Nearly all experimental data available implicate the role of anticoagulation in the mechanism of metastases. Using 3 separate experimental tumor-host systems in mice, the growth rate of an implanted, primary tumor can be significantly reduced with sodium warfarin induced anticoagulation. In all experimental test conditions a surgically safe prothrombin time prolongation of 2 to 3 times normal increased the long-term survival, cure rate and evidence of metastatic tumor dissemination in all experimental models. While complications from bleeding were occasionally encountered, surgery could be performed safely and anticoagulation levels satisfactorily monitored and stabilized. There is good experimental documentation to warrant a proposed clinical adjuvant anticoagulation study to determine if coumadin may effect a reduction in metastatic spread and subsequent cure-rate in patients who are to undergo surgery "for cure" of localized malignant disease, particularly the sarcomas which have such tremendous propensity for hematogeneous metastasis.

Animal studies involving prolonged anticoagulation have been few in number. In such studies the level of anticoagulation has infrequently been monitored and hemorrhagic complications have been common. A regimen of long‐term sodium warfarin anticoagulation was studied in several strains of mice using, with slight modification, the Miale micromethod of prothrombin time determination. The degree and course of the prothrombin time alteration with several warfarin dosages was determined over a 4‐week period. The method proved very sensitive to changes in warfarin dosage, and results obtained were reproducible to a degree sufficient to permit the selection of optimal warfarin dosages for normal and tumor‐bearing mice. On these dosages the mean prothrombin time remained within the range of 2–3 times the average normal value and hemorrhagic complications were entirely avoided. In studies with small animals these methods could facilitate the use of prolonged anticoagulation, permit accurate monitoring of the level of anticoagulation, and diminish the incidence of hemorrhagic complications.

Extensive studies on the survival rates of tumor-bearing mice in three induced and two spontaneous tumors show that there is a variation in the growth rate of the primary tumor and in pulmonary and extrapulmonary metastases over a period of several transplant tumor generations. From its inception, each tumor has its own malignant pattern, as determined by its growth rate and distant metastases. In general, induced tumors are more malignant than spontaneous tumors on the basis of these two parameters. Results may differ, depending upon the age of the tumor selected. After a number of transplant tumor generations the tumor behavior is more predictable. The time it takes to reach this relatively constant pattern varies with the individual tumor. The virulence of a tumor can be determined in both tumor-bearing and amputated tumor mice. The present investigation revealed that surgery reduced the number of mice with metastases and the number of metastases in each mouse. Contrary to results of previous reports, we found that original carcinogen-induced tumors and spontaneously arising tumors do metastasize when the animals survive long enough for cells to shed from the original tumor.