Imbalanced Sensory Eye Dominance of Surgically Aligned Late-onset 1 Acute Acquired Concomitant Esotropes with normal stereopsis 2

Background : Adults with late-onset acute acquired concomitant esotropia (AACE) have 29 chance to develop normal binocular functions including a balanced ocular dominance before 30 the onset of esotropia. For most patients, strabismus surgery re-establishing the ocular 31 alignment indeed effectively restore stereopsis and visual acuity to the normal level. However, 32 it is unclear whether they have already acquired balanced two eyes. 33 Methods : 11 surgically aligned patients with AACE (24.3 ± 1.5 years; mean ± SE) and 14 34 adults with normal vision (26.1±1.2 years) participated in our experiments. All patients had 35 normal binocularity and stereopsis. Using binocular phase combination paradigm, sensory eye 36 dominance was quantified as the interocular contrast ratio, termed balance point, at which the 37 contribution of each eye to the perception of cyclopean grating were equal. 38 Results : Normal controls had a mean balance point value close to unity (0.95±0.01), while 39 AACE group exhibited evident binocular imbalance (0.76±0.05), which was significantly 40 different from control group (t (10.45) = -3.485, p = 0.006) . The balance point value didn’t 41 depend on the interval from AACE onset to strabismus surgery ( r = -0.357, p = 0.281 ) or the 42 interval from the surgery to examination of sensory eye dominance ( r = -0.105, p = 0.759) . 43 Conclusions : Although strabismus surgery effectively straightened AACE patients’ ocular 44 alignment and even conferred them normal stereopsis, late-onset AACE patients’ two eyes 45 were still not balanced. These results indicated that binocular imbalance might be a risk factor for 46 adult AACE.


51
Ocular alignment is not only necessary for cosmetic use but also clearly the first step to 52 provide conditions crucial for re-establishing full binocular function in strabismus. Although there 53 were several therapies for the treatment of ocular misalignment of strabismus such as refractive 54 correction(1), eye exercises(2), botulinum toxin therapy(3), surgical alignment(4) and so on, 55 strabismus surgery is a generally applied treatment especially for large deviation angle(5, 6). The 56 benefit of surgical alignment to various binocular function have been reported for several types of 57 strabismus. For example, better fusion was gained in adults with congenital esotropia after the 58 prescribed optical correction, if needed, for data collections. 107

Apparatus 108
Visual stimuli used for measurement of sensory eye dominance were generated by a Mac 109 computer running Matlab with PsychToolbox 3.0.9 extensions (22, 23) and displayed on a head 110 mounted goggles (goovis G1, OLED) with 1024*768 pixels resolution and a vertical refresh rate 111 of 60 Hz. The mean luminance of the OLED goggles was 160 cd/m 2 . Luminance nonlinearities of 112 the screen were corrected with an inverse gamma lookup table derived from careful calibration 113 with a photometer, checked or recalibrated before each experiment. 114

Stimuli 115
The luminance profiles of the grating presented to the dominant and nondominant eyes as shown 116 in Fig. 1 can be defined as following: 117 The phase combination paradigm used for measuring the eye dominance has been described 126 previously (24). As illustrated in Fig. 1A, observers were asked to view dichoptically two 127 horizontal sinusoidal gratings with equal but opposite phase (± 22.5º) and estimate the perceived 128 phase of the cyclopean grating. This process was repeated for various contrast ratios of dominant to nondominant eye to evaluate the interocular ratio where two eyes made equal contributions to 130 binocular combination. The ratio was termed "balance point" which quantified sensory eye 131 dominance. In our study, the contrast of the grating presented to the nondominant eye was fixed at 132 100%, and the following interocular contrast ratios were used: 0, 0.2, 0.4, 0.6, 0.8 and 1. 133 To eliminate any potential bias, two configurations ( Fig. 1B) were used for each interocular 134 contrast ratio: in one configuration, the phase shift is +22.5º for grating to dominant eye and was 135 -22.5º for grating to nondominant eye; In the other configuration, vice versa. The perceived phase 136 was defined as half of the difference between perceived phases in these two configurations. Each 137 configuration was repeated eight times and there were 8 Trials * 2 configurations * 6 interocular 138 ratios, with 96 trials in total for each participant. All conditions were randomly intermixed. 139 Participants normally finished the test in 25 to 30 minutes. 140 convergence till the images seen by two eyes were successfully combined into one steady 143 cyclopean image. After the convergence was confirmed by pressing specified key, only 144 surrounding high contrast frame was presented for 500 milliseconds. This was followed by the 145 binocular phase combination task. Participants were instructed to indicate the perceived phase of 146 cyclopean grating by moving a reference line to align it to the center of the dark stripe of the 147 grating. The line was presented horizontally on both sides of monocular grating, with its initial 148 vertical position (-9 to 10 pixels, relative to the center of the frame) randomly assigned in each 149 trial. The line was moved up and down one pixel every step which corresponded to 4-degree phase 150 angle of the sinusoidal grating. The stimuli were displayed continually till the end of trial. The 151 next trial started immediately after participants confirmed the position of reference line by 152 pressing specified key. Before the formal test, participants were allowed to get familiar with the 153 task in a 5 to 10-minutes practice session. 154

Data Analysis 155
The perceived cyclopean phases for each interocular contrast ratio were calculated as the average 156 of eight repeated measurements and fitted to a modified contrast-gain control model developed by where φ is the perceived phase; is interocular contrast ratio; is interocular phase 160 difference (In our study, it is 45°); and are two free parameters, ( balance point ) 161 representing for the interocular contrast ratio at which the two eyes contributed equally to the 162 binocular combination and representing for transducer nonlinearity in the gain control pathway. 163 All the model-fitting programs were implemented in Matlab (Mathworks, Inc., Natick, MA, USA) 164 using the nonlinear least squares method to minimized ∑( ℎ − ) 2 . The 165 goodness-of-fit was evaluated by: 166 Independent-sample t test was used to test whether there was significant difference in balance 168 point between surgically corrected AACE group and normal control group. Correlation analysis 169 were also conducted to identify factors relating to AACE patients' postoperative balance point. All 170 statistical computations were done in SPSS 13.0 (SPSS, Inc., Chicago, IL, USA). 171

172
In total, 11 surgically aligned AACE patients (6 male; age, mean ± SE, 24.3 ± 1.5 years) which 173 exhibited clinically normal stereopsis (stereo acuity was within 100 arc sec) and 14 normal adults 174 (9 male; 26.1±1.2 years) were included in our study. These two groups were matched in sex (χ 2 175 (1)= 0.244, p = 0.622) and age (t(23) = -0.931, p=0.429). The clinical details of each patients in 176 our study were provided in Table 1. Note that myopia is present for all patients except P8 and P10 177 who had emmetropic eyes (less than ±0.75 diopters). The refraction error on average was equal to 178 a spherical equivalent of -3.47 diopters (range, -6.75 to 0.25 diopters; OD) and -3.41 diopters 179 (range, -6.5 to 0.25 diopters; OS). All patients included was on average 22.7 years old (range, 14 180 to 30 years) when the esotropia manifested. Before surgery, the mean initial angle of esotropia was 181 33.6 prism diopters (range, 10 to 55 prism diopters) at near and 30.5 prism diopters (range, 10 to 50 prism diopters) at distance. Furthermore, 4 of 11 patients have equal near and distance 183 esotropia. In the remaining 7 cases, the differences were within 5 prism diopters. The 184 characteristics of most patients in our study met the diagnostic criteria of Bielschowsky type 185 AACE defined by previous investigators (13, 19, 25), which was described as occurrence in 186 adolescents and adults, varying degree of myopia and nearly equal angle of deviation at far and 187 near distance (14, 20). 188 The perceived phase of cyclopean image versus interocular contrast ratio (PvR) functions for 189 each surgically corrected patients with AACE and their average were shown in separate panels in 190

219
Here we conducted a cross-sectional cohort study to examine whether adults with AACE had 220 balanced sensory eye dominance since the eyes have already been straightened through strabismus 221 surgery. We quantified the sensory eye dominance by the effective contrast ratio of images 222 presented dichoptically to two eyes, at which each eye made equal contribution to the binocular 223 percept in a binocular phase combination task. Our results showed that post-operative patients 224 with AACE yet have unbalanced eyes even when they had successfully corrected ocular alignment 225 and clinically normal stereoscope. 226 Surgical alignment commonly benefits various binocular functions for several type of 227 strabismus (7-10). Recent studies(16, 26) showed that, after strabismus surgery, patients with 228 intermittent exotropia although exhibiting visual acuity and even normal stereo still have 229 imbalanced sensory eye dominance. It should be noted that binocular imbalance might be due to 230 residual part-time ocular deviation of juvenile patients with intermittent exotropia. In contrary to 231 congenital intermittent exotropia, AACE is an esotropia which occurred suddenly in adolescents 232 and adults (25, 27). Apparently, binocular vision had developed normally before the onset of 233 esotropia (13). Thus, the postoperative imbalance of eye dominance observed here unlikely 234 originated from abnormal early visual experience. It has been demonstrated that adults with 235 anisometropia tend to have unequal eye dominance(28). Indeed, the participants in our 236 experiments almost all have myopia, a typical feature of Bielshcowsky type AACE, but none has 237 anisometropia. Fawcett(29) proposed a critical window for misalignment in adults beyond which 238 recovery of binocular function is not possible. It seems that the long-term esotropia without 239 surgical correction might make the interocular imbalance incurable. However, our results showed 240 that the balance points did not depend on the interval from onset of AACE to strabismus surgery. 241 This is consistent with previous studies which found that the duration of misalignment did not 242 predict failure to recover stereo acuity (8, 10). In addition, several work (10, 30) also suggested 243 that the recovery of binocularity may take several months to occur. Nevertheless, we found that 244 the balance point was independent of the interval from surgery to the measurement of sensory eye 245 dominance. Specifically, the patient whose sensory eye dominance was examined even two years 246 after surgery still had a balance point of 0.63. Previous study(26) on intermittent exotropes gave 247 the same results, in which the author examined the sensory eye dominance first 0.5 months after 248 surgery and then 5 months after surgery and found no difference of binocular balance measured at 249 different time. 250 The phase combination task have been applied in identifying the abnormality of sensory eye 251 dominance in amblyopia(24), anisometropia(28) and strabismus(31). On assessing eye preference 252 in binocular view, the traditional measurement e.g. the hole-in-the-card test (32) and the 253 Worth-4-dot test (33) was convenient in clinical practice but only able to provide qualitative 254 outcome of test. Thus the method we adopted here to some extent could detect binocular deficit 255 that would be ignored in traditional crude test. Similar to binocularity and stereopsis, the binocular 256 balance as reflected in the phase combination task has a cortical basis (34-37). It has been 257 demonstrated that distinct binocular processes sharing a similar interocular contrast-gain control can also be determined quantitatively based on the paradigm used in phase combination task. 264 Although results of these measurement were mostly consistent, there were still some difference 265 possibly due to distinct cortical mechanism involved (43). It is of great value to investigate