Acute Retinal Pigment Epitheliitis: Spectral-Domain Optical Coherence Tomography Findings in 18 Cases

Abstract

Purpose.: To describe the imaging characteristics and to investigate the prognostic factors of acute retinal pigment epitheliitis.

Methods.: In this retrospective observational case series, a total of 18 patients (18 eyes) with acute retinal pigment epitheliitis were included. The features of acute retinal pigment epitheliitis were analyzed by spectral-domain optical coherence tomography (SD-OCT).

Results.: Spectral domain–OCT images obtained at baseline revealed abnormal reflectivity in the RPE inner layer in every case (18 eyes, 100%). The line corresponding to the RPE inner layer, the inner segment ellipsoid (ISe), showed abnormal reflectivity in 16 cases (89%). Among patients with ISe abnormalities, three eyes (17%) also showed external limiting membrane (ELM) disruption and two expanded abnormal reflectivity in the outer nuclear layer (ONL). With time, SD-OCT images showed disappearance and gradual resolution of the ONL hyperreflectivity, and disruption of the ELM or ISe. Sixteen eyes (89%) fully recovered their BCVA within 2 months. However, two eyes (11%) with involvement of the ELM or ONL at baseline, with BCVA of less than 20/70 at baseline, showed incomplete BCVA recovery after more than 1 year of follow-up.

Conclusions.: Spectral domain–OCT findings indicated that the initial lesion in acute retinal pigment epitheliitis is located at the junction between the photoreceptor outer segments and the apical sides of the RPE cells. Poor visual acuity at baseline and involvement of the ELM or ONL on SD-OCT at baseline were associated with the incomplete recovery of visual acuity.

Introduction

Acute retinal pigment epitheliitis, first described in 1972 by Krill, ¹ is a rare, transient macular disorder affecting healthy young adults. ^2–4 Patients present with painless blurred vision or central scotoma. ^2,4,5 Although bilateral cases have been reported, ⁵ most patients are affected unilaterally. The diagnosis of acute retinal pigment epitheliitis depends on the presence of fine pigment stippling in the macular area, surrounded by hypopigmented halos. ^1–5 In most cases, the disease is self-limited with the recovery of good visual acuity within 6 to 12 weeks without treatment. ^1,3,6,7  

To date, the pathophysiology of acute retinal pigment epitheliitis is unknown, and only a few case reports have described the associated clinical findings. In addition, although inflammation is considered to play a role in acute retinal pigment epitheliitis development, little is known about its exact histologic localization and pathogenesis. Currently, high-resolution spectral-domain optical coherence tomography (SD-OCT) has been used to obtain quasi-histologic, in vivo cross-sectional images of the posterior ocular fundus. ^8,9 The purpose of this study was to investigate the histologic localization, natural course, and prognostic factors of acute retinal pigment epitheliitis using SD-OCT. 

Materials and Methods

We retrospectively reviewed the medical records of 18 consecutive patients diagnosed with acute retinal pigment epitheliitis at the Retina Center of Kim's Eye Hospital in Konyang University College of Medicine from May 2008 through November 2013. The three patients described in our previous case report were also included in this study, which was approved by the Institutional Review Board of Kim's Eye Hospital, Konyang University College of Medicine. ¹⁰ The clinical research in this study followed the tenets of the Declaration of Helsinki. 

The diagnostic criteria for acute retinal pigment epitheliitis were as follows: (1) presented with acute visual blurring or central scotoma; (2) localized areas of pigment stippling associated with yellowish, often halo-like, zones of hypopigmentation in the fovea; and (3) no sign of any other ocular disease including anterior uveitis or vitritis. Other cases of acute disease that manifested as stippled pigmented lesions at the macula, such as acute posterior multifocal choroiditis, multiple evanescent white dot syndrome, acute posterior multifocal placoid pigment epitheliopathy, acute macular neuroretinopathy, serpiginous choroiditis, punctate inner choroidopathy, multifocal choroiditis, and laser-induced retinopathy were excluded. All patients were followed up for at least 4 months, with follow-up intervals of 2 weeks within 2 months of baseline (acute phase) and 2 to 3 months thereafter (chronic phase). There was no specific treatment protocol during the follow-up period. 

Each patient underwent a complete ocular examination that included the assessment of best-corrected visual acuity (BCVA), slit-lamp examination, and fundoscopy. Fluorescein angiography (FA) and SD-OCT were performed at baseline and during the follow-up period with an SD-OCT device (Spectral OCT/SLO; OTI Ophthalmic Technologies, Inc., Miami, FL, USA; or Spectralis; Heidelberg Engineering, Heidelberg, Germany). Indocyanine green angiography (ICGA) was performed at baseline for six patients. Fluorescein angiography, ICGA, and OCT findings were related to clinical observations. The results of blood analysis including erythrocyte sedimentation rate, full blood count, blood chemistry, and C-reactive protein levels were available for six patients. 

Results

Eighteen patients (18 eyes) with acute retinal pigment epitheliitis were included in this study. All patients were South Korean, and seven patients (39%) were female. Patient ages ranged from 22 to 51 years, with a mean age of 38 years (median, 37 years). The mean duration of symptoms was 6.8 days (range, 1–14 days). The average duration of follow-up was 12.4 months (range, 4–34 months). The pigmented lesions observed on fundoscopy were located at the fovea in all cases. In 10 cases (56%), the lesion was located at the foveola. Three patients (17%) presented with prodromal influenza-like symptoms approximately 1 to 2 weeks before the onset of visual symptoms. No subject had any other ocular history (e.g., history of medico-surgical illness). The results of blood analysis (available for six patients) showed no significant findings. 

The average baseline logMAR BCVA was 0.31 ± 0.16 (Snellen equivalent: 20/40, median: 20/40, range: 20/100 to 20/30). Average logMAR BCVA at the final visit was 0.13 ± 0.20 (Snellen equivalent: 20/26, median: 20/20, range: 20/50–20/20). This degree of visual improvement was statistically significant (P = 0.0004, paired t test). Among all subjects, 16 eyes (89%) achieved a complete recovery of baseline BCVA within 2 months of follow-up. Nine eyes (50%) recovered their BCVA within 1 month of follow-up, and seven eyes (39%) recovered baseline BCVA within 2 months. However, despite recovering baseline BCVA, two eyes (11%) had not fully recovered their BCVA to 20/20. In the case of incomplete recovery, visual acuity did not improve beyond 12 months of follow-up (Table 1, patients 1 and 5; Fig. 1). 

Figure 1

 

Images from patient number 1 with acute retinal pigment epitheliitis. (A) Fundus photograph showing a stippled lesion with a halo at the fovea. (B) Fluorescein angiography and indocyanine angiography of the patient. Fluorescein angiography showed unremarkable findings, and ICGA revealed an area of hyperfluorescence with central hypofluorescence at the macula, leading to a central halo-like area of hyperfluorescence. (C) Spectral-domain optical coherence tomography at baseline revealed disruption of the inner layer of the RPE, ISe line, and ELM. Abnormal hyperreflectivity was observed to extend from the inner layer of the RPE to the ONL layer. (D) Spectral domain–OCT images obtained at 1 month after onset showed the disappearance of abnormal hyperreflectivity and ELM recovery. However, disruption of the ISe and decreased reflectivity of the inner RPE layer were observed to persist. (E) Spectral domain–OCT images taken after 12 months showed the restoration of normal reflectivity at the inner RPE layer but persistent ISe disruption. The patient still complained of decreased vision. Visual acuity and OCT findings remained below baseline after 34 months of follow-up.

Table 1

 

Baseline Characteristics of Patients With Acute Retinal Pigment Epitheliitis

Case	Age, y	Sex	Duration of Symptom, d	Prodromal Symptom	Lesion Location	Follow-up Period, mo	FA Finding	ICGA Finding	Retinal Layer Involvement on SD-OCT	Baseline BCVA	1 moBCVA	2 moBCVA	6 moBCVA	Final BCVA
1	35	Male	4	Flu-like symptoms	Foveola	34	Unremarkable	Hyperfluorescent halo with hypoF	RPE inner layer, ISe, ELM, ONL	20/70	20/30	20/30	20/30	20/30
2	33	Male	2	None	Foveola	20	Transmission hyperF	N/A	RPE inner layer, ISe	20/30	20/20	20/20	20/20	20/20
3	39	Female	7	None	Fovea	12	Transmission hyperF	Hyperfluorescent halo with hypoF	RPE inner layer, ISe	20/25	20/20	20/20	20/20	20/20
4	45	Female	10	None	Fovea	14	Transmission hyperF	N/A	RPE inner layer, ISe	20/40	20/25	20/20	20/20	20/20
5	25	Male	10	None	Foveola	16	Transmission hyperF	N/A	RPE inner layer, ISe, ELM, ONL	20/100	20/50	20/50	20/50	20/50
6	53	Male	3	Flu-like symptoms	Fovea	6	Transmission hyperF	N/A	RPE inner layer, ISe	20/40	20/25	20/25	20/25	20/25
7	48	Male	6	None	Foveola	11	Transmission hyperF	N/A	RPE inner layer	20/30	20/20	20/20	20/20	20/20
8	31	Male	7	None	Foveola	6	Transmission hyperF	N/A	RPE inner layer, ISe	20/40	20/30	20/20	20/20	20/20
9	36	Female	4	None	Foveola	8	Unremarkable	N/A	RPE inner layer, ISe	20/40	20/25	20/20	20/20	20/20
10	33	Male	6	None	Foveola	5	Transmission hyperF	Hyperfluorescent halo with hypoF	RPE inner layer, ISe, ELM	20/40	20/20	20/20	N/A	20/20
11	51	Male	9	None	Fovea	18	Transmission hyperF	N/A	RPE inner layer, ISe	20/40	20/25	20/20	20/20	20/20
12	39	Female	5	Flu-like symptoms	Fovea	5	Transmission hyperF	N/A	RPE inner layer, ISe	20/30	20/20	20/20	N/A	20/20
13	40	Male	12	None	Fovea	6	Transmission hyperF	N/A	RPE inner layer, ISe, ELM	20/40	20/20	20/20	20/20	20/20
14	23	Female	4	None	Foveola	8	Transmission hyperF	N/A	RPE inner layer, ISe	20/40	20/20	20/20	20/20	20/20
15	47	Male	14	None	Fovea	24	Transmission hyperF	Focal hyperF	RPE inner layer	20/30	20/20	20/20	20/20	20/20
16	50	Female	7	None	Fovea	4	Transmission hyperF	N/A	RPE inner layer, ISe	20/40	20/20	20/20	N/A	20/20
17	33	Male	3	None	Foveola	5	Transmission hyperF	Unremarkable	RPE inner layer, ISe	20/30	20/20	20/20	N/A	20/20
18	22	Female	1	None	Foveola	6	Unremarkable	Hyperfluorescent halo with hypoF	RPE inner layer, ISe	20/50	20/20	20/20	20/20	20/20

 hyperF, hyperfluorescence; hypoF, hypofluorescence; N/A, not available.

Spectral domain–OCT images obtained at baseline showed abnormal reflectivity in the RPE inner layer in every case (18 eyes, 100%, Table 2, Fig. 2). The RPE inner layer was disrupted (14 eyes, 78%) or irregular with decreased reflectivity (four cases, 22%). The line corresponding to the RPE inner layer, the inner segment ellipsoid (ISe) showed abnormal reflectivity in 16 cases (89%). The affected area of the inner RPE was similar in size or larger than the corresponding region of the ISe line on SD-OCT in every case (Fig. 2). Definite disruption of the ISe was observed in 11 eyes (61%); decreased and irregular reflectivity was observed in 5 eyes (28%). Among the patients with ISe abnormalities (16 patients), three patients also showed disruptions of the external limiting membrane (ELM). Two of these patients displayed expanded abnormal reflectivity of the outer nuclear layer (ONL) in areas of ISe and ELM disruption (Table 2; Fig. 1). 

Figure 2

 

Fundoscopy, FA, and SD-OCT images from several cases of acute retinal pigment epitheliitis. Abnormal reflectivity of the inner layer of the RPE was observed in every case (arrowhead). Affected sites of inner RPE were similar in size or wider than affected areas of the ISe in every case. (A) Images from patient 2. Despite the full recovery of visual acuity, inner RPE irregularity persisted. (B) Images from patient 7. The ISe seemed to be relatively intact at baseline. The only significant abnormal finding observed on SD-OCT was decreased reflectivity of the inner layer of the RPE. After 1 month, the SD-OCT findings showed full restoration of the inner layer of the RPE. (C) Images from patient 10. Despite the unremarkable fluorescein angiography findings, SD-OCT revealed disruption of the inner RPE, ISe, and ELM. After visual recovery, the ELM and ISe were restored, but inner RPE irregularity persisted. (D) Images from patient 14. Spectral domain–OCT showed disruption of the RPE inner layer and ISe at baseline. These disruptions had resolved by 2 months. However, minute irregularities of the inner RPE were observed after visual recovery. (E) Images from patient 18. Fluorescein angiography image showed unremarkable finding. Despite disruption of the inner RPE and ISe with abnormal dome-shaped hyperreflectivity, the ELM was intact. After visual recovery, the ISe was restored but the inner RPE disruption remained.

Table 2

 

Acute Retinal Pigment Epitheliitis Evaluation With SD-OCT in 18 Patients

Retinal Layer of the Fovea	Abnormal Reflectivity on SD-OCT at Baseline, n (%)	Abnormal Reflectivity on SD-OCT at the Final Visit, n (%)
RPE outer layer	0 (0)	0 (0)
RPE inner layer	Total: 18 (100) Presence of definite disruption: 14 (78)Irregularity with decreased reflectivity: 4 (22)	Total: 4 (22)Not completely restoration with decreased  reflectivity or irregularity: 4 (22)
ISe	Total: 16 (89)Presence of definite disruption: 11 (61)Irregularity with decreased reflectivity: 5 (28)	Total: 2 (11)Not completely healed with disruption: 2 (11)
ELM	Total: 3 (17)Disruption: 3 (17)	0
ONL	Total: 2 (11)Abnormal hyperreflectivity: 2 (11)	0

 ELM, external limiting membrane; ISe, inner segment ellipsoid; ONL, outer nuclear layer; RPE, retinal pigment epithelium; SD-OCT, spectral-domain optical coherence tomography.

Over time, SD-OCT images showed disappearance of the outer nuclear layer hyperreflectivity, and gradual resolution of the ISe line disruptions. Recovery of the RPE inner layer occurred gradually. More than 1 month after symptom onset, SD-OCT images showed a restored and continuous RPE inner layer and ISe line in 14 eyes (78%). External limiting membrane and ONL abnormalities had resolved in every case (100%, Table 2). However, RPE inner layer abnormalities with decreased reflectivity or irregularity persisted in four eyes (22%); two eyes among these displayed signs of persistent ISe disruption (Table 2; Fig. 1). These instances of RPE abnormality and ISe disruption had not recovered after more than 12 months of follow-up. The two eyes with persistent disruption of the ISe line showed incomplete BCVA recovery (Table 1, patients 1 and 5). 

The FA findings at baseline showed transmission hyperfluorescence in 15 eyes (83%). In every case, the leakage of dye into surrounding retinal tissue was absent during some phase of the examination. Fluorescein angiography yielded unremarkable findings in the remaining cases, despite characteristic fundoscopic findings of acute retinal pigment epitheliitis in three cases (17%). 

Indocyanine green angiography, which was performed for six eyes included in the study, demonstrated normal choroidal perfusion for all subjects. The early phase of ICGA was unremarkable in all subjects; however, a small foveal patch of hyperfluorescence was observed in one eye, and central halo-like hyperfluorescence was observed from the mid- to late phase of ICGA in four eyes. ICGA revealed no significant findings in one eye. In cases where ICGA did reveal a significant lesion, the size of such lesions was larger than that of the corresponding area of transmission hyperfluorescence on FA. 

Discussion

The pathophysiology of acute retinal pigment epitheliitis is unknown. Some electrophysiologic and ophthalmic evidence suggests that acute retinal pigment epitheliitis is an inflammatory disorder of the RPE. ¹¹ Previous studies have suggested that the associated inflammation involves primarily outer neurosensory retina. ¹² Several OCT-based studies have shown that acute retinal pigment epitheliitis typically involves outer neurosensory retina and associated areas of the RPE. ⁶ A recent case report of SD-OCT in four patients suggested the possibility that the junction between the photoreceptor outer segments and the apical side of the RPE cells is the initial area involved in acute retinal pigment epitheliitis. ¹³ However, due to the low incidence of this condition, the published literature comprises mostly case reports and small case series. There is limited information available for those looking to study this unique macular disease. 

The outermost highly reflective band of the outer retina on OCT is thought to be the RPE, with a potential contribution from Bruch's membrane and the choriocapillaris. ^14,15 Previous reports have supported this designation, including histologic-OCT correlation in animal eyes. ¹⁶ In the current study, all subjects showed abnormal reflectivity of the inner RPE at baseline. Moreover, the affected area of the inner RPE was similar in size or larger than the corresponding region of the ISe band on SD-OCT images. These findings suggest that acute retinal pigment epitheliitis primarily affects the inner RPE area. Inner layer reflectivity of the RPE as detected by commercial SD-OCT is commonly thought to represent the tips of outer segments of the photoreceptors ¹⁷ or Verhoeff's membrane. ^18,19 Verhoeff's membrane is now known to comprise junctional complexes between RPE cells. ²⁰ Therefore, a reflective band that appears to be physically separated from the RPE on OCT images cannot be Verhoeff's membrane. This would mean that the inner RPE reflectivity detectable by SD-OCT does not correspond to Verhoeff's membrane. Reflectivity of the RPE inner layer as detected by SD-OCT is commonly considered to correspond to the area of contact between the RPE apical process and the external portion of the cone outer segment (contact cylinder of the outer segment). ¹⁵ Therefore, it could be assumed that acute retinal pigment epitheliitis is the result of a transient dysfunction or inflammation at the interface between the RPE and photoreceptor outer segments. In the present study, the lesions associated with acute retinal pigment epitheliitis were all located juxtafoveally, involving the foveal center or at the foveola. The fovea is the area of greatest cone density and lowest rod density of the retina. ²¹ Additionally, the integrity of the ISe line at the fovea on SD-OCT was related with cone cell density on adaptive optics scanning laser ophthalmoscopy. ^22,23 Thus, acute retinal pigment epitheliitis could originate at the interface between RPE cells and cone photoreceptor outer segments. 

Focal inflammation originating at the RPE/cone photoreceptor outer segment interface might expand to the inner photoreceptor segments or disrupt the ISe. In cases of severe inflammation, photoreceptor cell bodies could swell or exhibit other signs of damage. Interestingly, the abnormal reflectivity revealed by SD-OCT was not found at the level of the RPE, Bruch's membrane, or the choriocapillaris. The progression of acute retinal pigment epitheliitis that can be visualized using SD-OCT seems to expand from the RPE/photoreceptor junction toward the inner photoreceptor layer in all cases. We hypothesized that inflammation could spread from the discs of cone photoreceptors or apical processes of the RPE, to extend through photoreceptor axons, such that inflammatory change would be confined to the cone photoreceptor layer. 

It has been reported that good visual acuity can be recovered in most acute retinal pigment epitheliitis patients within 6 to 12 weeks, without treatment. ^1,3,6,7 In the current study, 89% of patients (16 eyes) recovered good visual acuity, and 11% patients (two eyes) exhibited limited improvements in visual acuity despite participation in follow-up exams for more than 1 year. All patients with incomplete recovery exhibited visual acuity of less than 20/70 at baseline, which represented relatively reduced visual acuity compared to the overall mean visual acuity (20/40). In addition, the SD-OCT findings of incomplete visual recovery cases showed more extensive retinal involvement at baseline, involving the inner layer of the RPE, the ELM, the ONL and the ISe. Histological analysis has shown that the cell bodies of cone photoreceptors are located at the similar level as the ELM, external to the rod photoreceptors. ²¹ Thus, involvement of the ELM or ONL could be associated with the damage of cone photoreceptor nuclei, which could result in irreversible destruction. Our results indicate that poor visual outcome could be predicted in patients with poor visual acuity at baseline and/or involvement of the ELM or ONL as detected by SD-OCT at baseline. 

Considering its self-limiting nature, the accurate diagnosis of acute retinal pigment epitheliitis is essential for avoiding unnecessary treatment. In some of the cases presented here, the FA findings were unremarkable despite the presence of pigmented foveal lesions in 17% of eyes (three of eighteen), and ICGA showed no specific findings in 17% of eyes (one of six eyes imaged by ICGA). The reason why ancillary imaging tests showed unremarkable findings in some patients is unclear. Such unremarkable findings did not seem to be associated with disease severity or progression over time. In some patients, in spite of relatively poor baseline visual acuity, involvement of various retinal layers on SD-OCT images, and brief symptom duration, FA and ICGA results were unremarkable (Fig. 2, patient 18). However, abnormal reflectivity of the inner RPE was detected at the fovea with SD-OCT in every case (100%). Our results showed that SD-OCT provided enhanced diagnostic sensitivity for acute retinal pigment epitheliitis that was unmatched by any other examination modality. Spectral domain–OCT can thus be used to assist in the diagnosis of this unique retinal condition, particularly in the early stages of disease, or in patients with minimal changes to the retina. Therefore, SD-OCT should be considered preferentially over other examinations when acute retinal pigment epitheliitis is suspected. 

Our study is limited by its retrospective nature and the small size of the cohort. In addition, if other multimodal imaging techniques were available (e.g., near-infrared reflectance images), it would be possible to redefine the clinical presentation of acute retinal pigment epitheliitis and thereby to enhance our analysis. However, we focused on the SD-OCT findings to identify characteristics that could be used to diagnose this rare macular disease. Moreover, to the best of our knowledge, this is the largest case series devoted to the investigation of acute retinal pigment epitheliitis. 

Herein, on the basis of the morphologic changes shown by SD-OCT, we suggest that acute retinal pigment epitheliitis derives from the interface between the RPE and cone photoreceptor outer segments. Although visual acuity was fully recovered in most cases, poor visual acuity at baseline and involvement of the ELM or ONL on SD-OCT at baseline were associated with the incomplete recovery of visual acuity. Further investigations are required to elucidate the precise physiopathology of this rare and unique macular disease. 

Acknowledgments

The authors alone are responsible for the content and writing of the paper. 

Disclosure: H.J. Cho, None; S.Y. Han, None; S.W. Cho, None; D.W. Lee, None; T.G. Lee, None; C.G. Kim, None; J.W. Kim, None