Genetic testing for inherited ocular conditions in a developing country

Background: Inherited ocular conditions are a frequent cause of blindness. Gene therapy has encouraged the development of genetic testing, currently able to detect up to 80% of mutations in contrast to the 5% sensitivity achieved a few decades ago. Materials and methods: One hundred sixty-three patients with suspected genetic ocular disorders who were referred to a single clinician between August 2014 and August 2019 underwent a thorough ophthalmologic examination. Those diagnosed with congenital cataract, retinoblastoma, anterior segment dysgenesis, autoimmune retinal disease, posterior microphthalmia, or cobalamin C deficiency were excluded, along with patients who opted against genetic testing. Included probands were classified into a diagnostic clinical category and offered genetic testing. Blood samples were sent to foreign accredited diagnostic laboratories, followed by clinical interpretation of the results. Results: Of the 163 patients referred, 104 were enrolled in the study. Median age at disease onset was 2 years (range, 0 to 43 years). A molecular diagnosis was established at a median age of 10 years (range, 0.4 to 50 years). Disease-causing genotypes were identified in 82 of the probands, indicating a mutation detection rate of 78.8%. Mutations were identified in 38 genes, ABCA4 being the most commonly affected (23% of mutations), followed by CRB1 (13% of mutations). Whole-exome sequencing was performed in 6 patients, resulting in a definite diagnosis in 3 (50%). Conclusions: Molecular testing for inherited ocular conditions is feasible in developing countries by sending samples to certified foreign laboratories, with a mutation detection rate comparable to published values in developed countries. Further studies to identify more disease-causing genes may improve the overall sensitivity. ARTICLE HISTORY Received November 4, 2019 Revised December 30, 2019 Accepted February 16, 2020


Introduction
In the early days of genetic testing for inherited ocular conditions, the average chance that a molecular diagnosis could be determined to be causative was less than 5% (1)(2)(3). With the rise in genetic testing, the number has increased remarkably, up to 80% in some cases (4,5). The successful application of gene therapy for inherited retinal disease (6) has highlighted the importance of genetic testing. To effectively apply specific treatments, the clinical team must know the disease-causing gene or exact mutation. The expanding complexity of genetic tests has increased the need for detailed clinical phenotyping. The human genome has normal variations, gene interactions unique to each individual, and many different disease-causing mechanisms that can explain a specific phenotype.
An estimated 36 million people worldwide are blind. The agestandardized prevalence of blindness among developing countries ranges from 0.7% in Southern Latin America up to 5.1% in western sub-Saharan Africa (7). While the prevalence of genetic causes is unknown, one study suggests that an underlying genetic cause accounts for 29.6% of blindness in children (8).
The estimated incidence of childhood blindness is 1.2 per 1000 in undeveloped countries (5), and most cases occur as a result of congenital abnormalities and genetic conditions, including microphthalmia, anophthalmia, congenital cataracts, retinal dystrophies, and albinism, among others.
The main purpose of our study was to determine the sensitivity of genetic testing for inherited ocular disease in a developing country as Chile, which is considered as a country in a transitional stage of development (9,10).

Methods
This prospective, observational, multicenter study was approved by the local Ethics Committee and adhered to the guidelines of the Declaration of Helsinki. We enrolled 163 patients diagnosed with probable ocular genetic disorders. A single clinician thoroughly evaluated all referred patients between August 2014 and August 2019 and offered genetic testing (paid by the patient). Families with the following clinical diagnoses were excluded: retinoblastoma, autoimmune retinal disease, posterior microphthalmia, cobalamin C deficiency, anterior segment dysgenesis, and congenital cataract. In addition, patients that declined genetic testing were excluded. All clinical information from participants was reviewed and then classified into a diagnostic clinical category.
After the genetic counseling and obtaining informed consent, blood samples were obtained from each patient and their parents. Samples were sent to Clinical Laboratory Improvement Amendments (CLIA) certified laboratories (Molecular Vision Laboratory, Hillsboro, OR, USA; Invitae, San Francisco, CA, USA; Centogene, Germany), according to the price and type of testing required. After the results were received, all available literature was reviewed to identify specific genes and mutations previously shown to cause genetic ocular disease. Variants of unknown significance were also reviewed and categorized into pathogenic, likely pathogenic, uncertain significance, likely benign, and benign (11). Segregation analysis using available first-degree relatives was performed on specimens with a known, clinically relevant pathogenic finding.

Results
From the 163 referred patients, 104 were included in the study. Twenty-nine were excluded from the analysis (10 congenital cataract, 8 retinoblastoma, 6 anterior segment dysgenesis, 3 posterior microphthalmia, 2 cobalamin C deficit), and 30 did not want to perform the genetic testing after pretest counseling and costs information (17 retinitis pigmentosa, 4 cone and rod dystrophy, 4 albinism, 5 other). Among the 104 enrolled patients, 56 were male (53.8%). Median age at disease onset was 2 years (range 0 to 43 years) and median age of molecular diagnosis was 10 years (range 0.4 to 50 years), with a median diagnostic delay of 8 years. Overall, 82 of the probands had a specific genotype identified (mutation detection rate of 78.8%). The detailed clinical categories, diagnostic sensitivities, diagnostic delay, and legal blindness frequencies are shown in Table 1.
Overall, mutations were found in 38 genes, ABCA4 being the most common (18 cases, 23% of mutations) and CRB1 the second most common (11 cases, 13% of mutations). Detailed pathogenic variants are presented in Supplemental Table. Whole-exome sequencing (WES) was performed in 6 patients, with definite diagnostic results in 4 of them (Supplemental Table).

Six illustrative examples
Patient 1 was a 36-year-old asymptomatic woman who was evaluated for LASIK surgery. She had no family history of retinal disorders or any previous relevant medical condition. Her best-corrected visual acuity (BCVA) was 20/20 in both eyes. Ophthalmoscopy revealed a pericentric ring atrophy in the midperiphery, which was concordant with the optical coherence tomography (OCT) and Goldmann perimetry results. Work-up, which included audiometry, revealed subclinical hypoacusia. A retinal dystrophy panel revealed a compound heterozygous mutation in the USH2A gene (c.4377A>G c.1550 + 16 T > C). Over the next 2 years, she developed nyctalopia and visual field loss. One year later, she and her husband came in for genetic risk assessment. Testing revealed that he carried a USH2A mutation (c.15364T > C), leading the couple to a 50% risk of having a baby with Usher syndrome (pseudodominance).
Patient 2 was a 10-year-old girl with clinical findings and electrophysiology results compatible with LCA ( Figure 1). An LCA and retinal dystrophy panel were negative. WES revealed a mutation in the RP1L1 gene (c.1138G > A), which is not concordant with clinical findings, so the genetic cause of the phenotype is not yet identified.
Patient 3 was a 29-year-old woman with a history of peripheral retinal vasculopathy, skin dryness, hair loss, memory problems, short stature (10th percentile) and dental anomalies. There was no previous history of bone fractures. Her left eye was enucleated when she was 6 years old due to painful eye secondary to neovascular glaucoma. Her right eye underwent panretinal photocoagulation when she was 11 years old due to peripheral retinal ischemia and retinal neovascularization (Figure 2a,b). Her right eye BCVA was 20/20. IKBKG gene sequencing, retinal panel, and gene array were negative. WES revealed CTC1 gene mutations (c.2959C > T; c.248_251dupGCCA), suggesting Coats plus syndrome. Brain magnetic resonance showed leukodystrophy and cerebral microangiopathy.
Patient 4 was an 8-year-old boy with family history of X-linked retinitis pigmentosa. His BCVA was 20/20 in both eyes.  Ophthalmoscopy revealed mild arteriolar attenuation, initial retinal pigmentary changes (bone-spicule and pigment clumping), normal disc, and normal macula. RPGR sequencing confirmed a hemizygous pathogenic variation (c.2426_2427delAG). The patient was enrolled in a gene therapy trial. Patient 5 was a 10-year old boy with macular atrophy. His 34-year-old mother was affected by cone dystrophy. They lived in a region with a constricted gene pool. OCT showed macular atrophy, a full field electroretinogram showed normal rod function, and intravenous fluorescein angiography results were normal. A macular dystrophy panel revealed a novel homozygous missense variant of uncertain significance in the ABCA4 gene (c.1289C > A). The mother was homozygous for the same ABCA4 variant. The father was heterozygous for the ABCA4 variant.
Patient 6 was a 32-year-old man with fundus albipunctatus and cone dystrophy. Gene sequencing revealed he had a homozygous mutation of the RDH5 gene (c.712G > T).

Discussion
The purpose of the present study was to determine specific disease-causing genetic mutations in 104 patients affected by an inherited ocular disease in a developing country. In many cases, the correlation of clinical entities with specific mutations has been published, but in some cases, the literature was reviewed to determine the pathogenicity of a particular variant of unknown significance. Sometimes, informatic algorithms and other mutation databases were also reviewed. In our series, almost 79% of patients had a molecular diagnosis that was consistent with the literature (4,5). Testing should be directed by clinical findings, and molecular genetics should be evaluated in the context of the clinical phenotype to avoid errors in the diagnostic process.
Unless the patient has a very specific phenotype (e.g., Patient 6), the preferred strategy is to use customized genetic screening panels. Blood is drawn from the patient and sent to a CLIA-certified laboratory that provides the panel. For segregation analysis, both parents of the affected individual should ideally be tested to aid in establishing the inheritance pattern and increase the likelihood that a particular genotype is truly disease-causing. If the ophthalmological customized genetic screening panel test results are negative, the clinician may consider a gene array or WES, depending on clinical suspicion. Customized panels are more cost-effective and more focused on specific genes than WES (4). Sometimes, these panels may include non-exomic regions for genes that have known intronic mutations (e.g., ABCA4) (12). On the other hand, WES can analyze most transcribed sequences and the investigation can be tailored to the clinical findings (13), but it is more expensive and takes longer to receive the results. WES was specifically performed in cases in which the initial results were negative or for individuals with broad clinical findings suggestive of a single causative gene affecting multiple systems (e.g., Patient 3). In our series, WES revealed a definite diagnosis in 3 of 6 patients whose initial results were negative.
Diagnostic yield varied according to the various clinical categories. Childhood retinal diseases had the highest sensitivity (92%), comparable to that reported in the literature (5,14). It also had the smallest diagnostic delay, which is important considering that it was the most common cause of legal blindness in our series and had the second highest legal blindness rate (89%). In contrast, the microphthalmia, anophthalmia, and coloboma spectrum had only 30% sensitivity. This difference among groups highlights a clear disadvantage of a tiered testing strategy, because communicating an accurate clinical description of potential entities to the laboratory is necessary to increase the diagnostic yield. Additionally, it should be considered that in some genetic diseases we do not ask for molecular study, because phenotypically they are very characteristic and there is still no treatment (X-linked retinoschisis, vitelliform macular dystrophy or ocular albinism, among others).
Molecular diagnosis of inherited ocular diseases has several benefits not only for the patients but also for clinicians. For patients, it may help with genetic counseling by confirming the diagnosis and establishing the inheritance pattern (e.g., Patient 1), as well as defining the natural history of the disease and for evaluating different treatment options. For clinicians, being able to determine specific disease-causing mutations opens a window of opportunity for investigating gene therapy and eventually offering an effective treatment to patients. Molecular testing is costly for patients, especially considering that blood samples must be sent to foreign laboratories and in some cases, WES has to be performed. This makes it even more difficult for developing countries when comparing the price to that in developed countries. Narrowing the pretest hypothesis increases the sensitivity and reduces the overall cost (4). On the other hand, whether obtaining a definite diagnosis has a real impact on the natural history of the disease or whether treatment options are available remain unanswered questions. This may explain the large proportion of retinitis pigmentosa patients that did not want to perform the test, knowing that there is no specific treatment available for most cases.
Other forms of management that help these patients, such as visual aids (15,16) orientation and mobility training, community support, social inclusion (17), mental health therapies (18), among others, should be taken into account.
Twenty-four patients (21%) in our series remain undiagnosed. A plausible explanation could be that mutations reside in yet unreported genes, the phenotype is atypical for a clinical category and therefore the genetic testing was not correctly focused, or that other mechanisms, such as epigenetic factors or non-coding regions, are involved in the pathogenesis of these entities.
Our study may have a selection bias because only patients able to pay for the molecular testing were included. Another potential limitation is that a single clinician evaluated the patients: it would be interesting to know how the sensitivity is affected when multiple doctors evaluate the same patient and discuss the appropriate panel to be requested or to evaluate a series in which the patients are not referred to a single clinician.
The delay in molecular diagnosis should be addressed in future studies. It is currently unclear if it is due to the inability to afford molecular genetic testing, patient's delay in seeking for genetic counseling and assistance, slow progression in the disease, late onset of characteristic features of the different entities, or other possible explanations.
In conclusion, our findings indicated an overall mutation detection rate of 78.8% of genetic testing for inherited ocular diseases in a developing country, with mutations found across 38 different genes. To achieve this high sensitivity (1), an accurate clinical description is crucial so that a focused genetic test can be carried out and (2) multiple samples from the patient's family should be obtained for evaluation. Further studies might be able to identify more disease-causing genes, which will ultimately improve the sensitivity of genetic testing and elucidate the genotypes for currently undiagnosed families.