heterochromia

Congenital retinal nonattachment consists of a group of sometimes familial conditions for which no responsible gene has been identified.  In a genomic study of 21 consanguineous NCRNA Pakistani families 3 had mutations in ATOH7 and 10 had mutations in familial exudative vitreoretinopathy genes.  Genotyping did not reveal associated mutations in the remaining 38% of these families. It is likely that multiple entities are represented but until the molecular etiologies are identified, no more specific classification is possible.

Studies in mice document that the Atoh7 gene is important to retinal ganglion cell neurogenesis.  In humans, both autosomal recessive PHPV and congenital nonattachment of the retina are associated with microsatellite linkage and haplotype matching to a region at 10q21 adjacent to the ATOH7 gene but so far no causative mutation has been found in this region.  However, studies in large consanguineous kindreds in which a deleted DNA segment adjacent to ATOH7 segregated with the NCRNA phenotype suggest that a transcription regulator may be at fault in the timing and level of ATOH7 expression.

The disorder known as persistent hyperplastic primary vitreous is generally not considered hereditary since it usually occurs unilaterally and sporadically.  It is sometimes found in association with a number of syndromal conditions as well.  However, it has also been reported in familial patterns consistent with both autosomal recessive and autosomal dominant patterns.  DNA mapping of individuals with bilateral disease found in a consanguineous Pakistani kindred with presumed autosomal recessive disease suggests that a locus at 10q11-q21 may be responsible.

Evidence for autosomal dominant inheritance of persistent hyperplastic primary vitreous comes from rare families with an apparent vertical transmission of the condition.

Congenital nonattachment of the retina is also seen in the osteoporosis-pseudoglioma syndrome (250770).  However, this is a syndromal disorder with neurologic and joint disease in addition to porotic, thin, fragile bones (sometimes called the ocular form of osteogenesis imperfecta) resulting from mutations in LRP5 on chromosome 11.

Retinoblastoma is a malignant tumor of the developing retinal cells caused in most cases by mutations in both copies of the RB1 gene.  The RB1 gene is a tumor suppressor gene, located on chromosome 13q14 and is the first human cancer gene to be cloned. The gene codes for the tumor suppressor protein pRB, which by binding to the transcription factor E2F, inhibits the cell from entering the S-phase during mitosis.  Recent evidence suggests that post-mitotic cone precursors are uniquely sensitive to pRB depletion and may be the cells in which retinoblastoma originates.

However, more recent information suggests that the occurrence and viability of retinoblastic cells may be more complex than suggested by simple loss of function of the RB1 alleles.  There is increasing evidence for the role of epigenetic factors such as DNA methylation impacting the differential expression of more than 100 additional genes which may be influencing the retinoblastoma phenotype.  Among these is an upregulation of spleen tyrosine kinase (SYK) required for tumor cell survival which, if inhibited, leads to retinoblastoma cell death in vivo and in vitro.

Pedigrees of familial cases have an autosomal dominant pattern but the disease requires homozygosity of the RB1 mutation.  This complicates genetic counseling for retinoblastoma. One third of cases have a germline mutation with a mutation in only one of the two gene copies in every cell.  A somatic mutation in the second allele then leads to  homozygosity causing tumor development.  Since one of the parents contributed the germinal mutation, and there is high penetrance (as much as 85%), this leads to the autosomal dominant pattern in these families. In 6% of retinoblastoma cases with germline mutations the family history is positive. The risk for developing bilateral and multifocal retinoblastoma is high and the age of onset is around 14 months.  This is the case for virtually all bilateral tumors.  The mean number of tumors is about 5 in the two eyes.  The offspring of a parent with bilateral retinoblastoma have a 45% chance of developing a tumor (they have a 50% chance of inheriting the germline mutant allele).  Reduced penetrance of 10 to 15% lowers the expected occurrence of disease from 50% to 45%.

However, two thirds of cases are of non-germinal origin with both somatic mutations occurring in a single retinal progenitor cell.  Because this is a highly unlikely event, these cases are generally unilateral and unifocal with an average age of onset of 24 months. Sporadic cases constitute about 94% of all retinoblastomas, of which about 60% have unilateral disease with no germline mutations.  Individuals who acquire mutations in both alleles somatically (with single, unilateral tumors) do not have a mutation in their germ cells and therefore usually transfer no tumor risk to their offspring.  Laterality and number of tumors alone, however, cannot be used for accurate predictions in this case since about 15% of patients with unilateral and monofocal tumors actually have germline mutations.  This leaves a residual risk of transferring heritability of about 1-5% in unilateral patients without a family history.

To further complicate the story, recent evidence suggests that retinoblastoma is genetically heterogeneous.  About 6% of patients have no RB1 mutation.  In one study, about half of such individuals have up-regulation of the MYCN oncogene (2p24.3) suggesting a second mechanism leading to clinical retinoblastoma.  For unknown reasons, such tumors tend to  be larger, more aggressive, and discovered at an earlier age than unilateral non-familial RB1 tumors.  The MYCN gene product is a transcription factor important for organ development during embryogenesis.  Its amplification has been implicated in about 25% of neuroblastomas.