In this issue of the journal, Kreisle and associates (/) report that angiogenesis and growth of B16 melanoma decreased with age in C57BL/10 mice. They suggest that age-related reductions in tumor growth may be secondary to a diminished neovascular response by the host. In another mouse tumor system (SP1 fibrosarcoma), in which tumor growth did not decline with age, the overall neovascular reaction elicited by the tumor from the host was sufficient to support the growing tumor. However, the new vessels in the old mice arose mainly from peripheral nerves, whereas in younger mice, they arose from subcutaneous tissue. The behavior of tumors in both systems is consistent with the concept that" tumor growth is angiogenesis dependent"(2). This hypothesis, which was first proposed in 1971, can be stated in its simplest terms: Once tumor" take" has occurred, every increase in tumor cell population must be preceded by an increase in new capillaries converging on the tumor (3). The hypothesis has generated many studies leading to sequencing and cloning of angiogenic molecules (4); discovery of angiostatic steroids (5) as well as other angiogenesis inhibitors (6, 7); and elucidation of angiogenic diseases (4). The role of angiogenesis is now being explored in disciplines as diverse as developmental biology, cardiology, wound healing, ophthalmology, and dermatology. While the supporting evidence for this hypothesis has been accumulated over more than a decade, it is scattered in the literature of basic science and clinical research. Thus, the experimental basis for the hypothesis is not always clear to investigators in angiogenesis research. Consequently, I have assembled this evidence and examined the limitations of the hypothesis.

The hypothesis has certain qualifications. It does not apply to the early phase of solid tumor growth. In this" prevascular" phase, tumors are usually thin and cell population is limited; some early in situ carcinomas contain fewer than 106 cells. Tumor cells may grow in ascites without neovascularization even when the cells are angiogenic. Furthermore, the capacity of tumor cells to induce angiogenesis does not always correlate with malignancy, and there is considerable confusion in the literature about this fact. Adrenal adenoma, for example, is a benign tumor that is highly angiogenic. The onset of angiogenic activity in tumor cells is an independent event that may be expressed at very different times during neoplasia. Some examples are illustrated in table 1. These examples are not intended to imply