The Significance of Protein C Antigen in Acute and Chronic Liver Biliary Disease

Transcription

1 The Signifiane of Protein C Antigen in Aute and Chroni Liver Biliary Disease SILVANA VIGANO, M.D., PIER MANNUCCIO MANNUCCI, M.D., ARMANDO D'ANGELO, M.D., MARIA G. RUMI, M.D., PAOLO VIGANO, M.D., ERSILIO DEL NINNO, M.D., AONIETTA CARGNEL, M.D., MASSIMO COLOMBO, M.D., AND MAURO PODDA, M.D. Protein C, a naturally ourring inhibitor of blood oagulation, was measured immunologially in 160 patients with aute and hroni liver and biliary disease. In 31 patients with aute viral hepatitis serially studied from admission to disharge from hospital, protein C antigen (PC:Ag) was low on admission in a high proportion of ases (61%) but beame normal in 9 of them after two weeks at a time when the prothrombin time was still prolonged in 46% of the ases. PC: Ag was also low in 25 irrhoti patients and in 20 patients with hroni ative hepatitis. In hroni hepatitis and irrhosis, PC:Ag levels signifiantly orrelated with indexes of liver syntheti funtion. In primary biliary irrhosis (n:40), PC:Ag was low in patients with advaned disease (stages III IV) but high in the early phases, when holestasis was not yet aompanied by impaired protein synthesis. PC:Ag was also very high in 20 patients with large bile dut obstrution and highly orrelated with indexes of holestasis. The authors' findings indiate that PC:Ag is redued in liver disease proportionally to the impairment of the liver syntheti funtion and that its normalization after aute hepatitis might represent an early marker of reovery of this funtion. (Key words: Protein C; Liver disease; Liver tests) Am J Clin Pathol 1985; 84: PROTEIN C (PC) is a vitamin K-dependent protein 6 " 19 that, upon ativation by thrombin and the endothelial ofator thrombomodulin, 7 exerts anti-oagulant properties by inativating fators Va and Villa 15 ' 25 at the ratelimiting steps of thrombin formation. The pathophysiologi importane of PC has been established by demonstrating that its ongenital defiieny is responsible for a thromboti diathesis. 2,9 Aquired PC defiienies also have been found. For instane, plasma levels of PC antigen (PC:Ag) or ativity were low in a small series of patients with hroni liver disease. 5 ' 814 In this study, we have hosen to extend these findings by measuring PC:Ag in other liver diseases suh as aute viral hepatitis, primary biliary irrhosis, and extrahepati holestasis due to large bile dut obstrution. We have also attempted to larify the mehanism underlying the hanges of PC:Ag in liver and biliary disease by studying the relationship between PC: Reeived November 16, 1984; reeived revised manusript and aepted for publiation January 25, Supported in part by a CNR grant n , Progetto Finalizzato Mediina Preventiva e Riabilitativa. Address reprint requests to Dr. Mannui: Via Pae 9,20122, Milano, Italy. A. Bianhi Bonomi Hemophilia and Thrombosis Centre, and the Third Institute of Clinial Mediine, University and Maggiore Hospital, and Department of Infetious Diseases, L. Sao Hospital, Milano, Italy Ag levels and liver tests in a large series of patients with aute and hroni liver and biliary disease. Samples and Methods One-hundred sixty patients with liver and biliary disease were studied. Chroni persistent hepatitis (CPH), hroni ative hepatitis (CAH), and ative irrhosis (AC) were diagnosed on the basis of morphologi riteria 13 in 24, 20, and 25 patients, respetively. Twenty-two patients were HB s Ag positive (Ausria II, Abbott, North Chiago, IL). Eighteen patients with CAH and 14 with AC were under treatment with steroids (10 mg prednisone daily). The diagnosis of 40 patients with primary biliary irrhosis (PBC) was based on linial, laboratory, and histologi findings. Twelve of them had morphologi evidene of sarring or irrhosis (Stages III IV). Twenty patients with extrahepati holestasis, due to obstrution of the large bile duts by arinoma of the panreas, also were studied. All the patients in the latter two groups were reeiving a monthly intramusular injetion of 10 mg vitamin K). Samples from 31 patients with linial and laboratory evidene of aute viral hepatitis (AVH) (3 with type A, 18 with type B, 10 with type nona-non B) were tested for PC:Ag and liver tests on admission, after two weeks, and at the time of disharge from hospital (four to five weeks). PC:Ag was measured by quantitative eletroimmunoassay, employing a monospeifi anti-serum against purified PC. 14 Results were expressed as perentage of pooled normal plasma defined as. The prothrombin time (PT) was arried out with a human brain thromboplastin supplied by Dr. Poller (Manhester Comparative Reagent) and results expressed as ratios of patient's to ontrol lotting time. Serum alanine aminotransferase (ALT), bilirubin, pseudoholinesterase (CHE), and alkaline phosphatase (AP) were assayed with ommerial kits (Boehringer Biohemia, Mannheim, West Germany). 454

2 Vol. 84 No. 4 PROTEIN C IN LIVER DISEASE Table 1. Demographi Features and Liver Funtion Tests 455 Aute Viral (n = 31) Chroni Persistent (n = 24) Chroni Ative (n = 20) (n = 25) Primary Biliary (n = 40) Large Bile Dut Obstrution (n = 20) Normal Values Age (mean and range) Sex (male/female) No Hb -Ag positive Pseudoholinesterase (U/L) Albumin (g/dl) Prothrombin time (ratio) Alkaline phosphatase (U/L) Bilirubin (mg/dl) Alanine aminotransferase (U/L) 27 (16-46) 16/ ,870 (1,670-2,040) 39% 1.24 ( ) ( ) 94% 7.5 ( ) 1023 (813-1,287) 41 (18-69) 13/11 9 2,889 (2,580-3,237) 4.4 ( ) 1.07 ( ) 13% 87 (79-97) 4% 34 (25-47) 9 46 (24-66) 13/7 8 2,776 (2,438-3,161) 4.3 ( ) ( ) ( ) 1 43 (32-57) 9 53 (35-70) 18/7 5 1,670 (1,451-1,925) ( ) ( ) 51 (43-60) Mean values. 95% onfidene limits (in parentheses) and prevalene of patients with abnormal values are given for liver funtion tests. Serum albumin was evaluated by densitometri sanning after ellulose aetate eletrophoresis. All data were log-transformed before statistial analysis and geometri means were obtained by antilog transformation of the log means. Ninety-five perent onfidene limits of the mean were alulated by antilog transformation of the values obtained by subtrating or adding to the log mean the log standard error multiplied by Statistial analysis inluded analysis of variane, the Student's Mest for paired and unpaired data as appropriate, FIG. I. Protein C antigen in patients with aute and hroni liver disease. The solid horizontal lines represent the normal range; the horizontal bars represent the geometri means, and the vertial bars show the 95% onfidene limits of the means. Patients with primary biliary irrhosis with advaned disease (Stages III- IV) are indiated by triangles. 49 (38-70) 2/38 0 2,250 (1,500-2,930) 3.6 (3.5-) 58% 1.01 ( ) 11% 768 ( ) 2.0 ( ) 55% 55 (38-62) 13/7 0 2,961 (2,730-3,412) 4.5 ( ) 0.98 ( ) 996 (881-1,091) 9.5 ( ) 44 (35-59) 4 1,900-3, <20 and orrelation oeffiients (r). Sine simple orrelations do not take into aount the interrelationships existing between liver tests, regression tehnis were used to alulate the r values between PC:Ag and eah liver test. Results Table 1 shows the demographi data and liver tests in patients subdivided aording to the type of liver or biliary disease. Figure 1 shows the distribution of PC: Ag values. p<0.01 p<0.001 p<0.001 ACUTE VIRAL CHRONIC PERSISTE CHRONIC ACTIVE ACTIVE CIRRHOSIS PRIMARY BILIARY CIRRHOSIS LARGE BILE DUCT OBSTRUCTION In.31) ( n.24) (n.20) <n.40) In.20)

3 456 VIGANO ET AL. A.J.C.P. Otober 1985 Table 2. Correlations (r values) between Liver Funtion Tests and Protein C Antigen, after Allowing for the Interrelationship among the Liver Funtion Tests Aute Viral Chroni Persistent Chroni Ative Ative Primary Biliary Large Bile Dut Obstrution Pseudoholinesterase Albumin Prothrombin time ratio Bilirubin Alanine aminotransferase Alkaline phosphatase P < t/><0.0l. tp< * 0.41* -0.66^ -0.43* 0) o II) Q. <D *-* ro O Q) o a J on admission after two weeks PC.Ag was not signifiantly different from normal in CPH and PBC, with a low prevalene of low values (17% and 1, respetively). However, when patients with PBC were subdivided aording to morphologi riteria, 12 patients with sarring or irrhosis (Stages III and IV) had signifip<0.005 on disharge FIG. 2. Changes of protein C antigen in 19 patients with aute viral hepatitis with low values on admission. The solid horizontal lines represent the normal range. P values indiate differenes between levels on admission and after two weeks, and between two weeks and disharge from hospital (Student's /-test for paired data). 0.48* 0.55f 0.55* 0.60f : = orrelation not signifiant. = orrelation not tested. 0.40* 0.39* It 0.71J 0.59* antly low PC:Ag (mean and 95% CL: 66%, 52-82, P < 0.001), whereas patients without irrhosis (Stages I and II) had high PC:Ag (13, ; P < 0.01). Patients with large bile dut obstrution also had signifiantly high PC:Ag (P < 0.001) (Fig. 1). Conversely, low levels were observed in patients with CAH (P < 0.05), AC (P < 0.001), and AVH on admission (P < 0.001) (Fig. 1), with a prevalene of abnormally low values of 2, 88%, and 61%, respetively. There were no signifiant differenes in mean PC:Ag levels between HB s Ag positive and HB s Ag negative patients, neither in patients with or without ortiosteroid treatment. PC:Ag was signifiantly orrelated with PT, albumin, CHE, and ALT in AVH; with albumin and CHE in CAH; with albumin, PT and CHE in AC; with PT, albumin, CHE, and bilirubin in PBC; and with bilirubin and AP in bile dut obstrution (Table 2). None of the liver tests orrelated with PC:Ag in CPH. Patients with AVH were tested three times for PC:Ag on admission, after two weeks, and at the time of disharge from hospital (Fig. 2). Mean PC:Ag, whih was signifiantly low on admission (P < 0.001), returned to normal after two weeks. Only two patients had persistently low values when disharged from hospital (Fig. 2). This pattern ontrasted with that of the liver tests, whih still had abnormal results after two weeks, and, in partiular, with that of the PT ratio, whih was still abnormal in 46% of the patients after two weeks and in 24% of them on disharge (Table 3). Disussion It has been suggested that DIC, whih is aompanied by low PC:Ag levels, 1014 ours in patients with hroni liver disease. 22 This view is supported by the presene of a shortened half-life of radiolabeled fibrinogen, prothrombin, and plasminogen in irrhoti patients and by the reported ability of heparin to orret these alterations. 421 In our patients with CAH and AC, PC:Ag levels were signifiantly orrelated with tests exploring the syn-

4 Vol. 84 No. 4 PROTEIN C IN LIVER DISEASE 457 theti funtion of the liver, suh as serum albumin PT and CHE, indiating that redued liver synthesis is the more plausible mehanism for low PC:Ag. However, a role for DIC annot be ompletely exluded, beause the orrelation of PC:Ag with syntheti funtion might reflet the more ommon appearane of DIC as liver disease worsens. This problem an only be solved by PC turnover studies, whih are not yet feasible beause of the diffiulties enountered in purifying a suffiient amount of this trae protein. Patients with PBC do not onstitute a homogenous group, sine very different linial and laboratory pitures an be found at different stages of the disease. 20 This situation is also refleted by PC:Ag levels, whih were sattered over a wide range. Although the mean value of PC: Ag was normal in the entire series, patients with the more advaned stage of disease had low levels of PC:Ag, omparable to those observed in AC. Sine these patients were regularly supplemented with vitamin K, low PC:Ag annot be aounted for by vitamin-k. defiieny. The orrelation between PC:Ag and indexes of liver syntheti funtion, suh as albumin CHE and PT, indiates dereased protein synthesis by the liver as a more likely mehanism. The negative orrelation between PC: Ag and bilirubin is not in ontrast with this explanation. In the late stages of PBC, when sarring and irrhosis develop, bilirubin inreases in serum proportionally to the deline of the liver syntheti funtions Conversely, high PC: Ag levels were found in the early stages of PBC, when sarring or irrhosis are not present, bilirubin is still normal, and holestasis is only refleted by high serum levels of bile aids PC:Ag levels, inreased in early PBC and dereased as irrhosis develops, might be used in linial pratie as an additional weapon to stage and follow PBC. High PC:Ag was also found in patients with extrahepati holestasis due to large bile dut obstrution by panreas arinoma, with a highly signifiant orrelation between PC:Ag and indexes of the degree of obstrution suh as bilirubin and AP. We had previously established that PC: Ag was normal in patients with tumors 23 and was not an aute phase reating glyoprotein. 23 Therefore, the rise of PC:Ag observed in the early phase of PBC and in patients with extrahepati holestasis is unlikely to be a nonspeifi onsequene of inflammation, nerosis, or neoplasti growth and might be regarded as another expression of the inreased synthesis of vitamin K. dependent proteins ourring in liver diseases assoiated with holestasis. 1,316 When measured on admission to the hospital during the most aute phase of AVH, PC:Ag was low in a large proportion of patients, the degree of redution and prevalene of abnormal values being roughly similar to those observed in AC. The diagnosti sensitivity of PC:Ag (61%) was similar to that of the PT (7) but inferior to those of ALT (), bilirubin (), and AP (94%), sug- Table 3. Serial Changes of Protein C Antigen, Prothrombin Time, and other Liver Tests in 31 Patients with Aute Viral Studied on Admission to the Hospital, after Two Weeks and at the Time of Disharge Protein C antigen (%) Albumin (g/dl) Prothrombin time (ratio) Alkaline phosphatase (U/L) Bilirubin (mg/dl) Alanine aminotransferase (U/L) On Admission 58 (50-68) 61% 39% 1.24 ( ) ( ) 94% 7.5 ( ) 1,023 (813-1,287) After Two Weeks 94* (81-109) 1 ( ) 1.20$ ( ) 46% 240* ( ) 77% * ( ) 9 285* ( ) On Disharge 100* (90-111) 6% 4.0t (-4.1) 32% 1.08f ( ) 24% 176* ( ) 52% 1.2* ( ) 53* (42-67) 78% Mean values, 95% onfidene limits (in parentheses) and the prevalene of patients with abnormal values are given. Asterisks indiate signifiant differenes over admission values (Students' Mest for paired data). ' P< t / < o.oi. % P < gesting that this measurement is of little diagnosti value. Our serial studies indiate that two weeks after admission PC:Ag levels returned to within the normal range in 9 of the patients, at a time when the PT was still prolonged in about half the ases and ALT and bilirubin were still abnormal in 9 of them. PT, a global oagulation sreening test affeted by the plasma levels of Fators I, II, V, VII, and X, but not of PC, is onsidered a better index of prognosis than bilirubin, transaminases, and albumin in AVH. 12 The half-life of PC is similar to that of Fator VII (six to eight hours) but muh shorter than that of the other lotting fators affeting the PT 24 and is likely to explain the early return of PC:Ag to normal levels when the liver syntheti funtion improves in AVH. Hene, PC:Ag might be used as a marker of early reovery of liver syntheti funtion in AVH. Our findings of low levels of PC:Ag in aute and hroni liver disease might have some general impliations for the pathophysiology of hemostasis in liver disease. Low lotting fators and fibrinolysis inhibitors, thromboytopenia, and high fibrinolysis ativators potentially favor a bleeding tendeny. It must be kept in mind, however, that low levels of the two main anti-oagulant proteins (antithrombin III and PC) and of the zymogen of fibrinolysis (plasminogen) 12 affet hemostasis in the opposite way and so tend to reestablish a balane at lower levels.

5 458 VIGANO ET AL. A.J.C.P. Otober 1985 Referenes 1. Amman R, Iwarson S, Olsson R: The linial signifiane of inreased plasma levels of liver synthesized oagulation fators in liver disease. Sand J Gastroenterol 1977; 12: Broekmans AW, Veltkamp JJ, Bertina RM: Congenital protein C defiieny and venous thromboembolism: A study of three Duth families. N Engl J Med 1981; 309: Cederblad G, Korsan-Bengsten K, Olsson R: Observations of inreased levels of blood oagulation fators and other plasma proteins in holestati liver disease. Sand J Gastroenterol 1976; 11: Collen D: Metabolism of plasminogen and prothrombin in irrhosis of the liver: Plasminogen and prothrombin metabolism in man. Edited by Collen D. Ao, Leuven, 1974, p Comp PC, Nixon RR, Esmon CT: Determination of funtional levels of protein C, an antithromboti protein, using thrombin-thrombomodulin omplex. Blood 1984: 63: Esmon CT, Stenflo JA, Suttie JW, Jakson CM: A new vitamin K dependent protein. A phospholipid binding zymogen of a serine protease. J Biol Chem 1976; 251: Esmon CT, Owen WG: Identifiation of an endothelial ell ofator for thrombin atalyzed ativation of protein C. Pro Natl Aad Si USA 1981: 78: Franis RB, Path MS: A funtional assay for protein C in human plasma. Thrbmb Res 1983; 32: Griffin JM, Evatt B, Zimmerman TS, Kleiss AJ, Wideman C: Defiieny of protein C in ongenital thromboti disease. J Clin Invest 1981;68: Griffin JM, Mosher DF, Zimmerman TS, Kleiss AJ: Protein C, an antithromboti protein, is redued in hospitalized patients with intravasular oagulation. Blood 1982; 80:26 j Kisiel W: Human protein C. Isolation, haraterization and mehanism of ativation by a-thrombin. J Clin Invest 1979; 64: Lehner K, Niessner H, Thaler E: Coagulation abnormalities in liver disease. Semin Thromb Haemostas 1977; 4: Leevy CM, Popper H, Sherlok S: Disease of the liver and biliary trat: Standardization of nomenlature, diagnosti riteria and diagnosti methodology. Chiago, Year Book Medial Publishers, 1976, p Mannui PM, Vigano S: Defiienies of protein C, an inhibitor of blood oagulation. Lanet 1984; 2: Marlar RA, Kleiss AJ, Griffin JM: Human protein C: Inativation of fators V and VIII in plasma by ativated moleule. Ann NY Aad Si 1981;370: Ritland S, Skrede S, BlomhoffT, Gjone E: Coagulation fators and indiators of protein synthesis in hroni liver disease. Sand J Gastroenterol 1973; 8(suppl 19): Roll J, Boyer J, Barry D, Klatskin G: The prognosti importane of linial and histologi features in asymptomati and symptomati primary biliary irrhosis. N Engl J Med 1983; 308: Shapiro JM, Smith H, Shaffner F: Serum bilirubin: A prognosti fator in primary biliary irrhosis. Gut 1979; 20: Stenflo J A: A new vitamin K-dependent protein. Purifiation from bovine plasma and preliminary haraterization. J Biol Chem 1976:251: Stephen PJ, Hoofnagle JH, Strober W, Jones A: NIH Conferene. Primary biliary irrhosis: A model autoimmune disease. Ann Intern Med 1983; 99: Tytgat GN, Collen D, Verstraete M: Metabolism offibrinogenin irrhosis of the liver. J Clin Invest 1971;50: Verstraete M, Vermyien J, Collen D: Intravasular oagulation in liver disease. Ann Rev Med 1974; 25: Vigano S, Mannui PM, D'Angelo A, et al: Protein C antigen is not an aute phase reatant and is often high in ishemi heart disease and diabetes. Thromb Hemostas 1984; 52: Vigano S, Mannui PM, Solinas S, Bottasso B, Mariani G: Derease in protein C antigen and formation of an abnormal protein soon after starting oral antioagulant therapy. Br J Haematol, 1984; 57: Walker F, Sexton P, Esmon CT: Inhibition of bjood oagulation by ativated protein C through the seletive inativation of ativated fator V. Biohim Biophys Ata 1979; 471: