ophthalmologist with a number of challenging and difficult problems.
Much of what is known of the virus and its relationship to humans, its
natural host, has been accumulated as a result of intensive research
in the fields of virology and immunology. Gaps in our knowledge still
exist and as these gaps narrow, our principles and methods of
treatment will change.
In Western countries, infection with herpesvirus is almost universal.
By early adult life, neutralizing antibodies are present in up to 90%
of the population. Peaks of primary infections occur during infancy
and adolescence, but sporadic cases are seen in the neonatal period
and throughout adult life. In the majority, the primary infection is
subclinical or goes undiagnosed. The disease usually runs a
self-limited course but occasionally has a fatal outcome. With healing
of the primary infection, the body is apparently free of disease;
however, the virus has not been eliminated. Instead, having
established a foothold, it persists permanently in an almost perfect
symbiotic relationship that is marred by recurrent disease when the
virus is reactivated from its apparent latent state.
Primary ocular herpes usually occurs as an acute follicular
conjunctivitis with regional lymphadenitis and usually with vesicular
ulcerative blepharitis. Most patients also have an epithelial
keratitis, which persists somewhat longer than the conjunctivitis.
Only rarely is there significant stromal involvement.
Recurrent episodes are a different problem. In these, the cornea is
the principal target tissue. Males are infected twice as often as
females, and attacks, although occurring all year, tend to be more
frequent in autumn and winter. The most common form is the
morphologically characteristic epithelial keratitis (dendritic,
geographic, or punctate). Initially, there may be no serious sequelae
to infection, but with repeated attacks, stromal keratitis, and
associated uveitis may appear. Alternatively, disciform keratitis or
other more heavily infiltrated stromal keratitis may develop without
apparent preceding epithelial herpetic keratitis. When stromal
keratitis supervenes, permanent structural damage to the cornea and to
the rest of the eye exacts a heavy toll on vision. It is this effect,
coupled with chronicity and resistance to treatment, that makes herpes
simplex one of the most important viruses to affect the eye.
Back to Top
HERPES SIMPLEX VIRUS
Herpes simplex virus (HSV) is a member of the Herpesviridae family.
The virion is 180 nm in diameter. It is composed of four principal
components: the core, the capsid, the tegument, and the envelope (Fig.
Fig. 1 Schematic morphology of herpes simplex virus. (Published
courtesy of Liesegang TJ: Biology and molecular aspects of herpes
simplex and varicella-zoster virus infections. Ophthalmology 99: 781,
The viral core contains double-stranded DNA used for viral
replication. The viral genome has a molecular weight of 100 × 10,1,2
large enough to encode approximately 70 proteins and 72 genes.3 The
viral DNA molecule is arranged as a double helix composed of two
chains of repeating units of deoxyribose and phosphate, with purine
and pyrimidine bases extending sideways from the sugar. The purine
bases are guanine and adenine, whereas the pyrimidine are cytosine and
thymine. The chains are linked together by the pairing of protruding
bases to form the double helix. All the information required for virus
replication is inscribed on the DNA molecule in a code constructed
according to the order of repetition of the four bases. The physical
form of the DNA of the virion in the replicating and latent virus is
different. In the virion form of the virus, the DNA is linear but
after infecting a cell it becomes circular. The viral DNA contains
several classes of genes that encode regulatory proteins, structural
proteins, and enzymes that are spread in a highly regulated fashion
during the replicative cycle.4 Latency-associated transcripts (LATs)
represent limited transcription of viral DNA during latency and serve
as a marker for latently infected cells.3
The viral DNA is enclosed within a protein shell with an icosahedral
shape called a capsid. The capsid is composed of 162 five- or
six-sided subunits called capsomeres.3 In addition to protecting the
viral genome, the capsid allows entry of the viral DNA into the host
A region of amorphous protein called the tegument lies between the
capsid and the outer envelope. Tegument proteins, in association with
cellular factors, play a role in inducing transcription of viral
proteins5 and modulate host protein production.3
An essential component of the infective particle is the outer
envelope. The envelope is composed of lipoproteins, carbohydrates, and
lipids that have been derived from the host cell and have been
modified by the viral protein.3,4 Embedded within the envelope and
projecting from the external surface are glycoprotein subunits called
peptomers, which play a role in viral attachment and penetration into
the host cell.2,3
The virus shows a tropism for human tissues of ectodermal origin. In
addition to the ocular infection and the well-known skin and mucous
membrane lesions, herpes is responsible for a meningoencephalitis and
has been associated with trigeminal neuralgia. In disseminated
infections, it may replicate in liver, adrenal, and lung parenchyma.
HSV is an obligate intracellular parasite. Although it contains the
genetic material necessary to induce its own replication, it does not
have the metabolic machinery necessary for biomolecular synthesis. It
enters the host cell and uses the host cell metabolic pathways, often
causing destruction of the host cell. The replicative cycle has three
phases: entry, eclipse, and envelopment and release.
The host cell and the virus come in contact and bind by means of
specific cell surface glycosaminoglycans, principally heparan
sulfate.2,6 Lytic genes are expressed, and the virus enters the cell
fusion of the virion envelope with the cell’s plasma membrane. It then
penetrates the host cell cytoplasm in a pinocytotic vesicle, where the
envelope is removed. Enzymatic digestion results in uncoating of the
capsid. The bare viral capsid moves to the host cell nuclear pore,
where it is disassembled and the viral DNA released.
The eclipse phase of the replicative cycle is characterized by intense
molecular activity within the host cell nucleus and loss of
recognizable viral morphology. The viral DNA unwinds and is
transcribed by messenger RNA, which in turn directs viral protein
synthesis within the ribosome. Enzymes, structural proteins, and
regulatory proteins are produced. Most of the proteins are returned to
the nucleus. DNA replication and capsid reassembly occur there.
Envelopment and Release
Within 6 hours of entry, fully enveloped particles are detectable in
the cells. The new viral particle is enclosed by the envelope, which
is primarily derived from a nuclear membrane as it leaves the nucleus
and enters the cytoplasm. The mature infected virus negotiates the
cell membrane by a process of reverse phagocytosis to reach the extra
cellular space. The cycle is now complete.7
On the basis of site of isolation from the body and cell culture
characteristics, two types of herpes simplex virus can be
distinguished. HSV-1 characteristically produces oral, facial, and
ocular lesions. It is responsible for 85% of ocular isolates. HSV-2
serotype has conventionally been associated with the sexually
transmitted form of the disease. It is responsible for ocular disease
in neonatal herpes simplex keratitis but in only minority of adult
ocular infections. Simultaneous keratitis infection with both HSV-1
and HSV-2 has been described in a patient with acquired immune
deficiency syndrome (AIDS).8
Back to Top
HERPES SIMPLEX VIRUS INFECTION
Humans are the only natural host of HSV, although experimental
infection can be produced in a variety of animals, including rabbits,
mice, and primates. Persons infected with the virus constitute the
sole reservoir of infection. Humans are extremely susceptible to
infections with Herpesvirus. Given this highly susceptible host, poor
sanitary conditions and overcrowding greatly predispose to infection.
Studies of the presence of neutralizing and complement-fixing antibody
within different socioeconomic groups as an index of infection have
underlined this relationship. In the lower socioeconomic groups in the
United States, 80% of individuals have antibodies. This contrasts with
50% of those more economically advantaged.9 As standards of living
increase, an increase in the number of susceptible adults can be
anticipated, with a subsequent increase in the incidence of adult
primary herpes simplex.
Transmission of HSV-1, which is responsible for the vast majority of
facial and ocular herpetic infections, occurs either from direct
contact or via contaminated secretions. It has been shown that virus
particles are shed intermittently or chronically in tears,11 saliva,12
and respiratory secretions,13 as well as from the genital tract in the
absence of overt disease.
Mechanisms of spread from the portal of entry are not known precisely,
but a viremia appears likely and has been demonstrated in severe
conditions on a number of occasions. The incubation period of HSV-1 is
3 to 9 days.14 The overwhelming majority of primary infections occur
in infancy and adolescence, but sporadic cases of primary infection
occur among susceptible individuals throughout adult life. The primary
infection is subclinical in 85% to 90% of cases.15 A positive titer of
serum-neutralizing antibodies is noted by 1 week after the primary
infection. The titer then diminishes but remains positive throughout
life. Complement-fixing antibody follows a similar pattern but is more
variable during asymptomatic periods. Occasionally, primary HSV-1 is
ocular and causes lid vesicles, ulcerative blepharitis, keratitis, and
Infected persons become carriers of the disease by transneuronal
spread of the virus into the neural ganglia. The virus persists there
in a quiescent state called latency that may be interrupted by periods
of localized recurrence. A variety of endogenous and exogenous
stimuli, such as strong sunlight, fever, menstruation, and psychiatric
disturbances can serve as triggers of reactivation. In addition,
reactivation in the cornea can be precipitated by local factors, one
of which has been shown to be exposure to excimer laser irradiation
during refractive surgery.
Ocular herpetic involvement is less common than systemic infection.
Projections from a prevalence study in the northern United States to
the whole country suggest a total of approximately 20,000 new cases
per year and 400,000 with a diagnosis of ocular herpes simplex.18
Herpes simplex infection is the most common cause of corneal blindness
in developed countries.
An initial episode of HSV-1 epithelial keratitis has a 25% recurrence
risk within 2 years. A second episode has a 43% recurrence risk.13
Recurrence occurs in the eye originally infected in the majority of
cases, although involvement is bilateral in approximately 11% of
PATHOGENESIS AND LATENCY
The pathogenesis of human herpes simplex ocular infection exhibits two
critical features: complexity and diversity. This is demonstrated in
the marked differences in pathogenesis between primary and recurrent
ocular disease. It is also apparent in the many different forms of
ocular disease that result from the complex interplay of viral
replication, host defense systems, and tissue reparative responses.
Further confusion arises for the ophthalmologist in the management of
herpetic eye disease: appropriate therapy for one clinical form of
disease may be absolutely contraindicated in another clinical form.
The natural history of HSV-1 infections in humans is generally
characterized by an initiating childhood infection, which may present
as a nonspecific upper respiratory infection or may be totally
asymptomatic. Goodpasture20 in 1929 and subsequently others21,22
postulated that the virus gained access to the central nervous system
during the primary infection by moving centripetally along sensory
nerves to the sensory ganglia. In 1973, Cook and Stevens24 confirmed
the concept of retrograde axonal transport and latency in ganglia.
The process leading to latency is now understood to occur in three
stages: Entry, Spread and Establishment of latency. Entry defines the
time of the primary infection. Spread is the phase during which the
virus moves to the terminal axons of the sensory neurons and then, by
retrograde axonal transport to the neuronal cell bodies in sensory and
autonomic ganglia where there may be further viral replication. In the
final stage, Establishment of latency, lytic gene expression is
suppressed and virions cannot be detected. However, the viral genome
persists in the neuron.
Under the influence of various stimuli, control of latency breaks down
and viral replication begins again in the ganglia with spread to
peripheral sites where replication may also occur.25
The mechanisms by which the virus maintains latency and is ultimately
altered to cause recurrent disease are only partially understood.
While in the latent state, viral gene expression is suppressed almost
totally. Viral structural component and infectious viral particles are
not produced, although virus can be detected by cell culture
explantation techniques. However, viral RNA molecules called
LATs3,5,26 are transcribed. The LATs serve as useful markers for
latent HSV infection. LATs may play a role in reactivation.
There is also evidence to suggest that latent infection can occur at
ocular sites such as endothelial cells or keratocytes.27–30 This has
raised the question of whether extraganglionic latency occurs within
the cornea. HSV-1 DNA sequences have been identified in some human
corneas that do not have any history of herpetic eye disease5,27;
however, the existence of corneal latency has not been firmly
established. The possibility of corneal latency has important clinical
ramifications for it would allow for viral reactivation and
replication within the cornea without ganglionic HSV reactivation.
Recurrent clinical disease apparently occurs when local host defenses
in the eye are unable to control the virus, or there is a break in the
epithelial barrier function. It is clear that recurrent clinical
disease occurs despite systemic humoral and cell-mediated immunity
against the virus.
Strain variations in HSV appear to affect reactivation. Certain
strains are associated with high recurrence rates.31 Genetic
differences among strains appear to affect the clinical manifestations
of infection, including the morphology of an epithelial dendrite.
Certain strains are more likely to produce stromal disease and this
has been correlated with the amount of glycoprotein produced during
infection.32,33 The impact of corticosteroids on the course of the
epithelial infection also appears to be strain related.34
Corticosteroids may lead to an increased duration of herpetic disease
by interrupting the immune system, but it is unlikely that they induce
The host response to the virus plays a role in the disease process.
However, the importance of individual host differences in determining
the course of infection is unclear. For unknown reasons, herpetic
infection does appear to be more common in patients with atopic
Treatment of herpes simplex keratitis should be tailored according to
the clinical form of herpetic disease that is present. Purely
epithelial herpes simplex keratitis is typically managed with topical
antiviral agents with or without debridement. The management of
stromal and disciform endotheliitis is more complex and usually
involves both antiviral and anti-inflammatory measures. Surgery may be
necessary in more severe forms of this disease. Specific therapy for
each of these entities is discussed in detail in the following
In general, a rational approach to therapy for ocular herpes simplex
disease should include:
Minimize permanent ocular damage from each recurrent episode.
Avoid iatrogenic disease.
Counter the socioeconomic effects of a chronic debilitating disorder.
Such an approach is possible only if the ophthalmologist maintains a
clear perspective of the chronic, recurring, progressive course of
this disease. The nature of herpetic keratitis is such that these aims
are often in conflict. The topical antiviral agents used are
inherently toxic and the vigilance needed to manage the patient
carefully for protracted periods of time is demanding, as well as
potentially socially and economically crippling. Only currently
established methods of treatment are included here. Controversial
therapies and drugs, given the experimental stage of development, are
At one point, mechanical debridement was the only effective means of
treating epithelial herpes. Even with the advent of antiviral agents,
it remains a useful, safe, and sometimes preferred alternative.35 The
removal of virus-replicating epithelium abolishes the source of
infection for other cells and eliminates an antigenic stimulus to
inflammation in the adjacent stroma.
Debridement should be performed at the slit lamp or operating
microscope, with the use of topical anesthesia. Controlled removal of
the lesion is best achieved by gentle debridement along the margins of
the epithelial ulcer with a tightly rolled cotton-tipped applicator.
With this technique, known as minimal wiping debridement, the
virus-infected cells are removed while healthy epithelium is left
intact.36 Sharp knife blades should not be used because of the risk of
creating a portal of entry into the stroma through damage to
underlying Bowman’s layer. Recrudescence of viral replication
occasionally occurs and can be treated by repeat debridement or
administration of an antiviral agent. Chemical virucidal agents, such
as phenol 10%, have been advocated to sterilize the freshly debrided
ulcer margins but are unnecessary. Scrubbing the bare surface is
injurious, and iodine is damaging, especially to diseased corneal
Idoxuridine (IDU), the first antiviral drug to become available for
topical ophthalmic use, is a substituted pyrimidine nucleoside that
resembles thymidine (Fig. 2).It is phosphorylated to the nucleotide
and incorporated into the DNA of all cells, where it interferes with
Fig. 2 Chemical structures of A thymidine, B Idoxuridine, C
vidarabine, and D acyclovir.
Disadvantages include poor corneal penetration,37 lack of selectivity
for virus-infected cells, and toxicity. In the majority of patients,
the earliest signs of toxicity are recognizable after 2 weeks of
therapy. These include punctate keratoplasty, burning, injection,
irritation, lacrimation, hypersensitivity, and punctal stenosis (Table
TABLE 1. IDU Toxicity*
Fine punctate keratopathy
Cornea Corneal filaments
Retardation of epithelial healing
Superficial vascularization (late)
Superficial stromal opacification
Punctate staining with rose bengal
Follicles in lower tarsus
Lid margin Punctal edema → occlusion (may be irreversible)
Edema of orifices of meibomian glands
*Other currently available antiviral agents exhibit similar
toxicities, although trifluridine is less toxic than IDU or
vidarabine; contact allergy to each of these drugs is also possible.
Vidarabine, first synthesized in the early 1960s, is a purine
nucleoside analogue with in vitro activity against Herpesvirus and
certain other DNA viruses. Cellular enzymes convert vidarabine to the
triphosphate form, which acts as a competitive inhibitor of DNA
polymerase. Corneal penetration is poor but better than that of IDU.
Because vidarabine does not selectively inhibit virally induced
enzymes, there is potential for cellular toxicity. It is probably less
toxic than IDU. Vidarabine is available as a 3% ophthalmic ointment,
and the usual dose is five times per day. Collaborative studies have
indicated that vidarabine is effective in the therapy of epithelial
herpetic disease.38 As with IDU, resistant strains exist, but
cross-resistance has not been observed. Hypertrophic epithelial
changes similar to those seen with IDU occur, and some
hypersensitivity reactions have been reported.
Trifluridine, a thymidine analogue that inhibits thymidylate
synthetase, is incorporated in both viral and cellular DNA. It is semi
selective, interfering with viral metabolism in preference to normal
cellular metabolism, and is thus less toxic than IDU. Trifluridine is
10 times more soluble in water than IDU and is available as a 1% drop.
Studies have shown the healing time for active epithelial ulcers to be
better than that with IDU and comparable to that with vidarabine.12
When used in higher doses for prolonged periods of time, toxicity does
develop, producing changes similar to those seen with IDU, although
not as severe.
Acyclovir is an acyclic analogue of guanosine and is the prototype of
the generation of specific antiviral drugs that are activated by a
viral thymidine kinase to become potent inhibitors of viral DNA
polymerase. The selectivity of acyclovir for virus-infected cells is
approximately 200 times that for normal cells. Its antiviral spectrum
is limited to the herpes group and excludes vaccinia, adenovirus, and
RNA viruses.39 Acyclovir is available in the United States in oral and
intravenous forms, and as a topical dermatologic ointment. Topical 3%
acyclovir ointment for ophthalmic use (not commercially available in
United States) can penetrate the cornea to reach the anterior
chamber.40,41 It has been shown to be effective in the treatment of
HSV epithelial keratitis.40,41
Oral acyclovir in a dosage of 400 mg five times per day results in
therapeutic levels in the aqueous42 and tear fluid.43,44
In a recent analysis of 97 randomized treatment trials for herpes
simplex epithelial keratitis comparing the efficacy of topical or oral
antiviral agents with or without debridement, Wilhelmus45 concluded
that vidarabine, trifluridine and acyclovir are effective and nearly
equivalent. In contrast to treatment with idoxuridine, treatment with
vidarabine, trifluridine or acyclovir resulted in a significantly
greater proportion healing in one week. The combination of a
nucleoside and debridement seemed to hasten healing.45
Valcyclovir is the L-valyl ester prodrug of acyclovir with enhanced
bioavailability and significantly greater plasma concentrations of
acyclovir than can be achieved with oral acyclovir.46 Although there
are reports of animal studies, no case series or controlled trials
have been published.
Administration of topical corticosteroid is contraindicated in the
treatment of herpes simplex epithelial keratitis. In the management of
HSV stromal and disciform endotheliitis, topical corticosteroid
therapy combined with prophylactic antiviral cover is a typical form
of treatment. Controversial aspects of corticosteroid therapy is
discussed in detail below.
It is logical to limit inflammatory response in the cornea, because
this is largely responsible for the destructive effects of herpetic
infections. This inflammation has an immunologic basis and might
therefore be combated by systemic and local immunosuppressive
measures. For herpetic disease, the disadvantages and the dangers of
systemic immunosuppressive therapy make its use undesirable.
Corticosteroids can modify the immune response in a number of ways.
Applied locally in the cornea, their effect seems to be chiefly on the
efferent arc, possibly inhibiting chemotaxis and degranulation of
polymorphonuclear leukocytes. Although they also inhibit local
antibody production to some degree, their influence on the afferent
arc and central responses is probably less important. Corticosteroids
appear to have more effect on hypersensitivity reactions mediated by
humoral antibody than by cell-mediated immunity, although their action
in controlling corneal allograft reactions suggests that they may
block such reactions by causing destruction of sensitized lymphocytes.
Steroids are associated with a number of complications that tend to
diminish their effectiveness and at times prohibit their use:
Enhancement of viral replication. Steroids clearly foster Herpesvirus
replication in the corneal epithelium once this has been initiated.
For this reason they must never be used in the treatment of epithelial
herpes. The demonstration of replicating virus in corneal stroma and
deeper ocular tissues by electron microscopy suggests that steroids
may also enhance virus replication in these tissues. There is as yet
no conclusive evidence of this, and experiments in animals have
yielded conflicting results.47 Nevertheless, in the absence of an
antiviral agent that can effectively penetrate the corneal stroma, the
possibility of enhancement of virus replication must cause concern.
Secondary infection. The immunosuppressive effects of steroids may
allow bacteria and fungi to proliferate in the absence of specific
Elevation of intraocular pressure. Prolonged administration produces
an elevation in intraocular pressure in some patients. In our
experience, the latent period before the pressure begins to rise can
be quite variable and may be prolonged. The effects of an unrecognized
pressure elevation in an already diseased eye are devastating.
Cataract formation. Posterior subcapsular cataracts that may progress
to complete lens opacities have been associated with systemic
steroids. However, prolonged local administration is also a risk
factor for cataract.
In the presence of an inflammatory stimulus such as residual herpes
simplex antigen, a rebound in the inflammatory response almost
invariably follows the cessation or too rapid reduction in the topical
steroid therapy. As a consequence of the removal of steroid, immature
leucocytes proliferate and produce antibody in large amounts. Antibody
complexes with antigen and the resulting inflammatory cascade leads to
invasion of the cornea by a new wave of polymorphonuclear leucocytes.
This inflammatory rebound may lead to rapid deepening of corneal
ulceration and perforation. Clinically, the exacerbation in corneal
inflammation may be mistaken for deteriorating underlying disease.48
It is clear that steroids are dangerous preparations in inexperienced
hands because they introduce new hazards to an already complex and
difficult situation. Nevertheless, they are the only effective
anti-inflammatory agent available. It is mandatory that the clinician
be constantly aware of these hazards. The haphazard administration of
steroids in poorly monitored patients contributes significantly to the
disastrous sequelae of this disease.
Back to Top
Primary ocular herpes is predominantly a disease of infants and young
adults, but it can occur sporadically at all ages. Neonatal infection
is caused by HSV-2 in approximately 80% of cases. The scant emphasis
that primary herpetic infection has received in the literature is
regrettable in view of its importance as a cause of follicular
conjunctivitis or keratoconjunctivitis.11 These conditions remain
largely unrecognized, therefore, affected patients may be exposed
unwittingly to the hazards of corticosteroid administration.
Although this section is principally concerned with the keratitis that
often follows primary follicular conjunctivitis, it would be
unrealistic to consider it as an entity distinct from the primary
syndrome. Symptoms of infection appear 2 to 12 days after contact with
an infected individual (although not necessarily one with an active
lesion). In contrast to the recurrent form of the disease, there is
mild malaise and fever, indicating a constitutional illness.
Conjunctival injection, irritability and watery discharge are
typically unilateral and rarely severe. The patient or parents, whose
chief concern may be the skin lesions adjacent to the eye, may not
even mention the ocular disease.
The follicular conjunctivitis of primary herpes is associated with a
regional adenitis. Typically, the ipsilateral preauricular lymph node
is slightly enlarged and a little tender. Swollen lids and a primary
skin lesion are often readily apparent (Figs. 3 and 4), but on
occasion only a careful search will reveal the single or grouped
vesicles of crusted ulcers (Figs. 5 and 6) hidden among the lashes or
in the intermarginal strip. Similar lesions may be located elsewhere
on the face or at the mucocutaneous junction of the mouth, in the
nose, or on the trunk, and they may be easily missed unless a specific
search is made. In nearly one fourth of cases, no cutaneous lesions
are present.35 The conjunctiva is injected and edematous. Follicles
develop, especially in the fornices, and extend to the tarsal areas
(Fig. 7);they rarely occur at the limbus. Small subconjunctival
ecchymoses are not uncommon and phylectenule-like lesions may develop
on the globe (Fig. 8).
Fig. 3 Child with primary ocular herpes. (Courtesy of Dr. S. Darougar)
Fig. 4 Primary herpetic blepharoconjunctivitis in an adult.
Fig. 5 Herpetic ulcer on lid margin in a patient with primary herpes.
(Courtesy of Dr. S. Darougar)
Fig. 6 Umbilicated primary herpetic lesions at the inner canthal area.
(Courtesy of Dr. S. Darougar)
Fig. 7 Acute follicular conjunctivitis in primary herpetic infection:
A Upper tarsus, B Upper fornix, C Lower fornix. (Courtesy of Dr. S.
Fig. 8 Chemosis and ecchymosis of bulbar conjunctiva. (Courtesy of Dr.
Within 2 weeks, approximately half of these patients develop corneal
lesions associated with only relatively minor symptoms: a little
grittiness, photophobia, and blurring of vision. Initially these
lesions are epithelial and present a variety of appearances.
A fine punctate epithelial keratitis, consisting of tiny white flecks
in the superficial layers that stain poorly with fluorescein and
variably with rose bengal, may be present. These are transient spots,
only rarely progressing to larger lesions. As the flecks desquamate,
fluorescein stains the flecks more intensively during the healing
A coarse punctate epithelial keratitis presenting a variety of shapes
(circles, ovals, irregular elongated areas, and stellate figures) may
appear. Any of these lesions may progress to macroscopic dendritic
figures. They consist of slightly raised, closed clusters of opaque
epithelial cells, those in the periphery often being the most
regularly arranged. These swollen white cells stain well with rose
bengal but poorly with fluorescein. Typical herpetic intranuclear
inclusions can be demonstrated in these cells (Fig. 9).Initially,
there is no stromal reaction, but within 2 to 3 weeks, and sooner if
the lesions are peripheral (and regardless of epithelial healing),
subepithelial infiltrates appear.35
Fig. 9 Corneal epithelial cells stained in vivo with rose bengal in a
child with coarse punctate corneal epithelial lesions of primary
Herpesvirus infection. The cells were subsequently removed and
counterstained with hematoxylin. Many stained cells are swollen and
show eosinophilic intranuclear inclusions with margination of
DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS
The diagnosis can be made on clinical grounds alone in patients with
typical cutaneous lid lesions or typical herpetic corneal lesions. In
the absence of such lesions (approximately one fourth of all patients
with primary herpetic conjunctivitis), laboratory investigations are
essential for diagnosis. The differential diagnosis includes:
Keratitis with lid lesions: zoster, chickenpox, molluscum contagiosum,
and ulcerative blepharitis with keratitis caused by staphylococcal
Keratitis without lid lesions: vaccinia, adenoviral infections (types
3, 7, and 8 and 19), chlamydial infections, herpes zoster and
In most cases, laboratory confirmation of the clinical diagnosis is
unnecessary. In the remainder, an attempt should be made to isolate
the virus from untreated active lesions in skin and cornea and from
the conjunctiva. Positive cultures may take from 2 to 5 days to
Typical viral multinucleate giant cells may be demonstrated in
Giemsa-stained scrapings of the base of cutaneous lesions on the lid.
These are also seen in varicella or zoster.
The appearance of neutralizing and complement-fixing antibodies a week
after the onset, followed by a rising titer for the next few weeks, is
useful confirmatory evidence. The cytology of cornea and conjunctival
scrapings is useful in conjunction with antibody levels but is not
Therapy must be directed toward the elimination of virus from the
cornea and adjacent skin lesions. It is essential that lid vesicles
and ulcers be treated concurrently with the corneal disease because
they are a potent source of virus that, being continually shed, can
reinfect the cornea and vastly prolong the keratitis.
Trifluridine is instilled into the conjunctival sac fives times per
day. An antiviral ointment (acyclovir) can be applied to the eyelid
and adjacent skin lesions. Topical corticosteroids are
Systemic administration of acyclovir is recommended for neonatal
infections because of the enhanced risk of systemic disease as a
result of viral dissemination. In this age group, administration of
eye drops can be difficult. The addition of systemic therapy has the
added benefit of ensuring adequate therapeutic concentrations in the
eye. In older children and adults with a primary infection, systemic
therapy is usually unnecessary.
Since the advent of topical antiviral agents, debridement of the
corneal epithelium is seldom performed but it remains an effective
method of healing the corneal lesions. Solitary vesicles on the lids
may be removed by general debridement with a cotton-tipped applicator
moistened with phenol.
A cycloplegic may be prescribed when indicated to relieve photophobia
or ciliary spasm. The use of Atropine should be avoided because of the
risk of hypersensitivity. Scopolamine hydrobromide 0.25% twice daily
or cyclopentolate hydrochloride 1% to 2% three times daily is usually
effective. Patching is undesirable. However, wearing sunglasses may
give symptomatic relief.
A return visit is advisable within 2 to 3 days in all but the mildest
infections. Thereafter, patients can be seen weekly, provided recovery
is uneventful. The administration of an antiviral agent must be
continued until corneal and lid lesions are healed. Hospitalization is
rarely necessary, except in patients with severe bilateral and
The following steps are necessary only when the diagnosis is in doubt.
Gently swab the surfaces of the lesion with a dry, sterile,
cotton-tipped applicator. Place the applicator in a viral transport
medium for later inoculation into tissue culture. If cell lines of the
culture are not immediately available, the specimen can be frozen at
Gently swab (or preferably scrape) the conjunctiva in a similar manner.
Unroof ulcers and vesicles with a fine needle tip prior to taking the
specimen with a cotton-tipped applicator.
Gently scrape the opaque cells of the corneal lesion onto a glass
slide, using a platinum spatula, Beaver blade, or Bard-Parker knife.
This is best done under magnification, preferably at the slit lamp.
Stain with Giemsa.
On initial presentation, blood is drawn for neutralizing antibody
titers. Two to 3 weeks later, another sample is assayed to determine
whether the titer has risen.
Cultures of the eyelid lesions and conjunctival sac are desirable.
They should be accompanied by direct smear if bacterial infection is
NATURAL COURSE AND VARIATIONS IN CLINICAL PRESENTATION
The epithelial lesions tend to heal, but additional crops may appear
if active herpes persists untreated on the lids. The superficial
stromal infiltrates may persist for several weeks before gradually
resolving. Healing may be accompanied by superficial scarring in these
sites. Occasionally, these stromal lesions progress to a frank
disciform keratitis indistinguishable from the type usually associated
with recurrent herpes.
Primary herpes uncommonly presents as a bilateral ocular disease,
except in atopic individuals, who tend to have a more florid form of
disease.49 In rare instances, the course of primary infection is
severe: widespread herpetic infection in the face and trunk, often
pustular and accompanied by a severe systemic illness, may supervene
and is characteristic of Kaposi’s varicelliform eruption. Encephalitis
and hepatitis occur rarely but can be lethal, especially in infants.
Secondary bacterial infection occasionally supervenes. The appropriate
antibiotic, the selection of which is made from the results of culture
and sensitivity testing, best treats it locally. Severe cellulitis of
the lids may require systemic antibiotics. Prophylactic antibiotics
are unnecessary and may confuse the clinical picture. In atopic
individuals, management can be difficult, because topical antivirals
may not be tolerated. Careful debridement of the lesions is an
alternative in these situations.
Permanent damage to the cornea is uncommon. Most of the opacification
in the stroma seen during active infection is probably caused by
temporary edema rather than scarring. Should stromal keratitis
develop, it is treated in accordance with the principles of management
of recurrent disease (discussed later). Transient superficial stromal
infiltrates do not require treatment.
Most primary infections respond rapidly to antiviral therapy with
little or no sequelae and no loss of vision. Occasionally, the
development of antiviral toxicity necessitates cessation of the drug
therapy. Complications occur mainly in undiagnosed or ineffectively
Back to Top
RECURRENT HERPES SIMPLEX
Recurrent HSV infection occurs as a result of reactivation of the
virus in latently infected ganglia. Recurrent ocular HSV infection is
thought to be caused by reactivation of the virus in the trigeminal
ganglion. The virus travels down the nerve axon to the sensory nerve
endings, where it is transferred to corneal epithelial cells and
keratocytes. If favorable conditions exist in the epithelium, viral
replication and cell lysis ensue, producing clinical disease.
HSV-specific nucleic acid sequences have been detected, and HSV has
been organ-cultured from corneal buttons excised from patients with
chronic stromal keratitis. These patients had no active disease at the
time that penetrating keratoplasty was performed. These data suggest
that the human cornea may also be a site of latency and a potential
source of recurring clinical ocular disease.50
There are several different types of recurrent ocular HSV infection,
including dendritic and geographic epithelial keratitis, interstitial
and necrotizing stromal keratitis, disciform endotheliitis, and
In the corneal epithelium, normal epithelial cells are interspersed
with balloon cells (cytoplasmic vacuolation with marginated
chromatin). These balloon cells stain intensely with rose bengal in
vivo and contain replicating virus. Syncytial multinucleate giant
cells and occasionally epithelial cells with intranuclear eosinophilic
inclusions are also seen. The leukocytes that are present are
predominantly mononuclear. During a recurrent episode, the
inflammatory cell type in the conjunctiva is also predominantly
mononuclear with some admixture of polymorphonuclear leukocytes. Only
occasionally are giant cells and inclusions seen. By contrast, in the
initial stages of a primary infection, the predominant cell is the
polymorphonuclear leukocyte. Only after several weeks does the
mononuclear cell dominate the picture.
Initially, the epithelium is swollen along the margins of the
dendritic ulcer because of intercellular and intracellular edema. The
previously described general cytologic pattern is present and there
are necrotic cells in the ulcer bed. These changes gradually progress
to complete epithelial loss in some areas and the accumulation of
inflammatory cells and debris. In the geographic type of ulcer, the
changes are similar but more extensive.
In the early stages, the lesion is confined to the epithelium.
However, with time the process spreads to involve the anterior stroma.
Bowman’s layer and the immediately adjacent stroma become edematous,
necrotic in places, and infiltrated by a variable number of
inflammatory cells, predominantly polymorphonuclear leukocytes.
Purely epithelial lesions heal rapidly and with little scarring;
however, with stromal involvement superficial scarring occurs. Some
degree of faceting is inevitable when there has been loss of corneal
Patients with epithelial keratitis caused by HSV may be asymptomatic
or may experience mild to severe foreign-body sensation, photophobia,
redness, and blurred vision. After a number of episodes, the symptoms
of foreign-body sensation are commonly muted by cornea hypoesthesia.
Recurrent HSV epithelial keratitis typically has a classic dendritic
(dichotomously branching) shape. The pathogenesis of the branching
ulceration has not been elucidated. It may simply be a function of
viral linear spread by contiguous cell-to-cell movement.
Initially, a plaque of opaque cells appears on the epithelial surface.
Although usually dendritic, the shape may be coarsely punctate or
stellate. Within a few days, the center of the plaque desquamates to
form a linear, branching ulcer (Fig. 10) barely 0.1 mm wide with
overhanging margins of swollen opaque cells (Fig. 11). The dendritic
figure may be single or multiple; it can extend across the entire
cornea but is usually considerably smaller (Fig. 12). At the ends of
the branches terminal bulbs are typically seen. The cells lining the
edge of the ulcer are laden with virus and stain brilliantly with rose
bengal. Fluorescein stains the ulcer bed and seeps beneath the
adjacent cells (Fig. 13). Several days after the appearance of the
dendritic ulcer, infiltrate appears in the immediately subjacent
stroma. It usually remains superficial and localized. In addition to
these changes, scattered punctate epithelial erosions are common and
evidence of past attacks in the form of superficial scars and
superficial vascularization may be present (see Fig. 10). Although
dendritic keratitis can occur at any location on the cornea,
recurrences tend to affect the same areas noted in previous attacks.
Initially, corneal hypoesthesia is focal, so a great proportion of the
cornea appears unaffected. With repeated episodes, the loss of corneal
sensation becomes more profound. Ulcerations can occasionally occur
within 2 mm of the limbus. These lesions may not exhibit the typical
features of a dendritic ulcer and may be mistaken for staphylococcal
marginal keratitis. They tend to be more resistant to antiviral
therapy than more centrally located herpetic infections.
Fig. 10 Linear dendritic ulcer stained with rose bengal.
Fig. 11 Macrophotograph of portion of a dendritic ulcer by
retroillumination. Note opaque heaped-up cells along the ulcer margin.
(Magnification × 10) (Courtesy of Mr. N. Brown)
Fig. 12 Extensive dendritic ulcer, stained with rose bengal, running
around the margin of the corneal graft.
Fig. 13 Geographic ulcer stained with rose bengal and fluorescein.
Rose bengal especially stains cells lining the ulcer margin. Note the
greenish tinge in the central area of the ulcer and the green halo
around the ulcer margin. This effect is caused by fluorescein seeping
into bare stroma and under ulcer margins.
The disease process is usually confined to the cornea. However,
ciliary injection can be quite intense and frequently appears out of
proportion to the symptoms. Slight flare with an occasional cell is
indicative of a mild uveal reaction, but keratic precipitates (KP) are
uncommon. Concurrently with the corneal lesions, vesicles or ulcers
can develop on the lids, face, mucocutaneous junction of the mouth,
and nose or elsewhere.
Diagnosis and Differential Diagnosis
The true dendritic ulcer is pathognomonic and requires no laboratory
confirmation. In cases in which there is doubt, viral isolation should
be attempted. The differential diagnosis is extensive (Table 2).
TABLE 2. Differential Diagnosis of Dendritic Keratitis
Herpes zoster dendritic keratitis
Mucous plaques in herpes zoster ophthalmicus
Contact lens keratopathy
Healing ruptured bleb
Recurrent erosion syndrome
Epstein Barr keratitis
Tyrosinemia type 11 (rare)
In cases in which the diagnosis is in doubt, an attempt can be made to
recover the virus from untreated corneal lesions. However, viral
culture is expensive. It lacks some specificity because Herpesvirus
can be recovered from the tear film in the absence of corneal
epithelial disease. Examination of corneal scrapings may reveal
typical cytology. Herpetic antigen can be detected in corneal
scrapings by use of a fluorescein antibody staining technique. The
polymerase chain reaction can also be used to identify herpetic
nucleic acid. It is both sensitive and specific for Herpesvirus.
Conjunctival smears are not diagnostic, showing a nonspecific
inflammatory response that is predominantly mononuclear in some cases
but largely polymorphonuclear in others. Serum-neutralizing antibody
titers are elevated but do not rise further during the recurrent
episodes, whereas rising titers of complement-fixing antibody are
As in primary herpes, the aim is the speedy elimination of virus from
the epithelium to reduce the risk of significant stromal keratitis and
to minimize scarring. The steps to be taken are listed below:
In most cases, all that is required to confirm the diagnosis is to
stain the lesion with rose bengal, 1%.
In doubtful cases, consider scraping lesions for cytology, fluorescent
antibody studies, or viral isolation. If there is any question of
secondary bacterial infection, cultures can be made from the lid
margins, conjunctiva, and the lesion itself.
Antibody titers are of use only in the diagnosis of primary herpes.
Although most ulcers may be treated with an antiviral agent or minimal
wiping debridement, the former may be is preferred for convenience.
Debridement followed by topical antiviral therapy combines both
approaches and may lead to more rapid healing.45 Debridement should be
avoided when there is significant stromal involvement, when the
ulcerated area is large or when steroids have enhanced the ulcer.
Debridement may be impractical in children because of difficulties in
cooperation. On the other hand, when antiviral resistance, toxicity or
hypersensitivity is present, debridement is a good alternative.
Trifluridine, 1% drops, are administered every 2 hours until bedtime.
As an alternative, vidarabine or acyclovir (3% ointment if available)
is given five times daily. IDU can be administered as an ointment five
times daily or as drops every hour. However, its use is best avoided
because of toxicity.
With steroid-enhanced ulcers, the topical corticosteroid should be
progressively reduced as quickly as possible while instituting
antiviral therapy. Too-rapid reduction may lead to an intense
inflammatory rebound and it may be necessary to restart
corticosteroids temporarily. The goal is to eliminate the steroid
completely while there is active viral replication.
For patient with severe photophobia and ciliary spasm, a cycloplegic
can be prescribed as required. Scopolamine hydrobromide 0.25% twice
daily or cyclopentolate hydrochloride 1% to 2% is usually effective.
Analgesics may be necessary for pain relief, especially in the first
48 hours of treatment. Wearing sunglasses may give symptomatic relief,
but patching is rarely necessary.
Patients should be encouraged to remain at work if at all possible.
Even with the availability of much improved antiviral therapy, they
face the prospect of further episodes. The cumulative effects of
temporary incapacity are often psychologically and economically
Ulcers that are treated with antiviral agents tend to heal by breaking
up into islands so that the typical dendritic shape is lost. This
change is usually apparent within 2 to 3 days and thereafter these
patients are seen every 5 to 7 days until healing has occurred.
Steroid-enhanced ulcers should be examined more frequently until it is
certain that healing is occurring. Debrided cases should be seen every
2 to 3 days until healed. Repeated debridement is sometimes necessary.
At times, an epithelial lesion may lose its typical configuration to
form an ulcer that is usually round and of variable size (Fig. 14).
The edges are rolled in appearance and do not contain the “heaped-up”
opaque cells that stain brightly with rose bengal typical of a
dendritic or geographic herpetic ulcer. Evidence is lacking of viral
replication in the epithelium in such indolent, so-called metaherpetic
ulcers. However, the electron microscopic picture of Herpesvirus
replication may occasionally be demonstrated in indolent ulcers deep
in the stroma (Figs. 15 and 16). Often, there is considerable stromal
edema or infiltrate associated with these ulcers, or the ulcer may
represent breakdown over a previously scarred area. Indolent ulcers
are more common in stromal keratouveitis and will be discussed in more
Fig. 14 Indolent herpetic ulcer. This type of ulcer tends to be
circular with smooth, rolled margins that stain poorly, if at all,
with rose bengal (see Fig. 15). Electron microscopy revealed
Herpesvirus particles in keratocytes at all depths under this ulcer.
Fig. 15 Indolent herpetic ulcer in the same patient as in Fig. 14.
Rose bengal stains the base but not the epithelium at the edge of
Fig. 16 Dendritic-ameboid ulcer stained with rose bengal. Dendritic
shape is still discernible, but ulcer has widened considerably.
Antiviral toxicity is probably the most frequent immediate
complication of recurrent epithelial herpes and it may occur as soon
as 10 days after initiation of therapy. Because this is often the
period when healing is occurring, the appearance of new staining can
give rise to considerable confusion. It is sometimes impossible to be
sure whether the signs represent recrudescence of the infection or the
onset of toxicity. This dilemma may be resolved only in retrospect.
When there is doubt, cessation of antiviral therapy is often
necessary. Resolution of antiviral toxicity is extremely slow and may
take weeks. Toxic effects of the current topical antiviral agents51
are similar, although trifluridine seems to be the least toxic of the
Herpetic stromal keratitis is a serious complication. It develops in
approximately 3% of dendritic ulcers. Steroids are frequently needed
for its control. The subsequent section deals with this in detail.
Management of bacterial infection in these patients should always be
based on the results of isolate recovery.
Natural Course and Variations of Clinical Presentation
Untreated, the dendritic ulcer may spontaneously heal, but it usually
persists and may insidiously arborize to previously uninvolved
epithelium. Segments of the lesion may broaden, or large areas of
epithelium may desquamate so the ulcer assumes a geographic
configuration (see Figs. 13 and Figs. 16, 17, and 18). This is
particularly likely to happen when steroids have been administered.
The margins are similar in appearance to a dendrite and contain
actively replicating virus (see Figs. 16 and 17). With persistence of
the ulcer, especially if it enlarges, stromal involvement becomes more
marked and the uveitic reaction may become more intense. Eventually
the stromal keratouveitis may dominate the clinical picture (see Fig.
Fig. 17 Macrophotograph of the margin of a geographic ulcer. Note the
ragged appearance and the opaque swollen cells lining the ulcer
margin. (Magnification × 10) (Courtesy of Mr. N. Brown)
Fig. 18 Typical ameboid ulcer stained with fluorescein. Dye stains
whole ulcerated area. (Fig. 18, courtesy of Mr. A.J. Bron)
Healing is accompanied by a variable degree of scarring, dependent on
the severity of stromal infiltration. Portions of the ulcer may
persist for a considerable time as opaque, slightly elevated nodes
that stain with rose bengal. It is uncertain whether these represent
sites of continuing viral replication.
With recurrences of epithelial keratitis, superficial vascularization
is not uncommon. Corneal hypoesthesia in initial episodes may be
slight and only a portion of the cornea may be affected. However, with
repeated attacks it becomes a feature of the disease.52
Occasionally an apparent abortive form of herpes develops. Small
epithelial mounds appear that stain with rose bengal. Herpesvirus can
be cultured from these lesions. They may persist for a considerable
time before fading or transforming into the typical dendritic ulcer.
Most cases heal with remarkably little scarring. Unfortunately, with
repeated attacks there is an inevitable accumulation of superficial
scarring with the formation of corneal facets. Vision may be markedly
affected. In some instances, and especially with steroid-worsened
ulcers, healing is excessively prolonged and in others there is
gradual transition to a stromal keratitis. Occasionally, the ulcer
will heal partially and then worsen while treated with antiviral
therapy. This probably indicates the emergence of a resistant strain
of herpes18 or inadequate administration of the drug.
In children, dendritic keratitis is usually associated with a good
visual outcome. However, in a subset with geographic ulcers the
outlook is less optimistic. Vision tends to be worse with more
scarring, a higher degree of astigmatism and more recurrences. In this
group, an aggressive approach is necessary including the use of oral
acyclovir, prompt treatment of stromal keratitis should it develop and
close monitoring for the onset of amblyopia.53
The corneal stroma is the site of an inflammatory reaction that is
irregularly distributed, is of varying intensity and is accompanied by
an anterior uveitis. Epithelial keratitis, as previously described, is
variably present. In addition, there may be a more generalized
epithelial edema in which the epithelium may be separated from
Bowman’s layer by the edema fluid (Fig. 19).The corneal lamellae may
be necrotic in places, and inflammatory cells, predominantly
polymorphonuclear leukocytes, diffusely and focally infiltrate the
stroma. The process may involve the cornea at all levels from Bowman’s
layer to Descemet’s membrane. The endothelium may be edematous or
infiltrated by inflammatory cells and, especially beneath the stromal
lesion, may be replaced by a coagulated film of fibrin and
inflammatory cells. Keratic precipitates are prominent. The aqueous
contains fibrin and inflammatory cells (neutrophils, lymphocytes,
macrophages, and plasma cells). These cells frequently infiltrate the
angle and the trabecular bands are thickened and appear edematous, so
the term trabeculitis appears justified. Involvement of the iris is
variable, but it is frequently infiltrated by lymphocytes and plasma
cells and thickened by edema. The anterior ciliary body shows a
similar involvement. Posterior synechiae and anterior lens changes are
common, frequently in association with a fibrovascular membrane
extending across the pupil.
Fig. 19 Gross epithelial edema with bullae, resulting from severe
endothelial decompensation, in an eye that had previously been the
site of a severe herpetic uveitis. As edema resolved, endothelium was
seen to be studded with numerous secondary guttatae.
When ulceration occurs, Bowman’s layer and superficial lamellae are
replaced by debris and inflammatory cells. The ulcer may deepen, form
a descemetocele (Fig. 20), and ultimately perforate. If this is the
case, the immediately adjacent stroma is necrotic, edematous, and
densely infiltrated by acute and chronic inflammatory cells.
Fig. 20 Descemetocele in an eye with severe stromal keratitis.
Epithelium has been stained with rose bengal.
Repair is accompanied by scarring and vascular ingrowth (Fig. 21), but
foci of active inflammation may persist for a considerable period. The
endothelium has a remarkable propensity for recovery but may be
replaced by shrunken keratic precipitates.54
Fig. 21 Herpes simplex interstitial keratitis. Active inflammation has
resolved, leaving stromal scarring, thinning, and stromal
HSV is associated with several classes of antigens, including soluble
diffusible antigens that are released from an infected cell when it is
lysed, antigens fixed to the surface of the infected cells, and
insoluble large structural proteins that are capsid components. Any of
these classes of antigens can probably react with antibody,
complement, or sensitized cells and initiate the immune response.
Although viral particles have been demonstrated by electron microscopy
within the corneal stroma in cases of stromal herpes simplex
keratitis55–57 (Figs. 22 and 23), attempts to isolate infected virions
in tissue culture have been successful in only a minority of cases. In
most cases of stromal herpes simplex keratitis, the clinical disease
appears to be predominantly the result of an immunopathologic process,
which is a response to viral antigen rather than an active infectious
process caused by replicating virus. Immunocytochemical studies of
cornea tissue obtained from patients with herpes simplex stromal
keratitis at the time of penetrating keratoplasty have shown the
stromal infiltrate to be composed largely of macrophages and
lymphocytes.58 Controversy exists as to whether cytotoxic or helper T
cells play the major role in herpetic keratitis.58
Fig. 22 Electron micrograph of deep corneal stroma. Numerous
Herpesvirus particles can be seen lying in and around two degenerate
cells (probably keratocytes) between stromal lamellae. (Magnification
× 16,000) (Courtesy of Dr. R. Tripathi).
Fig. 23 Electron micrograph of keratocyte. This degenerating cell is
filled with Herpesvirus particles. Some of these have typical
morphology (arrow). Note the number of incomplete forms, empty capsids
and great variability in size (arrow). (Magnification × 100,000)
(Courtesy of Dr. R. Tripathi).
However, it is thought that some cases of herpes simplex stromal
keratitis may be caused by a combination of active viral replication
and the immune response.1 Specifically, some cases of necrotizing
stromal herpes simplex keratitis may be the result of this dual
Clinical Manifestations and Variation in Clinical Presentation
The clinical manifestations of stromal involvement with HSV are
protean. Patients exhibiting stromal keratouveitis commonly have a
history of previous attacks of epithelial herpes. The stroma may have
been involved to some degree in these episodes, but the emphasis for
the most part remains directed toward the epithelial keratitis. Then,
insidiously over a few episodes, but at times quite suddenly, the
pattern changes so that stromal disease becomes the dominant feature.
Occasionally, this time scale is shortened, and in extreme instances,
the transition is completed in the initial attack. Some patients will
develop stromal keratouveitis, having had an epithelial herpes in the
past, or will experience their first dendritic ulcer subsequently; in
others, stromal keratouveitis will follow an episode of dendritic
keratitis. It is important to realize that regardless of preceding
events, the onset of stromal disease is a serious portent because it
marks a new stage in the disease; deeper ocular structures are
involved, vision is seriously threatened and morbidity is
Apart from a complaint of blurred vision, the signs and symptoms are
nonspecific. The eye feels uncomfortable and tears excessively. The
pain experienced varies considerably from patient to patient. These
signs and symptoms are extremely variable and can be difficult to
interpret, especially when viewed against a background of previous
structural damage, secondary glaucoma, and endothelial dysfunction.
Evidence of coexisting inflammatory and reparative processes can be
recognized by slit lamp examination. Corneal edema, infiltration,
vascularization, ulceration, endothelial inflammation, and uveitis can
all occur to varying degrees in herpes simplex stromal disease. At
times, the degree of cellular infiltration and edema will indicate
that infiltration is the dominant process and, at other times,
scarring and neovascularization are more apparent.
Several common response patterns can be distinguished clinically,
which can aid in making the diagnosis and guiding treatment. These
include chronic interstitial keratitis and necrotizing stromal
When the predominant clinical findings include stromal infiltration
accompanied by an intact epithelium, the term interstitial keratitis
is appropriately used. The infiltration can present as single or
multiple patches of infiltrate and edema and involve the entire
stromal thickness or discrete lamellae. The infiltration tends to run
a chronic, indolent course that persists for many months. Superficial
and deep stromal vessels often accompany the infiltrate and can occur
early or late in the disease course. The infiltrates may resemble
those seen in infection with other viruses, bacteria, fungi, or
acanthamoeba but tend to be more indolent and with an intact
epithelium. This form of stromal inflammation is thought to represent
the inflammation often leads to the formation of a dense, white
vascularized scar (see Fig. 21).
When the stroma is edematous, it exhibits a ground-glass appearance
and is thickened. Edema commonly encompasses the infiltrate and, at
times, is the main feature of the disease and it is also an essential
component of the stromal inflammatory reaction. It may result in
endothelial dysfunction secondary to uveitis.
Limbal vasculitis and immune (Wessely) rings in the anterior stroma
are two other clinical manifestations of presumed immune stromal
disease.60–62 The Wessely ring is a partial or complete ring of
infiltrate in the stroma, surrounding the main stromal lesion and
separated from it by a relatively clear zone of cornea. It presumably
results from the inflammatory reaction to a ring or arc precipitate of
antigen-antibody complexes. Limbal vasculitis presents as an
edematous, hyperemic reaction. Although usually focal, more than one
quadrant may be involved. These vessels will often invade the cornea
while associated with stromal interstitial keratitis.
In the epithelium, a fine superficial edema, occupying a variable
surface area over the active stromal lesion, is common and is related
to endothelial dysfunction. At times the epithelium is grossly
edematous, with recurrent bullae appearing and sometimes breaking down
to form indolent ulcers that have to be distinguished from active
epithelial viral disease. Punctate erosions that stain well with rose
bengal and fluorescein are frequently seen.
The endothelial layer of the cornea is involved in all but the most
superficial lesions. Fine, white keratic precipitates may be scattered
over the surface or crops of discrete white precipitate may appear
(Fig. 24). Some of these lesions become pigmented. It is not unusual
for large endothelial plaques to develop in relation to the active
stromal lesion (Fig. 25). Secondary guttae are quite common but
Fig. 24 Typical keratic precipitates in an eye with a disciform
keratitis caused by herpes.
Fig. 25 Large endothelial plaque behind disciform keratitis.
Anterior uveitis is invariably present although it may be difficult or
impossible to assess because of the corneal opacification. In severity
it varies from the presence of an occasional cell and minimal flare to
the development of hypopyon. Posterior synechiae and rubeosis iridis
frequently complicate severe cases. The intraocular pressure may be
elevated in the acute stage as a result of an associated trabeculitis.
Vascularization can occur at any stage of the disease process. Vessels
penetrate the stroma from the limbus at all levels to invade the
active stromal lesion. They are often cuffed by fine granular
infiltrates while active but, as the inflammatory process subsides,
they lose the cuff of cells and may eventually become almost
bloodless. The first signs of a recrudescence of inflammation may be
the reactivation of these vessels.
During the acute stages of the inflammatory reaction, scarring may not
be obvious, but as the signs of inflammation subside, it becomes more
apparent. In the early stages the discrete white opacification may
easily be mistaken for infiltrate. Once significant scarring has
occurred, the clarity of the cornea is permanently impaired (see Fig.
Loss of corneal substance, ranging from minor faceting to gross
thinning or even perforation, is of variable occurrence. It often
relates to preceding dense infiltration and is most frequently seen in
the rebound phenomenon following withdrawal of topical steroid therapy
for stromal herpetic keratitis with dense infiltration (Fig. 26).
Fig. 26 Perforation of the cornea at the site of dense infiltration in
severe herpetic keratitis.
During the later stages of healing as the inflammation subsides, hard
white or sometimes yellowish lipid deposits, either crumbly dots or
fine crystals, may appear. Usually these are seen within vascularized
opacities and may be progressive. In some patients, they are
associated with demonstrable abnormalities of lipoprotein metabolism.
These deposits are not an indication for continuing energetic steroid
therapy. It is important to differentiate them from dense inflammatory
infiltrates, with which they are sometimes confused.
NECROTIZING STROMAL KERATITIS.
Necrotizing stromal keratitis is manifested clinically as a dense
yellow-white infiltration within the corneal stroma. The predominant
clinical pattern is stromal infiltration and necrosis (Fig. 27). This
more complicated manifestation of herpes simplex keratitis usually
occurs in corneas that have had recurrent episodes of herpetic eye
disease. Thus, it may follow chronic or recurrent epithelial disease,
disciform keratitis, superficial stromal disease, or recurrent disease
of any type. In a prospective study of 152 patients with either
dendritic or geographic epithelial keratitis, Wilhelmus and
coworkers63 noted that one-fourth of their patients developed
subsequent stromal inflammation. Of these, 37% presented with
necrotizing inflammation as the predominant pattern.
Fig. 27 Herpes simplex necrotizing stromal keratitis. A dense
yellow-white infiltrate occurs in the stroma with breakdown of the
In mild cases, infiltrates can be localized, but in more severe cases
a stromal abscess (Fig. 28) may develop, consisting of necrotic,
cheesy-white infiltrate that may occupy the entire cornea thickness.
The overlying epithelium often breaks down over the stromal
infiltrate. This can be followed by the appearance of edema,
ulceration, and stromal neovascularization. Ring infiltrates (Wessely
ring) may occasionally be seen surrounding the stromal infiltrate (an
of polymorphonuclear leukocytes. Uveitis is nearly always present and
may be severe, with retro corneal membrane, hypopyon, synechiae
formation, secondary glaucoma and secondary cataract. Stromal
perforation or super infection with fungi or bacteria can occur (see
Fig. 28 Large irregular stromal abscess underlying a large ameboid
ulcer, with a small hypopyon. Superficial and deep vascularization is
The frequent documentation of viral particles or antigen in herpes
simplex necrotizing stromal keratitis supports the belief that this
form of keratitis is a direct viral infection of the stroma with a
subsequent host immune response.1,55,62,64 Holbach and colleagues65
found that 91% of keratectomy specimens from patients with ulcerative
necrotizing keratitis displayed HSV antigens, compared to only 11% of
keratectomy specimens from patients with nonulcerative,
nonnecrotizing, or disciform keratitis. These antigens were located
primarily in stromal keratocytes and the extracellular stroma.
It is clear that there is a tremendous variation in the appearance of
the cornea, but if each case is approached with appreciation for these
features, activity of the disease can be assessed in a way that has
meaning both clinically and pathologically. Thus, a stromal keratitis
that is superficial and free of new vessels is still relatively mild,
whereas involvement of a full-thickness cornea, associated with a
significant uveitis and neovascularization, indicates severe disease.
Diagnosis and Differential Diagnosis
The diagnosis must be made on clinical grounds alone. In most cases
there is little problem, but occasionally the diagnosis can be in
doubt. Because the presentation of HSV stromal disease can be so
variable, many other conditions can result in similar clinical
presentations and must be considered in the differential diagnosis. A
history of recurrent disease and prior herpes simplex epithelial
keratitis can be helpful, but a history of herpetic epithelial
keratitis is not absolute evidence that the subsequent stromal
keratitis is herpetic in origin. The laterality of HSV stromal
keratitis may also be important in establishing a clinical diagnosis
because bilateral disease occurs rarely.66 Many potential causes of
stromal inflammation will have associated systemic disease that offers
clues to the etiology and helps differentiate it from HSV, such as
herpes zoster ophthalmicus, Cogan’s interstitial keratitis,
Epstein-Barr virus, and mumps. A history of previous corneal trauma or
contact lens wear, particularly associated with disruption of corneal
epithelium, should make one more suspicious of bacterial, fungal, or
acanthamoeba infection. Although HSV keratitis can present
predominantly in a perilimbal location, confinement of the stromal
inflammatory process to the peripheral cornea should alert the
examiner to associated eyelid disease (staphylococcal keratitis),
adjacent scleral inflammation, or possible collagen vascular disease.
The thrust of the investigation should be most appropriately directed
to a consideration of possible differential diagnoses. Stromal
herpetic eye disease is best diagnosed from the patient’s history and
the clinical appearance of the cornea.
If the epithelium is involved in a patient with stromal disease,
cytologic examination may reveal multinucleated giant cells (Giemsa
stain) and intranuclear and eosinophilic inclusions that are
infrequent in adults (Papanicolaou stain). These tests are easy to
perform but are relatively insensitive. Viral isolation in human cell
culture is unavailable to many and is expensive. A variety of
immunologic tests can be used to detect viral antigen in tissue
specimens. These include immunofluorescent staining, immunoperoxidase
staining, immunofiltra-tion techniques, enzyme-linked immunosorbent
assay (ELISA), and DNA probes. The most sensitive test readily
available commercially is the Herpchek, a simple kit based on the
Treatment of HSV stromal keratitis is considerably more controversial
than treatment of HSV epithelial disease. Mechanisms involved in the
pathogenesis of HSV stromal keratitis are complex and incompletely
understood. Both virus and host immune factors appear to be important
in the development and progression of HSV stromal keratitis.58 The
goal of management of HSV stromal keratitis is to guide the patient
through each episode while minimizing ocular damage, reducing
morbidity, and reducing the side effects of treatment.
In general, the most frequently used therapy for management of HSV
stromal keratitis currently includes the judicious use of topical
steroids with prophylactic antiviral cover. The dosing and frequency
of both steroid and antiviral cover are debatable. The Herpetic Eye
Disease Study (HEDS) was designed in an effort to resolve these
controversies, reach consensus on the management of herpes simplex
stromal disease, and establish therapeutic guidelines for antiviral
and anti-inflammatory agents.69,70 The series of double-blinded,
placebo-controlled, multicentered clinical trials included studies
designed to compare (1) the efficacy of topical corticosteroid, (2)
the efficacy of oral acyclovir combined with topical steroid for the
treatment of herpes simplex stromal keratitis and iridocyclitis, (3)
the efficacy of oral acyclovir in the prevention of stromal keratitis
or iridocyclitis in patients with HSV epithelial disease, and (4) the
efficacy of acyclovir in the prevention of recurrent HSV ocular
TOPICAL CORTICOSTEROID THERAPY.
Corticosteroids will suppress an immune response, resulting in reduced
corneal edema, inflammation, infiltration and neovascularization.
Therefore, many believe they are indicated in the treatment of HSV
stromal keratitis (with antiviral cover) to reverse the inflammatory
response, minimize permanent structural alteration, and improve
corneal clarity.71,72 Others caution that topical corticosteroids may
prolong the course of the disease and increase the severity of the
stromal keratitis.73,74 Although steroids do not experimentally induce
recurrent herpetic epithelial keratitis,75 they can predispose to an
increased susceptibility to recurrent infection76 and can exacerbate
active viral infection. Corticosteroids can also predispose to
secondary complications, including microbial super infection, stromal
melting, secondary glaucoma and cataract formation. Once
corticosteroids are begun, it is often difficult to discontinue them
and a marked rebound inflammatory response can ensue when withdrawal
is too abrupt. Nevertheless, in the HEDS controlled trial of topical
corticosteroid, given concomitantly with trifluridine, for herpes
simplex stromal keratitis, the topical steroid was significantly
better than placebo in reducing the persistence or progression of
stromal inflammation. The regimen also significantly shortened the
duration of the keratitis.77
The introduction of corticosteroids in the management of HSV stromal
keratitis is influenced by the need to control the inflammatory
process and to provide symptomatic relief to the patient. Effective
and safe administration of steroids requires close observation in
reliable patients. The dosage of steroid requires some judgment and
several options of corticosteroid administration are available.78 The
use of the lowest effective dose seems prudent because the goal of
therapy is to produce a clinically recognizable reduction in
inflammation while minimizing undesirable side effects. An antiviral
agent (trifluridine) should be administered concurrently to reduce the
chance of recurrence of live virus in the epithelium. Cycloplegics,
lubricants, and dark glasses can provide symptomatic relief.
Neither idoxuridine79 nor vidarabine37 is clinically effective
topically for herpes simplex keratouveitis or deep stromal disease.
Therapeutic levels of trifluridine can be obtained in the iris and
anterior chamber80 and may be of value in deep stromal disease and
uveitis, although this is unproven. Various investigators in small,
uncontrolled studies have reported a beneficial effect of topical
acyclovir combined with topical steroid.81,82 In general, however,
currently available topical antivirals have been disappointing in the
treatment of stromal keratitis.78
A potential use of oral acyclovir is in herpetic stromal disease and
keratouveitis. Several investigators have reported a beneficial effect
of systemic acyclovir when combined with topical corticosteroids in
the treatment of HSV stromal keratouveitis in small, uncontrolled
studies.42,82,84 Schwab used oral acyclovir (200 mg five times per day
for 14 to 21 days) in the treatment of 20 patients with active stromal
keratitis or keratouveitis who were not adequately controlled by
topical corticosteroids and antivirals.85 All patients improved while
using oral acyclovir, but one relapsed when the drug was tapered and 3
of 7 who had discontinued acyclovir had a prompt recurrence. Nineteen
of the 20 patients had concomitant epithelial keratitis when treated.
Therefore, it is not clear whether clinical improvement was related to
the effect of acyclovir on the stromal disease or secondarily to the
effect on the epithelial infection.
Two other uncontrolled studies have found no beneficial effect of
acyclovir in the treatment of active HSV stromal keratitis.86,87
Notably, both studies involved no adjunctive use of topical
corticosteroids. Sanitato and associates found the combination of
topical and oral acyclovir ineffective in the treatment of 17 patients
with disciform edema or necrotizing stromal keratitis who were not
receiving concomitant corticosteroids.87 However, at the dosage of
acyclovir used (200 mg five times per day), the authors point out that
the subtherapeutic drug levels may not have reached the stroma.
In the HEDS study, the only controlled trial published to date of the
treatment of HSV stromal keratitis, there was no significant
beneficial effect of oral acyclovir compared to placebo in patients
already receiving concomitant topical trifluridine and
The published data are confusing because of the lack of controls in
all but one study and the different clinical situations studied.
Nevertheless, collectively they do suggest that although viral
replication is a factor in stromal keratitis the immune response is
likely to be the dominant mechanism for the disease.
HSV Interstitial Keratitis.
Topical corticosteroids are currently the principal therapeutic
modality for HSV stromal keratitis, with topical antiviral drugs used
primarily as prophylaxis against recurrent epithelial disease. In mild
stromal keratitis that does not involve the visual axis, topical
lubricants and cycloplegics are often sufficient to manage the course
of the disease comfortably. In the HEDS controlled trial, postponing
the introduction of steroid therapy for a few weeks under careful
observation did not have a detrimental effect on the outcome at 6
months.77 If progression of the keratitis is noted, characterized by
progresses with increased infiltration and stromal vascularization, if
keratitis threatens to involve the visual axis, or if the patient’s
symptoms cannot be controlled comfortably, corticosteroids should be
initiated along with topical antiviral cover.
The initial starting dose and frequency of administration of the
topical corticosteroid should be dictated by the severity of the
inflammatory process. In general, the least amount of corticosteroid
necessary to produce a clinically detectable reduction in the
inflammatory response should be used. A typical regimen may begin with
prednisolone acetate 0.125% to 1% four to eight times per day with
adjunctive topical trifluridine two to four times per day (others will
match steroid and antiviral drop for drop). The frequency of the
steroid dose can then be decreased every 1 to 3 weeks, at a rate not
to exceed a 50% reduction in the dose at any given time. An even more
gradual reduction in the dose may be necessary if there is evidence of
recurrent stromal inflammation. When the frequency of administration
decreases to 1-drop daily of a 1% prednisolone concentration, the
concentration may be reduced (to 0.125%) or switched to a less
concentrated product (fluorometholone). Reducing the steroid dose can
take weeks to months and many individuals have required an extremely
low dose (prednisolone 0.125% one to two times per week) to prevent
recurrent stromal disease. The dose of the topical antiviral agent is
also correspondingly decreased in concert with the reduction in
steroid dosage. During this period, the patient is monitored carefully
for evidence of antiviral toxicity, which may necessitate more rapid
reduction of the topical antiviral agent or substitution of oral
acyclovir for the topical preparation. A similar approach is used for
the treatment of HSV-related limbal vasculitis. Antibiotic therapy as
a prophylactic routine is unnecessary.
If the epithelium ulcerates or active epithelial disease occurs, the
steroid should be sequentially reduced or discontinued. If concomitant
inflammation must be treated, systemic prednisone may be considered
(0.5 to 1 mg/kg per day) orally for 7 to 10 days and then gradually
tapered to control the inflammation until the epithelium is healed. In
general, it is best to avoid the use of any systemic immunosuppressant
in the management of these patients.
HSV Necrotizing Stromal Keratitis.
Stromal ulceration with infiltration as seen in necrotizing stromal
keratitis is by far the most difficult form of herpes simplex stromal
disease to manage. It appears that active viral replication may be
present in some cases, but the host immune response is believed to be
the principal destructive process. Cultures may be indicated if there
is concern regarding an infection caused by bacteria, fungi, or
acanthamoeba. If the clinical findings are suggestive of active viral
replication in the corneal epithelium, topical antiviral therapy
should be initiated (topical trifluridine up to five to nine times a
day) prior to institution of topical corticosteroids. After several
days, provided the clinical picture is not indicative of an active
viral ulcer, topical corticosteroids may be added with caution at low
dosage to reduce the inflammatory response and improve patient
comfort. The patient must be monitored frequently. Viral culture of
the lesion is an option if the diagnosis is in doubt. If ulceration
progresses with topical corticosteroids, they should be slowly
withdrawn and oral prednisone (0.5 to 1 mg/kg per day) instituted.
Alternatively, oral prednisone instead of topical corticosteroids can
be used initially when the epithelium is ulcerated. Topical
cycloplegics are invariably needed. Intraocular pressure should be
monitored carefully and treated as clinically indicated. As noted
above, the HEDS controlled trial of oral acyclovir in patients with
herpes simplex stromal keratitis being treated with topical
trifluridine and steroids failed to demonstrate a significant
beneficial effect of oral acyclovir.88
Back to Top
HERPETIC EPITHELIAL KERATITIS.
Epithelial recurrence can occur in the course of predominant stromal
disease. The most serious episode is the occurrence of epithelial
herpes complicating an already active stromal keratitis. These ulcers,
in addition to exacerbating the corneal inflammation, tend to be
refractory to treatment and can worsen and assume a geographic shape
or become truly indolent.
This complication may be avoided by the use of a topical antiviral
cover. In some cases, the epithelial keratitis recurs when the
antiviral agent is discontinued because of toxicity or because the
steroid dose is thought to be at a safe low level. It is important to
recognize the epithelial lesion early and treat it energetically
before it reveals a tendency to assume a geographic shape. Antiviral
therapy should be reinstituted or the dose temporarily increased, even
at risk of epithelial toxicity. The topical steroid dose must be
temporarily reduced. In some cases, temporary total cessation of local
steroid therapy will be necessary before epithelial healing occurs. In
the presence of topical antiviral toxicity, oral antiviral therapy
with acyclovir offers a therapeutic alternative. The overall
effectiveness of oral antiviral agents in this complicated situation
remains to be explored fully.
Clinically, it may be impossible to decide whether an ulcer is
indolent or contains actively replicating virus. Indolent ulcers tend
to persist and are refractory to treatment. In predominantly
epithelial disease, they may be a sign of antiviral toxicity,
responding slowly to temporary suspension of the antiviral drug. In
the majority of cases, however, indolent ulcers reflect the underlying
stromal inflammation and will heal only when this resolves. Therefore,
it is important to recognize these ulcers as indolent and to
distinguish them from lesions resulting from active replication of the
virus. The correct approach in treating an indolent ulcer is not to
stop steroid administration but rather to increase the dose under the
umbrella of antiviral therapy. In addition to these measures, patching
or use of a bandage contact lens may also encourage healing. If all
else fails, a temporary tarsorrhaphy or conjunctival flap may be of
As discussed in detail under “Necrotizing Stromal Keratitis,” the
epithelium may break down over a dense stromal infiltrate, forming a
superficial ulcer that may slowly or rapidly deepen, producing a
descemetocele (see Fig. 20), eventually progressing to a corneal
perforation (see Fig. 26). An indolent ulcer can follow a similar
course. The uveitis in such cases is usually more severe, and frank
hypopyon may develop.
Management with topical steroids and antiviral cover has been
described previously. Close supervision is essential because these
ulcers may perforate unpredictably. Abrupt discontinuation of topical
corticosteroid therapy should be avoided because this will invite
perforation. Keratoplasty is indicated when the ulcer fails to respond
to treatment or worsens, or if perforation is imminent. Once the
cornea has perforated, the most effective management is keratoplasty,
performed as expeditiously as possible to avoid the added complication
of chronic angle closure glaucoma from anterior synechiae formation.
Elevated pressure from a trabeculitis, probably on an immunologic
basis, can accompany the acute uveitis. Intraocular pressure should be
assessed at each examination and glaucoma therapy initiated as
indicated. Intraocular pressure control can usually be achieved by
medical measures, but in advanced cases surgery may be required. In
the presence of topical steroid administration, a
corticosteroid-induced elevation in intraocular pressure is an
Bacterial and fungal superinfections may be difficult to distinguish
from the underlying disease.54 A rapid increase in the inflammatory
signs in the cornea, associated with ulceration and hypopyon
formation, must alert the ophthalmologist to this possibility.
Secondary cataract from the continued effect of the disease and
steroid administration at times complicates the management. The
approach is essentially conservative, but emergency surgery may be
necessary for an intumescent lens, and a combined keratoplasty with
lens extraction is sometimes the appropriate therapy for visual
Stromal keratouveitis related to herpes simplex characteristically
runs a protracted and unpredictable course. The intensity of the
inflammatory process may remain unchanged, whereas scarring and
neovascularization gradually progress; or it may fluctuate greatly,
perhaps showing a tendency toward resolution, only to revert shortly
thereafter to its previous course. Occasionally there is an inexorable
march toward corneal abscess formation and this may also develop from
a relatively torpid inflammatory state. With each new episode, the
cornea may become further scarred and irregularly thinned and may
eventually be traversed by a network of new vessels.
The epithelial condition is variable, reflecting in part the changes
taking place in the stroma and anterior chamber. Fine epithelial edema
may become more generalized when the endothelium is diffusely
affected. Depending on the degree of endothelial dysfunction and
secondary glaucoma, bullae may form and rupture, giving rise to
indolent ulcers. In addition, epithelial breakdown may occur over
areas of scarring and active infiltration, particularly in anesthetic
The uveitis is usually of moderate severity but may be masked by the
corneal lesions and thus be impossible to assess. Hypopyon is uncommon
in the absence of epithelial breakdown or ulcer formation. Posterior
synechiae, rubeosis iridis, and anterior lens opacities progress
insidiously. In long-standing conditions the iris becomes atrophic.
Vision is commonly reduced early unless the lesion is peripheral and
circumscribed. In mild cases, there is often a remarkable degree of
recovery. With careful management, sequelae of repeated episodes of
inflammation can be controlled and useful vision retained. All too
frequently, however, the disease pursues a relentlessly destructive
course, seriously incapacitating the patient. With repeated scarring,
visual loss is permanent.
Histopathologically, disciform keratitis is characterized by the
presence of a mixture of sensitized lymphocytes, plasma cells,
macrophages, and polymorphonuclear leukocytes.55,89 The pathogenesis
of herpes disciform keratitis is unknown. It may represent a
cell-mediated immune reaction to herpetic antigen in the stroma (1) as
a byproduct of epithelial infection, (2) by viral DNA latent within
the stroma, (3) by antigenic residue of viral invasion, or (4) by
active viral infection.90–92 Others have proposed that the main
inflammatory reaction takes place in the endothelium.
Immunocytochemical studies have demonstrated HSV antigens in corneal
endothelial cells in patients with nonulcerative disciform
keratitis.65,93,94 However, herpesvirus particles have yet to be
demonstrated in the corneal endothelium in human cases although they
have been isolated form the aqueous.58,95
Disciform keratitis is a clinical inflammatory pattern that is usually
readily distinguishable from the various manifestations of stromal
keratitis related to HSV. Classically, disciform keratitis presents as
a focal, disc-shaped area of stromal edema that can be centrally or
eccentrically located in the cornea with fine keratic precipitates on
the endothelium just in the involved area (see Figs. 24, 25, and 29).
A fine granular infiltrate may be visible throughout the corneal
stroma. In milder forms, the overlying epithelium is intact, and there
is no necrosis or vascularization. Anterior segment reaction may be
absent or mild. In more severe forms, stromal edema is more
pronounced, with folds in Descemet’s membrane (Fig. 30), focal bullous
keratopathy, and development of superficial and deep vascularization.
Rarely, bullae may rupture, with ulceration and subsequent necrosis,
and melting of the cornea with iritis. There may have been a history
of prior dendritic keratitis, but frequently such a history is
lacking, although epithelial herpes may subsequently appear. Patients
will typically complain of acute onset of blurred vision associated
with tearing, photophobia and mild, dull orbital pain.
Fig. 29 Eccentric but otherwise typical disciform keratitis.
Fig. 30 Disciform keratitis showing folds in Descemet’s membrane.
Diagnosis and Differential Diagnosis
The diagnosis of herpes simplex disciform keratitis is purely clinical
and may be in doubt in the absence of a prior history of herpetic eye
disease. Other causes of disciform keratitis include herpes zoster,
vaccinia, mumps, varicella, Acanthamoeba, Epstein–Barr virus, and
chemical keratitis, although herpes simplex remains the most common.
Disciform keratitis is presumed to be a lymphocyte-mediated
inflammatory reaction. It is highly sensitive to topical
corticosteroids (Figs. 31 and 32). If the visual axis is not involved
and steroids have not been used previously, an effort should be made
to avoid their introduction if possible. Because disciform keratitis
is believed to be an immune-mediated inflammation, the role of topical
antiviral therapy is primarily prophylactic against recurrent
epithelial disease. Cycloplegics, topical lubricants, and dark glasses
can improve patient comfort. If steroid therapy is necessary because
of involvement of the visual axis or patient discomfort, the lowest
steroid dose needed to diminish the inflammatory response should be
used (prednisolone 0.125% to 1%, two to four times per day) with
antiviral cover. Recommended therapy is similar to that outlined for
HSV interstitial keratitis.
Fig. 31 Slit photograph of disciform keratitis due to herpes before
treatment with topical steroids under IDU cover. The stroma is
thickened mainly because of edema. Note folds in Descemet’s membrane
(see Fig. 24).
Fig. 32 Disciform keratitis in the same patient as in Fig. 23. One
month after topical steroid therapy under idoxuridine (IDU) cover, the
cornea is virtually clear.
In a placebo-controlled clinical trial of 40 patients with herpes
simplex disciform stromal keratitis, topical corticosteroids
(betamethasone 0.01%, five times daily) were compared to placebo under
antiviral cover with 3% acyclovir ointment.96 The
corticosteroid-treated group healed more rapidly than the placebo
group based on an evaluation of corneal thickness and iridocyclitis.
The epithelium may break down over the region of disciform keratitis
to form an indolent ulcer. A concurrent dendritic or geographic ulcer
will sometimes behave in a similar manner. Secondary fungal or
bacterial infection is always possible in this situation and must be
considered if there is an increase in the amount of infiltrate or if a
hypopyon develops. Secondary glaucoma can persist unrecognized for a
considerable period and may result in permanent visual loss from nerve
Natural Course and Variation in Clinical Presentation
The natural course of disciform keratitis is quite unpredictable. In
some instances, the keratitis will persist unchanged; in others, it
may worsen for a considerable period before slowly resolving with a
variable, but often severe, degree of scarring (Fig. 33). The affect
on vision may be more pronounced than might be expected from the
appearance, because of the induced irregular corneal astigmatism.
Occasionally, the amount of infiltrate will insidiously or suddenly
increase (Fig. 34), and leashes of new vessels may enter the stroma to
vascularize the lesion. In such instances, the appearance and course
are indistinguishable from stromal keratitis. Regular examination of
the cornea by sclerotic scatter is useful in observing the onset of
this transition. Further recurrences of a disciform or stromal
keratouveitis are likely in these patients and, in each episode, the
residual effects can exact their toll on vision.
Fig. 33 Dense stromal scar in the visual axis after an episode of
Fig. 34 Disciform keratitis that is starting to develop local areas of
An unusual form of disciform keratitis termed linear endotheliitis has
also been described. It presents clinically in a red and painful eye,
as a line of keratic precipitates on the endothelium. The precipitates
progress across the endothelium accompanied by the development of
stromal and epithelial edema. The features are thus not dissimilar to
those seen in endothelial rejection after corneal transplantation.
There is some evidence that this condition is less responsive to
treatment than other forms of disciform keratitis.97,98
Recently it has been suggested that all such manifestations of
herpetic disease, including disciform keratitis, be referred to as
herpetic endotheliitis rather than disciform keratitis, reflecting the
position that all represent primarily disease of the endothelial
layer.99 In this chapter, as the controversy is not yet resolved, the
term disciform keratitis has been retained for the clinical
description of these cases.
The outcome of an individual attack of disciform keratitis is
favorable in terms of preservation of vision and resolution of the
inflammation. However, disciform keratitis may merge into or be
followed by stromal keratouveitis and, therefore, shares its long-term
uncertainties in prognosis.
Conjunctival flaps have been helpful in the treatment of persistent
epithelial defects or corneal ulcerations secondary to herpes simplex
keratitis unresponsive to traditional treatment modalities. They can
provide a stable epithelial surface, stop ulceration, resolve the
inflammatory process and provide patient comfort.100 Recurrent active
stromal disease has been reported following conjunctival flaps and can
result in corneal perforation.101 Patients receiving a conjunctival
flap in this clinical setting should be monitored for recurrent
herpetic disease under the flap.
Penetrating keratoplasty for purposes of visual rehabilitation in
herpes simplex-scarred corneas can be performed with reasonable
success. Several studies have shown a wide disparity in the cumulative
survival rate, ranging from 40% to 83% over 5 years.102–105
The success rate is highest when keratoplasty can be performed in a
quiet eye.106–109 The principal factors accounting for failure appear
to be recurrence of herpetic infection and graft rejection. Ficker et
al.102 recommended prompt recognition and treatment of graft
rejection, treatment of recurrences and removal of loose sutures as
important measures to prolong graft survival. High-dose topical
corticosteroids used in the postoperative period and slowly tapered
over several months have also been found to be beneficial and
associated with an improved success rate for keratoplasty.108
Regardless of the amount of steroid used postoperatively, recurrence
of herpetic disease in the graft is relatively common and occurs in
approximately 15% of eyes in a 2-year period.106,108,112
Prophylactic oral acyclovir after penetrating keratoplasty for HSV
keratitis appears to play an important role for preventing HSV
reactivation113 and reducing viral shedding during postoperative
topical corticosteroid therapy.114,115 It significantly lowered the
incidence of recurrent HSV keratitis in a rabbit keratoplasty
model.117 Schwab used prophylactic oral acyclovir (200 mg five times
per day for 14 to 21 days) in three eczematoid patients with HSV
epithelial, stromal, or uveal diseases who were undergoing
keratoplasty or cataract extraction.85 None experienced recurrent
herpetic disease while on therapy for up to 18 months. Van Rooij et
al.,118 in a placebo–controlled multicenter trial of oral acyclovir
(400 mg twice daily) after penetrating keratoplasty found a
significant reduction in recurrent herpetic eye disease in the
acyclovir treated group. Patients were treated for 6 months. The
period of observation was two years. Further studies are needed to
determine the optimum dosage and duration of treatment with acyclovir.
Penetrating keratoplasty may be necessary in cases of chronic HSV
stromal keratitis that have resulted in descemetocele formation or
corneal perforation.110 In such patients the prognosis is much poorer,
especially if the eye is inflamed at the time of surgery. With the use
of glue to seal the actual or impending perforation it is now possible
to improve the prognosis for successful transplantation by delaying
surgery until the inflammation is under better control.
PREVENTION OF RECURRENCE OF HERPETIC EYE DISEASE
Vaccination against the herpes virus remains an elusive goal for the
foreseeable future. Eradication of the virus, once established in
latent form in ganglia, is also not achievable with currently
available antiviral agents. However, perhaps one of the most
encouraging developments in the therapy of herpes simplex infections
has been the efficacy of oral acyclovir in reducing recurrences of
infection. Building on experimental evidence and uncontrolled human
studies, a placebo controlled trial of acyclovir (400 mg, twice daily)
in patients with a past history of recurrent herpetic disease, has
demonstrated that the recurrences of stromal keratitis can be reduced
by almost 50%, half over the 12-month period of treatment. The
benefit, to a lesser degree, extended to surface infections and
iritis. In patients who had a past history of stromal keratitis, the
effect was particularly striking. In a 6-month follow-up period, there
was no evidence of a rebound in infections, nor was there evidence of
a continuing protective effect.119
Conclusions about the efficacy of long-term treatment are limited by
the short duration of the trial. However, in view of the high cost of
prolonged treatment with acyclovir, it seems logical to focus
preventive therapy on patients at greatest risk of recurrent
This approach will be helped by the findings of a HEDS study that
showed that while a history of a previous episode of epithelial
keratitis is not a risk factor for further episodes of epithelial
keratitis, in patients with a history of repeated attacks of stromal
keratitis, there is a greatly increased risk of future episodes of
Back to Top
1. Brik D, Dunkel E, Pavan-Langston D: Herpetic keratitis: Persistence
of viral particles despite topical and systemic antiviral therapy.
Report of two cases and review of the literature. Arch Ophthalmol
2. Mindel A: Epidemiology. In Herpes Simplex Virus. New York:
3. Liesegang TJ: Biology and molecular aspects of herpes simplex and
varicella-zoster virus infections. Ophthalmology 99:781, 1992
4. Corey L, Spear PG: Infections with herpes simplex viruses. N Engl J
Med 314:686, 1986
5. Cook SD: Herpes simplex virus in the eye. Br J Ophthalmol 76:365, 1992
6. Shieh MT, Spear PG: Herpesvirus-induced fusion that is independent
on cell surfaced heparin sulphate or soluble heparin. J Virol 68:1224,
7. Darlington RW, Moss LH: III: Herpesvirus envelopment. J Virol 2:48, 1968
8. Rosenwasser GO, Greene WH: Simultaneous herpes simplex types 1 and
2 keratitis in acquired immunodeficiency syndrome. Am J Ophthalmol
9. Kibrick S, Goodling GW: Pathogenesis of infection with herpes
simplex virus with special reference to nervous tissue in slow,
latent, and temporate virus infections. NINDB Monogr 2:143, 1965
10. Nahmias et al: Sero-epidemiological and sociological patterns of
herpes simplex virus infections in the world. Scand J Infect Dis
69(supp) :19, 1990
11. Kaufman HE, Brown DC, Ellison ED: Herpesvirus in the lacrimal
gland, conjunctiva and corneal of man—a chronic infection. Am J
Ophthalmol 65:32, 1968
12. Kaufman HE: In vivo studies with antiviral agents. Ann N Y Acad
Sci. 130:168, 1965
13. Lindgren KM, Douglas GR Jr, Couch RB: Significance of Herpesvirus
hominis in respiratory secretions of man. N Engl J Med 278:517, 1968
14. Rawls WE, Campione-Picardo J: Epidemiology of herpes simplex virus
type 1 and type 2 infections. In Nahmias AJ, Dowdle WR, Schinazi RF,
eds: The Human Herpesviruses. An Interdisciplinary Perspective. New
York: Elsevier, 1981:139
15. Arffa RC: Viral diseases. In Grayson’s Diseases of the Cornea, 3rd
ed. St. Louis: CV Mosby, 1991:238–294
16. Dhaliwal DK, Romanowski EG, Yates KA, et al: Experimental
laser-assisted in situ keratomileusis induces the reactivation of
latent herpes simplex virus. Am J Ophthalmol 131:506, 2001
17. Perry HD, Doshi SJ, Donnenfelf ED, et al: Herpes simplex
reactivation following laser in situ keratomileusis and subsequent
corneal perforation. CLAO J 28:69, 2002
18. Liesegang TJ, Melton J, Daly PF, et al: Epidemiology of ocular
herpes simplex.Incidence in Rochester, MN, 1950 through 1982. Arch
Ophthalmol 107:1155, 1989
19. Gordon YJ: Pathogenesis and latency of herpes simplex virus type 1
(HSV-1): an ophthalmologist’s view of the eye as a model for the study
of the virus-host relationship. Adv Exp Med Biol 278:205, 1990
20. Goodpasture EW: Herpetic infection with special reference to
involvement of the nervous system. Medicine 8:223, 1929
21. Roizman B: In Pollard M (ed): Perspectives in Virology. New York:
Harper & Row, 1965
22. Fenner F: The Biology of Animal Viruses. New York: Academic Press, 1968
23. Stevens JG, Cook ML: Latent herpes simplex virus in spinal ganglia
of mice. Science 73:740, 1971
24. Cook ML, Stevens JG: Pathogenesis of herpetic neuritis and
gangliomites in mice: evidence of intra-axonal transport of infection.
Infect Immunol 7:272–288, 1973
25. Pepose JS, Lieb DA, Stuart M, Easty D: Ocular infections and
immunity. In Herpes Simplex Virus Diseases: Anterior Segment of the
Eye. St. Louis: Mosby, 1995:908–909
26. Fraser NW, Spivack JG, Wroblewska Z, et al: A review of the
molecular mechanism of HSV-1 latency. Curr Eye Res 9:1, 1990
27. Cook D, Hill JH: Herpes simplex virus: molecular biology and the
possibility of corneal latency. Surv Ophthalmol 36:140, 1991
28. Rong BL, Kenyon KR, Bean KM, et al: Detection of the herpes
simplex virus genome in the human cornea. Ophthalmology 95(suppl):159,
29. Abghare SZ, Stulting RD: Recovery of herpes simplex virus from
ocular tissues of latently infected inbred mice. Invest Ophthalmol Vis
Sci 29:239, 1988
30. Cook SD, Batra SK, Brown SM: Recovery of herpes simplex virus from
corneas of experimentally infected rabbits. J Gen Virol 68:2013, 1987
31. Gerdes JC, Smith DS: Recurrence phenotypes and establishment of
latency following rabbit keratitis produced by multiple herpes simplex
virus strains. J Gen Virol 64:2441, 1983
32. Centifanto-Fitzgerald YM, Fenger T, Kaufman HE: Virus proteins in
herpetic keratitis. Exp Eye Res 35:425, 1982
33. Smeraglia R, Varnell ED, Centifanto UM, et al: The role of herpes
simplex virus secreted glycoproteins in herpetic keratitis. Exp Eye
Res 35:443, 1982
34. Kaufman HE, Varnell ED, Centifanto YM, et al: Effect of the herpes
simplex virus genome on the response of infection to corticosteroids.
Am J Ophthalmol 100:114, 1985
35. Jones BR: The management of ocular herpes. Trans Ophthalmol Soc UK
36. Coster DJ, Jones BR, Falcon MG: Role of debridement in the
treatment of herpetic keratitis. Trans Ophthalmol Soc UK 97:314, 1977
37. Pavan-Langston D, Dohlman CH, Geary P, et al: Intraocular
penetration of AraA and IDU—Therapeutic implication in clinical
herpetic uveitis. Trans Am Acad Ophthalmol Otolaryngol 77:455, 1973
38. Laibson PR, Kratchmer JH: Controlled comparison of adenine
arabinoside and idoxuridine therapy of human superficial dendritic
keratitis. In Pavan-Langston D, Buchanan RA, Alford GA, eds. Adenine
Arabinoside: An Anti-Viral Agent. New York: Raven Press, 1976:323
39. Teich SA, Cheung TW, Friedman AH: Systemic antiviral drugs used in
ophthalmology. Surv Ophthalmol 37:19, 1992
40. LaLau C, Oosterhuis JA, Versteeg J, et al: Acyclovir and
trifluorothymidine in herpetic keratitis—a multicenter trial. Br J
Ophthalmol 66:506, 1982
41. LaLau C, Oosterhuis JA, Versteeg J, et al: Multicenter trial of
acyclovir and trifluorothymidine in herpetic keratitis. Am J Med
42. Bialasiewicz AA, Jahn GJ: Systemische acyclovir—Therapie bei rezi
divierender durch herpes simplex virus bedingter keratouveitis. Klin
Monatsbl Augenheilkd 185:539, 1984
43. Collum LMT, Akhtar J, McGettrick P: Oral acyclovir in herpetic
keratitis. Trans Ophthalmol Soc UK 104:629, 1985
44. Collum LMT, McGettrick P, Akhtar J, et al: Oral acyclovir
(Zovirax) in herpes simplex dendritic corneal ulceration. Br J
Ophthalmol 70:435, 1986
45. Wilhelmus KR: The treatment of herpes simplex virus epithelial
keratitis. Trans Am Ophthalmol Soc 98:505, 2000
46. Beutner KR: Valcyclovir: a review of its antiviral activity,
pharmacokinetic properties and clinical efficacy. Antiviral Res
47. Bohigan G, Dawson C, Coleman V: Retrobulbar administration of
steroids in herpes simplex uveitis. Arch Ophthalmol 85:320, 1971
48. Liebowitz HM, Frangie JP: Inflammation of the cornea and its
management. In Leibowitz H, Waring GO, eds. Corneal Disorders:
Clinical Diagnosis and Management. 2nd ed. Philadelphia: WB Saunders,
49. Margolis TP, Ostler HB: Treatment of ocular disease in eczema
herpeticum. Am J Ophthalmol 110:274, 1990
50. Tullo AB, Easty DL, Shimeld C, et al: Isolation of herpes simplex
virus from corneal discs of patients with chronic stromal keratitis.
Trans Ophthalmol Soc UK 104:159, 1985
51. Dresner AJ, Seamans ML: Evidence of the safety and efficacy of
adenine arabinoside in the treatment of herpes simplex epithelial
keratitis. In Pavan-Langston D, Buchanan RA, Alford CA Jr, eds.
Adenine Arabinoside: An Antiviral Agent. New York: Raven Press,
52. Norn MS: Dendritic (herpetic) keratitis: IV. Follow-up examination
of corneal sensitivity. Acta Ophthalmol 48:383, 1970
53. Beiji B, Algawi K, Foley-Nolan A, et al: Herpes simplex in
children. Br J Ophthalmol 78:458, 1994
54. Hogan MJ, Kimura SJ, Thygeson P: Pathology of herpes simplex
kerato-iritis. Trans Am Ophthalmol Soc 61:75, 1963
55. Dawson C, Togni B, Moore TE: Structural changes in chronic
herpetic keratitis studied by light electron microscopy. Arch
Ophthalmol 79:740, 1968
56. Easty DL, Shimeld C, Claove CMP, et al: Herpes simplex virus
isolation in chronic stromal keratitis: Human and laboratory studies.
Curr Eye Res 6:69, 1987
57. Font RL: Chronic ulcerative keratitis caused by herpes simplex
virus. Electron microscopic confirmation in paraffin-embedded tissue.
Arch Ophthalmol 90:382, 1973
58. Pepose JS: Herpes simplex keratitis: Role of viral infection
versus immune response. Surv Ophthalmol 35:345, 1991
59. Pavan-Langston D: Herpetic infections. In Smolin G, Thoft RA, eds.
The Cornea. Scientific Foundation and Clinical Practice, 3rd ed.
Boston: Little, Brown & Co, 1994:169–215
60. Meyers R: Immunology of herpes simplex virus infection. Int
Ophthalmol Clin 15:37, 1975
61. Meyers-Elliott R, Elliott JH, Maxwell WA, et al: HLA antigens in
herpes stromal keratitis. Am J Ophthalmol 89:54, 1980
62. Meyers-Elliott R, Pettit T, Maxwell W: Viral antigens in the
immune rings of herpes simplex stromal keratitis. Arch Ophthalmol
63. Wilhelmus KR, Coster DJ, Donovan HC, et al: Prognostic indicators
of herpetic keratitis. Analysis of a five-year observation period
after corneal ulceration. Arch Ophthalmol 99:1578, 1981
64. Holbach LM, Font RL, Baehr W, et al: HSV antigens and HSV DNA in
avascular and vascularized lesions of human herpes simplex keratitis.
Curr Eye Res 10(suppl):63, 1991
65. Holbach L, Font R, Naumann G: Herpes simplex stromal and
endothelial keratitis. Ophthalmology 97:722, 1990
66. Wilhelmus KR, Falcon MG, Jones BR: Bilateral herpetic keratitis.
Br J Ophthalmol 65:385, 1981
67. Dunkel EC, Pavan-Langston D, Fitzpatrick K, et al: Rapid detection
of herpes simplex virus (HSV) antigen in human ocular infections. Curr
Eye Res 7:661, 1988
68. Gebhardt BM, Reidy J, Kaufman HE: An affinity membrane test for
superficial corneal herpes. Am J Ophthalmol 105:686, 1988
69. Dawson CR, Jones DB, Kaufman HE, et al: Design and organization of
the herpetic eye disease study (HEDS). Curr Eye Res 10:105, 1991
70. Dawson CR: The Herpetic Eye Disease Study. Arch Ophthalmol 108:191, 1990
71. Pavan-Langston D, Abelson MB: The role of steroids in ocular
herpes. In Boruchoff SA, Hutchinson BR, Lessell S, eds. Controversies
in Ophthalmology. Philadelphia: WB Saunders, 1970:438
72. Pavan-Langston D, Abelson MB: Glucocorticoid therapy in ocular
herpes simplex. II. Advantages. Surv Ophthalmol 23:43, 1978
73. Ostler HB: Glucocorticoid therapy in ocular herpes simplex. I.
Limitations. Surv Ophthalmol 23:35, 1978
74. Thygeson P: The unfavorable role of corticosteroids in herpetic
keratitis. In Boruchoff SA, Hutchinson BR, Lessell S, eds.
Controversies in Ophthalmology. Philadelphia: WB Saunders, 1970:450
75. Kibrick S, Takahashi GH, Leibowitz HM, et al: Local corticosteroid
therapy and reactivation of herpetic keratitis. Arch Ophthalmol
76. Easterbrook M, Wilkie J, Coleman V, et al: The effect of topical
corticosteroid on the susceptibility of immune animals to
reinoculation with herpes simplex. Invest Ophthalmol Vis Sci 2:181,
77. Wilhelmus KR, Gee L, Hauck WW, et al: The Herpetic eye disease
study. A controlled trial of topical corticosteroids for herpes
simplex keratitis. Ophthalmology 101:1883, 1994
78. Wilhelmus KR: Diagnosis and management of herpes simplex stromal
keratitis. Cornea 6:286, 1987
79. Kaufman HE, Martola EL, Dohlman C: Use of 5–iodo–2’deoxyuridine
(IDU) in treatment of herpes simplex keratitis. Arch Ophthalmol
80. Pavan-Langston D, Nelson DJ: Intraocular penetration of
trifluridine. Am J Ophthalmol 87:814, 1979
81. Collum LMT, Logan P, Ravenscroft T: Acyclovir (Zovirax) in
herpetic disciform keratitis. Br J Ophthalmol 67:115, 1983
82. VanGanswijk R, Oosterhuis JA, Swart-Van Den Berg M, et al:
Acyclovir treatment in stromal herpetic keratitis. Doc Ophthalmol
83. Colin J, Malet F, Chastel C, et al: Acyclovir in herpetic anterior
uveitis. Ann Ophthalmol 23:28, 1991
84. Porter SM, Patterson A, Kho P: A comparison of local and systemic
acyclovir in the management of herpetic disciform keratitis. Br J
Ophthalmol 74:283, 1990
85. Schwab IR: Oral acyclovir in the management of herpes simplex
ocular infections. Ophthalmology 95:423, 1988
86. Sundmacher R: Oral acyclovir—Therapie virologisch nachgewiesener
intraokularer herpes simplex virus infektionen. Klin Monatsbl
Augenheilkd 183:246, 1983
87. Sanitato JJ, Asbell PA, Varnell ED, et al: Acyclovir in the
treatment of herpetic stromal disease. Am J Ophthalmol 98:537, 1984
88. Barron BA, Gee L, Hauck WW, Kurinji N, et al: Herpetic eye disease
study. A controlled trial of oral acyclovir for herpes simplex stromal
keratitis. Ophthalmology 101:1871, 1994
89. Lass J, Pavan-Langston D, Berman M: Treatment of experimental
herpetic interstitial keratitis with medroxyprogesterone. Arch
Ophthalmol 98:520, 1980
90. Gordon YJ: Herpetic stromal keratitis: A new molecular model and
its clinical correlation. In Cavanagh HD, ed. The Cornea. Transactions
of the World Congress on the Cornea III. New York: Raven Press,
91. Metcalf JF, Reichert RW: Histological and electron microscopic
studies of experimental herpetic keratitis in the rabbit. Invest
Ophthalmol Vis Sci 18:1123, 1979
92. Oh JO: Endothelial lesions of rabbit cornea produced by herpes
simplex virus. Invest Ophthalmol 9:196, 1970
93. Sundmacher R, Neumann-Haefelin D: Herpes simplex virus-positive
and virus-negative keratouveitis. In Silverman AM, O’Connor GR, eds.
Immunology and Immunopathology of the Eye. New York: Masson, 1979:225
94. Vannas A, Ahoner R, Makitie J: Corneal endothelium in herpetic
keratouveitis. Arch Ophthalmol 101:913, 1983
95. Sundmacher R, Neumann-Haefelin D: Herpes simplex virus-positive
and negative keratouveitis. In Silverstein AM, O’Connor R, eds.
Immunology and Immunopathology of the Eye. New York: Masson,
96. Power WJ, Hillery MP, Benedict-Smith A, et al: Acyclovir ointment
plus topical betamethasone or placebo in first episode disciform
keratitis. Br J Ophthalmol 76:711, 1992
97. Olsen TW, Hardten DR, Meiusi RD, et al: Linear endotheliitis. Am J
Ophthalmol 117:468, 1994
98. Vogel A, Schneider H, Loffler KU: Histopathology of herpetic
corneal endotheliitis. Klinische Monatsblatter for Augenheilkunde
99. Holland EJ, Schwartz GS: Classification of herpes simplex
keratitis. Cornea 18: 144, 1999
100. Brown DD, McCulley JP, Bowman RW, et al: The use of conjunctival
flaps in the treatment of herpes keratouveitis. Cornea 11:44, 1992
101. Lesher MP, Lohman LE, Yeakley W, et al: Recurrence of herpetic
stromal keratitis after a conjunctival flap surgical procedure. Am J
Ophthalmol 114:231, 1992
102. Ficker LA, Kirkness CM, Rice NS, et al: The changing management
and improved prognosis for corneal grafting in herpes simplex
keratitis. Ophthalmology 96:1587, 1989
103. Ficker LA, Kirkness CM, Rice NS, et al: Long-term prognosis for
corneal grafting in herpes simplex keratitis. Eye 2(Part 4) :400, 1988
104. Larkin DFP: Corneal transplantation for herpes simplex keratitis.
Br J Ophthalmol 82:107, 1998
105. Halberstadt M, Machens M, Gahlenbek KH, et al: The outcome of
corneal grafting in patients with stromal keratitis of herpetic and
non-herpetic origin. Br J Ophthalmol 86:646, 2002
106. Cohen E, Laibson P, Arentsen J: Corneal transplantation for
herpes simplex keratitis. Am J Ophthalmol 95:645, 1983
107. Foster CS, Duncan J: Penetrating keratoplasty for herpes simplex
keratitis. Am J Ophthalmol 92:336, 1981
108. Langston R, Pavan-Langston D, Dohlman CH: Penetrating
keratoplasty for herpetic keratitis. Trans Am Acad Ophthalmol
Otolaryngol 79:577, 1975
109. Polack FM, Kaufman HE: Penetrating keratoplasty in herpetic
keratitis. Am J Ophthalmol 73:908, 1972
110. Patten JT, Cavanagh HD, Pavan-Langston D: Penetrating
keratoplasty in acute herpetic corneal perforations. Ann Ophthalmol
111. Pfister RR, Richards JS, Dohlman CH: Recurrence of herpetic
keratitis in corneal grafts. Am J Ophthalmol 73:192, 1972
112. Pfister RR, Richards JS, Dohlman CH: Recurrence of herpetic
keratitis in corneal grafts. Am J Ophthalmol 73:192, 1972
113. Beyer CF, Hill JM, Kaufman HE: Antivirals and interferons.
Ophthalmol Clin North Am 2:51, 1989
114. Falcon MG: Rational acyclovir therapy in herpetic eye disease. Br
J Ophthalmol 71:102, 1987
115. Green MT, Dunkel EC, Morris BL: Quantification of herpes simplex
virus type 1 shed in pre-ocular tear film of rabbits treated with
acyclovir. Antimicrob Agents Chemother 20:580, 1981
116. Barney NP, Foster CS: A prospective randomized trial of oral
acyclovir after penetrating keratoplasty for herpes simplex keratitis.
Cornea 13:232, 1994
117. Beyer CF, Arens MQ, Hill GA, et al: Oral acyclovir reduces the
incidence of recurrent herpes simplex in rabbits after penetrating
keratoplasty. Arch Ophthalmol 107:1200, 1989
118. Van Rooij J, Rijneveld WJ, Remeijer L, et al: Effect of oral
acyclovir after penetrating keratoplasty for herpetic keratitis. A
placebo controlled multicentered trial. Ophthalmology 110:1916, 2003
119. Herpetic Eye Disease Group: Acyclovir for the prevention of
recurrent herpes simplex virus eye disease. N Engl J Med 339:300, 1998
120. Lairson DR, Begley CE, Reynolds TF, et al: Prevention of herpes
simplex virus eye disease: a cost effectiveness analysis. Arch
Ophthalmol 122:108, 2003
121. Herpes Eye Disease Study Group: Predictors of recurrent herpes
simplex keratitis. Cornea 20:123, 2001
Acyclovir is a nucleoside analogue and antiviral agent used in therapy of herpes and varicella-zoster virus infections. Acyclovir has not been associated with clinically apparent liver injury.
Acyclovir (ay sye’ kloe vir) is an acyclic purine nucleoside analogue (acycloguanosine) which has antiviral activity against many herpes viruses, including herpes simplex 1 and 2, cytomegalovirus, Ebstein-Barr virus and varicella-zoster. Acyclovir is phosphorylated intracellularly by viral kinases and the resultant triphosphate competes with guanosine for incorporation into viral DNA blocking viral DNA polymerase activity. Acyclovir is indicated for therapy of localized as well as disseminated herpes simplex infections, both type 1 and 2. It is also used for varicella-zoster infections (chickenpox and shingles). Acyclovir was approved for use in herpes virus infections in the United States in 1982 and is still widely used in treatment and prophylaxis of genital and mucocutaneous herpes simplex infection with almost 5 million prescriptions filled yearly. Acyclovir is available as capsules of 200 mg, tablets of 400 and 800 mg, oral suspensions, creams, ointments, and parenteral preparations in several generic forms as well as under the brand name of Zovirax. The typical recommended oral dose in adults for genital or oral herpes simplex is 200 to 800 mg three to five times daily for 5 to 10 days; the usual prophylactic dose is 400 mg twice daily. The typical intravenous doses for severe infections is 5 to 10 mg/kg every 8 hours for 5 to 10 days. Side effects are uncommon with oral formulations but can include myalgias, rash, temors, lethargy and confusion. Rare side effects include bone marrow toxicity and Stevens Johnson syndrome.
Despite widespread use, there is little evidence that acyclovir when given orally causes significant liver injury. Serum enzyme levels generally do not change during oral acyclovir therapy. High dose intravenous administration of acyclovir is associated with renal dysfunction and thrombocytopenia and occasionally with transient mild-to-moderate elevations in serum ALT levels, which have been asymptomatic and self-limited. There have been no particularly convincing cases of acute liver injury with jaundice reported in patients receiving acyclovir. Some degree of liver injury and even jaundice can occur during the course of herpes simplex or varicella zoster infection, and these complications could be mistaken for drug-induced liver injury.
Mechanism of Injury
Acyclovir is metabolized intracellularly in viral infected cells and is minimally metabolized by the liver. Acyclovir is excreted largely unchanged by the kidneys, perhaps accounting for the absence or rarity of hepatic injury.
Drug Class: Antiviral Agents
Other acyclic nucleosidex used as antiherpes virus agents: Cidofovir, Famciclovir, Ganciclovir,Valganciclovir, Valacyclovir
REPRESENTATIVE TRADE NAMESAcyclovir – Zovirax®
REPRESENTATIVE TRADE NAMESAcyclovir – Zovirax®
DRUG CLASSAntiviral Agents
DRUG CLASSAntiviral Agents
Product labeling at DailyMed, National Library of Medicine, NIH
|DRUG||CAS REGISTRY NUMBER||MOLECULAR FORMULA||STRUCTURE|
References Last Updated: 28 May 2014
Zimmerman HJ. Antiviral agents. In, Zimmerman HJ. Hepatotoxicity: the adverse effects of drugs and other chemicals on the liver. 2nd ed. Philadelphia: Lippincott, 1999, pp. 621-3. (Expert review of antiviral agents and liver injury published in 1999: mentions that acyclovir has not caused “overt hepatic injury”).
Nunez M. Herpesviridae treatment. Hepatic toxicity of antiviral agents. In, Kaplowitz N, DeLeve LD, eds. Drug-induced liver disease. 3rd ed. Amsterdam: Elsevier, 2013, pp. 512-3. (Review of hepatotoxicity of antiviral agents; mentions that there have been post-marketing reports of ALT elevaitons, hepatitis and jaundice due to acyclovir).
Acosta EP, Flexner C. Antiviral agents(nonretroviral). In, Brunton LL, Chabner BA, Knollman BC, eds. Goodman & Gilman’s the pharmacological basis of therapeutics. 12th ed. New York: McGraw-Hill, 2011, pp. 1593-1622. (Textbook of pharmacology and therapeutics).
Straus SE, Takiff HE, Seidlin M, Bachrach S, Lininger L, Di Giovanna JJ, Western KA, et al. Suppression of frequently recurring genital herpes. A placebo-controlled double-blind trial of oral acyclovir. N Engl J Med 1984; 310: 1545-50. PubMed Citation (Placebo controlled trial of acyclovir in 35 patients with recurrent genital herpes; 1 patient had ALT elevations [88 U/L] that resolved even during continuation of therapy).
O’Brien JJ, Campoli-Richards DM. Acyclovir. An updated review of its antiviral activity, pharmacokinetic properties and therapeutic efficacy. Drugs 1989; 37: 233-309. PubMed Citation (Review of antiviral activity, mechanism of action, pharmacokinetics, clinical efficacy and adverse effects of acyclovir; side effects are usually mild but high doses intravenously are associated with nausea, vomiting, lightheadedness, neurologic symptoms, and renal dysfunction; no mention of hepatotoxicity or ALT elevations).
Tilson HH, Engle CR, Andrews EB. Safety of acyclovir: a summary of the first 10 years experience. J Med Virol 1993; Suppl 1: 67-73. PubMed Citation (Between 10-15 million persons have been treated with acyclovir, but the sponsor received only 923 adverse event reports: 129 considered serious, but no specific mention of hepatotoxicity or ALT elevations).
Styrt B, Freiman JP. Hepatotoxicity of antiviral agents. Gastroenterol Clin North Am 1995; 24: 839-52. PubMed Citation (Review of liver toxicity of antiviral agents; acyclovir may be associated with minor liver test elevations, but there have been no published reports of it causing clinically apparent liver injury).
Ormrod D, Scott LJ, Perrry CM. Valaciclovir: a review of its long term utility in the management of genital herpes simplex virus and cytomegalovirus infections. Drugs 2000; 59: 839-63. PubMed Citation (Review of efficacy and safety of long-term valaciclovir use; rates of side effects are similar in frequency to those in patients on placebo; a single case of hepatitis due to valaciclovir has been reported in abstract form).
Bodsworth NJ, Crooks RJ, Borelli S, Vejlsgaard G, Paavonen J, Worm A-M, Uexkull N, et al., International Valaciclovir HSV Study Group. Genitourin Med 1997; 73: 110-6. PubMed Citation (999 patients randomized at 48 sites to acyclovir or valacyclovir for 5 days for recurrent HSV infection; equivalent efficacy and “no clinically important changes from screening in any clinical chemistry variable”).
Renkes P, Trechot P, Blain H. Valaciclovir-induced hepatitis. Acta Clin Belg 1999; 54: 17-8.PubMed Citation (71 year old woman with shingles developed abdominal pain 7 days after starting valacyclovir and 3 g/day of acetaminophen [bilirubin 3.3 mg/dL, ALT 376 U/L, Alk P 246 U/L], resolving within 2 weeks of stopping).
Simpson D, Lyseng-Williamson KA. Famciclovir: a review of its use in herpes zoster and genital and orolabial herpes. Drugs 2006; 66: 2397-416. PubMed Citation (Review of famciclovir, oral prodrug of penciclovir, used in herpes zoster and simplex virus infections for limited period as therapy and extended use for suppression; in suppression studies, ALT elevations above twice ULN occurred in 3.2% of famciclovir vs 1.5% of placebo recipients; no hepatic serious adverse events reported).
Drugs for non-HIV viral infections. Treat Guidel Med Lett 2007; 5: 59-70. PubMed Citation (Review of status of non-antiretroviral antiviral agents for prevention and treatment of herpes, varicella-zoster, cytomegalovirus, influenza A and B, and hepatitis B and C; no mention of liver-related side effects for acyclovir).
Chalasani N, Fontana RJ, Bonkovsky HL, Watkins PB, Davern T, Serrano J, Yang H, et al.; Drug Induced Liver Injury Network(DILIN). Causes, clinical features, and outcomes from a prospective study of drug-induced liver injury in the United States. Gastroenterology 2008; 135: 1924-34. PubMed Citation (Among 300 cases of drug-induced liver disease in the US collected between 2004 and 2008, 8 were attributed to antiviral agents including one attributed to valacyclovir but none to acyclovir).
Reuben A, Koch DG, Lee WM; Acute Liver Failure Study Group. Drug-induced acute liver failure: results of a U.S. multicenter, prospective study. Hepatology 2010; 52: 2065-76. PubMed Citation (Among 1198 patients with acute liver failure enrolled in a U.S. prospective study between 1998 and 2007, 133 were attributed to drug-induced liver injury, 4 of which were due to antiretroviral agents, but none were attributed to an antiherpes virus agent).
Antiviral drugs. Treat Guidel Med Lett 2013; 11 (127): 19-30. PubMed Citation (Review of safety and efficacy of acyclovir treatment and prophylaxis against varicella and herpes zoster infections: mentions oral acyclovir has been linked to cases of Stevens Johnson syndrome, but does not specifically mention liver injury).
Gopal MG, Shannoma, Kumar B C S, M R, A S N, Manjunath NC. A comparative study to evaluate the efficacy and safety of acyclovir and famciclovir in the management of herpes zoster. J Clin Diagn Res 2013; 7: 2904-7. PubMed Citation. (Among 100 patients with herpes zoster treated for 7 days with either acyclovir or famciclovir, there was no “clincally significant difference” in serum biochemistry tests between the two groups).
Pugi A, Bonaiuti R, Maggini V, Moschini M, Tuccori M, Leone R, Rossi M, et al. Safety profile of antiviral medications: a pharmacovigilance study using the Italian spontaneous-reporting database. Am J Health Syst Pharm 2013; 70: 1039-46. PubMed Citation. (Analysis of adverse drug events reported spontaneous during a 22 year period in Italy identified 863 reports involving antivirals but liver related events were not among the 20 most reported reactions).
Tachibana T, Nozaki A, Enaka M, Yamamoto E, Kawasaki R, Koharazawa H, Hagihara M, et al. Drug-induced liver injury after allogeneic bone marrow transplantation. Int J Hematol 2013; 98: 499-503. PubMed Citation. (23 year old woman with acute leukemia and allogeneic bone marrow transplant developed jaundice [bilirubin 14.0 mg/dL, ALT 1379 U/L, Alk P 1389 U/L] 180 days after transplant, while receiving oral acyclovir (for 104 days), voriconazole (60 days) and rebamipide (97 days), as well as tacrolimus and lansoprazole; a positive lymphocyte stimulation test was found for acyclovir but not the other agents).
Kendrick JG, Ensom MH, Steer A, White CT, Kwan E, Carr RR. Standard-dose versus high-dose acyclovir in children treated empirically for encephalitis: a retrospective cohort study of its use and safety. Paediatr Drugs 2014; 16: 229-34. PubMed Citation. (Among 61 children with herpes simplex encephalitis treated with high or standard doses of intravenous acyclovir, “elevated liver enzymes” occurred in 4 [6.5%], 2 at each dose level; no mention of outcome, symptoms or jaundice).