Is hypermetropia regressing in accommodative esotropia ?

This content is licensed under a Creative Commons Attributions 4.0 International License. ABSTRACT | Purposes: To evaluate the relationship of chan­ ges in hypermetropia and ocular alignment in patients with accommodative esotropia. Methods: The medical records of consecutive patients diagnosed with refractive accommodative esotropia (esotropia eliminated or decreased to within 10 D with full hyperopic correction) were retrospectively reviewed. Cycloplegic refractions culled from medical records were converted into spherical equivalents. Presence of amblyopia, changes in refractive error and ocular alignment at admission and after the follow­up period were evaluated. Results: Seventy patients (mean age: 6.01 ± 5.41 years; female: 60.6%; mean follow­up: 5.8 ± 3 years) had corrected esotropia of 40 ± 20 prismatic diopters at admission. The average decrease per year in near and distance deviations with glasses was 1.71 ± 3.96 prismatic diopters/year and 1.09 ± 3.25 prismatic diopters/ year, respectively. The total myopic shift of the right and left eyes was 1.08 ± 1.35 D and 1.20 ± 1.40 D, respectively. Myopic shift/year was 0.22 D/year and 0.26 D/year, respectively. The correlation between the rate of myopic shift and rate of change in corrected near deviation was weak. The correlation for the rate of myopic shift was not high for the right and left eyes (r=0.18; p=0.15). Conclusion: The amount of deviation and hypermetropia gradually decreased in accommodative esotropia during follow­up. On the other hand, it may be incorrect to assure patients that the amount of deviation will decrease in parallel with the refractive error.


INTRODUCTION
Refractive accommodative esotropia is frequently associated with moderate or high hypermetropia (1) . Treat ment of cycloplegic refractive error is the mainstay of therapy for the prevention of amblyopia and mainte nance of strereopsis (2) . We frequently encounter ques tions from the families of these patients regarding the duration of spectacle wear and possibility of refractive changes with time. The abnormal accommodation reflex required to overcome blurring caused by undercorrec tion of hypermetropia may still persist after myopic shift in some patients (3) . Some patients may continue to use spectacles for maintenance of good ocular alignment, despite a decrease in their hypermetropia. On the other hand, some patients may require spectacles for better vision to correct the hypermetropia that has not regressed. This retrospective study was designed to evaluate the patterns of change in ocular alignment and refractive error in patients with refractive accommodative esotro pia during followup.

METHODS
The protocol for this research project was approved by the Ethics Committee of our institution within which the work was undertaken, and conforms to the tenets of the Declaration of Helsinki in 1995 (as revised in Edinburgh in 2000). The medical records of 70 patients with refractive accommodative esotropia, who attended the Pediatric Ophthalmology Clinic between 2005 and 2017, were reviewed. The best corrected visual acuity, cycloplegic refraction, degree of distance and near de viations with and without glasses, presence of amblyopia (diagnosed by a decrease in best corrected visual acuity not attributable to any clinically detected structural abnormality of the eye with an underlying amblyogenic factor, including strabismus, anisometropia [defined as ≥1.5 D difference in spherical equivalent or cylinder]), were recorded at the initial and final visits. Patients with hypermetropia >2 D whose esodeviation was corrected with glasses were included. The exclusion criteria were: presence of neurological problems; mental and motor retardation; other ocular pathologies; comitant vertical strabismus; and incomitant strabismus. All patients un derwent visual acuity measurement using the Snellen chart, cycloplegic refraction with 1% cyclopentolate, ophthalmological examination (including orthoptic and biomicroscopic examination), Worth's fourdot test, and TNO stereoacuity test (if cooperative). Spherical equiva lent was calculated by summing the spherical error with half of the cylindrical error. The myopic shift for each eye was calculated by subtracting the final refractive error from the initial refractive error. In cooperative children, the deviation was measured at 33 cm and 6 m with prisms and an alternate cover test. The Krimsky test was used in uncooperative children. The cyclopegic re fraction and deviations were evaluated every 36 months.
The KolmogorovSmirnov test was used to evaluate the normal distribution of data. The pre and post treatment values were compared by a paired ttest, and differences between patients with refractory amblyo pia and those without amblyopia was evaluated using Student's ttest. The pvalues <0.05 denoted statisti cally significant differences.

RESULTS
The mean followup period for the 70 patients was 70 ± 36 months (mean age at admission: 6.01 ± 5.41 years; females: 60.6%). The means of best corrected visual acuity (logMAR), spherical refractive error of each eye, as well as distance and near deviations with glasses at admission and final visit are depicted in table 1. During the followup, bifocals (11.4%, n=8) and surgery (20%, n=20) were used in addition to full cycloplegic refractive correction to control the deviation.
At the final visit, the spherical equivalent (p=0.001), near (p=0.001), and distance (p=0.011) deviations with glasses were significantly decreased compared with the initial values obtained at admission.
The mean total myopic shift was 1.08 ± 1.35 D and 1.20 ± 1.40 D for the right and left eyes, respectively. During the mean followup of 5.83 years, the myopic shift/year was 0.22 D/year and 0.26 D/year for the right and left eyes, respectively.
Although the correlation of total myopic shift of the right and left eyes was high (r=0.77; p<0.0001) the cor relation for their rate of myopic shift was weak (r=0.18; p=0.15). The rate of change in near and distance devia tions with glasses was 1.71 ± 3.96 PD/year and 1.09 ± 3.25 PD/year, respectively. The correlation of the rate of change in near and distance deviations was strong (r=0.90; p<0.0001).
The rate of change in distance and near deviations was weakly correlated with the rate of myopic shift of the right and left eyes ( Table 2).
Amblyopia was detected in 24% of patients, while 50% of the patients did not have amblyopia at admis sion. All patients with amblyopia had treatment invol ving refractive correction and patching. At the final visit, 23% of the patients had refractory amblyopia. The total myopic shift, and change in distance and near deviations were not significantly different between patients with refractory amblyopia and those without amblyopia at the final visit (p>0.05) ( Table 3). The rate of myopic shift and rate of change in distance and near deviations were also not significantly different between these groups (p>0.05) ( Table 4).

DISCUSSION
Prompt treatment of deviation in intermittent and continuous esotropia is crucial for improving motor and binocular function (2) . It has been shown that fair and poor compliance with spectacle use greatly increases Arq Bras Oftalmol. 2020 -Ahead of Print    the risk of poor sensory and motor outcomes in children with pure refractive accommodative esotropia (1) . Thus, all these patients had full cycloplegic correction to con trol esodeviation. On the other hand, controlled studies could not be performed in this patient group who had full cycloplegic correction in early stages. Consequently, families usually experience anxiety and have questions related to the improvement of hypermetropia, as well as deviation.
In this study, early changes in deviation and refraction in this patient group with spectacles were evalua ted. The mean hypermetropia at admission (4.20 ± 2.06 D for the right eye and +4.65 ± 2.22 D for the left eye; mean age: 6.01 ± 5.41 years) was relatively low com pared with that measured in other studies involving younger patients (+5.03 ± 1.87 D and +5.12 ± 1.54 D for the right and left eyes, respectively) (4) . After a followup of 70 ± 36 months, the mean spherical equivalent was decreased to +3.12 ± 1.82 D and +3.28 ± 1.89 D for the right and left eyes, respectively. At near, the esode viation with glasses was decreased at a mean rate of 1.71 ± 3.96 PD/year; at distance, this rate was 1.09 ± 3.25 PD/year. The myopic shift/year for the right and left eyes was 0.22 D/year and 0.26 D/year, respectively.
Birch et al. reported a spherical equivalent of +5.67 ± 1.26 D in the lateonset accommodative esotropia group, in whom deviation was initiated at 1848 months of age (5) . They could not observe myopic shift in the late onset group; however, they reported a myopic shift of 0.43 D/year in patients with earlyonset accommodative esotropia that was initiated at <1 year of age. In their study, the total myopic shift in the earlyonset group was 1.45 ± 2.16 D.
The population of our study was similar to the late onset group of this previous investigation in terms of age distribution, and had a myopic shift/year of 0.22 D/year and 0.26 D/year for the right and left eyes, respectively. The followup period was longer in the previous study and the initial hypermetropia was higher (patients with moderate and high hypermetropia were included). The lower rate of myopic shift observed in our study compa red with that noted by Birch et al. may be explained by the lower initial hypermetropia recorded in our patients. Notably, there are also studies reporting continuous myopic shift (6) .
Birch et al. suggested that, in patients with anisome tropia, highly hypermetropic eyes have more myopic shift, while amblyopic eyes exhibit a slower rate of myopic shift (5) . On the other hand, Park et al. reported that amblyopic eyes showed greater decreases in sphe rical equivalent compared with nonamblyopic eyes. In addition, most patients in the hyperopic group showed a great decrease in hyperopia over time, supporting our findings. We could not find a significant difference between patients with and without amblyopia in terms of total myopic shift, distance and near deviations, and their rate of change (p>0.05) (Tables 3 and 4). Our followup period was shorter and the differences in po pulations may explain the different observations versus the study conducted by Birch et al. (5) .
Park et al. reported a mean annual decrease in sphe rical equivalent (during a 2year followup period) of 0.30 D/y for the right eye and 0.40 D/y for the left eye (7) . Some previous studies of Western cohorts reported that hyperopia remained unchanged or increased until 510 years of age and decreased thereafter (811) .
Several studies proposed that refractive correction in esotropia prevents emmetropization (12,13) . Köse et al. reported that full or partial correction of refractive error in accommodative esotropia does not affect the final refractive status (14) . The amount of emmetropization was relatively low in the study conducted by Köse et al. (in the partial correction group from 5.29 D to 4.9 D). On the contrary, Hutcheson et al. proposed that the baseline and final refractive errors were significantly lower in the children successfully weaned from spectacles (p=0.014). Of note, the children who were successfully weaned from spectacles were older at the time of initial diagnosis with accommodative esotropia (4.6 vs. 2.5 years) (15) , inconsis tent with the results reported by Birch et al. (5) .
Consequently, the dilemma of emmetropization and the effect of early correction of full cycloplegic correction persists in this patient group. However, significant change in deviation is obvious. Decreasing the amount of devia tion is important for motor and sensorial development (2) .
Another finding in our study is the low correlation of the rate of change in hypermetropia of the right and left eyes. This may imply that we may need to decrease hypermetropia in different degrees in each eye. Ho wever, this approach is usually not practical in clinical practice.
There were several limitations in this study. All pa tients with accommodative esotropia, including partially refractive accommodative esotropia (since some patients required surgery in addition to refractive correction), were included. The number of patients with different types of accommodative esotropia can be increased to evaluate changes in each type. Since most patients had low and moderate hypermetropia, the results cannot be extrapolated to those with high hypermetropia. Further more, we did not include patients with earlyonset (<1 year of age) accommodative esotropia. Several factors, such as age at which the use of spectacles was initiated and the decreasing rate of hypermetropic correction, may influence the final results which should be evalua ted in prospective studies.
Finally, it is important to inform families that refrac tion may decrease in these patients with accommodative esotropia and lowtomoderate hypermetropia at the initial visit, although to a lesser degree compared with near deviation. It is also important to independently evaluate each eye for reductions in the degree of hyper metropia as the treatment evolves and decreases in de viation since their rates do not appear to be correlated.