Exploring the Divergence Range for Stereopsis Maintenance With a Computer-Simulated Troposcope in Patients With Intermittent Exotropia

Abstract

Purpose: To investigate whether sensory input or motor signal of the extraocular muscle is the main activator of suppression in human intermittent exotropia (X(T)).

Methods: A case-control study was performed. Ten subjects with X(T) and 10 control participants were enrolled. The divergence range between both eyes when binocular vision was maintained was measured by using stereotests with a self-written computer program mimicking a troposcope. The break point, defined as the deviation angle at which stereopsis broke during eye deviation, was compared between the experimental and control groups by using a t test.

Results: The median near deviation angle in the experimental group was 42.5 prism diopters (PD) (mean, 44.5 ± 10.82 PD). The mean break point was 40.45 ± 10.79 PD in the X(T) group and 26.86 ± 2.62 PD in the control group (P = 0.003). The mean ratio of the break point to the near deviation angle was 0.92 ± 0.23 in the X(T) group, with the ratio close to 1 in 7 of 10 subjects.

Conclusions: Binocular vision can be maintained if similar images are projected onto corresponding retinas during the tropic phase of X(T). The antidiplopic mechanism in X(T) patients (i.e., suppression) is evoked by sensory input from binocular rivalry rather than by motor signal of the extraocular muscle.

Exotropia (XT) is a common type of strabismus in children, and its prevalence is approximately 1% in children aged younger than 11 years.¹ Studies in Asia and the United States suggested that intermittent exotropia (X(T)) is the most common subtype of XT, and the prevalence was 3.24% in a study in China.^1–3 In patients with X(T), one eye intermittently deviates outward, most often when the patient is tired, ill, under stress, or in particular test situations.⁴ These patients may complain of blurry vision, fatigue, and photophobia, but rarely diplopia, even in the exotropic stage. The mechanism that keeps the eyes in the phoric position in those patients is considered to be strong fusional convergence, which may spontaneously break and manifest as XT.^5,6 Accommodation convergence also plays some roles in the maintenance of ocular alignment in these patients.⁷ Previous studies revealed that the deviation angle increased by more than 10Δ in 23.1% of patients at 5 years and in 52.8% of patients at 20 years.⁸ Romanchuk et al.⁹ found that 17% of patients experienced deterioration to constant distance XT in 5 years. The relationship between age-related decline in accommodation and X(T) progression remain undetermined; however, some studies suggested excessive accommodation was associated with a higher prevalence of myopia in X(T) patients.¹⁰ 

One mechanism proposed to prevent diplopia is suppression, and may be similar in X(T) and constant XT. A previous study showed that suppression existed during eye deviation in XT patients, and may develop from the rivalry between dissimilar images falling onto corresponding retinas.¹¹ However, stereovision is often preserved in X(T) patients when eyes are well aligned, especially when patients focus at a near target.¹² This finding might suggest that either only hemifield suppression exists or that suppression does not occur when the eyes are in the orthotropic position. The question arises as to how patients switch to the suppression mode when the eye deviates outward. The possible signal inputs are motor and sensory signals. The former may originate from proprioception of the extraocular muscle or efferent signals from the brainstem or any brain region involved in motor planning and execution.¹³ On the other hand, the latter may be induced by detecting different images on corresponding retinas. 

Serrano-Pedraza et al.¹⁴ demonstrated that antidiplopic mechanisms in X(T) can be triggered by purely retinal information with identical images on noncorresponding retinas and do not require an oculomotor signal to indicate that deviation has occurred. The study was conducted in X(T) subjects with orthotropic status. In this study, we aimed to investigate if a purely oculomotor signal without binocular rivalry could induce suppression in X(T) subjects with exotropic status. A troposcope is a device that enables an examiner to project images onto both corresponding retinas during ocular vergence. We designed a computer program to simulate the function of a troposcope by continuously projecting half of a stereograph onto both foveae during the exotropic phase in our participants. If patients with X(T) were able to keep perceiving stereopsis when the divergence range became larger, we would propose that sensory input was the main activator of suppression, instead of motor signal of the extraocular muscle. 

Methods

Participants

This case-control study was approved by the institutional review board of the National Taiwan University Hospital, Taipei, Taiwan, and complied with the Health Insurance Portability and Accountability Act. Written informed consent was obtained prospectively from the subjects after an explanation of the nature and possible consequences of the study according to a protocol conforming to the tenets of the Declaration of Helsinki. Subjects diagnosed with X(T) who passed the random-dot stereogram 300″ (NTU 300″ stereogram)¹⁵ at the National Taiwan University Hospital were recruited for this study as the experimental group. Subjects with a deviation angle less than 30 PD, aged less than 20 years, with suboptimal stereopsis or diplopia, with corrected visual acuity less than 20/25, or with a combined neurological deficit were excluded. We included age-matched participants without strabismus or amblyopia and with normal stereovision as the control group. A total of 10 subjects with X(T) and 10 normal participants completed the study. We collected data including age, sex, refraction error, best-corrected visual acuity, and pupillary distance for each participant, who then underwent the NTU 300″ test for confirmation of stereopsis. Prism cover tests were performed to measure both the distance and near deviation angles in X(T) subjects. In the control group, divergence tolerance was measured by placing a base-in prism in front of one eye while the participant fixated at a distant target, and the power of the prism was increased gradually. Divergence tolerance was defined as the maximal prism power that the subject could still fuse, which represents an additional range of divergence ability from the parallel eye position. 

Study Design and Procedure

Anaglyphic random-dot stereograms were designed on a personal computer by one of the authors (AHW; Fig. 1). Left and right half-stereograms could be scrolled horizontally with the keyboard, mimicking the horizontal movement of troposcopic arms. In this study, the half-stereograms were scrolled uncrossed to test the divergence limit at which the participants could hold their stereopsis. The participants wore red-blue glasses and were requested to view the stereogram at a working distance of 30 cm from the monitor. At the beginning of the examination, the monitor showed a red-blue random-dot stereogram without any disparity between the half-stereograms, which was defined as the zero point. The participants were asked to recognize the hidden shape in a 4-alternative forced-choice paradigm (a square, circle, triangle, or diamond). The examiner gradually increased the divergence disparity of the half-stereograms by using the keyboard. Fusional divergence of the eyes was increased gradually along with the uncrossed scrolling of the half-stereograms until the participants could no longer recognize the stereogram, which was recorded as the “break point.” The true distance of divergence disparity between the half-stereograms was shown in millimeters on the monitor (Fig. 1). The divergent angle of exotropia at this limit was then calculated from this distance and the testing distance (break point in PD = divergence distance [cm]/testing distance [m]). 

Figure 1

 

Half-stereograms for the left and right eyes were 70.3 mm uncrossed displaced.

Statistical Analysis

Statistical analyses were performed by using a spreadsheet program (Excel 2013; Microsoft Corp, Redmond, WA, USA). The t-test was used to compare baseline characteristics between the control and experimental groups and the significance of the difference in mean break points between the groups. Values of P < 0.05 were considered statistically significant. 

Results

The mean age was 26.8 ± 4.57 years in the experimental group and 28.2 ± 1.69 years in the control group (P = 0.38). A total of 5 (50%) of the 10 subjects and 5 (50%) of the 10 participants were women in the experimental and control groups, respectively. There was no significant difference between the groups with respect to sex or age. The median near deviation angle in the experimental group was 42.5 PD (mean, 44.5 ± 10.82 PD), and all participants in the control group were orthotropic on the alternate cover test. 

The mean break point in the experimental group was 40.45 ± 10.79 PD (median, 41.33 PD), and significantly larger than that in the control group (26.86 ± 2.62 PD; P = 0.003; median, 27.55 PD). The demographic data and test results from both groups are listed in Tables 1 and 2, respectively. The mean ratio between the break point and near deviation angle was 0.92 ± 0.23 in the X(T) group. Of the 10 subjects with X(T), seven had a ratio close to 1, two had a ratio much less than 1, and one had a ratio much greater than 1 (Fig. 2). Some participants complained of moderate eyestrain after the program, and often required rest before starting the second round. 

Table 1

 

Data From X(T) Subjects in the Experimental Group

 

Table 2

 

Data From Subjects in the Control Group

 

Figure 2

 

Scatter plot of the experimental data in the X(T) group, with the near deviation on the x-axis and break point on the y-axis. The blue area indicates the range of the average break point plus twice the standard deviation in the control group.

The relationships between the distant X(T) angle from the parallel position and the break point in the experimental group are listed in Table 3. 

Table 3

 

Comparisons of the Distant X(T) Angle and BP in the Experimental Group

 

The divergence tolerance while subjects fixated at a distant target was approximately 6 to 8 PD in the control group. The measurement of the divergence range resulted from the summation of the convergence need from the interpupillary distance (interpupillary distance divided by the working distance) and divergence tolerance from the parallel eye position (Table 4). The mean ratio between the break point and the divergence range was 0.97 in the control group. The data in Table 3 and Table 4 are showed in Figure 3. 

Table 4

 

Comparisons of the Divergence Tolerance and BP in the Control Group

 

Figure 3

 

Scatter plot of the experimental data in the two groups: divergence range on the x-axis X(T): the summation of the (interpupillary distance/working distance) plus the distant XT angle; control group: the summation of the (interpupillary distance/working distance) plus the divergence tolerance); and break point on the y-axis.

Discussion

Suppression has been proposed as the mechanism that prevents binocular diplopia in strabismic patients. Several studies have proven the existence of suppression via psychophysical or electrophysiological methods.^11,14,16 Maurits et al.¹⁶ demonstrated the existence of suppression by recording the visual evoked potential in strabismic patients, both in esotropia and exotropia. The visual evoked potential response was decreased in the binocular state in all participants, and suppression was irrelevant to the extent of the amblyopia. 

Economides et al.¹¹ demonstrated the existence of alternate suppression by presenting purple stimuli to participants who manifested exotropia while wearing red-blue glasses. The subjective perception of the colors of the stimuli was related to the location of the stimuli, and temporal suppression has been found. The presence of double vision in secondary strabismus caused by disease or trauma implies that suppression might have developed early in life or from an inborn neural difference in primary strabismus. The primary visual cortex is where binocular integration occurs. In an animal study, Scholl et al. found that strabismus increased monocularity in simple cells in the primary cortex owing to the difference in excitation inputs from both eyes.¹⁷ Binocular suppression also increased, and was blocked by an intracortical injection of a GABA antagonist in strabismic cats.¹⁸ As suppression seemed to alternate and was intermittent in subjects with X(T), the activating mechanism should be dynamic and rapidly adjustable.¹⁹ 

Serrano-Pedraza et al. conducted an experiment in which the same image was projected onto non-corresponding retinas.¹⁴ In normal participants, if the crossed or uncrossed disparities exceeded the range of Panum's fusional area, diplopia occurred. However, subjects with X(T) did not experience diplopia for objects with large crossed disparities, even under orthotropic conditions. The result suggested that suppression may exist without ocular misalignment. However, strabismic eyes are physically deviated in patients with X(T) in reality, and we should reinforce this by demonstrating that suppression does not exist if only motor signals are present. 

In our experiment, binocular vision remained intact during the exotropic phase in 9 of 10 (90%) subjects with X(T) if images were projected onto the corresponding retinas. The definition was that these break points in the X(T) group exceeded the average plus twice the standard deviation of that in the control group, which was 32.10 prism diopters. Moreover, the mean ratio of the break point to the near deviation angle was 0.92 ± 0.23, and the ratio was close to 1 in 7 of 10 subjects. Therefore, our results demonstrated that, in our experimental setting, binocular vision could remain intact until nearly the maximum exotropic phase in subjects with X(T). Suppression did not occur if only motor signals revealed the deviation of eyes when we projected rivalry-free images onto corresponding retinas. Together with study by Serrano-Pedraza et al.,¹⁴ the results represent the sensory mechanism of suppression. The findings correlated well with the clinical observation that X(T) patients who undergo successful strabismus surgery usually show good and persistent binocularity after surgery, even though the proprioceptive feedback and motor state in the brainstem is not changed. The results further support the hypothesis that binocular rivalry suppression and strabismic suppression are related. 

The evocation of suppression may remove or greatly weaken convergence movement, which arises after the detection of disparity.²⁰ Accordingly, visual therapies for X(T) are aimed at overcoming suppression. Based on our results, suppression may be avoided by projecting the same image onto corresponding retinas, even when the eyes are misaligned. Therefore, prism treatment might serve as a modality to avoid suppression in small-angle X(T) in the early stage. Early surgery in the initial stage of X(T) might also be beneficial in avoiding the development of deep binocular inhibition; however, this requires further study for confirmation. 

Interestingly, we found that subjects with X(T) maintained binocular vision in a range similar to the near deviation angle, rather than to the distant deviation angle from the parallel eye position, which is where maximum ocular deviation theoretically occurs (Table 3; Fig. 3). Only one subject (subject 7 in the experimental group) could maintain stereopsis to the extent of the distant deviation angle. However, the control group could maintain binocular vision nearly to the range of divergence tolerance from the parallel position at distance (Table 4; Fig. 3). These findings suggest that the power of accommodative convergence, considered one of the most important mechanisms for maintaining a phoric position, cannot be overcome easily in patients with X(T). Ahn et al.⁷ proposed a similar conclusion that these patients manifested binocular inhibition because of an increased accommodative response during binocular vision. However, the flat screen used in the experiment might have caused image distortion with increasing disparity, which would increase the difficulties of fusing images and obtaining stereopsis. The break points might be underestimated in the X(T) group. 

The limitations of our study were that the computer monitor used was flat, not curved like fusional space in reality; and the small sample size. 

In conclusion, we determined that the antidiplopic mechanism in human X(T) (i.e., suppression), is evoked by different images projected onto the corresponding retinas rather than by motor signal of the extraocular muscle. More studies regarding how the signal input from different images elicits suppression might be needed to further elucidate the neural mechanism of X(T). 

Acknowledgments

Supported by the Ministry of Science and Technology, R.O.C. (NSC 101-2410-H-002-095 and NSC 102-2420-H-002-016-MY2). 

Disclosure: C.-H. Huang, None; A.-H. Wang, None; F.-R. Hu, None; T.-H. Tsai, None