Alterations of coagulation and fibrinolysis in patients with angioedema due to C1-inhibitor deficiency

Authors

  • M. van Geffen,

    1. Laboratory of Hematology, Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands
    Search for more papers by this author
  • M. Cugno,

    1. Department of Internal Medicine, University of Milan, IRCCS, Fondazione Ospedale Maggiore Policlinico, Mangiagalli e Regina Elena, Milan
    Search for more papers by this author
  • P. Lap,

    1. Laboratory of Hematology, Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands
    Search for more papers by this author
  • A. Loof,

    1. Laboratory of Hematology, Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands
    Search for more papers by this author
  • M. Cicardi,

    1. Dipartimento di Scienze Cliniche ‘Luigi Sacco’ Università di Milano, Ospedale L. Sacco, Milano, Italy
    Search for more papers by this author
  • W. van Heerde

    Corresponding author
    1. Laboratory of Hematology, Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, the Netherlands
      W. L. van Heerde, Laboratory of Hematology, Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, PO Box 9101, 6500 HB Nijmegen, the Netherlands. E-mail: w.vanheerde@labgk.umcn.nl
    Search for more papers by this author

W. L. van Heerde, Laboratory of Hematology, Department of Laboratory Medicine, Radboud University Nijmegen Medical Centre, PO Box 9101, 6500 HB Nijmegen, the Netherlands. E-mail: w.vanheerde@labgk.umcn.nl

Summary

Patients with functional deficiency of C1-inhibitor (C1-INH) suffer from recurrent acute attacks (AA) of localized oedema associated with activation of the contact system, complement and fibrinolysis. To unravel further the role of coagulation and fibrinolysis in the pathophysiology of C1-INH deficiency, we performed simultaneous thrombin and plasmin generation measurements in plasma from patients with hereditary angioedema (HAE) due to C1-INH deficiency during AA (n = 23), in remission (R) (n = 20) and in controls (n = 20). During AA thrombin generation after in-vitro activation of plasma was higher than in controls, as demonstrated by shorter thrombin peak-time (P < 0·05), higher thrombin peak-height (P < 0·001) and increased area under the curve (AUC) (P < 0·05). Additionally, elevated levels of prothrombin fragment 1+2 (P < 0·0001) were observed in non-activated plasma from the same patients. In contrast, in activated plasma from patients during AA plasmin generation estimated as plasmin peak-height (P < 0·05) and plasmin potential (P < 0·05) was reduced, but non-activated plasma of the same patients showed elevated plasmin–anti-plasmin (PAP) complexes (P < 0·001). This apparent discrepancy can be reconciled by elevated soluble thrombomodulin (sTM) (P < 0·01) and thrombin activatable fibrinolysis inhibitor (TAFI) in patients during AA providing possible evidence for a regulatory effect on fibrinolysis. Plasminogen activator inhibitor-1 (PAI-1) was reduced in patients during AA indicating, together with the observed reduction of plasmin generation, the consumption of fibrinolytic factors. In conclusion, our results support the involvement of coagulation and fibrinolysis in the pathophysiology of HAE and show the possible application of simultaneous measurement of thrombin and plasmin generation to evaluate different clinical conditions in HAE patients.

Introduction

Hereditary angioedema (HAE) is caused by a functional deficiency of C1-inhibitor (C1-INH), which is inherited as an autosomal dominant trait and characterized by episodic swellings of subcutaneous tissues, bowel walls and upper airways [1]. Normal C1-INH plasma concentration is ∼2·5 µM [2]. Despite the presence of a normal allele, HAE patients have C1-INH levels markedly lower than 50% of normal because of a faster in-vivo consumption of C1-INH [3], mainly as a consequence of the formation of complexes with target proteases [4]. C1-INH has been shown to inhibit not only the first component of human complement, but also activated FXII [5,6], thrombin [7,8] and plasmin [8,9], thus it is involved in the control of several proteolytic systems which can be activated in conditions known to be potential triggers of angioedema attacks. In fact, physical trauma and surgery frequently induce attacks, perhaps because of their ability to activate the contact system, resulting in the release of bradykinin.

Two independent studies demonstrated the involvement of coagulation during HAE attacks by measuring increased plasma levels of activated FVII [10–12], and additionally one of these groups demonstrated downstream activation of coagulation factors by the measurement of prothrombin fragment 1+2. The generation of thrombin is particularly interesting, because experimental data indicate that thrombin increases vascular permeability [13,14], which occurs by cleavage and activation of a thrombin-susceptible receptor at the endothelial cell surface, thereby inducing intercellular gap formation [15]. Moreover, fibrinolysis also is involved in angioedema patients during attacks, as supported by the fact that anti-fibrinolytic agents such as ε-aminocaproic acid are suggested to be effective in the treatment of angioedema [16] and that contact activation may serve as an activator of fibrinolysis [17–19]. The fibrinolytic enzyme plasmin is released during episodes of swelling in patients with HAE [20] and, more importantly, plasmin is able to activate factor XII to factor XIIa and factor XIIf as an alternative to the kallikrein feedback, thereby elevating final bradykinin levels.

The involvement of plasmin in the pathogenesis of acute attacks is also supported by experimental in vitro data demonstrating that plasmin potentiates the generation of bradykinin from high molecular weight kininogen by kallikrein [21] and alternatively by Factor XII activation [22]. The disadvantage of measuring single components of coagulation and fibrinolysis is that the outcome of the assays does not reflect the real balance between activators and inhibitors that exist in the original blood sample. In the current study, we tried to overcome this problem by using the novel haemostasis assay (NHA) to evaluate haemostasis by simultaneous measurement of thrombin and plasmin generation in different clinical conditions in patients with C1-INH deficiency.

Patients and methods

Patients

We studied 43 patients with HAE; 20 patients {10 men and 10 women aged 51·4 ± 15·3 years [mean ± standard deviation (SD), range 24–75]} were during remission (R) (free of attacks for at least 1 month) and 23 patients [12 men and 11 women aged 58·0 ± 14·4 years (mean ± SD), range 24–84] had acute attacks (AA) involving oral submucosal tissues in six patients, subcutaneous tissues in 10 patients and the abdomen in seven. These two groups included five patients that were present in both groups. The diagnosis of abdominal acute attacks in the patients with known C1-INH deficiency was based on the presence of acute abdominal pain with or without vomiting and diarrhoea, the absence of any other cause of abdominal pain and prompt reversal upon the infusion of C1-INH or tranexamic acid. The control group consisted of 20 healthy subjects (10 men and 10 women) with a mean age of 41·7 ± 10·1 years (range 24–79).

Blood samples were collected in sodium citrate as anti-coagulant. The samples were centrifuged at 2000 g for 20 min at room temperature. The plasma obtained was quick-frozen immediately in liquid nitrogen and stored at –80°C until the time of study. All subjects were included after informed consent in accordance with the Declaration of Helsinki. The protocol was approved by the Institutional Review Board of Milan.

NHA

The NHA was performed according to the method of van Geffen et al. [23]. Patient plasma, 80 µl, was activated with 40 Tris-buffered saline (TBS) buffer containing crude cephalin (Roche, Basel, Switzerland), tissue factor (Innovin®) from Siemens Healthcare Diagnostics, Marburg, Germany), fluorescent thrombin substrate Bz-β-Ala-Gly-Arg-7-amino-4-methylcoumarin and fluorescent plasmin bis-(CBZ-l-phenylalanyl-l-arginine amide)-rhodamine substrate (both from Chiralix, Nijmegen, the Netherlands), tissue-type plasminogen activator (tPA) (Actilyse®; Boehringer Ingelheim, Ingelheim am Rhein, Germany) and calcium. Fluorescence was measured alternately every 30 s for 70 min in a 37°C thermostated fluorometer (Fluostar Optima Fluorometer; BMG Labtechnologies, Offenburg, Germany) using black polystyrene Fluotrac microtitre plates (Greiner Bio-One, Monroe, NC, USA) The first derivatives are calculated from the fluorescent, cumulative signals of both substrates and characterized by seven parameters.

The thrombin generation curve is represented by the following parameters: (i) lag-time, the time to the start of thrombin formation; (ii) thrombin peak time, i.e. the time when thrombin production reaches maximal velocity; (iii) thrombin peak height, the maximal velocity of thrombin generation; (iv) area under the curve (AUC); and total amount of thrombin formed. Plasmin generation is described by: (v) fibrin lysis time (FLT), the time between the initiation of thrombin generation and the time plasmin generation reaches maximal velocity; (vi) plasmin peak-height, the maximal velocity of plasmin production; and (vii) plasmin potential, area under the curve that represents the total amount of plasmin generated.

Laboratory measurements

The following measurements were performed in non-activated plasma: C1-INH activity was measured using a chromogenic assay (Technochrom C1-inhibitor; Technoclone GmbH, Vienna, Austria), and C1-INH antigen by means of radial immunodiffusion (RID) (NOR-Partigen, Behring, Marburg, Germany). Plasma prothrombin fragment 1+2 was measured using an enzyme-linked immunosorbent assay (ELISA; Enzygnost F1+2; Behring Diagnostics GmbH, Frankfurt, Germany; normal range 400–1100 pmol/l) with intra- and interassay coefficients of variation (CVs) of 5% and 8%. Plasma d-dimer was measured by ELISA (Zymutest d-dimer, Hyphen BioMed, Neuville-sur-Oise, France) with intra- and interassay CVs of 10% and 15%. Plasmin-α2–anti-plasmin (PAP) complexes were measured by an ELISA kit of Technoclone (Vienna, Austria; normal range: 0–514 ng/ml). Soluble thrombomodulin (sTM) was measured by a commercial sandwich ELISA (Asserachrom thrombomodulin; Diagnostica Stago, Asnieres, France). Thrombin activatable fibrinolysis inhibitor (TAFI) activity was measured by a commercial method (DSM Nutritional Products Ltd Branch Pentapharm, Basel, Switzerland). Plasminogen activator inhibitor-1 (PAI-1) activity was measured by a commercial kit (TriniLIZE PAI-1; Trinity Biotech, Jamestown, NY, USA).

Statistical analyses

Continuous variables were expressed as means ± standard error of the mean (s.e.m.) and tested for statistical significance with a non-parametric Mann–Whitney U-test. Correlation between values was assessed by means of Pearson's correlation test. All P-values are two-sided. P-values lower than 0·05 were considered statistically significant. Statistical analysis was performed using the GraphPad Prism 4·0 computer program (GraphPad, San Diego, CA, USA).

Results

Laboratory measurements of complement and markers of coagulation and fibrinolysis activation in patients with HAE during acute attacks and in remission are reported in Table 1. The changes in coagulation by examining thrombin generation parameters of the NHA are shown in Fig. 1. Three NHA parameters demonstrated increased thrombin generation. Reduced thrombin peak time was observed in patients with HAE and acute attacks (8·3 ± 0·3 min) compared to healthy controls (9·3 ± 0·3 min) (P < 0·05). Thrombin peak height differed significantly between patients with HAE and acute attacks (275 ± 18 nM) in relation to the healthy subjects (184 ± 11 nM) (P < 0·001) and to the patients in R (216 ± 12·0 nM) (P < 0·05). Controls and patients in R also demonstrated a difference (P < 0·05). Additionally, patients during AA had AUC values (2267 ± 118 nM/min) higher than controls (1943 ± 72 nM/min) (P < 0·05) and patients in remission (1870 ± 73 nM/min) (P < 0·05). F1+2 level is higher in patients with acute attacks (1449 ± 220 pmol/l) compared to controls (167 ± 22 pmol/l) (P < 0·0001) and to patients in remission (343 ± 87 pmol/l) (P < 0·0001) (Table 1).

Table 1.  Measurements of complement parameters and coagulation/fibrinolysis markers in 20 healthy control subjects and 43 patients with C1-inhibitor (C1-INH) deficiency (23 during acute attacks and 20 during remission).
Laboratory testsControls (n = 20)Acute attacks (n = 23)Remission (n = 20)
Mean ± s.e.m.Mean ± s.e.m.Mean ± s.e.m.
  1. *Significance P < 0·05; ***P < 0·001; ****P < 0·0001 compared to healthy controls. PAP: plasmin–anti-plasmin; s.e.m.: standard error of the mean.

C1-INH function (%)96 ± 2·616 ± 2·9****28 ± 2·2****
C1-INH antigen (%)90 ± 2·332 ± 5·2****31 ± 5·0****
F1+2 (pmol/l)167 ± 221449 ± 220****343 ± 87
PAP complexes (nM)2·3 ± 0·13·5 ± 0·3***2·7 ± 0·2
D-dimer (ng/ml)516 ± 1503111 ± 1162*1547 ± 650
Figure 1.

Thrombin generation in patients with C1-inhibitor (C1-INH) deficiency. Thrombin peak-time (min) (a), thrombin peak-height (nM) (b), and area under the curve (AUC, nM/min) (c), in healthy controls, patients during acute attacks (AA) and in remission (R). The horizontal lines indicate median values.

The relationship between HAE severity and thrombin generation was examined further and yielded a negative correlation for C1-INH activity (%) with AUC (nM/min) (r = −0·51; P < 0·05) in the patients during AA (Fig. 2a). A significant correlation was also observed between plasma levels of log F1+2 (pmol/l) and AUC (nM/min) (r = 0·42; P < 0·05), indicating that the thrombin generation in non-activated and activated plasma corresponded well (Fig. 2b). No correlations were observed in the patients in R for these results and nor was a significant correlation observed for any of the thrombin generation parameters with any of the other laboratory parameters in either groups.

Figure 2.

Correlation between coagulation parameters in patients with C1-inhibitor (C1-INH) deficiency. (a) Correlation between C1-INH activity (%) and area under the curve (AUC) (nM/min) (b) correlation between F1+2 (pmol/l) and AUC (nM/min). Solid line represents the correlation for the patient under acute attacks (AA). Filled circles refer to patients with AA and open circles refer to patients in remission (R). r, Pearson's correlation coefficient.

The effect of C1-INH deficiency on fibrinolysis parameters is demonstrated in Fig. 3. The plasmin peak-height in patients with acute attacks was reduced (8·6 ± 1·6 nM) compared to healthy subjects (13 ± 1·3 nM) (P < 0·05), and the same was true for the plasmin potential in patients with acute attacks (394 ± 51 nM/min) versus healthy controls (588 ± 63 nM/min) (P < 0·05). Interestingly, PAP complexes and d-dimer measured in plasma demonstrated an opposite effect. PAP complexes in both patients with acute attacks (3·5 ± 0·3 nM; n = 22) and those in remission (2·7 ± 0·2 nM) were elevated significantly compared to healthy subjects (2·3 ± 0·1 nM) (P < 0·001 for both) and d-dimer also showed a significant difference between patients with acute attacks (3111 ± 1162 ng/ml) and controls (516 ± 150 ng/ml) (P < 0·05) (Table 1).

Figure 3.

Evaluation of fibrinolysis in patients with C1-inhibitor (C1-INH) deficiency. Plasmin peak-height (nM) (a), and plasmin potential (nM/min) (b) in healthy controls, patients with C1-INH deficiency during acute attacks (AA) and during remission (R). The horizontal lines indicate median values.

Given the unexpected discrepancy between low fibrinolyic potential (supported by low plasmin peak-height and potential) in activated plasma and high levels of PAP complexes in non-activated plasma, we tested the involvement of sTM, TAFI and PAI-1 (Fig. 4). Plasma levels of sTM were increased during AA (33 ± 10 ng/ml; P < 0·001) compared to controls (23 ± 15 ng/ml; n = 15). TAFI activity was increased during AA (155 ± 9%; n = 15) and in R (137 ± 9%; n = 13), while PAI-1 levels were low during both AA (1·5 ± 0·6 IU/ml; n = 17) and in R (1·5 ± 0·4 IU/ml; n = 11). A statistically significant correlation between TAFI and AUC was present when AA and R were analysed together (n = 29; r = 0·57; P < 0·01).

Figure 4.

Soluble thrombomodulin (sTM) antigen, thrombin activatable fibrinolysis (TAFI) activity and plasminogen activator inhibitor-1 (PAI-1) activity levels in patients with C1-inhibitor (C1-INH) deficiency. sTM antigen, levels were measured in plasma of healthy controls and patients with C1-INH deficiency during acute attacks (AA) and during remission (R). Due to limited sample supply we could measure TAFI and PAI-1 only in a subgroup of patients. The dashed lines indicate upper and lower limit of the normal reference range. The horizontal solid lines indicate median values.

Discussion

The results of this study demonstrated increased thrombin generation in patients with acute attacks when compared to healthy controls. Data obtained in activated plasma paralleled those obtained in non-activated plasma with the thrombin generation marker, F1+2, clearly higher in patients with acute attacks compared to controls. These findings are in agreement with previous observations that activation of FXII and FVII occurs in HAE during attacks [10–12]. Activation of the two factors can occur through independent pathways, and in addition FXII is able to activate FVII [24]. Our present data confirm the previous studies demonstrating thrombin generation during attacks. In addition, we demonstrate that thrombin generation capacity is increased in HAE during attacks and that the balance between activators and inhibitors of coagulation shifts towards generation of thrombin. The inverse correlation between C1-INH plasma levels and both generation of thrombin and F1+2 plasma levels further supports the involvement of coagulation activation in the pathophysiology of HAE. However, the vasopermeability effect of thrombin may be reduced in vivo because of its rapid inactivation by anti-thrombin, whose levels are normal in HAE patients [25]; this inactivation may also explain why HAE patients do not develop thrombosis during acute attacks.

Because of a report of thromboembolic events in non-HAE paediatric patients treated with very high doses of C1-inhibitor concentrate [26], the risk of thromboembolic events is mentioned in the summary product characteristics of plasma derived C1-INH available for treatment of HAE (Berinert® and Cinryze®). Even if our data do not deal directly with the pathophysiological effects of C1-INH replacement, evidence that thrombin generation correlates inversely with C1-INH levels in HAE indicates that C1-INH is anti-thrombogenic rather than prothrombogenic. Thus, our data are partially reassuring regarding the potential thromboembolic risk of C1-INH therapy in HAE patients.

Instead, the data on fibrinolysis were apparently contradictory. Plasmin generation parameters (plasmin peak-height and plasmin potential) were reduced, indicating reduction of the fibrinolytic potential; while the increase of markers of fibrinolysis activation (PAP complexes and d-dimer) suggested the presence of a hyperfibrinolytic state. Elevated levels of PAP complexes during attacks confirmed previous data [12,20,27] The rapid formation of PAP complexes may explain why active fibrinolyis (plasmin) is not seen.

Fibrinolysis is also physiologically inhibited by PAI and by the thrombin-dependent inhibitor TAFI, whose activity is modulated by thrombomodulin. Hence, we measured plasma levels of sTM, TAFI activity and PAI-1 activity (Fig. 4) and found elevation of sTM and TAFI and decrease of PAI-1 during AA. TAFI is the main driver for fibrinolysis inhibition. Additionally, thrombin-TM activates the protein C system, thereby reducing thrombin generation. However, this occurs only at high concentrations. Previous research showed that addition of ∼ 5–10 ng/ml exogenous TM to our assay produced strong reduction of plasmin generation, whereas thrombin generation was only mildly affected [23]. Therefore, elevated levels of TAFI and sTM may indicate that the reduction of plasmin generation may be caused by their in vitro regulatory effect on fibrinolysis inhibition, as observed in our NHA assay.

An alternative explanation for the observed reduced plasmin generation could be the consumption of fibrinolytic factors as shown by reduced plasmin potential, apparent low levels of PAI-1 (Fig. 4c) and activation of fibrinolysis by high PAP and d-dimer levels in patients during AA. Low levels of PAI-1 may be a consequence of elevated thrombin generation associated with protein C activation, leading to consumption of PAI-1.

In conclusion, our findings demonstrate that the simultaneous measurement of thrombin and plasmin generation with the newly developed NHA provides novel information on the net effect of the different enzymes/inhibitors intervening in clot generation and clot lysis in HAE patients. This additional information may be relevant to the clinical and therapeutic monitoring of angioedema attacks.

Disclosure

W.v.H. is co-founder and co-owner of HaemoMagum, a Radboud University Nijmegen Medical Centre spin-off company. HaemoMagum exploits the NHA for preclinical and clinical studies.

Ancillary