Epidemiology of fibrous dysplasia

FD accounts for 5% to 10% of all bone tumors. Monostotic FD (MFD) is 6- to 10-fold more prevalent than polyostotic FD (PFD), and MAS is less common than the two. There is no gender predilection, and most cases of MFD present in the first 3 decades of life, whereas PFD and MAS tend to present earlier in childhood. Bone involvement varies and is usually unilateral, including mostly ribs and femur, whereas craniofacial involvement occurs in 50% to 100% of patients with the polyostotic form and in 10% with the monostotic variant. Among patients with skull lesions, the frontal bones are most commonly involved, followed by the sphenoid, ethmoid, parietal, temporal, and occipital bones. Advanced imaging methodologies demonstrated that ethmoid involvement is the most common, followed by the sphenoid and frontal bones. As such, the skull base is the most common site of involvement when the pathology involves the craniofacial skeleton.

Pathophysiology of fibrous dysplasia

Macroscopically, intramedullary tumors are well circumscribed and vary greatly in size. Large lesions can expand and distort the bone. Microscopically, FD comprises irregular trabeculae of woven immature bone, depicted as Chinese characters, surrounded by normal bone and invested within a cellular fibrous stroma with osteoblast progenitor cells resembling fibroblasts.

Since its first description by Lichtenstein in 1938 as a “perverted activity of the specific bone-forming mesenchyme,” great advancement has been made in our understanding of the pathophysiology of FD. In these tumors, the differentiation of bone marrow stromal cells is arrested, and the immature cells proliferate, giving rise to the fibro-osseous masses that make up FD. The molecular etiology, which leads to the arrest in differentiation, is a somatic missense mutation in the gene GNAS1 on chromosome 20. This gene encodes the alpha subunit of a stimulatory G protein-coupled receptor, Gsα. In FD tumors cells, there is a single amino acid replacement of arginine with either cysteine or histidine, which causes inhibition of the intrinsic GTPase activity of the Gsα protein. This process leads to constitutive, ligand-independent activation and accumulation of cAMP. In bone, it is as if the stromal cells are under constant stimulation, similar to the effect of a parathyroid hormone. A similar mechanism is responsible for the extraskeletal manifestations, including constitutive activation of skin melanocytes, ovarian stimulation, and hyperthyroidism. It was suggested that the burden of mutated cells in FD is influenced by growth factors and hormones and frequently declines with age, resulting in tumor arrest.

Pathophysiology of fibrous dysplasia

Macroscopically, intramedullary tumors are well circumscribed and vary greatly in size. Large lesions can expand and distort the bone. Microscopically, FD comprises irregular trabeculae of woven immature bone, depicted as Chinese characters, surrounded by normal bone and invested within a cellular fibrous stroma with osteoblast progenitor cells resembling fibroblasts.

Since its first description by Lichtenstein in 1938 as a “perverted activity of the specific bone-forming mesenchyme,” great advancement has been made in our understanding of the pathophysiology of FD. In these tumors, the differentiation of bone marrow stromal cells is arrested, and the immature cells proliferate, giving rise to the fibro-osseous masses that make up FD. The molecular etiology, which leads to the arrest in differentiation, is a somatic missense mutation in the gene GNAS1 on chromosome 20. This gene encodes the alpha subunit of a stimulatory G protein-coupled receptor, Gsα. In FD tumors cells, there is a single amino acid replacement of arginine with either cysteine or histidine, which causes inhibition of the intrinsic GTPase activity of the Gsα protein. This process leads to constitutive, ligand-independent activation and accumulation of cAMP. In bone, it is as if the stromal cells are under constant stimulation, similar to the effect of a parathyroid hormone. A similar mechanism is responsible for the extraskeletal manifestations, including constitutive activation of skin melanocytes, ovarian stimulation, and hyperthyroidism. It was suggested that the burden of mutated cells in FD is influenced by growth factors and hormones and frequently declines with age, resulting in tumor arrest.

Clinical features of fibrous dysplasia

The clinical picture of skull base FD depends on the pathologic bone’s compressive effectagainst adjacent structures. Patients are usually asymptomatic and present with a painless bony enlargement leading to deformity and asymmetry. The most common symptoms are facial asymmetry, bone mass, blurred vision, headache, epiphora, eyelid position abnormalities, loss of visual field, diplopia, sinusitis, epistaxis, and hearing loss ( Box 1 ). It is most convenient to categorize patients into 3 distinctive groups according to their symptoms: asymptomatic, headaches, and cranial nerve compression. Asymptomatic patients are usually diagnosed during the evaluation of head trauma or radicular cervical spine symptoms.

Evaluation and radiologic workup for fibrous dysplasia

Diagnosis of MFD is generally based on clinical symptoms and radiological evaluation. In contrast, patients with PFD often require bone biopsy for histopathologic diagnosis. Radiological workup usually includes computed tomography (CT) and magnetic resonance imaging (MRI). Scintigraphy can be used to assess the disease burden throughout the body. CT is efficient for assessing cranial nerve entrapment and ON compression ( Fig. 1 ). Three-dimensional bone reconstruction with helical CT gives optimal visualization of the extent of dysplastic bone in the skull base. FD has characteristic appearances on CT: a ground-glass pattern (50%), a homogeneously dense pattern (25%), and cystic variety (20%). On MRI, the signal intensity on T1- and T2-weighted images depends on several factors, including the amount of bone trabeculation and the degree of cellularity. Characteristically, lesions show low-signal intensity and well-demarcated borders on both T1- and T2-weighted images ( Fig. 2 ). The hypointense signal intensity on T2-weighted images is caused by the numerous bony trabeculae. Single photon emission CT has been reported to be sensitive in detecting the boundaries of FD involvement.

ON compression by fibrous dysplasia of the sphenoid bone. Axial CT image showing narrowing of the optic canal caused by enlarged lesser wing of sphenoid ( arrow ).

( A ) Coronal and ( B ) axial postcontrast T1-weighted MR images showing right supraorbital low intensity mass with characteristic ground glass appearance. Inhomogeneous areas might represent fibrous tissue, cyst formation, or hemorrhage.

Natural history of fibrous dysplasia

The natural history of FD has an important impact on decision making. It had been thought that FD becomes inactive after puberty, but recent publications reported that FD can occur after adolescence and progress into adulthood. Similarly, it was shown that one-third of the surgically treated patients had recurrence during adulthood. Although most tumors in patients with MFD arrest after puberty, there is a higher rate of disease progression in PFD, and most patients with MAS will demonstrate progression of bone lesion during adulthood. Lee and colleagues described 38 patients with sphenoid region FD who did not undergo surgery; 80% of the optic canals in the patients aged younger than 30 years were encased by FD, whereas only 44% of the optic canals in the patients aged 30 years or older were encased, suggesting a regression in the disease. The incidence of malignant transformation of FD is 0.4%. Sarcomatous transformation, mainly to fibrosarcoma or osteosarcoma, was reported in patients undergoing radiation therapy.

Medical treatment of fibrous dysplasia

In the absence of level-1 evidence, most data on the management of FD rely on expert opinion. Some studies investigated the utility of pharmaceutical treatment for FD. One of the most extensively investigated therapies is bisphosphonates, which inhibit osteoclastic bone resorption. Intravenous administration of pamidronate for 2 years was reported to induce rapid improvement in the degree of pain and disfigurement among 13 children with FD, but this treatment was stopped because of a severe increase in bone mass index in those patients. In another study, half of the patients with MFD reported significant improvement in symptoms and mass growth after 22 months. High-dose oral bisphosphonate was also reported to improve symptoms. Response is usually monitored with bone turnover markers, such as serum alkaline phosphatase and urinary hydroxyproline, which are not useful for diagnosis. Serial radiographs have also been used to assess treatment response, but the results are inconsistent. One study demonstrated response to treatment by the filling of osteolytic lesions and cortical thickening, whereas others showed no radiological response. Local bone mineral density has been found to be more consistent than serial radiographs in the monitoring of the response to treatment. The major drawback of these works is the short period of follow-up and the selection bias of easier cases (ie, those without major disfiguring lesions and neuropathies). The latter usually skip the consultation with endocrinologists and present directly to neurosurgeons, plastic surgeons, otolaryngologists, or ophthalmologists, depending on clinical picture. Hence, double blind, placebo-controlled, randomized trials are needed to demonstrate the efficacy of bisphosphonates for the palliation and control of tumor bulk. The role of steroid treatment has been described in a few cases and its effect was limited in time. It is recommended for use only in cases of acute visual deterioration with the intent of buying time before surgery.

Surgical intervention for fibrous dysplasia

It is generally acceptable that asymptomatic sphenoid FD does not necessitate biopsy or surgery, and that a clinico-radiological follow-up is sufficient. Surgery is advocated once patients become symptomatic (eg, cranial neuropathies, pain, or disfigurement). On the other hand, management of asymptomatic patients with radiologic evidence of ON compression is debatable (see later discussion).

Surgeons must be familiar with the course of disease and rate of recurrence before a decision is made regarding surgery. Earlier series mainly used debulking and recontouring of facial bones in patients with cosmetic deformity caused by FD. As a result, the reported recurrence rate was up to 25%. Because of such a high rate of recurrence after curettage, some investigators recommended radical excision and reconstruction for patients with orbital hypertelorism, dystopia, exophthalmos, or grotesque orbitofacial deformity. Because extensive removal of FD is associated with a low rate of tumor recurrence, this type of surgery was especially recommended for children with low predicted surgical morbidity and for symptomatic patients.

Recent advances in surgical technique have enabled more radical surgery and immediate reconstruction with good aesthetic and functional outcomes. As a result, the more conservative surgeries became less popular for the management of FD. Based on their experience with 28 patients with craniofacial FD (CFD) Chen and Noordhoff suggested classification of the disease status into 4 surgical groups: zone 1, which includes the fronto-orbital, zygomatic, and upper maxillary regions; zone 2, which represents the hair-bearing cranium; zone 3, which includes the central cranial base; and zone 4, which involves the teeth-bearing regions of the maxillary alveolus and mandible. They suggested a total tumor excision in zone 1 and conservative excision for lesions involving zones 2, 3, and 4. They recommended treatment only in the presence of symptoms for patients with lesions in zone 3. The surgical approach is determined according to the location and extent of the disease. Most lesions can be approached anteriorly, either endoscopically or via transfacial approaches or coronal flap. Solitary lesions of the ethmoidal, frontal, and sphenoid area can be safely resected by endonasal surgery.

Optic nerve involvement in fibrous dysplasia

A major issue in the management of patients with CFD is the role of optic canal decompression. In FD of the sphenoid bone, the rate of ON involvement is 50% to 90%. The degree of ON involvement varies between mild impingement of the nerve to a totally encased nerve. Cutler and colleagues reported 87 patients with CFD, all of whom had ON involvement; 83% of the nerves had more than 50% ON encasement. The high rate of ON involvement and the potential risk of visual impairment as a result of nerve compression have led many surgeons to suggest prophylactic decompression of the ON in patients who are asymptomatic. The proponents of prophylactic surgery claim that unfavorable postoperative results were obtained in patients who had long-standing ON compression and that they had atrophic optic discs. Another of their arguments was the rapid course of permanent visual deterioration in patients with radiologic evidence of ON encasement, with visual impairment having been reported in 20% to 80% of patients with ON encasement. Based on these results, Chen and colleagues argued that if the optic canal is radiologically involved, one-third of the patients will have visual deficits and up to two-thirds will report some degree of visual disturbance.

In contrast to earlier reports, later cross-sectional studies challenged the high risk of ON atrophy in patients with CFD. In 2 sentinel articles, Lee and colleagues and Culter and colleagues reported that the vast majority of asymptomatic patients with CFD will maintain normal vision despite radiological evidence of ON involvement. These unexpected results described normal vision even in patients with total encasement of the optic canal. The majority of the patients in this group had MAS; and Cutler and colleagues found significant correlation between increased growth hormone secretion and optic neuropathy (a relative risk of 3.8). ON compression can also result from secondary lesions, including cystic fibrous dysplasia, mucoceles, hemorrhage, and aneurismal bone cysts.

ON decompression is not risk free, and cases of blindness and visual deterioration in asymptomatic patients have been reported. The high risk of ON injury was attributed to the sensitivity of the compressed nerve to surgical insult. Therefore, partial decompression of the optic canal was suggested for reducing the pressure on the nerve in symptomatic patients. The vision of such patients is expected to remain stable or improve after surgery. Even in patients with bilateral FD and unilateral optic neuropathy, the risk for blindness in decompression of the asymptomatic side must be given serious consideration.

The authors analyzed all the relevant cases published in the English literature to evaluate the efficacy and outcome of decompression in asymptomatic patients with optic canal narrowing caused by FD. A PubMed, Google scholar, and the Cochrane Central Register of Controlled Trials search was conducted through September 2010 using the Medical Subject Headings terms (fibrous dysplasia) AND (orbit OR vision OR visual OR optic OR exophthalmus OR diplopia OR proptosis). Reference lists of selected articles were manually searched for additional publications. Two reviewers (M.A. and Z.G.) independently reviewed titles and abstracts of search results to determine if the content was related to the study outcome. Studies that were identified for potential inclusion were assessed by using a priori inclusion criteria and standardized forms.

The inclusion criteria according to the study design were any randomized controlled trial, a prospective cohort, a retrospective cohort, case-control study designs, case reports, and case series. Criteria for study population inclusion were

• 1.
  

  Preoperative or postoperative histopathology diagnosis of FD

  
  
• 2.
  

  Radiologically demonstrated optic canal narrowing

  
  
• 3.
  

  Pretreatment and posttreatment visual status (based on visual fields and visual acuity or patient visual status report)

  
  
• 4.
  

  At least 4 months of follow-up.

All studies were rated by the Oxford evidence-based guidelines. The authors excluded publications that did not report numeric data in a format conducive to a meta-analysis. The investigators of the included articles were contacted by email for further information about patient follow-up and outcome. Publications in a language other than English could not be translated because of resource constraints and were excluded. The two reviewers independently abstracted information from each included study. Discrepancies were resolved by discussion and consensus.

The search of the literature resulted in 248 articles. One hundred sixty-seven available full texts were screened. Forty-six relevant articles were reviewed. Seventeen articles were excluded because of insufficient data or for not meeting the inclusion criteria, 1 article was excluded because of duplicate publication, and 1 article in the conservative treatment group was excluded because of high-dose bisphosphonate treatment. A total of 198 patients were included, 88 of whom (44%) had MAS. The average age was 20 years (range 8–40), and the average follow-up was 67 months (range 6–228). Gender details were not available for the FD subgroups. A total of 295 ONs were involved, representing 92 cases of bilateral involvement and 5 cases of recurrence. One-half of the patient population (n = 99) had ON decompression (71 therapeutic and 28 prophylactic), and the other half were followed conservatively with serial radiographs and ophthalmologic evaluations.

Symptomatic patients who underwent surgical decompression had a 65% rate of stable vision during long-term follow-up. In asymptomatic patients, the results of this meta-analysis revealed stable vision in 87% of patients undergoing optic canal decompression and in 97% of the patients that did not undergo surgery ( P <.001). There was no significance difference in the age and mean follow-up period between the two groups of patients.