Optical quality assessment in patients with unilateral congenital ptosis: a matched case-control study

Background: To evaluate the differences in the corneal higher order aberrations (HOAs) and optical quality of the ptosis eyes compared with the normal fellow eyes in the unilateral congenital ptosis patients. Methods: A matched case-control study was performed in 17 pairs of eyes in 17 unilateral congenital ptosis patients. The ptosis eye was enrolled in the ptosis group while the normal fellow eye was enrolled in the normal group. The HOAs obtained from Pentacam HR and the optical quality parameters obtained from optical quality assessment system (OQAS) were compared between the two groups. Results: There were significant difference in vertical coma and vertical trefoil on the anterior corneal surface between the ptosis group and the normal group (p0.05). The OQAS parameters in the ptosis group were significantly different from that in the normal group (p0.05). The vertical fissure height (VFH) showed significant correlation with the total HOAs, 3 rd HOAs and vertical coma on the anterior corneal surface (p0.05). Conclusions: Compared with the normal fellow eyes, the ptosis eyes showed significantly degraded optical quality. The eyelid position was important for the changes of corneal HOAs in the ptosis eyes. The reasons for the discrepancy of the OQAS parameters between the ptosis eyes and normal fellow eyes were multiple.

concrete. Therefore, the preoperative evaluation of objective visual quality in ptosis eye is critical and could be an important approach to the visual function assessment.
Previous literatures have suggested that drooping upper eyelid pressure can cause changes in corneal curvature which played an important role on objective visual quality. Grey and Yap found the ocular with-the-rule astigmatism was significantly increased when the lid aperture was deliberately narrowed [4]. Buehren's study showed that the changes in corneal topography and HOAs were directly related to the force exerted by the lowered upper lid while reading [5].
Unlike the mechanism of upper eyelid drooping when reading or squeezing eye, the upper eyelid drooping and loosening in ptosis eyes are due to the loss of muscular or nerve function. Therefore, the influence of the upper eyelid pressure on cornea in ptosis eye might be different one. It might be mild but persistent compared with the conditions of squeezing or reading. To the best of our knowledge, there are little studies focusing on the objective visual quality changes in ptosis eyes.
Further studies to elucidate this issue would be desired.
In this study, we investigated the differences in the corneal higher order aberrations (HOAs) and optical quality assessment system (OQAS) parameters between the ptosis eyes and the normal fellow eyes. Furthermore, the correlation between the severity of ptosis and visual quality parameters was analyzed. To the best of our knowledge, this is the first study regarding on the investigation of the anterior and posterior corneal HOAs as well as OQAS parameters in the unilateral ptosis cases. Our work could help to further elucidate the causes of visual impairment in the ptosis patients and develop a new robust approach to evaluating the visual function in ptosis.

4
The study was approved by the Institutional Review Board. Written informed consent was obtained from legal parent/guardian for participants less than 18 years old and from participants directly for those 18 years or older. The tenets of the Declaration of Helsinki were followed for all study procedures.

Procedure
All subjects underwent the routine clinical assessments, topographic measurements, visual quality evaluations. All the measurements were completed by Jianqin Shen.

1.
Routine examinations involved uncorrected visual acuity, the best corrected visual acuity (BCVA), intraocular pressure, slit lamp examination, fundus examination and cycloplegic refraction (cycloplegic refraction measurement was performed after the examination of Pentacam HR and OQAS).

2.
Vertical fissure height (VFH) VFHs were obtained from the external eye photographs capturing by OCULUS Keratograph (Oculus, Wetzlar, Germany) (Figure 1). Visiometrics SL, Terrassa, Spain). The measurements were performed with a 4.0-mm central diameter, which was a standard size used in clinical double-pass studies. For uniform measurement, the refractive error of the eye was corrected by the spherical lens and the astigmatism was corrected by the cylindrical lens. All the measurements were conducted after blink twice in order to minimize the influence of tear film on the light scattering.

Corneal HOAs
The following parameters were measured to quantify the optical quality: (1) Objective scatter index (OSI): the ratio of the peripherally annular area light versus that of the central peak in the acquired double-pass image, which quantifies the intraocular scatter [6]. The lower OSI value indicates better optical quality.
(2) Modulation transfer function (MTF) cut-off: the cut-off point of the MTF curve on the x-axis, which is directly computed from the point spread function (PSF) [7]. The higher MTF cut-off value indicates better optical quality.
(3) Strehl ratio (SR): the ratio of peak focal intensities in aberrated PSF and ideal PSF, which ranges between 0 and 1.0. The larger SR indicates better optical quality [7,8].

Data Analysis and Statistics
All data were expressed as the mean ± standard deviation. Statistical analyses were performed using SPSS 22.0 software (SPSS, Chicago, IL, USA). The normality of all data samples was first checked using the Shapiro-Wilk test. When parametric analysis was possible, paired t-tests for paired data were used for the comparisons. When parametric analysis was not possible, Wilcoxon signed rank tests for paired data were used. For normal distribution, the correlation between variables was analyzed using Pearson bivariate correlation. For abnormal distribution, the correlation between variables was analyzed using Spearman bivariate correlation. The difference was considered statistically significant when the value is smaller than 0.05.

BCVA
The mean and standard deviations of best corrected logMAR visual acuity was 0.115±0.120 in the ptosis group and 0.017±0.038 in the normal fellow group. There was a significant difference in BCVA between the two groups(p0.05) ( Table 1).

Spherical equivalent (SE) and Astigmatism
The mean and standard deviations of SE was -2.07±3.61D in the ptosis group and -1.96±3.03D in the normal fellow group. There was no significant difference in SE between the two groups (p0.05) ( Table 1).
The mean and standard deviations of astigmatism was -0.94±0.95D in the ptosis group and -0.68±0.80D in the normal fellow group. There was no significant difference in astigmatism between the two groups (p0.05) ( Table 1).

VFH
The mean and standard deviations of VFH was 6.26±1.00mm in the ptosis group and 9.13±0.59mm in the normal group. There was significant difference in VFH between the two groups (p0.05) ( Table 1).

Corneal HOAs
The anterior corneal HOAs and Zernike coefficients (3 rd to 5 th order) were shown in Table 2.
There was no significant difference in total HOA RMS, 3 rd HOA RMS, 4 th HOA RMS and 5 th HOA RMS on the anterior corneal surface between the ptosis and the normal eyes (p>0.05) (Figure 2). There were significant differences in vertical coma ()vertical trefoil () on the anterior corneal surface between the ptosis and the fellow eyes (p<0.05) (Figure 3). The rest of the anterior corneal Zernike coefficients from the third to the fifth order showed no significant difference (p>0.05).
The posterior corneal HOAs and Zernike coefficients (3rd to 5th order) were shown in Table 2.
There was no significant difference in the total HOA RMS, 3rd HOA RMS4th HOA RMS and 5 th HOA RMS of the posterior corneal surface between the ptosis and the normal eyes (p>0.05) ( Figure 2). There were significant differences in vertical trefoil (), oblique trefoil (), vertical coma () and secondary astigmatismon the posterior corneal surface between the ptosis and the fellow eyes (p0.05) ( Figure   3). The rest of the posterior corneal Zernike coefficients from the third to the fifth order showed no significant difference (p>0.05).
[Please see the supplementary files section to access table 2.] Table 3 shows OQAS parameter comparisons between the ptosis group and the normal group.
The correlation between the OQAS parameters and VFH Table 5 showed that there was no significant correlation between the VFHs and the OQAS parameters (p>0.05).

Discussion
Objective visual quality is an important criterion for the visual function evaluation. The objective visual quality is mainly affected by aberrations, intraocular scatter and diffraction [12]. Aberration in ocular optical disorders has been well-documented in the previous studies. Eyelid position has been found to be an important factor contributing to the aberration changes [4,5,[13][14]. For objective visual quality changes in ptosis, nearly all studies have focused on the differences in lower-order aberration between ptosis eyes and normal eyes [15][16][17][18] and only a few researches have aimed at the HOAs changes in ptosis eyes. However, there are still no studies investigating the corneal HOAs and intraocular scatter changes in ptosis eyes.
Amblyopia and refractive errors have aroused much attention in ptosis over the past decades.
Amblyopia is one of the most common reasons for visual impairment in congenital ptosis, and the rate of amblyopia in congenital ptosis has been reported to be higher than the one in normal eyes [19].
Eyelid occlusion and significant refractive error were two leading causes of amblyopia in ptosis. Our results showed there was a significant difference in BCVA between the two groups, but no significant difference in spherical equivalent refraction. It can be speculated that the eyelid occlusion was the main cause of amblyopia in ptosis eyes. The association between BCVA and VFH were relatively weak and not statistically significant. This was probably because the degrees of most ptosis eyes in our study are moderate with the similar VFH values. As a result, the discrepancy of the pupil zone blocked by the ptosis eyes lid was too little to cause a significant different effect on amblyopia.
The pooled prevalence of myopia was much higher in ptosis population than the normal population [20]. The mechanisms of myopia in ptosis eyes are multifactorial and complicated. Form-deprivation by eyelid, defocused image on retinal and the eyelid pressure on eyeball might be the main causes [21][22][23]. Our results showed that there was no significant difference in refractive errors between the ptosis and the normal group, which was not in line with the literatures. The correlation between the refractive errors and the VFH were not statistically significant. This could also be ascribed to the little discrepancy of VFH in our patients.
The anterior corneal surface is a major contributor to the total HOAs of the eye. The effect on corneal shape of the eyelid mainly exerts on the anterior corneal surface. Thus we focused on the corneal HOAs in this study. In the present study, both the vertical coma and the vertical trefoil of the anterior corneal HOAs showed significant difference between the ptosis group and the normal group. Similar changes were found in other palpebral fissure narrowed conditions. Buehren's study showed corneal vertical coma and trefoil changed in both magnitude and direction after reading [5]. showed no significant difference between the two groups [24]. We speculated it was probably due to the different anatomical characteristics of eyelid between different races. Asian eyes have smaller vertical palpebral apertures, thicker upper eyelid and skin, all of which may contribute to the increased optical effects of the upper eyelid forces on cornea [25]. Changes in the ocular vertical coma while deliberately narrowing the palpebral fissure has been found significantly changed in our previous research [14], but the significance level is much stronger in present study. The reasons for the differences are multiple. Unlike the condition of deliberately squeezing eyelids, ptosis cases do not have an additional force from the contraction of muscle on the eyeball. In addition, the lid pressure distribution in congenital ptosis is uniform while the pressure distribution in deliberately squeezing state might be more localized. Moreover, congenital patients may develop head positions which could cancel the excess upper eyelid pressure to some degree.
Furthermore, we found that the VFH had a significantly negative correlation with the total HOAs RMS and the 3rd HOA RMS on the anterior corneal surface. The third-order HOAs are the dominant aberrations for most eyes in general. [26,27] Therefore, our results indicated that the eyes with severe ptosis may have a poorer optical quality induced by corneal HOAs. Moreover, there was a strong significantly correlation (p 0.01) between VFH and vertical coma (both magnitude and direction), which indicated that the eyelid position was the major factor impacting the corneal HOAs, particularly the vertical coma in ptosis eyes. The total HOAs, the 3rd HOA and the vertical coma might be a potential robust optical criterion to assess the visual function and the severity of the ptosis eyes.
It was also noticeable that the significant changes of HOAs in posterior corneal surface were observed in our ptosis group. The HOAs on posterior corneal surface have been rarely reported in either ptosis eyes or other narrowed palpebral fissure conditions. Vertical coma, vertical trefoil and horizontal trefoil and vertical secondary astigmatism were significantly different between two groups. The differences were basically in accordance with the anterior corneal HOAs, which indicated that the upper eyelid pressure might affect not only the HOAs on anterior surface but also the posterior surface. However, the exact and specific mechanisms were worthy to intensively investigate. We speculated that long-term and persistent pressure from the drooping eyelid might not only cause the instant distortion but also "remodel" the whole corneal contour, which involved the changes of posterior corneal surface. Therefore, HOAs on the posterior corneal surface should be an important factor to affect the ocular optical performance in ptosis.
The OQAS can evaluate the full information on aberration, diffraction and scattering [8,28]. OQAS has been successfully used to evaluate ocular optical quality in various eye diseases such as cataracts [29], pseudophakic eyes [11], dry eyes [30,31], corneal surface disorders [32], even macular disorders [33], and exhibited good repeatability. However, still no study was reported using the OQAS to assess the optical quality in ptosis eyes.
Our results showed significant higher OSI and lower MTF, SR, OV100%, OV20%, OV9% in the ptosis group, which indicated a comprehensive compromised optical quality for the ptosis eyes. Combined with the HOAs results, both the HOAs and intraocular scatter contributed to the degradation of the optical performances in ptosis eyes. There was no correlation between the VFH and neither of the OQAS parameters, which indicated that the reasons for poor optical quality in ptosis eyes might be multiple. The similar degree of VFH in our cases might be the major cause again here, as described above in the HOAs analysis. Nevertheless, the other factors, such as tear film, displacement of epithelial tissue might also be contributed to the reduction of objective visual quality in ptosis eye [34,35].
There were limitations in our study. The sample size was small and hence there was a possible bias in statistical conclusion. Tear film was also a factor affecting the HOAs in ptosis eyes, but was difficult to be measured for the ptosis eyes in the present study. Future studies regarding the tear film influence in the visual quality in ptosis eyes might be desired.

Conclusions
In conclusion, we found that the objective optical quality in the ptosis group was worse than that in the normal fellow group. Eyelid position is the critically important factor for the changes of HOAs on both the anterior and posterior corneal surfaces, particularly the vertical coma, in ptosis eyes. The   Comparisons of the Zernike coefficients (3rd to 5th order) of the anterior and posterior cornea surface between the ptosis group and the normal fellow group.

Figure 4
Comparisons of the OQAS parameter between the ptosis group and the normal group The correlations between VFH and total HOA RMS, 3rd RMS and vertical coma of anterior cornea surface. *p 0.05.

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download. Tables 2 and 4.docx