Tumor Invasion and Metastasis: An Imbalance of Positive and Negative Regulation1

A group of coordinated cellular processes, not just one gene product, is responsible for invasion and metastasis, the most life-threatening aspect of cancer. It is now recognized that negative factors may be just as important as positive elements. Genetic changes causing an imbalance of growth regulation lead to uncontrolled proliferation necessary for both primary tumor and metastasis expansion. However, unrestrained growth does not, by itself, cause invasion and metastasis. This phenotype may require additional genetic changes. Thus, tumorigenicity and metastatic potential have both overlapping and separate features. Invasion and metastasis can be facilitated by proteins which stimulate tumor cell attachment to host cellular or extracellular matrix determinants, tumor cell proteolysis of host barriers, such as the basement membrane, tumor cell locomotion, and tumor cell colony formation in the target organ for metastasis. Facilitory proteins may act at many levels both intracellulari) or extracellularly but are counterbalanced by factors which can block their production, regulation, or action. A common theme has emerged. In addition to loss of growth control, an imbalanced regulation of motility and proteolysis appears to be required for invasion and metastasis. Metastatic Cells Can Dominate the Primary Tumor Population Metastasis is a cascade of linked sequential steps involving multiple host-tumor interactions (1-3). To successfully create a metastatic colony, a cell or group of tumor cells must be able to leave the primary tumor, invade the local host tissue, enter the circulation, arrest at the distant vascular bed, extravasate into the target organ interstitium and parenchyma, and prolif erate as a secondary colony. Angiogenesis is required for the expansion of the primary tumor mass, and new blood vessels penetrating the tumor are frequent sites for tumor cell entry into the circulation (4, 5). Angiogenesis is also required for expansion of the metastatic colony. At any stage, tumor cells must overcome host immune cell killing (2). A very small percentage (<0.01%) of circulating tumor cells ultimately ini tiate successful metastatic colonies. Consequently, metastasis has been viewed originally by Fidler and Hart (1) as a highly selective competition favoring the survival of a subpopulation of metastatic tumor cells that preexist within the heterogeneous primary tumor. Fidler's metastatic subpopulation concept is well accepted, but the relative size of this subpopulation in the primary tumor may vary with time and between tumors. Estimating the size of the metastatic subpopulation is of clinical significance since a prognostic assay based on a sample of the primary tumor would be highly inaccurate if the aggressive subpopulation was only a very small proportion of the total number of tumor cells being sampled. Recently, Kerbel (6) has addressed the subpopulation question experimentally using genetic markers. He determined that the metastatic subpopulation dominates the primary tumor mass early in its growth. The dominance may be due to selective growth of the metastatic subpopulation in response to local cytokines. Thus, measurement of the average level of a molec' Presented at the Symposium "Discoveries and Opportunities in Cancer Research: A Celebration of the 50th Anniversar) of the Journal Cancer Research." May IS. 1991, during the 82nd Annual Meeting of the American Association for Cancer Research, Houston, TX. ular marker in a primary tumor sample is likely to reflect the general metastatic propensity of the entire tumor. Indeed, this has been borne out by a number of clinical studies indicating that the average level of specific protein markers or amplified oncogenes measured in the primary' tumor can be correlated with clinical aggressiveness parameters of metastasis and recur rence (7-10). Tumor Cell Interaction with the Extracellular Matrix During the development of invasive tumors, tumor cells disobey the social order of organ boundaries and cross into tissues where they do not belong. The mammalian organism is divided into a series of tissue compartments separated by the extracellular matrix unit consisting of the basement membrane and its underlying interstitial stroma (3, 11). The basal cells of the epithelium or organ parenchymal side of this unit are attached to the basement membrane. On the opposite side, the interstitial stroma contains stromal cells, fibroblasts, and myofibroblasts. The nervous system, muscle cells, and blood ves sels are also surrounded by a continuous basement membrane. During all types of benign tissue remodeling, proliferative dis orders, and carcinoma in situ, the cell populations on either side of this connective tissue unit do not intermix. Only during the transition from in situ to invasive carcinoma do tumor cells penetrate the epithelial basement membrane and enter the underlying interstitial stroma to interact with the stromal cells. Thus, a definition of the behavior of the metastatic tumor cell is the tendency to cross tissue compartment boundaries and intermix with opposite cell types (3, 12). The continuous basement membrane is a dense meshwork of collagen, glycoproteins, and proteoglycans which normally does not contain any pores large enough for passive tumor cell transversal. Consequently, invasion of the basement membrane must be an active process. Once the tumor cells enter the stroma, they gain access to lymphatics and blood vessels for further dissemination. Tumor cells must cross basement mem branes to invade nerves and muscle. During intravasation or extravasation, the tumor cells of any histolÃ3gica! origin must penetrate the subendothelial basement membrane. As a general feature of all types of carcinomas (12), defects in the basement membrane are associated with invasion. In contrast, benign proliferative disorders are all characterized by a continuous basement membrane separating the epithelium from the stroma. The general observation of defective basement mem branes associated with cancer invasion and progression indi cates that aggressive tumor cells may interact with basement membranes in a manner fundamentally different from that of normal cells. This general feature of malignant tumors provides the foundation for investigation of molecular mechanisms. Interactions of the tumor cell with the basement membrane can be separated into three steps: attachment; matrix dissolu tion; and migration. The first step is binding of the tumor cell to the basement membrane surface-mediated cell surface recep tors of the integrin (13, 14) and non-integrin (15, 16) variety. Matrix receptors recognize glycoproteins such as laminin, type IV collagen, and fibronectin in the basement membrane. Two