A. Greenspan et Rc. Stadalnik, BONE ISLAND - SCINTIGRAPHIC FINDINGS AND THEIR CLINICAL-APPLICATION, Canadian Association of Radiologists journal, 46(5), 1995, pp. 368-379
PURPOSE: TO evaluate bone islands that showed increased uptake of radi
otracer on skeletal scintigraphy and to present an algorithm for exami
ning such lesions to avoid misdiagnosis in difficult cases. PATIENTS A
ND METHODS: Over an 8-year period, 20 patients (10 men and 10 women ra
nging in age from 33 to 82 years) with bone islands that showed activi
ty on skeletal scintigraphy were examined with plain radiography (all
patients), computed tomography (CT; 5 patients) and magnetic resonance
imaging (MRI; 4 patients). For six of the patients the clinical prese
ntation and the radiologic studies suggested malignancy, which prompte
d biopsy and histopathologic examination. Histopathologic study was al
so performed for six other patients in whom the bone islands were foun
d incidentally during evaluation for joint replacement surgery for ost
eoarthritis. in the last eight patients the lesions exhibited the char
acteristic radiologic features of enostosis, and these patients were f
ollowed for up to 3 years without biopsy. RESULTS: In all cases plain
radiography showed the characteristic features of a bone island: a hom
ogeneously dense, sclerotic focus in the cancellous bone with distinct
ive radiating bony streaks (''thorny radiation'') that blended with th
e trabeculae of the host bone to create a feathered or brush-like bord
er. Histopathologic examination of scintigraphically active bone islan
ds showed increased osteoblastic activity, and the lesions were marked
by a mixture of compact and trabecular bone. In the patients who did
not undergo biopsy but were followed with radiologic examinations, the
re was no change in the size or appearance of the lesions. CONCLUSIONS
: The key to the correct diagnosis of bone island Lies in the distinct
ive radiographic features of enostosis. An asymptomatic, isolated scle
rotic bone lesion showing feathered or brush borders is most Likely an
enostosis, regardless of its size or its activity on scintigraphy. Th
erefore, a practical algorithm for examining bone islands should now f
rom their morphologic features as observed on radiographs and CT and M
RI scans, rather than from their activity on scintigraphy.