The crossroads of ocular inflammation and inherited retinal degenerations: a report of two cases

Summary

We describe the clinical and molecular diagnostic overlap between posterior uveitis and inherited retinal diseases (IRDs) in 2 cases of PRPF31 retinitis pigmentosa (RP) presenting with features of posterior uveitis. Both patients with PRPF31 retinopathy had typical inflammatory clinical features of anterior chamber (AC) cells, vitritis, retinal vasculitis, and cystoid macular edema (CME), which was resistant to carbonic anhydrase inhibitors, systemic steroids, and antimetabolites but eventually responded to intraocular fluocinolone implant. This improvement with local corticosteroids in these RP cases suggests an inflammatory component on top of retinal pigment epithelial pump dysfunction. Aggressive systemic anti-inflammatory/immunosuppressive treatment may cause ocular (eg, cataract, glaucoma) and/or systemic (eg, teratogenic effects, infection risk) morbidity; benefits and risks should be discussed with patients. Bilateral symmetrical reduced vision, nyctalopia, and pigmentary retinopathy in young patients suggests IRD, but significant inflammatory signs may be present and should be addressed. Genetic testing may be helpful in clarifying mixed IRD/uveitis phenotypes, although results must be interpreted cautiously.

Introduction

Inherited retinal diseases (IRDs) are a common cause of visual loss in young patients.1 IRDs are clinically and genetically heterogeneous with >300 known genetic associations.2 There are several “mimics” (eg, uveitis, trauma) in which late-stage features (eg, intraretinal pigment migration [IRPM], arteriolar narrowing) may simulate retinitis pigmentosa (RP) or other IRD phenotypes. Equally, early-stage disease (eg, cystoid macular edema [CME], reduced vision without classic pigmentary retinopathy) may suggest uveitis and lead to a delayed IRD diagnosis. In these cases, a curated set of investigations (eg, infectious, inflammatory, and genetic) can refine the diagnosis and provide guidance on prognosis and treatment.

Inflammatory signs (eg, vitreous cells, macular edema, cataract) and breakdown of the blood-ocular barrier are frequently seen in IRDs.3 Some IRD-associated genes have also been associated with a predisposition to ocular inflammation (eg, autosomal dominant neovascular inflammatory vitreoretinopathy [ADNIV] with the CAPN5 gene).4 We report a pedigree in which unclear signs and symptoms posed an etiological challenge.

Case 1

A 55-year-old woman presented to St. Paul’s Hospital, Vancouver, with bilateral painless reduction of visual acuity over the preceding 3 months. She had no history of ocular surgery or trauma and denied nyctalopia. Her past medical history was remarkable only for asthma. Her family history was positive for retinitis pigmentosa (RP), affecting her mother and maternal aunts and uncles.

On examination, visual acuity was 20/40 in the right eye and 20/80 in the left eye (emmetropia), with intraocular pressure (IOP) of 14 mm Hg in each eye. There were no anterior chamber (AC) cells or flare. She was phakic, without lens opacity. Funduscopic examination revealed nasal sectoral IRPM in each eye (Figure 1A–B). Fundus autofluorescence imaging (FAF) demonstrated evidence of symmetrical hypoautofluorescence (Figure 1C–D), and optical coherence tomography (OCT) showed marked CME (Figure 2) with central retinal thickness (CRT) of 629 μm in the right eye and 505 μm in the left eye. A presumptive diagnosis of RP with CME was made, and the patient was treated with topical carbonic anhydrase inhibitor (CAI, brinzolamide 1%) twice daily in both eyes.

Figure 1.

Case 1, pretreatment multimodal Imaging. A–B, Ultrawide-field color fundus photographs showing scant inferonasal bone spicules (white arrows). C–D, Fundus autofluorescence (FAF) showing areas of hypoautofluorescence inferonasally (white arrowheads) but also scattered throughout the superonasal region (black arrowheads). E–F, Late-phase fluorescein angiography (FA) demonstrating significant diffuse retinal pigment epithelium (RPE) “honeycomb” staining and vascular leakage (white arrow), angiographic cystoid macular edema (yellow arrow), and disc edema (red arrow).

Figure 2.

Case 1. A–B, Pre-treatment optical coherence tomography (OCT) showing significant bilateral cystoid macular edema (CME, white asterisks). C–D, OCT after insertion of fluocinolone implants demonstrating resolution of CME.

On review at 1 month, CRT had responded (right eye, 462 μm; left eye, 447 μm), but visual acuity deteriorated to 20/50 in the right eye and 20/200 in the left eye. Fluorescein angiography (FA) showed evidence of extensive vascular leakage (ie, retinal vasculitis), with angiographic CME and florid disc edema (right more than left) suggesting an inflammatory etiology (Figure 1E–F). The diagnosis was modified to retinal vasculitis / autoimmune retinopathy / intermediate uveitis, although investigations for infectious (including syphilis, tuberculosis, and toxoplasma) and inflammatory (eg, sarcoidosis) conditions were negative. The patient was started on topical difluprednate 0.05% and nepafenac 0.3% three times daily. Although some improvement occurred in the right eye (20/60; CRT, 362 μm) at 3 months, the left eye CME did not respond (20/200; CRT, 466 μm), prompting addition of systemic anti-inflammatory therapy (prednisone 60 mg and mycophenolate mofetil 1500 mg twice daily). Despite this combination of systemic and topical therapy, 6 months later, the aggressive features of uveitis and CME persisted, with visual acuity of 20/200 in both eyes. Intravitreal steroid injection in the left eye (triamcinolone 4 mg/0.1 ml) showed some anatomical (CRT, 310 μm) and functional (visual acuity, 20/80) improvement by 3 months, thus fluocinolone implants (Retisert 0.59 mg, Bausch and Lomb, Vaughan, Canada) were surgically inserted in both eyes. Control of CME was achieved (right eye, 217 μm; left eye, 240 μm) with stabilization of visual acuity of (20/30 in both eyes), which has lasted for 5 years (Figure 2C–D). Although she had no IOP response to topical steroids or intravitreal triamcinolone, after 18 months with fluocinolone implants in situ in both eyes, she developed cataract and glaucoma requiring surgical intervention.

Case 2

Her 32-year-old daughter presented 4 years later with a 1-year history of “terrible vision,” in particular nyctalopia (eg, needing flashlight to read a menu). She had no history of ocular pain, redness, or photophobia. Systemically, she had an atopic phenotype (ie, asthma, eczema), epilepsy treated with lamotrigine but no evidence of hearing loss or polydactyly. Her ophthalmic history included childhood strabismus and hyperopia (glasses from age 2 years) without amblyopia.

On examination, visual acuity was 20/40 in the right eye and 20/50 in the left eye, with low hyperopic correction (+2.00 DS in both eyes), and IOP was 13 mm Hg in the right eye and 14 mm Hg in the left eye. She was phakic and lenses were clear. On slit lamp examination, she had 1+ anterior chamber cell with 0.5+ anterior vitreous cell, without vitreous haze. On funduscopic examination, she was noted to have arteriolar sheathing and attenuation, with perivascular IRPM and peripheral retinal pallor (Figure 3A–B). Her maculae had a symmetrically thickened appearance, without hemorrhages or hard exudates. FAF showed highly symmetrical mid-peripheral hypoautofluorescence, with preserved posterior isoautofluorescence and macular hyperautofluorescence in the left more than right eye (Figure 3C–D). In addition to retinal pigment epithelium (RPE)-level staining (pattern reminiscent of a honeycomb), FA demonstrated marked large vessel perivascular and optic nerve head (ONH) hyperfluorescence, more marked in the right eye suggestive of vasculitis/phlebitis with disc edema (Figure 3E–F). OCT showed intraretinal and subretinal hyporeflective spaces, with CRT of 599 μm in the right eye and 611 μm in the left eye (Figure 4A–B).

Figure 3.

Case 2. Pretreatment multimodal imaging. A–B, Diffuse midperipheral RPE atrophic changes with mild-to-moderate bone spicules (black arrowheads) and vascular sheathing (white arrow). C–D, FAF showing peripheral hypoautofluorescence (white arrows) and central hyperautofluorescence (yellow arrows). E–F, Late-phase FA showing diffuse RPE staining in a “honeycomb” pattern, florid disc edema (red arrows), and vascular leakage (black arrows) in both eyes.

Figure 4.

Case 2 OCTs. Severe baseline CME (white asterisks) with associated subretinal fluid (red asterisks) in the right (A) and left (B) eyes. Significant improvement in CME post-treatment (C–D).

Investigations revealed negative infectious (ie, syphilis, tuberculosis) and inflammatory (eg, ANA, ACE) findings. The symmetrical presentation was highly suggestive of RP, so she was initially treated with topical brinzolamide 1% three times daily. However, visual acuity (20/40 in both eyes) and OCT (CRT of 694 μm in the right eye; 511 μm in the left eye) parameters failed to improve at 3 months’ follow-up, and topical difluprednate 0.05% and nepafenac 0.3% were added, leading to a partial functional (visual acuity of 20/30 in the right eye; 20/40 in the left eye) and anatomical (CRT of 246 μm in the right eye; 251 μm in the left eye) response (Figure 4C–D) 3 months later (Figure 4 A–B). She remains under active observation and has demonstrated recurrence of CME with withdrawal of treatment 6 months later. She currently remains on topical difluprednate and nepafenac without apparent ocular complications. Thus far, she has not required systemic or intravitreal anti-inflammatory medication, but we maintain a low threshold to escalate treatment due to her mother’s aggressive inflammatory phenotype.

Although no male-to-male vertical heritability was evident in the family, a three-generation pedigree suggested an autosomal dominant (AD) inheritance pattern. This, combined with the limited response to multiple anti-inflammatory therapies, both patients underwent genetic testing with a panel of 351 IRD-associated genes (Blueprint Genetics, Espoo, Finland). This revealed a pathogenic duplication/frameshift variant in PRPF31 (c.351dup, p.Tyr118Valfs*7) in both mother and daughter, supporting a diagnosis of AD RP11 (OMIM #600138).

Discussion

The inflammatory theory of IRD etiology has been studied for over four decades, with the “-itis” suffix of RP nomenclature presuming inflammation was the predominant destructive factor. Biomarkers supporting the role of inflammation in the degeneration of photoreceptors/RPE include circulating antiretinal antibodies (eg, anticarbonic anhydrase II), T cell activation to retinal antigens, and co-expression with other ocular and systemic autoimmune diseases.3,5

The features suggestive of uveitis in IRD cases vary between reports with contributions including AC cells, vitritis, vasculitis, and CME either clinically or angiographically.6 In reported series, the phenotype is referred to as uveitis with IRD features. However, with the confirmation of a genetic cause, or electroretinographic (ERG) features of IRD, perhaps the diagnosis should be refined to IRD with a prominent inflammatory phenotype (PIP). In the cases described herein, the mother’s presentation and course were not entirely typical of RP, prompting an aggressive anti-inflammatory management approach (escalating to intraocular surgery). Her daughter instead manifested an RP phenotype, although with a PIP component suggesting that this particular PRPF31 variant may have a more pro-inflammatory phenotype.

Many IRD-associated genes have been implicated in ocular inflammation and recalcitrant uveitis cases may be IRDs masquerading as uveitis (ie, PIP). These genes include, but are not limited to, ADGRV1, ALPK1 (ROSAH syndrome), CAPN5 (ADNIV), CBL, CDH23, CRB1, NLRC4, NLRP1, NLRP3, NOD2 (Blau syndrome), PRPF31, PTPN22, TYK2, and USH2A.3,7–9 However, approximately 1 in 3 of the global population carries a single allele for an autosomal recessive (AR) IRD; thus, genetic testing in an unwarranted or uncertain clinical situation may lead to spurious associations.10 For AR IRD, many interacting factors influence the phenotype, including the associated genotype, degree of deleteriousness of each variant, combination of variants, complex genetic background of interacting genes/proteins and acquired factors (eg, occupational light, trauma, and toxin exposure). For AD conditions, detection of a previously described pathogenic variant, as in the current cases, gives significantly greater confidence in the diagnosis. The risk of uncovering incidental genetic findings (ie, unrelated to the testing indication) should be considered before pursuing genetic testing but may be mitigated by testing technique (eg, gene panel limited to IRD-associated genes).11

The PRPF31 gene is an established cause of AD RP (OMIM#600138) that famously demonstrates incomplete penetrance and variable expressivity. The gene encodes a component of the “spliceosome” involved in post-translational RNA modification. The etiology of PRPF31-associated ocular inflammation is not established, being described in only isolated case reports.12 This may be explained by abnormal splice-modulating effects of mutated PRPF31, but modifying influences of variants in other genes may play a role. Immune reactions to retinal antigens, either primarily or after gene therapy, are thought to be primed by the innate immune system.13

CME is common (up to 50% of cases) in the setting of RP and is thought to be multifactorial, with contributions from blood-retinal-barrier breakdown, Müller cell and RPE pump dysfunction, altered visual cycle metabolism, and inflammation.14 Signs of chronic RPE dysfunction can be elicited on multimodal imaging (eg, macular honeycomb pattern seen on FA [Figures 1E–F and 3E–F]). In these cases, topical CAI monotherapy may have a limited or partial response for CME, because the mixed inflammatory and RPE degenerative mechanism may require combination with anti-inflammatory therapy.12 Preclinical data has shown beneficial structural and functional effects of systemic immunosuppression (ie, mycophenolate mofetil) in some murine RP genotypes (eg, PDE6A/PDE6B), and there is evidence that corticosteroids may treat TIMP3-associated macular neovascularization.15,16

FA is no longer a common imaging modality for IRDs, considering the volume of clinically relevant information that can gained from noninvasive techniques such as FAF and OCT.17 FA may highlight degenerative RPE changes in a mottled “honeycomb” pattern, as seen in our cases. Peripheral vascular leakage has been described in 11.5%–17% of RP cases, although no consistent leakage pattern has been identified.18,19 Park et al reported a series of ultrawide-field FA in a heterogenous IRD cohort.18 They noted “significant” leakage in 17% of patients, but this was subtle in comparison to the degree of leakage seen in our cases. Breakdown of the blood-retinal-barrier is thought to underlie vascular leakage, which can be confirmed and localized with FA, a hypothesis that is supported by elevated vitreous inflammatory markers and immune cells in RP eyes.18,20,21 The infrequency of routine FA use in IRD may be a barrier to accurate interpretation in the acute setting. FA may be best used to diagnose suspected complications of IRD (eg, Coats-like reaction, choroidal neovascularization) or conditions with similar signs (eg, posterior/intermediate uveitis).

The benefits and risks or complications of treatment must be carefully considered, because permanent blindness may be caused by both disease (eg, CME leading to retinal atrophy) and treatment (eg, steroid-induced glaucoma). Aggressive systemic anti-inflammatory and immunosuppressive treatment have morbidity and cost implications. In addition, there are family planning considerations, because many of these agents are teratogenic—a particular concern when treating women of childbearing age, as in our second case.

Late-stage disease of many pathologies, including posterior uveitis and resorbed retinal detachment may be indistinguishable from IRD without the diagnostic benefit of early disease features (eg, flecks, visible retinal detachment, inflammatory vasculitis). Unlike early-onset IRD (eg, nystagmus, very low vision, and extinguished ERG) investigated in children, teenage- to adult-onset RP shares some clinical features with chronic posterior uveitis. For example, vitreous cells may be seen in both conditions, although their nature differs (ie, pigment cells in RP vs leukocytes in uveitis).

Similarly, bone spicules are not pathognomonic for either condition, but a bilateral symmetrical distribution suggests RP. Specific features which favor uveitis include anterior chamber cells, a red, painful eye and the presence of one or more systemic autoimmune conditions (eg, sarcoidosis). The features favoring RP are bilateral symmetrical disease and nyctalopia. The constellation of symptoms can guide appropriate diagnostic tests. ERG can be indispensable in such diagnostic conundrums, clearly identifying rod and/or cone system dysfunction, which is not a classic feature of uveitis. Unfortunately, ERG was not available in the authors’ catchment area for a prolonged period. The decreasing cost of accurate IRD genetic testing should make this an accessible part of the diagnostic armamentarium for retina/uveitis/IRD specialists. Achieving a correct diagnosis has a positive impact on patients’ lives (eg, visual, socioeconomic, physical well-being). Accurately interpreted genetic testing in the context of the disease phenotype can be a useful adjunct to clinical care. Genetic testing is still warranted for cases in which family history and the constellation of symptoms and signs are consistent with IRD, though with appropriate guidance from experienced clinicians (eg, genetic counselors, geneticists).

Literature Search

A PubMed search was conducted on May 6, 2024, using combinations of the following keywords: inherited retinal disease, inherited retinal degenerations, retinitis pigmentosa, uveitis, retinal vasculitis, and PRPF31. No limits were placed on publication dates, types, or language.