We described the case of a child with 13q-mosaicism affected by retinoblastoma. The unilateral presentation agrees with previous data available for 13q deletions larger than 1 Mb including MED4 and SUCLA2 [6]. As in this case, retinoblastoma with both genes deleted is associated with less tumor aggressiveness compared with tumors whose genes are conserved [6].
Retinoblastoma seems to be caused by a double hit in RB1 approaching 98% cases (by mutation, deletion, promoter methylation or intra-genic chromothripsis) [12, 13]. Few retinoblastoma cases would start because of MYCN amplification [13]. 13q deletion syndrome patients would not be an exception. In fact, we confirmed a second RB1 hit (RB1 p.Arg320Ter) in the tumor.
However, double RB1 hit only gives rise to retinoma; therefore, subsequent epigenetic or genetic changes would give an advantage for tumor progression. The sequence of events capable of causing a malignant phenotype is only partially known. Epigenetic deregulation secondary to homozygous RB1 loss drives an increase in KIF14 and E2F3 levels [14] and could lead to the expression of the SYK oncogene as well. Moreover, cellular control mediated by p53 is inactivated as a result of high expression of MDM2 and MDM4 in retinoblastoma [14].
In addition, cytogenetic analysis has shown recurrent CNVs (copy number variation) among retinoblastoma tumors, which are mainly chromosomal gains at 1q, 2p, 6p, 13q and 19q and losses at 13q, 16q and 17p [15]. These recurrent aberrations allow to establish as a possible hypothesis that genes located at these loci could be related to retinoblastoma progression [15], yet no conclusive data are available about this at the moment. We looked for CNVs in the tumor and discovered a chromosomal gain in 6p12.3pter, which is one of the most frequently reported CNVs in retinoblastoma [15]. However, we also detected a less common deletion of 6q25.3qter. The deletion of this region has already been described among non-13q-deletion syndrome patients, although rarely [16]. Sixty OMIM genes are located in this region, and several of them are associated with different cancers, but none with retinoblastoma. A terminal 6q deletion may be present in ovarian cancer and neuroblastoma [17] and seems to be related to bad prognosis in neuroblastoma [17]. The fact that this deletion could play a role in retinoblastoma development in the context of 13q-syndrome is unknown.
Furthermore, NGS approaches have detected a low rate of mutations in retinoblastoma. Several studies support retinoblastoma as one of the less mutated human tumors. Only BCOR (mutated in 13% of tumors) and CREBPP mutations occur frequently in retinoblastoma [18]. Therefore, retinoblastoma presents a stable genome with few genetic events described and epigenetic deregulation appears to have a notable role [19]. Studies based on RNA-sequencing could continue to shed light on the genes and signaling pathways involved in retinoblastoma development [20]. In regards to common mutated genes in retinoblastoma, we determined BCOR and CREBPP status without detecting pathogenic variants. We did not find other variants considered pathogenic or likely pathogenic in 400 genes commonly mutated in pediatric cancer beyond RB1.
The patient carries the deletion 13q12.13–13q21.2 and, therefore, fits in Group 1 of the clinical classification for 13q-syndrome [7]. Patients with band 13q14 deleted typically present with mild facial anomalies such as high forehead, short nose, small upper lip, curly hair and down-turned corners of the mouth [6]. Our patient does not show these facial features. Furthermore, deletion of NUFIP1, located in 13q14.12, and PCDH8, in 13q21.1, may be crucial for developmental delay [6]. Both of them are deleted in our patient, but the degree of mosaicism in her central nervous system is unknown. In fact, she is neurologically normal. Moreover, other common abnormalities in Group 1 are micrognathia and microcephaly but these are related to loss in the 13q21.33q31.1 and 13q21.32q21.33 regions, respectively [6]. Our patient’s deletion finishes at 13q21.2; therefore, she does not present either micrognathia or microcephaly, because those regions are not affected. About 75% of patients with large deletions present short height, but this is not the case of our patient (50th percentile). Genes involved in short height have not been clearly defined.
The BRCA2 gene, located in 13q13.1, may be lost in some 13q-patients. Heterozygous mutations in this gene predispose to breast and ovarian cancer syndrome in adulthood [21] and a complete deletion of this gene might predispose to these tumors as well. However, the occurrence of these two tumors has not been reported in 13q-syndrome to date. Our patient loses BRCA2; therefore, she may benefit from risk-adapted surveillance strategies for breast/ovarian cancer.
After confirming retinoblastoma diagnosis in a child, genetic study of RB1 in the germline is mandatory. Any phenotypic manifestation, including minor peculiarities (clinodactyly of the fifth finger in our case) should raise suspicion of 13q-syndrome, and it should be studied, given the fact that mosaic forms exist.