Relationships Between Homocysteine, Factor VIIa, and Thrombin Generation in Acute Coronary Syndromes
Background—It has been suggested by clinical, epidemiological, and experimental in vitro studies that homocysteine potentiates thrombin generation. This prothrombotic effect however has not previously been demonstrated in patients presenting with acute coronary syndromes (ACS).
Methods and Results—Patients with ACS (n =117) presenting with confirmed acute myocardial infarction (MI) (n =57) or unstable angina pectoris (UAP) (n =60) were consecutively recruited together with patients (n =18) in whom the presenting chest pain was not of cardiac origin (NCP), included as controls. Plasma samples were collected on admission and before clinical intervention. Homocysteine was assayed by high performance liquid chromatography, and both Factor VIIa and prothrombin fragment F1+2 were analyzed by ELISA. There were significant elevations in F1+2 in MI (P<0.001) and UAP (P=0.003), and modest elevations in Factor VIIa in UAP (P<0.05) compared with NCP but no differences in homocysteine levels among those groups. On dividing patients with ACS into quartiles of homocysteine, there was a stepwise increase in F1+2 (P<0.0001) and of Factor VIIa (P<0.05). There were significant correlations in ACS between homocysteine and F1+2 (r=0.46, P<0.0001), homocysteine and Factor VIIa (r=0.24, P<0.01), and F1+2 and Factor VIIa (r=0.41, P<0.0001). There was no correlation between homocysteine and either F1+2 (r=−0.15, P=0.57) or Factor VIIa (r=0.22, P=0.37) in the NCP patients.
Conclusions—Elevated plasma homocysteine is associated with and may cause elevated Factor VIIa and thrombin generation in patients presenting with ACS. These findings suggest an explanation for the prothrombotic effect of homocysteine in ACS.
Recent epidemiological studies have shown that moderately elevated plasma homocysteine concentrations are associated with an increased risk for the development of atheromatous (cerebrovascular1 2 and peripheral vascular disease3 4 and thrombotic vascular diseases (deep vein thrombosis).5 Coronary artery disease may encompass both an atheromatous lesion and thrombosis during acute coronary events. There is a known association between elevated homocysteine levels and the development of coronary artery disease.6 7
Homocystinuria, the genetic form of hyperhomocystinemia, is associated with a marked rise of plasma homocysteine levels and an early risk of mortality from thrombotic vascular disease.8 Although the exact mechanism of atherothrombosis associated with hyperhomocystinemia is not clearly understood, several studies have pointed to an association with inhibition of thrombomodulin activity,9 reduction of protein C activation,10 11 increased platelet aggregation,12 and predisposition to endothelial cell injury.13 14 Most of these studies, however, were in vitro studies, where supraphysiological concentrations of homocysteine were used.
Disruption of coronary arterial atheromatous plaques is known to initiate a thrombotic process15 16 leading to complete or partial occlusion in the acute coronary syndromes (ACS) of myocardial infarction (MI) and unstable angina pectoris (UAP). Disrupted coronary plaques are thought to contain an abundance of tissue factor17 which, when exposed to circulating blood, forms a complex with Factor VII/VIIa initiating the coagulation cascade. Factor VIIa is generated by enzymatic cleavage of Factor VII by various circulating proteases, resulting in rapid generation of Factor IXa and Factor Xa. How these processes may be influenced by homocysteine in plasma in clinical syndromes characterized by artherothrombotic disease is uncertain.
We report an investigation of the relationship between plasma homocysteine and activation of coagulation in patients presenting with acute coronary ischemia, studied on admission and before clinical intervention.
Patients and Methods
Consecutive patients (n=135) were recruited on admission to the Coronary Care Unit at Charing Cross Hospital in London between October 1995 and January 1997. Written informed consent was obtained from all the patients with the approval of the local ethical committee at Charing Cross Hospital. The patients were given a provisional diagnosis to guide appropriate treatment, but this diagnosis was subsequently reviewed during follow-up, and final diagnoses (MI, UAP and noncardiac chest pain [NCP]) were used in all analyses.
Ml was diagnosed according to the World Health Organization criteria.18 UAP was diagnosed if the WHO criteria for MI were not met, all enzyme measurements (creatine kinase, aspartate transaminase, hydroxybutyrate dehydrogenase) were below twice their upper reference range throughout the routine sampling period (daily for 3 days). Evidence of ischemic heart disease was demonstrated by either follow-up cardiac event (death, coronary revascularization, or Ml), a positive exercise treadmill test (>0.1 mV ST segment depression 80 ms after the J point) or demonstration of ischemia on thallium radio-isotopic scanning. NCP was diagnosed if angina was excluded by an alternative source of pain, a negative exercise treadmill test, or normal coronary anatomy on subsequent angiography. Patients were excluded from the study if they were receiving anticoagulant treatment before admission.
Full clinical, biochemical, management and result of investigations were recorded on all patients. Clinical details included risk factor assessment for coronary artery disease. Smoking status was classified as current smokers and nonsmokers; current smokers included individuals who stopped <4 weeks before enrollment in the study. A known history of hypertension and/or diabetes mellitus and treatment received before admission were also recorded. Routine lipid profile estimation was performed on admission for all the patients documenting their total cholesterol, triglycerides, and HDL levels.
Venous blood samples were obtained on admission by trained medical or senior nursing staff from a cuffed antecubital vein before initiation of thrombolysis or anticoagulant treatment. Care was taken to obtain this sample with a fresh needle and syringe. Blood was transferred into bottles containing 0.105 mol/L sodium citrate for Factor VIIa and F1+2 assay and bottles containing EDTA (7.5%) for homocysteine assay (Becton Dickinson). Within 15 minutes of collection, platelet-poor plasma was obtained by centrifugation at room temperature for 15 minutes at 3000g, flash frozen on dry ice, and then transferred to a −80°C freezer. Blind analysis of all samples was performed in batches at completion of sample collection. The protocol for sample collection and storage was formulated during a prior study published elsewhere.19 Emphasis was put on clean venipuncture to minimize artifacts. This was a particular concern because sampling was performed in the environment of a busy coronary care unit. To provide reassurance that artifactual elevations in samples was minimal, blood samples were also taken from a group (n =11) of age-matched healthy controls; these samples were similarly processed and assayed. The results of the assay of these samples were not significantly different from those of the NCP group (which were collected in the coronary care unit and subsequently assayed blind), suggesting that the sampling strategy was adequate.
Plasma total homocysteine, which includes the sum of protein-bound and free homocysteine, was measured by high-performance liquid chromatography with fluorescence detection.20 21 The intra- and interassay coefficients of variation were ≤5%. Plasma homocysteine was recorded in micromoles per liter.
Factor VIIa and Prothrombin Fragment F1+2 Measurements
These were performed with specific ELISAs (developed in-house and fully reported elsewhere).19 22 23 The assay for F1+2 was performed as originally described23 and has intra- and interassay coefficients of variation of ≈10% and ≈15%, respectively. The normal laboratory mean (±SEM) is 31.0±3.7 ng/mL, which corresponds to 0.88±0.11 nmol/L. The Factor VIIa assay was modified, as described22 by replacement of the standard, which altered the normal mean to 0.25±0.10 ng/mL (0.005±0.003 nmol/L). The intra- and interassay coefficients of variation are ≈4% and ≈10%, respectively. The ELISA to Factor VIIa measures free Factor VIIa in plasma. Free Factor VIIa will have been generated and released from a cell surface on which tissue factor has been expressed. By its nature, the Factor VIIa ELISA cannot measure any Factor VIIa still bound to the cell surface.
Because the distribution of values for plasma homocysteine, F1+2, and Factor VIIa were positively skewed, the variables were logarithmically transformed; geometric means (95% CI), unless otherwise stated, are presented throughout the text, tables, and figures. Median and interquartile (25th to 75th) ranges of plasma homocysteine were used to categorize the patients into quartiles in MI (12.7, 11.6 to 16.8 μmol/L), UAP (13.1, 10.2 to 16.5 μmol/L), and ACS (12.9, 10.9 to 16.5 μmol/L). Patients with a final diagnosis of NCP were used as controls to establish that in those patients with definite ACS, coagulation activation had occurred and sampling artifacts had been kept to a minimum. The median homocysteine level of 12.0 μmol/L was used to categorize the NCP patients above and below this level. Group comparisons of the log-transformed values were made by independent-sample Student’s t test for 2 groups and ANOVA for >2 groups. Bivariate correlations were tested using Pearson’s method and correlation coefficients (r). All reported probability values are 2-sided. Analyses were performed using SPSS for Windows (SPSS Inc).
Clinical details of the 117 ACS and 18 NCP patients are given in Table 1⇓. Mean age, sex, and risk factor distribution of the patients with ACS was not statistically different from those with NCP.
The results for plasma homocysteine, Factor VIIa, and F1+2 concentration in ACS are presented in Table 2⇓, stratified by age, sex, smoking status, cholesterol, HDL, hypertension, and diabetes mellitus. Homocysteine showed significant difference with increasing age. Factor VIIa did not vary significantly with any of these factors. There was a highly significant difference in F1+2 levels with age, a finding that has been often reported. F1+2 was also noted to be slightly higher in women and in patients with hypertension.
The respective relationships of Factor VIIa and F1+2 to final diagnosis were examined. F1+2 levels were significantly (P<0.001) higher in patients with MI (44.8, 37.0 to 53.4 ng/mL) than in patients with NCP (25.1, 20.0 to 31.6 ng/mL), but not significantly different in patients with UAP (39.3, 33.2 to 46.6 ng/mL), Figure 1a⇓. Factor VIIa was significantly higher in patients with UAP (0.38, 0.29 to 0.51 ng/mL) than in patients NCP (0.22, 0.16 to 0.31 ng/mL) but not different in patients with MI (0.31, 0.23 to 0.41 ng/mL), Figure 1b⇓. Evidence of a relationship between changes in levels of F1+2 and Factor VIIa was then sought through correlation analysis. There were significant correlations between F1+2 and Factor VIIa levels when considering results in MI (r=0.38, P=0.01, n =58), in UAP (r=0.45, P=0.003, n =60), and when the ACS were considered as one group (r=0.41, P<0.0001, n =117). There was no correlation between F1+2 and Factor VIIa levels in the NCP group (r=−0.06, P=0.59). These observations confirm a hypercoagulable state in these patients presenting with ACS but not in patients with NCP.
Plasma homocysteine concentrations were similar in both MI (13.8, 12.5 to 15.3 μmol/L) and UAP (13.3, 12.2 to 14.5 μmol/L), but they were not significantly higher than NCP (12.0, 9.7 to 14.8 μmol/L), Figure 1c⇑.
The relationship between the coagulation activation markers (F1+2 and Factor VIIa) and homocysteine was studied by dividing the latter into quartiles. It can be seen from Figure 2a⇓ that in MI there is a stepwise increase in F1+2 with increasing quartiles (ANOVA, P=0.05): there is a significant difference (P=0.007) between F1+2 levels in the lowest (33.5, 27.2 to 41.2 ng/mL) and highest (60.1, 40.4 to 89.4 ng/mL) homocysteine quartiles. There was a similar stepwise increase in F1+2 levels (ANOVA, P=0.01) in UAP, with a significant difference (P<0.001) between lowest (32.1, 25.1 to 41.0 ng/mL) and highest (68.4, 45.6 to 102.7 ng/mL) homocysteine quartiles, Figure 2b⇓. When the results for all patients with ACS were combined (MI plus UAP), there was a significant graded increase in F1+2 between the homocysteine quartiles (ANOVA, P<0.0001). The difference between the upper (63.4, 49.0 to 81.9 ng/mL) and lower (31.1, 27.1 to 35.7 ng/mL) homocysteine quartiles was highly significant (P<0.0001; Figure 2c⇓). When all results for homocysteine and F1+2 in the ACS were compared by correlation analysis, there was a significant correlation (r=0.46, P<0.0001, n=117).
A similar analysis of relationship between quartiles of homocysteine and Factor VIIa levels showed no change in MI (Figure 3a⇓). There was, however, a significant increase overall in Factor VIIa levels in UAP (ANOVA, P<0.05) and between lowest levels (0.25, 0.16 to 0.37 ng/mL) and highest levels (0.71, 0.44 to 1.13 ng/mL, P<0.001; Figure 3b⇓). When all results in the ACS were included in a quartile analysis (Figure 3c⇓), there was a significant difference overall (P<0.05) and a significant difference (P<0.001) between lowest quartile Factor VIIa (0.22, 0.17 to 0.28 ng/mL) and highest quartile (0.49, 0.31 to 0.78 ng/mL). Correlation analysis of Factor VIIa with homocysteine reflected these results: when UAP results were correlated, there was a significant association (r=0.4, P=0.007, n=60). This was weakened by combining results in MI with UAP (r=0.24, P<0.01, n=118).
Because of the sample size of the NCP group, quartile analysis was inappropriate. Instead, F1+2 and Factor VIIa levels were grouped above and below the median homocysteine level of 12.0 μmol/L. F1+2 levels were not different between the groups (21.1, 15.2 to 29.2 and 30.0, 21.2 to 42.4 ng/mL, respectively, P=0.16). Similarly, there was no difference in Factor VIIa levels (0.29, 0.26 to 0.33 and 0.27, 0.21 to 0.34 ng/mL, respectively, P=0.59). Correlation analysis reflected these results with the absence of any correlation between homocysteine and F1+2 (r=−0.15, P=0.57, n=8) and Factor VIIa (r=0.22, P=0.37, n=8). It cannot be discounted, however, that this result could have been caused by the small sample size in this group.
The blood coagulation cascade is known to be activated in the acute phase of MI and UAP. Plasma levels of Fibrinopeptide A, F1+2, have been shown to be elevated, signaling activation of the end stages of the coagulation process. Plaque rupture and fissuring is thought to be a critical event in the pathogenesis of the ACS. This is associated with exposure of tissue factor, which can trigger the coagulation cascade via Factor VIIa generation. However, previous studies, which have investigated Factor VIIa levels, have generally failed to demonstrate changes that could be responsible for the readily detectable increase in thrombin generation (F1+2 elevation).24 An exception was a small previous study from this center, which showed elevated Factor VIIa levels, particularly in UAP.19
The design of our study differs from most investigations in that we have recruited all patients admitted to our coronary care unit and used patients diagnosed with NCP as a control group. This group was small because it comprised those patients who presented with suspected ACS but who were shown not to have the syndromes. It was mainly used as a control to establish that in those patients with definite ACS, coagulation activation had occurred and that sampling artifacts had been kept to a minimum. The main focus of this study was therefore not to study the relation between the ACS and NCP groups but the relationship between the degree of coagulation activation and homocysteine concentrations in patients in whom a thrombotic process had taken place. Although some formal analysis of results from the NCP group is presented, it must be stressed that the group size makes the interpretation of these results tentative.
We have been able to show a convincing association between Factor VIIa and F1+2 in the acute phase of the ACS. This association, manifest by correlation analysis, persists on subgroup analysis, despite the generally less clear elevation of Factor VIIa than F1+2 in MI. This latter finding could help explain the failure to demonstrate elevation of Factor VIIa in other studies of ACS. There is an important methodological aspect involved in the determination of Factor VIIa as an activation marker, which distinguishes it from other activation markers (eg, F1+2). Cellular expression of tissue factor is required for activation of Factor VII.15 16 17 An ability to detect Factor VIIa in plasma implies that sufficient Factor VIIa is released from the cell surface during the activation process to enable detection. Support for this possibility of detection in plasma is provided by results of in vitro activation experiments conducted in plasma. Addition of tissue factor and phospholipid vesicles to recalcified plasma results in generation in more Factor VIIa on a molar basis than the added tissue factor (H. Philippou, PhD, and D.A. Lane, PhD, unpublished observations, 1999). This suggests that each molecule of tissue factor can undergo multiple rounds of Factor VII activation, progressively releasing Factor VIIa. It might be that the low level activation of Factor VII observed in this clinical context might arise from the difficulty in its detection because of its slow release from the cell surface; this might also explain the lower levels of Factor VIIa observed in plasma in MI compared with UAP. There are additional methodological issues in the measurement of Factor VIIa levels. It is not clear whether our use of an ELISA for quantification of Factor VIIa levels was important for the demonstration of a relationship between Factor VIIa and F1+2 levels, because at the inception of this study we did not have available the functional assay for Factor VIIa. Regardless of these methodological points, our results provide an explanation for increased F1+2 levels in the ACS, by linking the initial cellular response (believed to be exposure of tissue factor) with the outcome (increased thrombin generation) through an essential proteinase (Factor VIIa) involved in activation of blood coagulation.
The main novel feature of our study has been to relate homocysteine levels to coagulation activation. A significant association between plasma homocysteine levels and coagulation activation has been demonstrated both in relation to changes in F1+2 and Factor VIIa levels. F1+2 plasma levels correlated positively and in a concentration-related manner to plasma homocysteine levels. This pattern was seen in patients with MI and UAP but not in patients with NCP (however, note the caution about the sample size). The association between F1+2 levels and homocysteine has also previously been shown by Kyrle et al in patients with prior history of venous thromboembolism.25 In a similar patient group, Cattaneo et al did not find any correlation.26 In the latter study, however, blood samples were taken at least 4 months after the last thrombotic event. Furthermore, there was no difference in F1+2 levels between patients with prior history of thromboembolism and healthy controls, suggesting no obvious activation of coagulation. This contrasts to the present study, in which we have sampled on presentation and before any intervention in an attempt to study patients in the acute hypercoagulable state. The association between homocysteine and Factor VIIa levels is less prominent in the current study, possibly because of the relatively low elevation of Factor VIIa levels arising for the reasons discussed above.
How could homocysteine interact with coagulation to enhance activation? Fryer et al have suggested that high levels of plasma homocysteine may initiate coagulation by enhancing endothelial cell expression of tissue factor activity,27 and this certainly seems plausible in the light of the present observation of a correlation between homocysteine and Factor VIIa. Inhibition of protein C activation and downregulation of thrombomodulin is also well-documented.9
This study has limitations: its power is limited by its size, this has prevented a detailed statistical analysis of possible confounders. The size of the NCP group has prevented firm conclusions from being drawn on differences between ACS and NCP patients regarding associations between coagulation activation markers and homocysteine.
Despite these limitations, this study has shown for the first time that elevated plasma homocysteine in patients presenting with ACS appears to be associated with increased coagulation activation as judged by increases in F1+2 and Factor VIIa levels. Although there is plausible experimental evidence to suggest that these are causally related, further experimental and larger scale clinical studies are required to confirm our findings. If such confirmation is obtained, it would suggest therapeutic modulation of plasma homocysteine as a means of reducing increased thrombin generation in ACS.
This study was supported by a Concerted Action Grant of the Biomed 1 Program of the European Commission (PL92005), and by a grant (PG/95128) from the British Heart Foundation.
- Received June 4, 1999.
- Revision received August 10, 1999.
- Accepted August 26, 1999.
- Copyright © 2000 by American Heart Association
Mudd S, Levy H, Skovby F. Disorders of transsulfuration. In: Scriver C, Beaudet A, Sly W, eds. Metabolic and Molecular Basis of Inherited Disease. New York: McGraw-Hill; 1995:1279–1327.
Hayashi T, Honda G, Suzuki K. An atherogenic stimulus homocysteine inhibits cofactor activity of thrombomodulin and enhances thrombomodulin expression in human umbilical vein endothelial cells. Blood. 1992;79:2930–2936.
Lentz SR, Sadler JE. Inhibition of thrombomodulin surface expression and protein C activation by the thrombogenic agent homocysteine. J Clin Invest. 1991;88:1906–1914.
Rodgers GM, Conn MT. Homocysteine, an atherogenic stimulus, reduces protein C activation by arterial and venous endothelial cells. Blood. 1990;75:895–901.
Durand P, Lussier-Cacan S, Blache D. Acute methionine load-induced hyperhomocysteinemia enhances platelet aggregation, thromboxane biosynthesis, and macrophage-derived tissue factor activity in rats. FASEB J. 1997;11:1157–1168.
Starkebaum G, Harlan JM. Endothelial cell injury due to copper-catalyzed hydrogen peroxide generation from homocysteine. J Clin Invest. 1986;77:1370–1376.
Toschi V, Gallo R, Lettino M, Fallon JT, Gertz SD, Fernandez-Ortiz A, Chesebro JH, Badimon L, Nemerson Y, Fuster V, Badimon JJ. Tissue factor modulates the thrombogenicity of human atherosclerotic plaques. Circulation. 1997;95:594–599.
WHO. Report No. Eur 8201 (5). 1971. Working group on the establishment of ischaemic heart disease registers: report of the fifth working group.
Philippou H, Adami A, Amersey RA, Stubbs PJ, Lane DA. A novel specific immunoassay for plasma two-chain factor VIIa: investigation of FVIIa levels in normal individuals and in patients with acute coronary syndromes. Blood. 1997;89:767–775.
Fiskerstrand T, Refsum H, Kvalheim G, Ueland PM. Homocysteine and other thiols in plasma and urine: automated determination and sample stability. Clin Chem. 1993;39:263–271.
Refsum H, Ueland PM, Svardal AM. Fully automated fluorescence assay for determining total homocysteine in plasma. Clin Chem. 1989;35:1921–1927.
Philippou H, Davidson S, Mole M, Pepper J, Burman J, Lane DA. Two chain factor VIIa generated in the pericardium during surgery with cardiopulmonary bypass: relationship to increased thrombin generation and to heparin concentration. Atheroscler Thromb Vasc Biol. 1998;19:248–254.
Merlini PA, Ardissino D, Oltrona L, Broccolino M, Coppola R, Mannucci PM. Heightened thrombin formation but normal plasma levels of activated factor VII in patients with acute coronary syndromes. Arterioscler Thromb Vasc Biol. 1995;15:1675–1679.
Cattaneo M, Franchi F, Zighetti ML, Martinelli I, Asti D, Mannucci PM. Plasma levels of activated protein C in healthy subjects and patients with previous venous thromboembolism: relationships with plasma homocysteine levels. Arterioscler Thromb Vasc Biol. 1998;18:1371–1375.
Fryer RH, Wilson BD, Gubler DB, Fitzgerald LA, Rodgers GM. Homocysteine, a risk factor for premature vascular disease and thrombosis, induces tissue factor activity in endothelial cells. Arterioscler Thromb. 1993;13:1327–1333.