Customize your JAMA Network experience by selecting one or more topics from the list below.
Acute retinal necrosis (ARN) is a viral retinitis occurring predominantly
in immunocompetent persons. Clinical diagnosis is based on the presence of
1 or more peripheral foci of retinal necrosis that show rapid circumferential
progression and often result in early retinal detachment. Characteristically,
the active phase is associated with anterior chamber inflammation and vitritis.
We report a series of cases in which recurrent anterior uveitis (AU) occurred
in the absence of recurrence of posterior segment inflammatory activity or
retinitis, after the initial episode of healed ARN. No patient was infected
with human immunodeficiency virus. All patients responded to treatment with
topical corticosteroids and mydriatics, without recourse to further antiviral
treatment; they did not commonly experience a subsequent attack of ARN. To
our knowledge, this is the first report of isolated recurrent AU following
Acute retinal necrosis is a fulminant viral retinitis first described
in 1971 by Urayama et al1 in Japan, where
it was initially known as Kirisawa uveitis. The American Uveitis Society's
diagnostic criteria for ARN include the following: 1 or more foci of retinal
necrosis with discrete borders commonly in the peripheral retina that show
rapid circumferential progression, dense vitritis, AU, and occlusive vasculitis.2 Other features that may be associated but are not
required for the diagnosis include optic neuritis or papillitis with subsequent
optic atrophy, early retinal detachment, and scleritis.2
The usual causative organism is varicella-zoster virus, but herpes simplex
virus (HSV) types 1 and 2 and cytomegalovirus have been implicated.3 The diagnosis may be confirmed by the detection
of viral DNA by polymerase chain reaction of a vitreous biopsy specimen or
by the identification of intraocular antibody.
Acute retinal necrosis may occasionally follow a mild self-limiting
course, but in general requires aggressive treatment with high-dosage intravenous
acyclovir or famciclovir, which may reduce the incidence of the development
of bilateral disease.4 High-dosage oral
prednisolone may be required, particularly if optic disc swelling is present.
To reduce the risk of retinal detachment, prophylactic laser demarcation may
be performed around necrotic foci. Despite such treatment, the prognosis for
maintaining vision is generally poor. Visual loss results from early retinal
detachment, papillitis with secondary optic atrophy, or extensive necrosis
involving the macula.
Recurrence of ARN is documented, and second-eye involvement may occur
several years later.5 However, the recurrence
of AU in the absence of active ARN, vitritis, or retinitis has not been reported,
to our knowledge.
Report of Cases
A retrospective case-note analysis of patients seen by the Moorfields
Eye Hospital's uveitis service, London, England, was undertaken. Between 1992
and 1998, 19 patients with classic appearance of ARN were identified. Ten
were women and 9 were men, with a mean age of 44 years (range, 18-79 years).
No patients were infected with human immunodeficiency virus. The causative
organism was determined in 5 patients by polymerase chain reaction of a vitreous
biopsy specimen. Of these, 1 was due to HSV-1; 2, to HSV-2; and 2, to varicella-zoster
virus. Four patients had bilateral disease.
Nine patients developed retinal detachment (2 bilateral), and all but
1 underwent retinal reattachment surgery. Six patients had optic disc swelling;
5 of these went on to develop a retinal detachment.
All were treated with a standard regimen of 7 days of intravenous acyclovir
(10 mg/kg of body weight 3 times a day) followed by either 800 mg of oral
acyclovir 5 times daily or 1 g of oral valcyclovir 3 times daily for 6 to
12 weeks, in conjunction with oral prednisolone at a starting dose of 40 to
80 mg. Patients with severe intraocular inflammation received cyclosporine
in addition to corticosteroids. In all cases, the initial inflammatory process
resolved within 3 to 6 months of treatment.
Of the 19 patients, 7 developed AU in the absence of vitritis or new
foci of retinitis after complete resolution of the initial inflammation. Two
patients had a history of herpetic recurrent AU before the development of
ARN, but the AU had been dormant for several years. Patient 5 had a history
of uveitis as a child between age 5 and 14 years and an isolated episode during
pregnancy at age 22. She had not had symptomatic AU for 33 years before the
development of ARN. Patient 6 had had recurrent AU following dendritic ulceration
between age 1 and 30 years. He had not had symptomatic AU for 13 years before
the episode of ARN. Another patient had recurrent AU after ARN associated
with recurrent severe scleritis.
One patient developed a second episode of ARN in the same eye after
the initial episode had healed and following a subsequent episode of the isolated
AU. However, this second episode of ARN occurred 4 months after complete resolution
of isolated AU. This was followed by another attack of isolated AU 3 months
after the second episode of ARN had completely resolved, lasting 1 month.
Five months later, she had another episode of isolated AU in the absence of
vitritis or retinitis, but this was accompanied by cystoid macula edema. Following
an orbital floor injection of methylprednisolone acetate, the AU and deficiency
resolved within 4 months. The cystoid macula edema recurred 1 month later,
but with further treatment resolved completely within another 6 months.
The AU seen after resolution of ARN was generally consistent with that
seen initially, with similar or less severe inflammation. Five of the 7 patients
had nongranulomatous AU in the initial attack and in all episodes of recurrent
AU. Two patients with granulomatous AU in the initial attack went on to have
nongranulomatous recurrent AU. Only 1 patient had 3+ cells and flare during
the initial episode of ARN. The isolated AU that followed was much less severe
(1+ cells and 1+ flare). Three patients had 2+ cells and flare during the
initial ARN, 1 subsequently developing AU with 2+ cells and 2+ flare but the
other 2 having a milder (1+ cells and flare) AU after resolution. The remaining
3 patients had mild (1+ cells and flare) AU associated with the initial ARN
and similarly mild (1+ cells and flare) subsequent isolated AU. One patient
had fine keratoprecipitate at the time of the ARN but not in subsequent attacks
of AU, and another patient developed fine keratoprecipitate with the subsequent
AU. With the recurrent AU, one patient developed mild posterior synechiae,
but another developed extensive posterior synechiae followed by iris bombé
and required laser peripheral iridectomy. The 2 patients with prior herpetic
anterior segment involvement had segmental iris atrophy.
All responded to treatment with topical corticosteroids and mydriatics
without additional antiviral treatment, and the AU did not herald a further
attack of ARN. Only 1 patient with AU after healed ARN had previously had
a polymerase chain reaction done on a vitreous biopsy specimen; this yielded
HSV-1 DNA. This was a patient with preceding herpetic anterior chamber involvement
(patient 6). The characteristics of the ARN are summarized for each patient
in Table 1. An outline of the
subsequent clinical course and episodes of AU is given in Table 2.
In this study, the occurrence of AU following healed ARN was not related
to the severity of the initial episode of ARN. In a multiple logistic regression
analysis, severity factors, including duration of ARN, presence or absence
of bilateral disease, relapse of ARN, optic disc swelling, and retinal detachment,
were studied. The occurrence of AU following healed ARN did not show a statistically
significant association with any of these factors, but a larger sample size
would be required to evaluate this conclusively.
In this retrospective clinical study, we found that healed ARN may be
followed by recurrent AU in the absence of posterior segment inflammation
or a recurrence of ARN. This phenomenon could be either a direct effect of
the virus itself, persistent or reactivated, or a secondary immunological
effect. In this study, 7 of 19 patients with ARN developed recurrent AU. Only
1 of the 7 developed a second episode of ARN. This observation does not favor
a hypothesis of recurrent low-grade herpetic infection or a latent herpetic
infection resulting in an intermittent inflammatory response analogous to
the recurrence of the common cold sore. Furthermore, in 1 patient, the anterior
chamber was tapped at the time of an episode of subsequent AU, and HSV DNA
was not detected by polymerase chain reaction. This patient did not have a
history of herpetic disease or AU before ARN. Similarly, this would not support
recurrent low-grade persistent herpetic infection as a likely cause, but clearly
a negative result in 1 case does not conclusively negate the low-grade infection
hypothesis in the other patients.
We therefore hypothesize that isolated AU following healed ARN could
be a secondary immune event. Similar to the sensitization to retinal S antigen
following panretinal photocoagulation, the destruction of retinal tissue may
release antigen that may incite the production of antibodies against ocular
tissue. Molecular mimicry between T cells specific for the virus could be
implicated, cross-reacting with ocular tissue, the molecular structure of
which may be native, altered, or rendered manifest by the initial tissue damage.
As this phenomenon does not occur in all patients with ARN, there may be a
genetic predisposition to the development of recurrent AU following healed
Two of the 7 patients in this series with recurrent AU had had anterior
segment herpetic disease that predated the ARN. As a history of ocular herpetic
disease or cold sores is rare in patients with ARN, it is possible that previous
herpetic anterior segment involvement, even several years earlier, could predispose
patients with ARN to the development of recurrent AU after resolution of the
Recurrent AU may follow healed ARN without evidence of recurrent herpetic
involvement. It is possible that the destructive retinitis results in an autoimmune
response, and recurrent anterior chamber immune-mediated inflammation can
Corresponding author: Susan Lightman, PhD, FRCP, FRCOphth, Department
of Clinical Ophthalmology, Institute of Ophthalmology, Moorfields Eye Hospital,
City Road, London EC1V 2PD, England (e-mail: email@example.com).
Comer M, Young S, Lightman S. Anterior Uveitis After Healed Acute Retinal Necrosis. Arch Ophthalmol. 2002;120(1):88–90. doi:
Monkeypox Resource Center