Leber congenital amaurosis (LCA, MIM 204000) represents a group of congenital retinal diseases that lead to blindness, with a worldwide prevalence of three in 100,000.³ Although rare, it accounts for at least 5% of all inherited retinopathies and approximately 20% of children attending schools for the blind.³ We estimate that 180,000 patients are affected worldwide.

In 1869, Leber defined LCA as a congenital form of retinitis pigmentosa, with severe visual loss at or near birth, wandering nystagmus, amaurotic pupils, a pigmentary retinopathy, and autosomal-recessive inheritance.^3,54 In 1956, Franceschetti and Dieterli reported a nondetectable electroretinogram (ERG) in the early course of LCA as essential in the diagnosis.²⁵ Dominant inheritance has also been reported, but this is thought to be rare.^{26,46,74,81,82} LCA is genetically heterogeneous, and since 1996, eight genes^{12,15,20,29,59,61,67,79,83} (six of which have been cloned) with disparate retinal functions have been implicated. Five of the LCA genes (tables 64.1, 64.2, and 64.3) are expressed in the photoreceptors, namely, retinal guanylate cyclase (GUCY2D), cone-rod homeobox (CRX), Aryl hydro-carbon receptor–interacting protein-like 1 (AIPL-1), retinitis pigmentosa (RP) GTPase interacting protein 1 (RPGRIP-1), and crumbs-like protein 1 (CRB-1), while one is predominantly expressed in the retinal pigment epithelium (RPE), the RPE65 gene.

The study of LCA is proving central to our understanding of normal retinal development and physiology and also improves our understanding of other retinal degenerations. In the near future, LCA may be treatable by pharmacological intervention and/or gene replacement therapy, as both mouse⁹⁰ and dog¹ LCA models showed dramatic short-term improvements in rod and cone physiology and restoration of vision, by ERG, pupillometry, and behavioral testing.¹ These future therapies for human LCA will likely be gene-specific, giving major significance to gene identification and genotype-phenotype studies.

The clinical variability of LCA is striking. We find variability in visual acuities (20/200 to no light perception), visual evolution (stable, deteriorating, or rarely improving),^{7,30,40,52,75} refractions (from high hyperopia to high myopia), retinal appearance (from near normal to severe pigmentary retinopathy), associated ocular features (keratoconus, cataracts), associated systemic features, and retinal histopathology (from essentially normal⁴³ to extensive degeneration.^{5,23,51,55,66,74,92,95} Some patients have an essentially normal retinal aspect; others may have yellow flecks, salt-and-pepper changes, a marbled pattern, atrophic changes, nummular pigment clumps, a “macular coloboma,” white dots, or preserved para-arteriolar RPE. Keratoconus and cataracts may develop in the course of the disease in some patients. LCA may rarely be associated with systemic disease, and this adds additional variability to the disease spectrum. Mental retardation, deafness, polycystic kidney disease (also known as Senior Loken disease), skeletal anomalies (also known as Saldino Mainzer disease), or osteopetrosis may be found in addition to the ocular disease.

In approximately 40% of LCA cases it is now possible to identify the causative mutations in one of the seven LCA genes. Several strategies are now available to provide a molecular diagnosis for a child affected with LCA. The most rapid and comprehensive is the new LCA genotyping array (LCA disease chip), which includes all 300 currently documented LCA mutations. In a period of four hours per sample it is possible to determine the genotype in ∼35% of the new cases.⁹⁷ Conventional SSCP, dHPLC combined with automated sequencing also allows identification of mutations but is much more cumbersome. Genotyping LCA patients is extremely helpful for providing 1) a more accurate clinical diagnosis, 2) a more accurate visual prognosis, 3) a molecular classification of disease, 4) a prenatal diagnosis in selected cases, and 5) a way of separating LCA patients who may be treatable in the near future and those that may be treatable later.

The clinical understanding of the diagnostic findings in LCA has been evolving as more patients have molecular diagnoses and it becomes possible to go back and correlate test and phenotype to a confirmed known type of early onset retinal degeneration. Because LCA represents a group of diseases with at least seven and potentially as many as 20 genes, it can be expected that there will be variation in severity at onset and in the severity of the disability. For many years, with the lack of availability of the electroretinogram (ERG) and a reluctance to test infants and young children, many patients did not have an ERG until they were older. By then the patients were frequently blind and had a non-detectable ERG signal. Foxman et al.²⁴ suggested ERG testing in blind infants before the age of 1 year to separate LCA from early onset RP. ERG testing is also crucial to differentiate albinism, complete and incomplete achromatopsia, and complete and incomplete congenital stationary night blindness from LCA.⁵⁰ Infants can now easily have standardized ERGs in a visual physiology laboratory, a site which is accustomed to managing infants.

Because the infant ERG is not equivalent to that of older children, caution is needed in interpreting these infantile ERGs. Fulton et al. have published standardized first-year values to assist in the interpretation of waveforms from this age group.³⁰ If it is necessary to perform ERG testing under anesthesia, then great care must be given to interpreting the resulting waveforms. Operating room conditions are seldom standardized, and more importantly some anesthetics may suppress or alter ERG waveforms. We recommend that a child suspected of having LCA or any other retinal dystrophy have an ERG at around age 6 months and then a repeat ERG at 1 year.

Repeat ERGs are important because infantile testing is so difficult—children frequently cry and the electrodes fall out, or the tears may interfere with the signals and the ERG signals still mature in the first year of life.³⁰ It is possible to have a small ERG signal early in the LCA disease process, and this certainly does not preclude the diagnosis of LCA. Also, a small number of children with early visual impairment have developmental delay and with repeat testing may have more robust or normal ERG signals on repeat testing.

The aim of this chapter is to discuss in detail, the seven types of LCA associated with the seven currently known LCA genes. Furthermore, we will summarize the management of the blind infant.