Osteogenesis imperfecta: translation of mutation to phenotype.

By the end of the recently completed Fourth International Conference on Osteogenesis Imperfecta (Pavia, Italy, 9-12 September 1990) more than 70 mutations in the two genes that encode the chains of type I collagen, the major protein of bone, had been identified as the molecular cause of different forms of osteogenesis imperfecta (OI). Although by no means complete, the set of mutations in hand provides a rough guide to how to predict the phenotypic effects of mutations in type I collagen genes, predicts that certain classes of mutations will give rise to very mild phenotypes that will blend with common disorders, such as osteoporosis, and clarifies the genetic aspects of the widely used clinical classification of O.1 Clinical classification of OI For almost a century clinicians have attempted to provide a rationale by which to understand the readily apparent clinical heterogeneity in OI. The early distinction of 'congenita' forms (those with fractures at birth) from 'tarda' forms (those who develop fractures at a later period),2 followed by acknowledgment of further phenotypic variation,3 provided early guides. About a decade ago, these early classifications were largely supplanted, at least for geneticists, by the scheme developed by Sillence et all that incorporated genetic, radiographic, and phenotypic criteria. Biochemical and genetic studies have further amended and refined this classification (table ). The value of the classification is often questioned because of the difficulty in deciding where