If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address for correspondence: Dr Simon Haefliger, Bone Tumour Reference Centre, Institute of Medical Genetics and Pathology, University Hospital Basel, University of Basel, Schoenbeinstrasse 40, 4031 Basel, Switzerland.
Cemento-osseous dysplasia (COD) belongs to the spectrum of benign fibro-osseous lesions occurring exclusively in the tooth-bearing areas of the jaws. Depending on site and extent of involvement, periapical, focal and florid subtypes can be distinguished that share an identical histomorphology. Most cases are asymptomatic and follow a self-limited course requiring no specific treatment. Over time, lesions progressively mineralise while the cellularity decreases. However, the molecular pathogenesis of COD, has not yet been explored. We analysed a series of 31 COD samples by targeted sequencing and detected pathogenic hotspot mutations involving the RAS-MAPK signalling pathway in 5/18 evaluable cases (28%). The mutations were found in the BRAF, HRAS, KRAS, NRAS, and FGFR3 genes. Our findings suggest that COD is driven by RAS-MAPK activation; however, the mechanism underlying the spontaneous growth arrest typically occuring in most of the lesions remains elusive.
Cemento-osseous dysplasia (COD) is a benign fibro-osseous lesion occurring exclusively in the tooth-bearing regions of the gnathic bones. Based on anatomical location and extent, COD can be divided into three variants: periapical COD which involves the apical areas of the mandibular anterior teeth, focal COD affecting single teeth not including the mandibular incisors, and florid COD which is associated with multifocal/multiquadrant involvement. CODs are generally asymptomatic lesions and discovered incidentally, although patients with florid COD may experience pain and discharge secondary to infection.
Radiographic assessment is essential for the diagnosis of COD and ideally, clinical presentation and imaging evaluation alone should be sufficient to reach a diagnosis without requiring biopsy confirmation. Radiographically, early CODs are primarily radiolucent but increasingly mineralise from the centre to the periphery over time.
From a histopathological point of view, all variants of COD have similar microscopic features, characterised by a variably cellular and fibroblastic appearing stroma intermixed with mineralised tissue consisting of osteoid, woven bone, and cementum-like material.
Patients with COD do not require any specific treatment or follow-up as far as they remain asymptomatic. Patients with florid COD may need to be monitored more closely as they are more prone to develop secondary osteomyelitis.
To date, molecular data on COD have not been published and COD is considered to represent a benign and generally self-limited lesion, most likely of reactive or hamartomatous nature. Since obtaining tissue for histological confirmation is generally neither required nor recommended, and when a biopsy is performed the tissue is usually scarce and decalcified, the availability of high-quality COD tissue samples for molecular analysis is limited. We assembled a cohort of 31 CODs and performed targeted DNA sequencing to search for mutations in common oncogenic drivers.
Material and methods
Tumour samples
Thirty-one formalin-fixed, paraffin-embedded (FFPE) tissue samples from patients with COD were included in the study. All cases were retrieved from the Bone Tumour Reference Center in Basel (CH). The diagnosis was confirmed by experienced pathologists with expertise in bone and soft tissue pathology. The study was approved at the University Hospital Basel, following the approval of the ethical committee for mutational analysis of anonymised samples (Ethikkommission beider Basel ref. 274/12).
Sequencing
Targeted sequencing (Oncomine Focus DNA Panel; ThermoFisher Scientific, USA) was carried out following routine protocols (for details see Supplementary Data, Appendix A) and included the hotspots of 35 genes (for the list of the genes covered with corresponding exons see Supplementary Data, Appendix A). Variant calling was performed using the Ion Reporter Analysis Software v5.16 (ThermoFisher Scientific). Copy number variants were not assessed. Polymorphisms were filtered against UCSC common single nucleotide polymorphism (SNP), ExAC, 1000 Genomes and 5000 Exomes databases. Detected sequence variants were evaluated for their pathogenicity based on previous literature, databases (COSMIC, ClinVar) and classified as pathogenic, variant of unknown significance (VUS) and benign.
Results
Clinical features
The average age at diagnosis was 39.2 years (range 20–66 years, median 39 years), and our cohort comprised 28 females and three males. Thirteen periapical CODs, 13 focal CODs, and five florid CODs were included in our study. Most lesions (27%) developed around the tip of the roots of the lateral incisors. Imaging showed primarily lytic, mixed as well as more mature and densely mineralised lesions. The average size was 21.8 mm (range 5–41 mm, median 21.3 mm). Based on the degree of mineralisation, we subdivided the samples for which imaging was available into early (n=6), intermediate (n=11), and mature (n=4) lesions. All clinical data are summarised in Table 1. Representative radiological aspects of the cases are shown in Fig. 1.
Table 1Summary of the clinical and pathological features of all cases (n=31)
Fig. 1Representative imaging findings of cemento-osseous dysplasia (COD). (A) OPG showing multiple osteolytic lesions in the entire mandible situated periapically, which show irregular mineralised internal structures. (B) CT scan of a focal cemento-osseous dysplasia, showing a periapical mixed radiolucent mass. (C) CT scan showing a florid cemento-osseous dysplasia showing multiple mixed osteolytic lesions, some have extensive mineralisation. (D) OPG of a periapical COD showing multiple mixed periapical radiolucent lesions in the mandible.
Histologically, the samples showed a similar appearance with a varying ratio of stroma and matrix components according to the state of maturation. Early lesions were characterised by a more cellular fibroblastic-appearing stroma and with rather limited osteoid, woven bone and mature lamellar bone formations. Aggregates of lobulated basophilic cementum-like matrix were also found. The more mature lesions typically revealed increasing calcification, reduced vascularity, and an only limited stroma component. Representative histological images are shown in Fig. 2.
Fig. 2Representative histology of cemento-osseous dysplasia (H&E). (A,B) Parts of a fibro-osseous lesion showing various proportions of woven bone trabeculae and cementum-like material (arrows) embedded in a moderately cellular and monomorphic fibroblastic stroma (A, low power; B, higher power). (C) The matrix of the lesion fuses with the pre-existing bone. (D) Over time, the matrix mineralises and coalesces, forming the so-called ginger root-like masses.
Of the 31 cases selected in our study, sequencing analyses failed in 13 cases because of insufficient DNA preservation, most likely due to prior acid decalcification. For the remaining cases, sequencing analyses revealed pathogenic hotspot mutations in five of 18 cases (28%). The alterations were found in five different genes, namely BRAF p.Gly466Glu, HRAS p.Gly13Val, KRAS p.Leu19Pro, NRAS p.Gly12Asp, and FGFR3 p.Gly380Arg with correspondent allelic frequency (AF) of respectively 23%, 7%, 6%, 30%, and 43%. No other genes were found to harbour mutations in our cohort. The COD cases carrying the BRAF, NRAS and FGFR3 mutations belonged to the focal subtype, while the HRAS and KRAS mutations were found in florid CODs. Of the five mutated cases, four were classified as lesions with intermediate maturation, while one lesion consisted of a mature lesion. All sequencing data are summarised in Fig. 3.
Fig. 3Summary of sequencing data. (A) Summary of the genetic alterations present in cemento-osseous dysplasia (COD) cohort (n=18), each column representing a patient and each row a gene of the RAS-MAPK pathway covered by the sequencing panel. (B–D) Illustration of three of the five genes (HRAS, NRAS, and FGFR3) altered in COD cohort with localisation of the mutations detected.
In this study we performed targeted sequencing in a cohort of patients with COD and found pathogenic mutations involving the RAS-MAP kinase pathway in 28% of cases.
Fibro-osseous lesions of the jaw include cemento-ossifying fibroma (COF), fibrous dysplasia (FD) and cemento-osseous dysplasia (COD). These entities have overlapping histological characteristics, but are usually easy to distinguish when morphology, imaging and clinical presentation are available. Despite sharing similar morphological features, the molecular landscape of these three lesions seems to be diverse. In COF, the Wnt and NOTCH pathway have been shown to be dysregulated, but without identification of recurrent genetic aberrations so far.
In contrast, FD is well known to harbour postzygotic activating mutations of the GNAS gene as the underpinning molecular event which can be used for diagnostic purpose as GNAS mutations are not present in other fibro-osseous lesions.
In another (very rare) fibro-osseous lesion of the jaw, namely gigantiform cementoma arising in the setting of gnatho-diaphyseal dysplasia, mutations in the anoctomin 5 (ANO5) gene have been detected.
Whole exome sequencing links dental tumor to an autosomal-dominant mutation in ANO5 gene associated with gnathodiaphyseal dysplasia and muscle dystrophies.
Overall, this suggests that, despite showing morphological overlap, the underlying molecular pathogenesis of fibro-osseous lesions of the jaw are distinct. Our data further support this observation as we found pathogenic driver mutations in RAS-MAPK pathway genes in 28% of our cases. This highly conserved signalling pathway is involved in a broad range of cellular functions, including cell proliferation and survival, and has not been shown to be altered in other fibro-osseous lesions of the jaws.
Remarkably, activating mutations of genes involved in this pathway are not only a feature of malignant neoplasms, but can also be seen in non-neoplastic conditions such as endometriosis and arteriovenous malformations.
In recent years, as DNA sequencing of neoplasms has become more widely used, genetic alterations that were assumed to be unique to cancer and its precursors are now being increasingly identified in lesions previously considered to be of developmental, reactive or hamartomatous origin. As a consequence, lesions like central giant cell granuloma are now mostly perceived as true neoplasms after recurrent driver mutations have been identified, although there is an ongoing debate if a driver mutation is sufficient to define neoplastic disease.
The data presented here point to a similar mechanism also in COD which had traditionally been considered a reactive/hamartomatous process but indeed seems to be driven by recurrent signalling alterations of the RAS-MAPK cascade.
A typical aspect of COD is the spontaneous growth arrest. Indeed, after an initial growth phase virtually all CODs invariably stop getting larger and progressively mineralise. Interestingly, there are other bone lesions that show a similar fate, notably non-ossifying fibroma (NOF) and FD.
Indeed, both lesions show a uniformly benign clinical course, and after an initial growth phase, they invariably cease growing and/or regress. In addition, in both conditions mutations in driver genes have been identified, GNAS alterations for FD and RAS-MAPK gene mutations in NOF. In this regard, COD shares evident similarities with these two lesions. Further, the uniformly benign and self-limiting clinical course of COD suggests that the driving mechanism—activation of the RAS-MAPK pathway—might be transient and could diminish over time. In FD it has been shown that cells carrying the driving mutation in the GNAS gene progressively undergo apoptosis over time thus explaining the difficulty to detect the mutation in more mature lesions.
A similar mechanism might account for the low frequency of mutated cases detected in our cohort. However, our data do not support a direct link between the degree of maturation and mutational status like FD as most mutated cases of our cohort exhibited features of intermediate maturation.
Another remarkable aspect of COD is its strong predilection for middle-aged women, particularly with brown skin, which has led to the hypothesis that a potential hormonal stimulus, notably oestrogen levels, may play a role in the pathogenesis of COD besides ethnic origin.
The importance of clinical and radiological findings for the definitive histopathologic diagnosis of benign fibro-osseous lesions of the jaws: study of 276 cases.
In this regard, COD share similarities with other bone lesions, e.g., NOF and osteochondroma, the clinical course of which also seems to be influenced by endocrine stimulation.
Similar observations have been made in osteochondromas, which cease growing after skeletal maturation, possibly due to a decrease in hGH and IGF1 levels.
In conclusion, our study is the first to provide sequencing data on COD. Our data suggest that COD, previously considered a reactive/hamartomatous disorder, should be regarded as a mutation driven lesion by activation of the RAS-MAPK pathway. The reason for the self-limiting course with progressive mineralisation remains unclear but might be caused by a diminishing number of lesional cells carrying the mutation over time, possibly similar to the mechanism proposed in FD.
Conflicts of interest and sources of funding
SH, DT, and DB were supported by the Foundation of the Basel Bone Tumour Reference Centre, the Gertrude von Meissner Stiftung, and the Stiftung für krebskranke Kinder, Regio Basiliensis. The authors state that there are no conflicts of interest to disclose.
Appendix A. Supplementary data
The following is the Supplementary data to this article.
Whole exome sequencing links dental tumor to an autosomal-dominant mutation in ANO5 gene associated with gnathodiaphyseal dysplasia and muscle dystrophies.
The importance of clinical and radiological findings for the definitive histopathologic diagnosis of benign fibro-osseous lesions of the jaws: study of 276 cases.
00322-1&id=gr1.jpg){kind=link}
00322-1&id=gr2.jpg){kind=link}
00322-1&id=gr3.jpg){kind=link}