Leber congenital amaurosis, from darkness to light: An ode to Irene Maumenee

Abstract

This article is dedicated to Irene Hussels-Maumenee, Professor of Human Genetics and Ophthalmology, Johns Hopkins’ Wilmer Eye Institute, Ocular Genetics Fellowship director in 1994–1995. Leber congenital amaurosis (LCA) has almost come full circle, from a profound and molecularly uncharacterized form of congenital retinal blindness to one in which a large number of causative genes and disease pathways are known; and the world’s first human retinal disease to be treated by gene therapy. Dr. Maumenee’s insights, efforts and leadership have contributed significantly to this remarkable scientific journey. In this manuscript, we present a short summary of the known LCA genes, LCA disease subtypes and emerging treatment options. Our manuscript consolidates previous knowledge with current findings in an attempt to provide a more comprehensive understanding of LCA.

In Ophthalmology and the study of visual sciences, there is nothing more profound and challenging than to diagnose and manage a baby with blindness and nystagmus. The process is filled with raw emotion and complexity due to the long differential diagnosis and the stakes involved. It is eventually extremely satisfying, as a clinically and molecular diagnosis are reached; allowing school, career, family and treatment planning.

In the summer of 1994 in Baltimore, a four-month old toddler sits proudly on his mother’s lap, chin down, eyes slowly beating, with sunken globes. Every now and then he pushes his index finger or hand into the eye, hard. We have no eye contact, but obviously the baby is aware of me and his surroundings. The room is silent, the expectations and anxiety of the family are high. They have already seen five or six other doctors, specialists, neurologists and ophthalmologists. They have undergone heart-wrenching, time-consuming, expensive tests, brain scan, MRI, CT scan, ERG and EEG, to rule out brain disease, tumors, seizures. Thus far the diagnosis is retinal dystrophy “not yet diagnosed”, rule out blindness, rule out brain disease.

The baby’s eye examination reveals absence of fixation, wandering nystagmus, +7.00 D cycloplegic refraction and a normal retinal appearance.

Dr. Maumenee walks into the room and introduces herself to the parents, and quickly assesses the child, listens to my assessment and asks a few questions to the parents. She says “Rob, this is Leber Congenital Amaurosis. Do you see the sunken eyes, the oculodigital sign of Franceschetti, the amaurotic pupils. No more testing is necessary to make the diagnosis.”

Four months of parental anxiety, guilt, and fear of brain pathology dissipate. No more testing, a spot-on diagnosis guided by a wealth of clinical experience and knowledge of inherited retinal degenerations, especially LCA. The parents and I realize and appreciate Dr. Maumenee’s experience. Dr. Maumenee continues; “We need to provide your boy with glasses, stimulate him visually, and prepare him for Braille and an otherwise as normal a life as possible. He will be very smart. Many LCA patients excel and develop an intense sense of touch, hearing and smell, she says. We need to be careful because if you have another child there is a 25% chance that they have the same condition. Unfortunately, nothing else is known about this disease, she laments. “For other retinal degenerations we have identified some genes recently; for LCA we have no known genes, no candidate abnormal proteins and no idea what goes wrong with the retina, because we don’t know the disease pathways yet. We will take a blood sample from him and you both, to prepare for the future, but we don’t have any genetic leads right now.”

Happily, things have changed. It is now 2016, and LCA has almost come full circle, from a profound and molecularly uncharacterized form of congenital retinal blindness to one in which a large number of causative genes and disease pathways are known; and the world’s first human retinal disease to be treated by gene therapy. Dr. Maumenee’s insight, efforts and leadership have contributed significantly to this remarkable scientific journey. We present a short summary of the genes, disease subtypes and treatment options for LCA. Dr. Maumenee’s fingerprints are apparent on many of them.

Gene discoveries

“In Leber congenital amaurosis, which probably comprises a number of genetically heterogeneous conditions, visual acuity remains stable despite progressive retinal pigmentary changes. The subgroup of patients with macular colobomas, however, may develop progressive decrease in vision.” (Heher, Traboulsi and Maumenee, 1992).

Leber congenital amaurosis (LCA) is the second most common inherited retinal dystrophy after retinitis pigmentosa (RP). The prevalence of LCA is estimated to be 2–3 per 100,000. LCA accounts for about 5% of all retinal dystrophies and about 20% of blindness in legally blind children¹.

LCA manifest itself in the first 6 months of life with poor visual acuity (20/200 to NLP), sensory nystagmus, high hyperopia (3D or more) and later keratoconus. Fundus characteristics are variable and in some cases the retina may be normal in appearance. In other cases, pigmentary retinal changes, vascular attenuation and optic pallor develop over time, accompanied in some macular changes.

Currently 20 genes have been identified to cause the phenotype of LCA, representing about 70% of the causal genetic variants. These genes, which are all needed for the normal functioning of the retina are: GUCY2D (LCA1), RPE65 (LCA2), SPATA7 (LCA3), AIPL1 (LCA4), LCA5 (LCA5), RPGRIP1 (LCA6), CRX (LCA7), CRB1 (LCA8), NMNAT1 (LCA9), CEP290 (LCA10), IMPDH1 (LCA11), RD3 (LCA12), RDH12 (LCA13), LRAT (LCA14), TULP1 (LCA15), IQCB1, CLUAP1, PRPH2, KCNJ13 and IFT140. Herein we provide a description of the major LCA genes. After discussing how each gene was discovered and the function of the protein, we will discuss how LCA disease subtypes can be classified and the progress of treatment studies.

1. GUCY2D (LCA1)

Using PCR and fluorescence in situ hybridization, the GUC2D gene was mapped to 17p13.1 in 1994². Guanylate cyclase 2D, aka GUCY2D, accounts for about 6–21% of all known mutations. It was the first gene to be discovered in LCA, by the Josselyn Kaplan group in Paris, in 1994. The gene is better known for its related protein: retinal guanylyl cyclase 1, (GC1) which inherently controls the levels of cyclic guanosine monophosphate (cGMP), which is the intracellular second messenger regulating phototransduction in mammals. GC1 is believed to be located in the nuclei and inner segments of the rod and cone photoreceptor cells. In particular, in 1999, Duda et al. reported that GC1 is a calcium-regulated module linked to retinal synaptic activity and thus phototransduction³. In the dark, cGMP levels are normally high and the cGMP-gated sodium channels are open. In this dark current, the photoreceptor cells are depolarized. In the light, a photon of light sets off the phototransduction cascade, which eventually activates phosphodiesterase, which lowers cGMP and closes the sodium channels, leading to hyperpolarization of the photoreceptor cells. Normally GC1 then replenishes cGMP, but not in LCA, with mutant GUCY2D, this does not occur. In GUCY2D-related LCA, the cells are low on cGMP, the sodium channels are closed and thus in a constant state of hyperpolarization. Currently, there are 132 different GUCY2D mutations⁴. The majority of causal mutations result in either protein truncation or complete functional loss⁵. Phenotypic manifestations of GUCY2D pathologic mutations in LCA include a visual acuity ranging from 20/200 to LP, an initial normal fundus appearance in the majority of cases⁶.

2. RPE65 (LCA2)

RPE65, stands for Retinal pigment epithelium-specific 65, is a common LCA gene with a mutation rate of about 3–16%⁷. This gene codes for the retinoid isomerohydrolase enzyme, which when it complexes with lecithin retinol acyltransferase (LRAT), isomerizes and hydrolyzes all-trans retinyl ester to 11-cis retinol each time a photon hits 11-cis retinal⁸. RPE65 is thus one of the crucial mediators of the retinoid (aka the visual or vitamin A) cycle. This cycle replenishes 11-cis retinal. Patients with RPE65 mutations cannot make 11-cis retinal, and therefore cannot see. Eventually the photoreceptor dies. In 1997, Marlhens et al. were the first to report that RPE65 mutations are causally linked to LCA⁹. Currently, 132 RPE65 mutations have been reported worldwide³. Fundus exam is characteristic for a blond translucent appearance and whitish drusen-like deposits in about 90% and for later developing chorioretinal atrophy peripheral patches in 64% of RPE65 LCA patients⁶. Mutations in RPE65 were also reported in RP patients having normal-looking fundi.¹⁰

3. SPATA7 (LCA3)

SPATA7 codes for the spermatogenesis-associated protein 7, aka spermatogenesis-associated protein HSD3 and localizes to the connecting cilium of the photoreceptors, but is ubiquitously expressed with high levels in brain and testis. SPATA7 is responsible for protein trafficking across the photoreceptor connecting cilium and for recruiting and localizing retinitis pigmentosa GTPase regulator interacting protein 1 (RPGRIP1) to the photoreceptor connecting cilium. In particular, Eblimit et al. reported not only a reduction in RPGRIP1 levels in the photoreceptor connecting cilium but also mislocalization to the inner segment in SPATA7 knockdown mutant mice¹¹. Its disease-causing rate is about 1.7%¹² and currently 14 SPATA7 mutations have been reported⁸. LCA patients with SPATA7 mutations have severe progressive rod cone degeneration with preserved foveal architecture. One case of CRD was described that converted to RP over time¹³. Most patients have isolated LCA, but rarely hearing loss and low sperm count have been documented.

4. AIPL1 (LCA4)

Aryl-hydrocarbon-interacting protein-like 1, aka AIPL1, is a major gene for LCA causing at least 7% of LCA cases¹⁴. Currently, there are 50 AIPL1 mutations in AIPL1⁴, Both GUCY2D and AIPL1 have different locus on the same chromosome. According to van der Spuy et al., AIPL1 is specific to rod photoreceptors in adult life only, showing a specific staining pattern extending from the inner segment to the synaptic junction in the outer plexiform layer¹⁵. AIPL1 specifically target processing of farnesylated proteins causing de-stabilization of phosphodiesterase 6 (PDE6), which is believed to be compromised in LCA leading to early-onset retinal degeneration¹⁶. Most patients have a significant decrease in their visual acuity which ranges from LP to NLP in 43% of cases during their first decade⁶. Fundus is characterized by drusen like deposits and a variable pigmentary atrophic changes with patches of hypopigmentation and pigmentary clumping⁶. A pathognomonic electroretinogram pattern characterized by slow insensitive scotopic responses¹⁶. Interestingly, mutations in AIPL1 also cause CRD and juvenile RP.

5. Lebercilin (LCA 5)

Irene always taught the fellows and residents that gene identification was much easier on islands. She had spent considerable time visiting and researching the Pingelap islands where severe colorblindness due to Achromatopsia is prevalent. The late Oliver Sachs worked in the Pingelab islands as well and visited Irene to prepare for his trips.

Religious groups represent genetic islands. Dr Maumenee’s group studied a religious isolate named the Old Order River Brethren in the farmlands of Pennsylvania and identified a new LCA locus on chromosome 6q11 and named it LCA5¹⁸. LCA5 is a 5 exon gene and accounts for about 1–2% of LCA mutations. Currently 12 LCA5 mutations have been documented². Lebercillin is located at the cilium between the inner and outer segments of photoreceptors¹⁹ and is known to interact with intraflagellar transport (IFT) machinery. Inactivation of Lebercillin results in photoreceptor degeneration due to failure in outer segment formation. This is believed to be due to the involvement of lebercillin in minus end-directed microtubule transport. Lebercillin pathologic mutations typically result in classic LCA findings including nystagmus, oculodigital sign, severe visual dysfunction despite a normal fundus. Atypically, macular colobomas were also reported²⁰.

6. RPGRIP1 (LCA6)

Retinitis pigmentosa GTPase regulator-interacting protein 1, aka RPGRIP1, accounts for about 5–6% of all LCA cases and 71 RPGRIP1 mutations are currently known^4,21. Most mutations in RPGRIP1 result in protein truncation and complete functional loss. Retinitis pigmentosa GTPase regulator (RPGR) and RPGRIP are co-localized to outer segments of human rod and cone photoreceptors as well as neuronal cells such as amacrine cells²². Both RPGR and RPGRIP proteins co-localize specifically to the photoreceptor connecting cilium where it is essential for disk morphogenesis and regulation of actin cytoskeleton dynamic²³. Similar to mutations in AIPL1, mutations in RPGRIP1 result in LP or NLP vision within the first decade of life and drusen-like retinal deposits in 60% and 80% of cases, respectively⁶.

7. CRX (LCA7)

CRX is most similar to the human OTX1 and OTX2 homeodomain proteins. CRX is expressed as an abundant 4.5-kb transcript in retina. CRX is transcription factor essential for both photoreceptor outer segment elongation and phototransduction cascade as it activates a specific sequence related to upstream processing of photoreceptor-specific genes²⁴. Furthermore, some components of the phototransduction cascade such as recoverin, were expressed by CRX ²⁵. Currently, 51 CRX mutations have been reported to cause stable but poor visual acuity^6,25 with diffuse pigmentary retinal changes and colobomatous-like atrophic macular changes²⁶. Mutations in CRX are mostly associated with autosomal dominant LCA.²⁷

8. CRB1 (LCA8)

CRB1, also known as Protein crumbs homolog 1, is an essential protein for retinal morphogenesis and localizes to the inner segment of mammalian photoreceptors. CRB1controls the photoreceptor outer segment orientation and its adherens junctions with Muller cells by stabilizing the membrane-associated spectrin cytoskeleton²⁸. CRB1 was identified as the gene for PPRPE and RP12, but also as a major gene for LCA with 166 mutations reported^4,29. The phenotypic particularity of CRB1 pathologic LCA mutations is an abnormally “unlaminated” thickened retina, almost resembling that of an immature state; this is thought to results from the interruption of the naturally occuring apoptic retinal processes³⁰.

9. NMNAT1 (LCA9)

In another “island population” in 2003, the research groups of Drs Maumenee and Inglehearn mapped a new LCA locus (LCA9 locus) to chromosome 1p36.22 using whole genome linkage analysis in a consanguineous Pakistani family³¹. Nicotinamide/nicotinic acid mononucleotide adenylyltransferase 1, aka NMNAT1, was initially not considered a good retinal and LCA candidate gene as it is ubiquitously and highly conserved protein that catalyzes the conversion of nicotinamide mononucleotide (NMN) and ATP into NAD+ as well as the cleavage of the latter³². NMNAT1 is essential for cell survival. NMNAT1 affects protein folding of the resultant NAD molecule³³. Recently NMNAT1 mutations were found in LCA patients and this is the LCA9 gene. Phenotypic manifestations of pathologic mutations in NMNAT1 result in a severe form of LCA characterized by progressive central macular “colobomatous” atrophic lesions with outer hyperpigmented borders as well as typical peripheral pigmentary changes, optic nerve pallor and retinal blood vessels attenuation³³.

10. CEP290 (LCA10)

Centrosomal protein Cep290, aka CEP290 or alternatively also called nephrocystin-6 (NPHP6), is involved in ciliogenesis. It is localized to centromer and cilia and is heavily concentrated in retinal photoreceptors cells³⁴. CEP290 involvement in cyclogenesis is based on its interaction with other building blocks, including pericentriolar material 1 (PCM1), to properly ensure centromer localization³⁵. CEP290 is likely the most common gene mutated in LCA with 153 causal mutations accounting for 21% of all LCA mutations^4,36. The c.2991+1655A→G mutation, which creates a splice-donor site, is the most common CEP290 mutation³⁶. Renal failure, hypotonia, ataxia and even Joubert syndrome were all reported as extraretinal findings in patients with mutations in CEP290³⁷. Despite the CEP290 related LCA severity, surprising foveal cones have been shown to be preserved by OCT¹³.

11. IMPDH1 (LCA11)

Inosine-5-prime-monophosphate dehydrogenase, aka IMPDH1, catalyzes the conversion of Inosine-5-prime-monophosphate into xanthine monophosphate. In fact, IMPDH1 affects the de novo synthesis of guanine nucleotides by being the rate-limiting enzyme³⁸. Currently, 13 IMPDH1 mutations are known and cause typical LCA phenotypic manifestations⁴.

12. RD3 (LCA12)

RD3 aka Retinal degeneration 3, is only expressed in the retina and functions as a chaperone protein required for the exit of guanylate cyclase from the endoplasmic reticulum of photoreceptors as part of its trafficking to photoreceptor outer segments³⁹. RD3 is necessary for vesicular trafficking of photoreceptor guanylate cyclases from the inner to outer segments³⁹. The disease burden of RD3 in LCA is low, with less than 1% of patients affected with RD3 mutations. Eight mutations are currently known⁴. Phenotypically, RD3 mutations results in characteristic retinal layers disorganization at a young age and thinning of the plexiform layers as well as the inner nuclear, ganglion cell and nerve fiber layers⁴⁰.

13. RDH12 (LCA13)

Retinol dehydrogenase 12, RDH12, is a dual specificity NADPH-dependent retinal reductase with catalytic activity towards retinoids. In the retina, RDH12 localizes to the inner segment base of photoreceptors⁴¹. RDH12 leads the transformation of 11-cis-retinol into 11-cis-retinal, which is the initial key step in phototransduction and vision sensation⁴². Currently there are 65 known RDH12 mutations⁴. Atrophic macular lesions and disorganization of the retinal architecture are often reported in LCA due to RDH12 mutations⁴³.

14. LRAT (LCA14)

LRAT, aka Lecithin retinol acyltransferase, catalyzes the esterification of all-trans-retinol in the RPE layer into all-trans-retinyl ester, which is a crucial step in vitamin A metabolism. It acts as a molecular switch necessary for the conversion of RPE65 into its membrane-associated form (mRPE65), which then enters the visual cycle for 11-cis-retinal processing⁴⁴. Currently 10 LRAT mutations are known⁴. Phenotypic manifestations of LRAT mutations in LCA included normal appearing posterior poles with sharp peripheral hypopigmenation outside the arcades⁴⁵.

15. TULP1 (LCA15)

Tubby like protein 1, aka TULP1, is exclusively expressed in the retinal photoreceptors. It contributes to protein transport between the inner and outer segments of photoreceptor cells, especially rhodopsin⁴⁶. It also stimulates the phagocytosis of RPE cells, which is necessary for RPE maintenance and regeneration. There are currently 31 TULP1 known mutations⁴. Funduscopic findings of patients with TULP1 LCA mutations vary significantly and include a normal fundus at early age progressing to indistinct foveolar reflex and pronounced maculopathy later on⁴⁷. The retinal normal layering was reported to be thinned and disrupted with absent outer retina structures⁴⁷.

16. IQCB1 (NPHP5)

IQ motif containing B1, aka IQCB1 or NPHP5, localizes to the connecting cilia of photoreceptors. NPHP5 plays a key role in ciliary function through its interaction with calmodulin⁴⁸. Currently 24 NPHP5 mutations are identified³. Pathologic mutations in NPHP5 traditionally cause Senior-Loken syndrome. LCA causal mutations are characterized by “lobular pattern of hypo- and hyperpigmentation around the vascular arcades” as well as retinal vessels straightening⁴⁹. Due to NPHP5 mutations high association with renal failure, it is recommended that all NPHP5 LCA patients undergo renal function monitoring.

17. CLUAP1

Clusterin associated protein 1, aka CLUAP1, is specifically implicated in intraflagellar transport due to its bidirectional movement along the axoneme⁵⁰. In fact, CLUAP1 is found to at the connecting cilium between the inner and outer segments of retinal photoreceptors⁵¹. Currently, there is only 1 mutation in CLUAP1has been described in LCA⁴.

18. IFT140

Intraflagellar transport 140, abbreviated IFT140, is involved in retrograde ciliary transport. Currently, 11 IFT140 mutations have been documented⁶. Phenotypic manifestations include RPE atrophy, hyperfluorescent macular ring, IS/OS disruption and an increased foveal thickness on OCT⁵². Interestingly, biallelic mutations in IFT140 were found in Saldino Mainzer Syndrome (SMS) which represents a syndrome with retinal dystrophy (like LCA), severe renal disease, cerebellar ataxia and skeletal dysplasia⁵³.

19. PRPH2 (previously RDS/peripherin)

Retinal degeneration slow, abbreviated RDS, is a tetraspanning membrane protein involved in disc morphogenesis and/or maintanance⁵⁴. All subsequent retinal degeneration phenotypes ranging from macular dystrophies to full-blown RP were found to harbor dominant (heterozygous) mutations in RDS, aka RDS/peripherin. PRPH2 is a major retinal gene with an unusual large range of retinal diseases and phenotypes. The gene and protein were renamed PRPH2. Currently there are 112 mutations reported in PRPH2 with only 2 reported to cause LCA; the other are associated with adRP, digenic RP as well as autosomal dominant macular degeneration^4,55.

20. KCNJ13 is a rare cause of LCA

KCNJ13 is a potassium channel and are members of inwardly rectifying potassium (Kir) channels which are expressed in a wide variety of cells involved in maintaining the resting membrane potential near the potassium (K) equilibrium potential. KCNJ13 is expressed in the apical microvilli of retinal pigment epithelia, but not in neural retina or in choroid. Mutations in KCNJ13 are found in Snowflake vitreoretinal degeneration¹. KCNJ13 is a very rare cause of LCA, with only 3 mutations known³. Sergouniotis et al. in 2011 and Pattnaik et al. found biallelic KCNJ13 mutations in 3 LCA patients^56,57. The phenotype was severe with nummular pigment and retinal degeneration and thinning, without lenticular and vitreal involvements.

CAPB4 and GDF6 are likely not LCA genes. CAPB4, aka Ca2+-binding protein 4, is a Cav1.4 channel regulator that plays an essential role in photoreceptor synaptic function⁵⁸. CAPB4 is localized to cone and rod photoreceptor synaptic vescicles. Littink et al. in 2009 found novel homozygous nonsense mutations in CABP4 in patients with an unusual ERG without night blindness and suggested a likely congenital cone-rod synaptic disorder⁵⁹. Aldahmeah et al., in 2010, then found CAPB4 mutations in a consanguineous Bedouin family and reported an LCA type phenotype characterized by normal appearing retinas even in early adulthood⁶⁰. LCA can have normal appearing retinas, but eventually subtle vessel narrowing and retinal pigmentation is almost always seen and usually LCA is progressive. One of the sibs had a decreased cone ERG and an almost normal rod ERG, which is not compatible with the diagnosis of LCA. Also CABP4 is expressed in the synapses, not the photoreceptor cell body, outer or inner segment where all the other LCA gene and proteins are found. Taken together, these are five strikes against LCA, and more likely a cone-rod synaptic disorder. GDF6 aka Growth/differentiation factor-6 is a member of the transforming growth factor-beta. In immunohistochemical analyses of mouse retina, Zhang et al. detected GDF6 in ganglion cell layer, inner plexiform layer, and retinal pigment epithelia, not the photoreceptors⁶¹. Duplication and deletions of GDF6 cause microphthalmia and colobomas (clefts of the iris and retina). Asai-Coakwell et al. found two LCA/juvenile RP patients with biallelic GDF6 mutations⁶². GDF6 was listed as a gene for LCA. Subsequently, these two patients were found to have PRPH2 mutations that co-segregated in the family and are more likely the cause of disease. Extensive WES studies did not document further GDF6 mutations in LCA in large cohorts. In conjunction with the fact that GDF6 is expressed in retina but not in photoreceptor, the LCA link is likely not tenable. GDF6 mutations are likely very important for retinal function, but unlikely to be causal for LCA by themselves.

LCA disease subtypes

“Photoaversion can be a prominent clinical feature in some patients with Leber congenital amaurosis. The ERG clinches the diagnosis. These patients may constitute a distinct genetic subtype of the disease and molecular genetic studies will help resolve this issue.” (Traboulsi and Maumenee 1995). Also, the ateliotic fovea defined as an unfinished primordial fovea, may represent a subtype of retinal disease (Maumenee, 2001).

Initially, at the time of first description by Theodor Leber in 1869, LCA was defined as congenital retinal blindness with nystagmus and poorly reactive pupils (“tapetoretinal degeneration mit amblyopie”) and thought to represent a homogenous form of congenital blindness⁶³ Subtypes were not considered. Histological studies (pre-genome) suggested several different morphological subtypes of LCA⁶⁴. The genetic and genomic era solidified that perception, as currently 20 genetic types of LCA are known⁴. The currently known 20 LCA genes participate in seven disparate retinal cycles/processes. These retinal cycles/processes are likely to harbour the still unknown LCA genes.

These seven retinal cycles/processes are interrupted resulting in one of the seven LCA disease pathways, and result in similar phenotypes and likely require similar treatments. LCA disease pathway 1 is in the phototransduction cascade, the retinal enzymatic process of capturing a photon of light, resulting in photoreceptor ion channel closure and hyperpolarization (GYCY2D, AIPL1, and RD3). Disease pathway 2 is in the retinoid cycle, to replenish 11 cis retinal after it is converted to all trans-retinal (RPE65, LRAT and RDH12). Disease pathway 3 is in retinal transcription factors (CRX). Disease pathway 4 is in the Intra Flagellar Transport (IFT), the anterograde and retrograde protein transport (conveyor belt) from inner to outer segment and back again (SPATA7, LCA5, RPGRIP1, CEP290, TULP1, NPHP5, CLUAP1 and IFT140). Disease pathway 5 is in photoreceptor structure, both external morphogenesis and internal cell structure (CRB1 and PRPH2). Disease pathway 6 is in metabolic enzymes for cellular survival (IMPDH1 and NMNAT1). Disease pathway 7 is in photoreceptor ion channels, such as Potassium ion channels (KCNJ13).

By clinical definition, LCA patients are phenotypically relatively homogeneous and invariably show severe visual loss early in life with nystagmus, sluggish pupils, oculodigital sign, and relatively normal appearing retinas (with later changes), but the retinal morphological basis of these visual malfunctions differs among genotypes^64,65.

Jacobson et al., in 2016, categorized the retinal morphological structural changes with in-vivo retinal architectural studies by optical coherence tomography (OCT) and concluded that some are associated with a specific genotype while others are indistinguishable¹³. Based on their classification, we can identify four major retinal structural categories (OCT types of LCA):

OCT type 1: Normal in vivo retinal architecture

Some OCT scans of LCA patients are remarkable for normal retinal layers thickness. Other young LCA patients may have subnormal retinal and ONL thicknesses within the central few degrees and normal thickness across the rest of the scan. This pattern was typical of LCA patients with GUCY2D mutations.

OCT type 2: Thickened retina

This pattern is specific for a reduced foveal ONL, limited extra-central ONL and thickened dysplastic-appearing retina across the remainder of the retina. Retained central ONL and better acuity in some patients may lead to a diagnosis of CRB1-RP⁶⁵. Whether LCA or RP, patients show extra-central coarse and abnormal lamination with thickening.

OCT type 3: Normal foveal architecture

Five genotypes showed a preserved central island of ONL that decreases in thickness eccentrically. The foveal ONL peak could be normal or reduced. A common mechanism resulting in this pattern is a ciliopathy, which includes Lebercilin (LCA5), RPGRIP1 (LCA6), CEP290 (LCA10), and TULP1 (LCA15), although the latter disorder may have a more complex mechanism⁶⁶. RPE65 (LCA2) could show a similar pattern. Retinal thickness was at the lower limit of normal or actually subnormal, in contrast, for example, to CRB1 (LCA8).

OCT type 4: Severe macular atrophy or foveal mal-development

This group includes two with macular disease (AIPL1 (LCA4) and RDH12 (LCA13)) and one that appears to be a developmental abnormality with lack of foveal formation and inner retinal laminae crossing the central retina (NMNAT1 (LCA9)). ONL thickness in all three is detectable, but reduced centrally. Retinal thickness varies in AIPL1 (LCA4) and RDH12 (LCA13), while he NMNAT1 (LCA15) patient had a thinned retina.

This last retinal subtype was predicted and suggested in 2001 by Irene Maumenee. “The ateliotic macula” coined by Irene Maumenee was defined and characterized by an unfinished or primordial appearance.

Pregenome histological studies of LCA eyes from cadavers or blind painful enucleated eyes revealed a wide diversity of morphological retinal subtypes. Koenekoop in 2004 summarized 13 retinal studies of these LCA patients eyes⁶⁷. These studies have led to a view that LCA may not be one disorder, but multiple disorders, which may be classified into three types of histological categories: Histological type 1: an aplasia (or dysplasia), with abnormal embryological formation of photoreceptors; Histological type 2: a retinal degeneration, with early, progressive photoreceptor cell death; and Histological type 3; a retinal dysfunction, in which the retinal anatomy remains essentially normal, but a key biochemical message is missing⁶⁷. These three histological types correlate with the OCT categories proposed by Jacobson et al. in 2016¹³. We may also predict that the retinal aplasia (or dysplasia) found by histology and by OCT may be represented by CRB1. Retinal degenerations are found in most LCA patients and are represented by RPGRIP1, LCA5, RPE65 and other genes. Retinal dysfunction is likely represented by GUCY2D.

Thus despite the relatively homogeneous and severe clinical phenotype of LCA, there appear to be at least three major retinal morphological groups, based on OCT and histological studies^13,64. Visual loss can be due to severe outer retinal degeneration (Group 1 called retinal degeneration) or retinal disorganization with loss of photoreceptors in early life (Group 2 called retinal dysplasia). A few molecular forms of LCA, however, have unexpectedly shown retained outer retinal structure and appear to be dysfunctions (Group 3 called retinal dysfunctions)^13,64.

Treatment options

In 1994, before the explosion of molecular, genetic and genomic discoveries of the disease pathways of LCA, no-one would have predicted 20 LCA genes (this review), seven LCA disease pathways (this review), three LCA histological types⁶⁸, four LCA OCT types¹³ and six potential treatments. Today, LCA remains officially untreatable, but at least six treatments for LCA are in human clinical trials. It is likely that at least for some types of LCA a treatment will be available to our patients soon.

The LCA patient and his family I saw with Dr Maumenee in 1994 may now be eligible for a clinical trial or may be authorized to an approved treatment soon. The exciting LCA treatments being tested in humans right now are gene replacement, stem cell replacement, retinal camera, CNTF, an oral drug and gene editing.

Excellent reviews of human RPE65 gene replacement in LCA are available^69,70. RPE65 gene replacement works to increase vision and is safe, in repeated trials, one or both eyes. The consensus on the durability of RPE65 gene replacement therapy is still debatable. On the one hand, Brainbridge et al. (the London group), in their trial, found that the effect wears off with subsequent follow-ups and the degeneration may not be halted, perhaps calling for earlier therapy^69–72. On the other hand, Maguire et al. and Testa et al. (the Philadelphia group) reported a three years maintained and stable improvement in visual acuity, area of the visual field enlargement and nystagmus^73,74. It is worth mentioning that both studies used different adeno-associated virus vectors, delivery techniques as well as different patients’ age ranges.

Other LCA gene replacement trials are likely to start soon (GUCY2D, AIPL1, RPGRIP1, LCA5, CEP290 and others)^75,76. Stem cell therapy to introduce retinal progenitor cells into the atrophic LCA retinas is now in phase I clinical trials⁷⁷. LCA patients with severe retinal cell loss and patients with maculopathies such as NMNAT1, AIPL1 and RDH12 are potential candidates. The Argus ll camera captures visual information through the inner retina and has been approved for adults with retinal degenerations. This device will likely be available to children in the near future. CNTF is a neurotrophic factor that has been shown to rescue photoreceptors in animal models of retinal degeneration. In human trials CNTF treatments show variable responses, from maintaining viable cone photoreceptors⁷⁸ to no response on visual fields⁷⁹. So far, CNTF studies were conducted only in adults. In the future, it is likely that CNTF will be tested in children but its effects could be less appreciated compared to adults as the age of patient markedly changes its predicted pathophysiological effect. An oral drug QLT 091001 was shown to be safe and efficacious in children with RPE65 and LRAT mutations⁸⁰ and in older adults with RP and RPE65 and LRAT mutations⁸¹, and this drug is now entering a phase III trial. An example of genetic editing is anti-sense oligonucleotide (ASO) treatment, targeting the very common LCA CEP290 mutation (c.2991 + 1655A > G). This splice mutation creates a cryptic splice donor site resulting in the insertion of a pseudo exon into the CEP290 mRNA. ASO delivery fully restored CEP290 pre-mRNA splicing, significantly increased CEP290 protein levels and rescued a ciliary phenotype present in LCA patient-derived fibroblast cells⁸². A human clinical trial is being planned.

It is very likely that matching the treatment with the LCA disease types; the LCA OCT types and/or the LCA histological types will increase treatment success. Gene replacement and gene editing must be performed in LCA retinas that harbour living cells, thus in LCA OCT type 1 and 3, but not likely in LCA OCT type 2 and 4. Success with gene replacement and editing is more likely for LCA histological type 2 (before the degeneration is too far advanced) and in type 3, but not likely in type 1. Stem cell replacement is likely reserved for LCA OCT type 4 and for advanced degeneration in all types. Oral drugs and CNTF is more likely to succeed in early degeneration and in the retinal dysfunctions. Argus ll is reserved for advanced LCA with some intact inner retinal architecture. Combination therapies are likely to be tested soon.

The future for LCA is very bright. From nothing to 20 genes and six treatments in 20 years. We can now genetically diagnose LCA in at least 70% of patients, guide their careers, aid their schooling and help plan their families. Some patients have entered gene specific treatment trials; others will soon follow. Due to Irene Maumenee’s visions and mentorships, she has stimulated, guided and participated in this remarkable evolution.