Peripheral Leakage on Ultra-Widefield Fluorescein Angiography in Patients With Inherited Retinal Degeneration

Abstract

Purpose:

This report illustrates that peripheral vascular leakage on ultra-widefield fluorescein angiography (FA) can occur in patients with inherited retinal degeneration (IRD) without evidence of a separate cause of leakage.

Methods:

We searched the electronic medical records of the Massachusetts Eye and Ear Infirmary from 2010 to 2019 for patients with an IRD diagnosis and examination with an ultra-widefield FA. Images from FAs were evaluated in masked fashion by 2 retina specialists. Documentation of an evaluation for alternative causes of vascular leakage was recorded, as well as results from electroretinography, Goldmann perimetry, and genetic testing.

Results:

A total of 305 patients with an IRD diagnosis and FA procedure code were identified. Of these, 26 patients had both a clinical diagnosis of IRD and ultra-widefield FA on detailed medical-record review. Three patients had FA to evaluate a Coats-like response and were excluded. Of the remaining 23, 4 patients (17%) had significant peripheral leakage on FA. Of these, 1 had pericentral retinitis pigmentosa (for which the genetic cause of disease was undefined), 1 had Refsum disease with confirmed biallelic PHYH mutations, 1 had a CRB1-associated macular dystrophy, and 1 had CERKL-associated macular dystrophy. There was no evidence of ocular inflammation from history, examination, or laboratory testing to account for the FA findings. Of the 19 patients without significant leakage, 4 had minimal leakage and 15 had no peripheral leakage.

Conclusions:

Peripheral retinal vascular leakage can be seen on ultra-widefield FA in patients with IRD that is likely due to the IRD disease process itself rather than to an additional, distinct eye condition.

Introduction

Inherited retinal degenerations (IRDs) represent a genetically and phenotypically heterogeneous group of disorders that affect approximately 1 in 3000 people and are a significant cause of vision loss worldwide. ¹ Retinitis pigmentosa (RP) and Stargardt disease, which are the 2 most common IRDs, typically have characteristic features on examination and clinical evaluation. A resulting clinical diagnosis can be further refined by genetic testing, and current genetic testing methods can reveal a molecular diagnosis in up to 60% of IRD cases with ongoing efforts to increase this success rate. ^2,3

Fluorescein angiography (FA) may be used when clinical features are less typical for an IRD, when there is concern for other diagnoses, or to add additional diagnostic information in conditions like Stargardt disease. FA also has the potential to demonstrate unexpected findings in patients with IRD. For example, among 78 consecutive patients with RP who had conventional 30° or 60° FAs, peripheral leakage was present in 9 eyes. ⁴ Ultra-widefield FA, which can capture 200° of the fundus in a single image, has come into widespread clinical use and can image peripheral retinal pathology that previously was difficult to capture. Therefore, it is important to also be aware of the ultra-widefield FA findings in IRDs so that diagnosis is not overlooked when peripheral leakage is seen. To our best knowledge, only 2 meeting abstracts and 1 case report have described ultra-widefield findings in IRDs; these featured patients with RP and CRB1-associated cone rod dystrophy. ^{5 -7} Therefore, the purpose of this study was to describe ultra-widefield FA findings across a spectrum of IRDs.

Methods

We performed a search of Partners HealthCare’s electronic medical record using the Partners Research Patient Data Registry to identify all patients seen by an ophthalmologist at the Massachusetts Eye and Ear Infirmary between 2010 and 2019 who had been assigned both a diagnosis code for an IRD and the Current Procedural Terminology code for FA (CPT 92235) at least once and not necessarily at the same visit. The International Classification of Diseases (ICD) codes used for IRD diagnoses were 362.70 to 362.77 (ICD, Ninth Revision) and H35.50 to H35.54 (ICD, Tenth Revision).

Patients identified from the database query then underwent a detailed review of their medical record. The specific IRD diagnosis and the supporting evidence for it were reviewed to determine if there was sufficient evidence for a diagnosis of IRD. Confirmation of an IRD diagnosis was based on data including clinical findings, retinal imaging, visual field findings, electroretinogram (ERG) results, and exclusion of alternative diagnoses. Identification of a genetic cause through genetic testing was not required. When available, genetic testing was performed primarily using diagnostic tests offered by commercial or academic laboratories with variant interpretation following standard guidelines. ⁸

The type of FA performed was reviewed, and patients who did not have ultra-widefield FA on the Optos platform were excluded. For the patients who had both an IRD diagnosis and an ultra-widefield FA, the FA images were graded independently by 2 retina specialists (R.M.H. and L.S.) for presence of leakage in the late frames of the angiograms in both eyes. Leakage was graded as none, minimal, or significant. The location and type of leakage (from large retinal vessels or deeper retinal capillaries) were recorded. Any disagreements between the 2 graders were arbitrated through re-review of the images until arriving at a consensus. For the patients with ultra-widefield FA, demographics, specific IRD diagnosis, and any genetic testing results were recorded.

Results

The initial query of the electronic medical records identified 305 patients (Figure 1). After detailed review of the medical records, 279 patients were excluded for having an unconvincing diagnosis of IRD or for lack of ultra-widefield FA. Therefore, 26 patients remained who had both a confirmed clinical IRD diagnosis and ultra-widefield FA (Table 1). Three of these patients had the FA to document a Coats-like reaction and thus had a secondary cause for their retinal vascular leakage. Among the remaining 23 patients without a Coats-like response, 4 (17%) patients had significant bilateral leakage on FA, 4 (17%) patients had minimal leakage, and 15 (65%) had no leakage. The following sections describe the 8 patients with significant leakage (Patient 1 to Patient 4) and minimal leakage (Patient 5 to Patient 8).

Figure 1.

Flowsheet showing method of patient identification. CPT indicates Current Procedural Terminology; FA, fluorescein angiography; ICD, International Classification of Diseases; IRD, inherited retinal degeneration; MEEI, Massachusetts Eye and Ear Infirmary.

^a ICD, Ninth Revision codes 362.70 to 362.77 and ICD, Tenth Revision codes H35.50 to H35.54.

^b CPT code 92235.

Table 1.

Patients Identified With a Clinical Diagnosis of Inherited Retinal Degeneration and Ultra-Widefield Fluorescein Angiography.

Age, y	Sex	Diagnosis	Genetic testing	Ultra-widefield FA findings	Location of leakage
47	M	Best disease	Not performed	No leakage	–
30	M	Cone-rod dystrophy	Negative	Minimal leakage	Temporal (right eye)
67	F	Macular dystrophy	ABCA4	No leakage	–
8	M	Macular dystrophy	Not performed	No leakage	–
6	M	Macular dystrophy	Not performed	No leakage	–
30	F	Macular dystrophy	Not performed	No leakage	–
36	F	Macular dystrophy	CRB1	Significant leakage	Temporal (both eyes)
45	F	Macular dystrophy	CERKL	Significant leakage	Temporal and nasal (both eyes)
43	M	Pattern dystrophy	Not performed	No leakage	–
48	F	Refsum disease	PHYH	Significant leakage	Diffuse (both eyes)
73	M	Retinitis pigmentosa	Negative	No leakage	–
49	M	Retinitis pigmentosa	Not performed	No leakage	–
36	F	Retinitis pigmentosa	PRPF31	No leakage	–
46	F	Retinitis pigmentosa	Negative	No leakage	–
40	F	Retinitis pigmentosa	RHO	No leakage	–
55	M	Retinitis pigmentosa	Negative	No leakage	–
50	F	Retinitis pigmentosa	Negative	Coats-like response	–
40	F	Retinitis pigmentosa	Negative	Significant leakage	Temporal and nasal (both eyes)
15	F	Stargardt disease	Negative	No leakage	–
60	F	Stargardt disease	ABCA4	No leakage	–
54	M	Stargardt disease	Negative	Minimal leakage	Temporal (left eye), nasal (both eyes)
24	M	Stargardt disease	ABCA4	Minimal leakage	Superior-temporal (both eyes)
21	M	Usher syndrome, type I	Not performed	No leakage	–
37	M	Usher syndrome, type I	MYO7A	Coats-like response	–
61	F	Usher syndrome, type II	Negative	Coats-like response	–
45	F	Usher syndrome, type II	USH2A a	Minimal leakage	Nasal (right eye)

Abbreviations: F, female; FA, fluorescein angiography; M, male.

^a One pathogenic variant and 1 variant of unknown significance in USH2A with segregation analysis pending.

Patient 1

A 40-year-old woman with a history of seronegative rheumatoid arthritis and progressive sensorineural hearing loss since the sixth grade was referred for an evaluation for possible retinal vasculitis. Her referring ophthalmologist had obtained an ultra-widefield FA to investigate peripheral visual field loss and noted far peripheral leakage. Her family history was notable for RP in her mother as well as adult-onset hearing loss in her mother and maternal grandmother. The referring doctor was unsure whether this family history was associated with the angiographic findings.

On presentation, the patient’s visual acuity (VA) was 20/20 –1 OD and 20/25 +3 OS. Slitlamp examination findings were negative for anterior chamber and vitreous cell. Fundus examination revealed subtle retinal pigment epithelium (RPE) mottling in the peripheral macula and arcades with attenuated vessels bilaterally (Figure 2, A and B). Spectral-domain optical coherence tomography (OCT) showed perifoveal loss of the outer retinal layers. Fundus autofluorescence (FAF) demonstrated subtle parafoveal hyperautofluorescence as well as granularity nasal to the nerve. Ultra-widefield FA exhibited significant multifocal deep leakage in the temporal and nasal periphery bilaterally (Figure 2, C and D). Goldmann perimetry showed bilateral relative perifoveal scotomas, and full-field ERG indicated reduced but recordable rod responses and delayed cone responses. This clinical constellation was consistent with pericentral RP, a subtype of mild RP in which the pathology is more posteriorly located than in typical RP.

Figure 2.

Patient 1: Ultra-widefield images show mildly attenuated retinal arteries in the (A) right and (B) left eyes. Ultra-widefield fluorescein angiography reveals significant multifocal leakage in the temporal and nasal periphery of the (C) right and (D) left eyes.

Results from a workup for inflammatory and infectious diseases by the referring doctor, including those for angiotensin-converting enzyme, lysozyme, fluorescent treponemal antibody absorption, anti-Treponema pallidum immunoglobulin M and immunoglobulin G, QuantiFERON-TB Gold (Qiagen), Lyme antibodies, and antinuclear antibodies (ANA), were negative, as were previous tests for cyclic citrullinated peptide, C-reactive protein, erythrocyte sedimentation rate, rheumatoid factor, complement 3, complement 4, antidouble-stranded DNA, U1 ribonucleoprotein, anti-Smith, anti-Jo1, and HLA-B27. Genetic testing using a commercially available retinal dystrophy gene panel showed 2 variants of uncertain significance in ARSG, which is a gene associated with Usher syndrome type IV, but no definite disease-causing variants. Mitochondrial testing returned negative results, and the genetic cause of disease remained undefined.

Patient 2

A 48-year-old woman was referred for evaluation of a suspected retinal dystrophy with awareness of abnormal peripheral vision beginning 1 year earlier. There was no family history of retinal disease. On presentation, her VA was 20/30 OD and 20/40 OS. Anterior segment examination showed only mild central thinning in the left eye that was consistent with her known diagnosis of keratoconus. Fundus examination revealed central pigment mottling, mild arteriolar attenuation, diffuse pigment clumps in the far periphery, and a few areas of cobblestone atrophy in both eyes (Figure 3, A and B). OCT illustrated generally intact photoreceptor layers with a few focal disruptions and trace cystoid changes. Ultra-widefield FAF presented areas of peripheral hypoautofluorescence and hyperautofluorescence. Ultra-widefield FA revealed diffuse retinal vascular leakage and mild optic nerve leakage in both eyes (Figure 3, C and D), which led to referral of the patient for evaluation of possible retinal vasculitis. However, the patient denied any episodes of photophobia, eye redness, or pain, and there were no inflammatory findings on examination such as vitreous or anterior chamber cell.

Figure 3.

Patient 2: Ultra-widefield images show mildly attenuated vessels and small peripheral pigment clumps in the (A) right and (B) left eyes. Ultra-widefield fluorescein angiography reveals diffuse leakage from the retinal vessels and mild optic nerve leakage in the (C) right and (D) left eyes.

Workup results for inflammatory and infectious diseases were negative or normal for anti-T pallidum immunoglobulin M and immunoglobulin G, QuantiFERON-TB Gold, angiotensin-converting enzyme, ANA, and Lyme antibodies. A chest radiograph showed no lymphadenopathy. Goldmann perimetry demonstrated mild bilateral peripheral constriction, and full-field ERG identified decreased rod and cone responses. Genetic testing identified 1 pathogenic (c.824G > A, p.[Arg275Gln]) and 1 likely pathogenic (c.497-2A > G) mutation in the gene PHYH, which is associated with Refsum disease. ⁹ This syndromic diagnosis unified several systemic findings and was accompanied by an elevated plasma phytanic acid level.

Patient 3

A 36-year-old woman was referred for worsening VA and nighttime vision over the past year. She had a history of bilateral cystoid macular edema identified in her early 20s and was treated with sub-Tenon steroid. On examination, best-corrected VA was 20/60 –1 OD and 20/60 +2 OS vs 20/25 OD and 20/30 OS 7 years prior. Anterior segment examination findings were normal without any anterior chamber or vitreous cell. On fundus examination, there were mild macular pigmentary changes bilaterally (Figure 4, A and B). OCT showed loss of the outer retinal layers centrally with hyperreflective deposits in the outer retina in both eyes but without cystoid changes. Ultra-widefield FAF revealed mild hypoautofluorescent mottling in the central macula bilaterally. Ultra-widefield FA was obtained to assess for inflammation and showed deep multifocal leakage in the far temporal periphery and peripheral vessel leakage bilaterally (Figure 4, C and D). No macular leakage was present.

Figure 4.

Patient 3: Ultra-widefield images show subtle macular retinal pigment epithelium changes in the right (A) and left (B) eyes. Ultra-widefield fluorescein angiography reveals mild deep multifocal staining and focal vessel leakage in the temporal periphery of the right (C) and left (D) eyes.

Previous inflammatory workup results were positive for HLA-B27 but were otherwise negative or normal for QuantiFERON-TB Gold, ANA, erythrocyte sedimentation rate, C-reactive protein, and comprehensive metabolic profile. Goldmann perimetry showed relative central scotomas, and multifocal ERG demonstrated bilateral depression. Full-field ERG was within normal limits. Genetic testing identified 2 mutations in CRB1: 1 pathogenic variant (c.2263_2272del, p.[Leu755ALAfs*10]) and 1 likely pathogenic variant (c.340T > G, p.[Cys114Gly]) with parental testing confirming that these variants were in trans. Her final diagnosis was CRB1-associated macular dystrophy with clinical similarity to previously reported cases. ¹⁰

Patient 4

A 45-year-old woman presented for evaluation of worsening central vision in the setting of a macular dystrophy diagnosed approximately 10 years earlier. There was no family history of retinal disease. Her VA was 20/125 –1 OD and 20/150 OS. Slitlamp examination showed no anterior chamber or vitreous cell. Fundus examination revealed central geographic-like atrophy with few overlying pigment clumps in both eyes accompanied by attenuated vessels and subtle midperipheral mottling (Figure 5, A and B). OCT showed extensive central outer retinal atrophy with accompanying macular hypoautofluorescence on FAF. Ultra-widefield FA exhibited multifocal peripheral leakage and poor temporal vascularity bilaterally (Figure 5, C and D). Subsequent Goldmann perimetry demonstrated bilateral central scotomas, and full-field ERG showed easily recordable but reduced rod and cone responses.

Figure 5.

Patient 4: Ultra-widefield images show mild arteriolar attenuation and central macular atrophy in the right (A) and left (B) eyes. Ultra-widefield fluorescein angiography reveals leakage in the nasal and temporal periphery and temporal nonperfusion in the right (C) and left (D) eyes.

Genetic testing identified 2 causative variants in CERKL: c.1540G > T, p.(Glu514*, pathogenic) and c.222del, p.(Gln74Hisfs*28, likely pathogenic). Several years later, the patient experienced polyarthralgias and myalgias leading to an extensive workup for inflammatory and infectious diseases that revealed a positive anti–cyclic citrullinated peptide antibody but no other contributory findings. The patient was suspected of having early rheumatoid arthritis without any symptoms or signs to suggest ocular inflammatory disease.

Patient 5

A 54-year-old man presented for evaluation of gradually worsening central vision over the past 5 years. There was no family history of retinal disease. Best-corrected VA was 20/30 –2 OU. Slitlamp examination showed no anterior chamber or vitreous cell. Fundus examination revealed perifoveal atrophy in the right eye as well as small yellow subretinal deposits in the macula and near periphery in both eyes (Supplemental Figure 1, A and B). OCT of each eye showed ellipsoid and RPE loss as well as degenerative cystic changes. FAF showed abnormal macular hyperautofluorescence and hypoautofluorescence sparing the peripapillary regions bilaterally with greater abnormalities in the right eye. Ultra-widefield FA demonstrated window defects and a dark choroid in both eyes (Supplemental Figure 1, C and D). In addition, there was focal leakage in the nasal periphery of the right eye (Supplemental Figure 1C) as well as the nasal and temporal periphery of the left eye (Supplemental Figure 1D). Goldmann perimetry identified paracentral scotomas in each eye with mild peripheral constriction, and full-field ERG showed mildly reduced rod and cone responses. The patient was diagnosed with macular dystrophy consistent with Stargardt disease. Genetic testing findings had been negative.

Patient 6

A 24-year-old man with poor central vision over the past year presented for evaluation. There was no known family history of retinal disease. Best-corrected VA was 20/50 OD and 20/80 OS. Anterior segment examination was unremarkable. Fundus examination showed macular atrophy with pigmentary changes and pisciform flecks bilaterally as well as capillary dropout in the periphery of both eyes (Supplemental Figure 2, A and B). Bilateral outer retinal atrophy involving the foveal and parafoveal regions was present on OCT with corresponding central hypoautofluorescence on FAF. Ultra-widefield FA exhibited a dark choroid, macular window defects and staining, and mild peripheral capillary dropout in the temporal periphery bilaterally (Supplemental Figure 2, C and D). Each eye had a single, focal, isolated area of superior-temporal leakage (Supplemental Figure 2, C and D). Goldmann perimetry demonstrated bilateral central scotomas. Full-field ERG found mildly reduced but easily detectable cone responses and near-normal rod responses.

Genetic testing identified 2 pathogenic variants in ABCA4 (c.6089G > A, p.[Arg2030Gln] and c.4195G > A, p.[Glu1399Lys]) as well as 1 variant of unknown significance in ABCA4 (c.2034G > T, p.[Lys678Asn]). Segregation analysis could not be performed, but these variants corresponded to the clinical diagnosis of Stargardt disease.

Patient 7

A 45-year-old woman with congenital sensorineural hearing loss and nyctalopia beginning in her teens was diagnosed with RP and presumed Usher syndrome type 2 at age 40 years. Her family history was negative for retinal degeneration and hearing loss. Full-field ERG at age 40 demonstrated nonrecordable rod responses with markedly reduced and delayed cone responses, and Goldmann perimetry showed bilateral constriction.

At age 45, her best-corrected VA was 20/25 OD and 20/50 OS. Anterior segment examination findings were normal in both eyes. Few anterior vitreous cells, mild vessel attenuation, and rare retinal pigmentary changes were noted in both eyes (Supplemental Figure 3, A and B). The left eye had a mildly vascular preretinal and retinal lesion around the nerve that was suspected to be an astrocytic hamartoma (Supplemental Figure 3B). Ultra-widefield FA was obtained for additional evaluation of this lesion, which did not show leakage (Supplemental Figure 3D). On retrospective review, there was minimal leakage from the far nasal vessels in the right eye (Supplemental Figure 3C). Genetic testing identified 1 likely pathogenic variant (c.11549_11552del, p.[Gly3850Valfs*33]) and 1 variant of unknown significance (c.8395G > C, p.[Gly2799Arg]) in USH2A with parental testing under way.

Patient 8

A 30-year-old man was referred to the IRD clinic for a suspected retinal dystrophy. His family history was notable for a sister with cone-rod dystrophy, but no other family members, including her sons or the patient’s brothers or children, were affected. His best-corrected VA was 20/300 OD and 20/60 –1 OS. Observations from the anterior segment examination were normal. Dilated fundus examination revealed symmetric and extensive sharply demarcated macular atrophy with a diffuse tapetal-like sheen in the peripheral retinas (Supplemental Figure 4, A and B). Macular OCT showed central loss of the photoreceptor and RPE layers. FAF of both eyes revealed macular hypoautofluorescence surrounded by a ring of hyperautofluorescence.

Ultra-widefield FA by the referring physician showed macular window defects bilaterally and minimal areas of multifocal leakage of the larger retinal vessels in the far temporal periphery of the right eye (Supplemental Figure 4C), but otherwise no leakage in either eye. Goldmann perimetry demonstrated central scotomas, and full-field ERG demonstrated reduced rod and cone responses with delayed cone implicit times in a pattern consistent with a diagnosis of cone-rod dystrophy. Genetic testing with a panel of IRD-associated diseases did not identify a genetic cause of disease.

Conclusions

This study found that approximately one-fifth of patients with an IRD diagnosis in whom ultra-widefield FA was performed demonstrated significant leakage. This information supports inclusion of IRD on the differential diagnosis in a patient with unexplained vision changes and leakage on ultra-widefield FA and adds to our understanding of the phenotypic heterogeneity of IRDs.

Vascular leakage on FA signifies a breakdown in the blood-retinal barrier, typically in the setting of retinal vasculitis from one of a myriad of infectious, autoimmune, inflammatory, and neoplastic disorders. ¹¹ A few authors have recorded peripheral vascular leakage with conventional, nonwidefield FA in patients with IRD. ^{4,12 -14} But conventional FA does not allow full evaluation of the retinal periphery. Two meeting abstracts from the same group of investigators, one in 2010 and one in 2013, have described FA findings with Optos ultra-widefield FA. The abstract from 2010 examined 16 patients with RP who had undergone ultra-widefield FA, and of these, 5 patients (31%) exhibited peripheral vascular leakage, with 3 of these 5 patients also showing cystoid macular edema on OCT and FA. ⁶ In the subsequent abstract by the same investigators, 15 of 25 patients with RP (60%) had peripheral vascular leakage in at least 1 eye. ⁵ The rate of FA leakage in the present study (17%) was closer to the findings from the initial 2010 abstract (31%). One difference was that our study included many different IRDs, not just RP. There may also have been other differences in methodology and definition of leakage between the studies, but we could not adequately assess these from the abstract text available to us.

Four of the patients we described showed significant leakage on their ultra-widefield FA. The pattern of leakage differed among the patients. For patient 1, who had pericentral RP, leakage was multifocal and from the deep retinal vessels in the midperiphery and far periphery and thus included areas without apparent retinal degeneration. Patient 2 had Refsum disease and showed leakage primarily from the large retinal vessels, including vessels in the midperiphery and posterior pole and not in more peripheral areas where degeneration was more advanced. Patient 3 had CRB1-associated maculopathy and leakage both from the larger retinal vessels and the deeper retinal capillaries, but this was confined to the far temporal periphery in each eye. Patient 4 had a CERKL-associated dystrophy and leakage in the temporal and nasal periphery of both eyes, possibly reflecting the extramacular photoreceptor involvement indicated by full-field ERG. The patients had negative findings from evaluations for systemic inflammatory diseases that can cause vasculitis. Although we cannot definitively rule out a second, primary inflammatory retinal vasculitis as the cause of the leakage, we feel this is unlikely.

Retinal vascular leakage has previously been described as a primary finding in IRDs, including patients with CRB1-associated cone-rod dystrophy. ^4,7,12,15,16 The etiology of the vascular damage that leads to leakage in IRDs is unclear and may or may not reflect an inflammatory process. Miller et al proposed that the changes may represent an early form of Coats-like response, although the leakage is not predominant in the inferior retina, as would be expected in that case. ⁶ Other studies in RP have shown evidence of an inflammatory component in photoreceptor degeneration in IRD. Yoshida and colleagues observed vitreous cell and increased levels of proinflammatory markers in 37.3% of 509 RP patients, and Newsome et al measured higher levels of immune cells in the vitreous of RP patients than in controls. ^17,18 Moreover, in an older study of 25 patients with RP and fluorescein leakage, all patients had elevated numbers of vitreous cells. ¹²

The leakage pattern in patient 1 and patient 4 would be consistent with what is observed as a diffuse vascular response to intraocular inflammation in primary inflammatory disorders such as pars planitis. ¹⁹ The leakage pattern in patient 2, with concentration of leakage along the major retinal vessels, suggested damage of the blood-retinal barrier at the level of larger vessels, which could have been due to inflammatory factors generated from deterioration of retinal tissue, particularly along the vessels. The leakage pattern in patient 3 was similar to patient 1 but milder and limited to the temporal peripheries. Of note, healthy eyes can also have peripheral abnormalities on ultra-widefield FA, but patient 3’s changes were different and more pronounced compared with what has been described in patients without peripheral disease. ²⁰

We described 4 additional patients—2 with Stargardt disease, 1 with Usher syndrome, and 1 with cone-rod dystrophy—who showed minimal leakage. Patients 5, 7, and 8 showed multifocal leakage from the larger retinal vessels, similar to the pattern seen in patient 2 but more limited and confined to the far nasal and temporal periphery. Patient 6, who reported symptoms starting just 1 year prior to the examination, showed only a single area of focal leakage in the far superior-temporal periphery in each eye. It is interesting that the 5 patients with primarily macular disease, patients 3 to 6 and patient 8, had leakage in the far periphery, remote from the site of most clinically appreciable retinal degeneration. This may suggest that there was milder retinal degeneration leading to localized inflammation even when the IRD seemed to be confined to the macula. A similar observation could be made in patient 1, in whom leakage was most pronounced in areas anterior to the regions most affected by pericentral RP.

Our study had limitations. First, the retrospective, cross-sectional design limited the precision of our estimate of the rate of FA leakage. To define the rate more precisely, consecutive patients with IRD would have to undergo ultra-widefield FA on a research basis, not for clinical indications. The patients included in our study had a clinical indication for their FA, and thus the rate of FA leakage may also be affected by this bias. Second, we included several types of IRDs instead of focusing on 1 type such as RP. This heterogeneity needs to be taken into account when comparing results to previous studies. One benefit of this diversity is the demonstration of leakage across different IRDs, including inherited maculopathies, which extends our knowledge about the scope of FA leakage among these disorders. Third, the sample size was small. This was in part because of the rare nature of the diagnosis, and because FA is infrequently used as part of the diagnostic workup for IRDs.

An additional limitation was our lack of a concurrent control group—patients without IRD who had an ultra-widefield FA performed and pathologies other than those that would be expected to cause peripheral leakage (eg, epiretinal membrane). Such a control group demonstrating no leakage would corroborate our findings in patients with IRD as abnormal and not simply part of expected variation. We were able, however, to compare our findings to previously published studies of ultra-widefield FA findings in “normal” eyes. ^{20 -25} We note that none of those studies imaged patients without a clinical indication, which is why we placed the word normal in quotation marks. Those studies ranged in their commentary on retinal leakage, with some that did not appear to observe any leakage, and others that identified leakage in up to 46.9% of patients. ^{20 -25} The studies that did find leakage did not describe the extent of the leakage in a systemic fashion as we did, or they discovered only very mild cases. ^{21 -23} The cases of far peripheral leakage shown in our patients appeared more extensive than most current descriptions in “normal” eyes, and instances of midperipheral or posterior pole leakage as seen in some of our cases were not described in case series of “normal” eyes.

FA plays a significant role in the diagnosis of retinal disease, and ultra-widefield FA offers an advantage over conventional FA by enabling visualization of previously uncharacterized changes in the retinal periphery. ²⁶ Conventional fundus cameras capture only the central 30° or 50° field of view; in contrast, widefield and ultra-widefield imaging allow up to 200° of view by leveraging the optics of an ellipsoid mirror. ²⁷ As ultra-widefield imaging becomes more common in practice, ophthalmologists may encounter more patients with undiagnosed IRD who have peripheral leakage and subsequently refer them for suspected vasculitis. Proper workup of these patients will depend on the clinician knowing that such findings can point to an inherited disease.

This case series illustrated that peripheral vascular leakage on FA may be a relatively common finding among patients with IRD. Patients who develop vascular leakage on FA without clear evidence of ocular or systemic inflammation should be evaluated for a possible IRD, especially in those who have other unexplained symptoms such as hearing loss or a family history of vision abnormalities. Timely and accurate diagnosis minimizes unnecessary exposure to harmful medications and enhances clinical care both for patients and their families.