Myopia progression following 0.01% atropine cessation in Australian children: Findings from the Western Australia – Atropine for the Treatment of Myopia (WA‐ATOM) study

A rebound in myopia progression following cessation of atropine eyedrops has been reported, yet there is limited data on the effects of stopping 0.01% atropine compared to placebo control. This study tested the hypothesis that there is minimal rebound myopia progression after cessation of 0.01% atropine eyedrops, compared to a placebo.


| INTRODUCTION
Myopia progression is generally fastest during childhood and stabilises around mid-adolescence, 1-3 although it continues to progress in about one-third of young adults, albeit at a slower rate. 4Myopia control is thus a mid-to long-term commitment.However, it remains unclear for how long treatment should be undertaken.Low-concentration atropine eyedrops are a method of myopia control that has been shown not only slow myopia progression, [5][6][7][8][9] but also to reduce myopia incidence. 10owever, as documented in the Atropine for the Treatment of Myopia (ATOM) trials in Singapore, 11 cessation of high concentration (0.5% and 1.0%) atropine eyedrops after 2 years of treatment was followed by a 'rebound effect', where the rate of myopia progression was faster compared to the placebo group.In contrast, the ATOM study of 0.01% atropine eyedrops did not result in such rebound myopia progression after stopping treatment, at least relative to higher concentrations. 11s later confirmed by the Low-concentration Atropine for Myopia Progression (LAMP) study in Hong Kong, 12 the amount of rebound myopia progression following treatment cessation is inversely linked to the concentration of the atropine eyedrops used, with concentrations of 0.05% linked to greater rebound myopia progression compared to concentrations of 0.025% and 0.01%.The LAMP study further reported that, following 2 years of initial low-concentration atropine therapy, continued usage of 0.05% atropine eyedrops for at least 12 months provided better myopia control outcome than complete cessation. 12ased on the findings of these studies, it has been generally assumed that stopping long-term use of 0.01% atropine eyedrops leads to minimal, if any, rebound myopia progression.However, there is a lack of studies that have directly investigated the effects of stopping longterm 0.01% atropine eyedrops compared to a placebo.In a Japanese study, Hieda et al. 13 found there was no difference in rate of myopia progression in the 12 months following cessation of eyedrops between children who used 0.01% atropine for 2 years compared to those who had used a placebo.However, that study is limited by the large number of participant withdrawals between the start and the end of the 12-month wash-out period ($68% lost to follow-up), which is likely to have impacted their statistical power.While the ATOM 11 and LAMP 12 studies concluded that rebound after stopping 0.01% was limited, neither study had a parallel placebo group to which the wash-out effects of 0.01% atropine could be compared.
Given that atropine drops are commonly prescribed for myopia control therapy, especially in Australia where a commercial version of the drug, Eikance 0.01% (Aspen Australia) has been approved for use since mid-2022 and many compounding pharmacies have dispensed 0.01% atropine eyedrops on prescription, the clinical course following cessation of long-term atropine eyedrops must be delineated.
We recently reported the results of the first 2 yearsthe treatment phase-of the Western Australia (WA)-ATOM study, 9 in which we found that 0.01% atropine eyedrops had a mild to moderate myopia control effect in Australian children after 18 months, compared to a placebo.However, this benefit waned to non-significant levels after 2 years of therapy.This study reports on the third year-the wash-out phase-of the WA-ATOM study, where we tested the hypothesis that rebound myopia progression is minimal for at least 1 year after ceasing application of 0.01% atropine eyedrops compared to a placebo.

| METHODS
The WA-ATOM study 14 is a single-centre, randomised double-masked placebo-controlled trial in which 0.01% atropine eyedrops were tested for myopia control efficacy and safety in parallel against a placebo.Its full design and protocol have been detailed previously. 14A total of 153 Australian children, 6-16 years of age, with documented myopia of ≤À1.50D and progression of ≥0.50D/year were recruited.To enhance recruitment rate, participants were randomised at a 2:1 ratio to receive either 0.01% atropine eyedrops or a placebo.The WA-ATOM study has three phases.The first 2 years of the study comprise the treatment phase, where participants received the allocated eyedrops on a nightly basis.The following 12 months was the washout phase, where eyedrop use ceased and wash-out effects were monitored.During these 3 years of the study, participants and investigators remained masked to the treatment allocation and participants were examined every 6 months.
At enrollment, following a full explanation of the nature of the study, participants and their parents or caregivers provided verbal and written consent, respectively.This trial was conducted in accordance with the Declaration of Helsinki and was approved by the University of Western Australia Human Research Ethics Committee.The use of the placebo and 0.01% atropine eyedrops was approved by the Therapeutics Goods Administration, Department of Health, Australia, and the trial was registered in the Australia and New Zealand Clinical Trials Registry (ACTRN12617000598381).

| Eye examination
Participants were examined at the start of the wash-out period (cessation of eyedrops; 24-month follow-up) and 6 and 12 months thereafter (30-and 36-month follow-ups, respectively).Distance and near visual acuities (logMAR charts) were measured monocularly with participants' habitual distance correction followed by pinholes over their correction.The better VA of the two measurements was taken as the best-corrected visual acuity (BCVA). 8,14ccommodative amplitude was measured using a Royal Air Forces rule (Good-Lite Elgin, Illinois, California), while anterior chamber depth (ACD) and axial length (AL) were measured using an IOLMaster V5 (Carl Zeiss Meditec AG, Jena, Germany).Pupillary measures, including mesopic pupil size and pupil light reactions, were measured in a dark room using a NPi-200 digital pupillometer (NeurOptics Inc., Laguna Hills, California).Participants were instructed to fixate at a small red target at a distance of $3 m for 5 s while a 50-milliWatts white light stimulus lasting 0.8 s triggered pupil light response and the device automatically measured pupil size, and the latency and velocity of constriction.The onset of constriction was defined as a decrease of 5% of the initial baseline pupil size (per confirmation from manufacturer).
One to three drops of 1% cyclopentolate were then instilled in each eye, depending on the amount of cycloplegia achieved, and autorefraction and autokeratometry (Nidek ARK-510A, NIDEK Co. Ltd, Japan) were performed at least 20 min after the last drop, with cycloplegia being confirmed through assessment of the light pupil response.The cycloplegic refraction was then quantified as the spherical equivalent (SphE).Additionally, at the 24-and 36-month visits, Scheimpflug imaging (Oculus Pentacam [software version 6.08r27; OculusOptikgerate GmbH, Wetzlar, Germany]) was conducted to measure the central corneal thickness and crystalline lens thickness after dilation.

| Statistical analysis
Analyses were conducted on an intention-to-treat basis.The main outcome measure during this wash-out phase was the changes in SphE and AL from during the wash-out phase (24-and 36-month).Linear mixed effect models were used to analyse the difference between groups.A random intercept term with nesting for eyes within individuals was included in the models to account for the repeated measurements of the two eyes and multiple visits.Baseline measurement of the ocular outcome was included in the models as a fixedeffect covariate, along with index age given the significant difference in age between groups at baseline (Table 1). 8We further explored for treatment group interaction effects with age and ethnicity.All analyses were conducted on R version 4.1.1(2021 The R Foundation for Statistical Computing Platform [https:// www.r-project.org/]) and the level of significance was set at p < 0.05.

| RESULTS
Of the 153 participants enrolled, 104 were randomised to receive 0.01% atropine eyedrops and 49 received a placebo.The demography and baseline ocular measures of each group are shown in Table 1.Compared to those in the atropine group, participants in the placebo group were on average 1 year older, had 0.1 mm deeper anterior chambers, 0.06 mm thinner crystalline lens, and 0.015 ms slower constriction onset ( p = 0.010 to 0.031; Table 1) at baseline.As reported previously, 9 during the 2-year treatment phase, significantly more participants withdrew from the placebo group than the atropine group ( p = 0.024 on Fisher exact test; Figure 1).During the wash-out phase, 2 participants withdrew from the study, both from the atropine group.The reasons cited for withdrawal are described in the footnotes of Figure 1.None of the participants were using any other forms of myopia control, with the exception of progressive addition or bifocal lenses, while they were in the 3-year study.

| Changes after stopping treatment
At the start of the wash-out phase, the median SphE and AL were À4.00 D (interquartile range [IQR] = À5.25 to À3.00) and 25.1 mm (IQR = 24.8 to 25.7) in the placebo group, and À3.88 D (IQR = À4.75 to À3.22) and 25.0 mm (IQR = 24.6 to 25.5) in the atropine group, respectively.Neither of these measures significantly differed between groups at the start of the washout phase (p = 0.57 for SphE and p = 0.19 for AL).
Figure 2 and Table 2 show the changes in SphE and AL during the wash-out phase.Children who had been receiving 0.01% atropine exhibited faster myopia progression after stopping the eyedrops compared to those who had been on a placebo.This difference in rate of myopia progression can be noted at both 6-and 12-months posttreatment but was only statistically significant at the latter visit.By the end of 2 years' treatment and 1-year wash-out periods, the cumulative myopia progression since baseline was similar between the two groups (Figure 2).
In both groups, the rate of SphE and AL change was significantly slower in the second half of the wash-out period (last 6 months) compared to the first half (p = 0.032 and p < 0.001, respectively).As also seen in Figure 2, the slowing appeared more pronounced in the placebo group than the 0.01% atropine group, but this difference in rate of slowing was not statistically significant (p-value for group by visit interaction effect = 0.91 for SphE and 0.13, for AL).
Within just the atropine group, myopia progressed significantly faster in terms of both SphE and AL change in the 12 months following eyedrop cessation compared to the 12 months immediately before.In contrast, progression rate in the placebo group was similar before and after eyedrop cessation (Table 3).
Additionally, a significant increase in ACD was noted in the atropine group compared to the placebo group at both 6 and 12 months after cessation of eyedrops (Table 2).Changes in pupillary measures did not differ significantly between the placebo and control groups after cessation of eyedrops.
In both groups, older age was associated with slower myopia progression, such that SphE and AL changes were slowed by 0.026 (95% CI = 0.001 to 0.051; p = 0.045) and 0.008 (95% CI = 0.001 to 0.017; F I G U R E 1 Sample size and retention.a Reasons for withdrawal: 3 did not like diagnostic or study drops; 4 had difficulty adhering to treatment regimen; 3 had difficulty attending appointments + concern of receiving placebo; 2 were uncontactable; 1 relocated; 3 wanted to seek myopia treatment (atropine or orthokeratology) privately; 2 was uncontactable; 4 cited personal reasons/ did not provide reason.b Reasons for withdrawal: 1 sought atropine and orthokeratology eyedrops privately, 1 was uncontactable.c Some participants did not attend the 30-month visit due to restrictions or concerns about the COVID-19 pandemic but remained in the study.
F I G U R E 2 Change in spherical equivalent (top row) and axial length (bottom row) since start of wash-out phase (left column) and since baseline (right column).Yellow highlighted region indicates the treatment phase, during which participants were using allocated eyedrops.p = 0.034) for each year older in age.There was a trend for less rebound axial elongation with older age, but this failed to reach statistical significance ( p = 0.16).There were no interaction effects between treatment group and age on change in SphE ( p = 0.74) or between group and ethnicity on either myopia progression measure ( p ≥ 0.32) during the wash-out period.As summarised by Brennan et al., 15 'rebound should be assumed unless proven otherwise'.There has been an implicit assumption that rebound myopia progression after stopping long-term use of 0.01% eyedrops is minimal, yet empirical evidence to support this assumption has been limited due to the lack of placebo-control trials.
Findings from the current analysis demonstrated that rate of myopia progression increases after cessation of long-term 0.01% atropine eyedrops.Ultimately, after 2 years of eyedrop use and a year of wash-out, there was no difference in cumulative myopia progression between the 0.01% atropine and placebo groups.
A main strength of this study is the inclusion of a placebo-control group to which the effects of treatment wash-out could be compared, as opposed to the ATOM study in Singapore 11 and the LAMP study. 12However, our study did not have a treatment group that continued 0.01% atropine treatment from 24 to 36 months, neither did we test other concentrations of low-concentration atropine eyedrops.A comparison between continuing and ceasing treatment has been done by the LAMP study, 12 which confirms a 'rebound' myopia progression effect after ceasing low-concentration atropine use, compared to continuing treatment.The same study, along with the Singapore ATOM study, also concluded that the amount of rebound is positively associated with the concentration. 11,12Another potential limitation of our study is that we lacked the power to detect statistical difference in some outcomes between groups, for example, a differential rate of slower in myopia progression throughout the wash-out period (i.e., first 6 months vs. the last 6 months).Furthermore, we did not evaluate the effects of tapering atropine treatment.Several researchers have highlighted the importance of evaluating the effects of tapering treatment, [16][17][18] but studies on this remain scarce.A study by Polling et al. 19 in a clinical setting reported that, in children with good myopia control response to 0.5% or 0.25% atropine eyedrops, reducing the concentration to 0.01% resulted in less rebound myopia, although there was no control group to which this could be compared.Nonetheless, their findings suggest that tapering could be an effective way of reducing rebound effects.
In contrast to the LAMP study, we did not find a significant effect of age on the amount of myopia progression after eyedrop cessation.This may be related to our smaller sample size and the older age of our participants, relative to the LAMP study.Because of these limitations of our study, it may not be appropriate to assume that age does not have an effect on rebound myopia progression.Further studies should explore this issue, which will inform on the age at which myopia control can cease without leading to a rebound effect.Studies have reported that the mean age of myopia stabilisation occurs during the mid-teenage years, ranging from 14 to 17 years old, 1,2 although this varies widely between individuals and according to sex, 2 ethnicity, 1 severity of myopia, 1 and amount of near work. 20The Correction of Myopia Evaluation Trial (COMET) reported that up to 23% and 10% of participants with myopia continued to have progression after 18 and 21 years of age, respectively. 1Thus, the age of natural myopia stabilisation, ergo the age at which myopia control can stop, should be tailored to each individual, although this may be difficult to assess without stopping treatment.
However, the current findings, along with the recent observations from the Atropine Treatment Long-term Assessment Study (ATLAS), 21 may have a more pessimistic conclusion-that the use of atropine eyedrops does not meaningfully alter final refractive error in the long term.The ATLAS is a long-term follow-up study of the ATOM 1 and 2 studies in Singapore.In the ATOM1 study, by the end of 2 years of treatment and 1 year of wash-out, the 1% atropine group had significantly less myopia progression in terms of SphE and AL, compared to those who were on placebo eyedrops.However, when the participants were re-examined 20 years later as part of the ATLAS, there was no significant difference in T A B L E 3 Estimated marginal mean (and 95% confidence intervals) myopia progression in the 12 months immediately before and after cessation of eyedrops.either measure between the two groups. 21Likewise, in the ATOM2 study, there was no significant difference in refractive error between groups who used different atropine concentrations (0.01%, 0.1%, and 0.5%) more than 10 years after the trial ended, 21 even though the majority of those participants underwent re-treatment with 0.01% atropine in the final 2 years of the 5-year trial. 22This highlights the possibility that lowconcentration atropine eyedrops only delays and do not control myopia progression.The LAMP2 study, 10 which evaluated the effects of atropine eyedrops on myopia incidence, similarly concluded that the lower myopia incidence in the treated group, relative to the placebo-control group, could simply be a delay in myopia onset, rather than prevention.This hypothesis that treatment only delays and does not control myopia progression is further supported by the current findings.While myopia progression was slower in the second half of the 12 months wash-out period compared to the first half in both groups, there was a trend for a more pronounced slowing in the placebo group, albeit not statistically significant possibly due to a lack of study power.Nonetheless, we may assume that the rate of slowing in myopia progression in the placebo group is natural and age-related, given the lack of intervention in this group.The attenuated slowing in the control group may be due to a tendency for myopia progression to 'catch up' following cessation of treatment.

While using eyedrops
Given the relatively recent growth in myopia control research and use in clinical practice, further longterm investigations are critical to test this hypothesis as it will have direct and highly impactful clinical significance in myopia control.There has been limited data on the long-term benefits, or lack thereof, of atropine therapy and other forms of myopia control.The ATLAS study, 21 which now comprises participants in their third and fourth decades of life, failed to find any difference in eye and refractive error outcomes, including rates of myopia-related complications, between groups using varying concentrations of atropine eyedrops.However, the ATLAS participants had generally been using mid-to-high concentrations, apart from the 0.01% atropine and the placebo group.The current study, along with those from the LAMP study in Hong Kong, 12 provides some insight on the short-to midterm effects of stopping low-concentration atropine eyedrops.
The current findings demonstrate rebound myopia progression after stopping 0.01% eyedrops, such that early cessation of 0.01% atropine eyedrops might negate the benefits of treatment in the preceding years.Future studies should explore the effect of tapering atropine dosage or frequency of instillation.
Participant demography and ocular measures at baseline.
T A B L E 1Note: Difference between groups analysed using a independent t-test for age, b chi-square test for categorical variables, and c linear mixed effect model for ocular measures.Abbreviations: BCVA, best-corrected visual acuity; IQR, interquartile range; SD, standard deviation.*Significantly different between groups at p < 0.05.