Oral Anticoagulants in the Prevention and Treatment of Vascular Disorders

Abstract

In 1940, Karl Link and his colleagues succeeded in synthesizing dicoumarin. They had identified and were able to manufacture the substance in spoiled sweet clover which had been incriminated as the cause of a severe bleeding disease in cattle. Within one year of its synthesis, dicoumarin was given to man (1), and numerous reports of its clinical application in thrombotic disorders soon followed. But why were clinicians so quick to perceive a therapeutic role for this recently identified cattle poison? The nineteenth century pathologists had emphasized that there were important structural differences between clots and thrombi, and had deduced that the processes of clotting and thrombosis were different, but this message was set aside in favor of the view that thrombosis was clotting in the wrong place (2). It seemed that a substance which could prevent clotting should prevent thrombosis. Moreover, there was a precedent. Heparin, the parenterally administered anticoagulant with a short duration of action, had been discovered earlier, and there was evidence from both animal experiments and clinical application that this anticoagulant could prevent both venous and arterial thrombosis (3). The newly synthesized orally active anticoagulant held out great promise and, initially, there was great enthusiasm for its use. However, the next twenty years were to result in much confusion about the role of oral anticoagulants in the treatment of thrombotic disorders. During this period, important advances in the field of coagulation researach were being made, but the clinical use of dicoumarol and the vitamin K antagonist drugs which followed tended to run ahead of these developments. Control of treatment was a particular problem. In 1935, Quick et al. (4) developed a test which was thought to measure prothrombin, and even before dicoumarin was synthesized, it was known that the severity of the bleeding disease in cattle correlated with the degree of prothrombin deficiency as judged by this test. Quick’s test was, therefore, used to control treatment when dicoumarin was used clinically, but many workers used modifications of the original test. There was uncertainty about the frequency with which tests should be performed, and it was not fully appreciated that thromboplastins from different sources would influence the test results. As a result, degrees of hypocoagulability used therapeutically varied from center to center and some workers abandoned control of treatment altogether (5). That hemorrhage is the main complication of treatment was widely recognized, but the frequency with which hemorrhagic complications were reported differed as did the methods of controlling treatment. Moreover, the contraindications to oral anticoagulant therapy and the effect of drug interactions that we recognize today were being learned by trial and error. Furthermore, during this period the efficacy of treatment was assessed mainly by clinical judgement; an awareness of the need for prospective control randomized trials with clearly defined end points in addition to more satisfactory methods of controlling treatment did not emerge until the late 1950’s. There is, therefore, a long period in the history of oral anticoagulant therapy during which uncertainty about the value and safety of the agents inevitably arose.