Roentgenologic Aspects of Metastases

THE roentgenologist has long sought a basis for rationalizing metastatic processes, particularly as applied to the skeletal and intrathoracic structures. Somehow, these efforts have been disappointing. Large series have been reviewed, and, together with innumerable case reports, constitute a formidable array of evidence from which some conclusion might be drawn regarding the incidence and behavior of metastases. Unfortunately, much of this evidence is lacking in thoroughness. Even in leading institutions it is exceptional to find a 50 per cent roentgenographic work-up; and pathologic confirmation is much less. Careful autopsy examination alone cannot give us an accurate estimate of the incidence of skeletal metastases, nor can the roentgenographic findings without pathologic confirmation serve as an entirely dependable criterion. Certainly, in many instances, the primary site remains a mystery to the roentgenologist. Such a correlation is singularly lacking in the recent literature. The roentgenologist in his analysis of a case is prejudiced by certain ideas, i.e., that skeletal metastasis from the gastro-intestinal tract is a negligible factor, that squamous-cell carcinomas of the dermis and of the uterus are similarly of no consequence. Many other sources of bone metastasis, such as lymphosarcoma, melanoblastoma, carcinoma of the urinary bladder, testes, and pancreas are considered rarities. The fact that osteogenic sarcoma may occasionally become manifest in distant bones, whether by primary or secondary intention, seems to have escaped attention. Furthermore, convenient rules are occasionally adopted such as, “In carcinoma of the prostate, pulmonary involvement is probably never present in the absence of skeletal metastases.” Consequently, only a film of the pelvis and spine is necessary. Or that, in the absence of regional lymph node involvement (i.e., carcinoma of breast), skeletal or pulmonary metastases are unworthy of serious consideration. Or that, skeletal metastases are late manifestations of malignant disease, etc. The pathologist has long been disillusioned as to the predictability of behavior of metastatic processes. Consider that cancerous neoplasms may be disseminated by the venous, arterial, or lymphatic systems; by direct extension; by drop metastases; by retrograde permeation; by arteriovenous communication; can pass from the lesser to the greater circulation through the pulmonary capillaries, and from the lymphatic to the venous circulation; that great differences in different tumors and in the same tumor all present their own peculiar characteristics; and, finally, that most of the tumor emboli are autolyzed and never take root. This gives the problem a complexion of such overwhelming intricacy as to make impossible even limited generalities.