The first group has its starting point from the osteoblastic layer lining the medullary canal, and is usually a giant-celled or myeloid tumor. Besides these, giant-celled sarcomas, spindle or round-celled sarcomas may also originate from the medullary tissue. These, however, are not common, particularly the pure spindle and round-celled tumors, but mixed tumors with a preponderance of giant cells are quite common.
The peripheral or periosteal tumors proceed from the layer of osteoblasts or embryonic cells that separate the periosteum from the bone.
These tumors are mostly spindle-celled sarcoma, while occasionally we find round-celled or mixed round and giant-celled tumors. By far the greatest number are spindle-celled or contain a large proportion of spindle cells.
Concerning the relative frequence of these different forms of sarcoma, Gross found, in reviewing one hundred and sixty-five cases of sarcoma of long bones, that seventy or 42 per cent. were giant-celled tumors; forty-five were spindle-celled osteoid sarcoma; sixteen central spindle-celled; thirteen periosteal round-celled; twelve central round-celled, and nine pure periosteal spindle-celled sarcomas.
Of these, 75 per cent. of the peripheral involved the lower epiphysis of the femur. There were seventy central giant-celled tumors of which nineteen or 27.01 per cent. involved the lower and two the upper extremity of the femur.
Central giant-celled or myeloid sarcoma are found principally in the lower epiphysis, and occasionally in the upper. They are of rather slow growth and represent the lowest degree of malignancy of any of the bone sarcomas. A few, no doubt, are benign and correspond to the class of tumors described by Bland Sutton as myelomata, and considered by him as being essentially benign.
Many of the earlier writers considered all pure myeloid tumors as innocent. Among those who regarded only the mixed tumor as malignant, are Nellaton, Wilks and Forster. It must be emphasized that in none do we find only giant cells, but all “mixed” tumors containing spindle and round cells and varying propositions. From a clinical standpoint, therefore, it is never safe to consider any of them innocent.
Undoubtedly, the degree of malignancy varies greatly in individual cases. The rapidly growing vascular and particularly the osteoid sarcomas of this type possess a great degree of malignancy as is evidenced by local infection of surrounding tissue, lymphatic involvement and metastasis in distant organs. Invasion of the
