalyses were also performed to account for possible confounders in the superior versus inferior cohort analysis. First, given that low baseline IOPs can lead to failure by the primary success criteria despite a stable postoperative clinical course, the primary success survival analysis was repeated, excluding patients with baseline IOPs <17 on no glaucoma medications, or single-digit IOPs on any medications (n = 231, 125 superior and 106 inferior). This subanalysis showed similar results, with nonstatistically significant crude and adjusted HRs of failure of 1.42 (95% CI = 0.89–2.26) and 1.45 (95% CI = 0.78–2.68) for the superior cohort, respectively. In addition, a first-eye analysis (n = 246, 132 superior and 114 inferior) was performed to investigate the potential effect of intraperson correlation for the few patients with the same surgical approach used on both eyes (n = 10 patients, 20 eyes). Again, there was an elevated, but the nonsignificant risk of failure associated with superior surgeries (crude HR = 1.38, 95% CI = 0.91–2.10; adjusted HR = 1.54, 95% CI = 0.91–2.60).

Cross-sectional Analysis: Intraocular Pressures and Intraocular Pressure-lowering Medications

The median baseline IOP was 15.0 (12.0–19.0) and 16.0 (12.0–19.0) for eyes receiving superior and inferior surgeries, respectively. At the last follow-up, the median IOP was significantly lower for both groups with identical median IOPs of 13.0 (11.0–15.0; P values <0.001) (Fig. 2). Excluding eyes that underwent major interventions, 66.4% of eyes receiving inferior GATT had last follow-up IOPs lower than baseline, compared with 61.8% of superior GATT eyes (Fig. 3; P value = 0.61). Similarly, both groups saw a statistically significant reduction from baseline at the last follow-up for medication use. For superior eyes, median baseline medication use was 2.0 (mean, 1.79; IQR, 1.0–3.0) with a reduction to a median of 0.0 (mean, 0.96; IQR, 0.0–2.0) at the last follow-up (P value <0.0001). For inferior eyes, median medication use was 2.0 (mean, 1.96; IQR, 0.0–3.0) at baseline and was reduced to a median of 1.0 (mean, 1.24; 0.0–2.0) at the last follow-up (P value <0.0001). 68.8% of superior GATT patients had a reduced or nil (if the baseline was zero) medication burden at the last follow-up, compared with 72.3% of inferior GATT patients (P value = 0.522); however, a higher, though the nonsignificant proportion of superior GATT patients were on no medications at last follow-up compared with the inferior cohort (55.2% vs 46.2%, P value = 0.1294). A cross-sectional subanalysis was performed of patients with baseline IOPs >21 mm Hg. This cohort had a collective median baseline IOP of 24 mm Hg (n = 54, superior; n = 29, inferior; n = 25; IQR, 22–26). This was reduced to 13 mm Hg (IQR, 11–15) at the last follow-up. There was not a significant difference for superior versus inferior baseline or last follow-up pressures in this subanalysis (P values = 0.52; 0.58, respectively).

FIGURE 2: IOPs for eyes receiving superior (red) and inferior (blue) 180 degrees hemi-gonioscopy-assisted transluminal trabeculotomy (GATT) surgeries at various postoperative follow-up intervals. IOP indicates intraocular pressure.

FIGURE 3: Scatterplot of baseline intraocular pressures (IOPs) (x-axis) versus last follow-up IOPs (y-axis) for eyes receiving superior (left) and inferior (right) 180 degrees hemi-GATTs. The colour of the data point represents the number of IOP-lowering medications at the last follow-up. Overlapping data points are shown with larger symbols and coloured for the highest medication. Eyes that underwent a major intervention in the postoperative period were excluded from this analysis. One data point is outside the axis limits on the inferior graph. [45, ¹⁴

Safety Analysis: Postoperative Interventions and Complications

Sixty-two total eyes (20.9%) underwent one or more postoperative interventions, for a total of 70 postoperative procedures (Table 5). This included 32 superior eyes (19.8%) and 30 inferior eyes (22.2%). The most common procedure was laser trabeculoplasty [either selective laser trabeculoplasty (SLT) or micropulse laser trabeculoplasty]. Median time to laser trabeculoplasty for these eyes was 12.0 months (5.5–15.6). Sixty-two eyes experienced an IOP spike in the early postoperative period (26 superior, 34 inferior; P value = 0.06). 38 IOP spikes (62.3%) were successfully medically managed, whereas 10 (16.0%) spontaneously resolved with conservative observation. Nineteen total AC paracenteses were performed on 12 eyes. All except for 2 of these were performed in the first postoperative month [median time to intervention=0.1 month (0.04–0.44)]. 3 eyes (1.0%) required an additional surgical intervention (Xen gel stent) for the management of postoperative intraocular pressures.

TABLE 5 - Postoperative Interventions and Reoperations in Eyes Receiving Superior and Inferior 180 Degrees Hemi-GATT Surgeries

	Superior 180	Inferior 180
	n=162; n (%)	n=135; n (%)
Intervention
Needling	1 (0.6)	0
AC Tap	5 (2.5)	8 (5.9)
Iridoplasty	4 (2.5)	8 (6.1)
Laser peripheral iridotomy	0	2 (1.5)
Laser trabeculoplasty	15 (9.2)	8 (6.1)
Reoperation
Cyclophotocoagulation	6 (3.7)	3 (2.3)
Xen	2 (1.2)	1 (0.8)

Laser trabeculoplasty included selective laser trabeculoplasty and micropulse laser trabeculoplasty; Xen, Xen Gel Stent.

AC indicates anterior chamber; GATT, gonioscopy-assisted transluminal trabeculotomy.

Postoperative complications were present in 139 total eyes (46.8%), including 67 inferior eyes (49.6%) and 72 superior eyes (44.4%; P value = 0.41) (Table 6). The most common complication was an early minor hyphema (<2 mm), which occurred in 23.3% and 25.9% of superior and inferior eyes, respectively (P value = 0.69). The majority of remaining complications were early postoperative iritis (9.8% of superior eyes; 11.1% of inferior eyes) and corneal edema (14.7% of superior eyes; 14.1% of inferior eyes). Collectively, early postoperative hyphema, iritis, and corneal edema account for 81.0% of all superior eye complications, and 72.3% of all inferior eye complications. Serious complications included 2 instances of suprachoroidal hemorrhage (1 superior, postoperative day 336; 1 inferior; postoperative day 1) and 1 retinal detachment (inferior; postoperative day 23). Both suprachoroidal hemorrhages resolved rapidly with medical management. The retinal detachment occurred in a complex monocular patient with a history of severe intraocular inflammation, poor medication compliance, and a baseline visual acuity of hand motions secondary to a dense posterior subcapsular cataract, which improved to a zenith of 20/60 postretinal detachment repair. Two postoperative eyes lost light perception in the postoperative period (1 superior, postoperative month 20; 1 inferior, postoperative day 2). The superior eye had advanced glaucoma preoperatively, with a visual field (VF) mean deviation of −32.5 dB and a baseline visual acuity of counting fingers. Light perception was lost at postoperative month 20 and did not return. The inferior eye had a baseline visual acuity of light perception primarily due to wet age-related macular degeneration. Light perception was lost for this eye on postoperative day 2 but recovered to light perception after postoperative month 4. It should be noted that gonioscopy was not performed routinely for all follow-up appointments at one center, so peripheral anterior synechiae (PAS) may be underreported.

TABLE 6 - Early (<3 mo) and Late (>3 mo) Postoperative Complications in Eyes Receiving Superior and Inferior 180 Degrees Hemi-GATT Surgeries

	Superior 180		Inferior 180
	Early (<3 mo)	Late (≥3 mo)	Early (<3 mo)	Late (≥3 mo)
Complication	n=163; n (%)	n=152; n (%)	n=135; n (%)	n=122; n (%)
Leak/dehiscense	0	0	0	0
Shallow AC	1 (0.6)	0	0	0
PAS	0	1 (0.7)	1 (0.7)	2 (1.6)
Hyphema	38 (23.3)	4 (2.6)	35 (25.9)	4 (3.3)
Hyphema >2 mm	3 (1.8)	0	4 (3)	0
Choroidal detachment	0	0	0	0
Iritis	16 (9.8)	4 (2.6)	15 (11.1)	5 (4.1)
Corneal edema	24 (14.7)	1 (0.7)	19 (14.1)	3 (2.5)
Dellen	0	0	0	0
Epithelial defect	0	0	1 (0.7)	1 (0.8)
Vitreous hemorrhage	0	2 (1.3)	0	2 (1.6)
Macular edema	0	1 (0.7)	3 (2.2)	0
Hypotony maculopathy	0	0	0	0
Suture abscess	0	0	0	0
Ptosis	0	1 (0.7)	0	0
Monocular diplopia	0	0	0	0
Binocular diplopia	0	2 (1.3)	0	1 (0.8)
Retinal detachment	0	0	1 (0.7)	0
Angle closure	0	0	0	0
Suprachoroidal hemorrhage	0	1 (0.7)	1 (0.7)	0
Malignant glaucoma	0	0	0	0
Endophthalmitis	0	0	0	0
No light perception	0	1 (0.7)	1 (0.7)	1 (0.8)

AC indicates anterior chamber; GATT, gonioscopy-assisted transluminal trabeculotomy; PAS, peripheral anterior synechiae.

Characteristics Associated With Surgical Failure

Three baseline factors were associated with a statistically significantly higher risk of failure of “primary success” in this cohort for either superior or inferior 180 degrees hemi-GATT on both a crude (univariable) and adjusted (multivariable) basis (Fig. 4). First, primary open angle glaucoma was associated with a significant HR of failure for both crude and adjusted analyses, with HRs of 1.87 (95% CI = 1.29–2.77) and 2.23 (95% CI = 1.39–3.57) for each, respectively. Advanced disease, defined as a VF mean deviation <12 decibels, was also associated with significantly increased failure, with a crude HR of 2.01 (95% CI = 1.28–3.15) and adjusted HR of 2.05 (95% CI = 1.28–3.28). Finally, age younger than 70 years at surgery was associated with a significantly increased risk of failure, with a crude HR of 1.64 (95% CI = 1.11–2.42) and adjusted HR of 1.84 (95% CI = 1.16–2.91). Patients with a baseline IOP <16 mm Hg had a crude HR of failure of 1.70 (95% CI = 1.14–2.54), though the adjusted HR did not meet statistical significance (1.41, 95% CI = 0.84–2.37). Similar variables were associated with increased failure when “complete success” criteria were utilized in the analysis in lieu of “primary success” criteria (data not shown).

FIGURE 4: Hazard ratios of failure for “primary success” for patient baseline and demographic characteristics with 95% confidence intervals. Advanced disease was defined as visual field (VF) mean deviation <−12 dB. BCVA indicates best-corrected visual acuity; HR, hazard ratio; IOP, intraocular pressure; POAG, primary open angle glaucoma.

Covariates with statistically significant HRs of failure were further investigated by finer segmented categorical survival analysis (Fig. 5). For each covariate, one subgroup has significantly lower survival by the primary criteria that the others, which were not significantly different from one another. Patients with advanced disease (VF mean deviation <−12 dB) had significantly higher rates of failure than those with mild or moderate disease (Log-rank P values = 0.02, 0.01). Conversely, patients with baseline IOPs below 14 were more likely to fail compared with higher subgroups (Log-rank P values = 0.01, 0.004). Consistent with the Cox regression analysis, POAG surgeries failed at significantly higher rates compared with other glaucoma types (Log-rank P values = 0.002, 0.04). However, primary angle closure glaucoma (PACG) was not significantly different from other glaucoma subtypes by this analysis (P value = 0.60). Finally, the effect of age on survival seems solely attributable to the subset of patients under 65, who were more likely to fail than 65–75 or >75-year-old cohorts (Log-rank P values = 0.04, 0.14).

FIGURE 5: Categorical Kaplan-Meier survival analyses for covariates identified as statistically significant risk factors for failure by the “primary success” criteria in Cox regression analysis. Advanced disease: VF MD <−12 dB; Moderate disease: VF MD −6 to −12 dB; Mild disease: VF MD >−6 dB. IOP indicates intraocular pressure; PACG, primary angle closure glaucoma; POAG, primary open angle glaucoma.

DISCUSSION

This multicenter, retrospective cohort analysis is the first to compare the relative efficacy and safety of superior versus inferior 180 degrees hemi-GATT surgeries with concomitant phacoemulsification cataract surgery as well as risk factors for failure of these surgeries collectively. Throughout most analyses and subanalyses there was no statistically significant difference in the rates of failure between superior and inferior suture hemi-GATTs, though there is a recurring trend and one criterion favoring the inferior approach (“20% reduction”). Despite its large size, it is possible that the present study is underpowered to detect significance for the relatively small HRs of failure between superior and inferior surgeries suggested by these data. Indeed, post hoc analysis for the primary outcome revealed 18.3% power, correlating with a required sample size of 1470 at this failure incidence and cohort proportions for 80% power. Although the “20% reduction” criteria revealed a significantly increased crude risk of failure for superior relative to inferior surgeries, it should be observed that the inferior cohort had a slightly higher median IOP (ie, easier to achieve an IOP reduction).

Approximately 65% of surgeries (61.7% superior and 68.1% inferior) were “successful” by the primary criteria, that is, IOPs below 17 mm Hg with either decreased medication burden, or a 20% decrease in IOPs on the same medications. However, it is difficult to summarize success by one measure in this study, given that there are different goals for different patients. Per the surgeons, the majority of included surgeries were undertaken as adjuncts to planned cataract surgery to reduce medication burden, lower IOPs in variably compliant patients, or lower IOPs in patients on no medications who have shown progression despite well-controlled IOPs. Thus, preoperative IOPs are significantly lower than in some previously published studies where uncontrolled glaucoma was the main indication for surgery.^11,23,28 Tracking outcomes for patients with a low baseline IOP is particularly challenging, especially when the criteria is an IOP reduction. For these patients, conventional outcomes used for subconjunctival surgery, for instance, “complete” success (achieving reasonable IOP without medications) or “qualified” success (achieving reasonable IOP with medications) may be more appropriate. Despite multiple criteria, similar results were observed across the myriad analyses.

Both superior and inferior surgeries were associated with similar and statistically significant reductions in IOP and medications at the last follow-up. There is limited published data on hemi-GATT and combined GATT/phacoemulsification surgeries for comparison. Thus, comparisons must be made primarily to the standalone GATT literature. The cumulative observed reduction in pre to postoperative IOPs in this study is the smallest yet reported, with a median reduction of only 2–3 mm Hg compared with the mean reduction of ~10 mm Hg in previous reports.^23,29,30 However, the baseline IOP in this study is—to the author's knowledge—also the lowest ever reported, >4 mm Hg less than the next lowest in the literature, Sato et al.³¹ Almost all published reports to date have preoperative IOPs >21 mm Hg, and therefore this analysis represents a novel addition to the literature both in data for this specific surgical technique and indication for GATT. In the subset of patients with baseline IOPs >21 mm Hg, the observed reduction in IOPs (11 mm Hg) was similar to other reports.

There were similar rates of postoperative complications between cohorts, including approximately half of the surgical eyes. Early hyphema, transient corneal edema, and iritis accounted for most of these complications. Five patients experienced serious complications in the observational period, though only 2 of these occurred within 30 days of surgery. The frequency of hyphema in this study—affecting ~1/4 patients—was slightly lower than the pooled rate of occurrence from a recent meta-analysis of GATT literature.²⁰ However, rates of corneal edema and iritis were slightly higher in this study. This may be attributable to the concomitant phacoemulsification cataract extraction, though rates of these complications in isolated cataract surgery are low.^32,33 The intervention and reoperation rate in this analysis was minimal, with only 12 patients requiring a major reoperation in the study period. This may, however, be attributable to the preoperative glaucoma stability of this cohort.

This study is the first to the author's knowledge to identify risk factors of failure for hemi-GATT/phacoemulsification surgeries, and among the first for GATT procedures of any kind. Collectively, 3 factors were associated with a significantly increased risk of failure during both crude and adjusted analyses within this cohort, and these factors had a stronger effect than the nonstatistically significant effect of the superior versus inferior approach. Patients with primary open angle glaucoma experienced significantly higher (2.23x) rates of failure for the primary criteria than those with primary angle closure or other glaucoma subtypes (mostly secondary open angle glaucomas). This does have some precedent in the literature, as in the 24-month retrospective analysis by Grover et al,²³ POAG patients had a smaller reduction in IOPs compared with other glaucoma types (9.2 vs 14.1 mm Hg), and higher rates of failure by their criteria—especially in patients with advanced disease characterized by high negative VF mean deviations. Accordingly, this study also independently identified advanced disease (VF mean deviation <−12 dB) as a significant risk factor for failure (2.05x). The increased failure for both advanced disease and POAG may be associated with increased occlusion downstream of the operative TM—in collector channels and the intrascleral plexus.^23,34 However, this interpretation is at least partially confounded by the disproportionate percentage of patients with PACG in this cohort (~40%) and the known IOP reduction associated with lens extraction.³⁵ Recent evidence does, however, support GATT in patients with PACG in previously pseudophakic patients and as an adjunct to phacoemulsification.³⁶ The other significant risk factor associated with surgical failure was age below 70 years. This has also been previously reported in studies analyzing alternative techniques including trabectome ab interno trabeculotomy with concomitant phacoemulsification³⁷ and trabeculectomy.³⁸ This finding is counter-intuitive, as it has been suggested that downstream collector channels are more likely to be salvageable in relatively younger patients with less accumulated age-related sclerosis.³⁹ In the present study, this finding may be attributable to the simultaneous phacoemulsification, which has been shown to have a significantly greater reduction in IOPs in older patients.^40–42 It should be noted that analyses of the effect of age on surgical success in the present study are limited by the relatively narrow distribution of age within this patient cohort (median age = 71; IQR, 65–77), and dearth of juvenile/young adult patients (only 9 patients <50 y (3.03%); 7/9 uveitic glaucoma). Accordingly, these findings should not be extrapolated to the entire age distribution of glaucoma patients. For all 3 factors identified (POAG, advanced disease, and younger age), it is also possible that more aggressive treatment by practitioners contributed to higher rates of failure, that is, by maintaining lower IOP targets and/or more cautiously weaning medications after surgery.

A major limitation of this study was a lack of systematic documentation of pre and postoperative angle characteristics by gonioscopy. The suggested supplementary IOP-lowering benefit of GATT in addition to lens extraction may be proportional to the extent and location of preoperative PAS in PACG and uveitic patients, and thus may be similar in effect to concurrent goniosynechialysis.⁴³ Further, there may be a disproportionate formation of postoperative PAS or reopposition of the trabeculotomy “shelf” following different hemi-GATT approaches.⁴⁴ Subsequent investigation for these pre and postoperative gonioscopic findings may reveal new covariates of failure, which were not borne out in the present work.

In conclusion, these results add to the existing literature supporting GATT as an effective, safe MIGS for the management of glaucoma, and suggest that hemi-GATT/phacoemulsification surgery may serve as an intermediary adjunct surgery in the stepwise management of appropriately chosen patients. Given the identified risk factors for failure, clinicians may consider more guarded expectations for patients at relatively younger ages, with POAG or advanced disease. These data suggest no significant difference in outcomes, complications, or postoperative interventions comparing superior versus inferior hemi-GATT approach, so surgeons may consider trialing both approaches and determining personal comfort and success with each to determine individual practice patterns pending further data. Future studies with increased power will be invaluable to independently validate these findings.