We have previously reported that the number of metastases resulting from the inoculation of TA3 ascites tumor cells into mice could be reduced by 50 per cent if the mice were first injected with neuraminidase.' It was assumed that this effect probably reflected the action of neuraminidase on the sialic acid coat of tumor or host endothelial cells, since it was believed that mutual stickiness between these cell types was important in metastasis formation. 'Since the uncoating of tumor cells by neuraminidase prior to injection into the host seemed to have very little antimetastatic effect,' attention was directed to the effects of neuraminidase on the host. The present paper reports the unexpected finding that neuraminidase in vivo produces thrombocytopenia, with a close correspondence between the decrease in platelet levels and the decrease in metastases. Further experiments, in which the concentration of circulating platelets was varied by other agents, confirm the association between platelet levels and metastases, and suggest that the previously observed antimetastatic effect of neuraminidase may be at least partially mediated by its effect on the platelets.Material8 and Methods.-TA3 ascites tumor cells2 were grown in the peritoneal cavity of CAFi/Jax female mice, each receiving 2.0-2.5 X 105 or 3.0-3.5 X 105 tumor cells suspended in 0.2 ml of medium. They were killed 14-16 days later, and ether and lung metastases were counted.Neuraminidase was given i.v. at different intervals before tumor inoculation; each mouse received a single dose of 0.2 ml of a commercial solution of neuraminidase from Vibrio cholera (Behringwerke) containing 500 units per ml. Each unit was capable of releasing from aL-glycoprotein 1 lig of sialic acid as N-acetylneuraminic acid at 370C in 15 min at pH 5.5. It is reported to be free of proteinases, aldolase, and lecithinase C. Controls were injected with the diluent of the commercial neuraminidase.When it became apparent that neuraminidase was preventing normal clot retraction and prolonging the bleeding timeI blood platelets were counted with a modified method of Nakeff and Ingram.4 Briefly, we proceeded as follows. Peripheral blood was obtained from the tail vein of mice in 10-IAl double oxalated capillary tubes (Scientific Products) and diluted to 20 ,l with a phosphate-buffered saline (PBS) containing 0.1% ethylenediaminetetraacetate (EDTA). This new suspension was taken up into a 75-41 heparinized standard microhematocrit capillary tube (Scientific Products) followed by approximately 10 jil of Dow Corning silicone fluid no. 550 (specific gravity, 1.07000). This tube was sealed with Critoseal (Scientific Products), and a small gap of air was left between the seal and the silicone fluid. The tubes were centrifuged in an IEC hematocrit centrifuge for 30 sec. Then each tube was broken at the plasma-silicone interface, and the layer of platelets was suspended in 20 ml PBS with 0.1% EDTA and counted electronically with a Coulter counter.In some experiments, platelets were transfused 24 h...

Thrombocytopenia reduces the number of metastases produced by a wide variety of murine tumors. Studies aimed at investigating interactions between tumors and platelets reveal that many tumors aggregated platelets in vitro andlor produced thrombocytopenia in vivo. I n some instances, tumor-cell-induced thrombocytopenia in vivo was accompanied by accumulation of platelets in the lung. Thrombocytopenia was most active against metastases produced by tumors with the capacity to aggregate platelets

Smooth muscle cells (SMCs) in the rat carotid artery leave the quiescent state and proliferate after balloon catheter injury. The precise signals responsible for this SMC mitogenesis need to be elucidated. Although platelet-derived growth factor (PDGF), a potent SMC mitogen, is released from activated platelets, damaged endothelium, and macrophages, it cannot be solely responsible for this proliferation. In search of other SMC growth factors, we have examined several proteins of the coagulation cascade. At nanomolar concentrations, factors X, Xa, and protein S promote cultured rat aortic SMC mitosis. In contrast, factor IX is only weakly mitogenic, whereas factor VII and protein C fail to stimulate SMC division. Protein S, the most mitogenic of these coagulation cascade factors, stimulates DNA synthesis in cultured SMCs with a time course similar to that of PDGF-AA and without the delay observed for transforming growth factor .3. Antistasin and tick anticoagulant peptide, two specific factor Xa inhibitors, inhibit SMC mitogenesis due to Xa and protein S. Coagulation factors that possess mitogenic activity may contribute to intimal SMC proliferation after vascular injury as a result of angioplasty or vascular compromise during atherogenesis.

This review studies interactions of tumor cells with a particular host system which is normally responsible for hemostasis and the physiological integrity of the blood vessel luminal surface. With malignancy components of this system are frequently activated, producing abnormalities of blood coagulation, increased platelet responses, and conditions favoring tumor growth and metastasis. Activation of the clotting cascade is mediated by tumor and macrophage procoagulants, acting via Factor X or VII. Thrombin and fibrin are formed. Thrombin also interacts with platelets and the endothelium, potentiating or decreasing coagulation. Generation of thrombin or other tumor mechanisms activate platelets, leading to direct aggregation or secretion of ADP, serotonin, and/or intermediates of the arachidonate metabolism. Vascular lesions caused by tumor attack, platelet secretion, or exogenous agents promoting metastasis may also activate the hemostatic system. It is not yet fully understood how activation of the clotting system, including platelets, contributes to metastasis. Secretion of platelet products appears, however, to be heavily involved. Based on putative mechanisms of action, anticoagulants, platelet inhibitors, thrombocytopenic or vascular repairing agents have been used to control tumor spread. Results depended on the agent and experimental model of metastasis used. Except for coumarin, which was beneficial even against spontaneous metastases, other anticoagulants and platelet inhibitors, excluding perhaps Nafazatrom, gave equivocal results. Thrombocytopenic agents, however, were effective in every tumor system and with any experimental model of metastasis, indicating that platelets play a role in this process. Also consistent were the inhibitory effects of leech salivary gland extract (probably a vascular repairing agent) against lung tumor colonization promoted by ionizing radiation, cyclophosphamide, and cortisone.