Main

Fibrous dysplasia and ossifying fibroma of the jaws are fibro-osseous lesions with different clinical course and treatment strategies.1 Ossifying fibroma is a benign tumor thought to arise from the periodontal ligament,2 which can occur in almost any bone in the craniofacial region, predominantly in the jaws. The clinical course of the tumor varies from indolent to aggressive progression.3 Generally, it is slow-growing and incidentally diagnosed by routine dental examinations, but in some instances, it can be destructive, causing facial deformity, sinus obstruction, proptosis, infection and intracranial complications, as a result of which complete surgical removal is needed.4 Fibrous dysplasia is a benign dysplastic disease of the bone, which occurs in three forms: monostotic fibrous dysplasia, which involves one bone; polyostotic fibrous dysplasia, which affects multiple bones; and McCune–Albright syndrome, in which polyostotic fibrous dysplasia is accompanied by cafe-au-lait spots or hyperfunctioning endocrinopathies.5 No bone can be spared.6 According to Hart et al’s report,7 90% of the total body disease skeletal burden is established by age 15 years, and most of the monostotic fibrous dysplasia tend to stop growing when skeletal maturity has been attained;1 thus, it is best to perform bone contouring subsequent to growth arrest of the lesion.8 Differential diagnosis of fibrous dysplasia and ossifying fibroma is of great importance for their treatment and prognosis.

According to the latest WHO classification, there are two important features that can be used to differentiate ossifying fibroma and fibrous dysplasia. First, ossifying fibroma is a radiologically and histologically well-demarcated lesion, whereas fibrous dysplasia lesional bones merge with its surroundings. Furthermore, fibrous dysplasia can be diagnosed by its typical histological characteristics: isolated trabeculae of woven bone generally without rimming of osteoblasts.9, 10 In addition to the WHO criteria, it was reported that bundles of collagen fibers oriented perpendicular to the bone surface, compatible with Sharpey’s fibers, were characteristic of the fibrous dysplasia lesion.11, 12 However, not all fibrous dysplasia or ossifying fibroma cases exhibit these classic features; instead, they can present overlapping clinical, radiographic and histological features posing diagnostic challenge to the clinicians and pathologists.4 Under such circumstances, investigation at the molecular levels may be useful.

It has been well established that fibrous dysplasia is associated with postzygotic activating mutations of the GNAS gene, encoding the α-subunit of the stimulatory G-protein Gs (Gsα).13 Somatic mutations at Arg201 and Gln227 codon of Gsα have been identified in many fibrous dysplasia lesions, but absent in ossifying fibroma lesions,14, 15 which points to a possible role of the mutational analysis in differentiating these two conditions. Furthermore, Toyosawa’s study also suggest that polymerase chain reaction (PCR) analysis with peptide nucleic acid (PNA) for GNAS mutations at the Arg201 codon is a useful method to differentiate between fibrous dysplasia and ossifying fibroma.1

To further explore the role of GNAS gene mutational analysis in differential diagnosis of fibrous dysplasia and ossifying fibroma, we examined both Arg201 and Gln227 codon in 30 patients with fibrous dysplasia and 21 with ossifying fibroma in the jaws using PCR and direct sequencing. A meta-analysis was also conducted from the published literature evaluating the GNAS mutations in fibrous dysplasia and ossifying fibroma in both the jaws and extragnathic bones. In addition, a mutation of GNAS gene was detected in one case with fibro-osseous lesions with overlapping clinical and pathological features of fibrous dysplasia and ossifying fibroma, thus confirming a diagnosis of fibrous dysplasia. Our results demonstrate that the mutational analysis of GNAS gene could be a clinically feasible method in differentiating fibrous dysplasia and ossifying fibroma of the jaws.

Materials and methods

Patients and Samples

A total of 30 cases of fibrous dysplasia and 21 cases of ossifying fibroma arising from the jaws were retrieved from the repository of the Department of Oral Pathology, Peking University School and Hospital of Stomatology from 2005 to 2011. Under an institutionally approved protocol, fresh tissues from the bone lesions were obtained during the surgical removal procedure. Once collected, all the specimens were kept at −80 °C. In addition, one extra case diagnosed as ‘fibro-osseous’ lesions with overlapping pathologic features of fibrous dysplasia and ossifying fibroma was also retrieved from our files. The formalin-fixed, paraffin-embedded tissues of the patient were obtained for mutational analysis. All of the cases were re-evaluated and confirmed by three experts according to the current histological, radiographic and clinical criteria for fibrous dysplasia and ossifying fibroma.9 The detailed information of these cases was listed in Tables 1 and 2.

Table 1 The clinical features and GNAS mutations in patients with fibrous dysplasia
Full size table
Table 2 The clinical features and GNAS mutations in patients with ossifying fibroma
Full size table

Mutational Analysis of GNAS Gene at Arg201 and Gln227 Codon

Genomic DNA was isolated from tissue samples as described above using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA, USA) according to the manufacturer’s instructions. For all patients, mutational analysis was undertaken by direct DNA sequencing of PCR-amplified target sequence of the GNAS gene. DNA (200 ng) was amplified in a standard 100-μl PCR reaction mixture using GoTaq Green Master Mix (Promega, Madison, WI, USA) according to the manufacturer’s instructions. A 270-bp fragment of the GNAS gene including the Arg201 codon was amplified using the following primers: forward, 5′-TGACTATGTGCCGAGCGA-3′ and reverse, 5′-AACCATGATCTCTGTTATATAA-3′,13 while another 316-bp sequence of the GNAS gene including the Gln227 codon was amplified using the following primers: forward, 5′-GACCTGCTTCGCTGCCGTGT-3′ and reverse, 5′-AGCCAAGAGCGTGAGCAGCG-3′. The optimized PCR procedure was as follows: denaturation at 94 °C for 15 min, 35 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C (for 270 bp sequence) or 65 °C (for 316 bp sequence) for 30 s and extension at 72 °C for 30 s, with a final extension at 72 °C for 7 min. The PCR products were purified by DNA purification system (Promega) and sequenced using an automated DNA sequencer model 373 (Applied Biosystems, Foster City, CA, USA).

Meta-Analysis

We searched the PubMed (National Library of Medicine) database from 1966 to July 2012 using the following terms: ‘GNAS’ OR ‘GNAS1’ and ‘fibrous dysplasia’, ‘GNAS’ OR ‘GNAS1’ and ‘ossifying fibroma’. Reference lists of retrieved articles were hand-searched for further publications. Two reviewers independently performed the literature search and evaluation. Papers were rejected at the initial screening if the articles were published in a language other than English or titles/abstracts showed that they were clearly irrelevant or the mutation analysis were not examined in the lesional bones. Full-text versions of potentially relevant articles were obtained and reviewed to assess their suitability for inclusion in this study. Study selection process was described in Figure 1.

Figure 1
figure 1

Flow chart of the selection process for meta-analysis.

Full size image

Statistical Analysis

Statistical analysis was performed using the SPSS for windows (version 11.0) statistical software package. Descriptive statistics were used as appropriate.

Results

Clinicopathological Features

The clinical characteristics of patients enrolled in this study were summarized in Tables 1 and 2. Of the 30 patients with fibrous dysplasia (13 females and 17 males), 13 occurred in the maxilla, 4 in the mandible, 8 patients presented both mandibular and maxillary lesions, 5 cases showed multiple bone lesions affecting both gnathic and extragnathic bones and 4 of whom were diagnosed as McCune–Albright syndrome. The onset age ranged from 5 to 40 years with a mean of 12.1±6.9 years, and the age at operation ranged from 9 to 47 years with a mean of 22.6±8.6 years. The duration ranged from 2 to 27 years with a mean of 10.5±5.5 years. The 21 cases with ossifying fibroma (7 males and 14 females) included 9 occurring in the maxilla and 12 in the mandible. The onset age ranged from 1 to 46.7 years old with a mean of 23.7±14.9 years, and the age at operation ranged from 1 to 47 years with a mean of 25.3±14.6 years. The duration ranged from 0 to 12 years with a mean of 1.6±2.7 years. Histologically, fibrous dysplasia was composed of fibrous stroma containing irregular-shaped woven bone generally without obvious osteoblastic rimming (Figure 2a). Ossifying fibroma featured as spherical and small bone spicules resembling normal cementicles, which were present in the periodontal ligament (Figure 2b).

Figure 2
figure 2

Histologic features of fibrous dysplasia and ossifying fibroma. (a) Fibrous dysplasia is featured as irregular trabeculaes of woven bone within fibrous stroma, no osteoblasts could be seen around the bone. (b) Ossifying fibroma showed calcified spherules similar to cementicles, which lie in a moderately cellular, dense fibrous stroma. Magnifications: × 40.

Full size image

GNAS Mutations

The results of GNAS mutational analysis were shown in Tables 1 and 2. A mutation of Arg201 codon of Gsα protein was found in 27 of the 30 (90%) cases of fibrous dysplasia, with a predilection of Arg-to-His (p.R201H) substitutions (Figure 3b, 19 cases, 70%) as opposed to Arg-to-Cys (p.R201C) substitutions (Figure 3c, 8 cases, 30%). The rarely reported mutation of Gln227 was not detected, and no mutation was detected in all 21 cases of ossifying fibroma.

Figure 3
figure 3

Mutational analysis of GNAS at the Arg201 codon. (a) The wild-type sequence of GNAS at the Arg201 codon. (b) The missense mutation at the Arg201 codon (arrow), CGT>CAT (p.R201H). (c) The missense mutation at the Arg201 codon (arrow), CGT>TGT (p.R201C).

Full size image

A case report

A 23-year-old woman was referred to our hospital complaining of an asymptomatic expansion of the left mandible for more than 10 years with a chronic progression. The patient took no medical treatment during the past 10 years. Physical examination revealed a painless, hard and immobile mass of size approximately 4.0 cm × 4.0 cm in the left mandible. The skin overlying the expansion appeared intact. Intraorally, a firm mass was present in the left mandible with buccal and lingual expansion, there was no sign of tenderness and the oral mucosa was normal. On the basis of the above clinical findings, a diagnosis of benign tumor was suggested.

Panoramic radiograph (Figure 4a) showed an enlargement of ramus and corpus of the left mandible with a ‘ground-glass’ appearance, which was consistent with fibrous dysplasia. However, the cystic change with a sclerotic margin in the corpus made the diagnosis of fibrous dysplasia difficult.

Figure 4
figure 4

GNAS mutational analysis in one case with overlapping clinicopathological characteristics of fibrous dysplasia and ossifying fibroma. (a) The panoramic radiograph revealed an expansion of ramus and corpus of the left mandible with a ‘ground-glass’ appearance. Cystic change with a sclerotic margin could be seen in the corpus. (b) Histologic features of the lesion: the low-power view ( × 40) showing small, round and disconnected bone lying within a cellular fibrous stroma, which was abundant compared with the area of the bone; osteoblasts could be seen around the bone surface as revealed in the high-power view (inset, × 200). (c) The sequence of polymerase chain reaction (PCR)-amplified product showed a mutation at the Arg201 codon (arrow), CGT>CAT (p.R201H).

Full size image

The patient underwent conservative surgery with trimming of the affected bone. The histological features of the removed lesion showed (Figure 4b) a cellular fibrous stroma within which were small, irregular and disconnected bone spicules (somewhat resembling cementicles), and part of these spicules were rimmed with osteoblasts, which was the characteristic of ossifying fibroma.

Owing to the overlapping features of fibrous dysplasia and ossifying fibroma described above (the radiologic appearance was suggestive of fibrous dysplasia, but the histopathology revealed features consistent with an ossifying fibroma), a diagnosis of ‘fibro-osseous’ lesion was then made. By GNAS mutational analysis of exons 8 and 9, we identified an R201H mutation in the lesional bone (Figure 4c), which confirmed a diagnosis of fibrous dysplasia. Because of the conservative surgery (trimming) of the lesion, the patient was followed up for one and half years postoperatively. The lesion was stable with no apparent enlargement.

Meta-analysis

The flow chart of the meta-analysis was present in Figure 1. Initially, 157 publications were retrieved, including 155 reports from PubMed and 2 additional articles from the reference lists. After a screening of the title and abstract available, 114 articles were excluded, of which 1 was excluded for the duplication, 101 due to irrelevancy, 2 published in languages other than English, 3 for unavailable full texts and another 10 were excluded because the mutations were not examined in the lesion tissues. After reviewing the remaining 43 full texts available, 24 articles including a total of 330 patients (307 cases of fibrous dysplasia and 23 cases of ossifying fibroma) were analyzed. The details of the 24 studies are summarized in Table 3.1, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35

Table 3 The characteristics of studies included in meta-analysis
Full size table

Among the 24 studies, the numbers of cases ranged from 1 to 64, various techniques have been used for the detection of GNAS mutations, such as reverse transcription PCR and clone sequencing, conventional PCR and direct sequencing, PCR-restriction fragment length polymorphism, pyrosequencing, PCR with mutation-specific restriction enzyme digestion and so on. A variety of materials were used for the extraction of genomic DNA, including fresh bone biopsy, fresh bone tissue, lesional stromal cell cultures, formalin-fixed, paraffin-embedded tissues and so on. All of these studies examined the exon 8 of GNAS, six of which also studied the exon 9.17, 18, 19, 21, 23, 34 One study examined the exon 7, as well as exons 10–13.19

Totally, the overall positive rate of GNAS mutation in fibrous dysplasia was 86%. The major types of GNAS mutation were R201H and R201C, except for one report22 that did not mention the type of mutation. The incidences of R201H and R201C were 53% and 45%, respectively. The other types of missense mutations rarely encountered were Q227L in three cases, R201S in one case and R201G in one case.

In the 13 studies that mentioned the types of the bone involved,1, 13, 16, 17, 19, 21, 23, 27, 28, 30, 31, 33, 34 the positive rate of GNAS mutation was 84% in the extragnathic bones and 78% in the craniofacial bones.

GNAS mutation analysis in ossifying fibroma was performed in three reports1, 14, 15 including a total of 23 cases, and no mutations were detected.

Discussion

According to the latest WHO classification,9 the fibro-osseous lesions in the oral and maxillofacial region referred to a group of entities including fibrous dysplasia, ossifying fibroma and osseous dysplasia. Usually, the diagnosis of these fibro-osseous lesions could be made solely based on a combination of clinical, radiological and histological estimation, but under some situations, it still can be challenging, especially for fibrous dysplasia and ossifying fibroma.

Our results demonstrated that mutational analysis of GNAS gene at codon 201 (exon 8) and codon 227 (exon 9) by direct sequencing is a rapid and effective method to differentiate ossifying fibroma and fibrous dysplasia of the jaws. GNAS mutations were specific to fibrous dysplasia and could be detected in most of the fibrous dysplasia cases (90%), whereas no mutations were detected in ossifying fibroma. Of the GNAS mutations in fibrous dysplasia, R201H was the primary mutation type with a ratio of 70%, whereas R201C mutation accounted for 30%. The results of meta-analysis further substantiated our findings, which revealed that an overall of 86% of the fibrous dysplasia cases from both the extragnathic and the craniofacial bones had GNAS mutations, including R201H, R201C, R201S and Q227L, with R201H and R201C being the two most frequent types. On the basis of the above findings, the molecular analysis might be a helpful method when the fibro-osseous lesions in the jaws were difficult to diagnose. By using this method, we successfully identified one case with overlapping clinical and histological features of fibrous dysplasia and ossifying fibroma, and the diagnosis of fibrous dysplasia was established based on the presence of GNAS mutation.

Notably, three cases with fibrous dysplasia in this study had no detectable mutation, which indicated that the diagnosis of fibrous dysplasia could not be ruled out when no mutation could be detected, mainly because of the technical concerns regarding regular PCR and direct sequencing, which requires high quality and quantity of DNA, and also a mutant threshold of about 20% in the total population;15 however, the somatic nature of the mutations in fibrous dysplasia may not meet this level of sensitivity in some cases, especially for the older ones, as reported by Kuznetsov et al13 that the percentage of mutated cells within a given lesion may decrease with age. Two of the three fibrous dysplasia cases with no detectable GNAS mutations in this study were older than 40 years of age, probably caused by the decreasing percentage of the mutant cells. To raise the detection rate of mutant GNAS in samples, a few methods had been employed, such as pyrosequencing, PNA-clamping PCR and multiple rounds of nested PCR in conjunction with restriction endonuclease treatment. However, because of the complexity, none of these mentioned methods could be a good candidate for routine clinical examination. Compared with the above methods, regular PCR and direct sequencing employed in this study was a relatively easier and more practical method for the detection of multiple mutations in routine practice. In some cases, this approach was proven to be sensitive enough even when DNA samples extracted from formalin-fixed, paraffin-embedded tissues were used.

Taken together, the present large series as well as a systematic review of the literature confirmed the role of GNAS mutational analysis (including both exons 8 and 9) in differentiating fibrous dysplasia and ossifying fibroma by using regular PCR and direct sequencing. However, considering the mosaic features of fibrous dysplasia and the limitations of the method, diagnosis of fibrous dysplasia could not be ruled out when no mutations are detected.