Printer Friendly

Nonneoplastic lesions that simulate primary tumors of bone.

The skeleton is affected by many nonneoplastic processes that may be interpreted clinically and radiographically as either benign or malignant primary bone tumors. Even if a benign or nonneoplastic process is suspected, however, it can sometimes present with histology containing unusual or unexpected elements, such as pathologic fractures, which could be unappreciated or unrecognized and risk misdiagnosis without an awareness of the accompanying clinical and imaging findings. For example, in cases with an associated fracture, a callus may resemble osteosarcoma. In such instances, imaging may resolve the dilemma and help to avoid more prolonged investigations or interventional procedures. Pathologists should be familiar with the clinical and radiologic, as well as the histologic, features of the most commonly biopsied nonneoplastic lesions of bone, along with the sex, age groups, and anatomical sites those lesions favor, all crucial data that must be considered when generating differential diagnoses.

Other frequently overlooked aspects in bone disease are the types and duration of symptoms. An acute onset of bone pain has different implications than does chronic or absent pain. Acute symptoms may be an early sign of an infectious or aggressive neoplastic process. A pathologic fracture in a preexisting, benign, and unsuspected lesion may also present acutely. Although a radiograph is not always helpful in differentiating acute osteomyelitis from an aggressive intramedullary neoplasm, such as Ewing sarcoma, it may be effective in distinguishing an associated fracture in an underlying benign lesion or in detecting a focal soft tissue mass. The precise anatomic sites of lesions (diaphyseal, metaphyseal, or epiphyseal and intramedullary versus cortical or periosteal) also require consideration. Lesions centered in the epiphysis will probably be different from radiographically similar lesions centered in the metaphysis or diaphysis.

Conventional radiographs remain the gold standard and first step in imaging a bone lesion. Besides revealing its position in situ, radiographs offer invaluable information, such as lesional interaction with host bone and its contents, including calcification. A bone lesion that radiographically has sclerotic, beveled, or otherwise well-defined borders is usually benign, whereas infiltrative and florid periosteal reactions are more common in malignant or infectious processes.

However, the inability of conventional radiography to detect early bone changes often requires the use of alternative investigative modalities, such as computed tomography (CT) and magnetic resonance (MR). The former, besides being able to display the cross-sectional anatomy of the affected region, also allows fine definition of cortical penetration and periosteal reaction, as well as permitting a more precise assessment of the lesional matrix and associated bone destruction. Magnetic resonance, on the other hand, offers better soft tissue contrast than do CTs and has better definition of perilesional bone marrow edema. Additionally, MR has the advantage of avoiding radiation exposure, which may be important in the pediatric population or when large areas of the body or an extremity need to be studied.


The pathologist's position is, in many ways, similar to that of a lens focusing on a single spot with multiple beams of light. A review of clinical information, a surgeon's personal impression before and during surgery, and imaging findings should be superimposed on the glass slide, thereby tempering microscopic findings to achieve the most meaningful final diagnosis. For example, from clinical and radiologic interpretations of cartilaginous neoplasm such as enchondroma versus well-differentiated chondrosarcoma, the pathologist should extract a diagnosis that includes prognostic implications as well as possible further treatment and follow-up.

This review centers on several nonneoplastic entities that pathologists routinely encounter under the microscope. It does not aim to be complete but emphasizes the experience that any process affecting bone may be a target for biopsy. For simplicity, we have based and divided the following discussion on the radiographic appearances of our most commonly encountered lesions.


Unicameral Bone Cyst

Unicameral bone cyst (UBC), also known as simple bone cyst or solitary bone cyst, as the names implies, is generally a unilocular and solitary lesion, which may appear multilocular. It represents approximately 3% of biopsied bone lesions and tends to affect the metaphysis of skeletally immature patients. More than 80% of cases are diagnosed in the first 2 decades of life. Males are more commonly affected than females (male to female ratio, approximately 3:1) (1) The etiology of UBC is uncertain; different hypotheses entertained include venous obstruction (2) and intraosseous synovial cyst. (3) However, because UBCs almost exclusively affect skeletal areas associated with the most rapid longitudinal growth and originate near or in contact with the epiphyseal plate, a local growth aberration is most likely. (4) Rare cases have been directly associated with trauma (eg, gunshot wounds with retained bullets), (5) and UBCs may involute and heal spontaneously. (1)

A UBC may be asymptomatic or discovered incidentally, but many are discovered because of pain, local swelling, or occasionally, a pathologic fracture. The UBCs have a predilection for the metaphyses of long tubular bones, with 75% of lesions seen in the humerus and femur.

Radiographically, UBC is a central, intramedullary radiolucent lesion, which thins the cortex symmetrically and is marginated by a thin zone of sclerosis (Figure 1, A). In the young, a UBC is near the epiphyseal plate, migrating distally in time to abut or involve the diaphysis. (4) This indicates that UBC is a dynamic lesion, arising and growing near the epiphysis in its "active" phase. It then enters a "latent" phase and is displaced away from the epiphysis by the normal growth process of a tubular bone. (4)

A UBC tends to have a conical shape, with a wider or "ice cream cone" base paralleling the epiphyseal plate. In major tubular bones, the transverse diameter of the cyst usually does not exceed the width of the growth plate. (6) This is not necessarily true in narrow tubular bones, like the fibula, where symmetrical expansion of its bone profile may be seen. A fracture may modify the radiologic picture. It can also assist in detecting a so-called fallen fragment sign, a detached bony fragment that migrates superiorly or inferiorly in a raised or lowered extremity to serve as direct evidence of the lesion's cystic nature. (7) Either a CT or an MRI scan may confirm a UBC's fluid-filled content, but the MRI is not as reliable in identifying subtle calcification.

Pathologically, in rare cases with resected specimens available, the cyst appears unilocular with prominent internal bony ridges that are filled with serous to blood-tinged fluid. Its wall is lined by a delicate, fibrous membrane (Figure 1, B) containing amorphous material that can calcify with a cementum-like appearance. (3) Cases that have undergone fracture may also contain hemosiderin pigment and osteoid and giant cells within the cyst wall. Most commonly, specimens consist of curetted bony fragments, whose only diagnostic clues are sparse fragments of thin, fibrous membranes adherent to bone. When the diagnosis is not straightforward, other cystic lesions, such as aneurysmal bone cysts (ABCs), intraosseous ganglia, and occasionally, fibrous dysplasia with cystic degeneration, enter the differential diagnosis.


Classically, ABC is a lytic, "blow-out," intramedullary bony lesion exhibiting explosive growth with eccentric expansion with its transverse diameter wider than the epiphyseal plate. Pathologically, an ABC is characterized by blood-filled spaces lined by a thick and more-cellular membrane, with a prominent multinucleated giant cell component. Conversely, a UBC membrane is delicate and composed of fibrous tissue often containing a cementum-like material, never seen in ABC.

Intraosseous ganglia are usually smaller radiolucent lesions in the epiphysis and subchondral region; areas not involved by UBC. They are commonly seen in the femoral head, proximal and distal tibia and fibula, and the medial malleolus in some series. (8,9) They are generally asymptomatic and discovered incidentally in young as well as mature skeletons. They contain mucoid material, not present in a UBC, which contains thin, serous fluid. Intraosseous ganglia when associated with osteoarthrosis may be difficult to divorce from large subchondral cysts. They are, therefore, more confidently identified radiographically in the young without joint space narrowing or productive changes. Other differential lesions to be considered are described below.

Fibrous dysplasia sometimes undergoes cystic degeneration, but the presence, at the periphery of the cystic space, of cellular fibrous tissue with sparse osteoid spicules and no osteoblastic rimming is characteristic of fibrous dysplasia and is never seen in UBC.

Treatment modalities vary from curettage and bone grafting to trephination with injection of allogeneic demineralized bone matrix and autogenous bone marrow. (1,10) Recurrence occurs in 20% to 50% of cases. (1)

Aneurysmal Bone Cyst

The ABC is a biologically benign lesion that may be associated with worrisome clinical and radiographic features. It is characterized by blood-filled spaces lined by fibrous septa containing osteoclast-like multinucleated giant cells and delicate spicules of osteoid.

The ABC was so named by Jaffe and Lichtenstein (4-11) in their article on UBC, however, it had first been described by Ewing in 1940, (77) who named it benign bone aneurysm and considered it a benign variant of aneurysmal giant-cell tumor. Numerous hypotheses on its pathogenesis had been proposed earlier including the concept of primary and secondary ABC. (12) Lichtenstein (11) favored its origin from a local circulatory disturbance. Jaffe (13) thought it was a modification of a bone lesion whose identity had been destroyed by hemorrhage.

Most affected patients are in the first 2 decades of life with 80% younger than 20 years. (14-16) The most common complaint is pain, usually of less than 6 months' duration, often with swelling and loss of function of the adjacent joint. Pathologic fracture can occur, which in the spine, may give rise to neurologic symptoms. (17)

An ABC may occur in almost any portion of the skeleton. Long tubular bones are affected in more than 50% of cases, with the region about the knee most commonly involved. The spine is affected in 12% to 30% of cases, (17) and the pelvis in about 50% of flat bone cases. (16) Radiographically, ABC appears as an expanded, radiolucent, eccentric, "blown out" lesion, in a long bone metaphyses (Figure 2, A), or in a vertebral body or its posterior arch. An ABC's margins are well defined by a more or less continuous shell of reactive periosteal bone. Less frequently, ABC presents with less-prominent bone expansion and appears symmetric. In short, tubular bones it may appear central because of their narrow diameters. Both CT and MRI scans may highlight internal septations and multiple fluid-fluid levels. (16) Bone scintigraphy usually shows increased radionuclide uptake.

Most reported cases are classified as primary, but approximately 20% to 30% is secondary to an identifiable, preexisting lesion. (15,18) Primary lesions most frequently associated with ABC (in order of frequency) include giant cell tumor of bone, chondroblastoma, chondromyxoid fibroma, and fibrous dysplasia, particularly in pregnancy. (15) Malignant bone tumors may also be associated with ABC-like features. An ABC has also been described in soft tissues. (19) The so-called solid variant of ABC, (20) is nothing more than a giant cell reparative granuloma.


Recently, recurrent karyotypic abnormalities in the short arm of chromosome 17 have been noted in primary ABC. (21) Fluorescence in situ hybridization analysis of 52 primary ABCs showed that 36 (69%) had CDH11 and/or USP6 locus rearrangement. (22) Similar USP6 abnormalities have been also described in nodular fasciitis, a self-limited lesion of the soft tissue that has some resemblance histologically with the solid portion of an ABC; this finding raises the possibility that both ABC and nodular fasciitis belong to a unique group of transient neoplasia with limited growth potential. (23) No karyotypic abnormalities were identified in secondary forms. (22) Other observed abnormalities include translocations with chromosome 16, and less frequently with other partners. Rare lesions have been reported with isolated abnormalities of chromosome 16 only. (24) The finding of nonrandom chromosomal abnormalities suggests that a somatic mutation may be involved in the pathogenesis of some aneurysmal bone cysts. If these preliminary findings are confirmed in more cases, they could prove useful in establishing the diagnosis in challenging cases.

Grossly, an ABC appears as a spongy mass composed of large, blood-filled cystic spaces separated by tan-pink, gritty, fibrous septa. In solid areas, careful sampling may reveal an underlying primary tumor.

Histologically, "primary" ABC shows cellular fibrous septa containing uniform fibroblasts with scattered osteoclast-like giant cells (Figure 2, B). Giant cells seen in ABC are mainly related to vascular spaces or hemorrhagic foci in the septa. In no instance are the giant cells of ABC as numerous, large, and evenly spaced as in giant cell tumor of bone. A delicate meshwork of osteoid spicules is usually present in the fibrous septa lining aneurysmal spaces. The osteoid is characteristically delicate and deposited parallel to the vascular space surface. In rare cases, chondroid foci may be present.

Most "secondary" ABCs are associated with benign neoplasms, which can be focal, requiring careful sampling of the most solid areas. In a few cases, the underlying tumor will be malignant, with osteosarcoma being the most common.

Telangiectatic osteosarcoma shares many similarities with classic ABC, particularly on imaging studies and even on gross morphology. However, in osteosarcoma, the cells in the septa are markedly atypical, with atypical mitotic figures. Sampling is crucial and should be extensive, especially in the more "aggressive-appearing" lesions. In cases involving unusual loci, such as the epiphysis, efforts should be made to identify an underlying giant cell tumor or chondroblastoma. When reasonable sampling excludes a malignant process, missing an underlying benign tumor is of no consequence because treatment and prognosis are the same.

The wall of an ABC shows many of the same histologic features as the entity giant cell reparative granuloma or "solid" ABC. The latter is a nonneoplastic, single to multifocal, osteolytic lesion initially described in the jaw, but now recognized in almost all bones, including lower and upper extremity long tubular bones (see following section for theoretical consideration).

"Primary" ABCs and "secondary" lesions associated with benign tumors are all benign processes. They are usually treated by curettage and bone grafting with varying recurrence rates of 20% to 70% in different series and with different treatment modalities. Spontaneous regression has been reported, as well as exceedingly rare cases of spontaneous malignant transformation. (25) The single most-common cause for malignant transformation is radiation therapy.

Intraosseous Ganglion

Lesions histologically similar to the soft tissue ganglion may occur in bone, usually at the ends of tubular bones, near the articular cartilage. Affected patients are skeletally mature, young to middle-aged individuals, with men slightly more affected than women. (9) The most-common site is in the medial malleolus in some series, (9) or the hip, followed by the proximal or distal ends of tibia and fibula, shoulder, wrist, and carpal bones. (9,26) Intraosseous ganglia are most confidently recognized radiographically in bones not affected by degenerative joint disease. However, considering that intraosseous cysts of degenerative joint disease occur in the same anatomic sites with the same relative frequency and are histologically indistinguishable from the "primary" intraosseous ganglia, logically, both "primary" and "secondary" ganglia may have the same pathogenesis. The cyst of degenerative joint disease is thought to be secondary to injection of synovial fluid into the underlying bone because of degeneration of the articular cartilage that normally acts as a barrier between the joint space and the bone. Some of these defects may also be related to degenerative events at the level of tendon joint capsule insertion.


Radiologically, intraosseous ganglia appear as rounded lucencies in bone with sharp, often sclerotic, borders (Figure 3, A). They vary in size from a few millimeters to several centimeters with most between 1 and 2 cm. The lesion usually consists of a fibrous wall of variable thickness composed of loosely arranged fibroblasts embedded in a myxoid matrix containing acellular mucin (Figure 3, B).

Many intraosseous processes contain mucin (chondromyxoid fibroma and chondrosarcoma, to name a few). The subchondral cyst of degenerative joint disease enters the differential diagnosis when joint space narrowing or osteophytes are simultaneously noted.


Giant Cell Reparative Granuloma/Solid ABC

Giant cell reparative granuloma (GCRG) was initially described by Jaffe (1953) as a reactive lesion limited to maxilla and/or mandible. It was thought to be secondary to intraosseous hemorrhage and histologically indistin guishable from the "brown" tumor of hyperparathyroidism. (27) Similar lesions were later described in short tubular bones of hands and feet. (28) Sanerkin (20) (1983) reported 4 lesions that he named solid ABC in the ethmoid and vertebral bodies. Interestingly, radiologists entertained a diagnosis of "usual" ABC in 3 of them. Apparently, Sanerkin (20) attributed the "solid" adjective exclusively based on their histology. The so-called solid ABC is a bone lesion that looks like the solid portion of an ABC throughout, that is, a solid to hemorrhagic fibrous lesion with scattered multinucleated giant cells, variable osteoid production, and variable vascularity. Interestingly, almost from the start, the solid portion of usual ABC was recognized as indistinguishable from the giant cell reparative granuloma. The 2 lesions are probably related. (29) Philosophically, it is not necessary to be Aristotle to appreciate the syllogism that if the solid component of the usual ABC is indistinguishable from a giant cell reparative granuloma, and the giant cell reparative granuloma is a solid ABC, then the only difference is the prominence of blood-filled spaces in the usual ABC. All the remaining elements, including clinical data, sites of involvement, and, often, imaging studies, are identical. The striking identity can be explained in 2, not necessarily mutually exclusive ways: (1) the 2 lesions are snapshots of the same lesion at different times of its evolution (early, solid; versus late, aneurysmal), or (2) usual ABC is an accident that occurs in a small subset of giant cell reparative granuloma/solid ABCs. Whichever hypothesis is valid, logically, the classic division between "primary" and "secondary" ABC is artificial, and all ABCs are in reality secondary processes. Interestingly, variant translocations involving 16q22 and 17p13 have been noted in solid and extraosseous forms of ABC, in keeping with a common pathogenesis of all forms of ABC, solid or not. (30)

Giant cell reparative granuloma affects patients in a wide age range, with most in their third decade. (31) It involves the jaw, the short tubular bones of hands and feet, and the long tubular bones. The femur is most commonly affected, followed by the tibia, fibula, humerus, spine, pelvic bones, and clavicle.

Radiographically, GCRG is a purely radiolucent process with sharp and occasionally sclerotic margins with some level of matrix mineralization noted radiographically. (32) The lesion, in most cases, expands the affected bone segment with thinning and occasional cortical erosion. (33) When affecting long bones, the lesion may be purely radiolucent without expansion. In the short tubular bones of the hands and feet, GCRG affects the metaphysis and diaphysis (Figure 4, A). In long tubular bones, it is primarily metaphyseal. In the spine, the neural arch is more commonly affected.

The microscopic features of GCRG are identical to those of ABC with proliferating spindle cells arranged in a haphazard fashion embedded in fibrous to myxoid matrix. (29) The lesional cells vary in size but lack significant cytologic atypia, although mitotic activity may be pronounced. Variable deposition of osteoid or bone matrix may be noted in the lesion usually with definite "zonation" similar to myositis ossificans, with the bone matrix at the periphery and the hypercellular areas in the center. (32) Multinucleated, osteoclast-like giant cells are uniformly present but in patchy distribution with focal clustering and frequent hemorrhagic areas in the form of hemosiderin-laden macrophages (Figure 4, B). (29)

The differential diagnosis for GCRG includes other giant cell-rich lesions, such as giant cell tumor of bone and low-grade intramedullary osteosarcoma. Giant cell tumor of bone generally affects the epiphyseal ends of skeletally mature patients. Only 1.8% of giant cell tumors are in skeletally immature patients with open growth plates. Nearly 50% of cases occur about the knee. (33) The distal radius, sacrum, and the vertebral bodies follow. Giant cell tumor of bone extends to the articular cartilage and tends to destroy the cortex with soft tissue involvement. Radiographically, it is purely lytic and lacks sclerosis at its interface with the host bone. Histologically, the typical giant cell tumor of bone is characterized by a proliferation of uniform, mononuclear stromal cells without discernible matrix production and evenly distributed, multinucleated giant cells. Spontaneous necrosis of the tumor and fracture may modify the picture, with secondary spindle cell proliferation giving a superficial resemblance to GCRG.

Low-grade osteosarcoma is a more ominous entity to be differentiated from GCRG. Like GCRG, low-grade osteosarcoma may occur in the patient's third decade and is commonly seen about the knee. Radiographically, most low-grade osteosarcomas involve the metadiaphysis with cortical bone destruction and a permeative growth pattern. A significant percentage has a deceptively bland radiographic appearance that can simulate GCRG. Grossly, low-grade osteosarcoma appears firm and whorled. Microscopically, it is characterized by a proliferation of uniform spindle cells with minimal cytologic atypia and scant mitotic activity and is arranged in fascicles with a close resemblance to aggressive fibromatosis. (31) The lesional cells demonstrate a permeative growth pattern with significant infiltration of the host lamellar bone. Tumor osteoid, generally prominent, is deposited in trabeculae scattered throughout the tumor and lacks the reactive "zonation" seen GCRG.

Giant cell reparative granuloma is a benign lesion with limited growth potential, usually treated with simple curettage. (31) Even with incomplete excision, it rarely recurs. Recurrence should alert the pathologist to the possibility of low-grade osteosarcoma.

Fibrous Dysplasia

Fibrous dysplasia is a benign fibro-osseous process that may be monostotic or polyostotic. Fibrous dysplasia has been known as osteitis fibrosa or generalized fibrocystic disease of bone. It was recognized in its current form by Lichtenstein in 1938 (78) and Lichtenstein and Jaffe in 1942. (79) In its polyostotic form, FD may be associated with neuroendocrine abnormalities and cutaneous pigmentation, as in the McCune-Albright syndrome. A small subset of patients with FD has intramuscular myxomas, an association known as Mazabraud syndrome. (34)

Recently, it was recognized that both localized and diffuse FD are nonneoplastic processes associated with postzygotic-activating mutations of signal-transducing G proteins encoded by GNAS1 on chromosome 20. (35,36) Osteoblasts carrying this mutation show increased proliferation and inappropriate differentiation. Extent and precocity of these acquired mutations may determine the severity of the disease. Despite this variability, most (70%) lesions are limited to a single bone. Polyostotic forms make up the rest. Approximately 3% of cases have the McCune-Albright syndrome.

Fibrous dysplasia affects both sexes equally. Most of the patients present early, with 75% of patients diagnosed in their first three decades, and more severe forms diagnosed in patients' first decade. Patients with the monostotic form may be entirely asymptomatic and are usually diagnosed later.

Monostotic and polyostotic forms share the same radiographic appearance, histologic features, and anatomic sites of involvement, but the mandible is often more-commonly affected in monostotic forms, and long bone involvement prevails in polyostotic FD. Although FD may be quite variable radiologically, lesions characteristically show an intramedullary radiolucency that may be eccentric or involve the entire width of a bone. (37) The presence of intralesional, delicate spicules of woven bone gives the affected segment a "ground-glass" appearance. Endosteal scalloping of an affected bone may be present with suggestive, coarse trabeculation within the lesion (Figure 5, A). An affected bone may appear expanded, particularly when in short, tubular, or flat bones. Independent of extent, the lesions of FD appear sharply circumscribed and/or surrounded by a rim of reactive bone. Both CT and MRI images, although not generally needed for diagnosis, can show subtle, cystic change with occasional fluid-fluid levels. (38) In the rare event of malignant transformation, cortical destruction and soft tissue extension are usually present with biopsy directed to these areas.

The jaw is the most-common site affected by FD, accounting for 35% of all lesions, a figure that may be artificially high because, in all probability, most monostotic lesions outside the gnathic area are asymptomatic and unrecognized. The long tubular bones of the lower extremity are affected in approximately 30% of reported cases. An additional 20% of cases involve ribs.

Resected lesions show an expanded bone that encases a firm, gritty, white to tan lesion. Cystic degeneration may be present, with the cyst containing serous fluid. Focally, islands of cartilage may be present (fibrocartilaginous dysplasia).


Histologically, the lesion characteristically appears well circumscribed and sharply delineated by the host lamellar bone. It is composed of uniformly cellular fibrous tissue containing a proliferation of bland and uniform spindle cells with sparse mitotic activity. Scattered across the fibrous matrix are lamellae or rounded nests of woven bone without significant osteoblastic rimming (Figure 5, B). Lesions of FD do not characteristically form lamellar bone. (39) There is some morphologic variability in the woven bone spicules. The classic, most commonly seen pattern is that of curvilinear, "Chinese alphabet" spicules of woven bone separated by abundant fibrous stroma. Less commonly, the woven bone may be deposited either in sclerotic, interconnected lamellae, cementoid bodies, or in orderly and parallel spicules. These less-frequent patterns are reported to be more common in the calvarium and gnathic bones (39,40) and may result from local factors, such as different structural characteristics of affected bones. Secondary changes, like hemorrhage and fracture, if seen, make the histologic diagnosis difficult. Superimposed findings may include ABC-like changes. Additionally, xanthoma and inflammatory cell infiltration and giant cell reaction may be prominent, necessitating differentiation of FD from giant cell tumor.

Malignant transformation occurs in about 0.4% to 4% of FD cases. (34,41) In the Mayo Clinic series, (41) twice as many patients with malignant transformation had monostotic, rather than polyostotic, FD, reflecting the more-common appearance of the former. A significant number of cases with malignant transformation had prior radiation therapy (46% in the Mayo Clinic series). Malignancies included osteosarcoma, fibrosarcoma, chondrosarcoma, and malignant fibrous histiocytoma. (41) Patients with Mazabraud syndrome have a higher risk of malignant transformation, (34) with a mean survival rate of 3.4 years after diagnosis.

When low-grade osteosarcoma is considered in the differential diagnosis, correlation of the histology and imaging studies is recommended. Histologically, low-grade osteosarcoma is more cellular; cytologically, it is more atypical; and mitotically, it is more active than FD. Moreover, the regularly spaced spicules of woven bone seen in FD are not present in osteosarcoma, where malignant osteoid is often deposited in broader and irregular trabeculae. Imaging studies are diagnostic when prominent, haphazard, intralesional calcification, cortical disruption, periosteal reaction, and soft tissue masses are noted. (41)

A benign fibro-osseous process that may be indistinguishable from FD is osteofibrous dysplasia. Histologically, it is very similar to FD, but is radiographically characterized by multiple, intracortical lucencies in the anterior tibia and occasionally in the fibula, with anterior bowing of the affected segment. Conversely, when FD involves the tibia, it is characteristically a single, intramedullary lesion. (34)

When FD is limited to the jaw, ossifying fibroma enters the differential diagnosis. The latter, however, is more consistently composed of a mixture of osteoid trabeculae and cementum-like, basophilic spherules that are not usually seen in FD.

Fibrous dysplasia lesions may remain stable over time or show progression, usually during skeletal growth. However, the disease is relatively stable. Monostotic forms do not progress to multiple bone involvement. Polyostotic forms may progress in affected segments but do not disseminate in the skeleton. In many cases, lesions stabilize when bone growth ceases so that surgical treatment is mainly conservative. (42) Radiation therapy has, for the most part, been abandoned because of its association with malignant transformation. (39) Bisphosphonates, because of their action in decreasing bone remodeling, have been used in more extensive forms, (34) but their use in children has resulted in osteopetrosis-like changes. (43)

Fibrous Cortical Defect/Nonossifying Fibroma

Fibrous cortical defect (FCD) and nonossifying fibroma (NOF) describe the same histologically benign entity. Radiographically, both FCD and NOF affect the metaphysis of long bones. Classically, the term FCD is applied to smaller lesions situated solely in the cortex on imaging. Nonossifying fibroma describes a larger lesion, which, although cortically based, involves the intramedullary cavity and parallels the long axis of the bone on imaging. (44) A few patients with multiple NOFs are associated with "cafe-au-lait" spots on the skin, reminiscent of the stigmata of neurofibromatosis, type 1; an association known as Jaffe-Campanacci syndrome. (45,46) Both lesions are associated with actively growing long bones. Their tendency to spontaneously resolve is in keeping with a reactive, rather than neoplastic, process. To our knowledge, chromosomal abnormalities have been described in 2 cases. (47,48)


Fibrous cortical defect is usually an incidental finding, affecting 30% to 40% of the pediatric population, (49) with a relatively high rate of bilaterality and multiplicity. Most FCDs are located in long bones. Fibrous cortical defect starts in the metaphysis near the epiphyseal plate but is usually "left behind" by the growing bone, giving the impression of migration, as in a UBC. Both FCD and NOF regress spontaneously within a few years of discovery.

Fibrous cortical defect has a characteristic radiographic appearance of a small, rounded or elliptical lesion (average 1-3 cm), eccentrically located in the metadiaphysis of the cortex of a long bone without thickening it. Those that persist longer are called NOFs and are usually diagnosed in the patient's first 2 decades (80% of cases). (45) Nonossifying fibromas involve the medullary cavity and are radiolucent, with their longest diameter cortically based, paralleling the long axis of the bone (Figure 6, A). They also originate near the metaphyseal side of the growth plate and appear to migrate downward into the metadiaphysis of long bones. In larger bones, like the femur and tibia, NOFs do not usually extend across the entire diameter of the affected shaft, but that is not the case for small-diameter bones, which may appear expanded. (45) The larger lesions may result in pathologic fractures.

Nonossifying fibroma accounts for up to 2% of biopsied bone lesions, but in all probability, most cases go undiagnosed. If untreated, a NOF may persist into early adulthood, and exhibit regressive changes. In cases with sequential radiographs, a NOF shows progressive ossification from the periphery inward. In many instances, a faint outline of the lesion may persist indefinitely.

Gross and histologic aspects of FCDs and NOFs are identical. They appear soft and tan and are composed of uniform, plump fibroblasts arranged in a whorled to storiform pattern. The nuclei are ovoid with fine chromatin and small nucleoli. The cytoplasm is eosinophilic with indistinct cell boundaries. Mitoses may be observed but are generally sparse and typical. Scattered lymphocytes and hemorrhagic foci with hemosiderin pigment are usually present, as well as multinucleated osteoclast-like giant cells, evenly distributed in the lesion (Figure 6, B). Although, in some cases, the giant cells may be numerous, their density is unlike that observed in giant cell tumor of bone. Moreover, giant cell tumor of bone affects the epiphyseal ends of the skeletally mature patient, as opposed to NOF, which never affects the epiphysis. Regressing lesions tend to have more foamy histiocytes with fibrosis of the matrix and peripheral reactive ossification. It is likely that some of the lesions diagnosed as fibrous histiocytoma and GCRG are really regressing NOF. Additional changes may be observed in cases undergoing fracture, including focal osteoid production, increased mitotic activity, and focal necrosis or complete infarction. Whatever the histologic findings, imaging studies will show the intact contour or borders of a NOF.

Fibrous cortical defects and NOFs require no treatment, although larger lesions may be symptomatic because of subtle fractures and may require curettage and packing with bone chips.


Bone Island (Enostosis)

The terms bone island or enostosis may be used interchangeably to describe the same lesion, namely, a localized area of compact bone within the medullary cavity of a bone. The area may be encompassed by the medullary cavity or show focal attachment to the inner cortex of the host bone. Bone islands are characteristically asymptomatic and usually discovered incidentally. They develop after puberty and may undergo prolonged growth. They are most common in the pelvis, proximal femora, and ribs. In tubular bones, they are most common in the epiphysis. (50) Not all bone islands are stable. Some, after a period of growth, may decrease in size and disappear. (51) Others may increase in size and concentrate radionuclide on bone scans.


Radiographically, bone islands are rounded to ovoid, may have speculated contours, and are generally solitary. Rarely, multiple lesions are noted, a condition known as osteopoikilosis. The typical bone island lesion varies from a few millimeters to a centimeter at its largest diameter. Rare cases may exceed 2 cm (giant bone island). Bone islands are usually situated at the ends of bones surrounding joints (Figure 7). They appear circumscribed, densely sclerotic, and may be bordered by spicules that blend with the neighboring trabeculae of the host bone. Occasionally, the largest lesions may concentrate radionuclide, probably reflecting increased metabolic activity. (52) In these larger lesions, the differential diagnosis may include benign lesions, such as osteopathia striata and osteoblastoma, and malignant lesions, such as sclerosing osteosarcoma.

Osteoblastomas, however, are largely metaphyseal, may expand the bone, and have a periosteal reaction, features not seen in bone islands. An osteoblastoma does not blend with host trabecular bone and may show reactive cortical destruction with cortical thickening.

Osteosarcomas, as opposed to bone island, show aggressive growth, with cortical destruction, interrupted periosteal reaction, a soft tissue mass, and permeative growth within the bone marrow. More important, primary bone tumors are usually symptomatic.

In the rare case of a biopsied bone island, the histology has shown compact lamellar bone with prominent Haversian canals. (50) Thickened trabeculae of lamellar bone, when radiating from the periphery of the lesion, usually blend with the bone trabeculae of the host bone. Treatment is not necessary.


Many types of nonneoplastic exophytic lesions of bone exist and include marginal osteophytes of degenerative joint disease; reactive lesions, such as subungual exostosis and bizarre parosteal osseous proliferation; and osteochondroma. The following discussion will be limited to osteochondromas.


The term osteochondroma refers to a bony projection, sessile or pedunculated. Osteochondromas originate from a cartilaginous stage and arise from the metaphyseal surface of a bone. They usually have a cartilage cap of uniform thickness and contain bone marrow at their base, which maintains direct continuity with the underlying bone marrow. The continuity between the lesional and host bone marrow helps to distinguish them from other exophytic lesions, including benign ones, such as bizarre parosteal osteochondromatous proliferation, and malignant ones, such as parosteal osteosarcoma. Both bizarre parosteal osteochondromatous proliferation and parosteal osteosarcoma lesions lie on the bone cortex surface without any continuity with the underlying marrow space. Dysplasia epiphysealis hemimelica also enters into the differential diagnosis for osteochondromas of the lower extremity; however, even if microscopically the 2 entities are indistinguishable, dysplasia epiphysealis hemimelica affects the epiphysis, whereas osteochondromas are never centered in the epiphysis but involve the metaphysis or metadiaphysis.

Osteochondromas are the most-common bone lesions (30%-50% of resected, benign lesions). Most (85%) are solitary and sporadic. Their true frequency is probably underestimated because most are asymptomatic. The 15% of osteochondromas with multiple lesions is indicative of a condition transmitted with a dominant pattern of inheritance known as hereditary osteochondromatosis. (53) When osteochondromas come to medical attention, the sporadic lesions are usually diagnosed by age 30. Patients with multiple exostoses, on the other hand, are diagnosed at a younger age. Both solitary and multiple forms have no sex predilection.


Osteochondromas may occur in any bone formed by enchondral ossification. The most common sites, in decreasing order of frequency, are the metaphyses of the distal femur, the upper humerus, the upper tibia, and the fibula. Flat bones are less commonly affected, with the ilium and scapula most often involved.

Osteochondromas are noticed because of their size or through such complications as bursitis, fracture of the stalk, or impingement on nerves or vessels resulting in ischemia or pseudoaneurysm formation.

An osteochondroma appears radiographically sessile or pedunculated. It is an exophytic lesion whose base is continuous with the neighboring bone marrow (Figure 8, A). Associated mineralization may appear irregular, but focal calcification is best appreciated on CT scans rather than MRIs. An MRI may be helpful in evaluating cartilaginous cap thickness as well as an associated bursa. (54,55)

Grossly, osteochondromas are sessile or pedunculated bony projections covered by a smooth, rounded to bosselated, cartilaginous cap of uniform thickness, usually a few millimeters thick (Figure 8, B and C). In some cases, caps may be thicker than 1 cm and up to 2 cm, particularly in actively growing patients. Microscopically, the cartilage cap is invested by perichondrium and composed of hyaline-type cartilage with clustered chondrocytes embedded in an abundant matrix. Its deepest portion is identical to growth plate cartilage and is arranged in columns that undergo enchondral ossification (Figure 8, D, E). Lesions in older individuals may have a very thin and discontinuous layer of cartilage or no cartilage at all. Rarely, caps in older patients can grow.

The etiology of osteochondroma is still being debated. The classic hypothesis considers them to be epiphyseal plate anomalies, occurring in the earliest phase of bone development, which results in the disruption, herniation, and rotation of a growth plate fragment. The displaced fragment of epiphyseal cartilage then grows at an angle to the bone and progressively grows away from the open physis during bone lengthening. Osteochondromas usually share the destiny of the epiphyseal plate, with loss of their cartilaginous cap at the same time as the epiphyseal plate closes. Characteristically, pedunculated lesions point away from the joint of origin because of the mechanical forces of adjacent tendon and muscles.

Cytogenetic aberrations have been identified in both sporadic and multiple osteochondromas, involving EXT genes, which are held to be tumor suppressors involved in normal chondrocyte proliferation and differentiation. Preliminary studies suggest that these chromosomal changes may predispose patients to location, extent, and type of osteochondroma and may be associated with the risk of malignant transformation. Additionally, mutations in EXT1 are associated with more lesions, higher risk of malignant transformation, and a higher rate of associated bone deformities, compared with mutations in EXT2.If these findings are confirmed, solitary and multiple osteochondromas may be considered true neoplasms. (53,56)

Surgical treatment, when required, consists of resection. Complicated cases, such as those associated with aneurysms, may require the assistance of a vascular surgeon.

Chondrosarcoma may arise in osteochondroma, more commonly in patients with multiple osteochondromas. (57) In a large published surgical series, (57) the rate of malignant transformation of solitary osteochondroma varied from 1% to 5%, but rose to 5% to 35% in patients with multiple osteochondromas. Selection factors are probably responsible for these results. The actual rate of transformation is lower because most osteochondromas are unrecognized or not treated.

Malignant transformation is frequently associated with clinical symptoms, such as pain of variable duration, neurologic symptoms, and growth of a previously stable lesion. The mean age of patients with malignant transformation of an osteochondroma is 35 years (1 to 2 decades younger than patients with primary chondrosarcoma). The average size of an osteochondroma exceeds 9 cm, with the thickness of the cartilaginous cap more than 3 to 4 cm.

Radiographically, sarcomatous transformation is suspected if, on serial images, the mass appears larger. Any increase in size of an osteochondroma after growth plate closure should be considered suspicious. Surface irregularities of osteochondromas with inhomogeneous mineralization are also ominous signs. Both CTs and particularly MRIs aid in determining the extent of soft tissue and underlying bone invasion, especially in deep locations, such as the pelvis and the spine. (55)

The most common sites for malignant transformation are the pelvic bones and proximal and distal femur for solitary lesions. In patients with multiple osteochondromas, the pelvic bones, scapula, and thoracic spine are the most common sites, particularly if they have been directly or indirectly included in radiation ports, especially during active skeletal growth. Although cartilaginous cap thickness is almost always increased, that element alone is insufficient for diagnosing malignancy. Malignancy is usually associated with soft tissue infiltration, most often by nodules separated from the main tumor mass, by bone stalk permeation, by host lamellar bone entrapment, by markedly increased cellularity with atypia, and by mitotic activity, myxoid stromal changes, and tumor necrosis. Treatment is largely surgical with wide resection of the lesion, including the host bone and an intact cuff of soft tissue. The prognosis of chondrosarcoma ex-osteochondroma is generally excellent, provided complete resection is possible. Most tumor deaths are due to uncontrollable local growth. Lung metastasis may occur. (57)


Chest Wall Hamartoma (Mesenchymal Hamartoma of the Chest Wall)

Chest wall hamartoma is a nonneoplastic, congenital lesion that affects the chest wall of infants. It was originally described as a mesenchymoma, but because it is nonneoplastic, the term hamartoma is more appropriate. (58) It is rare and, in approximately 40% of the cases, is present at birth or diagnosed in utero. (59) Most hamartomas are diagnosed in the infant's first year. Although generally intrathoracic, a hamartoma may protrude into the chest wall soft tissue. Hamartomas originate from the posterior or lateral aspect of a rib, and multiple ribs may be involved. (60) The most common symptom of a hamartoma is respiratory distress. More-severe cases require surgical resection. Asymptomatic cases may be managed conservatively in view of their tendency to regress spontaneously. Reports indicate that the size of chest wall hamartomas peak at birth and start to involute thereafter. (61)

On radiographs, a chest wall hamartoma appears as a mineralized mass that, if expansive, may partially destroy neighboring ribs. Both CT and MRI scans usually define solid and cystic areas with multiple fluid-fluid levels. (58,60)

Gross hamartoma specimens usually appear as circumscribed lesions composed of tan-white soft tissue, surrounding large, cystic, blood-filled spaces (Figure 9, A).


Microscopically, the solid areas consist of fibrous tissue containing mature hyaline cartilage (Figure 9, B) with variable enchondral ossification and areas showing features of ABC. (58,62) Microscopically, the lesion may resemble several neoplastic lesions, such as chondroblastoma and other cartilaginous neoplasms or ABC, but the age of the patient and location of the lesion make the diagnosis obvious. Symptomatic lesions may require resection. (57) Others may be followed and tend to regress. (61)


Langerhans Cell Histiocytosis

Langerhans cell histiocytosis (LCH) is a proliferation of histiocytes phenotypically identical to the dendritic Langerhans cells. This process encompasses several disorders of variable extent and severity, historically known as histiocytosis X, Letterer-Siwe disease, Hand-Schuller-Christian syndrome, eosinophilic granuloma of bone, and self-healing reticulohistiocytosis. (63) In the last few years, the debate whether LCH is a reactive or neoplastic process has been re-energized by the discovery that LCH is a clonal process. (65) The initial attempts to identify recurrent cytogenetic abnormalities, however, yielded mixed results, (66,83) but the recent discovery of consistent activating somatic mutations in all forms of LCH, including primary lung lesions, seems to support the neoplastic nature of Langerhans cell histiocytosis. (81) If these findings are confirmed, LCH probably represents a neoplasm of immature, marrow-derived myeloid dendritic cells. (82)

The severity and extent of LCH are age related. Older children and adults may be affected by self-limited, unifocal disease. Younger children may have multifocal disease, whereas widespread and multisystem forms are limited to infants. Factors adversely affecting the prognosis are young age at presentation (3 years or younger), hepatosplenomegaly, thrombocytopenia, and polyostotic bone disease. (64)

Langerhans cell histiocytosis is uncommon, and occurs in 5 people per million per year, most often in males. Whites of northern European descent are most commonly affected. Langerhans cell histiocytosis involving bone accounts for less than 1% of biopsied bone tumors. It has a broad age range, from early infancy to old age, with most cases diagnosed in a patient's first 3 decades. Monostotic lesions are more common than polyostotic lesions.

Although any bone can be affected, the flat bones are more common, and the skull is the most commonly affected bone (27% of lesions), (64) followed by the axial skeleton, including pelvis, spine, mandible, and ribs. Ribs are most commonly involved in older patients.

Langerhans cell histiocytosis may cause pain and swelling of the affected bones. Additionally, LCH may be associated with diabetes insipidus, compression fractures of vertebral bodies, and loosening and loss of teeth. Langerhans cell histiocytosis may also be an incidental finding, especially in isolated bone lesions.

Radiographs are often diagnostic, if well-established lesions are present, particularly in ribs when coexistent, classic, pulmonary lesions and spontaneous pneumothorax are noted on the same film. The most common radiographic finding is that of an intramedullary-based lytic lesion with a lobular, often beveled, but well-defined, peripheral contour (Figure 10, A and B). (67) However, less-typical LCH radiographic features include epiphyseal or transphyseal location and fluid-fluid levels. (68) Early lesions may appear radiographically aggressive and be mistaken for malignancy or infection, with radiographic appearance dependent on the bone involved. A prominent, periosteal reaction can simulate malignancy or infection, particularly in long tubular bones, but only rarely in flat bones. Langerhans cell histiocytosis in flat bones, however, may be associated with expansion or a soft tissue mass or both. A soft tissue mass, if present, can be easily palpated in the calvarium. In the spine, the thoracic segment is commonly involved with the vertebral body being the favored site, and the posterior arch is uncommonly affected. One or more vertebral bodies may undergo partial collapse with "anterior wedging" or near-total collapse, with a characteristic, but nonspecific, "vertebra plana." Mandibular involvement may include alveolar bone destruction, sometimes simulating "floating teeth." Bone scans, including the entire skeleton, often helps to identify additional lesions. A CT scan aids in documenting the extent of cortical destruction and/or the presence of a soft tissue mass. An MRI more reliably documents soft tissue masses and the extent of marrow involvement. Both modalities are nonspecific.


The histology of LCH includes characteristic large histiocytic cells with abundant, pale, eosinophilic cytoplasm and large, oval to indented nuclei embedded in a brisk inflammatory infiltrate containing eosinophils. The Langerhans cells are arranged in loose meshworks, clusters, and individual cells, with variable numbers of eosinophils usually representing the predominant inflammatory cell in all osseous cases (Figure 10, C, D). Lymphocytes and neutrophils are also noted. Foamy macrophages are less common. (69) Mitoses and patchy necrosis are often seen but are not associated with a more-aggressive clinical course. Langerhans cells stain for S100 protein and CD1a but are generally negative for CD45. Although not routinely used for diagnosis, electron microscopy may reveal diagnostic intracytoplasmic Birbeck granules. (69)

The histologic differential diagnosis includes lymphoma, osteomyelitis, extranodal Rosai-Dorfman disease, and Erdheim-Chester disease. (67) The morphology, immunophenotype, and associated eosinophilic infiltrate are helpful for arriving at the correct diagnosis. In selected cases, when the neutrophilic and lymphocytic infiltrate is predominant, differentiation from osteomyelitis may be difficult, with correlation between immunohistochemistry and clinicoradiographic findings required.

Treatment of LCH is not standardized but is dependent on the extent of the disease and the site of involvement. Solitary lesions may be treated with curettage, injections of methylprednisolone, or small doses of radiation to selected sites. Systemic chemotherapy has been used in some cases with multifocal bone involvement to reduce the risk of recurrence. (69)

Extranodal Rosai-Dorfman Disease of Bone

Rosai-Dorfman disease, also known as sinis histiocytosis with massive lymphadenopathy, (70) is an indolent histiocytic disorder of unknown etiology. The detection of polyclonal X-inactivation patterns, in some cases, is against a neoplasm. (71) Although originally described in lymph nodes, it can involve multiple extranodal sites, including bone, in approximately 30% of cases. Extranodal disease may occur alone or in association with nodal disease. Bone involvement is rare, occurring in approximately 10% of cases. Isolated bone involvement is exceptionally rare. (72)

Rosai-Dorfman disease may affect the entire skeleton but is most commonly seen in the long tubular bones, including the tibia, femur, and humerus. Flat bones, such as the ribs, skull, and sacrum, and the short tubular bones of the hand and maxilla may be involved.

Imaging studies of Rosai-Dorfman disease vary, showing purely lytic lesions with or without sclerotic borders (Figure 11, A). A soft tissue mass may be present along with a periosteal reaction. Some lesions may appear expanded or multiloculated.

Grossly bony lesions in Rosai-Dorfman disease may be soft to gritty and vary from white to tan. The histology in Rosai-Dorfman disease consists of a cellular process replacing the marrow. The cellular process is composed of a brisk, inflammatory infiltrate containing large histiocytes that characteristically demonstrate emperipolesis, the phagocytosis of neutrophils, lymphocytes, and plasma cells (Figure 11, B). The inflammatory infiltrate consists mainly of lymphocytes and plasma cells, with scattered neutrophils and eosinophils. The large histiocytes have abundant eosinophilic cytoplasm and nuclei that can vary in size and shape, with vesicular chromatin and conspicuous nucleoli. The lesions are [S100.sup.+] with variable positivity for CD68. CD1a is uniformly negative.


The differential diagnosis includes other histiocytic processes like LCH (see above) and Erdheim-Chester disease.

Erdheim-Chester disease is a rare, multisystem, non-Langerhans form of histiocytosis. It preferably affects adults in their 50s and 60s, with a slight male preponderance. Erdheim-Chester disease is characterized by the accumulation of large, lipid-laden, foamy histiocytes within the marrow spaces of long bones. (73) Involvement of long bones differs from Rosai-Dorfman disease. Erdheim-Chester disease is generally bilateral and symmetric and may be associated with interstitial lung disease, xanthogranulomas in the skin, exophthalmos, and diabetes insipidus. Bone lesions in Erdheim-Chester disease are mainly localized to the metaphysis and diaphysis of the long bones, which show areas of sclerosis alternating with lucencies. Bone scans show striking symmetrical tracer accumulation in the affected bone segments. In Erdheim-Chester disease, the bone marrow is replaced by a proliferation of large, foamy histiocytes embedded in dense fibrous tissue with sparse inflammatory cells and multinucleated, osteoclast-like giant cells. The lesional cells infiltrate the host trabecular bone, which appears thickened. The characteristic emperipolesis and immunophenotype of Rosai-Dorfman disease are not seen in Erdheim-Chester disease.

Rosai-Dorfman disease has no specific treatment, but it has an excellent prognosis. Some lesions resolve spontaneously. Curettage or resection of individual bone lesions is effective for local control. Chemotherapy and radiation in selected cases have had limited response. (72)


Osteomyelitis (OM) is defined as an infectious process that involves bone. If not promptly diagnosed and treated, it may destroy the affected bone and progress to a chronic phase that can persist despite medical treatment. Osteomyelitis can affect healthy individuals of all ages and those with numerous predisposing conditions, such as diabetes, hemodialysis, sickle cell disease, immunodeficiency states, and intravenous drug use. (74)

Causative organisms are usually bacteria that reach the bone via the bloodstream (hematogenous OM) or through a direct route, such as with a penetrating injury or operative procedure (secondary OM). When infection occurs, it may present with different clinical and radiographic patterns that classically are divided into acute, subacute, and chronic OM. The following discussion is focused on hematogenous OM, which is the form most commonly seen during biopsy because of its often ambiguous clinical and radiographic presentation.

Acute hematogenous OM usually affects the metaphysis of the long bones of the young. Most patients with acute hematogenous OM are in their teens. Staphylococcis aureus is the most commonly isolated organism. Haemophilus influenzae, however, is the most common organism in children younger than 3 years. Neonates may also be affected by group B streptococci and coliform bacteria. (74)

Clinically, OM may present with local pain and fever. Pain may be severe with loss of function of an adjacent joint that, in some instances, can be due to an associated septic arthritis. Laboratory values may show prominent leukocytosis and elevated erythrocyte sedimentation rate but may also be nonspecific.

When OM is suspected, imaging studies are crucial for confirming diagnosis. However, routine radiographs may show no evidence of bone involvement up to 2 weeks after onset of clinical symptoms. A loss of at least 30% to 50% of the bone matrix is required for detection of OM on radiographs (74) (Figure 12, A). In very early phases of infection, bone and soft tissue changes are often not apparent on radiographs. In the face of negative radiographs, imaging modalities should include MRI and/or bone scans, including blood pool studies. An MRI scan has been reported with positive findings within 24 to 48 hours of onset of symptoms; bone scintigraphy may show positive results within 48 to 72 hours. (75,76) The choice of modality may depend on the patient's ability to cooperate and the need for sedation. Scintigraphy is usually chosen if a whole-body skeletal survey is needed. However, recently, depending on equipment, an MRI was able to perform whole-body scanning, with the advantage of better documenting the extent of marrow and soft tissue involvement. If the extent of marrow and soft tissue involvement is known and the lesion is focal, a CT may serve as a more efficient guide for percutaneous or open biopsies because it is more readily available and doesn't require MR-compatible biopsy hardware.


Pathologically, OM is characterized by an inflammatory marrow infiltrate accompanied by fibrosis, neovascularity, and, in time, bone destruction (Figure 12, B, C). In its acute form, neutrophils predominate, accompanied by bone resorption due to numerous osteoclasts lining the host lamellar bone. In subacute and chronic forms of OM, bone necrosis, chronic inflammatory cells, fibrosis, and reactive bone formation dominate the picture. Generally, it is not always possible to confidently correlate histology with the stage of OM, so that a simple diagnosis of OM may have to be rendered. Relying solely on the histologic picture is not advisable before correlation with clinical and radiographic findings because an inflammatory infiltrate, accompanied by evidence of bone necrosis and remodeling, can be seen in many other conditions, including fracture, bone infarct, degenerative joint disease, and at the periphery of primary and metastatic neoplasms.

The differential diagnosis of OM in the presence of bone destruction and associated periosteal and soft tissue changes does include malignant processes. Therefore, although the young are mainly affected by OM, if a neoplasm is considered, lymphoma, Ewing sarcoma, and osteosarcoma should be included in the differential diagnosis. Nevertheless, it is important to remember that the opposite is also true. If a malignant bone process is suspected on clinical and radiographic grounds, it is good practice to consider OM in the differential.

Treatment modalities for hematogenous OM vary according to the stage of the disease. Acute forms, before extensive bone necrosis has developed, may be treated with antibiotics only. Subacute and chronic forms may also require surgical debridement of associated necrotic bone because microorganisms may persist in devitalized bone to cause flare-ups many decades after an initial attack. (74) Imaging studies play an important role. Radiographs in chronic OM document deformed bone, often with thick, undulating cortices, and bony sequestra entrapped by lucencies indicative of fibrous tissue and or pus. A CT scan improves documentation of subacute OM. However, an MRI scan is routinely more successful in defining draining sinus tracts and soft tissue extensions or neoplastic masses, which are rare sequelae of chronic osteomyelitis.


(1.) Wilkins RM. Unicameral bone cysts. J Am Acad Orthop Surg. 2000; 8(4): 217-224.

(2.) Cohen J. Etiology of simple bone cyst. J Bone Joint Surg Am. 1970; 52(7): 1493-1497.

(3.) Mirra JM, Bernard GW, Bullough PG, Johnston W, Mink G. Cementum-like bone production in solitary bone cysts (so-called "cementoma" of long bones): report of three cases--electron microscopic observations supporting a synovial origin to the simple bone cyst. Clin Orthop Relat Res. 1978; (135):295-307.

(4.) Jaffe HL, Lichtenstein L. Solitary unicameral bone cyst with emphasis on the roentgen picture, the pathologic appearance, and the pathogenesis. Arch Surg. 1942; 44(6):1004-1025.

(5.) Taxin RN, Feldman F. The tumbling bullet sign in a post-traumatic bone cyst. Am J Roentgenol Radium Ther Nucl Med. 1975; 123(1):140-143.

(6.) Lodwick GS. Juvenile unicameral bone cyst; a roentgen reappraisal. Am J Roentgenol Ther Nucl Med. 1958; 80(3):495-504.

(7.) Reynolds J. The "fallen fragment sign" in the diagnosis of unicameral bone cysts. Radiology. 1969; 92(5):949-953.

(8.) Sim FH, Dahlin DC. Ganglion cysts of bone. Mayo Clin Proc. 1971; 46(7): 484-488.

(9.) Feldman F. Johnston A. Intraosseous ganglion. AJRAm J Roentgenol. 1973; 118(2):328-343.

(10.) Lokiec F, Wientroub S. Simple bone cyst: etiology, classification, pathology, and treatment modalities. J Pediatr Orthop B. 1998; 7(4):262-273.

(11.) Lichtenstein L. Aneurysmal bone cyst: a pathological entity commonly mistaken for giant cell tumor and occasionally for hemangioma and osteogenic sarcoma. Cancer. 1950; 3(2):279-289.

(12.) Kransdorf MJ, Sweet DE. Aneurysmal bone cyst: concept, controversy, clinical presentation, and imaging. AJR Am J Roentgenol. 1995; 164(3):573-580.

(13.) Jaffe HL. Aneurysmal bone cyst. Bull Hosp Joint Dis. 1950; 11(1):3-13.

(14.) Ruiter DJ, van Rijssel TG, van der Velde EA. Aneurysmal bone cysts: a clinicopathologic study of 105 cases. Cancer. 1977; 39(5):2231-2239.

(15.) Martinez V, Sissons HA. Aneurysmal bone cyst: a review of 123 cases including primary lesions and those secondary to other bone pathology. Cancer. 1988; 61(11):2291-2304.

(16.) Vergel De Dios AM, Bond JR, Shives TC, McLeod RA, Unni KK. Aneurysmal bone cyst: a clinicopathologic study of 238 cases. Cancer. 1992; 69(12):2921-2931.

(17.) de Kleuver M, van der Heul RO, Veraart BE. Aneurysmal bone cyst of the spine: 31 cases and the importance of the surgical approach. J Pediatr Orthop B. 1998; 7(4):286-292.

(18.) Bonakdarpour A, Levy WM, Aegerter E. Primary and secondary aneurysmal bone cyst: a radiological study of 75 cases. Radiology. 1978; 126(1):75-83.

(19.) Rodriguez-Peralto JL, Lopez-Barea F, Sanchez-Herrera S, Atienza M. Primary aneurismal bone cyst of soft tissues (extraosseous aneurismal cyst). Am J Surg Pathol. 1994; 18(6):632-636.

(20.) Sanerkin NG, Mott MG, Roylance J. An unusual intraosseous lesion with fibroblastic, osteoclastic, osteoblastic, aneurysmal and fibromyxoid elements: "solid" variant of aneurismal bone cyst. Cancer. 1983; 51(12):2278-2286.

(21.) Althof PA, Ohmori K, Zhou M, et al. Cytogenetic and molecular cytogenetic findings in 43 aneurysmal bone cysts: aberrations of 17p mapped to 17p13.2 by fluorescence in situ hybridization. Mod Pathol. 2004; 17(5):518 525.

(22.) Oliveira AM, Perez-Atayde AR, Inwards CY, et al. USP6 and CDH11 oncogenes identify the neoplastic cell in primary aneurysmal bone cysts and are absent in so-called secondary aneurysmal bone cysts. Am J Pathol. 2004, 165(5): 1773-1780.

(23.) Erickson-Johnson MR, Chou MM, Evers BR, et al. Nodular fasciitis: a novel model of transient neoplasia induced by MYH9-USP6 gene fusion. Lab Invest. 2011; 91(10):1427-1433.

(24.) Panoutsakopoulos G, Pandis N, Kyriazoglou I, Gustafson P, Mertens F, Mandahl N. Recurrent t(16; 17)(q22; p13) in aneurysmal bone cysts. Genes Chromosomes Cancer. 1999; 26(3):265-266.

(25.) Kyriakos M, Hardy D. Malignant transformations of aneurismal bone cyst, with an analysis of the literature. Cancer. 1991; 68(8):1770-1780.

(26.) Schajowicz F, Clavel Sainz M, Slullitel JA. Juxta-articular bone cysts (intraosseous ganglia): a clinicopathological study of 88 cases. J Bone Joint Surg Br. 1979; 61(1):107-116.

(27.) Jaffe HL. Giant-cell reparative granuloma, traumatic bone cyst, and fibrous (fibro-osseous) dysplasia of the jawbones. Oral Surg Oral Med Oral Pathol. 1953; 6(1):159-175.

(28.) Lorenzo JC, Dorfman DD. Giant cell reparative granuloma of short tubular bones of the hands and feet. Am J Surg Pathol. 1980; 4(6):551-563.

(29.) Wold LE, Dobyns JH, Swee RG, Dahlin DC. Giant cell reaction (giant cell reparative granuloma) of the small bones of the hands and feet. Am J Surg Pathol. 1986; 10(7):491-496.

(30.) Dal Cin P, Kozakewich HP, Goumnerova L, Mankin HJ, Rosenberg AE, Fletcher JA. Variant translocations involving 16q22 and 17p13 in solid variant and extraosseous forms of aneurysmal bone cyst. Genes Chromosomes Cancer. 2000; 28(2):233-234.

(31.) Ilaslan H, Sundaram M, Unni KK. Solid variantofaneurysmalbonecystsin long tubular bones: giant cell reparative granuloma. AJR Am J Roentgenol. 2003; 180(6):1681-1687.

(32.) Bertoni F, Bacchini P, Capanna R, et al. Solid variant of aneurysmal bone cyst. Cancer. 1993; 71(3):729-734.

(33.) Oda Y, Tsuneyoshi M, Shinohara N. "Solid" variant of aneurismal bone cysts (extragnathic giant cell reparative granuloma) in the axial skeleton and long bones: a study of its morphologic spectrum and distinction from allied giant cell lesions. Cancer. 1992; 70(11):2642-2649.

(34.) DiCaprio MR, Enneking WF. Fibrous dysplasia: pathophysiology, evaluation, and treatment. J Bone Joint SurgAm. 2005; 87(8):1848-1864.

(35.) Weinstein LS, Shenker A, Gejman PV, Merino MJ, Friedman E, Spiegel AM. Activating mutations of the stimulatory G protein in the McCune-Albright syndrome. N Engl J Med. 1991; 325(24):1688-1695.

(36.) Cohen MM Jr. Fibrous dysplasia is a neoplasm. Am J Med Genet. 2001; 98(4):290-293.

(37.) Kransdorf MJ, Moser, RP Jr, Gilkey FW. Fibrous dysplasia. Radiographics. 1990; 10(3):519-537.

(38.) Shah ZK, Peh WCG, Koh WL, Shek TW. Magnetic resonance imaging appearances of fibrous dysplasia. BrJRadiol. 2005;78(936):1104-1115.

(39.) Dorfman HD. New knowledge of fibro-osseous lesions of bone. Int J Surg Pathol. 2010; 18(3)(suppl):62S-65S.

(40.) Riminucci M, Corsi A, Shenker A, et al. The histopathology of fibrous dysplasia of bone in patients with activating mutations of the Gsa gene: site-specific patterns and recurrent histological hallmarks. J Pathol. 1999; 187(2):249-258.

(41.) Ruggieri P, Sim FH, Bond JR, Unni KK. Malignancies in fibrous dysplasia. Cancer. 1994; 73(5):1411-1424.

(42.) Ippolito E, Bray EW, Corsi A, et al. Natural history and treatment of fibrous dysplasia of bone: a multicenter clinicopathologic study promoted by the European Pediatric Orthopedic Society. J Pediatr Orthop B. 2003; 12(3):155-177.

(43.) Whyte MP, Wenkert D, Clements KL, McAlister WH, Mumm S. Bisphosphonate-induced osteopetrosis. N Engl J Med. 2003; 349(5):457-463.

(44.) Mankin HJ, Trahan CA, Fondren G, Mankin CJ. Non-ossifying fibroma, fibrous cortical defect and Jaffe-Campanacci syndrome: a biologic and clinical review. Chir Organi Mov. 2009; 93(1):1-7.

(45.) Jaffe HL, Lichtenstein L. Non-osteogenic fibroma of bone. Am J Pathol. 1942; 18(2):205-221.

(46.) Campanacci M, Laus M, Boriani S. Multiple non-ossifying fibromata with extraskeletal anomalies: a new syndrome? J Bone Joint Surg Br. 1983; 65(5):627-632.

(47.) Tarkkanen M, Kaipainen A, Karaharju E, et al. Cytogenetic study of 249 consecutive patients examined for a bone tumor. Cancer Genet Cytogenet. 1993; 68(1):1-21.

(48.) Nelson M, Perry D, Ginsburg G, Sanger WG, Neff JR, Bridge JA. Translocation (1; 4)(p31;q34) in nonossifying fibroma. Cancer Genet Cytogenet. 2003; 142(2):142-144.

(49.) Arata MA, Peterson HA, Dahlin DC. Pathological fractures through non ossifying fibromas. J Bone Joint SurgAm. 1981; 63(6):980-988.

(50.) Gold RH, Mirra JM, Remotti F, Pignatti G. Case report 527: Giant bone island of tibia. Skeletal Radiol. 1989; 18(2):129-132.

(51.) Onitsuka O. Roentgenologic aspects of bone islands. Radiology. 1977; 123(3):607-612.

(52.) Greenspan A, Steiner G, Knutzon R. Bone island (enostosis): clinical significance and radiologic and pathologic correlations. Skeletal Radiol. 1991; 20(2):85-90.

(53.) Khurana J, Abdul-Karim F, Bovtee JVMG. Osteochondroma. In: Fletcher CDM, Unni KK, Mertens F, eds. Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press; 2002:234-236. World Health Organization Classification of Tumours; vol 4.

(54.) Lee JK, Yao L, Wirth CR. MR imaging of solitary osteochondromas: report of eight cases. AJR Am J Roentgenol. 1987; 149(3):557-560.

(55.) De Beuckeleer LHL, De Schepper AMA, Ramon F. Magnetic resonance imaging of cartilaginous tumors: is it useful or necessary? Skeletal Radiol. 1996; 25(2):137-141.

(56.) Hameetman L, Szuhai K, Yavas A, et al. The role of EXT1 in nonhereditary osteochondroma: identification of homozygous deletions. J Natl Cancer Inst. 2007; 99(5):396-406.

(57.) Ahmed AR, Tan T-S, Unni KK, Collins MS, Wenger DE, Sim FH. Secondary chondrosarcomas in osteochondroma: report of 107 patients. Clin Orthop Relat Res. 2003;(411):193-206.

(58.) McLeod RA, Dahlin DC. Hamartoma (mesenchymoma) of the chest wall in infancy. Radiology. 1979; 131(3):657-661.

(59.) Chu L, Seed M, Howse E, Ryan G, Grosse-Wortmann L. Mesenchymal hamartoma: prenatal diagnosis by MRI. Pediatr Radiol. 2011; 41(6):781-784.

(60.) Groom KR, Murphey MD, Howard LM. Mesenchymal hamartoma of the chest wall: radiologic manifestations with emphasis on cross-sectional imaging and histopathologic comparison. Radiology. 2002; 222(1):205-211.

(61.) Braatz B, Evans R, Kelman A, Cheng W. Perinatal evolution of mesenchymal hamartoma of the chest wall. J Pediatr Surg. 2010; 45(12):e37-e40.

(62.) McCarthy EF, Dorfman HD. Vascular and cartilaginous hamartoma of the ribs in infancy with secondary aneurismal bone cyst formation. Am J Surg Pathol. 1980; 4(3):247-253.

(63.) Lieberman PH, Jones CR, Steinman RM, et al. Langerhans cell (eosinophilic) granulomatosis: a clinicopathologic study encompassing 50 years. Am J Surg Pathol. 1996; 20(5):519-552.

(64.) Kilpatrick SE, Wenger DE, Gilchrist GS, Shives TC, Wollan PC, Unni KK. Langerhans' cell histiocytosis (histiocytosis X) of bone. Cancer. 1995; 76(12): 2471-2484.

(65.) Willman CL, Busque L, Griffith BB, et al. Langerhans'-cell histiocytosis (histiocytosis X)--a clonal proliferative disease. N Engl J Med. 1994; 331(3):154-160.

(66.) Murakami I, Gogusev J, Fournet JC, Glorion C, Jaubert F. Detection of molecular cytogenetic aberrations in Langerhans cell histiocytosis of bone. Hum Pathol. 2002; 33(5):555-560.

(67.) Fisher AJ, Reinus WR, Friedland JA, Wilson AJ. Quantitative analysis of the plain radiographic appearance of eosinophilic granuloma. Invest Radiol. 1995; 30(8):466-473.

(68.) Hindman BW, Thomas RD, Young LW, Yu L. Langerhans cell histiocytosis: unusual skeletal manifestations observed in thirty-four cases. Skeletal Radiol. 1998; 27(4):177-181.

(69.) Azouz EM, Saigal G, Rodriguez MM, Podda A. Langerhans' cell histiocytosis: pathology, imaging and treatment of skeletal involvement. Pediatr Radiol. 2005; 35(2):103-115.

(70.) Rosai J, Dorfman RF. Sinus histiocytosis with massive lymphadenopathy: a newlyrecognized benign clinicopathological entity. ArchPathol. 1969; 87(1):63-70.

(71.) Paulli M, Bergamaschi G, Tonon L, et al. Evidence for a polyclonal nature of the cell infiltrate in sinus histiocytosis with massive lymphadenopathy (Rosai Dorfman disease). Br J Haematol. 1995; 91(2):415-418.

(72.) Demicco EG, Rosenberg AE, Bjornsson J, Rybak LD, Unni KK, Nielsen GP. Primary Rosai-Dorfman disease of bone. Am J Surg Pathol. 2010; 34(9):1324-1333.

(73.) ShamburekRD, Brewer HB, GochuicoBR. Erdheim-Chester disease: arare multisystem histiocytic disorder associated with interstitial lung disease. Am J Med Sci. 2001; 321(1):66-75.

(74.) Lazzarini L, Mader JY, Calhoun JH. Osteomyelitis in long bones. J Bone Joint SurgAm. 2004; 86(10): 2305-2318.

(75.) Chung T. Magnetic resonance imaging in acute osteomyelitis in children. Pediatr Infect Dis J. 2002; 21(9):869-870.

(76.) Connolly LP, Connolly SA. Skeletal scintigraphy in the multimodality assessment of young children with acute skeletal symptoms. Clin Nucl Med. 2003; 28(9):746-754.

(77.) Ewing J. Neoplastic diseases; a treatise on tumors, 4th ed. Philadelphia: WB Saunders Co.; 1940; 323-324.

(78.) Lichtenstein L. Polyostotic fibrous dysplasia. Arch Surg. 1938; 36(5):874-879.

(79.) Lichtenstein L, Jaffe HL. Fibrous dysplasia of bone: a condition affecting one, several or many bones, the graver cases of which may present abnormal pigmentation of skin, premature sexual development, hyperthyroidism and still other extraskeletal abnormalities. Arch Pathol. 1942; 33:777-816.

(80.) RM, Annels NE. No genomic aberrations in Langerhans cell histiocytosis as assessed by diverse molecular technologies. Genes Chromosomes Cancer. 2009; 48(3):239-49.

(81.) Badalian-Very G, Vergilio JA, Degar BA, Rodriguez-Galindo C, Rollins BJ. Recent advances in the understanding of Langerhans cell histiocytosis. Br J Haematol. 2012; 156(2):163-72.

(82.) Allen CE, Li L, Peters TL, Leung HC, Yu A, Man TK, Gurusidappa S, Phillips MT, Hicks MJ, Gaikwad A, Merad M, McClain KL. Cell-specific gene expression in Langerhans cell histiocytosis lesions reveals a distinct profile compared with epidermal Langerhans cells. J Immunnol. 2010; 184(8):4557-4567.

(83.) da Costa CE, Szuhai K, van Eijk R, Hoogeboom M, Sciot R, Mertens F, Bjorgvinsdottir H, Debiec-Rychter M, de Krijger RR, Hogendoorn PC, Egeler RM, Annels NE. No genomic aberrations in Langerhans cell histiocytosis as assessed by diverse molecular technologies. Genes Chromosomes Cancer. 2009 Mar; 48(3):239-249.

Fabrizio Remotti, MD; Frieda Feldman, MD

Accepted for publication January 10, 2012.

From the Departments of Pathology and Cell Biology (Dr Remotti) and Radiology (Dr Feldman), Columbia University School of Medicine, New York, NY.

The authors have no relevant financial interest in the products or companies described in this article.

Reprints: Fabrizio Remotti, MD, Department of Pathology and Cell Biology, Columbia University Medical Center, VC14-215, 630 W 168th St, New York, NY 10032 (e-mail:
COPYRIGHT 2012 College of American Pathologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Remotti, Fabrizio; Feldman, Frieda
Publication:Archives of Pathology & Laboratory Medicine
Date:Jul 1, 2012
Previous Article:Primitive neuroectodermal tumor/Ewing sarcoma of the retina.
Next Article:Crystalline nephropathies.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters