SYSTEMIC MEDICATION USE AND THE INCIDENCE AND GROWTH OF GEOGRAPHIC ATROPHY IN THE COMPARISON OF AGE-RELATED MACULAR DEGENERATION TREATMENTS TRIALS

Abstract

Purpose:

To determine associations of systemic medications with the incidence and growth of geographic atrophy (GA) in participants of the comparison of age-related macular degeneration treatments trials.

Methods:

Participants of comparison of age-related macular degeneration treatments trials with new untreated choroidal neovascularization in the study eye (one study eye per participant) were randomized to receive treatment with bevacizumab or ranibizumab. Participants were released from clinical trial treatment at 2 years and examined at approximately 5 years. Color fundus photographs and fluorescein angiograms taken at baseline, Years 1, 2, and 5 were assessed for the presence and size of GA by two masked graders. Participants were interviewed about systemic medication use at baseline. Systemic medications previously reported to be associated with age-related macular degeneration were evaluated for associations with GA incidence in study eye using univariable and multivariable Cox models and for association with the GA growth using linear mixed effects models.

Results:

In multivariable analysis of 1,011 study eyes without baseline GA, systemic medications, including cholinesterase inhibitors, angiotensin-converting enzyme inhibitors, calcium channel blockers, beta-blockers, diuretics, aspirin, steroids, statins, hormone replacement therapy, antacids, and drugs targeting G protein-coupled receptors, were not associated with GA incidence in the study eye (all adjusted hazard ratios ≤1.86, P ≥ 0.18). In multivariable analysis of 214 study eyes with longitudinal GA size measurements, calcium channel blockers were associated with a higher GA growth rate (0.40 vs. 0.30 mm/year, P =0.02).

Conclusion:

None of the systemic medications analyzed were associated with GA incidence. However, calcium channel blockers were associated with a higher growth rate of GA in the study eye.

Age-related macular degeneration (AMD) is the leading cause of vision loss among people older than 50-year-old in the United States.¹ Geographic atrophy (GA) is one of the reasons for visual impairment in late-stage AMD in addition to the formation of choroidal neovascularization. Geographic atrophy is estimated to affect more than 5 million people globally, and its prevalence increases exponentially with age.² Geographic atrophy typically manifests as a progressive and irreversible loss of photoreceptors, retinal pigmented epithelium, and the choriocapillaris.³ Currently, no effective treatment for GA exists.

The underlying pathophysiology of GA remains unclear but is likely to involve oxidative stress⁴ and excessive accumulation of lipofuscin.⁵ Given that GA occurs in the elderly who often require systemic medications for other comorbidities, understanding the impact of these medications on the development or progression of GA is important. A secondary analysis of the rich data from the Comparison of Age-related Macular Degeneration Treatments Trials (CATT) was performed to determine whether systemic medications used at baseline are associated with the incidence and growth of GA in the study eye during 5-year follow-up. We evaluated systemic medications that had been previously reported to be associated with AMD including cholinesterase inhibitors,⁶ angiotensin-converting enzyme (ACE) inhibitors,⁷ calcium channel blockers,⁸ beta-blockers,^9,10 diuretics,¹¹ aspirin,^12,13 steroids,¹⁴ statins,¹⁵ iron medications,¹⁶ hormone replacements,^17,18 antacids,¹⁹ and drugs targeting G protein-coupled receptors.²⁰

Methods

Details on the CATT study design and methods for determining GA incidence and measuring GA size have been previously reported ^21,22 and on ClinicalTrials.gov (NCT00593450). Only the salient features related to this study are described here.

At enrollment, participants were interviewed about medication use, including the drug name, dose, frequency, route of administration, and dates of use. As previously described,²² color fundus photographs (CFPs) and fluorescein angiograms (FAs) taken at baseline, 1, 2, and 5 years were used for assessing GA. The diagnostic criteria for GA in the study eye and fellow eye included a partial or complete depigmentation of one or more patches ≥ 250 μ in the longest linear dimension within the macular vascular arcades in the CFP, with at least one of these additional characteristics: sharply demarcated borders seen in the CFP and/or FA, visibility of underlying choroidal vessels, excavated or punched out appearance on stereoscopy of CFP or FA, or uniform hyperfluorescence bounded by sharp borders on late-phase angiography. To differentiate fibrosis from GA, the fibrotic scars were defined as obvious white or yellow mounds of fibrousappearing tissue that were well defined in shape and appeared solid on color stereophotographic images.²³ Of note, despite spectral-domain optical coherence tomography was not used for determining and measuring GA area, a recent study found that the measurement of GA size and growth by CFP or FA is similar to that from spectral-domain optical coherence tomography in eyes without neovascular AMD.²⁴ Two trained and certified graders at the CATT Fundus Photograph Reading Center determined the presence of GA, and the discrepancies between the two graders were adjudicated. The area of each individual GA lesion was measured manually with ImageJ by two independent graders. When discrepancies were above 50% or 2 mm² between two graders, an adjudication was performed with either a new measurement or an average of measurements from the two graders. The total area of GA lesions in the study eye was calculated for statistical analysis. The square root transformation for the total area of GA lesions was used in calculating GA growth rate.

For each medication class, we evaluated the association with GA incidence in the study eye using univariable and multivariable Cox regression models and association with GA growth in the study eye using univariable and multivariable mixed effects models. These multivariable analyses for GA incidence included GA risk factors previously identified using the CATT data (age, study drug with either ranibizumab or bevacizumab, treatment regimen, hypercholesterolemia, baseline visual acuity in the study eye, baseline GA in the fellow eye, baseline choroidal neovascularization area, retinal angiomatous proliferation lesion, subretinal tissue complex, intraretinal fluid, and subretinal fluid).²¹ For the systemic medications that were only used by very small numbers of patients (e.g., <40), the multivariable analysis for GA incidence was only adjusted by baseline Age-Related Eye Disease Study (AREDS) simple severity scale, to avoid overfitting of the data. The multivariable analysis of GA growth included covariates that we previously found to be associated with GA growth in the CATT data (study drug with either ranibizumab or bevacizumab, treatment regimen, baseline GA in fellow eye, hemorrhage, and subretinal pigmented epithelium fluid).²²

All the statistical analyses were performed in SAS v9.4 (SAS Institute Inc, Cary, NC), and P ≤ 0.05 was considered to be statistically significant.

Results

Among a total of 1,185 patients enrolled into the CATT, 1,011 (85.3%) patients were eligible for the study eye GA incidence analyses (Figure 1). The crude incidence of GA was 265 (26.2%) patients in the study eye. The study eye GA growth analysis cohort included 214 patients who had GA size measured at two or more time points (baseline, Years 1, 2, or 5), including 81 with baseline GA, 114 with incident GA at Year 1, and 19 with incident GA at Year 2 (Figure 2). The median (first quartile and third quartile) of initial GA size was 1.2 (0.6, 2.9) mm² for 81 eyes with baseline GA and 1.2 (0.5, 2.7) mm² for 114 eyes with incident GA. The baseline characteristics of GA incidence cohort and GA growth cohort are summarized in Table 1. Specifically, the mean baseline age of patients was 79-year-old in the GA incidence cohort and 81-year-old in the GA growth cohort. The GA incidence cohort included 61% women, and the GA growth cohort included 64% women. Current smoking was reported for 8% of the GA incidence cohort and 7% of the GA growth cohort. Most patients had a baseline AREDS simple severity scale of 4 (approximately 83% in the GA incidence cohort and 92% in the GA growth cohort, Table 1). The use of each group of systemic medications at baseline was shown in online Supplemental Digital Content 1 (see Table 1, http://links.lww.com/IAE/B374). The most frequently used systemic medications were statin (49%), ACE inhibitors (46%), aspirin (42%), and diuretics (41%). Most of systemic medications were used with a median of 50 months or above.

Fig. 1.

Flowchart for the study eye GA incidence analysis cohort.

Fig. 2.

Flowchart for the study eye GA growth analyses cohort.

Table 1.

Baseline Characteristics of Study Eye GA Incidence Cohort and GA Growth Cohort

Baseline Characteristics	GA Incidence Cohort (N = 1,011)	GA Growth Cohort (N = 214)
Baseline age, mean (SD)	78.9 (7.6)	80.6 (6.4)
Gender, female (%)	623 (61.6)	136 (63.6)
Smoking status, n (%)
Never	436 (43.1)	94 (43.9)
Former	491 (48.6)	106 (49.5)
Current	84 (8.3)	14 (6.%)
Randomized treatment drug, n (%)
Ranibizumab	517 (51.1)	118 (55.1)
Bevacizumab	494 (48.9)	96 (44.9)
Treatment regimen, n (%)
Monthly for 2 years	259 (25.6)	61 (28.5)
Monthly in year 2, PRN in year 2	246 (24.3)	52 (24.3)
PRN for 2 years	506 (50.1)	101 (47.2)
Baseline visual acuity in study eye (letters), mean (SD)	60.7 (13.4)	59.4 (13.1)
Baseline visual acuity in fellow eye (letters), mean (SD)	67.2 (26.0)	58.2 (31.1)
Baseline GA in fellow eye: yes (%)	87 (8.7)	63 (29.7)
Baseline study eye choroidal neovascularization area (DA), mean (SD)	2.5 (2.5)	2.5 (2.6)
Retinal angiomatous proliferation lesion in study eye: yes (%)	112 (11.2)	38 (17.8)
Hemorrhage in study eye: yes (%)	606 (60.4)	155 (72.4)
Subretinal pigmented epithelium fluid in study eye: yes (%)	495 (53.0)	106 (53.8)
AREDS simple severity scale, n (%)
2	63 (6.2)	4 (1.9)
3	100 (9.9)	13 (6.1)
4	848 (83.9)	197 (92.1)

The results from univariable and multivariable analyses for the associations between each group of systemic medication and GA incidence in the study eye are shown in Table 2. In univariable analysis, the cholinesterase inhibitor use at baseline (n = 17) was associated with a higher GA incidence (hazard ratio [HR] = 2.58, P = 0.009), whereas this association was not significant in multivariable analysis (HR = 1.65, P = 0.18). All other systemic medications were not significantly associated with GA incidence in both univariable analysis (all P ≥ 0.17) and multivariable analysis (all P ≥ 0.18). A large hazard ratio was observed for the association between using iron medication at a daily dose of 50 mg or more (n = 10) and GA incidence in both univariate analysis (HR = 1.99, P = 0.17) and multivariable analysis (HR = 1.86, P = 0.22) (Table 2).

Table 2.

Univariable Analyses and Multivariable Analyses for the Association Between Baseline Medication Use With 5-Year Incidence of GA in the Study Eye (N = 1,011)

	Medications No	Medication Yes	Univariable Analysis		Multivariable Analysis*
Systemic Medications Taken at Baseline	Crude GA Incidence in Study Eye (%)	Crude GA Incidence in Study Eye (%)	Unadjusted HR (95% CI)	P	Adjusted HR (95% CI)	P
Cholinesterase inhibitors	257/994 (25.9)	8/17 (47.1)	2.58 (1.27–5.21)	0.009	1.65 (0.80–3.43)	0.18
ACE inhibitors	142/542 (26.2)	123/469 (26.2)	1.02 (0.80–1.30)	0.88	1.12 (0.87–1.45)	0.38
Calcium channel blockers	189/736 (25.7)	76/275 (27.6)	1.17 (0.90–1.53)	0.25	1.17 (0.89–1.54)	0.27
Beta-blockers	176/674 (26.1)	89/337 (26.4)	1.05 (0.81–1.35)	0.73	1.01 (0.78–1.32)	0.93
Diuretics	158/593 (26.6)	107/418 (25.6)	1.03 (0.81–1.32)	0.78	1.00 (0.77–1.29)	0.98
Aspirin	163/584 (27.9)	102/427 (23.9)	0.88 (0.69–1.12)	0.30	0.91 (0.70–1.18)	0.48
Steroids	232/885 (26.2)	33/126 (26.2)	1.09 (0.76–1.57)	0.64	1.29 (0.87–1.90)	0.21
Steroids (inhale)	253/977 (25.9)	12/34 (35.3)	1.33 (0.74–2.37)	0.34	1.44 (0.81–2.58)†	0.21†
Steroids (oral)	260/985 (26.4)	5/26 (19.2)	0.87 (0.36–2.10)	0.75	0.85 (0.35–2.05)†	0.71†
Steroids (topical)	245/933 (26.3)	20/78 (25.6)	1.12 (0.71–1.77)	0.63	1.20 (0.73–1.96)	0.47
Statin	130/515 (25.2)	135/496 (27.2)	1.08 (0.85–1.37)	0.55	1.19 (0.92–1.53)	0.18
Iron medication (all doses)	259/996 (26.0)	6/15 (40.0)	1.57 (0.70–3.52)	0.28	1.52 (0.68–3.42)†	0.31†
Iron medication (50 mg or more)	261/1,001 (26.1)	4/10 (40.0)	1.99 (0.74–5.35)	0.17	1.86 (0.69–4.99)†	0.22†
Iron supplements	228/868 (26.3)	37/143 (25.9)	1.03 (0.73–1.46)	0.85	1.15 (0.80–1.65)	0.46
Hormone replacement (females only)	253/977 (25.9)	12/34 (35.3)	1.35 (0.76–2.41)	0.31	1.35 (0.76–2.41)†	0.31†
Drug targeting G protein-coupled receptors	239/906 (26.4)	26/105 (24.8)	0.95 (0.63–1.42)	0.80	0.98 (0.64–1.51)	0.94
Antacid	247/954 (25.9)	18/57 (31.6)	1.24 (0.77–2.01)	0.37	1.20 (0.74–1.93)†	0.47†

Bold entries mean P < 0.05.

^*Using no mediation as a reference group, adjusted by age, smoking status, drug, treatment regimen, baseline visual acuity in study eye, baseline choroidal neovascularization area, RPA lesion, GA in fellow eye, subretinal tissue complex, intraretinal fluid, subretinal fluid, and AREDS severity scale.

^†Adjusted by an AREDS scale alone because of a small number of patients taking this specific medication.

The associations between each group of systemic medication and study eye GA growth are shown in Table 3. In multivariate analysis, the use of calcium channel blockers was associated with a higher GA growth rate in multivariable analysis (0.40 vs. 0.30 mm/year, P = 0.02), and aspirin use was marginally associated with a higher rate of GA growth with borderline significance (0.37 vs. 0.29 mm/year, P = 0.051). Only five patients took oral steroids at baseline; however, the use of oral steroids was associated with a higher GA growth rate (0.57 vs. 0.32 mm/year, P = 0.04). No statistically significant associations were found between other systemic medications and GA growth (Table 3). There were no statistically significant differences in baseline characteristics between patients with and without the use of calcium channel blockers at baseline (P ≥ 0.10, data not shown).

Table 3.

Univariable Analyses and Multivariable Analyses for the Association Between Baseline Medication With GA Growth in Study Eye During 5-Year CATT (N = 214)

			Univariable Analyses			Multivariable Analyses*
Systemic Medications Taken at Baseline		Frequency (%)	Mean GA Growth Rate mm/yr (SE)	Mean Difference mm/yr (95% CI)	P	Mean Growth Rate mm/yr (SE)*	Mean Difference mm/yr (95% CI)	P
Cholinesterase inhibitors	Yes	7 (3.3)	0.36 (0.13)	0.03 (−0.22, 0.29)	0.79	0.45 (0.16)	0.13 (−0.19 to 0.45)	0.43
	No	207 (96.7)	0.33 (0.02)			0.33 (0.02)
ACE inhibitors	Yes	95 (44.4)	0.35 (0.04)	0.03 (−0.04, 0.10)	0.39	0.34 (0.05)	0.03 (−0.05 to 0.10)	0.50
	No	119 (55.6)	0.31 (0.02)			0.32 (0.03)
Calcium channel blockers	Yes	58 (27.1)	0.38 (0.05)	0.08 (0.00, 0.16)	0.06	0.40 (0.05)	0.10 (0.02 to 0.19)	0.02
	No	156 (72.9)	0.31 (0.02)			0.30 (0.02)
Beta-blockers	Yes	63 (29.4)	0.36 (0.05)	0.04 (−0.03, 0.12)	0.27	0.36 (0.05)	0.04 (−0.04 to 0.12)	0.30
	No	151 (70.6)	0.31 (0.02)			0.31 (0.02)
Diuretics	Yes	98 (45.8)	0.32 (0.04)	−0.02 (−0.10, 0.05)	0.51	0.32 (0.05)	−0.02 (−0.10 to 0.06)	0.61
	No	116 (54.2)	0.34 (0.03)			0.34 (0.03)
Aspirin	Yes	91 (42.5)	0.36 (0.04)	0.05 (−0.02, 0.13)	0.14	0.37 (0.05)	0.07 (0.00 to 0.15)	0.051
	No	123 (57.5)	0.30 (0.02)			0.29 (0.02)
Steroids	Yes	36 (16.8)	0.31 (0.05)	−0.02 (−0.12, 0.07)	0.65	0.31 (0.05)	−0.02 (−0.12 to 0.08)	0.67
	No	178 (83.2)	0.33 (0.02)			0.33 (0.02)
Steroids (inhale)	Yes	12 (5.6)	0.26 (0.08)	−0.08 (−0.22, 0.07)	0.32	0.26 (0.08)	−0.08 (−0.23 to 0.07)	0.32
	No	202 (94.4)	0.33 (0.02)			0.33 (0.02)
Steroids (oral)	Yes	5 (2.3)	0.57 (0.12)	0.25 (0.01, 0.49)	0.04	NA	NA	NA
	No	209 (97.7)	0.32 (0.02)			NA
Steroids (topical)	Yes	23 (10.7)	0.28 (0.06)	−0.05 (−0.17, 0.06)	0.37	0.27 (0.07)	−0.06 (−0.19 to 0.07)	0.37
	No	191 (89.3)	0.33 (0.02)			0.33 (0.02)
Statin	Yes	103 (48.1)	0.32 (0.04)	−0.01 (−0.08, 0.07)	0.87	0.32 (0.05)	−0.01 (−0.08 to 0.07)	0.85
	No	111 (51.9)	0.33 (0.03)			0.33 (0.03)
Iron medication	Yes	8 (3.7)	0.39 (0.09)	0.06 (−0.12, 0.24)	0.51	0.39 (0.09)	0.06 (−0.12 to 0.24)	0.50
	No	206 (96.3)	0.33 (0.02)			0.32 (0.02)
Iron medication	Yes	7 (3.3)	0.36 (0.10)	0.03 (−0.16, 0.23)	0.74	0.36 (0.10)	0.03 (−0.16 to 0.23)	0.74
(50 mg or more)	No	207 (96.7)	0.33 (0.02)			0.33 (0.02)
Iron supplements	Yes	34 (15.9)	0.27 (0.05)	−0.07 (−0.16, 0.03)	0.18	0.28 (0.05)	−0.06 (−0.16 to 0.04)	0.24
	No	180 (84.1)	0.34 (0.02)			0.34 (0.02)
Hormone replacement (females only)	Yes	11 (5.1)	0.41 (0.09)	0.09 (−0.08, 0.26)	0.28	0.44 (0.09)	0.12 (−0.06 to 0.30)	0.18
	No	203 (94.9)	0.32 (0.02)			0.32 (0.02)
Drug targeting G protein-coupled receptors	Yes	24 (11.2)	0.32 (0.06)	−0.01 (−0.12, 0.11)	0.91	0.33 (0.06)	0.00 (−0.12 to 0.12)	0.98
	No	190 (88.8)	0.33 (0.02)			0.33 (0.02)
Antacid	Yes	14 (6.5)	0.37 (0.08)	0.04 (−0.11, 0.19)	0.58	0.39 (0.08)	0.06 (−0.09 to 0.22)	0.42
	No	200 (93.5)	0.33 (0.02)			0.32 (0.02)

Bold entries mean P < 0.05.

^*Adjusted by drug, treatment regimen, baseline GA in fellow eye, hemorrhage, and subretinal pigmented epithelium fluid. NA, not available because few patients had longitudinal GA measurement for multivariate modeling.

Discussion

In this secondary analysis of CATT data, we comprehensively evaluated many types of systemic medications for their associations with GA incidence and GA growth. We did not find any evidence of an association between multiple groups of systemic medications and GA incidence. However, we found calcium channel blockers were significantly associated with a higher growth rate of GA. None of the other systemic medications were associated with GA growth except that five patients with oral steroid use showed higher GA growth in univariate analysis.

The literature shows mixed and conflicting results regarding the effect of systemic medications on AMD. For example, previous studies showed statin use was protective for both early and late AMD;²⁵ however, other studies did not provide evidence of a beneficial effect.¹⁵ Similarly, the study showed an increased risk of AMD with beta-blockers,⁹ but other studies reported contradicting results.¹⁰ Although many studies evaluated systemic medication use with AMD or neovascular AMD, most studies did not evaluate systemic medication use with GA incidence, likely because GA is much less common than early AMD and neovascular AMD, and the cohort studies did not provide sufficient incident GA cases for statistical analysis. In this study, we evaluated associations with a comprehensive list of systemic medications with both GA incidence and growth and did not find any associations with GA incidence.

A recent published article showed that cholinesterase inhibitors may prevent or delay the development of neovascular AMD.²⁶ Possible mechanisms are that cholinesterase inhibitors promote the blood supply of retinal microvasculature and decrease the overproduction reactive oxygen species.⁶ In this study, among 17 patients who reported the use of cholinesterase inhibitor at baseline, cholinesterase inhibitor use was found to be associated with higher GA incidence in univariable analysis (HR = 2.58, P = 0.009), whereas the association was not significant in multivariable analysis (HR = 1.65, P = 0.18). Cholinesterase inhibitors were not associated with GA growth in both univariable analysis and multivariable analysis. One possible explanation is that patients who use cholinesterase inhibitors for cognitive impairment also suffer from GA. It is worth noting that only 17 patients were using a cholinesterase inhibitor at baseline, so that the statistical power is limited for evaluating associations between cholinesterase inhibitor use with GA incidence and GA growth. Therefore, larger studies are warranted to further evaluate these associations.

The contribution of iron to the AMD development has been previously described.¹⁶ Our previous analysis of CATT data showed that iron use (medication or supplement) was associated with a dose-dependent risk of retinal and subretinal hemorrhage in patients with neovascular AMD.²⁷ In this study with a low number of patients taking iron medications (n = 15), we found iron medication at a daily dose of 50 mg or more (n = 10) was associated with approximately a doubling of risk of GA in both study eyes (P = 0.22, Table 2) in multivariate analysis adjusted by the AREDS simple severity scale. Of note, anemia could confound the association between iron medication and GA because temporal macula atrophy due to anemia has been reported in patients with sickle cell disease.²⁸ Given the small number of patients taking iron, it is still an open question as to the effects of iron on GA, and a larger study is needed to better understand their association.

ACE inhibitors have showed inconsistent results for association with AMD incidence.⁷ Klein et al²⁹ reported that beta-blockers were associated with a higher incidence of neovascular AMD. However, results from the AREDS showed, relative to nonhypertensive patients, patients with hypertension were 1.5 times as likely to have neovascular AMD.¹⁹ Because hypertension is one of the significant risk factors for both cardiovascular disease and is often treated with beta-blockers, it is difficult to tease out the confounding effect of hypertension entirely. In our study, ACE inhibitors, beta-blockers, or diuretics were not associated with GA incidence or growth.

The Women’s Health Initiative Sight Exam Ancillary Study showed the use of calcium channel blockers was associated with a higher risk of late AMD and GA.⁸ Although our study did not find any significant association between calcium channel blockers and GA incidence, interestingly, calcium channel blockers were significantly associated with faster GA growth. This finding has not been reported before; caution must be taken in interpreting this result as it may have arisen by chance or be attributable to uncontrolled confounding; therefore, it requires validation in future studies.

Similarly, conflicting results have been reported in the literature regarding the use of aspirin and risk of AMD.¹² Analyses from the large scale studies AREDS and AREDS2 did not show any evidence of increased risk of AMD with aspirin use.¹³ Supporting that, our results did not show evidence that aspirin use was associated with increased GA incidence or GA growth. Individuals with or without AMD or GA should continue to take aspirin when it is medically indicated. In addition, our results showed none of the statins, hormone replacement therapy, antacids, and medication targeting G protein-coupled receptors was not statistically associated with GA development or growth, which agrees with the previous findings.^18–20,30

We hypothesized that the anti-inflammatory effect of steroids may have an impact on GA development or growth. Although our results did not show any evidence of steroids in any form, inhaled, oral, or topical were related with GA incidence, oral steroid use in five patients was associated with higher GA growth. The validity of this finding needs to be tested with further study.

Our study has several limitations. First, observational nature, rather than randomized nature of the study, precludes an interpretation of causality. Second, we used CFPs and FA rather than spectral-domain optical coherence tomography images or fundus autofluorescence images for identifying and quantifying GA. Although different image modalities have been reported to provide similar results for GA presence and size in the eyes without neovascularization, spectral-domain optical coherence tomography images may be better for identifying and measuring GA in some challenging cases. Third, the threshold of adjudication because of interreaders discrepancies was determined as 50% or 2 mm² in the CATT original study, which may have decreased the precision of our area measurements, although only approximately 15% GA measurements required adjudications. Finally, there was very limited number of patients in some medication groups, such as patients who used iron and oral steroids.

In conclusion, in this secondary analysis of CATT data, we did not find significant association between GA development with cholinesterase inhibitors, ACE inhibitors, calcium channel blockers, beta-blockers, diuretics, aspirin, steroids, statins, hormone replacement therapy, antacids, and drugs targeting G protein-coupled receptors. However, we found calcium channel blockers were associated with a higher growth rate of GA. Well-designed prospective studies are needed to validate these findings.