`Mylan Pharmaceuticals Inc. et al. v. Allergan, Inc.
`IPR2016-01127, -01128, -01129, -01130, -01131, & -01132
`
`
`
`
`
`Copyright © 1994 by Gilbert Smolin and Richard A. Theft
`
`ThirdEdfition
`
`Previous editions copyright © .1983, 1987 .by Gilbert Smolin
`and Richard A. Thofl;
`
`“
`
`All rights reserved. No part of this book may be reproduced in any
`form or by any electronic or mechanical means, including
`information Storage and retrieval systems, without permission in
`writing from'the publisher, except by a reviewer who may qiiote
`passages in a review.
`
`Library of Congress @ataloging—in-Eubfication Data
`
`The cornea : scientific foundations and clinical practice / edited by
`Gilbert Smolin, Richard A. Thoft.—3rd ed.
`p-
`cm.
`.
`Includes bibliographical references and index.
`ISBN Ell-3.1363627i)418
`A
`1. Cornea—Diseases.
`I. Smolin, Gilbert.
`Richard A.
`
`2. Cornea.
`
`II. Thoft,
`
`--=£DNLM: 1. Diseases.
`RE336.066 1994
`'
`'
`617.7'19—dc20
`DNLM/DLC
`
`for Library of Congress
`
`220 (38133 1994]
`'
`
`93-4460
`
`Printed in the United States of America
`EBM
`
`Editorial: Elizabeth Thompson, Robert J. Stuart
`Production Editor: Anne Holm
`
`Copyeditor: Mary Babcock
`indexer: Alexandra Nickeer
`
`ProductionSupervisorfDesigner: Michael A. Granger
`Cover Designer: Ethan Thomas
`
`
`
`
`
`
`
`
`
`
`
`
`
`Contents
`
`Contributing Authors
`Preface
`
`I. Basic Science
`1. Anatomy of the Conjunctiva,
`Cornea, and Limbus
`gene K. .Gépson
`2' ology
`Henry
`‘Ede’fiiauserfiDayle
`Gemski, and John L. Ubels
`3. Bioohemistry of the Cornea
`mend and
`HaSSeu
`John
`4; Morphology_and Pathologic
`Response of Gama} and
`Conjunctival Disease
`Kenneth R. Kenyon amt
`Henrique V. Chaves
`
`II. Clinical
`5. Bacterial Keratitis and
`
`,
`
`"
`
`'
`
`'
`
`Bacteriology
`Eileen M. Burd
`.
`Clinical Disease
`Gregory S. H. Og'éwa and
`Robert A. Hyndiuk
`6, Viral Keratiti's: and
`Conjunctivitis
`Wmlogy
`Paul R Kinchington
`Clinical Disease
`Herpetic Infections
`Deborah Pavan-Langston
`Adeva and Other
`'Nonherpetie Wrai‘Di-seases
`Jerold S. Gordon
`
`xi:
`xp
`
`3
`
`25
`
`47
`
`69
`
`115 "
`115 '
`.-
`125
`
`169
`169
`
`183
`
`$215
`
`3.
`
`7. Fungal Keratitis and‘
`Conjunctivitis
`Myeelegzy
`Denis M. O’Day and
`E11991“ M 3‘1”
`Clinical Dwease
`Rid-131‘?
`Forge? _
`- Female Keratms anti
`COIfjunéfifi'tiS
`quasltozogy
`.
`Mif’fiael 5:053“)
`Clinical Df‘sease
`Kirk Wilhelmus
`Keratitis and
`ConJWnths
`@‘medwtagy .
`REM S'IStephens
`01mm“; D5533?”
`John P: kadler .
`10. 811311th Infectious
`Diseases
`
`9-
`
`H. Bruce Ostler
`11. Basic Immunology of the
`_ Anterior Segment
`Gilbert Smolin
`12. Clinical Immunologic
`Diseases
`Ocular Allergy
`Peter C. Donshik and
`William H. Ehlers
`Rheumatoid Diseases
`Fuerst,
`David J. Schaninn,
`and Ronald E. Smith
`Non—rheumatoid Acquired
`Collagen Vascular Disease
`C. Stephen Foster
`Mooren’s Ulceration
`J.
`Derka Disease
`Barfly J. Mondino
`
`-
`229
`223
`
`239
`
`253
`
`262
`
`277.
`217
`
`282
`
`295
`
`305
`
`347
`347
`
`364
`
`385
`
`408
`
`414
`
`ix
`
`
`
`
`
`14.
`
`15.
`
`16.
`
`17.
`
`18.
`
`19.
`
`Physiolbgy of'the 'P‘ear
`David W. Lamberts
`Clinical Diseases of the
`Tear Film
`
`David
`
`Sjozgren’s Syndrome
`Khaiid F. Taliban
`
`Corneal Manifestations of
`Neurologic Diseases
`C. Stephen Foster
`Corneal Dystrophies and
`Degenerations
`Gilbert Smolin
`
`Congenital Anomalies of the
`Cornea anti Conjunctiva
`FredM. Wll's'on H-
`Metabolic Diseases
`Mitchell H. Friedlaender
`
`'Iiunors of the Conjunctiva
`am! Cornea
`Jerry A. Shiefls and
`Carol L. Shields
`
`Corneal and Conjnncfival
`Manifestations o'f'Die'tary
`Deficiencies
`Richard A. Thofl
`Corneal Trauma
`
`Physical Agents
`Robert G. Webster, Jr.
`
`439;
`
`457
`45'?
`
`477
`
`485
`
`535
`
`555
`
`519
`
`597
`
`605
`605-
`
`Contents
`
`Ckemieai
`James P. McCulley
`Endothelial Function
`Claes H. Dohlman
`
`22.
`
`Keratoplasty
`Lameliar and Penetrating
`S. Arthur Boruchofl' and
`
`Kemtopmsthesis
`James XL Aquavella
`Refractive Surgery
`Surgical Correction ofMyopia:
`Radial Kemtotomy and
`Excimer'Loser.
`_
`Photoreactive Keratectomy
`
`Roger F.“ Steinem
`Epikemtqnfiqkie- and“
`Keratomile‘usis
`David J. Schanzlin
`
`Correction ofAstigmatism
`Jay H. Krachmer
`Conjunctiva! and Limbo!
`Surgery for Corneal
`DiSeases
`
`26. Therapeutic Contact Lenses
`Mark J. Mannie
`
`Index
`
`51-?
`
`635
`
`645
`£545
`
`665
`
`673
`
`673
`
`696
`
`703
`
`709
`
`723
`
`739
`
` I3.
`
`21.
`
`
`
`
`
`
`
`
`
`
`
`Clinical Diseases of the
`Tear Film
`
`
`
`Keratoconjunctivitis Sicca
`
`David W. Lamberts
`
`This section considers four topics. The first is a
`rational method for dividing the dry eye syn-
`dromes into five categories. The second is the
`techniques available for diagnosing the dry eye,
`or Sicca, syndromes. The third is the currently or
`potentially available methods useful in treating
`these diseases. The complications of the dry eye
`syndromes are disoussed last. A dry eye or kera-
`toconjunctivitis sicca syndrome is said to exist
`when the quantity or quality of the precorneal
`tear film is insufficient to ensure the well-being
`of the ocular epithelial surface [39].
`
`Categories of Dry Eye Syndromes
`
`Dividing the dry eye syndromes into five varieties
`was originally suggested by Holly and Lemp [26].
`These divisions are based on specific deficiencies
`that may have been chosen somewhat arbitrarily;
`nevertheless, they have excellent clinical appli—
`cability The divisions are as follows: aqueous tear
`deficiency, mucin deficiency, lipid abnormalities,
`lid surfacing abnormalities, and epitheliopathies.
`
`Aqueous Tear Deficiency
`
`The aqueous layer forms the greatest bulk of the
`precorneal tear film. Aqueous tears are produced
`in the main lacri-mal glands, with a lesser contri-
`bution from the accessory glands of Wolfring and
`Krause. Aqueous deficiency is by far the most
`common of the dry eye syndromes. Various causes
`share responsibility for the aqueous deficiency
`syndromes (Table 14-1). In spite of their rather
`diverse origins, the clinical presentation of these
`diseases is similar. One entity, Sjogren’s syn—
`drome, is such an important disease it is dis-
`cussed separately in the second section of this
`chapter.
`
`457
`
`
`
`
`
`h'fihmlia‘uuwm‘a.-.“—m.
`
`
`."',-:A.aM‘thid-‘rv’!v
`:lb-‘l-.=
`355.];Mm
`
`
`2-15‘5:
`
`an
`
`
`
`458
`
`Part II Clinical Aspects
`
`Table 14-1. Causes of aqueous tear deficiency
`
`Table 14-2. Mucin deficiency diseases
`
`Congenital
`Riley-Day syndrome (familial dysautonomia) [57]
`Anhidrotic ectodermal dysplasia [59]
`Alacrima secondary to congenital absence of the
`lacrimal gland [7]
`Cri du chat syndrome [27]
`Absence of lacrimal nucleus [76]
`Congenital familial sensory neuropathy with
`anhidrosis [51]
`Adie’s syndrome [11]
`Multiple endocrine neopla‘sia [2]
`
`I
`Acquired
`Trauma to lacrirnal gland
`Surgical removal [62]
`Injury
`Radiation damage
`Inflammation of lacrimal gland
`Sjligren’s syndrome (see following section)
`Primary amyloidosis [65]
`Mumps
`Trachoma
`lnfiltrations
`Sarcoidosis [64]
`Lymphoma, leukemia [41]
`Amyloidosis
`Drugs
`Antihistamines [33]
`Thiabendazole [12]
`Antimuscarinics
`General anesthetics
`
`Neuroparalytic hyposecretiOn
`Brainstem lesions [6]
`Cerebellopontile angle and patrons bone lesions
`[52]
`Middle fossa floor lesions [76]
`Lesions of the sphenopalatine ganglion [76]
`
`Mucin Deficiency
`
`The mucin deficiency diseases are a more com-
`pact group of abnormalities than the aqueous de-
`ficiencies (Table 14-2). Mucin is produced primar-
`ily by the goblet cells of the conjunctiva. There is
`some evidence that at least a small amount of mu-
`
`cin is also produced by the main lacrimal gland
`[29]. With this in mind, it can be‘ said that those
`illnesses that damage the conjunctiva will result
`in mucin deficiency Whether or not the abnor-
`malities seen in these conditions are due to, or
`only associated with, the mucin deficiency is not
`yet clear, however.
`Kinoshita and colleagues [30] studied tear mu-
`cin levels in normal subjects and in patients with
`cicatricial pemphigoid. The results indicated that
`although there was a statistically significant de-
`
`Goblet cell dysfunction: vitamin A deficiency
`Goblet cell destruction
`Alkali burns
`
`Cicatricial pemphigoid
`SteVens-Johnson syndrome
`Trachoma
`
`Drug-induced
`Practolol [79]
`Echothiophate iodide [48]
`
`cream: in mucin levels in cicatricial pemphigoid, a
`substantial amount of mucin-like glycoprotein
`was still present in tears from these patients. In
`the pemphigoid patients, the goblet cell count was
`profoundly decreased, well out of proportion to
`the modest fall in mucin-like glycoprotein. The
`authors suggested that while moderate mucin de-
`ficiency may be associated with surface abnor-
`malities, such mucin deficiency may not be the
`only cause of the ocular surface problems. One
`might also conclude that the tear film is fragile
`enough to be susceptible to even slight falls in
`mucin level. Alternativer it is possible that mu-
`cin from different sources (lacrimal gland versus
`goblet cells) is qualitatively or functiOnally differ-
`ent and that a decrease in goblet cell mucin, while
`quantitatively small, may be important function-
`ally.
`
`Lipid Abnormalities
`
`Only in severe anhidrotic ectodermal dysplasia is
`there a true lipid deficiency. This is extremely
`rare and occurs when the meibomian glands are
`congenitally absent. A much more important as-
`pect of lipid abnormality involves changes found
`in the composition of meibomian secretions in
`various types of blepharitis. The bacteria that in-
`vade the meibomian glands secrete lipases that
`hydrolyze the normal lipids to produce various
`types of free fatty acids. These fatty acids are ex-
`tremely surface-active and are capable of ruptur-
`ing, on contact, an otherwise stable tear film.
`Whether these free fatty acids are directly toxic to
`the corneal epithelium or whether they only dam-
`age it via formation of dry spots is not known.
`Both mechanisms may play a role in the forma-
`tion of the superficial punctate staining com-
`monly seen in blepharitis.
`Gilbard and coworkers [19] closed the meibo-
`mian gland orifices in the right eyes of 11 rabbits
`by light cautery, and studied the changes for 20
`weeks. They found that tear film osmolarity was
`
`
`
`
`
`
`
`Chapter 14 Clinical Diseases of the Tear Film
`
`459
`
`increased throughout the observation period, con-
`junctival goblet cell density decreased, and cor-
`neal epithelial glycogen levels declined. They pos-
`tulated that the increase in tear film osmolarity
`was due to increased evaporation.
`
`Lid Surfacing Abnormalities
`
`.
`
`One of the most important functions of the lid is
`to resurface the eye with tears through constant
`blinking. In a blink, the eyelid sweeps the eye’s
`surface, brushing away exogenous contaminants
`as well as lipid-contaminated mucin strands.
`Fresh mucin is thereby distributed over the cor-
`neal and conjunctival surface. Any break in the
`integrity of the lid or its close apposition to the
`Ocular surface can produce areas of dryness. En-
`tities in this category are listed in Table 14-3. One
`of these items, dellen formation (from the German
`delle, meaning dent or depression), is unique and
`deserves separate comment. Dellen are areas of
`locally thinned cornea adjacent to limbal or con-
`junctiva] elevations. Dellen have been associated
`with muscle surgery, filtering blebs, and limbal
`and conjunctival tumors. The pathophysiology of
`dellen formation is shown in Figure 14-1. As the
`upper lid slides down over the limbal mass, it is
`unable to touch the adjacent “corneal valley.”
`There is a small area that is not resurfaced by the
`lid and does not receive a rejuvenated layer of mu-
`cin with each blink. This area is not able to sup-
`port a continuous tear film without the benefit of
`an adequate hydrophilic mucous layer. If allowed
`to persist, this area of dry cornea will dehydrate
`to the extent that a shallow crater will appear.
`These craters are dellen (Fig. 14-2). Fluorescein
`dye will pool in dellen, but the epithelium usually
`remains intact and will not stain. Most dellen will
`
`disappear if the eye is taped shut for 24 to 48
`hours. A similar pathophysiology explainsgthe 3
`
`Table 14-3. Lid surfacing abnormalities
`
`Lid problems
`Exposure keratitis
`. Entropion
`Ectropion
`Symblepharon _
`Large lid notches
`Lagophthalmos
`Keratinized lid margin
`Surface irregularities
`Dellen formation with limbal lesions
`3 and 9 o’clock stain in hard lens wearers
`
`Topical. anesthetic epitheliopathy
`
`
`
`Fig. 14-1. In the valley between any limbal mass and the
`angle of contact of the tear film with the cornea is a zone
`where the lid cannot wet the epithelium. It is here that a
`dellen forms. (Courtesy F. J. Holly, M.D., M.A. Lemp,
`\M.D., and C. H. Dohlman, MD.)
`
`
`
`Fig. 14-2. A dellen inside the limbus, adjacent to a swoln
`len and elevated coniunctival mass. (Courtesy David Don-
`aldson, MD.)
`
`an irregular or elevated area will predispose the
`
`and 9 o’clock staining syndrome commonly seen
`in hard contact lens wearers. In this case, the con-
`tact lens is the mass that interferes with the lid-
`
`globe integrity and thus prevents adequate wet-
`ting of the adjacent epithelium. This entity does
`not progress to the point of .stromal thinning,
`probably because of contact lens motion, but can
`be associated with superficial corneal vasculari-
`zation.
`
`Epitheliupathies
`
`An intact tear film, because it is very thin, is
`dependent on a smooth, uninterrupted epithelial
`surface. Any irregularity in the surface will cause
`an associated irregularity in the tear film. Such
`
`
`
`
`
`460
`
`Part II Clinical Aspects
`
`tear film to break up instantly at that spot. Since
`the actual cause of the corneal irregularity (e.g.,
`ulcer, erosion, scar) will usually be of primary
`concern, clinically this division of dry eye syn-
`dromes is the least important. It is included for
`the sake of completeness and to emphasize the
`point that the tear film is a fragile structure and
`subject to variances of the surface on which it
`rests.
`
`important reason for exploring skin problems
`during history taking is to discover entities that
`may mimic tear deficiency. A long list of skin dis-
`orders is associated with superficial punctate ker-
`atopathy and may present a picture that could be
`confused with sicca. Some of these-are seborrheic
`
`ichthyosis, and keratosis
`dermatitis, psoriasis,
`follicularis (Darier’s disease).
`
`“What medications do you take?” should also be
`asked. Very few adequately controlled double-
`blind studies have been done to determine the ef-
`
`
`
`Clinical Features
`
`History
`
`The patient’s history is occasionally forgotten in
`ophthalmology. Nowhere is it more important
`than in diagnosing dry eye syndromes. Several
`questions are so basic that they should be part of
`the routine examination. One such question is
`“What sorts of things bother your eyes?” Dry eye
`patients are exquisitely sensitive to drafts and
`winds. Often they willvolunteer information re-
`garding their intolerance to air conditioning or
`driving in the car with the windows rolled down.
`Reading is often difficult for dry eye sufferers.
`This probably occurs because the blink frequency
`decreases during tasks requiring concentration.
`As the blink frequency goes down, the length of
`time the eye is left exposed to the atmosphere
`may become longer than the tear breakup time
`(BUT) (discussed later), and so drying may in-
`crease. Patients often complain that nighttime or
`awakening is the worst part of their day. Sleep
`(like general anesthesia) decreases tear produc—
`tion. If the eye is already compromised with re-
`gard to tear flow, further reduction during sleep
`may be enough to produce nocturnal Symptoms. 1'
`This is especially true if concurrent blepharitis or
`lagophthalmos is present. Smoke is almost uni-
`versally intolerable to tear-deficient patients.
`Since smoke is actually a suspension of solid in
`air, the particulate bombardment of the ocular
`surface produces discomfort.
`“Do you have any skin diseases?” is a question
`that should always be asked. Although primary
`dermatologic illness is rarely the cause of a dry eye
`problem, looking for such illnesses often provides
`useful clues. Examples are numerous: sclero-
`derma, scurvy, thrombotic thrombocytopenic pur-
`pura, the facial rash in lupus (all these are seen
`in association with Sjogren’s syndrome), skin le-
`sions in pemphigoid, old scars from Stevens—John-
`son syndrome, and acne rosacea (associated with
`lipid abnormalities), to mention a few. An equally
`
`fects of systemic medications on tear flow. One
`study of birth control pills produced negative re-
`
`sults [14]. Chlorpheniramine was shown by Kof-
`fier and-'Lemp [33] to have a deleterious effect on
`tear production. Isotretinoin [15], hydrochloro—
`
`thiazide, and propranolol hydrochloride [32] have
`now been associated with decreased tear produfl‘
`tion. Eliciting other types of medication history:
`
`of help in learning about the patient’s gene"
`health.
`
`
`i
`
`"
`
`
`
`To ask about the patient’s other health prob—
`lems is perhaps so obvious as not to need me‘nJ
`tioning. It should be emphasized, however, that
`an entity such as -Sj(‘igren’s syndrome is truly a
`systemic disease that may involve any organ sys-
`tem, and that ophthalmologists see only a small
`part of an illness with protean manifestations.
`Symptoms desoribed or listed by the patient
`during the history taking are not particularly
`helpful in making the diagnosis of sicca. This is
`obviously due to the fact that there are no “dry
`7‘
`receptors” in the cornea. Instead, patients will
`usually complain of irritation, redness, or a vague '
`discomfort in the eye. Sometimes people will in— ,*
`' deed express the specific complaint of dryness, 1:,
`but this is not necessarily indicative of their un-
`derlying problem; it is merely another way to de-
`scribe discornfort. Often a patient with truly u .
`eyes will complain of tearing or excessive u
`tion in the eyes. These seemingly paradoxi
`complaints deserve a word of explanation.
`'
`i
`it is possible for the eye to get so dry that it
`comes irritated and tears reflexively, much as :
`eye might do if Suffering from a foreign body _
`abrasion. Second, as described in Chapter 13, _7
`eyes are particularly susceptible to excess mu
`precipitation. This can occasionally be such _
`annoying part of the symptoms that it will be .
`patient’s primary complaint (Plate 41).
`'
`
`i
`
`Physical Examination
`
`The objective physical examination is much u
`rewarding than the patient’s complaints in di
`
`
`
`
`
`Chapter 14 Clinical Diseases of the Tear Film
`
`461
`
`mg dry eye diseases. The following physical find,
`ings are discussed: filaments, meniscus height,
`meniscus floaters, mucous strands, and papillary
`conjunctivitis.
`Many factors have been implicated in the for-
`mation of filaments (Table 14—4), but by far the
`most common association is with dryness. Fila-
`ments (Plate 42) are short (usually < 2 mm long)
`“tails” that hang from the surface of the c0rnea.
`On sectioning a filament, one finds a periodic
`acid—Schiff (PASl—positive central core (mucin)
`surrounded by epithelium. Although the exact
`pathogenesis of filament formation is not known,
`the following represents a reasonable theory (Fig.
`14-3): When the cornea dries to a point that is in—
`compatible with a healthy epithelial layer, some
`surface cells will become desiccated and be shed.
`
`This creates a small pit on the corneal surface
`that is hydrophobic compared with the mucus-
`coated normal surface. Lipid-contaminated mu-
`cus will become attached to these pits by hydro-
`phobic bonding. Within a sh0rt time, surface
`epithelium will grow down these mucous cores,
`and a true filament will thus be born in_situ. Be—
`cause filaments are anchored to epithelial cells,
`pulling on them is very painful. Unfortunately,
`this is exactly what happens during blinking,
`with the resultant symptoms not unlike those
`produced by a foreign body.
`The normal human meniscus height is usually
`reported to be about 1 mm. In actual fact, normal
`subjects never have a meniscus height of that
`magnitude. Rather, a height of 0.2 to 0.3 mm is
`
`Table 14-4. Causes of corneal filaments
`
`Local
`
`Keratoconjunctivitis sicca
`Superior limbic keratoconjunctivitis
`Aerosol keratitis
`Beta radiation
`Herpes simplex viral infection
`Recurrent erosions
`
`Thygeson’s superficial punctate keratitis
`Cataract surgery
`-Neurotrophic and neurOparalylic keratitis
`Prolonged occlusion of the eyelids
`Retained foreign body beneath the upper lid
`Systemic
`Diabetes mellitus
`Psoriasis
`
`Ectodermal dysplasia
`Atopic dermatitis
`Osler—Weber—Rendujdisease
`
`'
`
`.
`
`'
`
`most common and will be found in about 85 per—
`cent of normal subjects [35].
`In a large population, no correlation was found
`between tear meniscus height and subsequent
`Schirmer test results [35]. In other words, people
`with a relatively high meniscus (0.4—0.6 mm) did
`not necessarily have a higher Schirmer test value.
`In a smaller study of nine patients with moderate
`to Severe dry eye (all had keratoconjunctivitis
`sicca or Sjogren’s syndrome confirmed by rose
`bengal staining and diminished Schirmer test
`values), I found that seven had a meniscus height
`of 0.1 mm and two had a meniscus height of 0.2
`mm (unpublished data, 1981). It seems then that
`in normal eyes the tear meniscus height is prob-
`ably subject to variables other than tear flow. Fac-
`‘tors such as the length of the lid, the location of
`the punctum,
`lid-globe apposition, and the dis-
`tance of the meibomian glands from the edge of
`the lid probably influence meniscus height. If tear
`flow is reduced beyond a certain level, this appar«
`ently is manifested by a reduced meniscus height.
`In this limited study, a meniscus height of only
`0.1 mm was present in 80 percent of the patients
`tested.
`
`Meniscus floaters are extremely common in
`dry eye patients. They can be seen as tiny bits of
`debris being carried along in the upper and lower
`tear menisci. Floaters probably have two origins.
`Some are dead epithelial cells that have fallen off
`the surface of the cornea, and some are small fi-
`brils of lipid-contaminated mucin. Although al-
`most always present in dry eyes, they are not
`pathognomom‘c, because patients with conjunc-
`tival
`infections or blepharitis may also show
`floaters.
`
`Mucous strands are actually strings of lipid-
`contaminated mucus that have rolled up and been
`pushed into the cul-de-sac by the shearing action
`of the lids. Although common in aque0us—defi-
`cient states, they can become rather spectacular
`in the mucin-deficient diseases (see'Plate 41). The
`
`i
`
`proposed dynamics of mucous strand formation
`are shown in Figure 14-4. If the aqueous layer of
`tears is thick enough to prevent diffusing lipid _
`from contaminating the mucous layer before a
`blink, or if mucin is available in excess to absorb
`the lipid molecules before a blink, then mucous
`strands will not become a problem. On the other
`hand, if mucin and excess lipid become intermin—
`gled, mucous strands may form.
`Papillary conjunctivitis is a rather n0nspecific
`reaction of the conjunctiva t0 irritation. It is com—
`monly seen in allergic eye disease, infections, and
`acne rosacea. Its presence in dry eye syndromes
`
`
`
`
`
`462
`
`
`
`
`
`Part ll Clinical Aspects
`
`
`
`Upper lid
`
`
`
`
`Mucous thread containing lipid
`moving downward
`
`
`Rolling up of mucus layer
`contaminated with lipid
`
`
`Dry 5P'3'l'—'b
`
`Lipid contaminated mucus layer
`Superficial lipid layer being
`compressed between the
`approximating lid edges
`
`
`
`
`
`
`' Cross section of mucous
`thread withrhigh lipid
`"" content in the lower
`fomix on the way to
`the puncta
`
`Fig. 14-4. Tentative mechanism for the formation of mu-
`Cous strands. As the tear film thins and breaks, sudaCe
`lipid may become apposed to the underlying mucin,
`forming lipid-mucin strands. (From F. J. Holly and M. A.
`Lemp, Tear physiology and dry eyes. Surv. Ophthalmol.
`22:69, 1977. With permission.)
`
`is mentionéd only to acknowledge its existence; it
`is of no value in making a diagnosis [39]. Staph-
`ylococcal blepharitis, which frequently accompa-
`nies a dry eye condition, can also result in punc-
`tate keratopathy and papillary hypertrophy of the
`conjunctiva.
`'
`
`_‘ Clinical Diagnostic Tests
`I I should openly confess at the start of this discus—
`sion that there is no reliable objective test to ren~
`der a firm diagnosis of dry eyes. Nevertheless,
`three tests are in common usage and are dis-
`cussed here. These are the rose bengal dye test,
`the measurement of tear BUT, and the Schirmer
`test.
`-
`r
`
`Fig. 14-3. A theory of filament formation. As a normal epi—
`thelial surface dries (A), some cells die and fall off, leaving
`a defect (B). Mucin may adhere to this high-energy pit (C),
`and eventual! epithelium may grow down over the mu—
`cin to form a illament (D). (From D. W. Lamberts, Dry Eye
`Syndromes. In C. 5. Foster [ed], Comeal and External
`Disease. Copyrighted by Year Book Medical Publishers,
`Inc, 1985. With permissiOn.)
`
`
`
`Chapter 14 Clinical Diseases of the Tear Film
`
`463
`
`Rose Bengal Stain
`Rose bengal, a red aniline dye, is a derivative of
`fluorescein but is different from fluorescein in
`
`several important ways. Rose bengal does not
`stain the precorneal tear film as does fluorescein.
`Instead, it seems to precipitate at the bottom of
`the meniscus. Unlike fluorescein, rose bengal
`does stain devitalized or abraded epithelial cells;
`this is probably due to rose bengal’s ability to
`stain cell nuclei and cell protoplasm. 0n the other
`hand, fluorescein is able to permeate through a
`disrupted epithelial layer and diffuse among the
`intercellular spaces. Rose bengal also stains mu-
`cus and filaments, whereas fluorescein stains
`only mucus.
`Rose bengal application (usually as a 1% solu-
`tion) is an extremely valuable test in the diagno-
`sis of sicca and deserves to be used more extent.-
`sively. It has only one disadvantage—irritation on
`instillation. The irritation seems to be directly re-
`lated to the amount of epithelial damage present
`on the corneal surface and, to some extent, the
`size of the drop. It is frequently possible to alle-
`viate some of this pain by drOpping the rose ben-
`gal onto the wooden end of a cotton-tip applicator
`and in turn touching the applicator to the inferior
`cul-‘de-sac. This decreases the amount of dye in-
`stilled in the eye. When rose bengal is used for
`evaluating dry eye patients, it is most helpful to
`divide each eye into three zones (medial, corneal,
`lateral) and reoord the amount of stain in each on
`a scale of 0 to 3 [73]. In this way, each eye could
`receive a maximum score of 9 if it stained maxi-
`
`mally in all three zones. A score of 3 or more for
`one eye is considered abnormal. Two false-posi~
`tives can be seen with rose bengal stain. A small
`amount of stain over the body of a pterygium or
`pinguecula is a common and normal finding. Also,
`if a Schirmer test has been performed before the
`use of rose bengal stain, the cenjunctiva will pick
`up the dye in the area of contact between the con-
`junctiva and the paper strip. It is important to re-
`member that in dry eye syndromes the pattern of
`'the stain, not merely its presence, is important.
`Anything that can damage the epithelium (infec-
`tions, trauma, toxic reactions) can cause rose ben-
`_. gal staining, but it is characteristic of sicca to pro-
`duce the pattern shown in Plate 43. This pattern,
`which is pathognomonic of keratoconjunctivitis
`sicca, consists of two triangles of stain on the
`conjunctiva with their bases on the limbus and
`in particularly severe cases, a connecting band
`across the cornea. This area corresponds to the
`exposed surface of the eye Within the palpebral
`fissure.
`‘
`
`_
`
`Norn [47] wrote a short but excellent text on
`examination of the external eye. He discussed the
`use of lissamine green in diagnosing dry eye.
`Norn stated that the dyeing quality of lissamine
`green is the same as that of rose bengal. Both so-
`lutions can be made in 1% concentration. It may
`be advantageous to use lissamine green rather
`than rose bengal because lissamine green causes
`no smarting. |
`
`.\
`
`Thar Breakup Time
`Tear BUT was originally described in 1969 by
`Norn [46], who referred to it as corneal wetting
`time. In this country, the concept of BUT was pop-
`ularized by Lemp and Hamill in 1973 [37]. Its
`measurement depends on the fact that given
`‘, enough time, the tear film will thin and eventu-
`ally rupture (even in normal eyes). The mecha-
`nism- of dry spot formation is discussed in Chap-
`ter 13. Its clinical measurement is accomplished
`as follows: A drop of fluoreScein is instilled in the
`eye. The patient is asked to blink two or three
`times to distribute the dye. The test is performed
`at the slit lamp by asking the patient to stare
`straight ahead and not blink. Without touching
`the patient’s eyelids, the examiner scans the cor-
`nea with the cobalt-blue light of the slit lamp,
`Watching for an area of tear film rupture mani-
`fested by a black island Within the green sea of
`fluorescein. The BUT is the time in seconds be-
`
`tween the last blink and the appearance of the
`dry spot. A normal BUT is considered to be 10 sec-
`onds or more. Several points are important when
`performing the test. First, the examiner must not
`touch the patient’s lids, nor should the patient ar-
`tificially elevate his or her lids. These maneuvers
`tend to expand the interpalpebral fissure and
`thus artificially shorten the BUT. Second, the
`areas of breakup must appear in a random pat-
`tern to be counted as tear film breakup. The rea-
`son for this is to avoid counting anepithelial ir—
`regularity or elevation (which will thin the tear
`film and cause a rapid breakup) as a representa—
`tive, random dry spot. Third,
`the solution in
`which the fluorescein is dissolved may theoreti- -
`cally play a role in BUT measm'ements. Some
`evidence suggests that the preserVatives in oph-
`thalmic irrigating solutions and even in the fluo-
`rescein paper strip itself may artificially lower
`the BUT (F. J. Holly, unpublished data, 1981).
`Some of these preservatives have detergent-like
`action and may break up the superficial lipid
`layer of the tear film, causing it to rupture pre-
`maturely. This may help to explain why BUT has
`fallen into some disrepute in recent years, since
`
`-
`
`
`
`
`
`
`
`464
`
`Part II Clinical Aspects .
`
`the nature of the fluorescein solution is a variable
`
`not usually controlled. It maybe better to use a
`nonpreserved type of fluorescein such as Fluore-
`soft (Holles Labs, Cohasset, MA) or one of the sev-
`eral brands of fluorescein for injection for BUT
`measurements. If one uses the latter, cauti0n
`must be taken to instill a very tiny drop (from the
`tip of a tuberculin syringe, for example), because
`these solutions are concentrated and burn on in-
`
`stillation. Because t0pical anesthetics also con— ‘
`tain preservatives that are surface-active and ca-
`pable of rupturing the tear film and adversely
`affecting the
`of the corneal surface [49],
`they should not be used for the test.
`A new method for measuring BUT was de-
`scribed in Chapter 13. I reiterate it briefly here.
`The test employs an instrument known as a to-
`poscope, which resembles a small visual field ap-
`paratus. By placing the patient’s head in the bowl,
`it is possible to reflect an optical grid pattern off
`the tear film. By observing this grid through a
`telescope, the observer can watch for a breakup to
`occur. A breakup is seen as a distortion of the n‘or-
`mally regular grid. The great advantage with this
`technique is that it is totally noninvasive. No
`fluorescein,
`topical anesthetic, or preServative
`needs to be placed in the tear film. In a study per-
`formed by the inventor of the toposcope (S. R.
`Tonge, personal communication, 1987),
`it was
`found that dry eye patients had shorter BUTs
`than did normal subjects, although all times were
`longer than previously described values using in-
`vasive tests.
`
`Schirmer Test
`
`In spite of its notorious variability, the Schirmer
`test remains the most commonly performed clini-
`cal test for sicca. For such a ubiquitous examina-
`tion, remarkably little basic science has been
`dune to explore the variabilities of the Schirmer
`test.
`It has been demonstrated in vitro that the vol-
`
`ume of water in milliliters applied to a paper strip
`has a linear relation to the length of wetting of
`the strip in millimeters. In vivo, the relationship
`is not so clear—cut. Figure 14-5 A presentsa graph
`of the wetting kinetics of Schirmer strips in nor-
`mal human volunteers. Figure 14-5 B is a com-
`panion graph of the rate of tear secretion plotted
`against time. The striking conclusion obtained
`from studying these curves is that tear production
`is extremely high in the first 2 minutes of the
`Schirmer test but levels off to a steady state at 4
`to 5 minutes. To be more explicit, these flow rates
`
`4r
`
`H D
`
`
` 'e’‘6’3Wettinglengthinmm
`rateinul/min 0
`Secretion
`
`0123456789101112
`Time in minutes
`
`A
`
`20
`
`15
`
`10
`
`0123456789101112
`Time in minutes
`
`B
`
`Fig. 14-5. A. When one examines the kinetics of a
`Schirmer test in vivo by plotting the wetting length versus
`time, one is struck by the rapid outflow of tears in the first
`3 or 4 minutes and by the leveling off to a steady state af-
`ter this time. B. If one plots rate of seCretion of tears (in
`vivo) versus time, one is again impressed by the precipi-
`tous fall in secretion