Abstract

Heterozygous COL1A1 and COL1A2 gene variants are known to cause osteogenesis imperfecta (OI) in 90% of the patients in the Western and Japanese populations. Two previous Indian reports, a total of 49 patients, showed their proportion in the Indian population to be 44% and 71%. We studied a population of 210 children with a clinical diagnosis of OI and focused on the cohort of children with (likely) pathogenic COL1A1/COL1A2 variants. The prospective study describes the clinical, radiological, and genetic findings of 100 COL1A1/A2 variants. Phenotypic evaluation included the type of OI, fracture history, extraskeletal and skeletal features, pamidronate treatment, and its effect on bone density. Genotyping was assessed by clinical or targeted exome sequencing and validated by Sanger sequencing. One hundred patients, including affected siblings, out of 210 had COL1A1/A2 gene variants. No genetic cause was found in 25, and the remaining had causative variants in other OI-associated genes. In the group of 100 children with (likely) pathogenic COL1A1/A2 variants, the proportions of phenotypes 1, 3, and 4 were 24%, 44%, and 32%, respectively. Consanguinity was 10%, which was less when compared to the other OI-associated genes. The age of the first fracture was primarily at birth or before 6 months, particularly in severe types of OI such as Type III. Blue sclerae were observed in 84 patients, most commonly associated with OI Type I. Dentinogenesis imperfecta (DI) was present in 47 patients, typically seen in Types III and IV. Radiological features included Wormian bones in 58 patients, vertebral fractures in 46, and spine deformities in 16. Lower limb deformities were present in 90% of cases, with a higher prevalence in more severe forms of OI, while upper limb deformities were noted in 39%. Among the 22 patients receiving regular pamidronate therapy, there was a noticeable improvement in bone density, emphasizing the therapy's effectiveness, particularly in managing moderate to severe OI types. Genotypically, we report (likely) pathogenic 56 COL1A1 and 46 COL1A2 variants, including 21 novel variants. Of the 66 missense variants, 29 were in COL1A1 and 37 in COL1A2 genes. Of these, 60 were glycine substitutions (28 in COL1A1 and 32 in COL1A2) in the triple helix, the commonest being glycine to serine, associated with OI Type 3 in 28 individuals. The report highlights that 47.6% (100 out of 210 patients) of COL1A1 and COL1A2 gene variants were found in our cohort who underwent genetic analysis. This percentage is notably lower compared to other populations, where the percentage of identified COL1A1/2 variants has been reported to be higher.