Descemet’s membrane endothelial keratoplasty for pseudoexfoliation syndrome: a case series

Background To evaluate the clinical outcomes and features of Descemet’s membrane endothelial keratoplasty (DMEK) for eyes with pseudoexfoliation syndrome (PEX). Methods In this retrospective study, 37 DMEK cases were reviewed from available medical records. Patients who exhibited endothelial dysfunction derived from PEX or Fuchs endothelial corneal dystrophy (FECD) and successfully underwent cataract surgery about four weeks before DMEK were enrolled. The best spectacle-corrected visual acuity (BSCVA), central corneal thickness (CCT), endothelial cell density (ECD), and incidence of intra-operative/post-operative complications of DMEK were analyzed. Results This study included 14 eyes of 14 patients (PEX: n = 6, FECD: n = 8). There was no primary graft failure. In the PEX group, BSCVA improved from 0.67 ± 0.28 at the preoperative point to 0.43 ± 0.14 at 1 month, 0.27 ± 0.10 at 3 months, and 0.19 ± 0.08 at 6 months after DMEK. The donor corneal ECD was 2704 ± 225 cells/mm2 at the preoperative point and decreased to 1691 ± 498 cells/mm2 at 1 month, 1425 ± 366 cells/mm2 at 3 months, and 1281 ± 340 cells/mm2 (52.7 ± 11.7% less than ECD of the donor graft) at 6 months after DMEK. None of the patients required rebubbling. When compared with the FECD group, no statistical difference was observed in CCT (p = 0.821); BSCVA (p = 0.001) and the reduction rate of ECD (p = 0.010) were relatively worse. Conclusions DMEK is effective for the treatment of endothelial dysfunction due to PEX. Electronic supplementary material The online version of this article (10.1186/s12886-019-1130-1) contains supplementary material, which is available to authorized users.


Background
Corneal transplantation is a common procedure. Well over 100,000 cases are performed annually worldwide. Approximately half of all corneal transplantations involve endothelial keratoplasty, which replaces the corneal endothelium with a monolayer of cells. Descemet membrane endothelial keratoplasty (DMEK) is a corneal endothelial keratoplasty newly introduced by Melles et al. that allows for a faster recovery of visual acuity, fewer higher-order aberrations, and lower immunological rejection rates compared to conventional penetrating keratoplasty such as Descemet's stripping automated endothelial keratoplasty (DSAEK) [1][2][3][4].
With the worldwide increase in number of DMEK surgeries, many papers regarding DMEK for Fuchs' endothelial corneal dystrophy (FECD) have been published. However, other causes of corneal endothelial dysfunction, such complications from cataract surgery (pseudophakic bullous keratopathy) or endotheliopathy in pseudoexfoliation syndrome (PEX), are poorly understood [5].
PEX is a genetically determined, age-related, and environmentally influenced disorder characterized by anomalous production and accumulation of abnormal fibrillar extracellular aggregates on anterior segment structures, most notably on the lens capsule and pupillary border of the iris [6][7][8]. The exfoliative material is often expressed as grey-white and dandruff-like, but its origin is still obscure [9]. The material is observed in multiple organs such as the heart, lung, liver, kidney, cerebral meninges and blood vessels [10. 11]. It is also observed in ocular structures such as the anterior capsule, iris, lens zonule, trabecular meshwork and corneal endothelium. It is the leading cause of glaucoma, cataracts, and bullous keratopathy (BK) [12][13][14].
Evidence has accumulated reporting the morphological alterations in almost all cell layers of the cornea in eyes with PEX. Eyes with PEX have been documented to have deposition of hyper reflective material on the endothelium, which is presumed to be PEX material, and to have significantly lower cell densities in the basal epithelium, anterior and posterior stromatolites, and endothelium compared to controls [15]. PEX can lead to corneal endothelial cell decompensation, which can result in severe BK, requiring keratoplasty [13].
To our knowledge, this is the first paper to focus on keratoplasty for PEX. Here we describe a case series in which we conduct DMEK for BK derived from PEX and compare the result with that derived from FECD.

Patients and Examinations
The surgical maneuvers and evaluation protocols used in this retrospective study were approved by the Institutional Review Board of Yokohama Minami Kyosai Hospital (Approval no. YKH_30_02_08). We carefully followed all ethical principles within the Declaration of Helsinki. Patients exhibiting endothelial dysfunction derived from PEX or FECD and cataract were enrolled after giving written informed consent. The diagnosis of PEX keratopathy was confirmed clinically, as well as by electron 4 microscopy. Eyes in the PEX group exhibited accumulation of exfoliative materials, which was characteristic of PEX; they did not exhibit other findings, such as guttata or history of past complicated cataract surgery, that might cause BK. A total of 37 surgeries were performed at the department of ophthalmology of Yokohama Minami Kyosai Hospital in Kanagawa, Japan, between April 1, 2016, and December 31, 2017. Fourteen eyes of 14 patients (five males and eight females) were considered eligible for the study. Six eyes revealed PEX syndrome (PEX group), and the other 8 eyes revealed FECD (FECD group).
We performed a standard ophthalmic examination and took the following measurements preoperatively and up to 6 months after the operation: best spectacle-corrected visual acuity (BSCVA), corneal endothelial cell density (ECD), central corneal thickness (CCT), and graft adaptation.
Graft adaptation was assessed via slit-lamp microscopy and anterior segment optical coherence tomography (AS-OCT, SS1000, Tomey, Aichi, Japan). Corneal thickness was also measured using AS-OCT. Preoperative ECDs were retrieved from donor eye bank records, and postoperative ECDs were measured with the aid of a specular microscope (FA3509, Konan Medical, Hyogo, Japan). To detect clinical/subclinical cystoid macular edema (CME), spectral-domain OCT (RS 3000, Nidek, Aichi, Japan) was performed at 1 month, 3 months, and 6 months after DMEK. CME was defined by the presence of intraretinal fluid spaces, which are seen in the fovea region using spectral-domain OCT.

Cataract surgery
Cataract surgery was scheduled approximately four weeks before DMEK. It was performed under sub-Tenon anesthesia. The pupil was preoperatively treated with mydriatic agents. Tropicamide and phenylephrine were used the same morning to achieve mydriasis. Maximum pre-operative pupil dilation was noted. Phacoemulsification was performed, and the foldable intraocular lens (IOL) was placed in the bag. Five PEX-syndrome patients who needed transscleral-sutured IOL implantation due to zonular dialysis were excluded from this study.

Surgical Procedure of DMEK
The graft edges were stained using 0.1% Brilliant Blue G (BBG) 250 (BBG; Sigma-Aldrich, St. Louis, MO, USA) (1.0 mg/mL) during peeling. A punch was gently placed on the endothelial surface to indent a circle 7.75, 8.0, or 8.25 mm in diameter. Next, 1.0-and 1.5-mm-diameter dermatological biopsy punches (Kai Industries, Seki, Japan) were used to place asymmetric marks on the edges of the identified circles [16]. Donor grafts were cut using the donor punch, stained with 0.1% BBG for 1 minute, and stored in a balanced salt solution (BSS) (BSS-plus; Alcon, Osaka, Japan) for approximately 30 min prior to insertion [17].
All surgeries were performed under local anesthesia. After establishing a retrobulbar block and a Nadbath facial nerve block, two paracenteses and a 2.8-mm upper corneal or corneoscleral incision were made for the recipient cornea. Peripheral iridotomy was performed at the 6-o'clock position using a 25-gauge vitreous cutter to prevent the development of a postoperative pupillary block. After central recipient descemetorhexis under air, the donor membrane graft was placed into an IOL injector (model WJ-60M; Santen Pharmaceuticals, Osaka, Japan) and inserted into the anterior chamber [5].
The inserted graft was unfolded using a no-touch technique with shallowing of the anterior chamber [18]. After the correct orientation was confirmed, the anterior chamber was filled with air to adhere the graft to the host cornea. Fifteen minutes later, the air was partially replaced with BSS. Finally, 0.4 mg of betamethasone (Rinderon; Shionogi, Osaka, Japan) was subconjunctivally administered in 1.5% (w/v) levofloxacin eye drops (Cravit; Santen Pharmaceuticals).

Statistical analysis
Male/female and right/left ratios were compared using the χ2 test. The paired t-test was used to compare preoperative and postoperative values; the unpaired t-test was used to compare the PEX and FECD groups. Moreover, multiple regression analysis was performed after age adjustment. All analyses were performed using JMP 13.2.0 (SAS Institute Inc., Cary, NC, USA). A p-value of <0.05 was considered to be statistically significant.

Patients
The preoperative patient profiles are summarized in Table 1. As shown in Supplemental Figure 1, even in PEX patients with severe corneal edema, the cornea edema disappeared and a completely clear cornea was achieved after cataract surgery and DMEK. The mean age of the PEX group was 79.7 ± 5.1 (from 75 to 85 years old); that of the FECD group was 70.4 ± 8.6 (from 55 to 81 years old). The mean age of the PEX group was significantly higher than that of the FECD group (p=0.037). Preoperative BSCVA and CCT (i.e., before cataract surgery) were not significantly different between the PEX and FECD groups (BSCVA; p=0.492, CCT; p=0.710). Preoperatively, none of the patients were diagnosed with secondary open angle glaucoma with optic nerve damage. Mean pupil diameter was smaller in the PEX group than in the FECD group (p=0.018) and three eyes in the PEX group required capsule expanders because of zonular weakness during cataract surgeries. All cataract surgeries were uneventful. months in the FECD group; the paired t-test was used for analysis of both groups), and there was no significant difference between the two groups at all examination points (p=0.71 preoperatively, and p=0.24, 0.86, and 0.82 at 1, 3, and 6 months, respectively; Figure 2).

Corneal Endothelial characteristics
In the PEX group, the coefficient of variation (CV) in cell area was 27.2 ± 7.0% at 6 months after DMEK. In the FECD group, CV was 33.1 ± 2.6% at 6 months after DMEK. There was no significant difference between the two groups (p=0.12). In the PEX group, cell hexagonality (HEX) was 46.3 ± 13.5% at 6 months after DMEK. In the FECD group, HEX was 55.9 ± 9.2% at 6 months after DMEK.
There was no significant difference between the two groups (p=0.21).

Complications after DMEK
None of the eyes showed intraoperative complications, and none revealed primary graft failure. Four eyes (50%) of the FECD group required rebubbling for partial detachment. CME was present in one eye (20%) of the PEX group and one eye (12.5%) of the FECD group. In all affected eyes, the CME resolved within 6 months after the surgery with topical 0.1% (w/v) bromfenac eye drops (Bronuck; Senju Pharmaceuticals) and sub-Tenon injection of 40mg/mL triamcinolone acetonide (Kenacort A; Bristol-Myers Squibb). None of the eyes demonstrated postoperative intraocular pressure elevation or exhibited glaucoma.

Discussion
The current study indicates that DMEK surgery can successfully be performed for eyes with PEX syndrome. Postoperative BSCVA and CCT were significantly improved in both the PEX and FECD groups, even though the final BSCVA was significantly worse and the final ECD was significantly less in the PEX group compared to the FECD group. To the best of our knowledge, this is the first study that 10 focused on the outcomes of DMEK for BK due to PEX and FECD. was significantly worse in the PEX group (52.7 ± 11.7%) [19].
Notably, the mean age of the PEX group was greater than that of the FECD group in the current study.
The prevalence of PEX increases progressively with age, and the diagnosis of PEX is rarely made in individuals younger than 50 [20,21]. However, the prevalence of FECD does not significantly increase with age [22]. Subclinical dysfunction of the macula, optic nerves, and brain due to increasing age may also contribute to the lower BSCVA observed in the PEX group. However, after we conducted age adjustment, BSCVA and ECD loss at 6 months after DMEK remained inferior in the PEX group, relative to the FECD group.
We speculated three possible causes for the relatively deteriorated postoperative BSCVA in the PEX group: PEX material may affect the centering of the IOL, posterior segment structures, and/or cognitive function.
It has been well-documented that PEX material deposits on the lens zonule cause zonular weakness and contributes to the dislocation of implanted IOLs. Ostern et al. investigated the long-term positioning of the posterior IOL following cataract surgery in eyes with and without PEX and reported that IOLs within the capsular bag were more prone to decentration in eyes with PEX [23]. The 11 dislocated IOL could lead to more higher-order aberration, resulting in the decreased BSCVA.
PEX material has also been reported to accumulate on posterior segment structures such as the choroid and optic nerve. In eyes with PEX, choroidal thinning related to increase vascular resistance and reduce blood flow has been reported [24]. The optic disc area has also been reported as being smaller than controls, both with and without glaucoma [25].
PEX material deposits have also been reported on the cerebral meninges. Magnetic resonance imaging of PEX patients with or without glaucoma showed a higher prevalence of white matter hyperintensities than controls [26]. Chronic cerebrovascular disease including senile dementia, cerebral atrophy and cerebral ischemia is reportedly more common in patients with PEX than patients with primary open-angle glaucoma [27]. These changes to the posterior structures of the visual pathway may deteriorate BSCVA in patients with PEX.
In a manner similar to that of BSCVA, ECD loss at 6 months after DMEK was worse in the PEX group.
The suggested causes of endotheliopathy include penetration of PEX material towards the Descemet's membrane [28,29] and changes in the blood-aqueous barrier [30,31]. Japan, under the reference number (YKH_30_02_08). Informed consent was obtained for each surgery.
We carefully followed all ethical principles within the Declaration of Helsinki.

Consent for publication
Not applicable Availability of data and material The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
All the authors declare that they have no competing interests.      group, the donor corneal endothelial cell density (ECD) decreased from 2,719 ± 222 cells/mm2 preoperatively to 1,272 ± 339 cells/mm2 at 6 months (53.3 ± 11.6% less than the preoperative value of the donor graft). In the Fuchs' endothelial corneal dystrophy (FECD) group, the donor corneal ECD decreased from 2,694 ± 116 cells/mm2 preoperatively to 1,954 ± 434 cells/mm2 at 6 months (27.5 ± 16.3% less than the preoperative value of the donor graft). There was no significant difference between the two groups at the preoperative point and at 1 month (p=0.84 and 0.084, respectively; Mann-Whitney U test), whereas the PEX group was significantly worse than the FECD group at 3 and 6 months (p=0.015 and 0.016, respectively; Mann-Whitney U test).

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.
Supplementary Figure 1 with legend.pdf