Fibrin d-Dimer and β-Thromboglobulin as Markers of Thrombogenesis and Platelet Activation in Atrial Fibrillation
Effects of Introducing Ultra–Low-Dose Warfarin and Aspirin
Background Previous studies have demonstrated increased markers of thrombogenesis in patients with atrial fibrillation (AF), suggesting the presence of a hypercoagulable or prothrombotic state. The objective of this study was to determine the effects of introducing ultra–low-dose warfarin (1 mg), conventional warfarin, and aspirin (300 mg) therapy on thrombogenesis and platelet activation in AF.
Methods and Results We measured sequential changes in plasma fibrin d-dimer (an index of thrombogenesis) and β-thromboglobulin (β-TG, a measure of platelet activation) in 51 patients with chronic AF before and at 2 and 6 weeks after randomization to either 1 mg warfarin or 300 mg aspirin (phase 1). Then all patients were started on conventional warfarin therapy (phase 2) with samples taken 2 and 6 weeks later. Pretreatment results were compared with those from 26 healthy control subjects in sinus rhythm. Baseline (pretreatment) β-TG and d-dimer levels in patients with AF were elevated compared with those of control subjects (P<.001). In phase 1, there were no significant changes in median levels of fibrin d-dimer or β-TG, despite warfarin 1 mg or aspirin 300 mg. With standard warfarin therapy (phase 2), there was a reduction in median β-TG at 6 weeks (P=.025) and a sequential reduction in median d-dimer levels at 2 (P=.001) and 6 (P<.001) weeks compared with baseline levels.
Conclusions Patients with AF have increased intravascular thrombogenesis and platelet activation compared with patients in sinus rhythm. Introduction of ultra–low-dose warfarin (1 mg) or aspirin 300 mg does not significantly alter these markers, although conventional warfarin therapy reduces β-TG and fibrin d-dimer levels. This is consistent with the beneficial effect of full-dose warfarin in preventing stroke and thromboembolism in AF and suggests that ultra–low-dose warfarin and aspirin may not exert similar beneficial effects.
Chronic AF is a common arrhythmia associated with a substantial risk of stroke and thromboembolism.1 Recent studies have shown that compared with persons in sinus rhythm, patients with AF have abnormalities in rheology and hemostatic factors that contribute to a prothrombotic or hypercoagulable state that may contribute to this risk of stroke.2 3 4
Many potential prothrombotic factors exist. One such plasma factor, the fibrin d-dimer fragment, reflects intravascular turnover of fibrin by identifying the presence of cross-linked fibrin degradation products without interference from fibrinogen or non–cross-linked fibrin.5 6 Elevated levels of fibrin degradation products are found in conditions associated with intravascular (and sometimes extravascular) activation of the coagulation system, including those with a predisposition to thrombosis.5 6 In cross-sectional studies, increased plasma levels of fibrin d-dimer have been found in patients with AF, regardless of whether coexisting structural heart disease is present, compared with subjects in sinus rhythm.2 4 In addition to fibrin formation, platelet activation and aggregation have an important role in the initiation (and maintenance) of the process of thrombogenesis. Abnormalities of plasma β-TG, a marker of in vivo platelet activity, have been found in patients with cardiovascular disease.7 Limited information, however, is available on this marker of platelet activation in patients with AF and the effects of introducing antithrombotic therapy.
Large-scale clinical studies have demonstrated that anticoagulation with warfarin significantly reduces the risk of stroke in patients with AF by two thirds.1 In cross-sectional studies, fibrin d-dimer levels also are reduced by about two thirds in patients with AF who are established on warfarin therapy.4 In a small pilot study, we previously showed that the introduction of anticoagulation with warfarin (aiming for a target INR of the prothrombin time of 2.0 to 3.0) resulted in normalization of plasma fibrin d-dimer levels.4 8 The effects of aspirin prophylaxis, however, are less certain, although clinical studies suggest that aspirin may be useful in certain lower-risk subgroups.9 10 Ultra–low-dose warfarin (1 mg) also is currently being evaluated as thromboprophylaxis in patients with AF; compared with conventional warfarin, ultra–low-dose warfarin minimizes the risks of bleeding and avoids regular monitoring of the INR.
Because no information is available on the effects of introducing aspirin or ultra–low-dose warfarin on markers of in vivo fibrin turnover and platelet activation (as indicated by plasma fibrin d-dimer or β-TG levels, respectively), we measured sequential changes in these markers in patients with chronic AF in whom aspirin or ultra–low-dose warfarin was initiated in a randomized study followed by conventional warfarin prophylaxis (INR, 1.5 to 2.5).
We conducted a prospective longitudinal study of consecutive male and female patients with chronic AF who were not receiving any antithrombotic therapy and were referred for anticoagulation. Chronic AF was defined as AF documented by ECG on two occasions on at least two visits to the outpatient clinic 6 weeks apart.
Patients were studied before treatment and 2 and 6 weeks after randomization to either ultra–low-dose warfarin (1 mg/d) or aspirin (300 mg/d) (phase 1). All patients then were started on standard warfarin therapy; the goal was to obtain a target INR of 1.5 to 2.5 (phase 2) with blood samples taken after an additional 2 and 6 weeks (Fig 1⇓). The INR was measured at each visit as an index of anticoagulation intensity. Pretreatment values of fibrin d-dimer and β-TG in patients with AF were compared with those derived from age- and sex-matched healthy subjects in sinus rhythm attending a nonacute preoperative ophthalmologic clinic for evaluation of minor eye conditions (eg, cataracts).
Excluded were those patients with significant systemic illnesses such as renal failure, liver impairment (defined as abnormal liver function tests with aspartate transaminase or alanine transaminase levels more than twice the upper limit of normal), chronic infections, collagen disease, or neoplastic disease. We also excluded patients with recent (within 2 months) myocardial infarction, unstable angina, or stroke to avoid the effect of any acute phase response in hemostatic tests; we also excluded those with limited venous access, patients requiring transfusion, or those who were actively bleeding.
Fibrin d-Dimer Measurement
Fibrin d-dimer was measured before treatment and 2 and 6 weeks after treatment. Blood samples were drawn from the antecubital vein by careful venipuncture, anticoagulated with trisodium citrate (0.11 mol/L, 9:1 vol/vol), and centrifuged. The platelet-free plasma was immediately separated, frozen at −70°C, and assayed for plasma d-dimer antigen levels with a commercially available ELISA method (AGEN). The reproducibility of this method allowed a coefficient of variation of <5%.
β-TG was measured before and 6 weeks after treatment. Blood for β-TG assay was collected into tubes containing EDTA and theophylline. After cooling in a crushed ice-water mix, platelet-poor plasma was prepared within 1 hour by centrifugation at 1800g for 30 minutes in a precooled centrifuge at 2°C. Midlayer plasma was collected and stored at −20°C until assay. Plasma β-TG was assayed in duplicate with a radioimmunoassay (Amersham International). Intra-assay and interassay coefficients of variation were <7.5% and <10%, respectively.
The INR was measured with Manchester Thromboplastin Reagent (ISI 1.08) and tested in an Amelung KC10 coagulometer (Brownes). The platelet count was obtained with a H2Tc auto-analyzer (Bayer plc), and the plasma viscosity was measured with a Coulter viscometer II (Coulter Electronics Ltd).
Values of fibrin d-dimer, β-TG, and INR were expressed as median±interquartile range; platelet count and plasma viscosity values were expressed as mean±SD. Statistical comparisons between patients and control subjects were performed by use of an unpaired t test for normal distributions and the Mann-Whitney two-sample rank test for nonparametric distributions. The 95% CIs were calculated for the point estimate of the difference between medians.
Paired comparisons of sequential changes in thrombogenic markers, platelets, and plasma viscosity were performed by use of a two-sample paired Wilcoxon test or a paired t test as appropriate. A value of P<.05 was considered statistically significant. All statistical calculations were performed on a microcomputer with a commercially available statistical package (MINITAB v8, Minitab Inc).
We studied 51 patients (34 men, 17 women; mean age, 70.4±7.4 years) with chronic AF. Pretreatment results were compared with values from 26 healthy control subjects in sinus rhythm (16 men, 10 women; mean age, 72.7±9.9 years).
Baseline (pretreatment) β-TG and d-dimer levels in patients with AF were significantly elevated compared with those of control subjects in sinus rhythm (β-TG: median difference, 79 ng/mL, Mann-Whitney test, P<.001; d-dimer: median difference, 106 ng/mL, Mann-Whitney test, P=.0001; Table 1⇓ and Fig 2⇓). In addition, mean plasma viscosity was significantly higher in patients with AF compared with subjects in sinus rhythm (unpaired t test, P=.0002). There was no significant difference in mean platelet count between the two groups (unpaired t test, P=.30; Table 1⇓).
Of the 51 patients, 25 were randomized to warfarin 1 mg/d and 26 to aspirin 300 mg/d (phase 1). There were no significant differences in age (unpaired t test, P=NS) and in the prevalence of underlying etiologic factors or smoking status between the two groups (χ2 test, P=NS; Table 2⇓). There also were no statistically significant differences in baseline levels of fibrin d-dimer, β-TG, plasma viscosity, and platelet count between patients randomized to aspirin or warfarin 1 mg (P=NS). In phase 1, there were no significant changes in median levels of fibrin d-dimer or β-TG after the introduction of ultra–low-dose warfarin (1 mg/d; Table 3⇓). After the introduction of aspirin 300 mg/d, there was no significant change in β-TG (169 versus 179 ng/mL, paired Wilcoxon test, P=NS) or fibrin d-dimer at 6 weeks (186 versus 211 ng/mL, P=NS) compared with baseline levels. However, there was a small decrease in median fibrin d-dimer levels at 2 weeks, which was statistically significant (186 versus 177 ng/mL, P=.02; Table 4⇓).
Because of patient withdrawals, only 43 patients entered phase 2 of the study. Of the 8 patients withdrawn from the study, 6 withdrew consent and 2 moved from Birmingham, England. With conventional warfarin therapy (INR, 1.5 to 2.5; phase 2), there was a reduction in median β-TG at 6 weeks (209 versus 173 ng/mL, P=.025) and a sequential reduction in median d-dimer levels that was significant at 2 weeks (220 versus 139 ng/mL, P=.001) and 6 weeks (220 versus 110 ng/mL, P<.001) compared with baseline levels (Table 5⇓). The mean warfarin doses when patients attended the 2- and 6-week visits in phase 2 were 3.24±1.08 and 3.23±1.41 mg, respectively. After stabilization on warfarin during phase 2, median INR increased correspondingly (Table 5⇓, Fig 3⇓).
There was no significant effect of warfarin 1 mg, aspirin 300 mg, or full-dose warfarin on plasma viscosity or platelet count (paired t test, P=NS; Tables 3 through 5⇑⇑⇑). There were no thromboembolic events or adverse events during the study period.
We have demonstrated high plasma levels of fibrin d-dimer, β-TG, and plasma viscosity in patients with AF. The presence of high fibrin d-dimer suggests ongoing fibrin formation and degradation in patients with chronic AF, and the high β-TG indicates increased platelet activation. These observations are consistent with the high thrombogenic state associated with this arrhythmia, which may contribute to the risks of stroke and thromboembolism.5 Our findings of high fibrin d-dimer in AF are consistent with previous reports2 4 and confirm high intravascular fibrin turnover in association with this arrhythmia. The abnormalities in fibrin d-dimer in patients with AF have been found to be independent of underlying origin or structural heart disease.2 4 In addition, high fibrin d-dimer levels are a strong predictor of both arterial disease progression and cardiovascular risk in normal subjects11 12 and patients with peripheral vascular disease.13 However, the value of fibrin d-dimer as a predictor of stroke and other types of cardiac thromboembolism in patients with AF merits exploration in large prospective studies.5
Anticoagulation with warfarin is effective thromboprophylaxis in patients with AF.1 However, there are two problems with full-dose conventional warfarin therapy, namely inconvenience and safety. This may account in part for the reluctance in clinical practice to prescribe anticoagulants for all patients with AF.14 15 Thus, efforts are being directed to find better, safer, and more convenient methods of providing thromboprophylaxis for such patients. Aspirin and simple, ultra–low-dose or low-intensity warfarin regimens are being evaluated as alternatives to conventional full-dose anticoagulation with warfarin, which carries a risk of hemorrhage and requires regular blood checks to maintain anticoagulation intensity at an average INR of between 2.0 and 3.0.10 In the five primary prevention trials, for example, the target INR ranged between 1.5 and 4.5.10
The effects of aspirin in preventing strokes in nonvalvular AF are inconsistent. The AFASAK study demonstrated no benefit from the use of aspirin at a dose of 75 mg/d.16 Although not one of the arms of the study, some patients in the placebo group of the BAATAF study took aspirin, with no beneficial effects seen on stroke rates.17 By contrast, the SPAF-I study suggested that aspirin 325 mg may be useful in patients <75 years of age, although aspirin failed to prevent severe strokes and reduce overall mortality.18 The recent SPAF-II study confirmed that aspirin 325 mg was useful in certain subgroups of patients, especially those <75 years of age with no cardiac risk factors.9 The pooled analysis by the Atrial Fibrillation Investigators10 reported that aspirin may be sufficient thromboprophylaxis for moderate-risk (annual stroke risk, 4%) patients with nonrheumatic AF.
In a previous cross-sectional study, patients with AF who were established on aspirin had a nonsignificant reduction in plasma fibrin d-dimer levels compared with those who were not on any antithrombotic therapy.4 In the present prospective trial, we also failed to show a significant reduction in plasma fibrin d-dimer or β-TG levels 6 weeks after the prospective introduction of aspirin 300 mg/d, although a small decrease in fibrin d-dimer was seen after 2 weeks of aspirin therapy. The lack of effect of aspirin 300 mg on either fibrin turnover or platelet activation in vivo in phase 1 may have been due to the high thrombogenic or hypercoagulable state in patients with AF and suggests that aspirin is a relatively weak inhibitor of thrombogenesis in such patients. Further trials are necessary to establish its antithrombotic effect.1 The lack of a long-term decrease in fibrin d-dimer levels may in part be a reflection of the older age of our population compared with that of the SPAF-II study and the high proportion of patients with associated risk factors. We also have shown increased platelet activation in patients with AF (as indicated by the high β-TG levels but similar mean platelet count compared with control subjects in sinus rhythm), which is in keeping with the observation that platelet activation and aggregation have an important role in the initiation and maintenance of thrombogenesis. Thus, because platelet activation appears to be significantly increased in this disorder, attention should be directed toward providing thromboprophylaxis in patients with AF through antithrombotic therapies that can reduce both intravascular clotting and platelet activation, possibly with treatment regimens combining both aspirin and warfarin.
Simpler, low-risk regimens of warfarin have been suggested as adequate thromboprophylaxis with minimal chances of bleeding and without the requirement for regular blood checks to monitor anticoagulant control. For example, ultra–low-dose warfarin (1 mg/d) has been used successfully to prevent thrombosis of central venous catheters19 and deep venous thrombosis in patients with malignancies.20 Ongoing clinical trials in AF, such as the AFASAK-II and SPAF-III studies, have included low-intensity warfarin therapy, aspirin-warfarin combinations, and ultra–low-dose warfarin regimens as part of their treatment arms.21 In phase 1 of the present study, we failed to find any statistically significant reduction of fibrin d-dimer or β-TG levels after the introduction of ultra–low-dose warfarin, although there was a trend toward a reduction in these markers at the 6-week follow-up. Our findings suggest that ultra–low-dose (1 mg) warfarin may be insufficient as thromboprophylaxis in patients with AF, and we await with interest the results of large clinical outcome studies investigating this regimen. We have, however, confirmed our previous observation that conventional warfarin normalized the increased fibrin d-dimer levels in patients with AF.4 Furthermore, we have shown for the first time that significant reductions in β-TG occurred after conventional full-dose warfarin therapy (phase 2). This observation may be a reflection of the effective reduction of thrombin formation by warfarin, which is a potent platelet activator. This dual effect of warfarin on two critical components of thrombogenesis (platelet activation and fibrin turnover) may account for its high efficacy in prophylaxis of cardiac thromboembolism in persons with AF.1
We have demonstrated that plasma viscosity is elevated in patients with AF. In large epidemiological studies, plasma viscosity levels have been related to cardiovascular risk.22 Although further studies are required on the factors influencing plasma viscosity and its potential roles in atherosclerosis, thrombosis, and ischemia,23 the finding of high plasma viscosity in patients with AF may be a reflection of the high levels of plasma fibrinogen in this condition.4 24 This observation is in keeping with the associations between plasma fibrinogen, cardiovascular disease, and stroke.24 25 Warfarin and aspirin are not known to influence plasma fibrinogen4 or plasma viscosity levels, which also is shown in this study.
This study is limited by the relatively short follow-up period in phase 1. With the trend (nonsignificant) toward a decrease in β-TG and fibrin d-dimer levels in patients given ultra–low-dose warfarin (1 mg), it is conceivable that with a longer follow-up period, statistically significant decreases in these markers may be demonstrated. In addition, the reduced number of patients entering phase 2 of the study illustrates the problems of regular hospital visits for anticoagulant monitoring; some patients requested to be withdrawn from the study because of the inconvenience of such visits. Finally, we did not relate abnormalities of these markers of thrombogenesis to detailed measurements of left atrial size or ventricular size or function on echocardiography because our previous work with larger numbers of patients with AF did not show any significant relationships.4 Nevertheless, the primary aim of the present study was to demonstrate the effects of introducing treatment with aspirin 300 mg or ultra–low-dose (1 mg) warfarin on these markers rather than to correlate the abnormalities with detailed measurements of cardiac chamber size or ventricular function.
This study shows that patients with AF have not only increased fibrin turnover but also increased platelet activation as reflected by high plasma levels of fibrin d-dimer and β-TG, respectively. Introduction of aspirin (300 mg/d) or ultra–low-dose warfarin (1 mg/d) did not reduce either of these markers, whereas conventional treatment with standard full-dose warfarin (aiming for an INR of 1.5 to 2.5) reduced both increased fibrin turnover and platelet activation. We therefore suggest that measures of platelet activation and fibrin turnover in vivo may be useful in assessments of the antithrombotic potential of prophylactic regimens in persons with AF.
This project was supported by the NHS Research & Development Programme on Cardiovascular Disease and Stroke (ref MC7). Dr Lip is a recipient of the Edith Walsh and Ivy Powell Awards for cardiovascular research from the British Medical Association. We thank M. Hart, C. Cox, and A. Rumley for expert technical assistance.
Selected Abbreviations and Acronyms
|AFASAK||=||Atrial Fibrillation, Aspirin, Antikoagulation study|
|BAATAF||=||Boston Area Anticoagulation Trial for Atrial Fibrillation|
|INR||=||international normalized ratio|
|SPAF||=||Stroke Prevention in Atrial Fibrillation study|
Corrrespondence to Dr G.Y.H. Lip, University Department of Medicine, City Hospital, Birmingham B18 7QH, England, UK.
- Received October 16, 1995.
- Revision received January 23, 1996.
- Accepted January 29, 1996.
- Copyright © 1996 by American Heart Association
Lowe GDO. Antithrombotic treatment and atrial fibrillation. Br Med J. 1992;305:1445-1446.
Gustafsson C, Blomback M, Britton M, Hamsten A, Svensson J. Coagulation factors and the increased risk of stroke in nonvalvular atrial fibrillation. Stroke. 1990;21:47-51.
Lip GYH, Lowe GDO, Rumley A, Dunn FG. Increased markers of thrombogenesis in chronic atrial fibrillation: effects of warfarin therapy. Br Heart J. 1995;73:527-533.
Lip GYH, Rumley A, Dunn FG, Lowe GDO. Plasma fibrin d-dimer levels in patients with atrial fibrillation and left ventricular aneurysms: effects of introducing warfarin therapy on thrombogenesis. J Am Coll Cardiol. 1994;23:227A. Abstract.
Lowe GDO, Rumley A, Yarnell JWG, Sweetnam PM, Thomas HF. Fibrin d-dimer, von Willebrand factor and tissue plasminogen activator antigens are predictors of major ischaemic heart disease: the Caerphilly Study. Blood Coagul Fibrinolysis. 1995;6:156-157.
Ridker PM, Hennekens CH, Cerskus A, Stampfer MJ. Plasma concentrations of cross-linked fibrin degradation product (d-dimer) and the risk of future myocardial infarction among apparently healthy men. Circulation. 1994;90:2236-2240.
Lip GYH, Tean KN, Dunn FG. Treatment of atrial fibrillation in a district general hospital. Br Heart J. 1994;71:92-95.
Bath PMW, Prasad A, Brown MM, MacGregor GA. Survey of use of anticoagulation in atrial fibrillation. Br Med J. 1993;307:1045.
Singer DE, Hughes RA, Gress DR, Sheehan MA, Oertel LB, Maraventano SW, Blewett DR, Rosner B, Kistler JP, for the BAATAF Investigators. The effect of aspirin on the risk of stroke in patients with nonrheumatic atrial fibrillation: the BAATAF study. Am Heart J. 1992;124:1567-1573.
Stroke Prevention in Atrial Fibrillation Investigators. Stroke Prevention in Atrial Fibrillation study: final results. Circulation. 1991;84:527-539.
Bern MM, Lokich JJ, Wallach SR, Bothe A, Benotti PN, Arkin CF, Greco FA, Huberman M, Moore C. Very low doses of warfarin can prevent thrombosis in central venous catheters. Ann Intern Med. 1990;112:423-428.
Major ongoing stroke trials. Stroke. 1994;25:541-545.
Lip GYH. Fibrinogen and cardiovascular disorders. QJM. 1995;88:155-165.