Secondary Logo

Share this article on:

How to Diagnose Malignant Giant Cell Tumor

Unni, K. Krishnan MB, BS


The term “malignant giant cell tumor” has been used traditionally to describe a sarcoma arising from a giant cell tumor. However, the term has led to much confusion resulting in the inclusion of many giant cell rich sarcomas not related to giant cell tumor in the group. The term “malignancy in giant cell tumor” has the advantage of overcoming this problem in semantics. It is defined as a sarcoma arising in association with a well documented giant cell tumor, either synchronously or at the site of a previously documented giant cell tumor.

From the Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN.

Address correspondence and reprint requests to K. Krishnan Unni, MB, BS, Professor of Anatomic Pathology, Department of Laboratory Medicine and Pathology, Mayo Clinic 200 First Street, S.W., Rochester, MN 55905.

Without question, the most important task facing a surgical pathologist is the distinction between a benign and a malignant tumor. A diagnosis of “cancer” evokes a sense of hopelessness in most patients and hence is fraught with anxiety. Both under-diagnoses and over-diagnoses may lead to disastrous consequences to the patient.

However, this task is not often straightforward. Examples abound in all organ systems in which the pathologist is unsure about the true nature of the lesion. Pathologists are forever seeking diagnostic criteria which would allow us to confidently separate a benign neoplasm with an expectation of cure from a malignant one which may (but not invariably) lead to the death of the patient. This difficulty in separating “benign” from “malignant” has lead to such unfortunate terms as “borderline malignant” and tumors of “uncertain malignant potential.” All tumors are of uncertain malignant potential.

Even the definition of malignancy is not straightforward. A benign tumor may actually be incurable (primarily because of the location, such as the brain) and kill the patient, whereas a malignant tumor may be curable. Most pathologists and oncologists would accept the capacity to produce distant metastases as a definition of “malignancy” in a neoplasm. However, even this definition is not infallible. There are rare but well documented examples of benign neoplasms which have produced distant metastases and even led to the death of the patient. These rare tumors are benign in their biologic evolution in the vast majority of cases, and the rare ones which produce metastases do not appear histologically different from those which do not. There are examples of benign metastasizing tumors in many organ systems such as salivary glands (benign mixed tumor) and the genitourinary system (leiomyoma).

One problem with the definition of malignant giant cell tumor is the occurrence of benign metastasizing giant cell tumors. Dahlin 1 described his experience with 407 cases of giant cell tumor seen at the Mayo Clinic (Figure 1A–C). Eight of these patients developed pulmonary metastases, and two of these died of tumor. Campanacci et al 2 reported on 293 patients with giant cell tumors seen at the Rizzoli Institute. Three patients developed histologically confirmed pulmonary metastases. No patient died of the disease. Three additional patients had roentgenographic evidence of pulmonary metastases, but these were not proven histologically. These patients were all alive and showed features of regression, clinically and roentgenographically. Siebenrock et al 3 reported on 31 patients with benign metastasizing giant cell tumors from the Mayo Clinic files. In this series, 33% of these patients had origin of the giant cell tumor in the radius. Local recurrence, either prior to or at the time of metastases, occurred in 88%. Latent period to metastases was 0 to 24 years. In 80% of patients, there were multiple metastatic nodules. Seven patients died; only four deaths were related to tumor. The reason for diagnosing a benign tumor, even with metastases, are: 1) the histologic features of the tumors which do metastasize are identical to those of the vast majority of tumors which do not metastasize; and 2) the death rate is low. Most patients are treated with surgical resection of the metastatic nodules. Pulmonary metastases from a sarcoma would almost certainly prove fatal.



Jaffe and Lichtenstein 4 tried to devise a histologic grading system which would predict prognosis in giant cell tumor. They divided giant cell tumor of bone into: grade 1 - no appreciable atypism of the stromal cells, mitoses are few, and none abnormal; grade 2–stromal cells may be slightly or strikingly atypical but not sufficiently to diagnose frank malignancy, abnormal mitotic figures may be found; and grade 3–frankly and obviously malignant with capacity to metastasize.

Sanerkin 5 performed a detailed study of 86 giant cell tumors from the Bristol Bone Tumor Registry and divided them into three groups: (1) conventional–according to the definitions provided by Drs. Jaffe and Lichtenstein and graded 1 according to these criteria); (2) borderline–tumors without frankly sarcomatous stroma but with vascular permeation and abnormal mitotic figures; (3) malignant tumors–grade 3 of Jaffe and Lichtenstein. Dr. Sanerkin found 16 tumors in the third category. Two of these were reclassified as benign giant cell tumor. Two were reclassified as osteoclast-rich osteosarcoma, and one was reclassified as a metastatic carcinoma with giant cells. Three tumors were considered to be borderline. Of the remaining eight, four were postradiation and hence, only four were considered intrinsically malignant giant cell tumor (Dr. Sanerkin does not count postradiation sarcoma in giant cell tumor as malignant giant cell tumor). All four of the patients with malignant giant cell tumor died of tumor. No benign tumor metastasized. Mitotic figures did not help in classifying the tumors.

I define malignancy in giant cell tumor as a sarcoma which arises in juxtaposition to a benign giant cell tumor or at the site of a previously documented giant cell tumor. Hence, these tumors can be classified as primary and secondary.

Most malignancies are secondary. In the series from the Mayo Clinic reported in 1994, 6 there were a total of 34 malignant giant cell tumors. Thirty were secondary. Twenty-four were postradiation (Sanerkin would not include these in the group). Six of the secondary malignant tumors occurred after surgical treatment alone (Figures 2a,b,c). There were only five primary malignant giant cell tumors. In these, clinical and roentgenographic features suggested a diagnosis of giant cell tumor (Figure 3a). Histologically, there were distinct areas of typical giant cell tumor juxtaposed to high grade spindle cell sarcoma (Figure 3b).





The largest series of primary malignant giant cell tumors was described by Nascimento et al. 7. They described eight cases defined as tumors showing characteristic features of giant cell tumor but with many fields showing sarcomatous features. Tumors in which only small foci showed atypical cells were not included. Only one of the patients died of tumor, and the authors suggested that the prognosis in the primary type of malignant giant cell tumor was better than for the secondary type.

Meis et al 8 described one case of a tumor in the ischium of a 44-year-old man which showed malignant fibrous histiocytoma juxtaposed to typical giant cell tumor. They suggested using the term dedifferentiated giant cell tumor.

Rock and coauthors 9 described 19 cases of secondary malignant giant cell tumor from the Mayo Clinic files (Figure 4). Twelve of these patients died of tumor. Gitelis and coauthors 10 described one case of fibrosarcoma occurring at the site of a previously surgically treated giant cell tumor. These authors pointed out that the latent period can be extremely long.



The problem is complicated by two factors. Some giant cell tumors will demonstrate the presence of anaplasia in the mononuclear cells. Usually, this is not very striking and can be safely ignored. Layfield and coauthors 11 described such a tumor in the sacrum of a young woman. Clinical and roentgenographic features supported a diagnosis of a giant cell tumor. Histologically, there was focal anaplasia of the mononuclear cells. However, they had the features of degenerative atypia and lacked mitotic figures. Follow-up information in this case suggested that the lesion was indeed a benign giant cell tumor.

Osteosarcomas can have giant cells–sometimes so many that the tumor resembles a giant cell tumor. However, they are usually situated metaphysially and show nuclear atypia and bone formation. There are two problems with osteosarcoma with giant cells. The nuclear atypia can be subtle and they can be situated epiphyseally (Figures 5A-B). The term malignant giant cell tumor has been used in soft tissues to describe a tumor with a mixture of benign giant cells and malignant mononuclear cells. This has been considered to be a type of malignant fibrous histiocytoma and probably not related to true benign giant cell tumors of soft tissue. 12 However, the term malignant giant cell tumor is confusing. Hence, the term malignancy in giant cell tumor is preferable. The majority of these are spindle cell sarcomas arising after radiation of a giant cell tumor. Although one may question the validity of these being classified as malignancy in giant cell tumor (as Sanerkin has), they are so classified by tradition. Malignancy arising in an untreated giant cell tumor is extremely unusual, and one must resist the temptation to classify all sarcomas with proliferation of giant cells as malignant giant cell tumor.



Back to Top | Article Outline


1. Dahlin DC. Giant cell tumor of bone: highlights of 407 cases. Am J Radiol 1985; 144: 955–960.
2. Campanacci M, Baldini N, Boriani S, Sudanese A. Giant cell tumor of bone. J Bone Joint Surg Am 1987; 69A: 106–114.
3. Siebenrock KA, Unni KK, Rock MG. Giant cell tumor of bone metastasizing to the lungs: a long-term follow-up. J Bone Joint Surg Br 1998; 80B: 43–47.
4. Jaffe HL, Lichtenstein L, Portis PB. Giant cell tumor of bone: its pathologic appearance, grading, supposed variants, treatment. Arch Pathol 1940; 30: 993–1031.
5. Sanerkin NG. Malignancy, aggressive, recurrence in giant cell tumor of bone. Cancer 1980; 46: 1640–1649.
6. Unni KK. Dahlin’s Bone Tumors. Philadelphia: Lippincott-Raven; 1996.
7. Nascimento AG, Huvos AG, Marcove RC. Primary malignant giant cell tumor of bone: a study of eight cases, review of the literature. Cancer 1979; 44: 1393–1402.
8. Meis JM, Dorfman HG, Nathanson SD, et al. Primary malignant giant cell tumor of bone: “Dedifferentiated” giant cell tumor. Mod Pathol 1989; 2: 541–546.
9. Rock MG, Sim FH, Unni KK, et al. Secondary malignant giant cell tumor of bone: a clinicopathological assessment of 19 patients. J Bone Joint Surg Am 1986; 68A: 1073–1079.
10. Gitelis F, Wang JW, Quast M, et al. Recurrence of a giant cell tumor with malignant transformation to a fibrosarcoma twenty-five years after primary treatment. J Bone Joint Surg Am 1989; 71A: 757–761.
11. Layfield LJ, Bentley RC, Mirra JM. Pseudoanaplastic giant cell tumor of bone. Arch Pathol Lab Med 1999; 123: 163–166.
12. Oliveira AM, Dei Tos AP, Fletcher CDM, et al. Primary giant cell tumor of soft tissues. A study of 22 cases. Am J Surg Pathol 2000; 24: 248–256.
© 2001 Lippincott Williams & Wilkins, Inc.