ObjectiveTo determine the influence of pentoxifylline on endothelial-associated coagulation.DesignProspective, randomized, placebo-controlled study.SettingA surgical intensive care unit of a university hospital.PatientsConsecutive patients (n equals 60) with trauma or sepsis secondary to major (nontrauma) surgery. All patients received controlled mechanical ventilation.InterventionsAccording to a randomized sequence, the patients either received pentoxifylline continuously over 5 days (1.5 mg/kg/hr iv) (trauma-pentoxifylline group [n equals 15], sepsis-pentoxifylline group [n equals 15]) or saline solution as placebo (trauma-control group [n equals 15], sepsis-control group [n equals 15]).Measurements and Main ResultsIn addition to the standard coagulation screen, thrombomodulin, protein C, (free) protein S, and thrombin-antithrombin plasma concentrations were measured by enzyme-linked immunosorbent assays. Intensive care therapy, hemodynamics, and changes of Acute Physiology and Chronic Health Evaluation II score were comparable for pentoxifylline-treated and nontreated patients. An average dose of 2.5 g/day of pentoxifylline (range 2.2 to 2.9) was infused into the pentoxifylline-treated patients. At baseline, plasma thrombomodulin concentrations were higher in the septic patients than in the trauma patients. Thrombomodulin plasma concentrations increased significantly more in the control patients (trauma: from 38.9 plus minus 10.5 to 59.9 plus minus 10.1 ng/mL; sepsis: from 49.7 plus minus 12.1 to 72.3 plus minus 11.2 ng/mL) than in the pentoxifylline-treated patients (trauma: from 37.9 plus minus 11.9 to 50.2 plus minus 9.2 ng/mL; sepsis: from 51.9 plus minus 10.1 to 63.3 plus minus 10.2). Starting from similar baseline values, protein C concentration increased significantly more in the sepsis-pentoxifylline patients (from 52.0 plus minus 11.1% to 69.1 plus minus 11.1%) than in the untreated septic patients (from 50.1 plus minus 10.0% to 52.5 plus minus 9.5%). There were no significant differences between the pentoxifylline-treated and nontreated groups for protein S and thrombin-antithrombin concentrations. Standard coagulation parameters (fibrinogen, activated partial thromboplastin time, antithrombin III) did not differ between these groups either.ConclusionsContinuous intravenous administration of pentoxifylline for 5 days beneficially influenced the thrombomodulin/protein C/protein S system in both the trauma and septic patients.(Crit Care Med 1996; 24:940-946)