Thrombosis Research 134 (2014) 803–806
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
Padua prediction score and thrombin generation in hospitalized medical patients Walid Saliba a,b,⁎, Wael Zahalka c, Lee Goldstein a,b,d, Gilat Ron a, Mazen Elias a,b a
Department of Internal Medicine C, Ha'emek Medical Center, Afula, Israel Bruce Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel Department of Internal Medicine A, Ha'emek Medical Center, Afula, Israel d Clinical Pharmacology Unit, Ha'emek Medical Center, Afula, Israel b c
a r t i c l e
i n f o
Article history: Received 29 April 2014 Received in revised form 12 July 2014 Accepted 14 July 2014 Available online 19 July 2014 Keywords: Padua score Thrombin generation Calibrated Automated Thrombogram
a b s t r a c t Introduction: The Padua prediction score is a risk assessment model used to identify medical patients at high risk for venous thromboembolim (VTE).We aimed to assess the relationship between the severity of Padua score and thrombin generation as a measure of overall thrombotic activity. Materials and Methods: A total of 253 patients hospitalized in the medical wards, at the Haemek Medical Center, Israel, were enrolled in the study. Patients treated with anticoagulation, and those admitted for VTE were excluded. Padua score was classified into two categories; low-risk for VTE (b 4 points), and high-risk for VTE (≥ 4 points). Thrombin generation was assessed by the Calibrated Automated Thrombogram (CAT) method. Results: Overall 187 (73.9%) patients had Padua score b 4, and 66 (26.1%) patients had Padua score ≥4. Comparison of the thrombogram parameters between the two Padua score categories showed no significant difference; lag time (P = 0.066), ETP (P = 0.266), peak height (P = 0.418), and time to peak (P = 0.415). Among the individual Padua score risk factors, only active cancer was significantly associated with peak height, myocardial infarction or stroke with lag time, and none of the risk factors was significantly associated with ETP. Because of their low frequency, the association with previous VTE, known thrombophilia, hormonal treatment, and recent trauma or/and surgery was not assessed. Conclusions: Single thrombin generation measurement obtained at the same time in acutely hospitalized patients didn’t bear any correlation with the Padua prediction score. This finding should be interpreted with caution considering the underrepresentation of risk factors that may influence thrombin generation. © 2014 Elsevier Ltd. All rights reserved.
Introduction Hospitalized patients are at increased risk for venous thromboembolism (VTE) [1]. The average age and sex adjusted incidence of VTE has been found to be more than 100 times greater in hospitalized patients compared to community residents [2], and more than one third of VTEs that developed in outpatient occurred in patients who have been hospitalized in the preceding 3 months [3]. It is now well established that a substantial proportion of hospitalized patients in medical wards carry a high risk of VTE, and several risk assessment models have been developed to identify high risk patients who may attain benefit from anticoagulant thromboprophylaxis [4–9]. The Padua prediction score is a risk assessment score that includes 11 risk factors for VTE, the score ranges from 0 to 20 points depending on the number and weight ⁎ Corresponding author at: Department of Internal Medicine C, Ha'emek Medical Center, 18101 Afula, Israel. Tel.: +972 4 6495132; fax: +972 4 6495243. E-mail address: [email protected] (W. Saliba).
http://dx.doi.org/10.1016/j.thromres.2014.07.022 0049-3848/© 2014 Elsevier Ltd. All rights reserved.
of the score risk factors [9]. Despite its limitations, the Padua prediction score has been recommended by the American College of Chest Physician (ACCP) as the preferred score for baseline risk stratification (low and high risk), as this model provides the best available basis for judging hospitalized patients risk for VTE [10]. It is conceivable that the increased risk of VTE, in medical ill patients, associated with high risk Padua prediction score (≥ 4 points) may be mediated by increased blood thrombogenicity. However, this association has not been assessed previously. In recent years the assessment of thrombin generation, an indicator of overall thrombotic activity, became available with the calibrated automated thrombogram (CAT) [11,12]. This study aimed to examine the relationship between Padua prediction score severity (b4 points vs. ≥ 4 points) and thrombin generation as a measure of overall thrombotic activity. This study may help to understand the underlying mechanism of increased risk of VTE in patients with high Padua score. It may open the way for future studies to assess the role of thrombin generation, either alone or in combination with Padua score, in predicting VTE in hospitalized medical patients.
804
W. Saliba et al. / Thrombosis Research 134 (2014) 803–806
Materials and Methods The study was conducted in Ha'emek Medical Center, in the northeastern area of Israel. The study was approved by the local institutional review board and ethics committee. All participants gave written informed consent prior to the study. The Study Population Candidates patients who were eligible for enrollment in the study were adult patients hospitalized in the medical wards at the Ha'emek Medical Center. Inclusions criteria were; age over 18 years and a written informed consent form. Exclusion criteria were; pregnancy, patients admitted because of VTE, and treatment with anticoagulants including; low molecular weight heparin (LMWH), unfractionated heparin, vitamin K antagonists, direct thrombin inhibitors, and direct factor Xa inhibitors. Study Variables and Definition of Terms Variables that were collected from the patients' interview and completed from the medical files included; demographic, anthropometric, and clinical variables that are components of the Padua prediction score. The Padua prediction score was calculated, for each patient at study entry, according to the weight and number of the following risk factors; active cancer (3 point), previous VTE (3 point), reduced mobility (3 point), already known thrombophilic condition (3 points), recent (≤month) trauma and/or surgery (2 points), elderly age (≥70 years) (1 point), heart and/or respiratory failure (1 point), acute myocardial infarction or ischemic stroke (1 point), acute infection and/or rheumatologic disorder (1 point), ongoing hormonal therapy (1 point), and obesity (BMI ≥ 30 Kg/m2) (1 point) [9]. Blood sample for thrombin generation was collected at the time of enrollment in the wards of internal medicine during the first 24 hours from the patient presentation to the emergency department. Thrombin generation was tested on platelet-poor plasma (PPP) with the calibrated automated thrombogram (CAT) technique using 5 picomolar tissue factor (TF) and 4 picomolar phospholipids. The thrombogram curves yielded four parameters; i) lag time, ii) endogenous thrombin potential (ETP), iii) peak height, and iv) time to peak. Short lag time/time to peak and high ETP/peak height point at hypercoagulable (pro-thrombotic) state and vice versa prolonged lag time/time to peak and decreased ETP/peak height indicate a hypoacogulabe (prohemorrhagic) state [12]. Statistical Analyses Continuous variables are summarized with means and standard deviation, and categorical variables are summarized with numbers and proportions. Padua prediction score was classified into two categories; i) Padua score b 4 points (low risk for VTE), and ii) Padua score ≥ 4 points (high risk for VTE) [9]. Thrombin generation parameters (lag time, ETP, peak height, and time to peak) were compared between the two Padua score categories using the unpaired student t test. Multiple linear regression analyses were used to assess the association between the individual risk factors of the Padua score and each of the thrombogram parameters (lag time, ETP, peak height, and time to peak). All covariates were checked against one another for collinearity. Pearson's r correlation coefficient was used to test the relation between continuous variables. P value of less than 0.05 for the 2-tailed tests was considered statistically significant. All statistical analyses were performed using IBM SPSS Statistics 21.0. Results A total of 253 patients were enrolled in the study with mean age of 59.3 ± 14.3 years, and 189 (74.7%) patients were males. The baseline
characteristics of the study population are summarized in Table 1. Seventeen (6.7%) had active cancer, 85 (33.6%) patients were obese (BMI ≥ 30 Kg/m2), only 3 patients had history of previous VTE, and 2 patients had already known thrombophilia (Table 1). The highest correlation between the thrombogram parameters was detected between lag time and time to peak (r = 0.928, P b 0.001), and between ETP and peak height (r = 0.881, P b 0.001) (Table 2). BMI had a significant positive, but weak correlation with ETP and peak height. Padua prediction score had a significant positive, but weak correlation with lag time. No significant correlation was detected between age and the thrombogram parameters (Table 2). Padua score ranged from 0 to 11 points; 58 (22.9%) patients had 0 points, 71 (28.1%) had 1 point, 38 (15%) had 2 points, and 20 (7.9%) had 3 points. Overall 187 (73.9%) patients had Padua score b 4 (low risk for VTE), and 66 (26.1%) patients had Padua score ≥ 4 (high risk for VTE). Comparison of the thrombogram parameters between the two Padua score categories showed no significant difference; lag time (P = 0.066), ETP (P = 0.266), peak height (P = 0.418), and time to peak (P = 0.415) (Table 3). In assessing the relationship between the individual risk factors of the Padua prediction score and the thrombogram parameters we have not included four risk factors in the multiple linear regression model due to their low frequency in our cohort; (3 patients had a previous history of VTE, 2 patients had known thrombophilia, 2 patients had recent trauma and/or surgery, and none was on hormonal treatment). Multiple linear regression analysis showed that only active cancer was significantly associated with peak height, acute myocardial infarction or ischemic stroke was associated with lag time, and none of the included Padua score risk factors was significantly associated with ETP (Table 4) Discussion Prediction of VTE risk among hospitalized medical patients continues to be an attractive research field. However, VTE is a multifactorial disorder and because of the large number of risk factors, assessing the risk of VTE in an individual patient is difficult and complex. It has been acknowledged that; ideally a single laboratory test that would detect multifactorial thrombophilia could help determine the overall risk of VTE. Thrombin generation which is a global coagulation assay seems to be promising for this purpose. Nonetheless, our study shows that thrombin generation was not increased in patients classified as high risk for VTE (Padua score ≥ 4 points) when compared to patients Table 1 Baseline demographic characteristics, Padua score components, and thrombogram parameters (n = 253). Gender Males Females
189 (74.7%) 64 (25.3%)
Padua prediction score component Active cancer Previous venous thromboembolism (VTE) Reduced mobility Known thrombophilic condition Recent trauma or/and surgery Age ≥ 70 years Heart and/or respiratory failure Acute myocardial infarction or stroke Acute infection and/or rheumatologic disorder Obesity (BMI ≥ 30Kg/m2) Hormonal treatment
17 (6.7%) 3 (1.2%) 58 (22.9%) 2 (0.8%) 2 (0.8%) 63 (24.9%) 62 (24.5%) 21 (8.3%) 64 (25.3%) 85 (33.6%) 0 (0%)
Thrombogram parameters Lag time (min) ETP (nM*min) Peak height (nM) Time to peak (min)
4.0 ± 1.6 1593 ± 500 296 ± 99 6.7 ± 1.9
Continuous variables are summarized with mean ± standard deviation, and categorical variables are summarized with numbers and percentages
W. Saliba et al. / Thrombosis Research 134 (2014) 803–806 Table 2 Pearson's r correlation coefficient for the relationship between the thrombogram parameters, Padua score, age, and BMI (n = 253).
Padua score Age BMI Lag time ETP Peak height Time to peak
Time to peak
Peak height
ETP
Lag time
0.021 -0.094 - 0.023 0.928** -0.159* -0.320** -
0.098 0.035 0.149* -0.125* 0.881** -
-0.050 -0.111 0.146* -0.072 -
0.139* -0.013 -0.002 -
*; P b0.05, **; P b0.01
classified as low risk for VTE (Padua score b 4 points). Furthermore, none of the individual risk factors of the Padua prediction score was found to be associated with ETP which is the most important measure of thrombin generation. Several studies have shown that increased thrombin generation is associated with increased risk of first episode and recurrent VTE [13–16]. However, no study has investigated the association of thrombin generation and risk of VTE in hospitalized acute medical ill patients. VTE within 90 days occurred in 0.3% of Padua score low risk patients (b 4 points), and in 11% of Padua score high risk untreated patients (≥ 4 points), while VTE developed in 2.2% of high risk patients who received thromboprophylaxis during hospitalization [9]. Randomized clinical trials showed that in spite of the short duration of thromboprophylaxis in hospitalized medical ill patients the observed benefit was maintained at 90 days [17,18]. It has been suggested that the correction of acute factors accounting for hospitalization even in chronically ill medical patients provide a plausible explanation for this finding [9]. Thus, the increased risk of VTE in hospitalized medical patients is likely to be transient, occurring during the acute illness that caused hospitalization. In 1884 Virchow postulated that thrombosis is caused by change in; blood flow (stasis), the state of vessel wall (endothelial activation), and/ or change in blood composition (thrombophilia) [19,20]. Based on the Virchow's triad and on the fact that thrombin generation is a global test that measures the overall tendency of a plasma sample to form thrombin after initiation of coagulation [12], it can be presumed that thrombin generation will increase only in the case of thrombophilia resulting from altered coagulation factor concentration, either genetic or acquired. Our findings indicate that thrombin generation may not be increased in hospitalized medical patients with high risk for VTE, and that the increased risk of VTE in these patients may be mainly caused by stasis and endothelial dysfunction. In such case the performance of a risk prediction score based on risk factor which are proxy for stasis and endothelial dysfunction like Padua score may be better than just assessing thrombin generation. Indeed studies have demonstrated the induction of tissue factor (TF) expression in various disease states some of which are component of the Padua prediction risk score. For instance, exposure of monocytes to bacterial lipopolysaccharide induces TF expression that may trigger the coagulation system [21]. A majority of DVT occur within the valve pocket of deep venous valves, which are exposed to periods of stasis and low oxygen levels [22]. Stasis and low blood flow in veins occur in conjunction with several classical risk factors of VTE, and during Table 3 Comparison of the thrombogram parameters between patients with Padua score b4 and patients with Padua score ≥4 (n = 253). Thrombogram parameters
Padua score b 4 (n = 187) Mean ± SD
Padua score ≥4 (n = 66) Mean ± SD
P value
Lag time (min) ETP (nM*min) Peak height (nM) Time to peak (min)
3.8 1614 293 6.6
4.4 1534 305 6.9
0.066 0.266 0.418 0.415
± ± ± ±
1.0 484 94 1.4
± ± ± ±
2.5 542 111 2.9
805
Table 4 Multiple linear regression analyses for the association of the individual risk factors of Padua prediction score with the thrombogram parametersa. Padua score risk factors
Age ≥ 70 (years) BMI ≥ 30 (Kg/m2) Active cancer Reduced mobility CHF and/or respiratory failure Acute MI or ischemic stroke Infection and/or rheumatologic disorder
Lag time
ETP
Betab
P
Beta
0.089 0.030 -0.078 0.035 0.049
0.21 -0.117 0.10 -0.057 0.42 0.038 0.60 0.64 0.101 0.12 0.088 0.17 0.010 0.87 0.24 0.091 0.17 0.136 0.04 -0.122 0.06 0.61 -0.077 0.25 -0.011 0.87 0.010 0.89 0.45 -0.013 0.84 0.056 0.39 0.021 0.75
P
Peal height
Time to peak
Beta
Beta
P
P
0.154 0.02 -0.001 0.99
0.040 0.53
0.074 0.25
0.103 0.11 -0.018 0.78
0.026 0.69
0.112 0.08
a; four risk factors of the Padua prediction score were not included because of their low frequency (only 3 patients had a previous history of VTE, 2 patients had known thrombophilia, 2 patients had recent trauma and/or surgery, and none was on hormonal treatment) b; the standardized regression coefficients are shown BMI; body mass index, CHF; congestive heart failure, MI; myocardial infarction
hospitalization for acute illness. In stasis hypoxia induces the expression of cytokines and adhesion receptors [23]. The resulting proinflammatory state of the endothelium supports the local recruitment of monocyte, granulocytes, platelets, and microparticles. The recruitment of these factors and their activation may lead to local exposure of TF [20]. In addition, blood coagulation reaction and platelet activation are in part dependent on blood flow. How and to what extent this dependence influences thrombotic risk when flow is impaired is unknown [23]. Furthermore, the Padua prediction score has not been well validated in large cohorts, and it has been found that Padua prediction score was not associated with VTE in patients with sepsis hospitalized in medical wards, and the authors concluded that Padua score lacks granularity in detecting patients at risk of acquiring VTE [24]. Although these explanations seem to be valid, it should be acknowledged that thrombin generation potential measures clot forming ability at a single time point and is not a dynamic measure. It therefore best measures underlying or pre-existing conditions that increase clot forming potential, but does not anticipate dynamic hemostatic changes which may occur in patients over the time they are admitted. In contrast, the Padua score attempts to be more anticipatory with multiple parameters including clinical ones which predict for immobilization, endothelial dysfunction, and thrombophilia. This primary difference may make comparisons between the two tools invalid and limit the ability to conclude the value of one over the other, particularly with single measurement of thrombin generation. Future studies with repeated measurement of thrombin generation at different times during admission may be helpful to address this issue. This study has other limitations; first the small number of included patients and particularly those with high risk for VTE (Padua score ≥ 4 points). In addition the data from this study come from a single medical center, which may affect the external validity of the results. Furthermore, we did not have follow up for the occurrence of VTE in the study population, hence we could not compare the performance of thrombin generation and Padua score in predicting future VTE, neither could we examine whether the combination of Padua predictive score and thrombin generation could increase the predictive accuracy of VTE. Selection bias may detract our study and may not be representative of the mix of hospitalized medical patients. Indeed, patients were young (mean 59.3 years), were more likely to be males (74.7%), had a very low frequency of previous VTE (3 patients), known thrombophilia (2 patients), recent trauma or/ and surgery (2 patients), and none of the patients were on hormonal treatment (Table 1). The younger age of our population may be explained by the exclusion of subjects who were not able to give informed consent because of cognitive impairment, more common in the elderly patients, resulting in underrepresentation of patients with advanced
806
W. Saliba et al. / Thrombosis Research 134 (2014) 803–806
age. Patients with previous VTE and known thrombophilia are more likely to be on anticoagulation treatment. Thus the lower frequency of these two conditions in our study may result from the exclusion of patients treated with anticoagulation. The exclusion of subjects treated with anticoagulants in combination with low frequency of previous VTE and known thrombophilia may result in underrepresentation of subjects who are particularly at increased risk of VTE and more likely to have increased thrombin generation. This may bias the results of our study toward the null, resulting in apparent lack of association between thrombin generation and Padua score. However, in spite of these limitations our findings may be relevant for hypothesis generation for more in depth future studies.
Conclusions Based on the lack of association between Padua score and thrombin generation, it may be suggested that the increased risk of VTE in hospitalized medical patients is mediated by stasis and endothelial activation. However, the findings of our study should be interpreted with caution considering the underrepresentation of risk factors known to be associated with increased thrombin generation. Hence, future studies are needed to replicate our findings and to examine the predictive value of thrombin generation for VTE in hospitalized medical patients and to compare its performance with Padua prediction score.
References [1] Cohen AT, Tapson VF, Bergmann JF, Goldhaber SZ, Kakkar AK, Deslandes B, et al. ENDORSE Investigators. Venous thromboembolism risk and prophylaxis in the acute hospital care setting (ENDORSE study): a multinational cross-sectional study. Lancet 2008;371:387–9. [2] Heit JA, Melton III LJ, Lohse CM, Petterson TM, Silverstein MD, Mohr DN, et al. Incidence of venous thromboembolism in hospitalized patients vs community residents. Mayo Clin Proc 2001;76:1102–10. [3] Spencer FA, Lessard D, Emery C, Reed G, Goldberg RJ. Venous thromboembolism in the outpatient setting. Arch Intern Med 2007;167:1471–5. [4] Alikhan R, Cohen AT, Combe S, Samama MM, Desjardins L, Eldor A, et al. MEDENOX Study. Risk factors for venous thromboembolism in hospitalized patients with acute medical illness: analysis of the MEDENOX Study. Arch Intern Med 2004;164:963–8. [5] Cohen AT, Alikhan R, Arcelus JI, Bergmann JF, Haas S, Merli GJ, et al. Assessment of venous thromboembolism risk and the benefits of thromboprophylaxis in medical patients. Thromb Haemost 2005;94:750–9. [6] Huang W, Anderson FA, Spencer FA, Gallus A, Goldberg RJ. Risk-assessment models for predicting venous thromboembolism among hospitalized non-surgical patients: a systematic review. J Thromb Thrombolysis 2013;35:67–80.
[7] Spyropoulos AC, Anderson Jr FA, Fitzgerald G, Decousus H, Pini M, Chong BH, et al. IMPROVE Investigators. Predictive and associative models to identify hospitalized medical patients at risk for VTE. Chest 2011;140:706–14. [8] Kucher N, Koo S, Quiroz R, Cooper JM, Paterno MD, Soukonnikov B, et al. Electronic alerts to prevent venous thromboembolism among hospitalized patients. N Engl J Med 2005;352:969–77. [9] Barbar S, Noventa F, Rossetto V, Ferrari A, Brandolin B, Perlati M, et al. A risk assessment model for the identification of hospitalized medical patients at risk for venous thromboembolism: the Padua Prediction Score. J Thromb Haemost 2010;8:2450–7. [10] Kahn SR, Lim W, Dunn AS, Cushman M, Dentali F, Akl EA, et al. American College of Chest Physicians. Prevention of VTE in nonsurgical patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141:e195S–226S. [11] Hemker HC, Giesen P, AlDieri R, Regnault V, de Smed E, Wagenvoord R, et al. The calibrated automated thrombogram (CAT): a universal routine test for hyper- and hypocoagulability. Pathophysiol Haemost Thromb 2002;32:249–53. [12] Castoldi E, Rosing J. Thrombin generation tests. Thromb Res 2011;127:S21–5. [13] van Hylckama Vlieg A, Christiansen SC, Luddington R, Cannegieter SC, Rosendaal FR, Baglin TP. Elevated endogenous thrombin potential is associated with an increased risk of a first deep venous thrombosis but not with the risk of recurrence. Br J Haematol 2007;138:769–74. [14] Ay C, Dunkler D, Simanek R, Thaler J, Koder S, Marosi C, et al. Prediction of venous thromboembolism in patients with cancer by measuring thrombin generation: results from the Vienna Cancer and Thrombosis Study. J Clin Oncol 2011;29:2099–103. [15] Tripodi A, Legnani C, Chantarangkul V, Cosmi B, Palareti G, Mannucci PM. High thrombin generation measured in the presence of thrombomodulin is associated with an increased risk of recurrent venous thromboembolism. J Thromb Haemost 2008;6:1327–33. [16] Hron G, Kollars M, Binder BR, Eichinger S, Kyrle PA. Identification of patients at low risk for recurrent venous thromboembolism by measuring thrombin generation. JAMA 2006;296:397–402. [17] Samama MM, Cohen AT, Darmon JY, Desjardins L, Eldor A, Janbon C, et al. A comparison of enoxaparin with placebo for the prevention of venous thromboembolism in acutely ill medical patients. Prophylaxis in Medical Patients with Enoxaparin Study Group. N Engl J Med 1999;341:793–800. [18] Leizorovicz A, Cohen AT, Turpie AG, Olsson CG, Vaitkus PT, Goldhaber SZ. PREVENT Medical Thromboprophylaxis Study Group. Randomized, placebo-controlled trial of dalteparin for the prevention of venous thromboembolism in acutely ill medical patients. Circulation 2004;110:874–9. [19] Prandoni P. Acquired risk factors for venous thromboembolism in medical patients. Hematology Am Soc Hematol Educ Program 2005:458–61. [20] Manly DA, Boles J, Mackman N. Role of tissue factor in venous thrombosis. Annu Rev Physiol 2011;73:515–25. [21] Pawlinski R, Pedersen B, Schabbauer G, Tencati M, Holscher T, Boisvert W, et al. Role of tissue factor and protease-activated receptors in a mouse model of endotoxemia. Blood 2004;103:1342–7. [22] Brooks EG, Trotman W, Wadsworth MP, Taatjes DJ, Evans MF, Ittleman FP, et al. Valves of the deep venous system: an overlooked risk factor. Blood 2009; 114:1276–9. [23] Reitsma PH, Versteeg HH, Middeldorp S. Mechanistic view of risk factors for venous thromboembolism. Arterioscler Thromb Vasc Biol 2012;32:563–8. [24] Vardi M, Ghanem-Zoubi NO, Zidan R, Yurin V, Bitterman H. Venous thromboembolism and the utility of the Padua Prediction Score in patients with sepsis admitted to internal medicine departments. J Thromb Haemost 2013;11:467–73.
Contents lists available at ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
Padua prediction score and thrombin generation in hospitalized medical patients Walid Saliba a,b,⁎, Wael Zahalka c, Lee Goldstein a,b,d, Gilat Ron a, Mazen Elias a,b a
Department of Internal Medicine C, Ha'emek Medical Center, Afula, Israel Bruce Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel Department of Internal Medicine A, Ha'emek Medical Center, Afula, Israel d Clinical Pharmacology Unit, Ha'emek Medical Center, Afula, Israel b c
a r t i c l e
i n f o
Article history: Received 29 April 2014 Received in revised form 12 July 2014 Accepted 14 July 2014 Available online 19 July 2014 Keywords: Padua score Thrombin generation Calibrated Automated Thrombogram
a b s t r a c t Introduction: The Padua prediction score is a risk assessment model used to identify medical patients at high risk for venous thromboembolim (VTE).We aimed to assess the relationship between the severity of Padua score and thrombin generation as a measure of overall thrombotic activity. Materials and Methods: A total of 253 patients hospitalized in the medical wards, at the Haemek Medical Center, Israel, were enrolled in the study. Patients treated with anticoagulation, and those admitted for VTE were excluded. Padua score was classified into two categories; low-risk for VTE (b 4 points), and high-risk for VTE (≥ 4 points). Thrombin generation was assessed by the Calibrated Automated Thrombogram (CAT) method. Results: Overall 187 (73.9%) patients had Padua score b 4, and 66 (26.1%) patients had Padua score ≥4. Comparison of the thrombogram parameters between the two Padua score categories showed no significant difference; lag time (P = 0.066), ETP (P = 0.266), peak height (P = 0.418), and time to peak (P = 0.415). Among the individual Padua score risk factors, only active cancer was significantly associated with peak height, myocardial infarction or stroke with lag time, and none of the risk factors was significantly associated with ETP. Because of their low frequency, the association with previous VTE, known thrombophilia, hormonal treatment, and recent trauma or/and surgery was not assessed. Conclusions: Single thrombin generation measurement obtained at the same time in acutely hospitalized patients didn’t bear any correlation with the Padua prediction score. This finding should be interpreted with caution considering the underrepresentation of risk factors that may influence thrombin generation. © 2014 Elsevier Ltd. All rights reserved.
Introduction Hospitalized patients are at increased risk for venous thromboembolism (VTE) [1]. The average age and sex adjusted incidence of VTE has been found to be more than 100 times greater in hospitalized patients compared to community residents [2], and more than one third of VTEs that developed in outpatient occurred in patients who have been hospitalized in the preceding 3 months [3]. It is now well established that a substantial proportion of hospitalized patients in medical wards carry a high risk of VTE, and several risk assessment models have been developed to identify high risk patients who may attain benefit from anticoagulant thromboprophylaxis [4–9]. The Padua prediction score is a risk assessment score that includes 11 risk factors for VTE, the score ranges from 0 to 20 points depending on the number and weight ⁎ Corresponding author at: Department of Internal Medicine C, Ha'emek Medical Center, 18101 Afula, Israel. Tel.: +972 4 6495132; fax: +972 4 6495243. E-mail address: [email protected] (W. Saliba).
http://dx.doi.org/10.1016/j.thromres.2014.07.022 0049-3848/© 2014 Elsevier Ltd. All rights reserved.
of the score risk factors [9]. Despite its limitations, the Padua prediction score has been recommended by the American College of Chest Physician (ACCP) as the preferred score for baseline risk stratification (low and high risk), as this model provides the best available basis for judging hospitalized patients risk for VTE [10]. It is conceivable that the increased risk of VTE, in medical ill patients, associated with high risk Padua prediction score (≥ 4 points) may be mediated by increased blood thrombogenicity. However, this association has not been assessed previously. In recent years the assessment of thrombin generation, an indicator of overall thrombotic activity, became available with the calibrated automated thrombogram (CAT) [11,12]. This study aimed to examine the relationship between Padua prediction score severity (b4 points vs. ≥ 4 points) and thrombin generation as a measure of overall thrombotic activity. This study may help to understand the underlying mechanism of increased risk of VTE in patients with high Padua score. It may open the way for future studies to assess the role of thrombin generation, either alone or in combination with Padua score, in predicting VTE in hospitalized medical patients.
804
W. Saliba et al. / Thrombosis Research 134 (2014) 803–806
Materials and Methods The study was conducted in Ha'emek Medical Center, in the northeastern area of Israel. The study was approved by the local institutional review board and ethics committee. All participants gave written informed consent prior to the study. The Study Population Candidates patients who were eligible for enrollment in the study were adult patients hospitalized in the medical wards at the Ha'emek Medical Center. Inclusions criteria were; age over 18 years and a written informed consent form. Exclusion criteria were; pregnancy, patients admitted because of VTE, and treatment with anticoagulants including; low molecular weight heparin (LMWH), unfractionated heparin, vitamin K antagonists, direct thrombin inhibitors, and direct factor Xa inhibitors. Study Variables and Definition of Terms Variables that were collected from the patients' interview and completed from the medical files included; demographic, anthropometric, and clinical variables that are components of the Padua prediction score. The Padua prediction score was calculated, for each patient at study entry, according to the weight and number of the following risk factors; active cancer (3 point), previous VTE (3 point), reduced mobility (3 point), already known thrombophilic condition (3 points), recent (≤month) trauma and/or surgery (2 points), elderly age (≥70 years) (1 point), heart and/or respiratory failure (1 point), acute myocardial infarction or ischemic stroke (1 point), acute infection and/or rheumatologic disorder (1 point), ongoing hormonal therapy (1 point), and obesity (BMI ≥ 30 Kg/m2) (1 point) [9]. Blood sample for thrombin generation was collected at the time of enrollment in the wards of internal medicine during the first 24 hours from the patient presentation to the emergency department. Thrombin generation was tested on platelet-poor plasma (PPP) with the calibrated automated thrombogram (CAT) technique using 5 picomolar tissue factor (TF) and 4 picomolar phospholipids. The thrombogram curves yielded four parameters; i) lag time, ii) endogenous thrombin potential (ETP), iii) peak height, and iv) time to peak. Short lag time/time to peak and high ETP/peak height point at hypercoagulable (pro-thrombotic) state and vice versa prolonged lag time/time to peak and decreased ETP/peak height indicate a hypoacogulabe (prohemorrhagic) state [12]. Statistical Analyses Continuous variables are summarized with means and standard deviation, and categorical variables are summarized with numbers and proportions. Padua prediction score was classified into two categories; i) Padua score b 4 points (low risk for VTE), and ii) Padua score ≥ 4 points (high risk for VTE) [9]. Thrombin generation parameters (lag time, ETP, peak height, and time to peak) were compared between the two Padua score categories using the unpaired student t test. Multiple linear regression analyses were used to assess the association between the individual risk factors of the Padua score and each of the thrombogram parameters (lag time, ETP, peak height, and time to peak). All covariates were checked against one another for collinearity. Pearson's r correlation coefficient was used to test the relation between continuous variables. P value of less than 0.05 for the 2-tailed tests was considered statistically significant. All statistical analyses were performed using IBM SPSS Statistics 21.0. Results A total of 253 patients were enrolled in the study with mean age of 59.3 ± 14.3 years, and 189 (74.7%) patients were males. The baseline
characteristics of the study population are summarized in Table 1. Seventeen (6.7%) had active cancer, 85 (33.6%) patients were obese (BMI ≥ 30 Kg/m2), only 3 patients had history of previous VTE, and 2 patients had already known thrombophilia (Table 1). The highest correlation between the thrombogram parameters was detected between lag time and time to peak (r = 0.928, P b 0.001), and between ETP and peak height (r = 0.881, P b 0.001) (Table 2). BMI had a significant positive, but weak correlation with ETP and peak height. Padua prediction score had a significant positive, but weak correlation with lag time. No significant correlation was detected between age and the thrombogram parameters (Table 2). Padua score ranged from 0 to 11 points; 58 (22.9%) patients had 0 points, 71 (28.1%) had 1 point, 38 (15%) had 2 points, and 20 (7.9%) had 3 points. Overall 187 (73.9%) patients had Padua score b 4 (low risk for VTE), and 66 (26.1%) patients had Padua score ≥ 4 (high risk for VTE). Comparison of the thrombogram parameters between the two Padua score categories showed no significant difference; lag time (P = 0.066), ETP (P = 0.266), peak height (P = 0.418), and time to peak (P = 0.415) (Table 3). In assessing the relationship between the individual risk factors of the Padua prediction score and the thrombogram parameters we have not included four risk factors in the multiple linear regression model due to their low frequency in our cohort; (3 patients had a previous history of VTE, 2 patients had known thrombophilia, 2 patients had recent trauma and/or surgery, and none was on hormonal treatment). Multiple linear regression analysis showed that only active cancer was significantly associated with peak height, acute myocardial infarction or ischemic stroke was associated with lag time, and none of the included Padua score risk factors was significantly associated with ETP (Table 4) Discussion Prediction of VTE risk among hospitalized medical patients continues to be an attractive research field. However, VTE is a multifactorial disorder and because of the large number of risk factors, assessing the risk of VTE in an individual patient is difficult and complex. It has been acknowledged that; ideally a single laboratory test that would detect multifactorial thrombophilia could help determine the overall risk of VTE. Thrombin generation which is a global coagulation assay seems to be promising for this purpose. Nonetheless, our study shows that thrombin generation was not increased in patients classified as high risk for VTE (Padua score ≥ 4 points) when compared to patients Table 1 Baseline demographic characteristics, Padua score components, and thrombogram parameters (n = 253). Gender Males Females
189 (74.7%) 64 (25.3%)
Padua prediction score component Active cancer Previous venous thromboembolism (VTE) Reduced mobility Known thrombophilic condition Recent trauma or/and surgery Age ≥ 70 years Heart and/or respiratory failure Acute myocardial infarction or stroke Acute infection and/or rheumatologic disorder Obesity (BMI ≥ 30Kg/m2) Hormonal treatment
17 (6.7%) 3 (1.2%) 58 (22.9%) 2 (0.8%) 2 (0.8%) 63 (24.9%) 62 (24.5%) 21 (8.3%) 64 (25.3%) 85 (33.6%) 0 (0%)
Thrombogram parameters Lag time (min) ETP (nM*min) Peak height (nM) Time to peak (min)
4.0 ± 1.6 1593 ± 500 296 ± 99 6.7 ± 1.9
Continuous variables are summarized with mean ± standard deviation, and categorical variables are summarized with numbers and percentages
W. Saliba et al. / Thrombosis Research 134 (2014) 803–806 Table 2 Pearson's r correlation coefficient for the relationship between the thrombogram parameters, Padua score, age, and BMI (n = 253).
Padua score Age BMI Lag time ETP Peak height Time to peak
Time to peak
Peak height
ETP
Lag time
0.021 -0.094 - 0.023 0.928** -0.159* -0.320** -
0.098 0.035 0.149* -0.125* 0.881** -
-0.050 -0.111 0.146* -0.072 -
0.139* -0.013 -0.002 -
*; P b0.05, **; P b0.01
classified as low risk for VTE (Padua score b 4 points). Furthermore, none of the individual risk factors of the Padua prediction score was found to be associated with ETP which is the most important measure of thrombin generation. Several studies have shown that increased thrombin generation is associated with increased risk of first episode and recurrent VTE [13–16]. However, no study has investigated the association of thrombin generation and risk of VTE in hospitalized acute medical ill patients. VTE within 90 days occurred in 0.3% of Padua score low risk patients (b 4 points), and in 11% of Padua score high risk untreated patients (≥ 4 points), while VTE developed in 2.2% of high risk patients who received thromboprophylaxis during hospitalization [9]. Randomized clinical trials showed that in spite of the short duration of thromboprophylaxis in hospitalized medical ill patients the observed benefit was maintained at 90 days [17,18]. It has been suggested that the correction of acute factors accounting for hospitalization even in chronically ill medical patients provide a plausible explanation for this finding [9]. Thus, the increased risk of VTE in hospitalized medical patients is likely to be transient, occurring during the acute illness that caused hospitalization. In 1884 Virchow postulated that thrombosis is caused by change in; blood flow (stasis), the state of vessel wall (endothelial activation), and/ or change in blood composition (thrombophilia) [19,20]. Based on the Virchow's triad and on the fact that thrombin generation is a global test that measures the overall tendency of a plasma sample to form thrombin after initiation of coagulation [12], it can be presumed that thrombin generation will increase only in the case of thrombophilia resulting from altered coagulation factor concentration, either genetic or acquired. Our findings indicate that thrombin generation may not be increased in hospitalized medical patients with high risk for VTE, and that the increased risk of VTE in these patients may be mainly caused by stasis and endothelial dysfunction. In such case the performance of a risk prediction score based on risk factor which are proxy for stasis and endothelial dysfunction like Padua score may be better than just assessing thrombin generation. Indeed studies have demonstrated the induction of tissue factor (TF) expression in various disease states some of which are component of the Padua prediction risk score. For instance, exposure of monocytes to bacterial lipopolysaccharide induces TF expression that may trigger the coagulation system [21]. A majority of DVT occur within the valve pocket of deep venous valves, which are exposed to periods of stasis and low oxygen levels [22]. Stasis and low blood flow in veins occur in conjunction with several classical risk factors of VTE, and during Table 3 Comparison of the thrombogram parameters between patients with Padua score b4 and patients with Padua score ≥4 (n = 253). Thrombogram parameters
Padua score b 4 (n = 187) Mean ± SD
Padua score ≥4 (n = 66) Mean ± SD
P value
Lag time (min) ETP (nM*min) Peak height (nM) Time to peak (min)
3.8 1614 293 6.6
4.4 1534 305 6.9
0.066 0.266 0.418 0.415
± ± ± ±
1.0 484 94 1.4
± ± ± ±
2.5 542 111 2.9
805
Table 4 Multiple linear regression analyses for the association of the individual risk factors of Padua prediction score with the thrombogram parametersa. Padua score risk factors
Age ≥ 70 (years) BMI ≥ 30 (Kg/m2) Active cancer Reduced mobility CHF and/or respiratory failure Acute MI or ischemic stroke Infection and/or rheumatologic disorder
Lag time
ETP
Betab
P
Beta
0.089 0.030 -0.078 0.035 0.049
0.21 -0.117 0.10 -0.057 0.42 0.038 0.60 0.64 0.101 0.12 0.088 0.17 0.010 0.87 0.24 0.091 0.17 0.136 0.04 -0.122 0.06 0.61 -0.077 0.25 -0.011 0.87 0.010 0.89 0.45 -0.013 0.84 0.056 0.39 0.021 0.75
P
Peal height
Time to peak
Beta
Beta
P
P
0.154 0.02 -0.001 0.99
0.040 0.53
0.074 0.25
0.103 0.11 -0.018 0.78
0.026 0.69
0.112 0.08
a; four risk factors of the Padua prediction score were not included because of their low frequency (only 3 patients had a previous history of VTE, 2 patients had known thrombophilia, 2 patients had recent trauma and/or surgery, and none was on hormonal treatment) b; the standardized regression coefficients are shown BMI; body mass index, CHF; congestive heart failure, MI; myocardial infarction
hospitalization for acute illness. In stasis hypoxia induces the expression of cytokines and adhesion receptors [23]. The resulting proinflammatory state of the endothelium supports the local recruitment of monocyte, granulocytes, platelets, and microparticles. The recruitment of these factors and their activation may lead to local exposure of TF [20]. In addition, blood coagulation reaction and platelet activation are in part dependent on blood flow. How and to what extent this dependence influences thrombotic risk when flow is impaired is unknown [23]. Furthermore, the Padua prediction score has not been well validated in large cohorts, and it has been found that Padua prediction score was not associated with VTE in patients with sepsis hospitalized in medical wards, and the authors concluded that Padua score lacks granularity in detecting patients at risk of acquiring VTE [24]. Although these explanations seem to be valid, it should be acknowledged that thrombin generation potential measures clot forming ability at a single time point and is not a dynamic measure. It therefore best measures underlying or pre-existing conditions that increase clot forming potential, but does not anticipate dynamic hemostatic changes which may occur in patients over the time they are admitted. In contrast, the Padua score attempts to be more anticipatory with multiple parameters including clinical ones which predict for immobilization, endothelial dysfunction, and thrombophilia. This primary difference may make comparisons between the two tools invalid and limit the ability to conclude the value of one over the other, particularly with single measurement of thrombin generation. Future studies with repeated measurement of thrombin generation at different times during admission may be helpful to address this issue. This study has other limitations; first the small number of included patients and particularly those with high risk for VTE (Padua score ≥ 4 points). In addition the data from this study come from a single medical center, which may affect the external validity of the results. Furthermore, we did not have follow up for the occurrence of VTE in the study population, hence we could not compare the performance of thrombin generation and Padua score in predicting future VTE, neither could we examine whether the combination of Padua predictive score and thrombin generation could increase the predictive accuracy of VTE. Selection bias may detract our study and may not be representative of the mix of hospitalized medical patients. Indeed, patients were young (mean 59.3 years), were more likely to be males (74.7%), had a very low frequency of previous VTE (3 patients), known thrombophilia (2 patients), recent trauma or/ and surgery (2 patients), and none of the patients were on hormonal treatment (Table 1). The younger age of our population may be explained by the exclusion of subjects who were not able to give informed consent because of cognitive impairment, more common in the elderly patients, resulting in underrepresentation of patients with advanced
806
W. Saliba et al. / Thrombosis Research 134 (2014) 803–806
age. Patients with previous VTE and known thrombophilia are more likely to be on anticoagulation treatment. Thus the lower frequency of these two conditions in our study may result from the exclusion of patients treated with anticoagulation. The exclusion of subjects treated with anticoagulants in combination with low frequency of previous VTE and known thrombophilia may result in underrepresentation of subjects who are particularly at increased risk of VTE and more likely to have increased thrombin generation. This may bias the results of our study toward the null, resulting in apparent lack of association between thrombin generation and Padua score. However, in spite of these limitations our findings may be relevant for hypothesis generation for more in depth future studies.
Conclusions Based on the lack of association between Padua score and thrombin generation, it may be suggested that the increased risk of VTE in hospitalized medical patients is mediated by stasis and endothelial activation. However, the findings of our study should be interpreted with caution considering the underrepresentation of risk factors known to be associated with increased thrombin generation. Hence, future studies are needed to replicate our findings and to examine the predictive value of thrombin generation for VTE in hospitalized medical patients and to compare its performance with Padua prediction score.
References [1] Cohen AT, Tapson VF, Bergmann JF, Goldhaber SZ, Kakkar AK, Deslandes B, et al. ENDORSE Investigators. Venous thromboembolism risk and prophylaxis in the acute hospital care setting (ENDORSE study): a multinational cross-sectional study. Lancet 2008;371:387–9. [2] Heit JA, Melton III LJ, Lohse CM, Petterson TM, Silverstein MD, Mohr DN, et al. Incidence of venous thromboembolism in hospitalized patients vs community residents. Mayo Clin Proc 2001;76:1102–10. [3] Spencer FA, Lessard D, Emery C, Reed G, Goldberg RJ. Venous thromboembolism in the outpatient setting. Arch Intern Med 2007;167:1471–5. [4] Alikhan R, Cohen AT, Combe S, Samama MM, Desjardins L, Eldor A, et al. MEDENOX Study. Risk factors for venous thromboembolism in hospitalized patients with acute medical illness: analysis of the MEDENOX Study. Arch Intern Med 2004;164:963–8. [5] Cohen AT, Alikhan R, Arcelus JI, Bergmann JF, Haas S, Merli GJ, et al. Assessment of venous thromboembolism risk and the benefits of thromboprophylaxis in medical patients. Thromb Haemost 2005;94:750–9. [6] Huang W, Anderson FA, Spencer FA, Gallus A, Goldberg RJ. Risk-assessment models for predicting venous thromboembolism among hospitalized non-surgical patients: a systematic review. J Thromb Thrombolysis 2013;35:67–80.
[7] Spyropoulos AC, Anderson Jr FA, Fitzgerald G, Decousus H, Pini M, Chong BH, et al. IMPROVE Investigators. Predictive and associative models to identify hospitalized medical patients at risk for VTE. Chest 2011;140:706–14. [8] Kucher N, Koo S, Quiroz R, Cooper JM, Paterno MD, Soukonnikov B, et al. Electronic alerts to prevent venous thromboembolism among hospitalized patients. N Engl J Med 2005;352:969–77. [9] Barbar S, Noventa F, Rossetto V, Ferrari A, Brandolin B, Perlati M, et al. A risk assessment model for the identification of hospitalized medical patients at risk for venous thromboembolism: the Padua Prediction Score. J Thromb Haemost 2010;8:2450–7. [10] Kahn SR, Lim W, Dunn AS, Cushman M, Dentali F, Akl EA, et al. American College of Chest Physicians. Prevention of VTE in nonsurgical patients: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest 2012;141:e195S–226S. [11] Hemker HC, Giesen P, AlDieri R, Regnault V, de Smed E, Wagenvoord R, et al. The calibrated automated thrombogram (CAT): a universal routine test for hyper- and hypocoagulability. Pathophysiol Haemost Thromb 2002;32:249–53. [12] Castoldi E, Rosing J. Thrombin generation tests. Thromb Res 2011;127:S21–5. [13] van Hylckama Vlieg A, Christiansen SC, Luddington R, Cannegieter SC, Rosendaal FR, Baglin TP. Elevated endogenous thrombin potential is associated with an increased risk of a first deep venous thrombosis but not with the risk of recurrence. Br J Haematol 2007;138:769–74. [14] Ay C, Dunkler D, Simanek R, Thaler J, Koder S, Marosi C, et al. Prediction of venous thromboembolism in patients with cancer by measuring thrombin generation: results from the Vienna Cancer and Thrombosis Study. J Clin Oncol 2011;29:2099–103. [15] Tripodi A, Legnani C, Chantarangkul V, Cosmi B, Palareti G, Mannucci PM. High thrombin generation measured in the presence of thrombomodulin is associated with an increased risk of recurrent venous thromboembolism. J Thromb Haemost 2008;6:1327–33. [16] Hron G, Kollars M, Binder BR, Eichinger S, Kyrle PA. Identification of patients at low risk for recurrent venous thromboembolism by measuring thrombin generation. JAMA 2006;296:397–402. [17] Samama MM, Cohen AT, Darmon JY, Desjardins L, Eldor A, Janbon C, et al. A comparison of enoxaparin with placebo for the prevention of venous thromboembolism in acutely ill medical patients. Prophylaxis in Medical Patients with Enoxaparin Study Group. N Engl J Med 1999;341:793–800. [18] Leizorovicz A, Cohen AT, Turpie AG, Olsson CG, Vaitkus PT, Goldhaber SZ. PREVENT Medical Thromboprophylaxis Study Group. Randomized, placebo-controlled trial of dalteparin for the prevention of venous thromboembolism in acutely ill medical patients. Circulation 2004;110:874–9. [19] Prandoni P. Acquired risk factors for venous thromboembolism in medical patients. Hematology Am Soc Hematol Educ Program 2005:458–61. [20] Manly DA, Boles J, Mackman N. Role of tissue factor in venous thrombosis. Annu Rev Physiol 2011;73:515–25. [21] Pawlinski R, Pedersen B, Schabbauer G, Tencati M, Holscher T, Boisvert W, et al. Role of tissue factor and protease-activated receptors in a mouse model of endotoxemia. Blood 2004;103:1342–7. [22] Brooks EG, Trotman W, Wadsworth MP, Taatjes DJ, Evans MF, Ittleman FP, et al. Valves of the deep venous system: an overlooked risk factor. Blood 2009; 114:1276–9. [23] Reitsma PH, Versteeg HH, Middeldorp S. Mechanistic view of risk factors for venous thromboembolism. Arterioscler Thromb Vasc Biol 2012;32:563–8. [24] Vardi M, Ghanem-Zoubi NO, Zidan R, Yurin V, Bitterman H. Venous thromboembolism and the utility of the Padua Prediction Score in patients with sepsis admitted to internal medicine departments. J Thromb Haemost 2013;11:467–73.
