First step towards a consensus strategy for multi-locus diagnostic testing of imprinting disorders.

ABSTRACT

BACKGROUND: Imprinting disorders, which affect growth, development, metabolism and neoplasia risk, are caused by genetic or epigenetic changes to genes that are expressed from only one parental allele. Disease may result from changes in coding sequences, copy number changes, uniparental disomy or imprinting defects. Some imprinting disorders are clinically heterogeneous, some are associated with more than one imprinted locus, and some patients have alterations affecting multiple loci. Most imprinting disorders are diagnosed by stepwise analysis of gene dosage and methylation of single loci, but some laboratories assay a panel of loci associated with different imprinting disorders. We looked into the experience of several laboratories using single-locus and/or multi-locus diagnostic testing to explore how different testing strategies affect diagnostic outcomes and whether multi-locus testing has the potential to increase the diagnostic efficiency or reveal unforeseen diagnoses. RESULTS: We collected data from 11 laboratories in seven countries, involving 16,364 individuals and eight imprinting disorders. Among the 4721 individuals tested for the growth restriction disorder Silver-Russell syndrome, 731 had changes on chromosomes 7 and 11 classically associated with the disorder, but 115 had unexpected diagnoses that involved atypical molecular changes, imprinted loci on chromosomes other than 7 or 11 or multi-locus imprinting disorder. In a similar way, the molecular changes detected in Beckwith-Wiedemann syndrome and other imprinting disorders depended on the testing strategies employed by the different laboratories. CONCLUSIONS: Based on our findings, we discuss how multi-locus testing might optimise diagnosis for patients with classical and less familiar clinical imprinting disorders. Additionally, our compiled data reflect the daily life experiences of diagnostic laboratories, with a lower diagnostic yield than in clinically well-characterised cohorts, and illustrate the need for systematising clinical and molecular data.