Aspirin, Platelets, and Bleeding

RECENT findings regarding the effect of acetylsalicylic acid on platelet function provide an interesting example of how astute clinical observation can lead to the discovery of basic physiologic and pathologic mechanisms. For many years Dr. Armand Quick had suspected that the ingestion of aspirin by hemophiliacs exaggerated their bleeding diathesis. We were interested to find that some of the ear, nose, and throat surgeons in our institution forbade patients' taking aspirin before an operation because they have associated its administration with increased bleeding at operation. The most impressive clinical evidence that aspirin causes bleeding, however, comes from its relationship to gastrointestinal bleeding. There are now many reports that the administration of aspirin causes occult gastrointestinal bleeding and, despite a few claims to the contrary, the evidence seems quite convincing. Blood loss from the gastrointestinal tract can be detected in 70% of normal subjects aftcr the administration of about 3 g of aspirin, and overt hemorrhage may occur. A local irritative effect on the gastrointestinal mucosa has been suggested as the cause of bleeding. Such a local effect is not the entire explanation, however, because intravenous administration of salicylates also produces a slight increase in occult gastrointestinal blood loss. Other etiologic mechanisms have been postulated, including an allergic reaction, increased gastric acidity, a central effect, and a cortisone-like effect. The clinical evidence that aspirin causes some derangement of the hemostatic mechanism is most suggestive and is supported by a number of experimental observations that seem to have been largely ignored until recent years. Salicylic acid and sodium salicylate cause a slight prolongation of the one-stage prothrombin time in rats maintained on a diet low in vitamin K. In humans, acetylsalicylic acid and sodium salicylate in doses of 1.3 to 5.3 g daily cause a similar mild lengthening of the prothrombin time, which is prevented by administration of vitamin K. The aspirin must be given for several days to prolong the prothrombin time for 1 to 2 seconds, and small doses of aspirin cause no detectable effect. It seems unlikely that the hemostatic abnormality produced by small doses of aspirin is due to any pronounced effect on the coagulation mechanism. Indeed, Quick described these findings as a "veritable red herring."'l Because the plasmatic coagulation mechanism did not seem to be a fruitful area for research, interest returned to the observation, made more than 10 years ago by Beaumont and colleagues and by Frick, that aspirin prolongs the bleeding time. Other workers have made similar observations, but their findings were overlooked until Quick's recent work' on the subject stimulated further investigation. Quick found that more than half of a group of normal subjects showed a small but significant increase of the Duke bleeding time 2 hours after ingestion of 1.3 g of aspirin. Even more sensitive were patients with von Willebrand's disease: They had prolongation of the bleeding time after 0.65 g (2 tablets) of aspirin. In recent years, research on the blood platelet has undergone a renaissance, and current investigations suggest that aspirn's effect on the bleeding time and on hemostasis may be due to a direct effect on platelet function. According to current concepts, the primary defense against bleeding appears to be platelet plugging at the site of injury. First, platelets are attracted to the injured vessel, probably to collagen. Other platelets attach to the trapped platelets, and an effective hemostatic plug is soon formed. Only later do fibrin threads develop and add permanence to the plug.