Perspectives in Biology and Medicine

TOWARD AN UNDERSTANDING OF ANGIOGENESIS: SEARCH AND DISCOVERY JUDAH FOLKMAN* Fortuitous Failure ofNeovascularization I was in the navy, assigned not to a ship but to the Naval Medical Research Institute in Bethesda. Frederick Becker and David Long were there also. The three of us had been invited to interrupt our respective residencies to fulfill 2 years of military obligation. It was 1960, and the navy was curious about the feasibility of blood substitutes for use in its hospitals aboard aircraft carriers. We experimented with hemoglobin solutions. Dog thyroid glands were perfused with hemoglobin solutions in isolated glass chambers for periods of more than a week. The perfused organs appeared to be viable, but could they support growth? We injected tumor cells from a mouse melanoma into the thyroid glands. Tiny tumors developed and grew rapidly at first but then stopped enlarging at diameters of 1-2 mm [I]. What was the explanation for the arrested growth of these tumors? Immune rejection seemed unlikely because these organs were perfused with acellular fluids. Furthermore, when tumors were transplanted to a freshly isolated thyroid gland, the size limit held. However, when reimplanted into the donor mice, these tumors grew rapidly, beyond 1 cm3, and killed their hosts. In the mice the tumors were vascularized; in the isolated perfused organs, they were not. (fig. 1). What we had witnessed, The author is grateful to the National Cancer Institute for nearly uninterrupted support since receiving a Research Career Development Award in 1967 and to Mary Fink, Barbara Sanford, and Colette Freeman of its administrative staff. Grants from the American Cancer Society and the Juvenile Diabetes Association and support from the Franzheim Synergy Trust, the Zaffaroni Foundation, and Harvey Arkin are gratefully acknowledged. Appreciation is expressed to Michael Klagsbrun, Robert Langer, Joanne Murray, and Bruce Zetter for their critical review of the manuscript and to Pauline Breen and Lisa Harrison for typing. And to the officers and scientists of the Monsanto Company, especially Monte Throdahl and Jack Hanley, a special expression of gratitude for their vision, patience, and gentlemanly cooperation. ?Department of Surgery, Children's Hospital, and the Departments of Surgery and Anatomy, Harvard Medical School, Boston, Massachusetts 02115.© 1985 by The University of Chicago. All rights reserved. 003 1-5982/86/2901-0460$01 .00 10 I Judah Folkman ¦ An Understanding of Angiogenesis Fig. 1.—Diagram of isolated perfused thyroid gland. Melanomas growing in the gland have reached a limit of 1-2 mm diameter and are not vascularized. When these tiny tumors were transplanted to mice, tumor vascularization and rapid growth ensued. although we did not realize it at the time, was a fortuitous failure of neovascularization. The remarkable outcome of this experiment was that a tumor implant was prevented from becoming vascularized and that the isolated organ survived just long enough so that the subsequent inhibition of tumor growth could be measured. In conventional studies of transplantable tumors in animals, the prevascular phase is brief and usually imperceptible . Once a tumor can be detected, it is already well vascularized. We soon found ourselves studying tumor neovascularization rather than hemoglobin perfusates. During the next 3 years of the completion of my surgical residency (at the Massachusetts General Hospital in Boston), the experience of this experiment fermented in my subconscious. I knew that we did not understand its full significance. When I operated on patients with cancer , their tumors were already large and well vascularized, like those in the mice. But on occasion I glimpsed tiny tumors in the prevascular stage, waiting for capillary blood vessels to connect to them, analogous to the melanomas in the isolated perfused organs. I cannot forget a woman with a large, vascularized carcinoma of the ovary; metastases had implanted on the peritoneal lining. The implants were white, avascular, and uniformly less than 2-3 mm in diameter. I also recall a man with colon cancer that had seeded the omentum. Some of the tumor implants were already vascularized and as large as 1 cm3, but many others were tiny, white, avascular nodules of a few millimeters in diameter. Perspectives in Biology and Medicine, 29, 1 ¦ Autumn 1985 | 1 1 The Idea That Tumors Are Angiogenesis Dependent I resumed the...