Choroidal lesions in varicella zoster virus uveitis

Abstract

Purpose:

To evaluate choroidal lesions with spectral domain optical coherence tomography (SD-OCT) scan in varicella zoster virus (VZV) uveitis.

Methods:

VZV-uveitis cases which underwent OCT scan for choroidal lesions were studied. SD-OCT scan passing through these lesions was studied in detail. Subfoveal choroidal thickness (SFCT) during active and resolved stages was studied. Angiogaphic features were studied where available.

Results:

Thirteen out of 15 cases had same-sided herpes zoster ophthalmicus skin rashes. All except three patients had old or active kerato-uveitis. All eyes demonstrated clear vitreous and a single or multiple hypopigmented orangish-yellow choroidal lesions. The number of lesions remained unchanged during the follow-up on clinical examination. SD-OCT over lesions (n = 11) showed choroidal thinning (n = 5), hyporeflective choroidal elevation during active inflammation (n = 3), transmission effects (n = 4), and ellipsoid zone disruption (n = 7). The mean change in SFCT (n = 9) after resolution of the inflammation was 26.3 μm (range: 3–90 μm). Fundus fluorescein angiography showed iso-fluorescence over lesions in all (n = 5), but indocyanine green angiography (n = 3) showed hypofluorescence at lesions. Mean follow-up was 1.38 years (range: 3 months–7 years). De-novo appearance of choroidal lesion during the first relapse of VZV-uveitis was captured in one case.

Conclusion:

VZV-uveitis can cause focal or multifocal hypopigmented choroidal lesions with thickening or scarring of choroidal tissue, depending on the disease activity.

Varicella zoster virus (VZV) is known to cause keratitis, scleritis, sclero-keratitis, kerato-uveitis, retinitis with or without vasculitis, and optic neuritis.[1,2] Choroidal involvement due to VZV posterior uveitis or panuveitis is rarely reported,[3-6] and even rarer is in kerato-uveitis.[7] Unchanged status of these lesions on clinical fundus examinations during the follow-up and limited imaging work-up in cases of VZV kerato-uveitis also raised doubts regarding the inflammatory activity of these lesions, then termed as “choroidal vitiligo in VZV kerato-uveitis”.[7] In the present series, we evaluated choroidal lesions in VZV-uveitis with multimodal imaging (MMI) and studied changes in choroidal thickness during follow-up.

Methods

This is a retrospective case study of patients diagnosed with VZV uveitis with choroidal lesions (n = 15) who underwent imaging studies (wide field fundus photography-Optos™, fundus fluorescein angiography (FFA), indocyanine green angiography (ICG), spectral domain optical coherence tomography (SD-OCT) (Heidelberg™) and OCT-angiography (OCT-A)) for the choroidal lesions. Cases were diagnosed clinically with a history of or evidence of herpes zoster ophthalmicus (HZO) skin lesions on the side of affected eye (n = 13). Eyes without a history of HZO (n = 2) were diagnosed based on decreased corneal sensation, dendritic ulcer, stromal keratitis, granulomatous pigmented central or diffuse keratic precipitates, hypertensive uveitis, sectoral iris atrophy, and presence of hypopigmented choroidal lesions unilaterally in the affected eye as described previously.[7] Cases with bilateral uveitis were excluded. The study was approved by the internal review board and adhered to the tenets of the Declaration of Helsinki. Patient’s demographic data and clinical and imaging findings were studied. The pattern of choroidal lesions on Optos™ scan was noted. Findings of SD-OCT scan passing through the lesion and subfoveal choroidal thickness (SFCT) during the active stage and after the resolution of inflammation were studied. FFA and ICG features of these lesions and OCT-angiography (OCT-A) (Optovue™) features were also studied where available.

Results

Fifteen eyes of 15 patients were studied. Seven were males and eight were females. The mean age at the presentation was 61.8 years (range: 26–75 years). Thirteen out of 15 cases had a recent or old history of typical HZO skin rash on the side of the affected eye. The other two cases without HZO had kerato-uveitis with decreased corneal sensation and unilateral hypopigmented choroidal lesions in the affected eye. Twelve patients were diagnosed as kerato-uveitis, 1 scleritis, 1 anterior uveitis, and 1 had choroidal lesions with the old history of HZO. Sectoral iris atrophy was seen in three cases. All eyes had clear vitreous and focal (n = 3) or multifocal (n = 12) choroidal lesions, with various patterns [Table 1]. A fellow eye examination was negative for choroidal lesions and uveitis in all. FFA showed isofluorescence over the lesions in all (n = 5), and no vascular leakage was observed except in a case of pseudophakic macular edema. ICG (n = 3) in all phases showed mild-moderate hypofluorescence at the choroidal lesions seen clinically. Scarred choroidal lesions appreciated as significantly decreased choroidal thickness on OCT scan also demonstrated hypofluorescence on ICG.

Table 1.

The pattern of choroidal lesions and choroidal thickness evaluative

Age/Sex	History of HZO	Anterior Segment	Pattern of choroidal lesions	SFCT (active stage)	SFCT (resolution)	SFCT Difference	OCT- over lesion*
77/M	Yes	Kerato- uveitis	Random- macular and extramacular, 3 quad, pre- and postequatorial	281	N.A.	N.A.	Choroidal scarring, Choroidal thickening during reactivation
70/F	No	Kerato- uveitis	Diffuse macular and extramacular, all quad, pre, and postequatorial	148	N.A.	N.A.	Normal
55/F	Yes	Kerato- uveitis	Diffuse extramacular, all quad, pre- and postequatorial	202	193	9	Choroidal thickening, mild choroidal elevation during the active stage, EZ loss
57/F	Yes	Kerato- uveitis	Random- extramacular, 1 quad, preequatorial	N.A.	N.A.	N.A.	N.A.
77/F	Yes	Kerato- uveitis	Random- extramacular, 3 quad, preequatorial	280	190	90	N.A.
63/F	Yes	Kerato- uveitis	single-extramacular, 1 quad, postequatorial	173	170	3	Mild choroidal thickening- unchanged, EZ disruption
63/M	Yes	Kerato- uveitis	Linear macular and extramacular, 1 quad, postequatorial	144	125	19	Choroidal scarring, choroidal elevation, transmission effects
67/M	Yes	Kerato- uveitis	Random macular and extramacular, all quad, pre, and postequatorial	189	177	12	Choroidal scarring, Choroidal elevations, EZ loss, transmission effects
61/F	Yes	WNL	Diffuse sectoral, 1 quad, pre- and postequatorial	N.A.	247	N.A.	Diffuse choroidal thinning, EZ loss, indistinct retinal layers
64/M	Yes	Anterior scleritis	Diffuse sectoral, 3 quad, pre- and postequatorial	221	209	12	Diffuse choroidal thinning, indistinct retinal layers, EZ loss
26/F	No	Kerato- uveitis	Single inf. Peripheral lesion, 1 quad, preequatorial,	240	N.A.	N.A.	Choroidal thickening, EZ disruption, transmission effects
60/M	Yes	Kerato- uveitis	Inf-temp Choroidal vitiligo, 1 quad, postequatorial	345	326	19	Choroidal thickening, EZ disruption
55/F	Yes	Kerato- uveitis	Single nasal peripheral lesion, 1 quad, preequatorial	424	400	24	N.A.
67/M	Yes	Anterior uveitis	Multiple lesions mid periphery, 1 quad, postequatorial	N.A.	110	N.A.	N.A.
66/M	Yes	Kerato- uveitis	Random macular and extramacular, 3 quad, postequatorial	293	244	49	Choroidal thickening, transmission effects

M=male, F=female, HZO=Herpes Zoster ophthalmicus skin rash, WNL=within normal limits, SFCT=subfoveal choroidal thickness, N.A. = not available/Not applicable, EZ=ellipsoid zone, quad: quadrant. *: OCT scan passing through multiple choroidal lesions in the same eye

SD-OCT passing through the lesions (n = 11) showed diffuse choroidal thinning (n = 5) suggestive of atrophy in the area of choroidal lesions [Fig. 1]. Hyporeflective choroidal elevation during the active stage was appreciated only in three eyes, which decreased after the resolution of the inflammation [Fig. 2]. The mean change in SFCT (n = 9) after the resolution was 26.3 μm (range: 3–90 μm). No subretinal fluid was detected over or around the lesions. But ellipsoid zone (EZ) loss or disruption was seen in seven eyes [Fig. 3]. Two eyes showed indistinct retinal layers over clustered multifocal choroidal lesions. Number of choroidal lesions, shape, and size of the lesions remained unchanged during follow-up on clinical examination. The mean follow-up was 1.38 years (range: 0.3–7 years).

Figure 1.

A 61/F, an old case of left HZO came for the routine eye checkup. Fundus examination revealed multiple hypopigmented choroidal lesions in the inferonasal quadrant (a). ICG showed hypofluorescent lesions corresponding to the fundus lesions (b). SD-OCT in the healthy quadrant showed normal choroidal thickness (c), whereas diffuse choroidal thinning was appreciated in the inferonasal quadrant and over the lesions (arrows) (d and e)

Figure 2.

A 67/M with a history of HZO presented with kerato-uveitis. Wide field fundus photo shows multiple hypopigmented peripheral choroidal lesions (arrows) (a). SD-OCT scan over the peripheral lesions showed choroidal elevation (b). Decrease in choroidal thickness over the same lesions can be seen after 1 month of treatment with antivirals and topical steroids (c)

Figure 3.

Fundus photo of the same patient in Fig. 3 shows choroidal lesions at the macula (a). SD-OCT scan through the temporal lesion shows loss of the ellipsoid zone (vertical arrow) and transmission effect below the lesion (horizontal arrow) (b). SD-OCT scan passing through the inferior lesion to the fovea shows thinning of the choroid (vertical arrow) (c)

In case 15, no fundus lesion was detected during the first episode of VZV-uveitis, but multiple hypopigmented choroidal lesions were appreciated during the relapse after 8 months [Fig. 4]. SD-OCT through the lesion showed hyporeflective choroidal stroma, while OCT-A showed normal vasculature of the chorio-capillaries at the site of the choroidal lesion [Fig. 4]. Whereas in case 7, chorio-capillary flow void area was noted on OCT-A, corresponding to choroidal thinning seen on OCT and the hypopigmented choroidal lesion seen on the fundus photograph.

Figure 4.

Wide-field fundus photo of the left eye at the presentation during the first episode of VZV-kerato-uveitis shows normal fundus (a). Eight months later, during the relapse of kerato-uveitis, the fundus showed multiple hypopigmented choroidal lesions (arrows) (b). SD-OCT scan over the choroidal lesion at the superior macula shows choroidal hyporeflectivity with increased transmission effect (arrow) (c). OCT-angiography (8 × 8 mm scan) of the macula showed normal vasculature of the chorio-capillaries at the site of the choroidal lesion (d)

Discussion

In this series, we studied SD-OCT characteristics and angiographic features of choroidal lesions in VZV infection. We noted that these lesions although appear unchanged on clinical fundus examination during the follow-up, mild changes in the choroidal thickness can be appreciated depending on the disease activity. The shape and the size of the lesions remained status quo once they appeared during the follow-up clinical examination, but the SD-OCT demonstrated hyporeflective choroidal elevations during the active stage, which resolved after the resolution of the inflammation in a few cases. Our study also revealed that some lesions were atrophic as the SD-OCT passing through these lesions showed severe thinning of choroid, which also corresponded with the hypofluorescent choroidal lesions seen on the ICG scan. SD-OCT also showed hyporeflectivity in the choroid without elevation but with mild choroidal thickening, which may suggest the resolving stage of the lesion. Whether hyporeflective choroidal lesions are dilated choroidal vessels or choroidal granuloma are debatable. Co-localization with unilateral hypopigmented choroidal lesions on fundus photo, and hypo-fluorescent lesions seen on ICG, as well as the “transmission effects” as described by Invernizzi et al.,[8] and thinning of choroidal tissue at the lesions seen on SD-OCT may point toward subtle inflammatory changes due to VZV infection than mere dilated choroidal vessels seen in normal eyes. We also studied SFCT during the active and resolved stages, which showed mild thickening during active inflammation. This may suggest overall choroidal thickening as a result of anterior segment inflammation, but the subtle and dome-shaped choroidal elevations appreciated in few cases [Fig. 3] may point toward additional inflammatory activity in these lesions. In addition to the choroidal changes, EZ disruption was also appreciated over the lesions in few eyes. It was more in the area of clustered lesions and along with indistinct retinal layers, mainly in the peripheral retina.

Choroidal lesions in VZV infection have been described in various reports. Punched-out, depigmented, oval areas of choroidal atrophy scattered around the midperiphery were described in three patients with the old history of HZO.[9] All their patients had sectorial iris atrophy. They suspected focal ischemia of the choroid and midperipheral fundus as an entry point to the neurotropic virus. Thus, one may expect choroidal lesions to appear corresponding with the sector of iris involvement, but no such correlation was observed in two of our cases with sectoral iris atrophy, while in the third case, iris atrophic patch and the single choroidal lesion were seen in inferior quadrant. Histopathological analysis of these lesions in the resolved state has also been done by McKelvie et al.[10] in a single case report. The authors observed significant choroidal thinning without an obvious decrease in choriocapillaris density. Patchy fibrosis of posterior ciliary nerves, along with perivascular mononuclear infiltration of posterior ciliary arteries, was also seen. Sparing of chorio-capillaries as described by McKelvie et al.[10] was in concurrence with the OCT-A findings seen in case 15 but not in case 7 in our series. Unfortunately, OCT-A was not available in other eyes.

Bissig et al.[11] have studied ICG in four retinitis patients and found patches of choroidal hypofluorescence in areas corresponding to the necrotic retina. Persistent hypofluorescence of choroidal lesions on ICG confirmed the presence of active choroiditis in another report.[12] Unfortunately, in our study, FFA and ICG were available only in few cases. As expected, ICG was superior to FFA in demarcating these lesions but was inferior to SD-OCT in the detection of activity in the lesion as scarred lesions also appeared hypofluorescent on ICG.

Although most of our patients had kerato-uveitis, we also had cases without corneal involvement and with scleritis and a case with normal anterior segment but with the old history of HZO. Recently, the association of choroidal lesions with acute retinal necrosis (ARN) has also been reported.[3] Thus, the choroidal lesions can be seen in various forms of VZV infection. One should be aware of this as they may mimic choroiditis lesions of other infectious or immune-mediated as well as noninflammatory conditions like tuberculosis, sarcoidosis, bird shot chorio-retinopathy (BCR), and focal or multifocal scleral nodules.[13,14] History of HZO and normal other eye examination, as well as the treatment outcomes and long-term follow-up, would give important differentiating clues in challenging cases. Interestingly, Bodaghi et al.[15] have described a case as bilateral asymmetric BCR which was positive for VZV on aqueous fluid evaluation for VZV-specific antibodies. Such situations are rare and would be challenging to differentiate.

Choroidal involvement in VZV was seen as a delayed feature and a delayed immune response by few authors, but concomitant choroidal involvement in ARN has been reported by Singh et al.[3] In most of our cases, it remained unknown if these lesions appeared at the first episode of VZV infection or later. Corneal haze due to keratitis hindered posterior segment examination in most of our cases during the active inflammation. But case 15 provided the evidence of “de-novo” appearance of a choroidal lesion, which may suggest that it is a delayed feature of VZV-kerato-uveitis.

Conclusion

In conclusion, we report the largest series on choroidal lesions in VZV-uveitis and demonstrate various phases of activity in the lesion using MMI with long term follow-up. VZV-uveitis can cause focal or multifocal hypopigmented choroidal lesions resulting in choroidal thickening or scarring depending on disease activity. The delayed appearance of these lesions and chorio-capillary sparing, as noted in a single case in our study, needs to be confirmed in further larger studies.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.