Conventional imaging methods (X-ray, CT and MRI) have been used to diagnose FDB. The imaging features of FDB are characteristic, although not specific and these lesions without periosteal reaction or soft tissue involvement generally look benign (
2,
3). Additionally, FDG PET/CT; Tc 99 mMDP; Ga 67 Citrate, In 111 pentreotide, In 111 leukocyte scintigraphy; and dual phase Tc 99 mMIBI scintigraphy imaging have also been used for the evaluation of FDB (
4). New bone formation and increased vascularity suggest an active osteoblastic state within the lesions of FDB, resulting in increased uptake of bone imaging agents (
5).
Several case reports have been described that FDB can have either intense FDG activity or can be metabolically normal without any increased FDG activity on PET-CT scan. Variability of FDG-uptake in various sites in one lesion or an individual may be due to differing numbers of actively proliferating fibroblasts (
6). The significantly increased FDG uptake may mimic bone metastasis or skeletal involvement of the primary malignancy in cancer patients with FDB. Most of the studies in the literature dealing with cancer patients with FDB are presented in
Table 1. These findings may change the stage of the cancer as in our case. Toba et al. concluded that the growth of FDB lesions needed the acceleration of bone mineral turnover without an increase in glucose metabolism (
7). Charest et al. reported a patient with synchronous liposarcoma and a monostotic FDB lesion; the SUV max of the monostotic lesion (proved by biopsy) was higher (
8). Su et al. reported that early SUV max values in the lesions of FDB ranged from 1.2 to 9.6 for 11 patients and this variability may cause higher SUV max values for FDB lesions compared to malignant lesions as in Charest report (
6,
8).
On the other hand, Shigesawa et al. showed that FDG PET/CT is useful for differentiating bone metastasis from FDB in patients with a malignancy (
9). Whole body bone scintigraphy remains the best method to identify the extent of skeletal involvement especially in polyostotic FDB. Zhibin et al. (
10) reported that the diagnostic specificity of ruling out metastases with radionuclide bone scanning may be improved in association with other imaging modalities (
10).
Malignant transformation including osteosarcoma, fibrosarcoma and chondrosarcoma may also be seen in some FDB patients and the risk increases by radiotherapy (
3). Berrebi et al. (
11) concluded that the local progressive increase of the SUV max index can contribute to an early detection of sarcomatous transformation. However, Bonekamp et al. (
12) described a case of biopsy-proven fibrous dysplasia of the skull in a colon cancer patient which changed its FDG activity and CT appearance within 10 months of follow-up (
12).
CT is the best technique for demonstrating the radiographic characteristics of FDB (
3). The most common appearance of FDB on CT is an expanded bone showing a ground-glass appearance (
3,
13,
14). Strobel et al. (
15) reported that dedicated CT interpretation led to the correct diagnosis of a benign lesion (
15). However, FDB may mimic other benign fibro-osseous lesions and may even be confused with certain types of malignancies (
3,
13,
16-
18). The CT appearance of FDB is in proportion with the extent of mineralization (
18). Lesions with increased CT density may be undergoing mineralization; while in lower CT density lesions, fibroblastic proliferation may predominate (
19). Su et al. (
6) reported that the peak SUV max values in FDB lesions were often located in the area with the lowest density lesions. Other diffusely ossifying or calcifying lesions may have a similar appearance on CT scans (
18). FDB usually has sclerotic margins on CT scans and may have a matrix of uniform density (
17). Additionally, CT can reveal areas of cortical breakthrough usually associated with cortical thinning and bone expansion that are not appreciated occasionally on radiographic reviews (
17,
20-
23). Distinction between FDB, adamantinoma and ossifying fibroma can be difficult and FDB can appear very similar to adamantinoma (
13,
14,
16,
24-
26). This distinction can be made pathologically (
13). Other radiologic differential diagnoses include simple bone cyst, nonossifying fibroma, paget disease, low-grade intramedullary osteosarcoma, giant cell tumor, neurofibromatosis and osteoblastoma (
3,
13,
18).
In the light of these literature findings, it might be concluded that FDG PET/CT scans need to be interpreted in the overall clinical context, on a patient-per-patient basis. Cho et al. (
27) concluded that MRI in addition to radiography may help differentiate fibrous dysplasia from metastasis in patients with malignancy. In our case, new bone scintigraphy was performed instead of conventional imaging methods to confirm the diagnosis. The recent bone scan showed the lesions that correlated with the ones on both the previous bone scan and FDG PET/CT scan. The CT characteristics of the lesions on FDG PET/CT were also compatible with FDB.
In conclusion, if the FDG positive bone lesions are the only positive findings on PET/CT scan, except for the primary tumor in cancer patients with a past medical history of FDB; they should not be misinterpreted as bone metastases. To improve the diagnostic accuracy, such lesions need to be correlated with the previous imaging studies and biopsied if they are equivocal.