Osteogenesis imperfecta: clinical, cephalometric, and biochemical investigations of OI types I, III, and IV.

The aim of the study was to analyze craniofacial development in 54 patients with osteogenesis imperfecta (OI), who were classified into OI types I, III, and IV according to clinical criteria, and to relate the findings to the abnormalities in collagen I production. In 33 patients, analysis of radioactively labelled procollagen was performed. Cephalometric radiographs, facial photographs, and CT-scans (a single case) were analyzed and mean facial diagrams for lateral and frontal films were produced based on registration of 221 reference points. Radiographs of 102 male and 51 female Danish students served as control material. In OI type I, size of the skull and jaws was generally slightly reduced, but morphology was within normal limits. In OI type IV and especially type III more severe abnormalities were found; the cranial base was flattened, the maxilla posteriorly inclined, and nearly all size-measurements were reduced. In OI type III the sagittal jaw relations were reduced and a mandibular overjet recorded. Three OI type I patients, whose fibroblasts produced structurally abnormal collagen I, had the stature and several features in the craniofacial region, which corresponded to those recorded for the OI type IV group. Also, in three OI type IV patients whose fibroblasts produced a reduced amount of normal collagen I, craniofacial morphology showed several features resembling type I patients. We conclude that structural abnormalities of collagen I generally give rise to more severe alterations of the craniofacial features than a quantitative defect of collagen I. OI type I patients are only slightly affected in their craniofacial region, while patients with OI type IV and especially type III are moderately to severely affected. The combined cephalometric and biochemical findings suggest that future classification of patients with osteogenesis imperfecta should be based on biochemical/molecular and radiological analyses in combination with clinical criteria rather than on clinical features alone.