- Email: [email protected]
von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated with oral anticoagulants
TR-06073; No of Pages 5 Thrombosis Research xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Full Length Article
von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated with oral anticoagulants Marcus Lind a,⁎, Lars Johansson a, Torbjörn K. Nilsson b, Jan-Håkan Jansson a, Martine J. Hollestelle c a b c
Department of Public Health and Clinical Medicine, Skellefteå Research Unit, Umeå University, Umeå, Sweden Department of Medical Biosciences, Umeå University, Umeå, Sweden Department Immunopathology and Blood Coagulation, Sanquin Diagnostic Services, Amsterdam, The Netherlands
a r t i c l e
i n f o
Article history: Received 27 May 2015 Received in revised form 18 August 2015 Accepted 22 August 2015 Available online xxxx Keywords: Anticoagulants Hemorrhage Biological marker Mortality VWF
a b s t r a c t Background: Serious bleeding is a frequent and feared treatment complication in patients treated with oral anticoagulants (OACs). Levels of von Willebrand factor (VWF) antigen have been linked to the risk of bleeding complications, mortality, and cardiovascular events. Objectives: In this longitudinal cohort study of evaluating patients treated with OACs, we aimed to evaluate the relationship between VWF displaying a glycoprotein Ib binding conformation (VWF activation factor) and the risk of cardiovascular events, bleeding complications, or all-cause mortality. Materials and methods: Blood samples were collected at baseline in 356 patients on OACs. Patients were followed for an average of 48 months and bleeding complications leading to admission to hospital or death, cardiovascular events (myocardial infarction, ischemic stroke, and peripheral arterial emboli), and all-cause mortality were recorded and classified. Results: During the study period, 47 bleeding complications, 84 cardiovascular events, and 97 deaths occurred. In multivariate Cox regression analyses, VWF activation factor was significantly associated with all-cause mortality (HR 1.62; 95% CI: 1.25–2.08) and cardiovascular events (HR 1.28; 95% CI: 1.01–1.63). There was no association observed between VWF activation factor and bleeding complications. Conclusions: Patients with high levels of VWF activation factor had an increased risk of cardiovascular events and allcause mortality during OAC treatment. The selectivity for thrombotic complications adds to the potential value of VWF activation factor as a biomarker or pharmacological target. © 2015 Published by Elsevier Ltd.
1. Introduction von Willebrand factor (VWF) is an essential player in primary hemostasis. Deficiency or structural defects of VWF lead to an increased bleeding tendency, which is seen in patients with von Willebrand disease [1]. In humans, most of the circulating VWF originates from Weibel–Palade bodies in endothelial cells [2]. Upon secretion, VWF is cleaved by ADAMTS13 into multimers [3]. In epidemiological studies, levels of VWF are most often measured as its antigen and are associated with myocardial infarction [4], stroke [5,6], intracranial bleedings [7], and mortality [8]. Levels of VWF antigen were proposed to be a risk marker, when used in clinical risk scores, for thromboembolic events in patients with atrial fibrillation [5].
⁎ Corresponding author at: Department of Medicine, Skellefteå County Hospital, 931 86 Skellefteå, Sweden. E-mail address: [email protected] (M. Lind).
High levels of VWF antigen are associated with both thrombotic events and bleeding complications in patients treated with oral anticoagulants (OACs) [9,10]. The mechanism for this dual association is not clear and different pathways could contribute to high levels of VWF antigen. Increased levels of VWF due to endothelial dysfunction [11] could reflect an underlying atherosclerotic process with a greater risk of thrombotic events. An increase in arteriosclerosis was observed in ADAMTS13 −/− mice, which still produce VWF [12]. It is also possible that VWF could be an effector in thrombotic events, possibly by increased binding to platelets. Measurement of VWF antigen or multimers might not reflect the platelet binding capacity of the circulating VWF since smaller multimers could still retain some platelet binding capacity [13]. The proportion of high-affinity VWF, which is able to bind spontaneously to platelets, could be measured using a specific antibody-based ELISA [14]. Our hypothesis is that the proportion of high-affinity VWF, also called active VWF, is increased in patients with thrombotic events. The primary aim of this study was to investigate possible associations between active VWF and cardiovascular
http://dx.doi.org/10.1016/j.thromres.2015.08.016 0049-3848/© 2015 Published by Elsevier Ltd.
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
2
M. Lind et al. / Thrombosis Research xxx (2015) xxx–xxx
events, bleeding complications, or all-cause mortality in patients on OAC treatment. 2. Methods 2.1. Patients The study cohort was recruited from the OAC clinic at Skellefteå County Hospital, Skellefteå, Sweden, in June 1996 and has in part been described in an earlier paper [10] Patients with a planned treatment time greater than 3 months were considered to be on long term treatment and eligible for the study; thus, consent forms were sent to them. Blood samples were obtained at inclusion for the 356 patients included in the final study cohort. All patients had been treated with OACs for at least 2 months prior to blood sampling. The indications for treatment and international normalized ratio (INR) at the time of blood sampling were available from registers at the OAC clinic. Additional data regarding diabetes, prior peptic ulcers, prior bleeding peptic ulcers, hypertension, weight, and height were available through questionnaires. The study was approved by the Research Ethics Committee of Umeå University. All patients provided written informed consent. 2.2. Blood sampling and laboratory procedures Venous blood samples were drawn after a minimum of stasis and collected in standard siliconized, plasma tubes containing 0.13 mol/L sodium citrate. After centrifugation the plasma samples were frozen and stored at −70 °C until analyzed. The study cohort was analyzed at the same time and location. The laboratory staff had no knowledge of event status. The levels of VWF with high affinity to platelets (VWF activation factor) were determined by measuring the levels of activated VWF, which represents the active, GPIba-binding conformation with the AU/VWFa-11 llama-derived nanobody, as described by Hulstein et al. [14] The VWF activation factor was calculated by dividing the absorbance slope of a patient's sample by the slope of the corresponding standard sample (normal pooled plasma [NPP] from a pool of more than 32 adult donors). The intra assay coefficient of variation was 7.1% and the inter assay coefficient of variation was 13.7%. The VWF antigen (kU/L) was measured using an enzyme-linked immunosorbent assay (ELISA) from DAKO in Denmark. High sensitivity C-reactive protein (hsCRP) was determined with an automated method (IMMULITE Diagnostic Products Corporation, USA). Creatinine was analyzed on a Hitachi 911 multi analyzer (Roche, Mannheim, Germany) with kits from Roche/Boehringer (Crea plus, enzymatic method). The estimated glomerular filtration rate (eGFR) was estimated using the four variable Modification of Diet in Renal Disease (MDRD) Study equation (GFR = 175 × standardized serum creatinine− 1.154 × age− 0.203 × 1.212 [if black] × 0.742 [if female]) [15]. The INR was determined using an Owren method with the reagent SPA 50 (Diagnostica Stago, Inc.) The INR was calibrated with a national standard. 2.3. Follow-up Study inclusion began on the date of blood sampling, with the earliest date being June 1st 1996. The patients were followed longitudinally from June 1st 1996 until death, bleeding, cessation of OAC treatment, or until January 1st 2002. All medical records were reviewed during the study period. Bleeding complications causing admission to hospital or death, myocardial infarction, ischemic stroke, and peripheral arterial emboli were recorded and classified by a panel of three researchers (ML, JHJ, LJ). Clinically relevant bleeding was defined as fatal bleeding and/or symptomatic bleeding in a critical organ, or overt bleeding based on objective investigations leading to hospital admission or prolonged hospital care. All other bleeding events were classified as minor and were excluded. Cardiovascular events and mortality included
myocardial infarction, stroke, transient ischemic attack, ruptured aortic aneurysm, and peripheral arterial emboli. Myocardial infarction and stroke were classified according to WHO criteria [16,17]. Peripheral arterial embolism and ruptured aortic aneurysm required clinical symptoms plus objective verification either on angiography, at operation or autopsy. The cause of death was registered and classified according to the death certificate. In all but one case, the cause of death could be classified. Investigators classifying events were blinded to the biochemical results. 2.4. Statistical analysis The distributions of VWF antigen, VWF activation factor, and hsCRP were skewed and, therefore, transformed using the natural logarithm (ln). Spearman correlation coefficients were used to evaluate potential relations between variables. Univariate Cox regression analysis was performed on each of the variables to estimate hazard ratios (HRs) and the 95% confidence interval (CI) with the increments of HRs presented for the standard deviations. Statistical analyses of VWF were performed using continuous variables and by categorizing the data into tertiles, with the lowest tertile as the reference group. Multivariate Cox regression analysis was performed to estimate the effects of different determinants when controlling for other factors. Age, hsCRP, and creatinine were possibly related to both VWF activation factor levels and outcome; therefore, they were considered to be potential confounders and were included in the multivariate analysis if the P-value in the univariate analyses was b 0.20. Separate multivariate Cox regression models were performed on each of the variables: VWF activation factor and VWF antigen. The assumption of proportional hazard was verified graphically using Kaplan–Meier survival curves. The VWF antigen and activation factor were dichotomized at the median and patients were classified into four groups according to high or low levels of antigen and activation factor. Direct age-adjustment was calculated using ten year intervals to present the effect of different levels of activity and antigen (Fig. 1). Cox regression analysis was used to test for significant differences between groups. A P-value b 0.05 (two-sided) was considered statistically significant. PAWS version 18 was used for all statistical analyses. Individuals with missing values (none for VWF antigen, 11 for VWF activation factor, 2 for CRP, and 29 for eGFR) were excluded from the statistical analyses. 3. Results Three-hundred-and-fifty six patients treated with OACs were enrolled in the study. Baseline characteristics of the patients are presented in Table 1. At the time of sampling, 9% of patients had an INR value b2.0, 5% had an INR value N 3.5, and 86% were within the therapeutic range used at the time, INR 2.0–3.5. During the study period, 74% of patients had a stable INR within the therapeutic range when treated with an OAC in the Skellefteå OAC clinic. Patients were followed on average for 1449 days (range 19–2028 days). The VWF activation factor significantly correlated with age (r = 0.16), eGFR (r = −0.18), VWF antigen (r = 0.43), and diabetes (r = − 0.14). There was no correlation between VWF activation factor and INR at the time of sampling. 3.1. All-cause mortality A total of 97 patients died during OAC treatment. VWF activation factor was available for 92 cases. Mean time to event was 818 days. Univariate and multivariate analyses are presented in Table 2. Hypertension and sex were not associated with the outcome and were, therefore, omitted from further analysis. The VWF antigen and activation factor, age, CRP, and eGFR were significantly associated with all-cause mortality in the univariate analysis. In the multivariate analysis, the highest tertile of VWF activation factor was significantly
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
M. Lind et al. / Thrombosis Research xxx (2015) xxx–xxx
3
Fig. 1. Age adjusted incidence at dichotomized levels of VWF antigen and VWF activation factor for the events. A — Incidence of all cause mortality according to levels of activation factor (Act) and antigen (Ag). B — Incidence of cardiovascular events according to levels of activation factor (Act) and antigen (Ag). C — Incidence of clinically relevant bleedings according to levels of activation factor (Act) and antigen (Ag).
associated with all-cause mortality (HR 3.06 [95% CI 1.56–6.00]) as was VWF activation factor as a continuous variable (HR 1.62 [95% CI 1.25– 2.08]). The VWF antigen, age, and CRP remained significantly associated with all-cause mortality after adjustment in multivariate analysis.
Diabetes and eGFR were attenuated and no longer significant in the multivariate analysis. A total of 66 patients died of cardiovascular causes. In multivariate analyses, the highest tertile of VWF activation factor was significantly
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
4
M. Lind et al. / Thrombosis Research xxx (2015) xxx–xxx
Table 1 Patient characteristics. n = 356 Age, median (SD), years Female, % Hypertension, % Diabetes, % VWF activation factor median (SD) Low (n = 115) Intermediate (n = 115) High (n = 115) VWF antigen, median (SD), kU dL−1 Low (n = 111) Intermediate (n = 115) High (n = 130) C-reactive protein, median (SD),mg L−1 eGFR, median (SD) mL/min/1.73 m2 Indications for treatment, % Prosthetic heart valve Atrial fibrillation Venous thromboembolism Peripheral arterial disease Ischemic heart disease Miscellaneous
72 ± 16 38 45 7 1.69 ± 0.72 0–1.43 1.44–1.91 1.92- ∞ 1.62 ± 0.63 0–1.38 1.39–1.82 ≥1.83 3.4 ± 18 77 ± 22 30 37 12 9 8 4
eGFR—estimated glomerular filtration rate. VWF—von Willebrand factor.
associated with cardiovascular death (HR 2.24 [95% CI 1.08–4.66]) as was VWF activation factor as a continuous variable (HR 1.43 [95% CI 1.06–1.92]). As shown in Fig. 1A, patients with high levels of both antigen and activation factor had the highest incidence of all-cause mortality, 11% per treatment year. Patients with low antigen and low activation factor had a significantly lower risk of mortality than patients with high levels of antigen and activation factor when tested in ageadjusted Cox regression (HR 2.41 95% CI 1.38–4.20). No synergistic interaction was found (data not shown). 3.2. Cardiovascular events In total, 97 cardiovascular events occurred during the period of the study. VWF activation factor was available for 90 cases. Age, CRP, diabetes, VWF activation factor, and VWF antigen were all associated with cardiovascular events in the univariate analysis. In multivariate analysis, the associations for age, sex, CRP, and VWF antigen were still significant. For VWF activation factor, a significant association was only observed for the continuous variable. For diabetes, the association was attenuated and no longer significant. Patients with high levels of both antigen and activation factor had the highest incidence of
cardiovascular events, 10.8% per treatment year (see Fig. 1B). Patients with low antigen levels and low activation factor had the lowest risk of cardiovascular events. No significant difference was found between the groups stratified by antigen and activation factor levels (HR 1.54 95% CI 0.92–2.58). 3.3. Clinically relevant bleeding In univariate analysis, only the highest tertile of VWF antigen and age were associated with clinically relevant bleeding. When adjusting for age in the multivariate analysis, the association with VWF antigen was attenuated and not significant. There was no association observed between VWF activation factor and clinically relevant bleeding. The incidences of clinically relevant bleeding at different levels of antigen and activation factor are presented in Fig. 1C. No significant association was found between the groups stratified by antigen and activation factor levels (data not shown). 4. Discussion The present study is the first studying the association between VWF activation factor and cardiovascular events and all-cause mortality in patients treated with OACs. In this study, VWF activation factor was significantly associated with all-cause mortality and cardiovascular events. Even after adjustment for age, inflammation (CRP), and kidney function (eGFR), a significant association was observed with all-cause mortality. This indicates that an increased level of VWF activation factor is not solely an expression of general inflammation and impaired kidney function. We did not find an association between VWF activation factor and bleeding complications. This result sets VWF activation factor apart from VWF antigen, which has been associated with bleeding complications as well as cardiovascular events [10]. It was also noted that in patients with high levels of VWF activation factor and antigen, the incidences of cardiovascular events and mortality were especially high, indicating that these patients are at a highest risk. The VWF activation factor measurement is determining the proportion of VWF with platelet activating ability [13]; while, the VWF antigen measures all plasma VWF fragments irrespective of the platelet activating ability. The VWF antigen is often regarded as a marker of endothelial dysfunction and arteriosclerosis [11]. It is possible that a high level of VWF antigen reflects a more severe form of endothelial dysfunction, in general; while, increased active VWF may reflect a decrease in the activity regulation due to reduced proteolysis (by ADAMTS13 or another protease) or diminished clearance by low-density lipoprotein receptor-related
Table 2 Uni- and multivariate analysis for different endpoints. All-cause mortality n = 97
VWF activation factor, per SD Low (n = 115) Intermediate (n = 115) High (n = 115) VWF antigen, per SD, kU dL−1 Low (n = 111) Intermediate (n = 115) High (n = 130) C-reactive protein, per SD, mg L−1 eGFR per SD, mL/min/1.73 m2 Female Diabetes Hypertension Age. per 10 years interval
Clinically relevant bleedings
Cardiovascular events
n = 47
n = 97
Univariate
Multivariate
Univariate
Multivariate
Univariate
Multivariate
1.94 (1.56–2.40) 1 2.51 (1.31–4.83) 5.05(5 (2.73–9.33) 1.92 (1.59–2.30) 1 2.87 (1.44–5.72) 5.98 (3.13–11.43) 1.52 (1.27–1.83) 0.77 (0.62–0.96) 0.79 (0.52–1.20) 0.59 (0.35–0.98) 0.77 (0.52–1.16) 1.94 (1.52–2.46)
1.62 (1.25–2.08) 1 1.90 (0.95–3.81) 3.06 (1.56–6.0) 1.62 (1.29–2.09) 1 1.97 (0.94–4.14) 3.17 (1.54–6.52) 1.31 (1.09–1.58) 1.03 (0.82–1.29) – 0.83 (0.47–1.45) – 1.75 (1.33–2.31)
1.22 (0.90–1.65) 1 0.83 (0.83–3.21) 1.23 (0.62–2.47) 1.25 (0.94–1.67) 1 1.55 (0.72–3.35) 2.27 (1.08–4.74) 0.91 (0.68–1.22) 1.01 (0.76–1.35) 0.86 (0.47–1.55) 0.92 (0.39–2.16) 0.90 (0.50–1.62) 1.81 (1.28–2.55)
1.10 (0.80–1.51) 1 0.70 (0.33–1.49) 1.34 (0.68–2.84) 1.14 (0.84–1.54) 1 1.38 (0.64–2.98) 1.78 (0.84–3.79) – – – – – 1.77 (1.24–2.51)
1.49 (1.20–1.84) 1 2.01 (1.16–3.48) 2.54 (1.47–4.39) 1.55 (1.28–1.88) 1 2.30 (1.28–4.16) 3.47 (1.96–6.14) 1.28 (1.06–1.55) 0.91 (0.73–1.22) 0.72 (0.47–1.10) 0.40 (0.30–0.80) 0.68 (0.46–1.02) 1.88 (1.48–2.39)
1.28 (1.01–1.63) 1 1.53 (0.87–2.71) 1.74 (0.97–3.14) 1.37 (1.10–1.71) 1 2.04 (1.06–3.93) 2.65 (1.39–5.09) 1.27 (1.04–1.56) – 0.58 (0.37–0.92) 0.67 (0.39–1.13) 0.84 (0.55–1.28) 1.93 (1.48–2.52)
eGFR—estimated glomerular filtration rate. VWF—von Willebrand factor. Factors were included in the multivariate analysis if p b 0.20 in the univariate analysis, von Willebrand factor activity and antigen were not included in the same multivariate analysis.
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
M. Lind et al. / Thrombosis Research xxx (2015) xxx–xxx
protein-1 [18]. Various findings support that VWF is not only a marker, but also an important effector in the pathogenesis of myocardial infarction [19]. Our finding that increased VWF activation factor levels are related to cardiovascular events and mortality could indicate a causal role of VWF in facilitating platelet activation. An intriguing finding is that patients with high levels of VWF antigen and activation factor seem to have a substantially higher incidence of cardiovascular events than other patients. It is possible that an increase in platelet activation could be especially deleterious in patients with more advanced atherosclerosis and endothelial dysfunction. Further studies are needed to investigate if the increased risk associated with VWF antigen is in part mediated by activation of platelets and, thereby, reflected in the levels of VWF activation factor. The levels of antigen were generally high in the studied population, which is most likely attributed to patients treated with OACs having a higher incidence of preexisting vessel wall damage. This leads to activation of the endothelium and/or exposure of the subendothelial matrix. 4.1. Strengths and limitations Samples were collected in a prospective manner, which limited the possible influence of acute phase reactions related to the studied events. The events were detected and classified in a retrospective manner by reviewing hospital records. To optimize the detection and minimize the risk of misclassification, all hospital records from all clinics were manually reviewed for each patient during the follow-up time. The levels of antigen are somewhat reduced by OAC treatment [20], and the effect on VWF activation factor is not known. We did not adjust for time in the INR range in this study. However, patients treated with warfarin in our clinic have a high time in range (74%) and we could not see any relationship between measured antigen or activity and levels of INR at sampling. In our study the samples were frozen and stored for some time before they were analyzed. The potential effects of this on levels of VWF antigen and VWF activation factor is not entirely clear. From other studies we know that a long storage time does not seem to affect the levels of fibrinolytic factors but that changes in the reagents kit used could introduce a bias [21]. Long term storage also seems to be acceptable when analyzing CRP [22]. The effects of long term storage on VWF is less studied but in a study of Rumley et al. a storage time of 6 to 12 years did not seem to affect the results for VWF antigen or factor VIIc [23]. It is reasonable to assume that storage of blood samples does not introduce a significant bias as long as all samples are analyzed at the same time, using the same reagents kits. Together, our data demonstrate that VWF activation factor in patients treated with OACs was associated with all-cause mortality and cardiovascular events particularly in patients with increased VWF antigen. There was no association with bleeding complications. The selectivity for thrombotic complications adds to the potential value of VWF activation factor as a biomarker and/or pharmacological target. Authors' contributions M. Lind collected the data designed the study, analyzed and interpreted the data and drafted the manuscript. L. Johansson, J.H Jansson, T.K Nilsson and M. J. Hollestelle designed the study, and participated in analysis and interpretation of data and in drafting of the manuscript. All authors approved the final version of the manuscript. Acknowledgements We gratefully thank prof. Philip de Groot, Department of Clinical Chemistry and Hematology, University Medical Center Utrecht, for kindly providing us with antibody AU/VWFa-11. Furthermore we like to acknowledge Marlies van Schagen for excellent technical support.
5
References [1] Z.M. Ruggeri, T.S. Zimmerman, von Willebrand factor and von Willebrand disease, Blood 70 (1987) 895–904. [2] T.C. Nichols, C.M. Samama, D.A. Bellinger, J. Roussi, R.L. Reddick, M. Bonneau, M.S. Read, O. Bailliart, G.G. Koch, M. Vaiman, et al., Function of von Willebrand factor after crossed bone marrow transplantation between normal and von Willebrand disease pigs: effect on arterial thrombosis in chimeras, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 2455–2459. [3] M. Furlan, R. Robles, D. Affolter, D. Meyer, P. Baillod, B. Lammle, Triplet structure of von Willebrand factor reflects proteolytic degradation of high molecular weight multimers, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 7503–7507. [4] J. Danesh, J.G. Wheeler, G.M. Hirschfield, S. Eda, G. Eiriksdottir, A. Rumley, G.D. Lowe, M.B. Pepys, V. Gudnason, C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease, N. Engl. J. Med. 350 (2004) 1387–1397. [5] G.Y. Lip, D. Lane, C. Van Walraven, R.G. Hart, Additive role of plasma von Willebrand factor levels to clinical factors for risk stratification of patients with atrial fibrillation, Stroke 37 (2006) 2294–2300. [6] D.S. Conway, L.A. Pearce, B.S. Chin, R.G. Hart, G.Y. Lip, Prognostic value of plasma von Willebrand factor and soluble P-selectin as indices of endothelial damage and platelet activation in 994 patients with nonvalvular atrial fibrillation, Circulation 107 (2003) 3141–3145. [7] L. Johansson, J.H. Jansson, B. Stegmayr, T.K. Nilsson, G. Hallmans, K. Boman, Hemostatic factors as risk markers for intracerebral hemorrhage: a prospective incident case-referent study, Stroke 35 (2004) 826–830. [8] J.H. Jansson, T.K. Nilsson, O. Johnson, von Willebrand factor, tissue plasminogen activator, and dehydroepiandrosterone sulphate predict cardiovascular death in a 10 year follow up of survivors of acute myocardial infarction, Heart 80 (1998) 334–337. [9] V. Roldan, F. Marin, B. Muina, J.M. Torregrosa, D. Hernandez-Romero, M. Valdes, V. Vicente, G.Y. Lip, Plasma von Willebrand factor levels are an independent risk factor for adverse events including mortality and major bleeding in anticoagulated atrial fibrillation patients, J. Am. Coll. Cardiol. 57 (2011) 2496–2504, http://dx.doi.org/ 10.1016/j.jacc.2010.12.033. [10] M. Lind, K. Boman, L. Johansson, T.K. Nilsson, L.S. Jarvholm, J.H. Jansson, von Willebrand factor predicts major bleeding and mortality during oral anticoagulant treatment, J. Intern. Med. 271 (2012) 239–246, http://dx.doi.org/10.1111/j.13652796.2011.02418.x. [11] A.D. Blann, Plasma von Willebrand factor, thrombosis, and the endothelium: the first 30 years, Thromb. Haemost. 95 (2006) 49–55. [12] C. Gandhi, A. Ahmad, K.M. Wilson, A.K. Chauhan, ADAMTS13 modulates atherosclerotic plaque progression in mice via a VWF-dependent mechanism, J. Thromb. Haemost. 12 (2014) 255–260, http://dx.doi.org/10.1111/jth.12456. [13] E. Groot, R. Fijnheer, S.A. Sebastian, P.G. de Groot, P.J. Lenting, The active conformation of von Willebrand factor in patients with thrombotic thrombocytopenic purpura in remission, J. Thromb. Haemost. 7 (2009) 962–969, http://dx.doi.org/10.1111/j. 1538-7836.2009.03433.x. [14] J.J. Hulstein, P.G. de Groot, K. Silence, A. Veyradier, R. Fijnheer, P.J. Lenting, A novel nanobody that detects the gain-of-function phenotype of von Willebrand factor in ADAMTS13 deficiency and von Willebrand disease type 2B, Blood 106 (2005) 3035–3042, http://dx.doi.org/10.1182/blood-2005-03-1153. [15] A.S. Levey, J. Coresh, T. Greene, L.A. Stevens, Y.L. Zhang, S. Hendriksen, J.W. Kusek, F. Van Lente, Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate, Ann. Intern. Med. 145 (2006) 247–254. [16] H. Tunstall-Pedoe, K. Kuulasmaa, P. Amouyel, D. Arveiler, A.M. Rajakangas, A. Pajak, Myocardial infarction and coronary deaths in the World Health Organization MONICA project. Registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents, Circulation 90 (1994) 583–612. [17] B. Stegmayr, K. Asplund, Measuring stroke in the population: quality of routine statistics in comparison with a population-based stroke registry, Neuroepidemiology 11 (1992) 204–213. [18] G. Rastegarlari, J.N. Pegon, C. Casari, S. Odouard, A.M. Navarrete, N. Saint-Lu, B.J. van Vlijmen, P. Legendre, O.D. Christophe, C.V. Denis, P.J. Lenting, Macrophage LRP1 contributes to the clearance of von Willebrand factor, Blood 119 (2012) 2126–2134, http://dx.doi.org/10.1182/blood-2011-08-373605. [19] A.O. Spiel, J.C. Gilbert, B. Jilma, von Willebrand factor in cardiovascular disease: focus on acute coronary syndromes, Circulation 117 (2008) 1449–1459, http://dx.doi.org/ 10.1161/CIRCULATIONAHA.107.722827. [20] F.L. Li-Saw-Hee, A.D. Blann, G.Y. Lip, Effects of fixed low-dose warfarin, aspirinwarfarin combination therapy, and dose-adjusted warfarin on thrombogenesis in chronic atrial fibrillation, Stroke 31 (2000) 828–833. [21] J. Hernestal-Boman, J.H. Jansson, T.K. Nilsson, M. Eliasson, L. Johansson, Long-term stability of fibrinolytic factors stored at −80 degrees C, Thromb. Res. 125 (2010) 451–456. [22] M.R. Lewis, P.W. Callas, N.S. Jenny, R.P. Tracy, Longitudinal stability of coagulation, fibrinolysis, and inflammation factors in stored plasma samples, Thromb. Haemost. 86 (2001) 1495–1500. [23] A. Rumley, G.D. Lowe, P.M. Sweetnam, J.W. Yarnell, R.P. Ford, Factor VIII, von Willebrand factor and the risk of major ischaemic heart disease in the Caerphilly heart study, Br. J. Haematol. 105 (1999) 110–116.
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Full Length Article
von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated with oral anticoagulants Marcus Lind a,⁎, Lars Johansson a, Torbjörn K. Nilsson b, Jan-Håkan Jansson a, Martine J. Hollestelle c a b c
Department of Public Health and Clinical Medicine, Skellefteå Research Unit, Umeå University, Umeå, Sweden Department of Medical Biosciences, Umeå University, Umeå, Sweden Department Immunopathology and Blood Coagulation, Sanquin Diagnostic Services, Amsterdam, The Netherlands
a r t i c l e
i n f o
Article history: Received 27 May 2015 Received in revised form 18 August 2015 Accepted 22 August 2015 Available online xxxx Keywords: Anticoagulants Hemorrhage Biological marker Mortality VWF
a b s t r a c t Background: Serious bleeding is a frequent and feared treatment complication in patients treated with oral anticoagulants (OACs). Levels of von Willebrand factor (VWF) antigen have been linked to the risk of bleeding complications, mortality, and cardiovascular events. Objectives: In this longitudinal cohort study of evaluating patients treated with OACs, we aimed to evaluate the relationship between VWF displaying a glycoprotein Ib binding conformation (VWF activation factor) and the risk of cardiovascular events, bleeding complications, or all-cause mortality. Materials and methods: Blood samples were collected at baseline in 356 patients on OACs. Patients were followed for an average of 48 months and bleeding complications leading to admission to hospital or death, cardiovascular events (myocardial infarction, ischemic stroke, and peripheral arterial emboli), and all-cause mortality were recorded and classified. Results: During the study period, 47 bleeding complications, 84 cardiovascular events, and 97 deaths occurred. In multivariate Cox regression analyses, VWF activation factor was significantly associated with all-cause mortality (HR 1.62; 95% CI: 1.25–2.08) and cardiovascular events (HR 1.28; 95% CI: 1.01–1.63). There was no association observed between VWF activation factor and bleeding complications. Conclusions: Patients with high levels of VWF activation factor had an increased risk of cardiovascular events and allcause mortality during OAC treatment. The selectivity for thrombotic complications adds to the potential value of VWF activation factor as a biomarker or pharmacological target. © 2015 Published by Elsevier Ltd.
1. Introduction von Willebrand factor (VWF) is an essential player in primary hemostasis. Deficiency or structural defects of VWF lead to an increased bleeding tendency, which is seen in patients with von Willebrand disease [1]. In humans, most of the circulating VWF originates from Weibel–Palade bodies in endothelial cells [2]. Upon secretion, VWF is cleaved by ADAMTS13 into multimers [3]. In epidemiological studies, levels of VWF are most often measured as its antigen and are associated with myocardial infarction [4], stroke [5,6], intracranial bleedings [7], and mortality [8]. Levels of VWF antigen were proposed to be a risk marker, when used in clinical risk scores, for thromboembolic events in patients with atrial fibrillation [5].
⁎ Corresponding author at: Department of Medicine, Skellefteå County Hospital, 931 86 Skellefteå, Sweden. E-mail address: [email protected] (M. Lind).
High levels of VWF antigen are associated with both thrombotic events and bleeding complications in patients treated with oral anticoagulants (OACs) [9,10]. The mechanism for this dual association is not clear and different pathways could contribute to high levels of VWF antigen. Increased levels of VWF due to endothelial dysfunction [11] could reflect an underlying atherosclerotic process with a greater risk of thrombotic events. An increase in arteriosclerosis was observed in ADAMTS13 −/− mice, which still produce VWF [12]. It is also possible that VWF could be an effector in thrombotic events, possibly by increased binding to platelets. Measurement of VWF antigen or multimers might not reflect the platelet binding capacity of the circulating VWF since smaller multimers could still retain some platelet binding capacity [13]. The proportion of high-affinity VWF, which is able to bind spontaneously to platelets, could be measured using a specific antibody-based ELISA [14]. Our hypothesis is that the proportion of high-affinity VWF, also called active VWF, is increased in patients with thrombotic events. The primary aim of this study was to investigate possible associations between active VWF and cardiovascular
http://dx.doi.org/10.1016/j.thromres.2015.08.016 0049-3848/© 2015 Published by Elsevier Ltd.
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
2
M. Lind et al. / Thrombosis Research xxx (2015) xxx–xxx
events, bleeding complications, or all-cause mortality in patients on OAC treatment. 2. Methods 2.1. Patients The study cohort was recruited from the OAC clinic at Skellefteå County Hospital, Skellefteå, Sweden, in June 1996 and has in part been described in an earlier paper [10] Patients with a planned treatment time greater than 3 months were considered to be on long term treatment and eligible for the study; thus, consent forms were sent to them. Blood samples were obtained at inclusion for the 356 patients included in the final study cohort. All patients had been treated with OACs for at least 2 months prior to blood sampling. The indications for treatment and international normalized ratio (INR) at the time of blood sampling were available from registers at the OAC clinic. Additional data regarding diabetes, prior peptic ulcers, prior bleeding peptic ulcers, hypertension, weight, and height were available through questionnaires. The study was approved by the Research Ethics Committee of Umeå University. All patients provided written informed consent. 2.2. Blood sampling and laboratory procedures Venous blood samples were drawn after a minimum of stasis and collected in standard siliconized, plasma tubes containing 0.13 mol/L sodium citrate. After centrifugation the plasma samples were frozen and stored at −70 °C until analyzed. The study cohort was analyzed at the same time and location. The laboratory staff had no knowledge of event status. The levels of VWF with high affinity to platelets (VWF activation factor) were determined by measuring the levels of activated VWF, which represents the active, GPIba-binding conformation with the AU/VWFa-11 llama-derived nanobody, as described by Hulstein et al. [14] The VWF activation factor was calculated by dividing the absorbance slope of a patient's sample by the slope of the corresponding standard sample (normal pooled plasma [NPP] from a pool of more than 32 adult donors). The intra assay coefficient of variation was 7.1% and the inter assay coefficient of variation was 13.7%. The VWF antigen (kU/L) was measured using an enzyme-linked immunosorbent assay (ELISA) from DAKO in Denmark. High sensitivity C-reactive protein (hsCRP) was determined with an automated method (IMMULITE Diagnostic Products Corporation, USA). Creatinine was analyzed on a Hitachi 911 multi analyzer (Roche, Mannheim, Germany) with kits from Roche/Boehringer (Crea plus, enzymatic method). The estimated glomerular filtration rate (eGFR) was estimated using the four variable Modification of Diet in Renal Disease (MDRD) Study equation (GFR = 175 × standardized serum creatinine− 1.154 × age− 0.203 × 1.212 [if black] × 0.742 [if female]) [15]. The INR was determined using an Owren method with the reagent SPA 50 (Diagnostica Stago, Inc.) The INR was calibrated with a national standard. 2.3. Follow-up Study inclusion began on the date of blood sampling, with the earliest date being June 1st 1996. The patients were followed longitudinally from June 1st 1996 until death, bleeding, cessation of OAC treatment, or until January 1st 2002. All medical records were reviewed during the study period. Bleeding complications causing admission to hospital or death, myocardial infarction, ischemic stroke, and peripheral arterial emboli were recorded and classified by a panel of three researchers (ML, JHJ, LJ). Clinically relevant bleeding was defined as fatal bleeding and/or symptomatic bleeding in a critical organ, or overt bleeding based on objective investigations leading to hospital admission or prolonged hospital care. All other bleeding events were classified as minor and were excluded. Cardiovascular events and mortality included
myocardial infarction, stroke, transient ischemic attack, ruptured aortic aneurysm, and peripheral arterial emboli. Myocardial infarction and stroke were classified according to WHO criteria [16,17]. Peripheral arterial embolism and ruptured aortic aneurysm required clinical symptoms plus objective verification either on angiography, at operation or autopsy. The cause of death was registered and classified according to the death certificate. In all but one case, the cause of death could be classified. Investigators classifying events were blinded to the biochemical results. 2.4. Statistical analysis The distributions of VWF antigen, VWF activation factor, and hsCRP were skewed and, therefore, transformed using the natural logarithm (ln). Spearman correlation coefficients were used to evaluate potential relations between variables. Univariate Cox regression analysis was performed on each of the variables to estimate hazard ratios (HRs) and the 95% confidence interval (CI) with the increments of HRs presented for the standard deviations. Statistical analyses of VWF were performed using continuous variables and by categorizing the data into tertiles, with the lowest tertile as the reference group. Multivariate Cox regression analysis was performed to estimate the effects of different determinants when controlling for other factors. Age, hsCRP, and creatinine were possibly related to both VWF activation factor levels and outcome; therefore, they were considered to be potential confounders and were included in the multivariate analysis if the P-value in the univariate analyses was b 0.20. Separate multivariate Cox regression models were performed on each of the variables: VWF activation factor and VWF antigen. The assumption of proportional hazard was verified graphically using Kaplan–Meier survival curves. The VWF antigen and activation factor were dichotomized at the median and patients were classified into four groups according to high or low levels of antigen and activation factor. Direct age-adjustment was calculated using ten year intervals to present the effect of different levels of activity and antigen (Fig. 1). Cox regression analysis was used to test for significant differences between groups. A P-value b 0.05 (two-sided) was considered statistically significant. PAWS version 18 was used for all statistical analyses. Individuals with missing values (none for VWF antigen, 11 for VWF activation factor, 2 for CRP, and 29 for eGFR) were excluded from the statistical analyses. 3. Results Three-hundred-and-fifty six patients treated with OACs were enrolled in the study. Baseline characteristics of the patients are presented in Table 1. At the time of sampling, 9% of patients had an INR value b2.0, 5% had an INR value N 3.5, and 86% were within the therapeutic range used at the time, INR 2.0–3.5. During the study period, 74% of patients had a stable INR within the therapeutic range when treated with an OAC in the Skellefteå OAC clinic. Patients were followed on average for 1449 days (range 19–2028 days). The VWF activation factor significantly correlated with age (r = 0.16), eGFR (r = −0.18), VWF antigen (r = 0.43), and diabetes (r = − 0.14). There was no correlation between VWF activation factor and INR at the time of sampling. 3.1. All-cause mortality A total of 97 patients died during OAC treatment. VWF activation factor was available for 92 cases. Mean time to event was 818 days. Univariate and multivariate analyses are presented in Table 2. Hypertension and sex were not associated with the outcome and were, therefore, omitted from further analysis. The VWF antigen and activation factor, age, CRP, and eGFR were significantly associated with all-cause mortality in the univariate analysis. In the multivariate analysis, the highest tertile of VWF activation factor was significantly
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
M. Lind et al. / Thrombosis Research xxx (2015) xxx–xxx
3
Fig. 1. Age adjusted incidence at dichotomized levels of VWF antigen and VWF activation factor for the events. A — Incidence of all cause mortality according to levels of activation factor (Act) and antigen (Ag). B — Incidence of cardiovascular events according to levels of activation factor (Act) and antigen (Ag). C — Incidence of clinically relevant bleedings according to levels of activation factor (Act) and antigen (Ag).
associated with all-cause mortality (HR 3.06 [95% CI 1.56–6.00]) as was VWF activation factor as a continuous variable (HR 1.62 [95% CI 1.25– 2.08]). The VWF antigen, age, and CRP remained significantly associated with all-cause mortality after adjustment in multivariate analysis.
Diabetes and eGFR were attenuated and no longer significant in the multivariate analysis. A total of 66 patients died of cardiovascular causes. In multivariate analyses, the highest tertile of VWF activation factor was significantly
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
4
M. Lind et al. / Thrombosis Research xxx (2015) xxx–xxx
Table 1 Patient characteristics. n = 356 Age, median (SD), years Female, % Hypertension, % Diabetes, % VWF activation factor median (SD) Low (n = 115) Intermediate (n = 115) High (n = 115) VWF antigen, median (SD), kU dL−1 Low (n = 111) Intermediate (n = 115) High (n = 130) C-reactive protein, median (SD),mg L−1 eGFR, median (SD) mL/min/1.73 m2 Indications for treatment, % Prosthetic heart valve Atrial fibrillation Venous thromboembolism Peripheral arterial disease Ischemic heart disease Miscellaneous
72 ± 16 38 45 7 1.69 ± 0.72 0–1.43 1.44–1.91 1.92- ∞ 1.62 ± 0.63 0–1.38 1.39–1.82 ≥1.83 3.4 ± 18 77 ± 22 30 37 12 9 8 4
eGFR—estimated glomerular filtration rate. VWF—von Willebrand factor.
associated with cardiovascular death (HR 2.24 [95% CI 1.08–4.66]) as was VWF activation factor as a continuous variable (HR 1.43 [95% CI 1.06–1.92]). As shown in Fig. 1A, patients with high levels of both antigen and activation factor had the highest incidence of all-cause mortality, 11% per treatment year. Patients with low antigen and low activation factor had a significantly lower risk of mortality than patients with high levels of antigen and activation factor when tested in ageadjusted Cox regression (HR 2.41 95% CI 1.38–4.20). No synergistic interaction was found (data not shown). 3.2. Cardiovascular events In total, 97 cardiovascular events occurred during the period of the study. VWF activation factor was available for 90 cases. Age, CRP, diabetes, VWF activation factor, and VWF antigen were all associated with cardiovascular events in the univariate analysis. In multivariate analysis, the associations for age, sex, CRP, and VWF antigen were still significant. For VWF activation factor, a significant association was only observed for the continuous variable. For diabetes, the association was attenuated and no longer significant. Patients with high levels of both antigen and activation factor had the highest incidence of
cardiovascular events, 10.8% per treatment year (see Fig. 1B). Patients with low antigen levels and low activation factor had the lowest risk of cardiovascular events. No significant difference was found between the groups stratified by antigen and activation factor levels (HR 1.54 95% CI 0.92–2.58). 3.3. Clinically relevant bleeding In univariate analysis, only the highest tertile of VWF antigen and age were associated with clinically relevant bleeding. When adjusting for age in the multivariate analysis, the association with VWF antigen was attenuated and not significant. There was no association observed between VWF activation factor and clinically relevant bleeding. The incidences of clinically relevant bleeding at different levels of antigen and activation factor are presented in Fig. 1C. No significant association was found between the groups stratified by antigen and activation factor levels (data not shown). 4. Discussion The present study is the first studying the association between VWF activation factor and cardiovascular events and all-cause mortality in patients treated with OACs. In this study, VWF activation factor was significantly associated with all-cause mortality and cardiovascular events. Even after adjustment for age, inflammation (CRP), and kidney function (eGFR), a significant association was observed with all-cause mortality. This indicates that an increased level of VWF activation factor is not solely an expression of general inflammation and impaired kidney function. We did not find an association between VWF activation factor and bleeding complications. This result sets VWF activation factor apart from VWF antigen, which has been associated with bleeding complications as well as cardiovascular events [10]. It was also noted that in patients with high levels of VWF activation factor and antigen, the incidences of cardiovascular events and mortality were especially high, indicating that these patients are at a highest risk. The VWF activation factor measurement is determining the proportion of VWF with platelet activating ability [13]; while, the VWF antigen measures all plasma VWF fragments irrespective of the platelet activating ability. The VWF antigen is often regarded as a marker of endothelial dysfunction and arteriosclerosis [11]. It is possible that a high level of VWF antigen reflects a more severe form of endothelial dysfunction, in general; while, increased active VWF may reflect a decrease in the activity regulation due to reduced proteolysis (by ADAMTS13 or another protease) or diminished clearance by low-density lipoprotein receptor-related
Table 2 Uni- and multivariate analysis for different endpoints. All-cause mortality n = 97
VWF activation factor, per SD Low (n = 115) Intermediate (n = 115) High (n = 115) VWF antigen, per SD, kU dL−1 Low (n = 111) Intermediate (n = 115) High (n = 130) C-reactive protein, per SD, mg L−1 eGFR per SD, mL/min/1.73 m2 Female Diabetes Hypertension Age. per 10 years interval
Clinically relevant bleedings
Cardiovascular events
n = 47
n = 97
Univariate
Multivariate
Univariate
Multivariate
Univariate
Multivariate
1.94 (1.56–2.40) 1 2.51 (1.31–4.83) 5.05(5 (2.73–9.33) 1.92 (1.59–2.30) 1 2.87 (1.44–5.72) 5.98 (3.13–11.43) 1.52 (1.27–1.83) 0.77 (0.62–0.96) 0.79 (0.52–1.20) 0.59 (0.35–0.98) 0.77 (0.52–1.16) 1.94 (1.52–2.46)
1.62 (1.25–2.08) 1 1.90 (0.95–3.81) 3.06 (1.56–6.0) 1.62 (1.29–2.09) 1 1.97 (0.94–4.14) 3.17 (1.54–6.52) 1.31 (1.09–1.58) 1.03 (0.82–1.29) – 0.83 (0.47–1.45) – 1.75 (1.33–2.31)
1.22 (0.90–1.65) 1 0.83 (0.83–3.21) 1.23 (0.62–2.47) 1.25 (0.94–1.67) 1 1.55 (0.72–3.35) 2.27 (1.08–4.74) 0.91 (0.68–1.22) 1.01 (0.76–1.35) 0.86 (0.47–1.55) 0.92 (0.39–2.16) 0.90 (0.50–1.62) 1.81 (1.28–2.55)
1.10 (0.80–1.51) 1 0.70 (0.33–1.49) 1.34 (0.68–2.84) 1.14 (0.84–1.54) 1 1.38 (0.64–2.98) 1.78 (0.84–3.79) – – – – – 1.77 (1.24–2.51)
1.49 (1.20–1.84) 1 2.01 (1.16–3.48) 2.54 (1.47–4.39) 1.55 (1.28–1.88) 1 2.30 (1.28–4.16) 3.47 (1.96–6.14) 1.28 (1.06–1.55) 0.91 (0.73–1.22) 0.72 (0.47–1.10) 0.40 (0.30–0.80) 0.68 (0.46–1.02) 1.88 (1.48–2.39)
1.28 (1.01–1.63) 1 1.53 (0.87–2.71) 1.74 (0.97–3.14) 1.37 (1.10–1.71) 1 2.04 (1.06–3.93) 2.65 (1.39–5.09) 1.27 (1.04–1.56) – 0.58 (0.37–0.92) 0.67 (0.39–1.13) 0.84 (0.55–1.28) 1.93 (1.48–2.52)
eGFR—estimated glomerular filtration rate. VWF—von Willebrand factor. Factors were included in the multivariate analysis if p b 0.20 in the univariate analysis, von Willebrand factor activity and antigen were not included in the same multivariate analysis.
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016
M. Lind et al. / Thrombosis Research xxx (2015) xxx–xxx
protein-1 [18]. Various findings support that VWF is not only a marker, but also an important effector in the pathogenesis of myocardial infarction [19]. Our finding that increased VWF activation factor levels are related to cardiovascular events and mortality could indicate a causal role of VWF in facilitating platelet activation. An intriguing finding is that patients with high levels of VWF antigen and activation factor seem to have a substantially higher incidence of cardiovascular events than other patients. It is possible that an increase in platelet activation could be especially deleterious in patients with more advanced atherosclerosis and endothelial dysfunction. Further studies are needed to investigate if the increased risk associated with VWF antigen is in part mediated by activation of platelets and, thereby, reflected in the levels of VWF activation factor. The levels of antigen were generally high in the studied population, which is most likely attributed to patients treated with OACs having a higher incidence of preexisting vessel wall damage. This leads to activation of the endothelium and/or exposure of the subendothelial matrix. 4.1. Strengths and limitations Samples were collected in a prospective manner, which limited the possible influence of acute phase reactions related to the studied events. The events were detected and classified in a retrospective manner by reviewing hospital records. To optimize the detection and minimize the risk of misclassification, all hospital records from all clinics were manually reviewed for each patient during the follow-up time. The levels of antigen are somewhat reduced by OAC treatment [20], and the effect on VWF activation factor is not known. We did not adjust for time in the INR range in this study. However, patients treated with warfarin in our clinic have a high time in range (74%) and we could not see any relationship between measured antigen or activity and levels of INR at sampling. In our study the samples were frozen and stored for some time before they were analyzed. The potential effects of this on levels of VWF antigen and VWF activation factor is not entirely clear. From other studies we know that a long storage time does not seem to affect the levels of fibrinolytic factors but that changes in the reagents kit used could introduce a bias [21]. Long term storage also seems to be acceptable when analyzing CRP [22]. The effects of long term storage on VWF is less studied but in a study of Rumley et al. a storage time of 6 to 12 years did not seem to affect the results for VWF antigen or factor VIIc [23]. It is reasonable to assume that storage of blood samples does not introduce a significant bias as long as all samples are analyzed at the same time, using the same reagents kits. Together, our data demonstrate that VWF activation factor in patients treated with OACs was associated with all-cause mortality and cardiovascular events particularly in patients with increased VWF antigen. There was no association with bleeding complications. The selectivity for thrombotic complications adds to the potential value of VWF activation factor as a biomarker and/or pharmacological target. Authors' contributions M. Lind collected the data designed the study, analyzed and interpreted the data and drafted the manuscript. L. Johansson, J.H Jansson, T.K Nilsson and M. J. Hollestelle designed the study, and participated in analysis and interpretation of data and in drafting of the manuscript. All authors approved the final version of the manuscript. Acknowledgements We gratefully thank prof. Philip de Groot, Department of Clinical Chemistry and Hematology, University Medical Center Utrecht, for kindly providing us with antibody AU/VWFa-11. Furthermore we like to acknowledge Marlies van Schagen for excellent technical support.
5
References [1] Z.M. Ruggeri, T.S. Zimmerman, von Willebrand factor and von Willebrand disease, Blood 70 (1987) 895–904. [2] T.C. Nichols, C.M. Samama, D.A. Bellinger, J. Roussi, R.L. Reddick, M. Bonneau, M.S. Read, O. Bailliart, G.G. Koch, M. Vaiman, et al., Function of von Willebrand factor after crossed bone marrow transplantation between normal and von Willebrand disease pigs: effect on arterial thrombosis in chimeras, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 2455–2459. [3] M. Furlan, R. Robles, D. Affolter, D. Meyer, P. Baillod, B. Lammle, Triplet structure of von Willebrand factor reflects proteolytic degradation of high molecular weight multimers, Proc. Natl. Acad. Sci. U. S. A. 90 (1993) 7503–7507. [4] J. Danesh, J.G. Wheeler, G.M. Hirschfield, S. Eda, G. Eiriksdottir, A. Rumley, G.D. Lowe, M.B. Pepys, V. Gudnason, C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease, N. Engl. J. Med. 350 (2004) 1387–1397. [5] G.Y. Lip, D. Lane, C. Van Walraven, R.G. Hart, Additive role of plasma von Willebrand factor levels to clinical factors for risk stratification of patients with atrial fibrillation, Stroke 37 (2006) 2294–2300. [6] D.S. Conway, L.A. Pearce, B.S. Chin, R.G. Hart, G.Y. Lip, Prognostic value of plasma von Willebrand factor and soluble P-selectin as indices of endothelial damage and platelet activation in 994 patients with nonvalvular atrial fibrillation, Circulation 107 (2003) 3141–3145. [7] L. Johansson, J.H. Jansson, B. Stegmayr, T.K. Nilsson, G. Hallmans, K. Boman, Hemostatic factors as risk markers for intracerebral hemorrhage: a prospective incident case-referent study, Stroke 35 (2004) 826–830. [8] J.H. Jansson, T.K. Nilsson, O. Johnson, von Willebrand factor, tissue plasminogen activator, and dehydroepiandrosterone sulphate predict cardiovascular death in a 10 year follow up of survivors of acute myocardial infarction, Heart 80 (1998) 334–337. [9] V. Roldan, F. Marin, B. Muina, J.M. Torregrosa, D. Hernandez-Romero, M. Valdes, V. Vicente, G.Y. Lip, Plasma von Willebrand factor levels are an independent risk factor for adverse events including mortality and major bleeding in anticoagulated atrial fibrillation patients, J. Am. Coll. Cardiol. 57 (2011) 2496–2504, http://dx.doi.org/ 10.1016/j.jacc.2010.12.033. [10] M. Lind, K. Boman, L. Johansson, T.K. Nilsson, L.S. Jarvholm, J.H. Jansson, von Willebrand factor predicts major bleeding and mortality during oral anticoagulant treatment, J. Intern. Med. 271 (2012) 239–246, http://dx.doi.org/10.1111/j.13652796.2011.02418.x. [11] A.D. Blann, Plasma von Willebrand factor, thrombosis, and the endothelium: the first 30 years, Thromb. Haemost. 95 (2006) 49–55. [12] C. Gandhi, A. Ahmad, K.M. Wilson, A.K. Chauhan, ADAMTS13 modulates atherosclerotic plaque progression in mice via a VWF-dependent mechanism, J. Thromb. Haemost. 12 (2014) 255–260, http://dx.doi.org/10.1111/jth.12456. [13] E. Groot, R. Fijnheer, S.A. Sebastian, P.G. de Groot, P.J. Lenting, The active conformation of von Willebrand factor in patients with thrombotic thrombocytopenic purpura in remission, J. Thromb. Haemost. 7 (2009) 962–969, http://dx.doi.org/10.1111/j. 1538-7836.2009.03433.x. [14] J.J. Hulstein, P.G. de Groot, K. Silence, A. Veyradier, R. Fijnheer, P.J. Lenting, A novel nanobody that detects the gain-of-function phenotype of von Willebrand factor in ADAMTS13 deficiency and von Willebrand disease type 2B, Blood 106 (2005) 3035–3042, http://dx.doi.org/10.1182/blood-2005-03-1153. [15] A.S. Levey, J. Coresh, T. Greene, L.A. Stevens, Y.L. Zhang, S. Hendriksen, J.W. Kusek, F. Van Lente, Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate, Ann. Intern. Med. 145 (2006) 247–254. [16] H. Tunstall-Pedoe, K. Kuulasmaa, P. Amouyel, D. Arveiler, A.M. Rajakangas, A. Pajak, Myocardial infarction and coronary deaths in the World Health Organization MONICA project. Registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents, Circulation 90 (1994) 583–612. [17] B. Stegmayr, K. Asplund, Measuring stroke in the population: quality of routine statistics in comparison with a population-based stroke registry, Neuroepidemiology 11 (1992) 204–213. [18] G. Rastegarlari, J.N. Pegon, C. Casari, S. Odouard, A.M. Navarrete, N. Saint-Lu, B.J. van Vlijmen, P. Legendre, O.D. Christophe, C.V. Denis, P.J. Lenting, Macrophage LRP1 contributes to the clearance of von Willebrand factor, Blood 119 (2012) 2126–2134, http://dx.doi.org/10.1182/blood-2011-08-373605. [19] A.O. Spiel, J.C. Gilbert, B. Jilma, von Willebrand factor in cardiovascular disease: focus on acute coronary syndromes, Circulation 117 (2008) 1449–1459, http://dx.doi.org/ 10.1161/CIRCULATIONAHA.107.722827. [20] F.L. Li-Saw-Hee, A.D. Blann, G.Y. Lip, Effects of fixed low-dose warfarin, aspirinwarfarin combination therapy, and dose-adjusted warfarin on thrombogenesis in chronic atrial fibrillation, Stroke 31 (2000) 828–833. [21] J. Hernestal-Boman, J.H. Jansson, T.K. Nilsson, M. Eliasson, L. Johansson, Long-term stability of fibrinolytic factors stored at −80 degrees C, Thromb. Res. 125 (2010) 451–456. [22] M.R. Lewis, P.W. Callas, N.S. Jenny, R.P. Tracy, Longitudinal stability of coagulation, fibrinolysis, and inflammation factors in stored plasma samples, Thromb. Haemost. 86 (2001) 1495–1500. [23] A. Rumley, G.D. Lowe, P.M. Sweetnam, J.W. Yarnell, R.P. Ford, Factor VIII, von Willebrand factor and the risk of major ischaemic heart disease in the Caerphilly heart study, Br. J. Haematol. 105 (1999) 110–116.
Please cite this article as: M. Lind, et al., von Willebrand activation factor as a marker of mortality, cardiovascular events, and bleeding complications in patients treated w..., Thromb Res (2015), http://dx.doi.org/10.1016/j.thromres.2015.08.016