Treatment of Severe Ocular Surface Disorders with Albumin Eye Drops

Abstract

Purpose:

Promising healing effects by albumin eye drops (AED) on the corneal epithelium have been demonstrated. The aim of this study was to analyze retrospectively if AED are an effective treatment for severe ocular surface defects and if recurrences can be reduced.

Methods:

Seventeen patients with persistent epithelial defects (PED) and 30 patients with sterile corneal ulcers (CU) were treated either with 5% AED or 0.1% hyaluronan eye drops (HED) 8 times daily until complete epithelium closure. Sizes of the corneal defects, length of treatment and follow-up period, as well as recurrence rate were evaluated. Follow-up of PED was carried out for 126±21 days in the AED group and 142±39 days in the HED group. For CU cases, regular follow-up visits were performed for 117±15 days in the AED group and 112±28 days in the HED group.

Results:

Nine patients in PED group were treated for 11±4 days with AED and the HED group (8 patients) was treated for 10±3 days until corneal epithelium was closed, with only one recurrence in the HED group (P=0.67). Fifteen patients in CU group were treated for 18±6 days with AED and the HED group (15 patients) was treated for 21±7 days until epithelial closure, with 1 recurrence in the AED group and 8 in the HED group (P=0.03).

Conclusion:

This comparative retrospective case–control study showed that AED are able to close corneal epithelial defects like CU and PED as good as HED. Even though times of healing were similar for both treatments, AED seem to reduce the recurrence of sterile CU. Our findings suggest that AED might be beneficial in the management of ocular surface epithelial damage; however, further larger studies are necessary to confirm these results.

Introduction

Defects of the corneal epithelium and stroma can persist for an extended time despite high-frequency therapy with eye drops and ointments for lubrication.^1–3 Beyond this, the included preservatives in some artificial tear solutions, such as benzalkonium chloride, can have toxic effects on the corneal epithelium.² Since first described by Fox et al. in 1984, eye drops made from autologous serum (ASED) have become an effective alternative to artificial tears in severe ocular surface disorders.^4–17 Several studies showed supportive properties of ASED for healing of persistent epithelial defects (PED) and corneal ulcers (CU).^18–20 Unfortunately, the production of ASED falls under the federal regulations on blood transmission in Germany, as it is also in many other European countries. Therefore, in Germany, the production of ASED can only be done with the support of a specialized and certified blood bank. This subsequently increases costs, complicates logistics, and limits the production of ASED to few hospitals after licensing the process of production.^21,22 Due to these facts, an effective alternative with similar action to ASED is eligible.

Tears are believed to be derived from serum.²³ Albumin is one of the main protein components in serum and tears with a concentration of albumin in tears of 10–12 μg/mL, whereas in serum the concentrations are much higher between 35 and 53 mg/mL.¹² Albumin is the prevalent plasmatic transport protein in blood and has the function of keeping the intravascular colloid osmotic pressure in balance. Physiologically, albumin is synthesized in the liver, hydrolyzed, and excreted through the kidneys, intestines, and liver. Main functions are antioxidative and it is also an important plasmatic transport protein for amino acids, fatty acids, and growth hormones, thus supporting the cellular metabolism.²⁴

Albumin obtained from human resources has now been available for more than 20 years as a ready-to-use medical product and there are published standards for its processing to eye drops as an alternative to ASED for the treatment of severe ocular surface diseases.²⁵

Shimmura et al. were the first to describe in 2003 the supportive effect of albumin eye drops (AED) on cultured corneal epithelial cells by demonstrating a decreased caspase-3 activity when adding albumin to the culture medium.²³ The first results on the usage of 10% albumin eye drops in a human cohort of 9 patients with severe dry eye disease due to Sjögren's disease showed after 4 weeks of application, a measurable reduction of discomfort and improvement of fluorescein and rose bengal staining and tear break-up time.²³ Later, Higuchi et al. showed that induced corneal erosions heal significantly faster when 10% albumin solution was regularly applied in comparison to the isotonic NaCl solution in a dry eye rat model.²⁶ Further studies on the specific use of AED in corneal defects are to the best of the authors' knowledge not published so far.

The aim of our study was to evaluate retrospectively if AED are an effective treatment for severe ocular surface defects like CU and PED and if recurrences can be reduced in comparison to standard of care with 0.1% hyaluronan eye drops (HED).

Methods

The University of Wurzburg Ethics Committee approved this study. Declaration of Helsinki was followed at all times. In this comparative, retrospective case–control study, records of patients treated for CU and PED at the University Eye Hospital in Wurzburg, Germany and at the Eye Clinic of the St. Johannes Hospital in Dortmund, Germany were reviewed. History was taken and the etiology of the corneal diseases was determined. CU was defined as tissue excavation associated with epithelial and stromal defect. Postinfectious CU was confirmed by a microbiological examination without evidence of germs before starting AED treatment.

All patients had undergone a complete eye examination at the beginning of treatment either with 5% AED or 0.1% HED. The epithelial defect size in PED and CU was measured in millimeters by corneal staining with sodium fluorescein solution and slit lamp examination with cobalt blue illumination recording the minimal and maximal diameter of the lesion. From these 2 values, the mean lesion diameter was calculated (adding both values and dividing by 2) for further evaluation.

Five percent albumin or hyaluronan 0.1% eye drops were prescribed 8 times daily from 8 AM to 22 PM (every 2 h) or for up to hourly use if necessary. All CU and all large PED were additionally treated with antibiotic eye drops (ofloxacin 3 mg/mL unpreserved 6 times daily). Two patients of each treatment group in PED (AED and HED) were treated without antibiotics due to the small size of the lesion.

Treatment period with AED or HED and the total follow-up interval were recorded. Treatment period was defined as the time from commencement of treatment until complete closure of corneal surface epithelium in CU or PED cases and the absence of fluorescein staining. Anterior eye segment evaluations were analyzed until last outpatient examination, defined as end of the follow-up period. Patients were advised to present immediately if disease symptoms occurred again. Relapses of CU or PED until the last follow-up visit were recorded.

The follow-up period in PED in the AED group was 126±20 days and in the HED group was 146±38 days. CU patients in the AED group were followed up for 117±15 days and in the HED group for 112±28 days.

Production of albumin eye drops

Five percent albumin solution for infusion can be purchased from the pharmacy as a quality control medical product in different quantities. Because 5% albumin solution is a ready-to-use medical product, no further production steps like serological testing or centrifugation have to be carried out before administration.

Due to the necessity for aseptical preparation, bottling has to be done at a sterile workbench. First, the 5% albumin solution has to be drawn into a syringe. The solution is then filled into the single eye drop containers through a 0.2-μm cellulose acetate filter. The eye drop containers are then closed and labeled. Five percent albumin eye drops prepared like this can be stored in the refrigerator (at 6°C–8°C) for up to 4 weeks. The single eye drop container has to be discarded 24 h after opening.

Statistical analyses

All statistical analyses were performed using SPSS Version 15 (IBM, Armonk, NY). The nonparametric Mann–Whitney U-test was performed for calculation of significant differences in the examined parameters in CU and PED treatment groups. A P-value<0.05 was considered significant.

Results

A total of 47 patient records in both centers could be identified as eligible for retrospective data analysis between 2006 and 2013. The history of reasons for PED or CU is shown in Table 1. Most often, corneal problems arose from trophic changes, secondary to rheumatic diseases, and only 3 patients showed evidence of postinfectious causes for corneal changes.

Table 1.

Etiologies of All Corneal Ulcers and Persistent Epithelial Defect Cases Within the 4 Examined Groups

	Albumin 5%		Hyaluronan 0.1%
	CU	PED	CU	PED
Trophic	6	2	6	2
Rheumatic	3	3	4	—
Toxic	3	1	2	4
Tear deficiency	2	3	1	2
Postinfectious	1	—	2	—
Total	15	9	15	8

CU, corneal ulcers; PED, persistent epithelial defects.

Persistent epithelial defects

A total of 17 patient records with PED were analyzed. Nine eyes of 9 patients were treated with AED and 8 eyes of 8 patients were treated with HED (Table 2). Mean age in the AED group was 64.4±12.2 years and in the HED group it was 69.4±15.6 years (P=0.37). In both groups, there was a narrow majority of male patients. The mean diameter of corneal defect in the AED group was 2.96±0.75 mm and in the HED group was 3.33±0.73 mm, with no significant difference in mean size of lesion (P=0.89). Treatment period in the AED group with albumin 5% was 11±4 days and in the HED group with hyaluronan 0.1% was 10±3 days (P=0.67). Within the follow-up period, only 1 patient in the HED group showed a recurrence after 65 days (13%); the treatment was repeated and no further recurrence occurred until end of follow-up. No significant difference between the outcome in AED and HED groups could be found (P=0.74).

Table 2.

Demographics and Results of Patients Treated for Persistent Epithelial Defects Within 2 Groups With Albumin 5% and Hyaluronan 0.1%

Persistent epithelial defects	Unit	Albumin 5% Value (SD)	Hyaluronan 0.1% Value (SD)	P
Number of eyes		9	8
Age	Years	64.4 (±12.2)	69.4 (±15.6)	0.37
Sex	Male:female	5:4	5:3
Mean diameter of corneal defect	mm	2.96 (±0.75)	3.33 (±0.73)	0.89
Treatment period	Days	11 (±4)	10 (±3)	0.67
Follow-up period	Days	126 (±20)	142 (±38)	0.31
Recurrence during follow-up		0 (0%)	1 (13%)	0.74

SD, standard deviation.

Corneal ulcers

Thirty patient records with CU could be identified. Fifteen eyes of 15 patients were treated with AED and 15 eyes of 15 patients were treated with HED (Table 3). The mean age in the AED group was 64.8±18.5 years and in the HED group it was 70.3±17.9 years (P=0.49). In both groups, there were a majority of female patients. Corneal defects in the AED group showed mean diameters of 3.29±0.75 mm and 2.94±0.81 mm in the HED group with no significant difference in mean size of lesion (P=0.12). Treatment period in the AED group with albumin 5% was 18±6 days and in HED group with hyaluronan 0.1% was 21±7 days (P=0.62). Within the follow-up period, 8 patients in the HED group (53%) and 1 in the AED group (7%) showed recurrences after 31–70 days after commencement of treatment. There was a significant difference in the recurrence rate between both groups (P=0.03). Analysis of the treatment regimen showed no effects of the frequency of eye drop application on recurrence rate.

Table 3.

Demographics and Results of Patients Treated for Corneal Ulcers Within 2 Groups With Albumin 5% and Hyaluronan 0.1%

Corneal ulcer	Unit	Albumin 5% Value (SD)	Hyaluronan 0.1% Value (SD)	P
Number of eyes		15	15
Age	Years	64.8 (±18.5)	70.33 (±17.9)	0.49
Sex	Male:female	6:9	6:9
Mean diameter of corneal defect	mm	3.29 (±0.75)	2.94 (±0.81)	0.12
Treatment period	Days	18 (±6)	21 (±7)	0.62
Follow-up period	Days	117 (±15)	112 (±28)	0.94
Recurrence during follow-up		1 (7%)	8 (53%)	0.03^a

Indicates statistical significance (P<0.05).

Discussion

Treatment of PED and CU is still a challenge for both ophthalmologists and patients.

ASED have demonstrated their ability and effectiveness for treating severe dry eye changes like in Sjögren's syndrome, trophic CU, and recurrent corneal erosions in many trials.²⁷ However, the use of ASED remains an experimental approach. General preparation and administration protocols are not available due to different regulations in several countries. Furthermore, ASED cannot be produced on a large scale. Recent trials tried to dilute ASED down to 50% or 20% to save serum for future use and to limit repeated venous puncture of patients.²⁸ One hundred percent ASED contain 5 times higher concentrations of transforming growth factor-beta (TGF-β) than tears, and it has been shown that TGF-β has an antiproliferative effect that is not desired in healing corneal defects.^27,29 ASED have to be frozen down to −20°C, whereas AED can be stored at 6°C–8°C in normal refrigerators.²⁵

Still being a product of human resources, 5% albumin solution for infusion bears the risk of bacterial, fungal, and viral contamination, although thorough serological testing by the providing company is readily carried out.

The product insert for 5% albumin solution for infusion states that 95% of the containing protein actually is albumin. The remaining 5% are other proteins found in the human serum, such as transferrin, ceruloplasmin, and prealbumin, which can also have wound healing or stabilizing properties on the corneal surface on their own. Transferrin, for example, may act bacteriostatic because of its high affinity to ferric oxide, thus binding serum ferric oxide making it inaccessible for bacterial metabolism.

Up to date, the approach of using AED for the treatment of ocular surface disorders has only been described in few reports.^23,26 It was shown that induced corneal erosions heal faster under therapy with AED, and analysis of function of AED showed antioxidative effects as well as suppression of apoptosis by downregulation of caspase-3 activity.^24,25 The precise mode of action of ASED and AED remains still unclear.

For the treatment of 6 PED patients, Schrader et al. used the combination of ASED 8 times a day and hydrogel contact lens. They could show after 14.2 (±8.9) days a complete closure of PED in 5 of 6 eyes. Only 1 eye needed 90 days until epithelial closure of PED. Ziakas et al. used ASED 6 times daily in 33 eyes with PED.³⁰ Epithelial defects closed after 3 days of treatment and ASED treatment was continued for 6 months. Eighty-five percent of the eyes showed no relapse and 15% only 1 single relapse within the remarkably long follow-up time of 30 (±6.3) months. Young et al. examined 8 patients with PED treated with ASED; 6 of them showed closure of PED within 2 weeks and 2 remained open after 1 month of follow-up.¹⁸ The size of initial PED was not measured in all of these studies.

The results of the hereby presented study are the first data on the comparison of AED and hyaluronan 0.1% therapy in PED and CU cases. Positive results with fast closure of corneal defects could be shown with the use of AED in this group of patients. Epithelial wound closure could be achieved after a mean of 10–11 days in both treatment groups with similar size of defects. AED and antibiotic treatment were stopped after the closure of the epithelium. During the mean follow-up time of nearly 4 months, in all groups we detected only 1 relapse in the HED group and no recurrence in the AED group. There were no adverse events noted in all treated cases.

Treatment of sterile corneal ulcerations includes the use of topical antibiotics and lubricants. Unpreserved antibiotics were used in all patients to prevent any effect of preservation agents on the healing of corneal defects. Nevertheless, Moreira et al. showed in an animal model that after excimer laser keratectomy in rabbits, the re-epithelialization is significantly prolonged in eyes treated with antibiotics like ofloxacin.³¹ Moreira et al. used ofloxacin with benzalkonium chloride for preservation, whereas we used unpreserved ofloxacin in this study; this may be the reason that patients treated without ofloxacin showed no trend to faster re-epithelization than patients using ofloxacin, but the number of patients in these groups is too low to state any comment on re-epithelialization time with unpreserved ofloxacin.

To the best of the author's knowledge, there are no studies published looking for treatment effects of AED or ASED on a higher number of patients with CU. Semeraro et al. examined 15 eyes with chemical corneal burns akin to CU.²⁷ All patients were treated with ASED for 16 (±6) weeks; at the end of the treatment period, all patients showed closure of epithelial defects and improvement of symptoms.

Our data showed closure of epithelium in CU patients after a mean treatment period of 18 days in AED and 21 days in HED group. The HED group showed a significantly higher recurrence rate after 4 months of follow-up with 8 relapses (53%), whereas in the AED group there was only 1 recurrence (17%). We therefore assume that AED treatment might provide a more distinctive healing of CU than lubricants like hyaluronan 0.1%. Mean time of relapse varied from 31 to 70 days after treatment start. Heyworth et al. examined the natural history and relapse rates of recurrent erosions and could show that recurrences occurred in nearly 60% of patients over a time period of 4 years with a mean time of 60 days between the single recurrence events.³² Despite the shorter follow-up time, the recurrence rate in our PED patients was low after AED or HED treatment. Exact data on the recurrence rate of CU are not found in literature.

A closer look at the frequency of eye drop application in all examined groups showed no trends toward a lower recurrence on the higher eye drop application rate either in the AED or HED groups. Treatment regimen with minimum of 8 times daily with AED therefore seems to be sufficient for lowering the recurrence rate in CU.

A limitation of our study is the retrospective design. Unfortunately, disease duration and treatment before start with AED were not recorded properly and could therefore not been analyzed. Furthermore, a higher number of patients and longer follow-up periods will be necessary for more detailed evaluation and distinction of the most reliable treatment option for PED and CU with AED, ASED, and lubricants like hyaluronan 0.1%. Also, variable healing patterns and time periods until complete defect closure may be present in PED and CU of different etiologies. Therefore, further evaluations should try to focus on similar entities of CU and PED. As shown in Table 1, CU and PED showed different underlying etiologies, but none showed signs of bacterial infection or lid closure defects. Also, the overall healing time did not differ significantly between the comparison groups.

Detailed positive or negative effects of high- or low-frequency daily application of AED cannot be found in literature. However, in our small series, there was no difference in the healing time of corneal defects regarding the frequency of AED application, suggesting that there may be no significant difference between the 2-hourly and hourly applications. Also, it cannot be excluded that patients did not use the eye drops exactly in the given frequency.

The major drawbacks of ASED are logistical and regulatory restrictions. Therefore, the production of AED in a higher volume may be a cheaper and more convenient possibility for treating severe dry eye conditions and corneal epithelial defects.

The price for the production of different eye drops is variable; detailed prices in literature are rare and depend on country-specific factors and laboratory settings as there are special instruments and workbenches necessary for production. Calculated ASED and AED price per bottle and day is quite similar with around 2.60 Euro in Germany.²⁵ Comparing this price to commercially available nonpreserved HED with about 15 Euros per 10-mL bottle for ∼20 days of use; AED and ASED are quite expensive (HED use for 20 days is about one-third the price of AED or ASED). If further studies could confirm that AED and ASED provide advantages in the treatment of CU and PED, reimbursement issues may have to be solved with health insurances.

Nevertheless, the application of AED as ASED is currently off-label use, and special attention is necessary for the patient information and written consent is still needed. AEDs were found to be equivalent to hyaluronan 0.1% eye drops in closing PED and sterile CU. Reduction of repeated recurrences by AED treatment could be shown in CU, but it must be pointed out that the overall number of patients in this retrospective study was low. Further studies comparing treatment outcomes with AED and ASED in larger populations are needed.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Poon

A.C.

, Geerling

, Dart

J.K.

, Fraenkel

G.E.

, and Daniels

J.T.

Autologous serum eyedrops for dry eyes and epithelial defects: clinical and in vitro toxicity studies. Br. J. Ophthalmol., 85:1188–1197, 2001.

Geerling

, Daniels

J.T.

, Dart

J.K.

, Cree

I.A.

, and Khaw

P.T.

Toxicity of natural tear substitutes in a fully defined culture model of human corneal epithelial cells. Invest. Ophthalmol. Vis. Sci., 42:948–956, 2001.

Kasper

, Kremling

, and Geerling

[Toxicity of a new moistening agent and preservative in vitro]. Ophthalmologe., 105:557–562, 2008.

Tananuvat

, Daniell

, Sullivan

L.J.

, et al. Controlled study of the use of autologous serum in dry eye patients. Cornea., 20:802–806, 2001.

del Castillo

J.M.

, de la Casa

J.M.

, Sardina

R.C.

, et al. Treatment of recurrent corneal erosions using autologous serum. Cornea., 21:781–783, 2002.

Jacobi

, Dietrich

, Cursiefen

, and Kruse

F.E.

[The dry eye. Current concepts on classification, diagnostics, and pathogenesis]. Ophthalmologe., 103:9–17, 2006.

Dogru

, and Tsubota

New insights into the diagnosis and treatment of dry eye. Ocul. Surf., 2:59–75, 2004.

Ferreira de Souza

, Kruse

F.E.

, and Seitz

[Autologous serum for otherwise therapy resistant corneal epithelial defects—Prospective report on the first 70 eyes]. Klin. Monatsbl. Augenheilkd., 218:720–726, 2001.

Fox

R.I.

, Chan

, Michelson

J.B.

, Belmont

J.B.

, and Michelson

P.E.

Beneficial effect of artificial tears made with autologous serum in patients with keratoconjunctivitis sicca. Arthritis Rheum., 27:459–461, 1984.

10.

Geerling

, Maclennan

, and Hartwig

Autologous serum eye drops for ocular surface disorders. Br. J. Ophthalmol., 88:1467–1474, 2004.

11.

Geerling

, and Hartwig

[Autologous serum-eye-drops for ocular surface disorders. A literature review and recommendations for their application]. Ophthalmologe., 99:949–959, 2002.

12.

Geerling

, Unterlauft

J.D.

, Kasper

, Schrader

, Opitz

, and Hartwig

[Autologous serum and alternative blood products for the treatment of ocular surface disorders]. Ophthalmologe., 105:623–631, 2008.

13.

Goto

, Shimmura

, Shimazaki

, and Tsubota

Treatment of superior limbic keratoconjunctivitis by application of autologous serum. Cornea., 20:807–810, 2001.

14.

Schulze

, Sekundo

, and Kroll

[Autologous serum versus hyaluronic acid eye drops for the treatment of corneal erosions after vitrectomy in diabetic patients. A prospective randomized study]. Ophthalmologe., 102:863–868, 2005.

15.

Schulze

S.D.

, Sekundo

, and Kroll

Autologous serum for the treatment of corneal epithelial abrasions in diabetic patients undergoing vitrectomy. Am. J. Ophthalmol., 142:207–211, 2006.

16.

Tsubota

, Goto

, Fujita

, et al. Treatment of dry eye by autologous serum application in Sjogren's syndrome. Br. J. Ophthalmol., 83:390–395, 1999.

17.

Tsubota

, Goto

, Shimmura

, and Shimazaki

Treatment of persistent corneal epithelial defect by autologous serum application. Ophthalmology., 106:1984–1989, 1999.

18.

Young

A.L.

, Cheng

A.C.

, Ng

H.K.

, Cheng

L.L.

, Leung

G.Y.

, and Lam

D.S.

The use of autologous serum tears in persistent corneal epithelial defects. Eye (Lond.), 18:609–614, 2004.

19.

Turkoglu

, Celik

, and Alagoz

A comparison of the efficacy of autologous serum eye drops with amniotic membrane transplantation in neurotrophic keratitis. Semin. Ophthalmol., 29:119–126, 2014.

20.

Schrader

, Wedel

, Moll

, and Geerling

Combination of serum eye drops with hydrogel bandage contact lenses in the treatment of persistent epithelial defects. Graefes Arch. Clin. Exp. Ophthalmol., 244:1345–1349, 2006.

21.

Geerling

, Grus

, Seitz

, Hartwig

, and Schirra

[Legal regulations to produce serum eye drops: when is it necessary, and how can it be obtained?]. Ophthalmologe., 105:632–638, 2008.

22.

Dietrich

, Weisbach

, Seitz

, et al. [Manufacture of autologous serum eye drops for out-patient therapy: cooperation between ophthalmic clinic and transfusion medicine department]. Ophthalmologe., 105:1036–1038, 2008.

23.

Shimmura

, Ueno

, Matsumoto

, et al. Albumin as a tear supplement in the treatment of severe dry eye. Br. J. Ophthalmol., 87:1279–1283, 2003.

24.

Peters

All About Albumin. Waltham, MA: Academic Press; 1996; p. 188–250.

25.

Unterlauft

J.D.

, Kohlhaas

, Hofbauer

, Kasper

, and Geerling

[Albumin eye drops for treatment of ocular surface diseases]. Ophthalmologe., 106:932–937, 2009.

26.

Higuchi

, Ueno

, Shimmura

, Suematsu

, Dogru

, and Tsubota

Albumin rescues ocular epithelial cells from cell death in dry eye. Curr. Eye Res., 32:83–88, 2007.

27.

Semeraro

, Forbice

, Braga

, Bova

, Di Salvatore

, and Azzolini

Evaluation of the efficacy of 50% autologous serum eye drops in different ocular surface pathologies. Biomed. Res. Int., 2014:826970, 2014.

28.

Cho

Y.K.

, Huang

, Kim

G.Y.

, and Lim

B.S.

Comparison of autologous serum eye drops with different diluents. Curr. Eye Res., 38:9–17, 2013.

29.

Gupta

, Monroy

, Ji

, Yoshino

, Huang

, and Pflugfelder

S.C.

Transforming growth factor beta-1 and beta-2 in human tear fluid. Curr. Eye Res., 15:605–614, 1996.

30.

Ziakas

N.G.

, Boboridis

K.G.

, Terzidou

, et al. Long-term follow up of autologous serum treatment for recurrent corneal erosions. Clin. Experiment Ophthalmol., 38:683–687, 2010.

31.

Moreira

L.B.

, Lee

R.F.

, de Oliveira

, et al. Effect of topical fluoroquinolones on corneal re-epithelialization after excimer laser keratectomy. J. Cataract Refract. Surg., 23:845–848, 1997.

32.

Heyworth

, Morlet

, Rayner

, et al. Natural history of recurrent erosion syndrom—a 4 year review of 117 patients, Br. J. Ophthalmol., 82:26–28, 1998.