The Relationship of Twins, Teratomas, and Ovarian Dermoids

The bizarre structure of teratomatous tumors has long excited interest. Unlike other neoplasms, the teratomas are composed, characteristically, of more than a single type of cell and are related, histogenetically, to more than one of the three embryonic germ layers. The constituent elements of these tumors may be quite foreign to their site of origin in the body and, in their arrangement, tend to mimic organs or even organ systems. The abortive or imperfect character of this resemblance finds reflection in the name, first applied by Virchow ( I 5 ) , Of "teratoma," a malformed or "monstrous" tumor. It is natural that the morphologic peculiarities of the teratomas should have led to the formation of theories of pathogenesis differing from those associated with other tumors. With its pleomorphic, organoid character, a teratoma represents in itself, as expressed by MacCallum (9), "a frustrated attempt at the formation of a human body in which the whole plan has failed through the lack of the necessary parts, and the distortion and disarrangement of those which were available." Two alternative mechanisms might reasonably explain the occurrence of such an ontogenic fiasco. First, there is abortive parthenogenesis. Second, there is inclusion of isolated blastomeres. A teratoma might be the end result of spontaneous development of unfertilized germ cells, ova or sperm. Several facts are in favor of such a view. The gonads are among the commoner sites occupied by tumors of this group. Dermoids comprise approximately I8 per cent of proliferative ovarian neoplasms (I4) , while tumors of the testis are quite commonly teratomatous. Michalowsky ( ix ) and Bagg ( i ) have produced teratomas of the testes in the rooster, by the injection of zinc chloride solutions. Bosaeus (2) demonstrated the production of teratomas in frogs by the reimplantation into the mother's body of eggs that had been removed and mechanically stimulated to develop parthenogenetically. Earlier, Loeb (7) described the occurrence of teratoid structures in the ovaries of virgin guinea pigs, and pointed out their importance as putative links between unfertilized ova and true teratomas, evolved through parthenogenesis. A teratoma might also be formed by the isolation and inclusion of a blastomere in the body of the embryo that develops into the host of the teratoma. Segregation of such a relatively undifferentiated cell, or group of cells, with temporary inhibition of its development, might result in the formation of a "cell rest," in the sense of Cohnheim (3) and Ribbert (I2), that would have sufficient multipotency to produce a teratoma, once it became reactivated. This theory, too, gains support from diverse facts. As emphasized by Marchand (IO), Schwalbe ( i3) , Greil (5), and others, it is possible to construct complete morphologic series of cases ranging from equal conjoined twins through irregular parasitic conjoined twins, and through the more complex embryomas to the characteristic teratomas. So many intergrades exist, anatomically, between twins and teratomas, that one might naturally expect to find common, or at least similar, modes of production of twins and teratomas. While knowledge of the actual mechanism of human twinning is still inferential, it seems most likely that the process involves a splitting of the zygote with developmental separation of component portions. Whether "identical" or "fraternal" twins are produced by similar or by dissimilar mechanisms is an important, but unsettled question. Curtius and yon Verschuer (4) believe the mechanisms are similar; most other authors believe them dissimilar. At least one type of twin, however, is formed by a process so closely analogous to that postulated by the inclusion theory of teratoma formation, that such variables as time of fission or relative proportions of the divided segments offer themselves as logical modifiers determining whether twin or teratoma is the end result. The teratomas encountered in extragonadal locations are most frequently situated in or near the midline of the host. The sacrococcygeal region is a classic site. It seems hardly coincidental that the midline is a common axis of symmetry in cosmobia (equal and unequal) or that the sacrococcygeal region is an equally classic site of fusion of conjoined twins.