Ab Interno Minimally Invasive Glaucoma Surgery Effectiveness in Black Patients: An IRIS Registry Study

Abstract

Précis:

MIGS in combination with cataract surgery resulted in clinically significant reductions in IOP and IOP-lowering medications up to 24 months in Black patients.

Purpose:

Describe outcomes following ab interno Minimally Invasive Glaucoma Surgery (MIGS) procedures US FDA cleared or approved for intraocular pressure (IOP) reduction in primary open angle glaucoma combined with cataract surgery and cataract surgery alone in Black patients.

Methods:

An observational, retrospective study of glaucoma in Black patients treated with MIGS (Hydrus®, iStent Inject®, OMNI® Surgical System) with cataract surgery or cataract surgery alone, in the American Academy of Ophthalmology IRIS® Registry (Intelligent Research in Sight). Deidentified data included glaucoma diagnosis, procedure data, IOP, and medication use from preoperatively through 24 months. The study period was 07/01/2016 and 12/31/2023. Eligible patients identified via Current Procedural Terminology codes coupled with electronic health records. Outcomes were changes in IOP, and medication class usage (months 6, 12, 18, and 24) stratified by baseline IOP (≤18; >18 mm Hg).

Results:

In 12,828 eyes of 12,828 patients, including 189 Hydrus, 491 iStent, 91 OMNI, and 12,057 cataract surgeries, mean IOP and medication use decreased significantly in each cohort. Mean IOP reductions were in the range of 1.2–2.7 mm Hg, and medication reductions were 0.4–1.0 medications across all cohorts and time points. In MIGS cohorts, IOP reductions were greater with higher baseline IOP, and medication reductions were greater with lower baseline IOP.

Conclusions:

Each MIGS procedure produced clinically and statistically significant reductions in both IOP and IOP-lowering medications up to 2 years postoperatively in Black patients. Surgeons could consider offering MIGS at the time of cataract surgery to this population.

Glaucoma is the leading cause of irreversible vision loss and blindness among Black patients in the United States.¹ Black patients are affected with glaucoma at higher rates than other racial/ethnic groups.² Further, at the time of diagnosis, glaucoma tends to be more advanced in Black patients than in other racial/ethnic groups.³ While we concur with the position of the AMA Style manual that race and ethnicity are “cultural constructs with biological contribution through genetic heritage, but without well-defined or clear scientific meaning,”⁴ it is clear that the impact of glaucoma is amplified in the Black population.

Outcomes of trabeculectomy have been frequently reported to be poorer in Black patients compared with White patients. In the Advanced Glaucoma Intervention Study, the 10-year cumulative incidence of trabeculectomy failure (as the first intervention) was 31.9% in Black patients versus 17.7% in White patients.⁵ In the Primary Tube Versus Trabeculectomy study, the 5-year surgical failure rate was 48% in Black patients versus 33% in White patients; this difference was not significant, but the study was not powered for this analysis.⁶

The development and proliferation of minimally invasive glaucoma surgeries (MIGS) has reinvented the glaucoma surgical landscape. Very little is known regarding the effect of race on MIGS outcomes. In Black patients, one study demonstrated that combined phacoemulsification and iStent implantation has been shown to reduce both intraocular pressure (IOP) and the need for IOP-lowering medications,⁷ but in another study, these reductions were no greater than with phacoemulsification alone.⁸ Given that Black patients with glaucoma are more likely to undergo concurrent MIGS procedures than White patients,⁹ there remains a significant unmet need for more data on MIGS outcomes in Black patients.

The Hydrus microstent (Alcon, Fort Worth, TX) is a nitinol stent placed in Schlemm’s canal designed to bypass the trabecular meshwork and to scaffold three clock hours of the canal. The iStent inject (Glaukos, San Clemente, CA) consists of 2 titanium microstents placed in the trabecular meshwork and at least two clock-hours apart and penetrating the trabecular meshwork to provide a bypass from the anterior chamber into Schlemm’s canal. The OMNI Surgical System (Sight Sciences, Menlo Park, CA) is used to perform ab interno canaloplasty and trabeculotomy. Multiple prior studies have demonstrated significant reductions in both IOP and IOP-lowering medication use following use of these devices combined with cataract surgery.^10–13

We conducted a database study of outcomes following phacoemulsification alone or combined with iStent, Hydrus, or OMNI glaucoma surgery in Black patients in the IRIS Registry to better understand the IOP and medication reduction outcomes for Black patients undergoing MIGS.

METHODS

Study Design

This was a retrospective, observational study investigating glaucomatous disease in Black patients that underwent FDA-approved/cleared ab interno MIGS combined with cataract surgery or cataract surgery alone. MIGS devices were selected for inclusion based on their FDA-approved/cleared indication. This was an analysis of existing and deidentified electronic health record (EHR)-derived data drawn from the IRIS Registry and linked with Komodo Health claims. The IRIS Registry is the largest comprehensive eye disease clinical database in the world and is comprised of real-world data extracted from the electronic medical records of 434 million encounters between 72.7 million unique patients and over 13,600 eye care clinicians between January 1, 2013 and October 1, 2023. This study was approved by the Western Institutional Review Board-Copernicus Group (WCG® IRB) and complied with the tenets of the Declaration of Helsinki. Patient consent was not required for this study due to the deidentified nature of the data extracted from the IRIS Registry. The database search and resulting data analysis were provided by Verana Health, the Academy’s designated analytics vendor.

Four cohorts of patients were evaluated: those undergoing Hydrus, iStent inject, or OMNI glaucoma surgery, each combined with phacoemulsification cataract surgery, and those undergoing cataract surgery alone. Each of these MIGS devices carries a US Food and Drug Administration (FDA)-approved or -cleared indication for reduction of IOP in patients with POAG.^14–16 The index period was from July 1, 2016 through December 31, 2020. The index date was the date of the earliest documented cataract surgery (with or without a MIGS device identified by additional CPT and/or HCPCS codes) in a patient eye within the IRIS Registry during the index period. The preindex look-back period was 12 months, and the postindex follow-up period was a minimum of 6 months of potential follow-up after the index date and a maximum of 24 months.

Patients

The eligibility criteria for this analysis were described previously.¹⁷ Eyes were included that received cataract surgery identified by presence of Current Procedural Terminology (CPT) code for cataract surgery (66982 or 66984) within the index period; were diagnosed with glaucoma [based on a qualifying International Classification of Disease (ICD) -9 or 10 code] in the same eye that received cataract surgery; had a documented IOP within 6 months before and including the index date (defined as baseline for the purpose of this study); and had at least one pharmacy claim for an IOP-lowering glaucoma medication preindex and at least one pharmacy claim for any medication postindex. The 4 cohorts were defined as follows: Hydrus included CPT code 0191T confirmed to be Hydrus and not any other MIGS procedure based on keywords/text notes; iStent inject included CPT code 0191T, similarly confirmed; OMNI included CPT code 66174, similarly confirmed; and cataract surgery included no additional CPT codes for any combined glaucoma procedure. Exclusion criteria included unknown laterality of surgery or missing demographic data, absence of a glaucoma diagnosis in the study eye, any prior incisional glaucoma surgery (such as trabeculectomy, tube-shunt, or MIGS procedure), or selective laser trabeculoplasty in the study eye within 90 days of the index date. For patients with 2 qualifying eyes within the index period, only the first-operated eye was included. As stated, identification of a single specific MIGS device was required and was based on mention of a single specific device within the electronic health record (EHR) between 7 days before or up to 14 days after the index date. Any terms in the HER related to a goniotomy (off-label) or more than one concomitant MIGS procedure excluded patients from the study. In addition, a custom set of regular-expression logic was used to prevent inclusion of false positives due to devices with names similar to medical terms using Python or SQL string searches for each MIGS group.

Study Devices

Three FDA-approved/cleared ab interno MIGS procedures were included: iStent Inject, OMNI Surgical System, and Hydrus. While there are other MIGS devices and procedures, the included devices have labeled indications for IOP reduction in patients with POAG; most others do not. The iStent Inject (Glaukos Corp., Laguna Hills, CA) is a micro-bypass system consisting of 2 stents, designed to bypass the trabecular meshwork (TM) and create a direct connection between the anterior chamber and Schlemm’s canal.¹⁸ The iStent Inject was FDA approved in 2018. The OMNI Surgical System (Sight Sciences, Menlo Park, CA) is an implant-free procedure using a microcatheter to dispense viscoelastic and perform canaloplasty followed by a trabeculotomy. It can be used as a standalone procedure or in combination with cataract surgery.¹⁹ OMNI first received FDA clearance in 2017. Hydrus Microstent (Alcon, Fort Worth, TX) is an aqueous drainage device designed to bypass the TM and scaffold Schlemm’s canal over 3 clock hours.²⁰ Hydrus was approved by the FDA in 2018. OMNI, iStent Inject, and Hydrus are each FDA approved/cleared for use in combination with cataract surgery.

Data Collection

Data collected included baseline demographic information (patient level), and eye-level baseline clinical characteristics; glaucoma diagnosis and severity (based on ICD-9/10 codes), index cataract procedure [routine (CPT 66984) vs. complex (CPT 66982)], visual acuity (VA), IOP, and IOP-lowering medication use. Glaucoma diagnoses were not mutually exclusive; an eye diagnosed with multiple forms of glaucoma was included in each corresponding glaucoma category; however, a POAG diagnosis was required before the index date. If an eye had multiple reported levels of glaucoma severity during the baseline period, the most severe recorded level was assigned.

IOP outcomes were collected at the eye level at 6, 12, 18, and 24 months postoperatively. Windows around study time points were +60 days at month 6 and +90 days at all subsequent time points. The most recent IOP measurement before surgery was considered as the baseline. Where there were multiple IOP measurements identified within a follow-up period, only the most recent IOP measurement during that follow-up period was considered. An average IOP (at baseline or follow-up) was used where more than one IOP was recorded per day, per eye.

Medication data were collected at the patient level from the linked claims dataset 12 months before and including the index date and for the same follow-up time points as for IOP (6, 12, 18, and 24 mo). Medication use was calculated as the number of different glaucoma medication classes being used at each time point. The claims database was comprised of closed sources (payer complete) and open sources (nonpayer complete) across commercial, governmental, Medicaid, Medicare, Medicare Advantage, and military payers, as well as those with other or no insurance. At least one claim for every patient was required during the follow-up period to minimize the risk of unobserved open-source claims. All medication outcomes were reported at the patient level; laterality data were not available within the claims dataset.

Outcomes and Statistical Analysis

For this analysis, patients whose race was identified as Black in the electronic health record were included. Key outcomes included mean IOP and mean medication use at each study time point for each of the 4 cohorts. Descriptive statistics were used to present demographic and baseline characteristics, IOP, and medication outcomes. Mean and SD are presented for continuous variables and percentages for categorical variables. Changes from baseline in both IOP and medication use were analyzed within each cohort using paired t tests. Continuous outcome measures were assessed for normality using Shapiro-Wilks tests. Subgroup analysis was performed based on baseline IOP (≤18 mm Hg vs. >18 mm Hg). The level of significance was taken to be 0.05. No formal sample size analysis was conducted; all qualifying eyes in the IRIS Registry were included. All statistical analyses were performed using PySpark and Google Sheets.

RESULTS

Data were drawn from a total of 12,828 eyes of 12,828 patients, including 189 in the Hydrus cohort, 491 in the iStent cohort, 91 in the OMNI cohort, and 12,057 in the cataract surgery only cohort. While we do not know precisely the proportion of Black patients in the overall POAG population undergoing MIGS + cataract surgery or cataract surgery alone this represents, a similar study with identical eligibility criteria and the same included devices had 13.8% Black patients.¹⁷ There are large differences in the sample sizes for each of the cohorts; this should be kept in mind when considering the outcomes results that follow. Demographic data and clinical glaucoma characteristics for each cohort are given in Table 1. The mean age in each cohort ranged from 70.1 to 71.0 years, and most (57.1%–62.2%) were female. Glaucoma severity was most severe in the OMNI cohort and least severe in the iStent inject cohort. Shapiro-Wilks tests confirmed the assumptions of normal distribution of the data.

TABLE 1.

Demographic Data for Each Study Cohort

	Hydrus	iStent inject	OMNI	Cataract surgery
N	189	491	91	12057
Age (y, mean) (SD)	70.1 (8.6)	70.2 (7.9)	70.9 (9.8)	71.0 (8.8)
Gender, n (%)
Male	113 (59.8)	300 (61.1)	52 (57.1)	7499 (62.2)
Female	76 (40.2)	191 (38.9)	39 (42.9)	4558 (37.8)
Glaucoma type, n (%)*
Primary open angle	187 (98.9)	489 (99.6)	91 (100)	11836 (98.2)
Secondary open angle	3 (1.6)	5 (1.0)	1 (1.1)	154 (1.3)
Primary angle closure	20 (10.6)	21 (4.3)	6 (6.6)	655 (5.4)
Secondary angle closure	0	0	0	2 (0.0)
Glaucoma suspect	6 (3.1)	33 (6.5)	3 (3.3)	731 (6.1)
Other	1 (0.5)	5 (1.0)	0	162 (1.3)
Glaucoma severity, n (%)
Mild	58 (30.7)	228 (46.4)	19 (20.9)	3571 (29.6)
Moderate	100 (52.9)	212 (43.2)	34 (37.4)	4161 (34.5)
Severe	27 (14.3)	40 (8.1)	35 (38.5)	3127 (25.9)
Unknown	4 (2.1)	11 (2.2)	3 (3.3)	1198 (9.9)
Index cataract procedure, n (%)
Routine (CPT 66984)	180 (95.2)	453 (92.3)	84 (92.3)	10610 (88.0)
Complex (CPT 66982)	9 (4.8)	38 (7.7)	7 (7.7)	1447 (12.0)

All subjects were Black/African American.

^*Totals may exceed 100% because some eyes carried multiple glaucoma diagnoses.

Mean IOP and mean change in IOP from baseline for each cohort are given in Table 2 and Figure 1. Mean baseline IOP (SD) was lowest in the OMNI cohort [15.8 (4.3) mm Hg] and highest in the Hydrus cohort [17.1 (6.4) mm Hg]. Mean IOP was decreased significantly from baseline in each cohort at each postoperative time point. Mean IOP reductions were in the range of 1.2 to 2.7 mm Hg across all cohorts and time points.

TABLE 2.

Mean IOP and Mean Change From Baseline at Each Time Point in Each Cohort

	Hydrus			iStent inject			OMNI			Cataract surgery
	N	Mean (SD)	P *	N	Mean (SD)	P *	N	Mean (SD)	P *	N	Mean (SD)	P *
N (total eyes)	189			491			91			12057
Mean IOP (mm Hg)
Baseline	189	17.1 (6.4)		491	16.6 (4.7)		91	15.8 (4.3)		12057	17.1 (5.5)
6 mo	111	15.0 (4.3)	0.000	319	14.9 (4.2)	0.000	66	13.6 (3.4)	0.000	7299	15.4 (5.0)	0.000
12 mo	129	15.0 (4.0)	0.000	330	15.0 (4.2)	0.000	59	13.7 (3.9)	0.021	7081	15.3 (4.7)	0.000
18 mo	112	15.4 (4.8)	0.003	303	15.2 (4.3)	0.000	53	14.0 (2.9)	0.002	6321	15.5 (4.7)	0.000
24 mo	102	15.1 (4.0)	0.000	286	15.1 (4.1)	0.000	52	13.9 (3.9)	0.011	5793	15.4 (4.8)	0.000
Mean change in IOP compared with baseline (mm Hg)
6 mo	111	−2.6 (6.5)		319	−1.7 (4.7)		66	−2.3 (4.8)		7299	−1.7 (6.2)
12 mo	129	−2.3 (6.2)		330	−1.7 (4.9)		59	−1.7 (5.4)		7081	−1.7 (6.0)
18 mo	112	−2.1 (7.4)		303	−1.5 (4.7)		53	−2.1 (4.5)		6321	−1.5 (6.0)
24 mo	102	−2.7 (6.8)		286	−1.2 (5.2)		52	−2.0 (5.4)		5793	−1.6 (6.2)

^* P-Value for mean IOP reflects 1-sample t test within given cohort, comparing follow-up period value versus baseline value.

FIGURE 1.

Mean IOP reduction in each cohort at each time point. All P≤0.021 versus baseline.

The proportions of eyes in each cohort achieving a >20% IOP reduction from baseline at each postoperative time point are given in Table 3. At all time points, the proportions were arithmetically higher in each of the MIGS cohorts compared with cataract surgery alone. At 24 months, arithmetically more eyes in the OMNI cohort (24.2%) than in any of the other cohorts (15.3%–19.1%) manifested an IOP reduction from baseline of 20% or more.

TABLE 3.

Proportions of Eyes in Each Cohort Achieving ≥20% IOP Reduction From Baseline at Each Postoperative Time Point

	N (%)
	Hydrus (N=491)	iStent inject (N=189)	OMNI (N=91)	Cataract surgery (N=12057)
6 mo	108 (22.0)	38 (20.1)	25 (27.5)	2353 (19.5)
12 mo	105 (21.4)	42 (22.2)	17 (18.7)	2213 (18.6)
18 mo	90 (18.3)	38 (20.1)	16 (17.6)	1979 (16.4)
24 mo	93 (18.9)	36 (19.1)	22 (24.2)	1846 (15.3)

Mean IOP and mean IOP change from baseline data by baseline IOP (≤18 mm Hg vs. >18 mm Hg) are given for each cohort in Table 4 and Figure 2. In the group with a lower baseline IOP (<18 mm Hg), the postoperative IOP at all time points was not significantly different from baseline for any of the MIGS cohorts and was slightly higher at all time points in the cataract-only group (by 0.4–0.6 mm Hg, P<0.001). In contrast, IOP reductions in eyes with a higher baseline IOP (>18 mm Hg) were larger and significant at all time points for all cohorts, including cataract alone.

TABLE 4.

Mean IOP and Mean Change From Baseline in High and Low Baseline IOP Cohorts

Mean IOP and mean change from baseline at each time point in each cohort among eyes with baseline IOP <18 mm Hg
	Hydrus			iStent inject			OMNI			Cataract surgery
	N	Mean (SD)	P *	N	Mean (SD)	P *	N	Mean (SD)	P *	N	Mean (SD)	P *
N (total eyes)	112			289			61			7166
Mean IOP (mm Hg)
Baseline	112	13.3 (2.8)		289	13.6 (2.4)		61	13.4 (2.3)		7166	13.7 (2.6)
6 mo	65	13.7 (3.8)	0.918	190	13.7 (3.6)	0.767	43	13.1 (3.3)	0.769	4304	14.2 (4.3)	0.000
12 mo	76	13.7 (3.0)	0.526	195	13.7 (3.4)	0.916	42	13.6 (4.0)	0.763	4198	14.2 (4.0)	0.000
18 mo	64	14.3 (5.2)	0.144	177	13.7 (3.2)	0.685	35	13.6 (2.5)	0.952	3766	14.4 (4.0)	0.000
24 mo	56	14.0 (3.4)	0.172	162	14.2 (3.7)	0.124	35	13.3 (4.2)	0.946	3460	14.4 (4.2)	0.000
Mean Change in IOP compared with baseline (mm Hg)
6 mo	65	0.1 (3.6)		190	0.1 (3.6)		43	−0.1 (3.1)		4304	0.4 (4.4)
12 mo	76	0.2 (3.0)		195	0.0 (3.4)		42	0.2 (4.1)		4198	0.5 (4.2)
18 mo	64	1.1 (5.7)		177	−0.1 (3.3)		35	0.0 (2.8)		3766	0.6 (4.2)
24 mo	56	0.7 (4.0)		162	0.5 (3.8)		35	−0.1 (5.0)		3460	0.6 (4.4)
Mean IOP and mean change from baseline at each time point in each cohort among eyes with baseline IOP >18 mm Hg
N (total eyes)	77			202			30			4891
Mean IOP (mm Hg)
Baseline	77	22.5 (6.2)		202	21.0 (3.8)		30	20.8 (2.7)		4891	22.0 (5.1)
6 mo	46	16.9 (4.2)	0.000	129	16.7 (4.4)	0.000	23	14.4 (3.3)	0.000	2995	17.1 (5.5)	0.000
12 mo	53	16.8 (4.6)	0.000	135	16.8 (4.5)	0.000	17	13.9 (3.7)	0.000	2883	16.9 (5.2)	0.000
18 mo	48	16.8 (3.8)	0.000	126	17.4 (4.8)	0.000	18	14.7 (3.4)	0.000	2555	17.1 (5.2)	0.000
24 mo	46	16.4 (4.3)	0.000	124	16.4 (4.3)	0.000	17	15.2 (3.0)	0.000	2333	16.8 (5.2)	0.000
Mean change in IOP compared with baseline (mm Hg)
6 mo	23	−6.3 (4.8)		129	−4.3 (5.0)		23	−6.3 (4.8)		2995	−4.8 (7.0)
12 mo	17	−6.3 (5.7)		135	−4.2 (5.6)		17	−6.3 (5.7)		2883	−4.9 (6.7)
18 mo	18	−6.0 (4.8)		126	−3.4 (5.6)		18	−6.0 (4.8)		2555	−4.7 (6.9)
24 mo	17	−6.0 (4.0)		124	−4.8 (5.2)		17	−6.0 (4.0)		2333	−5.0 (6.9)

^* P-value for mean IOP reflects 1-sample t test within given cohort, comparing follow-up period value versus baseline value.

FIGURE 2.

Mean IOP reduction in each cohort at each time point in eyes with baseline IOP ≤18 mm Hg (A) and in eyes with baseline IOP >18 mm Hg (B).

Mean medication use (by number of classes) and mean change in medication use from baseline for each cohort are given in Table 5 and Figure 3. Mean baseline medication use (SD) ranged from a low of 1.6 (0.8) medication classes in the iStent inject cohort to a high of 2.0 (1.0) in the OMNI cohort. Mean medication use within each cohort decreased significantly from baseline at each postoperative time point. Mean IOP-lowering medication use and mean change in medication use data by baseline IOP are presented in Table 6 and Figure 4. In general, medication reductions tended to be similar regardless of baseline IOP in all cohorts.

TABLE 5.

Mean IOP-lowering Medication Use and Mean Change From Baseline at Each Time Point in Each Cohort

	Hydrus			iStent inject			OMNI			Cataract surgery
	N	Mean (SD)	P *	N	Mean (SD)	P *	N	Mean (SD)	P *	N	Mean (SD)	P *
N (total eyes)	189			491			91			12057
Mean number of IOP-lowering medication classes
Baseline	189	1.9 (0.9)		491	1.6 (0.8)		91	2.0 (1.0)		12057	1.8 (0.9)
6 mo	155	1.1 (1.1)	0.000	355	1.1 (1.0)	0.000	66	1.4 (1.3)	0.000	9516	1.4 (1.2)	0.000
12 mo	144	1.0 (1.1)	0.000	351	1.1 (1.0)	0.000	60	1.4 (1.4)	0.001	9167	1.4 (1.2)	0.000
18 mo	139	1.0 (1.1)	0.000	336	1.0 (1.0)	0.000	58	1.4 (1.4)	0.001	8691	1.3 (1.2)	0.000
24 mo	134	0.9 (1.1)	0.000	330	1.0 (1.0)	0.000	61	1.2 (1.3)	0.000	8294	1.2 (1.2)	0.000
Mean change in number of IOP-lowering medication classes compared with baseline
6 mo	155	−0.9 (1.2)		355	−0.6 (1.0)		66	−0.6 (1.4)		9,516	−0.4 (1.1)
12 mo	144	−0.9 (1.2)		351	−0.6 (1.1)		60	−0.6 (1.4)		9,167	−0.4 (1.2)
18 mo	139	−0.9 (1.1)		336	−0.7 (1.1)		58	−0.6 (1.4)		8,691	−0.5 (1.2)
24 mo	134	−1.0 (1.2)		330	−0.7 (1.1)		61	−0.9 (1.4)		8,294	−0.6 (1.2)

^* P-Value for mean number of IOP-lowering medication classes reflects 1-sample t test within given cohort, comparing follow-up period value versus baseline value.

FIGURE 3.

Mean medication reduction in each cohort at each time point. All P≤0.001 versus baseline.

TABLE 6.

Mean IOP-Lowering Medication Use and Mean Change From Baseline in High and Low Baseline IOP Cohorts

Mean IOP-lowering medication use and mean change from baseline at each time point in each cohort in patients with baseline IOP <18 mm Hg
	Hydrus			iStent inject			OMNI			Cataract surgery
	N	Mean (SD)	P *	N	Mean (SD)	P *	N	Mean (SD)	P *	N	Mean (SD)	P *
N (Total eyes)	112			289			61			7166
Mean number of IOP-lowering medication classes
Baseline	112	1.9 (0.9)		289	1.6 (0.9)		61	2.1 (1.0)		7166	1.8 (0.9)
6 mo	92	1.1 (1.2)	0.000	216	1.1 (1.0)	0.000	45	1.5 (1.5)	0.002	5688	1.4 (1.1)	0.000
12 mo	86	1.0 (1.1)	0.000	207	1.1 (1.0)	0.000	39	1.4 (1.5)	0.002	5446	1.4 (1.2)	0.000
18 mo	82	0.9 (1.1)	0.000	201	1.1 (1.1)	0.000	40	1.3 (1.4)	0.000	5171	1.3 (1.2)	0.000
24 mo	79	0.8 (1.0)	0.000	196	1.0 (1.0)	0.000	44	1.1 (1.3)	0.000	4926	1.2 (1.2)	0.000
Mean change in number of IOP-lowering medication classes compared with baseline
6 mo	92	−0.8 (1.1)		216	−0.6 (1.0)		45	−0.7 (1.5)		5688	−0.4 (1.1)
12 mo	86	−1.0 (1.2)		207	−0.6 (1.1)		39	−0.7 (1.4)		5446	−0.4 (1.1)
18 mo	82	−1.1 (1.2)		201	−0.7 (1.1)		40	−0.8 (1.4)		5171	−0.5 (1.1)
24 mo	79	−1.2 (1.2)		196	−0.8 (1.1)		44	−1.1 (1.4)		4926	−0.6 (1.2)
Mean IOP-lowering medication use and mean change from baseline at each time point in each cohort in patients with baseline IOP >18 mm Hg
N (total eyes)	77			202			30				4891
Mean number of IOP-lowering medication classes
Baseline	77	2.0 (1.0)		202	1.6 (0.8)		30	1.7 (0.8)		4891	1.8 (0.9)
6 mo	63	1.0 (1.1)	0.000	139	1.1 (1.1)	0.000	21	1.3 (1.0)	0.086	3828	1.5 (1.2)	0.000
12 mon	58	1.1 (1.1)	0.000	144	1.1 (1.1)	0.000	21	1.3 (1.2)	0.154	3721	1.5 (1.2)	0.000
18 mo	57	1.1 (1.0)	0.000	135	1.0 (1.0)	0.000	18	1.7 (1.3)	0.717	3520	1.3 (1.2)	0.000
24 mo	55	1.1 (1.2)	0.000	134	1.0 (1.0)	0.000	17	1.4 (1.2)	0.275	3368	1.3 (1.2)	0.000
Mean change in number of IOP-lowering medication classes compared with baseline
6 mo	63	−1.0 (1.2)		139	−0.5 (1.0)		21	−0.5 (1.2)		3828	−0.4 (1.2)
12 mo	58	−0.9 (1.2)		144	−0.5 (1.1)		21	−0.4 (1.3)		3721	−0.4 (1.2)
18 mo	57	−0.8 (0.9)		135	−0.7 (1.0)		18	−0.1 (1.3)		3520	−0.5 (1.2)
24 mo	55	−0.8 (1.2)		134	−0.7 (1.1)		17	−0.4 (1.5)		3368	−0.6 (1.2)

^* P-Value for mean number of IOP-lowering medication classes reflects 1-sample t test within given cohort, comparing follow-up period value versus baseline value.

FIGURE 4.

Mean IOP-lowering medication reduction in each cohort at each time point in eyes with baseline IOP ≤18 mm Hg (A) and in eyes with baseline IOP >18 mm Hg (B).

DISCUSSION

This analysis of real-world data from the Academy’s IRIS Registry demonstrates that MIGS combined with cataract surgery significantly lowers both IOP and the use of IOP-lowering medications through up to 24 months postoperatively in Black patients with glaucoma. While IOP and medication reductions were seen for all 3 MIGS as well as for cataract surgery alone through 24 months postsurgically, the magnitude of these reductions was generally greatest for Hydrus and OMNI.

Glaucoma is a chronic disease, and long-term data supporting the effectiveness of therapy is essential in developing lifelong treatment strategies to delay or prevent disease progression and vision loss. This is particularly important in Black patients, who are more likely to have glaucoma,^2,21 to have more advanced glaucoma,^3,21 and to experience glaucoma progression over time^22,23 than patients of other races and ethnicities. Little is known regarding the long-term outcomes of MIGS procedures in Black patients. Previously published studies have limitations, including short follow-up, small samples from single centers, and/or single surgeons or surgeons in training, and may preclude generalizability.^24–28 This study addresses an unmet need by providing longer-term outcomes data (up to 2 y) in Black patients undergoing 4 different surgical procedures known to decrease IOP and the medication burden.

IOP reductions in all 4 cohorts were generally small, with average reductions of 1.2–2.7 mm Hg across cohorts and time points. Many eyes in this dataset likely underwent combined cataract and glaucoma surgery with the primary goal of medication reduction. Evidence for this assumption is the relatively low mean baseline IOP across the 4 cohorts (15.8–17.1 mm Hg), which was lowest in the OMNI cohort. In eyes with lower baseline IOP (≤18 mm Hg), the mean change in IOP was ±1.1 mm Hg across all cohorts and time points, with no cohort differences. In eyes with higher baseline IOP (>18 mm Hg), mean IOP reductions were generally in the range of 4–6 mm Hg and were similar across cohorts albeit numerically lowest for iStent and cataract surgery only (ranging from 3.4 to 5.0 mm Hg) compared with Hydrus and OMNI (ranging from 6.0 to 6.3 mm Hg). In summary, in Black patients undergoing surgery primarily for medication reduction, these 4 procedures produced similar IOP reductions.

In contrast, there were clinically and statistically significant differences in medication outcomes in this study, as expected given the lower baseline IOP and possible goal of medication reduction in most eyes. In the full dataset, medication use was reduced by a statistically significant amount in each of the 4 cohorts at every postoperative time point relative to their respective baseline. These reductions generally ranged from 0.5 to 1.0 medication and increased slightly over time, possibly as a manifestation of subsequent second-eye surgery given that medication use was a patient-level variable. This is a known challenge when analyzing patient-level claims data, particularly in specialties like ophthalmology, where paired organs require data granularity to the level of laterality.²⁹ Overall, at each time point, medication reductions were numerically greatest in the OMNI and Hydrus cohorts. To summarize the medication data in this study, medication reductions in Black patients undergoing these procedures were generally similar for each procedure through the first 2 years, but OMNI and Hydrus tended to yield modestly greater medication reductions.

The results of the present study are overall consistent with those from our recently published study of MIGS with cataract surgery, which included the entire patient population within the IRIS Registry.¹⁷ In that study, we also found that MIGS with cataract surgery provided greater IOP and medication reductions than cataract surgery alone, and of the included MIGS, OMNI generally showed the greatest reductions in each at the 24-month endpoint.

The decision to perform a MIGS procedure at the time of cataract surgery is for reduction of IOP, or the IOP-lowering medication burden, or both.³⁰ In eyes with medically well-controlled IOP, the goal is often to reduce the medication burden. There are several important reasons why reducing the medication burden is important. The most important of these is that many patients are poorly adherent to topical glaucoma medications,^31–33 and this nonadherence significantly increases the risk of disease progression over time.^31,32,34,35 Adherence worsens with increasing regimen complexity,^36,37 and simplification by decreasing the number of medications used per day would be expected to improve adherence and decrease the risk of progression. Topical medical therapy is also associated with numerous side effects, the most prevalent of which is ocular surface disease, which arises in 30%–70% of patients using chronic medical therapy.^38–41 The symptoms of ocular surface disease can lead to nonadherence to avoid these symptoms, setting up a cycle of nonadherence, IOP elevation, more medications, more symptoms, more nonadherence, and eventual disease progression.⁴² These issues are of great importance in Black patients, in whom nonadherence to glaucoma medications is generally more prevalent than in other racial and ethnic groups.^43,44 Even with perfect adherence, medical therapy has peak and trough effects such that IOP control is not uniform throughout the 24-hour cycle. Poor adherence exacerbates this problem with extended periods of little or no IOP control. MIGS has been shown to dampen diurnal IOP fluctuations and decrease the rate of disease progression.^45,46

As mentioned before, understanding treatment options, including surgical, and outcomes in the Black population is of paramount importance to address glaucomatous disease. In addition, the study and its findings are strengthened by the large sample size afforded through use of the IRIS Registry. Other benefits of drawing data from the IRIS Registry are the inclusion of many surgeons across all regions of the United States, improving the generalizability of the study’s findings. However, database studies have several important inherent limitations as well. Patients represented in the IRIS Registry may not be entirely similar to patients not included. However, because the IRIS Registry is representative of approximately >70% of US ophthalmology practices, we believe this risk will have minimal impact on the study findings. We are also reliant on the racial designation for patients that is provided in the IRIS registry—which is directly extracted from the patient electronic health records. There is some possibility that for some patients these designations might not be self-reported or self-identified; however in our previous similar study including all races, the proportion of Black patients was 13.8%—closely in line with current US Census estimates for those reporting “Black Only” in the US population. As mentioned previously, medication laterality is difficult to establish from the claims data, due to the underreporting of ICD codes in pharmacy fills data. The study therefore assumed that any medications filled were filled for the enrolled eye, which risked overcounting medication utilization both pre- and postintervention. We mitigated this risk by counting medication classes, rather than medications. The level of overcounting should not be differentially biased across cohorts. In addition, the study continued to follow patients after they received other interventions (such as secondary surgical interventions or the addition of new medications). These other interventions may have affected key outcomes for the patient, making it harder to attribute longer-term outcomes solely to the index procedure. However, this limitation applied to all cohorts and had the benefit of not limiting the cohorts to only those patients who had the most favorable outcomes. Sample sizes for the MIGS cohorts were relatively small, with the OMNI and Hydrus cohorts being substantially smaller than the iStent inject cohort. This may be a manifestation of the rate of adoption of each device by surgeons. OMNI was cleared by the FDA in 2017 and Hydrus and iStent inject were approved in 2018.^14–16 However, the inject is a second-generation iStent device, and at the time of its approval, American surgeons had 6 years of experience with the first-generation iStent and likely had a higher adoption rate for the new model upon its approval. Finally, the study was not designed to make between group statistical comparisons; descriptive results are provided, which should be interpreted in light of all of the caveats and limitations outlined above.

We have purposely not compared the results of the current retrospective registry analysis with those from prospective studies of MIGS in combination with cataract surgery. This is because the many differences in study design invalidate such comparisons. In prospective trials there are stringent eligibility criteria that are designed to minimize variability and maximize the observed treatment effect. Most have strict minimum baseline IOP criteria, and most have a preoperative medication washout. It is well known that the higher the baseline IOP, the greater the observed IOP reduction observed for a given treatment.⁴⁷ A medication washout isolates the IOP effect of the surgical treatment from the confounding contribution to IOP reduction of medications. These contributions can be substantial.⁴⁸ In contrast, real-world studies such as ours report real clinical outcomes observed in real clinical practice. There are no minimum IOP requirements—there is no medication washout.

In summary, MIGS in combination with cataract surgery produced clinically and statistically significant reductions in both IOP and the need for IOP-lowering medications through up to 2 years postoperatively in Black patients, which are generally greater than the outcomes for cataract surgery only, but likely dependent on the MIGS performed. Black patients with glaucoma undergoing elective cataract surgery may derive long-term reductions in both IOP and medication use from the addition of MIGS.