PURPOSE: Patients referred with an osteogenesis imperfecta (OI) phenotype exhibit marked genetic heterogeneity, and a substantial proportion may harbor variants associated with early-onset osteoporosis (EOO) or OI-mimicking disorders rather than classical collagenopathies. This study aimed to define the genetic spectrum and non-COL1A1/2 variant burden in a cohort clinically diagnosed with OI. METHODS: Ninety-eight unrelated patients referred with a clinical diagnosis of OI were analyzed using a targeted NGS panel on the DNBSEQ-400 platform and, when indicated, whole-exome sequencing. Variants were classified according to ACMG/AMP guidelines, and copy-number variants were assessed by microarray (Affymetrix CytoScan Optima, Thermo Fisher Scientific). Genes were categorized as COL1A1/COL1A2-related or non-collagen bone fragility genes. RESULTS: Seventy-two variants affecting 60 distinct loci were identified. While most pathogenic variants involved COL1A1 and COL1A2, a substantial proportion of patients (approximately one-quarter) carried variants in non-collagen genes including FKBP10, WNT1, P3H1, PLS3, and SERPINF1, which are known to be associated with early-onset osteoporosis and OI-overlap phenotypes. Several of these variants were detected in the heterozygous state and were incompatible with classical autosomal-recessive OI, supporting an osteoporosis-predominant or modifier-based disease model. CONCLUSIONS: A significant proportion of patients referred with an OI phenotype genetically represent early-onset osteoporosis or OI-overlap disorders rather than true collagenopathies. Comprehensive NGS-based testing, including copy-number analysis and non-COL1A1/2 genes, enables accurate classification across the bone fragility spectrum and improves diagnostic yield.
Building similarity graph...
Analyzing shared references across papers
Loading...
Fatma Nihal Öztürk
Haseki Eğitim ve Araştırma Hastanesi
Ece Keskin
Haseki Eğitim ve Araştırma Hastanesi
Haseki Eğitim ve Araştırma Hastanesi
Building similarity graph...
Analyzing shared references across papers
Loading...
Öztürk et al. (Thu,) studied this question.
synapsesocial.com/papers/69f5941871405d493affedd9 — DOI: https://doi.org/10.1007/s11657-026-01702-z