I. Introduction

Top Next References

Clinical signs and symptoms of several diseases result from the release and activation of endogenous inflammatory mediators. Among these mediators are plasma cascade systems such as the contact and complement system. Activation of these systems has indeed been demonstrated in a variety of human diseases. A typical feature of both the contact and the complement system is that on activation they give rise to vasoactive peptides such as bradykinin (contact system) or the anaphylatoxins (complement system), which in part explains the proinflammatory effects of either system (Mason and Melmon, 1965; Vogt, 1986).

C1-inhibitor (C1-Inh)² is a major inhibitor of both the complement and the contact system (Sim et al., 1979; Van den Graaf et al., 1983; Pixley et al., 1985; Wuillemin et al., 1995b), and, therefore, is endowed with anti-inflammatory properties (Table 1). The important physiological role of C1-Inh is best demonstrated by hereditary C1-Inh deficiency and its association with hereditary angioedema (HAE) (Landermann et al., 1962; Donaldson and Evans, 1963). The first detailed description of the clinical signs and the dominant inheritance of HAE was done by Quinke and Osler (Osler, 1888; Quinke, 1882). Lepow and colleagues demonstrated that human serum contained a heat-labile protein that inhibited the esterolytic activity of the first complement component C1, which they named "C1-Inhibitor" and subsequently achieved its partial purification (Lepow et al., 1956; Levy and Lepow, 1959; Pensky et al., 1961). Finally, it was shown that serum of patients with HAE was deficient in kallikrein inhibitory capacity and that patients with HAE had significantly decreased serum levels of C1-Inh (Landermann et al., 1962; Donaldson and Evans, 1963). HAE is characterized by episodes of painless local swelling of soft tissues resulting from a local increase of vasopermeability. Only heterozygous conditions are known to be associated with the disease and a homozygous deficiency is apparently lethal (Späth, 1997).

Attacks of HAE can be treated effectively by intravenous administration of C1-Inh purified from pooled human plasma. Long-term prophylactic substitution with pasteurized C1-Inh was demonstrated to be safe and of clinical benefit with few negative effects (Gadek et al., 1980; Bergamaschini et al., 1983; Bork and Witzke, 1989; Agostoni and Cicardi, 1992; Waytes et al., 1996).

Because of its anti-inflammatory (and anti-clotting) activity and the possibility of its large scale preparation with a high degree of purity, biological activity, and viral safety, C1-Inh concentrates may be useful for the treatment of other diseases as well (Poulle et al., 1994). Recent studies support this idea. Here, we will summarize the biochemistry and biology of C1-Inh. We then will review the results of experimental and clinical studies evaluating the therapeutic efficacy of C1-Inh therapy in several diseases other than angioedema. Finally, we will discuss a hypothesis that may explain the efficacy of C1-Inh in these diseases.

	II. Biochemistry and Biology of C1-Inh

Top Previous Next References

C1-Inh is a heavily glycosylated single-chain polypeptide of 478 amino acid residues, the protein portion of the molecule constituting only 51% of its molecular mass (Bock et al., 1986). It probably contains 13 carbohydrate groups, i.e., 6 glucosamine-based and 5 galactosamine-based, whereas the remainder is linked to threonine residues. Most carbohydrate groups are located at the N-terminal region (Perkins et al., 1990). The precise function of these carbohydrate groups is unknown. Removal of sialic acids from C1-Inh (asialo-C1-Inh) enormously enhances its clearance from the circulation yielding an apparent half-life time of 3 to 5 min in a rabbit model (Minta, 1981), presumably via binding to asialoglycoprotein receptors in the liver. The enhanced clearance of asialo-C1-Inh is due to exposure of penultimate galactosyl residues, because the subsequent removal of the latter prolongs the clearance rate up to a value similar to that of normal C1-Inh (Minta, 1981). Removal of sialic acid or galactose groups does not impair the functional activity of C1-Inh in vitro (Minta, 1981).

The molecular mass of C1-Inh is approximately 105 kDa, and plasma concentration ~240 mg/l, corresponding to 1 U/ml (Schapira et al., 1985; Nuijens et al., 1989).

By sequence homology C1-Inh belongs to the superfamily of serine proteinase inhibitors (serpins), which also includes, e.g., 1-antitrypsin, antithrombin, and plasminogen activator inhibitor type I (Travis and Salvesen, 1983; Carrell and Travis, 1985; Schapira et al., 1985; Carrell and Boswell, 1990). As with other serpins, the sequence homology is not distributed throughout the molecule but is restricted to the carboxyl-terminal end (serpin domain) (Carter et al., 1988; Coutinho et al., 1994). Near the carboxyl-terminal end of the serpin domain is the protease recognition region, which is termed the "reactive center loop" (Coutinho et al., 1994).

Several cells, including hepatocytes, fibroblasts, monocytes, macrophages, endothelial cells, amnionic epithelial cells, and perhaps cells in the mikroglia are capable of C1-Inh synthesis (Bensa et al., 1983; Yeung Laiwah et al., 1985; Katz and Strunk, 1989; Katz et al., 1995; Walker et al., 1995; Zahedi et al., 1997a). In rats, spontaneous expression of C1-Inh gene was found in Kupffer cells, whereas peritoneal macrophages and blood monocytes expressed C1-Inh only after treatment with interferon- (IFN-) (Armbrust et al., 1993). Human platelets contain C1-Inh in their -granules, and the platelet levels of C1-Inh correlate with plasma C1-Inh levels (Schmaier et al., 1993). Activated platelets can express C1-Inh on their external membrane (Schmaier et al., 1993). Because plasma C1-Inh does not bind and get absorbed onto platelets, it was suggested that the platelet C1-Inh level depends on synthesis of C1-Inh in megakaryocytes. In agreement herewith, human megakaryocytes contain C1-Inh mRNA (Schmaier et al., 1993).

Synthesis of C1-Inh is stimulated by IFN- and, to a lesser extent, by several other cytokines including tumor necrosis factor  (TNF), IFN-, monocyte colony-stimulating factor, and interleukin-6 (IL-6) (Lotz and Zuraw, 1987; Katz and Strunk, 1989; Heda et al., 1990, 1996; Zuraw and Lotz, 1990; Schmidt et al., 1991; Lappin et al., 1992; Schwogler et al., 1992). Studies on human erythroleukemia cells and on plasma specimens of patients receiving intravenous IFN- because of metastatic colorectal carcinoma indicated that IFN- can increase C1-Inh protein expression in vitro and in vivo (Heda et al., 1990). In purified blood monocytes, the release of functional C1-Inh was markedly increased in the presence of IFN-, but less with IFN- or IFN- (Lotz and Zuraw, 1987). It was shown that induction of C1-Inh mRNA in IFN--stimulated cells is primarily due to the enhanced transcription rate of its gene (Zahedi et al., 1994). The IFN--responsive element has been characterized and located in the 5'-flanking region of the C1-Inh gene (Zahedi et al., 1997b). Studies with human hepatoma cell line cultures (HepG2) indicated that phosphatase 2A is required to dephosphorylate a substrate to allow IFN- to induce transcriptional up-regulation of C1-Inh mRNA (Heda et al., 1996). However, in a small study with six volunteers and two patients with type I angioedema, a 4-day course of administration of IFN- failed to influence plasma levels of C1-Inh (Gluszko et al., 1994).

C1-Inh is an acute phase protein, the plasma levels of which may increase up to 2-fold during uncomplicated infections (Kalter et al., 1985). The synthetic rate of C1-Inh may increase to up to 2.5 times the normal rate in patients with rheumatoid arthritis (Woo et al., 1985). This increased synthesis during acute phase responses is probably the result of the release of IL-6, because in vitro studies with human HepG2 cells have shown IL-6, and to a lesser extent IL-1, to increase the biosynthesis of C1-Inh as well as of the complement components C3 and of factor B (Falus et al., 1990).

C1-Inh is encoded by a 17-kilobase single-copy gene on chromosome 11, which consists of eight exons separated by seven introns (Davis et al., 1986; Theriault et al., 1990; Carter et al., 1991). The first intron contains IFN--responsive elements that are functional in vitro and may play a role in IFN--mediated induction of C1-Inh synthesis (Zahedi et al., 1997a). The second exon contains the translation initiation site, whereas the DNA encoding the reactive center sequence is situated within exon 8 (Donaldson and Bissler, 1992). Approximately 20% of described mutations of the C1-Inh gene are large deletions or duplications, and a substantial proportion of mutations are localized within exon 8, the coding region for the reactive center of C1-Inh (Bissler et al., 1997).

Interaction of C1-Inh with target proteinases results in the formation of SDS-stable enzyme-inhibitor complexes and proteolytically cleaved C1-Inh (Schapira et al., 1988). Analogous to other serpins, C1-Inh inhibits a target proteinase by presenting a peptidyl bond (P1-P1') lying on an exposed loop within the reactive center that matches the substrate specifity of the proteinase. Attack on this peptidyl bond, connecting residues P1 (Arg-444) and P1' (Thr-445), results in the formation of a complex between inhibitor and proteinase (Sim et al., 1979; Van den Graaf et al., 1983; Pixley et al., 1985). The importance of the P1 residue and this peptidyl bond with regard to the binding capacity to target proteinases was demonstrated by construction of various P1 substitutions that resulted in nonfunctional molecules in the majority of cases (Eldering et al., 1992).

The complexes formed between C1-Inh and proteinase are removed from the circulation with an apparent half-life time of clearance ranging from 20 to 47 min (Nuijens et al., 1988; De Smet et al., 1993; Wuillemin et al., 1996a). The complexes were found to be removed via receptors specific for complexed serpins, such as the low-density lipoprotein receptor-related protein on hepatocytes or fibroblasts (Pizzo et al., 1988; Perlmutter et al., 1990; De Smet et al., 1993; Storm et al., 1997). Neutrophils and monocytes have been shown to express a not yet completely characterized serpin-enzyme complex receptor, which binds, internalizes, and degrades several serpin-proteinase complexes, including those with 1-antitrypsin, 1-antichymotrypsin, antithrombin, and to a lesser extend C1-Inh (Perlmutter et al., 1990).

C1-Inh, like most other serpins, can be inactivated by elastase released from activated neutrophils by limited proteolytic cleavage resulting in the production of several different and characteristic derivatives, so called "modified C1-Inh" (Brower and Harpel, 1982; Carrell et al., 1987; Weiss, 1989). C1-Inh mutants with a decreased susceptibility for inactivation by elastase have been developed, but their therapeutic efficacy remains to be established (Eldering et al., 1993). Human proteinase 3, isolated from human leukocytes, cleaves and inactivates human C1-Inh in a time- and dose-dependent manner (Leid et al., 1993). Likewise, bacterial elastases and proteinases were shown to proteolytically cleave and inactivate C1-Inh (Catanese and Kress, 1984). Finally, plasmin was found to play a role in the local cleavage and degradation of C1-Inh in inflammatory processes (Wallace et al., 1997). Also thrombin may inactivate C1-Inh (M. Cugno, I. Bos, and C. E. Hack, unpublished results), although the significance of this process remains to be established.

Thus, the inactivation of C1-Inh may predominantly occur locally in inflamed tissues and, therefore, contribute to increased local complement activation and consumption as well as to local potentiation of pathological proteolysis (Brower and Harpel, 1982; Leid et al., 1993). This conclusion is supported by the demonstration of increased plasma levels of modified C1-Inh in patients with sepsis (Nuijens et al., 1989).

In normal volunteers the fractional catabolic rate of C1-Inh is 2.5% of the plasma pool per hour, yielding an apparent plasma half-life time of clearance of ~28 h (Quastel et al., 1983; Woo et al., 1985). The half-life time of clearance of human C1-Inh in rabbits is comparable, i.e., 26 h (Minta, 1981), whereas in rats it is considerably shorter, i.e., ~4.5 h (De Smet et al., 1993). The apparent half-life time of clearance has been reported to be considerably longer in patients with HAE, in whom it may be 48 h (Agostoni et al., 1980; Gadek et al., 1980). It is to be noted, however, that the clearance half-life times of C1-Inh in these patients are often determined by assessing the course of plasma levels following the intravenous administration of ~1000 U. Presumably it is not correct to determine the half-life time of clearance in these patients in this way, because it is not considered that at lower plasma levels of C1-Inh (as occurs in untreated HAE patients), the first component of complement, C1, is autoactivated, which causes consumption of functional C1-Inh. At higher concentrations of C1-Inh (as occur after administration of C1-Inh), this autoactivation is inhibited leading to a decreased consumption of C1-Inh. Hence, following a therapeutic dose of C1-Inh, plasma concentrations of C1-Inh increase because of the administration of exogenous C1-Inh as well as a reduced consumption of endogenous C1-Inh.

We have administered high doses of C1-Inh (up to 12,000 U over a period of 2 to 5 days) to 12 patients with septic shock (Hack et al., 1993; Hack, 1996). Plasma C1-Inh concentration was measured at various time-points during the study period. To calculate the recovery of C1-Inh in these patients, we used a pharmacokinetic model assuming that, a) the fractional catabolic rate of C1-Inh is 2.5% of the plasma pool per hour; b) the C1-Inh concentration at a given time after administration of exogenous C1-Inh is described by the sum of a constant concentration due to endogenous production and a concentration increase resulting from the C1-Inh administration; c) the C1-Inh increase resulting from the administration of exogenous C1-Inh is equal to the summation of the concentration effects of each subsequent administered exogenous dose, distributed immediately in one central plasma compartment, and constantly eliminated from there following a first order process; and d) the plasma volume in patients with sepsis is approximately 45 ml/kg of body weight. The overall correlation between the course of C1-Inh levels after the various administrations of C1-Inh calculated according to this model and those actually measured in the patients was very significant (r = 0.7807, P < .0001; A. C. Ogilvie, C.E.H., L. G. Thijs, J. Wagsteff, unpublished observations), although on some occasions in individual patients, recovery was less than expected, possibly due to a higher fractional catabolic rate (Hack et al., 1993). These results indicate that the clearance data observed with radiolabeled C1-Inh in human volunteers (Quastel et al., 1983; Woo et al., 1985) may be used to calculate the dose of C1-Inh to be administered to patients. Recently, we have evaluated the effects of high doses of C1-Inh in septic baboons. C1-Inh was administered to yield 5- to 10-fold increased levels over a period of 8 h. The observed course of C1-Inh exactly matched that calculated from the clearance data described above (Jansen et al., 1998), again demonstrating the validity of the pharmacokinetic model.

	III. C1-Inh As Inactivator of Plasma Cascade Systems and Leukocytes

Top Previous Next References

C1-Inh is the only known inhibitor of the activated serine proteinases C1s and C1r from the classical pathway of complement, and is a major inhibitor of activated factor FXII (FXIIa) from the contact system, as well as an inhibitor of kallikrein and activated factor XI (FXIa) (Chan et al., 1977; Sim et al., 1979; Ziccardi, 1981; Schapira et al., 1982, 1985; Van den Graaf et al., 1983; Cooper, 1985; Pixley et al., 1985; Scherer et al., 1996; Wuillemin et al., 1995b) (see Fig. 1). C1-Inh is, therefore, an important regulator of inflammatory reactions and of the intrinsic pathway of coagulation.

View larger version (29K): [in this window] [in a new window]   Fig. 1.   Activation of plasma cascade systems. During host defense reactions, complement, coagulation, and the fibrinolytic system are activated. C1-Inh is the only known inhibitor of C1s and C1r of the classical pathway of the complement system, as well as a major inhibitor of FXIIa and kallikrein of the contact system and of FXIa, which links the contact activation system to the coagulation cascade. FI, Factor I; FH, Factor H; Ag, antigen; Ab, antibody; LPS, lipopolysaccharides; AT, antithrombin; TFPI, tissue factor pathway inhibitor; APC, activated protein C; t-PA, tissue plasminogen activator; 2-AP, 2-antiplasmin; shaded down arrow, action of C1-Inh (or other inhibitors).

The complement system consists of more than 30 serum and cellular proteins linked in three biochemical cascades, the classical and the alternative pathway (Makkrides, 1998) and the mannan-binding lectin (MBL) pathway (see Fig. 2). The classical pathway is usually initiated when a complex of antigen and immunoglobulin M (IgM) or IgG antibody binds to the first component of complement, C1. Activated C1 cleaves both C4 and C2 to generate C4a and C4b, and C2a and C2b, respectively. The C4b and C2a fragments combine to form the classical C3 convertase, which cleaves C3 to form C3a and C3b. The binding of C3b to the C3 convertase yields the C5 convertase, which cleaves C5 into C5a and C5b, the latter becoming part of the membrane attack complex (MAC; Makkrides, 1998).

View larger version (27K): [in this window] [in a new window]   Fig. 2.   The complement system. The complement system consists of three plasmatic cascades each leading to the formation of the C3 and C5 convertases. The classical pathway is initiated when antigen-antibody complexes bind to C1 or by cleavage of C1 by FXIIa. The alternative pathway is triggered by microbial surfaces and complex polysaccharides. The MBL pathway is triggered by polysaccharides in particularly those occurring on bacterial surfaces. The common final step of complement activation is the activation of C5 by the C5 convertases that leads to the formation of the MAC C5b-9, which finally inserts into target cell membranes and causes cell lysis. The peptides C3a, C4a, and C5a are known as anaphylatoxins and mediate several reactions in the inflammatory response, including smooth muscle cell contractions, changes in vascular permeability, chemotaxis for human mast cells, histamine release from mast cells, neutrophil chemotaxis, and platelet activation and aggregation. Ag-Ab complexes, antigen-antibody complexes; LPS, lipopolysaccharides; C, complement.

The alternative pathway of the complement system is triggered by microbial surfaces and a variety of complex polysaccharides. C3b, formed by the spontaneous low-level cleavage of C3, can bind to nucleophilic targets on cell surfaces and forms a complex with factor B that is subsequently cleaved by factor D. The resulting alternative C3 convertase is stabilized by the binding of properdin. Cleavage of C3 and binding of an additional C3b to the C3 convertase give rise to the C5 convertase. The C3 and C5 convertases of the alternative pathway are controlled by complement receptor type 1, decay-accelerating factor, membrane cofactor protein, and by factor H (Makkrides, 1998). The C5 convertase cleaves C5 to produce C5a and C5b. Thereafter, C5b sequentially binds to C6, C7 and C8 to form C5b-8 that catalyzes the polymerization of C9 to form the MAC, which inserts into target membranes and causes cell lysis (Hu et al., 1981; Podack et al., 1982; Tschopp et al., 1982). Vitronectin and similarly clusterin control fluid phase MAC by binding to the C5b-7 complex to prevent its insertion into membranes (Podack et al., 1977; Jenne and Tschopp, 1989). C8 binding protein (CD59) blocks MAC formation by binding to C8 and C9 (Rollins et al., 1991).

The MBL pathway is triggered by binding of MBL to polysaccharides of various microbes (Turner, 1996). Subsequently, MBL stimulates the activation of MBL-associated serine proteinase-1 (MASP-1) and MASP-2 (Matsushita and Fujita, 1992; Thiel et al., 1997). MASP-1 and MASP-2 can activate C4 (International Complement Workshop, Rhodes, Greece, October 1998), leading to classical pathway activation. MASP-1 and MASP-2 can be inhibited by C1-Inh.

The peptides C3a, C4a, and C5a are known as anaphylatoxins (Hugli and Müller-Eberhard, 1978). They mediate several reactions in the inflammatory response, including smooth muscle contractions, changes in the vascular permeability, chemotaxis for human mast cells, histamine release from mast cells, neutrophil chemotaxis, and platelet activation and aggregation (Hugli and Müller-Eberhard, 1978; Morgan, 1986; Gerard and Gerard, 1994; Hartmann et al., 1997). The anaphylatoxins are rapidly inactivated by carboxypeptidase N (Bokisch and Müller-Eberhard, 1970).

Antibody-mediated complement activation at the cell surface has been demonstrated to result in increased tissue factor activity, indicating that complement fixation on the cell surface can have a direct stimulatory effect on the coagulation cascade (Carson and Johnson, 1990).

Activation of the classical pathway of complement is regulated by C1-Inh. It is the only known inhibitor of the activated serine proteinases C1s and C1r from the classical pathway of complement (Sim et al., 1979; Schapira et al., 1985). C1-Inh either binds reversibly to proenzymic C1r and C1s within intact C1 to prevent the autoactivation of these proteinases (Bianchino et al., 1988) or binds to activated C1r and C1s and dissociates them from C1q in the form of a C1-Inh-C1r-C1s-C1-Inh tetramer (Sim et al., 1979; Liszewski et al., 1996). The rate of inhibition of C1r by C1-Inh is significantly slower than that of C1s (Sim et al., 1980). Interaction of C1-Inh with either activated C1r or C1s results in the formation of cleaved C1-Inh and an SDS resistant enzyme-inhibitor complex (Harpel and Cooper, 1975; Reboul et al., 1977; Arlaud et al., 1979; Ziccardi and Cooper, 1979; Chesne et al., 1982; Salvesen et al., 1985).

C1-Inh-C1s complexes were shown to be finally removed by the low-density lipoprotein receptor-related protein of murine fibroblasts and probably of hepatocytes but did not bind to the serpin-enzyme complex receptor of HepG2 cells, neutrophils, or monocytes nor to the hepatic asialoglycoprotein receptor (Storm et al., 1997).

FXII, prekallikrein, high-molecular weight kininogen (HK), and FXI are grouped together as "contact system," because they require contact with negatively charged surfaces for zymogen activation (Schmaier, 1997) (see Fig. 3). In vitro, FXII and prekallikrein reciprocally activate each other on contact with macromolecules such as kaolin, glass, celite, or dextran sulfate (Colman, 1984; Kaplan and Silverberg, 1987). In addition, FXII is able to autoactivate (Kaplan and Silverberg, 1987). Activation of FXII leads to the formation of the activated fragments FXIIa (or FXIIa) and FXII (or FXIIf). The strictly surface-dependent FXIIa converts FXI to FXIa, whereas FXIIa is an effective prekallikrein activator (Kaplan and Silverberg, 1987). Soluble FXIIa has been shown to activate the first component of complement (Ghebrehiwet et al., 1983). Activated FXII has the ability to cleave plasminogen, rendering it to a weak activator of the fibrinolytic system (Colman et al., 1975).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Contact activation system. The contact system consists of the proteins FXII, prekallikrein, HK, and FXI. FXII and prekallikrein activate each other upon contact with negative surfaces (kaolin, glass, dextran sulfate). The activation of C1 by FXIIa activates the classical pathway of the complement system. FXIIa activates prekallikrein and FXI, the latter forming the link to the intrinsic coagulation system. Kallikrein activates the fibrinolytic system by direct conversion of plasminogen to plasmin or by activation of u-PA. HK is a nonenzymatic cofactor for PK and FXI activation. Cleavage of HK by kallikrein liberates the potent vasoactive nonapeptide bradykinin. F, coagulation factor; "a", activated coagulation factor; C, complement; PK, prekallikrein; KK, kallikrein; Plg, plasminogen; Pl, plasmin.

HK is a nonenzymatic cofactor that augments reciprocal activation of FXII and prekallikrein as well as the rate of FXI activation by FXIIa (Griffin and Cochrane, 1976; Meier et al., 1977). It is the pivotal protein for contact protein assembly on endothelium (Schmaier, 1997).

The zymogen prekallikrein becomes activated to kallikrein when it binds to HK on endothelial cells. The resulting proteinase kallikrein is the major activator of FXII (Van den Graaf et al., 1982). Kallikrein activates the fibrinolytic system either by activation of single-chain urokinase or of plasminogen (Colman, 1969; Motta et al., 1998). Kallikrein has been shown to prime neutrophils for superoxide production (Zimmerli et al., 1989). Finally, kallikrein cleaves HK at two sides to liberate the potent vasoactive nonapeptide bradykinin (Kaplan and Silverberg, 1987; Schmaier, 1997). Bradykinin is known to stimulate endothelial cell prostacyclin synthesis, leading to inhibition of platelet function (Hong, 1980; Crutchley et al., 1983), to increase superoxide formation (Holland et al., 1990), to release tissue-type plasminogen activator (t-PA) (Smith et al., 1985), as well as to induce the formation of nitric oxide (Palmer et al., 1987). HK and bradykinin are selective inhibitors of -thrombin-induced platelet activation (Meloni and Schmaier, 1991; Hasan et al., 1996).

In an intact vessel the sum of bradykinin activities is to keep blood flowing and vessels patent; in the absence of endothelium, bradykinin stimulates repair of vessels, which leads to smooth muscle proliferation and intimal hypertrophy (Schmaier, 1997). Bradykinin is thought to play a major role in the symptomatology of acute attacks in patients with HAE (Fields et al., 1983; Al-Abdullah and Greally, 1985; Kaplan et al., 1989; Shoemaker et al., 1994; Cicardi et al., 1998; Nussberger et al., 1998). Activation of the kallikrein-kinin system during acute attacks of HAE indeed has been demonstrated by a significant increase in the cleavage of HK (Fields et al., 1983).

Degradation of bradykinin into inactive fragments depends on the activity of the angiotensin-converting enzyme, which in turn has been documented to be severely decreased in patients with septic shock and septic adult respiratory distress syndrome (Fourrier et al., 1985).

C1-Inh is a major inactivator of FXIIa and FXIIa and besides 2-macroglobulin a major inactivator of kallikrein (Nilsson, 1983a; Pixley et al., 1985; Kaplan and Silverberg, 1987; Schapira et al., 1988). Incubation of radiolabeled FXIIa with various plasma proteinases in purified systems and in human plasma demonstrated FXIIa-C1-Inh complex the predominant complex to be formed confirming the major inhibitory role of C1-Inh for FXIIa (Pixley et al., 1985). The majority of the additional inhibitory capacity toward FXIIa is provided by 2-macroglobulin (Davis, 1998).

C1-Inh and 2-macroglobulin are the predominant inhibitors of kallikrein (Schapira et al., 1988). In acute attacks of HAE, despite activation of the contact system, the C1-Inh-kallikrein complexes did not increase, leading to the suggestion that 2-macroglobulin may compensate, to a certain extent, for the lack of inhibitory capacity of C1-Inh toward kallikrein (Waage Nielsen et al., 1996).

FXI is the coagulation protein that links contact activation to intrinsic blood coagulation (see Fig. 4). FXI circulates in plasma complexed to HK (Thompson et al., 1977). It is activated by FXIIa into the active form FXIa (Bouma and Griffin, 1977; Kurachi and Davie, 1977). Because patients deficient in FXII, prekallikrein, or HK do not suffer from a bleeding tendency, contrary to the variable bleeding disorder of patients with FXI deficiency, an alternative pathway for the activation of FXI has been assumed (Gailani and Broze, 1991; Naito and Fujikawa, 1991). Recently it has been demonstrated that thrombin and FXa can activate FXI in the presence of dextran sulfate or activated platelets (Gailani and Broze, 1991; Naito and Fujikawa, 1991; Roberts et al., 1998).

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Coagulation system. A current concept of blood coagulation depicts the tissue factor-FVIIa complex to be the initiator of primary thrombin generation. This initial amount of thrombin is sufficient for the activation of platelets FV, FVIII, and FXI, but it is insufficient for the formation of a stable fibrin clot. Only the subsequent burst of thrombin generation by activated FXI and FIX leads to the appearance of a stable fibrin clot. F, coagulation factor; "a", activated coagulation factor; TF, tissue factor; Fbg, fibrinogen; Fbm, fibrinmonomer; Fbcx, cross-linked fibrin; Ca, calcium.

A current concept of blood coagulation depicts the tissue factor-FVIIa complex as the initiator of all subsequent reactions: The cell-based tissue factor-FVII complex activates both, FIX and FX, the latter interacting subsequently with FV. The FXa-FVa complex then converts prothrombin to thrombin. The amount of thrombin generated in this way is sufficient to activate platelets and to activate FV, FVIII and FXI. However, it is insufficient for the formation of a stable fibrin clot (Roberts et al., 1998). Only the subsequent burst of thrombin generation by activated FXI, FIX, and FVIII leads to quantitative thrombin generation and to the formation of a stable fibrin clot (Roberts et al., 1998) (Fig. 4). These data suggest that activation of FXI by thrombin is an alternative pathway for the activation of blood coagulation and that the role of FXI is to enhance and maintain thrombin generation.

Experiments in plasma milieu demonstrated C1-Inh to be the main inhibitor of FXIa, contributing to 47% of FXIa inactivation in EDTA plasma, followed by 2-antiplasmin (24.5%), 1-antitrypsin (23.5%), and antithrombin (5%) (Wuillemin et al., 1995b). This had been confirmed in chimpanzees given a bolus infusion of FXIa, where C1-Inh was shown to be the main inactivator of FXIa, followed by 2-antiplasmin, 1-antitrypsin, and antithrombin (Wuillemin et al., 1996c).

More evidence that C1-Inh plays a role in the regulation of blood coagulation raised from studies on patients with hereditary or acquired angioedema, where it has been shown that levels of coagulation FVIIa, prothrombin fragments F1 + 2, and thrombin-antithrombin complex significantly increased during acute attacks, although remaining normal in remission (Waage Nielsen et al., 1995, 1996).

The physiologic activators of the fibrinolytic system are t-PA and urokinase-like plasminogen activator (u-PA). Both serine proteinases are quite capable of converting plasminogen to plasmin (Vassalli et al., 1991). Cleavage of single-chain t-PA to the two-chain form increases its binding affinity for fibrin (Husain et al., 1989). Single-chain u-PA can be rapidly converted to two-chain u-PA by plasma kallikrein and plasmin. Two-chain u-PA has greater enzymatic activity but is less fibrin-specific (Zamarron et al., 1984). More recent studies suggested that the activation of single-chain u-PA by kallikrein can best occur on a platelet or endothelial surface (Gurewich et al., 1993; Loza et al., 1994). Kallikrein-activated single-chain u-PA is considered the major physiological activator of plasminogen (Hauert and Bachmann, 1985; Motta et al., 1998).

Kallikrein, FXIIa, and FXIa have the ability to cleave plasminogen directly, albeit much less efficiently than t-PA or u-PA (Schmaier, 1997). However, activation of plasminogen by FXIIa is considerably potentiated in the presence of cofactors (Schousboe, 1997) with the result that at plasma concentrations FXIIa is as potent as u-PA in clearing plasminogen. In agreement herewith, Levi et al. (1991) demonstrated that the plasminogen-converting activity induced by DDAVP is 50% dependent on t-PA, 25% on u-PA and 25% on FXIIa.

In patients with HAE, antigenic levels of t-PA and u-PA as well as t-PA-C1-Inh complexes remained normal during both remission and acute attacks. However, during acute attacks plasmin-antiplasmin complexes were significantly increased, supporting an enhanced activation of the fibrinolytic system via direct activation of plasminogen by kallikrein and FXIIa in a C1-Inh deficiency state (Waage Nielsen et al., 1996).

The main inhibitor of single-chain and two-chain t-PA and two-chain u-PA is plasminogen activator inhibitor type-1 (PAI-1) (Kruithof et al., 1986). PAI-2, originating from human placenta and macrophages and being distinct from PAI-1, inhibits two-chain t-PA and two-chain u-PA (Lecander and Astedt, 1986). The principal physiologic inhibitor of plasmin is 2-antiplasmin, which is present in plasma and in platelets. C1-Inh as well as 2-macroglobulin, antithrombin, and 1-antitrypsin only play a limited physiologic role as plasmin inhibitors (Levi et al., 1993).

C1-Inh contributes only slowly and to a minor extent to the inactivation of t-PA when t-PA levels are normal (Huisman et al., 1995). However, when t-PA circulates at high concentrations, e.g., during thrombolytic therapy or escapes rapid liver clearance, e.g., in case of venous occlusion, an increase of circulating t-PA-C1-Inh complexes has been demonstrated (Huisman et al., 1995).

C1-Inh has been shown to inhibit the activation of CD4- and CD8-positive T-lymphocytes by specific cleavage of the major histocompatibility complex class 1 molecules, whereas no effect on B-lymphocytes has been demonstrated (Eriksson and Sjögren, 1995). In a model of allogen- or mitogen-activated murine or human lymphocyte cultures the addition of C1-Inh was followed by the down-regulation of the activity and proliferation of cytotoxic T-lymphocytes (Nissen et al., 1998). The addition of C1-Inh altered the production of cytokines by T-lymphocytes, increasing the production of IFN-, IL-10, and IL-12 (Nissen et al., 1998). Because, inversely, IFN- is capable of inducing the production of C1-Inh, we concluded that C1-Inh and IFN- up-regulate each other during the maturation of the immune response, indicating a regulatory function of C1-Inh on T-cell-mediated immune functions.

	IV. Potentiation of C1-Inh Activity by Glycosaminoglycans

Top Previous Next References

The activity of a group of serpins, such as antithrombin, heparin cofactor II, and PAI-1 is potentiated by glycosaminoglycans (GAG) (Potempa et al., 1994). Early reports suggested the inhibitory activity of C1-Inh toward C1s also to be enhanced by heparin (Rent et al., 1976; Sim et al., 1980; Caughman et al., 1982; Nilsson and Wiman, 1983b; Lennick et al., 1986), whereas GAG had no effect on the interaction of C1-Inh with FXIIa (Pixley et al., 1987). Heparin and other GAGs have multiple inhibitory effects on the complement system, such as that on the binding of C1q to an activator and that on the formation of C3-convertases of the classical or the alternative pathway.

We, therefore, studied the influence of various physiological (heparin, heparan sulfate, dermatan sulfate, chondroitin sulfate) and nonphysiological (dextran sulfate) GAGs on the kinetics of the interaction of C1-Inh with its target proteinases C1s, FXIa, FXIIa, FXIIa, and kallikrein. First, we showed that the inactivation of C1s by C1-Inh is increased 6- to 130-fold in the presence of GAG, with dextran sulfate being the most effective GAG to enhance inactivation of C1s by C1-Inh (Wuillemin et al., 1997). Moreover, GAGs reduced the deposition of C3 and C4 on immobilized aggregated human IgG and also reduced the fluid phase formation of C4b/c and C3b/c in recalcified plasma upon incubation with aggregated IgG (Wuillemin et al., 1997). In similar experiments we demonstrated that in the presence of GAG the rate constant of the inactivation of FXIa by C1-Inh increased up to 117-fold compared with the rate of inactivation in the absence of any GAGs (Wuillemin et al., 1996b). Recently, low-molecular weight heparins, such as dalteparin, enoxaparin, and nadroparin and low-molecular weight dextran sulfate were found to increase the inactivation of FXIa by C1-Inh up to 39-fold. Moreover, in the presence of low-molecular weight heparin or low-molecular weight dextran sulfate, FXIa was inactivated in human plasma to more than 90% by C1-Inh (Mauron et al., 1998). In contrast, we found no significant influence of the tested GAGs on the inhibition of kallikrein by C1-Inh and about a 2-fold protection of FXIIa and FXIIa from inhibition by C1-Inh in the presence of dextran sulfate (Wuillemin et al., 1996b).

These findings definitely demonstrate that C1-Inh belongs to the group of GAG-sensitive serpins. Moreover, the results suggest the possibility of developing a C1-Inh preparation with strongly enhanced inhibitory activity. Interestingly, the influence of GAGs on C1-Inh function is not only a quantitative one but also modulates the inhibitory spectrum of C1-Inh; physiological GAGs selectively enhance the inactivation of the complement system (C1s) and the intrinsic coagulation (FXIa) but do not affect the activity of FXIIa and kallikrein, suggesting that GAGs may modulate the biological effects of contact activation by inhibiting intrinsic coagulation without affecting the fibrinolytic potential of FXIIa and kallikrein (Wuillemin et al., 1996b).

	V. C1-Inh Therapy in Animal Models and Clinical Disease

Top Previous Next References

The treatment of acute attacks in patients suffering from hereditary angioedema with C1-Inh purified from pooled human plasma is well established (Agostoni et al., 1980; Gadek et al., 1980; Bork and Witzke, 1989; Waytes et al., 1996) and will not be discussed further. Here we will elaborate on the therapeutic application of C1-Inh in other diseases (Tables 2 and 3).

Sepsis is often induced by bacterial infections and is a leading cause of mortality in noncardiologic intensive care units. Sepsis results from the excessive release and activation of endogenous inflammatory mediators, which include the complement and contact systems. In a study of experimental endotoxemia in healthy volunteers, activation of the contact system was suggested to occur due to endothelial injury during the septic process (De La Cadena et al., 1993). Another mechanism leading to activation of the contact factors was proposed to be a direct activation of the contact factors such as FXII and prekallikrein by bacterial lipopolysaccharides (Morrison and Cochrane, 1974; Kalter et al., 1983; ten Cate et al., 1993). Recently, as a clue to serious complications in infectious disease, the assembly and activation of the contact phase system on bacterial surfaces such as Escherichia coli and Salmonella typhimurium has been reported (Herwald et al., 1998). Although it has long been considered that the contact system is activated during sepsis, not until recently was this concept proven to be correct. Studies in primates showed decreasing levels of various contact system proteins during sepsis, accompanied by increasing levels of activation products of the contact system (Pixley et al., 1992). In patients with septic shock, significantly decreased activities of FXII have been demonstrated (Kalter et al., 1985; Nuijens et al., 1988) and in children with meningococcal septic shock the plasma levels of FXII, FXI, and prekallikrein were found to be reduced to ~50% of normal (Wuillemin et al., 1995a). Definite evidence for the activation of the contact system during the septic process was provided by a study in septic baboons treated (before the bacterial challenge) with a monoclonal antibody that blocks the activation of FXII. This treatment had no effect on clotting activation but largely prevented the irreversible hypotension and slightly improved the survival of baboons challenged intravenously with a lethal dose of E. coli (Pixley et al., 1993). Thus, FXII activation during sepsis does not primarily contribute to clotting derangements, but probably, via the generation of kallikrein and subsequently bradykinin, contributes to the formation of nitric oxide and to vasodilation (Vane et al., 1990).

The complement system can be activated by bacteria and their products such as endotoxin even in the absence of antibodies (Morrison and Cochrane, 1974; Morrison and Kline, 1977; Kalter et al., 1983). The role of complement activation during sepsis seems to be dual. Some activation is necessary for an efficient clearance of bacteria or their products as demonstrated in animal studies. Dogs and mice with a genetic C3 deficiency are more susceptible to endotoxin than healthy littermates due to an impaired clearance of endotoxin (Quezado et al., 1994; Fischer et al., 1997). On the other hand, inhibition of the biological effects of C5a in baboons suffering from sepsis attenuated lethal complications (Stevens et al., 1986), illustrating that the proinflammatory effects of complement activation, in particular those of C5a, may contribute to the complications of sepsis. Complement component C5a and the MAC have proinflammatory effects such as accumulation and stimulation of neutrophils and may increase the permeability of endothelial cells, mediated in part by histamine, and promote coagulation by inducing expression of tissue factor (Björk et al., 1985; Hack et al., 1993). In animal models, intravenous C5a can induce a fall in mean arterial blood pressure and leukopenia, the latter probably due to aggregation and subsequent sequestration of leukocytes (Lundberg et al., 1987). Finally, C5a is able to induce or enhance production of cytokines as IL-1, TNF, and IL-6 by monocytes (Cavaillon et al., 1990; Scholz et al., 1990). The proinflammatory effects of complement components during sepsis were supported by observations that C5-deficient mice tolerate endotoxin better than their C5-sufficient littermates (Olson et al., 1985). C5-deficient mice exhibit a 2-fold lower TNF response and a slower increase of pulmonary vasopermeability than C5-sufficient animals (Barton and Warren, 1993). Notably, these effects of complement on the release of cytokines have not been observed consistently. In an endotoxin model in rats inhibition of complement activation by administration of human soluble complement receptor-1 did not affect circulating TNF levels, although this treatment improved pulmonary responses during endotoxemia (Rabinovici et al., 1992). Thus, these studies suggest that during sepsis complement on the one hand is required for a rapid clearance of bacteria or their products, but on the other hand via the release of C5a and possibly other phlogistic fragments may enhance inflammatory reactions. Baboons challenged with lethal and sublethal doses of E. coli showed a biphasic pattern of complement activation consisting of a rapid initial activation and followed by a second more pronounced activation from about 6 h up to over 24 h (De Boer et al., 1993). Although the initial activation was probably due to a direct stimulation of the complement system, e.g., via IgG or IgM antibodies, the second phase of activation coincided with increasing levels of C-reactive protein (CRP), IL-2, and IL-6, suggesting a further complement activation via cytokines (De Boer et al., 1993).

Human studies revealed that plasma levels of native complement proteins are decreased in septic patients being the lowest in patients with fatal outcome (McCabe, 1973; Kalter et al., 1985). On the other hand, elevated plasma levels of C3a in patients with sepsis and septic shock were significantly correlated with mortality and patients with septic shock had significantly higher C3a levels than normotensive patients (Hack et al., 1989). The levels of C4a and C1-Inh complexes correlated with C3a levels and with the clinical outcome (Hack et al., 1989). It is likely that activated C5a plays the predominant role in the pathophysiology of the septic processes, because it greatly exceeds C3a in biologic activity. However, measurement of C5a is difficult due to its rapid binding to cellular receptors (Hack et al., 1989).

The role of C1-Inh in sepsis was investigated in several clinical studies. In 48 patients with sepsis, compared to healthy volunteers, a discrepancy was demonstrated in plasma levels of functional and antigenic C1-Inh that was mainly due to an increase of inactive cleaved C1-Inh (iC1-Inh) with molecular masses of 98, 96, and 86 kDa as assessed by SDS-polyacrylamide gel electrophoresis (PAGE), respectively (Nuijens et al., 1989). The extent of plasma C1-Inh proteolysis and the level of cleaved iC1-Inh appeared to be positively correlated with the mortality of the sepsis-patients. Functional C1-Inh was significantly reduced only in patients with septic shock (Nuijens et al., 1989). The similarity of the cleavage pattern of C1-Inh in vivo with the pattern of cleavage by plasmin led to the assumption that local degradation by plasmin may play a certain role in the loss of C1-Inh activity during inflammation (Wallace et al., 1997). However, in baboons challenged with lethal and sublethal doses of E. coli the peak values of iC1-Inh coincided with the peak values of elastase-antitrypsin complex levels and not with those of plasmin-antiplasmin complexes, suggesting that elastase may be predominantly responsible for the generation of iC1-Inh (De Boer et al., 1993).

Therapeutic C1-Inh administration incompletely blocked the activation of the classical pathway and did not interfere with the clearance of bacteria in primates suffering from lethal septic shock (Jansen et al., 1998). Hence, during C1-Inh administration some opsonization of the infecting micro-organisms or their products by the complement system will be preserved.

We have evaluated the effects of therapeutic administration of C1-Inh in a baboon model for lethal E. coli septic shock. Administration of C1-Inh at a dose that increased plasma levels 5- to 10-fold, reduced activation of C4 (and to a lesser extent C3) and improved mortality. Three of seven animals challenged with a lethal dose of E. coli survived 64 h, one of them was a permanent survivor (Jansen et al., 1998). Similarly, a favorable, although mild, effect of C1-Inh has been found in several endotoxin models in rats, dogs, rabbits, and in mice deficient for C4 and C3 (Guerrero et al., 1993; Scherer et al., 1996; Fischer et al., 1997). Notably, it has to be established whether the beneficial effects of C1-Inh in sepsis are due to its effect on the complement system, on the contact system, or on both.

Preliminary evaluation of C1-Inh therapy in patients with septic shock has been performed (Hack et al., 1992, 1993). Initially, five patients treated with mechanical ventilatory support, volume substitution, vasopressor, and positive inotropic drugs, received C1-Inh for 5 days, starting with a dose of 2000 U, subsequently followed by 1000 U every 12 h. No patient died during the study period of 5 days. Four of the patients needed less, and one patient needed more vasopressor therapy during this period. No side effects of C1-Inh treatment were observed. Both complement and contact system parameters were measured in the five patients who received C1-Inh. C3a levels tended to decrease in these patients (Hack et al., 1993), whereas FXII levels increased (Hack et al., 1992). Thus, high doses of C1-Inh are well tolerated by patients with sepsis and may attenuate ongoing complement and contact activation. We then administered C1-Inh to seven additional patients with septic shock, one of whom (with bacterial endocarditis) did not complete the study because of transfer to another hospital (for open heart surgery). Three patients received C1-Inh for 3 days, a starting dose of 4000 U followed by 2 doses of 2000 U and 4 doses of 1000 U each 12 h, the other three patients received 6000 U of C1-Inh followed by 3000, 2000 and 1000 U (all doses given at 12 h intervals). Comparable effects as with the other dose regimen were seen, i.e., no toxic side effects and a slight reduction of complement and contact activation. The overall results (no toxic side effects, no sepsis-related mortality during the study period, a possible attenuation of complement and contact activation, a possible beneficial effect on hypotension as reflected by a decreased need for vasopressor medication) were confirmed in several open uncontrolled studies in a limited number of septic shock patients, who all received C1-Inh according to the schemes outlined above (2nd workshop on C1-esterase inhibitor, Duesseldorf, Germany, 24-26 April 1997). Double-blind controlled studies in a larger number of patients are warranted to confirm these promising effects. Therefore, we initiated a prospective, randomized, double blind and placebo-controlled study investigating the clinical outcome and laboratory parameters of C1-Inh administration in patients with severe sepsis or septic shock. In addition to standard routine treatment on the intensive care unit, the patients were randomized to either high dose C1-Inh (starting dose of 6000 U, followed by 3000, 2000, and 1000 U at 12 h intervals) or placebo (same amount of a solution of albumin 5%). We expect the first results of this study at the end of 1999.

A vascular or capillary leak syndrome (VLS) may complicate sepsis (Nürnberger et al., 1992) but may also occur independently of this disease. For example, VLS is induced by therapy with cytokines, such as IL-2, or following a bone marrow transplant or open heart surgery (Gaynor et al., 1988; Ognibene et al., 1988; Nürnberger et al., 1993; Stieh et al., 1996), or develops in the absence of any known precipitating event. The pathogenetic mechanisms underlying VLS are increased vasopermeability and vasodilation. Hence, hypotension may complicate VLS. The molecular mechanisms causing these phenomena are poorly understood, although endothelial damage resulting from interactions with activated neutrophils and/or natural killer cells are likely to be at the basis (Damle et al., 1987; Damle and Doyle, 1989; Baars et al., 1992b). Studies of the VLS induced by IL-2 have suggested that this syndrome results from the release and activation of inflammatory mediators such as cytokines (Gemlo et al., 1988; Boccoli et al., 1990), activation of neutrophils (Baars et al., 1992b), complement (Thijs et al., 1989, 1990; Vachino et al., 1991; Baars et al., 1992b) and coagulation and fibrinolysis (Baars et al., 1992a). In addition, changes in the contact system proteins resembling those seen in sepsis, occur during IL-2 therapy (Hack et al., 1991). In IL-2-induced VLS, we have shown that the activation of the classical pathway of complement correlates with the development of side effects (Thijs et al., 1989; 1990; Baars et al., 1992b). Also in other forms of VLS, such as that following bone marrow transplantation, complement is activated via the classical pathway (Nürnberger et al., 1993). The mechanism of this activation is not known but may involve binding of CRP to IL-2 activated lymphocytes, which subsequently activate and fix complement (Vachino et al., 1991). The involvement of CRP in IL-2-induced activation of complement in vivo is discussed below. Regardless of the cause for classical pathway activation during IL-2 therapy, we decided to evaluate the effects of C1-Inh administration in patients receiving high doses of IL-2 (Ogilvie et al., 1994). Six patients with either metastatic melanoma or renal cell carcinoma received 72 × 10⁶ U of recombinant IL-2 (from the former Eurocetus, Amsterdam, The Netherlands) daily with exogenous C1-Inh given at a dose of 2000 U initially, subsequently followed by 1000 U every 12 h, for 4 days (treatment cycle). As controls, the same patients receiving a second cycle of IL-2 given at 4-fold lower doses 4 weeks after the first cycle (control cycle) were studied, as were 4 other patients who received escalating doses of IL-2, starting with 18 × 10⁶ U and increasing by 18 × 10⁶ U every 2 to 3 days (Thijs et al., 1990). Four of the six patients needed vasopressor medication during the treatment cycle (one patient for reasons not related to IL-2). The degree of hypotension was comparable to that observed during the control cycle and in the patients who received an escalating dose of IL-2. Thus, the clinical toxicity of IL-2 was comparable in all patients, despite the fact that the C1-Inh treatment group had received considerably more IL-2 (Ogilvie et al., 1994). Thus the results suggested that C1-Inh therapy is also able to reduce IL-2 toxicity, probably via inhibition of IL-2-induced complement activation. Moreover, various effects of C1-Inh substitution observed in the IL-2 treated patients (attenuated complement activation, less hypotension) were comparable to the effects seen in the patients with septic shock.

The effect of C1-Inh administration in other forms of VLS has also been evaluated in preliminary studies. A newborn baby with sepsis-associated VLS was treated with C1-Inh for 3 days (300, 100, and 50 U/kg on days 1, 2, and 3, respectively) (Nürnberger et al., 1992). Although the patient died 15 days later because of liver failure, the effect of C1-Inh administration was judged to be beneficial, because the patient no longer needed vasopressor medication, and his body weight normalized. The effect of C1-Inh has also been evaluated in VLS following bone marrow transplantation. In an initial study, two patients received 60 U/kg as a loading dose, then 2 doses of 30 U/kg given at 12-h intervals, and finally, 4 doses of C1-Inh of 15 U/kg (Nürnberger and Gobel, 1996). Body weight normalized in each patient, as did the increased levels of C4 days, which parameter reflects the extent of classical pathway activation. In a later report, the same authors describe 15 patients treated with C1-Inh because of VLS induced by bone marrow transplantation (Nürnberger et al., 1997). The one year-survival rate was 57% in the treated patients versus 14% in a control group consisting of seven patients (the study was not randomized and not placebo-controlled). Treatment was accompanied by normalization of circulating C4 days and C5a levels. These effects suggest a beneficial effect of C1-Inh therapy in this severe complication of bone marrow transplantation, but needs to be confirmed by a double-blinded placebo-controlled study.

Twenty-nine children with mild to severe VLS induced by open heart surgery were also treated with C1-Inh (starting dose of 300 U/kg, followed by 2 doses of 150 U/kg, 3 doses of 100 U/kg and finally 3 doses of 50 U/kg, each dose given at 8 h-interval). In most children the effect of C1-Inh therapy was judged to be favorable because hemodynamic, respiratory, and laboratory parameters improved. However, in 11 patients arterial blood pressure did not respond, in 3 leakage continued, and in 6 children diuresis was not ameliorated (Stieh et al., 1996). Notably, in two patients, possible adverse side effects were registered, superior vena cava thrombosis in one patient with a transposition of the great vessels, and extended renal vein thrombosis in a neonate.

Together these studies indicate that C1-Inh therapy is a promising approach for the management of patients with VLS; however, double-blinded placebo-controlled studies are needed to confirm this.

Acute myocardial infarction (AMI) is one of the major causes of mortality and morbidity in the western world. Mortality is due to arrhythmia, cardiac rupture, and acute heart failure, whereas morbidity often results from chronic heart failure. An important determinant in the development of heart failure is the amount of necrotic tissue in the jeopardized myocardium. In patients with unstable angina pectoris, evidence for the activation of the contact system of coagulation was provided by the stimulation of the kallikrein system and the generation of bradykinin in the acute phase (Hoffmeister et al., 1995). Studies in animals have shown that irreversible myocardial cell injury starts about 30 min after occlusion of coronary vessels and proceeds for hours. The later phase of myocardial cell injury likely results from an acute inflammatory reaction ensuing in the ischemic myocardium as it can be effectively reduced by anti-inflammatory agents. For example, corticosteroids given as late as 6 h after coronary occlusion reduce infarction size by about 35% compared with untreated control animals (Libby et al., 1973). The local inflammatory response ensuing in the infarcted myocardium is characterized by the local production of chemotactic factors, the infiltration and activation of neutrophils, the local production of cytokines (such as TNF- and IL-6), the expression of adhesion molecules, which enhance adherence of neutrophils to cardiac myocytes, and local activation of the complement system (Entman et al., 1991).

Complement activation by ischemic myocardium was first demonstrated by Hill and Ward who showed that complement activation products generated in the infarcted myocardium were responsible for the infiltration of neutrophils (Hill and Ward, 1971). Later studies in animals, as well as in patients, showed that several complement components become localized in the infarcted myocardium, independent of reperfusion, whereas membrane-bound complement inhibitors decrease (Pinckard et al., 1980; McManus et al., 1983; Rossen et al., 1985; Schafer et al., 1986; Crawford et al., 1988; Hugo et al., 1990; Vakeva et al., 1992, 1993, 1994; Lagrand et al., 1997). Furthermore, plasma levels of activated complement components are increased in patients with AMI and correlate with myocardial damage (Langlois and Gawryl, 1988; Yasuda et al., 1990). Although some studies claim that the activation of complement in ischemic myocardium occurs via the alternative pathway (Amsterdam et al., 1995