Despite recent advances in the understanding of the natural history and molecular abnormalities, many questions remain surrounding the progression and management of fibrous dysplasia (FD). In the absence of comorbidities, the expected behavior of craniofacial FD (CFD) is to be slow growing and without functional consequence. Understanding of the pathophysiologic mechanisms contributing to the various phenotypes of this condition, as well as the predictors of the different behaviors of FD lesions, must be improved. Long-term follow-up of patients with CFD is vital because spontaneous recovery is unlikely, and the course of disease can be unpredictable.
Craniofacial fibrous dysplasia is a genetic disorder resulting in slow growing lesions composed of immature bony and fibrous tissues that replace normal bone.
Ninety percent of lesions in craniofacial fibrous dysplasia are present before the age of five.
There is much unknown about the natural history of fibrous dysplasia.
Surgery is the mainstay of treatment in fibrous dysplasia, but the technique, timing and in some instances, indications, remain controversial.
Fibrous dysplasia (FD) is a nonheritable, genetic disorder characterized by the replacement of normal bone by immature, haphazardly distributed bony and fibrous tissues. The cause of this disorder is a gene mutation that prevents the differentiation of cells within the osteoblastic lineage. Despite these important advances in the understanding of FD, this disorder remains enigmatic, with many unanswered questions concerning its epidemiology, natural history, and management.
Monostotic and polyostotic forms of FD have classically been described, with the monostotic form representing about 70% of all patients. Most investigators consider the disease process to be monostotic even if contiguous bones are affected as long there is only 1 disease focus, as is the case in gnathic and craniofacial FD. A large European multicenter clinicopathologic study of FD determined that, of those patients with monostotic disease, the most common site of involvement was the femur. Other common sites of involvement included the ribs and the craniofacial skeleton. The incidence of craniofacial involvement in FD has been reported as 10% to 25% in the monostotic form and at least 50% to as high as 90% in polyostotic FD. Polyostotic FD is characterized by multiple foci involving several bones. The lesions more commonly affect 1 side of the body, but can be found on both. Various associated disorders are noted in the spectrum of polyostotic FD (PFD), including Jaffe- Lichtenstein syndrome (JLS), McCune-Albright syndrome (MAS), and Mazabraud syndrome. JLS is characterized by PFD and pigmented café-au-lait skin lesions, whereas MAS has the additional features of hyperfunctioning endocrinopathies. In children, the most commonly encountered endocrinopathy in MAS is precocious puberty, predominantly in girls. In adults, hyperthyroidism, acromegaly, and renal phosphate wasting are the most common endocrine complications of MAS. Mazabraud syndrome is defined by the association of FD and intramuscular myxomas.
The pathologic basis for FD is a postzygotic (ie, it occurs after fertilization in somatic cells) mutation in the GNAS-1 gene located on chromosome 20, which results in the formation of excess 3′,5′-cyclic adenosine monophosphate (cAMP) in mutated cells. The increase in cAMP is thought to prevent the differentiation of cells within the osteoblastic lineage. The mutation occurs in the 201 position, which is usually occupied by an arginine (R201) but, in FD, is replaced by either a cysteine (R201C), a histidine (R201H), or a guanine (R201G). The GNAS-1 mutation and subsequent cAMP excess impairs the ability of cells in the osteoblastic lineage to fully differentiate. The condition of cAMP excess contributes to the overexpression of interleukin (IL)-6 by mutated osteoblastic cells, which in turn activates surrounding osteoclasts, permitting these bony lesions to expand. The accepted explanation for the variability in the severity and extent of the disease is related to the stage at which the postzygotic mutation occurred. Severe disease is associated with an early mutational event such that all 3 germ cell layers are affected with a more widespread distribution of mutant cells, as seen in MAS. A mutational event occurring later in development is thought to result in a limited distribution of the mutant cells and the resulting phenotype is less severe, as seen in monostotic FD. This hypothesis also provides a biologic basis for the observation that monostotic FD does not convert to polyostotic disease over time.
The natural history of FD is difficult to predict, but a study of 109 patients over 32 years at the National Institutes of Health found that, in the craniofacial region, 90% of craniofacial lesions were present by 3.4 years. In the extremities, 90% were present by 13.7 years, and, in the axial skeleton, 90% were present by 15.5 years. The activity of lesions seems to decrease after puberty; however, there are several reported cases that FD remains or becomes active well into adulthood.
Craniofacial FD (CFD) is typically a slow-growing lesion that is often identified incidentally with routine imaging, particularly for monostotic disease, or when gradual swelling and facial asymmetry become noticeable. The behavior of the lesion may change if associated with other rapidly growing lesions such as an aneurysmal bone cyst, or in the setting of poorly controlled endocrinopathies, especially growth hormone (GH) excess. In such cases, referral to the appropriate craniofacial specialist and/or endocrinologist is recommended.
The diagnosis of FD should be based on the clinical history and physical, radiographic, and histopathologic findings. In those cases in which the location, age of patient, or other factors make it difficult to obtain a satisfactory biopsy (especially in polyostotic FD or MAS), the radiographic and clinical findings may be sufficient to provide a high probability of the diagnosis. A broad differential diagnosis may exist for monostotic FD that is influenced by the age of the patient and the variability in the radiographic appearance. In these circumstances, a bone biopsy may be prudent to confirm the diagnosis of FD. The risks of the procedure should be carefully weighed against the benefits, and an informed decision reached as to whether a biopsy is warranted. Should a biopsy be performed, it is imperative for the pathologist to have the imaging studies for review in order to differentiate the lesion from others that have similar histopathologic features. The differential diagnosis of FD is largely determined by the age of presentation, the clinical extent, and severity ( Table 1 ).
|Fibrous Dysplasia||SOD||ROD||Cherubism||JAOF||Jv Chronic Osteo||Renal OD|
|Chromosome||20q13||−||−||4p16||Few cases studied||−||−|
|Cause||Postzygotic mutation||UK||UK||SH3BP2||UK||Infectious? UK||Chronic renal fail|
|Age||Usually childhood||Childhood||Childhood||Childhood||Usually childhood||Childhood||Any|
|Site||Maxilla > mandible||Maxilla||Maxilla||Mandible > maxilla||Maxilla > mandible||Mandible||Any|
|Multifocal disease|| − in monostotic
+ in polyostotic
Radiography of lesion
|Vertically oriented trabecular pattern of bone; coarse trabeculae||Ghost teeth||Multilocular radiolucencies in posterior mandible > maxilla||Mixed opaque and radiolucent or predominantly opaque|| Periosteal new-bone reaction
Ground-glass or cotton wool
Loss of lamina dura
|Histopathology||BFOL|| Mosaic woven and lamellar bone
|Abnormal enamel and dentinal morphology||Giant cells in a fibrous proliferation|| BFOL, 2 variants:
|BFOL, occasional giant cells|
|Increased laboratory values|| +/− GH
|Extragnathic features||+/− CALMS||+/− Hypertrichosis||−||Possible if associated with other syndromes (Ramon, Noonan)||None||+ if syndrome associated (SAPHO)||Patient on dialysis|
|Alveolar ridge expansion||+||+||+/− may be soft tissue in origin (ie, gingival tissue)||+||+ if gnathic disease||+||+|
|Orodental features|| Tooth rotation
| Hypodontia within diseased site, usually premolars
Local gingival hyperplasia
Local gingival hyperplasia
| Unerupted teeth
Plain radiography, computed tomography (CT), magnetic resonance imaging (MRI), and bone scintigraphy are all modalities that have been used in the evaluation of FD. In general, the radiographic appearance can be variable because it is influenced by the proportion of mineralized tissue and fibrous tissue in the lesion. This variability results in a range of appearances from the classically described ground-glass appearance with ill-defined borders, to a mixed radiolucent-opaque lesion, to one that is primarily radiolucent ( Figs. 1 and 2 ).
A panoramic radiograph is a simple, inexpensive modality that has its greatest usefulness for those lesions occurring in the jaws ( Fig. 3 ). The classic radiographic findings include ground-glass trabeculation with loss of the lamina dura. One study suggested that superior displacement of the mandibular canal in the setting of a fibro-osseous lesion may be unique to FD. The cortex of the mandible is often described as thinned or indistinct.
CT without contrast is the best technique for showing the radiographic characteristics and extent of disease. The radiographic bone involvement on a CT scan has been characterized as ground-glass, sclerotic, pagetoid, and cystic, some of which are clearly seen in Fig. 1 . One study suggested that, regardless of the dominant radiographic pattern, there is at least some component of the classic ground-glass bone pattern that can be detected on CT. In a recent study, it was suggested that the variation of radiographic appearance of FD depends on the age of the patient. Within the first decade of life, the lesions most often appear as homogeneous, radiodense lesions on CT. As these patients enter the second decade of life, the lesions progress to a mixed radiodense/radiolucent appearance that stabilizes in adulthood but does not necessarily return to a ground-glass appearance ( Fig. 4 show this progression).
It is important to differentiate the natural progression of these lesions from the development of a secondary disorder, such as an aneurysmal bone cyst, which has been reported and may be associated with rapid swelling, vision changes, and hearing loss, in which case an updated CT should be obtained to rule out intralesional hemorrhage and compression. The head and neck, maxilla, orbit, frontal bones, and cranial base are commonly affected by FD. In the craniofacial region where overlap of anatomic structures limits the usefulness of plain radiography, CT is particularly helpful to show lesion extent and cortical thinning ( Fig. 4 ). FD abutting or compressing vital structures such as the orbit, optic nerve ( Fig. 5 ), epitympanum, and external auditory canal may result in neuropathy, obstruction, and functional compromise.
It is important to consider performing radionuclide bone scintigraphy or a skeletal survey at the initial presentation and when there is a history of fractures or endocrinopathies, in order to map the sites of the disease and survey for polyostotic involvement. Increased uptake in FD lesions is expected, but this uptake may become less intense as the lesion matures over time. Increased tissue activity is nonspecific and cannot be distinguished from the uptake seen in malignancy, which has been reported in FD. The specific diagnosis of CFD may be suggested by bone scintigraphy, particularly in cases involving the sphenoid wing or other associated facial bones as a result of the distinctive pattern of uptake, which has been described as resembling a pirate wearing an eyepatch.
Gadolinium-enhanced MRI is helpful in defining a lesion suspected to have undergone malignant transformation in the setting of FD. It is useful in the detection of irregular or focal contrast enhancement and these features may indicate malignant change within the dysplastic lesion.
Three distinct histologic patterns of FD have been described: the classic Chinese-character form, the pagetoid type, and the hypercellular type ( Fig. 6 ). Some have speculated that the different microscopic patterns may be site specific, whereas others have suggested that the differences are correlated with mechanical stress. A description of the differences between these histologic patterns is beyond the scope of this article and does not affect patient prognosis. The overall impression of classic FD histopathology is the presence of immature, non–stress-oriented, bony trabeculae enmeshed in a fibrous stroma of variable cellularity. The extensive proliferation of fibrous tissue is produced by the mutated cells of the osteoblastic lineage. The stromal cells tend to be plump and slightly spindled but do not show cytologic atypia, and variation in cellularity probably depends on the phase of the disease. Enmeshed in the fibrous stroma are thin, often curved, disconnected osseous trabeculae of woven bone with a characteristic absence of significant osteoblastic rimming ( Fig. 6 B). Because these trabeculae often fail to undergo remodeling, they seldom possess cement lines. Collagen fibers may emerge perpendicular to the bone surface along the bony trabeculae. Of surgical interest is the finding of multiple dilated vascular channels along the lesional bone trabeculae and interspersed in the stroma, which may result in brisk bleeding during surgery.
In the gnathic and craniofacial bones, there are at least 3 differences in histopathology compared with FD in other bony sites. First, it is unusual to find cartilaginous differentiation in FD of the jaw or craniofacial bones, unless the patient has a history of a fracture. Although pathologic fracture of the long bones is common in severe cases of FD, it is uncommon in gnathic or CFD. Second, in contrast with FD of long bones, the bony trabeculae in gnathic or CFD may acquire a lamellar configuration, rather than being strictly composed of woven bone. Third, it is possible to detect some evidence of osteoblastic rimming in CFD. These histologic differences do not affect prognosis, but may confound a pathologist, leading to delay or even misclassification in the diagnosis.
Through genetic amplification techniques such as polymerase chain reaction, it is now possible to test for the genetic mutation in peripheral blood samples and paraffin-embedded tissues, but the probability of detection may be proportional to the number of mutated cells and the severity of the disease. Detection of the activating GNAS mutation is useful to confirm a diagnosis in those cases in which there is diagnostic uncertainty surrounding the microscopic features. However, testing is not routine, and ideally requires samples that are either fresh or have not been subject to the routine harsh decalcification procedures. Furthermore, testing may not be easily available, which makes anticipating costs and coverage by insurance providers unpredictable.
The prognosis of FD is influenced by the clinical presentation and extent of the disease. Estimating prognosis can be difficult in FD and, in some cases, takes time to assess, but it is an important step in establishing an individualized plan of treatment. Monostotic lesions tend to enlarge in proportion to skeletal growth and, in general, the prognosis is good. Lesions are characteristically firm and expansile and, depending on their location, symptoms manifest because of compression and mass effect. Lesions surrounding foramina may show some sensory disturbances or pain. Lesions adjacent to the sinuses can grow into and completely obliterate the sinus resulting in congestion, headaches, and/or hyposmia.
In the jaws, dental shifting and malocclusion may be noted. Lesions involving the bones of the orbit are of particular interest and are discussed later. Most sites of FD become quiescent after cessation of skeletal growth, but some continue to enlarge after skeletal maturity and well into adulthood, with progressive deformity.
Prognosis in PFD is best considered as proportional to the extent of disease. The lesions in MAS, especially in the context of uncontrolled endocrinopathies, develop into larger and more persistent areas of disease and hav an increased frequency of complications. By adolescence, many patients with widespread PFD and MAS have severe deformities and functional impairment. Laboratory evaluation of biochemical markers of bone turnover, such as serum osteocalcin, total and bone-specific alkaline phosphatase, and C-terminal type I collagen cross-links have been advocated as a means of following activity and progression of disease. These markers are expected to decrease with medical therapy for FD.
Successful clinical management begins with an accurate diagnosis, careful documentation, interdisciplinary consultation (where appropriate), and thorough radiographic assessment. Estimating prognosis can be difficult in FD and, in some cases, takes time to assess, but is an important step in establishing an individualized plan of treatment. Patients and caregivers should understand that there is no cure for this disorder. A monostotic lesion is occasionally resectable; however, most treatment plans are variable, staged, and depend on the behavior of the lesions; long-term follow-up is paramount. Most lesions in the craniofacial structures deemed suitable for surgery are managed by debulking procedures. Depending on location, this can be accomplished by staged, serial shaving; more aggressive debulking maneuvers; or resection, contouring, and reconstruction. Fig. 7 shows an adult patient with a large FD lesion in the right mandible. After orthodontic preparation, he underwent a combined orthognathic surgery and debulking of the tumor.
Depending on the site of involvement, patients with CFD may benefit from the services of various subspecialists, and consideration should be given for follow-up through an interdisciplinary craniofacial team. In cases of orbital bone involvement, thorough ophthalmologic testing including a visual field assessment, color vision, visual acuity, and funduscopic examination by a neuro-ophthalmologist is necessary. Otolaryngologic and audiological assessments are necessary if there is temporal bone involvement. Patients with skull base involvement may also benefit from an evaluation by a neurologist and neurosurgeon. PFD may be associated with endocrine and metabolic abnormalities that require early diagnosis and appropriate treatment. Psychosocial well-being during the developmental years is vital, therefore children with a chronic and deforming disorder should receive pediatric psychological evaluation and prompt intervention if necessary.
Serial clinical observation may be indicated in patients in whom the lesions do not seem to progress, cause deformity, or functional impairment. Vigilant interval follow-up and radiographic assessment in an asymptomatic individual depend on the site(s) of bone involvement and the age of the patient. Those lesions involving the orbit, skull base, and temporal bone in children need particular vigilance and scrutiny that can only take place through an interdisciplinary effort.
No medical treatment is available to cure or definitively halt the progression of FD. As stated earlier, one of the effects of excess cAMP is the overexpression of IL-6 by abnormal osteoblasts. This overexpression leads to osteoclastic bone resorption, which is the rationale for treating patients with antiresorptive agents such as bisphosphonates (BP). Medical management has had a greater role in noncraniofacial FD, in which fractures and chronic pain are more common. There has been mixed success in craniofacial FD for pain reduction and to reduce the rate of growth of the lesion. With the increased risk of BP-related osteonecrosis, the unclear efficacy in the management of FD, and the unknown long-term effects of BP treatment in children, further studies are necessary before these medications can be recommended.
Surgery is the mainstay of treatment in FD, but the technique, timing, and, in some instances, indications remain controversial. In many cases, particularly in the jaws, the dysplastic bone can be successfully contoured by conservative surgical measures to approximate facial symmetry and/or relieve symptoms, without attempting complete resection ( Fig. 7 ). Surgical treatment performed at a young age may require revision surgery sooner than in cases performed on a postpubescent patient. This relative disadvantage is offset by the improvement in quality of life or relief of existing symptoms achieved through early surgical intervention. There is always some risk of regrowth, even when conservative therapy is performed after puberty. Kusano and colleagues noted that, in their long-term follow-up of 11 patients with FD, growth of monostotic FD arrested in adolescence, but PFD was less predictable.
In some anatomic locations, a shave procedure is technically challenging to accomplish, but the use of a surgical navigation system may assist in overcoming some of this difficulty. Other cases may require complete surgical excision and immediate reconstruction of the defect, such as those with an amenable monostotic lesion or those with a particularly aggressive lesion. Deciding between a procedure designed to contour versus an excisional surgery is influenced by the anatomic location of the disease, type of FD, rate of tumor growth and behavior of the lesion, nature of the symptomatic disturbance (aesthetic, functional, neurosensory), an assessment of the potential for reactivation or regrowth, consideration of the patient’s age and expectations, and the surgeon’s preference. The indications for surgery include:
Restoration of facial contour and aesthetics. Procedures to restore aesthetics can be accomplished at most ages; however, in younger patients, the risk of regrowth is greater than in older patients. Orthognathic surgery with or without contouring may be indicated in patients with facial deformity or a malocclusion with normal healing. Most studies recommend waiting until patients are beyond puberty, skeletally mature, and the lesions are quiescent before contouring and jaw surgery. In preparation for correction of malocclusion and asymmetry, presurgical orthodontic treatment can be undertaken without concern for increased activation of the disease state while plans are underway for orthognathic surgical correction at the cessation of growth. Fig. 8 show preoperative and postoperative orthognathic management.
Control of secondary disorder. Benign lesions (mucocoele, aneurysmal bone cyst ) and malignancy (usually sarcomas ) have the potential to develop in the primary lesion, and may contribute to rapid growth, symptoms, and complications. Many secondary benign entities (eg, sinus and lacrimal obstruction, nasal stenosis) require surgical treatment. Malignant entities require management through a pediatric cancer center. The treatment in these situations is dictated by the secondary disorder.
Removal of aggressive lesions. Lesions showing aggressive, accelerated behavior or those lesions associated with significant growth potential, particularly in the setting of endocrinopathies (as seen in MAS, and GH excess) are surgical candidates. There must be concomitant control and management of the underlying endocrinopathy. Complete excision is desirable, but may not be possible.
Restoration or preservation of function. FD is benign and natural progression occurs slowly. Adaptation of vital structures typically takes place and patients, particularly young children, do not often complain of symptoms. However, there are instances in which symptoms develop such as visual changes, conductive or sensorineural hearing loss, numbness, and pain. In such cases, surgical intervention, usually decompression of canal or nerve foramina, may be necessary but the risks associated with treatment must be reviewed. However, there is a dearth of studies reviewing the success of such procedures and the long-term sequelae.