Plasma Protein C Inhibitor Is Elevated in Survivors of Myocardial Infarction
Many studies have shown alterations of the hemostatic and fibrinolytic systems in patients with atherosclerotic disease, principally in levels of plasminogen activator inhibitor-1. However, in a large prospective study only fibrinogen, von Willebrand factor antigen, and tissue plasminogen activator antigen were found to be independent risk markers for acute coronary events. The present study evaluated the fibrinolytic system in coronary artery disease, paying particular attention to another inhibitor of fibrinolysis, plasminogen activator inhibitor-3, also called protein C inhibitor (PCI). One hundred fifteen nonanticoagulated male survivors of myocardial infarction were investigated for a range of hemostatic and fibrinolytic parameters that were compared with values in 87 age-matched healthy control male subjects. PCI active antigen was significantly (P<.03) elevated in the myocardial infarction group compared with the control group and was associated with the number of acute coronary events suffered (P=.005) but not with the severity of disease as determined by coronary angiography. Elevated PCI plasma levels can be considered as a risk marker for acute coronary events and might be of particular importance in the pathogenesis of this disease due to the interference of PCI in both the anticoagulant and fibrinolytic systems.
- myocardial infarction
- protein C inhibitor
- plasminogen activator inhibitor-1
- tissue plasminogen activator
Presented at the 37th Annual Meeting of the American Society of Hematology, Seattle, Wash, December 1-5, 1995.
- Received August 8, 1996.
- Revision received October 15, 1996.
Alterations of factors involved in the regulation of the hemostatic and fibrinolytic systems have been used extensively to describe various patient populations, including those with atherosclerotic disease,1 MI,2 3 4 5 6 angina pectoris,7 8 9 and peripheral arterial occlusive disease.10 The rationale for these approaches is that an impaired fibrinolytic system is thought to be responsible for delayed fibrin clot dissolution, thereby leading to intravascular fibrin deposition and thrombosis (for review, see Reference 11). Such an approach is related to both the clotting theory in the pathogenesis of atherosclerosis and to fibrin clot formation at the site of an atherosclerotic plaque being causative for the acute event (for reviews, see References 12 and 13).
In fact, disturbances of the fibrinolytic system have been shown in young survivors of MI,2 acute MI,4 14 peripheral arterial occlusive disease,10 and angina pectoris.7 In the latter, a large prospective study,7 fibrinogen, vWF:Ag, and TPA antigen were all independent predictors of subsequent coronary events. PAI-1 activity and antigen, however, were not significantly higher in the event group compared with the nonevent group in this cohort of patients, although PAI-1 levels have been shown to be increased in many other studies.1 2 3 4 A second important inhibitor, PAI-3, is also involved in the regulation of the fibrinolytic system. PAI-3 is also termed PCI due to its involvement in the protein C anticoagulant pathway as an inhibitor of activated protein C. PCI is a heparin-dependent serine protease inhibitor (serpin) with rather broad specificity (for review, see Reference 15). As yet, determination of PCI in patients with atherosclerotic disease is rare or has been undertaken in poorly defined patient groups.16 Therefore, it was the aim of the present study to analyze the hemostatic and fibrinolytic systems in 115 nonanticoagulated male survivors of MI, with particular focus on PCI measurement, and to compare the values with those in 87 age-matched male control subjects.
We here provide evidence that PCI levels are elevated in a group of MI survivors compared with levels in healthy control subjects and that the elevated PCI levels are significantly associated with the number of acute coronary events suffered by these patients.
The study group consisted of 115 Swiss, Caucasian, male patients who had survived at least one confirmed MI. These subjects were part of a cohort of 200 patients selected between March and June 1994 from records of the Cardiology Department, University Hospital of Bern. Survivors of MI that had occurred at least 2 months before the investigation were included in the present analysis. From the original 200-member cohort, patients receiving oral anticoagulation therapy and female patients were excluded due to possible confounding influences on PCI levels (V.A.C., unpublished data, 1996). Patients were free from malignancy, hepatopathy, or debilitating disease. All except one patient had undergone coronary angiography up to 1 year before blood sampling. Coronary heart disease was in a stable phase in all patients.
For the control group, 87 Swiss, Caucasian, age-matched male volunteers were simultaneously recruited to the study. Control subjects were not individually age-matched to the MI patients. Only recruits negative for a history of thromboembolic disease or bleeding tendency were accepted. The clinical and metabolic data for the two groups are given in Table 1⇓.
Informed consent was obtained from all subjects investigated. The study was approved by the local ethics committee of the University Hospital of Bern.
Blood collection was performed according to acknowledged procedures17 and rigorously standardized to avoid diurnal variations in fibrinolytic activity.18 Blood was drawn from an antecubital vein with a 19-gauge butterfly needle and anticoagulated by collection into Monovettes (Sarstedt) containing 0.106 mol/L trisodium citrate. To stabilize active TPA and PAI-1, blood was also collected into acidified sodium citrate (0.45 mol/L, pH=4.3). Plasma was prepared by centrifuging twice for 10 minutes each at 1500g at room temperature. Thereafter, citrated plasma and acidified citrated plasma were stored in polypropylene tubes at −70°C until analysis.
Functional Immunological Assay for PAI-1 Activity
PAI-1 active antigen was measured in acidified citrated plasma by using a modified ELISA (Technoclone Inc).19 Briefly, an excess of active TPA was immobilized on microtiter plates by an anti-TPA monoclonal antibody (3 VPA). Active PAI-1 present in the sample was allowed to bind to the TPA, thereby forming a complex that was immunologically detected by using an HRP-labeled anti–PAI-1 monoclonal antibody (3 PAI). For standardization purposes, purified melanoma PAI-120 was added to pooled normal plasma and calibrated against the PAI-1 reference reagent (87/512) from the National Institute for Biological Standards and Control (NIBSC), Potters Bar, Hertfordshire, UK.
Functional Immunological Assay for PCI Activity
PCI active antigen was measured in citrated plasma in an assay that is, in principle, similar to that for PAI-1 active antigen21 (Technoclone Inc); here, however, the protease used to bind PCI in plasma was UPA. Excess UPA was immobilized on microtiter plates by an using an anti-UPA monoclonal antibody (35 sc-UPA). Active PCI present in the sample was allowed to bind to the UPA, thereby forming a complex that was immunologically detected by using an HRP-labeled anti-PCI monoclonal antibody (4 PCI). Freeze-dried pooled plasma from healthy volunteers was used as a standard and defined as containing 100% of active PCI.
Total PAI-1 Antigen ELISA
Total PAI-1 antigen was measured in acidified citrated plasma in a sandwich ELISA (Technoclone Inc).19 Active and latent PAI-1 as well as TPA–PAI-1 complexes were recognized by the catching antibody (5 PAI). An HRP-labeled monoclonal antibody (3 PAI) was used for detection. Purified melanoma PAI-120 was used as a calibrator.
Functional Immunological Assay for TPA Activity and Antigen
TPA activity and total antigen were determined in acidified citrated plasma by using a bioimmunoassay (Technoclone Inc) as described by Wojta et al.22 This assay essentially consisted of two steps. First, TPA contained in plasma was bound to microtiter plates via an antibody (3 VPA) that did not interfere with the active site on the TPA molecule. After other plasma components were washed away, the functional activity of bound TPA was quantified by its plasminogen-activating capability. Second, total TPA antigen was immunologically determined by using an HRP-labeled monoclonal antibody (10 VPA). Recombinant TPA standardized against the Second International TPA standard (NIBSC 86/670) was used as a calibrator.
PAP Complex ELISA
PAP complexes were determined in citrated plasma (Technoclone Inc).23 Briefly, a monoclonal antibody (7 AP) directed against the neoantigen part of the PAP complex was used to specifically bind PAP complexes. Detection was achieved by using an HRP-labeled monoclonal antibody (2 PG) that recognized the kringle 1 through 3 region of the plasmin(ogen) part of the complex. To obtain a calibrator, maximum PAP complexes were generated in fresh, pooled, citrated plasma from healthy donors by adding saturating concentrations of UPA. The formed PAP complexes in the plasma standard were then calibrated against PAP complexes generated from purified components.
Metabolic parameters were determined according to standard laboratory procedures. Fibrinogen was determined according to the method of Clauss.24 FV and FVII clotting activities were analyzed by using prothrombin time–based assays using FV- and FVII-deficient substrate plasmas, respectively. vWF:Ag was determined by using an ELISA.
Possible differences between the control and MI groups were tested by the Mann-Whitney rank sum test to take into account nonsymmetrical distributions. For descriptive purposes, the MI group was subdivided into survivors of OAE and MAEs. Due to varying levels of skewness, a logarithmic transformation was performed on hemostatic and fibrinolytic variables, and a linear trend was calculated across three groups: normal (no acute events), OAE, and MAEs. Probability values of <.05 were considered significant for the Mann-Whitney analysis and the test for linear trend. Bivariate correlations were investigated in the normal and total MI groups separately according to Spearman; probability values of <.001 were considered significant.
For subsequent analysis of PCI levels, a cutoff point was determined by means of a probit plot.25 26 Briefly, the skewed frequency histogram distributions of the normal and total MI groups were log transformed to obtain more normal distributions. Thereafter, the area under the curve was integrated to achieve a straight line. The MI group was integrated from left to right, and the normal group was integrated from right to left. The point at which the two lines crossed was taken as the cutoff value (126%). This value was then used to obtain two groups of MI patients, with either low (≤126%) or high (>126%) PCI levels. This procedure provided the best separation between false positives and false negatives. The significance between the low and high PCI groups with respect to OAE versus MAEs was determined by means of a 2×2 contingency table with a two-sided Fisher's exact test. The above analyses were performed using the STATISTICA for Windows program, version 4.1.
PCI active antigen median (range) plasma levels were significantly (P<.03) increased in the MI group (123.0% [8.0% to 1000.0%]) versus 111.0% (16.0% to 377.0%) in the control group. PAI-1 active antigen was very significantly (P<.005) elevated in the patient group (8.9 U/mL [1.0 to 46.0 U/mL]) versus 6.3 U/mL (0.6 to 43.8 U/mL) in the control subjects, whereas the difference in total PAI-1 antigen between the MI group (26.5 ng/mL [2.0 to 113.6 ng/mL]) and the control group (22.4 ng/mL [0.5 to 116.2 ng/mL]) was not quite significant (P=.09). A significant (P<.01) increase was observed for total TPA antigen in the patient group (10.2 ng/mL [2.1 to 62.0 ng/mL]) compared with the normal volunteers (8.2 ng/mL [1.8 to 27.7 ng/mL]). PAP complexes were significantly (P<.05) decreased in the MI group (110.0 ng/mL [0 to 1500.0 ng/mL]) versus the control subjects (131.0 ng/mL [0 to 514.0 ng/mL]).
For further analysis, MI patients were subdivided into OAE and MAE groups. An association with events was determined for established clinical risk factors (Table 2⇓) and fibrinolytic variables (Table 3⇓) by means of a test for linear trends. HDL cholesterol showed a significant linear trend of decreased levels with increasing number of events (P<.01). vWF:Ag and FV and FVII clotting activities showed significant linear trends of increased levels with number of coronary events suffered (P=.006, P<.001, and P<.02, respectively). PAI-1 and PCI activities were the only fibrinolytic variables for which increased levels were significantly associated with acute coronary events (P<.05 and P=.005, respectively). A linear trend was observed for decreased PAP complexes with increased number of survived infarctions (P=.15).
PCI was correlated with all metabolic, hemostatic, and other fibrinolytic parameters measured (Table⇑s 1 through 3) using the Spearman nonparametric test in the control and total MI groups separately. In the control group, PCI significantly and positively correlated with vWF:Ag (r=.44, P<.0001). In the MI group, PCI did not significantly correlate with any parameters measured.
To further analyze the significance of PCI for sustained reinfarction, patients in the total MI group were split into high-PCI (n=56) and low-PCI (n=59) groups as determined according to the cutoff point between the normal and MI groups described above. The relative frequency of reinfarction was 10% in the high group and 4% in the low group (P=.11 by Fisher's exact test). However, when the MI group was further subdivided into three groups (one-, two-, and three-vessel disease as determined by angiography [Table 1⇑]), and the relative occurrence of reinfarction was plotted for the three groups, the results shown in the Figure⇓ were obtained. (One MI patient with no demonstrable CAD was omitted from this analysis). The occurrence of reinfarction increased with the number of affected vessels for the total MI group, but the relative chance of having had reinfarction for the low-PCI group was always below the total population and for the high-PCI group always above the total population. When the significance test was restricted to one- and two-vessel disease patients, the occurrence of reinfarction was significantly higher for the high-PCI group than the low-PCI group (P<.05 by Fisher's exact test; relative risk=4.2; 95% confidence interval=0.94 to 18.6). When the median PCI value in the MI group (123%) was used as the cutoff point, no alterations in the above outcomes were observed. There was, however, no correlation of PCI levels with severity of disease as determined by the number of affected vessels; median (range) PCI levels were 125% (48% to 1000%; n=43) for one-vessel disease, 124% (8% to 385%; n=38) for two-vessel disease, and 122% (29% to 264%; n=32) for three-vessel disease.
This study showed that male survivors of MI had elevated PCI levels compared with an age-matched male control group. Furthermore, the MI patients had significantly higher PAI-1 activity and TPA antigen and significantly lower PAP complexes than the control group, whereas the increase in PAI-1 antigen was not quite significant. These results are consistent with data reported from other studies concerning PAI-1 activity and TPA antigen1 2 3 4 7 8 9 10 and extend current insight into the fibrinolytic system in MI patients insofar as PCI, a second major inhibitor of the fibrinolytic system, was found to be increased and the overall plasmin formation in these patients was decreased as reflected by lower levels of circulating PAP complexes. The increased levels of TPA antigen found in MI patients most likely reflected elevated TPA–PAI-1 complex levels. This study therefore shows an effective decrease in the overall fibrinolytic activity in these patients.
Upon further analysis, a significant linear trend of increased levels of PCI across control, OAE, and MAE groups was noted. Among the hemostatic parameters measured, such a significant linear trend could only be found for FV clotting activity and vWF:Ag and to a lesser extent for FVII clotting activity. PAI-1 activity was the only other fibrinolytic factor measured that was associated with survived events. That the three groups differed in their relative risk is shown by the fact that significant linear trends were shown for established clinical risk factors, confirming the validity of the separation of patients into these groups. From these results, it can be concluded that PCI is related to CAD and among CAD patients linked to the acute events suffered.
To further prove the possible importance of PCI for either occurrence of reinfarction or severity of disease as determined by coronary angiography, the differences in the frequency of multiple infarctions were analyzed according to the number of affected vessels. The occurrence of reinfarction was significantly higher in those patients with one- and two-vessel disease and with elevated PCI levels compared with those with low PCI levels. In patients with three-vessel disease there was no significant difference in the occurrence of reinfarction between the high- and low-PCI groups. When PCI levels were compared in one-, two-, and three-vessel disease, no difference in PCI levels could be found, which indicates that high PCI levels are not related to the severity of disease as determined by coronary artery involvement but rather with the number of acute events suffered. Increased PCI levels might not only inhibit activated protein C directly, thereby impairing the protein C anticoagulant pathway, but may also inhibit activation of protein C by complex formation with thrombin in the presence of thrombomodulin.27 In addition, PCI is not only capable of inhibiting the activated protein C pathway but also interferes with activation of the fibrinolytic system by complex formation with urokinase.28 Therefore, higher PCI plasma levels would seem to be associated with infarction incidence by leading to an increased tendency for thrombus formation by inhibition of both the anticoagulant and fibrinolytic systems.
Selected Abbreviations and Acronyms
|CAD||=||coronary artery disease|
|ELISA||=||enzyme-linked immunosorbent assay|
|MAE||=||multiple acute event|
|OAE||=||one acute event|
|PAI||=||plasminogen activator inhibitor|
|PCI||=||protein C inhibitor|
|TPA||=||tissue plasminogen activator|
|UPA||=||urokinase plasminogen activator|
|vWF:Ag||=||von Willebrand factor antigen|
This work was supported by grant 32-36443.92 from the Swiss National Science Foundation and grants from the Austrian Fund for the Promotion of Scientific Research.
Cortellaro M, Cofrancesco E, Boschetti C, Mussoni L, Donati MB, Cardillo M, Catalano M, Gabrielli L, Lombardi B, Specchia G, Tavazzi L, Tremoli E, Pozzoli E, Turri M. Increased fibrin turnover and high PAI-1 activity as predictors of ischemic events in atherosclerotic patients: a case-control study. Arterioscler Thromb. 1993;13:1412-1417.
Jansson J, Olofsson B, Nilsson TK. Predictive value of tissue plasminogen activator mass concentration on long-term mortality in patients with coronary artery disease: a 7-year follow-up. Circulation. 1993;88:2030-2034.
Paramo JA, Colucci M, Collen D, van de Werf F. Plasminogen activator inhibitor in the blood of patients with coronary artery disease. BMJ. 1985;291:573-574.
Huber K, Rosc D, Resch I, Schuster E, Glogar DH, Kaindl F, Binder BR. Circadian fluctuations of plasminogen activator inhibitor and tissue plasminogen activator levels in plasma of patients with unstable coronary artery disease and acute myocardial infarction. Thromb Haemost. 1988;60:372-376.
Thomson JM. Blood collection and preparation: pre-analytical variation. In: Jespersen J, Bertina RM, Haverkate F, eds. ECAT Assay Procedures: A Manual of Laboratory Techniques. Dordrecht, Netherlands: Kluwer Academic Publishers; 1992:13-20.
Angleton P, Chandler WL, Schmer G. Diurnal variation of tissue-type plasminogen activator and its rapid inhibitor (PAI-1). Circulation. 1989;79:101-106.
Wagner OF, Binder BR. Purification of an active plasminogen activator inhibitor immunologically related to the endothelial type plasminogen activator inhibitor from the conditioned media of a human melanoma cell line. J Biol Chem. 1986;31:14474-14481.
Carroll V, Griffiths M, Geiger M, Merlo C, Lämmle B, Binder BR. Active protein C inhibitor (PCI) as an additional discriminatory factor for patients at risk of myocardial infarction. Blood. 1995;86:379a. Abstract.
Hattey E, Haumer M, Binder A, Binder BR. Plasmin-α2-antiplasmin complexes. In: Jespersen J, Bertina RM, Haverkate F, eds. ECAT Assay Procedures: A Manual of Laboratory Techniques. Dordrecht, Netherlands: Kluwer Academic Publishers; 1992:173-183.
Harris EK, Yasaka T. On the calculation of a ‘reference change’ for comparing two consecutive measurements. Clin Chem. 1983;29:25-30.
Rezaie AR, Cooper ST, Church FC, Esmon CT. Protein C inhibitor is a potent inhibitor of the thrombin-thrombomodulin complex. J Biol Chem. 1995;270:25336-25339.
Geiger M, Huber K, Wojta J, Stingl L, Espana F, Griffin JH, Binder BR. Complex formation between urokinase and plasma protein C inhibitor in vitro and in vivo. Blood. 1989;74:722-728.