How Thrombogenic Is Hypoxia?

Abstract

ON AUGUST 3, 1953, A US EXPEDITION CONSISTing of 7 men got caught in prolonged storms at an altitude of about 8200 m on the Abruzzi ridge of K2, the second highest mountain in the world. After 4 days, venous thrombosis was diagnosed in the leg of Art Gilkey followed 2 days later by coughing and dyspnea, presumably due to pulmonary embolism. During the retreat from the mountain, the whole party started to slide down the steep ridge as their ropes became entangled. All were stopped by the strong hand of the last man. This well-known incident, other anecdotal reports from Himalayan climbers, and the finding of thrombi in the pulmonary vasculature of individuals who succumbed to high altitude pulmonary edema led to the notion that hypoxia must be thrombogenic. However, the question of whether the thrombi may have been a result rather than a cause of whatever process produces high altitude pulmonary edema could not be answered by autopsies. Increased plasma levels of prothrombin fragment 1 2 and thrombin-antithrombin complex observed in healthy young men exposed to a simulated altitude of 2400 m over 8 hours suggested that the epidemiological observations of an association between long-distance flights and venous thromboembolism might in part be attributed to activation of blood coagulation by mild hypoxia resulting in an arterial oxygen saturation of 93%. The concept of activation of blood coagulation by mild hypoxia resulting in enhanced thrombin and fibrin formation was questioned because blood was sampled through an 18G catheter, which may explain the activation of coagulation observed in this study. In addition, a prospective study at an altitude of 4559 m did not find increased thrombin or fibrin formation during the early stages of high altitude pulmonary edema. A subsequent carefully controlled study demonstrated that 8 hours of normobaric isocapnic hypoxia corresponding to the ambient PO2 of an altitude of 3600 m did not increase plasma levels of prothrombin fragment 1 2 and thrombinantithrombin complex, or levels of D-dimer, a marker of fibrin degradation. The results were confirmed in hypobaric hypoxia corresponding to an altitude of 3600 m. All these findings are in accordance with earlier observations that the markers of thrombin and fibrin formation were not increased compared with control values at 550 m in mountaineers between 4 and 52 hours after ascending on foot to an altitude of 4559 m. Furthermore, exercise-induced activation of coagulation is not enhanced by hypoxia equivalent to an altitude of 4500 m. In young healthy athletes, exhaustive running on a treadmill over 1 hour at a fraction of inspired oxygen level of 0.12 resulting in a mean arterial oxygen saturation of 70% did not increase markers of thrombin and fibrin formation more than running to exhaustion in normoxia, while exhaustive running in normoxia in women taking oral contraceptives and in individuals with protein C or S deficiency resulted in significantly greater increases of these activation markers compared with the response in control individuals without oral contraceptives or thrombophilia. These data demonstrate that hypoxia at an altitude of 4500 m is not thrombogenic in an exercise model that detects additional activation of coagulation in individuals with thrombophilia. In this issue of JAMA, Toff and colleagues report the results of a crossover trial simulating the conditions of long-haul airplane flights in 73 healthy volunteers in whom the most frequent causes for thrombophilia (individuals with factor V Leiden and prothrombin G20210 mutations) were excluded. The participants included 12 older individuals ( 50 years) and 12 women taking oral contraceptives. The participants sat during 8 hours in a low-pressure chamber at pressures equivalent to an altitude of 2400 m and sea level. Coagulation, fibrinolysis, platelet function, and markers of endothelial activation after prolonged sitting over 8 hours in normoxia vs hypoxia did not differ in the entire group and in the 3 participant subgroups. Prolonged sitting in either condition leads to a small decrease of prothrombin fragment 1 2 and minor disparate changes of some other parameters partly attributable to diurnal variation. The study demonstrates, as suggested by a smaller, uncontrolled trial, that mild hypoxia in combination with prolonged sitting does not induce prothrombotic alterations in healthy individuals who are at low risk