ported as associated with liver cirrhosis. 4,5 Moreover, a vari-A defect in hemostasis has been repeatedly reported able degree of thrombopenia-caused mainly by an increased in patients with cirrhosis. However, the nature of this splenic pool of platelets-usually is present in cirrhotic padefect has not been fully characterized. We explored adtients and can also contribute to an impaired hemostatic rehesive and cohesive functions of platelets from cirrhotic sponse. 6,7 patients at different stages of disease development. The A complex defect of platelet function has been described in response of platelets to standard activating agents was these patients. 8-10 Impaired thromboxane A 2 (TxA 2 ) synthetested by aggregometric procedures. The interaction of sis, [11][12][13] defective signal transduction, 14 the presence of a platelets with subendothelial components was explored storage pool disease, 15 and quantitative defects in platelet in a perfusion system in which blood was exposed (shear glycoprotein Ib (GPIb), 16 among other defects, have been imrate, 800/s; 10 minutes) to denuded segments of rabbit plicated in cirrhotic thrombocytopathy (see Violi et al. 7 for aorta. Platelet interactions in these perfusions were anarecent review). Furthermore, alterations in coagulation lyzed morphometrically. Results were always compared mechanisms in cirrhotic patients can influence platelet funcwith those obtained in similar studies using blood obtion. Increased fibrinogen-depleted products, 8 acquired dysfitained from healthy subjects. Aggregation studies brinogenemia, 17 and changes in serum lipoproteins 18 have showed abnormal responses for single or several agobeen reported in patients with severe liver dysfunction. Hownists. Abnormalities in aggregation were more evident ever, the importance of the contribution of platelet dysfuncin patients with severe disease (Child-Pugh class C), altion to the hemostatic disturbance has not been elucidated though they occasionally were abnormal for single agocompletely. nists (ADP or U46619) in patients with less severe dis-Cirrhotic patients may have low hematocrit levels, another ease (Child-Pugh classes A or B). All the patient classes factor that can contribute to impaired hemostasis. The rheoshowed impaired platelet-subendothelial interactions (P logical effects of erythrocytes, although of critical importance õ .01 vs. healthy subjects) that were not justified by the for platelet function, are not taken into account in convenrelative thrombocytopenia present in the more severely tional laboratory tests used for the assessment of primary affected patients. Experimental increases in hematocrit hemostasis. in patients at stages B and C did not improve platelet-
The perfusion system devised by Baumgartner 19 provides subendothelial interactions. Platelets from cirrhotic paa useful model to investigate the interaction of platelets with tients interact defectively with subendothelial composubendothelial components in hemorrhagic disorders [20][21][22][23] and nents under flow conditions. The adhesion defect is to predict modifications in hemostasis by therapeutic procemore evident and consistent than the aggregation dedures. 21,24,25 Basically, this method provides a noninvasive fects and may already be present in patients with mild system that mimics the flow conditions taking place in vivo. liver failure. This adhesion defect may contribute to the A thrombogenic surface is exposed to circulating blood, and defective hemostasis observed in cirrhotic patients.
the platelet adhesive and cohesive functions can be quanti-(HEPATOLOGY 1996;24:1137-1142.)
fied morphometrically. The morphometric data obtained in perfusion studies parallel standard tests (i.e., bleeding time) Patients with cirrhosis frequently develop excessive bleedused to evaluate primary hemostasis. 21,26 ing during the course of the disease. The bleeding problem
In an attempt to clarify the nature of the platelet defect in cirrhotic patients has a multifactorial origin. 1,2 Gastropresent in cirrhotic patients, we explored adhesive and coheesophageal varices and portal hypertensive gastropathy repsive functions of platelets at different stages of the disease. resent a major factor in the bleeding history of these patients. The interaction of platelets with subendothelial components In addition, alterations of clotting factors are commonly seen was explored in a perfusion system in which blood was circuin these patients 3 ; a hyperfibrinolytic syndrome has been related at arterial rheological conditions (800/s). Parallel studies using standard aggregometric procedures were performed in the same patients. Results of previous studies were always Abbreviations: TxA2, thromboxane A2; GPIb, glycoprotein Ib; CPD, citrate-phosphate-compared with those obtained in studies using blood obtained dextrose; PRP, platelet-rich plasma; C, contact; A, adhesion; T, thrombi; CS, covered surface; from healthy donors. The possible contribution of relative vWF, von Willebrand factor. thrombocytopenia and anemia to the hemostatic defect was From the 1 Servicio de Hemoterapia y Hemostasia, 2 Servicio de GastroenterologıB a, and also considered.