THE PART PLAYED BY THE STROMA IN CARCINOGENESIS ARNOLD VAN DEN HOOFF* Introduction Fundamental cancer research has recently seen impressive breakthroughs . The cellular information in the genome that comes to expression in certain types of cancer has been elucidated by the discovery and detailed analysis of the oncogenes [1, 2]. In addition, the multistep character ofthe condition is better—though still far from completely—understood since it has been acknowledged that carcinogenesis includes at least two phases, initiation and promodon , as sequential processes [3]. In the wake ofsuch epoch-making discoveries, all earlier observations, ideas, postulates, and speculations on the cancer problem could run the risk ofbeing played down as being insignificant or obsolete by workers in the forefront of cancer research. I wonder whether on the issue of carcinogenesis a provocative pronouncement of Paul Weiss is not appropriate. In an essay, "The System of Nature and the Nature of Systems" [4], Weiss remarked "Not only novelties, but any propositions, even well-founded old ones, if they do not seem to fit into the canon of doctrine reigning at the time, are apt to be ignored, bypassed, or even discredited as spurious." In cancer research, one of the older propositions which I feel is well founded bears on the possible involvement ofthe underlying connective tissue stroma in the process of the malignant transformation of epithelium . It is my intention to argue in this essay that the possibility of such an involvement is by no means to be neglected and that old and new The author is grateful to Dr. Michael Petri of Sundby Hospital, Copenhagen, for his comment on the first draft ofthe manuscript, and to Mrs. T. M. S. Pierik for her assistance in the preparation of the article. ?Laboratory of Histology and Cell Biology, University of Amsterdam, 1st Const. Huygensstraat 20, 1054 BW Amsterdam, The Netherlands.© 1984 by The University of Chicago. AU rights reserved. 003 1-5982/84/2704-0403$01 .00 498 I Arnold van den Hooff · Stroma in Carcinogenesis observations and views on this involvement could fit quite well in the current doctrine on the carcinogenic process. Outline of the Argument The line of my reasoning, which is based on data in the literature and which I shall elaborate in this article, is the following. The first premise is that epithelium in its physiological growth, differentiation , and functioning is decisively influenced by stromal constituents , possibly collagen, its fragments or derivatives. To me the implication seems to be that epithelial cells carry receptors sensitive to these substances. The second premise is the plausibility of the conclusion reached by numerous histopathologists who studied the cancer process, namely, that alterations of a presumably degradative nature in the stroma are essential for the ultimate malignant proliferation of epithelium in carcinogenesis . The third premise is that current concepts of the carcinogenic process include at least two steps: initiation and promotion. Initiation is effected irreversibly by a carcinogenic agent that modifies the cell's genome. The altered cancer gene remains dormant until it is activated by the action of a promoter, which binds to a cell surface receptor, setting offan intracellular cascade resulting in derepression of the cancer gene. Considering these three premises together, I am led to a pertinent proposition: during carcinogenesis stromal influences, normally an element of a physiological stromal-epithelial interrelationship, are deranged in such a way that they constitute a factor in the multistep process of tumor promotion. Physiological Stromal-Epithelial Interactions An abundance of data has been obtained on the interactions between mesenchyme and epithelium in normal embryonic differentiation [5-7]. It was the development of methods allowing dissociation of embryonic organs into their constituent tissues, followed by in vitro culture methods , that has made detailed studies of such interrelationships possible. For a large number of fetal organs, it has now been well established that such interactions are an essential feature in normal development. Mesenchyme appears to play an important role in actively controlling the differentiation of the epithelium it supports. Experiments involving adult tissues have produced somewhat confusing results [8]. With the use of transplantation approaches, it has been shown that some stratified squamous epithelia conserve their characteristics when transplanted to foreign sites. Other epithelia...