© 2004 Schattauer GmbH, Stuttgart

Theme Issue Article

Elevated D-dimer level is an independent risk factor

for cardiovascular death in out-patients with symptoms
compatible with heart failure
Urban Alehagen1, Ulf Dahlström1,Tomas L. Lindahl2
1
Department of Cardiology, Heart Center, University Hospital of Linköping, Linköping, Sweden
2
Department of Clinical Chemistry, Laboratory Medicine Östergötland, and Department of Biomedicine and Surgery,
University Hospital of Linköping, Linköping, Sweden

Summary
D-dimer, a marker of fibrin turnover, exhibits many interesting function was defined as reduced E/A ratio and/or an abnormal
properties as a biological marker of thrombosis. Some of the pattern of pulmonary venous flow. Blood samples were drawn,
properties of D-dimer might also be used to provide addition- and BNP and D-dimer were analysed. D-dimer was analysed
al information about patients with heart failure. In this study, we using an automated micro-latex assay. A statistical analysis was
evaluate the prognostic information acquired from D-dimer performed to identify the prognostic value of increased plasma
concerning increased risk of cardiovascular mortality in an eld- concentration of D-dimer. Results showed that during a medi-
erly population with symptoms associated with heart failure. an follow-up period of 5.5 years, 68 (14%) patients died of car-
A cardiologist examined 458 elderly patients, out of 548 invit- diovascular disease. No gender difference was noted.A plasma
ed, attending primary care for symptoms of dyspnoea, fatigue concentration of D-dimer >0.25mg/L increased the risk almost
and/or peripheral oedema and assessed NYHA functional class 4-fold. In conclusion, D-dimer is an independent risk factor for
and cardiac function.Abnormal systolic function was defined as cardiovascular mortality that may be used to risk-stratify
EF <40% on Doppler echocardiography. Abnormal diastolic patients with heart failure.

Keywords
Heart failure, elderly patients, primary health care, D-dimer,
prognosis Thromb Haemost 2004; 92: 1250–8

Introduction identify patients with increased risk of cardiovascular (CV)

mortality in order to optimise their management.
Heart failure (HF) has become one of the most important prob- As degradation products of fibrin, fragments of varying
lems in health care in the western world today, described in a sizes containing the D-dimer epitope are markers of fibrin turn-
review by Rich (1). The prognosis for severe HF is worse than over (6). In the literature, reports have been published concern-
many malignant diseases, with less than 50% survival after ing possible prognostic markers for increased risk of cardiovas-
5 years (2). Even in patients with less severe HF, the prognosis cular mortality among HF patients. The best-known indicators
is poor (3-5). In an elderly population with symptoms associat- are impaired left ventricular function on Doppler echocardiog-
ed with HF, it is important to establish the diagnosis and to start raphy (7) and impaired functional capacity (8). Patients with
appropriate therapy as fast as possible. It is also important to increased plasma concentration of noradrenalin have also been

Correspondence to: Received May 5, 2004

Urban Alehagen Accepted after resubmission August 26, 2004
Department of Cardiology
Heart Center Prepublished online October 5, 2004 DOI: 10.1160/TH04-05-0278
University Hospital of Linköping
Linköping, SE-581 85 Sweden
Tel.: +46-13-22 20 00, Fax: +46-13-222224
E-mail: [email protected]

1250
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.
 Alehagen, et al.: D-dimer as risk factor for cardiovascular death

shown to have an increased risk (9). This is also the case for The aim of this is to study the prognostic value of D-dimer
patients with increased plasma concentrations of natriuretic concerning increased risk of cardiovascular mortality in elderly
peptides because of myocardial infarction (10, 11), acute coro- patients with symptoms associated with heart failure.
nary syndromes (12, 13), and heart failure (14, 15).
In patients with atrial fibrillation, the levels of D-dimer and Methods
fibrinopeptide A, a marker of thrombosis, are attenuated after
successful electroconversion into sinus rhythm. Hence, it is Patients
believed that the velocity of blood flow is of importance We studied 510 patients out of a cohort of 1,168 living in a rural
(16-18). municipality with 10,300 inhabitants located in the southeast of
Impaired blood flow could also contribute to elevated levels Sweden. They attended primary care because of symptoms
of D-dimer in heart failure despite sinus rhythm. Moreover, and/or signs that might be attributed to HF. The design of the
studies concerning the ability to predict clinical outcome after study has been described in detail in a previous report (22). In
myocardial infarction based on the measurement of D-dimer summary, we examined all records (n = 1,168, patient age
show that patients with an increased plasma concentration had 65-87 years) of patients with symptoms and/or signs associated
worse prognosis than those with lower plasma concentration with heart failure who, between 1995-96, contacted the only
(19). Galvani and co-workers used a new, rapid blood assay for primary health care centre in the municipality. The symptoms
D-dimer to show that risk stratification using D-dimer in and/or signs displayed were those that might be associated with
patients with acute coronary syndromes was useful (20). This heart failure (shortness of breath, and/or peripheral oedema,
has also been shown in other studies (21). and/or fatigue). Patients in whom heart failure could not be

Table 1: Characteristics of the study

population.

1251
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.
 Alehagen, et al.: D-dimer as risk factor for cardiovascular death

ruled out by carefully scrutinising the records were invited to

participate in our study (n = 548). Thirty-eight patients decided
not to participate because of long travelling distance, severe
illness, mental insufficiency, or incapacity. Thus, 510 patients
accepted (participation rate, 93%). Inclusion was completed in
1996.
We performed a three-step analysis of the study population
to simulate a situation in the clinical routine. We analysed the
prognostic information by a selection process that increased
stepwise to eliminate certain important factors, other than heart
failure, that influence plasma concentration of D-dimer (Fig. 1).
In the first step, we excluded those who declined to donate
blood samples, or where the image quality of the Doppler
echocardiography was not acceptable. None of the patients were
being treated with warfarin or coumarin. No information on
aspirin usage was available, but since many patients had
ischaemic heart disease we assume a significant number were
on low-dose aspirin, i.e. 75 mg. However, treatment with aspi-
rin alone or in combination with fixed low dose warfarin has
been shown to have no influence on D-dimer concentration in
clinical trials on patients with atrial fibrillation (23, 24). This
first-step study population consisted of 458 patients (Table 1).
In the next step, we also excluded those with atrial fibrilla-
tion or dilated left atrium on Doppler echocardiography, as the
literature suggests that this might increase the plasma concen-
tration of D-dimer. In this step, the study population included
383 patients.
In the final step, we also excluded those who developed
malignancy during the follow-up period of 6 years, as malig-
nancy might influence the plasma concentration of D-dimer.
Those with impaired renal function (s-creatinine >200 µmol/L)
were also excluded as D-dimer is eliminated through glomeru-
lar secretion. In this step, the study population consisted of
370 patients.
All patients visited one cardiologist (UA), who recorded
patient history including drug treatment, performed a clinical
examination and electrocardiography, and assessed New York
Figure 1: The stepwise exclusion process of the study
Heart Association functional class (NYHA class). population.

Doppler echocardiography
Doppler echocardiographic examinations (Accuson XP-128c) BNP analysis
were performed with the patient in the supine left position. Both Blood samples were taken while the patients were fasting, seat-
M-mode and 2D methodology were used. Values for systolic ed, and after 30 minutes rest. The samples were collected in pre-
function (14), expressed as ejection fraction (EF) (25, 26), were chilled plastic tubes, containing EDTA (Terumo EDTA K-3),
categorised into 4 classes with interclass limits of 30%, 40% placed on ice and centrifuged at 3000 g for 10 min at +4°C.
and 50%. A semiquantitative method of assessment was used. Plasma was then immediately transferred to -70°C and stored
Normal systolic function was defined as EF ≥50%. Severely until analysis.
impaired systolic function was defined as EF <30%. To assess BNP was measured by the non-extraction immunoradiomet-
diastolic function, mitral E/A ratio and pulmonary venous flow ric technique from Shionogi (Osaka, Japan). The normal refer-
pattern were analysed and compared with age-adjusted decision ence interval in middle-aged adults as stated by the manufactur-
limits. er is 0-5.3 pmol/L (0-18 ng/L). The detection limit at the labor-
atory was 1.2 pmol/L (zero + 2 SD). The total interassay coeffi-

1252
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.
 Alehagen, et al.: D-dimer as risk factor for cardiovascular death

cient of variation was 9.3% (mean 12.4 pmol/L, n = 10), 5.4% toris, treatment for angina pectoris, and/or a previous myocar-
(mean 45.2 pmol/L, n = 10). The intra-assay coefficient of vari- dial infarction. Impaired renal function was defined as a serum
ation was 3.3% (mean 17.4 pmol/L, n = 7), 2.2% (mean creatinine concentration exceeding 200 µmol/L.
108.1 pmol/L, n =7).
Statistical methods
D-dimer analysis Descriptive data are presented as percentage or mean and SD. In
Blood was collected in Vacutainer tubes containing 1/10 volume the case of continuous variables, analyses were performed using
sodium citrate 0.13 mol/L. Blood samples were sent to the hos- Student’s unpaired two-sided t-test, whereas the chi-square test
pital laboratory and directly analysed. Since D-dimer was not a was used for discrete variables. Since BNP and for D-dimer
part of the diagnostic strategy during the study period, the were not distributed normally, a 10 logarithmic transformation of
results were sent by the laboratory information system only to the values was performed.
the authors and not reported back to the clinicians, unless spe- Multiple regression analysis was undertaken to test the inde-
cifically requested. The samples were analysed by utilising pendent prediction of increased D-dimer concentration. Cox
an automated micro-latex assay, Tina-quant®, from Roche proportional hazard regression analyses as well as a Kaplan-
Diagnostics (Mannheim, Germany) using Hitachi 911 analyser Meier analysis were performed to analyse the risk of mortality
(Hitachi; Kobe, Japan). This assay has been shown to prefer during the follow-up period. A p-value less than 0.05 was con-
high molecular weight fibrin complexes containing dimerized sidered statistically significant.
D-domains (27). The same cut-off value was used as for All data have been analysed by using generally available
exclusion of venous thrombosis in symptomatic out-patients, statistical analysis software packages (Statistica v6.0, Statsoft
0.25 mg/L, which equals 0.5 mg/L fibrinogen equivalent units Inc, Tulsa, OK, USA). The Ethics Committee of the University
(28, 29). The precision was good for four different batches of Hospital of Linköping, Sweden, approved the study protocol.
low control at mean concentrations of 0.32-0.59 mg/L (n = 387),
and two different batches of high control at 1.05 and 1.23 mg/L Results
(n = 393). The total imprecision was 5.8-13.6%, 4.7 and 6.0%,
respectively. Follow-up period
The patients were followed up during a median period of
Co-morbidity 5.5 years (range 242-2222 days). During the follow-up period,
Diabetes mellitus was defined as a fasting blood glucose con- 74 patients suffered a total mortality and 51 patients a cardio-
centration ≥7.0 mmol/L or ongoing treatment for diabetes (diet, vascular mortality in the first-step analysis (study population
oral therapy or insulin). Hypertension was defined as a blood n = 458). In the non-survivor group, a median follow-up period
pressure of more than 160/95 mm Hg measured in the right arm of 3.8 years (range 242-2156 days) was found, whereas in the
with the patient in the supine position after at least a 30-minute survivor group the median follow-up period was 5.7 years
rest (the choice of blood pressure levels used took into account (range 1883-2222 days).
the fact that we studied elderly patients and that we only meas-
ured blood pressure on one occasion). Patients were defined as Relation to gender and age
hypertensive, if they had previously been diagnosed as hyper- The study population was analysed according to signs of
tensive and were receiving anti-hypertensive medication. impaired heart function by Doppler echocardiography. In males
Ischaemic heart disease was defined as a history of angina pec- and females with normal systolic and diastolic function

Table 2: Multiple regression analysis

using 10Log D-dimer as dependent
variable.

1253
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.
 Alehagen, et al.: D-dimer as risk factor for cardiovascular death

Table 3: Cox proportional hazard

regression analysis of various parame-
ters against cardiovascular mortality in
study population Step 1 (n = 458).

Table 4: Cox proportional hazard

regression analysis of various parame-
ters against total mortality in study
population Step 1 (n = 458).

Table 5: Cox proportional hazard

regression analysis of various parame-
ters against cardiovascular mortality in
study population Step 3 (n = 349).

1254
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.
 Alehagen, et al.: D-dimer as risk factor for cardiovascular death

Table 6: Cox proportional hazard

regression analysis of various parame-
ters against total mortality in study
population Step 3 (n = 349).

(n = 214), the plasma concentration of D-dimer was compared, D-dimer in relation to other variables
but no significant differences were found. The most common clinical variables in heart failure were tested
A univariate correlation analysis showed a significantly in a multiple regression analysis using D-dimer >0.25 mg/L as
higher concentration of D-dimer in the older patients with nor- dependent variable. The variables with independent predictive
mal cardiac function compared with the younger ones (r = 0.17; power when D-dimer plasma concentration increased were plas-
p <0.05). In a multiple regression analysis using plasma concen- ma concentration of C-reactive protein >30mg/L and plasma
tration of D-dimer as dependent variable and three age concentration of BNP >100 pmol/L (Table 2).
classes, among others, as independent variables (<70 years,
70-75 years, >75 years), both age 70-75 years as well as age D-dimer as a prognostic instrument for
>75 years significantly predicted increased plasma concentra- mortality in comparison with gender, diabetes
tion of D-dimer (Table 2). mellitus, NYHA functional class and BNP
A Cox proportional hazard regression analysis was performed in
which the most common clinical variables of heart failure, the
Cardiovascular mortality

Figure 3: A Kaplan-Meier survival analy-

sis comparing the group with a plasma
BN
concentration of D-dimer > 0.25 mg/L
P L) compared with D-dimer < 0.25 mg/L.
(p g/ Graph indicates the survival proportion after
mo (m
l/ er a specific follow-up time, and indicates those
L) d im
D- that died (= censored), as well as those sur-
viving (= completed) at the specific time.

1255
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.
 Alehagen, et al.: D-dimer as risk factor for cardiovascular death

plasma concentration of the natriuretic peptide BNP, and increased the risk 1.9 times, and NYHA functional class III
increased plasma concentration of D-dimer were analysed. Total increased the risk 3.6 times. No other variables were found to
mortality or CV mortality were used as censoring variables. increase the risk (data not shown).
In the step-one analysis (study population n = 458), a plas- In the step-three analysis (study population n = 349),
ma concentration of D-dimer greater than 0.25 mg/L increased increased plasma concentration of D-dimer (>0.25mg/L)
the risk of CV mortality 2.7 times, whereas the impaired physi- increased the risk of CV mortality 3.8 times. NYHA functional
cal capacity classified as NYHA functional class III increased class III increased the risk 4.7 times, whereas diabetes mellitus
the risk of CV mortality 4.6 times. A plasma concentration of increased the risk 2.2 times, and male gender increased the risk
BNP >100 pmol/L increased the risk of CV mortality 3.4 times 2.9 times (Table 5).
(Table 3). In an analysis of total mortality in the same study popula-
Performing the same analysis, but using total mortality as tion, a D-dimer concentration of >0.25 mg/L increased the risk
the censoring variable, a plasma concentration of D-dimer of mortality 3.2 times. NYHA functional class III increased the
>0.25 mg/L almost doubled the risk, NYHA functional class III risk 3.6 times, and diabetes mellitus increased the risk almost
increased the risk 3 times, and diabetes mellitus increased the 2.6 times (Table 6).
risk 1.7 times. No other variable analysed increased the risk of A separate analysis of men with increased plasma concen-
total mortality (Table 4). tration of D-dimer compared with men without increased plas-
In the step-two analysis (study population n = 383), the risk ma concentration showed 3.3 times increased risk for cardio-
of CV mortality increased almost 7 times if patients had a plas- vascular death during the follow-up period.
ma concentration of BNP higher than 100 pmol/L. A plasma To illustrate the relation between plasma concentration of
concentration of D-dimer >0.25 mg/L increased the risk of CV BNP, D-dimer and CV mortality, we have produced a surface
mortality 3.8 times, whereas male gender increased the risk 3.3 plot illustrating the relation in the step-three population (Fig. 2).
times. Severe impaired functional capacity (NYHA class III) The figure shows that some individuals had had low plasma
increased the risk 4.7 times, whereas the presence of diabetes concentration of BNP, but high plasma concentration of D-
mellitus increased the risk 2.4 times, for total mortality (data not dimer, and died of cardiovascular causes during the follow-up
shown). period.
Using total mortality as a censoring variable, male gender A Kaplan-Meier survival analysis of those with a plasma
was found to increase the risk 2.2 times, diabetes mellitus concentration of D-dimer >0.25 mg/L compared with those

Figure 2: A 3-
dimensional figure
showing the rela-
tion between plas-
ma concentration
of BNP, plasma
concentration of
D-dimer, and cardi-
ovascular mortality
in the step-three
study population
(n = 349).

1256
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.
 Alehagen, et al.: D-dimer as risk factor for cardiovascular death

with a plasma concentration below 0.25 mg/L revealed signifi- used a 3D surface plot analysis. It is not surprising that patients
cant difference in survival (Fig. 3). with elevated plasma concentration of BNP and elevated D-
dimer had higher cardiovascular mortality. However, important
Discussion information is gained from those with low plasma concentration
of BNP but elevated concentration of D-dimer with respect to
It was our aim to study the prognostic value of D-dimer con- cardiovascular mortality. This indicates that even in patients
cerning increased risk of cardiovascular mortality in elderly with a normal filling pressure of the left ventricle, expressed as
patients with symptoms associated with heart failure. As the normal plasma concentration of BNP, increased plasma concen-
population we studied had a mean age of 73 years, it comprised tration of D-dimer offers prognostic information on cardiovas-
patients who were older than those studied in the majority of cular mortality. Furthermore, the results shown in Figure 2 indi-
other heart failure studies. Our study population resembles the cate that elevated D-dimer increases the risk of cardiovascular
heart failure population in the community which has a mean age mortality at all BNP concentrations. The cause of elevated
of 75 years (31-33). It is not self-evident that the results gained D-dimer in patients with normal plasma concentration of BNP
from other epidemiological studies using younger patients can is not known, but is most likely not due to disturbed flow in
be extrapolated to the heart failure population in the community. the heart. One explanation could be a hypercoagulable state of
As our goal was to analyse the prognostic power of D-dimer in the blood and/or vessel wall changes. One such condition is
a population resembling that of the community, we believe the advanced arteriosclerosis. Obviously, even a slight elevation
results obtained are relevant in the management of elderly carries an increased risk, as does CRP for patients with ischaem-
patients in primary health care. ic heart disease. None of the patients with high plasma concen-
We used a three-step design for the study, representing three tration of D-dimer, but high or low plasma concentration of
different selection criteria, and we have stepwise eliminated BNP, died of pulmonary embolism.
factors that might influence the plasma concentration of D- The analyses displayed a correlation between age and plas-
dimer. In the first step, the study population was analysed irre- ma concentration of D-dimer. The reason for this is unclear. It
spective of factors influencing the D-dimer plasma concentra- might be explained by the fact that the vessels of the aging pa-
tion. In spite of this, our analyses showed the prognostic power tient become stiffer and the anticoagulant capacity of the endo-
of D-dimer in predicting cardiovascular mortality. thelium is impaired, which increases the fibrin formation. The
In the second step, those with atrial fibrillation and/or dilat- subsequent fibrinolysis increases the concentration of D-dimer.
ed left atrium were eliminated, as patients with atrial fibrilla- Besides confirming that D-dimer is present in patients with
tion, when evaluated with Doppler echocardiography, often dis- heart failure, as shown in the literature, and can be reduced by
play signs of dilated left atrium. This indicates areas within the treatment with low molecular weight heparin (34), this study
heart with impaired motion, and thus the possibility of increas- has shown that D-dimer provides prognostic information con-
ing plasma concentration of D-dimer, even if there is no objec- cerning the risk of cardiovascular mortality that might be useful
tive sign of heart failure. Consequently, the study group is less in prioritising patients with heart failure. The results are impor-
heterogenic, and the relation between BNP and cardiovascular tant. Therefore we believe that our findings need to be con-
mortality is not surprising. Moreover, as male gender is gener- firmed through further study.
ally more often associated with greater cardiovascular risk, the
increased hazard ratio is not surprising. Limitations
In the third step, we also eliminated patients who developed The major limitation of this study is the limited size of the study
malignant diseases during the follow-up period of six years. The population, which generated a small number of cardiovascular
theory behind this is that malignant diseases are often associat- mortalities. The study population was recruited from a primary
ed with increased plasma concentration of D-dimer, and the health care centre, and the patients included had light to moder-
patients, although undiagnosed, might have the malignant dis- ate symptoms of heart failure. If the total study population had
ease on inclusion in the study. We also eliminated those with been greater, and had included patients with more severe symp-
impaired renal function, as this might have interfered with the toms of heart failure, the level of cardiovascular mortality
elimination of D-dimer. In this study population, increased plas- would have been greater. The results obtained in the study have
ma concentration of D-dimer increased the risk of cardiovascu- been obtained for one specific D-dimer assay, and may not
lar mortality almost 4-fold. Therefore D-dimer is an important apply to other assays.
risk factor for cardiovascular death in this population, besides
male gender or diabetes mellitus. Conclusion
In order to illustrate the interrelationship between plasma
concentration of BNP, D-dimer, and cardiovascular mortality We conclude that elevated D-dimer is an independent risk fac-
during the follow-up period in the step-three population, we tor for cardiovascular mortality in elderly patients in primary

1257
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.
 Alehagen, et al.: D-dimer as risk factor for cardiovascular death

health care with symptoms and/or signs that might be attributed almost six years. The role of D-dimer in the management of
to heart failure. A plasma concentration of D-dimer >0.25 mg/L heart failure patients, however, has to be evaluated in further
increased the risk almost 4-fold for cardiovascular mortality, clinical studies.
and 3-fold for total mortality during a follow-up period of

References
1. Rich MW, Nease RF. Cost-effectiveness anal- tic peptide in patients with acute coronary 25. Jensen-Urstad K, Bouvier F, Hojer J, et al.
ysis in clinical practice: the case of heart fail- syndromes. N Engl J Med 2001; 345: Comparison of different echocardiographic
ure. Arch Intern Med 1999; 159: 1690-700. 1014-21. methods with radionuclide imaging for meas-
2. Stewart S, MacIntyre K, Hole DJ, et al. More 14. Choy AM, Darbar D, Lang CC, et al. uring left ventricular ejection fraction during
‘malignant’ than cancer? Five-year survival Detection of left ventricular dysfunction after acute myocardial infarction treated by throm-
following a first admission for heart failure. acute myocardial infarction: comparison of bolytic therapy. Am J Cardiol 1998; 81:
Eur J Heart Fail 2001; 3: 315-22. clinical, echocardiographic, and neurohormo- 538-44.
3. Stewart S, Horowitz JD. Home-based inter- nal methods. Br Heart J 1994; 72: 16-22. 26. van Royen N, Jaffe CC, Krumholz HM, et al.
vention in congestive heart failure: long-term 15. McDonagh TA, Robb SD, Murdoch DR, et al. Comparison and reproducibility of visual
implications on readmission and survival. Biochemical detection of left-ventricular echocardiographic and quantitative radionu-
Circulation 2002; 105: 2861-6. systolic dysfunction. Lancet 1998; 351: 9-13. clide left ventricular ejection fractions. Am J
4. Mosterd A, Cost B, Hoes AW, et al. The prog- 16. Lip GY, Rumley A, Dunn FG, et al. Plasma Cardiol 1996; 77: 843-50.
nosis of heart failure in the general population: fibrinogen and fibrin D-dimer in patients 27. Dempfle CE, Zips S, Ergul H, et al. The fibrin
The Rotterdam Study. Eur Heart J 2001; 22: with atrial fibrillation: effects of cardioversion assay comparison trial (FACT): correlation of
1318-27. to sinus rhythm. Int J Cardiol 1995; 51: soluble fibrin assays with D-dimer. Thromb
5. Rector TS, Cohn JN. Prognosis in congestive 245-51. Haemost 2001; 86: 1204-9.
heart failure. Annu Rev Med 1994; 45: 17. Kamath S, Chin BS, Blann AD, et al. A study 28. van der Graaf F, van den Borne H, van der
341-50. of platelet activation in paroxysmal, persistent Kolk M, et al. Exclusion of deep venous
6. Marder VJ. Identification and purification of and permanent atrial fibrillation. Blood thrombosis with D-dimer testing – comparison
fibrinogen degradation products produced by Coagul Fibrinolysis 2002; 13: 627-36. of 13 D-dimer methods in 99 outpatients sus-
plasmin: considerations on the structure of 18. Giansante C, Fiotti N, Miccio M, et al. pected of deep venous thrombosis using
fibrinogen. Scand J Haematol Suppl 1971; 13: Coagulation indicators in patients with parox- venography as reference standard. Thromb
21-36. ysmal atrial fibrillation: effects of electric and Haemost 2000; 83: 191-8.
7. Kupari M, Lindroos M, Iivanainen AM, et al. pharmacologic cardioversion. Am Heart J 29. Knecht MF, Heinrich F. Clinical evaluation of
Congestive heart failure in old age: preva- 2000; 140: 423-9. an immunoturbidimetric D-dimer assay in the
lence, mechanisms and 4-year prognosis in the 19. Dempfle CE, Kontny F, Abildgaard U. diagnostic procedure of deep vein thrombosis
Helsinki Ageing Study. J Intern Med 1997; Predictive value of coagulation markers con- and pulmonary embolism. Thromb Res 1997;
241: 387-94. cerning clinical outcome 90 days after anteri- 88: 413-7.
8. Cicoira M, Davos CH, Florea V, et al. Chronic or myocardial infarction. Thromb Haemost 30. Mastoroberto P, Chello M, Perticone F.
heart failure in the very elderly: clinical status, 1999; 81: 701-4. Elevated Circulating Levels of von Willebrand
survival, and prognostic factors in 188 patients 20. Galvani M, Ferrini D, Ottani F, et al. Early Factor and D-dimer in Patients with Heart
more than 70 years old. Am Heart J 2001; 142: risk stratification of unstable angina/non-Q Failure and Mechanical Prosthesis. Scand J
174-80. myocardial infarction: biochemical markers of Thor Cardiovasc Surg 1996; 30: 77-81.
9. Cohn JN, Levine TB, Olivari MT, et al. coronary thrombosis. Int J Cardiol 1999; 68 31. Pitt B, Segal R, Martinez FA, et al.
Plasma norepinephrine as a guide to prognosis Suppl 1: S55-61. Randomised trial of losartan versus captopril
in patients with chronic congestive heart fail- 21. Fiotti N, Di Chiara A, Altamura N, et al. in patients over 65 with heart failure
ure. N Engl J Med 1984; 311: 819-23. Coagulation indicators in chronic stable effort (Evaluation of Losartan in the Elderly Study,
10. Talwar S, Squire IB, Downie PF, et al. Profile angina and unstable angina: relationship with ELITE). Lancet 1997; 349: 747-52.
of plasma N-terminal proBNP following acute acute phase reactants and clinical out- 32. Swedberg K, Kjekshus J. Effects of enalapril
myocardial infarction; correlation with left come. Blood Coagul Fibrinolysis 2002; 13: on mortality in severe congestive heart failure:
ventricular systolic dysfunction. Eur Heart J 247-55. results of the Cooperative North Scandinavian
2000; 21: 1514-21. 22. Alehagen U, Eriksson H, Nylander E, et al. Enalapril Survival Study (CONSENSUS). Am
11. Hall C, Rouleau JL, Moye L, et al. N-terminal Heart failure in the elderly. Characteristics of J Cardiol 1988; 62: 60A-66A.
proatrial natriuretic factor. An independent a Swedish primary health care population. 33. Parameshwar J, Shackell MM, Richardson A,
predictor of long-term prognosis after myo- Heart Drug 2002; 2: 211-220. et al. Prevalence of heart failure in three gen-
cardial infarction. Circulation 1994; 89: 23. Kamath S, Blann AD, Chin BS, et al. A study eral practices in north west London. Br J Gen
1934-42. of platelet activation in atrial fibrillation and Pract 1992; 42: 287-9.
12. Jernberg T, Stridsberg M, Venge P, et al. N-ter- the effects of antithrombotic therapy. Eur 34. De Lorenzo F, Newberry D, Scully M, et al.
minal pro brain natriuretic peptide on admis- Heart J 2002; 23: 1788-95. Low molecular weight heparin (bemiparin
sion for early risk stratification of patients 24. Li-Saw-Hee FL, Blann AD, Lip GY. Effects of sodium) and the coagulation profile of patients
with chest pain and no ST-segment elevation. fixed low-dose warfarin, aspirin-warfarin with heart failure. Am Heart J 2002; 143: 689.
J Am Coll Cardiol 2002; 40: 437-45. combination therapy, and dose-adjusted war-
13. de Lemos JA, Morrow DA, Bentley JH, et farin on thrombogenesis in chronic atrial fib-
al. The prognostic value of B-type natriure- rillation. Stroke 2000; 31: 828-33.

1258
Downloaded from www.thrombosis-online.com on 2015-07-06 | IP: 180.251.68.93
For personal or educational use only. No other uses without permission. All rights reserved.