As early as 1943, Allen and Barker suggested that anticoagulants might be of value in the treatment of thrombosis of the cerebral arteries [12]. Fisher may have been the first to mention anticoagulation to prevent carotid artery occlusion in 1951, and he noted that recurrent symptoms disappeared after starting coumadin [13]. Perhaps one of the earliest reports of series of patients came from the Mayo Clinic and was based on several patients successfully treated with coumadin after presenting with recurrent transient neurologic symptoms attributed to an impending occlusion of the carotid or vertebrobasilar artery and considered affected by severe atherosclerotic disease [14, 15].

The neurologist Siekert later wrote:

My own work has been concerned principally with three groups of patients. Two of these concern states of intermittent insufficiency, involving, respectively, the basilar and the carotid arterial systems. The other is thrombosis within the basilar system. The laminated appearance of a thrombus in the basilar artery and the frequent stepwise progression of neurologic abnormalities when this artery becomes occluded suggested the usefulness of anticoagulant therapy. The transient episodes, so very common in patients who subsequently sustain thrombosis of the basilar, were then looked upon as precursors to actual occlusion.

Wright et al. [16] presented the findings of 57 patients with various types of cerebral thromboembolic episodes treated with anticoagulants. Of this total of 57 patients, 31 were patients with rheumatic heart disease and 19 were patients with cerebral thrombosis or embolism based on arteriosclerosis. The remaining seven patients had conditions of miscellaneous causes. The most impressive results were obtained in the patients with rheumatic heart disease. In a period of 464 months, 31 patients who were under observation developed 137 major episodes of thromboembolism, and among these, there were 46 cerebral vascular episodes. During the next several years, in which time they received anticoagulant therapy, there were only 11 thromboembolic episodes, of which 4 were cerebrovascular [16].

The details of a national cooperative controlled study of long-term anticoagulant therapy in cerebral ischemia and cerebral infarction were presented with seven clinical centers and extended over a period of 42 months, ending in 1961. This trial, which was reported in 1962, had 443 patients, and about half were anticoagulated. In the treated group, there were 12 fatal hemorrhagic complications, and hemorrhage contributed to a fatal outcome in a few further cases. There were two instances of nonfatal intracerebral bleeding in the treated group and one in the control [17]. They rightly determined that “judging from our data, it can be concluded that, from the standpoint of mortality, long term anticoagulant therapy plays no beneficial role in the treatment of thrombotic cerebrovascular disease and may be harmful” [17].

The clinical trial was not blinded, and the number of patients in each ischemia group was very few, diagnosis was clinical and without the opportunity to verify the nature of stroke by computed tomography scan. Use of heparin and coumadin remained a choice determined by clinicians; personal preference prevailed. The clinical trials that followed mostly looked at the benefit of anticoagulation in preventing stroke in patients with atrial fibrillation [18, 19].

The renowned stroke neurologist Louis Caplan later argued against its use in lacunar stroke but emphasized that it could prevent clot propagation in a basilar artery embolus and felt that under those circumstances, intravenous heparin was warranted. No data existed on the benefit of long-term treatment [20]. In 1999, before the new millennium, he appropriately stated that “use of heparin continues to be one of the most contentious and controversial issues in the fields of neurology and medicine… clinicians ask, ‘which product should be given to which patients, by what route, with or without an initial bolus, at what dose, monitored how, and when should the treatment be started?’” [21, 22]. Caplan further argued that “I use it as acute therapy in patients with recent large artery occlusion or tight stenoses of large arteries.” He also argued that the fact that we do not know whether we are dealing with “white” or “red” clots was part of the problem. White clots are composed of platelets and often mixed with fibrin, and red clots are fibrin-dependent thrombi. White clots could theoretically be prevented by agents that decrease platelet aggregation and agglutination, whereas heparin and warfarin might work against red clots. Now and then, we can argue that the evidence of heparin and coumadin in acute stroke is marginal at best. Personal preference has remained a main driver of its use in stroke, and many stay away from it.