Short‐term effects of cyclopentolate and tropicamide eye drops on macular choroidal thickness in myopic children

To investigate the short‐term effects of cyclopentolate and tropicamide eyedrops on choroidal thickness (ChT) in myopic children using placebo or low‐dose atropine eyedrops.

prior to axial elongation, while an increased axial length can also mechanically stretch and thin the choroid. 1,5horoidal thinning has been implicated as a risk factor for myopic maculopathy, the most severe ocular complication of myopia. 6Thus, measurement of choroidal thickness (ChT) may have clinical value as a potential indicator of myopic maculopathy risk.8][9] For example, myopia control spectacles, orthokeratology and atropine eye drops have all been reported to increase ChT or slow choroidal thinning during myopia treatment. 7,10,11herefore, analysis of the ChT response to myopia control intervention might provide a useful additional biomarker of treatment efficacy on the basis that the optimal treatment for myopia would positively impact both axial elongation and ChT in progressive myopes.
Despite the emergence of ChT as a potential biomarker in myopia management, there is no consensus on whether ChT should be measured with or without cycloplegia, and there is limited evidence to determine what impact cycloplegic regimes may have on ChT in myopic children.Pharmacological pupil dilation enhances the quality of optical coherence tomography (OCT) scans, particularly in children with smaller pupils.Additionally, cycloplegic refraction (specifically two drops of tropicamide 1% or cyclopentolate 1%) is a recommended component of myopia management 12 and routinely used when refracting children.Previous studies of adults have reported thickening, 13 thinning [14][15][16] or no significant change 13,17,18 in ChT in response to cyclopentolate 13,14,16 and tropicamide, 13,15,17,18 while choroidal thickening is seen in response to both short-and long-term topical application of another anti-muscarinic eye drop, atropine. 11,19The majority of studies investigating the effects of cyclopentolate and tropicamide have used adult subjects, and no investigation has yet examined the short-term effects of cyclopentolate or tropicamide on ChT in myopic children receiving long-term, low-dose atropine eye drop therapy.
Given the possibility that ChT measurements could be taken either pre-or post-cycloplegia in clinical practice, it is important to understand the effect of different cycloplegic agents on ChT, particularly in children undergoing active myopia management where ChT or changes in ChT might guide clinical decision-making.The purpose of this study was to investigate short-term changes in macular ChT following the instillation of cyclopentolate and tropicamide in myopic children undergoing treatment with atropine or placebo eye drops.

Study participants
This study recruited myopic participants (aged 6-16 years and 4-12 years for the cyclopentolate and tropicamide groups, respectively) involved in two clinical trials exploring the efficacy and safety of low-dose (0.01% or 0.02%) atropine at the Centre for Eye Research Ireland, as previously described. 20,21The studies were approved by the Research Ethics Committees at the Mater Misericordiae University Hospital, Ireland, Tallaght University Hospital, Ireland and Technological University Dublin, Ireland and adhered to the tenets of the Declaration of Helsinki.Participants and their parents provided written informed assent and consent, respectively, prior to enrolling in the studies.Participants who had ocular co-morbidities, astigmatism ≥2.50 D in either eye, anisometropia >1 D, strabismus, intraocular pressure ≥ 21 mmHg, hypersensitivity to atropine and other significant health problems were excluded.
Data for this analysis were collected at the 24-month visit from all participants involved in the two low-dose atropine trials.Cycloplegic refraction was performed with tropicamide 1% eye drops in 49 participants involved in one of the trials, and with cyclopentolate 1% eye drops in 198 participants involved in the second trial.Among the 198 participants who received cyclopentolate eye drops, 51 were included in a test-retest repeated measurement (test-retest group) analysis, which was conducted prior to cycloplegia.

Ocular examinations
Participants enrolled in the two clinical trials used either nightly atropine or placebo eye drops in both eyes, and ocular health examinations were performed in both eyes of all participants.Corrected distance visual acuity was measured using a randomised letterset of the MultiQuity (MiQ 720) computerised logMAR chart.Intraocular pressure was determined using the iCare tonometer (ic100, icare -world.com).Axial length and cycloplegic

Key points
• This study provides novel insights into the shortterm changes in choroidal thickness following the use of cycloplegic eye drops in myopic children undergoing long-term, low-dose atropine myopia management.Cycloplegia was confirmed 30 min after instillation of the last eye drop by measuring the pupil size (≥6 mm) with a pupil gauge and assessing the pupil reaction with a penlight.One ChT measurement was obtained from both eyes of participants just prior (pre-cycloplegia) to instilling the eye drops and a second scan at least 30 min (range 30-43 min) after instillation of the last drop (post-cycloplegia) using swept-source OCT (SS-OCT) (DRI-OCT Triton Plus, topconhealthcare.jp).To investigate the test-retest repeatability of ChT measurements, two successive repeated ChT measurements, approximately 2-3 min apart, were also taken prior to instilling the cyclopentolate eye drops in 51 participants in the cyclopentolate group.
The Triton SS-OCT technology uses 1050 nm wavelength light with a scanning speed of up to 100,000 A-Scan/sec and lateral and in-depth resolutions of 20 and 8 μm, respectively.To minimise the effect of involuntary eye movements, participants were asked to concentrate on a fixation target, and a three-dimensional 7 × 7 mm 2 scan centred on the fovea was acquired.The follow-up function was used to ensure the same macular locations were scanned.To minimise the influence of near-work activities and exercise on ChT, participants did not engage in intense near-work activities or rigorous exercise during the study examinations but were free to move around the examination room while their pupils were dilating.The total study examination time per visit for participants was 1.5 h.
The IMAGEnet 6 software version 1.34.1.19417(Topcon Corporation, topconhealthcare.eu) was used for automated fovea identification, choroidal segmentation and thickness analysis.Automated segmentation was reviewed, and errors due to algorithm failure were manually corrected (33/198 [16.7%] and 7/49 [14.3%] scans in the cyclopentolate and tropicamide participants, respectively, were manually corrected) using the layer modify function in IMAGEnet 6. ChT was measured as the vertical distance between Bruch's membrane and the choroidal-scleral interface, and images required an image quality rating ≥70 to be included in the analysis (Figure 1).All scans and ChT corrections were performed by a single trained examiner (EKA) between 09:00 and 16:00 h each day. 24,25The ChT measurements were extracted in a 256 × 512 array using the OCT Data Collector tool, topconhealthcare.eu,and a custom-written program (MATLAB R2021a, mathw orks.com) was used to adjust for the transverse magnification effects of axial length on image width following Bennett's modified version of Littman's formula. 26,27The Early Treatment of Diabetic Retinopathy Study (ETDRS) grid, centred on the fovea, was applied to each scan, with circle diameters of 1, 3 and 6 mm corresponding to the central fovea (subfoveal), parafoveal and perifoveal regions, respectively, and the mean ChT within each area calculated (Figure 1).

Statistical analyses
Statistical analyses were performed with R version 4.2.2 (R Foundation for Statistical Computing; R-proje ct.org/ ).The intra-class correlation coefficient (ICC) was used to assess the correlation between the right and left eyes and pre-and post-cycloplegia ChT measurements.Participant characteristics were presented as proportions, mean (standard deviation [SD]) and median (interquartile range [IQR]) for categorical, normal-and non-normally distributed continuous data, respectively.The normality of the data distribution was assessed using the Shapiro-Wilk test.
Repeated measures analysis of variance was used to compare the mean change in test-retest measurements and pre-and post-cycloplegia ChT in the nine macular ETDRS regions, with a Bonferroni correction for multiple comparisons applied (0.05/9 = 0.006).
The coefficient of repeatability (CoR) and the 95% CI of the CoR were calculated to assess the repeatability between test-retest and pre-and post-cycloplegia measurements.CoR describes the interval containing 95% of the test-retest and pre-and post-cycloplegia differences within individuals and was calculated following the recommendations of Bland and Altman (1.96 × sqrt (2) x the mean within-subject standard deviation). 28Agreement between test-retest measurements and pre-and post-cycloplegia measurements was analysed using Bland-Altman statistics by calculating the bias (mean difference between preand post-cycloplegia measurements [post-cycloplegia -pre-cycloplegia]) as well as the lower (mean difference -1.96 × the standard deviation of differences) and upper (mean difference + 1.96 × the standard deviation of differences) 95% limits of agreement (LOA).The respective 95% confidence intervals (CI) for the bias, lower and upper 95% LOA were also computed.
Multivariable linear regression was used to explore the influence of treatment group assignment (atropine or placebo), sex, age, axial length, use of second drop and intraocular pressure on the pre-and post-cycloplegia changes in ChT.

Participant characteristics
Table 1 summarises the characteristics of the study participants.Five children (2.5%) and 1 child (2.0%) were excluded from the cyclopentolate and test-retest groups, respectively, due to poor scan quality (n = 1) and uncorrectable segmentation errors (n = 5).A total of 50 participants were included in the test-retest repeated measurement analysis, T A B L E 2 Mean choroidal thickness for test-retest, pre-and post-cycloplegia measurements.

Choroidal thickness changes
In drop and two drops) at all macular locations (subfoveal, parafoveal and perifoveal regions, p < 0.001 for all) (Table 2).On average, ChT was thinner post-tropicamide instillation at all macular locations but not significantly different from pre-cycloplegia ChT (p > 0.05 for all) (Table 2).Pre-to post-cycloplegia ChT changes in the two drops of the cyclopentolate group were significantly greater than test-retest ChT changes at all macular locations, but compared to test-retest, changes in the tropicamide group were not significantly different (Figure 2).This finding could be due to participants in the two drops of cyclopentolate group being more myopic and older than those in the tropicamide groups.Choroidal thinning after cyclopentolate instillation was not significantly different between atropine and placebo treatment groups at all macular locations (p > 0.05 for all) (Figure 3).A higher proportion of individuals had a decrease F I G U R E 2 Difference between choroidal thickness (ChT) measurements at the nine macular locations.(Cyclo, cyclopentolate; SE, standard error).
As there were no significant pre-and post-cycloplegia differences in mean ChT for the tropicamide group, regression analysis was performed only for the cyclopentolate group to explore factors associated with the pre-and post-cycloplegia ChT changes.Multivariable regression analysis showed that long-term atropine eye drop use, sex, axial length and intraocular pressure were not significantly associated with pre-to post-cycloplegia changes in ChT after cyclopentolate instillation at any of the macular ETDRS locations (Table 4).Older age was significantly associated with short-term choroidal thickening only in the parafoveal inferior (β = 0.50, p = 0.004) and perifoveal superior (β = 0.54, p = 0.004) regions.Use of a second drop of cyclopentolate was significantly associated with choroidal thinning at all macular locations (p < 0.006 for all) except for the perifoveal inferior region (Table 4).

DISCUSSION
The current study showed that 1% cyclopentolate caused significant thinning of the choroid in all macular locations, but mean changes in ChT in response to 1% tropicamide eye drops were not significantly different from the test-retest ChT changes.Participants who received two drops of 1% cyclopentolate exhibited greater pre-to post-cycloplegia choroidal thinning.In addition, cycloplegic eye drops, particularly 1% cyclopentolate, produced greater CoR compared with test-retest suggesting that ChT measurements taken after administration of cyclopentolate were more variable.
Compared with the mean test-retest changes, which were significant in the para-and perifoveal regions but not in the subfoveal regions, one drop of cyclopentolate 1% caused significant thinning in the choroid at all macular locations, with the post-cycloplegia measurements being consistently thinner than pre-cycloplegia values.While the greatest choroidal thinning was observed in participants who received two drops of cyclopentolate, it is important to note that participants in the two-drop cyclopentolate group were both older and more myopic than those in the tropicamide group.However, the age and myopic differences, although statistically significant, may not fully explain the observed choroidal thinning; hence, underscoring the need for further studies to better understand the potential interplay between these demographic differences and the observed ChT changes.Further, the use of two drops of cyclopentolate was the only factor significantly associated with greater post-cycloplegia choroidal thinning at all but the perifoveal inferior macular locations on multivariable analysis.This may imply a dose-dependent effect of cyclopentolate on ChT, with administration being causally related to shortterm choroidal thinning.A limitation of the present study was that only participants with darker irides were given two drops of cyclopentolate; therefore, it would be interesting to know whether two drops of cyclopentolate would cause similar dose-dependent short-term choroidal thinning in individuals with lighter irides.The effects of cyclopentolate indicate that its application could potentially have a clinically meaningful impact on ChT measurements.This is important because the magnitude of the observed changes, particularly with two drops of cyclopentolate, was comparable to the ChT changes seen during myopia management. 7,11Choroidal thinning following the instillation of cyclopentolate was similar in the atropine and placebo treatment groups, suggesting that the short-term effects of cyclopentolate were independent of long-term myopia management.To the contrary, 1% tropicamide did not produce any significant changes in ChT following cycloplegia, relative to the testretest ChT changes.Although the 95% LOA were slightly wider, the CoR and mean ChT changes were similar to those observed in test-retest, suggesting a non-clinically meaningful effect of tropicamide on ChT measurements.
T A B L E 3 Coefficient of repeatability and 95% lower and upper limits of agreement for test-retest, cyclopentolate and tropicamide groups.

Subfoveal region Subfoveal
These findings are consistent with previous prospective studies that have reported short-term choroidal thinning after cyclopentolate instillation 14,16,19 but no significant changes with tropicamide. 13,17,18Conversely, other investigations have observed both thickening 13 and thinning 14,15 of the choroid following cyclopentolate and tropicamide instillation, which could be attributed to the study population (adult populations for all studies) frequency of instillation (three times at 5-10 min intervals).For example, variations in instillation protocols are likely to lead to differences in the intraocular concentration of the anti-muscarinic drug.While this may cause no practical difference in terms of the cycloplegic effect, it could theoretically affect biological activity at the retina and choroid. 31yopia management has evolved rapidly in recent years to include therapies that slow down progression and axial elongation.The need to understand the efficacy and mechanisms of action of such treatments has led to an increased emphasis on finding suitable biomarkers of the treatment response.9]11 Although the site of action for atropine in myopia control remains unresolved, 32 recent evidence suggests that atropine's myopia control effect may be correlated with changes in ChT. 11Consequently, ChT has been proposed as a useful biomarker in gauging long-term myopia control treatment. 11If indeed the choroid could be used as a marker for myopia treatment, then it is important to measure ChT accurately during long-term therapy.
The mechanism of action of cycloplegic eyedrops on ChT remains uncertain, but cyclopentolate 1% could potentially induce ChT changes through mechanical actions transmitted from the ciliary body within the anterior uvea posteriorly to the choroid. 33Another plausible explanation could be the binding of cyclopentolate to cholinergic receptors in the choroid. 34Anticholinergic drugs lead to vasoconstriction of perivascular plexuses and contraction of non-vascular smooth muscle that is innervated by parasympathetic fibres in the choroid.As a result, they may modulate changes in ChT by causing fluid efflux within the choroid, leading to its thinning. 33,35,36Atropine (also an anti-muscarinic agent) has been shown to increase ChT, potentially through nitric oxide 37 or dopamine 38 pathways.Cyclopentolate may affect the same pharmacological mechanisms, but analysis of the short-term ChT changes within the atropine and placebo groups showed that post-cyclopentolate choroidal thinning did not differ significantly between the atropine and placebo groups.While longer-term use of low-dose atropine eye drops may have led to an overall thickening of the choroid, multivariable regression analysis showed that long-term use of atropine was not significantly associated with shortterm cyclopentolate-induced changes in ChT.
The CoR statistic used in the test-retest analyses provided a probability of minimal detectable change beyond which any further changes were unlikely to be due to measurement error or noise. 39The increased CoR with cyclopentolate compared with the test-retest group suggests that cyclopentolate drops led to higher variability in the ChT measurements, further corroborated by the greater CoR observed in participants who received two drops of cyclopentolate.From a clinical perspective, these findings suggest that the use of cyclopentolate should be standardised during ChT measurements, particularly when two drops are used.Tropicamide should perhaps be considered as the drug of choice for studies involving ChT measurement unless cyclopentolate is indicated.If cyclopentolate is to be used in such investigations or in clinical practice, then consideration should be given to pre-cycloplegia ChT measurements and, most importantly, to the adoption of a consistent approach when ChT is monitored over time, such as during the course of myopia management.
1][42] It seems unlikely that factors such as exercise and near work would have influenced the findings, as participants tended not to engage in intense near work and only walked between examination rooms during the study visits.A further potential limitation was the inclusion of different treatment groups (i.e., atropine vs. placebo), which were pooled together.Given that there were no differences in ChT changes between the treatment groups, it is unlikely the pooled analyses were impacted.A significant strength of the study was the use of automatic segmentation to segment choroidal layers, which eliminated potential error in judgements during manual segmentation and provided correction for transverse magnification error.Recent advances in SS-OCT segmentation algorithm software have improved automatic choroidal segmentations and allowed for more repeatable ChT measurements. 43Only 16% of study participants required manual segmentation correction.
This study represents the first investigation into the short-term changes in ChT following the administration of cycloplegic eye drops in myopic children undergoing long-term atropine or placebo myopia management.By including both the atropine and placebo treatment groups, this investigation provides valuable insight into whether the long-term use of atropine eye drops impacts the short-term effects of cycloplegic eye drops on ChT measurements.The findings further emphasise the practical significance of exploring ChT as a valuable biomarker for myopia management, as emerging evidence continues to support its relevance.Future studies should aim to elucidate the mechanisms underlying these observed changes and establish a standardised protocol for measuring ChT metrics during myopia management.
In conclusion, in this cohort of myopic children, cyclopentolate eye drops caused significant thinning of the choroid and increased variation in ChT from pre-to post-cycloplegia, whereas administration of tropicamide did not induce significant changes in ChT.Therefore, ChT measurements are best performed prior to cycloplegia, particularly when 1% cyclopentolate is used.If a cycloplegic agent is administered prior to the measurement of ChT, then tropicamide 1% should be used.Advancements in ChT measurement technology are occurring rapidly, and while some confounders to ChT measurement are known, this study highlights the additional need for a standardised approach to the use of cycloplegic eye drops prior to measurement of ChT.

AC K N O W L E D G E M E N T S
Open access funding provided by IReL.

FU N D I N G I N FO R M AT I O N
Health Research Board and Fighting Blindness, Ireland (MRCG 2016-13).Topcon Europe Medical B.V. The sponsors or funding organisations had no role in the design or conduct of this research.

F
I G U R E 1 Optical coherence tomography scan of the retina and choroid.(a) Choroidal thickness (ChT) measured from Bruch's membrane to the choroidal-scleral interface (green horizontal lines).(b) Average ChT measurements within the nine Early Treatment of Diabetic Retinopathy Study locations after adjusting the grid boundaries for axial length-associated transverse magnification.

F I G U R E 3
Changes between pre-and post-cycloplegia subfoveal (top), para-(middle) and perifoveal (bottom) choroidal thickness measurements in atropine 0.01% and placebo eye drop treatment groups for one drop of cyclopentolate (a, c and e) and two drops of cyclopentolate (b, d and f).(ChT, choroidal thickness; ETDRS, Early Treatment Diabetic Retinopathy Study; SE, standard error).

F I G U R E 4
Bland-Altman plots of agreement between first and second, pre-and post-cycloplegia subfoveal (top), parafoveal (middle) and perifoveal (bottom) choroidal thickness measurements for the test-retest group (a, e and i), tropicamide (b, f and j), one drop of cyclopentolate (c, g and k) and two drops of cyclopentolate (d, h and l).The dashed lines in the purple-shaded area in each graph represent the bias between respective measurements, and the dotted purple-shaded area represents the 95% confidence interval (CI) of the bias.The dashed lines in the green-shaded area in each graph represent the 95% upper limits of agreement (LOA), and the dotted green-shaded area represents the 95% CI of the upper LOA.The dashed lines in the red-shaded area in each graph represent the 95% lower LOA, and the dotted red-shaded area represents the 95% CI of the upper LOA.(ChT, choroidal thickness).
Characteristics of study participants.
T A B L E 1*p-Values represent differences between cyclopentolate and tropicamide groups.