Non-collagen pathogenic variants resulting in the osteogenesis imperfecta phenotype in children: a single-country observational cohort study.

Abstract

In England, children (0-18 years) with severe, complex and atypical osteogenesis imperfecta (OI) are managed by four centres (Birmingham, Bristol, London, Sheffield) in a 'Highly Specialised Service' (HSS OI); affected children with a genetic origin for their disease that is not in <i>COL1A1</i> or <i>COL1A2</i> form the majority of the 'atypical' group, which has set criteria for entry into the service. We have used the data from the service to assess the range and frequency of non-collagen pathogenic variants resulting in OI in a single country. Children with atypical OI were identified through the HSS OI service database. All genetic testing for children with OI in the service were undertaken at the Sheffield Diagnostic Genetics Service. Variant data were extracted and matched to individual patients. This study was done as part of a service evaluation project registered with the Sheffield Children's Hospital Clinical Governance Department. One hundred of 337 children in the HSS met the 'atypical' criteria. Eighty have had genetic testing undertaken; 72 had genetic changes detected, 67 in 13 genes known to be causative for OI. The most frequently affected genes were <i>IFITM5</i> (22), <i>P3H1</i> (12), <i>SERPINF1</i> (8) and <i>BMP1</i> (6). Among children with more severe forms of OI (approximately one-third of all children with OI), around 20% have pathogenic variants in non-collagen genes. <i>IFITM5</i> was the most commonly affected gene, followed by genes within the <i>P3H1</i> complex. These data provide additional information regarding the likelihood of different genetic origins of the disease in children with OI, which may influence clinical care.