NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Bast RC Jr, Kufe DW, Pollock RE, et al., editors. Holland-Frei Cancer Medicine. 5th edition. Hamilton (ON): BC Decker; 2000.

Cover of Holland-Frei Cancer Medicine

Holland-Frei Cancer Medicine. 5th edition.

Show details

Chapter 121Bone Tumors

, MD, , MD, , MD, and , MD.

Benign and malignant bone tumors are relatively rare forms of cancer. A great deal of attention has been focused on the malignant bone tumors in recent years because of the great progress made in the multi-disciplinary management of the more common ones including osteosarcoma and Ewing’s sarcoma. 1, 2 The most common malignant primary bone tumor encountered in the general population is multiple myeloma, but this tumor will not be discussed here. Primary non–Hodgkin’s lymphoma (NHL) of bone will be only briefly mentioned, as it pertains to the differential diagnosis. Similarly, rhabdomyosarcoma and Ewing’s sarcoma which may involve bone may be considered in the differential diagnosis. 3

The identification of specific genetic alterations in bone tumors, such as in Ewing’s tumor and rhabdomyosarcoma, has led to improved diagnosis and classification, as well as defining the pathogenesis (see below). 4– 6

The differential diagnosis and staging become increasingly important as multi-disciplinary therapy has progressively improved the prognosis in patients with osteosarcoma, particularly rhabdomyosarcoma and Ewing’s sarcoma. Thus, there has been a progressive improvement in the 5-year survival from diagnosis of all three of these tumors from 20% or less in 1965 to 60 to 70% by 1990. 7

Although many of the benign bone tumors are not generally considered to be cancers, their discussion here is essential because they frequently enter into the differential diagnoses of the malignant bone tumors. Indeed, identification of some of the benign bone tumors is difficult, and differentiation between the benign and malignant bone tumors may require expert pathologic consultation. The clinician dealing with neoplastic diseases should be familiar with benign lesions that can simulate malignant bone tumors.

Classification of Benign and Malignant Bone Tumors

It is convenient to classify benign and malignant bone tumors into lesions that produce bone or osteoid, lesions that produce collagen (or have a predominant fibrous stroma), lesions that produce cartilage, and lesions of vascular or uncertain histogenesis (Table 121.1). Many malignant bone tumors share certain characteristics, for example, a malignant spindle cell stroma.

Table 121.1. Common Benign and Malignant Bone Tumors.

Table 121.1

Common Benign and Malignant Bone Tumors.

The general principles of diagnosis, work-up, and treatment of malignant bone tumors, will be elucidated in this chapter. It is sometimes unnecessary and indeed dangerous to delay definitive treatment planning for a malignant bone tumor in order to obtain pathologic consultation precisely defining the histogenesis of the tumor (e.g., osteosarcoma versus malignant fibrous histiocytoma), since the treatments for the latter types of tumor are identical. With modern effective multi-disciplinary treatment, the prognosis may depend on the rapid institution of treatment.

Staging of Bone Tumors

Bone tumors are rare and are principally treated in a few centers in the United States and abroad. A careful preliminary strategy after multi-disciplinary study of the patient with a suspected primary bone lesion is essential, if one is to arrive at the correct diagnosis and plan out the therapeutic approach that has the best chance of achieving long survival.

Most commonly, all individuals suspected of having a bone tumor should be staged using the Enneking-Musculoskeletal Tumor Staging System. 8 As noted in Tables 121.2 to 121.6, this requires obtaining data to establish the grade of the lesion (benign G0, low-grade G1, and high-grade G2); the anatomic site of the lesion (totally within a bone T1, outside the bone or in a bone very close to other structures such as the popliteal space, hands, or feet T2); and the presence or absence of metastases (M0 or M1).

Table 121.2. Varieties of Osteogenic Sarcoma.

Table 121.2

Varieties of Osteogenic Sarcoma.

Table 121.6. Malignant Round Cell Tumors of Bone.

Table 121.6

Malignant Round Cell Tumors of Bone.

Table 121.3. Chemotherapy for Osteogenic Sarcoma.

Table 121.3

Chemotherapy for Osteogenic Sarcoma.

The preliminary work-up of a patient with a bone tumor requires careful history taking and physical examination; imaging studies (usually radiography) of the affected part as well as the chest; and a technetium-99 ( 99 Tc) bone scan. In addition, most centers perform a laboratory screen to include a complete blood count (CBC), sedimentation rate, calcium, phosphorus, alkaline phosphatase, blood urea nitrogen (BUN) and creatinine, and sugar and protein immunoelectrophoresis. If metastatic disease is considered, additional appropriate studies such as a carcinoembryonic antigen (CEA) or prostate-specific antigen (PSA) may be indicated.

Once the preliminary studies are obtained, it should be possible to make a judgment as to (1) whether the lesion is primary in bone, and (2) whether it is suspected of being a malignant or benign lesion. If it is considered to be malignant, additional staging studies are performed to assess the exact anatomical distribution of the lesion (T1 or T2 and the extent within the bone for the purpose of planning a resection). These studies are principally computed tomography (CT) and magnetic resonance imaging (MRI) of the lesion. The former is best for bony lesions involving the cortex and without a soft tissue mass, while the latter usually clearly illuminates the extent of the marrow involvement and a soft tissue mass outside the bone. A CT of the chest at this point is very helpful in defining metastasis (M), as is the bone scan for bony metastases. If the lesion is suspected of being a round cell tumor (see below), it is essential to add a staging abdominal CT, particularly if lymphoma is suspected, and a bone marrow aspiration. Since 25% of myelomatous lesions are not detectable by bone scanning, a skeletal survey may be useful, if this entity is suspected. Occasionally, angiography is useful in assessing the location of the lesion in relation to major vessels. This may be particularly important in planning resective surgery.

To obtain grade (G), a biopsy is necessary. Before embarking on this last part of the staging procedure, it is essential to assess the data with the radiologist and pathologist to be certain that they concur with the approach and proposed tentative diagnoses. The biopsy should be planned carefully in light of the potential for subsequent resective or ablative surgery. Errors in the planning of the biopsy can be costly to the patient’s welfare. CT-directed needle biopsy may be the method of choice for some lesions and can provide a diagnosis in a high percentage of the patients, but it often yields insufficient amounts of tissue for special studies and is particularly problematic for cartilage tumors or for those tumors that are likely to have necrotic areas.

An open biopsy should be performed through as small an incision as possible and placed in a site that can be resected by either subsequent ablative or resective surgery. The biopsy should be directed through muscles rather than along fascial planes (it is much simpler to resect a muscular segment than to have to excise an entire compartment exposed to a malignant lesion by biopsy through a fascial plane). If a soft tissue mass is encountered outside the wound, it is usually unnecessary to biopsy both the mass and the subjacent bone. If the bone is biopsied, it is often prudent to place a plug of polymethylmethacrylate at the site because it limits leakage of cells into the soft tissues and helps prevent pathologic fracture. The wound should always be cultured; it is too late to do that when the pathologist suggests that the pattern is not one of neoplasm but of infection. A frozen section should always be obtained at the time of biopsy, chiefly to ascertain that one has representative tissue.

Occasionally, for small lesions or for those that are certainly benign, excisional biopsy is a useful technique, but one should be cautious of this, since it may lead to significant problems if the diagnosis is not as suspected.

Once the biopsy is performed and the lesion identified as malignant, then ideally, the surgeon, radiation oncologist, and medical oncologist, along with the radiologist and pathologist, should meet to discuss the most appropriate treatment plan. As will be noted below, this frequently involves some combination of radiation, chemotherapy, and surgery, but the order and extent of these are critical and should be clearly defined and then discussed with the patient and the family.

Bone-Producing Tumors

Benign Bone-Producing Tumors

Stress Fracture

Stress fractures are sometimes called fatigue fractures, which may be a better term in trying to understand the pathogenesis of these common lesions. They represent perhaps the most common pathology of the skeleton to be brought to the attention of the clinician, and although not a true tumor, their radiographic appearance and histopathology can closely simulate osteosarcoma.

Stress fracture is a benign lesion of bone (Figs. 121.1 and 121.2), common in the adolescent and young adult age group. It is usually produced by various forms of athletic injury or prolonged marches (hence the other term “march fracture”) in individuals who are not well conditioned; most commonly a stress fracture is in the long bones (see Fig. 121.1) or metatarsals but can occur in any bone in the body and, at least in some individuals, appears to predilect the proximal and midfemur. The trauma of exertional weight bearing is transmitted through the medial proximal femur. Not only is this a common site of stress fracture in athletic individuals, but it is also a common site in patients who have undergone amputation of the contralateral limb. Since young amputees frequently do not use a prosthesis when they are active, and tend to hop on their one normal leg, a stress fracture can occur in the medial diaphyseal area of the femur. Differential diagnosis from osteosarcoma is always critical, and it is even more challenging if the contralateral amputation was performed for that cause. Stress fractures are also common in the metatarsal bones in the foot (see Fig. 121.2) and are frequently referred to as march fractures. Young female ballet dancers frequently develop stress fractures in the anterior cortex of the tibia due to the abnormal stress of toe dancing.

Figure 121.1. A benign stress fracture in an athletic adolescent (16-year-old female).

Figure 121.1

A benign stress fracture in an athletic adolescent (16-year-old female).A. This patient experienced pain and swelling after an active workout at gymnastics. X-ray appearance of the mid-femoral shaft shows periosteal elevation with a soft tissue mass. (more...)

Figure 121.2. Callus formation around a “large” stress fracture in the 4th metatarsal of an adolescent male.

Figure 121.2

Callus formation around a “large” stress fracture in the 4th metatarsal of an adolescent male. A dense callus formation mimics an extra-osseous calcified mass similar to that seen in osteogenic sarcoma.

The radiographic appearance of the stress fracture can be somewhat disconcerting (see Figs. 121.1 and 121.2). Callus formation within the soft tissues can simulate the appearance of an osteosarcoma with a soft tissue mass. This appearance in an adolescent is always of concern, and it should lead to other imaging studies. CT will be helpful in delineating a stress fracture (see Fig. 121.1), provided that bone windows are obtained to demonstrate a fracture line, which is seldom visible on a plain radiograph. We have found imaging with thallium-201 (a common radionuclide used in cardiac imaging) to be useful in differentiating benign from malignant bone lesions. Although the technetium bone scan will be very hot in a stress fracture and usually eccentrically placed because of the involvement of the cortex, the thallium-201 scan in a mature stress fracture is negative (see Fig. 121.1), whereas it is almost always positive in osteosarcoma.

The histopathology of a stress fracture can be confusing if the lesion is biopsied very early. The initial stages of callus formation include the proliferation of primitive mesenchyme, which can resemble a malignant osteosarcoma. It is only after a period of 2 to 3 weeks that callus formation takes on its typical benign characteristics. After 2 to 3 weeks of laying down a primitive mesenchyme and woven bone, the bone matures into a typical reactive callus formation, typified by bony trabeculae lined with normal benign osteoblasts. Therefore, if a stress fracture is suspected, even though CT scan may not have definitively demonstrated a stress fracture line, the patient should be observed with immobilization (if indicated) for a period of 2 to 3 weeks before considering biopsy. Patients with malignant tumors can, and do, present with pathologic fractures. If a suspected stress fracture does not show evidence of healing after a period of 2 to 3 weeks, a biopsy should be undertaken. Stress or fatigue fractures can usually be distinguished from osteosarcoma in adults, but in children with exuberant callus, the differentiation may be difficult to interpret and the MRI is sometimes misleading, particularly in terms of the marrow changes. A normal alkaline phosphatase, negative thallium scan, and especially the CT are helpful. The CT often shows the extent of the lesion and the relationship of the callus to the linear fracture line, which supports the diagnosis. It is rare that a biopsy is required, but if one is done, application of a short plate is sometimes indicated to reduce the likelihood that a more extensive displaced fracture will occur. Limiting sporting activities, particularly, avoiding long distance running, and immobilization in a cast or brace are almost always effective in controlling symptoms.

Osteoid Osteoma

Osteoid osteoma is a benign bone lesion characterized by pain and sometimes swelling in the affected area. The tumor has a peak incidence in the second decade of life. It is very rare in children age < 10 years of age and extremely rare in adults age > 30 years. The male-to-female ratio is 2.3:1. The most common sites of osteoid osteoma include the femur and tibia, but osteoid osteomas can occur anywhere in the skeleton. Osteoid osteomas of the spine and pelvis can cause considerable pain before they are diagnosed, since routine radiographs of these areas seldom demonstrate the diagnostic radiographic findings. 9

The radiographic appearance of an osteoid osteoma is quite characteristic (Fig. 121.3): a sclerotic lesion somewhat eccentrically placed in the long bones that has a lucent nidus. This “nidus” is the actual tumor that produces a great deal of reactive sclerosis. Sometimes, there is calcification within the tumor or nidus itself, giving the nidus a sclerotic appearance. There is usually a lucent rim around the nidus within the sclerotic bone, however, because of the vascularity of this tumor.

Figure 121.3. X-ray of an osteoid osteoma in a 17-year-old male.

Figure 121.3

X-ray of an osteoid osteoma in a 17-year-old male. The lesion produces dense cortical sclerosis; however, a characteristic “nidus” can be seen as a central lucent area (arrow) diagnostic of an osteoid osteoma.

Histopathologically, the tumor consists of reactive bone surrounding a nidus of fibrovascular tumor (Fig. 121.4). On gross inspection, the nidus of the tumor appears to be “cherry red” because of its rich blood supply. The tumor is richly endowed with blood vessels (accounting for the lucent appearance of the nidus) and nerve fibers (accounting for the characteristic symptomatology of pain). 10

Figure 121.4. The characteristic histologic appearance of a resected osteoid osteoma.

Figure 121.4

The characteristic histologic appearance of a resected osteoid osteoma. The tumor or nidus located in the lower right hand portion of the specimen contains richly anastomosing vascular channels as well as nerve fibers. There is calcification within this (more...)

The clinical history of an osteoid osteoma involves the abrupt onset of pain in the lesional area, which usually awakens the patient at night. Of patients presenting with osteoid osteoma pain, 75% have reported relief of their pain with the ingestion of aspirin, a higher frequency than occurs for malignant bone tumors.

Surgery is recommended for those patients who are not sufficiently relieved with aspirin or other nonsteroidal medication. The only part of an osteoid osteoma that needs to be removed or ablated is the nidus. This is the central radiolucent component of the lesion. In the past, block resections were commonly done, but currently either a burring of the nidus with a high-speed mechanical burr or a percutaneous radiofrequency ablation is recommended. 11


Osteoblastoma, also known as osteogenic fibroma or giant osteoid osteoma, is a rare bone-producing lesion that may affect the long bones or vertebrae. The histologic picture of osteoblastoma is often quite similar to that of osteoid osteoma, but the resemblance ends there. The clinical picture is different, with a considerably less characteristic pain pattern that only rarely is totally respondent to aspirin. Imaging studies show considerably greater variation in the appearance of the lesion, which may be cortical or medullary, lytic or ossified, erosive or marginated, but almost always > 2 cm in longest diameter. 12

Osteoblastoma has a tendency to occur in the spine and, in particular, in the posterior elements of a vertebral body in young adults. As with osteoid osteoma, approximately 75% occur in males, usually in the second or third decade of life.

The radiographic appearance of an osteoblastoma is that of a benign tumor (Fig. 121.5). It is usually an expansile lesion eccentrically placed in the bone, frequently a vertebra, with clearly defined smooth margins. Although the margins of the tumor may expand the bone, they usually do not break the cortex. Tumors occurring in the spine or pelvis often require CT scanning to better define the qualitative nature of the lesion. Most osteoblastomas are cured surgically with excision and/or curettage, which sometimes requires packing with bone chips.

Figure 121.5. Osteoblastoma is a benign lesion of bone that may have a number of presentations depending on the site and how ossified the lesion is.

Figure 121.5

Osteoblastoma is a benign lesion of bone that may have a number of presentations depending on the site and how ossified the lesion is. A. Lesion is in the shaft of the humerus. B. Lesion is in the 5th metacarpal. C and D. Lesion is in the 3rd lumbar vertebra. (more...)

The pathology of an osteoblastoma shows vast amounts of benign bone production with osteoblastic rimming of the bony trabeculae produced by this tumor. Although usually easily identified as a benign fibrous stroma, a great deal of atypia within osteoblastomas occasionally occurs. 13 There is considerable confusion about those osteoblastomas that have an atypical pattern, particularly when they recur locally, break out of the bone, and even metastasize. Such lesions have been termed “aggressive” osteoblastoma or even have gained the appellation “malignant,” but regardless of name, they should be considered to be a form of osteosarcoma (perhaps atypical, in that they may not have as malignant a course) and treated as such.

Osteoblastoma is much rarer than osteosarcoma. A problem may arise in the rare osteosarcoma that does occur in the spine because of a tendency to make the diagnosis of benign osteoblastoma based on the site. The diagnosis depends on histology: a biopsy is always indicated for an osteoblastoma. The rare cases of osteoblastoma that have been termed “aggressive” and then metastasize are really misdiagnosed osteosarcomas. 14 The variety of osteosarcoma that tends to occur in the spine and can be mistaken for an osteoblastoma may be a different clinical or histologic entity than the true classic osteosarcoma of the long bones. In our experience with two patients who were mistakenly diagnosed as having osteoblastoma, both metastasized to bones only and did not show the typical progression of the classic osteosarcoma, which metastasizes to lungs first in the vast majority of patients. The two patients had a more protracted course than usual but eventually succumbed to metastatic osteosarcoma. Experts in the field of skeletal pathology should be consulted about any atypical biopsies of osteoblastoma.

The efficacy of radiation therapy remains uncertain for osteoblastoma. The literature is replete with anecdotal reports of long-term local control or freedom from progression following radiotherapy for incompletely resected/unresectable disease. 15– 19 Doses in these reports vary from 1,000 to 5,500 cGy, casting some doubt on the therapeutic value of radiotherapy. Radiotherapy has been most commonly recommended for incompletely resected vertebral osteoblastoma. 20 Jackson et al. 20 reviewed the osteoblastoma literature in 1977. Eighteen recurrences were noted in a total population of 181 patients. 20 All recurrences occurred in those patients initially incompletely resected. Seven recurrent patients had been electively irradiated after incomplete tumor removal, while 11 received no adjunctive treatment. Marsh and co-workers 19 also documented progressive osteoblastoma despite postoperative radiotherapy. The natural history of osteoblastoma is unpredictable. Several authors have reported patients free of clinical recurrence years following incomplete curettage alone. 18 Malignant transformation of osteoblastoma to osteosarcoma has also been reported for recurrent tumor. Osteosarcoma development has been reported both with and without prior radiotherapy use. 20, 21 These data imply little therapeutic benefit to radiotherapy for typical osteoblastoma. Radiotherapy should be reserved for those patients with unresectable tumor presenting in a dire clinical condition. Campanna et al. 18 successfully managed a large sacral osteoblastoma following preoperative radiotherapy. Several authors report that irradiation may result in an extremely radiodense lesion. 19, 22

Radiation oncologists must be aware of putative histologic variants of osteoblastoma apparently prone to local invasion and postoperative relapse. These variants are termed “aggressive osteoblastoma” and “malignant osteoblastoma.” These terms are frequently used synonymously, but some consider them unique tumors, and the nosology remains controversial. 23 These variants may, in fact, represent low-grade osteosarcoma resembling osteoblastoma. 24 Anecdotal cases of variant osteoblastoma have been controlled by radiotherapy, but this is not a universal finding. Schajowicz and colleagues reported a patient free of disease 2 years following radiotherapy (unstated dose) for tumor cut-through. 25 The same group reported progressive tumor despite radiotherapy in another patient. Mitchell et al. 17 reported progressive tumor growth despite 6,000 cGy for recurrent tumor. This lesion, however, proved to be osteosarcoma at necropsy. In view of the mixed results for radiotherapy and the possibility that these variants are osteosarcoma, the radiation therapy guidelines for malignant/aggressive osteoblastoma follow those for typical osteoblastoma.

Surgical removal is recommended for osteoblastoma. When it is possible, it is best to excise the entire lesion. If a resection would result in permanent loss of function due to the anatomic location of the lesion, an extensive curettage is preferred. Only if a lesion proves to be particularly aggressive and recurs should a resection that produces permanent loss of function be done.

Fibrous Dysplasia

Although monostotic fibrous dysplasia can affect any bone in the body, the ribs and the long bones are the most common sites. The radiographic appearance of fibrous dysplasia is characteristic of a benign bone tumor. 26 It usually produces an expansile lesion in the bone that does not break through the bony cortex. The lesion itself appears to be lucent and is frequently described as having a “ground-glass” appearance because of bone production within the tumor. When there is not much bone production within the fibrous dysplasia, the lesion may have a “soap-bubble” appearance (Fig. 121.6).

Figure 121.6. Fibrous dysplasia may be monostotic or polyostotic and limited or florid in extent.

Figure 121.6

Fibrous dysplasia may be monostotic or polyostotic and limited or florid in extent. The proximal radius shows a remarkable degree of expansion with marked thinning of the cortices and a ground glass appearance to the medullary cavity. The distal humerus shows (more...)

The pathology of fibrous dysplasia is quite characteristic. The lesion contains a low-grade benign fibrous stroma with bone production. Bone production is somewhat random and consists of tiny spicules of woven bone. The spicules of woven bone frequently are quite small; microscopically, they frequently look like what has been termed “alphabet soup,” with little “c’s,” little “j’s,” and little “y’s.”

Fibrous dysplasia frequently presents with a pathologic fracture, and the differential diagnosis then includes pathologic fracture from a malignant bone tumor, particularly low-grade osteosarcoma. The characteristic dysplastic bone production that occurs in fibrous dysplasia is usually diagnostic, but confusion with a low-grade osteosarcoma is possible if the tumor does not produce much dysplastic bone. The absence of anaplasia and variation in the size and shape of the nuclei in the fibrous stroma weigh against a low-grade osteosarcoma.

In the absence of fractures, fibrous dysplasia does not ordinarily produce pain, and frequently these lesions are incidentally diagnosed after radiographs are taken for other reasons. Fibrous dysplasia is also “hot” on bone scan and can be discovered on bone scans done for staging other diseases. When fibrous dysplasia occurs in weight-bearing extremities, it can frequently undergo multiple pathologic fractures before radiographs are taken and diagnosis is finally made. Multiple pathologic fractures with subsequent healing are what account for the “shepherd’s crook” appearance of the proximal femur when fibrous dysplasia occurs in that area. Fibrous dysplasia often presents as a painless swelling on a rib. The characteristic radiographic appearance of the benign tumor allows a presumptive diagnosis of fibrous dysplasia. 27

The treatment of monostotic fibrous dysplasia in adults is almost always observation, unless fracture supervenes or the lesions become painful (suggesting impending fracture). The exception is the proximal femur, particularly in a child. Multiple osteotomies are sometimes necessary to prevent the shepherd’s crook deformity. Treatment of polyostotic disease is principally one of avoiding complications, such as dislocated hips, nonunion of fractures, and significant deformities of the extremities. The threat of malignancy in the absence of radiation (formerly given to patients with cranial deformity to slow the process) is so low that unless one sees obvious structural change, there is no reason to resect or even biopsy the typical lesions of fibrous dysplasia. Patients with polyostotic disease should always be screened for occult hormonal or metabolic abnormality.

Radiation therapy should be avoided in the management of fibrous dysplasia. Although experience is limited, there is no convincing proof that irradiation alters the clinical course of fibrous dysplasia. 28, 29 Tanner et al. irradiated 14 patients with craniofacial fibrous dysplasia to uncertain doses in the kilovoltage era. 28 Eight patients remained unchanged after 5 to 10 years of follow-up. Bony abnormalities progressed despite radiotherapy in 3 patients, and the remainder were lost to follow-up.

Sarcoma complicating fibrous dysplasia is a rare but reported event. 30 Ruggieri and colleagues 29 reviewed 1,122 Mayo Clinic patients with fibrous dysplasia and reported sarcoma development in 28 (2.5%). Most authors report sarcoma development, particularly osteosarcoma, in 0.5% of typical fibrous dysplasia. 30 The incidence of sarcoma approaches 4% in Albright’s syndrome. The role of prior irradiation in sarcoma induction remains controversial for fibrous dysplasia. 31– 33 Yabut 30 reviewed the literature and collected 83 cases of sarcomas arising in the setting of fibrous dysplasia. Of those patients with available clinical information, 46 had received no prior radiotherapy and 23 had been previously irradiated to unstated doses. Others have reported sarcoma development following doses varying from 750 to 6,000 cGy. 28, 31 In the experience of Ruggieri et al., 29 13 patients had received radiotherapy 3 to 52 years earlier (mean, 19). These data imply that radiation exposure is not necessary for sarcoma induction in fibrous dysplasia. Radiation therapy, nevertheless, should not be utilized for fibrous dysplasia since it is potentially harmful. 32, 33

Malignant Bone-Producing Lesions


Osteosarcoma is a malignant spindle cell tumor that produces neoplastic osteoid. Osteosarcoma occurs as several different entities (see Table 121.2) all of which warrant consideration. The classic variety of primary osteosarcoma most often of spindle cells occurs most commonly in the second decade of life. It is slightly more common in males than in females, approximately (M:F 5 = 1.5:1). Osteosarcoma is the most common primary malignant bone tumor, with the exception of myeloma, and it reaches its peak frequency in the growing adolescent. Prior to the advent of adjuvant chemotherapy, this tumor carried an 80% mortality rate despite surgery. Some of the different subtypes listed in Table 121.2 have a more benign clinical course, and inclusion of some of these more benign clinical entities in reported series of osteosarcoma may have resulted in an erroneously high cure rate with surgery alone in the past. Recent clinical trials of adjuvant chemotherapy in osteosarcoma have clarified and confirmed the poor prognosis of the classic variety (or high-grade) osteosarcoma treated with surgery alone. 7, 34, 35

Although osteosarcoma is a common primary malignant bone tumor, it is still rare, with less than 1,000 new cases reported per year in the United States. 7

The etiology of osteosarcoma is obscure, but it appears in some forms to have an oncogenic predilection and in others to be related to chronic (or acute) bone injury. The oncogenic focus was first noticed in association with retinoblastoma (Rb gene abnormality) 36, 37 but more recently with errors in the p53 and p16 gene structures, both of which produce protein products that exert a regulatory effect on DNA synthesis and, at least for p53, promotes apoptosis (programmed cell death). A mutation or genetic error in one copy of these genes makes the cell vulnerable to a second random hit or chromosomal loss. Unremittent cell cycling then compounds DNA mutations, leading to the development of a sarcoma (or, for the p53, a number of carcinomas as well). Proponents of the theory that chronic bone injury is a cause of osteosarcoma point out that patients with Paget’s disease, radiation injury, and certain forms of hardware have vastly increased frequencies of occurrence of osteosarcoma and malignant fibrous histiocytoma, compared with normal individuals of the same age group.

Other etiologic clues to the development of osteosarcoma can be found in studying its epidemiology. Patients with hereditary retinoblastoma have a high risk of second cancers, 50% of which are osteosarcomas. Osteosarcoma can arise in patients with multiple exostoses, Paget’s disease of bone, and fibrous dysplasia. 7 Osteosarcoma occurs in adolescents at the time they are most rapidly growing in their most rapidly growing bones. Rapid bone growth may, thus, play some role in the development of osteosarcoma. Frequently, one elicits a history of trauma from patients presenting with osteosarcoma. Probably, however, trauma merely calls attention to a pre-existing lesion. Osteosarcoma does occur in rapidly growing bones that undergo the most stress of weight bearing, however, with the sites about the knee (distal femur and proximal tibia) being the most common. The role of weight-bearing trauma as a possible etiologic factor is suggested by the fact that osteosarcoma is 80 times more common in large dogs weighing over 20 lb than in smaller breeds. In addition, the most common site of osteosarcoma in dogs is also the bones that bear the most weight, the forelimbs.

Classic osteosarcoma can have various appearances depending on whether the tumor produces a great deal of bone (an osteoblastic osteosarcoma) or whether it is a predominantly lytic lesion (telangiectatic osteosarcoma). The osteoblastic variety of osteosarcoma has a characteristic radiographic appearance (Fig. 121.7) in the metaphyseal area. It has the appearance of a malignant tumor that produces a mottled sclerotic and lytic lesion in the medullary cavity of the bone as well as a soft tissue mass. The soft tissue mass in sclerotic lesions is calcified and has a typical “sunburst” appearance (see Fig. 121.7). Less typical lesions can occur in any area of the bone. A lytic osteosarcoma, a common type known as the telangiectatic variety, can occur in the diaphysis, where pathologic fracture can take place (Fig. 121.8). Osteosarcoma is “hot” on the bone scan, as are many other lesions. We have found imaging with thallium-201 to be very specific for the imaging of malignant tumors of bone and soft tissue, and we have also utilized the thallium scan to follow the patient’s progress while on preoperative chemotherapy (Fig. 121.9). The thallium scan may also be useful in differentiating pathologic fractures from benign stress fractures (see Fig. 121.1). 38

Figure 121.7. Osteogenic sarcoma of the distal femur in an adolescent male.

Figure 121.7

Osteogenic sarcoma of the distal femur in an adolescent male. There is a mixed lytic and sclerotic lesion in the medullary cavity of the bone. The soft tissue mass contains radiating spiculated calcification typical of a malignant osteogenic sarcoma. (more...)

Figure 121.8. A lytic or telangiectatic osteogenic sarcoma presenting as a pathologic fracture in the diaphysis of the femur.

Figure 121.8

A lytic or telangiectatic osteogenic sarcoma presenting as a pathologic fracture in the diaphysis of the femur.

Figure 121.9. Serial thallium 201 scans performed before and after preoperative chemotherapy for an osteogenic sarcoma of the proximal tibia.

Figure 121.9

Serial thallium 201 scans performed before and after preoperative chemotherapy for an osteogenic sarcoma of the proximal tibia. Note improvement of the intense thallium 201 uptake 2 months later following chemotherapy. This improvement in the thallium scan (more...)

Classic osteosarcoma in its several varieties is composed of a predominant malignant spindle cell stroma. The spindle cells show variation in the size and shape of the nuclei and of the cells (pleomorphism) and frequent mitotic figures. Neoplastic bone formation in hematoxylin and eosin–stained sections takes the form of pink staining amorphous material or osteoid. This osteoid can go on to calcify (Fig. 121.10) and to produce the characteristic radiographic appearance at the time of diagnosis. In addition to malignant spindle cells, telangiectatic osteosarcoma contains aneurysmally dilated vascular spaces filled with blood (Fig. 121.11). These spaces are not lined by normal endothelium, but rather by tumor cells. Other varieties of osteosarcoma produce predominantly fibrous stroma with very little tumor osteoid and are given the designation of fibroblastic osteosarcomas. Still others produce predominantly neoplastic cartilage and are called chondroblastic osteosarcomas. By definition, all osteosarcomas produce some degree of osteoid or bone.

Figure 121.10. Osteosarcoma.

Figure 121.10

Osteosarcoma. The tumor cells are spindle shaped and polyhedral and their nuclei are markedly pleomorphic and hyperchromatic. Several foci of red-pink lacelike and microtrabecular osteoid are interposed between the tumor cells (252 ×). (courtesy (more...)

Figure 121.11. Telangiectatic osteosarcoma.

Figure 121.11

Telangiectatic osteosarcoma. Left , low-power section reveals that the tumor cells are arranged in a few solid nests within large vascular sinusoids (12.5 ×). Right , a higher power photograph demonstrates a focus of early micrabecular osteoid (more...)

Two rare clinical subtypes, of parosteal osteosarcoma and periosteal osteosarcoma may have a lower risk of metastasis and improved prognosis, compared with classic osteosarcoma. 7

The differential diagnosis of osteosarcoma includes benign conditions that can mimic osteosarcoma on the radiograph. This includes other bone-producing lesions, stress fractures, osteoblastomas, and other less malignant varieties of osteosarcoma, such as the parosteal osteosarcoma. It is sometimes difficult to distinguish a low-grade fibrous lesion, such as fibrous dysplasia, from a low-grade fibroblastic osteosarcoma.

Patients with osteosarcoma usually present with pain in a lesional area. This frequently consists of pain around the knee in patients with femur or tibia primaries. The pain is usually progressive and is not related to the time of day. Most lesions start to produce pain as the tumor starts to expand the bony cortex and stretch the periosteum. Swelling and tenderness occur late in the course of osteosarcoma, and then a large soft tissue mass may appear. The serum alkaline phosphatase is elevated in 60% of patients with osteosarcoma at the time of diagnosis. The patient is usually afebrile. 7, 39, 40

The differential diagnosis of osteosarcoma includes rhabdomyosarcoma and Ewing’s sarcoma. Rhabdomyosarcoma may occur at any site, most particularly the orbit, and the bladder and/or prostate, where it may produce urinary obstruction. Most tumors of the orbit are embryonal and have a high rate of cure. Tumors of the limbs are alveolar. They commonly spread to lymph nodes and have a much less favorable prognosis.

Ewing’s sarcoma and rhabdomyosarcoma often metastasize to bone marrow, which is not true for osteosarcoma. Genetic changes in rhabdomyosarcoma include chromosomal translocation T(2:13) and less commonly T(1:13). This leads to a chimeric protein. Use of in situ hybridization and real-time polymerization chair reaction (RTPCR) allow for molecular, and therefore, much more sensitive determination of these translocations. Ninety-five per cent of patients with Ewing’s sarcoma have 11/22 or 21/22 translocation. These changes involve the ETS gene family and infusion of the EWS gene on 22 with the FL11 gene on chromosome 11. 7, 37

Osteosarcoma must be diagnosed by biopsy. From the various considerations discussed above, it is not safe to rely solely on a typical appearing radiograph to make the diagnosis. Clinical staging consists of a CT scan of the lungs to rule out pulmonary metastases and a bone scan to rule out the presence of bone metastases. Once the biopsy confirms a fully malignant osteosarcoma, we do a thallium-201 scan prior to starting treatment to identify the lesional area and to serve as the zero time point to follow the course of thallium uptake while the patient is on treatment. 38 The biochemistry profile should be performed, including a serum alkaline phosphatase. The serum alkaline phosphatase can be used as a tumor marker to monitor the patient’s progress through preoperative chemotherapy, but the alkaline phosphatase must always be considered in relationship to the other liver enzymes, since it is not always practical to perform bone isoenzymes. Furthermore, various chemotherapeutic agents, in particular high-dose methotrexate with leucovorin rescue, can alter liver functions and serum enzymes, including the total alkaline phosphatase.

Because at present the diagnosis of osteosarcoma carries with it a commitment to preoperative chemotherapy, patients should also have a creatinine clearance to determine the adequacy of their renal function. Most of the drugs utilized in the treatment of osteosarcoma depend on renal excretion. Because modern surgery for osteosarcoma usually consists of limb salvage surgery, it is also important to do various imaging studies to determine the extent of tumor within the bone and soft tissues. The definitive surgery should be based on the original extent of disease prior to the shrinkage that may occur from preoperative chemotherapy. Full-length scanograms of the affected extremities should be done to accurately measure the patient for an endoprosthesis. CT scans should be done to determine the extent of bone destruction within the primary bone lesion, and an MRI should be done to best determine the extent of medullary involvement of the tumor as well as the extent of soft tissue involvement. 38, 39

Limb Salvage Surgery and the Chemotherapy of Osteosarcoma

Prior to 1972, the cure rate of osteosarcoma (or osteogenic sarcoma, OS) was less than 20%. 7, 39 The primary tumor most commonly presented on the extremities and was, therefore, subject to local control by amputation. However, 80% of such patients have micrometastases in the lungs at the time of diagnosis, such that relapse with overt pulmonary metastases occurs usually within 6 to 12 months and almost always by 3 to 4 years. 7

Chemotherapy was occasionally employed in the adjuvant situation, but more commonly in patients with overt lung metastases. A number of agents were studied prior to 1970. Only the alkylating, or related agents, such as cyclophosphamide, melphalan, mitomycin C and (DTIC) exhibited activity, albeit low level, with response rates in the range of 15%. Moreover, responses were partial and of short duration. 39, 40

Since 1970, three agents have been found to be reproducibly active in patients with metastatic osteosarcoma. These include high-dose methotrexate with leucovorin rescue (HD-MTX) and doxorubicin which were discovered in the early 1970s and cisplatin with was discovered in the late 1970s. Doxorubicin and cisplatin, in various studies, consistently produced response rates in the range of 30% most of which were partial. Reported response rates to HD-MTX were more variable ranging from 10 to 70%. This variation probably reflected heterogeneity of HD-MTX protocols in terms of dose, time to rescue, hydration, monitoring of MTX pharmacokinetics and renal function (see below). 41– 46

It has been proposed that a major mechanism of response of osteosarcoma to chemotherapy may be the result of differentiation and not necrosis/apoptosis. Indeed, metastases over time have been observed to show increasing osteoid and bone formation. The extent to which this compromises the interpretation of chemotherapy effects is not known but probably is not major.

A complete surgical removal of the osteosarcoma is necessary for control of the tumor. In most cases, this requires complete removal of the end of the bone involved including the articular surface. The tumor is removed with a covering of soft tissue to obtain what is called a “wide” surgical margin. The combination of better radiographic studies (CT and MRI) and preoperative chemotherapy have allowed surgeons to safely do adequate surgical resections without amputating the extremity. Recent studies have demonstrated that the surgical margin and percent necrosis is predicative of the risk of a local recurrence. 46 This report indicates that if there is better than 95% necrosis, the risk of a local recurrence is less, compared with the situation where the percent necrosis is less than 95%. The assumption from the data is that a closer margin is needed for the patient whose tumor has 95 to 100% necrosis, compared with those with less necrosis.

Amputation has some advantages, compared with limb salvage resection, and can occasionally be better for the patient. When the patient has an extremely large tumor, which does not respond well to chemotherapy, amputation is often recommended. In the patient < 5 years of age, and in the patient who is not willing or does not want to have multiple operations over their lifetime, amputation is better. A variation on a conventional amputation called a rotationplasty has been used for the patient < 10 years of age. This operation offers an excellent method of total removal of the tumor and better function than a conventional amputation.

There are numerous methods available for reconstruction of the extremity after a limb salvage resection. Initially, the reconstruction was done with an arthrodesis of the adjacent joint, usually the knee, or a custom-made metallic endoprosthesis. Currently, the two most common methods of reconstruction are an allograft bone transplant and a modular endoprosthetic device. Both are acceptable. The advantage of allograft reconstruction is that it replaces the resected bone with a biologic material that has the ability to replace the resected bone. The disadvantage is that it takes longer to heal and the early function of the extremity is not as good, compared with an endoprosthetic replacement. The advantage of an endoprosthetic replacement is that the patient recovers quickly and the early function is superb. The disadvantage is that the endoprosthesis wears out over time and most will need to be replaced between 15 and 20 years. Each has it proponents. Patients should be offered both. (For details see “Limb Salvage Surgery” below).

The patient with substantial growth left in the resected bone will have a limb length discrepancy if a conventional reconstruction is done. Lewis 47 developed an endoprosthesis which has a mechanism which allows “growth.” This concept has been accepted and numerous surgeons have used a variety of techniques to lengthen the young patient’s extremity. Repeated surgical procedures are necessary over the growth period of the patient.

Adjuvant Chemotherapy

It has been demonstrated in preclinical in vivo models that the antitumor effect, particularly the curability, of model tumors, such as mouse leukemia L1210 and breast cancer MCF7, is increased if treatment is directed at minimal, ideally microscopic, tumor burdens. 46, 47 Also, the microenvironment, such as oxygen tension, vascularity, and growth fraction, of larger tumors is adverse with respect to chemotherapy. 48 Accordingly, in 1972, studies were initiated involving the use of chemotherapy immediately following amputation (adjuvant chemotherapy) directed at eradicating micrometastatic tumors in the lungs (see above and below). 44, 45

The first two adjuvant chemotherapy studies of osteosarcoma were reported in 1974. 49, 51 The sequence of relevant positive adjuvant and neoadjuvant studies in osteosarcoma are presented in Table 121.7. 49– 65 Study #1, which employed HD-MTX in the adjuvant setting, produced a 40% relapse free survival. 49 Osteosarcoma is a kinetically aggressive tumor such that relapse, if it is to occur, does so 90% of the time in the first 6 months and decreasingly so in up to 2 years, after which relapses are rare. All the studies in this report have a minimum of 3 years follow-up and usually much longer. Consequently, percent relapse-free survival is a definitive determinant of response .

Table 121.7. Representative Trials of Adjuvant and Neoadjuvant Chemotherapy of Osteosarcoma.

Table 121.7

Representative Trials of Adjuvant and Neoadjuvant Chemotherapy of Osteosarcoma.

Study #2 was a later trial performed by the National Cancer Institute (NCI) which confirmed study #1. 50

Study #3 performed by the Cancer and Leukemia Group B (CALGB) involved doxorubicin as adjuvant. This resulted in a 39% relapse-free survival. 51 In an independent but co-temporal study, performed at the (DFCI), a 42% disease-free survival with HD-MTX was achieved. 54

Study #4 performed by the CALGB involved alternating or cycling HD-MTX and doxorubicin on a monthly basis for a maximum of 6 months. 52 The DFCI employed the same two programs given concurrently (study #5 in Table 121.7). 53 This was possible because with appropriate monitoring of renal function and pharmacology (see below), HD-MTX can be delivered without myelosuppression or mucositis and, thus, at twice the dose rate, that is, twice the summation dose intensity of the rotating combination.

The relative dose rates for the concurrent and sequential (cycling) programs are presented schematically and hypothetically as follows:

Image ch121e1.jpg

Most subsequent studies tended to confirm the above and the general conclusion that such adjuvant chemotherapy produced a 40% disease-free survival plateau, compared with the historical control of 15%.

But, a major challenge to the significance of these findings was mounted particularly by the Mayo Clinic (study #8). 56 Investigators there and at two institutions in Europe found that in the absence of adjuvant chemotherapy, an increasing proportion of patients had increasing relapse-free survival up to 45 and 50%. 56 Suggested explanations included prognostic factor drift, changing referral patterns, the introduction of computer scanning, and other variables. Because of this and because of the cost in morbidity and dollars, many patients were not offered adjuvant chemotherapy.

Accordingly, a randomized comparative study which included a “no chemotherapy” control arm was proposed (study #9, Table 121.7). Opponents of the comparative study felt that the effectiveness of adjuvant chemotherapy had been established, and that a “no chemotherapy” arm was therefore neither justified nor ethical. This position was strongly supported by the results of the neoadjuvant trials at the Memorial Sloan-Kettering Cancer Center (MSKCC) where relapse-free survival rates of 70% were being reported.

Advocates of the comparative study cited the failure of the comparative studies to confirm the noncontrolled studies, the Mayo Clinic data as above, and, most particularly, the fact that many patients were not being offered chemotherapy because of the Mayo Clinic data.

Two no-treatment control studies (study 9 and study 10 in Table 121.7) were conducted. 57, 58 The treatment arms consisted of HD-MTX and doxorubicin given concurrently. The results of the two studies were similar, each demonstrating in a small number of patients (about 40) a highly statistically significant superiority for adjuvant chemotherapy. Thus, the relapse-free survival was 60 to 70% in the chemotherapy arms, compared with 15 to 25% in the control arms. 57, 58 This difference was present for both randomized patients and for patients treated “off” protocol. These definitive results settled the issue. That adjuvant chemotherapy was effective, there is no doubt. The question was how could it be made more effective?

The importance of the dose and schedule as well as combination chemotherapy was examined in the context of larger studies.

With respect to dosage of doxorubicin, there was suggestive evidence that 90 mg/m2 q3 weeks was superior to 60 mg/m2. For HD-MTX, Rosen provided evidence that 8 to 12 g/m2 q3 weeks was superior to lower doses and DFCI investigators found that HD-MTX administered once weekly was superior to q3 weeks. 59– 61

With this emphasis on dose intensity, particular attention was given to supportive care and to methods of reducing toxicity. This was particularly important with the HD-MTX program, wherein the gram quantity doses had the potential of major, even fatal, toxicity. It was found that HD-MTX could be delivered safely, indeed, with minimal toxicity, if pharmacokinetic and toxicity studies were applied. In brief, HD-MTX may produce transient renal damage, in part by precipitation in the renal tubule, which, because MTX is excreted by the kidneys, results in high and prolonged blood levels with resultant serious and potentially fatal toxicity. Hydration and alkalanization will prevent such damage. If such damage has occurred, it can be detected early, that is, 24 hours after the start of HD-MTX by an elevated serum creatinine and/or an elevated blood MTX level. This calls for “super-rescue,” that is, an increased dose of antidote, leucovorin, adjusted to the elevation of the serum MTX at 24 hours and over the next 3 days (see Bleyer for details). 61– 64

Neoadjuvant Chemotherapy

While adjuvant chemotherapy clearly increased relapse-free survival for osteosarcoma from 20 to 60%, major challenges remain. Amputation required for control of the primary tumor adversely affects the long-term quality of life. Moreover, up to 40% of patients still died of their disease. To address these and other limitations of adjuvant chemotherapy, the strategy of moving chemotherapy “up front,” that is, to 1 to 3 months preceding surgical treatment of the primary, was instituted. The initial studies were undertaken at the MSKCC in the mid-1970s. These studies were motivated by the orthopedic surgeons there who needed 1 to 2 months to select, modify, and/or fabricate a prosthesis. Thus, “holding the tumor in check” by chemotherapy for 1 to 2 months provided the necessary time. 64, 65

This novel therapeutic strategy became known as neoadjuvant chemotherapy. This strategy resulted in the discovery at the MSKCC, and independently at the DFCI, that the primary tumor could be profoundly sensitive to chemotherapy. 66 In addition to osteosarcoma, neoadjuvant chemotherapy was producing provocative effects in other tumors, such as head and neck cancer, as well. Indeed, the uniqueness of neoadjuvant chemotherapy as a treatment strategy, the integration of clinical and basic science rationale—and thus opportunities—and the approach was recognized, organized, and presented for the first time as an integrated innovative treatment strategy that was deserving of study.

Selected therapeutic opportunities provided by the neoadjuvant approach and the application to osteosarcoma include the following: 64– 67


Early treatment of micrometastatic tumor. Chemotherapy of metastases is moved forward in time 1 to 3 months. Cytokinetic and drug resistance studies indicate that resistant cell lines may be selected during that period. The development of resistance within a given time relates directly to the cytokinetic thrust of the tumor. Osteosarcoma is one of the faster growing of human tumors. While noncontrolled studies, particularly at the MSKCC, indicated that neoadjuvant chemotherapy might improve the control of micrometastatic metastases, such was not the case in subsequent randomized comparative studies, where the only variable was moving the chemotherapy up front. Indeed, follow-up studies at the MSKCC did not confirm their earlier results. 59, 61, 66, 68


Neoadjuvant chemotherapy–induced regression of the primary tumor. In osteosarcoma, head and neck cancer, and perhaps other tumors, the primary may be substantially more chemosensitive than the regional or distant metastases. Tumor cells capable of metastasis have a higher mutation rate than the primary tumor, a feature that accelerates resistance development. 61, 68


Reduction of the primary tumor by chemotherapy may permit local control of the primary with less radical procedures. For osteosarcoma, neoadjuvant chemotherapy response of the primary tumor was associated with the successful use of limb preservation procedures, which became possible in over 60% of patients (see above). Orthopedic surgeons emphasized that it is the delay in surgery which permits fabrication of the endoprosthesis that is most important to limb preservation, although reduction of soft tissue invasion at the primary site also facilities limb presentation 69, 70 (see “Surgical Aspects of Limb Preservation” below).


In vivo assay for response. Response to new chemotherapy may produce effects of subsequent adjuvant chemotherapy and predict for survival, and related to lack of response to neoadjuvant chemotherapy, it may allow adjustment of the adjuvant chemotherapy to a different and more sensitive regimen.

Treatment of Overt Pulmonary Metastases in Osteosarcoma

The Biology of Lung Metastases

Long-term disease-free survival can be achieved in 30 to 40% of patients with osteosarcoma metastatic to the lungs. Thus, the physician should “continue to think cure” in such patients. Factors that are favorable for pulmonary research of metastases include the following.

Pulmonary metastases can be removed by wedge resection with good margins. In fact, local recurrence at a site of lung metastasis is uncommon. The fewer the tumors, the better is the prognosis. But, cure can be achieved with surgical research of as many as 15 to 20 pulmonary metastases. Local recurrence at the site of removal occurs infrequently, if margins are adequate. Involvement of the adjacent structures, such as the pleura and mediastinum, render the tumors nonresectable. The growth rate of the lung metastases is an important prognostic determinant. Metastases that grow to detectable size on chest film in 6 months have a substantially worse prognosis, compared with those that appear after 1 year.

There is, however, considerable variation in the impact of numbers of tumors and growth rates such that treatment with curative intent is rational for many patients with significant numbers of metastases, regardless of number and growth rate. It has been proposed that a less aggressive approach to patients with multiple fast-growing lesions may explain the poorer prognosis in this group.

There is evidence that patients who relapse after cytoreductive surgery do so because of micrometastatic tumor present at the time of diagnosis or metastatic from metastases. Thus, a second and third or more resections may be indicated. As with the treatment of primary osteosarcoma, the integrated delivery of chemotherapy to eradicate micrometastases, while the surgeon removes the macrometastatic tumors, represents rational interdisciplinary therapy. Whether chemotherapy should precede surgery, be given at the time of surgery, or follow (analogous to neoadjuvant, concurrent, or adjuvant) remains an open question. However, it is hard to obtain evidence that chemotherapy is active in this setting. Given the modern concepts of drug resistance which emphasize a final common pathway and given the cross-resistance in these patients who have already received the gamut of known active agents, it seems likely that they would be chemoresponsive after relapse.

Limb Salvage Surgery

Limb salvage surgery for osteosarcoma historically began with the efforts of Eilber and Marcove in the United States and by Salzer in Austria, who introduced metallic devices; of Enneking and Campanacci, who described methods for autograft reconstruction; and of Parrish, Ottolenghi, Gross, and Mankin, who utilized frozen cadaveric massive allografting, all as techniques for replacement of a resected part. 39– 41 Early limb salvage surgery for osteosarcoma of the distal femur consisted of replacement of an entire femur and knee joint for fear of skip lesions in the primary tumor-bearing bone. As experience was gained in the use of limb salvage surgery and with the continued improvement in preoperative chemotherapy as well as the improved survival with adjuvant chemotherapy for osteosarcoma, limb salvage surgery became more popular. Limb salvage surgery for osteosarcoma was originally limited to patients who had achieved most of their linear growth at the time of diagnosis and to patients who did not have large soft tissue masses. The introduction of the expandable prosthesis by Lewis 47 made it possible to do limb salvage surgery in younger children who had not achieved their maximum growth, since the new prosthesis could be expanded at periodic intervals (Fig. 121.12).

Figure 121.12. A total humerus replacement with an expandable prosthesis used for a limb salvage surgery in an 8-year-old female with a malignant tumor (in this case a Ewing’s sarcoma) that involved the entire humeral shaft.

Figure 121.12

A total humerus replacement with an expandable prosthesis used for a limb salvage surgery in an 8-year-old female with a malignant tumor (in this case a Ewing’s sarcoma) that involved the entire humeral shaft. With the use of such new devices, (more...)

The refinement of chemotherapy given preoperatively to patients with osteosarcoma has led to dramatic shrinkage in soft tissue masses as well as complete histologic responses in many of the tumors. This has also encouraged more surgeons to become interested in limb salvage surgery. Today, most patients can successfully undergo limb salvage surgery rather than amputation, and most have an almost equivalent chance of being cured. In some patients, limb salvage surgery is still not 100% as safe as amputation. Amputation is no guarantee against local recurrence, since local recurrences are known to take place in the stump. Indeed, today, many patients who undergo amputations do so because of the very large size of their tumor and certain other high-risk situations. These patients are at a greater risk for local recurrence even with an amputation. Had limb salvage surgery been done on these high-risk patients, the local recurrence rate would be much higher.

Factors that seem to predispose to a local recurrence following limb salvage surgery include venous invasion by tumor, noted at the time of surgery; very large soft tissue masses; tumor eroding into the muscle through the fascial plane, where the entire muscle group is not removed from origin to insertion; and a poor histologic response to preoperative chemotherapy (i.e., 95% necrosis attributable to chemotherapy) noted after histologic review of the entire resected specimen. Even without these noted risk factors, patients who undergo limb salvage surgery are still at a slight risk for local recurrence, and clearly amputation is still the safest procedure to perform. However, because of the feasibility of limb salvage surgery, many surgeons are doing it, and most patients opt for limb salvage surgery when initially given a choice. However, the surgeon and oncologist must keep in mind the risk factors involved and present a unified consensus to the patient. At the very least, limb salvage surgery for osteosarcoma should be performed by surgeons who are experienced in it. Tumors of the proximal femur seem to have a greater propensity to recur locally because of the large soft tissue mass that develops prior to diagnosis of osteosarcoma in that area. The frequent occurrence of pathologic fractures in proximal femur lesions and the lack of a readily palpable soft tissue mass with which to judge the efficacy of preoperative chemotherapy are additional risk factors for a poor prognosis in proximal femur lesions.

The results of experienced surgeons in doing limb salvage surgery represents a remarkable advance in the treatment of osteosarcoma. Most patients who undergo successful limb salvage surgery have function remarkably superior to that obtained by amputation and, of course, superior cosmetic results. Most can now have limb-sparing surgery with a low rate of local recurrence (less than 5% for most series) and, depending on the device selected, more or less long periods of incident-free existence. The devices are far from perfect, however, and for the metallic systems, early failure because of infection accounts for up to 10% of the failures, and over 50% show loosening at 5 years. In contrast, for the allografts, approximately 20% of the patients lose their grafts in the first 3 years as a result of fracture or infection but achieve good stability of the system for years following that.

It should be noted that not all patients’ tumors are suitable for limb salvage. This category includes very large tumors, those that involve major nerves or vessels, those where the resection is intralesional, those located below the midtibia where the results of limb-sparing surgery and replacement are not as functional as a below-the-knee prosthesis, and those in very young children who will face up to 6 or more inches of limb length inequality as a result of growth after standard procedures. The expandable metallic devices may be helpful in this latter group, but most pediatric orthopedic oncologists prefer to perform turnabout procedures for lesions of the distal femur, which, in fact, leave the patient with a functional “knee” joint (in fact, the ankle turned 180°) and a useful hip and “femoral shaft” (the tibia).

The newest drug of importance against osteosarcoma is ifosfamide (IFF). This alkylating agent was available and on the market in Germany as early as 1968. It is only recently that it has been popularized in the treatment of malignant bone tumors, both Ewing’s sarcoma and osteosarcoma. The use of IFF is made possible by the commercial introduction of the uroprotective agent sodium mercaptoethylsulfonate (MESNA). When MESNA is given at equal doses with IFF, it neutralizes active metabolites in the urine, thereby protecting the bladder from hemorrhagic cystitis, which had been the limiting toxicity. Initial phase II studies of IFF combined with MESNA in the treatment of metastatic osteosarcoma led to response rates of approximately 20 to 30%, utilizing doses of IFF of 6 to 10 g/m2. 53– 55 For the treatment of the adult spindle cell sarcomas and metastatic osteosarcoma, we have utilized IFF with MESNA at the dose of 14g/m2 as a single agent. 56 At this higher dose, we obtained a major response rate (complete and partial remissions) of 67% in 21 patients with metastatic osteosarcoma (whose tumors were resistant to combination therapy with cisplatin and doxorubicin as well as to HD-MTX). Thus, with IFF, it appears that there is also a dose–response curve that is rather steep after one exceeds the dose of 10g/m2. At the dose of 14 to 18 g/m2, IFF/MESNA is well tolerated and produces acceptable toxicity. Although myelosuppression is significant, recovery usually occurs within 3 weeks of the first day of administration of the drug. IFF/MESNA is much more readily tolerated than high-dose cisplatin combined with doxorubicin. We administer high-dose IFF by giving 2g/m2 as a 4-hour infusion followed by 2g/m2/d by 24-hour infusion for consecutive days to achieve the dose of 14 to 18g/m2. Doses as high as 20g/m2 have been given without the need for extraordinary support measures (such as autologous bone marrow transplantation). During the 7-day infusions, potassium and bicarbonate replacement must be meticulously done, since they are excreted in excess in the urine. The major toxicity of IFF/MESNA is the production of Fanconi’s renal tubular acidosis.

Current studies are directed at defining the role of IFF. IFF, particularly in combination with etoposide, has produced a high response rate in patients with metastatic osteosarcoma. An ongoing phase III randomized study compared HD-MTX + doxorubicin + cisplatin with or without etoposide + IFF. The study has major statistical power and should be decoded and presented within the next 2 years.

Parosteal Osteosarcoma

Parosteal or juxtacortical osteosarcoma is a low-grade osteosarcoma that arises on the surface of the bone. 65 It rarely penetrates to the intramedullary compartment and rarely demonstrates transformation to highly malignant osteosarcoma.

Parosteal osteosarcoma typically arises in an age group different from the classic osteosarcoma of adolescents (see Table 121.2). It is more common in females than males, is most common in the third and fourth decades of life (ages 20–40), and typically arises in the posterior aspect of the distal femur, although any bone in the skeleton can be affected. The tumor arises on the surface of bone in association with the periosteum (Fig. 121.13).

Figure 121.13. The AP and lateral radiographs demonstrate the characteristic presentation of a parosteal osteosarcoma.

Figure 121.13

The AP and lateral radiographs demonstrate the characteristic presentation of a parosteal osteosarcoma. As can be seen on the CT scan, the lesion is located in the popliteal fossa and grows on the surface of the bone extending in irregular fashion posteriorly. (more...)

The radiographic appearance is that of a heavily ossified exophytic tumor that is closely applied to the cortex of the bone. The medullary cavity of the bone is invaded not infrequently, and although it was thought in the past that this was a poor prognostic sign, recent studies have not supported this view, and the histologic grade of the tumor (usually low grade) appears to be a more significant predictor of outcome and the need for adjuvant therapy. Normally, the histologic appearance of these tumors is quite striking, in that they are almost always heavily ossified and show a sparse, or at most moderately abundant, spindle cell population, which appears to be making the bone (Fig. 121.14).

Figure 121.14. Parosteal osteosarcoma.

Figure 121.14

Parosteal osteosarcoma. Amedium power photograph of the typical low-grade tumor shows, rather abundant immature microtrabeculae of tumor bone and an intertrabecular fibrous stroma which is the actual tumor stroma. The histological changes are rather bland and (more...)

The differential diagnosis of parosteal osteosarcoma included periosteal osteosarcoma (see below), osteocartilaginous exostosis, and myositis ossificans traumatica or circumscripta. It is important to note that osteocartilaginous exostoses “share a cortex” with the bone, rather than arising from the surface of the bone, as does parosteal osteosarcoma. Similarly, myositis ossificans can be distinguished on imaging studies on the basis of the frequent appearance of a clear space between the lesion and the cortex rather than being firmly attached, as is noted with parosteal osteosarcoma. Another important differential is that osteosarcoma that invades the cortex destroys the bone and has a similar appearance and the same potential to metastasize as a central osteosarcoma.

The clinical history presented by patients with parosteal osteosarcoma is one of minimal pain and tenderness over the site and, if in its characteristic location, progressive limitation of flexion and ultimately extension of the knee joint. The tumor may be present for long periods prior to discovery and grow to rather spectacular sizes, with pain only supervening in the late phases. The treatment is for the most part surgical, with wide-of-marginal surgery usually being curative. The most difficult site for surgical resection is the popliteal space, where the proximity to the artery and nerves makes a wide resection more problematic. It is usually sufficient to take only the underlying cortex of the affected bone with some of the medullary cavity, making it possible to reconstruct the shaft with an autograft or, more frequently, a hemicortical frozen cadaveric allograft. Occasionally, the lesion recurs after such surgery and requires either another attempt at removal (often successful) or an amputation. As indicated above, if the lesion is of a high grade or is grossly destructive of the cortex and invades the medullary cavity to a great degree, adjuvant chemotherapy should be used. 14 The prognosis even for these cases appears to be somewhat better than for a classic central osteosarcoma.

Both parosteal and periosteal osteosarcomas are uncommon lesions that are designated as surface osteosarcomas because they arise on the cortex of a bone, in this case most often the tibia. 65 In contrast with parosteal osteosarcoma’s pattern of low-grade spindle cells forming bone, the periosteal lesion is usually of a higher grade and more often primarily cartilaginous in structure. It occurs in a younger age group than the parosteal lesions and slightly more frequently in females. The lesion most often presents as a calcified or ossified mass on the anterior aspect of the proximal tibial diaphysis, which is firm and tender.

Histologically, the tumor may be confused with a fully malignant chondrosarcoma. If the specimen is examined carefully, however, malignant osteoid being produced by the malignant spindle cells can be seen. Because the periosteal osteosarcoma is more likely to metastasize than the parosteal osteosarcoma, some oncologists prefer to treat this lesion in the same manner as a central osteosarcoma. The likelihood of “cure” under these conditions is fairly good. If the lesion is small and is easily resectable, this may be sufficient treatment, unless the lesion is histologically of very high grade. Because of the site of the tumor and its proximity to the cortex, the resection should include a segment of bone adjacent to the tumor mass and, if like the parosteal osteosarcoma, is sometimes best treated by an allograft implant. Of some concern in these cases is the soft tissue coverage, which may require a gastrocnemius flap and skin graft.

Multi-focal Sclerosing Osteosarcoma

Multi-focal sclerosing osteosarcoma is an extremely rare variety of osteosarcoma. It is a fully malignant osteosarcoma that tends to present in one primary site, but on careful examination of the patient, it is evident that there are multiple osteosarcomas throughout most of the skeletal system (Fig. 121.15). 61, 62

Figure 121.15. Multi-local sclerosing osteogenic sarcoma in a 10-year-old female.

Figure 121.15

Multi-local sclerosing osteogenic sarcoma in a 10-year-old female. This figure demonstrates both the right and left lower extremities. Note the multiple sclerotic lesions appearing in the metaphyseal as well as diaphyseal areas of all four bones. Other (more...)

Multi-focal sclerosing osteosarcoma tends to occur in children < 10 years of age. It represents less than 3% of all primary osteosarcomas. Because of the presentation with multiple bone lesions, it has been postulated that the etiology of this disease could be a blood-borne infectious agent; no proof of this hypothesis has been provided. Because of the rarity of this tumor and its rapid 100% mortality rate, it has been impossible to study the tumor in further detail. The occurrence of this highly malignant form of osteosarcoma might indeed be some “experiment of nature,” whose victims could give us some clue to the etiology of osteosarcoma if the disease were more common or the survival of the rare patient who develops this disease more prolonged.

The patient with multi-focal sclerosing osteosarcoma usually presents with a primary osteoblastic osteosarcoma that is quite typical on radiographic examination. The clinical history is similar to most other primary osteosarcomas, except for the fact that most patients are young children. Usually, the patient has a biopsy of the suspected osteosarcoma and is referred for a staging work-up. Other multiple lesions are often asymptomatic at that time. The bone scan may not give a clue to the multi-focal nature of this disease, since the disease attacks the growing ends of the bones symmetrically. One can be very suspicious of multi-focal osteosarcoma in the proper clinical setting, if the bone scans show extremely hot metaphyseal ends of all bones. However, younger children tend to have bone scans that show increased uptake in the ends of all bones, since they are growing rapidly. Thus, in this rare entity, the bone scan may be of little use. One should be suspicious of multi-focal osteosarcoma in patients who have extremely elevated levels of serum alkaline phosphatase. The finding of multiple sclerotic areas in the lesional area of the primary site might give a clue to the nature of the disease, since the primary tumor sometimes is made up of many discrete round foci of sclerotic lesions that seem to coalesce into one big area. Sometimes the inclusion of other bones in the routine radiograph of the primary lesion will show additional lesions.

In the case of multi-focal sclerosing osteosarcoma, a skeletal survey is helpful in defining the entity and demonstrating the multiple lesions (see Fig. 121.15). Since all the bones in the body are involved, surgery is usually futile. Patients can go into remission on aggressive chemotherapy. Sometimes, it is necessary to perform surgery on the primary tumor because of uncontrollable pain. Patients should be managed with chemotherapy and pain medication. However, even with the most aggressive chemotherapy, multiple tumors always eventually progress, leading to death, usually within 1 year of diagnosis.

Osteosarcoma of the Jaw and Skull

Osteosarcoma of the craniofacial bones represents a clinical syndrome somewhat different from the classic osteosarcoma. Histologically, these tumors are identical to the classic variety of osteosarcoma. They tend to occur in an older age group (20–40 years) and, because of their location, are more difficult to resect surgically.

Osteosarcomas of the craniofacial bones are relatively rare and represent less than 10% of all osteosarcomas. The male-to-female ratio is close to one. The most commonly affected bones of the head and neck are the mandible, followed by the maxilla. Osteosarcomas of the other skull bones are extremely rare. The radiographic appearance of osteosarcomas in the head and neck is characteristic of bone-producing osteosarcomas in general. The disease carries a mortality similar to classic osteosarcoma. Because of its location, it tends to be diagnosed earlier, and the tumor can remain localized for a longer period than the classic fully malignant osteosarcoma of adolescents. Its location precludes very radical surgery, and there is a high incidence of local recurrences following primary surgery. Osteosarcoma of the skull, other than the mandible and maxilla, seems to carry with it an even worse prognosis due to the difficulty in radically resecting the tumor. Multiple local recurrences often occur prior to the appearance of distant metastases.

The differential diagnoses of a small jaw lesion include various forms of tooth abscess, osteomyelitis of the jaw, and some of the rare benign tumors, such as ossifying fibromas and brown tumors of hyperparathyroidism. In patients older than 50 years, Paget’s disease may produce a jaw lesion that mimics early osteosarcoma. In patients with known Paget’s disease of the skull, Paget’s sarcoma may arise.

The clinical history of the more common tumors of the skull include multiple local recurrences following inadequate surgery and the eventual metastasis of disease to the lungs and to other bones similar to the classic variety of osteosarcoma.

Treatment for a fully malignant osteosarcoma occurring in the skull bones is similar to that for the classic osteosarcoma. It should include preoperative chemotherapy, which in this setting takes on additional importance because of the limited margins for radical surgical excision. Osteosarcoma of either the mandible or maxilla seems to have a better prognosis than lesions of the skull bones and indeed better than the more standard types of classic medullary osteosarcoma. Some believe this is because of the earlier discovery of the lesions because they are painful and deforming early in the course; but others suggest that these lesions are lesser in malignancy and slower to metastasize. In any event, the treatment by hemimandibulectomy or segmental resection of the maxilla is successful in eliminating the primary tumor in most cases. Reconstruction of the defect is technically demanding but possible, and most patients achieve a satisfactory result from such surgery. High-dose chemotherapy is advocated for these patients in an effort to prevent metastatic lesions and local recurrence. Lesions of other bones of the skull are frequently considered to have a worse prognosis than those of the mandible and maxilla. This may be associated with the difficulties of gaining a wide margin at surgery and the frequency of the local recurrences. In our experience, the only survivors of osteosarcoma of the parietal bone or occipital bones have been patients who have had extremely wide surgical excisions and have escaped local recurrences. These patients have also undergone preoperative and postoperative chemotherapy. 63

Osteosarcoma Arising in Paget’s Disease

Osteosarcoma arising in patients with Paget’s disease has a very poor prognosis. The lesions are large at the time of their appearance and very destructive, making resective surgery difficult, and pulmonary metastases are often present at the outset. The lesions almost always arise in patients with polyostotic Paget’s disease and, in view of the frequency of Paget’s disease in the general population (estimated for monostotic or polyostotic disease at 3% of the population over the age of 60 years), are quite rare. The tumors have a predilection for the pelvis, proximal femur, and proximal humerus and are difficult to treat surgically, principally because of the patient’s age and the size of the lesion. Amputative surgery is sometimes necessary, especially in the face of a pathologic fracture, which is a frequent occurrence.

The pathology of osteosarcoma arising in Paget’s disease is usually that of a fully malignant osteosarcoma. 64, 65 Mortality associated with Paget’s sarcomas is extremely high due to their location, which often precludes radical removal; the fact that patients with polyostotic Paget’s disease sometimes develop multiple synchronous primary osteosarcomas; and the fact that older patients have not, in the past, been treated with aggressive chemotherapy.

The clinical history usually includes increased pain and swelling in the lesional area. Sometimes, it is known that the patient has Paget’s disease, but frequently radiography reveals for the first time a suspected tumor that looks like it is arising in a Pagetoid bone. The serum alkaline phosphatase level is almost always elevated in patients with Paget’s disease to begin with, and the superimposition of osteosarcoma may serve to elevate the alkaline phosphatase level even more. Because of this, alkaline phosphatase is not valuable as a tumor marker. In addition, in patients with polyostotic Paget’s disease it is sometimes impossible to tell whether or not the patient has metastatic bone lesions with osteosarcoma or simply polyostotic Paget’s disease, with one primary osteosarcoma.

Patients with confirmed osteosarcoma arising in Paget’s disease should have a bone scan and radiographic examination of all identified lesions to determine whether changes are due to osteosarcoma or multi-focal Paget’s disease. A CT scan of the chest should also be done to rule out pulmonary metastases. Osteosarcoma arising in Paget’s disease should be treated according to identical principles as those for fully malignant osteosarcoma arising in a younger population. Great care must be taken during the administration of chemotherapy, since many older patients have impaired renal function. Patients who have osteosarcomas arising in Pagetoid lesions appear to have an improved prognosis through the use of preoperative chemotherapy, radical surgical excision of the tumor, and continued postoperative chemotherapy. Tumors of the axial skeleton, including the skull and spine, are sometimes inoperable. In this circumstance, patients should be given chemotherapy and radiation therapy to try to control the primary lesion and delay the appearance of metastatic disease. The patient depicted in Figure 121.16 with osteosarcoma arising in Paget’s disease of the ilium was successfully able to undergo preoperative chemotherapy, en bloc resection of the ilium (internal hemipelvectomy) and postoperative adjuvant chemotherapy. She is surviving free of disease more than 3 years after treatment.

Figure 121.16. Paget’s sarcoma occurring in the pelvis of a 60-year-old female.

Figure 121.16

Paget’s sarcoma occurring in the pelvis of a 60-year-old female. The radiographic appearance of the right ilium is characteristic of advanced pagetoid changes. The benignity of the Paget’s disease can be seen by the lack of involvement (more...)

Postirradiation Osteosarcomas

Osteosarcomas arising in irradiated bones represent fully malignant osteosarcomas. They arise in the irradiated area following therapeutic radiation therapy for other diseases.

The median time to development of a postirradiation osteosarcoma is 10 years. 66 We have observed postirradiation osteosarcomas as early as 4 years following exposure. Because of the genetic propensity to develop osteogenic sarcoma in patients with familial retinoblastoma due to reduction to homozygosity of their Rb genes, postirradiation sarcomas are more common. The younger the age at the time the patient receives radiation therapy, particularly in infancy, the higher is the incidence of osteosarcoma. In the general population of individuals cured of cancer by radiation therapy, postirradiation sarcomas are rare and develop in less than 5%. Whether or not these patients have some genetic predisposition to developing a second cancer similar to that which has been demonstrated for retinoblastoma is unknown. Evidence suggests that osteosarcomas occur more frequently in patients who receive higher doses of radiation therapy.

The differential diagnosis of postirradiation osteosarcoma is usually that from recurrence of the original irradiated primary tumor. Postirradiation osteosarcomas usually appear on radiographs as osteoblastic or mostly lytic lesions arising in the irradiated site. Some lesions appear to be multi-focal osteosarcomas because of separate satellite areas arising within the irradiated site. On histologic examination, in addition to a fully malignant spindle cell tumor producing osteoid, postirradiated osteosarcomas may also contain areas of radiation osteitis in the non-neoplastic surrounding bone.

The clinical presentation of secondary postirradiation osteosarcoma is one of painful swelling in the irradiated area. After establishing the diagnosis of a postirradiated osteosarcoma by open biopsy, staging is similar to that for other osteosarcomas and should include a bone scan, CT scan, or MRI of the lesional area and a CT scan of the lungs to rule out the presence of metastases.

Postirradiation sarcomas are treated in a manner similar to other fully malignant osteosarcomas, with preoperative chemotherapy, radical surgical excision of the primary tumor site, when possible, and postoperative adjuvant chemotherapy. The prognosis for postirradiation sarcomas is worse than that for classic osteosarcomas because most are in the axial skeleton following corrective radiotherapy for lymphomas. Osteosarcomas may occur in the sternum, vertebral bodies, shoulder, and bones of the hip. If treatment is to be successful, surgery must be radical.

Radical excision of axial lesions is difficult to perform and almost impossible without the use of preoperative chemotherapy. Preoperative chemotherapy may shrink the primary lesional area and make radical surgery possible. Tumors of the sternum need to be resected en bloc with the adjacent sternoclavicular joints, which otherwise would represent areas destined for local recurrence. The majority of osteosarcomas of the sternum tend to recur in the ribs and to produce malignant pleural effusions. The reconstruction required following radical sternal excision is difficult but may be achieved by implantation of methylmethacrylate marlex mesh inserts or transplantation of autograft segments from the iliac wings. We give a protracted course of preoperative chemotherapy and do not perform surgery until good response to preoperative chemotherapy is established. Since chemotherapy in the immediate postoperative period may be precluded by surgical complications, a protracted course of preoperative chemotherapy seems prudent, if the patient is responding well to treatment, prior to subjecting the patient to an attempt at curative surgery.

Fibrous Lesions of Bone

Giant Cell Tumor

The giant cell tumor of bone, also called osteoclastoma, is a benign but aggressive neoplasm representing approximately 5% of all primary bone tumors. 68 The peak incidence of this tumor is in the third decade of life. The tumor is extremely rare in prepubescent adolescents, and the diagnosis of giant cell tumor in this age group should raise suspicion of misdiagnosis. The tumor is slightly more common in females. Giant cell tumors are typically epiphyseo-metaphyseal tumors, with the majority involving the distal femur and proximal tibia. Giant cell tumor can arise in any site in the skeleton, however. Giant cell tumors of the spine or pelvis can cause significant morbidity, due to the difficulty in resecting them and the large amount of bleeding that occurs at the time of biopsy due to the vascularity of the tumor.

The radiograph of giant cell tumor is quite typical and includes a well-demarcated lesion that appears to arise in the epiphysis, usually of the distal femur or proximal tibia. The tumor can produce considerable swelling due to expansion of the bony cortex, but this tumor seldom breaks through the cortex to produce an extraosseous soft tissue mass (Fig. 121.17).

Figure 121.17. A.

Figure 121.17

A. A classic giant cell tumor of the distal femur, purely lytic and thinning the cortex. It is eccentrically located and, as seen on the MRI (B), has broken into the joint. C. The lesion was curetted, the cavity washed with phenol, and after reconstruction (more...)

Histologically, giant cell tumors are composed of a benign spindle cell stroma with numerous multi-nucleated giant cells. Characteristically, the nuclei of the benign spindle cells are identical in appearance to the nuclei in the multi-nucleated giant cells. Frequently, there is a component of aneurysmal bone cyst associated with the giant cell tumor.

The differential diagnosis of a giant cell tumor includes other rare benign conditions, such as reparative granuloma of bone, brown tumors associated with hyperparathyroidism, and fibrous dysplasia. Problems in differential diagnosis arise when a giant cell tumor has extended outside the bone or when areas in the tumor show a sparsity of giant cells with a predominance of a spindle cell stroma. In such instances, it may be difficult to differentiate benign giant cell tumor from the so-called “malignant giant cell tumor of bone,” which is, in fact, either a telangiectatic osteosarcoma or a malignant fibrous histiocytoma.

The clinical history of a giant cell tumor usually includes pain and limitation of motion because of the tumor’s proximity to the joint space. Swelling occurs late in the course of the giant cell tumor. Pathologic fracture may bring the previously asymptomatic patient to the surgeon, who then makes the diagnosis and is faced with the necessity of caring for what clearly can be a difficult clinical situation. It is best under these circumstances to immobilize the limb and wait until the fracture heals before performing resective or intralesional surgery. Giant cell tumors may grow inordinately rapidly during pregnancy; for the vast majority of the patients so afflicted, protection until the time of viability of the fetus is logical, and then early induction followed by appropriate surgery seems to be the best cure.

Because benign giant cell tumors have been noted to metastasize, patients should be staged with a CT scan of the chest in addition to radiography of the lesional area and a CT scan to help determine the extent of the local disease. Giant cell tumors are locally aggressive (although benign tumors), and treatment once entailed little more than just curettage and packing with bone chips. With the latter treatment, however, local recurrence occurs in half the cases. To prevent local recurrences, more aggressive local surgery is usually necessary. This may consist of thorough curettage and cryosurgery or burring of the cavity with or without phenol installation. The lesional area may then be packed with allograft or, as has recently been suggested, polymethyl methacrylate. Such treatment reduces the local recurrence rate to less than 20%. Resection of the lesion and autograft arthrodesis, allograft, or metallic implant have all been advocated in the past but now seem excessive, except when the deformity of the joint is so extreme as to require some sort of arthroplastic procedure to allow competent function. Lesions of bone sites that may be resected with minimal disability, such as the ilium or fibula, should be removed by surgical extirpation of the site. The highest rate of local recurrence is in the bones of the hand and distal radius (which also have the highest rate of metastasis), and these should also be considered for surgical excision of the lesion and appropriate replacement by autograft (i.e., replacement of a metacarpal by a metatarsal or a distal radius by a proximal fibula) or allograft.

The prognosis for giant cell tumor is excellent if it does not recur locally. Local recurrences are usually treated with even more aggressive surgery and in rare instances have led to amputation.

Radiation therapy has been used for the treatment of giant cell tumor of bone since 1906. 68, 69 During the early decades of this century, the limitations of surgery led many notable authors to conclude that radiation therapy was the treatment of choice for this benign but locally aggressive epiphyseal tumor arising in young skeletally mature patients. 71– 82 These investigators generally delivered from one to several courses of 2,000 cGy in 10 days. The time period between irradiation courses was often many weeks, and accurate imaging studies were unavailable for portal planning. Nevertheless, only rare local failures were reported even after follow-up as long as 25 years.

Despite its historic success, radiation therapy has been largely abandoned for giant cell tumors. Several authors state that this lesion is radioresistant because of frequent tumor progression following primary or postoperative radiotherapy. Dahlin reported recurrence in 47% receiving irradiation following simple curettage, compared with 42% following curettage alone. 69 The failure of elective irradiation to improve control rates after incomplete excision has been noted elsewhere. 70, 83 Tumor progression after primary radiotherapy was reported in 7 of 10 patients by Goldenberg, 83 6 of 7 by Dahlin, 84 and 6 of 12 by McGrath 70 reported progression in 13 of 16 patients receiving radiotherapy in any setting. These authors, however, typically fail to report radiotherapy details, including total dose or the selection criteria for choosing radiotherapy.

Modern megavoltage irradiation remains a viable therapeutic modality for the giant cell tumor. Bennett, 85 reporting the University of Florida experience and reviewing the recent world literature, noted an overall local control of 77% following radiotherapy in 97 cases. The majority of these patients were irradiated primarily for gross tumor. Radiotherapy has been reported successful for giant cell tumor of the skull, 70 spine, 86– 88 nonepiphyseal long bone locations, and metastatic deposits. 70

Radiotherapy is best reserved for those giant cell tumors not amenable to modern resection or curettage with aggressive chemical installation. Radiation therapy should be avoided for the “giant cell” lesion of the jaw. This entity, while histologically resembling the osteoclastoma, actually represents a far more indolent giant cell condition properly termed giant cell reparative granuloma. 89, 90 Recurrence of this rare lesion is unusual following even simple curettage. An extragnathic form, often termed “solid aneursymal bone cyst,” has been reported. 91 Recurrence in these locations is also unusual after curettage except for the small tubular bones of the hand and foot. 92, 93

Total dose determines success of radiotherapy. Chen 94 reported local progression in 3 of 17 primarily irradiated patients receiving > 3,500 cGy, compared with 5 of 8 receiving < 3,000 cGy. 95 Bennett 85 reported no local control differences over the range 3,500 to 5,500 cGy. Virtually all authors now recommend 4,500 to 5,000 cGy. There is no longer any role for multiple courses of low-dose radiotherapy. Chen reported control following a single protracted course of irradiation in 11 of 15 primarily treated patients, compared with 5 of 10 patients treated in the historic fashion. Particular attention must be directed to any soft tissue extension of giant cell tumor during portal design. Bennett recommended 3- to 5-cm margins around the entire lesion.

Increased tumor rarefaction is commonly noted immediately following radiotherapy. 85, 94 This radiographic exacerbation must not be ascribed to true tumor progression. Increased bony lysis noted after radiotherapy is followed by slow radiographic resolution of tumor. Maximum osseous healing requires 12 to 24 months. Bennett described complete radiographic resolution after 9 years. Osseous reconstitution, however, may never be complete. 90 Healing is said to occur more rapidly in the young patient. 68 Pain responds quickly to radiotherapy, prior to conclusion of treatment.

The impact of various tumor/patient characteristics on the results of radiotherapy remains uncertain. Bennett reported 85% control of giant cell tumors in membranous bone, compared with 70% in long bones (p = .11). Control was reported in 75% of females and 86% of males (p = .24). Although soft tissue invasion has been noted to adversely affect surgical control of giant cell tumors, this does not appear to be the case with radiotherapy. 97 Bennett reported control in 3 of 5 tumors confined to the site of origin, compared with 9 of 11 for those with soft tissue extension in the Florida experience. Patient age was not associated with the outcome in the series of Bradshaw. 84 The control rate by radiotherapy of giant cell tumors of the limbs equals that of the spinal column. 84, 93 Vascular space invasion, tumor ploidy, symptom duration, and histopathologic grade have not been found to reliably predict clinical course in surgical series of giant cell tumors. 95, 96 These factors have not been evaluated in radiotherapy series.

Recurrent giant cell tumors may demonstrate malignant osteoclastoma or frank sarcomatous change to fibrosarcoma or osteosarcoma. Many investigators believe that prior exposure to irradiation accounts for the majority of these worrisome recurrences. 70 Although malignant transformation or sarcoma induction has admittedly been associated with ionizing radiation and giant cell tumors, this outcome has also occurred without prior exposure. 87, 98 Moreover, the vast majority of malignant recurrences after radiotherapy have been reported from the kilovoltage era, often after quite excessive total doses. In the review of the megavoltage era by Bennett, there was but a single case of post-treatment sarcoma in 97 irradiated patients, 24 of whom were monitored for at least 10 years. Giant cell tumors associated with Paget’s disease are reportedly sensitive to irradiation. 87, 99 This specific giant cell lesion, however, may represent a reparative granuloma. 71 Malignant transformation has also been reported for giant cell lesions arising in Paget’s disease. 72, 73 For these reasons, it is prudent to avoid irradiating this giant cell entity.

Benign giant cell tumors have been noted to metastasize to the lungs. In the case of a solitary pulmonary metastasis, surgical resection of the pulmonary metastasis leads to a cure in most instances. Rarely does a patient develop multiple unresectable pulmonary metastases. This usually occurs in patients who have undergone multiple inadequate surgeries for the primary tumor. A trial of chemotherapy may be warranted, but because the tumors are slow growing, chemotherapeutic effect may be difficult to evaluate.

Giant cell tumor of bone is a locally aggressive tumor that is managed by surgery. Because it arises immediately adjacent to the articular cartilage, a resection requires removal of the articular surface. When the adjacent articular surface is unnecessary (e.g., proximal fibula), a resection is recommended. Otherwise, curettage is better. A simple curettage is insufficient and is followed by an unacceptable incidence of local recurrence (approximately 50%). A more aggressive curettage is associated with a lower incidence of recurrence, even though a recurrence rate of approximately 20% is considered acceptable. The simplest method for doing an “aggressive” curettage is to use a large window in the bone and burring the cavity with a high-speed burr. Phenol has been used to kill any residual cells and polymethyl methacrylate. Others have used cryosurgery, which reduces the incidence of recurrence lower than other methods but has its own potential complications. Most local recurrences can be treated with a second curettage with good results. There is no evidence that local recurrence leads to an increased risk in metastasis.

Adamantinoma of Bone

Adamantinoma is a rare low-grade malignant tumor that usually arises in the tibia in young male adults, although it also occurs less frequently in females and, very rarely, in the femur, humerus, and ulna. The tumor represents less than 1% of primary bone tumors (see Table 121.4). The radiographic appearance of the adamantinoma is quite characteristic and usually presents with a mixed lytic and sclerotic lesion eccentrically placed in the tibia (Fig. 121.18).

Table 121.4. Malignant Fibrous Bone Tumors.

Table 121.4

Malignant Fibrous Bone Tumors.

Figure 121.18. An adamantinoma of the tibia showing the extent of the lesion and the classic modest anterior bow, often associated with this rare tumor.

Figure 121.18

An adamantinoma of the tibia showing the extent of the lesion and the classic modest anterior bow, often associated with this rare tumor.

Histologically, adamantinoma appears to have a low-grade malignant fibrous stroma with epithelial nests of cells that may resemble basal cell carcinoma or carcinoma of the breast scattered throughout the fibrous stroma.

Although not perplexing when typical radiographic and histologic findings are present, the absence of a typical epithelial-like component raises the differential diagnosis of other benign lesions, such as fibrous dysplasia, and of ossifying fibroma.

Since adamantinoma metastasizes to the lungs, late in the course, staging should include a CT scan of the chest as well as adequate imaging studies of the primary lesional area and careful follow-up for many years.

Because adamantinoma is a low-grade malignant tumor, surgical treatment should be more radical than for benign bone lesions. Wide excision of the lesional area is indicated. The procedure lends itself well to allograft replacement of an intercalary segment. If this procedure does not produce adequate negative soft tissue and bony margins, then amputation may be indicated. With adequate aggressive surgical treatment of the primary lesional area, the prognosis for cure is excellent. However, patients should be monitored with periodic examinations and chest radiographs. In the rare instance when this tumor has metastasized, a trial of chemotherapy may be warranted. Because of the rarity of this phenomenon, there are no data on the effects of various chemotherapeutic modalities on metastatic adamantinoma. 74

The role of radiotherapy in the treatment of adamantinoma of long bone and the closely related ameloblastoma of the jaw remains uncertain. Radiotherapeutic experience to base recommendations on has been precluded by the rarity of these lesions as well as disagreement over their histogenesis, nosology, and malignant potential.

The ameloblastoma is a neoplasm of odontogenic epithelium with a fibrous background typically arising from tissue resembling the embryonic enamel organ (e.g., ameloblasts). 75, 76 Other proposed origins include the basal layer of the dental lamina and the lining of dentiginous cysts. 77 The origin of the extragnathic form of this tumor is greatly debated. Although resembling the expansile lesion of the mandible or maxilla, most authors consider the long bone lesion to be of other than odontogenic origin, hence the rejection of the term ameloblastoma in favor of adamantinoma. The long bone tumor has variously been considered of synovial, vascular, and epithelial origin. 78 Use of the term “malignant angioblastoma” attests to possible vascular derivation of this tumor. 79 Recent analysis suggests that adamantinoma displays both mesenchymal and epithelial differentiation. 80 There is frequent association of adamantinoma with adjacent fibrous dysplasia or ossifying fibroma (osteofibrous dysplasia). 81 Those tumors with few neoplastic cells and a prominent component of osteofibrous dysplasia may represent a unique form of adamantinoma. The osteofibrous dysplasia may represent a secondary phenomenon associated with spontaneously regressing adamantinoma. 82 Moon 100 has strongly advocated separating ameloblastoma and adamantinoma, since there remains no histologic proof that these two entities are identical. Whatever the tissue origin, both ameloblastoma and adamantinoma are misnomers. 75, 101 Ameloblasts are not invariably present in the jaw lesion, and neither tumor produces enamel or adamantine substance. Both lesions bear a striking resemblance to basal cell carcinoma. 102– 104 A rare mucosal form of ameloblastoma arising on the gingival surface has been termed “basal cell carcinoma of the oral cavity” or peripheral ameloblastoma.

These bone tumors are histologically benign. They are locally destructive, typified by extensive invasion of cancellous bone and occasional cortical disruption with soft tissue extension. Cortical disruption appears more common with adamantinoma than with ameloblastoma. 105 Metastases are reported in 5% of ameloblastomas. According to Weiss, 105 metastases occur in up to 15% of adamantinomas. Local relapses and soft tissue extension may predispose to metastases. The median disease-free interval prior to metastatic ameloblastoma is 9 years. The average interval prior to modal metastatic adamantinoma is 6 years and to pulmonary disease is 8 years. 97 Ameloblastic carcinoma and ameloblastic sarcoma have been described. 97

An enormous number of histologic variants of ameloblastoma/ adamantinoma have been described: follicular, pigmented, acanthomatous, plexiform, granular cell, stellate, basaloid, ameloblastic fibroma, and adenoameloblastoma. 106 The histologic pattern of these tumors, however, does not reliably predict their clinical course or serve as a guide to therapeutic decision making. Morphologic subtypes described include unicystic, multi-cystic, solid, and peripheral. 107 Several investigators state that the unicystic and peripheral tumors are well circumscribed and amenable to conservative surgery, while the solid and multi-cystic forms are poorly circumscribed and may require more radical procedures or adjunctive therapies. 108

The ameloblastoma/adamantinoma are commonly labelled “radioresistant,” and consequently, radiotherapy is not considered to have a role in their management. This conclusion is based on poorly documented experiences primarily from the kilovoltage era. 76, 97, 109, 110 The historic failure of radiotherapy to control or induce response in the majority of cases has been properly ascribed to a variety of factors that are of less importance in the modern era: physical dose limitations of low-energy irradiation, primitive tumor imaging resulting in inadequate portal design, and failure to appreciate the naturally slow response time of these lesions. 111

An analysis of documented radiotherapy experience demonstrates that ameloblastoma/adamantinoma are not inherently radioresistant. 112 Response to primary radiotherapy was reported in 8 of 11 patients by Sehder, 113 6 of 7 by Atkinson, 114 and 4 of 5 by Gardner. 111 There are other anecdotal reports of tumor regression following modern radiotherapy. 101, 104, 115 Full expression of response may require several years, indicating the necessity for extended follow-up. Two years were required for complete healing of an oral ulcer and bone reconstitution following radiotherapy of ameloblastoma, according to Hair. 116 Atkinson reported continued tumor response for 10 years following irradiation of an incompletely resected ameloblastoma. 114 Complete radiographic resolution of adamantinoma of the humerus following irradiation may require 2.5 years. Palliative effects are commonly noted even with minimal clinical or radiographic tumor response. Although Ikemura 117 advised abandonment of radiotherapy for treatment of metastatic disease, gratifying response of metastatic deposits has been described. 102

Permanent tumor control is unlikely following radiotherapy alone. Persistent disease may remain quiescent for many years after treatment. All 8 responses reported by Sehder were partial, but eventual tumor progression was delayed for 2 to 15 years. Others have reported partial response of ameloblastoma and adamantinoma that were maintained following radiotherapy for 3 to 11 years before ultimate relapse. 108, 114 Local recurrence has been reported after 15 years of disease-free survival, implying that even complete response is no assurance of permanent tumor eradication. 101

Primary radiotherapy should be reserved for inoperable patients or those with unresectable disease. Postoperative radiotherapy may be indicated for those patients with less than optimal surgical clearance of tumor. Patients free of disease for 3 to 20 years have been reported following curettage or partial excision and radiotherapy. 104, 113 Local relapse as long as 26 years after incomplete surgery and irradiation implies that prolonged follow-up is also required for patients irradiated adjunctively. 113 Pandya 115 reported responses in two patients irradiated with preoperative intent.

The response of ameloblastoma and adamantinoma to radiotherapy is unpredictable. Tumor responses have been reported after a variety of radiotherapy doses and fractionation schemes. Atkinson 114 reported disease-free survival for 6 months to 10 years in 5 of 8 patients with jaw lesions receiving between 30 Gy in 6 fractions and 47 Gy in 20 fractions as primary treatment. Gardner 111 reported disease-free survival in three patients receiving 40 to 55 Gy. Eight-year disease-free survivals were reported following 54 Gy for ameloblastoma 117 and 55 Gy for adamantinoma. 116 Total resolution of pain due to osseous metastases has been noted after 30 and 39.6 Gy. Laughlin 103 reported complete response of metastatic chest wall disease after 27 Gy. Although some investigators have employed 60 to 65 Gy, higher doses do not appear to yield superior results, either primarily or adjunctively. Most authors now recommend 45 to 50 Gy delivered over 4 to 5 weeks. 116 Proper field size remains undefined. The propensity for these lesions to invade widely through the medullary cavity must be recognized in portal design. Particular attention must be directed to portal design for the maxillary ameloblastoma. A 2-cm margin around clinically evident tumor has been proposed. 117 That recommendation appears logical since the surgical literature advises resection margins of 0.5 to 1.0 cm. 93, 113– 117

Malignant Fibrous Histiocytoma and Fibrosarcoma of Bone

These malignant bone tumors represent similar, and probably identical, entities. The terminology “malignant fibrous histiocytoma” (MFH) has been used more frequently since this malignant primary bone tumor was extensively studied by Huvos 118 in the mid-1970s. Malignant fibrosarcomas and MFHs of bone are composed of a malignant spindle cell stroma that, in many instances, produce abundant collagen. Multi-nucleated giant cells are frequently found within this tumor that presumably represent malignant histiocytes in often bizarre forms. These tumors are rare, occur with equal frequency in males and females, and do not occur in any characteristic age group. Many people feel that the MFH of bone represents the response of the tissue to chronic injury and hence appear with great frequency at the site of radiation, at bone infarcts, in Pagetoid foci, or even as dedifferentiated chondrosarcoma (see below).

The radiographic appearance of an MFH or malignant fibrosarcoma of bone is predominantly a lytic lesion. The distribution of skeletal sites of these tumors is similar to that of osteosarcoma, with a majority of lesions occurring in the femur, tibia, and pelvis, although any bone in the skeleton can be affected.

MFH and fibrosarcoma can occur as low-grade as well as high-grade malignant lesions. They may also occur as periosteal lesions similar to parosteal osteosarcoma; indeed, parosteal MFHs are of low-grade malignancy. MFH tend to have abundant giant cells present as well. The fibrous stroma of the MFH tends to produce a circular or storiform appearance, whereas that of a malignant fibrosarcoma tends to produce a linear or herringbone pattern of interweaving fascicles of malignant fibrous cells, although most often they cannot be distinguished from one another.

The main problem in differential diagnosis is from osteosarcoma. Indeed, many of these tumors that occur in the younger age group can be found to produce osteoid. Malignant fibrous tumors that produce small amounts of osteoid in some areas are classified as osteosarcomas by many pathologists; other pathologists may allow osteoid to be produced in small amounts and still classify the tumor as a fibrosarcoma or MFH. The histogenetic name tag given to the tumor is primarily of taxonomic interest; since it is a fully malignant spindle cell neoplasm of bone, it is treated in a way similar to osteosarcoma and carries with it a similar poor prognosis without adequate treatment.

The clinical history is similar to that for an osteosarcoma: pain and swelling in an area, which usually leads to a radiograph that demonstrates a malignant-appearing bone lesion. Because of the lytic nature of this tumor, patients may sometimes present with a pathologic fracture (Fig. 121.19). In this instance, as in any pathologically fractured malignant tumor, the prognosis for performing limb salvage surgery and indeed for ultimate survival appears to be worse.

Figure 121.19. Pathologic fracture in a patient with a malignant fibrous histiocytoma of bone.

Figure 121.19

Pathologic fracture in a patient with a malignant fibrous histiocytoma of bone. This lesion is usually a lytic lesion, and therefore when it occurs in the weight-bearing lower extremities, the presentation with a pathologic fracture is not uncommon.

Staging work-up of this malignant tumor is similar to that for osteosarcoma and includes imaging studies of the lesional area with CT and MRI, a bone scan, and CT scan of the chest.

Treatment is identical to that given for osteosarcoma. Low-grade lesions need only undergo radical en bloc surgical excision, but fully malignant lesions require the use of preoperative and postoperative chemotherapy. In our experience, MFH of bone has a slightly higher complete response rate to preoperative chemotherapy than does the classic variety of osteosarcoma. 119 Indeed, the prognosis for fully malignant fibrous lesions of bone, as in osteosarcoma, seems to be best predicted from the patient’s response to preoperative chemotherapy.

Available data do not permit a conclusion regarding the efficacy of radiotherapy for this unusual primary bone tumor. 120 Persistent or locally recurrent MFH has been described after 60 to 80 cGy. 122 Morphologically identifiable tumor has been reported after preoperative doses of 26 to 55.8 Gy. 123 Recent series demonstrate that the MFH of bone, like osteosarcoma, responds to modern preoperative chemotherapy. 124, 125 Nonrandom use of various drug regimens has resulted in an improvement in survival rate, compared with historical experience with surgery alone. 126 Specific chemotherapeutic agents to be given as postoperative adjuvants may depend on the degree of histopathologic response to neoadjuvant chemotherapy. Radiotherapy, therefore, should be judiciously employed in selected situations so as not to interfere with the analysis of resected specimens.

Osseous MFH that is not amenable to neoadjuvant chemotherapy or to surgery may respond to radiotherapy. Campanna et al. 126 primarily irradiated 8 patients to doses of 40 to 80 Gy. All four with long bone tumors were locally disease free from 3 to 8 years. Two patients with axial lesions demonstrated “probable” local relapse at 14 and 23 months. Two others with axial tumors succumbed rapidly to metastatic disease. Dahlin and associates 127 also reported significant results of radiotherapy, although details of treatment were not provided. Of those primarily irradiated, 1 patient was palliated for 6 years, 2 were cured, and 1 relapsed locally. One other patient was disease-free 18 years after excision and radiotherapy. Spanier et al. 128 described a patient receiving radiotherapy for lung metastases and a stump recurrence who was free of tumor in all irradiation sites at necropsy 32 months later. The authors hypothesized that an MFH with an abundance of histiocytes was radiosensitive, while a predominantly fibroblastic tumor was relatively radioresistant. Marked pain reduction after irradiating vertebral MFH has been described. 129

Postoperative radiotherapy may be indicated for MFH, should margins be less than radical. Campanna reported local relapse with wide margins in 17% of resected patients and 21% of amputated patients, compared with 6.5% with a radical margin. Adjuvant chemotherapy did not influence local relapse rates regardless of margin status in their series. Neoadjuvant chemotherapy followed by limb-sparing surgery with adequate margins is also associated with a low risk of local relapse in the absence of routine radiotherapy. 130

The majority of patients with MFH present with high-grade, extracompartmental tumors. Neither the intraosseous nor the extraosseous components of MFH are microscopically well circumscribed. 126, 128– 130 Radiotherapy field margins must necessarily be generous to encompass areas of potential soft tissue and intramedullary extension. The role of regional lymphatic irradiation remains uncertain. Frequent lymphatic involvement was documented by Feldman, 129 but most series do not substantiate this finding. 123, 127

Cartilage Tumors of Bone

Lesions of bone that contain cartilage are among the most difficult to diagnose and treat. The principal reason for this is the fact that with the exception perhaps of the chondromyxoid fibroma and, with rare exceptions, the chondroblastoma, each of the lesions, particularly the hyaline tumors, have benign and malignant representatives. Furthermore, each of these may be graded from 0 (benign) to 4 (highly malignant), and the biologic behavior varies considerably depending partly on the grade but also on other factors. Thus, management depends on a set of “rules” that apply particularly to the hyaline lesion and that may be helpful in regard to these difficult tumors. The rules include the following: (1) Large tumors are more likely to be malignant than small ones. (2) Chondrosarcoma is rare under the age of 25 and increases in frequency as the years advance. (3) No matter how aggressive they appear on imaging studies, lesions of acral parts (hands and feet) are almost never malignant, while those of the axial skeleton, and more particularly the pelvis, should be viewed with suspicion, regardless of how benign they appear. (4) Lesions that occur in patients with hereditary multiple osteocartilaginous exostoses, Ollier’s enchondromatosis, or especially Maffucci’s syndrome are more likely to be malignant than those osteocartilaginous exostoses or enchondromas that occur as solitary lesions. (5) A cartilage lesion that is “cold” on scan is rarely malignant. One that is “hot” may be malignant or benign, and the test really does not discriminate. (6) The more aggressive a lesion is, the smaller will be the extent of the calcification, the more likely the cortex will be eroded, and the larger will be the soft tissue extension (for exostotic lesions). (7) Regardless of all of the above, a patient who has persistent pain over the lesional site, particularly at night or at rest, should be viewed as having a potentially malignant lesion.

Thus, a 5-year-old child with a painless small enchondroma of the fifth toe as a solitary lesion that is cold on scan surely has a benign enchondroma, while a 60-year-old woman with Maffucci’s syndrome who presents with a large painful mass in the ilium that is hot on scan almost certainly has a malignant one. Those are obvious; it is the ones in between that are difficult!

Benign Cartilage-Producing Bone Tumors


Osteochondroma is a benign cartilaginous lesion usually referred to as a benign osteocartilaginous exostosis. Osteochondromas are developmental defects probably caused by cartilaginous nests that are left behind in the advancing growing epiphyseal plate. There is a hereditary syndrome of multiple osteochondromata (or hereditary multiple cartilaginous exostoses).

Osteochondromas are the most common benign bone tumors seen. They represent approximately 50% of all benign bone tumors and approximately 15% of all primary skeletal lesions. As noted, they usually are discovered in growing adolescents when both the normal long bones as well as the lesion are still growing. They occur with equal frequency in males and females. The tumor is more common in the long bones and usually starts to grow in the metaphysis adjacent to the epiphyseal cartilage plate. As the normal bone continues to grow, the osteochondroma appears to be located farther away from the epiphyseal cartilage plate but strikingly sharp in the metaphysis or metaphyses diaphyseal junction. The majority of osteochondromas occur about the knee joint, with the femur a slightly more common site than the tibia. Osteochondromas have also been reported in the humerus and the short tubular bones of the distal extremities as well as the spine and pelvis.

The radiographic appearance is quite typical, and in general, biopsy is not necessary since the radiographic diagnosis of an osteochondroma is usually obvious to most skeletal radiologists (Fig. 121.20).

Figure 121.20. A.

Figure 121.20

A. Osteochondroma (osteocartilaginous exostosis) is a benign lesion that may be solitary of multiple and extends from the metaphysis of the long bones into the adjacent soft tissues. B. The lesion shares a cortex with the bone. C. Such a lesion arising (more...)

As noted in the discussion of parosteal lesions above, osteocartilaginous exostoses, whether solitary or in a patient with multiple lesions, are always seen to “share a cortex” with the bone rather than being engrafted tightly on it, like a surface osteosarcoma, or with a clear space between it and the bone, as occurs with a myositis ossificans. Histologically, the osteochondroma consists of a cartilage cap of hyaline cartilage, which is more cellular than articular cartilage and more distorted in appearance, but the cells are small and the nuclei uniformly dark on hematoxylin and eosin staining. In young children, the cap is a site of endochondral ossification leading to the mass of bone that forms the exostosy. As the individual ages, the cartilage becomes considerably thinner and often heavily calcified. One should view lesions in adults with thick cartilage caps seen on imaging studies with some degree of suspicion.

The clinical history of osteochondromata is that of a benign tumor. It usually manifests in the early growth years of adolescence and may be an incidental finding discovered after an unrelated trauma. It rarely produces symptoms of pain, and a characteristic smooth bump can often be felt near the distal ends of the affected bones. Uncommonly, pain and tenderness can arise from irritation in the nearby tendons or injury to nerves or blood vessels. Only in this latter instance may surgical treatment be indicated. Transformation into malignant chondrosarcoma is an extremely rare event, more common in the syndrome of multiple exostosis than in solitary osteochondroma but substantially less frequent than would justify cancer-preventive surgery. 131

The syndrome of hereditary multiple osteocartilaginous exostoses is one of the most common genetic errors affecting the skeleton. The male-to-female ratio is approximately 3:1. The syndrome is inherited as an autosomal dominant trait. In patients with multiple osteocartilaginous exostoses, follow-up examinations are indicated at frequent intervals during the adolescent years. The finding of one exostosis that appears to be growing at a much faster rate than others could alert the clinician to the rare possibility of malignant degeneration. This is particularly true if the lesion continues to grow progressively after the cessation of the normal period of growth. If this is the case, the enlarging lesions should be removed surgically for both diagnostic and cosmetic considerations. In addition, should any osteochondroma be removed for cosmetic purposes, a recurrence of the lesion would be highly suspicious for the possibility of malignant degeneration, since benign lesions do not normally recur after simple excision.


Enchondromas are benign cartilaginous growths that develop within the medullary cavity of the bone, affect both sexes with equal frequency, and usually become clinically apparent during adolescence. They are usually not diagnosed at the time they arise, however, since they produce no symptoms and are incidentally discovered following radiographs or bone scans that are performed for other reasons, such as trauma, staging, or diagnoses of other diseases. Enchondromas can arise in any bone; they also have a propensity to arise in the short tubular bones of the distal extremities. Enchondromas are the most common bone tumor found in the hand. They occur in both the long bones and the flat bones and occur in the bones of the skull as well as the pelvis.

The radiographic appearance of an enchondroma is that of a benign tumor (Fig. 121.21). Lesions that are located in the long bones are usually in the diaphyseal portion. They usually show a slightly expansile character with thinning of the cortex; however it is usually not breached by the tumor, unless there is a pathologic fracture in an enchondroma arising in an area of unusual mechanical stress. Many enchondromas show characteristic calcifications that appear as rounded or stippled areas of dense calcium deposition.

Figure 121.21. A.

Figure 121.21

A. Enchondromas are benign lesions in which the marrow cavity is replaced by often heavily calcified cartilage. B. Note that the cortex is intact and the lesion is not actively destroying the bone.

Histopathologically, the cartilaginous lesions appear quite characteristic. One can see benign nests of cartilage that have benign ossified rims around them. The normal surrounding marrow spaces are not infiltrated or destroyed by sheets of cartilage, as is seen in chondrosarcomas.

The differential diagnoses of an enchondroma is usually from a low-grade chondrosarcoma. While small enchondromas that occur in the long bones are usually radiographically characteristic and seldom require a biopsy (unless symptomatic due to a pathologic fracture), lesions in the pelvis can grow to a considerable size, and usually wind up being biopsied to rule out a low-grade chondrosarcoma, particularly if the patient is in the older age group, where chondrosarcomas are more common. The low-power histologic examination of an enchondroma is quite characteristic and is easy at times to differentiate from a low-grade chondrosarcoma with the criteria developed by Mirra. 132 These include the presence of benign bony rimming of the cartilaginous nodules, and the lack of an infiltrative marrow pattern in the surrounding bone marrow and between the normal bony trabeculae, as occurs in chondrosarcoma. The cellular elements, however, should still be carefully studied for evidence of pleomorphism, cellular atypism, hypercellularity, increased lucency of the nuclear material on hematoxylin and eosin staining, prominent nucleoli and the presence of double nuclei, bizarre forms, or rarely mitotic figures. It is on the basis of these characteristics that the cartilage lesion is most often graded as 0 to 3 to 4.

Clinically, enchondromas are usually diagnosed when patients are staged for other diseases with a bone scan since they show considerable uptake on the technetium bone scan. They also may be diagnosed incidentally when radiographs of the affected limb are taken for other reasons. The differential diagnosis between an enchondroma and a chondrosarcoma includes the fact that enchondromas are totally asymptomatic, and chondrosarcomas usually produce pain. In patients with multiple enchondromatosis (Ollier’s disease), multiple lesions may cause considerable deformity. This is particularly true in the later years, when there is excessive involvement of the bones of the hand.

Most benign enchondromas do not require surgical treatment. Patients with multiple enchondromatosis may require corrective surgery for both functional and cosmetic deformities produced by the multiple lesions. The major indication for treatment is the possibility of degeneration of the benign enchondroma into a chondrosarcoma, or the differential diagnoses of a large solitary enchondroma by ruling out a chondrosarcoma. The development of pain without a pathologic fracture is an indication for biopsy; the large size of some lesions in the pelvis requires a surgical biopsy to rule out a malignant chondrosarcoma. The incidence of malignant degeneration into chondrosarcoma is much higher in the syndrome of multiple enchondromatosis, and especially Maffucci’s syndrome, than in solitary enchondroma. In addition to the development of pain, the rapid growth of a lesion that has been under observation (i.e., in patients who are routinely monitored for multiple enchondromatosis or patients who have a suspected benign enchondroma who have follow-up examinations) is also an indication for a biopsy.

When biopsying an enchondroma for the possibility of chondrosarcoma, it is important to realize that further treatment for a chondrosarcoma may require a radical en bloc resection. The placement of the biopsy scar should, therefore, be done as it would relate to a possible future radical en bloc resection, if indeed a chondrosarcoma were discovered. In addition, care must be taken to carry out the biopsy in an area where there is suspected malignant degeneration, since malignant degeneration within an enchondroma may take place in only one position of the tumor. CT and MRI imaging are helpful in guiding the surgeon to the area of rapid suspected growth, which is where the biopsy should be done. When a biopsy is done, it is best to do a complete curettage, if the frozen section is benign. This eliminates the risk of a sampling error and removes the entire lesion.

If benign appearing enchondromas that are incidentally found are not biopsied, they should have some regular form of follow-up, since a finite low frequency of malignant degeneration occurs even in solitary benign enchondromas. If a benign enchondroma is biopsied and removed, the prognosis is excellent. However, as in the case of benign osteochondromas that are surgically treated, benign enchondromas that recur or continue to grow must be looked at suspiciously as potential chondrosarcomas.

Chondromyxoid Fibroma

The chondromyxoid fibroma is a rare benign primary tumor of bone, which is composed of a fibrous stroma demonstrating myxoid differentiation throughout the stroma with scattered foci of chondroid differentiation as well. The tumor can occur in any age group, but it has been reported most frequently during the second decade of life. It is slightly more common in females, and although it can occur in all the long bones, short tubular bones of the extremities, and flat bones, it is most common in the tibia.

The radiographic appearance of the chondromyxoid fibroma is usually that of the benign tumor, that is, the lesions usually have sharply defined margins within the bone, with a somewhat sclerotic border forming the bony margin. However, there can be expansile growth pushing out the cortex of the bone. It has been said that the radiographic appearance of the typical chondromyxoid fibroma in the tibia is that of a bone in which a “bite” has been taken out, since the margins are frequently somewhat irregular although well defined.

Histopathologically, the lesion is composed of a benign fibrous stroma growing on a myxoid background. There may be scattered foci of cartilaginous material that are usually in well-defined lobules within the lesion.

The major differential diagnosis of a chondromyxoid fibroma is from a low-grade myxoid chondrosarcoma, which tends to occur more frequently in the older age group. The finding of a typical lesion in the tibia of an adolescent usually leads to a diagnosis of the benign chondromyxoid fibroma. Should a question of the histology arise, however, consultation with a skeletal pathologist is recommended.

The clinical history of a chondromyxoid fibroma is usually that of slowly developing dull pain and swelling in the lesional area. Because chondromyxoid fibromas are frequently eccentrically placed within the bone and do not usually involve the entire medullary cavity, pathologic fractures in this disease are rare.

Treatment for chondromyxoid fibroma is surgical excision of the lesion. With simple curettage of the lesion, there can be a 10 to 20% local recurrence rate, which does not signify malignant degeneration, but rather incomplete removal of a lesion which is somewhat more aggressive than benign osteochondromas, or enchondromata. Therefore, complete curettage of the chondromyxoid fibroma is recommended, with either phenol instillation or cryosurgery. En bloc resection of the tumor is the preferred treatment if the lesion is placed eccentrically enough to remove the lesional area surgically without sacrificing normal joint function. 7


Benign chondroblastoma is a primary bone lesion consisting predominantly of immature cartilage, scattered calcification, and a benign plump spindle cell stroma. Chondroblastomas are tumors believed by some to be of epiphyseal origin. The tumor is relatively rare and represents less than 1% of all primary bone tumors. Nevertheless, it is slightly more common than the chondromyxoid fibroma. Like the chondromyxoid fibroma, the chondroblastoma reaches its peak incidence in the second decade of life. The tumor is about twice as common in males. Benign chondroblastoma occurs in the epiphyses of growing bones and is most common in the proximal humerus, followed by the epiphyseal centers of the femur and tibia. It can occur in any bone where there is epiphyseal cartilage present. Its etiology is a suspected aberration of cartilage production, representing a proliferation of immature chondrocytes.

The radiographic appearance of benign chondroblastoma is usually that of a well-circumscribed small lesion appearing in the epiphysis of a growing adolescent. The lesion is usually well marginated and frequently shows stippled calcification (Fig. 121.22). Histopathologically, the lesion is quite typical and consists of the background stroma of polyhedral cells with small, dark nuclei. In some areas, multi-nucleated giant cells can be quite numerous. The tumor exhibits a calcification, which is in a typical pattern sometimes referred to as “chicken wire” or spectacle calcification. There may also be areas of chondroid differentiation found within the tumor. Rarely is the tumor associated with aneurysmally dilated blood vessels somewhat resembling an aneurysmal bone cyst.

Figure 121.22. A.

Figure 121.22

A. A classic chondroblastoma in a typical location in a 13-year-old girl. The radiograph shows the lesion located in the proximal humeral epiphysis, and the CT scan (B) demonstrates the extent and the nature of the benign cartilage tumor.

The clinical history of chondroblastoma is usually of slowly developing pain and swelling associated with the end of a growing bone. If the tumor progresses, it may rarely produce a joint effusion secondary to growth and irritation of the tendons and joint space.

Treatment for benign chondroblastoma, as that for a chondromyxoid fibroma, should be somewhat more aggressive than simple curettage. Chondroblastomas have a recurrence rate of about 10 to 30%, rarely behave as locally aggressive neoplasms, and can recur in the soft tissues following simple curettage. It is, therefore, recommended that complete curettage or en bloc excision of the lesional area be performed when possible. If chondroblastomas recur locally in the soft tissues, they can metastasize, albeit rarely. The authors have seen five patients with recurrent chondroblastomas that metastasized to lung and subcutaneous tissues leading to death. Complete surgical treatment of the primary tumor and even more aggressive surgical treatment for locally recurrent chondroblastomas is imperative, since completely excised chondroblastomas have an excellent prognosis. 133

Malignant Cartilage-producing Bone Tumors

The malignant cartilage-producing lesions include the periosteal osteosarcoma (which is discussed under “Osteosarcoma”) and chondrosarcomas (see Table 121.5). The classic chondrosarcoma that occurs in the older age group is usually a low-grade malignant tumor that requires locally aggressive treatment. Not infrequently, these tumors undergo what has been termed dedifferentiation into highly malignant tumors. Mesenchymal chondrosarcoma, another form of chondrosarcoma, is a rare malignant tumor found in the younger age group and requires multi-disciplinary treatment as discussed below.

Table 121.5. Malignant Cartilage-Producing Tumors.

Table 121.5

Malignant Cartilage-Producing Tumors.

Chondroblastoma is a locally aggressive lesion which is best treated with a curettage. A complete curettage will control more than 85% of lesions. The lesion that recurs can be treated with another curettage. If the lesion occurs in a bone that can be resected without any functional deficit, resection is acceptable. Chondroblastoma has been reported to metastasize, but it is so uncommon that the concern over this risk should not influence the initial surgical management.

Classic Chondrosarcomas

Chondrosarcoma is a malignant primary tumor of bone characterized by malignant cartilaginous proliferation. Chondrosarcoma of bone is second in frequency to osteosarcoma of bone among malignant bone lesions.

Chondrosarcomas represent approximately 20% of all primary bone lesions. They can occur de novo as a primary tumor or as a secondary tumor following malignant degeneration of a benign osteochondroma or a benign enchondroma. The tumor usually occurs in patients > 40 years of age; however, it also occurs in the younger age group, and when it does, it tends to be of a higher-grade malignancy capable of metastases. It occurs with nearly equal frequency in both sexes. Chondrosarcomas can occur in any bone, but there is a preponderance of lesions found in the pelvis and femur. Chondrosarcomas arising in other flat bones, such as the scapula, ribs, and skull, are not uncommon.

The radiographic appearance of a chondrosarcoma is that of a malignant bone lesion (ill-defined borders, permeation of the bone, and the production of the soft tissue mass). There is usually extensive calcification found within both the bony lesion and the soft tissue mass (Fig. 121.23).

Figure 121.23. a.

Figure 121.23

a. A typical chondrosarcoma of the upper end of the femur. B. The lesion is much more destructive than that shown in Figure 121.21, and on both lateral radiograph and CT scan (C), the tumor lies outside the bone as a soft tissue extension.

Histopathologically, chondrosarcoma is composed of malignant cartilage production. The malignant chondrocytes within the cartilage show a variation of size and shape of the nuclei, and frequently there are binucleated and multi-nucleated cells found in the single cartilaginous lacunae. The classic chondrosarcoma that occurs in the older age group ranges from a very low-grade lesion that can be effectively treated by curettage and phenolization of the cavity to one that is so destructive and extensive that widely resective surgery and adjuvant radiation and chemotherapy may be advocated. One of the problems with cartilage tumors is that the site biopsied is a limited biopsy of the lesion and may be deceptively benign or low-grade in appearance, while the more aggressive focus is the one that dominates the picture. Some chondrosarcomas show higher degrees of anaplasia, nuclear atypia, and multiplicity of binucleated and multi-nucleated cells within the lacunae, however, and are classified as fully malignant (grade 3) lesions. 134 The latter malignant lesions are more commonly seen in classic chondrosarcomas that arise in the younger age group (Fig. 121.24) in proximal sites or in patients with Ollier’s or Maffucci’s syndrome.

Figure 121.24. An advanced high-grade (chondrosarcoma) occurring in the ilium of a 21-year-old male.

Figure 121.24

An advanced high-grade (chondrosarcoma) occurring in the ilium of a 21-year-old male. Note the extensive involvement of the ilium, calcifications in the soft tissues, and involvement of the sacrum and vertebral bodies by direct extension of this tumor which (more...)

The clinical history of chondrosarcoma is characterized by pain in the lesional area. Since the pelvis is a common site of chondrosarcoma, symptoms usually go on to include the manifestations of nerve root involvement, including pain and bowel and bladder dysfunction.

Staging for a malignant chondrosarcoma should include, in addition to adequate imaging of the primary lesion (CT scan or MRI), a CT scan of the chest to rule out pulmonary metastases and a bone scan to rule out the presence of other bony metastases.

Treatment for a malignant chondrosarcoma includes wide en bloc excision of the lesion area, which sometimes means amputation. In many of the pelvic lesions, amputation or radical en bloc excision is not possible because of extension into the sacrum or spine. In the latter instance, preoperative radiation therapy and/or chemotherapy may be indicated to try to make the primary lesion more surgically operable. It is commonly thought that most chondrosarcomas are resistant to both radiation therapy and chemotherapy. We have noted from examining specimens resected after preoperative chemotherapy and radiation therapy, however, that many of the malignant chondrocytes have been destroyed by the preoperative therapy. Preoperative treatment of some patients with high-grade lesions may also lead to a decrease of symptomatology. Indication of objective clinical regression may be elusive because most of the lesional matrix is made of cartilage, and the radiograph or CT appearance of the lesion following preoperative therapy may not change substantially. Preoperative therapy of some of the more marginally resectable chondrosarcomas might make the local recurrence rate lower where en bloc surgery or limb salvage surgery is utilized rather than the classic form of surgical treatment, which is radical amputation.

Following adequate surgery, the major risk for the majority of patients with low-grade chondrosarcomas is local recurrence. Even the fully malignant (grade 3) lesions represent more of a risk of local recurrence than they do of metastases. The occurrence of pulmonary metastases is often quite late and may only include a few solitary metastases that are amenable to surgical treatment. If the surgically removed metastases demonstrate a high-grade spindle cell stroma rather than a predominantly cartilaginous matrix, this may be an indication that the metastases of the original primary tumor have undergone “dedifferentiation” to a more highly malignant lesion, that is, cytologically more primitive less differentiated cells have metastasized. This event may also be an indication that at least one region of the primary lesional area has also undergone “dedifferentiation,” that is, cytologically more primitive and less differentiated cells have replaced part of the more profuse chondroblasts. The majority of low-grade chondrosarcomas that are amenable to complete surgical excision, however, have a relatively good 10-year survival rate. Even lesions of the pelvis that are difficult to resect can have a median survival time as high as 10 years with repeated surgeries and the use of radiation therapy. 135

The surgical procedures currently in vogue for resection of chondrosarcoma vary considerably, depending on the site, the size of the lesion, the presence of metastases, the grade of the tumor, and the age of the patient. In general, the lesion should be widely excised and the specimen carefully studied for the proximity of the tumor to the margins. Amputation, particularly for lesions below the knee, or internal or occasionally external hemipelvectomy for big tumors of the pelvis is a useful approach. Adjuvant radiation before and after resection may be helpful, as might be chemotherapy for high-grade tumors. Once the lesion is removed, allograft replacement has been advocated by some, while metallic implants are in vogue in other centers. Both have high initial infection rates, but the metallic implants do better initially and then have a high rate of loosening or breakage at 5 to 10 years. Allografts have more problems early with nonunion and fracture (both in the 20% range), but once the 3-year period of high incidence for these complications is over, the grafts function well for at least 10 to 20 years. For patients with chondrosarcomas who are more likely to live, or for very young people, the allograft (at least in some centers) appears to be the treatment of choice. For older patients, the metallic implants allow a more rapid restoration and are obviously the better of the two approaches.

Dedifferentiated Chondrosarcoma

“Dedifferentiated” chondrosarcoma is a rare highly malignant tumor characterized by the finding of the typical low-grade chondrosarcoma of bone that has undergone malignant degeneration and produces a fully malignant soft tissue mass, the latter being indistinguishable from an MFH. 136

Dedifferentiated chondrosarcomas occur rarely and are more common in the older age group (i.e., after 60 years of age). Their incidence is about equal in males and females. They occur most commonly in the areas where chondrosarcomas are common, including the pelvis, femur, and other flat bones, such as the ribs and scapula. In most series that have been reported, the disease is uniformly fatal, with patients succumbing to pulmonary metastases early in the course of the disease.

The radiographic appearance of dedifferentiated chondrosarcoma is that of a malignant chondrosarcoma of bone, that is, diffuse permeative bony involvement and calcifications within the bone. A large soft tissue mass usually exists, which lacks calcifications within it.

The histopathology of dedifferentiated chondrosarcoma is that of a fully malignant fibrous stroma, similar to that of an MFH or fibrosarcoma, in the soft tissue mass. The intramedullary or bony portions of the lesions usually contain low-grade cartilaginous areas. The differential diagnosis is from other malignant tumors, including a fibrosarcoma or MFH of bone.

The clinical history of dedifferentiated chondrosarcoma is usually that of long-standing pain in the lesional area. Often, there is then a rapid development of a soft tissue mass that finally forces the patient to seek medical attention. Following the diagnosis and even radical surgical treatment of dedifferentiated chondrosarcoma, there is usually rapid dissemination of the tumor to both the lung and other bones.

Staging of patients with a dedifferentiated chondrosarcoma should include imaging of the primary lesional area with CT scanning and MRI to determine the extent of both the soft tissue mass and the extension of the tumor within the medullary canal. Treatment for dedifferentiated chondrosarcomas is similar to that for osteosarcoma. Because of the rarity of dedifferentiated chondrosarcomas, data are lacking to indicate that HD-MTX with leucovorin rescue is as effective in this disease as it is in osteosarcoma, and thus they are usually omitted. Preoperative treatment with drugs such as cisplatin combined with doxorubicin and high-dose ifosfamide, as is given for osteosarcoma, frequently results in the dramatic shrinkage of the soft tissue mass. Following preoperative chemotherapy, wide resection or amputation of the primary lesional area is performed. Postoperatively, patients can continue on adjuvant chemotherapy with the same agents initially proven active in the patient. Within the past 5 years, three patients with dedifferentiated chondrosarcoma treated with high-dose cisplatin (120 mg/m2) combined with doxorubicin (60 mg/m2) had complete histologic and clinical responses noted in the soft tissue mass of the primary tumor. Following surgery, the only residual tumor found was the more resistant residual low-grade chondrosarcoma within the medullary cavity of the bone. Two patients who underwent wide surgical excision (proximal humerus and distal femur) are surviving free of disease over 3 years from the time of diagnosis. The third patient who did not have complete excision of the residual low-grade chondrosarcoma in a pelvic lesion developed a local recurrence and rapid dissemination of the tumor resulting in death. This emphasizes the need to remove the residual chondrosarcoma as well the entire tumor. Dedifferentiated chondrosarcoma occurs in the older age group, and the two surviving patients who tolerated both preoperative and postoperative chemotherapy with cisplatin and doxorubicin were 83 and 74 years of age, respectively. 137

Mesenchymal Chondrosarcoma

Mesenchymal chondrosarcoma is a rare malignant tumor of bone (that can also occur as a soft tissue primary tumor) composed of an undifferentiated round or small cell stroma, which can contain spindle cell elements and neoplastic cartilage formation.

Mesenchymal chondrosarcoma occurs in the young adult group and is relatively rare below the age of 20 years or above the age of 40 years. It has been shown to affect males and females equally in the few series that have been reported. The most common site is the femur, followed by the flat bones, including the pelvis, ribs, skull, and spine. It is said to have a propensity to occur in the parameningeal areas in the head, neck, and spine.

The radiographic appearance of mesenchymal chondrosarcoma is that of the malignant bone tumor with an ill-defined border in the medullary canal of the bone, and a production of a soft tissue mass with tumor penetrating the bony cortex. The soft tissue mass may contain stippled calcification similar to that seen in classic chondrosarcomas. However, the finding of this characteristic type of malignant lesion in the younger age group and, in particular, in the parameningeal areas, such as the spine or skull, could raise the possibility of a mesenchymal chondrosarcoma. The diagnosis is usually arrived at following biopsy of the suspected malignant lesion.

Histopathologically, the tumor is composed of sheets of undifferentiated malignant round cells sometimes quite similar to those seen in a Ewing’s sarcoma of bone or small cell osteosarcoma. However, the tumor may also contain more differentiated or plump spindle cells similar to that seen in osteosarcomas. In addition, neoplastic cartilage is abundantly seen within most tumors.

The differential diagnosis of a mesenchymal chondrosarcoma is that of any malignant bone or cartilaginous tumor. Histologic analysis, however, is critical to the diagnosis. The finding of a predominant small round cell component, which is usually the case in this tumor, means that the tumor is a mesenchymal chondrosarcoma. The neoplasm is usually more responsive to radiotherapy and to chemotherapy than the average osteosarcoma or chondrosarcoma. Because of the malignant undifferentiated round cell component to this tumor, the tumor is highly malignant. Its differentiation from a classic chondrosarcoma is important because preoperative chemotherapy and radiation need to be initiated. They frequently result in significant shrinkage of the primary tumor, making it more amenable to radical surgical excision. This is particularly important in tumors of axial location, such as the spine, skull, scapula, ribs, and pelvis, where complete en bloc excision is otherwise seldom possible. It is important to differentiate true osteosarcoma from mesenchymal chondrosarcoma since radiation therapy is seldom indicated for the preoperative treatment of osteosarcomas.

The clinical history of mesenchymal chondrosarcoma is one of pain and swelling in the lesional area. The onset of symptoms is usually rapid, since this is a highly malignant tumor. Because of the axial location of many of these tumors, neurologic symptoms may be the first to occur. Early diagnosis of pelvic, sacral, and paraspinal primaries may be possible because of the neurologic pain syndromes produced by these tumors.

Staging for mesenchymal chondrosarcoma includes (in addition to adequate imaging studies of the lesional area) a bone scan to rule out bony metastases, more common with mesenchymal chondrosarcoma than with other chondrosarcomas or osteosarcoma. A CT scan of the lungs to rule out pulmonary metastases is also included in the preoperative staging.

The treatment of mesenchymal chondrosarcoma should be similar to that discussed below for Ewing’s sarcoma. The majority of the malignant stroma of mesenchymal chondrosarcoma consists of small round cells that are usually sensitive to both radiation therapy and chemotherapy with traditional agents, such as doxorubicin, and a high-dose alkalizing agent, such as cyclophosphamide or ifosfamide. Radiation therapy is usually given to a dose of between 40 and 45 Gy during the course of preoperative chemotherapy. This usually results in rapid shrinkage of the primary tumor, prevention of early dissemination to bones and lungs, and the tumor being rendered approachable by wide en bloc excision. Postoperatively, if there is significant tumor necrosis (90%) in the resected specimen, the patient is continued on chemotherapy similar to that used to treat Ewing’s sarcoma. Because mesenchymal chondrosarcoma is a pleomorphic tumor sometimes containing more resistant spindle cells, the finding of residual spindle cell stroma in the resected specimen and 90% necrosis suggests resistance to the radiation therapy and alkylating agents used preoperatively. Postoperative adjuvant chemotherapy should then be changed to include cisplatin combined with doxorubicin together with high-dose ifosfamide as used in the treatment of osteosarcoma. 137

Prior to the use of combination multi-modality therapy for the treatment of mesenchymal chondrosarcoma, this tumor metastasized in almost 100% of the reported cases. With modern multi-disciplinary treatment as described above, the prognosis for mesenchymal chondrosarcoma should be similar to that for Ewing’s sarcoma with similar treatment. A caution is that most mesenchymal chondrosarcomas arise in the axial skeleton and are difficult to resect surgically, making the overall prognosis more similar to that of axial Ewing’s sarcomas, which is worse than that for extremity lesions. In addition, the occasional finding of more resistant spindle cell stroma in the resected specimen indicates that this tumor is somewhat more resistant to combined modality treatment, and the risk for metastases developing late in the course may be somewhat higher than that for the more common Ewing’s sarcoma that is similarly treated.

Chondrosarcoma has a reputation for being a radioresistant neoplasm of bone. Morphologically, intact tumor has been reported after 10,000 cGy. 138 Close analysis of the few well-documented series from the megavoltage era reveal that long-term progression-free survival following radiotherapy is possible for selected patients with chondrosarcoma of bone. The largest experience with modern radiotherapy for chondrosarcoma has been reported from the Princess Margaret Hospital. Harwood et al. 139 initially described the results of palliative or radical irradiation for 31 patients. These authors, as well as others, 140 admit that analysis of tumor response to radiotherapy is difficult for chondrosarcoma of bone. Harwood stressed that in no case did the radiographic appearance of an irradiated bone return to normal despite prolonged progression-free survival. “Response” was defined strictly by clinical estimates of tumor size and intensity of symptoms. Stable or decreasing tumor size or symptomatic reduction were defined as responses. Patients reported by Harwood and associates were selected for several poor prognostic features. Twelve patients had mesenchymal/dedifferentiated tumors, 24 had axial lesions, and 20 presented with painful masses. These features are associated with poor prognosis in surgical series of chondrosarcoma patients. Twelve patients with well- or moderately differentiated chondrosarcomas were irradiated with curative intent following biopsy or grossly incomplete resection. Six of these 12 patients remained in complete remission for 3 to 16 years following treatment. Six patients had local progression despite radical radiotherapy. Four of these relapsing patients, however, demonstrated a response to radiotherapy that was maintained for 2.5 to 8 years before the eventual progression of their disease. Typically, many months were required for maximum expression of response to radiotherapy. The authors noted 1 patient, furthermore, demonstrating apparent radiographic tumor progression several months after irradiation that was followed by slow recalcification. Eight patients with poorly differentiated, dedifferentiated, or mesenchymal chondrosarcoma were radically irradiated. Two were alive, free of progression 3 and 8 years after treatment. Two had local disease progression 6 and 12 months following radiotherapy. The remaining 11 patients reported by Harwood received radiotherapy of purely palliative intent to various metastatic deposits. None of these patients derived substantive benefit from low-dose irradiation. A subsequent report of 11 mesenchymal chondrosarcomas confirmed these initial findings. 141 Four of six patients responded to radical irradiation. One patient remained free of progression 1.5 years following therapy. The other responding patients had local relapses 1.5 to 5 years after treatment. One palliatively irradiated patient responded for 2 years before eventual local relapse. Mesenchymal chondrosarcoma with a prominent component of small undifferentiated cells may be more radioresponsive than those variants with a differentiated “hemangiopericytomatoid” pattern. 142 The Princess Margaret Hospital experience with radical irradiation for chondrosarcoma was updated in 1983. Five-year actuarial survival for 25 patients with favorable histology was 48%, compared with 22% for 13 patients with unfavorable tumors. Progression-free survival was reported in 13 of 25 versus 2 of 13 (p = .01), respectively. Local progression occurred in 11 of 25 cases with favorable histology, compared with 8 of 13 unfavorable cases. Isolated local relapse accounted for 17 of the 19 local progressions.

McNaney et al. 140 irradiated 20 patients with chondrosarcomas at the M.D. Anderson Hospital. Fourteen patients were treated solely with radiotherapy for primary 11 or recurrent 3 tumors. Twelve responded to radiotherapy. Response was maintained for 12 to 87 months (median 30) in those with de novo disease and 19 to 77 months for those with recurrent tumor. Five responding patients eventually relapsed at the primary site 26 to 156 months after treatment. Significantly, 5 of the patients remaining in remission at the time of analysis were treated with both photons and neutrons. The world experience with high (LET) neutron radiotherapy for chondrosarcoma was summarized by Richter et al. 143 Local control was obtained in 23 of 41 (56%) cases treated in six different centers. Analysis of neutron therapy for chondrosarcoma is difficult, since these centers employ different selection criteria, definitions of control, and follow-up. McNaney also reported 6 patients with primary chondrosarcoma treated in combination with either partial excision or chemotherapy. Four were free of progression for 15 to 52 months.

Mark and co-workers 144 retrospectively reviewed the University of California, Los Angeles, (UCLA) experience with incompletely resected chondrosarcomas of the head and neck. Three of 6 patients irradiated as part of primary management were progression-free after 6 to 8 years. Radiotherapy was used as part of salvage therapy for 5 patients. Three were progression-free for 3 to 6 years before dying of metastatic disease, and 1 had local relapse 5 years after treatment. Since 7 UCLA patients had high-grade tumors, the apparent beneficial effects of radiotherapy may not apply to low-grade chondrosarcomas in this region. Paddison and Hanks 145 reported 2-year control by radiotherapy alone for low-grade maxillary sinus chondrosarcoma. In several surgical series, there has been no significant relationship between the grade of typical chondrosarcoma and the local recurrence rate. 146

Although the majority of chondrosarcomas recurring locally after surgery are detected within 5 years, the natural history of this tumor is notoriously capricious. 138 Assumptions concerning the efficacy of any local therapy based solely on short-term follow-up must be considered tenuous. Evans et al. 138 reported local relapse of low-grade chondrosarcoma 10 years after resection. Late-occurring local relapses hamper analysis of proper dose and technique for irradiating chondrosarcoma. Field margins of at least 5 cm have been advocated. Elective irradiation of the regional draining lymphatics should be considered for mesenchymal chondrosarcoma. Huvos et al. 142 reported metastases to regional or distant lymph nodes in 16 of 35 patients with this neoplasm. The Princess Margaret Hospital investigators used 50 Gy in 20 fractions radical treatment. All but one patient reported by Mark et al. received between 49.5 and 65 Gy. At the M.D. Anderson Hospital, patients with locally controlled disease received an equivalent dose of 60 to 70 Gy. Efforts to escalate dose even further require specialized techniques. Investigators in Boston and San Francisco have each reported 5-year actuarial local control rates of 78%, using protons and helium ions, respectively. 148 These centers attempted to deliver charged particle doses varying between 65 and 80 Gy, but tumor dose heterogeneity was common due to constraints imposed on treatment planning by the tolerance of surrounding normal tissues. CT, rather than MRI, was also used for treatment planning in many of the patients reported from these centers. The majority of local relapses reported after charged particle treatment of skull base tumors were the result of regions of tumor underdosage or progression of disease at field margins. 149 Treatment planning utilizing MRI has resulted in an improvement in the local control rate of skull base neoplasms, following helium ion irradiation. Local progression of chondrosarcoma within an area of prescribed particle dose is uncommon but has been reported by Austin et al. 149 The average tumor dose in cases relapsing within the prescribed dose region was 69 Gy. Tumor volume likely explains local relapse following such high-dose radiotherapy. 150 The average tumor volume for patients with recurring disease in a prescribed region in Austin’s review was 126 cc, compared with 75 cc for patients with recurrences due to other technical reasons. These experiences imply that even extraordinary radiotherapy doses may not permanently eradicate bulky chondrosarcoma. That serious complications were noted in 23 of 84 progression-free survivors by Castro et al. 151 implies that charged particle radiotherapy should be employed with caution for skull base tumors and only when sophisticated treatment planning is available.

Bone Tumors of Vascular or Uncertain Histogenesis

Aneurysmal Bone Cyst

Aneurysmal bone cyst is a benign bone lesion characterized by aneurysmally dilated vascular spaces within the bone. Aneurysmal bone cyst is a rare tumor that usually occurs within the first and fourth decades of life. The lesion occurs in the metaphyses of long bones, but it can also affect the bones of the axial skeleton including the vertebral body.

The radiographic picture of an aneurysmal bone cyst is quite typical (Fig. 121.25). It has the appearance of a well-demarcated expansile lesion that thins the cortex and appears to “blow out” the bone, producing a “bubble” appearance.

Figure 121.25. An aneurysmal bone cyst arising in the distal fibula of a 14-year-old female.

Figure 121.25

An aneurysmal bone cyst arising in the distal fibula of a 14-year-old female. Note the clear and sharply defined margins defining the extent of disease in this lesion. The tumor has thinned and expanded the cortex but has not broken through it. Because (more...)

Histologically, aneurysmal bone cysts consist of large vascular channels lined with plump cells that may not be of endothelial origin. Many of the lesions show fibroblastic proliferation, histologic infiltration, and focal areas of giant cell proliferation.

Because aneurysmal bone cysts occur in the metaphyseal areas of long bones in children and adolescents, it is important to differentiate these tumors from telangiectatic osteosarcoma, which can present with a somewhat similar radiographic picture. Telangiectatic osteosarcoma histologically also contains aneurysmally dilated, blood-filled spaces, but the blood-filled spaces are malignant tumor cells, not the plump modified fibroblasts, and sometimes endothelial cells of aneurysmal bone cysts. Aneurysmal bone cysts usually produce painless swelling of the affected area. Pathologic fracture secondary to marked thinning of the cortex in a weight-bearing bone can occur.

Adequate biopsy is necessary to exclude telangiectatic osteosarcoma. Aneurysmal bone cysts are then treated by thorough curettage. Occasionally they recur, and further curettage with either cryosurgery or phenol installation may prove effective. Multiple recurring aneurysmal bone cysts in adolescent patients should raise a suspicion of misdiagnosed telangiectatic osteosarcoma or transformation to it, and a thorough pathologic examination of currently curetted material should be undertaken.

Radiation therapy is an effective treatment for aneurysmal bone cysts. A compilation of literature series, usually consisting of several cases from any particular institution, indicates that 90% of aneurysmal bone cysts are controlled by irradiation. 152 Follow-up in these series varies from 1 to 32 years. Radiotherapy is equally effective for primary disease and for lesions recurrent after surgery. 151 There is no apparent advantage to incomplete excision prior to radiotherapy, compared with biopsy alone. Aneurysmal bone cysts rarely respond completely or even partially to radiotherapy. Rather, the expanded segment of bone recalcifies and persists for years in a densely sclerotic condition. 153 Local relapse of an aneurysmal bone cyst after primary surgical therapy generally occurs within 2 years. Relapses after radiotherapy have been reported as long as 6 years after treatment. 151 Delayed relapses imply the requirement for prolonged follow-up after irradiation. The rate of tumor sclerosis due to radiotherapy is variable. Aneurysmal bone cysts typically ossify slowly, but more rapidly than the true giant cell tumor. 153 Dabska and Buraczewski 153 stated that occasionally early recalcification may be noted at the conclusion of radiotherapy. Complete tumor response and reconstitution of normal osseous architecture has been reported, particularly for vertebral lesions. 154 Complete alleviation of neurologic deficits has been reported following irradiation of spinal lesions. Transient cyst enlargement after radiotherapy followed by progressive bony sclerosis has been described.

Attributing cystic osteosclerosis and long-term control of an aneurysmal bone cyst to irradiation may be spurious. Radiation oncologists must remember that many authors have documented resolution or long-term control of aneurysmal bone cysts following incomplete excision alone. 153– 155 This bone lesion may have a propensity for spontaneous regression following a surgical manipulation. While control after incomplete removal has been reported, these cases remain largely anecdotal, and many authors stress that spontaneous resolution is never assured. 156 Radiotherapy remains an appropriate treatment option for difficult aneurysmal bone cysts. Biesecker et al. 157 and Tillman et al. 158 reported superior disease-free survival following radiotherapy for this lesion, compared with all methods of curettage alone.

The minimum effective dose of radiotherapy for aneurysmal bone cysts has not been established. Much of the radiotherapy experience with this entity was accumulated in the kilovoltage era, accounting for uncertainty over treatment details. Previous authors, when commenting on technique at all, often confusingly expressed dose as “estimated tumor dose in Roentgens” or “air dose in rad.” 159 These kilovoltage era investigators employed doses varying from 1,000 to 3,000 R and typically recommended 1,200 to 2,000 R. 154, 159 Tillman et al. 158 reported local relapse after 800 R.

Experience from the megavoltage era is scanty. Doses varying from 1,100 to 4,000 cGy have been utilized. 157, 158 Although most authors now recommended 1,400 to 2,000 cGy, Parrish 160 reported relapse after 3,850 cGy. Marks and co-workers 152 calculated that 2,000 R of kilovoltage irradiation was equivalent to 2,855-cGy megavoltage treatment. These authors concluded that the minimum effective dose for aneurysmal bone cysts might approach 3,000 cGy, a dose known to be effective for other benign vascular disorders, such as juvenile nasopharyngeal angiofibroma. Long-term control of locally recurrent aneurysmal bone cysts by reirradiation has been reported. 157, 159

Fields for irradiating aneurysmal bone cysts should include the radiographic extent of disease without excessive margins of uninvolved tissue. The aneurysmal bone cyst is a strictly circumscribed process despite often enormous cystic expansion beyond normal bone cortex. Microscopic invasion of surrounding normal tissue is limited by a thin rim of subperiosteal new bone formation. 157, 158 In a very early phase of growth, a thin bony shell may not be appreciated. 153 Radiotherapy, however, is rarely administered at this stage of aneurysmal bone cyst evolution. Invasion of the medullary cavity may occur but, typically, not beyond the radiographic outline of the lesion. Lichtenstein 161 noted that a neglected lesion invaded nearby joint spaces. Pathologic fracture of the periosteal new bone may allow soft tissue invasion, and in this unusual situation a margin of apparently uninvolved soft tissue should be provided. Careful attention is required for field planning for aneurysmal bone cysts of the spine, since multiple vertebral bodies may be involved. 160, 162

Potential late effects of radiotherapy have led several authors to erroneously advise against this form of treatment. Atrophic skin changes are a result of kilovoltage equipment and are of no concern to the modern radiotherapist. Suppression of bone growth by irradiation is unlikely, given the natural history of the aneurysmal bone cyst. 158 This lesion occurs in adolescents and young adults, many of whom have minimal remaining bone growth. The majority of lesions affecting long bones, moreover, arise in the metaphysis. Epiphyseal involvement by an aneurysmal bone cyst is distinctly unusual, the majority of such patients demonstrating an associated pathologic process of bone (e.g., giant cell tumor, chondroblastoma). 157, 163 The epiphyses may be spared the full dose of therapeutic irradiation. Finally, sarcoma induction after radiation therapy had been reported for aneurysmal bone cysts. 164 Although patients with this disorder are irradiated at an admittedly young age, the incidence of later malignancy is distinctly uncommon. Tillman and co-workers 158 reported 3 sarcomas in a literature review of 95 cases of aneurysmal bone cyst.

Eosinophilic Granuloma

Eosinophilic granuloma or, more appropriately, Langerhans’ cell eosinophilic granuloma, is a benign bone lesion that can occur in either solitary or multi-focal forms. 158, 165 It is a tumor composed of histiocytes and eosinophils. Classically, the histiocytes that are the definite cell, contain Langerhans’ cell granules, which identify the lesion. These are best seen on electron microscopy. Recent evidence suggests that the lesion is a neoplasm rather than a response to a virus infection, as was previously thought.

Eosinophilic granuloma enters into the differential diagnosis of other bone lesions discussed in this chapter. Eosinophilic granuloma predominately occurs in childhood and is rarely seen above the age of 30 years. The multi-focal variety tends to occur in younger children and may be progressive, requiring systemic therapy as well as local therapy. The solitary disease is always benign and nonprogressive, and the prognosis is excellent for cure.

The radiographic appearance of eosinophilic granuloma in flat bones is usually quite typical and produces a “punched out” lesion. Eosinophilic granuloma can occur in any bone in the skeleton, but the skull, including the jaw bones, is a particularly common location. Although the radiographic appearance is ordinarily characteristic, eosinophilic granuloma can grow very rapidly and appear extremely destructive on radiographs. Some lesions in the jaw can produce complete destruction of all bony landmarks, giving the radiographic appearance of a malignant tumor.

Pathologically, eosinophilic granuloma shows a histiocytic stroma diffusely infiltrated with eosinophils. There may be multi-nucleated phagocytic giant cells in some areas. Sometimes the eosinophilic infiltrate is most prominent around blood vessels.

The clinical history of eosinophilic granuloma of bone is one of pain and discomfort in the lesional area. Frequently, a palpable soft tissue mass can be felt. This usually leads to a biopsy and definitive treatment. Patients with suspected eosinophilic granuloma of bone should have a bone scan or skeletal survey to exclude multi-focal disease.

After biopsy, the treatment for benign solitary eosinophilic granuloma of bone is surgical curettage of the lesion. Occasionally, large solitary lesions in the weight-bearing bones in the lower extremities or in the spine can be adequately treated with low doses of radiation therapy (6–12 Gy). Even lesions that are very destructive of bone go on to heal completely with reformation of the normal bony architecture. The persistence of a lesion after curettage, which usually is curative in solitary eosinophilic granuloma, should make one consider the diagnosis of Hodgkin’s disease either metastatic or (very rarely) primary to bone (Fig. 121.26).

Figure 121.26. A mixed lytic and sclerotic destructive lesion with periosteal elevation and a soft tissue mass.

Figure 121.26

A mixed lytic and sclerotic destructive lesion with periosteal elevation and a soft tissue mass. This lesion is very characteristic of a malignant lesion on radiographs. However, multiple biopsies were originally thought to represent benign eosinophilic (more...)

Bone lesions associated with Langerhans’ cell histiocytosis (LCH) have been regarded as exquisitely radiosensitive since the early report of Sosman. 166 The extremely high response rate of these lytic lesions to irradiation led many previous investigators to conclude that radiotherapy was the treatment of choice for bone lesions. 167 Analyzing the impact of treatment on LCH is difficult, however, because spontaneous resolution of lytic bone lesions has repeatedly been described after incomplete surgery, biopsy, or observation alone. 168, 169 Radiotherapy is best reserved for selected cases: disease recurrent after other therapies, symptomatic disease not amenable to curettage, or locations where curettage would be cosmetically deforming or result in structural instability. There remains no documented role for elective irradiation after adequate curettage. Nesbit 170 advocated radiotherapy for partially collapsed vertebrae to prevent complete collapse, which may, rarely, result in neurologic sequelae. No treatment is advocated for complete vertebral collapse that is asymptomatic.

The world literature on treatment of LCH bone lesions was reviewed in 1980 by Slater and Swarm. 171 The overall local recurrence rate for 545 analyzable sites was 2.8%. Local control following radiotherapy with or without supplemental surgery was reported for 105 of 108 sites and 76 of 80 sites, respectively. Local control following curettage or excision was reported in 90 of 91 and 116 of 118, respectively. Biopsy alone resulted in healing of 13 of 15 lesions. Perslegin and associates 172 subsequently reported a large series from the Soviet Union. Local control following radiotherapy with or without additional surgery was reported for 115 of 118 sites and 63 of 68 sites, respectively. There is no significant difference in the local control rate of bone lesions associated with monostotic LCH, compared with polyostotic presentations. 173, 174 Patient age and the presence of extraosseous LCH, however, may influence radiation response of bone lesions. In a series from UCLA, 35 of 40 bone lesions were controlled by irradiation. 175 Recurrent bone lesions were noted only in adult patients who had presented with multi-system LCH. All lesions occurring in children and in adults with disease limited to bone were controlled by radiotherapy. The impact of the location of the bone lesions of LCH on control rate remains controversial. Perslegin noted no difference in control of lesions affecting flat bones, compared with long bones, but advocated curettage plus radiotherapy for jaw lesions. In the UCLA series, 9 of 11 long bone lesions were controlled by radiotherapy, compared with 26 of 29 flat bone lesions. Control of jaw lesions, however, was obtained in 6 of 8, compared with 29 of 32 sites other than the jaw. Hartman et al., 176 in a series of 88 jaw lesions, reported local control rates following surgery, radiotherapy, or combined therapy of 12%, 25%, and 19%, respectively. Jaw lesions recurring after radiotherapy have been reported elsewhere. 177 Aronson and associates 178 from Stanford University reported 100% control by radiotherapy of LCH affecting craniofacial bones. Few of these patients had jaw disease, and many received additional chemotherapy. Radiation response of LCH involving the mastoid has been doubted by some, but there is little clinical data to support this contention. The impact of lesion location on local control rate has not been assessed independent of other potential prognostic factors. Duration of symptoms has no impact on the radiation response of bony LCH. 175

Bone pain due to LCH lesions responds rapidly to irradiation. 172 Nesbit described resolution of neurologic dysfunction associated with vertebral collapse. 170 Bone reconstitution occurs almost uniformly after irradiation, although at a slower pace than either symptomatic response or resolution of soft tissue deposits of LCH. 179 Partial bone healing is noted 3 to 4 months after treatment. 179 Complete reconstitution occurs over a reported range of 6 to 24 months, with a 10- to 12-month median duration. 169, 180 The appearance of a sclerotic rim heralds the resolution of a lytic LCH lesion. Several authors note a slower rate of healing of lesions in adult patients, compared with those in children. 180 The impact of the age of the lesion on the duration required for healing remains controversial. Vertebral lesions recalcify and partially reconstitute, even if completely collapsed (e.g., vertebra plana) prior to radiotherapy. Complete restoration of vertebral height, however, has not been reported after irradiation. There is no significant difference in the duration required for healing after radiotherapy, compared with other forms of management. McCullough et al. 181 reported mean intervals of 12 and 11 months to complete healing after irradiation or simple curettement, respectively. Wormer et al. 169 analyzed the results of 42 lesions nonrandomly receiving surgery, radiotherapy, chemotherapy, steroid injection, or observation. No significant differences were found in median times to partial or complete healing. Nesbit et al. 170 reported no differences in the eventual height of collapsed vertebrae receiving radiotherapy or chemotherapy. Two untreated vertebrae in their experience also demonstrated some reconstitution.

The minimum effective dose to control LCH has not been established. The results of many small series indicate that total doses of 300 to 1,800 cGy are equally effective. 168, 180, 182 Smith et al. 183 reported control of 60 of 69 irradiated lesions. There were no clear differences in the total dose range for controlled lesions (375–1,360 cGy), compared with those recurring after irradiation (270–900 cGy). In that series, however, the majority of relapses were reported after doses ,500 cGy. In the UCLA experience, there were no significant differences in dose range for controlled (60–1,500 cGy) and relapsing (800–1,500 cGy) lesions. Median doses were 900 and 1,000 cGy, respectively. Slater and Swarm 171 found no clear evidence of a dose–response relationship over 100 to 7,500 cGy in their large review. The wide range of doses used in the literature and the very low local relapse rates led these authors to conclude that virtually any dose schedule would be effective, and that local relapse may be more a result of geographic miss than underdosage. Greenberger et al. 184 reviewed 380 bone lesions receiving total doses varying from 200 to 1,000 R and could discern no significant differences in local control rate between any regimen. The majority of authors now recommend total doses of 600 to 900 cGy delivered in 3 to 5 fractions. Righter and D’Angio 185 have cautioned that patients in the literature may have received irradiation for bone lesions that were destined to heal or were already spontaneously healing, resulting in the erroneous conclusion that low-dose radiotherapy caused remission. Higher doses have been advocated by some for specific locations or for lesions in adult patients. Perslegin recommended 400 to 600 cGy for calvarial lesions in children and 800 to 1,000 cGy for similar lesions in adults. The recommended doses for other skeletal lesions arising in childhood were 600 to 800 cGy and 1,000 to 1,400 cGy in adults. Cassady 186 also believes that lesions in adults have a higher local relapse rate than in the pediatric population and advises ranges of 1,500 to 2,000 cGy and 600 to 1,000 cGy, respectively. Dose recommendations above 1,000 cGy for specific age groups and locations have never been subjected to prospective analysis. Advanced LCH has responded to fractionated half-body radiotherapy plus vincristine/prednisone. 187

According to Slater and Perslegin, the majority of local relapses after radiotherapy occur within 1 year of treatment. Relapses in the UCLA series occurred 14 to 36 months after irradiation. Reirradiation of recurrent disease has been successful according to some authors 172 and of no benefit according to others. 175 Occasional reports of second malignancies occurring in irradiated LCH patients indicate high-dose radiotherapy or re-treatment should be employed cautiously for this condition. 187

Primitive Neuroectodermal Tumor of Bone

Primitive neuroectodermal tumor of bone may simulate Ewing’s sarcoma. It has been recently differentiated from Ewing’s sarcoma of bone on the basis of the finding of neurosecretory granules on electron microscopy. Occasionally, characteristic rosette formation can be noted on light microscopic examination. In the past, this tumor has been confused with metastatic neuroblastoma because of its tendency to arise in younger children. These tumors usually arise in the chest wall, and they behave in a fashion similar to Ewing’s sarcoma, that is, they produce locally recurrent tumors following radiation therapy for the treatment of the primary tumor. Although most common in the rib, these tumors can occur in the pelvis and long bones. 188

This tumor responds to the same treatment as does Ewing’s sarcoma. In addition, the prognosis is similar, in that if the resected specimen following combination preoperative chemotherapy and radiation therapy shows considerable or complete necrosis, the patient has good prognosis if continued on the same treatment.

Because primitive neuroectodermal tumor frequently arises in the rib (Fig. 121.27), it is worth noting some of the special therapeutic considerations given to rib primaries.

Figure 121.27. Primitive neuroectodermal tumor arising in the rib in a 19-year-old male patient.

Figure 121.27

Primitive neuroectodermal tumor arising in the rib in a 19-year-old male patient. Rib lesions are at a higher risk than other primary lesions because of the frequent finding of “dropped” metastases around the diaphragm as well as chest (more...)

Small Cell Sarcomas of the Rib

Primary small cell sarcomas of the rib have a tendency to recur locally in portions of the primary lesional bone, if the entire bone is not resected. Thus, Ewing’s sarcoma or primitive neuroectodermal tumor of the distal end of the rib has a tendency to recur in the proximal stump of the rib even if a wide surgical excision has been performed, with normal bony margins obtained. Recurrence in the proximal portion of the rib usually becomes manifest as the tumor grows across the costovertebral margin and causes neurologic symptoms of root or cord compression. For that reason, the surgical treatment of primary small cell sarcomas of ribs should include resection of the entire rib from the costochondral to costovertebral junction.

In addition, primary small cell sarcomas of the rib tend to have late metastatic disease develop in the chest wall, on pleural surfaces, and at sites caudal to the primary tumor, particularly the diaphragm. For that reason, we prefer to have preoperative radiation therapy for rib lesions, including a wide field beginning one rib above the lesion and including the entire inferior chest wall, with a boost to the diaphragm as well. With these special considerations for the radiation therapy of Ewing’s sarcomas of the rib, the prognosis should be no different from Ewing’s sarcoma in other areas. The above type of recurrences are what led people to believe that primitive neuroectodermal tumors (usually rib lesions) had a worse prognosis than other typical Ewing’s tumors.

Primary Non–Hodgkin’s Lymphoma of Bone

The diagnosis and treatment of non-Hodgkin’s lymphoma is thoroughly discussed. Primary NHL of bone will be briefly mentioned in this chapter because it is prominent in the differentiation of bone tumors.

Approximately 10% of primary NHLs arise in extranodal sites. Primary malignant lymphomas of bone represent approximately 5% of all malignant bone tumors. The tumor is rare below the age of 20 years and reaches its peak incidence in the fifth, sixth, and seventh decades. NHLs of bone do occur in younger children below the age of 10 years and behave extremely aggressively as do NHL of other sites in childhood. Lymphomas of bone occur more commonly in males, and although lymphoma of bone can occur anywhere in the skeletal system, it is most common in the femur and pelvis. The radiographic appearance of lymphoma of bone is usually that of a mixed lytic and sclerotic lesion. The sclerosis seen in the majority of lymphomas of bone probably constitutes reactive bone formation to the malignant process. In contrast to Ewing’s sarcoma, NHL of bone usually is a metaphyseal lesion in the long bones. At the time of presentation, there is usually a sparsity of soft tissue mass in contrast to primary Ewing’s sarcomas of bone (Fig. 121.28).

Figure 121.28. Non–Hodgkin’s lymphoma of bone presenting with a primary lesion in the distal femur in a 19-year-old female.

Figure 121.28

Non–Hodgkin’s lymphoma of bone presenting with a primary lesion in the distal femur in a 19-year-old female. This is a mixed lytic and sclerotic lesion with a preponderance of sclerosis, which can be typical of lymphomas of bone. Note (more...)

The histopathology of lymphoma of bone is similar to the histologic varieties of NHL encountered in other sites. When lesions are present in bone, however, by definition, they are of a diffuse, aggressive nature. Lymphoblastic morphology is characteristic of lymphoma of bone in the pediatric age range. Most lymphomas of bone are B-cell lymphomas. The histologic appearance may be confused with Ewing’s sarcoma since primary lymphomas also appear as small round cell tumors. With modern immunohistochemical techniques, the pathologic identification of NHL of bone and its differentiation from Ewing’s sarcoma is easily made. Lymphoma of bone should stain positively for common leukocyte antigen. Improper fixation can sometimes lead to erroneously negative histochemical staining, however.

The diagnosis of Ewing’s sarcoma in the older age group should be looked upon with suspicion. It is not uncommon for primary lymphoma of bone to present with regional nodal involvement. Ewing’s sarcoma of bone never has regional nodal involvement. Common leukocytic antigen on immunopathology is a distinguishing feature. Frequently, Ewing’s sarcomas contain small lymphocytic infiltrates that are positive for common leukocyte antigen in the normal lymphocytes that are incidentally found in the specimen, while the tumor cells are negative. This can serve as an internal control for the technical procedure.

The clinical presentation of NHL of bone is that of other malignant bone tumors and usually involves persistent pain as the predominant symptom. A large palpable mass is seldom present in extremity lesions, but pelvic lesions can produce large masses by the time they are diagnosed. Lesions in the metaphyseal area that are close to joints may present with a painful effusion in the joint. Even though primary bone lymphomas produce lytic lesions most of the time, they rarely present with a pathologic fracture, compared with other predominantly lytic bone lesions.

The staging for lymphoma of bone is identical to that of NHL of other sites. Primary NHL of bone, particularly in the younger age group, has a tendency to present with bone metastases at diagnosis. A total body bone scan is required to search for other skeletal metastases. In the younger age group, skeletal metastases tend to be symmetrical and most commonly involve the bones about the knee. Older patients may have regional nodal involvement at the time of diagnosis.

The treatment of lymphoma of bone is similar to that of NHL of other sites. The tumor is exquisitely sensitive to radiation therapy and chemotherapy. The frequency of local recurrence following radiation to the primary bone tumor is much lower than that found in Ewing’s sarcoma of bone. In most cases, primary radiation therapy in the range of 40 to 45 Gy is adequate for permanent local control.

NHL of bone is more malignant than was originally thought. An early report of 43% cure with local therapy alone dealt with a highly selected group, which excluded all but local lesions that had not spread in 6 months. 189 The unselected disease does not appear curable 43% of the time with local treatment.

Nevertheless, even though NHL of bone may be disseminated at the time of diagnosis, the prognosis with multi-disciplinary treatment including chemotherapy and local irradiation is excellent. Local irradiation should be given to the primary bone lesion, since the propensity for local recurrence is highest in that area if left unirradiated. Occasionally, resection of the involved segment, especially if it is weakened or fractured, is appropriate.

Malignant Vascular Lesions of Bone

The malignant vascular lesions of bone are rare, but they make up three distinct clinical entities, each of which has a different prognosis and requires different treatment.

Hemangiopericytoma of Bone

Malignant hemangiopericytoma of bone is characterized by a low to intermediate grade spindle cell stroma arranged in an irregular pattern consisting of proliferation around blood vessels.

The majority of patients are < 40 years of age. The tumor has a predilection for the long bones, pelvis, and scapula.

The radiographic appearance of a hemangiopericytoma is that of a malignant bone lesion but is nonspecific. It can occur in the diaphyseal as well as the metaphyseal area of the long bone.

It is important to differentiate this tumor histologically because it is malignant, but usually of only intermediate grade. It does not metastasize as readily as do other malignant vascular tumors of bone. On low-power microscopy, multiple irregularly shaped blood vessels can be seen, sometimes referred to as a “stag horn” vascular pattern. The stromal cells seem to proliferate around these blood vessels. The blood vessels are lined by normal endothelium in contrast to malignant angiosarcomas, where the vascular spaces are lined by malignant tumor.

These tumors usually do not respond well to chemotherapy or radiation therapy. 190 Surgery is the primary modality of treatment. Preoperative treatment in the form of embolization may shrink the tumor and make it less vascular, facilitating surgical resection.

Malignant Hemangioendothelioma of Bone

We prefer to use the term “malignant hemangioendothelioma of bone” for the form of angiosarcoma that is characterized by multi-focal occurrence usually in bones and soft tissue of the affected extremity. This tumor is rare below the age of 50 years. It is more common in the distal extremities and occurs multi-focally, usually only in the bones of the involved extremity, perhaps by retrograde venous regional metastases. Involvement of other bones in the skeletal system is uncommon in this multi-focal variety of angiosarcoma.

The radiographic appearance is usually that of a malignant-appearing lytic bony lesion that is hot on bone scan.

Histopathologically, these lesions are composed of plump, sometimes epithelioid, stromal cells. The stromal cells form anastomosing vascular channels. The lining of the vascular channels are the malignant stromal cells themselves and not normal endothelium. Sometimes, one can see malignant cells being shed into the vascular channels. The stromal cells of the tumor stain positively for Factor VIII. The differential diagnosis of weakened or fractured malignant hemangioendothelioma of bone includes other malignant lesions. However, if the disease is suspected in a very young child, it may be merely a proliferative hemangioma, a more common lesion in younger patients. Proliferative or active benign hemangiomas can have a very cellular phase, which closely mimics that of a malignant hemangioendothelioma. In particular, in very young children, the syndrome of hereditary multiple hemangiomatoses can give rise to what appear to be “metastatic bone lesions.” These lesions tend to regress spontaneously. Hemangiomas tend to have a sclerotic central appearance, while malignant hemangioendotheliomas are usually only lytic on radiographs.

The clinical history of malignant hemangioendothelioma is usually quite typical in that the patient develops recurrent lesions in the affected extremity. Eventually, after multiple recurrences, this tumor can give rise to distant metastases.

Staging of malignant hemangioendothelioma of the adult usually includes careful imaging of the entire affected extremity. For staging of this tumor, MRI is the procedure of choice. CT scans of the chest should also be done to rule out pulmonary metastases.

This tumor is relatively resistant to both combination chemotherapy and irradiation. The treatment of choice is surgical ablation. If radical surgery or amputation (if indicated by the presence of multi-focal lesions in the extremity) is performed early, the prognosis is excellent. 191

Malignant Angiosarcoma of Bone

The authors prefer to reserve this terminology for the rare highly malignant vascular tumors that can usually be recognized histopathologically. The malignant angiosarcoma of bone occurs predominantly in patients under 40 years of age, without predilection for gender. This lesion can affect any bone in the skeletal system but tends to arise in the axial skeleton, including the flat bones of the pelvis and skull, as well as the ribs and scapula in affected individuals.

The radiographic appearance of malignant angiosarcoma is that of a lytic punched out lesion in bone. Sometimes, the malignant vascular tumors of bone can be differentiated on radiographs from the more proliferative hemangiomas that sometimes produce stippled calcifications within the bony lytic lesion that give it a “sunburst” appearance. Malignant angiosarcomas of bone produce an almost purely lytic lesion in the bone. They expand and destroy the cortex, producing a soft tissue mass, which is not the case in benign hemangiomas of bone.

Malignant angiosarcomas of bone are highly malignant spindle cell, epithelioid, or small round cell tumors, the latter sometimes mimicking the appearance of Ewing’s sarcoma. There may only be subtle signs that one is not dealing with primary Ewing’s sarcoma: the vascular channels seen in angiosarcomas are composed of malignant tumor cells lining the vascular channels (rather than benign endothelium). Frequently, tumor cells can be seen to be shedding into the small vascular channels individually or in clumps (Fig. 121.29). These tumors stain positively for Factor VIII, which helps to differentiate them from Ewing’s sarcoma.

Figure 121.29. Angiosarcoma of bone.

Figure 121.29

Angiosarcoma of bone. The tumor consists of solid sheets of vasoformative polyhedral and spindle shaped cells (note numerous intracellular erythrocytes and small lumina). The mitotic rate is very high and numerous bizarre mitotic figures are identified (more...)

Angiosarcoma of bone is highly malignant. It is treated in a similar fashion to Ewing’s sarcoma of bone, with combination preoperative chemotherapy and radiation therapy, followed by en bloc excision of the lesional area and by postoperative adjuvant chemotherapy.

The malignant vascular tumors constitute three separate clinical entities. 190 The older literature usually lumped them together under the heading of angiosarcoma of bone. They require separate and distinct treatment, and only the highly malignant round small cell angiosarcoma of bone is obligatorily treated with systemic chemotherapy.


Link MP, Eilber E. Osteosarcoma. In: Pizzo PA, Poplack DG, editors. Principles and practice of pediatric oncology, 3rd ed. Philadelphia, PA: Lippincott-Raven; 1997. p. 889–920.
Grier H E. The Ewing family of tumors: Ewing’s sarcoma and primitive neuroectodermal tumors. Pediatr Clin North Am. 1997;44:991–1004. [PubMed: 9286296]
Pappo A S, Shapiro D N, Crist W M. Rhabdomyosarcoma: biology and treatment. Pediatr Clin North Am. 1997;44:953–972. [PubMed: 9286294]
Pappo A S, Shapiro D N, Crist W M, Maurer H N. Biology and therapy of pediatric rhabdomyosarcoma. J Clin Oncol. 1995;13:2123–2121. [PubMed: 7636557]
Horowitz MS, Malawer MM, Woo SY, Hicks HM. Ewing’s sarcoma family of tumors: Ewing’s sarcoma of bone. In: Pizzo PA, Poplack DG, editors. Principles and practice of pediatric oncology, 3rd ed. Philadelphia, PA: Lippincott-Raven; 1997. p. 831–888.
Meyers P A, Gorlick R. Osteosarcoma. Pediatr Clin North Am. 1997;44:973–989. [PubMed: 9286295]
Arndt C A S, Crist W M. Common musculoskeletal tumors of childhood and adolescence. N Engl J Med. 1999;341:342–352. [PubMed: 10423470]
Enneking W F, Spanier S S, Goodman M A. A system for the surgical staging of musculoskeletal sarcoma. Clin Orthop. 1980;(153):106–120. [PubMed: 7449206]
Freiberger R, Loitman B, Helpern M, Thompson T. Osteoid osteoma—a report of 80 cases. AJR Am J Roentgenol. 1959;82:194. [PubMed: 13670322]
Israeli A, Zwins S, Horosznowski H, Farine I. Use of radionuclide method in preoperative and intraoperative diagnosis of osteoid osteoma of the spine. Case report. Clin Orthop. 1983;175:194. [PubMed: 6220854]
Rosenthal D I, Hornicek F J, Wolfe M W. et al. Percutaneous radiofrequency coagulation of osteoid osteoma compared with operative treatment. J Bone Joint Surg (Am) 1998;80:815–821. [PubMed: 9655099]
Schajowicz F, Lemos C. Osteoid osteoma and osteoblastoma. Closely related entities of osteoblastic derivation. Acta Orthop Scand. 1970;41:272–291. [PubMed: 5486182]
Marcove R C, Alpert M A. Pathologic study of benign osteoblastoma. Clin Orthop. 1963;30:175–181. [PubMed: 5240469]
Marsh H O, Choi C B. Primary osteosarcoma of the cervical spine originally mistaken for benign osteoblastoma: a case report. J Bone Joint Surg (Am) 1970;52:1467–1471. [PubMed: 5273071]
Albright F, Buttler A M, Hampton A O, Smith P. Syndrome characterized by osteitis fibrosa disseminata, areas of pigmentation and endocrine dysfunction with precocious puberty in females. N Engl J Med. 1937;216:527.
Jaffee H L. Benign osteoblastoma. Bull Hosp Joint Dis. 1956;17:141–151. [PubMed: 13413389]
Mitchell M L, Ackerman L V. Metastatic and pseudomalignant osteoblastoma. A report of two unusual cases. Skeletal Radiol. 1986;15:213–218. [PubMed: 3704675]
Capanna R, Ayala A, Bertoni F. et al. Sacral osteoid osteoma and osteoblastoma: a report of 13 cases. Arch Orthop Trauma Surg. 1986;105:205–210. [PubMed: 3753174]
Marsh B W, Bonfiglio M, Brady L P, Enneking W F. Benign osteoblastoma: range of manifestations. J Bone Joint Surg Am. 1975;57:1–9. [PubMed: 1123354]
Jackson R P. Recurrent osteoblastoma: a review. Clin Orthop. 1978;(131):229–233. [PubMed: 657629]
Dorfman H D. Proceedings: Malignant transformation of benign bone lesions. Proc Natl Cancer Conf. 1972;7:901–913. [PubMed: 4764948]
Gibbons J M, Hammond G. Benign osteoblastoma. Lahey Clin Bull. 1963;13:97–103. [PubMed: 14132598]
Marcove R C, Weis L D, Vaghaiwalla M R. et al. Cryosurgery in the treatment of giant cell tumors of bone. A report of 52 consecutive cases. Cancer. 1978;41:957–969. [PubMed: 638982]
Kenan S, Floman Y, Robin G C, Laufer A. Aggressive osteoblastoma. A case report and review of the literature. Clin Orthop. 1985;(195):294–298. [PubMed: 3978963]
Schajowicz F, Lemos C. Malignant osteoblastoma. J Bone Joint Surg Br. 1976;58:202–211. [PubMed: 932083]
Mirra JM. In: Bone tumors: clinical, radiologic and pathologic correlations. Philadelphia, PA: Lea & Febiger; 1989. p. 155–173.
Mirra JM. In: Bone tumors: clinical, radiologic and pathologic correlations. Philadelphia, PA: Lea & Febiger; 1989. p. 191–225.
Tanner H C, Dahlin D C, Childs D S. Sarcoma complicating fibrous dysplasia. Probable role of radiation therapy. Oral Surg. 1961;14:837–846. [PubMed: 13775207]
Ruggieri P, Sim F H, Bond J R, Unni K K. Malignancies in fibrous dysplasia. Cancer. 1994;73:1411–1424. [PubMed: 8111708]
Yabut S M Jr, Kenan S, Sissons H A, Lewis M M. Malignant transformation of fibrous dysplasia. A case report and review of the literature. Clin Orthop. 1988;(228):281–289. [PubMed: 3277758]
Huvos A G, Higginbotham N L, Miller T R. Bone sarcomas arising in fibrous dysplasia. J Bone Joint Surg (Am) 1972;54:1047–1056. [PubMed: 4506620]
Perkinson N G, Higginbotham N L. Osteosarcoma arising in polyostotic fibrous dysplasia. Report of a case. Cancer. 1955;8:396–402. [PubMed: 14352179]
Schwartz T, Meyer A. The malignant transformation of fibrous dysplasia. Am J Med Sci. 1964;247:1–20. [PubMed: 14106874]
Eilber F, Giuliano A, Eckardt J. et al. Adjuvant chemotherapy for osteosarcoma: a randomized prospective trial. J Clin Oncol. 1987;5:21–26. [PubMed: 3543236]
Link M P, Goorin A, Miser J. et al. The effect of adjuvant chemotherapy on relapse-free survival in patients with osteosarcoma of the extremity. N Engl J Med. 1986;314:1600–1606. [PubMed: 3520317]
Dryja R P, Friend S, Weinberg R A. Genetic sequences that predispose to retinoblastoma and osteosarcoma. Symp Fundam Cancer Res. 1986;39:115–119. [PubMed: 3480547]
Miller C W, Aslo A, Won A. et al. Alterations in p53, Rb and MDM2 genes in osteosarcoma. J Cancer Res Clin Oncol. 1996;122:559–656. [PubMed: 8781571]
Ramanna L, Waxman A, Binney G. et al. Thallium-201 scintigraphy in bone sarcoma: comparison with gallium-67 and technetium-MDP in evaluation of chemotherapeutic response. J Nucl Med. 1990;31:567–572. [PubMed: 2341892]
Marcove R C, Mike V, Hajek J. et al. Osteosarcoma under the age of 21—a review of 145 operative cases. J Bone Joint Surg (Am) 1970;52:411–423. [PubMed: 5269156]
Malawer MM, Link MP, Donaldson SS. Sarcoma of bone. In: DeVita, Hellman and Rosenberg, editors. Cancer: principles and practice of oncology, 4th Edition, J.B. Lippincott; 1993. p. 1509–1567.
Jaffe N. Recent advances in the chemotherapy of metastatic osteosarcoma. Cancer. 1972;30:1627–1631. [PubMed: 4539306]
Rosen G, Nirenberg A, Caparros B, et al. Cisplatinum in metastatic osteosarcoma. In: Prestayko AW, Crooke ST, Carter SK, editors. Cisplatinum—current status and new developments. New York, NY: Academic; 1980. p. 465–475.
Cortes E P, Holland J F, Wang J J, Sinks L F. Doxorubicin in disseminated osteosarcoma. JAMA. 1972;221:1132–1138. [PubMed: 4512088]
Bacci G, Ferrari S, Mercuri M. et al. Predictive factors for local recurrence in osteosarcoma: 540 patients with extremity tumors followed for minimum 2.5 years after neoadjuvant chemotherapy. Acta Orthoped Scand. 1998;69:230–236. [PubMed: 9703394]
Jaffe N, Frei E, Traggis D, Watta H. Weekly high dose methotrexate-citrovorum factor in osteosarcoma. Cancer. 1977;39:45. [PubMed: 299828]
Skipper H E, Schabel F M Jr, Wilcox W S. Experimental evaluation of potential anti-cancer agents. XIII. On the criteria and kinetics associated with “curability” of experimental leukemia. Cancer Chemother Rep. 1964;35:1. [PubMed: 14117037]
Lewis M M. An approach to the treatment of malignant bone tumors. Orthopedics. 1985;8:655–656. [PubMed: 3867863]
Jain R K. Delivery of novel therapeutic agents in tumors: physiological barriers and strategies. J Natl Cancer Inst. 1989;81:570–576. [PubMed: 2649688]
Jaffe N, Frei E, Traggis D, Bishop Y. Adjuvant methotrexate and citrovorum-factor treatment of osteogenic sarcoma. N Engl J Med. 1974;291:994–997. [PubMed: 4606174]
Rosenberg S A, Chabner B A, Young R C. et al. Treatment of osteogenic sarcoma. I. Effect of adjuvant high-dose methotrexate after amputation. Cancer Treat Rep. 1979;83:739–751. [PubMed: 313245]
Cortes E P, Holland J F, Wang J J. et al. Amputation and Adriamycin in primary osteosarcoma. N Engl J Med. 1974;291:998–1000. [PubMed: 4528415]
Cortes E, Necheles TF, Holland JF, et al. Adjuvant chemotherapy for primary osteosarcoma: a Cancer and Leukemia Group B experience. In: Salmon S, Jones S, editors. Adjuvant therapy of cancer. III. New York, NY: Grune & Stratton; 1981. p. 201–210.
Goorin A, Delorey M, Gelber R D. et al. The Dana-Farber Cancer Institute/The Children’s Hospital adjuvant chemotherapy trials for osteosarcoma: three sequential studies. Cancer Treat Rep. 1986;3:155–159.
Goorin A, Perez-Atayde A, Gebhardt M. et al. Weekly high dose methotrexate and doxorubicin for osteosarcoma. The Dana-Farber Cancer Institute/The children’s Hospital—study III. J Clin Oncol. 1987;5:1178–1184. [PubMed: 3476688]
Pratt C B, Champion J E, Fleming I D. et al. Adjuvant chemotherapy for osteosarcoma of the extremity. Long-term results of two consecutive prospective protocol studies. Cancer. 1990;65:439–445. [PubMed: 2297634]
Link M P, Goorin A M, Miser A W. et al. The effect of adjuvant chemotherapy on relapse-free survival in patients with osteosarcoma of the extremity. N Engl J Med. 1986;314:1600–1606. [PubMed: 3520317]
Link MP, Shuster JJ, Goorin AM, et al. Adjuvant chemotherapy in the treatment of osteosarcoma: results of the Multi-institutional Osteosarcoma Study. In: Ryan J, Baker LO, editors. Recent concepts in sarcoma treatment. Proceedings of the International Symposium on Sarcomas. Tarpon Springs, Florida. October 8–10, 1987. Dorecht, The Netherlands: Kluwer Academic Publishers; 1988. p. 283–290.
Eilber R, Giuliano A, Eckardt J. et al. Adjuvant chemotherapy for osteosarcoma: a randomized prospective trial. J Clin Oncol. 1987;5:21–26. [PubMed: 3543236]
Rosen G, Murphy M L, Huvos A G. et al. Chemotherapy, en bloc resection and prosthetic bone replacement in the treatment of osteogenic sarcoma. Cancer. 1976;37:1–11. [PubMed: 1082364]
Kalifa C, Mlika N, Dubousset J. et al. Experience with the T-10 protocol in the pediatric service at the Gustave-Roussy Institute. Bull Cancer (Paris) 1988;75:207–211. [PubMed: 2451956]
Rosen G. Preoperative (neoadjuvant) chemotherapy for osteogenic sarcoma: a ten year experience. Orthopedics. 1985;8:658–664. [PubMed: 3867864]
Rosen G, Nirenberg A. Neoadjuvant chemotherapy for osteogenic sarcoma: a five-year follow-up (T-10) and preliminary report of new studies (T-12). In: Wagner DJ, Blijham GH, Smeets JBE, Wils JA, editors. Primary chemotherapy in cancer medicine. New York, NY: Alan R. Liss; 1985. p. 39–52. [PubMed: 3867902]
Huvos A G, Rosen G, Marcove R C. Primary osteosarcoma: pathologic aspects in 20 patients after treatment with chemotherapy, en bloc resection and prosthetic bone replacement. Arch Pathol Lab Med. 1977;101:14–18. [PubMed: 299812]
Stoller R G, Hande K R, Jacobs S A. et al. Use of plasma pharmacokinetics to predict and prevent methotrexate toxicity. N Engl J Med. 1977;297:630–634. [PubMed: 302412]
Ahuja S C, Viljacin A B, Smith J. et al. Juxtacortical (parosteal) osteosarcoma. Histological grading and prognosis. J Bone Joint Surg (Am) 1977;59:632–647. [PubMed: 267053]
Rosen G. Preoperative (neoadjuvant) chemotherapy for osteosarcoma. A ten-year experience. Orthopedics. 1985;8:659–664. [PubMed: 3867864]
Herendeen R E. Results in the roentgen-ray therapy of giant-cell tumor of bone. Ann Surg. 1931;93:398–411. [PMC free article: PMC1398781] [PubMed: 17866487]
Phaler G F, Parry L D. The treatment of giant-cell bone tumors by roentgen irradiation. AJR Am J Roentgenol. 1932;28:151–166.
Herendeen R E. Roentgen-ray treatment of giant cell tumor. AJR Am J Roentgenol. 1924;12:117–125.
McGrath P J. Giant-cell tumor of bone: an analysis of fifty-two cases. J Bone Joint Surg Br. 1972;54:216–229. [PubMed: 5034822]
Dahlin D C. Caldwell Lecture. Giant cell tumor of bone: highlights of 407 cases. AJR Am J Roentgenol. 1985;144:955–960. [PubMed: 3872579]
Upchurch K S, Simon L S, Schiller A L. et al. Giant cell reparative granuloma of Paget’s disease of bone: a unique clinical entity. Ann Intern Med. 1983;98:35–40. [PubMed: 6848040]
Russell D S. Malignant osteoclastoma and the association of malignant osteoclastoma with Paget’s ostitis deformans. J Bone Joint Surg (Am) 1949;31B:281–290. [PubMed: 18145165]
Huvos A G, Marcove R C. Adamantinoma of long bones. A clinico-pathological study of 14 cases with vascular origin suggested. J Bone Joint Surg (Am) 1975;57:148–154. [PubMed: 1112840]
Robinson H B G. Ameloblastoma. Survey of three hundred and seventy-nine cases from the literature. Arch Pathol. 1937;23:831–862.
Lubschitz K. Adamantinomas of the jaw, with reference especially to their management. Acta Radiol. 1945;26:441–445.
Fitzgerald G W, Frenkel S, Black M J. et al. Ameloblastoma of the jaws: a 12 year review of the McGill experience. J Otolaryngol. 1982;11:23–28. [PubMed: 7077721]
Keeny G L, Unni K, Beabout J W. et al. Adamantinoma of long bones. A clinicopathologic study of 85 cases. Cancer. 1989;64:730–737. [PubMed: 2743266]
Huvos A G, Marcove R C. Adamantinoma of long bones. A clinico-pathological study of 14 cases with vascular origin suggested. J Bone Joint Surg (Am) 1975;57:148–154. [PubMed: 1112840]
Pieterse A, Smith P S, McClure J. Adamantinoma of long bones: clinical, pathological and ultrastructural features. J Clin Pathol. 1982;35:780–786. [PMC free article: PMC497777] [PubMed: 7201477]
Campanacci M, Guinti A, Bertoni F. et al. Adamantinoma of the long bones. The experience at the Istituto Ortopedico Rizzoli. Am J Surg Pathol. 1981;5:533–542. [PubMed: 7325273]
Czerniak B, Rojas-Corona R R, Dorfman H D. Morphologic diversity of long bone adamantinoma. The concept of differentiated (regressing) adamantinoma and its relationship to osteofibrous dysplasia. Cancer. 1989;64:2319–2334. [PubMed: 2804923]
Goldenberg R R, Campbell C J, Bonfiglio M. Giant-cell tumor of bone. An analysis of two hundred and eighteen cases. J Bone Joint Surg Am. 1970;52:619–664. [PubMed: 5479455]
Dahlin D C, Cupps R E, Johnson E W. Giant cell tumor: a study of 195 cases. Cancer. 1970;25:1061–1070. [PubMed: 4910256]
Bennett C J, Marcus R B, Millian R R, Enneking W F. Radiation therapy for giant cell tumor of bone. Int J Radiat Oncol Biol Phys. 1993;26:299–304. [PubMed: 8491687]
Bertoni F, Unni K K, Beabout J W, Ebersold M J. Giant cell tumor of the skull. Cancer. 1992;70:1124–1132. [PubMed: 1515987]
Berman H L. The treatment of benign giant cell tumor of the vertebrae by irradiation. Radiology. 1964;83:202–207. [PubMed: 14207410]
Friedman M, Pearlman A. Benign giant-cell tumor of bone: radiation dosage for each type. Radiology. 1968;91:1151–1158. [PubMed: 5699616]
Fain J S, Unni K K, Beabout J W, Rock M G. Non-epiphyseal giant cell tumor of the long bones. Cancer. 1993;71:3514–3519. [PubMed: 8490899]
Schwartz L H, Okunieff P G, Rosenberg A, Suit H D. Radiation therapy in the treatment of difficult giant cell tumors. Int J Radiat Oncol Biol Phys. 1989;17:1085–1088. [PubMed: 2808042]
Oda Y, Tsuneyoshi M, Shinosharia N. Solid variant of aneurysmal bone cyst (extragnathic giant cell reparative granuloma) in the axial skeleton and long bones. Cancer. 1992;70:2642–2649. [PubMed: 1423194]
Markow R L, Boneal M, Ingles A E. Giant cell reparative granuloma in the hand: report of three cases and review of the literature. J Hand Surg (Am) 1985;10:733–739. [PubMed: 4045162]
Wold L E, Dobyns J H, Swee R G, Dahlin D C. Giant cell reaction (giant cell reparative granuloma) of the small bones of the hands and feet. Am J Surg Pathol. 1986;10:491–496. [PubMed: 3728789]
Chen Z X, Gu D Z, Yu Z H. et al. Radiation treatment of giant-cell tumor of bone. J Bone Joint Surg (Am) 1984;66A:2698.
Rock M G, Pritchard D J, Unni K K. Metastases from histologically benign giant-cell tumor of bone. J Bone Joint Surg Am. 1984;66:269–274. [PubMed: 6693454]
Sara A S, Ayala A G, El Naggar A. et al. Giant cell tumor of bone. A clinicopathologic and DNA flow cytoorectiv analysis. Cancer. 1990;66:2186–2190. [PubMed: 2224773]
Pantazopolous P E, Nomikos N. Adamantinoma of the maxillary sinus. Ann Otol Rhinol Laryngol. 1966;98:1160–1168. [PubMed: 5957984]
Hutter R V P, Worcester J N, Frances K C. et al. Benign and malignant giant cell tumors of bone: clinicopathological analysis of the natural history of disease. Cancer. 1962;15:653–690. [PubMed: 14450279]
Dahlin DC Giant cell tumor of cancer.1977;39:1350-1356. [PubMed: 912665]
Moon N F, Mori H. Adamantinoma of the appendicular skeleton—updated. Clin Orthop. 1986;(204):215–237. [PubMed: 3514033]
Chaudhuri P. Ameloblastoma of the upper jaw. J Laryngol Otol. 1975;89:457–464. [PubMed: 1127339]
Sampson H E. Basal-cell carcinoma and peripheral ameloblastoma. Oral Surg Oral Med Oral Pathol. 1974;38:241–253. [PubMed: 4528580]
Laughlin E H. Metastasizing ameloblastoma. Cancer. 1989;64:776–780. [PubMed: 2663133]
Shaw H J, Katsikas D K. Ameloblastoma of the maxilla. A clinical study with four cases. J Laryngol Otol. 1973;87:873–884. [PubMed: 4728301]
Weiss S W, Dorman H D. Adamantinoma of long bone. An analysis of nine new cases with emphasis on metastasizing lesions and fibrous dysplasia-like changes. Hum Pathol. 1977;8:141–153. [PubMed: 852865]
Vickers R A, Gorlin R J. Ameloblastoma. Delineation of early histopathologic features of neoplasia. Cancer. 1970;26:699–710. [PubMed: 5458275]
Gardner D G, Pecak A M. The treatment of ameloblastoma based on pathologic and anatomic principles. Cancer. 1980;46:2514–2519. [PubMed: 7438024]
Bredenkamp J K, Zimmerman M C, Mickel R A. Maxillary ameloblastoma. A potentially lethal neoplasm. Arch Otolaryngol Head Neck Surg. 1989;115:99–104. [PubMed: 2642382]
Bonta J A, James A G. Adamantinoma. Ann Surg. 1955;142:1007–1012. [PMC free article: PMC1464976] [PubMed: 13269061]
Small I A, Waldron C A. Ameloblastoma of the jaws. Oral Surg Oral Med Oral Pathol. 1955;8:281–297. [PubMed: 14356722]
Gardner D G. Radiotherapy in the treatment of ameloblastoma. Int J Oral Maxillofac Surg. 1988;17:201–205. [PubMed: 3135350]
McWhirter R. Symposium on the treatment of adamantinoma. Proc R Soc Med. 1952;45:701–707. [PMC free article: PMC1987607] [PubMed: 13003976]
Sehder M K, Huvos A G, Strong E W. et al. Ameloblastoma of maxilla and mandible. Cancer. 1974;33:324–333. [PubMed: 4812754]
Atkinson C H, Harwood A R, Cumings B J. Ameloblastoma of the jaw. A reappraisal of the role of megavoltage irradiation. Cancer. 1984;53:869–873. [PubMed: 6420036]
Pandya N J, Stuterville O H. Treatment of ameloblastoma. Plast Reconstr Surg. 1972;50:242–248. [PubMed: 4115148]
Hair J A G. Radiosensitive adamantinoma. Br Med J. 1963;1:105–106. [PMC free article: PMC2122589] [PubMed: 13951996]
Ikemura K, Tashiro H, Fujino H. et al. Ameloblastoma of the mandible with metastasis to the lungs and lymph nodes. Cancer. 1972;29:930–940. [PubMed: 5017362]
Huvos A G. Primary malignant fibrous histiocytoma of bone; a clinicopathologic study of 18 patients. N Y State J Med. 1976;76:552–559. [PubMed: 177910]
Urban C, Rosen G, Huvos A G. et al. Chemotherapy of malignant fibrous histiocytoma of bone. A report of 5 cases. Cancer. 1983;51:795–802. [PubMed: 6295615]
McCarthy E F, Matsuno T, Dofman H D. Malignant fibrous histiocytoma of bone: a study of 35 cases. Hum Pathol. 1979;10:57–70. [PubMed: 218883]
Mirra J M, Bullough P G, Marcove R C. et al. Malignant fibrous histiocytoma and osteosarcoma in association with bone infarcts. J Bone Joint Surg Am. 1974;56:932–940. [PubMed: 4367886]
Galli S J, Weinstraub H P, Proppe K H. Malignant fibrous histiocytoma and pleomorphic sarcoma in association with medullary bone infarcts. Cancer. 1978;41:607–619. [PubMed: 204409]
Huvos A G. Primary malignant fibrous histiocytoma of bone. A clinico-pathologic study of 18 patients. N Y State J Med. 1976;76:552–559. [PubMed: 177910]
Urban C, Rosen G, Huvos A G. et al. Chemotherapy of malignant fibrous histiocytoma of bone. A report of 5 cases. Cancer. 1983;51:795–802. [PubMed: 6295615]
Chawla SP, Benjamin RS, Abdul-Karim FW, et al. Adjuvant chemotherapy of primary malignant fibrous histiocytoma of bone—prolongation of disease-free survival and overall survival. Adj Therapy Cancer 1984;V:621–629.
Campanna R, Bertoni F, Bacchini P. et al. Malignant fibrous histiocytoma of bone. The experience at the Rizzoli Institute: report of 90 cases. Cancer. 1984;54:177–187. [PubMed: 6326996]
Dahlin D C, Unni K K, Matsuno T. Malignant (fibrous) histiocytoma of bone—fact or fancy? Cancer. 1977;39:1508–1516. [PubMed: 192432]
Spanier S, Enneking W F, Enriquez P. Primary malignant fibrous histiocytoma of bone. Cancer. 1975;36:2084–2098. [PubMed: 173460]
Feldman F, Lattes R. Primary malignant fibrous histiocytoma (fibrous xanthoma) of bone. Skel Radiol. 1977;1:145–160.
Bacci G, Avella M, Picci P. Primary chemotherapy and delayed surgery of malignant fibrous histiocytoma of bone in the extremity. Tumori. 1990;76:537–542. [PubMed: 2178284]
Solomon L. Chondrosarcoma in hereditary multiple exostoses. S Afr Med J. 1974;48:671–676. [PubMed: 4545083]
Mirra J M, Gold R, Downs J, Eckardt J J. A new histologic approach to the differentiation of enchondroma from chondrosarcoma of the bones. A clinicopathologic analysis of 51 cases. Clin Orthop. 1985;(201):214–237. [PubMed: 4064409]
Dahlin D C, Ivins J C. Benign chondroblastoma. A study of 125 cases. Cancer. 1972;30:401–413. [PubMed: 5051664]
Evans H L, Ayala A G, Romsdahl M M. Prognostic factors in chondrosarcoma of bone. A clinico-pathologic analysis with emphasis on histologic grading. Cancer. 1977;40:818–831. [PubMed: 890662]
Marcove R C, Mike V, Hutter R V P. et al. Chondrosarcoma of the pelvis and upper end of the femur. An analysis of factors influencing survival time in 113 cases. J Bone Joint Surg (Am) 1972;54:561–572. [PubMed: 5055153]
Mirra J M, Marcove R C. Fibrosarcomatous dedifferentiation of primary and secondary chondrosarcoma. A review of 5 cases. J Bone Joint Surg (Am) 1974;56:285–296. [PubMed: 4452687]
Rosen G. Spindle cell sarcomas of bone, dedifferentiated, conventional and mesenchymal chondrosarcomas. In: Schein P, editor. Decision making in oncology. Philadelphia, PA: Decker; 1989. p. 170–173.
Evans H L, Ayala A G, Romsdahl M M. Prognostic factors in chondrosarcoma of bone. A clinico-pathologic analysis with emphasis on histologic grading. Cancer. 1977;40:818–831. [PubMed: 890662]
Harwood A R, Krajbich J I, Fornasier V L. Radiotherapy of chondrosarcoma of bone. Cancer. 1980;45:2769–2777. [PubMed: 7379008]
McNaney D, Lindberg R D, Ayala A G. et al. Fifteen year radiotherapy experience with chondrosarcoma of bone. J Radiat Oncol Biol Phys. 1982;8:187–190. [PubMed: 6806219]
Harwood A R, Krajbich J I, Fornasier V L. Mesenchymal chondrosarcoma: a report of 17 cases. Clin Orthop. 1981;(158):144–148. [PubMed: 7273513]
Huvos A G, Rosen G, Dabska M, Marcove R C. Mesenchymal chondrosarcoma. A clinico-pathologic analysis of 35 patients with emphasis on treatment. Cancer. 1983;51:1230–1237. [PubMed: 6825046]
Richter M P, Laramore G E, Griffin T W, Goodman R L. Current status of high linear energy transfer irradiation. Cancer. 1984;54:2814–2822. [PubMed: 6437661]
Mark R J, Tran L M, Fu Y S. et al. Chondrosarcoma of the head and neck. The UCLA experience, 1955–1988. Am J Clin Oncol. 1993;16:232–237. [PubMed: 8338058]
Paddison G M, Hanks G H. Chondrosarcoma of the maxilla. Report of a case responding to supervoltage irradiation and review of the literature. Cancer. 1971;28:616–619. [PubMed: 4999570]
Pritchard D J, Taylor W F, Dahlin D C, Medley B E. Chondrosarcoma: a clinicopathologic and statistical analysis. Cancer. 1980;45:149–157. [PubMed: 7350999]
Suit H D, Gotein M, Munzenrider J. et al. Definitive radiation therapy for chordoma and chondrosarcoma of base of skull and cervical spine. J Neurosurg. 1982;56:377–385. [PubMed: 7057235]
Suit H D. Progress in low LET heavy particle therapy: intracranial and pericranial tumors and uveal melanomas. Radiat Res. 1985;104:S219–S226. [PubMed: 3003784]
Austin J P, Urie M M, Cardenosa M D, Munzenrider J E. Probable causes of recurrence in patients with chordoma and chondrosarcoma of the base of skull and cervical spine. Int J Radiat Oncol Biol Phys. 1993;25:439–444. [PubMed: 8436522]
Berson A M, Castro J R, Petti P. et al. Charged particle irradiation of chordoma and chondrosarcoma of the base of skull and cervical spine: the Lawrence Berkeley Laboratory Experience. Int J Radiat Oncol Biol Phys. 1988;15:559–565. [PubMed: 3138208]
Castro J R, Lindstadt D E, Bahary J P. et al. Experience in charged particle irradiation of tumors of the skull base: 1977–1992. Int J Radiat Oncol Biol Phys. 1994;29:647–655. [PubMed: 8040010]
Marks R D, Scruggs H J, Wallace K M, Fenn J O. Megavoltage therapy in patients with aneurysmal bone cyst. Radiology. 1976;118:421–424. [PubMed: 814583]
Dabska M, Buraczewski J. Aneurysmal bone cyst. Cancer. 1968;23:371–384. [PubMed: 5764977]
Lichtenstein L. Aneurysmal bone cyst. Observation on 50 cases. J Bone Joint Surg (Am) 1957;39A:873–882. [PubMed: 13438943]
Brock E H, Bogart F B. Benign giant cell tumor of the spine. Report of a case occurring in the cervical spine. AJR Am J Roentgenol. 1945;54:512–518. [PubMed: 21005150]
Oberman H A. Idiopathic histiocytoses. A clinicopathologic study of 40 cases and review of the literature on eosinophilic granuloma, Hand-Schuller-Christian disease and Letterer-Siwe disease. Pediatrics. 1961;38:307–328. [PubMed: 13730162]
Biesecker J L, Marcove R C, Huvos A G, Mike V. Aneurysmal bone cysts. A clinico-pathologic study of 66 cases. Cancer. 1970;26:615–625. [PubMed: 5458269]
Tillman B, Dahlen D C, Lipscomb P R, Stewart F R. Aneurysmal bone cyst. An analysis of 95 cases. Mayo Clin Proc. 1968;43:478–495. [PubMed: 5710584]
Unni KK. Dahlin’s bone tumors: general aspects and data on 11,080 cases, 5th ed. Philadelphia, PA: Lippincott-Raven; 1996. p. 144–250.
Parrish F F, Pevey J K. Surgical management of aneurysmal bone cyst of the vertebral column. J Bone Joint Surg Am. 1967;49:1597–1604. [PubMed: 6065453]
Lichtenstein L. Aneurysmal bone cyst. Further observations. Cancer. 1953;6:1228–1237. [PubMed: 13106840]
Donaldson W F. Aneurysmal bone cyst. J Bone Joint Surg (Am) 1962;44A:25–40.
Martinez V, Sessone H A. Aneurysmal bone cyst. A review of 123 cases including primary lesions and those secondary to other bone pathology. Cancer. 1988;61:2291–2304. [PubMed: 2835141]
Saltzman E I, Jun M Y. Aneurysmal bone cyst of the mandible: report of a case. J Surg Oncol. 1981;17:385–394. [PubMed: 7265978]
Komp D M. Historical perspectives of Langerhans’ cell histiocytosis. Hematol Oncol Clin North Am. 1987;1:9–21. [PubMed: 3312151]
Sosman M C. Xanthomatosis (Schuller’s disease; Christian’s syndrome). Report of three cases treated with roentgen rays. AJR Am J Roentgenol. 1930;23:581–597.
Wallace W S. Reticulo-endotheliosis: Hand-Schuller-Christian disease and the rarer manifestations. AJR Am J Roentgenol. 1949;62:189–207. [PubMed: 18136832]
McGavran M H, Spady H A. Eosinophilic granuloma of bone. A study of twenty-eight cases. J Bone Joint Surg (Am) 1960;424:974–992.
Wormer R B, Raney R B, D’Angio G J. Healing rates of treated and untreated bone lesions in histiocytosis X. Pediatrics. 1985;76:286–288. [PubMed: 3875075]
Nesbit M E, Kieffer S, D’Angio G J. Reconstitution of vertebral height in histiocytosis X: long-term follow-up. J Bone Joint Surg Am. 1969;51:1360–1368. [PubMed: 5823199]
Slater J M, Swarm O J. Eosinophilic granuloma of bone. Med Pediatr Oncol. 1980;8:151–164. [PubMed: 6999317]
Perslegin I A, Ustinove V F, Podlyashuk E L. Radiotherapy for eosinophilic granuloma of bone. Int J Radiat Oncol Biol Phys. 1981;7:317–321. [PubMed: 7275709]
Greenberger J S, Crocker A C, Vawter G. et al. Results of treatment of 127 patients with systemic histiocytosis (Letterer-Siwe syndrome, Schuller-Christian syndrome and multifocal eosinophilic granuloma) Medicine. 1981;60:311–338. [PubMed: 6974293]
Ridley M M, Raney R B, Thawerani H, Meadows A T. Histiocytosis X in children: patterns of disease and results of treatment. Med Pediatr Oncol. 1983;11:99–105. [PubMed: 6601232]
Selch M T, Parker R G. Radiation therapy in the management of Langerhans’ cell histiocytosis. Med Pediatr Oncol. 1990;18:97–102. [PubMed: 2304426]
Hartman K S. Histiocytosis X: a review of 114 cases with oral involvement. Oral Surg Oral Med Oral Pathol. 1980;49:38–54. [PubMed: 6965322]
Chase D C, Eversol L R, Hall H D. Histiocytosis X with jaw involvement. J Oral Surg. 1974;32:494–502. [PubMed: 4525417]
Aronson C L, Donaldson S S. Langerhans’ cell histiocytosis of the head and neck. Laryngoscope. 1987;97:537–542. [PubMed: 3494895]
Avery M E, McAlfee J C, Guild H G. The course and prognosis of reticuloendothelioses (eosinophilic granuloma, Schuller-Christian disease, Letterer-Siwe disease) Am J Med. 1957;22:636–652. [PubMed: 13410955]
Sbarboro J L, Francis K C. Eosinophilic granuloma of bone. JAMA. 1961;178:114–118.
McCullough C J. Eosinophilic granuloma of bone. Acta Orthop Scand. 1980;51:389–398. [PubMed: 6969513]
Raney R B, D’Angio G J. Langerhans’ cell histiocytosis experience at the Children’s Hospital of Philadelphia, 1970–1984. Med Pediatr Oncol. 1989;17:20–28. [PubMed: 2783619]
Smith D G, Nesbit M E, D’Angio G J D, Levett S H. Histiocytosis X. Role of radiation therapy in management with special reference to dose levels employed. Radiology. 1973;106:419–422. [PubMed: 4684485]
Greenberger J S, Cassady J R, Jaffee N. et al. Radiation therapy in patients with histiocytosis: management of diabetes insipidus and bone lesions. Int J Radiat Oncol Biol Phys. 1979;5:1749–1755. [PubMed: 316811]
Righter M P, D’Angio G J. The role of radiation therapy in the management of children with histiocytosis X. Am J Pediatr Hematol Oncol. 1981;3:161–163. [PubMed: 7030100]
Cassady J R. Current role of radiation therapy in the management of histiocytosis-X. Hematol Oncol Clin North Am. 1987;1:123–129. [PubMed: 3312144]
Griffin T W. The treatment of advanced histiocytosis-X with sequential hemebody irradiation. Cancer. 1977;39:2435–2436. [PubMed: 301431]
Askin F B, Rosai J, Sibley R. et al. Malignant small cell tumor of the thoracopulmonary region in childhood. Cancer. 1979;43:2438. [PubMed: 222426]
Winkler K, Beron G, Delling G. et al. Neoadjuvant chemotherapy of osteosarcoma: results of a randomized cooperative trial (COSS-82) with salvage chemotherapy based on histologic tumor response. J Clin Oncol. 1988;6:329–337. [PubMed: 2448428]
Wold L E, Unni K K, Sim F H, Dahlin D E. Hemangiopericytoma of bone. Am J Surg Pathol. 1982;6:53. [PubMed: 7081534]
Huvos AG. Bone tumors—diagnosis, treatment, and prognosis. Philadelphia, PA: WB Saunders; 1979. p. 358–370.
© 2000, BC Decker Inc.
Bookshelf ID: NBK20801


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...