Fibrinolytic Variables and Cardiovascular Prognosisin Patients With Stable Angina Pectoris Treated With Verapamil or Metoprolol
Results From the Angina Prognosis Study in Stockholm
Background Disturbed fibrinolytic function may influence the progression of coronary atherosclerosis and contribute to thrombotic cardiovascular (CV) events.
Methods and Results In the Angina Prognosis Study In Stockholm (APSIS), patients with stable angina pectoris were studied prospectively during double-blind treatment with metoprolol or verapamil. Various measures of fibrinolytic function were studied in 631 (of 809) patients. During a median follow-up time of 3.2 years (2132 patient-years), 32 patients suffered a CV death, 21 had a nonfatal myocardial infarction (MI), and 77 underwent revascularization. Plasma levels of tissue plasminogen activator (TPA) activity and antigen (ag), plasminogen activator inhibitor (PAI-1) activity at rest, and TPA responses to exercise were determined at baseline and after 1 month's treatment and were related to subsequent fatal and nonfatal CV events. Univariate Cox regression analysis revealed that elevated levels of TPA-ag at rest (P<.05), high PAI-1 activity (P<.05), and low TPA-ag responses to exercise (P<.05) were associated with increased risk of subsequent CV death. After adjustment for baseline risk factors, TPA-ag independently predicted CV death or MI. In addition, PAI-1 activity independently predicted CV death or MI in male patients. Verapamil treatment was associated with a 10% decrease of TPA-ag levels and metoprolol treatment with a 2% increase (P<.001 for treatment difference).
Conclusions Plasma TPA-ag levels at rest, and among male patients PAI-1 activity as well, independently predict subsequent CV death or MI in patients with stable angina pectoris. Impaired fibrinolytic reactivity to exercise is a novel factor related to CV prognosis. Effects of verapamil or metoprolol treatment on fibrinolytic function did not importantly influence CV prognosis.
The fibrinolytic system is of considerable interest in coronary artery disease as part of the natural defense against thrombosis as well as a putative factor in the atherosclerotic process.1 Impaired fibrinolysis is considered to be a risk factor for coronary artery disease and MI.1 Disturbances of fibrinolysis have been found in cross-sectional studies of patients with MI2 3 and angina pectoris,4 5 6 7 8 but few studies have examined the prognostic impact of fibrinolytic variables prospectively. High PAI-1 level was a risk factor for recurrent MI in young post-MI patients,9 and studies on patients with stable angina pectoris have shown that high TPA-ag levels predict mortality or MI.10 11 However, possible influences of drug treatment are rarely examined in epidemiological studies of risk factors, and there are indications that calcium antagonism and β-blockade may have differential effects on fibrinolysis.12 Thus, treatment may be of importance for the prognostic impact of fibrinolytic parameters. Indeed it has been speculated that improved fibrinolysis may be a beneficial mechanism during β-blocker treatment of coronary heart disease patients.13
APSIS is a prospective trial evaluating the influence of treatment with metoprolol or verapamil on long-term prognosis in stable angina pectoris.14 The two drugs were found to have similar effects on CV end points,14 but treatment effects may differ in subgroups of patients.15 We have previously shown that these patients had a reduced fibrinolytic capacity at rest and an attenuated fibrinolytic response to exercise compared with matched healthy control subjects,7 but the prognostic importance of fibrinolysis was not analyzed.
The present report evaluates the associations between fibrinolytic function and CV events during follow-up of the APSIS study, primarily on the basis of baseline measurements. Attention was paid not only to fibrinolytic capacity at rest but also to TPA responses to exercise. We also investigated the effects of metoprolol and verapamil on fibrinolytic parameters to evaluate the possible impact of treatment on the prognostic information afforded by the fibrinolytic variables.
The APSIS study is a prospective, randomized, single-center trial involving double-blind treatment with either metoprolol or verapamil, performed at Danderyd Hospital in Stockholm. The design and main treatment results of the study have been described in detail elsewhere.14 Informed consent was given by all subjects before entering the study, which was approved by the Ethics Committee of the Karolinska Institute.
A total of 809 patients (248 women) with a typical history of angina pectoris (effort-induced, angina at rest, or mixed angina), <70 years of age, were included in the APSIS study. The present substudy concerns 631 patients (173 women) in whom fibrinolytic variables (both TPA and PAI-1) are available. The median follow-up time was 3.2 years, comprising 2132 patient-years on an intention-to-treat basis. The diagnosis was based on history and physical examinations. Thorough examinations were performed when indicated to exclude noncardiac causes of chest pain. Exclusion criteria were previous MI within 3 years, unstable angina or severe angina with an anticipated need for revascularization within 1 month, CABG or PTCA within the last year, cardiac failure despite optimal treatment, hypotension (systolic blood pressure <100mm Hg), obstructive pulmonary disease, significant valvular disease, or obvious risks for poor compliance. Baseline findings in the population have been described previously.7 14
The patients were randomized to double-blind treatment with either metoprolol (target dose, 200 mg once daily) or verapamil (target dose, 240 mg BID). The initial dose was half of the target dose, and treatment was titrated to full dose after 2 weeks unless adverse effects appeared. Half of the patients were receiving the full dose in both treatment groups at the end of study, and the remaining were on the half dose. To avoid rebound phenomena or severe deterioration, patients previously treated with β-blockers or calcium antagonists were switched to minimal doses of the study drug (25 to 50 mg/d metoprolol or 40 to 80 mg/d verapamil, depending on which type of drug had been used) during a 2-week run-in period before the baseline examinations. Treatment with long-acting nitrates, acetylsalicylic acid, ACE inhibitors, and lipid-lowering drugs were allowed in the study.
Follow-up and Definitions of End Points
The patients returned for a visit after 1 month of treatment, at which all baseline examinations were repeated. The patients were then followed with regular visits at 6-month intervals. The follow-up time varied between 6 and 75 months. Primary end points for follow-up were death and nonfatal cardiovascular events defined as acute MI, incapacitating or unstable angina, cerebrovascular events (including transitory ischemic events), or peripheral vascular events (threatening or overt gangrene or surgery for aortic aneurysm). The criteria for MI were a typical clinical picture, a significant rise in cardiac enzymes, and/or development of Q waves on the ECG. Also, patients who developed a new significant Q wave without hospitalization were classified as having had an MI. Incapacitating angina was defined as symptoms that were insufficiently relieved by optimal doses of the study medication, complemented with long-acting nitrates, and severe enough to compromise ordinary life and indicate a revascularization procedure after due investigation by angiography. In the present report we limit the analysis to relationships between fibrinolytic and CV death, nonfatal MI, or revascularization.
Venous blood sampling for measurements of fibrinolytic variables was performed between 8 and 11 am after 15 minutes of rest in the supine position and was repeated immediately after exercise testing as described below. The levels of TPA-ag were determined on citrated samples by enzyme immunoassay (Tintelise, Biopool AB). Venous plasma samples for determinations of TPA activity in blood were collected in commercially available tubes containing acidified citrate (Stabilyte, Biopool) and analyzed essentially as described by Chmielewska and Wiman.16 PAI-1 and TPA activities were assessed by commercially available kits (Spectrolyse PL and Spectrolyse fibrin, respectively, Biopool AB). Assessments of TPA-ag levels were performed in 631 patients and of TPA activity in a subgroup of 336 patients. PAI-1 activity was assessed in 642 patients. The main reason for not obtaining TPA-ag and PAI-1 activity measurements in all 809 patients was the breakdown of one of the −80°C freezers. Assessments of TPA activity were, however, incorporated into the study at a later stage. Assessments were performed at baseline and repeated after 1 month's treatment with the study drugs to evaluate the short-term effects on fibrinolysis.
Symptom-limited exercise tests were carried out at baseline and after 1 month on a computer-assisted bicycle ergometer (Siemens Elema AB), starting at a workload of 30 W and with increments of 10 W/min. The perceived levels of chest pain and exertion during exercise were estimated according to 10- and 20-grade Borg scales, respectively.
Variables in contingency tables were compared by the χ2 test or, in the case of small expected frequencies, by Fisher's exact test. Within- and between-group comparisons were made by ANOVA. Coefficients of skewness and kurtosis were used to test deviations from a normal distribution. If the underlying normality assumption was violated, logarithmic transformation was performed before statistical analysis. Statistical comparisons of continuous variables in different groups were made by independent two-tailed t tests or Mann-Whitney U tests. Relationships between continuous variables were tested by calculation of Pearson correlation coefficients. In addition, descriptive statistics and graphical methods were used to characterize the data. Analyses were carried out with the use of the SAS version 6.0817 or Statistica, statistical package version 5.0 (StatSoft). A value of P<.05 was considered significant. Means and 95% confidence limits for the means are given unless otherwise stated.
To investigate associations between fibrinolytic variables and coronary events, univariate Cox regression analyses were performed as a first step. Follow-up times until CV death, nonfatal MI, or the end of the study were used. Patients were, however, censored at the actual dates of revascularization, since the proportional risk changes considerably by the procedure. In a second step, fibrinolytic variables that showed some relationship to coronary events (P<.1) were further evaluated with adjustments for factors known (and verified in the present database) to influence prognosis. These factors were a history of previous MI, diabetes mellitus, hypertension, and smoking habits. Analyses were performed both in the entire patient cohort and in men separately. The event rate among women was very low, thus not allowing such a subgroup analysis.
During follow-up, 39 patients in this cohort died and 153 suffered nonfatal CV events, as specified in Table 1⇓. Seven patients who died of malignancies, 29 with aggravating angina but without revascularization, 18 with cerebrovascular events, 5 with peripheral vascular disease, and 3 with unstable angina pectoris were excluded from the present analyses regarding CV prognosis. Fibrinolytic variables and disease severity were compared in the following groups: CV death, CV death or nonfatal MI, patients who were revascularized, and event-free patients. Patient characteristics known to influence risk and results concerning fibrinolytic variables at baseline for these groups are shown in Table 2⇓.
Short-term Effects of Verapamil and Metoprolol
Resting levels of TPA-ag did not differ between the treatment groups at baseline, but the TPA-ag responses to exercise were greater in the metoprolol group compared with the verapamil group (4.5 [4.0 to 5.0] versus 3.7 [3.3 to 4.1] ng/mL; P<.001). In patients treated with verapamil, TPA-ag levels at rest decreased by 10%, whereas a 2% increase was observed in the metoprolol group (P<.001 for treatment difference) (Table 3⇓ and Fig 1⇓). None of the other fibrinolytic variables were significantly affected by treatment. TPA activity at rest and responses to exercise decreased in both treatment groups, although there was a trend toward a greater decrease among metoprolol-treated patients (Table 3⇓).
Since a majority of the patients were on minimal doses of verapamil or metoprolol during the run-in period, we also investigated the short-term effects of treatment in the subgroup of patients without such run-in medication (n=242). The previously untreated patients had somewhat milder angina pectoris.18 They had similar levels of TPA-ag at rest as those on run-in medication but greater TPA-ag and TPA activity responses to exercise and lower PAI-1 levels. Resting TPA-ag levels decreased with verapamil treatment compared with metoprolol treatment also among untreated patients (P<.05 for treatment difference). No other fibrinolytic variable was significantly affected by short-term treatment with either drug in the subgroup without run-in treatment (Table 3⇑).
Fibrinolytic Variables in Patients With Different Outcome
Patients suffering cardiac death had higher levels of TPA-ag at rest (P<.05) and PAI-1 (P<.05) than event-free patients (Table 2⇑). The TPA-ag responses to exercise were lower among patients suffering cardiac death (P<.05), but TPA activity levels at rest and responses to exercise were similar. When patients with CV death or nonfatal MI were combined, differences remained for TPA-ag at rest (P<.05) and TPA-ag responses to exercise (P<.05). In addition, TPA activity at rest was lower (P<.05) (Table 2⇑). In patients who underwent revascularization, only a low TPA-ag response to exercise (P<.05) compared with the event-free was found. Thus, there was a decreasing gradient for the severity of fibrinolytic disturbances from subgroups with CV death to CV death or nonfatal MI to patients who were revascularized to event-free patients. The levels of TPA-ag at rest were correlated to PAI-1 activity (r=.72, P<.05) and inversely correlated to TPA activity at rest (r=−.64, P<.05) among patients with a CV death or nonfatal MI. The corresponding relationships for TPA-ag levels at rest were weaker among patients who had any CV event (r=.30, P<.05 for PAI-1 activity and r=−.11, P=NS for TPA activity at rest). Among event-free patients, TPA-ag at rest correlated to PAI-1 activity (r=.37, P<.05) and to TPA activity (r=−.18, P<.05).
Table 4⇓ shows fibrinolytic variables during treatment in the above-mentioned subgroups. These results differ somewhat from those based on baseline assessments (Table 2⇑). CV death was no longer associated with significantly elevated PAI-1 levels or significantly attenuated TPA-ag responses to exercise, whereas TPA activity responses to exercise were attenuated among patients with CV death alone or combined with MI.
Fig 2⇓ shows fibrinolytic variables at baseline, divided into tertiles, in relation to the incidence of CV death and the combination of fatal and nonfatal CV events. High TPA-ag levels were associated with a higher incidence of CV death or combined events, as was a low TPA-ag response to exercise. For PAI-1 activity there was a nonlinear trend for CV death but no significant gradient for combined CV events (Fig 2⇓). The incidence of combined CV events was also related to TPA-ag levels at rest after 1 month's treatment (73 versus 42 events in the highest versus lowest tertile; P<.001), but the incidence of CV deaths did not differ significantly (13 versus 5, NS). More combined events were observed among those in the lower tertile of TPA-ag responses to exercise (79 versus 42, P<.01). The three tertiles of PAI-1 activity, TPA-activity at rest, or TPA activity responses to exercise did not differ with regard to the incidence of CV events when assessed during treatment.
Prognostic Evaluation of Fibrinolytic Variables: Univariate Analyses
A Cox regression model was used to assess the impact of fibrinolytic parameters at baseline on survival. Kaplan-Meier plots (Fig 3⇓) showed that TPA-ag levels at rest above median (P<.05), TPA-ag responses to exercise below median (P<.05), and PAI-1 levels above median (P<.05) were all related to a higher CV mortality in univariate analyses. The curves began to separate early during follow-up for both TPA-ag at rest and the TPA response to exercise, whereas the separation started after 1 year of follow-up for PAI-1 activity. Corresponding analyses with on-treatment data showed somewhat different results; only TPA-ag levels above median were related to poor survival (P<.05).
Prognostic Evaluation of Fibrinolytic Variables Adjusted for Baseline Factors
The Cox regression model was also used to evaluate the prognostic importance of the fibrinolytic variables after adjustment for known risk factors, ie, a history of previous MI, hypertension, diabetes mellitus, and smoking. Among subgroups with different smoking habits, we found that ex-smokers were those with the best prognosis, followed by nonsmokers. Thus, in adjustments for smoking habits, smoking and nonsmoking were combined as a factor versus ex-smokers. These adjusted analyses revealed that baseline levels of TPA-ag at rest were independently related to the incidence of CV death or nonfatal MI during follow-up (P<.05). Table 5⇓ shows the positive relationship between TPA-ag levels and the relative risk of suffering a CV death or nonfatal MI (P<.05 for trend). Among male patients, both TPA-ag and PAI-1 activity were independently related to prognosis (P<.05). There were too few female patients suffering a CV death or MI to enable a statistical evaluation.
When the prognostic importance of fibrinolytic variables during treatment (ie, 1-month values) was evaluated, only TPA-ag at rest was independently related to the incidence of a CV death or nonfatal MI (P<.05), and PAI-1 levels were no longer predictive of such events in men.
Influence of Verapamil and Metoprolol Treatment on Associations Between Fibrinolysis and Cardiovascular Outcome
Since the two drugs influenced fibrinolysis somewhat differently, patients were also analyzed according to randomized treatment. In the verapamil group, the on-treatment levels of PAI-1 were higher (P<.05) and TPA-ag (P<.05) and TPA-activity responses to exercise were lower (P<.05) among those who suffered a subsequent CV death or a nonfatal MI (unadjusted for baseline factors). However, high TPA-ag at rest (P<.01) and low TPA-ag (P<.05) and activity (P<.01) responses to exercise were associated with a higher incidence of fatal and nonfatal CV events combined (ie, MI, revascularization, angiography without revascularization, cerebrovascular disease, peripheral vascular disease). In the metoprolol group, on the other hand, the only fibrinolytic factor associated with CV death or nonfatal MI was high levels of TPA-ag at rest (P<.001). When we evaluated all fatal and nonfatal CV events combined, TPA-ag responses to exercise were lower (P<.05) than among the event-free in the metoprolol group.
When confined to either of the two treatment groups (verapamil or metoprolol), Cox regression survival analysis could not detect any independent relationship between fibrinolytic variables and prognosis. The probability values for the influence of TPA-ag on prognosis were .49 and .08 in verapamil- and metoprolol-treated patients, respectively. These subgroups were, however, substantially smaller than in the overall analysis.
The present prospective study shows that disturbed fibrinolysis is associated with an increased risk of suffering CV events in patients with stable angina pectoris. Thus, the fibrinolytic alterations we found when comparing these patients with healthy control subjects7 may be of importance for disease progression. Elevated level of TPA-ag was the strongest independent fibrinolytic predictor of poor outcome (CV death or nonfatal MI). Among male patients, PAI-1 also was an independent predictor for CV events. Reduced fibrinolytic reactivity, defined as a low TPA response to exercise, was another interesting and novel factor that predicted a poor cardiovascular prognosis. Verapamil and metoprolol were found to have differential short-term effects on fibrinolysis, but the impact of fibrinolytic variables on prognosis was little influenced by treatment.
Our data support the results of two recently published prospective studies on fibrinolysis in patients with stable angina pectoris.10 11 In the ECAT study,10 TPA-ag, fibrinogen, and von Willebrand factor were found to be independent predictors of subsequent acute coronary syndromes. In the study of Jansson et al,11 TPA-ag independently predicted mortality in patients with angina pectoris and angiographically verified coronary artery stenoses. Furthermore, euglobulin clot lysis time after venous occlusion has been shown to independently predict sudden death in patients with stable angina pectoris.19 These prospective studies did not involve randomized drug treatment, as was the case in the APSIS trial. With regard to healthy individuals, fibrinolytic activity as measured by dilute clot lysis time was a strong independent predictor of the risk of developing ischemic heart disease in the Northwick Park Heart Study,20 and TPA-ag levels were associated with the risk of MI in the Physicians Health Study.21 Thus, our results concerning TPA-ag levels are consistent with findings from several previous studies.
We found an independent relationship between PAI-1 activity and the risk of CV death or nonfatal MI among male angina patients but not in the entire material, in congruence with previous findings in young male post-MI patients.9 Elevated PAI-1 activity levels were also found in high-risk patients suffering an atherothrombotic event in a small case-control analysis of the PLAT study.22 The positive correlations between TPA-ag and PAI-1 activity and the inverse correlation to TPA activity suggest that TPA-ag levels to a large extent reflect circulating TPA-PAI-1 complexes, as has been suggested by others.10 11 The greater variability of PAI-1 levels compared with TPA levels (eg, as the result of diurnal variability23 ) probably explains why TPA was a more robust predictor than PAI-1 in the present study.
Fibrinolytic reactivity to exercise has not been investigated in the previous studies. In the APSIS study, an attenuated TPA response to exercise indicated an increased risk for subsequent CV events. Several mechanisms may be responsible for the attenuation of fibrinolytic responses to exercise, as discussed elsewhere.7 Maximal workload was, of course, one major determinant, but differential workloads did not fully explain the differences in TPA responsiveness between patients and control subjects. Endothelial dysfunction with defective TPA release and elevated PAI-1 levels are other plausible explanations. Evaluation of TPA-ag levels both before and after a diagnostic exercise test may provide additional prognostic information. It is tempting to speculate that impaired fibrinolytic reactivity might be an important feature in patients with a dysfunctional fibrinolytic system and that this may contribute to the triggering of an MI during physical exercise.24 25
Verapamil and metoprolol had different short-term effects on fibrinolysis, with a lowering of TPA-ag levels at rest among verapamil-treated patients. The changes were relatively small and did not affect prognosis. The baseline run-in medication received by many patients did not invalidate our conclusion regarding the short-term effects of treatment with verapamil or metoprolol, as suggested by a subgroup analysis of untreated patients.
Improved fibrinolysis may be a mechanism contributing to beneficial effects of β-blockade in coronary heart disease.13 However, conflicting results have been presented regarding the effects of metoprolol on fibrinolysis in smaller studies, as PAI-1 activity has been reported to be either reduced26 or unchanged.27 28 29 Propranolol has been shown to decrease TPA activity levels in healthy volunteers,12 whereas metoprolol did not influence TPA-ag or activity levels in a placebo-controlled study in volunteers.28 The present study clearly shows that metoprolol does not influence the prognosis of angina patients via improved fibrinolysis. Effects of verapamil on fibrinolytic parameters in patients with coronary arterydisease have not been reported previously, to our knowledge. Why verapamil but not metoprolol lowered TPA-ag levels in our study is unknown. A speculation might be different effects on endothelial function.
Our findings support and extend the hypothesis that disturbed fibrinolytic function is related to cardiovascular prognosis in patients with stable angina pectoris. Thus, elevated TPA-ag levels at rest, and in male patients PAI-1 activity as well, were independent predictors of CV death or nonfatal MI. In addition, impaired fibrinolytic reactivity during exercise was related to poor prognosis. It is still not known whether disturbed fibrinolysis is a marker for the underlying atherosclerotic disease or if there is a causal relationship between fibrinolytic function and cardiovascular events. Effects of drug treatment should be considered when evaluating the prognostic importance of risk factors in multifactorial diseases such as coronary artery disease and arteriosclerosis. However, our analysis suggests that the effects of metoprolol or verapamil treatment on fibrinolytic function are not likely to be of major importance for prognosis in coronary artery disease.
Selected Abbreviations and Acronyms
|APSIS||=||Angina Prognosis Study in Stockholm|
|CABG||=||coronary artery bypass graft|
|PAI-1||=||plasminogen activator inhibitor-1|
|PTCA||=||percutaneous transluminal coronary angioplasty|
|TPA||=||tissue plasminogen activator|
This study was supported by grants from the Swedish Heart and Lung Foundation, the Swedish Medical Research Council (5193, 5930), King Gustav V and Queen Victoria's Foundation, the Swedish Society of Medicine, Stockholm, Sweden, Knoll AG, Ludwigshafen, Germany, and Astra Hässle, Gothenburg, Sweden. The help from and overall support of our research nurses Inger Bergbom, Ewa Billing, Ann-Mari Ekman, and Britt Rydén and technicians Maj-Christina Johansson, Maud Daleskog, Christina Perneby, Wiveka Ring-Persson, Ann-Cathrine Kjerr, and Margareta Ring are gratefully acknowledged. We also thank Per Näsman, at the Royal Institute of Technology, and Ulf Brodin, at the Department of Medical Information and Educational Development at the Karolinska Institute, for data management and statistical services and Margret Lundström for invaluable help with data-entering management.
- Received September 17, 1996.
- Revision received December 9, 1996.
- Accepted December 14, 1996.
- Copyright © 1997 by American Heart Association
Hamsten A, Eriksson P. Fibrinolysis and atherosclerosis: an update. Fibrinolysis. 1994;8(suppl 1):253-262.
Olofsson BO, Dahlén G, Nilsson TK. Evidence for increased levels of plasminogen activator inhibitor and tissue plasminogen activator in plasma of patients with angiographically verified coronary artery disease. Eur Heart J. 1989;10:77-82.
Aznar J, Estellés A, Tormo G, Sapena P, Tormo V, Blanch S, Espana F. Plasminogen activator inhibitor activity and other fibrinolytic variables in patients with coronary artery disease. Br Heart J. 1988;59:535-541.
Held C, Hjemdahl P, Rehnqvist N, Wallén NH, Forslund L, Björkander I, Angelin B, Wiman B. Hemostatic markers, inflammatory parameters and lipids in male and female patients in the Angina Prognosis Study In Stockholm (APSIS): a comparison with healthy controls. J Intern Med. 1997;241:59-69.
ECAT Angina Pectoris Study Group. ECAT angina pectoris study: baseline associations of haemostatic factors with extent of coronary arteriosclerosis and other coronary risk factors in 3000 patients with angina pectoris undergoing coronary angiography. Eur Heart J. 1993;14:8-17.
Thompson SG, Kienast J, Pyke SDM, Haverkate F, van de Loo JCW, for the European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group. Hemostatic factors and the risk of myocardial infarction or sudden death in patients with angina pectoris. N Engl J Med. 1995;332:635-641.
Jansson JH, Olofsson BO, 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.
Winther K, Gleerup G, Hedner T. Platelet function and fibrinolytic activity in hypertension: differential effects of calcium antagonists and β-adrenergic receptor blockers. J Cardiovasc Pharmacol. 1991;18(suppl 9):S41-S44.
Teger-Nilsson AC, Larsson PT, Hjemdahl P, Olsson G. Fibrinogen and plasminogen activator inhibitor-1 levels in hypertension and coronary heart disease: potential effects of β-blockade. Circulation. 1991;84(suppl VI):VI-72-VI-77.
Rehnqvist N, Hjemdahl P, Billing E, Björkander I, Eriksson SV, Forslund L, Held C, Näsman P, Wallén NH. Effects of metoprolol vs verapamil in patients with stable angina pectoris: the Angina Prognosis Study in Stockholm (APSIS). Eur Heart J. 1996;17:76-81.
Chmielewska J, Wiman B. Determination of tissue plasminogen activator and its ‘fast’ inhibitor in plasma. Clin Chem. 1986;32/3:482-485.
SAS Institute Inc. SAS/STAT Version 6. 4th ed. Cary, NC: SAS Institute; 1989.
Billing E, Hjemdahl P, Rehnqvist N. Psychosocial variables in female versus male patients with stable angina pectoris and matched healthy controls. Eur Heart J. In press.
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, for the PLAT Group. 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.
Mittleman MA, Maclure M, Tofler GH, Sherwood JB, Goldberg RJ, Muller JE, for the Determinants of Myocardial Infarction Onset Study Investigators. Triggering of acute myocardial infarction by heavy physical exertion: protection against triggering by regular exertion. N Engl J Med. 1993;329:1677-1683.
Wright RA, Perrie AM, Stenhouse F, Alberti KGMM, Riemersma RA, MacGregor IR, Boon NA. The long-term effects of metoprolol and epanolol on tissue-type plasminogen activator and plasminogen activator inhibitor 1 in patients with ischaemic heart disease. Eur J Clin Pharmacol. 1994;46:279-282.