The deduction of tumor histogenesis, with special reference to teratomas and ovarian tumors

The secure deduction of tumor histogenesis in vivo is notoriously difficult, owing to the retrospective form of the inquiry, which implies that no histogenetic theory makes certain-enough predictions. Deduction of a tumor's parent tissue currently relies on 1) its location relative to that of small early tumors, 2) correlation of its relative availability with tumor frequency, 3) its resemblance to the tumor tissue in various respects, and 4) its microscopic continuity with the tumor through transitional forms. Each of these criteria has pitfalls, owing to possible failure of the relevant factor to persist or to subsequent mimicry of it by other processes. This makes reliance on only one or two criteria undesirable. The application of these criteria to teratomas shows that there are no convincing data provided by the fourth. A dysgerminoma containing multiple small teratoid foci, which was studied in detail, is described, and the origin of the second tissue from the first is argued. The apparently somatic origin of some of the mesenchyma and the role of the mesenchyma in promoting teratoma development and segregation are discussed.