Genetic phenotypic characteristics and inheritance patterns of patients with achromatopsia at a large academic institution and a review of the literature and gene therapies

• Introduction
• Methods
• Results
• Discussion

Abstract

Purpose: Achromatopsia (ACHM) is a cone dysfunction syndrome associated with low color vision, photophobia, and congenital nystagmus. Pathogenic variants in the CNGA3 and CNGB3 genes are the most common causes of ACHM. Identifying the underlying genetic etiology in patients with clinical findings of ACHM is critical to establishing the diagnosis and selecting targets for gene therapies. This study utilizes the inherited retinal disease (IRD) database at the University of Minnesota (UMN) to understand clinical features of achromatopsia, potential diagnostic methods, and genetic variability at one large academic institution.

Methods: This retrospective cohort study examined patients with a genetically confirmed diagnosis of ACHM evaluated at the UMN/M Health IRD clinic between May 2015 and August 2022. Data were collected through UMN’s electronic health record system and included information on demographics, history of ocular illness, visual acuity, and diagnostic exam results, such as electroretinogram, fundus autofluorescence, and optical coherence tomography (OCT) imaging. Genetic testing was performed using next-generation sequencing panels of varying sizes, and variants were interpreted according to American College of Medical Genetics guidelines.

Results: A total of 21 patients with a genetically confirmed ACHM were included in the study. Most patients were female (67%) and white (76%). The cohort included one pair of twin siblings and one group of three siblings. The median age of genetic testing conducted was 12 years (interquartile range, 12-20 years). One-third of patients had nyctalopia. All patients experienced difficulties with color vision, 20 of 21 (95.2%) had photosensitivity, and 18 of 21 patients reported congenital nystagmus. Diagnostic imaging showed most patients with a normal scotopic response and absent photopic response on electroretinogram, macular hyperfluorescence on fundus autofluorescence, and normal optical coherence tomography imaging.

Conclusion: This study contributes to an understanding of the phenotypes and distribution of genetic variants associated with achromatopsia at a single academic institution in the state of Minnesota, particularly those involving the CNGA3 and CNGB3 genes. Future research could explore the prevalence of these variants in larger cohorts to enhance clinical management and treatment strategies for patients with achromatopsia.

Achromatopsia (ACHM) is an inherited retinal disease (IRD) associated with reduced visual acuity from birth, low color discrimination, photophobia, pendular nystagmus, and a small central scotoma on visual field testing [1]. ACHM is the most common cone dysfunction syndrome, with a prevalence of 1/30,000; its pathophysiology involves defects to the cone photoreceptors, which process daylight and color vision [2,3]. In comparison to color blindness, which entails the loss of only spectral vision, ACHM impacts short, medium, and long cone photoreceptors, leading to complete reliance on rod function, also known as monochromatic vision [4]. Presentation often begins at birth with low visual acuity, congenital nystagmus, and a paradoxical pupil, or constriction of the pupil in the dark as opposed to the light, as some of the first findings [5]. On examination and imaging, ACHM presents with normal fundus, normal scotopic response with no photopic response, fundus autofluorescence (FAF), and optical coherence tomography (OCT) ranging from normal to hyper- or hypofluorescence and foveal atrophy, respectively [1,6].

Complete ACHM is inherited in an autosomal recessive pattern. Causative genes include cyclic nucleotide gated channel subunit alpha 3 (CNGA3, Gene ID: 1261), cyclic nucleotide gated channel subunit beta 3 (CNGB3, Gene ID: 54714), G protein subunit alpha transducin 2 (GNAT2, Gene ID: 2780), phosphodiesterase 6C (PDE6C, Gene ID: 5146), phosphodiesterase 6H (PDE6H, Gene ID: 5149), and activating transcription factor 6 (ATF6, Gene ID: 22926) [7]. Except for ATF6, all genes are involved in the cone photoreceptor signal transduction pathway. CNGA3 and CNGB3 encode the alpha and beta subunits for the cone cyclic nucleotide-gated (CNG) channel [8,9]. GNAT2 encodes the alpha subunit of the cone photoreceptor transducin [10]. PDE6C and PDE6H are subunits of the phosphodiesterase (PDE) protein [11,12]. CNG channels are located in the cellular membrane of the outer photoreceptor membrane. In dark conditions, guanylate cyclase converts guanosine triphosphate to cyclic guanosine monophosphate (cGMP), a ligand for CNG channels. High levels of cGMP allow Na⁺ and Ca²⁺ to enter the cell and keep it in a depolarized state. Glutamate is released to downstream second-order neurons such as bipolar and horizontal cells. When light enters the eye, individual photons bind with opsins on the cell membrane, causing a change in conformation. Opsin activation releases the G-protein transducin, which activates the PDE to mediate cGMP hydrolysis. Lower levels of cGMP cause the CNG channels to close, and the photoreceptor cell is hyperpolarized, leading to a signal being sent to the brain [7,13]. The absence or dysfunction of the genes encoding this process leads to dysfunction of the photoreceptor pathway. ATF6 is a protein involved in the unfolded cellular response, managing enzymes involved in protein folding in the endoplasmic reticulum (ER) [14]. This mechanism is present within many different cell types. It is thought that endoplasmic reticulum stress and damage during retinal development could lead to ACHM in patients with pathogenic variants in this gene, although the exact mechanism is still unclear. The prevalence of causative gene variants varies by population. CNGA3 and CNGB3 variants make up 90% of gene variants found in patients with ACHM, PDE6H and PDE6C make up <5%, GNAT2 represents <3%, and AFT6 variants are found in <2% of the population [7].

Electrophysiological and imaging methods for ACHM include full-field and multifocal electroretinogram (ERG), fundus photos, OCT, and FAF. Electrical activity can provide information on the functionality of different parts of the retina. On ERG, patients with ACHM are typically found to have absent or markedly reduced cone response with normal response of rods [15]. FAF uses the fluorescent properties of lipofuscin present within the lysosomal breakdown of photoreceptor segments to produce a photo of the retinal pigment epithelium (RPE). Retinal pathologies often cause RPE dysfunction and accumulation of lipofuscin in response, leading to abnormalities seen on FAF [16]. For patients with ACHM, FAF typically shows macular hyperfluorescence or an oval-shaped area of hypofluorescence at the fovea with a surrounding ring of hyperfluorescence [17,18]. OCT is used as a method to look at high-quality retinal images. In particular, the ellipsoid zone (EZ), the hyperreflective OCT band in the junction of the outer and inner photoreceptor, is a marker of photoreceptor health and mitochondrial function [19]. Findings on OCT range from normal to foveal atrophy and an absent or disrupted EZ band [18,20]. As ACHM is a cone dysfunction disorder, EZ band disruption begins centrally and extends outwardly in contrast to rod conditions, such as retinitis pigmentosa, which is characterized by peripheral loss of the EZ band [21].

New gene therapies are continuing to be developed to treat ACHM. One example is the utilization of viral vectors such as adeno-associated virus; a transgene, along with a cone-specific promoter, is injected in the subretinal space [22]. With the functional gene, cones can restore photoreceptor function. As the potential for gene therapies increases, it is critical to understand the genetic variety seen within populations, common methods of diagnosis, and phenotypic variability to allow for safer and more targeted treatments. This study utilizes the IRD database at the University of Minnesota (UMN) to understand clinical features of ACHM, potential diagnostic methods, and genetic variability in Minnesotan patients.

In this retrospective cohort study, patients with clinical features and confirmed genetic diagnoses of ACHM evaluated at the UMN/M Health IRD Clinic between May 1, 2015, and August 5, 2022, were included. Data were collected using UMN’s electronic health care record system, EPIC. REDCap, an online software to securely collect and manage data, was used to create a database of patients with IRD. This review was performed in accordance with the Health Insurance Portability and Accountability Act of 1996 and the Declaration of Helsinki. The institutional review board at UMN approved the study protocol. Written informed consent was obtained from all patients whose images were included.

Patients with a clinical diagnosis of ACHM, cone dystrophy, or other variations and confirmed genetic diagnosis of ACHM were included in the study. ACHM was considered genetically confirmed if patients had pathogenic/likely pathogenic variants of CNGA3, CNGB3, GNAT2, PDE6C, PDE6H, or ATF6. Patients without confirmatory genetic testing were excluded. A data collection form was created to record demographic information, history of ocular illness, and diagnostic examination results for each patient based on chart review. Data that could not be directly found through chart review were labeled as “unknown” and denominators reflect only cases with data present. Demographic questions consisted of age, sex, and self-identified race. History of illness included age of onset, visual acuity for both eyes, presence of visual field loss, and symptoms including photosensitivity, nyctalopia, and color vision loss. Visual acuity was stratified into Snellen scores of above 20/40, between 20/40 and 20/80, and below 20/80. Visual field loss was assessed using Goldmann field testing. Color vision was assessed by Ishihara color plates. Diagnostic electrophysiological and imaging studies included ERG, FAF, and OCT, all available in the IRD clinic. ERG testing was used to assess the degree of cone versus rod photoreceptor responsiveness. LKC (LKC Technologies, Gaithersburg, MD) and Diagnosys (Diagnosys LLC, Lowell, MA) were the two types of ERG used in the clinic. Tested values included dark-adapted response, light-adapted flicker, and light-adapted single-flash response. Each ERG device had its own range of reference values to assess a reduced or absent response. FAF allowed for analysis of macular or peripheral ring hyper- or hypoautofluorescence, assessing retinal damage. FAF was conducted using Optos ultrawide field imaging, which provides a 200-degree view of the retina. Spectralis OCT imaging was also used for both FAF and measurement of EZ band loss and retinal atrophy. Loss of EZ banding was manually measured using the embedded measurement tool. EZ band loss was considered a sign of structural retinal damage.

Genetic testing was offered to all patients during the IRD evaluation. Family members of patients were also invited to complete genetic counseling and testing at the clinic. Genomic data were analyzed by a variety of methods ranging from single-gene targeted testing to panels of over 260 genes reported to cause isolated or syndromic retinal disease, including cone-rod dystrophy or retinitis pigmentosa. For patients with known familial variants, single-gene testing was conducted. Testing was performed via next-generation sequencing. Genetic testing laboratories used included Invitae Laboratory (San Francisco, CA) and the University of Minnesota Molecular Diagnostic Laboratory (Minneapolis, MN). Outcomes of genetic testing included pathogenic or likely pathogenic (P/LP) variants for genes known to cause the phenotype, P/LP variants for genes associated with phenotypes not present in the patient, variants of uncertain significance (VUS), or no variants. Variants were interpreted by the respective laboratories according to the American College of Medical Genetics guidelines. Results were reviewed with a genetic counselor and ophthalmologist. Autosomal recessive pattern of inheritance, as expected for patients with ACHM, was confirmed by the presence of P/LP variants in two copies of the gene.

The UMN IRD database included 21 patients with clinically and genetically confirmed achromatopsia. We identified 16 patients with CNGB3-related achromatopsia and 5 with CNGA3-related achromatopsia. The cohort included two sets of siblings—one comprised of three individuals (Appendix 1, patients 370-372) and the other with two (Appendix 1, patients 380 and 381). Each group of siblings demonstrated the same variants as their family members. Most of the patients in our database self-identified as white. Previous studies of northern European populations, including one in the Netherlands and one in Hungary, noted a high prevalence of CNGB3 and CNGA3 variants, without a significant phenotypic difference [23,24]. Our cohort’s genetic distribution aligns with the expected genetic spectrum associated with these conditions. All patients in our cohort with CNGA3 dysfunction showed missense variants, one of which was a novel variant, p.Tyr443Cys, similar to trends seen in the literature, indicating little tolerance for substitutions in the functionality of this receptor [9,25,26]. Three related patients displayed a novel splice site variant, c.395+1G>A, in a highly conserved donor splice site consequence sequence. A previous study has noted the importance of splice site variants in the pathogenesis of CNGA3 and reported variants c.395+1G>T and c.395+9C>T, which were found to cause splice defects [27]. Splice defects can lead to frameshift variants, insertions, and deletions that may cause dysfunction in the final transcript of the gene. The likely pathogenicity of both novel variants warrants further investigation to confirm their role in achromatopsia and as a potential target for gene therapies. Another patient was found to have a CNGA3 variant currently classified as a VUS, c.1981C>T p.Arg661Cys. Given the clinical phenotype of ACHM and finding an additional pathogenic CNGA3 variant in this patient, these results were interpreted as diagnostic. In the CNGB3 gene, frameshift variants were found in all patients. In particular, the c.1148del (p.Thr383Ilefs*13) transcript is considered the most recurrent pathogenic variant noted in patients with autosomal recessive ACHM, with studies showing the prevalence of this variant in around 66% of patients with CNGB3 [28,29]. There were 15/16 patients in our cohort who demonstrated this variant, again highlighting its significance as a key target for gene therapies.

Presenting symptoms in our patient cohort included congenital nystagmus, diminished color vision, photosensitivity, low visual acuity, and paradoxical pupil. The lack of photopic response with an intact scotopic response on ERG was also a significant prompt for initiation of genetic testing. Most patients with available data displayed macular hyperfluorescence on FAF, suggesting increased levels of RPE lipofuscin due to RPE dysfunction. Previous studies have noted the finding of macular hyperfluorescence as a sign of early ACHM [18]. Notably, patients in our cohort with hypofluorescence on FAF also had a ring of hyperfluorescence surrounding it. One study further describes this as a “punched-out lesion” or an area of hypofluorescence surrounded by an area of hyperfluorescence that may be a signature finding of ACHM, specifically associated with increased age of the patient [17]. Additionally, the presence of predominantly macular changes versus peripheral changes typically seen on FAF for patients with retinitis pigmentosa is also notable as a unique feature [21]. The presence of hyper- or hypofluorescence did not appear to be correlated with vision.

OCT findings ranged from normal to foveal atrophy and EZ band disturbance. Most our cohort displayed a normal appearance of OCT imaging, without EZ band disruption. One patient with CNGA3-related ACHM and one patient with CNGB3-related ACHM had evidence of EZ band disruption. The patient with CNGA3-related ACHM had a P missense variant and a VUS (Appendix 1, patient 373). The patient with CNGB3-related ACHM had a frameshift variant and a splice donor site variant and was the oldest member of the cohort (Appendix 1, patient 375). Both patients were the only members of the cohort to show hypofluorescence on FAF. In the literature, EZ band disruption in ACHM has not shown clear associations with either age or genomic markers [30]. In comparison, patients with retinitis pigmentosa, for example, typically have initial peripheral loss of EZ banding with progressive loss until complete blindness. More severe forms of the disease are noted with X-linked retinitis pigmentosa [21]. Further research with a larger cohort would be helpful in solidifying relationships between genotypes and phenotypes in relation to imaging findings for patients with ACHM. Notably, the high proportion of patients with normal OCT findings and no photopic function suggests that foveal structure is largely maintained even after functional loss. As such, patients with ACHM may have a larger therapeutic window to receive gene therapies. However, close monitoring of patients is necessary as foveal atrophy and EZ band disruption, when it does begin, can be fatal for vision.

Initial gene therapy testing has been conducted in mouse, canines, and sheep models [31-33]. The presence of light-adapted ERG response after therapies is considered restorative. Thus far, it was noted that mice treated at older ages have a poor response to treatment [31]. Prompt genetic testing when there is clinical suspicion of ACHM can aid in early diagnosis and potentially increase a patient’s opportunities for gene therapies. Current registered clinical trials for ACHM have focused on CNGB3 and CNGA3-related ACHM due to their relatively higher prevalence in the population (Table 2). Replacements of CNGB3 and CNGA3 genes are delivered using an adeno-associated virus capsid through a subretinal injection [7,22]. As of June 2025, there are three active ACHM gene therapy clinical trials. Genetic sequencing can expand prior understanding of ACHM distribution and assist in the selection of genetic marker targets for clinical trials.

There are a few limitations to this study. This was a retrospective study with a small sample size and lack of available diagnostic and ocular history data for each patient. Due to the low incidence of ACHM, it was difficult to identify many patients from a single institution. Additionally, all family members of patients diagnosed with an IRD at the clinic are invited to undergo genetic counseling and testing. However, many family members do not choose to undergo genetic testing, likely due to the cost and the lack of additional clinical utility, which limits our available information regarding familial structure. These limitations should be considered when interpreting the results of this study.