Delayed‐onset cord1 progressive retinal atrophy in English Springer Spaniels genetically affected with the RPGRIP1 variant

Abstract

Objective

Cone‐rod dystrophy (cord1) is a form of progressive retinal atrophy. It is linked to an RPGRIP1 genetic variant which is the third most common canine disease variant thus far. While the variant affects various breeds, it is highly prevalent in English Springer Spaniels (ESSs). Yet its clinical and pathological implications remain equivocal. Herein, we study the retinal phenotype in ESSs genetically affected with the RPGRIP1 variant.

Animal Studied

Over 4 years, 494 ESSs (123 affected) were enrolled.

Procedure(s)

Owner‐perceived vision was collected via a questionnaire. Ophthalmic examination included fundus photography. In selected ESSs, retinal function and structure were assessed using electroretinography (ERG, 148 dogs) and optical coherence tomography (OCT, 4 dogs).

Results

Ophthalmoscopic changes included peripheral hypo‐reflective lesions often with distinct borders progressing centripetally culminating in generalized retinal atrophy. Cross‐sectional study revealed declining photopic ERG amplitudes with age in the affected group but not in controls. OCT indicated progressive photoreceptor loss. Despite ophthalmoscopic, ERG, or OCT abnormalities, most affected dogs were not visually impaired per their owners. In a fraction of afflicted ESSs, vision/globe‐threatening complications were documented including cataracts, lens luxation, and glaucoma.

Conclusions

In ESSs, the RPGRIP1 variant is associated with insidious pathology with delayed‐onset visual defects. The subtle phenotype without apparent visual deficit until the final years of life, if at all, may have caused underdiagnosis of cord1. Still, DNA testing remains informative, and ERG and OCT indicate progressive pathology. Peripheral fundus examination and photopic ERG are particularly useful for early detection and monitoring of cord1.

1. INTRODUCTION

Progressive retinal atrophy (PRA) consists of a group of inherited retinal diseases characterized by gradual loss of photoreceptor function and structure, ultimately leading to blindness. PRA is a term used in veterinary medicine and is considered homologous to retinitis pigmentosa in humans. ¹ Decades of clinical and genetic studies in a variety of canine breeds have revealed independent forms of PRA segregating in specific breeds, and molecularly associated with one of over 30 genetic variants identified to date. ² , ³ The vast majority of PRAs studied follow Mendelian modes of inheritance consistent with a monogenic etiology. ² , ³ Clinically, PRA may be categorized based on the age of onset when the decline in vision is first noticed and/or the disease is diagnosed. The age of onset generally reflects the causal genetic variant as well as the genomic background (e.g., breed) and is loosely described as early or late onset, allowing us to characterize, associate, or distinguish different forms of PRA. In early‐onset PRAs (e.g., rcd1‐PRA), clinical abnormalities may be detectable within the first few months of life and progress rapidly, while late‐onset PRAs (e.g., prcd‐PRA) may not become evident until mid‐ages and are often accompanied by slower progression.

Variants in the gene RPGRIP1 have been associated with a range of inherited retinal diseases in humans including juvenile retinitis pigmentosa, Leber congenital amaurosis (LCA6), and cone‐rod dystrophy. ⁴ , ⁵ , ⁶ , ⁷ RPGRIP1 encodes the protein RPGRIP1 (retinitis pigmentosa GTPase regulator‐interacting protein 1) which localizes at the photoreceptor connecting cilia contributing to photoreceptor ciliary transport, disc morphogenesis, and outer segment formation. ⁸ , ⁹ The absence of functional RPGRIP1 in mice has been associated with outer segment disc swelling and subsequent photoreceptor death. ⁹

Cone‐rod dystrophy 1 (cord1) was first described in Miniature Longhaired Dachshunds (MLHDs) as a form of PRA ¹⁰ associated with a homozygous 44 bp exonic insertion in RPGRIP1. ¹¹ This disease has a cone‐led phenotype in which reduction in cone function long proceeds that of rods. ¹² Affected dogs in the early disease stage show diminished photopic electroretinogram (ERG) along with normal scotopic ERG.

Structurally, truncation of cone photoreceptor inner and outer segments is observed at this early stage. ¹³ Paradoxically, the common initial clinical presentation of cord1 is nyctalopia without day vision deficits. Such discordance may suggest an asymmetric effect on the different types of photoreceptors, or a multifactorial pathologic pressure on cones enough to manifest in the photopic ERG phenotype but not in behavior. ¹⁴ As such, this disease has variably been referred to as PRA, ¹⁰ cord1, ¹¹ , ¹⁵ cord1/crd4, ¹⁶ or RPGRIP1‐CRD. ¹⁴

Subsequent studies in expanded populations of MLHDs found that genetically affected dogs (i.e., RPGRIP1 44 bp insertion homozygous, RPGRIP1 ^ins/ins ) showed a broad range of phenotypes suggesting the effects of modifiers on disease expression. ¹³ , ¹⁷ , ¹⁸ This led to the identification of a MAP9 variant ¹⁹ , ²⁰ , ²¹ and Locus 3 ¹⁴ as recessively inherited modifiers. If an RPGRIP1‐affected dog is homozygously affected for these additional variants, the disease results in earlier onset and faster progression.

Additional factors that may be affecting disease expression have also been described; complex splicing of RPGRIP1 leading to multiple protein isoforms including those that skipped the 44 bp insertion site ¹⁵ , ²² ; molecular slippage associated with the polyA tract within the 44 bp insertion leading to a readthrough, rescuing the premature stop codon. ²² Some or all of these factors were thought to contribute to the variable expression of cord1 in MLHDs.

Meanwhile, DNA testing for the RPGRIP1 variant has been available since the 2000s through multiple commercial sources as single variant tests, and more recently, as part of panel testing along with 250+ known variants associated with inherited canine diseases. Interestingly, studies from screening the known disease‐related genetic variants in dogs revealed that the RPGRIP1 variant was present across many diverse breeds. ¹⁸ , ²³ Furthermore, screening via the panel testing platform found that the RPGRIP1 variant had the second to third highest allele frequency among all canine traits, making it one of the most frequent ophthalmic traits in dogs across mixed and purebred dogs screened. ²³ , ²⁴ The widespread and highly prevalent nature of the RPGRIP1 variant across dog breeds highlights the need to clarify its clinical significance and the effect on canine populations.

The English Springer Spaniel (ESS) is one of such breeds previously found to carry the RPGRIP1 variant at relatively high frequency. ¹⁸ , ²⁵ Meanwhile, the clinical affliction of PRA has not been as prevalent in the breed with only late‐onset and mild clinical signs reported. ²⁵ Narfström et al. ²⁵ first looked into a limited number of ESS dogs affected with the RPGRIP1 variant, noting that affected individuals might not necessarily display clinical signs of disease. Hence, the need to further characterize this disease and to examine genotype–phenotype correlation in ESSs sets the premise of this study. According to the recently published consensus guidelines for the nomenclature of companion animal inherited retinal disorders, ²⁶ this disease may be described as PRA‐RPGRIP1(cord1)‐English Springer Spaniel. We will refer to the disease as cord1.

Herein, an extensive cross‐sectional evaluation was conducted across nearly 500 individual ESS dogs recruited for the study. In addition, several informative ESS lineages were followed over multiple years to investigate changes and progression of the retinal phenotype during the 4‐year study period. The purpose of the current study was to examine the retinal phenotype in RPGRIP1‐affected ESSs across ages and to better understand the genotype–phenotype correlation of cord1 in ESSs.

2. MATERIALS AND METHODS

2.1. Animals

Following a national call to ESS breeders and owners through the breed clubs, dogs were recruited to participate in the prospective study. The majority of the dogs were enrolled through eye clinics provided by the author (KM) at venues including national and local dog show sites within the United States. Other dogs were enrolled directly through the submission of blood and/or buccal cell swab samples along with data reporting through an online interface designed for the study. Both bench‐ and field‐bred ESSs were recruited for this study. In addition to the dogs that were examined, clinical data from historically documented ESSs with DNA samples archived at national repositories were acquired through the same online survey. Some of the owner‐provided data were retrospective data on deceased dogs and were cross‐checked where possible with available CERF/OFA eye examinations carried out by board‐certified ophthalmologists.

2.2. Owner‐reported visual behavior

The vision‐guided behavior of RPGRIP1‐affected ESS dogs was investigated using an online survey noted above or an equivalent questionnaire. The survey included questions regarding the dog's vision perceived by the owners to be recorded in four increments: normal, mild decrease (e.g., minor hesitation), moderate decrease (e.g., obvious difficulty navigating including collisions), to blind (e.g., complete vision loss night and day). To analyze meaningful data that could point toward the vision‐based age of onset, the latest age known to be normal or the age at which the first sign of visual deficit was noted was recorded. For categorical data analysis, response entries were grouped into four age brackets 0–3, 4–8, 9–11, and 12+ years of age. In parallel, using the same interface, retrospective information was collected on historical ESS dogs previously identified to be affected via DNA testing or based on parentage.

2.3. Ophthalmic examination

All dogs clinically examined for the study received at least one ophthalmic examination performed by the same board‐certified veterinary ophthalmologist (KM) where possible for consistency. In rare instances, examinations were done by other clinicians who reported the ophthalmic findings. Mydriasis was achieved via topical 1% tropicamide (Akorn, Gurnee, IL). ERG (detailed below) was carried out, if applicable, before the ophthalmic examinations or photography to avoid photobleaching which could affect the ERG responses. Slit lamp biomicroscopy was then performed to examine the eyes from the anterior segment to the anterior vitreous, followed by indirect ophthalmoscopy to examine the fundus. Upon detailed fundus examination, particular attention was given to the peripheral areas where the early changes in cord1 were typically observed. For peripheral examination of the retina, a 40D condensing lens was primarily utilized along with a 60D lens for increased field of view and a 2.2 pan retinal lens for magnification (Volk Optical, Mentor, OH). Finally, fundus photographs were captured in selected cases using the Genesis‐D Portable Retinal Camera (Kowa, Torrance, CA). For ESSs of Lineage 1 that underwent detailed phenotyping, fundus photography was carried out using the RetCam® Shuttle retinal camera (Natus Medical Inc., Pleasanton, CA) following topical anesthesia with 0.5% proparacaine (Bausch and Lomb, FL).

2.4. Electroretinography

The RETevet™ (LKC Technologies Inc., Gaithersburg, MD) handheld ERG device was used to assess retinal function. ERG was performed with the animal awake in sitting or sternal recumbency positions. Sedation or general anesthesia was not used to eliminate their negative effect on ERG amplitudes, ²⁷ , ²⁸ as well as for feasibility when performed outside of hospital settings. Where applicable, dogs received either a full ERG (i.e., both scotopic and photopic, ECVO 5‐step single‐flash protocol of the device) or photopic ERG only (i.e., cone function testing protocol of the device). Dogs undergoing a full ERG were placed in dark adaptation for at least 20 min prior to scotopic ERG. Scotopic ERG was recorded with increasing single‐flash stimuli at 0.01, 3, and 10 cd·s/m². All flashes were white light with no background lighting. Next, photopic ERG was recorded under a light‐adapted state. The photopic protocol consisted of 3 cd·s/m² white flash stimuli at 2 and 28.3 Hz to generate single‐flash and flicker ERG traces, respectively. All photopic tests had a background lighting of 30 cd/m². For each ERG step which itself was an average of multiple stimuli, 1–3 outputs were obtained per eye. Where multiple outputs from the same step were available, the highest amplitude of the two eyes was extracted for data analysis. The choice was made to circumvent artifactual within‐ and inter‐eye differences in amplitudes caused by the lower amplitudes from suboptimal recordings due to factors such as the dogs being awake. Following the ERG recording, the negative a‐wave and the positive b‐wave were automatically determined by the RETevet™ system, deriving the implicit time and amplitude for each peak. The correct assignment of the peaks was confirmed by visual inspection of the ERG trace. The trough‐to‐peak flicker ERG amplitude was also automatically calculated by the RETevet™ system.

2.5. Statistical analyses

All statistical analyses were performed using GraphPad Prism 10 (GraphPad Inc). Owner‐reported visual behavior was analyzed between the four groups of dogs (i.e., normal vision, mild decrease, moderate decrease, and blind as reported by owners) using a one‐way ANOVA, where p < .05 was deemed statistically significant. For each group, data from individual dogs were plotted and overlayed by a box plot that showed the median, mean, quartiles, and minimal/maximal age values. For both scotopic and photopic ERGs, after the amplitudes (μV) were plotted against the age (years) at the time of recording, a linear regression analysis was performed to determine the regression formulas and coefficients of determination (R ²), revealing the data trends. Mann–Whitney U test was applied to verify significant differences (p < .05) in the distribution of photopic ERG data between affected and control (i.e., carrier and clear) dogs.

2.6. Optical coherence tomography (OCT)

At the conclusion of the clinical examinations, OCT was performed on four affected dogs under general anesthesia. Dogs were premedicated via 0.2–0.5 mg/kg acepromazine (Boehringer Ingelheim Animal Health, St. Joseph, MO) subcutaneously prior to induction by IV propofol to effect and then inhalational isoflurane as anesthetic maintenance. Using an HRA + OCT Spectralis® machine (Heidelberg Engineering Inc., MA) with a 130° field‐of‐view lens, 55° infrared confocal scanning laser ophthalmoscopy (cSLO) images, and then OCT volumetric scans were taken spanning the fundus. Conjunctival stay sutures were placed at the limbus to acquire images of the far peripheral retina. High‐resolution OCT single scans were later analyzed across different time points to assess changes in outer nuclear layer (ONL) thickness. Furthermore, blue‐light autofluorescence (BAF) cSLO images were acquired to screen for additional lesions not detected through infrared imaging.

2.7. DNA collection

Whole blood samples were collected into EDTA tubes for DNA sampling. In some cases, cheek swabs (CytoSoft™ cytology brush, Medical Package Corporation, Camarillo, CA) were performed to collect buccal cells either by the owner or by the authors as an alternative or supplementary sample for DNA. Genomic DNA was extracted using routine spin‐column (QIAamp DNA mini blood kit, Qiagen, Germantown, MD), chloroform‐ethanol precipitation (Illustra DNA extraction kit, GE Healthcare, Chigaco, IL), or salt‐precipitation (Gentra Puregene buccal cell kit, Qiagen) methods. DNA samples were used to confirm the RPGRIP1 variant status and to search for disease modifiers via genome‐wide association study and whole‐genome sequencing in an ongoing study. Genotyping of the RPGRIP1 insertion variant was carried out as previously described. ¹³

3. RESULTS

3.1. Study population

A total of 494 ESS dogs (434 bench‐bred, 53 field‐bred, 7 field‐ and bench‐bred crosses) were enrolled in the study over 4 years (2019–2023) receiving at least one ophthalmic examination and/or their vision/ophthalmic examination status reported. Of the dogs examined by the author (KM), 148 dogs received ERGs; 58 dogs had both scotopic and photopic ERGs recorded while 90 dogs received photopic ERG only (Table 1). An additional 166 primarily deceased dogs had owner‐provided phenotype information of the dog's vision status. All 494 dogs prospectively enrolled were genotyped for the RPGRIP1 variant (123 affected, 220 carrier, 151 clear). Affected dogs examined ranged from 0.1 to 14.7 years of age (mean 6 years; median 5.7 years). The group of carrier and clear dogs ranged between 0.1 and 15.5 years of age (mean 4.1 years; median 2.9 years).

TABLE 1.

Number of dogs that received the different ERG testing protocols (n = 148).

	Scotopic and photopic		Photopic only
	One‐time	Repeated	One‐time	Repeated
RPGRIP1 affected dogs	34	7	33	9
Control dogs	13	4	45	3
Total	47	11	78	12

3.2. Owner‐reported vision status via an online questionnaire

Of a total of 164 online questionnaire entries, 123 dogs with confirmed RPGRIP1 genotypes were analyzed. These comprised 96 RPGRIP1‐affected dogs (n = 96; age 0.2–17.6 years) and 27 unaffected (clear or carrier) dogs (n = 27). As shown in the plotted graph of affected dogs (Figure 1A), those with reportedly “normal” vision are distributed across all ages while any vision deficits were only reported in the geriatric age range of 12 years and above. There was no significant difference in the distribution of age between the mild decrease (n = 4, mean ± SD = 14.1 ± 1.9 years), moderate decrease (n = 3, mean ± SD = 14.2 ± 0.8 years), and blind (n = 10, mean ± SD = 13.9 ± 1.4 years) groups, all averaging around 14 years of age, reflecting the age at the end of life. Indeed, half of the dataset represented deceased dogs where the owners reported the age and the vision statuses when the dog passed away, typically spanning 12–16 years of age. Across the geriatric population, the owner‐reported vision at the end‐of‐life timepoint was anywhere from normal to blind with variably reduced vision in between (Figure 1A). Note that owner‐reported vision from in‐person ophthalmologic examination by the authors was handled separately from the data acquired online as the detailed cross‐examination regarding subtle changes in vision often altered the initial impressions on vision reported by the owners.

FIGURE 1.

Owner‐reported vision in RPGRIP1‐affected ESS dogs. (A) Owner‐reported visual status of ESS dogs genetically tested affected (homozygous) for the RPGRIP1 variant. Dogs reported to exhibit visual deficits ranging from mild (minor hesitation/decline in night vision, blue box) or moderate (obvious difficulty navigating with night ± day vision, yellow box) decrease, to blind (complete vision loss, red box) were all within the old to geriatric ages (10+ years). In contrast, dogs reported to be of no vision issues (apparently normal vision, green box) were found across all age points including the geriatric age range (p ≤ .05, multivariate ANOVA). (B) Fundus scores based on the fundus photographs taken during eye examinations show the expected trend of dogs with lower cone flicker ERG amplitudes having higher fundus scores. A few outliers can be seen, especially within the group that received a fundus score of 3. The overlap in similar ERG amplitudes in dogs that received different fundus scores highlights the importance of a multimodal approach in assessing clinical progression.

3.3. Ophthalmoscopic changes in RPGRIP1‐affected ESSs

Ophthalmoscopic examinations were carried out with particular attention to the peripheral retina along the far superior tapetal‐non‐tapetal border. Fundus changes were scored based on the criteria described previously for the disease stages of cord1. ¹⁴ Briefly, the fundus was scored from 0 (normal), 1 (“incipient”: mild darkening and mottling of the peripheral tapetum), 2 (“early”: central extension of the dark mottling reaching the mid‐tapetum), 3 (“mid”: dark mottling and/or mild tapetal hyper‐reflectivity encompassing two‐thirds of the tapetum), 4 (“advanced”: hyper‐reflectivity and dark mottling encompassing nearly the entire tapetum), to 5 (“end‐stage”: marked generalized hyper‐reflectivity).

Increasing levels of fundic changes were observed in RPGRIP1‐affected dogs in relation to age (Figure 2D). Initial changes included variably demarcated, often subtle zones of dark discoloration or hypo‐reflectivity along the superior peripheral tapetal fundus (Figure 2A₁ : score 1) while the central fundus remained normal (Figure 2A₂ ). With disease progression, a clear border of darkened peripheral tapetum could be seen (Figure 2B: score 2). In advanced cases, there was a more generalized pattern of retinal atrophy with diffuse retinal vascular attenuation (Figure 2C: score 4). While there were some variations to the fundic lesions, the typical changes started as the darkening of the peripheral fundus which descended centrally toward the optic nerve head. Unlike other established forms of PRA where the observed fundic changes are uniform and generalized, cord1 shows a distinct centripetal development of retinal atrophy. Ultimately, the end stage (score 5) resembles the advanced phase of the generalized form of PRA although very few affected ESSs seem to reach this stage.

FIGURE 2.

Ophthalmoscopic findings of cord1 in RPGRIP1‐affected ESS dogs and the progression with age. (A: Score 1) Early changes include far peripheral zones of hypo‐reflectivity and hyper‐reflectivity (A₁). In this dog, the central fundus is within normal (A₂). (B: Score 2) With progression, a well‐demarcated zone of atrophy of hypo‐ or hyper‐reflectivity descends centripetally often with a distinct border. (C: Score 4) With advanced disease, hyper‐reflectivity and vascular attenuation become diffuse. Wide‐field retinal imaging using RetCam (B, C) allowed for a broad evaluation up to the far periphery of the fundus while routine indirect ophthalmoscopy with a 30D condensing lens (A) might miss subtle peripheral lesions. (D) Distribution of fundus scores according to the age range of RPGRIP1‐affected dogs shows an overall increased severity in score with ascending age group.

3.4. Electroretinographic changes in RPGRIP1‐affected ESSs

To examine the subclinical evidence of retinal atrophy and to quantify the reduction in retinal function associated with cord1, ERG was carried out in 83 RPGRIP1‐affected along with 65 unaffected control (49 carrier and 16 clear) ESSs. All ages were represented in both affected (age 0.9–14.7 years) and control (age 0.4–15.5 years) groups. There were 41 affected and 17 control dogs that received both scotopic and photopic ERG while 42 affected and 48 control dogs received photopic ERG only (Table 1). Repeat ERG at multiple time points were recorded in 16 affected and 7 control (6 carrier, 1 clear) ESSs. In the cross‐sectional analysis of RPGRIP1‐affected dogs across all ages, photopic cone ERG amplitudes showed a decreasing trend with age for both single‐flash and flicker recordings (Figure 3A). In contrast, unaffected controls showed relatively steady photopic ERG amplitudes across all ages. On regression analysis, the affected group followed a linear regression trend (single flash, $y=-4.01x+58.10,$ R ² = .47; flicker, $y=-4.80x+70.44$, R ² = .38), indicating a functional decline with age while no trend was observed in the control group (single flash, $y=-1.15x+63.93,$ R ² = .08; flicker, $y=0.22x+78.78$, R ² = .00).

FIGURE 3.

Photopic ERG responses across all ages in affected and control ESSs. (A) RPGRIP1‐affected dogs show decreased ERG amplitudes with increasing age, whereas unaffected (carrier or clear) dogs have steady ERG amplitudes at all ages. There is no notable difference in the distribution of the amplitudes between the carrier and clear ESS dogs which is consistent with the recessive segregation of the RPGRIP1 variant. (B) Cone single‐flash and flicker ERG amplitudes of affected ESSs that received multiple ERG indicate the overall trend of progressive decline at varying rates.

In the subset of affected dogs that had repeated ERG recordings, photopic ERG amplitudes invariably decreased with age (Figure 3B). In contrast, scotopic ERG amplitudes in the affected dogs showed a milder decline over time (Figure S1). Of the increasing light intensities used for scotopic stimulus, the highest light intensity (10 cd·s/m²) resulted in the steepest regression line, reflecting the increased contribution of cone photoreceptors with this mode of recording (Figure S1). Of note, among the unaffected controls, the distribution of normal ERG amplitudes did not distinguish carriers from clear dogs, supporting the recessive nature of this disease in that carriers do not develop a phenotype.

3.5. Case studies

A major goal of the current study was to investigate the full spectrum of the cord1 phenotype in ESSs spanning from the mildest (i.e., older dogs with minimal clinical abnormality) to the most severe (i.e., younger dogs afflicted with PRA). Our aim also was to identify subsets of RPGRIP1‐affected dogs that represented the extreme ends of the phenotypes which will facilitate mapping of modifiers. Modifiers may either rescue or exacerbate disease expression. Closely related ESSs that share the genomic backgrounds but express differing phenotypes can be particularly informative for mapping modifiers. To this end, we examined subsets of related ESSs that contained multiple affected dogs across generations or as littermates (Figure 4). Phenotypic findings from two lineages are described.

FIGURE 4.

Pedigrees of two ESS lineages in the multi‐year studies. (A) Pedigree tree of ESS lineage 1, containing RPGRIP1‐affected ESSs including an affected dam and her two litters of affected dogs. OCT imaging was performed in selected affected dogs alongside ERG. (B) Pedigree tree of ESS lineage 2 containing five littermates including three RPGRIP1‐affected dogs that received eye examinations and ERG.

3.5.1. Lineage 1 (OCT imaging)

A group of 11 closely and distantly related ESSs (7 affected and 4 carriers) was examined annually over three consecutive years (Figure 4A). All affected dogs received both scotopic and photopic ERGs while carriers received photopic ERG with or without scotopic ERG. In addition, OCT was performed in four affected dogs and repeated annually for a total of two (n = 2) or three (n = 2) imaging sessions. The reported vision of the affected dogs per the owner varied with a declining trend with age, where signs of nyctalopia such as hesitation under dim light conditions or preference for lit conditions were reported as early as 5 years of age. The two oldest dogs examined at 11.7 and 14.6 years of age were reported to be non‐visual. All affected individuals showed declining photopic ERG amplitudes (Figures 3B and 5A) while some showed characteristic fundus changes progressively (Figure 5A). In contrast, carriers showed little to no fundus or ERG changes. Cone ERG amplitudes were always greater than that of affected individuals at comparable ages. Notably, the reduced photopic ERG in the affected dogs was markedly lower than in geriatric control dogs (Figure 5B) suggesting that the decline in retinal function in the affected dogs cannot be explained by natural aging alone.

FIGURE 5.

Structural and functional discordance of cord1 phenotype in ESS dogs. Wide‐field fundus photographs (upper row) and corresponding cone flicker ERG waveforms (lower row) in selected RPGRIP1‐affected (A) and control (B) ESSs. (A) Left: Within normal vasculature, possible peripheral fundus changes (exaggerated with Retcam imaging and less obvious with indirect ophthalmoscopy) along with moderately reduced cone ERG at 4.2 years of age (ES193). Middle: Mild vascular attenuation, subtle hint of peripheral hypo‐reflectivity, along with mildly reduced cone ERG at 5.8 years of age (ES391, female). Right: A more advanced disease stage with severe vascular attenuation and diminished cone ERG in an older dog at 9.9 years of age (ES268, female). Note variable functional decline loosely correlating with age or fundus appearance. Also note that the early signs of fundic abnormalities in affected dogs could be subtle and often may only be validated later after progressive changes have developed from the initially subtle changes. (B) In contrast, control carrier dogs at 3.9, 11.2, and even 13.5 years of age exhibit within normal fundus and cone ERG amplitudes greater than those of affected ESS.

Infrared cSLO imaging was particularly useful in delineating the border of atrophic lesions that advanced centripetally over time (Figure 6) and identifying the developing lesions of retinal atrophy in the non‐tapetal fundus which could not be readily assessed by ophthalmoscopy (Figure 7E,F). OCT revealed reduced ONL thickness in all affected dogs first imaged at ages ranging from 3.3 to 9.9 years (Figure 7A,B, Figure S2). Subsequently, there was a gradual reduction in ONL thickness particularly at the periphery from which ophthalmoscopic signs of disease would begin (Figure 7C,D, Figure S2). Individual clinical findings are summarized in Table S1.

FIGURE 6.

Progression of retinal atrophic lesions are detectable via ophthalmoscopy and infrared cSLO. Sequential wide‐field fundus photographs (A, C) and infrared cSLO scans (B, D) of two RPGRIP1‐affected dogs shown for each right (A1–D1) and left (A2–D2) eye. (A, B) In this younger dog (4.8–6.6 years of age), expanding areas of increased reflectivity is visualized in the superior (yellow arrowheads) and inferior (red asterisk, contrasted by lesser affected region in green asterisk) fundi. The inferior lesions of retinal atrophy were undetectable via routine ophthalmoscopy and could only be visualized using infrared cSLO. (C, D) In the older dog (9.0–10.7 years of age) which is the dam of the younger dog, the expression of retinal atrophy in the superior fundus is more diffuse. An advancing border of retinal atrophy can be seen in the inferior fundus (blue arrows) only via infrared cSLO.

FIGURE 7.

Progressive ONL thinning via OCT and sub‐ophthalmoscopic retinal lesions are visualized by cSLO. (A–D) Sequential cSLO and OCT scans of an affected dog (ES269) over 4.9 (I), 5.8 (II), and 6.9 (III) years of age. (A) Sequential single scans of the superior central fundus. The yellow outline indicates the location of magnification in (B). (B) There is a gradual decrease in outer nuclear layer (ONL) (spanned by red lines) thickness. (C, D) Sequential cSLO and OCT scans of the superior (C) and inferior (D) peripheral fundus. The progressive collapse of ONL from the periphery corresponds to the border (purple arrow) of the lesion seen in the en face image on the left. With increasing age, there is separation of the posterior hyaloid membrane (yellow arrowheads) and increased signal in the pre‐retinal vitreous (*Mild, **Moderate). (E) Retcam (E1) and juxtaposed infrared cSLO (E2) images of an affected ESS (ES269) at 6.9 years of age. The infrared image allows visualization of the distinct inferior non‐tapetal atrophic lesions (red asterisk) contrasted by the lesser affected area (green asterisk). (F) Retcam (F1) and juxtaposed infrared cSLO (F2) and blue autofluorescence (BAF) cSLO (F3) images of an affected ESS (ES193) at 4.2 years of age. Infrared reveals coalescing islands of peripheral atrophy (blue arrows) while BAF scan reveals small areas of autofluorescence (yellow arrow heads) within those islands.

3.5.2. Lineage 2

A second lineage of directly related ESS dogs was examined up to three separate time points at approximately 1‐year intervals. A total of 11 dogs were examined of which six were littermates that included three affected, two carrier, and one clear dogs, providing an age‐matched cohort (Figure 4B). Functionally, the affected siblings showed varying severity in cone ERG reduction. Of the two affected dogs that were examined thrice at 5.5, 6.7, and 7.8 years of age, one had a more severe ophthalmoscopic abnormality and reduction in cone ERG (Figure 8A), while the other consistently showed milder phenotype both functionally and structurally (Figure 8B). A third affected littermate examined twice at 6.7 and 7.8 years of age was the least affected, having no ophthalmoscopic lesions but with reduced cone ERG (Figure 8C). Over time, ophthalmoscopic disease progression was minimal in the affected dogs over the 3‐year time span. However, there was a trend of mild reduction in cone ERG amplitudes despite their stable fundus appearances, indicating that cone ERG is more sensitive in identifying disease progression than ophthalmoscopy. Interestingly, none of the three affected dogs were reported to have noticeable visual deficits. Still, cone flicker ERG was consistently lower in all affected dogs compared to the unaffected littermates (Figure 8D,E) at all age points indicating the insidious development of retinal atrophy progressing subclinically.

FIGURE 8.

Variable phenotypic severity and progression in affected and control ESS littermates. Fundus photographs (A₁–C₁) and cone flicker ERG traces (A₂–C₂, D, E) of five littermates examined at 5.5, 6.7, and 7.8 years of age. (A–C) Representative ophthalmoscopic images and cone flicker ERG traces are shown for the three RPGRIP1‐affected dogs. While progression assessed via ophthalmoscopy is limited, the declining trend of cone ERG amplitude over time is more evident. (D, E) Cone ERG in the controls (normal ophthalmoscopy, not shown) consistently exhibited higher photopic response compared to the affected littermates at all ages.

3.6. Ocular complications in afflicted RPGRIP1‐affected ESSs

Among the RPGRIP1‐affected dogs examined, cataracts of varying degrees were common findings, particularly in cases where ophthalmoscopic signs of retinal atrophy were also present. While cataracts (n = 13) were found to develop in any location of the lens, some early stage cataracts were often found to develop in the posterior lens often exhibiting vacuolar changes (Figure 9A,B). Vitreous changes such as vitreous degeneration or asteroid hyalosis could also be visualized in the affected dogs (Figure 9B).

FIGURE 9.

Ocular complications observed in RPGRIP1‐affected ESSs. (A) Anterior (A1) and posterior (A2) cortical cataracts in an 8.2‐year‐old affected ESS (ES182, left eye) with diminished cone ERG but no visual deficit reported. Note that A1 and A2 are the same eyes captured at the same examination but focused on the lens anterior and posterior, respectively. The inserts are magnified images of the cataract details. Bilateral findings. (B) Posterior cortical cataracts and nuclear sclerosis (B1), asteroid hyalosis (B2), and generalized retinal atrophy (B3) in a 12.3‐year‐old affected ESS (ES245, left eye) with advanced PRA. This dog started to lose vision at 9 years of age. Bilateral findings. (C) Anterior lens luxation of a hypermature cataract in the right eye (C1) of an 11.9‐year‐old affected ESS. The contralateral eye exhibited signs of mild to moderate retinal atrophy (C2).

Furthermore, based on ophthalmic examinations, clinical records, or surveys submitted by the owners, a number of affected ESSs presented with ocular complications some of which were found to compromise the animals' quality of life. These included lens luxation (n = 7) (Figure 9C) and glaucoma (n = 5) which often required enucleation. A historical affected dog was documented to have suffered from painful sequelae following affliction with PRA, having developed cataracts and then glaucoma, leading to bilateral enucleation at 13 years of age. This dog produced an affected dog in the current prospective study which was deemed afflicted with PRA at 12 years of age based on ophthalmoscopy, and becoming blind by 13 years of age with no ocular complications.

4. DISCUSSION

In the current study, the phenotypic spectrum found in the RPGRIP1‐affected ESSs ranged from no observable phenotype to complete blindness, with a positive correlation between age and the development of disease phenotypes (Figures 1 and 3). The earliest sign of cord1 is changes in cone ERG amplitude which gradually deteriorates over the typical ESS lifespan of 12–14 years as suggested by the cross‐sectional analysis. Other clinical changes assessed by OCT, ophthalmoscopy, and behavior develop subsequently.

Dog DNA tested as homozygous for the RPGRIP1 variant is labeled as genetically “affected.” However, only a subset of the affected dogs seem to develop obvious signs of PRA and are additionally referred to as clinically “afflicted.” Some dogs that were reported with “normal” vision may in fact had some visual deficit that was not evident to the owner due to compensation or the environment that masked detection. In these dogs, abnormalities may become apparent with a thorough ophthalmic examination including ERG. Meanwhile, we have previously shown that RPGRIP1‐affected dogs with flat photopic ERG could still navigate normally at all light levels as tested in a standardized obstacle course up to 5 years of age (Das et al. 2017), highlighting the elusiveness of diagnosing cord1 clinically. It is therefore critical to perform a comprehensive assessment including the history of vision, ERG, and a thorough fundic examination.

Unlike most forms of PRA where the ophthalmoscopic changes are relatively diffuse and generalized, the initial changes in cord1 are limited to the peripheral fundus while the central fundus remains normal. The peripheral changes often develop a distinct border of atrophy that progresses centripetally, giving the appearance of a dorsal “drape” of hypo‐reflectivity descending over the normal tapetal fundus. As such, the initial fundus changes in cord1 may only be visualized along the superior peripheral tapetal‐non‐tapetal junction. Therefore, a thorough ophthalmoscopic examination of the far peripheral fundus is critical for identifying the early stages of cord1. Pharmacologic mydriasis is necessary, and the usage of higher diopter condensing lenses with a greater field of view can be helpful to evaluate the peripheral fundus fully. We found that the subtle peripheral hypo‐reflective changes were best appreciated by tilting the lens back and forth.

Cord1 is a cone‐rod dystrophy in which cone function declines first. Whenever possible, comprehensive scotopic and photopic ERGs should be performed for a complete functional assessment of both rods and cones. This allows for the evaluation of the disease stage and ongoing monitoring of disease progression. However, if a comprehensive ERG cannot be pursued, photopic ERG will be the single most informative diagnostic to assess cord1, in addition to a thorough fundic examination and DNA testing of the RPGRIP1 variant. All the dogs we performed ERGs in the study minimally received photopic ERGs for screening of early disease and for monitoring of disease progression. We found that performing both the single‐flash and 29 Hz flicker ERGs was informative and complementary. The flicker ERG amplitudes were more consistent between repeated recordings within the same session. Still, the cross‐sectional ERG amplitude plotted against age showed that the single‐flash ERG values stayed closer to the trend line (R ²  = .47), indicating that it can reliably inform the trend. Ultimately, performing both single‐flash and flicker photopic ERGs would help determine the dog's place on the cone phenotypic scale.

To examine if ERG amplitudes are impacted by natural aging, ²⁹ carrier and clear control ESSs of all ages including those with advanced ages over 10 years underwent ERG for comparison. In the controls, photopic ERG amplitudes remained constant except for a subtle declining trend in the single‐flash ERG. In contrast, the RPGRIP1‐affected group showed a much steeper progressive reduction in cone ERG amplitudes, leaving limited measurable amplitude beyond 10 years of age. While the factor of advanced age could exacerbate cord1, the overwhelming cord1 phenotype seemed to diminish any age‐related effects. Of note, while cord1 in ESSs may be considered a much delayed late‐onset PRA and one may question its significance, the impact of vision loss on the dog's quality of life in the final years of life when other senses and functions are also declining is not trivial.

The use of OCT allowed the identification of ONL thinning before ophthalmoscopic changes were evident. In addition, cSLO imaging revealed the early changes in retinal atrophy as well as lesions in the non‐tapetal fundus not readily visualized by ophthalmoscopy. In some cases, lesions that are not visible on fundus examination can only be visualized with infrared cSLO. In the affected dogs from Lineage 1, the dark peripheral lesions were “highlighted” under infrared. This is particularly helpful in examining the advancing border of retinal atrophy in the far inferior periphery. Other imaging modalities such as blue‐light autofluorescence imaging also revealed lesions that were not visible by routine ophthalmoscopy.

The ESS consists of two subtypes of distinct features: bench‐ and field‐bred ESSs. The majority of the ESSs examined in this study were bench bred. Interestingly, the RPGRIP1 variant was identified in both types of ESSs but less frequently in the field‐bred ESSs. The differences in the prevalence of the RPGRIP1 variant between the two ESS subtypes could simply be a consequence of population bottleneck and subsequent founder effect (i.e., genetic drift). Meanwhile, although ESSs are not considered to be part of the sighthound group, selective breeding pressure may have played a role in the field‐bred lines due to the functional demand to visualize the game and the hunter in day light. Importantly, the bench and field ESSs may be interbred, meaning that the gene flow between the two groups are ongoing. Therefore, there is a need to continue monitoring the disease in both types of ESS.

Histopathological analysis was not performed in the current study of privately owned ESSs. Nevertheless, cross‐sectional images of the retinal cell layers captured via non‐invasive OCT in the affected dogs provided insights into the attenuating ONL (i.e., nuclei of photoreceptors) over time as we and others have previously shown in research colonies of MLHD ³⁰ and their descendants. ¹³ , ¹⁴ In future studies, there is a need for further morphological characterization of this disease in ESSs on a cellular level.

The age at which genetically affected dogs are recognized to be clinically afflicted may vary depending on the timing and means of assessment (ERG vs. ophthalmoscopy vs. behavior) which should be noted. As discussed above, photopic ERG is the most sensitive modality to evaluate cord1 in the early phase of the disease and can be used to establish the baseline cone function in known affected dogs. Given the phenotypic variability in this disease, ERGs should be repeated as often as annually to track disease progression. In this study, 84% of dogs were examined once, contributing to the cross‐sectional study. In addition, repeat examinations in a subset of affected (n = 16) and control (n = 7) dogs have provided insights into the variable rates of disease progression. To better understand the individual variation of cord1 disease expression, a longitudinal study of a large group of affected ESSs is needed.

PRAs have generally been considered a group of blinding but quiet diseases without further complications other than cataracts. Of critical interest that could impact how aggressively cord1 should be controlled genetically as well as managed clinically are the ocular complications which could be secondary to retinal atrophy. It is known that the intraocular microenvironment may be changed by the inflammatory reaction of retinitis pigmentosa in people. ³¹ Toxins released from dying photoreceptor cells can lead to inflammatory and degenerative changes in the eye, including vitreous degeneration and cataracts. ³² Cataracts observed in the affected ESSs showed the tendency to begin posteriorly while some affected dogs presented with asteroid hyalosis without cataracts. The posterior–anterior pattern of these ocular complications suggests a retinal origin of the sequelae, although further investigation is needed to confirm the causal link with cord1, any triggers, and preventative options. The protracted nature of retinal atrophy of cord1 in ESSs could be a factor in the development of ocular complications not typically seen with other PRAs.

Our findings indicate that the RPGRIP1‐affected ESSs undergo pathophysiologic changes consistent with slowly developing PRA. Modifiers that could separate the more severely afflicted cases from the subclinical cases of the same age and are being investigated in our ongoing study. Identification of such modifiers can be a therapeutic target as well as used for selective breeding toward a milder phenotype while avoiding an abrupt elimination of the highly prevalent RPGRIP1 allele thus preserving the diverse ESS gene pool.

5. CONCLUSION

Routine ophthalmic examination of ESSs affected with the RPGRIP1 variant including ophthalmoscopy alone often fails to capture the early changes in cord1 in the far peripheral fundus. Additional diagnostic modalities including ERG and OCT demonstrate that most if not all genetically affected dogs will invariably be afflicted. In particular, photopic single‐flash and flicker ERGs revealed reduced cone function in apparently normal genetically affected dogs. In the current cross‐sectional study, examination of the affected ESSs at variable ages indicated that there is a progressive disease phenotype with advancing age. However, the disease severity varied among age‐matched affected dogs across any given age. At present, it is not possible to pinpoint whether an affected dog will end up on the mild or severe end of the disease spectrum. Ongoing effort is focused on identifying modifiers that can predict disease severity. Based on detailed phenotypic characterization, the RPGRIP1 variant is found to be associated with a pathologic process most easily assessed by photopic ERG. DNA testing of this variant is hence still recommended in ESSs to slowly move away from the variant over time. However, given the high allele frequency, it is not possible and further, not recommended to abruptly eliminate all affected dogs or even carriers from the breeding pool. Further studies are required to demonstrate the molecular players impacting the variable disease expression and their long‐term and population‐wide consequences.

AUTHOR CONTRIBUTIONS

Jennifer C. Kwok: Data curation; formal analysis; investigation; project administration; writing – original draft; writing – review and editing. Yu Sato: Data curation; formal analysis; investigation. Jessica K. Niggel: Data curation; formal analysis; investigation; writing – review and editing. Emma Ozdogan: Formal analysis; investigation. Leonardo Murgiano: Investigation; writing – review and editing. Keiko Miyadera: Conceptualization; data curation; formal analysis; funding acquisition; investigation; methodology; project administration; resources; supervision; validation; writing – review and editing.

FUNDING INFORMATION

American Kennel Club – Canine Health Foundation (grant # 02572‐MOU), National Eye Institute/NIH (grant # R01 EY‐006855).

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

This study was conducted following the recommendations of the Association for Research in Vision and Ophthalmology (ARVO) Resolution on the Use of Animals in Ophthalmic and Vision Research. Anesthetic procedures were conducted under the host institute's IACUC protocol (#806301). Care was taken to pre‐screen the dogs for health implications prior to anesthetic procedures and to avoid any potential discomfort during the process.

Supporting information

Data S1

Kwok JC, Sato Y, Niggel JK, Ozdogan E, Murgiano L, Miyadera K. Delayed‐onset cord1 progressive retinal atrophy in English Springer Spaniels genetically affected with the RPGRIP1 variant. Vet Ophthalmol. 2026;29:e13290. doi: 10.1111/vop.13290

DATA AVAILABILITY STATEMENT

Data available on request from the author.