In the recently published one-year results of the AMIXT (myopia and intermittent exotropia) Clinical Trial conducted by researchers in Nanjing, China, participants aged six to 12 years who received 0.01% atropine drops nightly had slower spherical equivalent progression and slower axial length elongation than those in placebo. However, a published commentary questions the study’s baseline IXT control scores. Photo: Getty Images.

Intermittent exotropia (IXT) is one of the most common types of strabismus, and affected children are also likely to be myopic. Studies have identified myopia as a risk factor of concomitant exotropia, with one population-based study linking myopia to a 5.2-fold increased risk of developing IXT. However, the growing body of literature on myopia control only infrequently addresses the binocular vision status of the kids being studied and any treatment effects—for good or ill—on strabismus arising from myopia interventions.

This has lately spurred research into management strategies for this specific population, particularly the use of low-concentration atropine, which hasn’t been studied on myopic children with strabismus until now.

The AMIXT (myopia and IXT) study is a two-year, placebo-controlled, double-masked, randomized trial evaluating the safety and efficacy of 0.01% atropine eye drops on myopia progression, exotropia conditions and binocular vision in children aged six to 12 years with myopia and IXT. The one-year results, published last week in JAMA Ophthalmology, show a slower spherical equivalent progression (-0.51D vs. -0.75D) in the atropine group compared to placebo, as well as slowed axial length (AL) elongation (0.31mm vs. 0.42mm). The study also looked at accommodative amplitude (AA), and found a mean change of -3.06D v. .012D in the atropine and placebo groups, respectively.

The study included 300 children with basic-type IXT and myopia of -0.50 to -6D after cycloplegic refraction in both eyes. The trial was conducted at the first Affiliated Hospital with Nanjing Medical University in China from December 2020 to September 2023. All participants were of Han ethnicity due to geographical distribution. Participants were randomly assigned in a 2:1 ratio to 0.01% atropine or placebo eye drops administered in both eyes once a night for 12 months.

In addition to the primary outcomes, researchers in AMIXT reported no aggravated exotropia conditions in the atropine group. They also said the atropine group didn’t appear to have worse binocular vision, and the use of the therapy appeared safe with a mild decrease in AA and a mild increase in pupil size, stability of BCVA, near vision, IOP and corneal endothelial cell density.

There was some initial concern about the decrease in AA, noted the authors. “Previous randomized clinical trials have reported that 0.01% atropine administration led to annual reductions in AA ranging from 0.26D to 4.4D,” they wrote. “In our study, the mean decrease in AA in the 0.01% atropine group was 3.06D, appearing at the two-month visit and stabilized thereafter. We were concerned that the decrease in AA might reduce accommodative convergence and cause aggravation of IXT,” but were pleased to not find evidence of this. “Our results showed no definitive decrease in the near magnitude of exodeviation with 0.01% atropine and no differences for changes in the distant magnitude of exodeviation and distant or near exotropia control between groups, supporting the safety of 0.01% atropine in this trial.”

Several limitations exist with this clinical trial, said the authors, including the single-center study design. They also cite the “lack of correction for multiple comparisons from many secondary outcomes, thus some of the significant differences can be due to chance.” Although COVID lockdowns resulted in some missing data, study drugs were delivered to patients who had missed or delayed visits and researchers kept in close contact with participants, which presumably improved compliance. “Although this study supported the use of 0.01% atropine for the first year, it is uncertain whether efficacy and safety will be maintained in the second year; this will be reported when participants complete two years of follow-up,” wrote the authors.¹

However, it’s the aforementioned trial’s limitation involving the standard deviation of baseline distant exotropia control that is the subject of a commentary, also published in JAMA Ophthalmology. In the study, the AMIXT authors wrote: “The SD of baseline distant exotropia control in this trial is small compared with some other published data, and most participants had relatively good control at baseline. It is possible that our participants did not represent the full clinical spectrum of IXT because those with poorer or worsening control may be more easily noticed, and their guardians were more likely to ask for surgery.”

The commentary author says the concerns regarding IXT control scores, at baseline and outcome, “cast a shadow” over AMIXT’s primary conclusions that 0.01% atropine drop have no meaningful effect on IXT.

“Before we consider Wang and colleagues’ results, we need to look at their baseline data, and here we see something unusual about their control scores,” wrote the commentary author. “Using the published control scale, a much higher proportion of children in the Wang et al. study had a control score of 2 than any previously published cohort; 86% of the Wang et al. cohort had a distance control score of 2 at baseline and only 7% had scores of either 1 or 3 (combined). The minute-to-minute variability of control would seem to render this biologically impossible. Children who have a control score of 2 at one moment often have a control score of 1 or 3 or 4 or 5 a minute or so later.”

The author goes on to say that it’s unclear why the AMIXT trial had so few participants with baseline distance control scores of 1 or 3, and says it’s possible that the trial researchers weren’t assigning control scores using the same methods as described in the literature.

“In summary, it is hard to imagine a cohort of children with IXT (even with constrained ages and refractive errors) who had Wang and colleagues’ baseline distribution of distance control scores,” said the commentary author. “Readers find themselves in a difficult place. Unresolved questions regarding the baseline control scores carry forward to the outcome control scores, and indeed the authors’ reported standard deviation of change in control score is notably less than published cohorts.

“We must at least question the generalizability of [their] conclusion, and this discussion may remind us to pay attention to the baseline characteristics of study cohorts, asking ourselves ‘can I apply the results?’” concluded the commentary.²