tear film–oriented diagnosis and tear film–oriented ... · special issue tear film–oriented...

Special Issue

Tear Film–Oriented Diagnosis and Tear Film–OrientedTherapy for Dry Eye Based on Tear Film Dynamics

Norihiko Yokoi1 and Georgi As Georgiev2

1Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan2Department of Optics and Spectroscopy, Faculty of Physics, St. Kliment Ohridski University of Sofia, Sofia, Bulgaria

Correspondence: Norihiko Yokoi,Department of Ophthalmology, Kyo-to Prefectural University of Medi-cine, 465 Kajii-cho, Hirokoji-agaru,Kawaramachi-dori, Kamigyo-ku,Kyoto 602-0841, Japan;[email protected].

Submitted: December 22, 2017Accepted: March 12, 2018

Citation: Yokoi N, Georgiev GA. Tearfilm–oriented diagnosis and tear film–oriented therapy for dry eye based ontear film dynamics. Invest Ophthal-

mol Vis Sci. 2018;59:DES13–DES22.https://doi.org/10.1167/iovs.17-23700

In December 2010 and January 2012, 3% diquafosol sodium ophthalmic solution and 2%rebamipide ophthalmic suspension, respectively, appeared first in Japan as prescription drugsfor the treatment of dry eye (DE). Since then, not only the diagnosis and treatment but alsothe understanding of the pathophysiology of DE have greatly advanced, and a new concept oflayer-by-layer diagnosis and treatment for DE, respectively termed ‘‘tear-film–orienteddiagnosis’’ (TFOD) and ‘‘tear-film–oriented therapy’’ (TFOT) was born. This new concept iscurrently in the process of expanding from Japan to other Asian countries. TFOD is themethod used for the differential diagnosis of DE, which includes aqueous-deficiency DE(ADDE), decreased wettability DE (DWDE), and increased evaporation DE (IEDE), through thedynamics of tear film (TF) and breakup patterns (BUPs) after the eye is opened. BUPs and/oreach diagnosed DE subtype are/is able to distinguish the insufficient components of theocular surface that are responsible for each BUP in a layer-by-layer fashion. Aqueous fluid,membrane-associated mucins (especially MUC16), and the lipid layer and/or secretory mucinsmust be insufficient in ADDE, DWDE, and IEDE, respectively, and this allows for a layer-by-layer treatment to be proposed for each BUP via the supplementation of the insufficientcomponents, using the topical therapy currently available. In Japan, TF breakup is regarded asa visible core mechanism for DE, and an abnormal breakup time (i.e., �5 seconds) andsymptoms are currently used for the diagnosis of DE. Therefore, TFOD and TFOT could be anideal and practical pathway for clinicians to manage DE.

Keywords: dry eye, tear film oriented diagnosis, tear film oriented therapy, tear film breakup,breakup pattern

When healthy eyes are compared with eyes with dry eye(DE) disease through the staining of tears with fluores-

cein, tear-film breakup (TFBU) may not be observed in a healthyeye until around 10 seconds, even when the eye is kept open.In contrast, in DE eyes, when the eye is kept open, fluoresceinbreakup (BU) can generally be observed within 5 seconds asdark spots, often together with epithelial damage. This fasterfluorescein BU of the precorneal tear film (TF), as well as theassociated epithelial damages, has long been regarded in Japanas a sign of visible abnormalities in DE to differentiate DEs fromnormal eyes. In addition, TFBU has been regarded in Japan as avisible core mechanism of DE, and great emphasis has beenplaced on abnormal fluorescein BU time (BUT) (i.e., �5seconds) and epithelial damage of the ocular surface. Moreover,further understanding of short-BUT–type DE (SBUTDE)1,2 haspushed our emphasis in DE more toward the abnormal BUT,because in this type of DE, in spite of minimal association withepithelial damage,1,2 symptoms are equivalent to DE withepithelial damage.3 The common understanding of SBUTDE inAsian countries has become the common definition of DEadopted by the Asia Dry Eye Society,4 and has also producedgreat impact on the definition and diagnostic criteria for DE inJapan.5 On the other hand, according to the Dry Eye Workshop(DEWS)6 or DEWSII7 reports, based on the fact that TFBUcauses hyperosmolarity of tears—which also causes ocularsurface inflammation that ultimately leads to TFBU, thusconstituting a vicious cycle in these abnormalities—hyperos-

molarity and inflammation, respectively, are now regarded asthe key points for the diagnosis and the treatment for DE inother countries.7,8

In regard to the differences between countries as to the‘‘points of emphasis’’ for the diagnosis and treatment for DE,differences in the available prescription drugs are alsopresumably related. In Japan, eye drops that can enhance TFstability are available, while in Western countries, eye dropshaving an anti-inflammatory effect other than that produced bysteroids are available. However, in Japan, although theimportance of inflammation in DE is acknowledged, theinflammation is regarded as a result of the vicious cyclebetween TFBU and damaged corneal surface epithelium, notthe cause of the vicious cycle in DE (Fig. 1).

Recent advancements in eye drop treatments for DE haveresulted in a strengthening of our concept and approach to DEvia the greater emphasis on TFBU as a core mechanism of DE,thus revising the current definition and diagnostic criteria forDE.5 According to our 2006 definition,9 DE is defined as achronic disease of tears and ocular surface epithelium. Incontrast, according to our 2016 criteria,5 DE is a diseasecharacterized by unstable TF. It should be noted that in the2016 criteria,5 the Schirmer 1 test and evaluation of ocular-surface epithelial damage are excluded. Moreover, a probablediagnosis for DE is not included. Hence, the examination solelyinvolves the detection of abnormalities in fluorescein BUT (i.e.,�5 seconds) and positive eye symptoms related to discomfort

Copyright 2018 The Authors

iovs.arvojournals.org j ISSN: 1552-5783 DES13

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Downloaded from iovs.arvojournals.org on 05/14/2019

https://creativecommons.org/licenses/by-nc-nd/4.0/

and visual disturbance. Through this revision, SBUTDE1,2 wasformerly diagnosed as a probable DE owing to negative ocular-surface epithelial damage and Schirmer 1 test scores, but thecompatible extent of symptoms with definite DE3 is diagnosedas definite DE requiring a treatment via the improvement of TFstability. Thus, the recent advancements of eye drop treatmentsin Japan that enhance TF stability have not only made thediagnosis of SBUTDE definitive, but also helped promote theJapanese concept that the instability of TF or TFBU is a visiblecore mechanism of DE.

TF-ORIENTED DIAGNOSIS AND TF-ORIENTED

THERAPY FOR DE

The corneal surface is composed of epithelium and TF, and TFis composed of the lipid layer and aqueous layer. In thisstructure, lipid layer, aqueous fluid, secretory mucins (espe-cially MUC5AC10), and membrane-associated mucins (especial-ly MUC16, the longest11) are the essential components formaintaining TF stability. Therefore, an insufficiency in any oneof these components (i.e., the lipid layer, aqueous fluid,MUC5AC, or MUC16) is thought to result in TF instability, thusleading to TFBU.12–15 As stated above, TFBU must be one of themost important and visible core mechanisms in DE. Thus, ifinsufficient components of the corneal surface can bediscovered, and if the insufficient components are supple-mented by topical therapy, these diagnostic and therapeuticmethods are thought to become an ideal and practicalapproach to DE.16,17

The formation of stable TF is a result of the cooperation ofmany components, each with its own set of functionalities thatmutually complement each other. Thus, when the dynamicbehavior of the TF and its layers is observed until theestablishment of stable TF or until its premature BU, thisinformation would report on the function of each component.If any component is dysfunctional, TFBU would occur (1)before the complete establishment of TF or (2) earlier thannormal, even if the TF is completely established. Based on thisidea, and as previously reported,16,17 the classification of the

fluorescein BU patterns (BUPs) of the precorneal TF can becategorized into five independent essential patterns17 withpathophysiologically different mechanisms depending on theinsufficiency of the ocular surface components. Therefore,through the evaluation of TF dynamics when taking BUPs intoconsideration, the insufficient TF components that areresponsible for each BUP can be found. Furthermore, it wasalso found that through the classification of BUPs, DE can beclassified into three independent subtypes, that is, aqueous-deficient DE (ADDE), decreased wettability DE (DWDE), andincreased evaporation DE (IEDE), and that aqueous fluid,membrane associatedmucins (especially MUC1611), and lipidlayer and/or secretory mucins (especially MUC5AC10) must beinsufficient in ADDE, DWDE, and IEDE, respectively. Based onthe BUPs and/or the DE subtypes, it will be possible to proposewhich TF components should be replenished/supplementedby the currently available topical formulations. This novelconcept of a layer-by-layer diagnosis and therapy for DE iscoined ‘‘TF-oriented diagnosis’’ (TFOD) and ‘‘TF-orientedtherapy’’ (TFOT), respectively.4,16,17 The goal of TFOD is toclassify DE subtypes via the observation of TF dynamics untilTFBU appears, the BUPs, and whether or not the TFBU rapidlyexpands while the eye is kept open.

PROCESS FOR THE ESTABLISHMENT OF PRECORNEAL TFIN NORMAL EYES

To properly understand the TF behavior (TF dynamics) andTFBU after the eye is opened, it is essential to understand theprocess for the establishment of precorneal TF in normal eyes.When the eye is opened, the upper tear meniscus (TM) pullsup the aqueous tears retained at the lower TM18 and depositsthem at the corneal surface.16,17 The deposition process isassisted via the hydrophilic nature of the membrane associatedmucins (especially MUC 16, the longest among them).16,17 As asecond step after the eye is opened, the TF lipid layer (TFLL)spreads upwards, driven by the surface tension gradientbetween the lipid-covered surface near the lower TM and theTFLL-deficient upper tear surface.19–22 However, this upward

FIGURE 1. A simplified scheme of the different ideas about the mechanism of the vicious cycle of dry eye, stemming from Japan and from thehyperosmolarity/inflammation school of thought. In Japan, more attention is paid to the tear-film instability or breakup when considering thevicious cycle. In the hyperosmolarity/inflammation school of thought, hyperosmolarity and resultant inflammation are included in the vicious cycle.In Japan, inflammation is generally regarded as a result, not as a cause, of the vicious cycle. Those differing ideas may influence the emphasizedfindings in the diagnosis and therapy of dry eye. GC, goblet cell; LG, lacrimal gland; MG, meibomian gland.

TFOD and TFOT IOVS j Special Issue j Vol. 59 j No. 14 j DES14


spread of the TFLL simultaneously drags the underlyingaqueous tears upward.19,20 Combined with the suctionpressure from the lower TM (meniscus-induced TF thinning18),this results in temporarily thinner aqueous TF at the inferiorpart of the cornea. Within this thinner TF area at the inferiorpart of the cornea, when TFBU is not triggered, precorneal TFis subsequently completely established. In normal eyes, thecomplete establishment of precorneal TF takes approximately2 seconds.23 During the upward spread of the TFLL, due to theupward drag of the underlying aqueous tears, just behind theleading edge of the spreading TFLL, a dimple (i.e., a type oftransient thinning) is expected to be formed24–27 that graduallydisappears until the establishment of the TF. It agrees well withthe experimentally measured ~1-lm decrease of aqueous tearthickness over the central cornea induced by the upward dragof aqueous tears by the spreading TFLL.28,29 After the completeestablishment of TF, a black line26 (meniscus-induced TFthinning18) perches the tears at the menisci, and stable, gel-like23 precorneal TF is formed.26

TF BUPS IN DE

Even in normal eyes, TFBU occurs after the completeestablishment of TF when the eye is kept open. However, itshould be noted that in DE cases, there are four otherfundamental types of TF BUPs17 that occur, based on thepathophysiologically different mechanisms, and they charac-terize the difference of DE subtype (Fig. 2). As was previouslyreported,17 to effectively diagnose the BUPs, before theobservation of fluorescein BUPs, the following steps shouldbe performed: (1) Not to increase tear volume, it is essential toperform a less invasive method for staining tears, that is, afluorescein strip being vigorously shaken and just touching thecentral top of the strip to the lower lid margin. (2) After severalblinks, verbally instruct the patient to briskly open the eyeafter gently closing the eye as a kind of provocative test todiscover the hidden BU; it should be observed whether or notrapid expansion of BU can be seen when the eye is kept open.In the classification of BUPs, reproducible BUPs must beregarded as more important, which is more related to theirpathophysiology.

The presented classification of BUPs in DE is based on thedetailed biophysical and surface chemistry concepts developedover decades of basic science research and on the statisticalanalysis of extensive clinical data, as discussed in detail in ourprevious study.17 The clinical study involved 106 DE patientsfor which the following assessments were performed: DE-related symptoms when using the visual analog scale (100 mm¼ maximum), tear meniscus radius (mm), TF lipid layerinterference grade (grades 1–5; 1 ¼ best) and spread grade(grades 1–4; 1 ¼ best), noninvasive BU time (seconds) of TF,fluorescein BU time (seconds), corneal-epithelial damage score(15 points¼maximum), ocular surface epithelial damage score(9 points ¼ maximum), and the Schirmer 1 test (mm). Thecategorization of the patients into five characteristic BUPs wasachieved via a statistical approach using discriminant analysis.

In our previous study,17 as well as in this current reviewsummary, only ‘‘pure’’ BUPs are discussed from the analyzedpatient sample. In addition, in our previous study, meibomiangland dysfunction (MGD) patients were not enrolled, since thesymptoms of MGD may not be derived only from ‘‘pure’’ BUPsand IEDE, but from associated marginal blepharitis andaccumulated lipids within the ducts of the meibomian glands.Thus, the combination of different DE subtypes can sometimesexist in a single individual (e.g., ADDE with increasedevaporation due to MGD), resulting in several BUPs beingmanifested in the patient’s eyes depending on the severity of

the underlying DE core mechanisms. Although the analysis ofsuch cases is ongoing research and will be reported in futurestudies, it should be noted that the clinically importantinterpretation for DE based on BUPs is included in the sectionson TF BUPs in ADDE and IEDE shown below in order to avoidany misinterpretation of the correct classification of DEsubtype based on the BUP.

TF BUPS IN ADDE

In the most severe ADDE cases, owing to the deficiency of theaqueous component, tear fluid cannot be uniformly depositedacross the cornea, which results in an area lacking in aqueouscoverage corresponding to characteristic TFBU. This BUP,which occurs during eye opening, is termed ‘‘area break’’(AB).16,17 Using fluorescein, in the severest form of ADDE,upward movement of fluorescein-stained aqueous tears cannotbe confirmed, while only the severe ocular surface epithelialdamage and the lower height of the TM can be observed.However, in relatively less severe cases, upward movement offluorescein-stained aqueous tears can be observed just withinthe inferior part of the cornea (may be appropriately coined as‘‘partial AB’’). Also, in partial AB as well, severe punctatestaining of the ocular surface epithelium within the palpebralzone may be seen.

In mild to moderate ADDE cases, at the thinner aqueous TFarea in the lower part of the cornea that is susceptible to TFBU,TFBU is likely to occur during the upward movement ofaqueous tears owing to the simultaneous action of this upwardmovement and TF thinning induced by the lower TM,18 whichappeared as a line-like BUP when using fluorescein. We coinedthis BU, which is seen after the eye is opened, as ‘‘line break’’(LB).16,17 Theoretically, this BU is facilitated by the greatersuction effect18 and thinner aqueous TF30 in cases with lessaqueous tear volume. Around LB, corneal epithelial damage isgenerally observed together with conjunctival epithelialdamage within the interpalpebral zone. In SBUTDE,1,2 a casewith LB sometimes occurs with rapid expansion of the BUregion, together with no or minimal corneal epithelial damageand apparently normal height of the lower TM. In our currentthinking, this BUP might be associated with decreased cornealwettability31–33 rather than aqueous tear deficiency, and DEwith this BUP should be properly classified as DWDE, not asADDE. In this type of DE, owing to the decreased wettability ofpossibly the inferior part of the cornea, aqueous deposition isless than normal, in spite of normal meniscus tear volume, andthis may be why LB with its rapid expansion is likely to occur.However, further study is needed to support this theory.

TF BUPS IN DWDE

Even when the aqueous tear volume is sufficient, acceleratedTFBU can occur, probably in relation to decreased cornealwettability.31–33 This BU can be seen (1) instantaneously duringeye opening at the deposition process of aqueous tears on thecornea, and/or (2) during the upward spread of the TFLL afterthe eye is opened, during which process a dimple24,25 passesover the cornea and BU occurs when the dimple passes overthe corneal surface, at which point the wettability is impaired.The former BUP is coined as ‘‘spot break’’ (SB)16,17 owing tothe spot-like appearance of the BU, and the smaller SB is likelyto be erased during the upward movement of aqueous tearsafter the eye is opened. The latter BUP is coined as ‘‘dimplebreak’’ (DB),17 because the BU occurs at the dimple site. Theunderlying mechanism of SB and DB is presumably thecontamination of the corneal surface by the TFLL that resultsin decreased wettability.31–33 Therefore, SB and/or DB can be



seen together with LB or partial AB in ADDE, because in ADDE,during deposition and/or the upward movement process ofaqueous tears, the TFLL is closer to the corneal surface thannormal TF owing to diminished aqueous TF thickness inADDE.30

TF BUPS IN IEDE

Even in normal eyes, when the eye is kept open for a longertime, after the complete establishment of TF, TFBU can occur,probably owing to evaporation.31–35 TFLL34–38 and secretorymucin10,38 (MUC5AC) are thought to be components of the TF,contributing to the suppression of evaporation and stability ofTF after its establishment, and insufficiency of those compo-nents would result in earlier TFBU, even after the completeestablishment of TF owing to facilitated evaporation. It shouldbe kept in mind that currently, the capability of TF to resistevaporation and the evaporation-suppressive action of theTFLL are controversial, both in clinical and in vitro studies, andprecise clarification of these phenomena is a topic for futureresearch.39,40 We coined this BUP as ‘‘random break’’ (RB),16,17

because in RB the portion and shape of the BU is not likely tobe reproducible. In RB, it should be noted that the BU isobserved after the cessation of the upward movement offluorescein-stained aqueous tears after the eye is opened,which corresponds to the timing after complete establishmentof precorneal TF after the cessation of upward spread of theTFLL.20,22,23 Even when RB is seen, if the BU area expandsrapidly, it is reasonably suggested that decreased wettabili-ty,31–33 probably due to the impairment of membrane-associated mucins (especially MUC1611), is also associated;

and if we consider the treatment for RB with rapid expansion,not only the facilitated evaporation but also the decreasedwettability should be considered. The BUPs, their character-ization as a DE subtype, insufficiency of components, and thepossible selection of topical treatments are summarized in theTable.

TFOD is a diagnostic method based on tear dynamics. Thegoal of TFOD is to classify DE into one of three subtypes, thatis, ADDE, DWDE, and IEDE, taking into considerationfluorescein BUPs and the expansion of their BU. Through thedetermination of BUPs and DE subtypes, insufficient compo-nents of the TF and corneal epithelium necessary to stabilizeTF can be discovered, and topical therapy to stabilize TF can beproposed as TFOT4,16,17 with reference to the availability oftopical therapy in each country (Fig. 3).

RECENT ADVANCEMENTS IN EYE DROPS FOR TFOT

It should be noted that 3% diquafosol sodium (DQS) eyedrops,41 which produced a paradigm shift in our approach toDE in Japan, can enhance the production of aqueous fluid,42

evaluated for the first in the human eye via meniscometry,43–45

from the conjunctival epithelium and secretory mucin10,46

(MUC5AC) from the conjunctival goblet cells. Moreover, DQSeye drops can enhance the expression of membrane-associatedmucins10,11,47 (MUC1, MUC4, and MUC16) of the cornealsurface epithelium. Considering that artificial-tear eye dropsand hyaluronic-acid eye drops can only increase tear volumefor up to 5 and 10 minutes, respectively,45 the effect of DQS forincreasing tear volume as long as 30 minutes42 is expected tobe effective for treating ADDE via the longer enhancement of

FIGURE 2. Representative fluorescein breakup patterns (FBUPs). (A) Area break (AB). (B) Spot break (SB). (C) Random break (RB). (D) Line break(LB). (E) Dimple break (DB). (F) LB with rapid expansion. FBUPs were classified with reference to (1) when fluorescein BU occurs in relation toupward movement of fluorescein (UMF)–stained aqueous tears, which was observed after the subject’s eye is opened; (2) where fluorescein BUoccurs within the interpalpebral zone of the cornea; and (3) the shape of the BU. AB is diagnosed when UMF is not observed, or limitedly observed,within the lower part of the cornea; this AB is thought to be associated with severe ADDE. SB is diagnosed as a spot-like shape immediately after eyeopening and at least one SB is not erased during UMF; SB is thought to be associated with DWDE. LB is diagnosed as a vertical line-like shape duringUMF at the lower part of the cornea, within which fluorescein intensity becomes decreased with time until the cessation of UMF; LB is thought tobe associated with mild to moderate ADDE. DB is diagnosed as an irregular but vertical line–like shape during UMF within the zone closer to thecentral part of the cornea, within which fluorescein intensity increases with time until the cessation of UMF; DB is thought to be associated withDWDE. RB is diagnosed as an irregular and indefinite shape whose typical place for the BU to occur generally differs with cases and with each blink;RB is thought to be associated with increased evaporation DE. RB must occur after the cessation of UMF. LB is generally accompanied by superficialcorneal epithelial damage at the lower part of the cornea. However, LB can occur with rapid expansion of the FBU with minimal or no superficialcorneal epithelial damage, which is generally thought to be associated with DWDE.



TF stability. In addition, its effect has been reported in ADDEassociated with Sjogren’s syndrome.48 Moreover, it has beenreported that the effect of DQS can possibly increase tearvolume, irrespective of the lacrimal gland function,49,50 whichmust also be favorable when treating ADDE in Sjogren’ssyndrome cases.48,50 In TFOT using 3% DQS eye drops, theimprovement of TF stability can be expected, not only throughthe supplementation of aqueous fluid and MUC5AC to the TF,but also through the supplementation of the membrane-associated mucins (especially MUC1611) to the corneal surfaceepithelium to enhance its wettability; and those two actionspresumably contribute to the reported continuous improve-ment of DE in the long-term clinical use of DQS.51 MUC16reportedly plays a key role in ensuring the low corneal contactangle,52 and its spatial distribution reportedly differs betweenhealthy eyes and dry eyes.53–55 Alternatively, the glycocalyxmay become contaminated with lipids, for example, due todimple formation below lipid ‘‘globs,’’37 and lipid particlessupposedly precede the spreading of the major part of theTFLL. A synergistic action of both mechanisms is also possible.An interesting hindsight in relation to these points is providedby the impact of DQS (P2Y2 purinergic receptor agonist) eyedrops, which reportedly rapidly (i.e., within 15 minutes postinstillation) increase the volume of aqueous tear in normal42 or‘‘dry’’ human eyes,49 secretory mucin content in normalhuman eyes,46 and gene expression for membrane-associatedmucins (MUC1, MUC4, and MUC16) in cultured humancorneal epithelial cells.47 Clinically, the use of 3% DQS eyedrops gradually recovers the SB pattern to normal after monthsof treatment,48 closely matching the anticipated course of itsMUC16-recovering action.41,47 The enhanced production ofsecretory mucin MUC5AC can be also very important, as it canpromote the mucoaqueous gel formation that will providemechanical stability of the TF in an open eye and will act as asurface chemical trap shielding the corneal glycocalyx fromlipid contaminations.31,32,56

As previously reported, DE can be classified into ADDE andSBUTDE, and the latter can be further classified into DWDE andIEDE.16,17 Therefore, eye drops that can simultaneouslyreplenish all or most of the TF layers and/or ocular surfaceepithelium can be of great value owing to their potentialimplementation as broad-spectrum DE treatments. A formula-tion such as DQS, which enables the supplementation ofaqueous fluid,42,49 secretory mucins,46 and membrane-associ-ated mucins,47 can be regarded as a perspective compositionfor a layer-by-layer treatment of aqueous tear deficiency andmucin deficiency–related pathologies. In fact, it has beenreported that DQS is effective not only for ADDE,48,50 but alsofor SBUTDE, possibly via the enhancement of mucins.15,57

In addition, in Japan, 2% rebamipide (RBM) ophthalmicsuspension is available, which can enhance the production ofsecretory mucins by increasing the number of conjunctivalgoblet cells,58 and it can also supplement membrane-associatedmucins59 of the corneal surface epithelium. Therefore, DQSand RBM are excellent eye drops for the treatment of DErelated to abnormalities in secretory and membrane-associatedmucins.

TFOD AND TFOT IN ASIAN COUNTRIES AND

CONCOMITANT THERAPY TO TFOT

In Asian countries where DQS has gradually become availableas prescribed eye drops, in which not only ADDE48,50 but alsoSBUTDE,57 including DWDE and IEDE, has become the targetfor TFOT,4,16,17 the interest in TFOD and TFOT seems to haveincreased. Related to this finding, a new standard DE definitioncommon to Asian countries4 has been established as well as theT

AB

LE.

Sum

mar

yo

fFlu

ore

scein

-Sta

ined

Tear

Film

Beh

avio

r,B

reak

up

Pat

tern

s,an

dth

ePat

tern

so

fO

cu

lar-

Surf

ace

Ep

ith

elial

Dam

age

and

the

Inte

rpre

tati

on

of

the

Insu

fficie

nt

Co

mp

on

en

tsas

Pat

ho

ph

ysio

logy

and

Dry

Eye

Sub

typ

e,

asW

ell

asth

eP

oss

ible

Sele

cti

on

Fro

mTo

pic

alT

reat

men

t

Cla

ssifi

cati

on

of

BU

Ps

Sh

ap

eo

fB

U

UM

FA

fter

Eye

isO

pen

ed

(Up

per)

Tim

ing

Ep

ith

eli

al

Dam

age

Path

op

hy

sio

logy

Dry

Ey

e

Su

bty

pe

Dir

ecti

on

for

Th

era

py

(Lo

wer)

Co

rneal

Po

rtio

n

for

BU

tob

eSeen

Co

rnea

Co

nju

ncti

va

Are

ab

reak

Are

a-like

X~

n*I

mm

ed

iate

lyaf

ter

eye

iso

pen

ed

(þþþ

)(þþþ

)Se

vere

aqu

eo

us

defi

cie

ncy

AD

DE

Sup

ple

men

tati

on

of

aqu

eo

us

flu

idW

ho

leco

rnea

(typ

ical

cas

e)

Lin

eb

reak

Lin

e-li

ke

*D

uri

ng

UM

F(þ

~þþ

)(þ

~þþ

)M

ild

tom

od

era

te

aqu

eo

us

defi

cie

ncy

AD

DE

Sup

ple

men

tati

on

of

aqu

eo

us

flu

idLo

wer

par

t

Spo

tb

reak

Cir

cle

-like

}†Im

med

iate

lyaf

ter

eye

iso

pen

ed

(�~þ

)(�

~þ

)D

ecre

ased

co

rneal

wett

abilit

y

DW

DE

Sup

ple

men

tati

on

of

mem

bra

ne-a

sso

cia

ted

mu

cin

Cen

tral~

up

per

par

t

Dim

ple

bre

akLin

e-li

ke

}D

uri

ng

UM

F(�

~þ

)(�

~þ

)D

ecre

ased

co

rneal

wett

abilit

y

DW

DE

Sup

ple

men

tati

on

of

mem

bra

ne-a

sso

cia

ted

mu

cin

Lo

wer

par

t

Ran

do

mb

reak

Ind

efi

nit

e}

Aft

er

the

cess

atio

no

fU

MF

(�~þ

)(�

~þ

)In

cre

ased

evap

ora

tio

nIE

DI

Sup

ple

men

tati

on

of

lip

ids

and

/or

secre

tory

mu

cin

Ind

efi

nit

e

Lin

eb

reak

or

ran

do

m

bre

akw

ith

rap

id

exp

ansi

on

RE

of

the

bre

aku

p

afte

reye

iskep

t

op

en

}A

fter

the

cess

atio

no

fU

MF

(�~þ

)(�

~þ

)D

ecre

ased

co

rneal

wett

abilit

y

DW

DE

Sup

ple

men

tati

on

of

mem

bra

ne-a

sso

cia

ted

mu

cin

Lo

wer

par

to

rin

defi

nit

e

UM

F:}

,q

uic

k;*

,sl

ow

;n

,sl

ow

and

lim

ited

wit

hin

the

low

er

par

to

fth

eco

rnea;

X,

can

no

tb

eo

bse

rved

.E

pit

helial

dam

age:�,

no

ne;þ,

mild

;þþ

,m

od

era

te;þþþ,

seve

re.

RE

,ra

pid

exp

ansi

on

.*

Incas

eo

fp

arti

alA

B,

SBm

ayb

ese

en

imm

edia

tely

wh

en

eye

iso

pen

ed

and

DB

may

be

seen

wh

en

the

eye

iskep

to

pen

.†

InSB

,at

leas

to

ne

SBm

ust

no

tb

eera

sed

un

til

the

cess

atio

no

fU

MF.



2016 version of a new definition of DE5 in Japan. It is atranslation in Japanese from that of the Asia Dry Eye Society. InWestern countries, on the other hand, DQS and/or RBM eyedrops are not available, and the idea of TFOD and TFOT mightpossibly be less favored. However, in Western countries, therehas been an advancement in prescription anti-inflammatoryeye drops other than steroids to treat DE. In the United Statesand in various European countries, cyclosporine eye drops60

are available, and in the United States, in addition tocyclosporine, lifitegrast, an inhibitor of intercellular adhesionmolecule-1,61 has become available for the treatment of DE.

In Japan, we do not deny the importance of anti-inflammatory therapy for DE. The Japanese theory is basedon the idea that the inflammation is not the cause, but theresult, of a vicious cycle (Fig. 1) between TFBU and damagedcorneal surface epithelium. Therefore, steroid eye drops areoften used together with the TFOT, for example, at the start ofTFOT and at exacerbation of the symptoms. As an alternativetreatment to TFOT, for severe ADDE cases presenting AB,16,17

even at present, punctal occlusion of both the upper and lowerpuncta is essential, together with artificial tear eye drops, andthis results in stable TF via the establishment of precornealTF.45 Moreover, for DE cases with meibomian gland dysfunc-tion (MGD), for which RB16,17 is expected, topical therapy62 asTFOT and treatment for MGD (such as using warm compress,lid hygiene, and antibiotic eye drops) must be adopted.

FUTURE DIRECTIONS

In Japan, great attention has been paid to the instability of TF asa visible core mechanism of DE. This concept has beenadopted by Asian countries,4 and it is clinically and practicallyuseful to diagnose and treat DE. TFOD and TFOT are the idealmethods for a clinician to diagnose DE subtype through BUPswith a different pathophysiology, only using sodium fluores-cein, and to choose the best treatment for DE available in eachcountry. In future studies, it should first be validated whetheror not TFOT proposed by theoretically supported TFOD is alsopractically useful, even in the other countries, and the possiblelimitations of the methods should be elucidated and improve-ments proposed. Moreover, if this concept works properly, anoninvasive method for TFOD other than using fluorescein,such as the use of an interferometer,20,22,23,63 should beexplored for easier screening of DE subtype.

As another direction of TFOD, the relationship betweenTFOD and blink-related friction should be elucidated. TFinstability is the manifestation of DE in an open eye, and it isusually only viewed from the perspective of TF’s preventingthe desiccation of the ocular surface epithelium. However, tocomprehensively understand the pathophysiology, ocularmanifestations, and symptoms of DE,64 attention should befocused on blink-related friction as another important mech-anism of DE (Fig. 4). Tears are known to act as a lubricant, withtheir shear-thinning property65 being important to reduce

FIGURE 3. The detailed concept of TFOT version 1 produced by the Dry Eye Society of Japan. TFOT illustrates the target for therapy and possiblecandidates for topical eye therapy currently available in Japan and elsewhere to improve TF stability. However, for the selection of the most effectivetopical therapy, TFOD is essential to detect the insufficient component needed for the ocular surface to stabilize the TF and to classify the DEsubtype. Reprinted from http://www.dryeye.ne.jp/tfot/index.html, with permission from Dry Eye Society.



friction during blinking. However, in DE, owing to quantitative

and/or qualitative abnormalities in tears, increased blink-

related friction may also become the cause for a vicious cyclebetween the lid-wiper region66,67 and the eyeball surface.

Friction-related ocular surface diseases (OSDs), such as lid-

wiper epitheliopathy,66 superior limbic keratoconjunctivi-

tis,68–71 and filamentary keratitis,72,73 are known to be

sometimes associated with DE, especially in ADDE cases.

Moreover, other than these blink-related OSDs, conjunctivo-chalasis,74–76 which is highly prevalent in elderly people,75 is

known to enhance not only TF instability, but also blink-related

friction, and is thought to modify the ocular surface

FIGURE 4. Stratified structure of DE. To comprehensively understand DE, TF breakup and increased friction are taken into consideration as themajor mechanisms of DE. The former mechanism is common to any type of DE when the eye is kept open. The latter mechanism is that whichduring blinking may be more important in aqueous-deficiency DE. Various intrinsic and extrinsic risk factors flow into the two mechanisms, whichform the vicious cycle (VC), from which symptoms result while presenting ocular surface manifestations of DE.

FIGURE 5. An example of implementation of TFOD and TFOT in Japan. In Japan, DE is diagnosed from symptoms and abnormal fluorescein breakuptime (FBUT, �5 seconds). For TFOD, the first step is the classification of the DE subtypes (ADDE, DWDE, or IEDE) based on major BUPs. In thisclassification, SBUTDE is thought to correspond to DWDE and IEDE. Based on BUPs and/or DE subtypes, components that are insufficient andshould be supplemented can be suggested as the target for TFOT. AT, artificial tears; HA, hyaluronic acid; PP, punctal plug; RE, rapid expansion.



manifestations and symptoms in DE. In our recent report onTFOD,17 we have excluded DE cases accompanied by theabove-described friction-related OSDs in order to analyze onlythe relationship between BUPs and objective signs andsubjective symptoms. Therefore, in a future study, theassociation of friction-related OSDs with DE subtype shouldbe elucidated. Furthermore, blink-related friction must berelated to ocular surface inflammation. Therefore, a compre-hensive understanding of the relationship between TFinstability, increased friction, and inflammation must also bea target for further investigation.

In TFOT,5,16,17 DQS and RBM eye drops are useful, as theyboth increase secretory mucins. However, the action by whichthis increase is achieved may differ. DQS can increase MUC5ACwithin 5 minutes.46 In contrast, RBM may increase thesecretory mucin gradually via the increase of goblet cells,58

and this might be the reason why there have been reports thatRBM can effectively treat friction-related OSDs.71,73,77 Consid-ering that the increased secretory mucin may increase theviscosity of tears, there must be cases in which the use of DQSresults in increased friction, leading to the exacerbation of theaccompanied friction-related OSDs in ADDE. Therefore, fromthe point of blink-related friction, there must be cases in whichconcomitant use of a TF stabilizer such as DQS and a lubricantsuch as RBM results in improvement in signs and symptoms ofDE. Further study is necessary to elucidate the optimaltreatment for DE, while being based on TFOD via theconsideration of increased friction as another mechanism ofDE.

CONCLUSIONS

The new concepts of TFOD and TFOT opened another field forthe diagnosis and therapy for DE, based on the dynamics ofprecorneal TF. According to this concept, using fluorescein isall that is needed to look through the DE subtype via theclassification of BUPs and to propose the appropriate choice oftopical therapy based on the instability of TF as a coremechanism for DE (Fig. 5). Moreover, this concept appears tobe very useful and practical for clinicians. Although TFOD andTFOT are concepts first established in Japan, the commercialavailability of both DQS and RBM has been found to befavorable, and the availability of DQS is now expanding fromJapan to Asian countries in combination with the concept ofTFOT.4 Currently, this concept still requires further investiga-tion; however, future study is expected to deepen ourunderstanding of DE.

Acknowledgments

The authors thank John Bush for reviewing the manuscript.

Supported in part by Grants-in-Aid for scientific research from theJapanese Ministry of Education, Culture, Sports, Science andTechnology (Grant No. 16K11269). Funding of the publication feeand administration was provided by the Dry Eye Society, Tokyo,Japan. The Dry Eye Society had no role in the contents or writingof the manuscript.

Disclosure: N. Yokoi, None; G.A. Georgiev, None

References

1. Toda I, Shimazaki J, Tsubota K. Dry eye with only decreasedtear break-up time is sometimes associated with allergicconjunctivitis. Ophthalmology. 1995;102:302–309.

2. Yamamoto Y, Yokoi N, Higashihara H, et al. Clinicalcharacteristics of short tear film breakup time (BUT)-type

dry eye [in Japanese]. Nippon Ganka Gakkai Zasshi. 2012;116:1137–1143.

3. Yokoi N, Uchino M, Uchino Y, et al. Importance of tear filminstability in dry eye disease in office workers using visualdisplay terminals: the Osaka study. Am J Ophthalmol. 2015;159:748–754.

4. Tsubota K, Yokoi N, Shimazaki J, et al. New perspectives ondry eye definition and diagnosis: a consensus report by theAsia Dry Eye Society. Ocul Surf. 2017;15:65–76.

5. Shimazaki J, Yokoi N, Watanabe H, et al. Definition anddiagnosis of dry eye in Japan, 2016 [in Japanese]. Atarashii

Ganka (J Eye). 2017;34:309–313.

6. The definition and classification of dry eye disease: report ofthe definition and classification subcommittee of the Interna-tional Dry Eye WorkShop (2007). Ocul Surf. 2007;5:75–92

7. Bron AJ, de Paiva CS, Chauhan SK, et al. TFOS DEWS IIpathophysiology report. Ocul Surf. 2017;15:438–510.

8. Wei Y, Asbell PA. The core mechanism of dry eye disease isinflammation. Eye Contact Lens. 2014;40:248–256.

9. Shimazaki J, Tsubota K, Kinoshita S, et al. Definition anddiagnosis of dry eye 2006 [in Japanese]. Atarashii Ganka (J

Eye). 2007;24:181–184.

10. Gibson IK. Distribution of mucins at the ocular surface. Exp

Eye Res. 2004;78;379–388.

11. Argueso P. Glycobiology of the ocular surface: mucins andlectins. Jpn J Ophthalmol. 2013;57:150–155.

12. Craig JP, Tomlinson A. Importance of the lipid layer in humantear film stability and evaporation. Optom Vis Sci. 1997;74:8–13.

13. Golding TR, Bruce AS, Mainstone JC. Relationship betweentear-meniscus parameters and tear-film breakup. Cornea.1997;16:649–661.

14. Lemp MA, Dohlman CH, Holly FJ. Corneal desiccation despitenormal tear volume. Ann Ophthalmol. 1970;284:258–261.

15. Shimazaki-Den S, Dogru M, Higa K, Shimazaki J. Symptoms,visual function, and mucin expression of eyes with tear filminstability. Cornea. 2013;32:1211–1218.

16. Yokoi N, Georgiev GA. Tear-film-oriented diagnosis andtherapy for dry eye. In: Yokoi N, ed. Dry Eye Syndrome:

Basic and Clinical Perspectives. London: Future Medicine;2013:96–108.

17. Yokoi N, Georgiev GA, Kato H, et al. Classification offluorescein breakup patterns: a novel method of differentialdiagnosis for dry eye. Am J Ophthalmol. 2017;180:72–85.

18. McDonald JE, Brubaker S. Meniscus-induced thinning of tearfilms. Am J Ophthalmol. 1971;72:139–146.

19. King-Smith PE, Fink BA, Hill RM, et al. The thickness of thetear film. Curr Eye Res. 2004;29:357–368.

20. Yokoi N, Yamada H, Mizukusa Y, et al. Rheology of tear filmlipid layer spread in normal and aqueous tear-deficient dryeyes. Invest Ophthalmol Vis Sci. 2008;49:5319–5324.

21. Berger RE, Corrsin S. A surface tension gradient mechanismfor driving the pre-corneal tear film after a blink. J Biomech.1974;7:225–238.

22. Bron AJ, Tiffany JM, Gouveia SM, et al. Functional aspects ofthe tear film lipid layer. Exp Eye Res. 2004;78:347–360.

23. Yokoi N, Bron AJ, Georgiev GA. The precorneal tear film as afluid shell: the effect of blinking and saccades on tear filmdistribution and dynamics. Ocul Surf. 2014;12:252–266.

24. Afsar-Siddiqui AB, Luckham PF, Matar OK. The spreading ofsurfactant solutions on thin liquid films. Adv Colloid Interface

Sci. 2003;106:183–236.

25. Bull JL, Grotberg JB. Surfactant spreading on thin viscousfilms: film thickness evolution and periodic wall stretch. Exp

Fluids. 2003;34:1–15.



26. Miller KL, Polse KA, Radke CJ. Black-line formation and the‘‘perched’’ human tear film. Curr Eye Res. 2002;25:155–162.

27. Jones MB, Please CP, McElwain DL, et al. Dynamics of tear filmdeposition and draining. Math Med Biol. 2005;22:265–288.

28. King-Smith PE, Fink BA, Hill RM. Evaporation from the humantear film studied by interferometry. Adv Exp Med Biol. 2002;506(pt A):425–429.

29. Benedetto DA, Clinch TE, Laibson PR. In vivo observation oftear dynamics using fluorophotometry. Arch Ophthalmol.1984;102:410–412.

30. Creech JL, Do LT, Fatt I, Radke CJ. In vivo tear-film thicknessdetermination and implications for tear-film stability Curr Eye

Res. 1998;17:1058–1066.

31. Sharma A. Breakup and dewetting of the corneal mucus layer:an update. Adv Exp Med Biol. 1998;438:273–280.

32. Sharma A. Surface-chemical pathways of the tear filmbreakup: does corneal mucus have a role? Adv Exp Med

Biol. 1998;438:361–370.

33. Tiffany JM. Measurement of wettability of the cornealepithelium, II: contact angle method. Acta Ophthalmol

(Copenh). 1990;68:182–187.

34. Mishima S, Maurice DM. The oily layer of the tear film andevaporation from the corneal surface. Exp Eye Res. 1961;1:39–45.

35. Iwata S, Lemp MA, Holly FJ, Dohlman CH. Evaporation rate ofwater from the precorneal tear film and cornea in the rabbit.Invest Ophthalmol. 1969;8:613–619.

36. Tomlinson A, Doane MG, McFadyen A. Inputs and outputs ofthe lacrimal system: review of production and evaporativeloss. Ocul Surf. 2009;7:186–198.

37. King-Smith PE, Reuter KS, Braun RJ, et al. Tear film breakupand structure studied by simultaneous video recording offluorescence and tear film lipid layer images. Invest Oph-

thalmol Vis Sci. 2013;54:4900–4909.

38. Mantelli F, Tiberi E, Micera A, et al. MUC5AC overexpressionin tear film of neonates. Graefes Arch Clin Exp Ophthalmol.2007;245:1377–1381.

39. Willcox MDP, Argueso P, Georgiev GA, et al. TFOS DEWS IItear film report. Ocul Surf. 2017;15:366–403.

40. Georgiev GA, Eftimov P, Yokoi N. Structure-function relation-ship of tear film lipid layer: a contemporary perspective. Exp

Eye Res. 2017;163:17–28.

41. Keating GM. Diquafosol ophthalmic solution 3%: a review ofits use in dry eye. Drugs. 2015;75:911–922.

42. Yokoi N, Kato H, Kinoshita S. Facilitation of tear fluidsecretion by 3% diquafosol ophthalmic solution in normalhuman eyes. Am J Ophthalmol. 2014;157:85–92.

43. Yokoi N, Bron AJ, Tiffany JM, Kinoshita S. Reflectivemeniscometry: a new field of dry eye assessment. Cornea.2000;19(3 suppl):S37–S43.

44. Yokoi N, Bron AJ, Tiffany JM, et al. Relationship between tearvolume and tear meniscus curvature. Arch Ophthalmol.2004;122:1265–1269.

45. Yokoi N, Komuro A. Non-invasive methods of assessing thetear film. Exp Eye Res. 2004;78:399–407.

46. Shigeyasu C, Yamada M, Akune Y. Influence of ophthalmicsolutions on tear components. Cornea. 2016;35(suppl):S71–S77.

47. Takaoka-Shichijo Y, Nakamura N. Stimulatory effect ofdiquafosol tetrasodium in the expression of membrane-binding mucin genes in cultured human corneal epithelialcells [in Japanese]. Atarashii Ganka (J Eye). 2011;28:425–429.

48. Yokoi N, Sonomura Y, Kato H, Komuro A, Kinoshita S. Threepercent diquafosol ophthalmic solution as an additionaltherapy to existing artificial tears with steroids for dry-eye

patients with Sjogren’s syndrome. Eye (Lond). 2015;29:1204–1212.

49. Yokoi N, Kato H, Kinoshita S. The increase of aqueous tearvolume by diquafosol sodium in dry-eye patients withSjogren’s syndrome: a pilot study. Eye (Lond). 2016;30:857–864.

50. Koh S, Ikeda C, Takai Y, Watanabe H, et al. Long-term resultsof treatment with diquafosol ophthalmic solution for aqueous-deficient dry eye. Jpn J Ophthalmol. 2013;57:440–446.

51. Yamaguchi M, Nishijima T, Shimazaki J, et al. Clinicalusefulness of diquafosol for real-world dry eye patients: aprospective, open-label, non-interventional, observationalstudy. Adv Ther. 2014;31:1169–1181.

52. Yanez-Soto B, Mannis MJ, Schwab IR, et al. Interfacialphenomena and the ocular surface. Ocul Surf. 2014;12:178–201.

53. Danjo Y, Watanabe H, Tisdale AS, et al. Alteration of mucin inhuman conjunctival epithelia in dry eye. Invest Ophthalmol

Vis Sci. 1998;39:2602–2609.

54. Liotet S, Van Bijsterveld OP, Kogbe O, Laroche L. A newhypothesis on tear film stability. Ophthalmologica. 1987;195:119–124.

55. Yanez-Soto B, Leonard BC, Raghunathan VK, Abbott NL,Murphy CJ. Effect of stratification on surface properties ofcorneal epithelial cells. Invest Ophthalmol Vis Sci. 2015;56:8340–8348.

56. Fatt I. Observations of tear film break up on model eyes.CLAO J. 1991;17:267–281.

57. Shimazaki-Den S, Iseda H, Dogru M, Shimazaki J. Effects ofdiquafosol sodium eye drops on tear film stability in shortBUT type of dry eye. Cornea. 2013;32:1120–1125.

58. Kase S, Shinohara T, Kase M. Effect of topical rebamipide onhuman conjunctival goblet cells. JAMA Ophthalmol. 2014;132:1021–1022.

59. Itoh S, Itoh K, Shinohara H. Regulation of human cornealepithelial mucins by rebamipide. Curr Eye Res. 2014;39:133–141.

60. Wan KH, Chen LJ, Young AL. Efficacy and safety of topical0.05% cyclosporine eye drops in the treatment of dry eyesyndrome: a systematic review and meta-analysis. Ocul Surf.2015;13:213–225.

61. Keating GM. Lifitegrast ophthalmic solution 5%: a review indry eye disease. Drugs. 2017;77:201–208.

62. Geerling G, Tauber J, Baudouin C, et al. The internationalworkshop on meibomian gland dysfunction: report of thesubcommittee on management and treatment of meibomiangland dysfunction. Invest Ophthalmol Vis Sci. 2011;30;52:2050–2064.

63. Yokoi N, Takehisa Y, Kinoshita S. Correlation of tear lipid layerinterference patterns with the diagnosis and severity of dryeye. Am J Ophthalmol. 1996;122:818–824.

64. Yokoi N. Frontiers in dry eye therapy-TFOT (tear film orientedtherapy) [in Japanese]. Med Sci Digest. 2014;40:112–115.

65. Tiffany JM, Pandit JC, Bron AJ. Soluble mucin and the physicalproperties of tears. Adv Exp Med Biol. 1998;438:229–234.

66. Korb DR, Greiner JV, Herman JP, et al. Lid-wiper epitheliop-athy and dry-eye symptoms in contact lens wearers. CLAO J.2002;28:211–216.

67. Knop N, Korb DR, Blackie CA, et al. The lid wiper containsgoblet cells and goblet cell crypts for ocular surfacelubrication during the blink. Cornea. 2012;31:668–679.

68. Yang HY, Fujishima H, Toda I, et al. Lacrimal punctalocclusion for the treatment of superior limbic keratoconjunc-tivitis. Am J Ophthalmol. 1997;124:80–87.

69. Yokoi N, Komuro A, Maruyama K, et al. New surgicaltreatment for superior limbic keratoconjunctivitis and its



association with conjunctivochalasis. Am J Ophthalmol.

2003;135:303–308.

70. Yokoi N, Inatomi T, Kinoshita S. Surgery of the conjunctiva.

Dev Ophthalmol. 2008;41:138–158.

71. Takahashi Y, Ichinose A, Kakizaki H. Topical rebamipide

treatment for superior limbic keratoconjunctivitis in patients

with thyroid eye disease. Am J Ophthalmol. 2014;157:807–

812.

72. Tanioka H, Yokoi N, Komuro A, et al. Investigation of the

corneal filament in filamentary keratitis. Invest Ophthalmol

Vis Sci. 2009;50:3696–3702.

73. Ikegawa W, Yamaguchi M, Shiraishi A, et al. Efficacy of

rebamipide ophthalmic solution for treatment resistant

filamentary keratitis: three case reports [in Japanese].

Atarashii Ganka (J Eye). 2014;31:1369–1373.

74. Meller D, Tseng SC. Conjunctivochalasis: literature reviewand possible pathophysiology. Surv Ophthalmol. 1998;43:225–232.

75. Mimura T, Yamagami S, Usui T, et al. Changes of conjuncti-vochalasis with age in a hospital-based study. Am J Oph-

thalmol. 2009;147:171–177.

76. Yokoi N, Komuro A, Nishii M, et al. Clinical impact ofconjunctivochalasis on the ocular surface. Cornea. 2005;24(suppl 8):S24–S31.

77. Itakura H, Kashima T, Itakura M, et al. Topical rebamipideimproves lid wiper epitheliopathy. Clin Ophthalmol. 2013;7:2137–2141.

Corrected March 4, 2019: In Figure 2, reading left to right,top to bottom, the labels were changed from A, B, E, C, D, F toA, B, C, D, E, F. The order of the six images was not changed.



tear film–oriented diagnosis and tear film–oriented ... · special issue tear film–oriented...

Documents