Complement system is not activated in primary biliary cirrhosis

CLINICAL IMMUNOLOGY AND IMMUNOPATHOLOGY Vol. 87, No. 3, June, pp. 297–303, 1998 Article No. II984542 Complement System Is Not Activated in Primary Biliary Cirrhosis Marco Gardinali,* Luisa Conciato,* Cristina Cafaro,* Andrea Crosignani,† Pier Maria Battezzati,† Angelo Agostoni,* and Mauro Podda† *Institute of Internal Medicine, IRCCS Ospedale Policlinico, and †Institute of Internal Medicine, Ospedale San Paolo, University of Milan, 20100 Milan, Italy There is controversial evidence suggesting that the classical pathway of complement system is chronically activated in primary biliary cirrhosis (PBC) and that complement activation may be important in development of bile duct injury. We have reevaluated this issue by measuring by-products of complement activation such as C4a, C3a, Bb, and terminal complement complexes (SC5b-9) in plasma of 44 PBC patients with sensitive methods not previously used to detect complement activation in this disease. Agematched healthy women and patients with chronic hepatitis of different etiology were studied as controls. We found that PBC patients have normal C4a concentrations. This finding argues strongly against chronic classical pathway activation. Although a minor increase of C3a levels was observed in a minority of PBC patients, the C3a/C3 ratio, an index used to evaluate the extent of native protein conversion, was remarkably similar in all groups. Potentially lytic terminal complement complexes were not increased. PBC patients had normal Bb plasma levels, indicating that the alternative pathway is also not activated. C3 concentration was higher in PBC patients than in healthy subjects and in chronic hepatitis patients, particularly in the early stages of the disease. C3 and C4 concentrations became lower in PBC and chronic hepatitis with the progression of the disease. The increase of C3 concentration in PBC does not reflect liver inflammation, since serum levels of Creactive protein are normal. We found high serum C3 levels in patients with rare chronic cholestatic syndromes without superimposed infections and observed that serum C3 levels paralleled those of bilirubin in a patient with benign recurrent intrahepatic cholestasis. In conclusion, our data indicate that complement is not activated in PBC and that the increase of serum C3 levels is related to cholestasis. q 1998 Academic Press Key Words: complement system; primary biliary cirrhosis; chronic hepatitis; cholestasis. INTRODUCTION Primary biliary cirrhosis (PBC) is a chronic, cholestatic liver disease of unknown etiology which may eventually progress to liver failure and death (1, 2). Its frequent association with immune-mediated diseases and the finding of several disorders of both cellular and humoral immunity have suggested that some derangement of immune regulation is involved in the pathogenesis (4–9). Accordingly, a number of abnormalities of complement metabolism have been described in PBC patients (10–18). Earlier studies detected reduced levels of C4 in PBC patients, suggesting that the classical pathway is chronically activated, possibly by immune complexes (11, 12, 14–20). However, no study so far has directly measured the catabolic product of C4, C4a, in PBC. Moreover, plasma levels of some complement proteins (C1q, C2, C3, C5) were elevated rather than reduced as would be expected in case of activation (11, 16). In addition, other investigators did not find any circulating immune complexes in serum (21) or complement terminal complexes in bile ducts from PBC patients (22). There are several possible interpretations of these data. Methods to detect complement activation have variable accuracy and sensitivity. More important, the small numbers of previously investigated populations of PBC patients prevented taking into consideration the progress of the disease. It is well known, in fact, that patients with early-stage PBC show a cholestatic picture with limited necroinflammatory hepatocyte injury, while more advanced disease is characterized by progressive impairment of the liver’s synthetic function (23). Plasma levels of complement proteins depend on liver synthesis. Low serum levels of complement proteins are, therefore, expected in patients with late-stage PBC. Unfortunately, despite its potential interest, there is no available information regarding complement protein plasma levels in different stages of PBC. In the present study we have reevaluated the issue of complement system activation in a large number of PBC patients by measuring, in addition to plasma levels of C3 and C4, products of complement activation (C4a, C3a, Bb, SC5b-9), employing more sensitive methods than those used in earlier studies (24–26). These methods can detect levels of complement activation as low as 2–5%, as demonstrated in several studies 297 AID Clin 4542 / a520$$$101 06-04-98 17:48:14 0090-1229/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved. clina AP: Clin 298 GARDINALI ET AL. from different laboratories (27–29). To evaluate the effect of disease progression on complement activation, PBC patients were analyzed according to their clinical and histological stages (30, 31). We included in the study an age-matched control group consisting of apparently healthy subjects. To disentangle the effects of necroinflammatory liver damage and chronic cholestasis on complement parameters, two additional disease control groups were studied, the first consisting of patients with chronic hepatitis of different etiology and the second of patients affected by other cholestatic syndromes of different origin, with no features of autoimmunity and without superimposed infections (32–35). MATERIALS AND METHODS Patients and Study Design Forty-four women with PBC who consecutively attended the outpatient liver clinic at the San Paolo University Hospital in Milan during a period of 3 months were studied. An informed consent was obtained from each patient. Diagnosis of PBC was based on a combination of the following, widely accepted criteria: presence of cholestatic liver disease for at least 6 months, liver biopsy compatible with the diagnosis, a positive test for antimitochondrial antibodies, serum alkaline phosphatase levels twice the upper limit of normal values, absence of biliary obstruction as assessed by endoscopic cholangiography, and/or ultrasonography. Liver biopsy specimens obtained during the year preceding the study were classified according to Scheuer (31). Four patients had antibodies to hepatitis B virus (HBV) core antigen and 5 against hepatitis C virus (HCV) (EIA 3). Seventeen had antinuclear antibodies; 2 had anti-smooth-muscle antibodies. PBC patients frequently had autoimmune features. Nineteen had sicca syndrome; 5 had Raynaud’s phenomenon. One patient had scleroderma; 1 had systemic lupus erythematosus. Three control groups were studied. The first group consisted of 17 apparently healthy, age-matched (mean age, 51 { 10 years) women. The second group consisted of 23 age-matched (mean age, 52 { 12 years) patients with chronic hepatitis of viral (HCV, 10 patients, or HBV infection, 2 patients) or alcoholic (11 patients) etiology. Fourteen of the patients with chronic hepatitis had liver cirrhosis: 10 had Class A and 4 Class B or C cirrhosis, according to the Child–Turcotte classification as modified by Pugh et al. (30). The third control group consisted of 15 patients with different cholestatic conditions. Two of these patients had consecutive episodes of benign recurrent intrahepatic cholestasis (Summerskill–Walshe syndrome), 3 syndromic paucity of interlobular bile ducts (Alagille syndrome or arteriohepatic dysplasia), 5 idiopathic ductopenic syndrome of the adult, and 5 primary sclerosing cholangi- AID Clin 4542 / a520$$$102 06-04-98 17:48:14 tis (32–35). Serum autoantibodies were negative in all patients of this group. The patients with sclerosing cholangitis had no associated inflammatory bowel disease. Liver histology did not show inflammation in patients with idiopathic ductopenic syndrome of the adult. Blood samples were drawn from the patients with benign recurrent intrahepatic cholestasis during and between two consecutive episodes of jaundice. In the analysis involving the entire control group of cholestatic syndromes, the contribution of these 2 patients was limited to the data obtained at the cholestatic peak. Table 1 summarizes the main clinical and laboratory data of the different patient groups. Due to the rarity of some of the cholestatic conditions included in this control group, no attempt was made to match patients belonging to the PBC group for age. Since complement is activated in sepsis (27), we excluded patients for whom concomitant bacterial infections could be suspected on the basis of history, clinical examination, and routine laboratory tests. Methods of Analysis Serum levels of native proteins C3 and C4 and plasma levels of anaphylatoxins C4a and C3a were measured in all groups of patients. Plasma levels of the catabolic product of Factor B, Bb, and of the terminal complement components, SC5b-9, were measured in PBC patients and in healthy subjects. Plasma levels of C-reactive protein (CRP) were determined as an index of acute-phase reaction (36). Blood samples were collected in the morning after overnight fasting. Blood for anaphylatoxin measurement was collected in polystyrene tubes containing disodium ethylenediaminotetracetic acid (Na2EDTA; 7.5 mg/5 mL of blood). After centrifugation at 2500g for 15 min at 47C, plasma was adjusted to 0.2 mM phenylmethylsulfonyl fluoride to prevent in vitro activation. Plasma and serum samples were stored at 0807C until they were assayed. C3 and C4 were measured in serum by radial immunodiffusion (Nor Partigen, Behringwerke AG, Marburg Lahn, Germany). C3a and C4a were measured in plasma by radioimmunoassay (Amersham, England). The C3a/C3 and C4a/C4 ratios were calculated to evaluate the extent (in percent) of native protein conversion, according to the equation Ratio % Å A 1 MW N , N 1 MW A [1] in which A and N are the anaphylatoxin and corresponding native protein levels. Molecular weights (MW) of 9 (C3a and C4a), 205 (C4), and 185 kD (C3) were used (24). Plasma levels of the catabolic product of factor B, Bb, and the terminal complement components (SC5b- clina AP: Clin 299 COMPLEMENT IN PRIMARY BILIARY CIRRHOSIS TABLE 1 PBC No. of subjects: Male/female: 56 Agea (years): Laboratory dataa Albumin (g/dL) 3.9 Alkaline phosphatase 598 (U/L) Bilirubin (mg/dL) 1.0 IgM (mg/dL) 405 Child–Turcotte classificationb No cirrhosis Class A Class B–C Healthy women 44 0/44 { 12 (28–77) 51 17 0/17 { 10 (38–76) Chronic hepatitis 52 23 15/8 { 12 (25–75) Cholestatic syndromes 39 15 4/11 { 11 (23–55)** { 0.4 (3.2–4.8) { 346 (124–1436)* 4.2 { 0.6 (3.5–5.0) 210 { 80 (105–290) 3.9 { 0.4 (3.2–4.8) 223 { 92 (80–540) 4.3 { 0.5 (3.5–5.2) 447 { 285 (177–1192)* { 1.1 (0.3–6.7) { 193 (110–789)** 0.8 { 0.5 (0.4–1.5) 148 { 95 (70–210) 1.0 { 1.1 (0.3–6.7) 194 { 103 (47–418) 1.3 { 1.1 (0.6–5.1) 224 { 73 (128–383) 20 19 5 9 10 4 13 2 0 a Mean { SD; range in brackets. Number of subjects. * P õ 0.05 vs healthy women and chronic hepatitis. ** P õ 0.05 vs all other groups. b 9) were measured by enzyme immunoassay commercial kits (Quidel, San Diego, CA) according to manufacturer’s instructions. CRP levels were measured by radial immunodiffusion (Nor Partigen, Behringwerke AG). The method allows the detection of the protein when the concentration is higher than 0.5 mg/dL. groups of healthy subjects and patients with chronic hepatitis are shown in Figs. 1 and 2. C4a/C4 and C3a/ C3 ratios are also reported. Serum levels of C4 and C4a Statistical Analysis Data are reported as means { standard deviations ({SD) unless stated differently in the text. All complement data were log-transformed in order to achieve normal distributions and homogeneity of variances. The Student t test for independent samples was used to determine the significance of differences between continuous variables for the groups. Differences in laboratory data between the PBC and the control groups were analyzed by one-way analysis of variance. When the overall F test in the analysis of variance was significant, the Scheffè test was used to compare different pairs of means (37). The combined effects of liver disease (i.e., PBC or chronic hepatitis) and of disease stage, as defined by the Child–Turcotte classification (30) on complement parameters, were evaluated by two-way analysis of variance. When a significant effect of disease stage was found, the Scheffè test was used to determine significant differences between pairs of classes. P values õ0.05 were considered significant. Tests for two-tailed data were used. RESULTS Plasma levels of native protein C4 and C3 and of anaphylatoxin C4a and C3a in PBC patients and in the AID Clin 4542 / a520$$$102 06-04-98 17:48:14 FIG. 1. C4, C4a, and C4a/C4 ratios in patients with PBC, in age-matched healthy women, and in patients with chronic hepatitis. There are no significant differences between PBC and control groups. clina AP: Clin 300 GARDINALI ET AL. FIG. 2. C3, C3a, and C3a/C3 ratios in patients with PBC, in age-matched healthy women, and in patients with chronic hepatitis. Significance of differences between PBC and control groups are indicated. and C4a/C4 ratio were similar in PBC patients and in the two other groups. C3 and C3a were significantly higher in PBC, whereas the C3a/C3 ratio did not differ in the groups. Plasma levels of SC5b-9 (388 { 218 ng/mL vs 360 { 70 ng/mL) and Bb (1.08 { 0.37 vs 1.16 { 0.45 mg/ mL) were similar in patients with PBC and in healthy subjects. No differences were found between plasma levels of C4a, C3a, and SC5b-9 in PBC patients with and without autoimmune features (C4a, 500 { 366 vs 361 { 265 ng/mL; C3a, 425 { 236 vs 331 { 117 ng/mL; SC5b-9, 355 { 260 vs 414 { 180 ng/mL). However, particularly high levels of C4a (1735 ng/mL), C3a (1062 mg/mL), and SC5b-9 (672 ng/mL) were detected in one patient with PBC and scleroderma. Had this patient been excluded from the statistical analysis, C3a levels in PBC would not have been significantly different from those observed in patients with chronic hepatitis. Mean ({SD) serum C4 and C3 levels of PBC and chronic hepatitis patients at different clinical stages of the disease, as assessed by the Child–Turcotte classification, are shown in Fig. 3. With the progression of disease stage, there were significant decreases in both C4 and C3. This reduction was similar in PBC and AID Clin 4542 / a520$$$102 06-04-98 17:48:14 FIG. 3. Serum C3 and C4 mean (/SD) levels in patients with PBC (gray columns) and chronic hepatitis (black columns) by Child– Turcotte classification. There was no statistically significant interaction between disease groups and stages of disease. Significant differences among Child–Turcotte classes were demonstrated for C3 (P Å 0.001) and C4 (P õ 0.001). Asterisks indicate values significantly different from class B–C of corresponding disease (P õ 0.05). There were significant differences in C3 between disease groups (P õ 0.001). in patients with chronic hepatitis, as indicated by the absence of any significant statistical interaction between disease group and stage of disease. Patients with PBC, however, had significantly higher C3 levels, particularly in the early stages of disease. The mean ({SD) anaphylatoxin concentrations and native protein/anaphylatoxin ratios in patients with different stages of disease are reported in Table 2. The mean C4a and C3a levels were similar in the different Child–Turcotte classes of patients. A slight, but sig- TABLE 2 No cirrhosis (n Å 20) C4a (ng/mL) C4a/C4 ratio (%) C3a (ng/mL) C3a/C3 ratio (%) 441 4.4 349 0.8 { 166 { 7.1 { 203 { 0.6 Child A (n Å 19) 599 3.8 382 0.9 { 399 { 2.5 { 145 { 0.4 * P õ 0.05 vs no cirrhosis and Child A groups. clina AP: Clin Child B–C (n Å 5) 320 4.5 493 1.7 { 183 { 1.8 { 298 { 0.4* 301 COMPLEMENT IN PRIMARY BILIARY CIRRHOSIS FIG. 4. Serum levels (mg/dL) of C3 (gray circles) and C4 (black squares) in a patient with benign recurrent intrahepatic cholestasis, during and between consecutive episodes of jaundice. Dashed columns indicate total serum bilirubin levels (mg/dL). nificant increase in C3a/C3 ratio (but not C4a/C4) ratio was observed in Child B–C group (P õ 0.05). In the group of patients with other cholestatic syndromes, plasma levels of C3 and C4 were 109 { 36 and 35 { 12 mg/dL. While C4 levels were not different from those observed in the other groups, C3 levels were higher than in healthy subjects, patients with chronic hepatitis (P õ 0.001 for both comparisons), and PBC (P Å 0.02). The C3 concentration was weakly but significantly correlated with alkaline phosphatase levels in PBC (r Å 0.43, P Å 0.004) and in patients with other cholestatic syndromes (r Å 0.51, P Å 0.01). Figure 4 shows serum C3 and C4 levels in samples collected from a patient with benign recurrent intrahepatic cholestatis during and following a cholestatic episode. C3, but not C4, levels were increased when jaundice was present. Approximately 30% of all the plasma tested had detectable levels of CRP (ú0.5 mg/dL). Plasma CRP levels in PBC and control groups are reported in Table 3. There were no significant differences among mean levels of CRP of the four different groups. No correlation between C3 and CRP levels was observed in PBC patients as a group (P Å 0.15, NS) or in the ‘‘no cirrhosis’’ subset (P Å 0.28, NS). No correlation was found between C3 and CRP levels either in the cholestatic syndrome group (P Å 0.49, NS). DISCUSSION To reach a definite conclusion about the existence of any abnormality in complement metabolism in PBC we have used sensitive methods which have been shown to be suitable for detection of slight degrees of activation in other clinical conditions (27–29). PBC is characterized by different degrees of cholestasis, portal in- AID Clin 4542 / a520$$$102 06-04-98 17:48:14 flammation, fibrosis, and impairment of liver function (1, 2). To assess the influence of cholestasis separately from that of necroinflammatory liver damage, we also studied a group of patients affected by other cholestatic syndromes with no features of autoimmunity and a group of patients with chronic hepatitis of several origins. C4 and C4a levels were similar in PBC and in control subjects. This strongly argues against the hypothesis that there is a chronic activation of the classical complement pathway in PBC. The study of other catabolic products of complement activation points in the same direction. The mean levels of C3a were slightly elevated compared to healthy women and chronic hepatitis groups; however, C3a concentration was within the normal range in most of the plasma from PBC patients (Fig. 2). The complement system was overtly activated only in a woman with PBC and associated scleroderma. High levels of anaphylatoxins C4a and, more rarely, of C3a in patients with scleroderma have been previously reported by Wild et al. (38). Had this patient been excluded from the statistical analysis, C3a mean levels in PBC would not have been different from those observed in chronic hepatitis. Moreover, the ratio of C3a to C3, an index used to evaluate the extent of native protein conversion, was very similar in all groups. Normal levels of SC5b-9 definitely indicate that, should there be any activation of the complement system in PBC, it does not continue to involve potentially lytic terminal complement components. Finally, the determination of plasma levels of the catabolic product of factor B, Bb, rules out that there is alternative pathway activation in PBC, as previously suggested (10). Discrepancies between our studies and previous reports suggesting that complement is activated in PBC are, in our opinion, largely dependent on methodology. Most of these studies, in fact, demonstrated C3 conversion by counter- or rocket-immunoelectrophoresis, which are qualitative or, at best, semiquantitative methods. Split products of C3 (or C4) with different electrophoretic mobility were demonstrated in serum (18), but not in EDTA–plasma (15), suggesting that the TABLE 3 CRP (mg/dL) PBC Healthy women Chronic hepatitis Cholestatic syndromes No. of samples ú0.5 mg/dL Mean { SD Range 16/44 0.6 { 0.8 NDa —3.2 4/17 0.2 { 0.4 ND—1.2 4/23 0.3 { 0.7 ND—2.9 6/15 0.6 { 1.0 ND—2.7 Note. No significant difference among groups by one-way analysis of variance. a Nondetectable (i.e. õ0.5 mg/dL). clina AP: Clin 302 GARDINALI ET AL. complement activation observed in the former study occurred in vitro after clotting. Our study is the first one in which complement data were analyzed taking into consideration the stage of PBC. Figure 3 clearly shows that C3 and C4 levels decrease with the worsening of the disease. Potter et al. previously noted that the C3 and C4 levels were lower in PBC patients with than in those without ascites (14). The decrease of plasma concentration of C3 and C4 is due to the reduction of synthesis rather than to complement activation, as an effect of progressive necroinflammatory liver damage. Plasma levels of anaphylatoxins C4a and C3a, in fact, did not increase in the late stage of the disease (Table 2). A slight, albeit significant, increase in mean C3a/C3 ratio (but not in C4a/C4 ratio), however, has been observed in the Child B–C group. In three of these five patients a limited activation of C3 cannot be excluded, as the increase of C3a/C3 ratio is the result of increase of C3a to approximately twice the upper limit of the normal range, with contemporary decrease in C3 concentration. In two patients we observed exclusively a relevant decrease in C3 levels, with C3a levels within the normal range. The finding of high levels of C3 in the early phases of PBC remains to be explained. It has been shown that C3 behaves like an acute-phase reactant and that its plasma levels can increase as much as twofold in inflammatory diseases. CRP on the other hand is a much more sensitive marker of acute-phase reaction, increasing 100- to 1000-fold after tissue injury or inflammation (36). Most of the PBC patients had levels of CRP lower than 0.5 mg/dL; those who had detectable levels of the protein by radial immunodiffusion had concentrations usually between 1 and 2 mg/dL. As gene expression in hepatocytes of both CRP and C3 is responsive to stimulation by interleukin-1 and interleukin-6 (39, 40), the finding of normal levels of CRP in PBC argues against the hypothesis that the C3 increase is the result of acute-phase reaction. In the present study, we tested the hypothesis that the increase in serum C3 levels is associated with cholestasis. Elevated serum C3 levels have been reported in patients with cholestasis secondary to large bile duct obstruction by gallstones (16). However, this form of cholestasis develops acutely and is frequently associated with infection; therefore, increased levels of C3 may reflect inflammation rather than cholestasis. Inclusion of the group of patients with rare cholestatic syndromes and no superimposed infection or autoimmune features (32–35) enabled us to further investigate the relationship between alterations of the complement profile and cholestasis. The severity of cholestasis was similar in these patients and in PBC, as judged by serum alkaline phosphatase and bilirubin levels, and, as expected, we ruled out a significant inflammatory involvement on the basis of normal CRP values. We did find that serum C3 AID Clin 4542 / a520$$$102 06-04-98 17:48:14 levels are also elevated in these cholestatic conditions and are correlated with alkaline phosphatase levels. Furthermore, striking elevations of C3, but not of C4, levels were found during cholestatic episodes in benign recurrent intrahepatic cholestasis. These observations strongly support the hypothesis of a specific cholestasis-induced elevation of C3. Cholestasis may induce liver synthesis of other proteins (41–43). It is also possible that cholestasis impairs secretion of C3 into bile, although no study has so far identified complement proteins in bile, with the exception of C1 protease-inhibitor catabolic fragments (44, 45). Whatever the mechanism leading to C3 elevation in PBC, we have found that both serum C3 and C4 levels decrease with worsening of liver disease, when impairment of liver synthesis occurs. In conclusion, the complement system is not specifically involved in the pathogenesis of PBC. High serum C3 levels, related to cholestasis, may be found in PBC, particularly in the early stage of disease. The potential role of C3 and C4 as markers of disease progression still need to be evaluated in properly designed future studies. REFERENCES 1. Kaplan, M. M., Primary biliary cirrhosis. In ‘‘Disease of the Liver’’ (L. Schiff and E. R. Schiff, Eds.), pp. 377–410, Lippincott, Philadelphia, 1993. 2. Sherlock, S., Primary biliary cirrhosis: Clarifying the issues. Am. J. Med. 96(1A), 27S–33S, 1994. 3. Culp, K. S., Fleming, C. R., Duffy, J., Baldus, W. P., and Dickson, E. R., Autoimmune associations in primary biliary cirrhosis. Mayo Clin. Proc. 57, 365–369, 1982. 4. James, S. P., Hoofnagle, J. H., Strober, W., and Jones, E. A., Primary biliary cirrhosis: A model of autoimmune disease. Ann. Intern. Med. 99, 500–506, 1983. 5. Bahn, A. K., Dienstag, J. L., Wands, J. R., Schlossman, S. F., and Reinerz, E. L., Alterations of T-cell subsets in primary biliary cirrhosis. Clin. Exp. Immunol. 47, 531–537, 1982. 6. James, S. P., Elson, C. O., Jones, E. A., and Strober, S., Abnormal regulation of immunoglobulin synthesis in vitro in primary biliary cirrhosis. Gastroenterology 79, 242–246, 1989. 7. Lindgren, S., and Eriksson, S., IgM in primary biliary cirrhosis. Physicochemical and complement activating properties. J. Lab. Clin. Med. 99, 636–638, 1982. 8. Roberts-Thompson, P. J., and Sheperd, K., Low molecular weight IgM in primary biliary cirrhosis. Gut 31, 88–91, 1990. 9. Gershwin, E. C., and MacKay, I. R., Primary biliary cirrhosis: Paradigm or paradox for autoimmunity. Gastroenterology 100, 822–833, 1991. 10. Wands, J. R., Dienstag, J. L., Bahn, A. K., Feller, E. R., and Isselbacher, K. S., Circulating immune complexes and complement activation in primary biliary cirrhosis. N. Engl. J. Med. 298, 233–237, 1978. 11. Schlesinger, M., Benbassat, C., and Schoenfeld, Y., Complement profile in primary biliary cirrhosis. Immunol. Res. 11, 98–103, 1992. 12. Munoz, L. E., De Villiers, D., Markham, D., Whaley, K., and clina AP: Clin COMPLEMENT IN PRIMARY BILIARY CIRRHOSIS 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Thomas, H. C., Complement activation in chronic liver disease. Clin. Exp. Immunol. 47, 548–554, 1982. Danielsson, A., Nillson, T. K., and Uddenfeldt, P., Alterations in C1 inhibitor and clotting factor concentrations in primary biliary cirrhosis and other chronic liver diseases. Scand. J. Gastroenterol. 25, 149–154, 1990. Potter, B. J., Trueman, A. M., and Jones, E. A., Serum complement in chronic liver disease. Gut 14, 451–456, 1973. Lindgren, S., Laurell, A. B., and Erikkson, S., Complement components and activation in primary biliary cirrhosis. Hepatology 4, 9–14, 1984. Jones, E. A., Frank, M. M., Jaffe, C. J., and Vierling, J. M., Primary biliary cirrhosis and the complement system. Ann. Intern. Med. 90, 72–84, 1979. Potter, B. J., Elias, E., and Jones, E. A., Hypercatabolism of the third component of complement in patients with primary biliary cirrhosis. J. Lab. Clin. Med. 88, 427–439, 1976. Teisberg, P., and Gjone, E., Circulating conversion products of C3 in liver disease. Evidence for in vivo activation of the complement system. Clin. Exp. Immunol. 14, 509–514, 1973. Jones, E. A., and James, S. P., Circulating immune complexes and the pathogenesis of primary biliary cirrhosis. Gastroenterology 83, 709–711, 1982. Thomas, H. C., Potter, B. J., and Sherlock, S., Is primary biliary cirrhosis an immune complex disease? Lancet 2, 1261–1263, 1977. Goldberg, M. J., Kaplan, M. M., Mitamura, T., Anderson, C. L., Matloff, D. S., Pinn, V. W., and Agnello, V., Evidence against an immune complex pathogenesis of primary biliary cirrhosis. Gastroenterology 83, 677–683, 1982. Garred, P., Lyon, H., Christoffersen, P., Mollnes, T. E., and Trenum-Jensen, J., Deposition of C3, the terminal complement complex and vitronectin in primary biliary cirrhosis and primary sclerosing cholangitis. Liver 13, 305–310, 1993. Popper, H., Morphological and immunological studies on chronic aggressive hepatitis and primary biliary cirrhosis. In ‘‘Immunology of the Liver’’ (M. Smith and R. Williams, Eds.), p. 17, Heinemann, London, 1971. Wagner, J. L., and Hugli, T. E., Radioimmunoassay for anaphylatoxins: A sensitive method for determining complement activation products in biological fluids. Anal. Biochem. 136, 75–88, 1984. Kolb, W. P., Morrow, P. R., and Tamerius, J. D., Ba and Bb fragments of factor B activation: Fragment production, biological activities, neoepitope expression and quantitation in clinical samples. Complement Inflammation 6, 175–204, 1989. Schafer, H. J., Mathey, D., Hugo, F., and Bakhdi, S., Deposition of the terminal C5b-9 complement complex in infarcted area of human myocardium. J. Immunol. 137, 1945–1949, 1986. Gardinali, M., Padalino, P., Vesconi, S., Calcagno, A., Ciappellano, S., Conciato, L., Chiara, O., Agostoni, A., and Nespoli, A., Complement activation and polymorphonuclear neutrophil leukocyte elastase in sepsis: Correlation with severity of diseases. Arch. Surg. 127, 1219–1224, 1992. 28. Chenoweth, D. E., Cooper, S. W., Hugli, T. E., Stewart, R. W., Blackstone, E. H., and Kirklin, J. W., Complement activation during cardiopulmonary bypass: Evidence for generation of C3a and C5a anaphylatoxins. N. Engl. J. Med. 304, 497–503, 1981. 29. Frangi, D., Gardinali, M., Conciato, L., Cafaro, C., Pozzoni, L., and Agostoni, A., Abrupt complement activation and transient neutropenia in patients with acute myocardial infarction treated with streptokinase. Circulation 89, 76–80, 1994. 30. Pugh, R. N. H., Murray-Lyon, Dawson, J. L., Pietroni, M. G., and Williams, R., Transection of the esophagus for bleeding oesophageal varices. Br. J. Surg. 60, 646–649, 1973. 31. Scheuer, P. J., Primary Biliary Cirrhosis. In ‘‘Liver Biopsy Interpretation,’’ p. 47, Williams & Wilkins, Baltimore, 1986. 32. Lee, Y. M., and Kaplan, M. M., Primary sclerosing cholangitis. N. Engl. J. Med. 332, 924–933, 1995. 33. Summerskill, W. H. S., and Walshe, J. M., Benign recurrent intrahepatic ‘‘obstructive’’ jaundice. Lancet 2, 686–690, 1959. 34. Woolf, G. M., and Vierling, G. M., Disappearing intrahepatic bile ducts: The syndromes and the mechanisms. Semin. Liver Dis. 13, 261–275, 1993. 35. Alagille, D., Estrada, A., Hadchouel, M., Gautier, M., Odievre, M., and Domergues, J. P., Syndromic paucity of interlobular ducts (Alagille’s syndrome or arteriohepatic dysplasia): Review of 80 cases. J. Pediatr. 110, 195–200, 1987. 36. Dowton, S. B., and Colten, H. R., Acute phase reactants in inflammation and infection. Semin. Hematol. 25, 84–90, 1988. 37. Winer, B. J., ‘‘Statistical Principles in Experimental Design,’’ 2nd. ed., pp. 196–201, McGraw–Hill, New York, 1971. 38. Wild, G., Watkins, J., Milford, Ward, A., Hughes, P., Hume, A., and Rowell, N. R., Complement activation in systemic sclerosis. J. Clin. Lab. Immunol. 31, 39–41, 1990. 39. Wilson, D. R., Juan, T. S. C., Wilde, M. D., Fey, G. H., and Darlington, G. J., A 58-base-pair region of the human C3 gene confers synergistic inducibility by interleukin-1 and interleukin-6. Mol. Cell. Biol. 10, 6181–6191, 1990. 40. Ganter, U., Arcone, R., Toniatti, C., Morrone, G., and Ciliberto, G., Dual control of C-reactive protein gene expression by interleukin-1 and interleukin-6. EMBO J. 8, 3773–3779, 1989. 41. Kaplan, M. M., and Righetti, A., Induction of rat liver alkaline phosphatase: The mechanism of the serum elevation in bile duct obstruction. J. Clin. Invest. 49, 508–516, 1970. 42. Schulz, B. O., and Schlager, R., Evidence for an enhanced de novo synthesis of alkaline phosphatase in cholestatic rat liver using immunotitration and precursor incorporation techniques. Enzyme 24, 173–180, 1979. 43. LaRusso, N. F., Proteins in bile: How they get here and what they do. Am. J. Physiol. 247, G199–G205, 1984. 44. Dive, C., and Heremans, J. F., Nature and origin of the proteins of bile. I. A comparative analysis of serum and bile proteins in man. Eur. J. Clin. Invest. 4, 235–239, 1974. 45. Johansson, J., Grondal, S., Sjovall, J., Jornvall, H., and Curstedt, T., Identification of hydrophobic fragments of a1 antitripsin and C1 protease inhibitor in human bile, plasma and spleen. FEBS Lett. 2, 146–148, 1992. Received June 24, 1997; accepted with revision February 16, 1998 AID Clin 4542 / a520$$$103 06-04-98 17:48:14 303 clina AP: Clin