Utility of No-Charge Panel Genetic Testing for Inherited Retinal Diseases in a Real-World Clinical Setting

Introduction

Inherited retinal diseases (IRDs) are among the most common causes of progressive bilateral vision loss. Historically, the management of IRDs has involved obtaining clinical genetic testing if available, offering genetic counseling, and managing complications such as macular edema. However, genetic testing is not routinely performed, in part because of insufficient access, the impracticality of testing, and the lack of understanding the diagnostic value of testing.

Over the past decade, interest in genetic testing has increased as a result of newly available gene therapy (eg, voretigene neparvovec-rzyl [LUXTURNA], an RPE65 gene therapy) and other emerging gene therapies in clinical trials.^1,2 Testing technologies have rapidly evolved, ranging from single-gene detection to panel testing for multiple genes with next-generation sequencing, whole-exome sequencing (WES), or whole-genome sequencing (WGS). ³ Although testing options continue to increase, the cost of genetic testing for IRDs, in particular panel testing, is rarely covered by health insurance, possibly because of the perception that the results might not immediately affect clinical management or the long-term prognosis.

Recently, 2 independent, no-charge panel genetic testing programs for IRD were developed, providing a means by which retinal specialists can offer genetic testing at no financial cost to patients. They are the My Retina Tracker program provided by Blueprint Genetics and sponsored by The Foundation Fighting Blindness and the ID Your IRD program provided by Invitae Corp and sponsored by Spark Therapeutics. Implementation of these programs has substantially increased the rate of and access to testing, suggesting that patients and physicians are increasingly interested in genetic testing. ⁴

The ID Your IRD program provides a panel testing kit to physicians for patients who desire testing. It assays for the presence of more than 200 IRD-associated gene mutations and provides a streamlined process of sample acquisition and results reporting. However, many retina specialists are still unfamiliar with the diagnostic value of this approach. Several studies have reported yields of genetic diagnosis with a variety of testing platforms, some within highly idealized environments, such as in an academic research facility. These studies report a wide range (25%-76%) of diagnostic yield.^{4 –12} These reports might not reflect the expected yields or challenges of testing in the setting of everyday clinical practice. The present study aimed to show the clinical utility and diagnostic yield of no-charge panel genetic testing from the ID Your IRD program at a single academic center.

Methods

The Institutional Review Board, University of Oklahoma Health Sciences Center, approved this study before its commencement. The research adhered to the tenets of the Declaration of Helsinki.

All patients were seen by the same clinician (S.Y.L.) in the Retina Clinic, Dean McGee Eye Institute, from August 1, 2019, to April 31, 2021. A clinical diagnosis of IRD was judged by the clinician based on family and historical data, a clinical examination, Goldmann perimetry, and multimodal imaging analyses including color fundus photography, fundus autofluorescence photography, spectral-domain optical coherence tomography, and an electroretinogram test. Any patient with a questionable diagnosis (eg, autoimmune retinopathy, age-related macular degeneration) or with potential confounders (eg, long-term history of retinotoxic medications such as pentosan polysulfate or hydroxychloroquine) were excluded. A clinical diagnosis of IRD was defined as a single most probable disease and/or assigned into subcategories of IRD, including nonsyndromic or syndromic retinitis pigmentosa (RP), cone or cone-rod dystrophy, rod-predominant dystrophy, macular disease (eg, Stargardt, pattern and vitelliform dystrophies), and other retinopathies (eg, choroideremia, X-linked retinoschisis) according to a well-established IRD classification scheme. ¹³

Once the most likely clinical diagnosis was determined, patients were offered genetic tests by saliva sampling using no-charge panel testing by the Invitae IRD panel. The panel initially used in this study tested for 248 gene mutations and was later upgraded to testing for 293 mutations in 2020 (Figure 1). Those who opted to undergo testing provided written informed consent. Results were provided by the testing panel in a comprehensive report of detected mutations along with interpretations of pathogenicity based on a proprietary variant classification algorithm. ¹⁴

OPEN IN VIEWER

Figure 1.

Genes tested in the Invitae inherited retinal disorders panel.

Genetic testing results were retrospectively reviewed and categorized as positive when pathogenic mutations were identified and undetermined or negative when pathogenic mutations were not identified. Undetermined results included cases in which only a single recessive pathogenic mutation was identified for a clinical disease known to have autosomal recessive inheritance or in which only variants of uncertain significant (VUS) were identified. Negative results included cases in which no mutations of any kind were detected. All pathogenic mutations and VUS from each patient were reviewed, and the published literature was reviewed to confirm that mutations designated as pathogenic have been shown to cause phenotypes commensurate with the patient’s clinical diagnosis.

Results

Demographics

Post hoc quality control excluded 5 results from inclusion in data analyses. Of the 59 patients included in the study, 39 were tested for 248 mutations and 20 were tested for 293 mutations. The mean age of the 34 women (57.6%) and 25 men (42.4%) was 47.5 years (range, 16-77 years). Twenty-eight patients (47.5%) self-identified as non-Hispanic White, 5 (8.5%) as American Indian or Hispanic, and 4 (6.8%) as African American or Asian; 13 patients (22%) declined to specify their ethnicity.

Table 1 shows the subcategories of IRD in the present study. The most common was nonsyndromic RP followed by macular dystrophies.

OPEN IN VIEWER

Table 1.

Subcategories of Disease in the Present Study.

Classification of IRD/Diagnosis 13	Number (%) a
Nonsyndromic retinitis pigmentosa	28 (47.5)
Syndromic retinitis pigmentosa	8 (13.6)
Cone and cone-rod dystrophy	4 (6.8)
Rod-predominant dystrophy	1 (1.7)
Macular disease	13 (22.0)
Stargardt disease	9 (15.3)
Pattern dystrophy	2 (3.4)
Vitelliform dystrophy (including Best)	2 (3.4)
Other retinopathy	5 (8.5)
Choroideremia	3 (5.1)
X-linked retinoschisis	2 (3.4)

Abbreviation: IRD, inherited retinal disease.

a

Percentages depicted as proportion of the total included in the study population (N = 59).

Diagnostic Yield

Genetic testing results were returned after a mean of 20 days (range, 14-28 days); this included sample shipping time. The overall rate of obtaining positive genetic testing results (ie, identifying disease-causing mutations) was 50.8% (30/59). Figure 2 shows the breakdown of the type of positive results. Testing yielded undetermined genotypes in 47.5% of patients (28/59) as a result of 3 possible scenarios: (1) detection of a single recessive pathogenic mutation, (2) detection of 1 recessive pathogenic mutation and at least 1 VUS, or (3) detection of a VUS only. One patient (1.7%) had negative testing results with no mutations or VUS identified.

OPEN IN VIEWER

Figure 2.

Testing yield of Invitae inherited retinal disease panel. Percentages represent proportion of all 59 patients. Neg, negative; rec mut, recessive mutation; RP, retinitis pigmentosa; VUS, variant of uncertain significance. Bold outline shows the proportion of cases solved or with positive diagnostic yield from genetic testing.

Rates of obtaining a genetic diagnosis varied further by disease subcategory. Figure 3 shows the positive test rates by subcategory. Higher rates of detection of a pathogenic mutation were found for specific clinical diagnoses, with 100% (3/3) detection in patients with choroideremia and a 55.6% (5/9) detection in patients with Stargardt disease.

OPEN IN VIEWER

Figure 3.

Rates of positive testing results by disease category. Other, other retinopathies including choroideremia and X-linked retinoschisis; RP, retinitis pigmentosa; Solved, confirmed diagnosis.

Overall, 17 genes with disease-causing mutations were identified across the study population. Mutations in USH2A and ABAC4 (each representing 16.7% or 5/30 of all pathogenic mutations identified) were the most frequently identified disease-causing mutations (Table 2). In patients with syndromic or nonsyndromic RP, the most frequent genes in which mutations were identified were USH2A, CLN3, RHO, and RP2. Altogether, mutations in these genes accounted for 57.9% (11/19) of the disease-causing mutations identified in patients with clinically diagnosed RP, cone or cone-rod, or rod-predominant dystrophies. Mutations in ABCA4 accounted for the majority (4/7; 57.1%) of the disease-causing mutations identified in patients with clinically diagnosed macular dystrophies. As expected, pathogenic mutations in CHM were identified in all 3 cases of choroideremia.

OPEN IN VIEWER

Table 2.

Frequency of Identified Disease-Causing Gene Mutations.

Genes	Number a (%)
ABCA4	5/59 (8.5)
USH2A	5/59 (8.5)
CHM	3/59 (5.1)
RP2, CLN3, or RHO	2/59 (3.5)
ADGRV1, BBS1, BEST1, CABP4, CEP83, EYE, NRL, PROM1, PRPH2, RP1, or RS1	1/59 (1.7)

a

Numbers in column 2 identify how many patients exhibited a pathogenic mutation in the relevant gene.

At least 1 VUS was present in 55 patients (93.2%). In the study population with positive and undetermined testing results, 326 VUS were identified across 250 genes. The number of VUS ranged from 1 to 13 in patients with positive testing results and from 2 to 8 in patients with undetermined testing results.

Discussion

Approval of voretigene neparvovec-rzyl gene therapy for RPE65-mediated IRD and several other investigative gene-based treatments have prompted increased attention to genetic testing for patients with IRD.^1,2 Accordingly, the recommendation from the American Academy of Ophthalmology for genetic testing in patients with IRDs changed from avoidance of routine testing in 2012 to offering testing for patients with clinically suggested diseases with Mendelian inheritance in 2014. ¹³

Options for genetic testing continue to increase for patients with IRD. However, there is a wide range of cost ($120-$7000), turnaround time (1 week-24 weeks), and diagnostic yield (20%-70%) depending on the testing method of choice. Considerations of whether and how to use genetic testing for IRD patients are complicated by the lack of access to testing and a lack of understanding of the diagnostic value of testing. No-charge panel testing programs, such as the ID Your IRD program, have improved access and participation in testing. The present study aimed to provide an estimated likelihood of obtaining confirmatory genetic diagnosis using this testing approach.

Several others have reported diagnostic yields of panel testing, some using earlier versions of kits provide by the RD Your IRD program. Li et al ⁶ reported a yield of 25% in 48 patients with nonsyndromic RP using a 31-gene panel testing kit via the program. In their study, disease-causing mutations were identified in 6 genes, with dominant mutations in RHO and recessive mutations in USH2A being the most common. Using a 248-gene panel test, Seraly et al ¹² found a 28.2% yield in a study of testing on 46 patients, while McGowan et al ¹⁵ reported a 48.6% detection rate in a 37-patient cohort using a 31 gene-panel test. Invitae has successively updated its testing panel from 31 genes to 328 genes (at the time of this writing) as more information emerges on genetic variants associated with IRDs. Using the 248-gene and 293-gene panel, we obtained a genetic diagnosis in 30 of 59 patients, representing a sensitivity of 50.8%. Continued expansion of the ID Your IRD gene panel and of other available panel testing kits will presumably increase the probability of detecting disease-causing mutations.

Published rates of diagnosis for genetic testing depend not only on the methodology of testing but also on the disease categories represented and classification schemes selected in each study. A previously reported large cohort study (1000 consecutive families) included patients with photoreceptor diseases, macular diseases, and so-called “third-branch” disorders. ⁹ As Stone et al ⁹ suggested, the large diagnostic categories within this clinical classification system are not a standardized system by which to classify the clinical diagnosis of IRD and differ from the IRD classifications represented here.

Nevertheless, it might be helpful to compare independent cohorts comprising heterogeneous clinical entities of IRD classified by alternate methods because the diagnostic yield of testing can be influenced by the selection of patients or by the proportion of represented cases attributable to known disease genes. ⁹ For example, peripheral photoreceptor diseases, including syndromic RP and nonsyndromic RP, often display a wide genotype-to-phenotype spectrum of diversity; therefore, the yield of confirmatory diagnosis can be lower for these than for macular dystrophies including Stargardt or for more stereotyped disorders such as choroideremia.

If the present study had selected an IRD classification system similar to that published by Stone et al, ⁹ we would have observed a 46.3% solve rate (19/41) in patients with peripheral photoreceptor diseases such as syndromic and nonsyndromic RP, a 46.2% (6/13) solve rate in patients with macular diseases, and an 80.0% (4/5) solve rate in patients with third-branch disorders, such as choroideremia and X-linked retinoschisis. Overall, the 50.8% rate of diagnosis in our study would have remained consistent regardless of the use of an alternate disease-classifying system. Stone et al ⁹ reported a higher rate of pathogenic mutation detection (76%; 760/1000) than in our study; however, attention must be paid to the differences in the testing platform. In the Stone et al study, disease-causing mutations were identified in 57.6% of patients (576/1000) using next-generation sequencing, in 18.2% (182/1000) using WES, and in 0.2% (2/1000) using WGS.

There is no single ideal genotyping strategy to identify disease-causing mutations because of the different strengths and weaknesses of each approach. However, panel testing appears to be a practical initial approach for genotyping for IRD patients in the clinical setting. It provides a reasonable diagnostic yield when there is limited access to extended genetic testing modalities, such as WES or WGS. If such modalities are accessible, panel testing might be a cost-effective means of identifying cases whose pathogenic mutation(s) are less common, unknown, or otherwise better candidates for identification by WES or WGS. In our cohort, 1 patient with clinically diagnosed Stargardt disease was found to have a mutated copy of the ABCA4 gene and was later found to have a second ABCA4 gene mutation through extended testing by another provider. Other advantages of panel testing include a relatively quick turnaround time, easy access, and low to no financial cost to the patient or physician, with a fair focus on common disease-causing mutations.

One potential disadvantage of such panels is that when results are negative or undetermined, they do not allow wider analytic exploration because they do not reflexively pivot to WES or WGS. ⁹ One strategy to mitigate this drawback is for genetic counselors and retinal specialists to reference alternative databases, such as the Genome Aggregation Database, to further contextualize indeterminate findings. In addition, panel testing assays only for mutations that have been included in custom kits. This approach, without additional deeper sequencing methods, excludes newly discoverable mutations that could contribute to disease. Furthermore, as with nearly all methods of genetic testing, detection of multiple VUS is common and often complicates the interpretation of results. In our cohort, 17 patients (28.8 %) were found to have only VUS. Current understanding of the contributions of the VUS to disease is limited. It is anticipated that the ability to interpret genetic testing results will improve with the expansion of genomic databases, functional studies of VUS, and more detailed characterization of the combined effects of these variants with other well-known mutations.

The high rate of indeterminate results (47.5% in our study) suggests that panel genetic testing should not be used as a screening tool or to rule out IRD for questionable clinical diagnoses. Indeterminate results include scenarios in which the returned testing results do not appear to be consistent with the clinical phenotype, which can often confuse interpretation and could require consultation with medical geneticists or counselors to determine the implications for patient care. Furthermore, it is important to acknowledge the limitations of each particular panel test. The version of the ID Your IRD panel test used in the present study does not include mutations in the RPGR gene (which accounts for the majority of cases of X-linked RP), in mitochondrial genes, or in the OPN1LW/OPN1MW gene (which underlies blue-cone monochromacy). Newer versions of the Invitae kit include mutations in the RPGR gene (ORF15) and are likely to include an increasing number of mutations across the spectrum of IRDs. For the aforementioned reasons, the ideal approach to panel genetic testing is to first obtain as precise a clinical diagnosis of IRD as possible based on family and personal history and extensive structural and functional testing and then to select a panel test that assays for pathogenic mutations that are of highest clinical suspicion.

Other areas of challenge to genetic testing include lower reported rates of participation among different ethnic groups. ⁴ The involvement of a variety of patients is critical to gaining insights into the genetics of IRDs; therefore, the opportunities to intervene should remain a high priority for retina specialists who manage IRDs. Moreover, because commercial vendors remain integral in the provision of no-charge kits and sample analysis, it is important to understand the theoretical risk to all patients of lost control of personal genetic data. Although the financial cost of no-charge panel testing is minimal, this does not capture potential harm in the form of identity theft or data theft extortion and must be a consideration in the disclosure to patients considering genetic testing.

In summary, no-charge panel testing for IRD has improved access to and participation in testing by patients and retinal specialists. Panel genetic testing using the Invitae IRD panel gives a reasonable diagnostic yield and therefore represents a valuable tool for providing diagnostic confirmation and improved genetic counseling as an initial genetic testing platform. These efforts might counterbalance the technological limitations of a panel testing approach with better access and an improved sequencing analysis with a larger database.