Long-Term Outcomes and Risk Factors for Severe Vision Loss in Autosomal Dominant Neovascular Inflammatory Vitreoretinopathy (ADNIV)


Autosomal dominant neovascular inflammatory vitreoretinopathy (ADNIV) is a rare disorder characterized by uveitis, retinal neovascularization, and retinal degeneration. We sought to describe the course of treated and untreated ADNIV and to identify risk factors for severe vision loss.


Observational case series.


Clinical data from ADNIV patients from 4 families seen from 1967 through 2019 at a single academic, tertiary referral center were reviewed. The main outcome measures were visual acuity at baseline and follow-up, as well as risk factors for vision loss.


A total of 130 eyes from 65 ADNIV patients (45 female, 20 male; mean age 40.8 years, range 6-77 years) were included. Mean best corrected visual acuity (BCVA) at presentation was LogMAR 0.59 (about Snellen 20/80). Longitudinal analysis included 84 eyes from 42 patients (31 female, 11 male), with mean follow-up of 17.3 years (range 2-43.6 years). Mean BCVA at last follow-up was LogMAR 1.48 (about Snellen 20/600). The disease accelerated in the fifth decade of life, during which the majority of eyes went from normal vision or mild vision loss to at least moderate vision loss (20/70 Snellen equivalent); 25 eyes from 16 patients (29.8%;) showed a steep trajectory of vision loss to no light perception. Tractional retinal detachment was the greatest risk factor for severe vision loss (BCVA <20/200) on multivariable analysis ( P < .05).


Patients with ADNIV have a high lifetime risk of severe vision loss. Tractional retinal detachment is an important risk factor for poor vision.

A A utosomal dominant neovascular inflammatory vitreoretinopathy (ADNIV), first described by Bennett et al in 1990, is a rare inherited disorder characterized by uveitis, retinal neovascularization, and retinal degeneration. The clinical features mimic those of noninfectious panuveitis, proliferative diabetic retinopathy, and retinitis pigmentosa. Blindness due to complications including mixed-mechanism glaucoma and relentless tractional retinal detachments is common. In 1992, soon after the initial clinical description, the genetic locus for ADNIV was mapped to chromosome 11q13. Two decades later, the causative gene was identified as CAPN5, which encodes calpain-5, a calcium-activated cysteine protease mainly expressed in the outer segments of retinal photoreceptor cells, as well as in the photoreceptor synapses, the inner plexiform layer, mitochondria, and retinal ganglion cells. Mutations in CAPN5 lead to mislocalization of calpain-5 from the photoreceptor cell membrane to the cytosol and increase expression of numerous autoinflammatory factors, including tumor necrosis factor−α (TNFα), interleukin-1-α (IL1α), and toll like receptor 4 (TLR4), as well as downregulation of versican, which is involved in inherited vitreoretinopathies such as erosive vitreoretinopathy and Wagner syndrome.

In similarity to other autosomal dominant conditions, ADNIV patients from the same pedigree with the same genotype can possess a wide range of phenotypes, with varying severities of vision loss. Based on cross-sectional data from the initial ADNIV families, Mahajan et al proposed a staging system, , which was developed using cross-sectional data only and was intended as a classification scheme rather than a means to provide prognostic information. Currently, there are relatively few publications on longitudinal changes in ADNIV over time, and little is known regarding the natural history of the condition, rate of severe vision loss, risk factors for poor vision, or efficacy of various treatments.

In this study, we describe the clinical features of ADNIV seen in a large cohort of patients from 4 distinct families. We use longitudinal data to provide insight into the natural history of both treated and untreated ADNIV. We evaluate the rate of sight-threatening complications (eg, glaucoma, tractional retinal detachment) and risk factors for poor outcomes (eg, severe vision loss, phthisis, enucleation).



This is a retrospective chart review of patients seen at the University of Iowa with ADNIV from 1967 to November 21, 2019, including some members of previously published multigenerational families. , The research was approved by the Institutional Review Board at the University of Iowa and adhered to the tenets of the Declaration of Helsinki, and was conducted in accordance with regulations set forth by the Health Insurance Portability and Accountability Act.


Demographic data, including age at presentation and last visit, sex, and genotype, were recorded for each patient. Results from clinical examination at each visit, including best corrected visual acuity (BCVA), intraocular pressure (IOP), as well as slit lamp and dilated fundus examination findings were recorded. Available multimodal imaging from color fundus photography and optical coherence tomography (OCT) imaging were reviewed. Note was made of key clinical features including the presence of vitreous cells, anterior segment neovascularization (NV) including NV of the iris or angle, peripheral retinal NV, vitreomacular traction (VMT), epiretinal membrane (ERM), full thickness macular hole (FTMH), cystoid macular edema (CME), vitreous hemorrhage (VH), peripheral pigmentary changes, tractional (TRD), and/or rhegmatogenous retinal detachment (RRD) with or without macular involvement.

Medical management, including local therapy such as intravitreal steroid or anti−vascular endothelial growth factor (VEGF) injection, or systemic oral steroids or immunosuppression, was recorded for each patient. Similarly, surgical interventions, including cataract extraction with intraocular lens placement (CEIOL), glaucoma surgery (eg, tube shunt, or filtering procedure), fluocinolone acetonide 0.59 mg intravitreal implant (Retisert, Bausch and Lomb), panretinal photocoagulation (PRP), or pars plana vitrectomy (PPV, eg, for epiretinal membrane peel, retinal detachment repair) were recorded for each patient. Poor outcomes were defined as those with (1) severe vision loss (BCVA <20/200) or worse, including those with (2) phthisis (no light perception [NLP] vision and confirmatory clinical exam/echography findings), or (3) enucleation, and the age at which these poor outcomes occurred was noted. Patients with ocular disease from other unrelated causes (eg, ocular trauma) were excluded from analysis.


Snellen BCVA for each eye and visit was converted to a logarithm of minimal angle of resolution (logMAR) value, and the following logMAR values for lower levels of vision, as described previously: counting fingers (CF; 1.9), hand motions (HM; 2.3), light perception (LP; 2.7), and no light perception (NLP; 3.0). The BCVA was plotted per patient by age at time of first and last visit. In addition, eyes were stratified by severity of vision loss based on the International Council of Ophthalmology visual standards, and the number per category was compared across decades of life.

To evaluate risk factors for poor outcomes (BCVA <20/200) for each eye, we analyzed the association between potential risk factors and visual acuity at the final follow-up visit. The following variables were coded as present or absent in each eye: anterior segment NV, cataract, CME, FTMH, glaucoma, peripheral NV, peripheral TRD, pigmentary changes in retina, presence of vitreous cell, sex (male), VMT/ERM, and VH. Five individuals had missing data for certain parameters in both eyes and were excluded, leaving 37 individuals (74 eyes) for analysis. To account for correlation in the error term between eyes of the same individual, we fit generalized estimating equations with a Firth-type penalty using the geefirthr package (version 0.1.0) for R (version 4.0.2; R Foundation for Statistical Computing), including a variable for each person. The Firth-type penalty reduces large estimates that arise when binary variables show separation or near-to-separation with the outcome. Although both eyes within a person are affected by the same genotype, treatment and other events can vary between eyes; therefore, we chose the independence covariance structure when fitting the model, reflecting our uncertainty of exactly how eyes might be correlated within a person. Standard error (SE) and 95% confidence intervals (95% CIs) (1.96 × SE) are reported for all estimates, and a P value of <.05 was considered statistically significant for all variables.

To evaluate the progression to NLP over time, we constructed Kaplan-Meier survival curves for time to NLP using the survminer package. We constructed a Cox proportional hazard model to compare the hazard rate between categorized baseline age using the coxph function from the survival package. Within the model, observations were clustered by patient ID. We computed the 95% confidence intervals using robust standard errors.


In total, clinical data from 121 affected individuals across 4 families (pedigrees shown in Figure 1 ) were reviewed. Of these, sufficient clinical information meeting inclusion criteria as above was available from 65 patients (45 female, 20 male) for baseline analysis (Supplemental Figure 1). Nearly all included patients were molecularly confirmed (97%; 63/65), and testing of the remaining 2 individuals was deferred because of their status as minors in known affected families. Patients were from 4 distinct families: family 1: [46 patients (32 female, 14 male), Arg243Leu (96% genetically confirmed)]; family 2 [12 patients (10 female, 2 male), Leu244Pro (100% genetically confirmed)]; family 3 [5 patients (2 female, 3 male), Arg243Leu (100% genetically confirmed)]; and family 4 [2 patients (1 female, 1 male), Lys250Asn (100% genetically confirmed)].


Pedigrees of 4 multigenerational families with autosomal dominant neovascular inflammatory vitreoretinopathy (ADNIV). A. Family 1: 7-generation family (Arg243Leu). B. Family 2: 7-generation family (Leu244Pro). C. Family 3: 6-generation family (Arg243Leu). D. Family 4: 6-generation family (Lys250Asn). Affected individuals are represented by filled-in symbols. Unaffected individuals are represented by open symbols. Deceased individuals are marked with a slash.


The mean age at first visit to our institution was 40.8 years (range 6-77 years). Of the 130 eyes from 65 patients analyzed at baseline, the mean visual acuity was LogMAR 0.59 (about Snellen 20/80), and the majority of patients (63.1%, 41/65) presented with excellent visual acuity (better than LogMAR 0.3 [Snellen 20/40]) in at least 1 eye. However, 14 patients (21.5%) had severe visual impairment on presentation (BCVA worse than logMAR 1.0 [Snellen 20/200]) in at least 1 eye. Of these, 9 patients (13.8%) had bilateral, severe visual impairment on presentation, and 5.4% of eyes (7/130) were phthisical or had been enucleated. Anterior segment NV (8.7%; 11/126) was relatively uncommon. About one-half of the eyes (62/122, 50.8%) had vitreous cells, and 20.3% (24/118) had cystoid macular edema. Vitreomacular traction or epiretinal membranes were seen in 27.9% of eyes (34/122), and TRD was seen in 10% of eyes (13/130). A summary of baseline characteristics is shown in Table 1 .


Demographic Data and Clinical Features of ADNIV Patients at Baseline and Follow-up

Baseline Follow-up
No. of patients (eyes) 65 (130) 42 (84)
Sex, n (%) M: 20 (30.8%) F: 45 (69.2%) M: 11 (26.2%) F: 31 (73.8%)
Age, y, mean (median; range) 40.8 (42; 6-77) 56.5 (55.5; 10-89)
OU BCVA (LogMAR): mean (median; range) 0.59 (0.18; −0.12 to 3) 1.48 (1.20; 0−3)
Characteristics, n/eyes (%)
Anterior neovascularization 11/126 (8.7%) 21/82 (25.6%)
Glaucoma 7/130 (5.4%) 27/84 (32.1%)
Cataract 28/130 (21.5%) 57/84 (67.9%
Presence of vitreous cells 62/122 (50.8%) 63/81 (77.8%)
Vitreous hemorrhage 13/124 (10.5%) 27/84 (32.1%)
Cystoid macular edema 24/118 (20.3%) 48/80 (60.0%)
Macular hole 5/124 (4.0%) 5/82 (6.1%)
Vitreomacular traction/epiretinal Membrane 34/122 (27.9%) 55/83 (66.3%)
Peripheral neovascularization 14/121 (11.6%) 23/81 (28.4%)
Pigmentary changes 67/121 (55.4%) 61/81 (75.3%)
Peripheral tractional retinal Detachment 7/122 (5.7%) 26/83 (31.3%)
Phthisis or enucleation 7/129 (5.4%) 15/82 (18.3%)

ADNIV = autosomal dominant neovascular inflammatory vitreoretinopathy, BCVA = best corrected visual acuity, LogMAR = logarithm of the minimum angle of resolution, OU = both eyes.


A total of 84 eyes from 42 patients (31 female, 11 male) had follow-up data available for longitudinal analysis. The mean duration of follow-up was 17.3 years (range 2-43.6 years; standard deviation 11.6 years). The mean BCVA at last follow-up visit was LogMAR 1.48 (Snellen 20/604) (range LogMAR 0-3). When evaluating BCVA by age, most patients had at least mild vision loss by the third decade of life (Figure 2) . The disease accelerated in the fifth decade of life, during which the majority of eyes went from normal vision or mild vision loss to at least moderate vision loss (20/70 Snellen equivalent). Many eyes showed a steep trajectory of vision loss to no light perception (29.8%; 25 eyes from 16 patients). A Kaplan-Meier curve for time to no light perception vision is shown in Figure 3 , A. Survival curves for NLP, categorized by age at baseline visit, are shown in Figure 3 , B. Individuals in the age range of >60 to 80 years at baseline showed a substantially higher hazard rate for NLP than did the youngest individuals (age range of 0-20 years at baseline; hazard ratio = 34.5, 95% CI 6.4-188.0; P < .001) ( Figure 3 , C).

Jan 3, 2022 | Posted by in OPHTHALMOLOGY | Comments Off on Long-Term Outcomes and Risk Factors for Severe Vision Loss in Autosomal Dominant Neovascular Inflammatory Vitreoretinopathy (ADNIV)

Full access? Get Clinical Tree

Get Clinical Tree app for offline access