TFOS DEWS II Report Executive Summary
Jennifer P. Craig
a
,
1
, J. Daniel Nelson
b
,
c
,
1
, Dimitri T. Azar
d
, Carlos Belmonte
e
,
f
,
Anthony J. Bron
g
,
h
, Sunil K. Chauhan
i
, Cintia S. de Paiva
j
, Jos
e A.P. Gomes
k
,
Katherine M. Hammitt
l
, Lyndon Jones
m
, Jason J. Nichols
n
, Kelly K. Nichols
n
,
Gary D. Novack
o
,
p
, Fiona J. Stapleton
q
, Mark D.P. Willcox
q
, James S. Wolffsohn
r
,
David A. Sullivan
i
,
*
a
Department of Ophthalmology, New Zealand National Eye Centre, The University of Auckland, Auckland, New Zealand
b
Department of Ophthalmology, HealthPartners Medical Group and Clinics, St Paul, MN, USA
c
Department of Ophthalmology, University of Minnesota, Minneapolis, USA
d
University of Illinois at Chicago College of Medicine, Chicago, IL, USA
e
Instituto de Neurociencias de Alicante, University Miguel Hernandez-CSIC, Spain
f
Instituto Fernandez-Vega, Oviedo University, Spain
g
Nuffield Department of Clinical Neurosciences, University of Oxford, Oxford, UK
h
Vision and Eye Research Unit, Anglia Ruskin University, Cambridge, UK
i
Schepens Eye Research Institute, Massachusetts Eye and Ear, and Department of Ophthalmology, Harvard Medical School, Boston, MA, USA
j
Department of Ophthalmology, Baylor College of Medicine, Houston, TX, USA
k
Department of Ophthalmology and Visual Sciences, Federal University of Sao Paulo/Paulista School of Medicine, Sao Paulo, Brazil
l
Sj
€
ogren's Syndrome Foundation, Bethesda, MD, USA
m
Centre for Contact Lens Research, School of Optometry and Vision Science, University of Waterloo, Waterloo, Ontario, Canada
n
University of Alabama at Birmingham School of Optometry, Birmingham, AL, USA
o
Pharma Logic Development, San Rafael, CA, USA
p
Departments of Pharmacology and Ophthalmology, University of California, Davis, School of Medicine, USA
q
School of Optometry and Vision Science, University of New South Wales, Sydney, New South Wales, Australia
r
School of Life and Health Sciences, Aston University, Birmingham, UK
article info
Article history:
Received 2 August 2017
Accepted 4 August 2017
abstract
This article presents an Executive Summary of the conclusions and recommendations of the 10-chapter
TFOS DEWS II report. The entire TFOS DEWS II report was published in the July 2017 issue of The Ocular
Surface. A downloadable version of the document and addition al material, including videos of diagnostic
and management techniques, are available on the TFOS website: www.TearFilm.org.
© 2017 Elsevier Inc. All rights reserved.
1. Introduction
Dry eye disease (DED) affects hundreds of millions of people
throughout the world and is one of the most frequent causes of
patient visits to eye care practitioners. It is a symptomatic disease,
characterized by a vicious cycle of tear film instability and hyper-
osmolarity, which leads to increased ocular surface inflammation,
damage and neurosensory abnormalities. Moderate to severe DED
is associated with significant pain, limitations in performing daily
activities, reduced vitality, poor general health, and often
depression.
To increase our understanding of DED, the Tear Film & Ocular
Surface Society (TFOS), a non-profit organization, launched the
TFOS Dry Eye Workshop II (TFOS DEWS II) in March 2015 [1]. This
initiative reflected the TFOS mission, which is to advance the
research, literacy, and educational aspects of the scientifi c field of
the tear film and ocular surface. The goal of the TFOS DEWS II was to
achieve a global consensus concerning multiple aspects of DED.
More specifically, TFOS DEWS II sought to: 1) Update the definition
and classification of DED; 2) Evaluate critically the epidemiology,
pathophysiology, mechanism, and impact of this disorder; 3)
Develop recommendations for the diagnosis, management and
therapy of this disease; and 4) Recommend the design of clinical
* Corresponding author. Schepens Eye Research Institute, 20 Staniford Street,
Boston, MA 02114, USA.
E-mail address: david.sullivan@schepens.harvard.edu (D.A. Sullivan).
1
Co-first author.
Contents lists available at ScienceDirect
The Ocular Surface
journal homepage: www.theocularsurface.com
http://dx.doi.org/10.1016/j.jtos.2017.08.003
1542-0124/© 2017 Elsevier Inc. All rights reserved.
The Ocular Surface 15 (2017) 802e812
trials to assess future interventions for DED treatment.
The TFOS DEWS II involved the efforts of 150 clinical and basic
science research experts from around the world, who utilized an
evidence-based approach and a process of open communication,
dialogue and transparency to increase our understanding of DED.
This process required more than 2 years to complete.
The entire TFOS DEWS II report was published in the July 2017
issue of The Ocular Surface. A downloadable version of the docu-
ment and additional material, including videos of diagnostic
and management techniques, are available on the TFOS website:
www.TearFilm.org. It is anticipated that translations of the report
will be offered in many languages, including, but not limited to,
Chinese, French, German, Italian, Japanese, Korean, Polish, Portu-
guese, Romanian, Spanish, Turkish and Vietnamese. These trans-
lations, when completed, will be available on the TFOS website.
An Executive Summary of the conclusions and recommenda-
tions of the TFOS DEWS II report is presented in this article. The
material is abstracted from the reports of ten TFOS DEWS II Sub-
committees, which were Definition and Classification; Epidemi-
ology; Sex, Gender, and Hormones; Pathophysiology; Tear Film;
Iatrogenic Dry Eye; Pain and Sensation; Diagnostic Methodology;
Management and Therapy; and Clinical Trial Design. Additional
details and all references can be obtained in the open access, online
version.
2. Definition and classification [2]
The goals of the TFOS DEWS II Definition and Classification
Subcommittee were to create an evidence-based definition and a
contemporary classification system for DED. The new definition is
as follows:
“Dry eye is a multifactorial disease of the ocular surface char-
acterized by a loss of homeostasis of the tear film, and accom-
panied by ocular symptoms, in which tear film instability and
hyperosmolarity, ocular surface inflammation and damage, and
neurosensory abnormalities play etiological roles.”
The terminology in this definition, including diction, word order,
emphasis, and accepted meaning, was critical in creating an inter-
nationally accepted definition. The term “multifactorial disease”
recognizes DED as a significant and complex, functional disorder
that cannot be characterized by a single process, sign or symptom.
The term “ocular surface” is defined as comprising the structures of
the ocular surface and adnexa, including the tear fi lm, lacrimal and
meibomian glands, cornea, conjunctiva and eyelids. “Homeostasis”
describes a state of dynamic equilibrium in the body with respect to
its various functions, and to the chemical composition of the fluids
and tissues. Disruption of homeostasis is considered to be the
unifying characteristic that encompasses the myriad of signs of tear
film and ocular surface imbalance that might be observed in DED.
The term “
symptoms” embraces a broad range of possible patient-
reported experiences associated with DED including, but not
limited to, discomfort and visual disturbance. The key elements
contributing to the pathophysiological process, including tear film
instability, hyperosmolarity, inflammation and damage, recognized
as etiological triggers of the vicious circle, were deemed important,
along with neurosensory abnormalities, which have featured
increasingly in the recent literature, for inclusion in the definition.
In the classification of DED, the latest evidence supports a
scheme based on its pathophysiology in which aqueous deficient
dry eye (ADDE) and evaporative dry eye (EDE) exist as a continuum,
such that elements of each need to be considered in diagnosis and
management. This approach is not intended to override clinical
assessment and judgment but to help guide clinical management
and future research.
The Subcommittee's recommended classification of DED is
shown in Fig. 1. The upper portion of the figure represents a clinical
decision algorithm, beginning with the assessment of symptoms,
and followed by review for signs of ocular surface disease. DED
exhibits both symptoms and signs, and can be differentiated from
other ocular surface disease with the use of triaging questions and
ancillary testing. It is to this DED group that diagnostic subtyping,
and conventional DED management strategies, apply. Symptomatic
patients without demonstrable clinical signs do not fall into the
DED group, but are differentiated into pre-clinical ocular surface
disease or neuropathic pain (non-ocular surface disease).
Conversely, asymptomatic patients exhibiting signs are differenti-
ated into patients with poor corneal sensitivity, or those with
prodromal signs, who may be at risk of developing manifest DED
with time or provocation, for example following ophthalmic sur-
gery or contact lens fitting. Finally, the option exists for patients
without either signs or symptoms to be classified, according to the
flow chart, as ‘normal’.
The lower portion of Fig. 1 represents the etiological classifica-
tion of DED, and highlights the two predominant and non-mutually
exclusive categories; ADDE and EDE. Epidemiological and clinical
evidence suggest that the preponderance of DED is evaporative in
nature, which is reflected by devotion of a greater proportion of
Fig. 1 to EDE than to ADDE. While it is possible that ADDE can occur
without obvious signs of EDE and vice versa, as DED progresses, it is
increasingly likely that characteristics of both ADDE and EDE will
become evident. Further subclassification of ADDE and EDE is not
detailed in Fig. 1, but is acknowledged to relate to a vast range of
conditions, as described in the TFOS DEWS II Pathophysiology
report. ADDE describes conditions affecting lacrimal gland func-
tion. EDE is recognized to include both lid-related (for example,
meibomian gland dysfunction [MGD] and blink-related) and ocular
surface-related (such as mucin and contact lens-related) causes.
3. Epidemiology [3]
The TFOS DEWS II Epidemiology report examines literature on
the prevalence, incidence, risk factors, natural history, and
morbidity and reviewed questionnaires used in epidemiological
studies of DED. The report focuses on epidemiological studies
published since the previous TFOS DEWS report in 2007. A meta-
analysis of all published prevalence data was undertaken to esti-
mate the impact of age (Table 1) and sex on symptoms and signs of
DED. Global mapping of DED prevalence was undertaken using
geospatial analysis. The report summarizes the available evidence
on the epidemiology of DED and provides recommendations for
future needs and opportunities.
DED epidemiology continues to be challenged by the failure for
a standardized de
finition and diagnostic criteria to be used.
Consequently, the report describes prevalence based on commonly
used diagnostic criteria, including those based on symptoms, on
self-report of a practitioner diagnosis, and on DED signs.
While much new information has been published in the last
10 years, no population studies have reported on prevalence of
DED for popul ations south of the equator. Much of the atten tion
has focuse d on Asia and Europe. The prevalence of DED, with and
without symptoms, ranged in prevalence from 5 to 50%. DED
preval ence based on signs alone was generally higher and more
variable, reaching up to 7 5% in some populations. Criteria for
positive DED signs vari ed be twee n studies a nd it was acknowl-
edged that some signs may reflect secondary outcomes or may be
related to normal aging. Very few studies were c ond ucted in
youn ger populations (les s than 40 years of age ) but indications
are that DED is also prevalent in these populations. The evidence
J.P. Craig et al. / The Ocular Surface 15 (2017) 802e812 803
for Asian race as a risk factor for DED now appears mostly
consistent.
The meta-analysis confirmed that symptomatic disease and
signs of DED increase wit h age, however prevalen ce of signs
showed a greater increase per decade than symptomatic disease.
Higher rates of DED are reported in women than men, althou gh
the differences generally become significant only w ith increasing
age.
Risk factors were categorised as consistent, probable, and
inc onclusive, in line with the previous TFOS DEWS report [4]. Age,
sex, race, MGD, connective tissue disease, Sj
€
ogren syndrome,
androgen deficiency, computer use, contact lens wear, estrogen
replace ment therapy, hematopoie tic stem cell transplantation,
certain environ mental conditions (such as poll ution, low humid-
ity, and sick building syndrome) and medication use (for example,
antihistamines, antidepressants, anxiol ytics, a nd isotretino in)
were identified as consisten t ris k factors. Probable risk factors
inc luded diabetes, rosa cea, viral infe cti on, thyroid disease, psy-
chi atr ic conditions, pterygium, low fatty acid intake, refractive
surgery, allergic con jun ctivitis, and additio nal medications (e.g.
anti-cholinergic, diuretics,
b
-blockers). Inconclusive DED risks are
Hispanic ethnicity, menopause, acne, sarcoidosis, s moking,
alcohol, pregnancy, demodex infestation, botulinu m toxin injec-
tion, multivit amins and oral contraceptives.
The economic burden on society and impact of DED on the in-
dividual, through its detrimental effect on vision, quality of life, and
work productivity, as well as the psychological and physical impact
of pain, are considerable. The most significant costs are indirect
costs due to reduced work productivity. Questionnaires used to
evaluate DED vary in their utility for epidemiological studies and
further evidence for normative ranges and clinically significant
changes are required.
Future research needs include better evaluation of the preva-
lence of DED of varying severity and in youth, the incidence of
disease in different populations, and the impact of modifiable risk
factors such as mobile device usage. Geographical mapping ap-
proaches will further allow the impact of climate, environment and
socioeconomic factors on DED to be elucidated. There has been
limited study of the natural history of both treated and untreated
DED and this remains an important area for future research.
4. Sex, gender, and hormones [5]
One of the most compelling features of DED is that it occurs
more frequently in women than in men. In fact, the female sex is a
significant risk factor for the development of DED. That such a sex-
related variation exists in the prevalence of an eye disease, or any
other ocular function, should not be a surprise, as sex-related
Fig. 1. DED classification scheme. Please see the original report for a complete description of this figure [2].
J.P. Craig et al. / The Ocular Surface 15 (2017) 802e812804
differences are present in almost every cell, tissue and organ system
of the body. Indeed, since 1945, more than 575,000 scientific re-
ports have been published which address the basic and/or clinical
impact of sex on human physiology and pathophysiology.
The TFOS DEWS II Sex, Gender, and Hormones report details
numerous sex-related differences that have been identified in the
eye. Many of these differences have been attributed to the effects of
sex steroids (e.g. androgens and estrogens), hypothalamic-pituitary
hormones, glucocorticoids, insulin, insulin-like growth factor 1 and
thyroid hormones. For example, androgens are extremely impor-
tant in the regulation of the ocular surface and adnexa. They appear
to mediate many of the sex-related differences in these tissues.
Androgen deficiency, in turn, predisposes to lacrimal gland
dysfunction, serves as a risk factor for MGD, and is associated with
the development of both ADDE and EDE. In contrast to androgens,
the role of estrogens at the ocular surface is less well defined, with
effects that appear to be sex-, tissue-, and dose-specific.
In addition, sex-related differences may arise from the sex
chromosome complement, including differences in parent-of-
origin effects, X chromosome gene dosage (e.g. X-inactivation)
and genes in the non-recombining region of the Y chromosome, as
well as from sex-specific autosomal factors and epigenetics (e.g.
microRNAs, DNA methylation and acetylation, histone
modifications).
It is important to note that the word “sex” is used for a reason.
Although “sex” and “gender” are often used interchangeably, they
have distinct meanings. As stated in a 2001 report by the Institute
of Medicine [6], “sex” refers to the classification of living things,
generally as male or female, according to their reproductive organs
and functions assigned by chromosomal complement. “Gender”
refers to a person's self-representation as a man or woman, or how
social institutions respond to that person based on the individual's
gender presentation. Gender is rooted in biology, but is shaped by
environment and experience. In other words, sex distinguishes
males and females based on their biological characteristics. Gender,
in turn, reflects socially constructed characteristics such as behav-
iors and expectations related to being a man, masculine, or being a
woman, feminine. Furthermore, gender is dynamic, context-related
and operates on a spectrum.
In effect, both sex and gender affect health and disease, as well
as patients' perceptions about their health. Gender also affects in-
dividuals' access to and interactions with the health care system.
Many health disparities are associated with gender. Disparities
arise from a range of influences that are biological, behavioral/
perceptual, cultural, and societal. Therefore, both sex and gender-
dterms that are distinguishable, but intertwined, should be
considered, as they both have pronounced effects on health and on
health disparities. Gender and biological sex affect DED risk, pre-
sentation of the disease, immune responses, pain, care-seeking
behaviors, service utilization, and a myriad of other facets of eye
health.
Overall, sex, gender and hormones play a major role in the
regulation of ocular surface and adnexal tissues, and in the differ-
ence in DED prevalence between women and men.
5. Pathophysiology [7]
On the basis of peer-reviewed literature, the TFOS DEWS II
Pathophysiology Subcommittee concluded that the core mecha-
nism of DED is evaporation-induced tear hyperosmolarity, which is
the hallmark of the disease. It damages the ocular surface both
directly and by initiating inflammation. The cycle of events,
described as the Vicious Circle of DED, is shown at the center of
Fig. 2.
Two forms of DED are recognized, ADDE and EDE. In ADDE, tear
hyperosmolarity results when lacrimal secretion is reduced, in
conditions of normal evaporation from the eye. In EDE, tear
hyperosmolarity is caused by excessive evaporation from the
exposed tear film in the presence of a normally functioning lacrimal
gland. Since tear osmolarity is a function of tear evaporation in
either ADDE or EDE, tear hyperosmolarity arises due to evaporation
from the ocular surface and, in that sense, all forms of DED are
evaporative. In other words, EDE is more accurately considered a
hyper-evaporative state.
In DED, tear hyperosmolarity is considered to be the trigger for a
cascade of signaling events within surface epithelial cells, which
leads to the release of inflammatory mediators and proteases. Such
mediators, together with the tear hyperosmolarity itself, are un-
derstood to cause goblet cell and epithelial cell loss and damage to
the epithelial glycocalyx. Inflammatory mediators from activated T-
cells, recruited to the ocular surface, reinforce damage. The net
result is the characteristic punctate epitheliopathy of DED and a
tear film instability which leads at some point to early tear film
breakup. This breakup exacerbates and amplifies tear hyper-
osmolarity and completes the vicious circle events that lead to
ocular surface damage. Ultimately this is thought to lead to self-
perpetuation of the disease.
Tear film instability can be initiated without the prior occur-
rence of tear hyperosmolarity, by conditions that affect the ocular
surface, including xerophthalmia, ocular allergy, topical preserva-
tive use and contact lens wear. In this case, early tear film breakup is
hypothesized to be the primary basis for tear film hyperosmolarity
initially experienced locally at the site of breakup, and with
increasing severity, at some point becoming detectable in tear
meniscus samples. This represents an ocular surfaceerelated form
of EDE. In MGD-related EDE tear hyperosmolarity results from a
tear film lipid layer deficiency. In ADDE the onset of early breakup
during the evolution of the disease, may add a secondary evapo-
rative element to the DED.
There are various causes of ADDE. It may result from blocking
the sensory drive to the lacrimal gland that is essential to maintain
tear film homeostasis. Bilateral topical anesthesia can cause both a
Table 1
Regression analysis of prevalence data by age for each diagnostic subgroup.
Diagnostic subgroup N of studies Slope estimate (per decade of age) Std error of slope estimate p-value (H
0
: slope ¼ 0) R
2
1. Symptoms or OSDI 23 8 3.43 0.57 0.001 0.858
2. Self-report of a clinician diagnosis of dry eye 8 2.01 0.73 0.034 0.556
3. WHS Criteria 8 0.44 0.57 0.475
a
0.088
b
4. Schirmer 5 10.55 1.78 0.010 0.921
b
5. Tear break up time 5 9.71 1.20 0.004 0.956
b
6. Corneal staining 5 7.63 1.67 0.020 0.875
b
7. MGD 5 5.23 1.44 0.036 0.815
OSDI - Ocular Surface Disease Index;WHS- Women's HealthStudy; MGD - meibomian gland dysfunction. Please see the original report for a complete description of this figure [3].
a
Indicates that there is no change in prevalence by age for the WHS criteria.
b
Regression analyses are based on estimates of prevalence from age 40e49 and beyond (i.e., missing values for prevalence for ages 15e18, 19e29, and 30e39).
J.P. Craig et al. / The Ocular Surface 15 (2017) 802e812 805
reduction in tear secretion and blink rate. DED due to a block in
reflex tearing can be caused by chronic abuse of topical anesthetics,
trigeminal nerve damage and refractive surgery including LASIK
surgery. The delivery of aqueous tears to the tear sac can also be
reduced by obstruction to the lacrimal ducts, which might occur in
any form of cicatricial conjunctival disease, such as trachoma,
ocular cicatricial pemphigoid, erythema multiforme, graft-versus-
host-disease and chemical burns. A number of drugs in systemic
use, such as antihistamines,
b
-blockers, antispasmodics, diuretics
and some psychotropic drugs, can cause a reduction in lacrimal
secretion and are risk factors for DED [3,8]. Also, tear secretion rate
falls in later life.
In the Western world the most common cause of ADDE is in-
flammatory infiltration of the lacrimal gland, encountered most
severely in the DED associated with autoimmune disorders such as
Sj
€
ogren syndrome (SSDE) and, with lesser severity, in non-Sj
€
ogren
syndrome (NSDE). Inflammation causes both acinar and ductal
epithelial cell dysfunction and/or destruction and a potentially
reversible neurosecretory block. Circulating antibodies to the
muscarinic (M3) receptor may also cause a receptor block. Low
tissue androgen levels may predispose to lacrimal gland
inflammation.
Epithelial injury and defective glycocalyx, loss of tear volume
and of goblet cell mucin, lead to increased frictional damage and
friction-related symptoms. The tear hyperosmolarity and epithelial
injury caused by DED stimulates corneal nerve endings, leading to
symptoms of discomfort, increased blink rate and potentially, to a
compensatory, reflex increase in lacrimal tear secretion. This
compensatory secretion is more likely in EDE, where lacrimal gland
function is potentially normal.
A schematic diagram to show the etiology and mechanism of
MGD, which is the major cause of EDE, is shown in Fig. 3. Although
many mechanistic aspects are not yet understood, the figure at-
tempts to summarize the current view. The upper part of the figure
illustrates the etiology of the two forms of MGD that result in low
delivery of meibum, cicatricial and non-cicatricial MGD.
With age, there is an increase in meibomian gland dropout,
particularly after the age of 50 years, which correlates with the
appearance of primary MGD. A fall in bioavailable androgens may
contribute to these events. In youth, treatment of acne vulgaris
with cis-retinoic acid may induce gland atrophy and MGD, while in
an older age group, androgen receptor insensitivity or blockade
may induce signs of MGD. The anti-glaucoma drugs pilocarpine and
timolol also have direct effects on human meibomian gland
epithelial cells that may influence their morphology, survival and/
or proliferative capacity, and possibly promote MGD. Poly-
chlorinated biphenyls may cause a systemic disorder that includes
MGD-like features. Certain skin disorders, such as acne rosacea,
atopic dermatitis, seborrheic dermatitis and psoriasis are associ-
ated with non-cicatricial MGD, while cicatricial conjunctival dis-
eases such as trachoma, erythema multiforme and pemphigoid,
lead to cicatricial MGD.
A key event in non-cicatrici al MGD is hyperkeratinization of
the terminal ducts, leading to duct obstruction, duct dilatation
and disuse atrophy of the gland s. La ter, obliteration of the gland
orifices may occur. Obstruction may be exacerbated by changes in
oil composition that increase m eibum viscosity. The degree to
which inflammatory changes are fou nd around affected glands
varies in different reports, but signs of inflammation are commo n
at the lid margin. Inflammatory mediators and lipids may be
released into tears and onto the ocular surface to cau se epithelial
damage. In cicatricial MGD, submucosal conjunctival scarring
drags the meibomian orifices, terminal ducts and mucocutaneous
junction posterio rly, across the posterior lid border and onto th e
Fig. 2. Pathophysiology of DED. Please see the original report for a complete description of this figure [7].
J.P. Craig et al. / The Ocular Surface 15 (2017) 802e812806
![Fig. 3. Pathophysiology of MGD. Please see the original report for a complete description of this figure [7].](/figures/fig-3-pathophysiology-of-mgd-please-see-the-original-report-3t0i9uvc.webp)
![Fig. 1. DED classification scheme. Please see the original report for a complete description of this figure [2].](/figures/fig-1-ded-classification-scheme-please-see-the-original-2073mvj7.webp)
![Fig. 5. Recommended diagnostic approach for DED. Please see the original report for a complete description of this figure [11].](/figures/fig-5-recommended-diagnostic-approach-for-ded-please-see-the-1o6ngkxa.webp)