Abnormalities in Tooth Formation after Early Bisphosphonate Treatment in Children with Osteogenesis Imperfecta

Osteogenesis imperfecta (OI) is a clinically and genetically heterogeneous group of heritable disorders of connective tissue which is caused by dominant mutations in collagen type I, encoded by COL1A1 and COL1A2 in up to 90% of cases [1]. Clinical signs may include bone fragility, short stature, joint laxity, hearing loss, tendency to prolonged bleeding, bruising, blue sclera, and dentinogenesis imperfecta (DGI) [2] as well as other dental abnormalities like tooth agenesis and malocclusions [3, 4]. OI has traditionally been classified in four main types based on primary clinical and radiological findings combined with pattern of inheritance: type I (mild OI with blue sclera), type II (pre- or perinatal lethal), type III (severe deforming), and type IV (intermediate severity and white sclera after infancy) [5]. This classification is now expanded into five different phenotype groupings type 1–5 where type 5 is characterized by calcification of the interosseous membranes and/or hypertrophic callus [6]. The prevalence of OI types I, III, and IV in Sweden has been estimated at 7.4/100,000 [7].

Treatment with intravenous pamidronate (APD) started in Sweden in 1991 in children and adolescents with OI with repeated fractures, vertebral compression, and pain [8]. Treatment was initially given to adolescents with OI type III. Later, younger children were successively included and treatment indications over time changed to include infants with more severe forms of OI (repeated fractures and acquired vertebral compressions) and also children with milder OI with progressive vertebral fractures, low bone density, and pain. Similar treatments have been started elsewhere and treatment with bisphosphonates (BPs) in patients with OI has expanded and has become an important symptomatic therapy especially in moderate and severe OI [9]. The rationale for BP therapy of OI is predominantly inhibitory effect on osteoclasts which results in a net increase in bone mass and mineralization [10]. A substantial body of work has demonstrated the efficacy of these agents in preventing bone fractures, decreasing pain, and increasing the quality of life [11]. Intravenous, i.v., pamidronate, neridronate, or zoledronate is the treatment of choice for pediatric patients with moderate-to-severe OI, whereas BP treatment for patients with mild forms of OI is still discussed [12‐14]. In most previous studies, the average age of children with OI at onset of BP therapy is around 4 years, although there are reports of children who started treatment as early as 2 weeks of age [15]. During the first four years of age, the permanent teeth are under intense development and theoretically BP therapy could affect the tooth formation.

Earlier, we reported a high prevalence of tooth agenesis in individuals with OI (17%) [4] and impairment of the maturation and eruption of permanent teeth when pamidronate has been administered from infancy [16]. To our knowledge, no human studies have been published on disturbances of tooth mineralization or tooth morphology due to BP administration.

The aim of the present study was to evaluate the effect of BP therapy (i.v. pamidronate) on the development and mineralization of permanent teeth. We hypothesized that BP treatment increases the prevalence of tooth agenesis and causes disturbances in tooth formation/morphology and mineralization when administration is begun in early childhood.

The mean age at imaging in study groups 1, 2, and 3 was 10.1, 8.8, and 10.9, respectively, and in the control group, 10.7 years. Mean ages of BP treatment onset for study groups 1, 2, and 3 were 0.6, 4.1, and 7.8 years, while mean treatment duration was 9.4, 4.6, and 3.1 years, respectively. Table 1 presents the group characteristics.

Our main findings are that BP administration can affect tooth development when treatment starts before 2 years of age. Prevalences of tooth agenesis, malformed teeth, and tooth mineralization disturbances were significantly higher in children who had received BP treatment from infancy compared to children who had begun BP treatment first after the age of 2 years and those who had not received BP therapy. Tooth agenesis and DGI is most frequently seen in individuals with OI type III. DGI is not influenced by BP treatment. As OI type III is the most severe type of OI, BP treatment usually starts early in these patients. We found the prevalence of tooth agenesis significantly higher in children with OI type III who had begun treatment before the age of 2 years compared to those treated after 2 years of age or not treated.

In most previous studies, the average age of children with OI at onset of BP therapy was around 4 years. However, there are reports of children who began treatment as early as 2 weeks of age [15]. The most widely used BPs in pediatric patients are pamidronate and zoledronate. They have different in vitro potencies with pamidronate having the lowest relative potency (100) and zoledronic acid the highest relative potency (10,000). At the present time, possible negative effects of BPs on tooth development are being debated. Based on reported in vitro data relevant for dentistry, clinical use of etidronate with the lowest relative potency should cease due to its impact on tooth development, whereas use of alendronate with the relative potency between 1000–2000 seems to be safe [20]. To our knowledge, the present study is the first to focus on the impact of BP therapy on tooth development in humans.

Third molar agenesis is the most common form of agenesis and being the most variable tooth in the dentition with its formation time and morphology. The permanent third molars are excluded in all Nordic prevalence studies and in most European. Nordic studies have reported tooth agenesis prevalences, excluding third molars, of between 6 and 8% [21] and oligodontia prevalences of about 0.1–0.2% [22]. Mutations in several genes have been associated with familial tooth agenesis, among others MSX1, PAX9, AXIN2, EDA, EDARADD, and EDAR [23, 24]. Oligodontia has been reported in individuals with OI [4, 25, 26]. In a Finnish study on patients with OI before BP therapy had been initiated, Lukinmaa et al. [25] reported tooth agenesis in 18.4% of the cohort and oligodontia in 2%. The Malmgren et al. [4] study found hypodontia in 11% and oligodontia in 6% of their patients with OI. In that study, treatment with BPs was not considered; instead, the very high prevalence of hypodontia and oligodontia prompted the group to investigate whether mutations in genes other than those earlier associated with OI (COL1A1, COL1A2, and CREB3L1) were responsible for the disturbances [27]. Except for the known variants, we were unable to identify any other mutual variant related to collagen type I that could explain the phenotype of OI associated with hypodontia/oligodontia. In the present study, we found that 31 (14%) of all patients with OI had tooth agenesis, and 13 of the 31 had been treated with BPs before the age of 2 years. Furthermore, eight (4%) of all patients with OI exhibited oligodontia, and six of them had begun treatment with BPs before the age of 2 years. Oligodontia is a serious condition and negatively affects quality of life [28, 29]. Since onset of BP treatment in early childhood may cause this condition, their use should be carefully considered before BP therapy is begun in newborns and infants.

Dens invaginatus (DI) and Dens evaginatus (DE) are developmental aberrations often causing pulp necrosis if not prophylactically treated. DI forms by invagination of the enamel organ into the dental papilla prior to calcification of the dental tissues [30]. We found DI in 20% of group 1, children treated before the age of 2 years. This is high compared to a study of 3020 Swedish schoolchildren where DI was found radiographically in the upper incisors in 2.7% of the group; in 43% of these cases, DI was bilateral [31] and predominantly type I according to Oehlers [17]. DI rarely occurs in premolars. Ridell et al. [32] collected all dental records from the Department of paediatric dentistry at the Eastman Dental Institute in Stockholm with a diagnosis of DI between 1969 and 1997, (n = 131 teeth). Only one invagination was observed in a premolar (0.8%) compared with 20% in group 1 in the present study. Since DI in permanent premolars is rare, only a few case reports have been published [33].

DE is a rare developmental anomaly occurring on the crown of the tooth as an extra bump or cusp, often with a fine pulpal extension. Premolars are more frequently affected than other teeth [18]. Lin et al. [34] studied two groups of age- and sex-matched individuals, 147 Taiwanese and 147 Spanish students. DE was found in 4.1% in the Taiwanese group but none in the Caucasian group, which the authors thought could possibly suggest that DE may be common in Asian populations. We found DE in 25% in group 1 and none in the other groups. This high number of DE in Caucasians has not been previously reported.

Interestingly, besides DI and DE, striking morphological defects were observed in the second premolars of patients with OI who had received BP therapy before the age of 2 years. The cusps of the second premolars were abnormal in size and patterning, with mineralization defects of the enamel a common occurrence. To our knowledge, a similar phenotype has not been reported previously. The serial formation of teeth during development, where the second premolars are the last teeth to form and their crown development is incomplete at age 2 years, would explain the exceptional vulnerability of the second premolars [35]. In our study, BP treatment of all children in group 1 began before 2 years of age, which is earlier than in the study cohorts of most published studies. While the morphogenesis of most organs occurs during the embryonic stage, tooth formation continues postnatally, which would explain why BP therapy affects tooth morphogenesis but has no detected effect the development of other organs.

The observed associations of dens invaginatus (DI) and dens evaginatus (DE) with dentinogenesis imperfecta (DGI) are likely explained by the disturbed structure of forming dentin in DGI. This abnormal dentin may have affected the next dental epithelium and its folding which determines the tooth cusp patterns. The resulting abnormal epithelial invaginations and evaginations would finally develop to DI and DE, respectively.

The presence of developmental defects of the enamel was determined clinically, and the defects were classified as hypoplastic or hypomineralized. The buccal surface of the tooth crown was the site that was most affected. Animal studies have reported indications of disturbances in tooth eruption and mineralization due to BP administration [36‐39], but to our knowledge our study is the first to report the effect of pamidronate on enamel formation in humans.

Tooth eruption requires osteoclastic bone resorption [40]. Hiraga et al. [38] suggested that apoptosis induction of the osteoclasts by nitrogen-containing BPs (NBPs) impairs tooth eruption in zoledronate-treated rats. The study also found a possible relationship between zoledronate administration and induction of several types of dental abnormalities; it demonstrated that un-resorbed bone particles occasionally came into contact with or even compressed the cells of the enamel organ in zoledronate-treated rats. This pressure was suggested to affect ameloblasts and enamel matrix formation in the incisors, leading to enamel defects.

Developmental disturbances in tooth morphology and mineralization due to administration of etidronate and alendronate have also been reported in animal studies [36, 38, 39, 41, 42]. In a scanning electron microscope study [36] of 72 alendronate-treated newborn rats, morphological alterations such as depressions were detected along the entire enamel surface and at the cervical portion. In rats, etidronate treatment [39] was shown to cause disturbances in the ameloblasts and developing enamel, resulting in enamel hypoplasias.

Tooth development comprises highly regulated complex events characterized by cell–cell and cell–extracellular matrix (ECM) interactions [43]. These interactions regulate tooth shape and size as well as the differentiation of odontoblasts and ameloblasts, which form dentin and enamel, respectively [20]. Tooth development and eruption are accompanied by remodeling of the adjacent bone, which is deposited by osteoblasts and resorbed by osteoclasts, and serum calcium and phosphate are attracted to form hydroxyapatite in the developing tooth structure [40]. Between 30 and 70% of the absorbed i.v. administered dose of BP is deposited in the skeleton at sites of active bone remodeling [44]. For these reasons, BPs may affect tooth development and eruption.

In conclusion, BP therapy has improved the quality of life considerably [45], but the treatment does not seem to have a positive effect on dental aberrations, which are common in individuals with OI. The results of the present study indicate that pamidronate treatment started before the age of 2 years increases the risk of disturbances in tooth development. We have followed up 22 children, treated with BP before 2 years of age, until all permanent teeth could be evaluated. Although the group is small, the results indicate that serious disturbances in tooth formation occur. Collaborations with other centers could further support the findings. Individuals with OI are in need of early multidisciplinary treatment planning; therefore, disturbances in dental development should be diagnosed as soon as possible.