Case 1

A 54-year-old pastor with retinitis pigmentosa (RP), who was otherwise healthy, consulted his retinal specialist for more information about his diagnosis. Fundus examination showed typical features of RP with bone spicule formation in the periphery; investigations revealed constricted visual fields and a generalized reduction in all amplitudes of the full-field electroretinogram. His deceased father was not thought to have been affected with RP; his mother was unaffected; his 3 children, aged 26-20, had no children; 2 had been screened for signs of RP and were reportedly normal. A 90-minute counseling session reviewed the modes of inheritance—dominant, recessive, X-linked, and mitochondrial; as there was no family history, no specific risks based on the pattern of inheritance could be outlined for his children. He requested genetic testing to potentially resolve the uncertainly around the risk of transmission of his trait; clinical testing was requested with an RP panel (parallel testing) that accomplishes sequencing of 91 genes and deletion/duplication testing. This approach is commonly adopted for cases of nonsyndromic RP when the phenotype does not allow a single gene to be considered. The panel was selected based on cost with the limitation of testing only 91 genes and would not have tested all the possible genes that may underlie RP.

Two variants of uncertain significance (ie, variants that do not yet have sufficient evidence to show that they are either disease-causing or benign) were found—1 in NPHP4 (c.3292G>A) and another in ZNF513 (c.886G>A); further testing confirmed that there was no deletion or duplication of the other allele, which may have masked the results to appear as if only 1 mutation was present. Biallelic pathogenic variants (ie, variants with sufficient evidence to show that they are disease-causing and identified in each of the 2 alleles) in NPHP4 are seen in autosomal recessive conditions, nephronophthisis and Senior-Løken syndrome, which would not have been considered initially as a possible clinical diagnosis in our patient. Also, biallelic pathogenic variants in ZNF513 are known to be associated with autosomal recessive RP. In many cases, when a variant is detected, segregation analysis is performed to determine if the variant was inherited from either parent, especially if the parent has a similar or related disorder. If the variants of unknown significance in NPHP4 and ZNF513 , identified in the patient, were to be later reclassified as pathogenic variants, we still could not presume that we have identified the genetic cause for the patient’s symptoms, as the patient would only have the single pathogenic variant (presuming the patient has a recessive disorder). In this case, the results were further reviewed by a local medical genetics diagnostic laboratory that concurred with the reference laboratory’s reported results. Whereas genetic testing could have led to a specific diagnosis and then allowed prognostication for the children, it did not. With no direction on how to proceed in this case, routine clinical follow-up of his children for signs of RP seemed to be the most appropriate advice to the family, without further genetic testing at this time.

When undertaking testing for genetic disease, pretest counseling is recommended, as that testing may uncover variants of unknown significance, may find mutations in genes other than the targeted common gene when panel testing is performed, or may not even find a pathogenic variant. The patient may not recognize that when a variant is found, further testing of affected and unaffected family members may be required in order to resolve its importance. The following case has a distinct phenotype, allowing targeted genetic investigation with Sanger sequencing of specific genes (a tiered approach to testing).

Case 2

A 4-month-old boy was referred by a retinal surgeon and a pediatric ophthalmologist, both of whom had examined him under anesthesia. In both eyes, the anterior segment was shallow and there was a temporal mass posterior to the lens dragging the macula. His birth history was unremarkable. The mother’s pregnancy was normal. There was no family history of a similar condition. He would not fixate on objects or follow them, his pupils constricted to light and both eyes were of normal size. A presumptive diagnosis of familial exudative vitreoretinopathy (FEVR) was suggested and molecular genetic testing was ordered with sequencing of genes known to be associated with FEVR: FZD4 , NDP , LRP5 , and TSPAN12 .

Heterozygous variants in LRP5 (c.713C>T: Thr238Met) and TSPAN12 (c.434G>A: Trp145X) were found and no variants were detected in FZD4 or NDP . Segregation analysis was then requested and some 4 months later, this showed that the infant had inherited both variants from the mother and not the father. An examination of his mother revealed that, in the temporal periphery, she had areas of vascular nonperfusion and abnormal vessels. The reference laboratory reported the TSPAN12 mutation as a definitive, disease-causing pathogenic variant and the LRP5 mutation as a variant of uncertain significance.

A reference laboratory should cite appropriate published articles that form the basis of the interpretation of the results. The clinician can use the Internet to check whether the mutation has been seen before and the associated phenotype through ClinVar as part of the suite of resources available through the National Center of Biotechnology Information website ( www.ncbi.nlm.gov ). Counseling this family may appear to be straightforward if one accepts the report of the reference laboratory at face value. However, how are we to understand the relative risk of the mutation in LRP5 , as mutations in this gene also have been associated with loss of bone mass in late adolescence? Perhaps the best course of action would be to continue to provide care within the pediatric period and monitor growth and development of the affected child, and to caution the family about recurrence in a future pregnancy. As well, bone density screening could be considered for the mother, although no recommendation was made in her case.

Genetic testing does not replace accurate phenotyping by careful clinical examination, the identification of the pattern of inheritance, and clinical experience. For example, the recognition of a dominant family history of RP and a phenotype of sector RP would suggest that the underlying causative mutation would be in the rhodopsin gene. A candidate gene approach would then be the most cost-effective approach, rather than depending on a panel of genes that may generate additional information that is difficult to interpret if variants are uncovered that would appear to be unassociated with the phenotype. There are, however, instances in which a parallel approach, testing multiple genes, would be more effective than a candidate gene or tiered testing approach. Such would be the case for Bardet-Biedl syndrome or Usher syndrome, in which pathogenic variants in many genes may underlie the clinical phenotype. These issues and others are aptly discussed in the Guidelines for Genetic Testing of Inherited Eye Diseases developed by the American Academy of Ophthalmology Task Force on Genetic Testing. Case 2 was considered to be a Mendelian disorder and the family was referred to a physician and counselor who were familiar with the disease. Pretest counseling was obtained such that the parents were aware of the possibility of the need to test their DNA for inheritance of the variants. A Clinical Laboratory Improvement Amendments (CLIA) certified laboratory was accessed that would provide an opinion of the pathogenicity of the variants uncovered in the genes tested. Parallel testing was not undertaken; tests of specific genes were requested.

Though genetic testing has proven to be essential in establishing a molecular diagnosis, it has limitations. To reiterate, pretest counseling of the patient should be provided to discuss the purpose and benefits of testing, as well as to emphasize the risks and limitations of testing, as these latter 2 topics are often glossed over. If 2 mutations are found in a gene that could be the putative cause of a recessive disorder, parental testing should be undertaken to understand if the mutations are on 1 chromosome (cis) or on separate chromosomes (trans) and therefore inherited from each parent. Unfortunately, biological relationships of relatives are assumed to be correct, and the surprise of nonpaternity needs to be considered. Finally, clinical testing may be unsuccessful and the patient may need to depend on research testing, which has no obligation to report results in a reasonable time frame. Research takes time and resources; nevertheless, the advent of newer technologies such as next-generation sequencing (NGS) has enabled the resolution of some unsolved cases. The potential for targeted NGS and whole genome sequencing to resolve undiagnosed cases of inherited retinal disorders is discussed by Ellingford and associates. The significant investment of time and resources in this exercise is not trivial; thus for the moment, the investigation of cases through these routes must be very selective and by nature remains within the research realm. Research results still require clinical validation in a diagnostic laboratory. Undoubtedly, there is a wealth of continuing new information on genetic eye disease and we are very privileged to have access to testing that can better inform our patients.