- Email: [email protected]
Corneal higher-order aberrations in eyes with chronic ocular graft-versus-host disease
The Ocular Surface xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
The Ocular Surface journal homepage: www.elsevier.com/locate/jtos
Original Research
Corneal higher-order aberrations in eyes with chronic ocular graft-versushost disease Eisuke Shimizua,∗, Naohiko Aketaa, Hiroyuki Yazua,b, Miki Uchinoa, Mizuka Kamoia, Yasunori Satoc, Kazuo Tsubotaa, Yoko Ogawaa,∗∗ a
Department of Ophthalmology, Keio University School of Medicine, Tokyo, Japan Department of Ophthalmology, Tsurumi University School of Dental Medicine, Kanagawa, Japan c Department of Preventive Medicine and Public Health, Biostatistics at Clinical and Translational Research Center, Keio University School of Medicine, Tokyo, Japan b
A R T I C LE I N FO
A B S T R A C T
Keywords: Graft-versus-host disease Dry eye disease Higher-order aberration Visual acuity Irregular astigmatism
Purpose: Chronic ocular graft-versus-host disease (GVHD) is a long-term complication after hematopoietic stem cell transplantation (HSCT) and leads to irreversible visual morbidity due to severe ocular surface impairments including visual disfunction. However, knowledge about the optical function in chronic ocular GVHD is limited because it is difficult to assess quantitative optical function objectively. The development of anterior segment optical coherence tomography has allowed objective quantification of optical function by evaluating corneal higher-order aberrations (HOAs). Therefore, we applied this quantification in chronic ocular GVHD patients and verified the correlation between corneal HOAs and visual acuity. Methods: We retrospectively reviewed chronic ocular GVHD patients and the recipients after HSCT. Then, analyzed the relationship between visual function and the severity of chronic ocular GVHD. Results: The eyes of patients with chronic ocular GVHD had higher corneal HOAs than those of non-GVHD patients (HOAs; 0.481 ± 0.306 vs 0.254 ± 0.084, and 0.917 ± 0.609 vs 0.529 ± 0.130. P < 0.001, and 0.002. 4-mm and 6-mm diameters respectively. Corneal HOAs were correlated with the severity of chronic ocular GVHD (r = 0.436. P < 0.001). Moreover, corneal HOAs were correlated with visual acuity, especially in eyes with severe chronic ocular GVHD cases (HOAs; 4-mm r = 0.636. P = 0.036. Total 6-mm r = 0.871. P = < 0.001). Conclusions: We succeed to assess the objective value in the optical function of the chronic ocular GVHD. Quantification of corneal HOAs could be an objective evaluation to assess optical function in eyes with chronic ocular GVHD.
Introduction Chronic graft-versus-host disease (GVHD) is a major and lethal complication after hematopoietic stem cell transplantation (HSCT) for hematologic malignancies. For chronic GVHD in the eye, the main and the most common phenotype is dry eye disease (DED) with keratoconjunctivitis sicca [1]. The recipients are diagnosed as showing chronic ocular GVHD when the severe DED outbreak occurs after HSCT [2], and half of the recipients develop DED within 2 years after HSCT [3]. Although chronic ocular GVHD occurs in a large number of recipients, it typically does not lead to visual loss or blindness and the recipients generally have a good visual acuity [4]. Moreover, the revised criteria by the National Institutes of Health did not mention about
∗
any visual function loss in chronic ocular GVHD [5]. However, some studies have demonstrated that ocular involvement is a poor prognostic factor for mortality in GVHD [6,7]. Furthermore, it is known that ocular GVHD impairs recipients’ health-related quality of life [8,9]. For example, Saboo et al. demonstrated that chronic ocular GVHD patients often experience impairment of vision-related quality of life [10]. Pezzotta et al. showed similar results and claimed that it is important to recognize and treat all GVHD manifestations, including ocular involvement, because the total number and life expectancy of recipients treated with bone marrow transplantation has increased [11]. The Worldwide Network for Blood and Marrow Transplantation announced that by the improvements in the state-of-the-art technique, over 60,000 recipients receive HSCT per year [12]. Moreover, as the life
Corresponding author. Department of Ophthalmology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo, 160-8582, Japan. Corresponding author. Department of Ophthalmology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo, 160-8582, Japan. E-mail addresses: [email protected] (E. Shimizu), [email protected] (Y. Ogawa).
∗∗
https://doi.org/10.1016/j.jtos.2019.10.005 Received 6 August 2019; Received in revised form 4 October 2019; Accepted 8 October 2019 1542-0124/ © 2019 Published by Elsevier Inc.
Please cite this article as: Eisuke Shimizu, et al., The Ocular Surface, https://doi.org/10.1016/j.jtos.2019.10.005
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
conjunctivitis, blepharitis, ocular surface infectious disease, allergic conjunctivitis, ocular cicatricial pemphigoid, Stevens-Johnson syndrome, and Sjögren's syndrome), (2) eyes with other corneal diseases (ie, keratoconus, corneal dystrophies, bullous keratopathy, infectious keratitis, Pellucid marginal corneal degeneration, and corneal perforation), (3) history of corneal surgeries (ie, after laser in situ keratomileusis, radial keratotomy, and corneal transplantation), because a history of keratoconus, corneal surgery, or perforation can influence the corneal shape, and (4) presence of acute GVHD. On the basis of these exclusion criteria, 33 patients are excluded, and 97 patients remained in our study. Next, to confirm the eligibility of the cases for further visual function analysis, cases with a history of cataract, glaucoma, retinal disease, and optic nerve disease were thought to be inadequate. Thus, 7 patients were omitted from further analyses. Next, the patients were divided into 2 groups: (1) chronic ocular GVHD cases, and (2) non-GVHD (post-HSCT and not defined as chronic ocular GVHD) cases, according to the diagnostic criteria proposed by ICO GVHD Consensus group [19]. Finally, to adjust the effect of the intervention by patient background factors, propensity score matching was performed. Ultimately, 30 eyes of 15 GVHD cases and 30 eyes of the 15 non-GVHD cases were matched for the main analysis. In addition, 29 eyes of 20 age-matched subjects were included as normal controls.
expectancy continues to increase, the total number of survivors worldwide is estimated to be over 300,000 in 2030 [13]. Despite the importance of visual function in chronic ocular GVHD patients, previous studies only covered questionnaire-based analyses and did not include measurements with any quantitative examination. Our previous study used a novel calculation method for evaluating corneal higher-order aberrations (HOAs) in the transplanted cornea by using anterior segment optical coherence tomography (AS-OCT) [14]. This quantitative technique was applied to other corneal disorders, such as herpetic keratitis [15], corneal dystrophies [16], and corneal scars after perforation [17]. Moreover, this method demonstrated a strong correlation between the increase of corneal HOAs and decrease of visual acuity [15–18]. Therefore, we expect that the quantitative analysis technique such as corneal HOAs evaluation using AS-OCT could be a new optical function assay for chronic ocular GVHD patients. Although chronic ocular GVHD was said to be affected to the ocular surfaces such as the cornea and conjunctiva [1], the exact cause of the visual impairment is still unknown. Therefore, we hypothesize that corneal HOAs are deteriorated in the chronic ocular GVHD eyes and the corneal HOAs are correlated with visual acuity as the several corneal diseased which we demonstrated in the past [14–18]. Hence, the aim of this study is to elucidate the relationship between the severity of chronic ocular GVHD and quantitative measures of visual and optical function, such as visual acuity and corneal HOAs.
Data analysis Routine examinations, including slit-lamp microscopy, clinical ocular evaluation, fundus examinations, and best corrected visual acuity (BCVA) measurement, were conducted at the time of diagnosis for all patients. Clinical ocular parameters such as tear film breakup time (TFBUT), corneal fluorescein staining score (CFS), corneoconjunctival lissamine green staining score (LG), degree of the meibomian gland dysfunction (MGD), degree of fibrosis, degree of filamentosa, and degree of hyperemia were examined by multiple DED specialists (E.S., Y.O., M.U., and M.K.) according to the diagnostic criteria proposed by the ICO GVHD Consensus group [19]. Schirmer's test (Sterilized Tear Production Measuring Strips. 4987896590227. Ayumi Pharmaceutical Corporation. Tokyo, Japan) was performed prior to ocular evaluations to avoid the effect of the fluorescein/lissamine green solution. In addition, during several examinations, participants answered the Ocular Surface Disease Index (OSDI) questionnaire. Visual acuity was measured using a standard Snellen chart, and the BCVA with spectacle correction was recorded. The results were measured in decimal acuity and converted to logarithm of the minimal angle of resolution (logMAR) units using a visual acuity conversion chart.
Materials and methods This retrospective study adhered to the tenets of the Declaration of Helsinki. All procedures were performed in compliance with the protocol (#20170350) approved by the institutional ethics review board of Keio University School of Medicine (Tokyo, Japan). Ethical guidelines for clinical studies from the Japanese Ministry of Health, Labour and Welfare indicate that for studies that do not involve biological tissue and which involve reviewing medical records retrospectively, researchers do not need to obtain written informed consent from each patient. On the basis of the guidelines of ethics committees, we have provided detailed written guidelines and the ethical statement of the present study on the website [20]. Informed consent was obtained by notices including the following factors; background of the study, purpose of the study, study design, privacy policy, freedom to withdraw, inclusion and exclusion criteria, the factors assessed in the medical records, advantages and disadvantages of participating the study, disclosure of the data, presentation of the data at a conference or in a journal, and contact information. Patient data were anonymized before access and/or analysis.
Anterior segment optical coherence tomography
Study design
Eyes in all the cases were routinely examined using anterior segment optical coherence tomography (AS-OCT). All subjects were examined until at least two sets of sufficient images were obtained. Sixteen rotating AS-OCT scans were used to reconstruct three-dimensional models of the entire corneal structure. The CASIA system (SS1000, Tomey, Nagoya, Japan) corrected distortions in the AS-OCT images based on the refractive index of the anterior surface. Two corneal specialists (N.A. and H.Y.) carefully checked all AS-OCT images to ensure that the surface digitalization recognized by the automated inbuilt software was correct. Zernike coefficients were calculated using Zernike analysis as previously reported [21]. In brief, the anterior and posterior corneal surfaces were reconstructed as a three-dimensional model from the corneal height data. The anterior, posterior, and total corneal aberrations at diameters of 4 mm and 6 mm were calculated separately with the installed ray tracing software, version 5.1. The refractive indices of the cornea and aqueous humor were set to 1.376 and 1.336, respectively. The wavefront aberration was expanded with normalized Zernike polynomials up to the 8th order. HOA was defined as the root mean square (RMS) of the 3rd-to the 8th-order Zernike
This retrospective study was performed according to the flowchart in Fig. 1. The Dry Eye Disease (DED) specialists (E.S. and Y.O.) and the corneal specialists (N.A. and H.Y.) carefully reviewed the medical records of all patients who visited the DED outpatient clinic in the Keio University hospital from April 2018 to March 2019. We included Japanese adult males and females (over 20 years of age) who underwent hematopoietic stem cell transplantation (HSCT) at the Department of Hematology, Keio University hospital between April 1995 and December 2017 and presented with new onset of chronic ocular GVHD after HSCT. Chronic ocular GVHD was defined on the basis of the diagnostic criteria proposed by the International Chronic Ocular GVHD Consensus Group [19]. Also, the severity of the chronic ocular GVHD was defined by the quantitative scale from 0 through 11, based on the GVHD-related DED severity score (ICO severity score) [19]. In total, 130 patients conformed to the inclusion criteria. Next, patients who showed at least one of the following findings were excluded: (1) common findings of other ocular surface diseases (ie, 2
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Fig. 1. Study chart. The retrospective study was performed at the dry eye clinic in the Keio University hospital from April 2018 to March 2019.
coefficients. Spherical aberration (SA) was defined as the RMS of Z40 (spherical aberration) and Z60(secondary spherical aberration). Coma aberration (COMA) was defined as the RMS of Z3−1 and Z31 using the same method as in previous studies [21].
deviation (SD). P-values < 0.05 were considered to indicate statistical significance.
Results Statistical analysis Patient demographics The data were analyzed using Prism software (ver. 6.04 for Mac; GraphPad Software Inc, San Diego, CA, USA) and Excel for Mac (ver. 15.32; Microsoft Corporation, Redmond, WA, USA). Sample size was difficult to define because there was no past reference that evaluated HOAs of the chronic ocular GVHD patients. Therefore, the HOA data of patients with similar ocular surface diseases such as Stevens-Johnson syndrome was selected for the sample size calculation [22]. A total sample size of 18–19 was calculated to be adequate based on the effective size (1.10), statistical power (0.95), and the significance level (0.05). The D'Agostino and Pearson omnibus normality test was used to assess whether the data showed a normal distribution. To compare differences in clinical parameters between the GVHD cases and the nonGVHD cases, unpaired t-test or Mann-Whitney U test was used. To reduce the effect of patient selection bias and potential confounding effects, we performed adjustments for differences in baseline characteristics by means of propensity score matching. Patient selection was performed by using the propensity score matching method with the Greedy 5-to-1 digit-matching algorithm for baseline characteristics, i.e., age, gender, and post-HSCT duration at baseline. To compare the differences in HOAs, SA, and COMA among the GVHD cases, the nonGVHD cases, and the controls, Dunn's multiple comparison test was used. The HOAs were transformed into natural logarithm values for the correlation analysis including HOAs, visual acuity, and the ICO GVHD Consensus Group Proposed Severity Scale (ICO severity score). Then, Spearman's correlation coefficients were determined. For the correlation analysis between HOAs and visual acuity according to the ICO severity score (severe; n = 11, moderate; n = 24, and none; n = 25), Pearson's correlation coefficient and Spearman's correlation coefficient were selected. To exclude the bias which using bilateral eyes in a single analysis, we compared the ipsilateral eyes between the GVHD and the non-GVHD eyes. All data were expressed as the mean ± standard
Mean ages were 54.20 ± 11.33 years (range, 29–70 years) in the GVHD group and 52.93 ± 10.02 years (range, 33–70 years) in the nonGVHD group. There was no significant intergroup difference in patient age (P = 0.748. Table 1). Mean follow-up duration after HSCT was 1689 ± 1032 days (range, 199–3695 days) in the GVHD patients and 1773 ± 1671 days (range, 205–5374 days) in the non-GVHD patients, with no significant intergroup difference in the mean follow-up duration (P = 0.869. Table 1). Mean ICO severity score was 8.00 ± 1.89 in the GVHD patients and 2.93 ± 1.70 in the non-GVHD patients, with the difference being significant (P < 0.001. Table 1). Mean TFBUT was 3.13 ± 1.73 in the GVHD patients and 5.53 ± 1.85 in the non-GVHD patients. The GVHD patients showed a significantly shorter TFBUT (P = 0.001. Table 1). The GVHD patients also showed significantly worse phenotypes with respect to the LG score, MGD grade, degree of fibrosis, and degree of filamentosa (GVHD vs non-GVHD, P value: LG score, 3.87 ± 2.07 vs 0.47 ± 0.64, P < 0.001; MGD grade, 2.13 ± 0.83 vs 1.27 ± 0.80, P = 0.007; degree of fibrosis, 1.07 ± 0.80 vs 0.33 ± 0.49, P = 0.012; degree of filamentosa, 0.27 ± 0.46 vs 0.00 ± 0.00, P = 0.032; Table 1) Mean visual acuity was −0.02 ± 0.12 in the GVHD patients and −0.08 ± 0.02 in the non-GVHD patients (logMAR). The GVHD patients had significantly worse visual acuity (P = 0.004. Table 1). However, there were no statistically significant differences in the corneal pachymetry and intraocular pressure between the GVHD patients and the non-GVHD patients (corneal pachymetry, 542.4 ± 37.8 vs 540.8 ± 31.3, P = 0.631; intraocular pressure, 13.87 ± 2.86 vs 13.13 ± 2.99, P = 0.782). Moreover, in the analysis using ipsilateral eyes, estimated value differences of ICO score, TFBUT, CFS, LG, MGD, fibrosis, filamentosa, and hyperemia were similar to the result of the bilateral eye analysis. (Supplementary Table 1 and Table 1). 3
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
and 6-mm diameters (P < 0.05. Table 2 and Supplementary Fig. 1). Moreover, for COMA, statistically significant differences were only shown between the GVHD patients and the non-GVHD patients in total and anterior COMA at the 4-mm diameter, and between the GVHD patients and the normal controls in total COMA at the 6-mm diameter (P < 0.05. Table 2 and Supplementary Fig. 2). Moreover, we analyzed the corneal HOAs in the same lateral eyes, the eyes with the GVHD had more significantly high HOAs compared to the non-GVHD in both right or left eyes (Supplementary Table 2).
Table 1 Demographic of the subjects.
Eyes, n Cases, n Male Female Age, years old * Post HSCT, days* ICO severity score, points* OSDI, points** Schirmer's test, mm* TFBUT, seconds** CFS, points** LG, points* MGD, points* Fibrosis, points** Filamentosa, points* Hyperemia, points** Corneal pachymetry, μm** Visual Acuity, LogMAR** Intra Ocular Pressure, mmHg**
GVHD
nonGVHD
P Value
30 15 10 5 54.20 ± 11.33 1689 ± 1032 8.00 ± 1.89 38.64 ± 29.84 4.13 ± 3.56 3.13 ± 1.73 5.07 ± 2.31 3.87 ± 2.07 2.13 ± 0.83 1.07 ± 0.80 0.27 ± 0.46 0.93 ± 0.59 542.4 ± 37.8 −0.02 ± 0.12 13.87 ± 2.86
30 15 9 6 52.93 ± 10.02 1773 ± 1671 2.93 ± 1.70 9.44 ± 13.84 7.13 ± 4.93 5.53 ± 1.85 0.60 ± 0.91 0.47 ± 0.64 1.27 ± 0.80 0.33 ± 0.49 0.00 ± 0.00 0.13 ± 0.35 540.8 ± 31.3 −0.08 ± 0.02 13.13 ± 2.99
0.748 0.869 < 0.001 0.001 0.066 0.001 < 0.001 < 0.001 0.007 0.012 0.032 0.001 0.631 0.004 0.782
Correlations between corneal HOAs and visual acuity Fig. 3 shows the correlation between HOAs and visual acuity. There was a statistically significant positive correlation between HOAs and visual acuity in all HOAs, including total, anterior, and posterior HOAs at the 4-mm and 6-mm diameters. (HOAs vs visual acuity: total 4 mm, r = 0.518, P < 0.001; total 6 mm, r = 0.430, P < 0.001; anterior 4 mm, r = 0.516, P < 0.001; anterior 6 mm, r = 0.442, P < 0.001; posterior 4 mm, r = 0.385, P = 0.003; and posterior 6 mm, r = 0.337, P = 0.011; Fig. 3).
Data were expressed as the mean ± standard deviation (SD). *Unpaired t-test.**Mann–Whitney test. HSCT: hematopoietic stem cell transplantation. ICO severity score: International Chronic Ocular GVHD. Consensus Group Proposed Severity Scale. OSDI: ocular surface disease index. TFBUT: Tear Film Breakup Time. CFS: Corneal Fluorescein Score. LG: Lissamine Green staining score. MGD: Meibomian Gland Dysfunction.
Correlations between ICO severity score and visual acuity Fig. 4 shows the correlation between ICO severity score and visual acuity. There was a moderate positive correlation between ICO severity score and visual acuity (ICO severity score vs visual acuity; r = 0.436, P < 0.001.).
Corneal higher-order aberrations, spherical aberrations, and coma aberrations
Correlations between corneal HOAs and ICO severity score
Table 2 and Fig. 2 show the mean HOAs, SA, and COMA of Zernike terms at 4-mm and 6-mm diameters in each group. The total and anterior HOAs at both 4-mm and 6-mm diameters in GVHD patients were significantly higher than those in the non-GVHD cases and the healthy controls (P < 0.05; Table 2 and Fig. 2). However, no differences were observed in posterior HOAs between the groups. This trend was only shown in the HOAs and not in the SA and COMA. For SA, statistically significant differences were only shown between GVHD patients and the healthy controls in total and anterior SA at both 4-mm
Fig. 5 shows the correlation between HOAs and ICO severity score. There was a statistically significant positive correlation between HOAs and ICO severity score in all HOAs, including the total and anterior HOAs at 4-mm and 6-mm diameters (HOAs vs ICO severity score: total 4 mm, r = 0.594, P < 0.001; total 6 mm, r = 0.565, P < 0.001; anterior 4 mm, r = 0.594, P < 0.001; anterior 6 mm, r = 0.581, P < 0.001; Fig. 5). There was a weak positive correlation between HOAs and ICO severity score in posterior HOAs at the 4-mm and 6-mm diameters (HOAs vs ICO severity score: posterior 4 mm, r = 0.353,
Table 2 Comparison of the corneal higher-order aberrations, spherical aberrations, and coma aberrations. n
GVHD
nonGVHD
Normal
P value*
30
30
29
GVHD vs nonGVHD
GVHD vs Normal
nonGVHD vs Normal
< 0.001 0.002 < 0.001 0.001 0.114 0.399
0.025 0.017 0.022 0.015 > 0.999 > 0.999
0.941 > 0.999 0.928 > 0.999 0.490 0.070
HOA
Total 4 mm Total 6 mm Anterior 4 mm Anterior 6 mm Posterior 4 mm Posterior 6 mm
0.481 0.917 0.506 0.959 0.060 0.153
± ± ± ± ± ±
SA
Total 4 mm Total 6 mm Anterior 4 mm Anterior 6 mm Posterior 4 mm Posterior 6 mm
0.027 ± 0.189 0.124 ± 0.214 0.031 ± 0.192 0.146 ± 0.213 −0.021 ± 0.010 −0.112 ± 0.026
0.124 ± 0.073 0.225 ± 0.144 0.130 ± 0.071 0.243 ± 0.144 −0.024 ± 0.009 −0.110 ± 0.019
0.158 ± 0.067 0.257 ± 0.108 0.168 ± 0.068 0.273 ± 0.113 −0.026 ± 0.009 −0.104 ± 0.031
0.415 0.136 0.411 0.160 0.746 > 0.999
0.004 0.006 0.002 0.012 0.582 0.865
0.250 0.811 0.145 > 0.999 > 0.999 > 0.999
COMA
Total 4 mm Total 6 mm Anterior 4 mm Anterior 6 mm Posterior 4 mm Posterior 6 mm
0.187 0.411 0.190 0.423 0.014 0.045
0.079 0.245 0.082 0.265 0.013 0.036
0.110 0.209 0.112 0.229 0.018 0.050
0.015 0.564 0.015 0.828 > 0.999 0.311
0.178 0.032 0.181 0.078 0.214 > 0.999
> 0.999 0.640 > 0.999 0.753 0.066 0.255
± ± ± ± ± ±
0.306 0.609 0.307 0.597 0.023 0.037
0.189 0.425 0.188 0.433 0.010 0.036
0.254 0.529 0.271 0.567 0.049 0.141
± ± ± ± ± ±
± ± ± ± ± ±
0.084 0.130 0.087 0.140 0.011 0.018
0.043 0.121 0.044 0.129 0.007 0.026
0.280 0.560 0.297 0.605 0.054 0.155
GVHD: Graft-versus-host disease. HOA: higher-order aberration. SA: Spherical aberration. COMA: Coma aberration. *Kruskal-Wallis test. 4
± ± ± ± ± ±
± ± ± ± ± ±
0.096 0.149 0.093 0.160 0.015 0.024
0.106 0.148 0.104 0.156 0.010 0.032
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Fig. 2. Corneal higher-order aberrations in the GVHD, non-GVHD, and normal eyes. The anterior HOAs are higher at both 4-mm and 6-mm diameters in the GVHD patients compared to those in the non-GVHD cases and the healthy controls (all P < 0.05). In contrast, no differences were observed in posterior HOAs between the groups.
r = 0.616, P = 0.044; anterior 6 mm, r = 0.859, P < 0.001; Table 3). However, there was no correlation in other measurements.
P = 0.006; posterior 6 mm, r = 0.294, P = 0.023; Fig. 5). Table 3 shows the results of the correlation analysis between HOAs and visual acuity according to the ICO severity score (severe, n = 11; moderate, n = 24; and none, n = 25). There was a strong positive correlation between HOAs and visual acuity in total, and anterior HOAs at 4-mm and 6-mm diameters (HOAs vs visual acuity; total 4 mm, r = 0.636, P = 0.036; total 6 mm, r = 0.871, P < 0.001; anterior 4 mm,
Discussion In the current study, patients with chronic ocular GVHD had higher corneal HOAs than the non-GVHD patients and the normal controls 5
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Fig. 3. Correlation between corneal higher-order aberrations and visual acuity. Statistically significant positive correlation was shown between corneal HOAs and visual acuity in the eyes with the GVHD and the nonGVHD groups. (n = 30, all P < 0.05). The strong correlation was seemed between total and anterior 4-mm cornea and the visual acuity (r = 0.518 and 0.516).
compared to the normal. Unlike these whole corneal morphological changes caused by infections, inflammations [8], degeneration, and trauma, chronic ocular GVHD affects the ocular surface through inflammation and cicatricial change [1]. Therefore, it is conceivable that the surface impairment caused by chronic ocular GVHD increases corneal HOAs, especially in the anterior and total cornea (Fig. 2, Table 2).
(Fig. 2, Table 2). Similar trends were observed in SA (Supplemental Fig. 1) and COMA (Supplemental Fig. 2). However, these trends were not as specific as the differences observed in HOAs. Several corneal and ocular surface disorders such as infectious keratitis [15,21], corneal dystrophies [16], corneal perforations [17], and Stevens-Johnson's syndrome [22] showed significantly higher HOAs, SA, and COMA
6
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
and increases only anterior and total HOAs, but not the posterior HOAs. With regard to the application of our findings in the clinical settings, the reduction of corneal HOAs is known to contribute to an increase in patients' visual acuity. Rigid gas-permeable contact lenses (RGPCL) are frequently used to increase the visual acuity in several corneal diseases. We previously reported that the visual acuity recovers with the use of RGPCL in eyes with high corneal HOAs due to the offset of the anterior HOAs [21,27,28]. Interestingly, our demographic case (Fig. 6) shows a similar result. Despite the high ICO severity scores (10/11 points), RGPCL was effective in improving visual acuity (logMAR 0.6 to 0.1, before and after wearing RGPCL). RGPCL can only correct anterior HOAs; therefore, residual HOAs such as posterior HOAs cannot be corrected and the visual acuity will not improve with an irregular posterior surface. As we hypothesized in the past, chronic ocular GVHD mainly impairs anterior HOAs. Therefore, RGPCL could be a good option to improve the patients' visual acuity. However, the eyes with severe chronic ocular GVHD has a low affinity to RGPCL due to several RGPCL-related complications. Therefore, instead of RGPCL, a gas-permeable scleral lens with a larger diameter will be the choice. The scleral lens has been effective in mitigating subjective symptoms and healing ocular surface by a liquid bandage in chronic ocular GVHD eyes [29]. Furthermore, Magro L. et al. demonstrated that the scleral lenses are effective in the severe ocular GVHD cases with keratoconjunctivitis sicca [30]. Moreover, recent evidence shows that the effect of the scleral lens to the corneal HOAs had no significant differences with the user of RGPCL [31]. Therefore, scleral lenses could be a good option to increase visual acuity in chronic ocular GVHD patients. To optimize the efficacy of the RGPCL as well as the scleral lens, information regarding the HOAs is essential to predict the visual acuity after the lens correction [18]. Examination of visual acuity is a subjective process and requires time. In contrast, corneal HOAs evaluation using AS-OCT is an objective technique with a total duration of only a few seconds [32]. We believe that subjective plus objective measurements would cooperate in improving the accuracy in visual function examinations. We found that there is a typical topographical pattern of the chronic ocular GVHD cases represented in Fig. 6. The topographical pattern shows a lower part of the paracentral cornea flattering which often develops corneal thinning and leads to perforation [1]. Interestingly, 63.6% (7 out of 11 eyes) of the severe chronic ocular GVHD eyes develops this topographic pattern. Moreover, comparing HOAs between the lower flattening pattern and the other pattern, anterior and total HOAs were higher in the lower flattening pattern eyes (Table 5). Similar topographical patterns were reported in the various corneal diseases which might be related to the pathophysiology of the causative disease [15,17,18,21]. Therefore, evaluation of the topographical patterns in the chronic ocular GVHD cases could be an important imaging biomarker for the additional evaluation. There are several limitations of the current study. First, we did not evaluate complete visual and optical parameters such as corneal scatter [33] contrast sensitivity [34], or functional visual acuity [35], which may be affected by DED. Therefore, future studies with more exhaustive visual and optical function assessments are required. Second, this was a retrospective, cross-sectional study that evaluated corneal HOAs, visual acuity, and ICO severity scores. Although we did perform propensity score matching to minimize the effect of selection bias, we did not evaluate the effect of systemic treatment or/and perform systemic evaluation in other chronic GVHD target organs. Thus, assessment of systemic management will be essential in further studies. Moreover, we included bilateral eyes in several analyses which might cause bias. To exclude this bias, we performed an additional analysis using ipsilateral eyes. The result of the comparative analysis of the ocular parameters (Supplementary Table 1) and the HOAs between the GVHD and the non-GVHD eyes (Supplementary Table 2), were similar with the bilateral eyes analysis (Table 1 and Fig. 2). Third, this study focused on chronic GVHD but not on acute GVHD. The mechanisms underlying
Fig. 4. Correlation between visual acuity and the International chronic ocular GVHD consensus group proposed severity scale. Statistically significant positive correlation was shown between ICO severity score and visual acuity in the eyes with GVHD and the non-GVHD group (n = 30, P < 0.001). The correlation coefficient was moderate (r = 0.436).
A healthy ocular surface including the tear film maintains a smooth refractive surface for good quality of vision, and the surface regularity in the central part of the cornea over the pupil is especially important in terms of visual function [23]. Kaido et al. demonstrated that the HOAs were higher in the eyes with superficial punctate keratitis (SPK) than in those without SPK, and HOAs were correlated with the severity of the epithelial damage [24]. Thus, we additionally analyzed the correlation between the corneal HOAs and CFS in terms of corneal surface impairment to check the influence of the epithelial damage. HOAs and CFS showed a significant positive correlation (total 4 mm, total 6 mm, anterior 4 mm, and anterior 6 mm; Table 4), which was similar to the correlation between HOAs and ICO severity score (Fig. 5). The explanation for this phenomenon is that the ICO severity score reflects the combined grade of multiple subjective and objective scales such as CFS, Schirmer's test, OSDI, and degree of hyperemia [19]. Therefore, CFS could influence the entire combined grading. However, the correlation coefficient between the ICO severity score and HOAs was stronger than that between CFS and HOAs (Table 4, Fig. 5). Thus, we can conceive that a combined scale such as the ICO severity score is more accurate than CFS and that the epithelial damage is not the only factor increasing the HOAs in eyes with chronic ocular GVHD. In our retrospective study, moderate correlations were observed between HOAs and visual acuity and between visual acuity and the ICO severity score (Figs. 3 and 4). Interestingly, the correlation coefficient between HOAs and visual acuity became much stronger only in the severe chronic ocular GVHD eyes (Table 3). This result suggests that none to moderate chronic ocular GVHD only shows a minimal effect on the recipient's visual function, as reported previously [10]. However, with more severe chronic ocular GVHD, the surface damage impairs the recipient's eyes to a greater degree, increases the corneal HOAs, and finally decreases the visual acuity irreversibly. Moreover, this trend was only seen in the total and anterior HOAs, but not in the posterior HOAs. This can be explained by the pathophysiology of chronic ocular GVHD, which is mediated by severe DED and keratoconjunctivitis sicca [1]. In the morphological evaluations, immune cells in the epithelia have been shown to play a role in inflammatory ocular diseases [25]. He J et al. demonstrated that the density of immune cells is higher in chronic ocular GVHD human corneas by using confocal microscopy [26]. Considering the pathophysiology of chronic ocular GVHD and the findings of previous studies, surface and epithelial impairments could be a major explanation for this finding. Thus, it is speculated that chronic ocular GVHD mainly impairs the anterior segment of the cornea 7
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Fig. 5. Correlation between corneal higher-order aberrations and the International chronic ocular GVHD consensus group proposed severity scale. Statistically significant positive correlation was shown between corneal HOAs and ICO severity score in the eyes with the GVHD and the non-GVHD group (n = 30, all P < 0.05). The strong correlation appeared to be between the total and anterior cornea HOAs and the visual acuity (total 4 mm, r = 0.594; total 6 mm, r = 0.565; anterior 4 mm, r = 0.594, and anterior 6 mm, r = 0.581).
8
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Table 3 Correlation between corneal higher-order aberrations and Visual Acuity by the International Chronic Ocular Graft-versus-host disease Consensus Group Proposed Severity Scale.
Severity Severea (n = 11) Moderateb (n = 24) Noneb (n = 25) a b
Total 4 mm
Total 6 mm
Anterior 4 mm
Anterior 6 mm
Posterior 4 mm
Posterior 6 mm
0.636 (0.036) 0.235 (0.293) 0.354 (0.098)
0.871 ( < 0.001) 0.059 (0.795) 0.257 (0.236)
0.616 (0.044) 0.246 (0.271) 0.354 (0.098)
0.859 ( < 0.001) 0.078 (0.730) 0.257 (0.236)
0.287 (0.391) 0.287 (0.195) 0.354 (0.097)
0.431 (0.186) 0.208 (0.354) 0.225 (0.302)
Pearson's correlation coefficient. Sperman's correlation coefficient. r (P).
Table 4 Correlation between corneal higher-order aberrations and Corneal fluorescein score.
r P value
Total 4 mm
Total 6 mm
Anterior 4 mm
Anterior 6 mm
Posterior 4 mm
Posterior 6 mm
0.550 < 0.001
0.523 < 0.001
0.550 < 0.001
0.525 < 0.001
0.340 0.008
0.158 0.228
Sperman's correlation coefficient. r (P).
Fig. 6. Representative case showing the effectiveness of rigid gas-permeable contact lens. A 44-year-old Japanese woman presenting with symptoms 1191 days after undergoing hematopoietic stem cell transplantation due to acute myeloblastic leukemia. The figure shows her right eye which had diagnosed with chronic ocular GVHD. OSDI, in terms of her subjective symptoms was 50/100, Schirmer's test was 0 mm, corneal fluorescein score was 8/9, and the grading of injection was moderate (top left and top right). The topography map showed an irregular corneal surface and matched thinning (bottom left and right). Anterior HOAs were 1.27 and 2.89, posterior HOAs were 0.14 and 0.27, and total HOAs were 1.27 and 2.99 (at 4 mm and 6 mm, respectively). Her visual acuity improved from 0.6 to 0.1 (logMAR) by rigid gas-permeable contact lens correction.
years [3], which will be needed for further studies as well. To conclude this study, this is the first report that quantified the detailed optical function such as anterior, posterior, and total corneal HOAs in the eyes of patients with chronic ocular GVHD, and analyzed
acute and chronic GVHD are different and the pathophysiology of acute GVHD is well known [1]. Therefore, we only focused on the chronic phase. Nevertheless, we need to follow the quantitative visual function before HSCT and follow the patients’ post-HSCT course for at least 2 9
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Table 5 Comparison of corneal higher-order aberrations according to patterns.
HOAs
a
lower flattening pattern
other pattern
n
7
4
Total 4 mm Total 6 mm
1.29 ± 0.871 2.01 ± 0.867
0.456 ± 0.143 0.850 ± 0.297
0.044 0.012
Anterior 4 mm Anterior 6 mm
1.328 ± 0.907 2.134 ± 0.854
0.495 ± 0.155 0.908 ± 0.306
0.052 0.009
Posterior 4 mm Posterior 6 mm
0.109 ± 0.072 0.223 ± 0.094
0.064 ± 0.013 0.160 ± 0.024
0.155 0.134
Hematol 2002;75:332–4. [7] Holler E. Risk assessment in haematopoietic stem cell transplantation: GvHD prevention and treatment. Best Pract Res Clin Haematol 2007;20:281–94. [8] Srinivasan M, Mascarenhas J, Prashanth CN. Distinguishing infective versus noninfective keratitis. Indian J Ophthalmol 2008;56(3):203–7. [9] Sun YC, Chai X, Inamoto Y, et al. Impact of ocular chronic graft-versus-host disease on quality of life. Biol Blood Marrow Transplant 2015;21(9):1687–91. [10] Saboo US, Amparo F, Abud TB, Schaumberg DA, Dana R. Vision-related quality of life in patients with ocular graft-versus-host disease. Ophthalmology 2015;122(8):1669–74. [11] Pezzotta S, Rossi GC, Scudeller L, et al. A cross-sectional study on vision-related quality of life in patients with ocular GvHD. Bone Marrow Transplant 2015;50(9):1224–6. [12] Gratwohl A, Pasquini MC, Aljurf M, et al. Worldwide Network for Blood and Marrow Transplantation (WBMT). One million haemopoietic stem-cell transplants: a retrospective observational study. Lancet Haematol 2015;2(3):e91–100. [13] Majhail NS, Tao L, Bredeson C, et al. Prevalence of hematopoietic cell transplant survivors in the United States. Biol Blood Marrow Transplant 2013;19(10):1498–501. [14] Yamaguchi T, Ohnuma K, Tomida D, et al. The contribution of the posterior surface to the corneal aberrations in eyes after keratoplasty. Investig Ophthalmol Vis Sci 2011;52(9):6222–9. [15] Kashizuka E, Yamaguchi T, Yaguchi Y, Satake Y, Shimazaki J. Corneal higher-order aberrations in herpes simplex keratitis. Cornea 2016;35(12):1562–8. [16] Yagi-Yaguchi Y, Yamaguchi T, Okuyama Y, Satake Y, Tsubota K, Shimazaki J. Corneal higher order aberrations in granular, lattice and macular corneal dystrophies. PLoS One 2016;11(8):e0161075. [17] Shimizu E, Yamaguchi T, Satake Y, Tsubota K, Shimazaki J. Corneal higher-order aberrations in eyes after corneal perforation. Eye Contact Lens 2019;45(2):124–31. [18] Yamaguchi T, Shimizu E, Yagi-Yaguchi Y, Tomida D, Satake Y, Shimazaki J. A novel entity of corneal diseases with irregular posterior corneal surfaces: concept and clinical relevance. Cornea 2017;36(Suppl 1):S53–9. [19] Ogawa Y, Kim SK, Dana R, et al. International chronic ocular graft-vs-host-disease (GVHD) consensus group: proposed diagnostic criteria for chronic GVHD (Part I). Sci Rep 2013;3:3419. [20] Department of Ophthalmology, Keio University School of Medicine. Opt out of the retrospective study [cited 2019 June 22]. Available from:. http://ophthal.med.keio. ac.jp/research-info/ri022.html. [21] Shimizu E, Yamaguchi T, Yaguchi Y, et al. Corneal higher-order aberrations in infectious keratitis. Am J Ophthalmol 2017;175:148–58. [22] Ibrahim OMA, Yagi-Yaguchi Y, Noma H, Tsubota K, Shimazaki J, Yamaguchi T. Corneal higher-order aberrations in Stevens-Johnson syndrome and toxic epidermal necrolysis. Ocul Surf 2019 Jul 17. pii: S1542-0124(18)30383-5. [23] Koh S, Maeda N, Ikeda C, et al. The effect of ocular surface regularity on contrast sensitivity and straylight in dry eye. Investig Ophthalmol Vis Sci 2017;58(5):2647–51. [24] Kaido M, Matsumoto Y, Shigeno Y, Ishida R, Dogru M, Tsubota K. Corneal fluorescein staining correlates with visual function in dry eye patients. Investig Ophthalmol Vis Sci 2011;52(13):9516–22. [25] Mastropasqua L, Nubile M, Lanzini M, et al. Epithelial dendritic cell distribution in normal and inflamed human cornea: in vivo confocal microscopy study. Am J Ophthalmol 2006;142(5):736–44. [26] He J, Ogawa Y, Mukai S, et al. In vivo confocal microscopy evaluation of ocular surface with graft-versus-host disease-related dry eye disease. Sci Rep 2017;7(1):10720. [27] Yamaguchi T, Satake Y, Dogru M, Ohnuma K, Negishi K, Shimazaki J. Visual function and higher-order aberrations in eyes after corneal transplantation: How to improve postoperative quality of vision. Cornea 2015;34:S128–35. [28] Shimizu E, Yamaguchi T, Tomida D, et al. Corneal higher-order aberrations and visual improvement following corneal transplantation in treating herpes simplex keratitis. Am J Ophthalmol 2017;184:1–10. [29] Takahide K, Parker PM, Wu M, et al. Use of fluid-ventilated, gas-permeable scleral lens for management of severe keratoconjunctivitis sicca secondary to chronic graftversus-host disease. Biol Blood Marrow Transplant 2007 Sep;13(9):1016–21. [30] Magro L, Gauthier J, Richet M, et al. Scleral lenses for severe chronic GvHD-related keratoconjunctivitis sicca: a retrospective study by the SFGM-TC. Bone Marrow Transplant 2017 Jun;52(6):878–82. [31] Vincent SJ, Fadel D. Optical considerations for scleral contact lenses: a review. Contact Lens Anterior Eye 2019 May 1. pii: S1367-0484(19)30036-0. [32] Nakagawa T, Maeda N, Higashiura R, Hori Y, Inoue T, Nishida K. Corneal topographic analysis in patients with keratoconus using 3-dimensional anterior segment optical coherence tomography. J Cataract Refract Surg 2011;37(10):1871–8. [33] Koh S, Maeda N, Ikeda C, et al. Ocular forward light scattering and corneal backward light scattering in patients with dry eye. Investig Ophthalmol Vis Sci 2014;55(10):6601–6. [34] Koh S, Maeda N, Ikeda C, et al. The effect of ocular surface regularity on contrast sensitivity and straylight in dry eye. Investig Ophthalmol Vis Sci 2017;58(5):2647–51. [35] Kaido M. Functional visual acuity. Investig Ophthalmol Vis Sci 2018;59(14):DES29–35.
P valuea
Mann–Whitney test.
the correlation with visual function. We found that the chronic ocular GVHD eyes had higher corneal HOAs than the non-GVHD and normal eyes. Moreover, HOAs are correlated with visual acuity, especially in severe cases using quantitative ICO severity scores. The quantification of corneal HOAs can be an objective factor to assess corneal optical function, especially in eyes with severe chronic ocular GVHD. Declaration of competing interest There is no conflict of interest to declare associated with this manuscript. Acknowledgement Funding information: This work was supported by the Japanese Ministry of Education, Science, Sports and Culture, #18K09421 and The Uehara Memorial Foundation. No other funding statement to declare. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jtos.2019.10.005. Author contributions ES, NA, HY, TY, and YO conceived the experiments, ES, YO, YS, MY, MU, MY, MF, JH, FY, and SM conducted the experiments, ES, and YO analyzed the results. All authors have read and agreed on the final version. YO and KT supervised this study. References [1] Shikari H, Antin JH, Dana R. Ocular graft-versus-host disease: a review. Surv Ophthalmol 2013;58(3):233–51. [2] Townley JR, Dana R, Jacobs DS. Keratoconjunctivitis sicca manifestations in ocular graft versus host disease: pathogenesis, presentation, prevention, and treatment. Semin Ophthalmol 2011;26:251–60. [3] Uchino M, Ogawa Y, Uchino Y, Mori T, Okamoto S, Tsubota K. Comparison of stem cell sources in the severity of dry eye after allogeneic haematopoietic stem cell transplantation. Br J Ophthalmol 2012;96:34–7. [4] Sales CS, Johnston LJ, Ta CN. Long-term clinical course of dry eye in patients with chronic graft-versus-host disease referred for eye examination. Cornea 2011;30:143–9. [5] Jagasia MH, Greinix HT, Arora M, et al. National Institutes of health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I. The 2014 diagnosis and staging working group report. Biol Blood Marrow Transplant 2015;21(3):389–401.e1. [6] Saito T, Shinagawa K, Takenaka K, et al. Ocular manifestation of acute graft-versushost disease after allogeneic peripheral blood stem cell transplantation. Int J
10
Contents lists available at ScienceDirect
The Ocular Surface journal homepage: www.elsevier.com/locate/jtos
Original Research
Corneal higher-order aberrations in eyes with chronic ocular graft-versushost disease Eisuke Shimizua,∗, Naohiko Aketaa, Hiroyuki Yazua,b, Miki Uchinoa, Mizuka Kamoia, Yasunori Satoc, Kazuo Tsubotaa, Yoko Ogawaa,∗∗ a
Department of Ophthalmology, Keio University School of Medicine, Tokyo, Japan Department of Ophthalmology, Tsurumi University School of Dental Medicine, Kanagawa, Japan c Department of Preventive Medicine and Public Health, Biostatistics at Clinical and Translational Research Center, Keio University School of Medicine, Tokyo, Japan b
A R T I C LE I N FO
A B S T R A C T
Keywords: Graft-versus-host disease Dry eye disease Higher-order aberration Visual acuity Irregular astigmatism
Purpose: Chronic ocular graft-versus-host disease (GVHD) is a long-term complication after hematopoietic stem cell transplantation (HSCT) and leads to irreversible visual morbidity due to severe ocular surface impairments including visual disfunction. However, knowledge about the optical function in chronic ocular GVHD is limited because it is difficult to assess quantitative optical function objectively. The development of anterior segment optical coherence tomography has allowed objective quantification of optical function by evaluating corneal higher-order aberrations (HOAs). Therefore, we applied this quantification in chronic ocular GVHD patients and verified the correlation between corneal HOAs and visual acuity. Methods: We retrospectively reviewed chronic ocular GVHD patients and the recipients after HSCT. Then, analyzed the relationship between visual function and the severity of chronic ocular GVHD. Results: The eyes of patients with chronic ocular GVHD had higher corneal HOAs than those of non-GVHD patients (HOAs; 0.481 ± 0.306 vs 0.254 ± 0.084, and 0.917 ± 0.609 vs 0.529 ± 0.130. P < 0.001, and 0.002. 4-mm and 6-mm diameters respectively. Corneal HOAs were correlated with the severity of chronic ocular GVHD (r = 0.436. P < 0.001). Moreover, corneal HOAs were correlated with visual acuity, especially in eyes with severe chronic ocular GVHD cases (HOAs; 4-mm r = 0.636. P = 0.036. Total 6-mm r = 0.871. P = < 0.001). Conclusions: We succeed to assess the objective value in the optical function of the chronic ocular GVHD. Quantification of corneal HOAs could be an objective evaluation to assess optical function in eyes with chronic ocular GVHD.
Introduction Chronic graft-versus-host disease (GVHD) is a major and lethal complication after hematopoietic stem cell transplantation (HSCT) for hematologic malignancies. For chronic GVHD in the eye, the main and the most common phenotype is dry eye disease (DED) with keratoconjunctivitis sicca [1]. The recipients are diagnosed as showing chronic ocular GVHD when the severe DED outbreak occurs after HSCT [2], and half of the recipients develop DED within 2 years after HSCT [3]. Although chronic ocular GVHD occurs in a large number of recipients, it typically does not lead to visual loss or blindness and the recipients generally have a good visual acuity [4]. Moreover, the revised criteria by the National Institutes of Health did not mention about
∗
any visual function loss in chronic ocular GVHD [5]. However, some studies have demonstrated that ocular involvement is a poor prognostic factor for mortality in GVHD [6,7]. Furthermore, it is known that ocular GVHD impairs recipients’ health-related quality of life [8,9]. For example, Saboo et al. demonstrated that chronic ocular GVHD patients often experience impairment of vision-related quality of life [10]. Pezzotta et al. showed similar results and claimed that it is important to recognize and treat all GVHD manifestations, including ocular involvement, because the total number and life expectancy of recipients treated with bone marrow transplantation has increased [11]. The Worldwide Network for Blood and Marrow Transplantation announced that by the improvements in the state-of-the-art technique, over 60,000 recipients receive HSCT per year [12]. Moreover, as the life
Corresponding author. Department of Ophthalmology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo, 160-8582, Japan. Corresponding author. Department of Ophthalmology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo, 160-8582, Japan. E-mail addresses: [email protected] (E. Shimizu), [email protected] (Y. Ogawa).
∗∗
https://doi.org/10.1016/j.jtos.2019.10.005 Received 6 August 2019; Received in revised form 4 October 2019; Accepted 8 October 2019 1542-0124/ © 2019 Published by Elsevier Inc.
Please cite this article as: Eisuke Shimizu, et al., The Ocular Surface, https://doi.org/10.1016/j.jtos.2019.10.005
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
conjunctivitis, blepharitis, ocular surface infectious disease, allergic conjunctivitis, ocular cicatricial pemphigoid, Stevens-Johnson syndrome, and Sjögren's syndrome), (2) eyes with other corneal diseases (ie, keratoconus, corneal dystrophies, bullous keratopathy, infectious keratitis, Pellucid marginal corneal degeneration, and corneal perforation), (3) history of corneal surgeries (ie, after laser in situ keratomileusis, radial keratotomy, and corneal transplantation), because a history of keratoconus, corneal surgery, or perforation can influence the corneal shape, and (4) presence of acute GVHD. On the basis of these exclusion criteria, 33 patients are excluded, and 97 patients remained in our study. Next, to confirm the eligibility of the cases for further visual function analysis, cases with a history of cataract, glaucoma, retinal disease, and optic nerve disease were thought to be inadequate. Thus, 7 patients were omitted from further analyses. Next, the patients were divided into 2 groups: (1) chronic ocular GVHD cases, and (2) non-GVHD (post-HSCT and not defined as chronic ocular GVHD) cases, according to the diagnostic criteria proposed by ICO GVHD Consensus group [19]. Finally, to adjust the effect of the intervention by patient background factors, propensity score matching was performed. Ultimately, 30 eyes of 15 GVHD cases and 30 eyes of the 15 non-GVHD cases were matched for the main analysis. In addition, 29 eyes of 20 age-matched subjects were included as normal controls.
expectancy continues to increase, the total number of survivors worldwide is estimated to be over 300,000 in 2030 [13]. Despite the importance of visual function in chronic ocular GVHD patients, previous studies only covered questionnaire-based analyses and did not include measurements with any quantitative examination. Our previous study used a novel calculation method for evaluating corneal higher-order aberrations (HOAs) in the transplanted cornea by using anterior segment optical coherence tomography (AS-OCT) [14]. This quantitative technique was applied to other corneal disorders, such as herpetic keratitis [15], corneal dystrophies [16], and corneal scars after perforation [17]. Moreover, this method demonstrated a strong correlation between the increase of corneal HOAs and decrease of visual acuity [15–18]. Therefore, we expect that the quantitative analysis technique such as corneal HOAs evaluation using AS-OCT could be a new optical function assay for chronic ocular GVHD patients. Although chronic ocular GVHD was said to be affected to the ocular surfaces such as the cornea and conjunctiva [1], the exact cause of the visual impairment is still unknown. Therefore, we hypothesize that corneal HOAs are deteriorated in the chronic ocular GVHD eyes and the corneal HOAs are correlated with visual acuity as the several corneal diseased which we demonstrated in the past [14–18]. Hence, the aim of this study is to elucidate the relationship between the severity of chronic ocular GVHD and quantitative measures of visual and optical function, such as visual acuity and corneal HOAs.
Data analysis Routine examinations, including slit-lamp microscopy, clinical ocular evaluation, fundus examinations, and best corrected visual acuity (BCVA) measurement, were conducted at the time of diagnosis for all patients. Clinical ocular parameters such as tear film breakup time (TFBUT), corneal fluorescein staining score (CFS), corneoconjunctival lissamine green staining score (LG), degree of the meibomian gland dysfunction (MGD), degree of fibrosis, degree of filamentosa, and degree of hyperemia were examined by multiple DED specialists (E.S., Y.O., M.U., and M.K.) according to the diagnostic criteria proposed by the ICO GVHD Consensus group [19]. Schirmer's test (Sterilized Tear Production Measuring Strips. 4987896590227. Ayumi Pharmaceutical Corporation. Tokyo, Japan) was performed prior to ocular evaluations to avoid the effect of the fluorescein/lissamine green solution. In addition, during several examinations, participants answered the Ocular Surface Disease Index (OSDI) questionnaire. Visual acuity was measured using a standard Snellen chart, and the BCVA with spectacle correction was recorded. The results were measured in decimal acuity and converted to logarithm of the minimal angle of resolution (logMAR) units using a visual acuity conversion chart.
Materials and methods This retrospective study adhered to the tenets of the Declaration of Helsinki. All procedures were performed in compliance with the protocol (#20170350) approved by the institutional ethics review board of Keio University School of Medicine (Tokyo, Japan). Ethical guidelines for clinical studies from the Japanese Ministry of Health, Labour and Welfare indicate that for studies that do not involve biological tissue and which involve reviewing medical records retrospectively, researchers do not need to obtain written informed consent from each patient. On the basis of the guidelines of ethics committees, we have provided detailed written guidelines and the ethical statement of the present study on the website [20]. Informed consent was obtained by notices including the following factors; background of the study, purpose of the study, study design, privacy policy, freedom to withdraw, inclusion and exclusion criteria, the factors assessed in the medical records, advantages and disadvantages of participating the study, disclosure of the data, presentation of the data at a conference or in a journal, and contact information. Patient data were anonymized before access and/or analysis.
Anterior segment optical coherence tomography
Study design
Eyes in all the cases were routinely examined using anterior segment optical coherence tomography (AS-OCT). All subjects were examined until at least two sets of sufficient images were obtained. Sixteen rotating AS-OCT scans were used to reconstruct three-dimensional models of the entire corneal structure. The CASIA system (SS1000, Tomey, Nagoya, Japan) corrected distortions in the AS-OCT images based on the refractive index of the anterior surface. Two corneal specialists (N.A. and H.Y.) carefully checked all AS-OCT images to ensure that the surface digitalization recognized by the automated inbuilt software was correct. Zernike coefficients were calculated using Zernike analysis as previously reported [21]. In brief, the anterior and posterior corneal surfaces were reconstructed as a three-dimensional model from the corneal height data. The anterior, posterior, and total corneal aberrations at diameters of 4 mm and 6 mm were calculated separately with the installed ray tracing software, version 5.1. The refractive indices of the cornea and aqueous humor were set to 1.376 and 1.336, respectively. The wavefront aberration was expanded with normalized Zernike polynomials up to the 8th order. HOA was defined as the root mean square (RMS) of the 3rd-to the 8th-order Zernike
This retrospective study was performed according to the flowchart in Fig. 1. The Dry Eye Disease (DED) specialists (E.S. and Y.O.) and the corneal specialists (N.A. and H.Y.) carefully reviewed the medical records of all patients who visited the DED outpatient clinic in the Keio University hospital from April 2018 to March 2019. We included Japanese adult males and females (over 20 years of age) who underwent hematopoietic stem cell transplantation (HSCT) at the Department of Hematology, Keio University hospital between April 1995 and December 2017 and presented with new onset of chronic ocular GVHD after HSCT. Chronic ocular GVHD was defined on the basis of the diagnostic criteria proposed by the International Chronic Ocular GVHD Consensus Group [19]. Also, the severity of the chronic ocular GVHD was defined by the quantitative scale from 0 through 11, based on the GVHD-related DED severity score (ICO severity score) [19]. In total, 130 patients conformed to the inclusion criteria. Next, patients who showed at least one of the following findings were excluded: (1) common findings of other ocular surface diseases (ie, 2
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Fig. 1. Study chart. The retrospective study was performed at the dry eye clinic in the Keio University hospital from April 2018 to March 2019.
coefficients. Spherical aberration (SA) was defined as the RMS of Z40 (spherical aberration) and Z60(secondary spherical aberration). Coma aberration (COMA) was defined as the RMS of Z3−1 and Z31 using the same method as in previous studies [21].
deviation (SD). P-values < 0.05 were considered to indicate statistical significance.
Results Statistical analysis Patient demographics The data were analyzed using Prism software (ver. 6.04 for Mac; GraphPad Software Inc, San Diego, CA, USA) and Excel for Mac (ver. 15.32; Microsoft Corporation, Redmond, WA, USA). Sample size was difficult to define because there was no past reference that evaluated HOAs of the chronic ocular GVHD patients. Therefore, the HOA data of patients with similar ocular surface diseases such as Stevens-Johnson syndrome was selected for the sample size calculation [22]. A total sample size of 18–19 was calculated to be adequate based on the effective size (1.10), statistical power (0.95), and the significance level (0.05). The D'Agostino and Pearson omnibus normality test was used to assess whether the data showed a normal distribution. To compare differences in clinical parameters between the GVHD cases and the nonGVHD cases, unpaired t-test or Mann-Whitney U test was used. To reduce the effect of patient selection bias and potential confounding effects, we performed adjustments for differences in baseline characteristics by means of propensity score matching. Patient selection was performed by using the propensity score matching method with the Greedy 5-to-1 digit-matching algorithm for baseline characteristics, i.e., age, gender, and post-HSCT duration at baseline. To compare the differences in HOAs, SA, and COMA among the GVHD cases, the nonGVHD cases, and the controls, Dunn's multiple comparison test was used. The HOAs were transformed into natural logarithm values for the correlation analysis including HOAs, visual acuity, and the ICO GVHD Consensus Group Proposed Severity Scale (ICO severity score). Then, Spearman's correlation coefficients were determined. For the correlation analysis between HOAs and visual acuity according to the ICO severity score (severe; n = 11, moderate; n = 24, and none; n = 25), Pearson's correlation coefficient and Spearman's correlation coefficient were selected. To exclude the bias which using bilateral eyes in a single analysis, we compared the ipsilateral eyes between the GVHD and the non-GVHD eyes. All data were expressed as the mean ± standard
Mean ages were 54.20 ± 11.33 years (range, 29–70 years) in the GVHD group and 52.93 ± 10.02 years (range, 33–70 years) in the nonGVHD group. There was no significant intergroup difference in patient age (P = 0.748. Table 1). Mean follow-up duration after HSCT was 1689 ± 1032 days (range, 199–3695 days) in the GVHD patients and 1773 ± 1671 days (range, 205–5374 days) in the non-GVHD patients, with no significant intergroup difference in the mean follow-up duration (P = 0.869. Table 1). Mean ICO severity score was 8.00 ± 1.89 in the GVHD patients and 2.93 ± 1.70 in the non-GVHD patients, with the difference being significant (P < 0.001. Table 1). Mean TFBUT was 3.13 ± 1.73 in the GVHD patients and 5.53 ± 1.85 in the non-GVHD patients. The GVHD patients showed a significantly shorter TFBUT (P = 0.001. Table 1). The GVHD patients also showed significantly worse phenotypes with respect to the LG score, MGD grade, degree of fibrosis, and degree of filamentosa (GVHD vs non-GVHD, P value: LG score, 3.87 ± 2.07 vs 0.47 ± 0.64, P < 0.001; MGD grade, 2.13 ± 0.83 vs 1.27 ± 0.80, P = 0.007; degree of fibrosis, 1.07 ± 0.80 vs 0.33 ± 0.49, P = 0.012; degree of filamentosa, 0.27 ± 0.46 vs 0.00 ± 0.00, P = 0.032; Table 1) Mean visual acuity was −0.02 ± 0.12 in the GVHD patients and −0.08 ± 0.02 in the non-GVHD patients (logMAR). The GVHD patients had significantly worse visual acuity (P = 0.004. Table 1). However, there were no statistically significant differences in the corneal pachymetry and intraocular pressure between the GVHD patients and the non-GVHD patients (corneal pachymetry, 542.4 ± 37.8 vs 540.8 ± 31.3, P = 0.631; intraocular pressure, 13.87 ± 2.86 vs 13.13 ± 2.99, P = 0.782). Moreover, in the analysis using ipsilateral eyes, estimated value differences of ICO score, TFBUT, CFS, LG, MGD, fibrosis, filamentosa, and hyperemia were similar to the result of the bilateral eye analysis. (Supplementary Table 1 and Table 1). 3
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
and 6-mm diameters (P < 0.05. Table 2 and Supplementary Fig. 1). Moreover, for COMA, statistically significant differences were only shown between the GVHD patients and the non-GVHD patients in total and anterior COMA at the 4-mm diameter, and between the GVHD patients and the normal controls in total COMA at the 6-mm diameter (P < 0.05. Table 2 and Supplementary Fig. 2). Moreover, we analyzed the corneal HOAs in the same lateral eyes, the eyes with the GVHD had more significantly high HOAs compared to the non-GVHD in both right or left eyes (Supplementary Table 2).
Table 1 Demographic of the subjects.
Eyes, n Cases, n Male Female Age, years old * Post HSCT, days* ICO severity score, points* OSDI, points** Schirmer's test, mm* TFBUT, seconds** CFS, points** LG, points* MGD, points* Fibrosis, points** Filamentosa, points* Hyperemia, points** Corneal pachymetry, μm** Visual Acuity, LogMAR** Intra Ocular Pressure, mmHg**
GVHD
nonGVHD
P Value
30 15 10 5 54.20 ± 11.33 1689 ± 1032 8.00 ± 1.89 38.64 ± 29.84 4.13 ± 3.56 3.13 ± 1.73 5.07 ± 2.31 3.87 ± 2.07 2.13 ± 0.83 1.07 ± 0.80 0.27 ± 0.46 0.93 ± 0.59 542.4 ± 37.8 −0.02 ± 0.12 13.87 ± 2.86
30 15 9 6 52.93 ± 10.02 1773 ± 1671 2.93 ± 1.70 9.44 ± 13.84 7.13 ± 4.93 5.53 ± 1.85 0.60 ± 0.91 0.47 ± 0.64 1.27 ± 0.80 0.33 ± 0.49 0.00 ± 0.00 0.13 ± 0.35 540.8 ± 31.3 −0.08 ± 0.02 13.13 ± 2.99
0.748 0.869 < 0.001 0.001 0.066 0.001 < 0.001 < 0.001 0.007 0.012 0.032 0.001 0.631 0.004 0.782
Correlations between corneal HOAs and visual acuity Fig. 3 shows the correlation between HOAs and visual acuity. There was a statistically significant positive correlation between HOAs and visual acuity in all HOAs, including total, anterior, and posterior HOAs at the 4-mm and 6-mm diameters. (HOAs vs visual acuity: total 4 mm, r = 0.518, P < 0.001; total 6 mm, r = 0.430, P < 0.001; anterior 4 mm, r = 0.516, P < 0.001; anterior 6 mm, r = 0.442, P < 0.001; posterior 4 mm, r = 0.385, P = 0.003; and posterior 6 mm, r = 0.337, P = 0.011; Fig. 3).
Data were expressed as the mean ± standard deviation (SD). *Unpaired t-test.**Mann–Whitney test. HSCT: hematopoietic stem cell transplantation. ICO severity score: International Chronic Ocular GVHD. Consensus Group Proposed Severity Scale. OSDI: ocular surface disease index. TFBUT: Tear Film Breakup Time. CFS: Corneal Fluorescein Score. LG: Lissamine Green staining score. MGD: Meibomian Gland Dysfunction.
Correlations between ICO severity score and visual acuity Fig. 4 shows the correlation between ICO severity score and visual acuity. There was a moderate positive correlation between ICO severity score and visual acuity (ICO severity score vs visual acuity; r = 0.436, P < 0.001.).
Corneal higher-order aberrations, spherical aberrations, and coma aberrations
Correlations between corneal HOAs and ICO severity score
Table 2 and Fig. 2 show the mean HOAs, SA, and COMA of Zernike terms at 4-mm and 6-mm diameters in each group. The total and anterior HOAs at both 4-mm and 6-mm diameters in GVHD patients were significantly higher than those in the non-GVHD cases and the healthy controls (P < 0.05; Table 2 and Fig. 2). However, no differences were observed in posterior HOAs between the groups. This trend was only shown in the HOAs and not in the SA and COMA. For SA, statistically significant differences were only shown between GVHD patients and the healthy controls in total and anterior SA at both 4-mm
Fig. 5 shows the correlation between HOAs and ICO severity score. There was a statistically significant positive correlation between HOAs and ICO severity score in all HOAs, including the total and anterior HOAs at 4-mm and 6-mm diameters (HOAs vs ICO severity score: total 4 mm, r = 0.594, P < 0.001; total 6 mm, r = 0.565, P < 0.001; anterior 4 mm, r = 0.594, P < 0.001; anterior 6 mm, r = 0.581, P < 0.001; Fig. 5). There was a weak positive correlation between HOAs and ICO severity score in posterior HOAs at the 4-mm and 6-mm diameters (HOAs vs ICO severity score: posterior 4 mm, r = 0.353,
Table 2 Comparison of the corneal higher-order aberrations, spherical aberrations, and coma aberrations. n
GVHD
nonGVHD
Normal
P value*
30
30
29
GVHD vs nonGVHD
GVHD vs Normal
nonGVHD vs Normal
< 0.001 0.002 < 0.001 0.001 0.114 0.399
0.025 0.017 0.022 0.015 > 0.999 > 0.999
0.941 > 0.999 0.928 > 0.999 0.490 0.070
HOA
Total 4 mm Total 6 mm Anterior 4 mm Anterior 6 mm Posterior 4 mm Posterior 6 mm
0.481 0.917 0.506 0.959 0.060 0.153
± ± ± ± ± ±
SA
Total 4 mm Total 6 mm Anterior 4 mm Anterior 6 mm Posterior 4 mm Posterior 6 mm
0.027 ± 0.189 0.124 ± 0.214 0.031 ± 0.192 0.146 ± 0.213 −0.021 ± 0.010 −0.112 ± 0.026
0.124 ± 0.073 0.225 ± 0.144 0.130 ± 0.071 0.243 ± 0.144 −0.024 ± 0.009 −0.110 ± 0.019
0.158 ± 0.067 0.257 ± 0.108 0.168 ± 0.068 0.273 ± 0.113 −0.026 ± 0.009 −0.104 ± 0.031
0.415 0.136 0.411 0.160 0.746 > 0.999
0.004 0.006 0.002 0.012 0.582 0.865
0.250 0.811 0.145 > 0.999 > 0.999 > 0.999
COMA
Total 4 mm Total 6 mm Anterior 4 mm Anterior 6 mm Posterior 4 mm Posterior 6 mm
0.187 0.411 0.190 0.423 0.014 0.045
0.079 0.245 0.082 0.265 0.013 0.036
0.110 0.209 0.112 0.229 0.018 0.050
0.015 0.564 0.015 0.828 > 0.999 0.311
0.178 0.032 0.181 0.078 0.214 > 0.999
> 0.999 0.640 > 0.999 0.753 0.066 0.255
± ± ± ± ± ±
0.306 0.609 0.307 0.597 0.023 0.037
0.189 0.425 0.188 0.433 0.010 0.036
0.254 0.529 0.271 0.567 0.049 0.141
± ± ± ± ± ±
± ± ± ± ± ±
0.084 0.130 0.087 0.140 0.011 0.018
0.043 0.121 0.044 0.129 0.007 0.026
0.280 0.560 0.297 0.605 0.054 0.155
GVHD: Graft-versus-host disease. HOA: higher-order aberration. SA: Spherical aberration. COMA: Coma aberration. *Kruskal-Wallis test. 4
± ± ± ± ± ±
± ± ± ± ± ±
0.096 0.149 0.093 0.160 0.015 0.024
0.106 0.148 0.104 0.156 0.010 0.032
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Fig. 2. Corneal higher-order aberrations in the GVHD, non-GVHD, and normal eyes. The anterior HOAs are higher at both 4-mm and 6-mm diameters in the GVHD patients compared to those in the non-GVHD cases and the healthy controls (all P < 0.05). In contrast, no differences were observed in posterior HOAs between the groups.
r = 0.616, P = 0.044; anterior 6 mm, r = 0.859, P < 0.001; Table 3). However, there was no correlation in other measurements.
P = 0.006; posterior 6 mm, r = 0.294, P = 0.023; Fig. 5). Table 3 shows the results of the correlation analysis between HOAs and visual acuity according to the ICO severity score (severe, n = 11; moderate, n = 24; and none, n = 25). There was a strong positive correlation between HOAs and visual acuity in total, and anterior HOAs at 4-mm and 6-mm diameters (HOAs vs visual acuity; total 4 mm, r = 0.636, P = 0.036; total 6 mm, r = 0.871, P < 0.001; anterior 4 mm,
Discussion In the current study, patients with chronic ocular GVHD had higher corneal HOAs than the non-GVHD patients and the normal controls 5
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Fig. 3. Correlation between corneal higher-order aberrations and visual acuity. Statistically significant positive correlation was shown between corneal HOAs and visual acuity in the eyes with the GVHD and the nonGVHD groups. (n = 30, all P < 0.05). The strong correlation was seemed between total and anterior 4-mm cornea and the visual acuity (r = 0.518 and 0.516).
compared to the normal. Unlike these whole corneal morphological changes caused by infections, inflammations [8], degeneration, and trauma, chronic ocular GVHD affects the ocular surface through inflammation and cicatricial change [1]. Therefore, it is conceivable that the surface impairment caused by chronic ocular GVHD increases corneal HOAs, especially in the anterior and total cornea (Fig. 2, Table 2).
(Fig. 2, Table 2). Similar trends were observed in SA (Supplemental Fig. 1) and COMA (Supplemental Fig. 2). However, these trends were not as specific as the differences observed in HOAs. Several corneal and ocular surface disorders such as infectious keratitis [15,21], corneal dystrophies [16], corneal perforations [17], and Stevens-Johnson's syndrome [22] showed significantly higher HOAs, SA, and COMA
6
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
and increases only anterior and total HOAs, but not the posterior HOAs. With regard to the application of our findings in the clinical settings, the reduction of corneal HOAs is known to contribute to an increase in patients' visual acuity. Rigid gas-permeable contact lenses (RGPCL) are frequently used to increase the visual acuity in several corneal diseases. We previously reported that the visual acuity recovers with the use of RGPCL in eyes with high corneal HOAs due to the offset of the anterior HOAs [21,27,28]. Interestingly, our demographic case (Fig. 6) shows a similar result. Despite the high ICO severity scores (10/11 points), RGPCL was effective in improving visual acuity (logMAR 0.6 to 0.1, before and after wearing RGPCL). RGPCL can only correct anterior HOAs; therefore, residual HOAs such as posterior HOAs cannot be corrected and the visual acuity will not improve with an irregular posterior surface. As we hypothesized in the past, chronic ocular GVHD mainly impairs anterior HOAs. Therefore, RGPCL could be a good option to improve the patients' visual acuity. However, the eyes with severe chronic ocular GVHD has a low affinity to RGPCL due to several RGPCL-related complications. Therefore, instead of RGPCL, a gas-permeable scleral lens with a larger diameter will be the choice. The scleral lens has been effective in mitigating subjective symptoms and healing ocular surface by a liquid bandage in chronic ocular GVHD eyes [29]. Furthermore, Magro L. et al. demonstrated that the scleral lenses are effective in the severe ocular GVHD cases with keratoconjunctivitis sicca [30]. Moreover, recent evidence shows that the effect of the scleral lens to the corneal HOAs had no significant differences with the user of RGPCL [31]. Therefore, scleral lenses could be a good option to increase visual acuity in chronic ocular GVHD patients. To optimize the efficacy of the RGPCL as well as the scleral lens, information regarding the HOAs is essential to predict the visual acuity after the lens correction [18]. Examination of visual acuity is a subjective process and requires time. In contrast, corneal HOAs evaluation using AS-OCT is an objective technique with a total duration of only a few seconds [32]. We believe that subjective plus objective measurements would cooperate in improving the accuracy in visual function examinations. We found that there is a typical topographical pattern of the chronic ocular GVHD cases represented in Fig. 6. The topographical pattern shows a lower part of the paracentral cornea flattering which often develops corneal thinning and leads to perforation [1]. Interestingly, 63.6% (7 out of 11 eyes) of the severe chronic ocular GVHD eyes develops this topographic pattern. Moreover, comparing HOAs between the lower flattening pattern and the other pattern, anterior and total HOAs were higher in the lower flattening pattern eyes (Table 5). Similar topographical patterns were reported in the various corneal diseases which might be related to the pathophysiology of the causative disease [15,17,18,21]. Therefore, evaluation of the topographical patterns in the chronic ocular GVHD cases could be an important imaging biomarker for the additional evaluation. There are several limitations of the current study. First, we did not evaluate complete visual and optical parameters such as corneal scatter [33] contrast sensitivity [34], or functional visual acuity [35], which may be affected by DED. Therefore, future studies with more exhaustive visual and optical function assessments are required. Second, this was a retrospective, cross-sectional study that evaluated corneal HOAs, visual acuity, and ICO severity scores. Although we did perform propensity score matching to minimize the effect of selection bias, we did not evaluate the effect of systemic treatment or/and perform systemic evaluation in other chronic GVHD target organs. Thus, assessment of systemic management will be essential in further studies. Moreover, we included bilateral eyes in several analyses which might cause bias. To exclude this bias, we performed an additional analysis using ipsilateral eyes. The result of the comparative analysis of the ocular parameters (Supplementary Table 1) and the HOAs between the GVHD and the non-GVHD eyes (Supplementary Table 2), were similar with the bilateral eyes analysis (Table 1 and Fig. 2). Third, this study focused on chronic GVHD but not on acute GVHD. The mechanisms underlying
Fig. 4. Correlation between visual acuity and the International chronic ocular GVHD consensus group proposed severity scale. Statistically significant positive correlation was shown between ICO severity score and visual acuity in the eyes with GVHD and the non-GVHD group (n = 30, P < 0.001). The correlation coefficient was moderate (r = 0.436).
A healthy ocular surface including the tear film maintains a smooth refractive surface for good quality of vision, and the surface regularity in the central part of the cornea over the pupil is especially important in terms of visual function [23]. Kaido et al. demonstrated that the HOAs were higher in the eyes with superficial punctate keratitis (SPK) than in those without SPK, and HOAs were correlated with the severity of the epithelial damage [24]. Thus, we additionally analyzed the correlation between the corneal HOAs and CFS in terms of corneal surface impairment to check the influence of the epithelial damage. HOAs and CFS showed a significant positive correlation (total 4 mm, total 6 mm, anterior 4 mm, and anterior 6 mm; Table 4), which was similar to the correlation between HOAs and ICO severity score (Fig. 5). The explanation for this phenomenon is that the ICO severity score reflects the combined grade of multiple subjective and objective scales such as CFS, Schirmer's test, OSDI, and degree of hyperemia [19]. Therefore, CFS could influence the entire combined grading. However, the correlation coefficient between the ICO severity score and HOAs was stronger than that between CFS and HOAs (Table 4, Fig. 5). Thus, we can conceive that a combined scale such as the ICO severity score is more accurate than CFS and that the epithelial damage is not the only factor increasing the HOAs in eyes with chronic ocular GVHD. In our retrospective study, moderate correlations were observed between HOAs and visual acuity and between visual acuity and the ICO severity score (Figs. 3 and 4). Interestingly, the correlation coefficient between HOAs and visual acuity became much stronger only in the severe chronic ocular GVHD eyes (Table 3). This result suggests that none to moderate chronic ocular GVHD only shows a minimal effect on the recipient's visual function, as reported previously [10]. However, with more severe chronic ocular GVHD, the surface damage impairs the recipient's eyes to a greater degree, increases the corneal HOAs, and finally decreases the visual acuity irreversibly. Moreover, this trend was only seen in the total and anterior HOAs, but not in the posterior HOAs. This can be explained by the pathophysiology of chronic ocular GVHD, which is mediated by severe DED and keratoconjunctivitis sicca [1]. In the morphological evaluations, immune cells in the epithelia have been shown to play a role in inflammatory ocular diseases [25]. He J et al. demonstrated that the density of immune cells is higher in chronic ocular GVHD human corneas by using confocal microscopy [26]. Considering the pathophysiology of chronic ocular GVHD and the findings of previous studies, surface and epithelial impairments could be a major explanation for this finding. Thus, it is speculated that chronic ocular GVHD mainly impairs the anterior segment of the cornea 7
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Fig. 5. Correlation between corneal higher-order aberrations and the International chronic ocular GVHD consensus group proposed severity scale. Statistically significant positive correlation was shown between corneal HOAs and ICO severity score in the eyes with the GVHD and the non-GVHD group (n = 30, all P < 0.05). The strong correlation appeared to be between the total and anterior cornea HOAs and the visual acuity (total 4 mm, r = 0.594; total 6 mm, r = 0.565; anterior 4 mm, r = 0.594, and anterior 6 mm, r = 0.581).
8
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Table 3 Correlation between corneal higher-order aberrations and Visual Acuity by the International Chronic Ocular Graft-versus-host disease Consensus Group Proposed Severity Scale.
Severity Severea (n = 11) Moderateb (n = 24) Noneb (n = 25) a b
Total 4 mm
Total 6 mm
Anterior 4 mm
Anterior 6 mm
Posterior 4 mm
Posterior 6 mm
0.636 (0.036) 0.235 (0.293) 0.354 (0.098)
0.871 ( < 0.001) 0.059 (0.795) 0.257 (0.236)
0.616 (0.044) 0.246 (0.271) 0.354 (0.098)
0.859 ( < 0.001) 0.078 (0.730) 0.257 (0.236)
0.287 (0.391) 0.287 (0.195) 0.354 (0.097)
0.431 (0.186) 0.208 (0.354) 0.225 (0.302)
Pearson's correlation coefficient. Sperman's correlation coefficient. r (P).
Table 4 Correlation between corneal higher-order aberrations and Corneal fluorescein score.
r P value
Total 4 mm
Total 6 mm
Anterior 4 mm
Anterior 6 mm
Posterior 4 mm
Posterior 6 mm
0.550 < 0.001
0.523 < 0.001
0.550 < 0.001
0.525 < 0.001
0.340 0.008
0.158 0.228
Sperman's correlation coefficient. r (P).
Fig. 6. Representative case showing the effectiveness of rigid gas-permeable contact lens. A 44-year-old Japanese woman presenting with symptoms 1191 days after undergoing hematopoietic stem cell transplantation due to acute myeloblastic leukemia. The figure shows her right eye which had diagnosed with chronic ocular GVHD. OSDI, in terms of her subjective symptoms was 50/100, Schirmer's test was 0 mm, corneal fluorescein score was 8/9, and the grading of injection was moderate (top left and top right). The topography map showed an irregular corneal surface and matched thinning (bottom left and right). Anterior HOAs were 1.27 and 2.89, posterior HOAs were 0.14 and 0.27, and total HOAs were 1.27 and 2.99 (at 4 mm and 6 mm, respectively). Her visual acuity improved from 0.6 to 0.1 (logMAR) by rigid gas-permeable contact lens correction.
years [3], which will be needed for further studies as well. To conclude this study, this is the first report that quantified the detailed optical function such as anterior, posterior, and total corneal HOAs in the eyes of patients with chronic ocular GVHD, and analyzed
acute and chronic GVHD are different and the pathophysiology of acute GVHD is well known [1]. Therefore, we only focused on the chronic phase. Nevertheless, we need to follow the quantitative visual function before HSCT and follow the patients’ post-HSCT course for at least 2 9
The Ocular Surface xxx (xxxx) xxx–xxx
E. Shimizu, et al.
Table 5 Comparison of corneal higher-order aberrations according to patterns.
HOAs
a
lower flattening pattern
other pattern
n
7
4
Total 4 mm Total 6 mm
1.29 ± 0.871 2.01 ± 0.867
0.456 ± 0.143 0.850 ± 0.297
0.044 0.012
Anterior 4 mm Anterior 6 mm
1.328 ± 0.907 2.134 ± 0.854
0.495 ± 0.155 0.908 ± 0.306
0.052 0.009
Posterior 4 mm Posterior 6 mm
0.109 ± 0.072 0.223 ± 0.094
0.064 ± 0.013 0.160 ± 0.024
0.155 0.134
Hematol 2002;75:332–4. [7] Holler E. Risk assessment in haematopoietic stem cell transplantation: GvHD prevention and treatment. Best Pract Res Clin Haematol 2007;20:281–94. [8] Srinivasan M, Mascarenhas J, Prashanth CN. Distinguishing infective versus noninfective keratitis. Indian J Ophthalmol 2008;56(3):203–7. [9] Sun YC, Chai X, Inamoto Y, et al. Impact of ocular chronic graft-versus-host disease on quality of life. Biol Blood Marrow Transplant 2015;21(9):1687–91. [10] Saboo US, Amparo F, Abud TB, Schaumberg DA, Dana R. Vision-related quality of life in patients with ocular graft-versus-host disease. Ophthalmology 2015;122(8):1669–74. [11] Pezzotta S, Rossi GC, Scudeller L, et al. A cross-sectional study on vision-related quality of life in patients with ocular GvHD. Bone Marrow Transplant 2015;50(9):1224–6. [12] Gratwohl A, Pasquini MC, Aljurf M, et al. Worldwide Network for Blood and Marrow Transplantation (WBMT). One million haemopoietic stem-cell transplants: a retrospective observational study. Lancet Haematol 2015;2(3):e91–100. [13] Majhail NS, Tao L, Bredeson C, et al. Prevalence of hematopoietic cell transplant survivors in the United States. Biol Blood Marrow Transplant 2013;19(10):1498–501. [14] Yamaguchi T, Ohnuma K, Tomida D, et al. The contribution of the posterior surface to the corneal aberrations in eyes after keratoplasty. Investig Ophthalmol Vis Sci 2011;52(9):6222–9. [15] Kashizuka E, Yamaguchi T, Yaguchi Y, Satake Y, Shimazaki J. Corneal higher-order aberrations in herpes simplex keratitis. Cornea 2016;35(12):1562–8. [16] Yagi-Yaguchi Y, Yamaguchi T, Okuyama Y, Satake Y, Tsubota K, Shimazaki J. Corneal higher order aberrations in granular, lattice and macular corneal dystrophies. PLoS One 2016;11(8):e0161075. [17] Shimizu E, Yamaguchi T, Satake Y, Tsubota K, Shimazaki J. Corneal higher-order aberrations in eyes after corneal perforation. Eye Contact Lens 2019;45(2):124–31. [18] Yamaguchi T, Shimizu E, Yagi-Yaguchi Y, Tomida D, Satake Y, Shimazaki J. A novel entity of corneal diseases with irregular posterior corneal surfaces: concept and clinical relevance. Cornea 2017;36(Suppl 1):S53–9. [19] Ogawa Y, Kim SK, Dana R, et al. International chronic ocular graft-vs-host-disease (GVHD) consensus group: proposed diagnostic criteria for chronic GVHD (Part I). Sci Rep 2013;3:3419. [20] Department of Ophthalmology, Keio University School of Medicine. Opt out of the retrospective study [cited 2019 June 22]. Available from:. http://ophthal.med.keio. ac.jp/research-info/ri022.html. [21] Shimizu E, Yamaguchi T, Yaguchi Y, et al. Corneal higher-order aberrations in infectious keratitis. Am J Ophthalmol 2017;175:148–58. [22] Ibrahim OMA, Yagi-Yaguchi Y, Noma H, Tsubota K, Shimazaki J, Yamaguchi T. Corneal higher-order aberrations in Stevens-Johnson syndrome and toxic epidermal necrolysis. Ocul Surf 2019 Jul 17. pii: S1542-0124(18)30383-5. [23] Koh S, Maeda N, Ikeda C, et al. The effect of ocular surface regularity on contrast sensitivity and straylight in dry eye. Investig Ophthalmol Vis Sci 2017;58(5):2647–51. [24] Kaido M, Matsumoto Y, Shigeno Y, Ishida R, Dogru M, Tsubota K. Corneal fluorescein staining correlates with visual function in dry eye patients. Investig Ophthalmol Vis Sci 2011;52(13):9516–22. [25] Mastropasqua L, Nubile M, Lanzini M, et al. Epithelial dendritic cell distribution in normal and inflamed human cornea: in vivo confocal microscopy study. Am J Ophthalmol 2006;142(5):736–44. [26] He J, Ogawa Y, Mukai S, et al. In vivo confocal microscopy evaluation of ocular surface with graft-versus-host disease-related dry eye disease. Sci Rep 2017;7(1):10720. [27] Yamaguchi T, Satake Y, Dogru M, Ohnuma K, Negishi K, Shimazaki J. Visual function and higher-order aberrations in eyes after corneal transplantation: How to improve postoperative quality of vision. Cornea 2015;34:S128–35. [28] Shimizu E, Yamaguchi T, Tomida D, et al. Corneal higher-order aberrations and visual improvement following corneal transplantation in treating herpes simplex keratitis. Am J Ophthalmol 2017;184:1–10. [29] Takahide K, Parker PM, Wu M, et al. Use of fluid-ventilated, gas-permeable scleral lens for management of severe keratoconjunctivitis sicca secondary to chronic graftversus-host disease. Biol Blood Marrow Transplant 2007 Sep;13(9):1016–21. [30] Magro L, Gauthier J, Richet M, et al. Scleral lenses for severe chronic GvHD-related keratoconjunctivitis sicca: a retrospective study by the SFGM-TC. Bone Marrow Transplant 2017 Jun;52(6):878–82. [31] Vincent SJ, Fadel D. Optical considerations for scleral contact lenses: a review. Contact Lens Anterior Eye 2019 May 1. pii: S1367-0484(19)30036-0. [32] Nakagawa T, Maeda N, Higashiura R, Hori Y, Inoue T, Nishida K. Corneal topographic analysis in patients with keratoconus using 3-dimensional anterior segment optical coherence tomography. J Cataract Refract Surg 2011;37(10):1871–8. [33] Koh S, Maeda N, Ikeda C, et al. Ocular forward light scattering and corneal backward light scattering in patients with dry eye. Investig Ophthalmol Vis Sci 2014;55(10):6601–6. [34] Koh S, Maeda N, Ikeda C, et al. The effect of ocular surface regularity on contrast sensitivity and straylight in dry eye. Investig Ophthalmol Vis Sci 2017;58(5):2647–51. [35] Kaido M. Functional visual acuity. Investig Ophthalmol Vis Sci 2018;59(14):DES29–35.
P valuea
Mann–Whitney test.
the correlation with visual function. We found that the chronic ocular GVHD eyes had higher corneal HOAs than the non-GVHD and normal eyes. Moreover, HOAs are correlated with visual acuity, especially in severe cases using quantitative ICO severity scores. The quantification of corneal HOAs can be an objective factor to assess corneal optical function, especially in eyes with severe chronic ocular GVHD. Declaration of competing interest There is no conflict of interest to declare associated with this manuscript. Acknowledgement Funding information: This work was supported by the Japanese Ministry of Education, Science, Sports and Culture, #18K09421 and The Uehara Memorial Foundation. No other funding statement to declare. Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jtos.2019.10.005. Author contributions ES, NA, HY, TY, and YO conceived the experiments, ES, YO, YS, MY, MU, MY, MF, JH, FY, and SM conducted the experiments, ES, and YO analyzed the results. All authors have read and agreed on the final version. YO and KT supervised this study. References [1] Shikari H, Antin JH, Dana R. Ocular graft-versus-host disease: a review. Surv Ophthalmol 2013;58(3):233–51. [2] Townley JR, Dana R, Jacobs DS. Keratoconjunctivitis sicca manifestations in ocular graft versus host disease: pathogenesis, presentation, prevention, and treatment. Semin Ophthalmol 2011;26:251–60. [3] Uchino M, Ogawa Y, Uchino Y, Mori T, Okamoto S, Tsubota K. Comparison of stem cell sources in the severity of dry eye after allogeneic haematopoietic stem cell transplantation. Br J Ophthalmol 2012;96:34–7. [4] Sales CS, Johnston LJ, Ta CN. Long-term clinical course of dry eye in patients with chronic graft-versus-host disease referred for eye examination. Cornea 2011;30:143–9. [5] Jagasia MH, Greinix HT, Arora M, et al. National Institutes of health consensus development project on criteria for clinical trials in chronic graft-versus-host disease: I. The 2014 diagnosis and staging working group report. Biol Blood Marrow Transplant 2015;21(3):389–401.e1. [6] Saito T, Shinagawa K, Takenaka K, et al. Ocular manifestation of acute graft-versushost disease after allogeneic peripheral blood stem cell transplantation. Int J
10