Discussion

The main feature of this case was the appearance of peripheral acute retinal necrosis (ARN) after an initially isolated foveal lesion. Aqueous humor was obtained at our first examination, when no peripheral lesion was seen on dilated examination or widefield fundus photography. Four days later, peripheral necrotizing retinitis developed, and multiplex polymerase chain reaction (PCR) of the previously collected aqueous sample was positive only for Varicella-zoster virus (VZV) DNA. These findings suggest that the foveal lesion was an early manifestation of VZV-associated ARN rather than an unrelated macular disorder.

The optical coherence tomography (OCT) morphology should be distinguished from paracentral acute middle maculopathy (PAMM). PAMM is an ischemic OCT finding characterized by a placoid or band-like hyperreflective lesion centered in the outer plexiform and inner nuclear layers, often followed by thinning of the inner nuclear layer [10]. In this patient, the OCT abnormality involved nearly the entire inner retina, produced substantial posterior shadowing, and was accompanied by ellipsoid zone disruption. The lesion was also clinically visible as a whitish foveal focus in an eye with ocular hypertension, fine keratic precipitates, and increased aqueous flare. Anterior chamber cells, anterior vitreous cells, and vitreous opacity were absent at the first examination at our hospital. Taken together, the morphology and inflammatory findings were not typical of isolated PAMM. A focal macular ischemic process was initially considered because no peripheral lesion or angiographic vasculitis was documented; however, the subsequent clinical course and VZV-positive aqueous PCR strongly favored viral retinitis.

Macular or posterior pole involvement at the onset of ARN has been described previously [4-9]. These reports show that the first clinically apparent lesion may be located close to the fovea or within the posterior pole before peripheral necrosis becomes obvious. The present case provides serial multimodal imaging together with VZV-positive PCR findings from an aqueous sample collected before a peripheral lesion became visible. Widefield fundus photography at the referring hospital showed no visible peripheral lesion within the photographed field. However, the method of peripheral retinal examination at the first local ophthalmology clinic was not documented, and scleral depression was not performed at the referring hospital or at our first examination. Therefore, a very small far-peripheral lesion cannot be completely excluded.

Aqueous humor PCR was important in establishing the viral diagnosis. PCR testing is particularly useful when the clinical presentation is incomplete or atypical [12,13]. In this case, a 24-target multiplex panel detected VZV DNA alone in an aqueous sample collected before a peripheral lesion became visible. Herpes simplex virus and cytomegalovirus were negative, supporting VZV as the cause of the intraocular infection. The patient had already started oral valacyclovir before presentation to our hospital, which may have modified the subsequent clinical course. Early antiviral therapy may have contributed to the limited extent of retinal necrosis and the absence of retinal detachment during 42 weeks of follow-up.

The combination of ocular hypertension and subtle anterior uveitis was diagnostically important. Herpes simplex virus, VZV, and cytomegalovirus can cause hypertensive anterior uveitis with keratic precipitates, and these anterior segment findings may provide an early clue to herpetic disease [14,15]. In the present case, the intraocular pressure of 30 mmHg at the referring hospital, fine keratic precipitates, and increased aqueous flare were not typical of a purely ischemic macular lesion. These findings supported continued observation for viral retinitis and early antiviral treatment before peripheral ARN became visible.

Progressive outer retinal necrosis (PORN) was also considered because VZV retinitis can involve the posterior pole early. PORN usually occurs in severely immunocompromised patients and is characterized by multifocal deep or outer retinal necrosis with relatively little intraocular inflammation [16,17]. In this patient, HIV serology was negative, and no evidence of a systemic autoimmune or connective tissue disease, diabetes mellitus, or malignancy was found. Ocular hypertension, keratic precipitates, increased aqueous flare, and the subsequent development of peripheral necrotizing retinitis with vascular sheathing, punctate retinal hemorrhages, and vitreous opacity favored ARN over a classic PORN presentation.

This report has several limitations. First, it describes a single patient. Although VZV DNA was detected in aqueous humor collected before peripheral retinitis became visible, this does not prove with certainty that the foveal lesion itself was caused by VZV. Second, the method of peripheral retinal examination at the first local ophthalmology clinic was not documented. Dilated examination and widefield fundus photography at the referring hospital and our hospital showed no peripheral lesion, but scleral depression was not performed; therefore, a very small far-peripheral lesion cannot be completely excluded. Third, the earliest OCT image was obtained using a Topcon system, whereas subsequent OCT images were obtained using CIRRUS HD-OCT. Direct comparison between scans obtained using different platforms was therefore limited. Fourth, OCT angiography was limited to a 6 × 6-mm macular scan; vascular abnormalities outside the scanned area could not be assessed. Finally, valacyclovir started before presentation to our hospital may have altered the subsequent clinical course.