To be certain of the diagnosis of malignant giant cell tumor, the pathologist must demonstrate zones of typical benign giant cell tumor in the malignant neoplasm under appraisal or in previous tissue obtained from the same neoplasm. When confronted with an obviously malignant growth that contains a few or many benign osteoclast-like giant cells, the pathologist can prove a relationship to benign giant cell tumor in no other way because other neoplasms of bone, including many of the osteosarcomas, contain a scattering or many of these benign giant cells. Even classic low-grade parosteal osteosarcoma can recur as a highly malignant sarcoma with such an abundance of benign giant cells that, without reference to the original neoplasm, the recurrent lesion may be mistaken for a malignancy in giant cell tumor. Some of the osteosarcomas of soft-tissue origin also contain numerous benign giant cells but obviously bear no relation to giant cell tumor of bone. For any given neoplasm, the stromal cells, not the benign multinucleated cells, must determine its classification. In 1960, Troup and coworkers provided clinicopathologic correlations to support this view.

With this strict definition of malignant giant cell tumor, 39 examples were found in the Mayo Clinic files. Of these, 34 were judged to occur after treatment of typical benign giant cell tumors—tumors that contained no features differentiating them from the rest of the giant cell tumors. These are referred to as secondary malignant giant cell tumors. Of the 34 cases, 26 occurred after treatment of a benign giant cell tumor which had included radiation. The other eight tumors occurred after surgical treatment of a giant cell tumor. In 6 of the 34 cases, the clinical features suggested that the original diagnosis was giant cell tumor, although this material has not been reviewed at Mayo Clinic. The interval from the diagnosis of giant cell tumor to the diagnosis of sarcoma varied from 1 to 42 years. In only one case was residual giant cell tumor present at the time the sarcoma was diagnosed. The other five primary tumors contained foci of sarcoma along with a giant cell tumor at the time of diagnosis.

Evidence is increasing that radiation may trigger the malignant transformation of various osseous lesions, especially giant cell tumor. As indicated above, however, radiation was not implicated in 13 of the 39 tumors.

Analysis of the literature on malignant giant cell tumor is virtually impossible because of lack of strict definition. The subject is further clouded by the finding of an extremely rare benign metastasizing giant cell tumor. As indicated in the preceding chapter, 20 of the 671 benign giant cell tumors produced pulmonary metastasis.