VA Cooperative Study #75 was established to test in a controlled, randomized trial the hypothesis that warfarin anticoagulation would favorably affect the course of certain types of malignancy. No differences in survival were observed between warfarin‐treated and control groups for advanced non‐small cell lung, colorectal, head and neck and prostate cancers. However, warfarin therapy was associated with a significant prolongation in the time to first evidence of disease progression (P = 0.016) and a significant improvement in survival (P = 0.018) for patients with small cell carcinoma of the lung, including the subgroup of patients with disseminated disease at the time of randomization (P = 0.013). A trend toward improved survival with warfarin treatment was observed for the few patients admitted to this study with non‐small cell lung cancer who had minimal disease at randomization. These results suggest that warfarin, as a single anticoagulant agent, may favorably modify the course of some, but not all, types of human malignancy, among which is small cell carcinoma of the lung. Further trials of warfarin may be indicated in patients with limited disease who have cell types that failed to respond when advanced disease was present.

Routine blood coagulation tests were performed on 431 consecutive patients enrolled in a study of the role of anticoagulation in cancer treatment (VA Cooperative Study #75). Two hundred sixteen control patients were treated with standard therapy, and 215 patients were treated with standard therapy plus sodium warfarin. At the time of entry into the study, the most common abnormalities were elevated fibrinogen levels, platelet counts, and fibrinopeptide A levels. Serial studies demonstrated a steady increase in platelet count and fibrinogen levels before death. Anticoagulation lowered FPA levels but had no significant effect on fibrinogen levels, platelet counts, or euglobulin clot lysis times. An unexpected finding was a dramatic increase in fibrin split product levels after institution of anticoagulation (means +/- SEM = 42.6 +/- 116.4 vs. 2.9 +/- 7.0 mg/L in control subjects; P less than 0.02). This study supports the presence of subclinical activation of blood coagulation in most patients with cancer. Moreover, the preferential activation of fibrinolysis in anticoagulated patients suggests a role for a vitamin K-dependent factor(s) in the regulation of fibrinolysis in patients with cancer.